FIRSTPHE?SSN

Game Design asNarrativeArchitectureHenry JenkinsThe relationship between games and story remains adivisive question among game fans, designers, andscholars alike. At a recent academic Games Studiesconference, for example, a blood feud threatened toerupt between t}e self-proclaimed ludologists, whowanted to see the focus shift onto the mechanics ofgame play, and the narratologists, who were interestedin studying games alongside other storytelling media.rConsider some recent statements made on this issue,

Interacti 'r ity is almost the opposite ofnarrative; narrative flows under thedirection of the author, while interactivitydepends on the player for motive power.(Adams 1999)

There is a &rect, immediate conflictbetween the demands of a story and thedemands of a game. Divergence from a

storys path is likely to make for a lesssatisfying story; restricting a player'sfreedom of action is likely to make for aless satisfying game. (Costikyan 2000,44-s3)

Computer games are not narratives....Rather the narrative tends to be isolatedlrom or even work against the computer-game-ness of the game. (Juul f gg8)z

Outside academic theory people areusually excellent at making distinctionsbetween narrative, drama and games. If Ithrow a ball at you I don't expect you todrop it and wait until it starts tellingstories. (Eskelinen 2001)

I find myself responding to this perspective with mjxedfeelings. On the one hand, I understand what thesewriters are arguing against - various attempts to maptraditional narrative structures ("hypertext,""lnteractive Cinema," "nonlinear narrative") onto gamesat the expense ofan attention to their specificity as anemerging mode of entertainment. You say "narrative"

to the average gamer and what they are apt to imagineis something on the order of a choose-your,own

Response by Jon McKenzieThe model of creativity often associated with digitalmedia is not that of originality and uniqueness butrecombination and multiplicity, a model hardwired tothe computer's uncanny ability to copy and combineimages, sounds, texts, and other materials from anendless array ofsources. Indeed, in different thoughrelated ways, both &gital media and poststructuralisttheory teach us that it is impossible to create and studythe new without drawing at times on forms andprocesses taken from what is already around us. Fromthis perspective, no genre, work, or field is unique andself-contained: each is a specific yet fuzzy combinationof other things that are tJremselves diverse andnonunique. In short, what makes something "unique" is

not so much its make up but its "mix,up."

For practical, conceptual, and institutional reasons,any formation of a field of "ludology" may inevitablyinvolve arguing for that field's uniqueness andoriginality, its clear-cut distinction from other fields:thus,

'games are not narratives, not films, not plays,

etc." Yet I'm willing to gamble that if a formal disciplineof ludology ever does emerge, it will sooner or laterdiscover what other disciplines have learned:&scoveries are triggered by the oddest (and oldest) ofsources.

As Henry Jenkins suggests, games are indeed notnarratives, not films, not plays - but theyre also not-not-narratives, not-not-films, not-not-plays. Gamesshare traits with other forms of cultural production,

Game Theories 'Jenkins McKenzie Eskel inen. l r r ' r ' . i "c l . : ; i i a , :

l,'.:,::,,,.,, l';:::; :;,;i''*t "adventure book, a form noted for its lifelessness andmechanical exposition rathe r than entJ-rrallingentertainment, thematic sophistication, or charactercomplexiry. And game industry executives are perhapsjustly skeptical that they have much to learn from theresolutely unpopular (and often overtly antipopular)aesthetics promoted by hypertext theorists. Theapplication of film theory to games can seem hea',y-handed and literal-minded, often failing to recognizethe profound differences between the two me&a. Yet,at the same time, there is a tremendous amount thatgame designers and critics could learn through makingmeaningful comparisons with other storytelling media.One gets rid of narrative as a framework for thinkingabout games only at ones own risk. In this short piece, Ihope to offer a middle-ground position between theludologists and the narratologists, one that respects theparticularity of this emerging medium - examininggames less as stories than as spaces ripe with narrativepossibility.

Let's start at some points where we might all agree:

1. Not all games tell stories. Games may be an abstract,expressive, and experiential form, closer to music ormodern dance t}ran to cinema. Some ballets (Tfte

Nutcracker for example) tell stories, but storytelling isn't

IV. Game Theories

an intrinsic or defining feature of dance. Similarly,many of my own favorite games - Tetis, BIix, Snood -

are simple graphic games that do not lend themselvesvery well to narrative exposition.3 To understand suchgames, we need other terms and concepts beyondnarrative, including interface design and expressivemovement for starters. The last thing we want to do isto reign in the creative experimentation that needs tooccur in the earlier years of a medium's development.

2.Many games do have narrative aspirations.Minimally, they want to tap the emotional residue ofprevious narrative experiences. Often, they depend onour familiarity with the roles and goals of genreentertainment to orient us to the action, and in manycases, game designers want to create a series ofnarrative experiences for the player. Given thosenarrative aspirations, it seems reasonable to suggestthat some understanding of how games relate tonarrative is necessary before we understand tleaesthetics of game design or the nature ofcontemporary game culture.

3. Narrative analysis need not be prescriptive, even ifsome narratologists - Janet Murray is the most oft-cited example - do seem to be advocating for gamesto pursue particular narrative forms. There is not one

although reducing them to any one of these comes at acertain cost. Jenkins rightly contends that gamedesigners should therefore seek to expand the formsand processes from which to draw rather t}'ran reducethem. He is also right to point out that someIudologists are themselves much too quick to reducenarrative to overly simplistic models (e.g., strictly linearstructures). Most importantiy, his exploration ofspatially oriented narrative forms prorrides provocativeapproaches to contemporary game design. At the sametime, however, Jenkins's stated goal to offer a "middle

ground" between ludologists and narratologistsremains slanted toward the narratological end ofthings. This is indicated in his essay's title, "Game

Design as Narrative Architecture." A more pla14u1

ludologist might have offered a response titled"Narrative Architecture as Game Design." JohanHuizinga, after all, analyzed law, war, poetry, andphilosophy'as" play, and across d.rverse culturaltraditions storytelling has complex agonistic&mensions.

Another middle ground for ludology might be"experience design," a notion and practice that runs indifferent ways from Brenda Laurel to Donald Normanto Eric Zimmerman. Experience design refers to the

generation and shaping of actions, emotions, andthoughts. How one operates a kitchen appliance, takes ina sophisticated science exhibition, or becomes enmeshedin a role-plapng game - or for that matter shops in a

store, reads a novel, or visits a polling booth - all this

Game Design as Narrative ArchitectureHenry Jenkins

future of games. The goal should be to fosterdiversification ofgenres, aesthetics, and audiences, toopen gamers to the broadest possible range ofexperiences. The past few years have been ones ofenormous creative experimentation and innovationwithin the games industry, as might be represented bya list of some of the groundbreaking titles. The Sims,Black andWhite, Majestic, Shenmue: each representsprofoundly different concepts of what makes forcompelling game play. A discussion of the narrativepotentials of games need not imply a privileging ofstorytelling over all the other possible t}ings games cando, even if we might suggest that if game designers aregoing to tell stories, they should tell them well. In orderto do that, game designers, who are most oftenschooled in computer science or graphic design, need tobe retooled in the basic vocabulary of narrative theory.

4. The experience of playing games can never be simplyreduced to the experience of a story. Many other factorsthat have little or nothing to do with storytelling per secontribute to the development of great games and weneed to significantly broaden our critical vocabulary fortalking about games to deal more fully with those othertopics. Here, the ludologists insistence that gamescholars focus more attention on the mechanics ofgame play seems totally in order.

FIRSTPERSON

5. If some games tell stories, they are unlikely to tellt-hem in the same ways t-hat other media tell stories.Stories are not empty content tlat can be ported fromone media pipeline to another. One would be hard-pressed, for example, to translate the internal dialogueof Proust's Remembrance of Things Past into acompelling cinematic experience, and the tight controlover viewer experience that Hitchcock achieves in hissuspense films would be directly antithetical to theaesthetics of good game design. We must, therefore, beattentive to the particularity of games as a medium,specifically what distinguishes them from othernarrative traditions. Yet, in order to do so requiresprecise comparisons - not the mapping of old modelsonto games but a testing of those models againstexisting games to determine what features they sharewith other media and how they differ.

Much of the writing in the ludologist tradition rsunduly polemical: they are so busy trying to pull gamedesigners out of their "cinema enry or define a fieldwhere no hypertext theorist dares to venture that theyare prematurely dismissing the use value of narrativefor understanding their desired object of study. For mymoney, a series of conceptuai blind spots prevent themfrom developing a full understanding of the interplaybetween narrative and games.

First, the discussion operates with too narrow a

From Markku Eskelinen's 0nline ResponseFor some reason Henry Jenkins doesn't define thecontested concepts (narratives, stories, and games) socentral to his argumentation. That's certainly aneffective way of building a middle ground (or aperiphery), but perhaps not the most convincing one.

can be approached in terms of experience design. Howare interactions organized and solicited? How does oneevent flow into another? How does the overallexperience "hang together"? Although Laurel t}eor2esexperience design using the model of fuistoteliant-heater (arguing that it has been shaping au&ences'experience for centuries), there are in practice an almostunlimited set of performative models to draw upon:sports, rituals, sagas, popular entertainments, novels,jokes, and so on.

Perhaps what's really at stake in ludology is less theright model and more a sense of tone and attitude - awillingness to mix it up, to entertain many possibilities,to play with lots of different models.

Jenkins also misrepresents a &spute (on theusefulness of narratology), important parts of which heseems to be unaware of It has its roots both in EspenAarseth's Cybertext (which deals extensively with therelationship between stories and games, showingelementary differences in communicative structures ofnarratives and adventure games) and Gonzalo Frasca'sintroduction of ludology to computer game stu&es. A

Game Theories > Jenkins McKenzie Esketinen,: r ir j l i |-r PlarCe

;:;;"",",.- ::il,ffi i;l*-'-model of narrative, one preoccupied with the rules andconventions of classical hnear storytelling at theexpense of consideration of other kinds of narratives,not only the modernist and postmodernistexperimentation that inspired the hlpertext theorists,but also popular traditions that emphasize spatialexploration over causal event chains or which seek tobalance the competing demands of narrative andspectacle.a

Second, the &scussion operates with too Limited anunderstan&ng of narration, focusing more on theactivities and aspirations of the storyteller and toolittle on the process of narrative comprehension.s

Third, the discussion deals only with the question ofwhether whole games tell stories and not whethernarrative elements might enter games at a morelocalized level. Finally, the discussion assumes tJratnarratives must be self-contained rather thanunderstanding games as serving some specificfunctions within a new transme&a storytellingenvironment. Rethinking each of these issues mightlead us to a new understanding of the relationshipbetween games and stories. Specifically, I want tointroduce an important third term into this discussion- spatiality - and argue for an understanding ofgame designers less as storytellers and more asnarrative architects.

IV. Game Theories

Spatiat Stories andEnviron mentaI StoryteltingGame designers don't simply tell stories; they designworlds and sculpt spaces. It is no accident, for example,that game design documents have historically beenmore interested in issues of levei design than onplotting or character motivation. A prehistory of videoand computer games might take us through t}eevolution of paper mazes or board games, bothpreoccupied with the design of spaces, even where theyalso provided some narrative context. Monopoly, forexample, may tell a narrative about how fortunes arewon and lost; the individual Chance cards may providesome story pretext for our gaining or losing a certainnumber of places; but ultimately, what we remember isthe experience of moving around the board and landingon someone's real estate. Performance theorists havedescribed role-playing games (RPGs) as a mode ofcollaborative storytelling, but the Dungeon Mastersactivities start with designing the space - thedungeon - where the players' quest will take place.Even many of the early text-based games, such as Zork,whlch could have told a wide array of different kinds ofstories, centered around enabling players to movethrough narratively compelling spaces: "You are facingthe north side of a white house. There is no door here,and all of t}re windows are boarded up. To the north a

discussion of the present topic, which ignores theseworks, cannot hope to break new ground. A few facts ofcultural history wouldn't hurt either: as the oldestastragals (forerunners of dice) date back to prehistory,I'm not so sure'games fit within a much older tra&tionof spatial stories."

http://wwuetectronicboo kreview. com/th read/fi rstperson/eskeli nen r1

Jenkins RespondsI feel a bit like Travis Bickle when I ask Eskelinen,

'Are

you talking to me?" For starters, I don't consider myselfto be a narratologist at all.

http ://www. e [ectro nicb o o kreview. co m/th read/fi rstperson/jen ki nsr2

GameHenry

Design as Narrative ArchitectureJenkins

70.7. Civilizotion 3. (Atari)

narrow path winds tlrough tie trees." The earlyNintendo games have simple narrative hooks - rescuePrincess Toadstool - but what gamers foundastonishing when they first played them were tleircomplex and imaginative graphic realms, which were somuch more sophisticated than the simple grids thatPong or Pac-Man had offered us a decade earlier.

When we refer to such influential early works asShigeru Miyamoto's Super Mario Bros. as "scroll games,"we situate them alongside a much older tradition ofspatial storytelling: many Japanese scroll paintings map,for example, the passing of the seasons onto anunfolding space. When you adapt a film into a game, theprocess gpically involves translating events in the 6lminto environments witlin the game. When gamermagazines want to describe the experience of gameplay,they are more likely to reproduce maps of the gameworld than to recount their narratives.6 Before we cantalk about game narratives, then, we need to talk aboutgame spaces. Across a series of essays, I have made thecase that game consoles should be regarded as machinesfor generating compelling spaces, that their virtualplayspaces have helped to compensate for tle decliningplace of tle traditional baclcyard in contemporary boyculture, and that t}re core narratives behind many gamescenter around the struggle to explore, map, and mastercontested spaces (Fuller and Jenkins 1994; Jenkins1998). Here, I want to broaden that discussion furtherto consider in what ways the structuring of game space

FIRSTPHRSON

facilitates different kinds of narrative experiences.As such, games fit within a much older tradition of

spatial stories, which have often taken the form ofhero's odysseys, quest myths, or travel narratives.T Thebest works of J.R.R. Tolkien, Jules Verne, Homer, L.Frank Baum, or Jack London fall loosely within thistradition, as does, for example, the sequence inWar andPeace that describes Pierre's aimless wanderings acrossthe battlefield at Borodino. Often, such works exist onthe outer borders of literature. They are much loved byreaders, to be sure, and passed down from onegeneration to another, but they rarely figure in thecanon of great literary works. How often, for example,has science fiction been criticized for being preoccupiedwitl-r world-making at t}re expense of characterpsychology or plot development?

These writers seem constandy to be pushing againstthe limits of what can be accomplished in a printed textand thus their works fare badly against aestheticstandards defined around classically constructednovels. In many cases, the characters - our guidesthrough these richly developed worlds - are strippeddown to the bare bones, description displacesexposition, and plots fragment into a series of episodesand encounters. When game designers draw storyelements from existing film or literary genres, tley aremost apt to tap those genres - fantasy, adventure,science fiction, horror, war - which are most investedin world-making and spatial storytelling. Games, inturn, may more firlly realize the spatiality of thesestories, giving a much more immersive and compellingrepresentation of their narrative worlds. Anyone whodoubts that Tolstoy might have achieved his true callingas a game designer should reread the final segment ofWar and Peace where he works through how a series ofa-lternative choices might have reversed the outcome ofNapoleon's Russian campaign. The passage is deadweight in the context of a novel, yet it outlines ideasthat could be easily communicated in god-games suchas those in the Civilizahon series (figure 10.1).

Don Carson, who worked as a Senior Show Designerfor Walt Disney Imagineering, has argued that gamedesigners can learn a great deal by studying techniquesof "environmental storytelling," which Disney employs

Game Theor ies > Jenkins McKenzie Eskel inen

;,1i... t. ,r1",- **r,.,or.r ,l : i : ; r r : , t*r , t : L jeLi .?fr-r id i " i . i r* i

in designing amusement park attractions. Carsonexplains,

The story element is infused into thephysical space a guest walks or ridesthrough. It is the physical space that doesmuch of the work of conveying the storytlre designers are trytngto tell.... fumedonly with their own knowledge of theworld, and those visions collected frommovies and books, the audience is ripe tobe dropped into your adventure. The trickis to piay on those memories andexpectations to heighten the thrill ofventuring into your created universe.(Carson 2000)

The amusement park attraction doesn't so muchreproduce tle story of a literary work, such as The Windin the Willows, as it evokes its atmosphere; the originalstory provides

'a set of rules that will guide the design

and project team to a common goal" and that will helpgive structure and meaning to the visitors experience.Ifl for example, the attraction centers around pirates,Carson writes, "every texture you use, every sound youplay, every turn in the road should reinforce theconcept of pirates," while any contradictory elementmay shatter the sense of immersion into this narrativeuniverse. The same might be said for a game such as SeaDogs, which, no Iess tlan Pirates of the Caibbean,depends on its ability to map our preexisting piratefantasies. The most significant difference is thatamusement park designers count on visitors keepingtheir hands and arms in the car at all times and t}ushave a greater control in shaping our total experience,whereas game designers have to develop worlds wherewe can touch, grab, and fling things about at wiil.

Environmental storytelling creates the preconditionsfor an immersive narrative experience in at least one offour ways: spatial stories can evoke pre-existingnarrative associations; they can provide a stagingground where narrative events are enacted; they mayembed narrative information within their mise-en-scene; or they provide resources for emergent narratives.

IV. Game Theories

10.2. American McGee's Alice (Rogue Entertainment, Etectronic Arts)

Evocative SpacesThe most compelling amusement park attractions buildupon stories or genre tra&tions already well-known tovisitors, allowing them to enter physica-lly into spacesthey have visited many times before in their fantasies.These attractions may either remediate a preexistingstory (Back to the Future) or draw upon a broadly sharedgenre tradition (Disney's Haunted Mansion). Suchworks do not so much tell self-contained stories asdraw upon our previously existing narrativecompetencies. They can paint their worlds in fairlybroad outlines and count on the visitor /player to do therest. Something similar might be said of many games.For example, American McGee's Alice* is an originalinterpretation of Lewis Carroll's Alice inWonderland(figure 10.2). Alice has been pushed into madness afteryears of living with uncertainty about whether her

Game Design as Narrative ArchitectureHenry Jenkins

Wonderland experiences were real or hallucinations;now, shes come back into this world and is looking forblood. McGees wonderland is not a whimsicaldreamscape but a dark nightmare realm. McGee cansafely assume that players start t|e game wit} a prettywell-developed mental map of the spaces, characters,and situations associated with Carrolls fictionaluniverse and that they wilJ read his distorted and oftenmonstrous images against the background of mentalimages formed from previous encounters withstorybook illustrations and Disney movies. McGeerewrites Alice's story in large part by redesigning Alice'ssPaces.

fuguing against games as stories, Jesper Juulsuggests that, "you clearly can't deduct the story of StarWars from Star Wars the game," whereas a film versionof a novel will give you at least the broad outlines of tlreplot (Juul 1998). This is a pretty old-fashioned modelof the process of adaptation. Increasingly, we inhabit aworld of transmedia storytelling, one that depends lesson each individua-l work being self-sufficient than oneach work contributing to a larger narrative economy.The Star Wars game may not simply retell the story ofStar Wars. but it doesn't have to in order to enrich orexpand our experience of the Star Wars saga.

We already know the story before we even buy thegame and would be frustrated if all it offered us was aregurgitation ofthe original film experience. Rather,the Star Wars game exists in dialogue with the films,conveying new narrative experiences through itscreative manipulation of environmental details. Onecan imagine games taking their place within alargernarrative system with story informationcommunicated through books, film, television, comics,and otler medja, each doing what it does best, each arelatively autonomous experience, but the richestunderstanding of the story world coming to tlose whofollow the narrative across the various channels. Insuch a system, what games do best will almost certainlycenter around their ability to give concrete shape to ourmemories and imaginings of the story'world, creatingan immersive environment we can wander through andinteract with,

FIRSTPERSON

Enacting StoriesMost often, when we &scuss games as stories, we arereferringto games that either enable players to perform

or witness narrative events - for example, to grab a

light-saber and dispatch Darth Maul in aStar Warsgame. Narrative enters such games on two levels - rn

terms of broadly defined goals or conflicts and on thelevel of localized incidenrs.

Many game critics assume that all stories must beclassically constructed with each element tightlyintegrated into the overall plot trajectory Costikyan(2000) writes, for example, that

'a story is a controlled

experience; the author consciously crafts it, choosingcertain events precisely, in a certain order, to create astory with maximum impact."s

Adams (1999) daims, "a good story hangs togethertlre way a good jigsaw puzzle hangs together. When youpick it up, every piece is locked tightly in place next to

its neighbors."Spatial stories, on the other hand, are often

&smissed as episodic - that is, each episode (or setpiece) can become compelling on its own termswithout contributing significantly to the plotdevelopment, and often t}e episodes could be reorderedwit}out significantly impacting our experience as awhole. There may be broad movements or series ofstages within the story as Troy Dunniway suggestswhen he draws parallels between the stages in theHero's journey (as outlined by Joseph Campbell) andthe levels of a ciassic adventure game, but within eachstage, the sequencing of actions may be quite loose.Spatial stories are not badly constructed stories; rather,they are stories that respond to alternative aestheticprinciples, privileging spatial exploration over plotdevelopment. Spatial stories are held together bybroadly defined goals and conflicts and pushed forwardby the characters movement across the map. Theirresolution often hinges on the player reaching his orher final destination, though, as Mary Fuller notes, notall travel narratives end successfully or resolve thenarrative enigmas that set them into motion. Onceagain, we are back to principles of "environmental

storytelling." The organization of the plot becomes amatter of designing the geography of imaginary worlds,

Game Theories > Jenkins McKenzie Esketinenj ' . iu[ i ic Pearce

:;;;:,*." ::;:il,u;,li'*"'"so that obstacles thwart and affordances facilitate theprotagonists forward movement towards resolution.Over the past several decades, game designers havebecome more and more adept at setting and varyrngthe rhythm of game play through features of the gamespace.

Narrative can a-lso enter games on the level oflocalized incident, or what I am calling micronarratives.We might understand how micronarratives work bythinking about the Odessa Steps sequence in SergeiEisenstein's B attle ship Potemkin. First, recogni ze that,whatever its serious moral tone, the scene basicallydeals with the same kind of material as most games -

the steps are a contested space with one group (thepeasants) trying to advance up and another (theCossacks) moving down.

Eisenstein intensifies our emotiona-l engagementwith this large-scale conflict through a series of shortnarrative units. The woman with the baby carriage isperhaps the best known of those micronarratives. Eachof these units builds upon stock characters orsituations drawn from t}e repertoire of melodrama.None of them last more than a few seconds, thouqhEisenstein proiongs them (and intensifies theiremotiona,l impact) through cross-cutting betweenmultiple incidents. Eisenstein used the term "attraction"

to describe such emotionally packed elements in hiswork; contemporary game designers might call them"memorable moments." Just as some memorablemoments in games depend on sensations (the sense ofspeed in a racing game) or perceptions (the suddenexpanse of slcy in a snowboarding game) as well asnarrative hooks, Eisenstein used the word "attractions"

broadly to describe any element within a work thatproduces a profound emotional impact, and theorizedthat the themes of the work could be communicatedacross and through these discrete elements. Even gamesthat do not create large-scale plot trajectories may welldepend on these micronarratives to shape the player'semotiona-l experience. Micronarratives may be cut,scenes, but they don't have to be. One can imagine asimple sequence of preprogrammed actions throughwhich an opposing player responds to your successfiJtouchdown in a football same as a micronarrative.

IV. Game Theories

Game critics often note that the player's participationposes a potential t}reat to the narrative construction,whereas the hard rails of the plotting can overlyconstrain the "freedom, power, and self-expression"associated with interactivity (Adams 1999). Thetension between performance (or game play) andexposition (or story) is far from unique to games. Thepleasures of popular culture often center on spectacularperformance numbers and self-contained set pieces. Itmakes no sense to describe musical numbers or gagsequences or action scenes as disruptions of the film'splots: the reason we go to see a kung fu movie is to seeJackie Chan show his stuff.e Yet, few films consistsimply of such moments, typically falling back on somebroad narrative exposition to create a frameworkwithin which localized actions become meaningfirl.10

We might describe musicals, action films, or slapstickcomedies as having accordion-like structures. Certainplot points are fixed, whereas other moments can beexpanded or contracted in response to audiencefeedback without serious consequences to the overallplot. The introduction needs to establish the character'sgoals or explain the basic conflict; the conclusion needsto show the successful completion of those goals or thefinal defeat of the antagonist. In commedia dell'arte, forexample, the masks define tle relationships betweenthe characters and give us some sense of their goals anddesires.r 1

The masks set limits on the action, even though tlreperformance as a whoie is created throughimprovisation. The actors have mastered the possiblemoves, or lazzi, associated with each character, much asa game player has mastered the combination of buttonsthat must be pushed to enable certain characteractions. No author prescribes what the actors do oncethey get on t}re stage, but the shape of the storyemerges from this basic vocabulary of possible actionsand from the broad parameters set by this theatricaltra&tion. Some of the lazzi can contribute to the plotdevelopment, but many of them are simple restagingsof the basic oppositions (the knave tricks the master orgets beaten).

These performance or spectacle-centered genresoften display a pleasure in process - in the

Game Design as Narrative ArchitectureHenry Jenkins

experiences along the road - that can overwhelm anystrong sense of goal or resolution, while exposition canbe experienced as an unwelcome interruption to t}repleasure of performance. Game designers struggle witllthis same balancing act - trying to determine howmuch plot will create a compelling framework and howmuch freedom players can enjoy at a local level withouttotally derailing the larger narrative trajectory. Asinexperienced storytellers, they often fall back onrather mechanical exposition through cut scenes, muchas early filmmakers were sometimes overly reliant onintertides rather than learning the skills of visualstorytelling. Yet, as with any other aestietic tra&tion,game designers are apt to develop craft through aprocess of experimentation and refinement of basrcnarrative devices, becoming better at shaping narrativeexperiences without unduly constraining the space forimprovisation within the game.

Embedded NarrativesRussian formalist critics make a useful distinctionbetween plot (or syuzhet) that refers to, in KristenThompsons (198S) terms, "t-he structured set of allcausal events as we see and hear them presented in thefilm itselfl" and srory (or fabu]a), whiclL refers to tJ.reviewers mental construction of the chronology ofthose events (Thompson 1988,39-40). Few films ornovels are absolutely linear; most make use of someforms of backstory t}at is revealed gradually as wemove tlrough the narrative action. The detective storyis the classic illustration of this principle, telling twostories - one more or less chronological (the story ofthe investigation itself) and the other told radically outof sequence (the events motivating and lea&ng up tothe murder).

According to this model, narrative comprehension isan active process by whlch rriewers

".r.-tl. and make

hypotieses about likely narrative developments on thebasis of information drawn from textualcues andclues.l2 As they move through the film, spectators testand reformulate their mental maps of the narrativeaction and the story space. In games, players are forcedto act upon those mental maps, to literally test themagainst the game world itself. If you are wrong about

FIRSTPERSShI

whether the bad guys lurk behind the next door, youwill find out soon enough - perhaps by being biown

Sway and having to start the game over. The heavy-handed exposition that opens many games serves auseful function in orienting spectators to the corepremises so that they are less lil<ely to mal<e stupid andcostly errors as tiey first enter into t}re game world.Some games create a space for rehearsal, as well, so thatwe can make sure we understand our character'spotential moves before we come up against thechallenges of navigating narrational space.

Read in this light, a story is less a temporal srructurethan a body of information. The authoiof a fil- or abook has a high degree of control over when and if wereceive specific bits of information, but a game designercan somewhat control the narrationai process by&stributing the information across the game space.Within an open,ended and exploratory narrativestructure like a game, essential narrative informationmust be presented redundantly across a range of spacesand artifacts, because one cannot assume the playerwill necessarily locate or recognize the significance ofany given element. Game designers have developed avariety of kludges that allow them to prompt prayers orsteer them towards narratively salient spaces. yet, thisis no different from t}le ways that ."du.,d"n.v is builtinto a television soap opera, where the assumption isthat a certain number of yiewers are apt to miss anygiven episode, or even in classical Hollywood narrative,where the law of three suggest.s that any essential plotpoint needs to be communicated in at least three ways.

To continue with the detective example, then, oneca-n imagine the game designer as developing two kindsof narratives -_ one relatively unstructured andcontrolled by the player as they explore the game spaceand unlock its secrets; the other prestructured burembedded within the mise-en-scene awaiting discovery.The game world becomes a kind of informaii,cn space, amemory pala ce. Myst is a highly successfirl

"*ample of

this kind of embedded narrative, but embeddednarrative does not necessarily require an emptying oftJle space of contempora ry narrative activities, as agame such as Half.Life might suggest. Embeddednarrative can and often does occur within contested

Game Theor ies 'Jenkins McKenzie Eskel inen

;;:::." ,,1^1;=,;"o,u,.",,*,.: i iqi i;:{: i::r; i, C r:tv.jori Jrtr.t l

spaces. We may have to battle our way past antagonists,navigate through mazes, or figure out how to pick locksin order to move through the narratively impregnatedmise-en-scene. Such a mixture of enacted andembedded narrative elements can allow for a balancebetween the flexibility of interactivity and thecoherence of a pre-authored narrative.

UsingQuake as an example, Jesper Juul argues thatflashbacks are impossible within games, because thegame play always occurs in real-time (Juul 1998). Yeqthis is to confuse story and plot. Games are no morelocked into an eternal present than films are alwayslinear. Many games contain moments of revelation orartifacts that shed Light on past actions. Carson (2000)

suggests that part of the art of game design comes infinding artfr.rl ways of embedding narrativeinformation into tie en'v-ironment without destroyingits immersiveness and without giving the player asensation ofbeing drug around by the neck:

Staged areas... [can] lead the game player tocome to their own conclusions about aprerrious event or to suggest a potcntiajdanger just ahead. Some examplesinclude... doors that have been brokenopen, traces ofa recent explosion, acrashed vehicle, a piano dropped from agreat height, charred remains of a fire.

Players, he argues, can return to a familiar space later inthe game and discover it has been transformed bysubsequent (off-screen) events. Clive Barker's Undying,for example, creates a powerfirl sense of backstory inprecisely this manner. It is a story of sibling rivalry thathas taken on supernatural dimensions. As we visit eachcharacters space, we have a sense of the human theyonce were and the demon they have become. In PeterMolynenx's Black and Wh it e, the player's ethical choiceswithin the game leave traces on the landscape orreconfigure the physical appearances oftleir charactersHere, we might read narrative consequences off mise-en-scene the same way we read Dorian Gray'sdebauchery off ofhis portrait. Carson describes suchnarrative devices as "following Saknussemm," referring

IV. Game Theories

to the ways that the protagonists of Jules Verne'sJourney to The Center of the Earth keep stumbling acrossclues and artifacts left behind by the sixteenth-centuryIcelandic scientist/explorer fune Saknussemm, andreaders become fascinated to see what they can learnabout his ultimate fate as the travelers come closer toreaching their intended destination.

Game designers might study melodrama for a betterunderstanding ofhow artifacts or spaces can containaffective potential or communicate significant narrativeinformation. Melodrama depends on the externalprojection of internal states, often through costumedesign, art direction, or lighting choices. As we enterspaces, we may become overwhelmed with powerfirlfeelings of loss or nostalgia, especially in thoseinstances where the space has been transformed bynarrative events. Consider, for example, the moment inDoctor Zhivago when the characters return to themansion, now completely deserted and encased in ice,or when Scariett O'Hara travels across the scorchedremains of her famiiy estate in Gone With the Windfollowing the burning of Atianta ln Alfred HitchcocklsRebecca, the tide character never appears, but she exertsa powerfirl influence over the other characters -

especially the second Mrs. DeWinter, who must inhabita space where every artifact recalls her predecessor.F]itchcock creates a number of scenes of his protagonistwandering through Rebeccas space, passing throughlocked doors, staring at her overwhelming portrait onthe wall, touching her things in drawers, or feeling thetexture of fabrics and curtains. No matter where shegoes in the house, she cannot escape Rebeccas memory.

A game such as Neil Young's Majestic pushes thisnotion of embedded narrative to its logical extreme.Here, the embedded narrative is no longer containedwithin the console but rather flows across multipleinformation channels. The player's activity consists of

sorting through documents, deciphering codes, makingsense of garbled transmissions, moving step-by-step

towards a firller understanding of the conspiracy that is

the game's primary narrative focus. We follow links

between web sites; we get information throughwebcasts, faxes, e-mails, and phone calls. Such anembedded narrative doesn't require a branching story

Game Design as Narrative ArchitectureHenry Jenkins

structure but rather depends on scrambling the piecesof a Linear story and allowing us to reconstruct the plotthrough our acts of detection, speculation, exploration,and decryption. Not surprisingly, most embeddednarratives, at present, take the form of detective orconspiracy stories, since these genres help to motivatethe players active examination of clues and explorationof spaces and provide a rationale for our efforts toreconstruct the narrative ofpast events. Yet. as thepreceding examples suggest. melo&ama providesanother - and as yet largely unexplored - model forhow an embedded story might work, as we read lettersand diaries, snoop around in bedroom drawers andclosets, in search of secrets that might shed light on therelationships between characters.

FIRSTPERSON

Emergent NarrativesThe Sims represents a fourtl model of how narrativepossibfities might get mapped onto game space (figure

10.3). Emergent narratives are not prestructured orpreprogrammed, taking shape through the game play,yet they are not as unstructured, chaotic, andfrustrating as life itself. Game worlds, ultimately, arenot real worlds, even those as densely developed asShenmue oras geographically expansive as Everquest.Will Wright frequently describes The Sims as a sandboxor dollhouse game, suggesting that it should beunderstood as a kind of autloring environment withinwhich players can define their own goals and writetheir own stories. Yet, unlike Microsoft Word, the gamedoesn't open on a blank screen. Most players comeaway from spending time with The Sims with somedegree of narrative satisfaction. Wright has created aworid ripe with narrative possibilities, where eachdesign decision has been made with an eye towardsincreasing the prospects of interpersonal romance orconflict.

The ability to design our own "skins" encouragesplayers to create characters who are emotionallysignificant to them, to rehearse their own relationshipswith friends, family, or coworkers or to map charactersfrom other fictional universes onto The Sims. A glanceat the various scrapbooks players have posted on theweb suggests that they have been quick to takeadvantage ofits relatively open-ended structure. Yet,let's not underestimate the designers' contributions.The characters have a will of their own, not alwayssubmitting easily to the players control, as when adepressed protagonist refuses to seek employment,preferring to spend hour upon hour soaking in theirbath or moping on the front porch.

Characters are given desires, urges, and needs, whichcan come into conflict with each other, and thusproduce dramatically compelling encounters. Charactersrespond emotionally to events in their enrrironment, aswhen characters mourn the loss of a loved one. Ourchoices have consequences, as when we spend all of ourmoney and have nothing left to buy them food. Thegibberish language and flashing symbols allow us tomap our own meanings onto tJre conversations, yet the

10.3 The Sims. (Maxis. Etectronic Arts)

Game Theories 'Jenkins McKenzie Eskel inen

i:::,. ill"l"1'j *-.-,.,u.,"Zimnierman Crawford . Iuul

tone of voice and body language can powerfully expressspecific emotional states, which encourage us tounderstand those interactions within familiar plor

situations. The designers have made choices aboutwhat kinds of actions are and are not possible in thisworld, such as aliowing for same-sex kisses, but limitingthe degree of explicit sexual activity that can occur.(Good programmers may be able to get around suchrestrictions, but most players probably work within thelimitations of t}re system as given.)

Janet Murray's Hamlet on the Holodecl< might describesome of what Wright accomplishes here as proceduralauthorship. Yet, I would argue that his choices godeeper than this, working not simply through theprogramming, but also through the design of the gamespace. For example, just as a dollhouse offers astreamlined representation that cuts out much of theclutter of an actual domestic space, the Sims' houses arestripped down to only a small number of artifacts, eachof which perform specific kinds of narrative functions.Newspapers, for example, communicate jobinformation. Characters sleep in beds. Bookcases canmake you smarter. Bottles are for spinning and thusmotivating lots of kissing. Such choices result in ahighly legible narrative space. In his classic study TfreImage of The City, Kevin Lynch made the case thaturban designers needed to be more sensitive to thenarrative potentials of city spaces, describing cityplanning as "t-he deliberate manipulation of the worldfor sensuous ends" (Lynch 1960, 116).

Urban designers exert even less control than gamedesigners over how people use the spaces they create orwhat kinds of scenes they stage there. Yet, some kindsof space lend themselves more rea&ly to narrativelymemorable or emotionally meaningfirl experiencesthan others. Lynch suggested that urban plannersshould not attempt to totally predetermine the usesand meanings of the spaces they create: "a landscapewhose every rock tells a story may make difficuit thecreation offresh stories" (Lynch 1960,6). Rather, heproposes an aesthetic of urban design that endows eachspace with "poetic and symbolic" potential "Such asense of place in itself enhances every human activitythat occurs there, and encourages the deposit of a

IV. Game Theories

memory trace" (Lyrrch 1960, 119). Game designerswould do well to study Lynch's book, especially as theymove into t}le production of game platforms whichsupport player-generated narratives.

In each ofthese cases, choices about the design andorganization of game spaces have narratologicalconsequences. In the case of evoked narratives, spatialdesign can either enhance our sense of immersionwithin a familiar world or communicate a freshperspective on that story through the altering ofestablished details. In the case of enacted narratives, thestory itself may be structured around the character'smovement through space and the features of theenvironment may retard or accelerate that plottrajectory.ln the case of embedded narratives, t}re gamespace becomes a memory palace whose contents mustbe deciphered as the player tries to reconstruct the plot.And in the case of emergent narratives, game spaces aredesigned to be rich with narrative potential, enablingthe story-constructing activity ofplayers. In each case,it makes sense to think of game designers less asstorytellers than as narrative architects.

Notes1. The term "Ludo[ogy" was coined by Espen Aarseth, who advocatesthe emergence of a new fieLd of study, specificatty focused on thestudy of games and game pLay, rather than framed through theconcerns of pre-existing disciplines or other media. (Editors' note:Markku Eske[inen, in his response to this essay, points out that theterm was introduced to computer game studies by Gonzalo Frasca.This introduction, according to Frasca, was in the Cybertext Yearbook- a pubtication coedited by Esketinen and named for Aarseth'sCybertext [1997].)2. For a more recent formulation of this same argument, see JesperJuul. (2001), "Games TeL[ing Stories?"

3. EskeLinen (2001) takes Janet Murray to task for her narratjveanalysis of lefns as "a perfect enactment of the overtasked lives ofAmericans in the 1990s - of the constant bombardment of tasksthat demand our attention and that we must somehow f i t into ourovercrowded schedu|'es and clear off our desks in order to make roomfor the next ons[aught." Eskelinen is conect to note that theabstraction of lefns wou[d seem to defu nanative interpretation, butthat js not the same thing as insist ing that no meaningful anaLysiscan be made of the game and its fit within contemporary culture.Ietru might well express something of the frenzied pace of modern[ife, just as modern dances might, without being a story.

4. 'A story is a collection of facts in a time-sequenced order thatsuggest a cause and effect relationship" (Crawford 1982). "The storyis the antjthesis of game. The best way to tel.[ a story is in [inearform. The best way to create a game is to provide a structure withinwhich the p[ayer has freedom of act'ion" (Costikyan, 2000).

Game Design as Narrative Architecture FIRSTPERSONHenryJenkins

5. "In its richest form, storytelling - narrat'ive - means thereade/s surrender to the author. The author takes the reader by thehand and leads him into the wortd of h is imaginat ion. The readerhas a rote to pLay, but it's a fairly passive roLe: to pay attention, tounderstand, perhaps to th ink. . . but not to act" (Adams 1999).

6. As I have noted elsewhere, these maps take a distinctive form -

not objective or abstract top-down views but composites ofscreenshots that represent the game wortd as we will encounter it inour travels through its space. Game space never exists in abstract,but always experientiatLy.

7. My concept of spatiaI stories is strongly influenced by MicheI deCerteau (1988) The Pradice of Everyday Life and Henri LeFebvre(1,991), The hoduction of Space.

8. For a futler discussion of the norms of c[assicalty constructednarrative, see Bordwetl, Staiger, and Thompson (L985), The ClassicalHollywood Cinema.

9. For usefuI d iscussion of th is issue in f iLm theorv. see Dona[dCraf ton (1995), "Pie and Chase: Gag, SpectacLe and Narrat ive inSLapst ick Comedy," in Kr ist ine Brunovska Karnick and Henry Jenkins(eds.), Classical Hollywood Comedy; Henry Jenkins (1991), WhatMade Pistachio Nuts?: Early Sound Comedy and The VaudevilleAesthetic; Rick Attman (1999), The American.Filrn .rYusical; TomGunning (1990), "Ihe Cinema of Attractions: Earty Fitm, Its Spectatorand the Avant Garde" in Thomas Etsaesser wi th Adam Barker (eds-) ,Early Cinema: Space, Frame, Narrative; Linda Wi[iams (1999), HardCore: Power, Pleasure and "The Frenzy of the Visible."

10. "Games that just have nonstop act ion are fun for a whi le butoften get boring. This is because of the lack of intrigue, suspense,and drama. How many act ion movies have you seen where the heroof the story shoots his gun every few seconds and is always on therun? People lose interest watching this kind of movie. Playing agame is a bit different, but the fact is the brain becomes overst imulated af ter too much nonstop act ion" (Dunniway 2000).

11. See, for example, John Rudl in (1994), Commedia Del l 'Ar te: AnActor's Handbook for a detailed inventory of the masks and lazzi ofth is t radi t ion.

12. See, for example, David Bordwell (1989), lVanahon in the FidionFilm, and Edward Branigan (1992), Narrative Comprehension and FiIm.

ReferencesAarseth, Espen (1997). Cybertext: Perspectives on Ergodic Literature.BaLtimore: Johns Hopkins University Press.

Adams, Ernest (1999). "Three Problems For Interactive Storytelters."Gamasutra. December 29. 7999.

Attman, Rick (1,999). The American Film Musical. B[oomington:Indiana University Press.

Bordwell, David, Janet Staiger, and Kristen Thompson (1,985). TheClasstcal Hollywood Cinema. New York: Columbia University Press.

BordwelL, David (1989). Narration in the Fidion Film. Madison:University of Wisconsin.

Branigan, Edward (1992). Narrative Comprehension and Fflm. NewYork: Routledge.

Carson, Don (2000). "Environmental Storytelt ing: Creating Immersive3D Worlds Using Lessons Learned From the Theme Park Industry."Gamasutra, March 1, 2000..htIp: / /www.gamasutra.com,/featu res/20000301 / carson_pfu. htm>.

Costikyan, Greg (2000). "Where Stories End and Games }egin." Game

Developer, September 2000.

Craf ton. Donald (1995). "Pie and Chase: Gag. Spectacle and Narrat ivein Sl.apstick Comedy." In Classical Hollywood Comedy, edited byKr ist jne Brunovska Karnick and Henry Jenkins. New York:Rout[edge/American Fitm Institute.

Crawford, Chris (1982). The Art of Computer Game Design.<http://www.vancouver.wsu.edu/f ac/ peabody / gane-book/Coverpage. htm L>.

de Certeau, Michet (1988). The hactice of Everyday lrle. Berkeley:University of Ca[ifornia Press.

Dunniway, Troy (2000). "Using the Hero's Journey in Games."Gamasutra, November 27. 2000..http://www. gamasutra.com/feattres / 20001127 /du n n iway_pfv. htm'.

Esketinen, Markku (2001). "The Gaming Situat'ion." Game Studies 1,,no.1 (Juty 2001). <http:./ / cmc.uib.no/gamestudies/0t01/eske[inen'.

Frasca, Gonzalo (1999). "Ludotogy Meets Narratology: Simit i tude andDifferences between (Video) Games and Narrative."<http://www.jacara nda. org/frasca/Ludoto gy. htm'.

FuL[er , Mary, and Henry Jenkins (1994). "Nintendo and New WortdNarrative." In Communications in Cyberspace, edited by Steve Jones.New York: Sage.

Gunning, Tom (1990). "The Cinema of Attractions: Earty Fi|.m, ItsSpectator and the Avant Garde." In Early Cinema: Space, Frame,Narrat[ve, edited by Thomas Elsaesser with Adam Barker. London:British Fil.rn Institute.

Jenkins, Henry (1991). What Made Hstachio Nuts?: Early SoundComedy and The Vaudeville Aesthetic. New York: Columbia UniversityPress.

Jenkins, Henry (1993). "x Logic: Placing Nintendo in Chi ldren'sLives." Quarterly Review of Film and Vfdeo, August 1993.

Jenkins, Henry (1998). 'Complete Freedom of Movemenf: Video

Games as Gendered Playspace." In From Barbie to Mortal Kombat:Gender and Computer Games," edited by Justine Cassell and HenryJenkins. Cambridge: The MIT Press.

Juu[ , Jesper (1998). 'A C[ash Between Games and Nanat ive." Paper

presented at the Digital Arts and Culture Conference, Bergen,November 1998.<http://www.jesperju ut.dkltext/DAC%20Papef/"20 1998. htm L>.

Juu[, Jesper (2001). "Games Telting Stories?" Game Studies 1, no.7(Juty 2001). <ht tp: / /cmc.uib. nolgamestudies/0101/uul-gts>.

LeFebvre, Henri (1991). The hodudton of Space.

Lynch, Kevin (1,960). The Image of the Cify. Cambridge: The MITPress.

Munay, Janet (1997). Hanlet on the Holodeck: The Future ofNaffative in Cyberspace. Cambridge: The MII Press.

Ritvo, Harrjet (1998). the Platypus and the l4ermaid, and ltherFigments of the Classifuing Imagination. Cambridge: HarvardUniversity Press.

Rud[in, John (7994\. Commedia Dell'Arte: An Actor's Handbook. NewYork: Routledge.

Thompson, Kristen (1988). Breahng the Glass Armor: NeoformalistFilm Analysis. Princeton: Princeton University Press.

WjLtjams, Linda (1999). Hard Core: Power, Pleasure and "The Frenzyof the Visible. " Berketey: University of Catifornia Press.