February 24, 2006

Dramatic Paidia

by Andrew Stern · , 7:57 pm

I wanted to re-read Jean-Paul Sartre’s No Exit before commenting on Patrick Dugan’s article from a few weeks ago in The Escapist, “An Exit“. I’m glad I did, it’s a fascinating play that I hadn’t read in a long while.

When developing Façade we took direct inspiration from Edward Albee’s Who’s Afraid of Virginia Woolf; we actually hadn’t thought about No Exit. But in hindsight Façade does seem to share as much thematically with Sartre’s play as Albee’s. Façade‘s characters aren’t that likable, more like Sartre’s and less like Albee’s outrageously hilarious George and Martha. Also, No Exit and Façade are both one-act plays. So thanks to Patrick for making that connection.

Patrick points to Façade as an example of paidia — unstructured and open-ended play — and to the Sims as the most commercially successful paidic interactive entertainment. (This is opposed to ludic play, which is your typical rule-based, goal-oriented game.) Patrick surmises there’s an untapped market demand for paidia. (I keep wanting to type paella, but maybe just because I haven’t eaten dinner yet.)

It’s interesting — while I certainly don’t think of interactive drama a la Façade as rule-based or strictly goal-oriented play, it’s certainly not unstructured either like the Sims. Interactive drama, to earn its title and actually offer the player some of the pleasures of drama, needs to retain some degree of economy and efficiency, pacing, tension building, etc. — qualities I find the Sims sorely lacks. (I get paidic enjoyment from the Sims, not dramatic enjoyment.)

What we’re striving for here, with Façade as an early example, is collaborative generativity. In such an experience, the player and the system each act creatively in real-time through meaningful dialog and action. If each party is skillful, a moderately well-formed drama results. Players can form and pursue a variety of goals if they wish. In its ideal form, this is not unstructured play, but cooperative, open-ended creative performance — creating a structured play, actually. It’s less about emergent behavior and more a connection between two creative agents, the human player and the artificially intelligent drama manager, who make and take offers to and from each another, like in structured improvisation.

I see Façade borrowing the open-ended aspect of paidia, but not as much the unstructured part. I think people seek structure in art, interactive or not, and we’re actively trying to give it to them, to build those smarts into the system. At the same time, we want to offer players the open-endedness to say or do anything they want at any time, to support and reward a variety of player goals, for players to not be constrained to operate only within a particular moment’s mini-game or small set of options in a multiple-choice dialog menu. An open-ended interface is one of primary ways Façade distinguishes itself from adventure games like Indigo Prophecy (aka Fahrenheit), Grim Fandango, or command-based IF.

Of course interactive drama can be built to optionally allow players to have a chaotic, unstructured experience. Sometimes that’s what the players want to have, which offers the fragmentary pleasures of emergent behavior, e.g. Trip and Grace attempting to play it straight when players act like zombies or murderers. Screwing around, acting inconsistently or trying to break the system is allowed in Façade, but probably is less rewarding ultimately than actively collaborating with the system. Note, as we mention in “Behind the Façade“, collaboration with the system does not mean you should just agree all the time with Grace and Trip; actually it often means conflicting or disagreeing with them, but in ways that you think might propel the drama forward in productive ways (as opposed to trying to “break it”).

15 Responses to “Dramatic Paidia


  1. mark Says:

    Obligatory pedantic comment: paidia (more commonly spelt paideia, but either transliteration is fine) is Greek, not the Latin that Patrick claims (the ai is the giveaway—compare Greek paidagogos with Latin paedagogus). In English it also, confusingly, already means something else.

    In any case, I prefer the fancy terms “open-ended play” or “non-goal-oriented play” myself. :-)

  2. Adam Russell Says:

    Patrick has clearly borrowed the term from Roger Caillois’ terminology in his book Man, Play and Games. By that rationale, it seems ironic Andrew that you would like to associate Facade with paidia to some extent, but add that “Screwing around, acting inconsistently or trying to break the system is allowed in Façade, but probably is less rewarding “, when others might argue that this sort of behaviour (which we saw *a lot* of when Facade was released, if I recall) is exactly the sort of thing that Caillois was talking about when he contrasted the ludic urge to the paidic! If we wanted to continue with Caillois’ terms, I would have to classify your work as almost entirely about mimicry, with little to no agon, alea or ilinx elements. With the invitation to ‘cooperate’ with the intended mimesis, not to subvert it, as you emphasise here, the whole endeavour also takes a distinctly ludic stance, not in fact supporting paidia as Patrick hoped.

    In summary, Facade’s attempted move towards interactive drama in the tradition of Brenda Laurel’s 1986 PhD thesis comes across under Caillois’ terms as a highly structured act of mimicry, not supporting 90% of what Caillois (and presumably Huizinga before him) would have termed ‘play’ or ‘game’. Perhaps this is why so many traditional computer game enthusiasts and developers seemed to find it frustrating?

    Having said all that, I would certainly agree that Facade is an important work, expanding our notions of interactive media. The heavy emphasis on mimicry to the exclusion of almost all other aspects of play is rare. I’m sure RPG fanatics would reply ‘we’re all about mimicry!’ but in practice the agonistic elements of computer RPGs leave mimicry out in the cold.

  3. Patrick Dugan Says:

    I admit I was a bit high on the term at the time I wrote that article, since then I’ve come to recognize mimicry as more essentail to the new wave and the “reimagined” challenge of interactive storytelling. Open ended play is probably a better term than paidia.

    Actually, a big point of the article, maybe made inexplicitedly, was the structure of the social heterachry, that is, how three equal elements, in this case social actors, composed that collaborative generativity. Emergence, like ludus and paidia, is a problematic term I suppose, but the intransitive relationship is a major building block design pattern that yields dynamics which could be interpreted by users as “emergent behavior”, for instance the Residential, Commercial, Industrial zoning verbs in SimCity, which was definetly a more open game than Facade. The idea that such a basic game design building block can be brought to a social simulation demonstrates how important a benchmark Facade was.

    Interestingly, the wikipediac (wikipaidic?) definition of paidia is like an encyclopedia, hence Raph Koster’s quote of paidic meaning “very big rule-sets”. He later elucidated me as to his meaning, describing paidic rule-sets as imported, for instance in a simualtions model or a table-top RPG’s extensive set of rule books. In Facade’s case, the imported rules are the social moray’s we’ve been dealing with all our lives.

  4. andrew Says:

    Refreshing my memory of the defintion of agon (1 2), it seems Facade has that in spades:

    ag·on (ag’on, -on, ä-gon’)
    1. A conflict, especially between the protagonist and antagonist in a work of literature.
    2. The part of an ancient Greek drama, especially a comedy, in which two characters engage in verbal dispute.

    Agon is the ancient greek word meaning contest or challenge. it is a formal debate which takes place between two characters, usually with the chorus acting as the judge.
    … where the player takes on the role of both antagonist and chorus for the protagonists Grace and Trip. (Grace and Trip are each other’s antagonists as well.)

    Yes, most players screwed around with Facade, versus constructively collaborating with it to make a more well-formed performance. Perhaps it’s too soon to tell whether subverting vs. collaborating with interactive drama will be more common / pleasurable. If subversion was more fun in Facade, my theory is that’s because the collaboration wasn’t supported well enough. Once interactive drama matures over time, I’d hope that players will find constructive collaboration as or more pleasurable as subversion. (But subversion should always be supported and rewarded as well.)

    In terms of interactive drama as mimicry, presumably meant to be mimicry of reality, I have a couple of responses. First, Facade is intended to be drama, which is not realism. It’s not exactly realistic that a married couple would end their marriage only 15 minutes after their friend walked in the door and provoked them with a few bits of marital advice; such accelerated action is dramatic. Drama is reality with the boring bits taken out; doing so skillfully is an artistic act, not mimicry. Two, even if interactive drama a la Facade were considered mimicry, mimicry can be achieved in either representation of reality (a fixed script) or simulation of reality; interactive drama strives for simulation far more than representation, an important distinction to be made.

    Adam, to me your most interesting point is questioning if constructive collaboration is ludic or not. Just because collaboration (between player and AI drama manager) is a creative act with the goal to generate a structure, e.g. a well-formed dramatic performance, it doesn’t quite qualify as a game with formal rules and given goals. Creative collaboration to write/perform a decent story is more open-ended, diverse and playful than what’s allowed / possible within the activity space of most games, I’d suggest. Players can be creative when playing games, to be sure, but the range of allowed behavior is pretty constrained compared to creative writing / improvisation.

  5. Raph Says:

    Ah, those social morays… evil green things hiding in the shadowy parts of culture… you’ll be swimming along merrily when one of them jumps out of a crevice to bite you. They travel in packs, too, being social…

    (sorry, couldn’t resist)

  6. Adam Russell Says:

    lol @Raph

    Andrew, while quite true that Facade echoes many meanings of the word agon, I don’t think it has much to do with Caillois’ term agon:

    “like a combat in which the equality of chances is artificially created, in order that the adversaries should confront each other under ideal conditions, susceptible of giving precise and incontestible value to the winner’s triumph. It is therefore always a question of a rivalry [...] in such a way that the winner appears to be better than the loser in a certain category of exploits”

    Summarised in Rules of Play as: “competition and competitive struggle”

    Now, I have no interest in protacted debates about what term a word should refer to, but I do find Caillois’ term a useful one in describing an important family resemblance amongst experiences people have when they play (or pleasures given by play, as Eric and Katie would have it). Imho Facade strikingly avoids providing this experience / pleasure, to the frustration of many games players approaching it looking for the ‘win conditions’, ‘optimal strategies’ etc. The fact that there is no objective measure of success in Facade is an important signal that there is no experience of agon on offer (in the sense of Caillois’ term only).

    By contrast, many of your descriptions of the intended experience of playing Facade sound exactly like Caillois’ term mimicry:

    Mimicry is incessant invention. The rule of the game is unique: in consists in the actor’s fascinating the spectator, while avoiding an error that might lead the spectator to break the spell. The spectator must lend himself to the illusion without first challenging the decor, mask, or artifice which for a given time he is asked to believe in as more real than reality itself.

    I don’t think mimetic play in this sense need strictly mirror natural life, as you suppose. It seems perfectly desirable that mimicry conjur up unlikely or contrived situations, or that it create carefully arranged situations, as in drama. The only desired constraint on the ‘open-ended play’ of Facade’s free text interface is that we try to keep it ‘in character’, not that we make it ‘plausible’, whatever that means. This is quintessential mimicry in my mind.

    Sorry I haven’t actually addressed the ‘interesting’ ludus/paidia question at all in this post… :)

  7. andrew Says:

    Adam, yes, you’re right about Caillois’ uses of those terms — pretty different than the standard definitions of those words! I’m glad our discussion ended up defining each use of the words, that’s instructive.

    I’ll be curious if you or others have any thoughts on to what extent collborative writing/performance is ludic. Again, my take is that it’s open-ended like pure play, and therefore unlike games — but with the goal to create a reasonably well-formed structure (a drama), that pure open-ended play does not require. Further, this dramatic structure to be created during play is sort of “apples and oranges” when compared to the pre-supplied explicit rules and goals (the structure) of games.

    In other words, (collaborative) writing is neither pure play, nor game.

  8. Patrick Dugan Says:

    I think Adam made a good distinction about Mimicry and Agon.

    Ludus and Paidia exist on a continuum. Facade is more paidic than %90 of the games on the market, the open nature of the primary interface ensures this, though less so than The Sims or maybe even Utopia. I’d say its more paidic than Storytron, based on the fact that Facade’s interface involved user-provided text, and Storytron depends on a fixed logographic dictionary.

    I think you need the encyclopedic nature of paidia in order to support mimicry, which is the essence of social challenge. The difference is that interactive drama requires active creation of belief in the user; while the limited information channel of linear media makes passive suspension of disbelief the best possiblity, interactivity allows the exploration of data-bases of information. Mimicry requires giving the user lots of information in order to “fill-out” the role they are given. The benifit of paidia, of importing information and cultural references and rule-sets, is that you can have your dramatic conflict resonate a challenge that lots of people can relate to, but you don’t have to come up with all of the content explicitedly by yourself.

    For example, in Star Wars Galaxies or WoW or any of these other agon dominated games trying to encourage mimicry, you use a franchise or lots of money and time to come up with loads of background, history, fictional culture and geography and let people disseminate all that in playing a Jedi or a level 60 Orc. But nobody really cares about the mimicry in those games, because the agon IS the game. In Facade, I would argue, the primary play loop is a mimetic one, and Facade was able to sustain this with only so much embedded content. The rest of the active creation of belief was supported by the notion that you are having a real conversation with these people and you could talk like you do to anyone else. This wasn’t perfectly feasible, but a more powerful form of langauge processing and improvisational AI would make that sensibility robust instead of an illusion.

    As I understand it, Facade isn’t really agon in the sense that there is no explicit competition, on which agon depends. You really “won” the drama after you figured out how to get these people to communicate, at which point the conflict is difused. Collaborating with the drama subverted the conflict. Hence the exit, thats social challenge, thats the exit from agonic ludes, or goal-oriented competition, a form which caters to a quarter of our possible audience, yet has dominated the landscape of games for decades.

    But maybe you can have a mimicry ludus. Maybe ludus and paidia are one axis, and agon and mimicry are another. Its the circle of play.

  9. Adam Russell Says:

    Patrick, I certainly see ludus/paidia as one axis, and agon/mimicry as another. Caillois even drew a table with agon/alea/mimicry/ilinx shown as orthogonal to ludus/paidia.

    Andrew, having revisited Caillois’ text, I am prepared to revise my rejection of paidia in improvisational drama :)

    “a primary power of improvisation and joy, which I call paidia, is allied to the taste for gratuitous difficulty that i propose to call ludus, in order to encompass the various games to which, without exaggeration, a civilising quality can be attributed.”

    Having said that, on the ludus-paidia column under mimicry in his table of classifications, Caillois states ‘children’s initiations’ and ‘masks, disguises’ at the paidia end of things, and ‘theater’ and ‘spectacles in general’ at the ludus end. Perhaps the best way to approach Facade under this question then, is not to examine the work standing by itself (‘is it ludic?’), but rather to employ Michael’s conversation metaphor as in his papers on expressive AI. The dialectical negotiation of meaning certainly allows for the artist to ‘attempt to influence this negotiation’ as Michael put it, but it surely also allows for the audience to influence the negotiation also? Given this, then we can make sense of there being a range of modes of play in which Facade can be approached by an audience – some more ludic than others. To adhere very stringently to the authors’ vision in the piece and attempt to smoothly achieve one of the (very finite number of) preauthored outcomes, seems to me a ludic mode of engagement with Facade. At the opposite extreme of excessive paidia would be the attempts at subversion discussed previously. But any simultaneously exploratory and cooperative engagement with the work will wander somewhere between these poles.

    Continuing with the expressive AI approach, one might say that different works afford more ludic or more paidic audience engagements. It seems fair even to apply this idea to different features of one work. The free text interface in Facade for example, affords plenty of paidia if the audience wishes to engage in this manner. The underlying drama manager and its finite selection of preauthored beats, on the other hand, affords less paidia and better supports ludic engagement.

    Does this make any sense?

  10. Patrick Dugan Says:

    I think I see what you’re saying, to borrow a bit of theory from another of Micheal’s papers, you could say the material constraints of Facade are open and paidic, while the formal constraints are very ludic, to the point where curve-ball syntax in the user entry will prevent the semantic from being recieved as a valid “move” in the drama.

    Going with the circular concept of agon/mimicry and ludus/paidia (I’d argue illinx and alea are “spin” attributes that can both be applied in different ratios to different points in the circle) its possible that ludic mimicry requires a paidic space to support the sense of agency, and that a good interactive drama, one with a larger scale structure than a one Act play, will fluctuate between both. As things approach a climactic event, the discourse becomes more ludic and global agency becomes maginified in immediate descisions. In between plot arc climaxes, things grow more relaxed and paidic, and local agency becomes more expressive while being less influential at the global level. This freedom sets the stage for the next build-up, with an accompanying tightening of focus. Facade seems to follow this pattern in a single plot arc, so that at first things are legitimately more paidic (you can approach the couple through a host of different topics and angles) and become more constrained and ludic as the arguments start flying.

  11. mark Says:

    In response to Patrick’s (first) comment: The use of paidia/paedia as part of the word encyclopaedia doesn’t have anything to do with large rule-sets, but simply the more generic meaning of paedia as education, particularly education suited to the elite (a Roman aristocrat might display his paedia—as opposed to mere wealth—through his ability to engage in philosophical debate, for example). An encyclopedia is certainly useful in that pursuit (hence the name), but it hardly defines it.

  12. andrew Says:

    First I should point out for posterity’s sake, the earliest (first?) use I know of a conversational metaphor for interactive art/entertainment was Crawford (1993) “A Better Metaphor for Game Design: Conversation” in his journal Interactive Entertainment Design. (I riffed on that in a 2001 paper.) Also, a better link for a Michael document discussing the conversational metaphor is his dissertation (the DiGRA paper linked to above doesn’t have it that I can find).

    Adam wrote, To adhere very stringently to the authors’ vision in the piece and attempt to smoothly achieve one of the (very finite number of) preauthored outcomes, seems to me a ludic mode of engagement with Facade. … The underlying drama manager and its finite selection of preauthored beats, on the other hand, affords less paidia and better supports ludic engagement.

    Hmm. In a collaborative writing situation between human player and AI system, if the player knows beforehand the system only has a finite number of story endings it can produce, I suppose the player could try to “game the system” to force one of them, which is a sort of ludic meta-game. But in Facade’s case, that’s an artifact of the relative simplicity of the AI; I’m not sure that helps us answer my larger question, is collaborative writing ludic. If the system were a bit more generative with 10x or 100x as many variations/endings, could you still treat it as a ludic?

    Patrick wrote, while the formal constraints are very ludic

    How so? Again, I see dramatic writing, while structured, so open-ended that it’s hard to frame it as ludic.

  13. Patrick Says:

    Mark: I think the different meanings tie together. Robust paidia, as in Second Life, requires lots of rules, including meta-rules, and is in that expansive sense both encyclopedic and learning inductive. (All games are learning inductive, as Raph thesized in his book, (clever joke Raph) but maybe ludic lessons are more direct, like a worksheet, and paidia is like surfing hyperlinks on wikipedia.) You start in Second Life totally clueless, fumbling with the interface, then you learn how to buy or make your own clothes, home ect. Then you figure out basic scripting and eventually you might make a replicating bot army that crashes the server.

    Andrew: A sentance from your “Deeper Conversations” paper cues me in on something:

    “Computer games are currently the dominant interactive format — not because they are goal-oriented experiences, but because they are closer than any other format to fully using the capabilities of the medium.”

    If we’re defining ludic as more strictly goal-oriented, and paidia as allowing the user to bring their own goals (be they aesthetic or achieving a particular game-state) then Facade seems to be aiming for the middle. Whatever success it achieves in this regard comes from harmony between its material and formal constraints. Now that I think about it however, while the open parser seems like absolute paidia, its constrained by the contextual language cues you implemented in your DM and agents. So, Facade is in a sense a ludic game about inferring and manipulating a constrained set of sociolinguistic codes, to what end is up to the player, and this importing of goals is what gives Facade its paidia, not so much its interface. For example, a player can try to sleep with Grace, cajole a threesome, be Darth Vader ect. Thats raw paidia, though largely unsupported, the ludic formal constraints heavily bottle-neck the workable goals. But the fact that the user had to infer their own goals, from a system which essentailly references the entire social symbolic scope of interpersonal experience, puts Facade in an entirely new design category. If we had langauge and contextual processing robust enough to support more generativity, then we’d give people this more personal goal-orientation while also utilizing the potential of the medium.

    I’ve looked at three different drama-engines, Facade’s, Storytron and Utopia’s, and what strikes me as common to the three is that the “gamplay” or “storyplay” depends on the user inferring their own goals. Thats the exit. Maybe I shouldn’t have dragged Callois into this and focused on the main point: we need to facilitate people’s dreams, not make them run around a wheel for a pre-rendered carrot.

  14. andrew Says:

    Here’s a long piece by Chris Bateman on ludus and paidia.

  15. Time Travel: What Wave and Braid have in Common Says:

    [...] the player through unanticipated time bending experience via its paidia during play, minus the dramatic agency of actual time [...]

Powered by WordPress