Multimedia is not, as so many have suggested, an art form in its infancy.  In the opinion of this writer, the medium is as yet unborn, and the efforts we've seen thus far are mere kickings in the womb.  The day the first consumer is moved to tears by an interactive experience will be the day of the medium’s birth.  It will draw its first breath when the first person is compelled by a multimedia program to re-evaluate the meaning of his or her life.

If this seems too lofty a standard, it will be admitted that every other significant art-form has met it time and again.  Had they not, art would not be the mark of culture that it is, and we would not hold art or artists in the high esteem we do.  And by this measure it is clear that the multimedia industry is thus far fetal, and further, that it is yet in danger of miscarriage.

Certainly the computer brings dazzling "new functionalities" to traditional media.  Linking and indexing tools, and live-action media clips make encyclopedias, and all sorts of non-fiction and educational works more useful and meaningful.  And  time-driven activities, requiring real-time interactions, make games more exciting.  But games and encyclopedias have never been particularly compelling, psychologically or spiritually, and the new media doesn't change that.

Any industry insider will readily admit - in private - that our offerings to date have fallen far short of our hopes, and of our hype. Yet widespread certainty of the potential of the medium remains undiminished, even among, if not especially among, insiders.   To become viable, our industry must now come to understand and actualize the potential of the new media to express profound emotional and spiritual ideas.

Interactive narrative is the Holy Grail of the new-media industry.  It is through narrative - through drama - that humanity explores, and seeks to understand the workings of the human psyche:  our aspirations, our short-comings, and our contradictions.  Interactive narrative is the industry’s Grail because of the great promise that it holds, and because it has unexpectedly proved to be elusive.

So we must ask ourselves what does "new functionality" offer to the process of storytelling?  What new tools for the elucidation of theme does it provide?

In this article I will offer a number of reasons for why our attempts to tell stories interactively have failed.  I will then propose an expanded dramatic paradigm which puts the new functionality of the medium to the service of thematic expression.  But before proceeding further, it would be helpful to define a few of the terms we will be working with.
 
 

Because our industry is new, and highly technical, and driven for the most part by large, market-oriented, publishing corporations, multimedia jargon is developing rapidly, and often in directions which run counter to our artistic purpose.  Indeed, one cannot be taken seriously in this business without frequent references to "content development", "titles", "new-functionality", and "media assets".  Terms such as these detach us from our primary goal which is to create and communicate.

"Title," for example, is a marketer's term.  It clearly insinuates little interest in the "content" beyond the title given to it.  It implies that the program is just one among many titles which a marketer has amassed to a product roster for no purpose more constructive than to build market share and to crowd out competition for channels of distribution.  "Titles" are not presumed to move people;  but rather, like the ever-expanding multitude of variations of soda pop, they merely take up retail shelf space.

And when we refer to the substance of our creations as "content",  we sound as if the program itself, beyond the title, was only an afterthought.  We need to remind ourselves that our purpose is to create media which will move people emotionally.  "Content" is what one expects to find in a can of soup.

And nowhere in the lexicon of the new-media industry is our cluelessness made more plain than in our naming of the person with whom we are trying to communicate - the "user"…

Before this nomenclature becomes entrenched, allow me to propose a couple of alternatives.  Our terminology should emphasize interaction above all else; not just the interaction between human and machine, but more significantly, between artist and audience.  The term "multimedia" is today used interchangeably with "interactive media", but from the perspective of the consumer, multimedia without interactivity would be indistinguishable from television.

I propose to refer to interactive programs as "interplays", and to the person interacting with them as the "respondent."   The use of these terms will help to keep us focused on our actual product and purpose, and I will use them throughout the rest of this article.
 
 

The paradigm for drama is as well established as any of our most basic mental constructs.  By contrast, logic almost certainly has a shorter history than storytelling.  Human beings recognize the pattern of conflict, crisis, and resolution as a meaningful way of organizing experience as readily as we recognize food for food.  Drama is innate to us.  The principles of drama were not invented by Aristotle, as many believe, but were merely codified by him in the 4th century BC, long after they were established.  Aristotle's accomplishment was to categorize, and to record for posterity, the elements he consistently observed in every effective dramatic work.

And what are those elements?  In brief, the dramatic principals identified by Aristotle are that a theme may be expressed by the movement of a character through a plot, and that a plot is the interaction of characters, or forces, in conflict.  The hero’s quest to overcome the antagonist is marked by complications, and reversals of fortune, and an overall escalation of tension and suspense leading to a crisis, and further, to a climactic contest which determines the nature of the conflict’s ultimate resolution. Audience identification with the hero’s struggle produces an emotional catharsis, and new understanding.

With the technological advance that has brought us a new medium will naturally come a time of great experimentation with dramatic form.  But we dispense with the basic paradigm at our peril.
 
 

At present, there are two prevailing conceptions of interactive storytelling which are opposite to each other, and mutually contradictory.

The first holds that the respondent must be given maximal control over the progress of the plot, and complete freedom of access to the separate elements of the story.  This approach inevitably fails because random access to story elements causes them to be arbitrarily juxtaposed.  It thus obscures the clarity of causal links, reveals the outcome of what otherwise would be suspenseful, and destroys the dynamic escalation of conflict that makes resolution meaningful and releasing.  Interplays of this type are usually organized by a spatial, or architectural metaphor.   The 7th Guest is an example.

Art, it is said, imitates life.  But, of course, art, unlike life, is consciously ordered for a meaningful purpose, and in this respect, it diverges spectacularly from reality.  In life, two things may be next to each other without expressing any relationship between them.   Two houses may exist side by side, and no particular meaning may be inferred from the fact.

But two paragraphs are presumed to follow from one another in a meaningful way.  In a unified work of art, things juxtaposed are presumed to be related, either by chronology,  analogy, cause and effect, emotional dynamics, or what have you.  Juxtaposition implies linkage.

So if scenes in an interplay can be sliced and diced, and randomly reordered, then any inference to be derived from their interconnectedness crumbles.  A ramshackle collection of interesting but unrelated events cannot be called a story, any more than a pile of bricks may be called a building.

The opposing school of interactive design holds essentially that interactive stories work to precisely the degree to which they are not interactive.  The respondent is given narrow options for control, and is allowed to choose one among perhaps eight, or ten, or a dozen outcomes to a basic story.  These interplays typically exhibit a more fully developed plot than those of the random-access school, but give little emotional payoff for the extra effort and distraction of interacting.  An example is Silent Steel.

Interacting, in itself, makes multimedia inherently less seamless as a medium than film or television.  What subtitles do to one’s concentration and involvement in a foreign-language film, interface tools and queries do more so in an interplay.  The "willing suspension of disbelief" which is so essential to the dramatic experience - the feeling of being swept up by the story - is interrupted by interacting.  So unless meaningfulness is greatly enhanced through the encumbrance of interaction, stories are better expressed linearly.

The basic misconception underlying both schools of thought, and which has led to the failure of every attempt at interactive narrative thus far is the idea that the greatest value of interactivity is that it "gives control to the user", or worse, that it allows the user to "write the story."  This is ridiculous both in theory and practice.

There has yet to be an interplay in which the path taken by one user cannot be repeated exactly, branch by branch, by another.  The fact that every path must be predefined should be proof enough that the "user" is not the creator.  It is ironic and insulting that we would imagine that the user is inventing the story, when in reality the preparation of a typical interactive narrative requires the writer to create a dozen or so closely interrelated stories, instead of only one.  Many interactive scripts exceed a thousand pages.  The user is supposedly doing the work, yet the burden upon the writer has exponentially increased!

Great stories invariably take us where we don't want to go.  This is precisely why they are described as "compelling".  If the audience had controlled the story, would Oedipus ever have screwed his mother, killed his father, and put out his own eyes?  Who would be moved by the pap that we would create for ourselves?  And what meaning could such a solipsism carry?

Great art is an intimate communication between an artist and an audience.  "Control" and "choice" may be wonderful (and quintessentially American) social values, but they are the antithesis of artful communication.  Art depends upon the willingness of audiences to give themselves over - at least momentarily - to an artist’s vision.  When we appreciate a work of art we relinquish control, take a leap of faith, and entrust our psyche to the person who would communicate with us.

And this requires no small act of trust.  Consider the photographic work of Robert Mapplethorpe.  The commitment required of us to look at a photograph may at first seem trivial.   But an image may exalt, or it may disappoint us.  Or it may so deeply shock us that even in turning away we are not spared the persistence in our minds of the disturbing vision.  This is why art so often inspires outrage in people who are ideologically brittle, because, having entrusted themselves to the artist, they feel betrayed.
 
 

There is a third approach to the design of interactive narrative which is prevalent of late.  This, in essence, is to create games, and to call them stories.  There are some who argue that gaming is the future of storytelling.  This is bunk,  but it is not easily dismissed because, to date, game-makers have exploited the functionality of multimedia far more successfully than storytellers have.  The current trend is to blur the distinction between the two entertainment forms.  This approach has yielded limited results because games cannot be used to express emotionally relevant themes.  Understanding the difference between games and stories is absolutely critical.

Games fix a player’s attention on a goal, or on a series of goals.  Stories, by contrast, fix our attention on the unfolding of a process - on the growth of a character, or the evolution of a relationship between people or forces in opposition.

Games give tasks to the player, either mental or motor, and when we play a game, we are the central actor.  Games test us.

Stories give tests to a character with whom we may identify, but who we know that we are not.  They force us to contrast and compare ourselves - our values, abilities, and foibles - with those of the character; to acknowledge our commonalties with the character in the face of our presumption of our otherness.  Characters interest us precisely because they are not us.

Game designers set the parameters for activity, but the locus of action during play is within the player, which is why playing games may provide a sense of personal achievement.

In narrative, it is the storyteller who initiates the action, and who animates the characters, which is why stories enable communication.

Communication - if it is intimate and heartfelt - will evoke in us an emotional response.  A story may elicit from its audience the full range of human emotions.

Games, by contrast, evoke a very limited range of feelings.  These include excitement and curiosity, pride in accomplishment or frustration at failure, and very little else.

This describes the typical reaction to the game Myst, which is the cleverest effort yet to blur the game/story distinction.  Myst is a game, the object of which is to determine what the story was.  Its creators designed a story, the action of which is supposed to have left artifacts which we must gather together, and from which we, like archeologists, must reconstruct the plot.  It is exceptionally intricate, and astoundingly beautiful, but it is nevertheless unmoving.
 
 

Clearly, interactive narrative demands the application of an expanded dramatic paradigm.  What follows are six aesthetic principles by which the New-Drama will be informed.
 

The First Principle:

The emotional impact of an interplay depends upon the maintenance of its dramatic throughline.

Regardless of the order in which the events in an interplay are presented to the respondent, tension must escalate by means of conflict, toward a crisis, and ultimately to a resolution.  The order must sustain that emotional dynamic.  Any program so loosely designed as not to maintain a dramatic throughline will fail to move our emotions.
 

The Second Principle:

Structure carries meaning, and must correspond with the theme.

The architecture of an interplay conveys meaning, and as such, must correspond with the intended theme.  Because the structure of a non-linear narrative is necessarily complex, and because the respondent must consciously navigate through it, the respondent’s attention will be drawn to the implications, and metaphoric resonance, of its particular construction.  Therefore, that structure must harmonize with all other story elements to elucidate the theme.

In any effective narrative, linear or otherwise, the structure relates to the theme.  If, for example, we wish to express the idea that over-ambition leads to short-term achievement, but also, inevitably, to long term destruction, we could make use of the standard rise-and-fall structure of classical tragedy.

Other structures suggest other themes.  Consider the storyline of the movie Jagged Edge, in which a woman trial-lawyer falls in love with the accused killer whom she must defend.  Their deepening romance is impinged upon by her vacillating doubts as to his guilt or innocence.  The murder weapon was a jagged-edged knife, and the storyline takes us through a series of sharp emotional jags, back and forth, between her loving certainty of his innocence, and her terrified certainty of his guilt.

Akira Kurosawa’s famous film Rashomon expresses the theme that a singular reality may evoke multiple "truths"; that different witnesses may perceive the same event in radically different ways which are not reconcilable.  To this end, Kurosawa recounts the story of a rape and murder, from beginning to end, four times in succession, from four points of view.  Rather than intercut the separate perspectives, as would be the common approach, Kurosawa repeats the complete dramatic form - from conflict, to crisis, to resolution - for each witness.  He keeps their differing points of view entirely separate, and so emphasizes their incompatibility.

The dynamic structures of interactive storylines provide significant advantages over linear structures for the elucidation of thematic meaning.   But to tap this capacity, new-media designers will have to control the respondent’s control more playfully.  The current industry ideal, that the respondent must feel that (s)he is inventing the storyline, is not only unobtainable, it severely limiting artistically.

Before proceeding to the other principles, let’s develop some specific examples of the special expressiveness of interactive structures.

Structure Example 1:   Forced Loop.

Suppose we want to create an interplay examining the never-ending quality of certain arguments between spouses.  We could present a marital fight, from both the husband, and the wife’s point of view, and let the respondent make choices for both of them.   At key points in the argument the respondent could choose to escalate, or de-escalate the disagreement, or to take it in a new direction.   Now suppose that we construct the program so that regardless of what choices the respondent makes, no matter how far (s)he follows out any particular branch in the line of the argument, the discussion always loops back to the original point of contention.  Such a design would send the respondent looping around and around, searching fruitlessly for a way out, and would give her a direct experience of futility and frustration that would put her in close identity with the characters.  Would that not say more about the inextricable circularity of certain marital fights than is possible in a linear medium?

Structure Example 2:   Indifferent Program Response.

To go further with this, suppose instead that our subject is domestic violence, and that our intention is to express the sense of impotence that victims of spousal abuse typically feel.  Here, the hypothetical argument erupts into a beating.  Interactivity, and point-of-view, could be used to sharpen sympathy for the victim.  But to make the specific point about the wife’s sense of powerlessness, we could design the interplay so that after every choice the respondent makes for her, no matter whether to defend herself, or to placate her husband, the beating intensifies.

Structure Example 3:   Unavailable or Disabled Options.

Now suppose that we want to point up the psychological interdependence of the abuser and the abused.  When the beating ends, the respondent, acting on behalf of the abuser, could be given three options:

1. Tell her to "Get out!"
2. Act like nothing special just happened.
3. Beg her to stay.

Structure Example 4:    Multiplexed Node.

By contrast, if we wanted to make the opposite point, we could present the victim with a list of 15 or 20 options, any one of which would be superior to agreeing to stay.

Structure Example 5:    Imaginary Option.

As a simple way to express the victim’s feelings about her predicament, we could give the respondent a button to press which would make the abuser burst into flames.
 

Some might object that the third example above purports to be non-linear, but that only one of the options presented may in fact be enacted.  And that’s correct.  In fact, four of the five techniques just postulated make use of constrictions to non-linearity to make their emotional point.  In the context of an interplay that is presumed to be non-linear, a narrowly linear node becomes especially meaningful, and the more so if it purports to be broader than it is.  And even the narrowest of non-linear interactive structures can be more meaningful than a structure which is not interactive at all, precisely because the respondent does not have the option not to respond.
 

The Third Principle:

The most significant source of meaning in interactive media is not its non-linearity, but rather its non-passivity.

The distinctions between the example structures given above may at first seem subtle, but they are significant.  They all exploit the tensions between the respondent’s apparent control, and actual control, and between her ability and responsibility.  But they don’t all share the same meaning.  Mixing and matching the proposed techniques with the various plot points would drastically alter their emotional impact.

We must more playfully control the respondent’s control.  Perhaps the reason interactive media has thus far exclusively employed the straightforward and rather banal notion of giving maximal and predictable control to the user is because interplay designers do not yet appreciate the distinction between sound application programming technique, and the art of storytelling.   Hence the term "user."

A well designed application, such as a word-processor, provides its user with positive control, and maximum flexibility.  The appropriate input must be easy to choose, and must elicit a direct and predictable response from the program.  Movement within the architecture of the program, to access its various capabilities, must be as free as possible.

But art is rarely so utilitarian.  In interactive art, a random response to an input may be the most meaningful.  So too might be blind choices, or unresponsive buttons, or buttons that do the opposite of their labels.  Freedom of movement within the branching structure may be thematically relevant, or it may be self-contradictory:  in art, as in life, some choices are more significant than others precisely because they are not reversible.  And the most profound and character forming choices are often those in which every option we are given leads to loss or destruction.  Recall, for example,  the emotional impact of Sophie’s Choice.

In addition to structural considerations, the examples above also illustrate the two most potent tools of interactivity in narrative - that it intensifies both the first and third-person perspectives, and polarizes them.  The mechanisms by which it does so are made plain in the following two principles.
 

The Fourth Principle:

The respondent’s identification with a character arises from sharing commensurate options and capabilities.

Of course, a character may incorrectly assess the options which are available to him.  He may ignore, or fail to consider certain options.  Or he may fantasize having greater power than he possesses.  All such possibilities can be revealed to the respondent by presenting her with options and abilities which are congruent with those of the character.

This is the first-person perspective of interactive storytelling.  The technique serves the same purpose as the "point of view" shot in film,  but takes it further.  By the use of the first person, the respondent comes to know the character more intimately.  (S)he is compelled to share a sense of responsibility for the character’s actions, and to deepen her emotional investment in the character’s fate.
 

The Fifth Principle:

The more control the respondent has over the storyline, or the more freedom to move within its architecture, the greater the respondent’s omniscience, and detachment from the characters.

The respondent’s omniscience - her detachment from the "reality" of the story - is determined by the degree to which the design empowers her to move freely within the story’s structure, or to alter it’s structure.  Interactivity affords storytellers the opportunity to show causal relationships, which the respondent may explore by playing out alternative scenarios and contrasting the outcomes.  Interplays can examine, for example, how differently a character would behave if given different circumstances.  Or how different characters would behave given the same circumstances.

This is the third-person perspective of interactive narrative.  The method removes the respondent from identity with individual characters and allies her instead with the narrator whose distance from the story allows for a detached perspective on its interpretation.

For example, consider the end of the film BraveHeart in which Sir William Wallace is tortured to make him confess to treason against the English crown and beg forgiveness.  Wallace refuses to submit, and is beheaded.  But suppose the film was interactive, and that we could return to the point of Wallace’s decision, and chose instead for him to confess, and beg.  Then, if he were killed, despite having begged, we would be compelled to consider the difference between the two endings.  Either way he dies, but in the first scenario he keeps his dignity, whereas in the second, he loses it.  The meaning which this interaction would convey is intellectual in nature - it is an exploration of the relationship between a cause (begging) and an effect (keeping or losing one’s dignity).  It would not invite us into identity with Wallace’s passion and righteousness.  In fact, the "reality" of his story would be undermined by this technique because that reality would be shown to be fundamentally fungible - so unreal that it can be replayed, and re-done differently.  The usefulness of omniscience in narrative cannot be understated, but the technique exacts a price in loss of empathy.
 

The Sixth Principle:

The designer’s theme is communicated through an active counter-engagement between the respondent and the interplay.

An effective interplay engages the respondent in a kind of Socratic dialogue with its designer.  It employs a system of thematically expressive querying in which non-interactive material frames "questions" which are openings for interaction.  Hence the term "respondent."

Specifically, meaning is conveyed through a synergy between three elements:

1. the content of the non-interactive segments presented,
2. the form of the options given for a response, and
3. the nature of the program’s reaction when an option is selected.

1.    The non-interactive, linear segments are constructed in accordance with all of the established principles of drama, and are presented to the respondent exactly as they would be in film or television.

2.    When the story reaches a node for the respondent’s interaction, the form in which the option to interact is presented - what options are made available, how those options are labeled, what type of interface "tool" is provided, etc. - determines the thematic meaning of the node.

One must recognize that the proposal of an option, in itself, imparts meaning.  For example, Abbie Hoffman’s primer on radicalism, "Steal This Book", is more famous for its title than for its content.  For a would-be reader in a bookstore this unexpected invitation to steal rather than buy alters the whole social paradigm.  It proposes a new societal standard of behavior, and forces us to question the authority of the old.  And, of course, this directly illustrates Hoffman’s theme.  Even if the potential buyer forgoes the option to steal, (as most do) and either purchases the book like any other, or rejects it, that experience will be altered for that person, and in the direction Hoffman intends. Not stealing the book, in the face of its title, will necessarily speak worlds to the purchaser about her own attitudes towards society’s values, towards property and the propriety of stealing, and about her reverence for, or fear of, the criminal justice system.  The invitation Hoffman poses makes the thematic statement:  "I'm radical.  So how radical are you?"

3.    The nature of the program’s reaction to the respondent’s input may take a number of different forms:

• It may be logical and direct, and provide a predictable reaction.
• It may be random, and wholly unpredictable.
• It may be proportionately weighted so as to provide a response according to predetermined odds.
• It may be fuzzily logical, altering its reaction in accordance with a pattern in the respondent’s interactions.
• It may be nonsequitur.
• It may be antithetical, and deliberately do the opposite of whatever the option was labeled to do.
• It may be non-functioning, providing no response at all.

We can see then that clarity as to the respondent's control, or lack thereof, is essential to meaningful interactivity.  Developers have long been tempted to create interplays in which at least two people interact with the computer simultaneously, and, via the computer, with each other.  (There already are health club fitness machines where one stationary biker "races" another on a virtual track.  And, of course there are the Multiple User Dungeons, and such, on the net.  But these, remember, are games and not stories.)  From the analysis just presented we can predict that any such attempt to link multiple respondents within an interactive narrative will inevitably fail.  Interplays for multiple users would be inherently unsatisfying because the clarity of each user's control, or lack of control of the storyline would be muddied.  For interactivity to be comprehensible, each participant must be able to relate his or her specific input to a specific response from the program.  This is why interplays are more isolating than movies or television.
 
 

 The foundation of the new dramatic paradigm is an ethic of participation, as opposed to control.  The basic idea is that "interaction" is not synonymous with "control", in art, or in life.  Interactivity provides extraordinary new methods by which a dramatist may sharpen our identification with characters, and broaden our understanding of a dramatic situation.  But, if these tools are thoughtlessly or haphazardly applied, they will create only an artless chaos.

Applying the six aesthetic principles above, here is the prescription, in brief, for the design of an emotionally effective interplay:
 

• Keep the dramatic throughline.
• Build meaningful structures which reinforce the theme.
• Compel the respondent to make active, emotionally meaningful choices, regardless of the breadth or narrowness of the non-linear structure.
• Unify the respondent’s feeling of control with that which the character feels so as to compel identification.  (first-person)
• Empower the respondent to change the storyline to afford a detached perspective.  (third-person)
• Engage the respondent in a thematically expressive dialogue of query and response.

There is a question looming over our industry, "Does the public want to interact?"  Will they consider interacting to be entertaining?  Or would they simply prefer to be acted upon?
 Studies have shown that for as much as one third of all the hours that the average television is on in the average home, it is not directly attended by a viewer.  The TV is used as a background to other activities such as eating, or housework.  Clearly, interactive media has no inroad to make in to that territory, since it requires at least some attention and input.  But certainly not all television is used as background. Sometimes we want to zone-out.  At other times we crave something more engrossing.

I submit that the public will want anything which provides a more intense sense of being communicated with, anything which gives new insight or understanding, or which moves them deeply.  A well-designed interactive narrative has unique capacities to deliver these experiences.
 

Copyright 1996 Jeff Goldsmith
(510) 530-0358
4665 Dolores Ave.
Oakland, CA  94602
business1@expressive.info