To understand how to develop interactive fiction you should
not only look at currently extant interactive games as models,
but you should also consider the properties inherent in -- and
pleasures obtained from -- interactive fiction's parents: specifically,
the dramatic narrative (including theater and film) and the
written narrative (including novels, stories, and poems). Of
those properties, the nature of the narrative voice, the nature
of fictional time, and the nature of the reader's emotional
engagement seem particularly important in producing the peculiar
pleasures of those arts.
Part One -- Games and Narratives
Written narratives almost exclusively present themselves in
one of two voices: the first-person narrative ("Midway in our
life's journey, I went astray from the straight road and woke
to find myself alone in a dark wood..."), and the third-person
narrative ("Mrs. Dalloway said she would buy the flowers herself.");
the second-person narrative is seldom employed in other than
an occasional literary experiment. Dramatic narratives employ,
perforce, the third-person narrative almost exclusively, whether
it is a play (and its players) presented before, and hence,
separate from, its audience; or whether it is a film, where
the camera represents the invisible third-person narrator.1
By contrast, the second-person is the narrative voice most often
employed in the interactive adventure game ("You are standing
at the well-house of a small spring. . .").
More exactly, the interactive game takes the form of a second-person
conversation rather than a narrative as such, a series
of descriptions provided by the game which are interspersed
with (usually telegraphic) imperatives supplied by the player:
>You are standing at the well-house.
[YOU] DESCRIBE [THE WELL-HOUSE TO ME]
>It is the well-house of a small spring. There is a metal
grate at your feet. There
>are some keys here.
Read over after the fact (in those games that are text-based
and allow the player to capture the game moves) the interactive
conversation does form a sort of narrative, but the narrative
experience is really a secondary one, and not an essential component
of the game itself. Those games that are more visual and auditory
(e.g. Myst, The Journeyman Project) follow a similar
dynamic: the images and sound presented to the player are direct
mappings of a second-person description ("You see this. . ."),
and the commands given to the game (mouse-clicks, menu choices)
are direct mappings of second-person imperatives (e.g. a click
on the left side of the screen = "You, there, show me what is
to the left of my current position. . .").
The opposition of conversation and narrative is directly related
to the nature of time as experienced and conceived by the player/reader.
Narratives presuppose a flow of time, a fate -- even though
the order in which the specific events occur in narratives may
be given out of chronological sequence (e.g. Catch-22),
or told and retold from different points of view (e.g. Rashomon).
But no matter the time scheme and shifting viewpoints employed,
narratives are fixed in time: when one is reading a book from
cover to cover, or watching a play from opening to final curtain,
the events occur in an order that is out of one's control: the
author is the timelord. The audience is bound to ask, and stays
interested only as long as it feels compelled to ask, "What
happens next?"
The interactive game, by contrast, establishes its control over
space. The player explores a game-space in a sequence of actions
that is only partially predetermined: one may encounter the
thieving dwarfs in the Colossal Cavern before or after one has
vanquished the snake. Although the configuration of a game's
space can change over time or as a result of the player's moves,
it is still, when all is said and done, a finite space. Game
time is more malleable -- and more uncertain: there are
few interactive games that do not rely on random-number generators
at one level or another. The interactive-game player does not
ask, "What happens next?" but "What do I do next?"
By taking time out of the reader's control, the narrative promises
a certain satisfaction and comfort to its audience. Closure
of some sort (even in many modern and technically open-ended
narrative works) is assured: though the fate of some characters,
and, possibly, the significance of the events portrayed, may
be left in doubt, narrative works do end -- and, as importantly,
they do end predictably, by the closing of the curtain, or by
the reader's reaching the book's final page. Closure, in that
sense, is guaranteed to all; or, if not guaranteed, is within
the reach of all. The audience can reach the end of the
work by dint of sheer endurance if nothing else. And if the
artist's craft is sound, and the particular reader/audience
no more than normally perspicacious, the main effects intended
in the work can be achieved: sorrow, laughter, fear, regret,
hope, despair, inspiration, confusion.
In an interactive game, however, where time is fluid and its
course dependent upon the player's actions, there must be a
goal (or why bother playing?), and it is not for all players,
or even most players, to achieve that goal. Hence, the more
common emotional effects of the interactive game are frustration
and boredom. The duration of the game and the sequence of events
depends on the actions of the player rather than on the art
of the author, and the randomness inherent in most interactive
games exacerbates the temporal indeterminacy: one can be stuck
in the twisty little passages of the Colossal Cavern for hundreds
of moves, and the length of time one is stuck depends on the
random numbers generated at the game's core as much as on the
specific moves one makes. While it is true that all art, or,
at least, the appreciation of all art, is collaborative -- a
classic read by a dull reader will be forever dull, and a symphony
heard by the tone-deaf will ever remain a cacophony -- the closure
promised by an interactive game tends to be much more exclusionary,
for only the best players will "win" or finish before boredom
and frustration set in. For the less gifted player, the emotional
response evoked by interactive games tends toward a sense of
failure and inadequacy -- hardly feelings that are eagerly sought
after by most people.
Related as well to the narrative voice is the emotional engagement
that the reader/audience/player will experience in the course
of the work. Narrative fictions, whether in first- or third-person,
whether visual, written, or oral, take the audience out of themselves,
let them identify with a life or lives other than their own.
A reader can be Moll Flanders or David Copperfield or Frodo
Baggins or Anna Karenina: caught in the inexorable narrative
stream, the reader lives their lives, feels their hopes and
fears, even while remaining apart, judging and appreciating
and comparing their lives and experiences to his or her own.
In the second-person conversation of an interactive game, the
main focus of concentration is on how to achieve the
next experience, rather than on the experience itself: to find
true love does one click here? go there? push that button? open
that door? It is not the identification with a fictive life
that is the issue here, but the technique one must employ to
manipulate that life. The emotional impact of that vicarious
experience is secondary: few players will weep for the death
of their interactive-alter-ego in the bowels of the Colossal
Cavern. Such deaths are not tragic or pathetic. They simply
mean it is time to play again, to achieve instant resurrection
by choosing "New Game," to spin around one more time on the
electronic interactive Karmic Wheel and use the lessons of the
just-experienced death to prolong and enrich the next life-in-the-game.
If we are to have interactive fictions that promise the same
depth of emotional and intellectual impact that narrative works
do, and have done for many hundreds of years, it may be that
the current crop of interactive games are not an appropriate
model.
 |
Part Two -- The Interactive
Narrative
How can an interactive fiction evoke as rich an experience as
do the more traditional fictional forms? We must, of course,
realize that we can never enumerate the all ways in which it
can happen; artistic creativity delights in outstripping
critical analysis. Nonetheless, we can imagine practical guidelines
for some ways, at least, to create complex and emotionally satisfying
interactive fictions. The place to start looking is at the class
of effects we wish to achieve and let that suggest the techniques
we may employ; the how depends on the what.
Narrative fictions, as we have seen, usually require a surrender
to the author, a giving up of self in some degree. It is Coleridge's
old "willing suspension of disbelief" and more: the listener
to the tale must not just trust the tale-teller, the listener
to some extent must become the tale-teller, must let
the tale hijack the memory, the senses, the imagination. The
surrender need not be (and seldom is) total and absolute; the
act of reading, for example, is a continual oscillation between
considered reflection and helpless possession. In fact, the
effortless and often rapid flickering between the "I" of the
story and the "I" of the beholder is the very source of much
of the pleasure of narrative art.
For interactive fiction to provide a similar pleasurable oscillation,
it must allow for helpless possession. At first, this seems
to deny the very concept of interactivity: an interaction requires
an "I" to interact, and the surrender to a tale requires giving
up the "I." But surrender is only one side of the oscillation:
opportunities for interaction arise on the backswing. Text-video-audio
on the computer is fully capable of possessing the reader-viewer-listener,
and it goes without saying that computers allow for interaction.
What is important in interactive fiction is to avoid breaking
the spell of possession by requiring interaction at inopportune
moments. A few specific (albeit hypothetical) examples might
help illustrate this.
Suppose we have a story involving three or four people, each
of whom have their own view of an important event and what it
means. In more traditional media, the author of such a story
would have to come up with an arrangement that reveals all the
viewpoints, but that arrangement would have to be embedded in
a strictly linear form: any given page in a book always and
forever follows the page before, any given frame in a film always
and forever follows the frame before. An interactive version
of the same tale, however, could allow for the movement between
the available viewpoints.
There are several different ways to allow for that movement.
The interactive version could, at specific points, require the
reader to choose a different thread of the tale (say, a button
which, when clicked, gave a window with the available narrative
path options). Of course, such "decision nodes" would have to
be carefully placed so as not to disrupt the narrative at its
most possessive; after all, making the user choose between Arnold's
and Maria's viewpoints in the middle of a love scene would certainly
disrupt any identification the reader had with Arnold or Maria.
Or one could make the choice of pathways more optional and less
obtrusive (say, an unobtrusive palette with the currently active
pathways always available, dynamically modifying itself when
appropriate). This would allow the reader at any time to switch
between Arnold and Maria. But there are consequences for this
strategy as well: what would be lost here is the artist's ability
to present irony, or supply sudden revelation, by the judicious
juxtaposition of the narrative threads.
So, given one hypothetical narrative model (the multiple viewpoint
story), we have (at least) two strategies for conveying it (static
decision nodes versus dynamic navigation palette). We have the
how, but which how we choose depends on what
we want our tale to achieve. Do we want distance or intimacy?
Are Arnold and Maria specimens under glass, puppets in our little
human comedy? If so, maybe the decision node strategy is the
answer: it highlights the artifice, it makes us look at
the characters as characters. Or are Arnold and Maria fully
realized individuals, with powerful perceptions of their own,
independent intelligences running in parallel, and yet in constant
interplay and interaction with each other? To portray that case,
a more voyeuristic approach might be more appropriate: let the
audience choose at any time whose thoughts will be overheard,
whose life will be vicariously lived.
For other narrative models, looking at just what "hooks" us
in a story is a profitable exercise. In many stories, the "what
happens next?" that keeps the reader turning the pages is joined
by the "what is going on?" -- maybe a puzzle is to be solved
(e.g. detective stories), or relationships to be discovered
and disentangled (almost anything by Harold Pinter), or prior
histories to be reconstructed out of the sparse clues given
in the narrative (e.g.The Great Gatsby). In such stories,
the interactivity is wholly mental: the information is given
on the page, or shown on the screen, but the significance
of that information must be inferred.
In a detective novel -- in a "fair" detective novel, at least
-- all the clues needed to discover whodunit ("what is going
on?") must be given to the reader, although the clues may be
veiled, surrounded by red herrings. And, in a novel, those clues
are always available on the page; a reader can go back and re-read
the important passages. In a detective movie, the clues must
be shown, though once shown, they cannot be retrieved (at least,
not until the film is available on tape or laserdisc). A detective
film might contain a scene that places important clues in the
background, or there might be a seemingly irrelevant cutaway
shot, or an odd camera angle. Different media, different hows
to serve the whats.
Multimedia increases the number of available hows to
handle the detective story what. An interactive, multimedia
detective story might exist as text and as movie; it
can supply the media that seem appropriate at any particular
point. Some clues might reside in the text, some in video illustrations:
such a story, presented as text on a page, can have pictures
that move, or open up into crime scenes that can be explored.
The reader can read the story, but can also stop and
really look around that dingy third-floor apartment in the less
desirable part of Brooklyn where the foul deed was done. Here,
the interaction is not used to control the direction of the
story -- the murderer will not change no matter what the reader
does -- but to help the reader interpret the importance of individual
story elements, to figure out whodunit.
From just these two examples some central guidelines have emerged.
First, the artist must decide what the story is and what it
will do. Aristotle may have been onto something when he identified
plot as the most important element of tragedy. Second, the artist
must realize that the viewer-reader-listener's role is at times
active, at times passive: sometimes the self is swept away on
the tide of narrative; sometimes the self is taking notes and
pressing the "rewind" button. Require too much interaction,
or interaction at the wrong time, and the story's intended effects
might not be achieved. Require too little, or not allow it when
the audience desires it, and the result will be frustration,
bad reviews and low sales figures. Third, the artist must realize
that interactive fiction may be a new thing, but it will derive
much of its power from the very human love of the tale, which
has existed since before movies, before printed books, before
manuscripts, and maybe even before cave paintings. Knowing how
stories work is the key. From that knowledge, all technique
flows.
1
Of course, film is not without its attempts at first-person
narrative, usually implemented as a voice-over, but the result
is a hybrid, where the spoken first-person narrative is woven
around a series of images and episodes that are still regarded
from the camera's eye, and in which the narrator is almost always
beheld from the outside no differently from the other characters
in the fiction, and thus does not occupy the sort of privileged
psychological and emotional position held by the first-person
narrator in prose or poetry. Very few films, and those experimental,
attempt to equate the camera's view with that of the narrator.
|
