2 Aristotle
2.1 Aristotelian stories
I do not want to use the term "story" in the everyday sense: one
telling, one or many listening - what is being told and listened to is a
story. Instead I want to broaden the concept like this: anything that
unfolds itself over time is a story. There are things that you can
instantly understand, and or be aware of; similarly there are some
things that only unfold over time. Story line is essentially something
that needs to take some time in order to be understood. One perceives
them in time like navigational paths, which can be interpreted in terms
of Aristotelian notions of ontological causes, story line and structure
of drama.
Picture 9: Any story model can serve as a tool both in analysing,
synthesising and producing structures over time.
In algorithmic design I used to promote strategy of dividing any
programming problem into begin-middle-end sequence. Recursive
application of the strategy worked out as program refinement. I called
this format "The Japanese Model" because it is simple and you can apply
it flexibly to any sequential structure.
Hollywoodian movie format is based and built on Aristotelian tragedy and
classic short-story. In this type the story is largely related trough
the fates and fortunes of main character (Aristotle).
The brechtian model, where play is a succession of independent acts,
that are joined by narrative commentary of songs and visuals (Brecht).
As a fourth example of story formulae are algorithmic outlines either
for computer programs or work and job descriptions. They have a start
and they usually have an end point. The action of algorithm is
described precisely in steps leading to the desired result. There may be
repetitions or alternative routes in the flow of algorithm but it is
essentially like a story unfolding over time.
When doing hypermedia you are working on and with computers and the end
product will be viewed on computers, so inevitably you will end up
dealing with programing and algorithms, or somebody in your group will.
The idea is that story model for hypermedia storyboarding should be
applied to algorithmic design too.
2.2 Navigational paths
We made in spring 1994 together with photographer Jari Arffman three
installations of an interactive piece "Crusade". We had seven grayscale
photographs and three screenfuls of highly abstract text which were
interlinked for audience exploration on a Macintosh with a mouse.
Extracts of poetry read by actors were added to enrich the links. The
idea was that the exhibition visitor could look at the high quality
photographic prints and then he could, in a very postmodern manner,
deprive images of their visual meaning and start exploring them in
association with a very conceptual text framework.
Picture 10: Crusade installation in Alvar Aalto museum in Jyväskylä,
Summer -94.
We wanted to have a few people at a time engaged in exploring images and
the text preferably engaged in mutual conversation. We did not put in
any navigational clues, because contemplation over our material through
free exploration was one of the points of our piece. To get those minds
moving and the conversation going. I wanted to leave it to the dark
attractors of chaos in the work to make and keep audience interested.
I call this form of hypermedia hyperpoetry, because it thins out the
narrative and does not necessarily have story lines as such. At best
this leads to contemplative interaction between the explorer and work
and is an attempt in lyrical hypermedia expression.
I was very happy to see and learn later on that roughly about a third of
the visitors were pleased with the experience. They said it was
something adding to the images on the wall. Of course, one third said it
was completely beyond their understanding. Perhaps that evens the
success out a bit. This was our crusade into a storyless composition -
conclusion was that the lack of narration personalises the experience -
not so much diminishing it but rather rendering it incommunicable
further.
A navigational path lays out a story, the flow of which can be
user-controlled, marked by attractors, outlined by agents or led by a
guide. Crusade took the user control to the extreme offering no cues and
no stories. A navigational path can be marked out for the user by icons
that connote the line of narration.
An agent can be represented by anything - in hypermedia documents it
means that any programmable object can call up operations that complete
the task of called agent. In modest environments text labels and icons
can represent agents.
One term has been coined: knowbots - software robots, task performers,
programmed entities that exist only inside the computer. They may have
some defined task-handling skills, even learning skills and so on.
A primitive example of an agent controlling any arbitrary linear story
could be a leafer agent that is implemented essentially as two routines
that use an index - a list of is-a-kind-of components in their
navigating order. The two routines locate the name of the component the
navigator is at from the index and according to the call either choose
the name above or below it and take the navigator there. In our first
experiment, Codex B, the hypernovel we placed these routines into an
animated icon of a book that leafed itself back and forth and took the
navigator back and forth in our stories. I know that this is stretching
the agent concept, but the rationale was to offer a context free service
(an agent) that anyone would intuitively be able to use. This
arrangement also enabled us to modify and rearrange the stories freely
once we kept the index in order.
Similarly it is relatively easy to construct services that select items
or indexes or both. With the selection method they can impose their
intentions on the navigator, and as such assume the role of a narrator
or instructor. Use of agents, especially intelligent agents or agencies
(collections of agents) in storytelling - laying out navigational paths
- is definitely one of the main techniques in future hypermedia.
Through artificial intelligence techniques you can device intelligent
entities that represent or draw your story line, who are guardians of
your story, or they tell or act you the story. I have often called
agents with distinct intentions guides - one example might be learning
companions in educational titles. More about agents in (Maes) or
(Minsky), Apple Computer's Guide -project is discussed in (Laurel 90).
2.3 Aristotle's four causes
Aristotle uses a metaphor of a chair making when he explains his ideas
of the four causes of all existence. Same ontological principles apply
in and on all design levels, I will elaborate this on story's
is-a-kind-of components.
Formal cause: There is a form of chair, a universal chair form that all
chairs wish to fulfil. A table does not fulfil the form of chair, so
nobody calls it a chair.
Material cause: The matter out of which the chair is made sets
constraints to its properties. It makes a difference if the chair is a
wooden or a metal one.
Efficient cause: Aristotle means by them the skills and the tools
available for the chair maker.
End cause: Not least even if last, the end cause. Why chairs? For
princes or paupers?
Let us apply these principles in hyperboarding: With formal cause I mean
that there are certain ways (data formats for the items) you need or
have to represent certain is-part-of contexts. One has to be aware of
what one is describing in writing - basic linear text, dialogue,
something one needs to represent visually or with simulations. Certain
messages and statements crave certain forms of expression. Also some of
the formats (videoclips) are overappreciated and others underrated
(animation).
Material cause is the physical data that hyperdocument incorporates. At
first most of it is described in writing that evolves towards the end
layout and representations gradually throughout the project. One might
say that text is the stuff hypermedia is made of and it follows that
hyperboarding is writing hypertext for the hypermedia application. As
the design progresses and better understanding of size and media
constraints is achieved, decisions on the number, size, binary type of
the composite data elements are made.
Efficient cause means the tools and know-how. When one is embarking on a
hypermedia project one has to analyse the possible audience, its size,
the variety of its computers. Then one has to make allowances for what
machines and environments one really has at hand in the production
phase etc. The skills and methodologies one acquires and accumulates
during the project may have great influenece on the end title. The
efficient cause working on one's title is something you cannot really
fix to any given time. The process and equipment develops so fast that
you may want to keep your design very modifiable at this point in order
to take advantage of technological development. The development of both
job and task routines and skills together with group work dynamics and
tools is vital in securing quality and timing in all hypermedia
projects.
End cause - the reason. The principal constraint one needs to impose on
one's work is the message or collection of messages you want to convey
to the navigator. This is best formulated in a couple of concrete and
descriptive sentences. This definition serves as the ultimate measure in
deciding which design propositions are accepted and which rejected,
therefore it is utterly important that the team agrees and understands
the document's purpose - its end cause.
2.4 Tableau
Next I will break stories into components that can simultaneously belong
to a number of stories. A story is a sequence of tableaux that are in an
is-a-kind-of relation to each other. One is going from one screen to
another when one navigates through a hypermedia title. The screens frame
out the document for you in tableau by tableau. Our national poet
Aleksis Kivi uses word "asuma", when he goes out in the summer by a lake
and gazes at the sunset and what he sees, hears, smells and feels he
calls asuma: a moment of stillness, something that one can stand and
take a look at, understand and contemplate. For Finns there are
connotations of "asuma" - to live, habitus, attire, inhabit, home. My
idea is that a tableau should not be conceived unfolding over time but
something that can be grasped instantly as a whole.
We all know that aesthetic appreciation sometimes requires training,
education, understanding, experience and so on. It stands to reason that
some people understand some wholes (and tableaux) more readily than
others. When we enter a room we instinctively conceive it as a unit 'X's
room, Y's office etc.' - our knowledge of the content changes as we
visit the room more often - but it still is understood as some one
thing.
Concrete realisation of a tableau may be a layout, just a simple
newspaper page, for instance, it can be an ideal tableau for joining
several story lines. Or it may be a painting, a still life, or it can be
any collection - a collection of things. A snapshot of me lecturing at
the podium here counts for a tableau. Situation... a pose, a scene - we
approach drama and film.
Wittgenstein in his "Tractatus Logicus Philosophicus" offers a formal
view to tableau as he explains how facts are comprised of states of
affairs. He sees behind linguistic concepts that can be understood as
facts bundles of facts that are connected by grammars (Wittgenstein).
One combination of facts produces a new fact and further on another
combination (to another level of abstraction). Wittgenstein is very
useful when you deal with computers and computer programming because he
says you should shut up if you do not know what to say, which is
precisely what you have to do with computers. Wittgenstein's "Tractatus
Logicus Philosophicus" is a very useful piece of philosophy if you are
dealing with computer programming, program design or algorithmic design.
2.5 Aristotle revisited
I will apply Aristotle's ontological principles on tableau description:
in hyperdocument one has a number of stories that are comprised of
tableaux. One starts hyperboarding by outlinig one's stories - and it is
very advisable to use several authors, so one gets the personal flair to
each story. As a team you decide and work on the tableaux that are
shared by individual stories. The process shifts between your own desks
and meetings where you refine the tableaux for your stories. It is
useful to start describing tableaux in terms of the end cause. Define
the meaning, the fact you want the navigator to grasp instantly. A short
and precise definition of its purpose and/or aim - for the team it is
also a goal. Because you are working in a team, you have to make
compromises, you have to make common decisions on what ideas go in and
what is left out. The end cause serves as the measurement in making
those decisions. It is vital that one does that definition otherwise one
ends up having horrible arguments with nothing getting done at some
point in the project.
And then the material cause. By and large this means the text.
Description of the content of tableau in text format gives the authors a
better understanding to the quality, quantity and scope of information
desired. This is the basis of its material cause, the substance of our
end title. The text may be decomposed into several items.
Picture 11: A template for tableau description
The formal cause defines the vessels (dataformats) for realising the
content. By that I mean that the left column (picture 11) contains the
material - e.g. a piece of dialogue - as its formal cause one defines
in the right column correspondingly "how one does it", e.g. "Two icons
having a dialogue in text fields, so as to give a 'semi-animated comic
strip' effect". Formal cause is how one chooses to define the way one is
going to do it when one gets the equipment and the time etc., etc...
Even if one cannot go into details yet, the important thing is that one
tries to describe the corresponding format for every text-item in the
material cause. Then one analyses, synthesizes and produces the final
tableau filtered by the efficient cause.
For an efficient cause one needs to analyse what is feasible and lay
down a production schedule and DO IT. This means the tools, facilities,
skills, personnel and time available for the final tableau from the
production point of view. In order to understand the effects and
efficiencies shaping your work you need to have an understanding of your
audience and the equipment available to them. It may even be reasonable
to invest project money in improving infrastructure (network
organisation or database management) rather than producing a single
glamorous hyperdocument.