Re: Higer-order plot abstractions

["B. Fallenstein" <b.fallenstein@gmx.de>]

"Brandon J. Van Every" wrote:
> Welcome aboard!  And I see you've adopted to local culture nicely by making
> a terribly long-winded 1st post.  :-)  

Um, the complete body of the first post from you I can find in my
archives reads:
     Does anybody exist on this list?
;o)

> Cutting to the essential of your
> concept, regarding your abstractions (which I do agree are necessary):

<< snip myself >> 

> But what are these meta-plot conflicts *worth* ?  What do they *mean* to the
> audience?  Really, they don't mean anything.  They are completely empty
> semantic placeholders.  They have not done the essential job of
> storytelling, which is getting an audience to *care* about your story.  To
> recognize your story as something meaningful and worthwhile.
>
> I think the abstractions that you/we are going to have to concentrate on,
> are abstractions of meaning, motive, value, and audience buy-in.  Without
> such entities you do not have drama.  You have a Punch and Judy pantomime
> where someone screams "I accuse you of adultery!"  "Well what if I did
> bugger your wife??!?"  "Have at thee!"  Events are reduced to obvious,
> mercenary, wooden economic transactions, it is boring simulationism.  If we
> do not have a reason to care, then who cares?

*g*

In short: I agree that on the level I have presented them, my examples
of plot abstractions are semantic placeholders, not meaningful units. I
will say that they are intended that way: as the short-hand speak an
author uses when thinking about their own 'dream.' I will also say that
I believe for the author, plot abstractions similar to the ones I gave
can hold meaning. I do agree that we have to think about just *how* an
author can communicate this meaning to their audience. And in order to
get productive (in a way), I shall spend the rest of this post pondering
this issue. :)

I imagine that the level on which we communicate with our audience is
the level of events, i.e. we tell them what happens and then ask them
what their character shall do next (i.e., we give them the opportunity
to decide upon the next event happening in the story). On this atomic
level, I think the Erasmatronic tinkertoy text will work, i.e. we write
a text with a few gaps which are in turn filled in by the computer. This
requires that the event template associated with such a text actually
has one single "feeling" attached to it; I mean, it is easy to say "X
threatens to do A if Y does not do B," but the act of threatening can be
on very different levels, with different intensities. Let me take a stab
at two such texts-with-blanks, communicating different flavors of the
above concept (each in the "you, the player, are threatened" perspective).

 #1: (X) goes to the wall and appreciates one of the knives hanging there.
     (She) slowly takes it in her hand and examines the hilt. "You could
     do me a favor," (she) says coolly without even looking at you. (She)
     probes the blade with (her) index finger. "You could (do B) for me."
     (She) makes a dramatic pause, then suddenly looks at you, pointing at
     you with the knife. "Because if you don't... well, you don't want
     (A to happen), do you?"

(Note that this assumes a specific setting. I think that this being
obviously a central plot point in the story, this may be justifiable--
maybe the author has to provide two or three versions in different
settings, though.)

 #2: (X) sticks (her) tongue out at me. "Bah, you know what I'll do if
     you don't (do B)? I'll (do A)!" (She) storms away, (her) nose
     high in the air.


> You can't just state that somebody cares.  "X is angry at Y" is not a
> motive.  It doesn't sell anybody, it doesn't set the audience up to *feel*
> anger at Y.  I think what we need to do, is look at cinema/drama in general,
> and encapsulate the devices by which they get an audience to care.  I don't
> have any precise answers right now on how to do this.  But generally there
> are 3..5 semantic things that we have to know about a character/situation to
> care, and those 3..5 things also have to be delivered with a certain timing
> in order to build our interest/acceptance in them.

<<snip>>

> "Caring" is not and can never be an abstraction, except perhaps for schlocky
> genre characters like Steven Segal, and only if you actually enjoy that
> whole paradigm.  "Caring" is a cultural or ideological specific, that you as
> an author must provide.  This is where the human author *has* to do
> something with the software.  The software might provide a framework for how
> the "caring" objects will be delivered, but it is a human being who must
> decide what we are going to care about.  Will we care?  Only if the author
> has compelling concepts *and* those concepts are deliered with good timing.
> A good authoring system would assist with the timing.  A bad one would leave
> timing to chance.  The mere presence of materials does not drama make!  They
> must be timed.

Okay, right, so we need a way to line up and time things. I agree that
there is some kind of magic that makes some sequences of events powerful
and others not compelling; as 'dreamcrafters,' we will need to find some
kind of abstraction over these. I have been pondering these issues a bit
today (which is why I haven't replied earlier). The issue reminds me of
Walt's description of conflicts, each conflict being resolved through a
number of subconflicts. Mmaybe extended to the notion of "developments"
(e.g. falling in love-- I'm not sure whether this would be called a
*conflict*...), it seems that these conflicts are a good level for
modelling timing and the lining up of those 3..5 things.

I am imagining that a conflict/development would be, on the lowest
level, be essentially an abstraction over one scene. It would specify
the different ways a scene could go, maybe have a number of state
variables of where the scene is going; it would also make sure that in
each case, the scene is dramatically satisfying. Here's an example.

Since she was ten or so, Alanna of Trebond has been trained as a knight
at the court of Tortall, the country she's living in. Girls are not
allowed to become knights, so she's been masquerading as 'Alan,' a boy.
Now, it's around her 18th birthday, and she's become a knight; at this
point, she has decided a long time ago, she will reveal her real gender
and sex to the court. Now, the development we're concerned with is the
scene where she makes that announcement (in the story this is taken
from, it's actually finally revealed against her will, but that's
another story). In general, the end result can range from accepting to
stripping her from her title or so. During the scene, the resonances she
gets from the crowd change as well as her own nervousness, and the
dynamics between these have a large impact on the outcome.

Now, the development is structured in four stages:
1. Alanna decides to make her announcement now, thus starting the development.
2. She makes clear that she has something to say and waits until the
crowd is listening to her.
3. She makes the announcement.
4. There is some reaction.

The ways these individual steps can be filled out are very much varied;
for example, she can make her announcement nervously or proudly; the
king or someone else in the crowd can give the first reaction, and the
crowd can follow them or be just shocked, etc. Again, to archieve the
combinatorial explosion that gives us a feeling of agency, I think the
development should be higher-order, i.e. accept different events as the
actual four steps, thus giving the player a higher degree of freedom.
But frankly, I don't quite see how this can work. Then, of course, I
have only today started to really think about this issue :o)

I dunno, I feel this can lead somewhere, but there are too many open
questions and I'm too tired to think them through now. Any ideas?
- Benja