Visit Our Web Site at

Monday, March 2, 2009

What Determines Plot Progression Sequences?

by Melanie Anne Phillips
creator StoryWeaver, co-creator Dramatica

Rich asks:

The one thing that I am having trouble understanding is the plot rotations. Why does choosing the rotation in one Domain sometimes chose them in others and sometimes not? And what relation does one rotation have to the other?


As many of you may have noticed, choosing items in the Plot Progression doesn’t work the same way for all four throughlines. Some seem to have much more impact, control, or power on the overall progression then others, and in fact, they do!

Now this immediately smacks of some inconsistency or inaccuracy in the software and/or theory. After all, why should one throughline be inherently more structurally "important" than another? Well, conceptually, one throughline is not more important than another, but in practice one MUST be more important than another.

I know that sounds trite. Let me explain with a brief visualization, then describe how "plot rotation" works as a mechanism in the software.

First, the visualization:

Think of a globe of the world. Now, try to draw it on a flat piece of paper. You’ve all seen the different kinds of projection we end up. Some make Greenland HUGE, but the USA small. Others make the USA large, but split the map, as if you’ve flattened out the peel of an orange. In fact, there are many different projections of the globe, but each has a different kind of distortion, due to trying to project a 3 dimensional object onto a two dimensional surface.

The Dramatica Structure suffers the same problem. It is SUPPOSED to represent a model of the mind, as called for by the theory. The mind itself is a FOUR dimensional object. That fourth dimension is Time. To be accurate, time cannot be broken into a series of increments but must flow continuously and simultaneously throughout the model. The problem is, that a computer cannot create a truly unbroken "flow."

In computer programming, every operation is a series of steps, be it a function or sequence of operations. As a result, to create a model of the four dimensional mind in a computer, you need to "project it" onto three dimensions, then "move" it through time in steps. That is not completely accurate, just as any projection of the globe is not completely accurate on a flat surface. Still, in this way, the first three dimensions are VERY close to accurate, but the fourth dimension is where you pick up the distortion.

In the software model of the Story Mind, this distortion will show up with the Plot Progression.

Now, as you might expect, there are three other projections of the Story Mind which might be created: One in which the distortion shows up in CHARACTER, one with a distorted THEME, and one with a distorted GENRE. Each has a different strength and a different weakness.

Ultimately, it is our hope to program the other three as well, so that authors have a choice of where to sweep the distortion under the carpet. Unfortunately, each requires the creation of a completely different model with its own unique algorithms. The original model took four years to build and two more to perfect. It was also VERY expensive, costing over one million dollars in R & D before the FIRST version of the software was released. As you may imagine, it will be many years before we can offer another projection of the Story Mind (especially being intellectually burned out by the mind-warping contortions of visualizing the first model!)

Okay, so this simple visualization gives an overview of the problem. It tells us why the distortion will show up in plot. But what is actually going on in the software that makes that distortion give more "power" to one throughline over the other?

The simple answer is that the same bias that makes Plot Progression distorted also favors the Main Character and Objective Story throughlines at the expense of the Obstacle Character and Subjective Story throughlines. As a result, more power is assigned to them, over the others.

Here’s where we have to get a bit more technical...

You may be familiar with my analogy of "winding up" the structure to create a storyform, as if the structure were a Rubik’s cube. This is a surprisingly accurate visualization. In the form you see the structure on the chart, it is neutral and at rest. In other words, there is no dramatic tension in the resting model. This is because all the quads are balanced and consistent in both the vertical and horizontal planes. This can be seen by nothing that on the chart, "Past" is to "Universe" as "Memory" is to "Mind" This shows that identical vertical distance in the creates identical semantic differences in meaning. Horizontally, "Being" is to "Becoming" as "Doing" is to "Obtaining." This indicates that identical horizontal distances create identical differences in meaning. In other words, in the at rest model, identical vectors in the three dimensional matrix represent identical differences in meaning, so that the relationships among any story points plotted on the matrix can be determined by their semantic distance.

Sorry about that!

Now, on to the next technical information necessary for the answer to your question...

When the model is "twisted and turned" it moves items out of alignment, altering their relative semantic distances and creating a tension or distortion based on the degree of misalignment. This is what happens when you answer questions in the Dramatica software.

In fact, there are two kinds of wind-ups which occur. One is applied to the Main Character Domain and then ripples out over the entire structure. The other is applied to the Objective Story Domain and then ripples out.

The eight questions you answer about Main Character Dynamics and Plot Dynamics (Resolve, Growth, Approach, Mental Sex, Driver, Limit, Outcome, and Judgment) determine many things about those two wind-ups.

For example, because Time is not free flowing in the model as it would be in a real mind, one of the windups (Main or Objective) will be applied first to the neutral model, the other will then be applied to an already twisted model. Which comes first creates the feel in a story as to which is more "screwed up" - the Main Character or the world at large. In this way, the story develops a dynamic imperative indicating that a Main Character must change or must remain steadfast if success in the Objective Story is to be achieved.

The real question is, how does the mechanism of the wind-up actually work?

Okay, the wind-up in each of the two throughlines begins at the bottom and works its way up. Why? Because that way it screws more with time (the horizontal plane) than with space (the vertical plane) in keeping with a consistent projection or bias to the model overall. (The bias must remain consistent in both structure and dynamics or the distortion will drift and create apparently chaotic inaccuracies rather than limiting them to one area for the benefit of all the others.)

To wind up the very bottom quad of elements, the software must know the problem element for that throughline. That can either be chosen directly by the author, or the story engine will eventually work it out as a cross-reference of the effects of other choices.

Once the problem element is known, it becomes the pivot point or "seed" of the throughline’s wind-up. Now, on that first quad, there are two kinds of wind-ups which may be applied: "Flips" and "Rotates."

A flip will swap the positions of two elements in a diagonal relationship, such as "Faith" and "Disbelief." Why would this happen? In a real mind, when we have one of our elemental sensibilities rubbed raw by experience, one of two things happens - we become ultra sensitive to that topic when it comes up or we become insensitive to it (scab it over). A flip containing the problem element itself represents a scabbing over by moving the problem out of harms way. A flip along the other axis (between the other two elements not containing the problem element) represents an increased sensitivity by leaving the problem in place.

Of course, when one becomes overly sensitive to an item, the items around it become less sensitive to pinpoint the irritation and make it easier to avoid further injury. But, if one scabs over, then the surrounding items become more sensitive to make up for the loss and also as a sensitive perimeter that warns the mind something is approaching which might rip off the scab.

In contrast, one might "rotate" elements rather than "flip" them. Why? Because in our own minds, we sometimes don’t just become biased by experience to make things more or less sensitive, but we also move items up and down in the pecking order or sequence of consideration depending on their endlessly adjusting priority.

So, in a "rotate," we move the items in a quad circularly, like a turning a knob. This also has two version, clockwise and counter-clockwise. This creates a different kind of tension determining whether or not the problem element is being moved up or down in priority.

Once we have flipped and rotated (twisted and turned) the first quad in the first throughline, we move up to the variation level (issue or range). The same kinds of dynamics are at work here too, but not necessarily the same arrangement as in the quad of elements below.

The upper quads have an additional aspect - they might "carry the children" or not. This means, when the variations flip and/or rotate, for example, do they drag their underlying elements along with them or leave them behind. Why? Because justifications (biases) can enter a real mind at any level and may or may not affect the levels above and below.

You can see this flipping and rotating at work in actual stories. To do this, find some dialog that deals with thematic issues. ("Witness" is a good example). Find a quad of variations that deals with those issues. Plot the sequential progression of the issues that occurs in the story. After plotting a number of different quads you’ll find sequential patterns that appear as "U" shapes, "Z" patterns, and "hairpins." All these patterns can be created by the sequential application of flips and rotates to any quad.

Ultimately, you work your way up to the top level of the structure. Here, flipping and/or rotating moves the problem from an interior position (Mind, Psychology) to and exterior position (Universe, Physics) or vice versa. This is the model’s accurate description of the psychological process of "projection," where one comes to feel that "I’m not the problem, it’s everybody else" when it really is the person or conversely, "I guess I’m the problem," when is really is everybody else. Ironic that the psych term for that is "projection" - not unlike the projection maps we have been talking about.

Now, I could go on endlessly about this mechanism, but we now have enough to answer the questions: "Why does choosing the rotation in one Domain sometimes chose them in others and sometimes not? And what relation does one rotation have to the other? "

The Dramatica software story engine actually predicts the best order for not only acts, but sequences, scenes, and events as well. Early on, we realized this information would amount to "micro-managing" the plot, so we "suppressed" it. It’s still in there on every storyform, but not presented in output. We did output it for a few sample storyforms, and it amounted to literally hundreds and hundreds of pages of progressions for every quad and "quad of quads" in the entire structure. Ultimately, we only kept the "act level" progressions, as they seemed truly useful without being overly binding.

The first two versions did not allow plot progression choices so the nature of the distortion was not apparent. But when we added it in version 3, it came right up to the surface. We actually considered not including that feature to avoid the sense that the software was not accurate, even though it was just the projection distortion described above. But, the desire to provide all possible useful tools prevailed, so we put it in with great trepidation.

I think we have seen why one throughline has more power than another, but what is the relationships among the four plot progressions? In the structure without plot progression, each throughline represents a different angle on the same issues. In one sense, they represent the I, You, We, and They points of view. In another sense, they represent Knowledge, Thought, Ability, and Desire in the Story Mind (more about this in another post sometime down the line).

Once "wound-up" they create structural differential or dramatic potential among them. In motion over time, they create resonance and dissonance (harmony and disharmony). Both the dramatic potential and the interference patterns of the flow must work in conjunction so that the space-sense and time-sense of the storyform serve to carry the same message. The trick is to make the "particle" and "wave" work together. Because the structural bias exists due to the projection of the mind on three dimensions, there must be an identical bias to the temporal progression.

Taken altogether, the Plot Progression simply does not allow certain sequences because, although possible, they cannot occur in this projection without interjecting inaccuracies BETWEEN the structure and the progression.

As it stands, every available progression consistent with the model’s necessary bias IS available, so that the progressive harmony and discord of the flow of the four throughlines creates an interference pattern in which the nodal points intersect with the story points in a synthesized four-dimensional space.

In other words, the plot progression of all four throughlines will wrap around each other as the story proceeds so that it creates the spatial meaning of the story in much the same way that the scanning lines on a TV screen work together to create the greater mosaic of the Big Picture.

Thanks for asking!