Predicting Human Behavior with Narrative

By Melanie Anne Phillips
Co-creator, Dramatica Theory of Narrative Structure

Human behavior cannot be predicted by observation alone.  No matter how deep the statistical database, no matter how sophisticated the algorithms, accuracy derived from observation falls short because it is unable to see the inner mechanism of the mind itself.  All that can be catalogued is simply the external impact of internal mental processes, and therefore observation can only chart the progress of ripples in the pond and speculate as to the nature of the pebble that produced them.

Human behavior cannot be predicted solely by internal self-examination.  No matter how deep we focus our inner eye, no matter how extensive our thoughts, accuracy derived from self-examination falls short because it is unable to see the mechanism of its own sentience.  All that can be grasped is simply the results of inner mental processes, and therefore self-examination can only map our attitudes and speculate as to the nature of the feelings that produced them.

To predict human behavior, a true model of the mind is required – one not derived from external observation nor internal self-examination.  The question arises as to how such a model can be created.  The answer is that such a model already exists.  It is called Dramatica and it was discovered in the structure of narrative.

The creation of narratives – both as stories and in the real world – is a uniquely human endeavor with two primary purposes:  one, to move an audience to adopt an attitude or point of view and, two, to describe human truth as best we can, so that we might better know ourselves and understand our relationships with others.  The first purpose is directed toward subject matter – the real world issues about which an author might wish to move an audience.  But the second purpose is accomplished below the level of subject matter for it documents human nature itself.

When you strip away the subject matter, the structure is laid bare and reveals itself as a model of the mind.  Why should this be?  Because when humans gather in groups to address a common issue, they tend to self-organize into specialties that represent different attributes of the human mind.   For example, one will emerge as the voice of reason while another will express skepticism and yet another might express the considerations of conscience.  In this way, each specialist is able to bring greater depth to the collective discussion than if each individual was a general practitioner, all trying to do the same job – a shallow exploration of every perspective.  It is a simple societal survival technique.

Simple stories, the first stories, addressed this and established the archetypal characters and how the fundamental human attributes they represented interrelated.  In fact, the interaction of one character with another is analogous to the way these attributes interact in the mind of an individual, as if our own mental processes had been projected outward and made tangible in a macroscopic manner.

When groups grow even larger, the fundamental attributes attract additional followers so that they become sub-groups within the larger group.  In this manner, each perspective on the problem is represented by many individuals.  And, when a sub-group grows to a critical mass, it will itself self-organize, just as did the original master group.  One member of the specialty group will emerge as the leader with the others falling into the roles of the other human attributes.   And, similarly, if a number of master groups come together to address and even larger issue of common interest, each master group will shift off center as they all self-organize into specialty roles as well.

As thousands of generations of storytellers documented what they saw in the way people and groups of people organized themselves, though trial and error they gradually refined the conventions of story structure until it accurately represented the functioning of the mind itself.  Recognizing the correlation of structure to the mind, Dramatica further refined the structural elements and the dynamics that drive them.  Conceptually, this model of the mind is the substance of the Dramatica theory and, practically, it is re-created in the Dramatica software.

Dramatica’s model of the mind is comprised of two principal components.  The first is a periodic table of narrative elements in which the nested nature of human attribute self-organization is presented as families within families, much as the periodic table in physics gathers elements into families such as the rare earth elements or the noble gasses.  The second is a set of algorithms that describe the manner in which these mental attributes interact and interrelate.

In combination, the algorithms describe mental dependencies in which the action of every human dynamic has impact or influence upon other closely related dynamics.  The dynamics form a web that can be interpreted to reveal the tensions and forces at work in the mind and how they warp the shape of the mind and focus motivation in predictable directions.

In conjunction with the table of narrative elements, these algorithms of dynamics can pinpoint the sources of motivation and, conversely the location of blind spots into which one’s own consciousness, or that of a group, cannot see.

The model as whole is able to determine the relationship between a state of mind and the sequential progression of considerations, both conscious and subliminal, which must follow from such an arrangement.  This process is commutative, for if one knows the order in which a sequence of considerations occurred one can regressively ascertain in great detail the mental arrangement that must have existed to drive such a progression.

In theory, if one identifies in the real world an individual or group mind of any number of nested levels of sub-groups, one can accurately determine the drives, areas of focus, purposes and methods of that mind.  By applying the sequential algorithms, one can also predict the progressive behavior of the mind under study.

In practice, both individual and group minds are constantly coming into conjunction, in conflict or collusion.  Since they are not joining in a common purpose for a sufficient period of time, they do not reach the flash point at which they would self-organize into a single predictable system.  Rather, their interactions are only partial facets of a mental model and operate more according to the procession of chaos than an orderly progression.  Nonetheless, the impact of such encounters leaves an identifiable impact upon both parties that enables a revised accurate assessment of each altered mind and its future behavior.

The domain of chaos can be somewhat reduced by the application of those same standard statistical and algorithmic approaches that have been unable to predict human behavior, for they take into account environmental considerations, essentially the subject matter of medium in which the minds encounter one another.

In conclusion, human behavior can be predicted with significant accuracy through narrative modeling.  But when narratives are only partially present in volatile scenarios, statistical modeling can be used in conjunction with narrative to achieve the best possible predictive algorithm.

Contact me about Narrative Science

Learn more about Narrative Science