sbenthall.net

Narrative is a High Level Data Structure

was the first thing I read this morning. It was a text message that Ethan Jucovy had sent the night before.

“Or…is data a high level narrative structure?” I responded, pleased with myself for having found something to say on the topic at 7:14 am. I was preoccupied with the question for most of the early morning before I left my apartment to go to work.

So you can imagine my joy when I bumped into Ethan on the Park Place platform of the Franklin Avenue Shuttle.

We discussed the matter. In short order it was clear that Ethan’s original comment had depth deserving of a more careful response than my cheeky one a few hours earlier.

To sum up, his contention was this: he had used to believe that narratives were primarily a way of packaging a single idea, or what he sometimes calls “conceptual atoms”. But he has recently thought that they may be data structures, like stacks and queues.

Data structures are designed to optimize certain kinds of data input and access. So what kind of operations are narratives optimized for? Ethan had struggled with these questions and hypothesized that the narrative structure might be defined by a geometric braiding of ideas.

At this point I should have admitted that I didn’t fully understand what he meant by braiding. But I was already suspicious of the contention that narratives could be defined in such geometric terms.

We debated the topic for the duration of the subway ride. At one point, we were interrupted by a panhandler who announced:

I am homeless veteran. I came home with post-traumatic stress disorder. I have been waiting on the results of a specialized treatment. I don’t know why its taken this long, but it has. I am out of money and haven’t eaten in two days. If any of you could spare an apple, a bottle of water, a quarter, a dime, a penny, I would deeply appreciate it. I hope you will find that charity or love of our country in our hearts, and blessed you all.

I asked Ethan what he thought the constituent ideas of that narrative were, and how they were arranged. He hesitated, then said that the difficulty in extracting the entangled constituent ideas from such a story was precisely why he had proposed the braid structure. The most compelling Art, he claimed, expressed just one idea–the idea of the work of art itself. So a good narrative was, in this way, like good music.

We arrived at my stop on Canal Street. I asked him to hold the thought.

Later, we would continue the discussion on IRC:

<sbenthall> ejucovy: so, what I was going to say before being so rudely interrupted by arriving at my stop was that I was concerned that that description of what makes a narrative compelling wouldn’t generalize culturally

He agreed and said he had the same reservations, and began to refine the definitions of his terms. But as we discussed them, they showed themselves to be increasingly entangled. This was frustrating because it took the discussion away from the motivation behind our interest in narratives, which, between the two of us, were threefold:

• Interest in the potential for technology to capture and communicate the processes implicit in the creation of various forms of human expression
• The conviction that the establishment of narrative is essential for building political movements (or more generally, essential for broadcasting actionable ideas)
• Curiosity about the role of narrative in knowledge legitimization

Ethan pointed out that the discussion was getting lost and that we needed to regroup. He asked how we should proceed. I hemmed and hawed, saying that it depended on what we were trying to accomplish in the first place. But eventually I came around to the conclusion that we should begin to accumulate examples of narratives in order to provide a data set against we could test any proposed definition of narrative. I with the follow examples:

Narrative	Not Narrative
Epic poetry	Wallace’s Infinite Jest
A lawyer’s description of a sequence of events around a crime	The conversation we were having at that very moment
The Cat in the Hat	One Fish, Two Fish, Red Fish, Blue Fish

But I surprised myself when I proposed as an example the following sequence of words in quotation marks:

“There once was a small boy named Peru who was fond of foreign languages. He had learned about five or six before he was hit by a bus. His gravestone read, ‘He never learned Spanish.”

And suddenly my way of looking at the problem changed radically.

What does the narrator in the preceding narrative mean when he concludes cryptically that his “way of looking at the problem changed radically”?

Luckily, I am that narrator and can explain everything.

If one wants to go about identifying the necessary structure of a narrative, then the best place to start is with concrete examples and look for parsimonious hypotheses that accommodate them. This is the methodology that linguistic syntacticians use to derived models of formal syntax. In retrospect it seems like a no-brainer to consider narrative to be a linguistic phenomenon and treat it as such.

I am sure that just such a sociolinguistic analysis has been done already. But rather than looking those up (because what fun would that be?), we can already draw substantive conclusions from the narrative provided in the first half of this post.

First, and most obviously, the narrative structure integrates well with other data structure. The above narrative has, nested within it, several other narratives as well as more formal structures such as the list structure and the tabular structure.

Moreover, the narrative structure nests easily into non-narrative forms, such as this expository post on the nature of narrative structure.

This method is especially informative when applied to extreme cases, such as the minimal narrative. For example, the following:

I met a girl yesterday. I got her number.

Can this be further decomposed into narratives? Does “I met a girl yesterday” suffice? Or is some introducing and easing of import or tension necessary for mere statements of fact to combine into narratives?

We can also use this method to find counterexamples that break the structure of narratives (and the structures within which they are embedded) to test their limits. Because Peru’s grave was secretly empty; he had been stolen from the hospital and raised in Argentina, where he learned Spanish after all.

To conclude, I’d like to provide another example that demonstrates another interesting feature of narrative structure: the possibility of a sequel.

I rambled on in IRC about possible examples on narratives until it was clear that I had killed the conversation. All for the best, perhaps, because there was work to do. By the time I came home and consumed my burrito from Taqueria De Los Muertos, I was thoroughly distracted by more urgent matters.

But shortly after dinner I found myself pacing, unable to stop puzzling over the conversation of the day.

I invited Ethan to go on a walk with me, but he did not answer. Perhaps he knew my intentions and had grown bored of the subject. Or perhaps he knew, precognitively, that it was about to rain.

Plodding through the drizzle down Eastern Parkway, I began to plot this very post as a structural knot for him to untie, should he accept my methodological proposal.

So ends the story of my day.