Narrative Structure
There is a structure to ideas. In general, I’m quite far from knowing what that structure is, though I’ve come to have some insight on what it’s not. I spent a good chunk of the last five years writing academic papers, an exercise, like so many exercises, which I found deeply dissatisfying.
A paper, as a piece of prose, is essentially linear in form with a start and an end and a preferred order in which the middle bits reveal themselves. The experience of doing research is nothing like this. I’m struggling for something to compare it to, but for the sake of moving forward let’s say it’s something like sculpture. We have a vague idea at first, enough to make some general plan, and select a suitable material but once the form is outlined, we start in on the details. Now that the bust has a head, we impress a face, but not being done with that we move to the details of each feature, not one by one, systematically finishing one and moving to the next from left to right, but by fudging the nose a bit, removing some stuff from the cheek, sanding at the lip, scraping at the nostril and constantly stepping back to see the total effect before moving in again to tweak some detail. This iterative, incremental accretion of artifice goes on for ages. Arguably, it never stops. Arguably it was itself inevitably born out of an earlier process whose completion was unsatisfying just as this one will be.
After this process, or since it never really finishes, during it, a paper is begun. Its linear form begets a narrative, a story of the research. As all good students know, we start this story with a research question and in the second act turn dutifully to The Literature to discover the work of our great ancestors who have always left us, we are pleased to discover, with some little lacuna in which to settle down and further develop our plot. The story of course can have no characters — only an omniscient narrator who we trust has known all along how things would unfold. We are in good hands says the faceless voice before guiding us to the inevitable conclusion. God has a plan.
Now meanwhile, the research is still going on. Little bits are being constantly resculpted. The anonymous reviewers, when they are finally helicoptered in, lay siege to the fortress we’ve been building and launch their ballistics into the citadel, requiring great feats of reconstructive masonry. This constant adding and removing, accretion and erosion, is the actual process of research. But it is not the research that is demanded of us now, only the conventional story about it. So like someone caught in a lie, we change our story in response to the reviewers hoping to present a coherent and impenetrable front. An antagonistic review system must lead to dishonest and obscure writing.
…
It seems to me that knowledge accumulated, accreted in this way has something more like a network than a narrative structure. Wikipedia is a helpful guide here. One concept, more or less developed itself, links to another and another. One node in the graph, becoming too big itself splits into several interlinked. The process of learning is that of enlarging and splitting the nodes in our developing web.
A paper then, a focused research project, is the development of a small subset of the huge, boundless graph representing and connecting all knowledge. To narrativize this, we must string out the nodes in this subset in sequence. The sequence we choose though is only ever one possibility among many, the subset itself being again only one.
This is why the narrative for me has always seemed so squirrelly. If the topic is a subgraph and the narrative is an arbitrary ordering then the critique confronts a story which was only ever a make-do contrivance. Picture a graph flattened into a sequence. See the tension among the edges now stretched. See that an optimal arrangement may exist but yet understand that tension can be minimized but only minimized and not reduced beyond that point.
