Metis-First Editing

From Sam[ ]zdat to Ribbonfarm, one of the ideas underlying much of the criticism of rationalists (and by extension, the modernists) is James C. Scott’s concept of metis, advanced in 1998’s Seeing Like A State.

The idea is grounded in two priors: first, that the world is complicated; second, that the complexity of worldly domains is substantially hidden to outsiders. Because of this hidden complexity, governing bodies impose systematic order on localities; this new order appears “rational,” on its face, due to its internal coherence of logic, but will inevitably aggravate the problem it set out to solve because it did not take into account this local complexity—its “unknown unknowns” or hidden variables. Scott refers to this orderly, top-down, abstract knowledge as episteme. Metis, meanwhile, is generated bottom-up by locals more aware of the environmental constraints which determine the success or failure of an attempted solution. Often, it arises in an evolutionary process where it is repeatedly tested against reality to ensure proper functionality, and in the event of failures, has been modified accordingly. This process makes metis solutions appear messy, and ad-hoc; often, locals cannot even fully describe or justify the rationale behind it, since many of its complexities have been evolved by previous generations. (A good solution “hides” the problem it solves, since the undesirable outcomes no longer surface.) This “unjustifiable nature” of metis solutions leads to their casting as irrational by outsiders.

Some relationships which produce this metis-episteme dynamic include that between architect and inhabitant, government and citizenry, empire and colony, & parent and child. A further instance is the relationship between editor and writer; here, I sketch out a few ways that an editor might approach editing in a metis-friendly fashion, avoiding the pitfalls that come with fetishizing episteme.

Why should I be a metis-first editor?

In most cases, the writer will have spent an order of magnitude more time than his editor considering the work’s subject matter and structural organization, as well as his general authorial style. There are clear benefits to outside views in escaping biases and expanding frame of reference, but this discrepancy in time still means that tampering with a piece of writing in significant ways should be approached with caution—a natural respect for the complexity of the problem and its inevitable hidden constraints.

Not all writing deserves this amount of caution. Much writing, especially on the blogosphere, is similar to the piece you’re reading now: thematically simple, linguistically uncomplicated, and without creative or artistic flourish. Most of what’s happening in such pieces of writing is apparent on the surface; the emergent complexity of parts and whole is lower. But this is decidedly not the case for writers such as Randall McCloud. (McCloud’s “Fiat Flux,” linked, is one of the most incredible pieces of literary criticism I’ve come across; I had no interest in poetry, let alone medieval poetry, when I first encountered it, but spent the rest of the evening reading it in entirety.)

1. Read the damn thing first.

Some editors begin immediately marking a text upon starting to read it—not just descriptive marks, which can be helpful in recording a first impression, but prescriptive marks— attempts to shape and sand away at a form they do not yet have any real concept of. These prescriptive marks commit an editor, psychologically; they begin a pattern of shoulds that will be unconsciously perpetuated & tiled throughout the piece as a matter of taste. Or, perhaps, the editor will begin unstitch the thing so immediately on first pass that he never actually understands how it was put together in the first place. To phrase the problem in the terminology of Christopher Alexander: without a concept of the gestalt whole, the fitness of any individual part cannot be assessed.

2. Writing is also subject to Chesterton’s Fence.

Some of a writer’s decisions are made automatically, by default, or because the writer knows no other way. For an even moderately skilled writer, however, many of the decisions are decisions proper, made for reasons which are not readily transparent to an editor. Reading thoroughly and closely is one way to figure out the “why” behind a “what.” Communicating directly with the writer is another, even better way. Misspellings can be intentional puns; talky-casual style can be ideologically grounded. “If I can’t tell it’s intentional, it isn’t effective!” Such an argument forgets that published, edited work is typically ascribed by readers with significantly more intentionality than drafts are read by editors.

3. Make the editing history available to, and verified by, the writer.

This is often avoided by editors trying to avoid conflict, either because they lack the time for it (e.g. they work for a publication on deadline), or because they want to avoid the emotional and cognitive work of 1) explaining the rationale behind an editorial decision, 2) being open to their edit having inadvertently diminished the piece, and 3) finding instance-specific resolutions when intractable differences in motivation & taste arise between parties. Still, due to the complexity of a well-written piece of sufficient length, errors are inevitable and should be caught.