Almost all contemporary writing advice is based on the idea, stated otherwise as "writing is rewriting," that writing is an open-ended process — that it begins with something rough and then turns out different versions that are increasingly polished. Eventually, it reaches a point where rewriting yields diminishing returns — or the writer hits a deadline and runs out of time — and the revision process stops. The work is published, and then it hopefully attracts enough attention to be read and re-read for as long as possible.
Writing has probably always involved some degree of rewriting. But as we have moved to digital media, rewriting plays an even more central role in the writing process than it did before.
First, of course, is how easy the computer makes it to revise and rewrite text. Whereas in the past, rewriting meant shuffling pieces of paper around, cutting things up and crossing them out, laborious copying, and so forth, rewriting is physically now just as easy as writing a first draft.
Besides this, though, other developments have accelerated the trend of writer versioning. For one, digital publishing means that you can update something as many times as you want — even after it has been published in a more traditional sense. It's trivially easy, for example, to make changes to a blog article once you've published it online.
At the same time, the creation of text content has become increasingly social. Creators and authors are becoming more comfortable with the idea of remixed or resampled content — content where different authors are involved in creating different versions. This extends, of course, beyond the written word, to other media that depend upon it, such as comics, podcasts, and movies.
We're living in an age where everyone from independent writers to companies like Disney are creating complicated shared universes and media franchises. Some of these properties have even come to mean the management of parallel universes and timelines, all being run simultaneously and side by side. And while they encompass a broad range of media, they all are defined by outlines, scripts, and characters that must, at some point, be rigorously defined and understood in the written word. The most obvious and successful example of this is the Marvel Cinematic Universe, a new version of the superhero universes made for comic books. There's the Kelvin and Prime timelines of Star Trek. There's even some suggestion that Star Wars will follow a similar path. To some degree, these properties represent an evolution of the worlds established in William Faulkner's Yoknapatawpha County or in J. R. R. Tolkien's Middle-Earth — especially in how they seek to establish rigorous continuities while allowing for multiple ongoing projects with many different contributors.
So how can writers understand their work in an age of endless versions? There are two key strategies for adapting to this age.
The first strategy involves reimagining the writing process as an act of creating a series of dependent layers of outlines and tables. This not only helps with the integration of longer pieces but also supports the production of companion content and even collaboration between writers.
The traditional model of writing successive drafts of a single piece conceals the fact that in every rewrite core things usually stay the same (or change less frequently) and superficial items are more likely to change.
Thus, any versioning process, even if you track all the changes in git or some other version control system, will only give you a superficial idea of what has changed.
Thinking about writing as a layered process allows for greater visibility into what revision actually means.
When we start writing, we often begin by writing an outline. The outline is thought to be a kind of simplified, easy-to-manipulate map of the content, laying out the most essential pieces and putting them in order. But even there it fails because it doesn't give a three-dimensional picture of what is important and what isn't.
Ideally, though, an outline like this would let us see more than the order we need to put our ideas. It would be separable into different tables, which could reveal not just what things are likely to change based on each draft, but what dependencies exist between different aspects of the composition. This can help writers understand the vital relationships between different parts of the project — as well as what dependencies need updating when deeper, more fundamental parts of the text are changed.
Working in this way makes it easier for writers to navigate the writing process, and provides support not unlike the meta-data which can be stored in complicated writing apps like Scrivener and Ulysses. Besides helping organize larger writing projects like a novel or a thesis, however, it can also help unify the ideas of something less traditionally integrated like a blog, a set of essays, an anthology, or even a media franchise.
Having content documented in layers can also help with the development of new companion pieces that can also be more fully integrated with each other.
The second strategy is to keep track of document versions by counting down instead of up. Doing so helps writers get more perspective and control over the versioning process without relying so much on deadlines. It also affords readers more transparency when an already published work is still undergoing a versioning process.
For those who want to produce timeless writing, the goal is to get a piece of writing to the point where it can no longer be improved. However, recognizing when this happens is easier said than done.
When using a computer, you can potentially create an infinite number of drafts when writing and rewriting, and it can be hard to know when to stop.
Of course, at some point, writers reach a point of diminishing returns. It doesn't make sense to continue polishing something indefinitely. At some point, you don't make things better, you just make them different. Writers are expected, one supposes, to rely on intuition to reach this point. But even experienced writers can have trouble letting go of a work like this.
Because of this, writers frequently turn to deadlines to help them bring an end to the writing process. This can be an effective solution, but it's also external to the act of writing. Instead, it sometimes amounts to an abandonment of the writing project, rather than something that involves an assessment of just how much a work needs to be polished.
As an alternative, I propose limiting the number of drafts one is allowed to make at the beginning of the writing process — that is, by establishing a kind of draft budget so that, once your writing has gone through a predetermined number of iterations, further drafts or versions would not be allowed. Doing so would have the benefit of making writers more conscious of how they work through each iteration. It would force them to think, to reflect, to make sure that what they're working on in this current draft will be enough to ensure that the writing project could be completed, not within a specific period of time, but through the remaining possible versions.
In the digital realm, of course, revisions can be made after publication, but by exposing the remaining budget of versions, it enables a more robust post-publication revision process. If readers can see how many versions are still possible, writers can share their work earlier in the writing process — and revise previously published work in an open, candid way.
Mechanically, a system like this could work similarly to the so-called "Kelvin Versioning" system that's being used in the Urbit project.
At the beginning of the project, you set your version budget. For example you might say that a certain blog article should be allowed to be revised 8 times. After this, each time you complete a major draft of the article, you reduce this number by one.
In a layered writing project, as described above, tables, outlines, and other documents that are lower in the dependency chain will also be tracked by a version budget. When you make a new draft of one of these lower layers, you'll also reduce its version number — and the version number of any layer or piece of content that depends upon it.
Once numbers get low, the revision process will likely slow down. You'll be spending more time polishing superficial things. Changes may still be possible, but it's not critical to make new versions of the document.
Once this number reaches zero, however, the piece of writing or the layer below it is frozen. It can be changed no more.
Writing, of course, is a complicated process and is often driven by inspiration as much, if not more, than the left-brained demands of outlines and format writing. Any approach to writing needs to take this into account — and a three-dimensional writing process like the one we described is as likely to mean outlines and tables are created after early drafts as they are beforehand.
That said, this approach addresses many of the needs of traditional publishing as it adapts to the digital era. It not only gives writers more control over the entire writing and publishing process, but, by exposing more of the writing process to collaborators, makes it easier for writers to work together.