Rolling revision is another one of those controversial writing techniques that many people discourage but a few people swear by. I am a rolling reviser, but I recognize that for a lot of writers, it has serious drawbacks, so I’ll try to talk about it from both sides.
First off, a definition: what I term “rolling revision” means that rather than writing a first draft straight through, making few to no changes once a scene has been completed, the writer is constantly fiddling with already-written sections, adding bits of suddenly-required foreshadowing or deleting planted information that has turned out to be unnecessary, and so on. It covers everything from massively rearranging scenes to minor tidying-up of awkward sentences; what makes it “rolling revision” is that the writer does it multiple times before the first draft is ever finished.
The disadvantage to this should be obvious: it is perilously easy for the writer’s forward progress to slow down and stop due to Endless Revision Syndrome. This happens often enough that many, many writing-advice-givers warn new writers very strongly to never, ever go back over their manuscripts until their first draft is entirely completed.
The equally obvious advantage is that the writer’s first draft, once it is finally completed, usually doesn’t need a whole lot of additional revising to get it to submission quality. This is not generally considered a strong enough reason to risk the aforementioned Endless Revision Syndrome, except in the case of those few writers who absolutely cannot bring themselves to go back over a draft once they have written “The End.”
I’ve been a rolling-reviser since my earliest writing, back in the Jurassic Era before computers and word processors (and believe me, it is No Fun At All being a rolling-reviser when “cut and paste” means spending half an hour physically cutting your pages apart and then taping them back together with the paragraphs in a different order). It is part of my process because my backbrain simply will not cooperate if it isn’t really, really sure that what I have already written is a solid foundation for whatever is currently at the leading edge of the story.
For instance, in the current WIP, I started a new scene in Chapter 11 and a piece of unexpected backstory showed up for one of the characters. I realized instantly that a) this was a really cool idea and fit right into the story and solved a bunch of plot problems, and b) if this backstory was true, a character back in Chapter 2 should have reacted very differently during their conversation. Theoretically, I could have made a note to “Change Ch 2 conversation between A and B to reflect backstory from Ch 11 p.1” and gone on working on Chapter 11…except that I couldn’t. Oh, I could force myself another couple of sentences forward, but none of them felt right.
So I went back and spent ten minutes fiddling with the conversation in Chapter 2, so that the reactions and the dialog were consistent with the backstory in Chapter 11 (not revealing it, but also not something that a reader would hit on their second time through and go “Hey, doesn’t it turn out later that A knows B’s secret? So why is A talking and reacting as if he/she doesn’t know it?”). And then I could go on.
While I was back there, I tightened up some other parts of the conversation and added some stage business, neither of which was strictly necessary to get that bit of backstory in, but as long as I was there and saw the opportunity, I just did it. I did not go over the whole scene looking for other potential revisions, and the whole fix didn’t take more than ten minutes to do and get back to work on the leading edge of the story.
Most of my rolling revisions are like this: they’re matters of plot, characterization, setting, or backstory that I realize are inconsistent with what I am currently writing, and that I have to fix before I go on. I can let everything else go (I didn’t have to tighten the rest of the conversation or add the stage business). Or there was the time when I called a character Andrew for three chapters, then switched to Anthony and didn’t realize it until five more chapters were done. There was no reason not to do a quick search-and-replace, so I did. I will take quick fixes if the opportunity arises, but I don’t go looking for them…unless I’m stuck.
If I’m stuck, then I will go back over earlier parts of the manuscript and fiddle with them, tightening and smoothing and improving the style and pace and so on, while my backbrain works out what to do next on the current scene. Usually, this is rather like warming up; it’s fifteen or twenty minutes at the start of a writing session, and then I’m ready to go. If I’m seriously stuck – like, for weeks – it’s usually because a) I’ve gone badly wrong somewhere, and I have to really dig into the earlier stuff to figure out where and fix it, or b) I’ve gone so badly wrong that I’m going to have to rip back several chapters, and I really don’t want to, so I’m resisting admitting what I secretly already know. For me, this manifests as a massive reluctance to write at all, rather than as a desire to fiddle with earlier chapters (because my backbrain knows that I’m going to have to rip that stuff up eventually, and doesn’t want to waste time fiddling with already-written stuff, but also refuses to waste time writing new leading-edge stuff that will just have to be ripped up because I haven’t yet fixed the really-wrong bit).
The key, for me, is that in all cases, going back over the story and fixing stuff results in more forward progress, faster, than trying to force myself to move forward without doing those revisions, and that forward progress keeps happening steadily. I don’t do rolling revisions on style and grammar unless they’re fast and easy/obvious; that way lies Endless Revision Syndrome. Consistency, plot points, characterization, backstory – those, I have to fix.