Open a working manuscript at the stage before anyone else has seen it and the opening rarely resembles what ends up printed. There is often a paragraph of throat-clearing before the writer finds the sentence that should have started the chapter. A character says something in a voice that will later belong to someone else. A scene runs two pages past the point where it should have ended, because the writer was still finding out what the scene was for. None of this survives to the final draft. All of it was necessary to get there.
Editors see this constantly, because we are often the first people, after the writer, to see a manuscript in something close to its raw state. What looks like carelessness on page one is usually the opposite: a writer who understood that the first pass through a piece of writing is not for readers. It is for finding out what the piece is actually about.
The draft that isn’t for anyone
Anne Lamott gave this idea its most durable name in Bird by Bird, her 1994 book on writing, where she describes the “down draft” (get it onto the page, however badly), the “up draft” (fix it up) and the “dental draft” (check every sentence the way a dentist checks every tooth, for what is loose, cramped or decayed). The phrase that stuck from that book is blunter: the shitty first draft. Lamott’s point was not that bad writing is fine. It was that bad writing is where writing starts, and pretending otherwise is what stalls people before they finish anything.
That distinction matters more than it sounds. A writer who believes the first draft has to be readable will spend an afternoon on one paragraph, trying to make it presentable before moving on. A writer who has accepted that the first draft is not readable, by design, moves through the same afternoon and comes out the other side with a full chapter that says roughly what it needs to say. The second writer has more to work with. The first has a beautifully polished opening and nothing after it.
What the research on revision actually shows
This isn’t just writing-workshop folklore. In 1980, the composition researcher Nancy Sommers published a study, “Revision Strategies of Student Writers and Experienced Adult Writers,” comparing twenty university freshmen with twenty working writers, journalists, editors and academics, from Boston and Oklahoma City. Each wrote three essays and revised each one twice, producing nine drafts per writer in total, along with interviews about what they were doing at each stage.
The difference wasn’t polish. It was what revision was even for. The students mostly treated it as word-swapping: finding a “better” synonym, cutting a repeated word, tidying a sentence that already said what they meant. Sommers called this their “thesaurus philosophy of writing.” The experienced writers did something structurally different, they used early drafts to discover what they thought, then revised at the level of argument and structure, not just vocabulary, often rewriting entire sections because the draft had shown them the piece needed to say something else. Revision, for them, was recursive: it happened throughout the process, not as a cleanup stage tacked onto the end.
That gap explains a lot of what separates writers who finish long projects from writers who don’t. The ones who finish have made peace with producing something unusable in the first pass, because they know the unusable version is what the usable one gets built from.
Permission is not the same as sloppiness
Where this gets misread is in treating “write badly on purpose” as license to never revise at all. It isn’t. The discipline in a shitty first draft is that it still gets finished, still gets reread with real scrutiny, and still gets rebuilt, sometimes from the ground up. What makes it a discipline rather than an excuse is the writer’s willingness to do the harder second half of the work: reading their own bad sentences honestly and cutting the ones that don’t earn their place.
In manuscripts we work on at Global English Editing, the tell isn’t rough language in an early draft, that’s expected and, frankly, welcome, because it usually means the writer was moving forward instead of stalling. The tell of a writer who never learned this discipline is different: a manuscript where the first three chapters are meticulously worked and the back half is thin, rushed or unfinished, because all the effort went into making the beginning presentable rather than getting the whole thing down.
Why the rough version reads as more trustworthy now
There’s a reason this permission to write badly has taken on new weight recently, and it has to do with what readers are now suspicious of. As AI-generated text has become harder to distinguish from human writing at a glance, small imperfections have started to function as evidence rather than embarrassment. A University of Michigan School of Information piece from May 2026 quotes UMSI professor Nicole Ellison, noting that as AI-generated writing becomes more common, small mistakes may read less like carelessness and more like proof of human writing.
It’s worth being precise about what that trend does and doesn’t support. A typo left in a finished, published piece is a different thing from a first draft that nobody else reads, one is a signal aimed at an audience, the other is a private working method. But the underlying instinct is the same one Lamott and Sommers were describing decades earlier: writing that has visible seams, false starts and evidence of a mind working through a problem reads as more credible than writing that arrived too smoothly. What’s shifted is that this preference, once mostly an internal permission writers gave themselves to get past page one, has become something audiences actively look for in finished work. The private discipline and the public taste have started to point the same direction.
What the discipline looks like on the page
Practically, this shows up as a specific, learnable habit rather than a personality trait. Writers who finish tend to draft forward without stopping to fix the sentence behind them, accepting that the sentence might be wrong, or that the paragraph might get cut entirely once the fourth chapter reveals what the first chapter should have said. They separate the act of generating text from the act of judging it, and they do this on purpose, not because they lack standards. The standards get applied later, in full, often more severely than a writer who edited as they went would ever apply them to themselves.
What distinguishes this from genuine carelessness is sequencing. The bad draft comes first and gets treated as raw material. The judgment comes second and gets applied without sentimentality. Writers who skip the first step often produce careful, lifeless openings they can’t bear to build on. Writers who skip the second step produce drafts that never become anything else. The ones whose work editors end up admiring have simply refused to collapse the two steps into one.
That rough page one, still sitting somewhere in a writer’s files, thrown out before anyone else saw it, was never a failure of the finished book. It was the only way to find out what the finished book was going to be.
