The manuscripts that take longest to edit are rarely the sparse ones. They are the ones where every paragraph carries the same weight, background, context, caveat, data point, aside, so that nothing on the page signals which sentence is the point and which is scaffolding. The writer has done the research. The writer has done the thinking.
What’s missing is the harder decision: which of it the reader actually needs.
The instinct to over-explain
Writers who know a subject well tend to over-explain it, not under-explain it. This looks like generosity, more context, more nuance, more of the reasoning laid bare, but it usually has the opposite effect. Chip Heath and Dan Heath named this pattern the “curse of knowledge” in a widely cited 2006 Harvard Business Review piece, describing how executives who have spent years immersed in a strategy’s logic default to abstract language, because to them the abstraction is already loaded with meaning. Employees hearing the same phrase get nothing, because they weren’t there for the years of context that made it concrete.
Editors see a version of this in almost every long manuscript, academic or commercial. The writer has internalized ten years of a field’s assumptions, and the draft reads as if the reader has too. Steven Pinker made the same argument the foundation of his writing guide The Sense of Style, telling a Harvard audience in a 2012 lecture that the curse of knowledge is “the failure to understand that other people don’t know what we know.” His remedy wasn’t to add more explanation. It was to test the draft against an actual reader and cut whatever that reader didn’t need to follow the argument.
What working memory actually allows
There’s a hard ceiling behind all of this, and it isn’t about attention span or patience, it’s about capacity. George Miller’s classic 1956 paper, still cited in cognitive psychology as one of its foundational findings and archived by the Max Planck Institute for Psycholinguistics, established that people can hold roughly seven items, plus or minus two, in working memory before earlier ones start dropping out to make room for new ones. Every sentence a reader processes is competing for a small, fixed amount of mental space.
That has a direct editorial consequence. A writer who adds a fourth qualifying clause to an already complex sentence isn’t adding precision, they’re pushing something else out of the reader’s grasp. In manuscripts we work on, the sections readers describe as confusing are almost never the ones missing information. They’re the ones carrying more active threads than a reader can track at once, so the thread that mattered gets crowded out by three that didn’t.
Detail as evidence of effort, not clarity
There’s a specific version of this in business and data-heavy writing: detail used to signal rigor rather than to inform. A report with every available figure, every caveat, every sub-finding reads as thorough to the person who wrote it, because they know which numbers matter and which are there for completeness. The reader doesn’t have that map. Cole Nussbaumer Knaflic’s Storytelling with Data built an entire framework around this problem for data visualization specifically, arguing that every additional element on a chart, every gridline, every label, every extra data series, imposes cognitive load on the person trying to read it, and that clarity comes from deciding what to remove, not what to add.
The same logic applies to prose that reports on data rather than displaying it. A paragraph that lists six supporting statistics when the argument only needs two doesn’t read as more convincing. It reads as unsorted. The writer has done the work of gathering the material and skipped the harder work of ranking it.
The question that replaces completeness
The editing decision underneath all of this is not “is this true and relevant,” which almost everything a competent writer includes will pass. It’s “what will this specific reader do with this specific fact.” Will they act on it, remember it, use it to follow the next paragraph, or is it simply true and adjacent to the topic? Manuscripts improve fastest when a writer starts asking that question during drafting rather than leaving it for revision, because it changes what gets written down at all, not just what gets crossed out later.
This is different from simplifying. Simplifying strips out complexity that the reader would benefit from. Selecting for use keeps the complexity that changes what the reader does or believes, and cuts the complexity that only demonstrates the writer did their homework. A legal memo, a research summary, and a client-facing report all fail for the same reason when they fail: they were organized around what the writer knew rather than what the reader needed to act.
The manuscripts that arrive already edited
Writers who finish clean, publishable drafts tend to have made this decision before they started typing, not after. Their manuscripts don’t have less material behind them, often they have more, because they’ve done enough research to know which ten percent of it changes the reader’s understanding. What separates them from writers who stall in endless revision isn’t discipline in the abstract. It’s that they’ve already asked, sentence by sentence, whether the reader will use this or simply receive it.
The manuscripts that take longest to edit still arrive with everything the writer learned. What they’re missing is the one decision that would have made the research worth reading: not what to include, but what the reader is going to do with it.
