1.13 Revision Based Craft

Revision-based craft is the understanding of revision not as correction of mistakes but as the primary arena in which a novel is genuinely written. In this understanding, the first draft is raw material — necessary but not sufficient — and the art of novel writing is exercised principally in the work of revision: the sustained, intelligent reassessment and rebuilding of draft material to close the gap between what was intended and what was achieved, between what exists on the page and what the novel needs to be.

The distinction between draft-writing and revision is crucial. Draft-writing is generative — it produces raw narrative material, often through a process of discovery in which the writer does not fully know what the novel is until it has been written. Revision is reflective and constructive — it assesses that material from the perspective of a reader, identifies its strengths and weaknesses, and rebuilds it according to an increasingly clear understanding of what the novel should be. These are different cognitive activities requiring different orientations, and conflating them — trying to revise while drafting, or trying to generate while revising — typically impairs both.

The Conception of Revision as Re-Seeing

The word "revision" derives from the Latin roots meaning "to see again," and this etymology points to its essential nature. Revision is not proofreading, not mechanical correction, not the smoothing of rough edges. It is genuinely seeing the material again — returning to the draft with fresh eyes and discovering what is actually there, as opposed to what the writer intended or imagined was there.

This re-seeing is difficult precisely because the writer who produced the material knows it too well. They know what was meant, what was intended, what was imagined in the writing. This prior knowledge makes it easy to read into the draft meanings that are not present on the page, to hear voices that are not audible without the authorial prior knowledge, to find coherence in a structure that a reader without that prior knowledge would find disordered.

Effective revision requires developing the ability to read one's own draft as a stranger would — to experience it as a reader encounters it, without the author's prior knowledge of its intentions and meanings. Strategies that support this include:

Temporal distance: Reading the draft after a significant period (weeks or months for a novel) allows the writer's detailed memory of producing it to fade enough that they can read it more nearly as a stranger. Many novelists report that returning to a draft after several months reveals problems they could not see when they were closer to the writing.

Reading aloud: Reading the prose aloud slows the reading process enough to hear problems that the eye skips over, and it engages auditory perception in a way that surface reading does not. Prose that sounds awkward when read aloud is typically awkward in ways that readers also feel, even when reading silently.

Physical medium change: Reading a manuscript on paper rather than on screen, or printing it in a different font, creates enough perceptual novelty to interrupt the automatic filling-in of intention that occurs when reading in the familiar writing environment.

Reader responses: Hearing how actual readers respond to the draft — where they are engaged, where they are confused, where they are bored or unconvinced — provides information that no amount of internal re-reading can replicate. The reader's experience is the ultimate measure of revision success.

The Hierarchy of Revision

Revision-based craft operates across multiple levels, and effective revision attends to these levels in sequence rather than simultaneously. Attempting to address all levels of the manuscript at once typically produces revision that is thorough at the sentence level while blind to larger structural problems — a common failure mode in which writers polish sentences that will ultimately need to be cut.

Structural revision addresses the manuscript as a whole: its overall shape, the proportions of its parts, the trajectory of its character arcs, the architecture of its plot. Structural revision asks: Does the story begin in the right place? Is the central conflict clearly established early enough? Are the character arcs fully developed? Does the pacing serve the story — neither too slow in setup nor too rushed toward climax? Does the subplot integrate meaningfully with the main plot? Does the ending pay off everything the novel has established?

Structural revision typically requires the most significant alterations to the manuscript — cutting scenes or chapters, adding new material, reorganizing the sequence of events, fundamentally rethinking character development. It should be addressed before scene-level or prose-level revision, because addressing structural problems may require cutting material that would have been polished unnecessarily, and because understanding the structure clearly changes what scene-level and prose-level work is needed.

Scene-level revision addresses each scene individually: its function, length, point of view, internal structure, and effectiveness. For each scene the revising novelist asks: Is this scene necessary? What does it accomplish — does it advance the plot, develop character, deepen theme, or some combination? Does something change during this scene? Is the point of view the most effective choice? Is the scene the right length — neither too brief to accomplish its purpose nor too extended relative to its importance in the narrative? Does the scene end in a way that propels the reader forward?

Scene-level revision may involve significant rewriting of individual scenes — altering their structure, changing their point of view, adding or removing material — but operates at a more local scale than structural revision and typically occurs after the manuscript's overall shape has been established.

Prose revision (also called line editing) addresses the manuscript at the level of the sentence and the word: clarity, precision, rhythm, concision, the elimination of redundancy, the pursuit of the exact word rather than an approximate one. Prose revision asks: Is this sentence clear? Is this the most effective order for these words? Is this sentence longer or shorter than it should be? Is this word doing the work it needs to do, or can it be replaced with something more precise? Is this paragraph rhythmically varied, or has it settled into a monotonous pattern?

Prose revision is the most locally focused level of revision work and is typically addressed last — after structural and scene-level concerns have been resolved — because prose improvements in scenes that will ultimately be cut are wasted effort.

Proofreading is the final check for mechanical errors — typos, grammatical errors, punctuation inconsistencies, formatting problems — that have survived the revision process. It is distinct from revision proper in that it addresses surface correctness rather than craft.

Diagnostic Revision

Revision-based craft includes a specific diagnostic skill: the ability to identify what is wrong with a piece of writing — to name the problem accurately enough that effective solutions become apparent. Many revision struggles occur not because the writer cannot improve their prose but because they have diagnosed the wrong problem and are therefore applying solutions that address something other than the actual difficulty.

Common diagnostic challenges include:

Confusing symptom with cause: A scene that feels dull and inert is a symptom; its cause may be any of several things — lack of clear character desire, insufficient conflict, point of view too distant from the character's interiority, setting that does not contribute to the scene's meaning, dialogue that tells rather than shows, or simply a scene that should not exist at all. Identifying which cause produces the symptom is the diagnostic work; rewriting the scene without diagnosing the cause typically produces a differently written but equally inert scene.

Local revision for global problems: Revising sentences in a scene whose fundamental problem is structural (the scene shouldn't exist, or should be in a different place in the novel) is applying a local solution to a global problem. Structural problems require structural solutions; local polish cannot repair them.

Missing additions: Some manuscripts need revision that adds material rather than improves existing material. A character who is underdeveloped — who appears in scenes without sufficient interiority, motivation, or behavioral specificity — needs more scenes, not better-written scenes. A theme that is stated rather than dramatized — asserted rather than demonstrated through character action and consequence — needs additional scenes that dramatize it. Revision focused only on what exists can miss what needs to be created.

Revision as Craft Development

Revision is not only how novels are improved; it is how novelists develop their craft. The process of identifying what is wrong with a piece of writing, understanding why it is wrong, and developing an effective solution builds craft knowledge and judgment that transfers to future projects.

The writer who has revised a novel through multiple structural passes develops an increasingly refined understanding of structure — of how long different kinds of scenes can sustain, of how much setup is necessary before a reveal lands effectively, of how much parallel development a dual storyline requires to feel balanced. This understanding becomes available as intuition in drafting: the writer who has revised structural problems extensively begins to draft with an implicit sense of structure that requires less explicit revision.

Similarly, the writer who has revised prose extensively — who has worked at the sentence level to eliminate redundancy, increase precision, and improve rhythm — develops stylistic judgment that shapes the drafting prose from the outset. Craft knowledge built through revision progressively reduces the revision burden by improving the quality of the raw material that revision must work with.

The Emotional Dimension of Revision

Revision-based craft is not purely technical; it has a significant emotional dimension that experienced writers learn to manage. The experience of returning to a draft and seeing its problems clearly can be deflating — particularly when structural problems are significant, when large amounts of material must be cut, or when the gap between intention and achievement is large. The equanimity to look at one's own work clearly, assess its actual qualities honestly, and address its actual problems rather than the problems one wishes it had is a form of emotional craft as important as any technical skill.

Revision also requires the capacity to release attachment to specific material. The scene that was written with great pleasure and that contains some beautiful prose may nonetheless need to be cut because it doesn't serve the novel's structure. The passage that felt deeply true to the writer's personal experience may need to be removed because it serves the writer's needs rather than the novel's. The ability to cut material — even excellent material — when the novel's needs require it is a craft skill developed through practice, and one that most writers find difficult at first.

Darlings — the term often used for passages a writer loves but that don't serve the work — need to be "killed": cut from the manuscript. This doesn't mean the prose is destroyed; it can be saved in a cut file and may find its place in another context. What it means is that the writer's attachment to specific material must not override their judgment about what the novel needs. The capacity to make this distinction, and to act on it reliably, is one of the marks of a mature revision practice.

Revision Across Multiple Projects

Revision-based craft develops most significantly across multiple completed projects rather than within a single one. Each novel revised builds additional understanding of how novels work, what kinds of problems recur, and what kinds of solutions are effective. The third novel revised is approached with significantly more craft knowledge than the first, because three projects' worth of diagnostic and corrective experience has been accumulated.

This cumulative development is one of the arguments for completing projects even when they are not, in the end, publishable: the revision of a flawed novel builds craft knowledge that would not be available from revising only successful novels, because failure typically reveals the craft's demands more clearly than success. The novelist who abandons projects at the first-draft stage denies themselves the revision experience that is the primary engine of craft development.