The instinct when a section isn’t working is to edit it. Add a sentence to clarify. Move a paragraph. Find a better word for the idea that isn’t landing. This instinct is almost always wrong.

Editing a weak section improves its sentences. It rarely improves its logic. The result is a better-written version of the same unclear argument—which reviewers still flag, editors still hesitate over, and authors still don’t understand why.

The rewriting strategies that actually work start from a different assumption: if a section isn’t working after normal revision, the problem is structural, not verbal. You don’t need to find better words. You need to figure out what the section is actually supposed to do, and then rebuild it from that function outward.

Why Tweaking Sentences Doesn’t Fix Weak Sections

A weak section in a manuscript typically has one of three underlying problems.

The first is a missing claim. The section contains relevant information but no stated argument. It tells the reader what was found or what was done without telling them what it means. Editing improves the prose while the missing claim remains absent.

The second is structural incoherence. The section covers multiple ideas that don’t follow from each other. The paragraph sequence feels arbitrary because it is—the content was assembled rather than constructed. Editing individual paragraphs doesn’t address the sequence.

The third is scope mismatch. The section is trying to do too much, or is carrying content that belongs elsewhere in the manuscript. Every paragraph may be individually defensible; the problem is that they don’t belong together. Editing can’t resolve this because the problem is architectural.

In all three cases, the fix isn’t editing. It’s tearing down the section and rebuilding from its purpose. Why Academic Writing Is a System, Not a Skill describes why writing decisions that feel local—a sentence, a paragraph—are usually symptoms of structural decisions made upstream. The same logic applies to rewriting.

The Rewrite Workflow

When a section isn’t working after two passes of revision, the workflow shifts.

Step 1: Stop editing. Close the section. Don’t try to fix it by looking at it. The act of looking at existing prose anchors your thinking to what’s already there.

Step 2: State the section’s function in one sentence. What is this section supposed to do for the reader? Not what information it contains—what logical work does it perform? A Discussion section doesn’t exist to report findings again. It exists to interpret them: to explain what they mean, what they change, and what they don’t resolve. If you can’t write that function sentence, the section has no governing logic and editing won’t create one.

Step 3: List the three to five points that must appear. Given the function, what specific claims or moves are necessary? If the Discussion needs to interpret the primary finding, acknowledge the key limitation, and position the result relative to prior evidence—those are three distinct logical moves. Write them as a list, not as prose.

Step 4: Delete the original section. Move it to a holding document rather than erasing it—you may recover sentences from it—but remove it from view. Write the section fresh from the list of necessary moves.

Step 5: Check against the original. After drafting, compare against the deleted version. The goal is not to recover what was lost but to confirm that nothing essential was missed. Usually, the new version is shorter, clearer, and contains more actual argument.

The full process takes longer than another revision pass. It reliably produces better output.

The Section Teardown Framework

Use this before rewriting any section that has survived multiple revision passes without improving:

Section Teardown Framework

• [ ] Name the section’s function: Write one sentence describing what logical work this section must do (not what it contains)
• [ ] List required moves: Identify 3–5 specific claims or arguments the section must make to fulfill its function
• [ ] Diagnose the failure: Which problem describes the current draft? (a) Missing claim — (b) Structural incoherence — (c) Scope mismatch
• [ ] Move the original: Copy the existing section to a separate document; do not delete permanently
• [ ] Rewrite from the list: Draft from the required moves, not from the existing prose
• [ ] Recover useful sentences: Go back to the original and extract any sentences or evidence that belong in the new draft
• [ ] Verify the function is fulfilled: Does the final section do the logical work named in step 1?

What This Changes in Practice

The Real Reason Papers Feel ‘Fragmented’ identifies fragmentation as a structural problem, not a writing problem. A fragmented Discussion isn’t fragmented because the writing is poor—it’s fragmented because the logical moves weren’t identified before the section was written.

The teardown workflow forces that identification after the fact. It’s slower than editing. It’s also the only approach that addresses the actual problem rather than its surface symptoms.

The willingness to delete a section you spent time writing is one of the harder disciplinary habits to develop in academic writing. It becomes easier once you’ve done it enough times to see that the replacement is almost always better—and that the original was preserved, not lost.