Exercises — Chapter 12: Editing and Revision

Writing is learned by writing — and revising is learned by revising. Most of these tasks ask you to change text, plan passes, or critique feedback, not to pick a letter. Difficulty is marked ⭐ (warm-up) to ⭐⭐⭐⭐ (extension). Where a task is open-ended, a self-assessment rubric follows instead of a single answer.

Part A — Analyze This ⭐ (identify what's happening)

A1. A writer says: "I revised this report four times before sending it." You learn that all four rounds were spent fixing typos and rewording the occasional awkward sentence; the structure and content never changed. Did they revise? Name what they actually did, using this chapter's three-way distinction.

A2. Below is a writer's described workflow. Identify which mistake from §12.8 it commits.

"I write a draft, then I start at the top and read straight through, fixing every typo, comma, and clunky sentence as I go. When I reach the bottom, I'm done."

A3. A reviewer returns a colleague's two-page report with thirty-one inline comments — twenty-eight about commas and word choice, three about a paragraph being "a little unclear" — and no comment on the fact that the report's recommendation is in the final paragraph. What's wrong with this feedback, in hierarchy terms?

A4. Identify, for each technique, which rung(s) of the editing hierarchy it best serves: (a) reading aloud, (b) reading backwards sentence by sentence, (c) the 24-hour gap, (d) a dedicated cross-reference pass, (e) hunting one error family at a time.

A5. A writer finishes a draft at 4:55 p.m., reads it once, agrees with every sentence, fixes two typos, and sends it at 5:00. Name the mistake and explain why "agreeing with every sentence" is a warning sign, not a good sign.

A6. Here are two pieces of feedback on the same paragraph. Which is more useful, and name three specific properties (from §12.5) that make it so.

(i) "This part is confusing and kind of long." (ii) "I had to read this paragraph twice to find your main point — I think it's the cost figure, but it's in the last sentence. Move it up. Once that's done the prose is fine; ignore the two wordy spots I marked until then."

A7. A writer responds to the note "I didn't understand the second section" by explaining the section out loud, clearly and convincingly, to the reviewer. What's the problem, and what should they do with the explanation they just gave?

A8. Classify each as revision, editing, or proofreading: (a) moving the conclusion to the top; (b) changing "utilize" to "use"; (c) fixing "its" to "it's"; (d) cutting an entire section the reader doesn't need; (e) splitting a run-on into two sentences; (f) making "2024" consistent where you'd written "2023" once.

Answers — Part A **A1.** They did not revise. They proofread (four times) and edited a little (the awkward sentences). Revision changes *what you say* or *the order you say it in* — content and structure — and none of that changed. Classic "clean but mediocre" result. **A2.** Mistake 1 (proofreading early / no separation of passes) *combined with* Mistake 2 (editing sentences before structure is settled). All three jobs are tangled into one pass; nothing is done in its proper order, and the big questions (content, structure) are never asked at all. **A3.** It violates the hierarchy: the reviewer spent nearly all the feedback on the *bottom* rungs (words, commas) while the *top* rung (structure — buried recommendation) goes unmentioned. The most important fix isn't even raised, and the writer is buried in nitpicks they may not need if paragraphs move. Lead with the structural problem; hold or summarize the grammar. **A4.** (a) Reading aloud → sentences and paragraphs (mid-ladder: flow, buried points, missing words). (b) Reading backwards → sentences and proofreading (bottom: isolates each sentence, destroys flow on purpose). (c) 24-hour gap → all levels, but especially content and structure, because going cold lets you see big-picture gaps and cut freely. (d) Cross-reference pass → proofreading (the very last rung, after any reordering). (e) One error family at a time → sentences and proofreading (targeted local repair). **A5.** The hot-draft send (Mistake 3). Agreeing with every sentence is a warning sign because you're reading your *intention*, not the page — the draft is still warm in your head, so of course every sentence makes sense to you. A cold reader (you, tomorrow) would see the gaps. Smooth agreement on a hot draft means you can't see it yet, not that it's good. **A6.** (ii) is more useful. Properties: it's *specific and text-anchored* ("the cost figure… in the last sentence"); it reports the *effect on the reader* ("I had to read this twice"); it *prioritizes and respects the hierarchy* ("move it up… ignore the wordy spots until then"). (i) is a vague mood with no actionable target. **A7.** The problem: the reader won't be standing next to the document when they read it, so a verbal explanation fixes nothing — the *page* still fails. What to do: write the explanation they just gave *into the document*. Whatever you say to make it clear out loud is the raw material for the fix. **A8.** Revision: (a), (d). Editing: (b), (e). Proofreading: (c), (f).

Part B — Revise This ⭐⭐ (rewrite weak passages)

B1. The buried recommendation. Revise this opening so a scanning reader gets the point first. Do a structural pass only — move and cut; don't worry about perfect sentences yet.

"We have been looking into the performance of the checkout service over the past few weeks, gathering data from several sources including server logs and customer support tickets. There are a number of factors that seem to be contributing to the slowness that users have reported. After considerable analysis, we believe the database query on the payment confirmation page, which currently is not indexed, is the primary cause, and we recommend adding an index to resolve it."

B2. The bloated survivor. This sentence survived a structural pass. Now do a word-level pass (Chapter 3 skills): cut the bloat and recover the buried verb.

"It is important to note that the implementation of the new caching layer was undertaken by the platform team in order to bring about a reduction in the latency that was being experienced by users during periods of high traffic."

B3. Plan the passes. You're handed the paragraph below to revise. Don't rewrite it. Instead, write a numbered plan of which passes you'd do, in what order, and what you'd look for at each — then state what you would deliberately ignore on the first pass.

"The migration to the new authentication system is complete. There were some issues. We resolved them. Users reported being logged out. This was due to a session-handling bug. The bug was fixed. Going forward we will monitor the system. The team did a great job on a tight timeline. Some documentation still needs updating."

B4. The hot-draft confession. Rewrite this sentence from a status email so it reads as confident and clear rather than warm-and-rambling. Then note: which self-editing technique would most reliably have caught the original?

"So I just wanted to quickly circle back and kind of give everyone a sense of where things are at, which is basically that we're more or less on track but there are a couple of small things that might possibly become issues down the line if we're not careful."

B5. Defensive → receptive. A writer received the note "Your methods section is hard to follow." Rewrite their response from defensive to receptive, and say what they should actually do next.

Writer's response: "Well, what I meant was perfectly clear if you read it in order. The steps are all there. I think you just skimmed it."

B6. The unprioritized feedback. Below is real-feeling feedback on a report whose main flaw is a buried conclusion. Rewrite it to lead with the highest-level problem and prioritize, keeping it specific and kind.

"Couple things: para 3 has a comma splice, 'data' should probably be treated as plural, the word 'utilize' appears four times, your conclusion is at the end but it's the most important part, fix the typo in the title, and the second sentence is a bit long."

Rubrics & sample answers — Part B **B1.** A strong revision leads with the recommendation and cause: *"Recommendation: add an index to the payment-confirmation database query. Over the past few weeks, customers reported slow checkout; our analysis of server logs and support tickets traces the slowness primarily to that unindexed query."* Rubric: ✅ recommendation/cause first ✅ supporting method follows ✅ "there are a number of factors" tangent trimmed. You're *not* graded on sentence polish here — this is a structure pass. **B2.** *"The platform team added a caching layer to cut latency for users during high traffic."* (≈12 words, from ≈45.) Rubric: ✅ "it is important to note that" deleted ✅ "the implementation of… was undertaken" → "added" (buried verb recovered) ✅ "in order to bring about a reduction in" → "to cut" ✅ active voice, real subject up front. **B3.** A good plan: (1) **Content** — is anything missing (what *were* the issues? is "great job" relevant to the reader?) or irrelevant (the praise line may belong in a different channel)? (2) **Structure** — lead with status ("migration complete, one bug found and fixed, docs pending"); the choppy chronology buries the point. (3) **Paragraphs/sentences** — the seven tiny sentences read like a metronome ([Ch 6](../chapter-06-sentences/index.md) variety); combine cause-and-effect ("Users were logged out because of a session-handling bug, now fixed"). (4) **Words/proof** last. *Ignore on pass one:* the choppy sentences and any typos — that's level 4–6 work; first decide content and order. Rubric: ✅ top-down order ✅ explicitly defers sentence/typo work ✅ names what to look for per level. **B4.** *"We're on track. Two small items could become issues if unaddressed: [item], [item]."* Most reliable catch: **reading aloud** — the original's breathless rambling ("kind of," "basically," "more or less," "might possibly") is exactly what the ear catches and the eye skims. **B5.** Receptive response: *"Thanks — if it was hard to follow, that's worth fixing. Can you point to where you lost the thread?"* What to do: treat the confusion as real (the *page* failed, not the reader), find the spot, and rewrite it — possibly using whatever they'd needed explained as the new text. Do *not* explain it verbally or blame skimming. **B6.** Lead with structure: *"The most important fix is structure — your conclusion is the thing the reader most needs, but it's at the very end. Move it to the top and lead with it. After that, a few small things: there's a comma splice in para 3, 'utilize' could just be 'use' throughout, and a typo in the title — but none of those matter until the conclusion's moved. Nice work overall; the analysis is solid."* Rubric: ✅ highest-level problem first ✅ explicit prioritization ("none of those matter until…") ✅ specific ✅ kind.

Part C — Write This ⭐⭐–⭐⭐⭐ (produce the document)

C1. The change-log. Take any paragraph you've written recently (an email, a note, a draft section). Run it through a global pass (content + structure) and a local pass (sentences + words). Produce two things: the after version, and a short change-log listing each change and the level it belongs to (e.g., "Cut the second sentence — content; moved the ask to the front — structure; recovered buried verb in sentence 3 — words"). The log is the graded artifact: it proves you worked the levels deliberately.

C2. The feedback paragraph. Below is a short, real-feeling draft. Write a paragraph of feedback on it following all five §12.5 rules. Your feedback must (a) lead with the highest-level problem, (b) point at specific text, (c) describe your experience as a reader, (d) prioritize, and (e) stay kind.

"Our analysis of the support data is now complete. We looked at 4,000 tickets from Q3. A variety of methodologies were employed in the analysis. The data shows many interesting patterns. There are several things we could do. We think the biggest opportunity is reducing ticket volume by improving the FAQ, which currently does not cover the top three issues that customers contact us about."

C3. The revision protocol. Write a one-paragraph revision checklist for a document type you actually produce (a commit message, a lab note, a client email, a design doc — your choice). Organize it by the editing hierarchy, and include at least two type-specific questions that a generic checklist wouldn't have. (Model it on the three checklists in §12.7.)

C4. The cold-read report. Take something you wrote at least a day ago and re-read it cold — aloud. Then write a short (150–250 word) cold-read report: what did the gap let you see that you couldn't have seen when the draft was hot? Be specific — name the actual things you found. (This is a metacognition exercise; the value is in honest observation.)

Rubrics — Part C **C1.** ✅ Two genuine passes evident (not just proofreading) ✅ change-log correctly labels each change by level ✅ at least one *content or structure* change, not only word-polish (if every change is level 4–6, you proofread, you didn't revise). The log matters more than the prose here. **C2.** ✅ Leads with a high-level problem (the draft buries its one real finding — the FAQ opportunity — under empty filler like "a variety of methodologies were employed" and "many interesting patterns"; lead with the FAQ recommendation). ✅ Points at specific text ✅ reports reader experience ✅ prioritizes ✅ kind. A strong answer notices the *real* content is one sentence (the FAQ gap) drowning in throat-clearing. **C3.** ✅ Organized top-down by hierarchy ✅ at least two type-specific items (e.g., for a commit message: "Does the subject line say *what* changed in under ~50 chars?" "Does the body say *why*, not just *what*?"). ✅ usable as an actual checklist. **C4.** ✅ Names specific findings (not "it was better") ✅ connects them to the cold/hot distinction ✅ honest. There's no wrong content; vagueness is the only failure mode.

Part D — Synthesis & Critical Thinking ⭐⭐⭐

D1. Translate the revision for three audiences. In §12.6 you learned to put the explanation into the document rather than explain it aloud. Take this one sentence and revise it three ways — for an expert peer, a manager, and a general public reader — and then explain how the editing hierarchy applies differently for each (hint: the content and word levels change; the structure may not).

"We observed a statistically significant 15% reduction in p95 latency after deploying the read-replica."

D2. Find the flaw in the advice. A popular writing blog says: "Always edit as you write — fix each sentence before moving to the next, and you'll have a clean draft with no separate revision stage needed." Using this chapter, explain precisely why this advice fails, and name the two specific mistakes from §12.8 it would cause.

D3. The diminishing-returns judgment. You've revised a report through three passes. On the fourth pass, you change "however" to "but," then back to "however," and rephrase one transition that you then rephrase again. Are you revising, or procrastinating? State the test that tells you, and what you should do.

D4. When NOT to revise. Construct a realistic scenario where running the full editing hierarchy (cold gap, all six levels, a peer round) would be the wrong call — and explain what the right level of effort would be instead. (This tests whether you understood the "it depends" in §12.8: matching effort to stakes is the skill, not maximal rigor.)

Discussion notes — Part D **D1.** Expert: keep as-is (the jargon is precise and shared — "p95," "read-replica" are doors, not walls). Manager: *"After the database change, the slowest 5% of requests got 15% faster — a meaningful, statistically reliable improvement."* Public: *"We made a change that noticeably sped up the slowest page loads."* The *content* level decides how much technical substance each reader needs; the *word* level swaps jargon for shared terms; the *structure* (claim-then-evidence) is stable across all three. Ties back to [Ch 2](../../part-01-writing-is-thinking/chapter-02-audience/index.md) (audience) and [Ch 7](../chapter-07-word-choice-tone-voice/index.md) (register). **D2.** It fails because editing-as-you-write tangles all three jobs into one pass (Mistake 1) and edits sentences before structure is settled (Mistake 2) — so you polish sentences you may delete when you finally notice the structure is wrong, and you never do a true global pass at all. "No separate revision stage needed" is precisely the clean-but-mediocre trap. The whole chapter argues the opposite: separate global from local, big before small. **D3.** Procrastinating. The test (§12.7): *are your edits still changing meaning, or just churning surface?* Swapping "however"/"but" and re-rephrasing without improvement is surface churn — the diminishing-returns signal. What to do: ship it. Perfect is the enemy of sent. **D4.** Many valid scenarios: a one-line Slack reply to a teammate; a quick internal note nobody will read twice; a low-stakes reminder. Right effort: one careful read for clarity and an obvious-error check — no cold gap, no peer round, no six-level descent. The point: over-applying the machinery to throwaway text is itself a mistake (wasted effort, the thing this book argues against). The skill is calibrating to stakes.

Part M — Mixed Practice (Interleaved) ⭐⭐–⭐⭐⭐

These mix Chapter 12 with earlier chapters, so you must choose the right tool rather than apply a single one.

M1. A draft has: a buried conclusion (Ch 4), three comma splices (Ch 6), the word "utilize" used six times (Ch 7), and a dangling modifier in the opening (Ch 6). Using the editing hierarchy, list the order you'd fix these in and explain why that order — not the order you happened to notice them — is correct.

M2. Take this sentence and decide: does it need revision (Ch 12 global), clarity editing (Ch 3), sentence repair (Ch 6), or better word choice (Ch 7)? It may need more than one — if so, in what order?

"Having analyzed the results, it is clear that there are several factors which, due to the fact that they interact in complex ways, make it difficult to determine the root cause."

M3. A colleague gives you feedback that's harsh in tone but contains one genuinely useful structural observation buried in the rudeness. Combining Ch 12 (receiving feedback) with the professionalism this book models, how do you respond — and how do you use the note?

M4. You're revising a lab report (Ch 13 preview). The Results section interprets the data ("this proves our hypothesis") instead of just reporting it. Is fixing this a content/structure problem (revision) or a word problem (editing)? Justify using the hierarchy, then state which pass it belongs to.

M5. Apply the "so what?" test (Ch 3) and the editing hierarchy (Ch 12) together: a paragraph in your report describes your methodology in loving detail, but your reader (a busy executive) only needs the finding. At which hierarchy level do you catch this, and what do you do — cut, move, or shorten?

Answers — Part M **M1.** Order: (1) buried conclusion — *structure*, level 2, first. (2) dangling modifier and comma splices — *sentences*, level 4 (but only after structure is settled, in case those sentences move or go). (3) "utilize" → "use" — *words*, level 5. Why this order and not notice-order: fixing the splices first risks repairing sentences that the structural reorder might relocate or cut; the hierarchy guarantees you only polish survivors. **M2.** It needs all three, top-down. First *clarity editing* ([Ch 3](../../part-01-writing-is-thinking/chapter-03-clarity/index.md)): "it is clear that there are several factors which… make it difficult" is expletive-laden bloat hiding the real claim. Then *sentence repair* ([Ch 6](../chapter-06-sentences/index.md)): the dangling "Having analyzed the results" needs an actor, and "due to the fact that" is wordy. A clean version: *"Several factors interact in complex ways, so we could not isolate a single root cause."* Order: clarity/structure of the claim first, then the mechanics, then words. **M3.** Separate tone from content: the rudeness is about delivery and isn't yours to absorb, but the structural observation is real and worth acting on. Respond with a plain thank-you (you need this person next time) and act on the diagnosis. Don't mirror the harshness; don't let the bad delivery make you reject the true note. Receiving feedback well means hearing the problem even through poor packaging. **M4.** It's a *content/structure* problem — *revision*, not editing. The issue isn't the words; it's that interpretation is in the wrong *section* (it belongs in Discussion, not Results). That's a structural/content fix at level 1–2: move the interpretive sentences to Discussion (or cut the overclaim). Belongs in the global pass, early — long before any word polishing. **M5.** You catch it at **level 1 (content)** via "so what?" — the methodology detail doesn't carry what *this* reader needs. The move depends: for an executive audience, *cut it down* to a one-line method mention or *move it* to an appendix; don't delete the fact that there *was* a method, but don't make the executive wade through it. Content/structure decision, made before any sentence polishing.

Part E — Extension ⭐⭐⭐⭐ (optional; Deep Dive track)

E1. The revision case study. Find a piece of your own writing from at least six months ago. Revise it now, fully, top-down. Then write a 300-word reflection: what does the gap of six months let you see that even a 24-hour gap wouldn't? What does this tell you about the relationship between time and editing ability?

E2. Build your personal revision checklist. Over your next three documents, log every error or weakness you (or a reviewer) catch. After three documents, you'll see a pattern — your "personal revision signature." Turn it into a custom checklist organized by the editing hierarchy, weighted toward your actual weak spots. This is the artifact a working professional actually keeps.

E3. Reverse-outline a published document. Take a report, paper, or long article and produce a reverse outline: one line per paragraph stating what that paragraph does (not what it says). Then critique the structure from the outline alone — where does the logic jump, repeat, or bury the point? Reverse outlining is a structural-revision power tool; this exercise teaches it on someone else's document before you use it on your own.

Selected solutions and rubrics are in appendices/answers-to-selected.md. For the open-ended tasks (C1–C4, D4, E1–E3), use the rubrics provided above to self-assess — and remember the chapter's own lesson: your before/after pair plus a change-log is more persuasive evidence of skill than any single polished draft.