Exercises — Chapter 5: The Writing Process

These exercises are about doing the process, not just describing it. Several of them ask you to write badly on purpose, then revise — which is the whole point. Keep your fast first drafts; you'll need them for the revision tasks. Where a task is open-ended, a self-assessment rubric is given instead of an answer.

Part A — Analyze This ⭐

Read each scenario and identify what's happening through the lens of the five-stage process.

A1. Priya opens a blank document to write a project proposal. She types a sentence, frowns, deletes it. Types another, deletes it. Twenty minutes later the page is empty and she feels stupid. Name the failure precisely. Which two stages is she collapsing into one, and which one is sabotaging the other?

A2. Marcus says, "I revised my report — I went through it twice, fixed all the typos, and smoothed out a dozen clunky sentences." What stages did Marcus actually perform? What did he skip? What's the tell?

A3. A writer spends four hours making the opening paragraph of a report perfect. In hour five, drafting the body, she realizes the entire report should be organized around cost, not chronology — which means the perfect opening is now wrong and must be cut. Which process mistake caused the wasted four hours? State the principle that would have prevented it.

A4. Below is the purpose statement a writer wrote before drafting. Is it a real purpose statement or just a topic? Fix it if it's broken.

"This document is about our team's Q3 testing results."

A5. Two writers describe their habits. Writer 1: "I outline for an hour, then draft in one focused pass, then revise the next day." Writer 2: "I just start typing and fix everything as I go until it's perfect, usually the night before it's due." Predict which writer hits writer's block more often and which produces better final documents — and explain why using §5.2 and §5.5.

A6. A teammate sends you a draft and says, "Thoughts?" the night before it's due. Identify two separate process mistakes in that one sentence-and-timing.

A7. Someone claims: "I write better under pressure — my best work is always the night before." Using §5.6, explain what's probably actually true about their situation.

Answers to Part A **A1.** She's collapsing **drafting** (generating) and **editing** (judging) into one act. The editing impulse — judging each sentence as she writes it — is sabotaging the drafting, which needs momentum and permission to be bad. This is the classic blank-page death spiral (§5.2). Fix: bench the editor, set a timer, write the worst version. **A2.** Marcus **edited** (smoothing clunky sentences) and **proofread** (typos). He **skipped revision** — he never asked whether the content and structure were right, and made no big moves (no cutting, reordering, or adding sections). The tell: he changed how sentences *read* but never touched what the document *says* or its order. Calling editing "revision" is the chapter's headline mistake (§5.4, §5.6). **A3.** She **polished before revising** — editing (perfecting the opening) before the structure was settled. The principle: *do structural experimentation in the outline/revise stages, where it's cheap, and edit late so you don't fall in love with words you'll have to cut* (§5.3, §5.4). An outline would have surfaced the cost-vs-chronology decision before any prose existed. **A4.** It's a **topic**, not a purpose — it says what the document is *about* but nothing about what it should *do* or who reads it. A fixed version: "After reading this, the engineering lead will understand that our Q3 pass rate dropped because of the flaky test suite, and approve a sprint to stabilize it." Now it names a reader, a goal, and an argument (§5.3). **A5.** Writer 2 hits writer's block far more (drafting and editing at once = the death spiral) and produces worse final documents (drafts once under deadline, so never truly revises — only edits the first draft). Writer 1 rarely blocks (stages separated) and produces better work (a real revision pass plus a 24-hour gap). See §5.2 and §5.5. **A6.** (1) Feedback is requested **too late** — the night before leaves no time to act on structural advice (§5.5). (2) The request is **undirected** ("Thoughts?") — it invites comma-level nitpicks instead of the structural feedback the writer presumably needs; a good request names the stage ("Is the structure clear?"). **A7.** They almost certainly don't write *better* under pressure — they only write *at all* under pressure, so they have nothing to compare it to. Under deadline they draft once and submit, which means they skip revision entirely. The night-before draft isn't their best work; it's their *only* work, mislabeled as best (§5.6).

Part B — Revise This ⭐⭐

Each task gives you weak material. Your job is to revise (the big moves), not just edit. Make structural decisions.

B1. Below is a complete, fast, ugly first draft of a Slack message asking a colleague to review a pull request before end of day. Revise it into a clean message. Decide what the reader actually needs and cut the rest.

"hey so i finished the thing finally, it was annoying because the test fixtures were all broken and i had to rebuild like half of them which took forever, anyway the PR is up, its the auth refactor one, i think its fine but theres one part in the token validation i wasnt sure about, could you look before eod? if you can't thats ok i'll find someone else. its #4412."

B2. Here is a paragraph from a report. It contains the right information but in the wrong order for a busy manager. Reorder it so the conclusion comes first (revision, not editing).

"We surveyed three vendors. Vendor A quoted $40,000 with a 6-week timeline. Vendor B quoted $55,000 with a 4-week timeline but better support. Vendor C quoted $38,000 but had poor reviews and a 10-week timeline. After weighing cost, timeline, and risk, we recommend Vendor B despite the higher price, because the faster timeline lets us hit the product launch and the support reduces our maintenance burden."

B3. A writer wrote this outline for a "why we should adopt automated testing" proposal. Revise the outline's order so it leads with what a skeptical engineering manager cares about most, and explain your reordering in one sentence.

  1. History of automated testing
  2. Types of tests (unit, integration, end-to-end)
  3. Tools available (Jest, Pytest, etc.)
  4. The cost: 3 weeks of setup
  5. The problem: we shipped 4 critical bugs last quarter
  6. Expected benefit: ~60% fewer escaped bugs

B4. Below is a too-long, unstructured draft of an email. It has no clear purpose. First, write the purpose statement it's missing ("After reading this, the reader will ___"). Then revise the email to serve that purpose, cutting anything that doesn't.

"Hi everyone, I wanted to reach out about the situation with the staging server. As many of you know it's been having issues lately and yesterday it went down again during the demo which was embarrassing. I've been looking into it and I think it's a memory issue but I'm not 100% sure, it could also be the disk. We've had this server for like 4 years so it's pretty old. Anyway I think we should probably consider doing something about it. Let me know your thoughts when you get a chance, no rush."

B5. This procedure step buries a critical warning. Revise so the warning can't be missed.

"Run the migration script with the --apply flag once you've confirmed the connection string points to the right database, because if it points to production and you haven't taken a backup you could lose data, and the script doesn't ask for confirmation."

Self-assessment rubric for Part B You revised well if, in each case, you made **structural** changes, not just sentence polish: - **B1:** Did you *cut* the backstory about broken fixtures (the reader doesn't need it), lead with the request, and surface the specific ask (review the token-validation part of PR #4412 by EOD)? A strong revision is ~2 short sentences. If you only fixed the lowercase and typos, you edited — you didn't revise. - **B2:** The recommendation (Vendor B, and why) should now be the **first** sentence; the per-vendor numbers follow as support. If the conclusion still comes last, you didn't reorder. - **B3:** Strong order: **5 (the problem) → 6 (the benefit) → 4 (the cost) → 3 (tools) → 2 (types) → 1 (history, or cut it)**. Lead with the pain the manager already feels, then the payoff, then the honest cost. One-sentence rationale: *a skeptical manager cares first about the problem and the ROI, not the history.* - **B4:** A reasonable purpose: "After reading this, the team will agree to allocate budget to replace the staging server." The revision should *make a clear recommendation* (replace it), state the impact (failed demo, recurring downtime, 4-year-old hardware), and ask for a specific decision — not end with a vague "thoughts? no rush." - **B5:** The **warning goes first**, visually separated: e.g., a `⚠️ WARNING:` line stating "This script applies changes immediately with no confirmation and can destroy data. Back up first and verify the connection string is NOT production," *then* the step. If the warning is still embedded mid-sentence, revise again.

Part C — Write This ⭐⭐–⭐⭐⭐

These require producing text. For the "draft fast" tasks, time yourself — the constraint is the lesson.

C1. (Draft fast.) Pick any topic you know well. Set a timer for eight minutes and write a fast, ugly, complete first draft of a one-paragraph explanation of it for a smart non-expert. Rules: no backspacing, no stopping, placeholders allowed. When the timer ends, stop. Do not clean it up. The deliverable is the messy draft itself — proof you can generate without judging.

C2. (Then revise hard.) Take the messy draft from C1. In a separate pass (ideally after a short break), revise it: cut what doesn't serve the explanation, reorder for clarity, fill any gap. Then edit the sentences. Submit both versions side by side and write one sentence naming the biggest structural change you made (not a sentence fix — a structural one).

C3. Write a complete planning package for a real document you actually need to write soon: (a) a one-sentence purpose statement, (b) a three-line audience note, (c) a working outline. This is the planning stage, done for real.

C4. Write a purpose statement for each of these documents, in the form "After reading this, [reader] will ___": - (a) A README for an open-source command-line tool. - (b) An email to your manager requesting three days off next month. - (c) An incident report after a service outage.

C5. Take the migration-warning sentence from B5 and write the full short procedure it belongs to: a numbered list of 3–5 steps with the warning properly placed as a callout. (This previews the instructions chapter, Chapter 22 — for now, just apply the process: plan the steps, draft, then revise so the warning is unmissable.)

Rubrics for Part C - **C1:** Pass = a *complete* paragraph produced in the time limit with no editing. If you stopped to fix things or didn't finish, you edited while drafting — run it again and resist. - **C2:** Strong = at least one genuine structural change (cut/reorder/add), not only sentence polish. The one-sentence note must describe a *structural* change; if it describes a comma fix, you revised at the wrong level. - **C3:** Purpose names a specific reader + a goal + (ideally) an argument. Audience note covers know/want/how-they-read. Outline shows the *shape* in reader-order, not discovery-order. - **C4:** Each statement names a concrete reader and a concrete outcome. "(a) After reading this, a developer who just found the tool will be able to install it and run their first command in under five minutes." "(b) …my manager will approve June 12–14 off, confident my work is covered." "(c) …the team will understand what failed, why, and what we changed so it can't recur." - **C5:** Warning appears *before* the dangerous step as a visually distinct callout; steps are numbered, parallel, and each is a single action.

Part D — Synthesis & Critical Thinking ⭐⭐⭐

D1. The chapter claims revision matters more than editing, and Chapter 3 was entirely about sentence-level clarity (which is editing). Are these in tension? Reconcile them in a short paragraph: how can sentence-level clarity be both "the core skill" (Ch 3) and "less important than revision" (Ch 5)?

D2. (Find the flaw.) A productivity blogger writes: "The secret to fast writing is to edit as you go — fix each sentence before moving on, so you never have to come back. One pass, done." Diagnose the flaw using this chapter. Under what narrow conditions might the blogger's advice not be terrible?

D3. Explain, in your own words and in no more than four sentences, why separating generating from judging unlocks writing — the cognitive mechanism, not just the rule. (This is the elaborative-interrogation move: explain the why.)

D4. The chapter's threshold concept is "revision is the work, not cleanup." Describe how a writer thinks about a first draft before internalizing this idea versus after. What concretely changes in their behavior?

D5. (Translate for audience.) Take this one sentence and rewrite the process advice in it for three readers: a stressed first-year student, a busy senior engineer, and a manager who writes one email a week.

"Separate drafting from editing to avoid writer's block."

Discussion notes for Part D - **D1:** Not in tension — they operate at *different stages*. Clarity ([Ch 3](../chapter-03-clarity/index.md)) is the craft you apply in the **editing** stage; revision (Ch 5) decides *what to edit and in what order*. Both are essential, but logically revision comes first: clear sentences in the wrong order still fail. "More important" here means "higher-leverage and more-often-skipped," not "you can ignore clarity." - **D2:** The flaw: editing-as-you-go *is* the blank-page death spiral — it collapses generating and judging and causes writer's block; it also wastes effort polishing sentences that revision may delete, and it prevents you from seeing the whole structure (you can't revise what isn't fully drafted). Narrow exception: a very short, low-stakes, well-understood message (a one-line Slack reply) where there's effectively nothing to revise — the process scales down to a single pass. - **D3:** Look for: generating and judging are opposite mental modes that compete for working memory; running them together splits attention so neither works; the judging mode actively kills the generating mode's momentum (the death-spiral loop); separating them lets full attention go to one mode at a time, which is faster overall despite being "two passes." - **D4:** *Before:* the draft is the product; revision is a quick typo-tidy; the writer fears the blank page and polishes prematurely. *After:* the draft is disposable raw material that exists to be changed; revision is the main event; the writer drafts fearlessly (it's all going to be reworked) and budgets real time for revision. Behavior change: drafts faster, edits later, and protects revision time. - **D5:** Student: "Don't try to write it perfectly the first time — bash out a rough version fast, *then* fix it. Trying to do both at once is why you stare at a blank screen at midnight." Engineer: "Treat the doc like code: get a working (ugly) version committed first, refactor after. Editing while drafting is premature optimization." Manager: "Brain-dump the email in one pass, then reread once before sending. Don't polish word one before word two exists."

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

These mix Chapter 5 with earlier chapters, so you have to choose the right tool.

M1. You're handed a dense, jargon-filled paragraph written for experts and asked to make it work for a general audience and it's structured badly and it's due in two hours. In what order do you tackle audience (Ch 2), structure (Ch 4), clarity (Ch 3), and the process (Ch 5)? Justify the order.

M2. A colleague's document has clear, polished sentences but you finish it with no idea what they want you to do. Is this a clarity problem (Ch 3), a structure problem (Ch 4), or a purpose/planning problem (Ch 5)? Defend your diagnosis.

M3. You wrote a report by starting at the keyboard with no plan, and three pages in it's a mess. Diagnose which Part I principle you violated, and describe the recovery move — do you push forward, or stop and do something first?

M4. Rewrite this sentence (Ch 3 clarity), but also say which writing-process stage you'd be in when you make this change (Ch 5):

"The utilization of the new framework resulted in a reduction in the occurrence of errors."

M5. A teammate asks you to review their draft. It's well-organized and the sentences are clean, but it's aimed at the wrong reader — it's written for executives when the audience is end users. Which stage should they return to, and is fixing this "revising" or "editing"?

Answers to Part M - **M1:** Plan → revise structure → edit for clarity → (audience runs through all of it). Concretely: (1) **planning/audience** first — decide who the reader is and the purpose, because that governs every later choice; (2) **structure** next (revision-level) — reorder for the new reader; (3) **clarity** last (editing-level) — cut jargon and bloat sentence by sentence. Doing clarity edits first would waste effort on sentences you'll restructure or cut. The process (Ch 5) sequences the other three. - **M2:** A **purpose/planning problem (Ch 5)**. The sentences are clear (so not [Ch 3](../chapter-03-clarity/index.md)) and presumably organized; what's missing is a clear *purpose* — the writer never decided what the reader should *do*, which is a planning-stage failure. The fix is a purpose statement, then a revision to surface the ask. - **M3:** You violated **plan-before-you-draft** (§5.3) — you started at the keyboard. Recovery: **stop drafting.** Don't push forward polishing a mess. Step back, write the purpose statement and a quick outline, then either restructure what you have (revision) or re-draft fast against the new outline. Pushing forward just produces more mess to untangle later. - **M4:** Rewrite: "The new framework reduced errors." (Killed two nominalizations — "utilization," "reduction in the occurrence" — and the dead verb "resulted in.") You'd make this change in the **editing** stage (stage 4), after the structure is settled. - **M5:** They should return to **planning/audience** (the wrong-reader problem starts there), and the fix is **revising** — adapting content and structure for end users is a big-moves change, not sentence polish. Clean sentences aimed at the wrong audience still fail; that's a revision-level issue.

Part E — Extension ⭐⭐⭐⭐ (optional, Deep Dive)

E1. Keep a process log for your next three real documents. For each, record: how long you spent in each stage, where writer's block hit (if it did), and whether you actually revised (made a big move) or only edited. After three documents, write a one-paragraph analysis of your process — where your time actually goes versus where it should go (§5.5's thirds). This is metacognition applied to your own habits.

E2. Find a piece of your old writing — something from a few months ago. Revise it now, with fresh eyes, applying this chapter. Document every change as either "revision" (structural) or "editing" (surface). Count them. The ratio tells you something about what the original draft most needed — and what you're now able to see that you couldn't then.

E3. Interview someone whose writing you respect (a professor, a senior colleague, a writer you know). Ask them three questions: How many drafts do you write? What does your revision actually look like? When do you get feedback? Compare their real process to this chapter's model and note where it matches and where it differs — and what their differences might teach you.

Selected solutions and rubrics for this book live in appendices/answers-to-selected.md. For the open-ended writing tasks above, use the self-assessment rubrics provided inline — the goal is to judge your own work against the criteria, which is itself a revision skill.