Exercises — Chapter 1: Why Writing Matters More Than You Think

Writing is learned by writing. None of these exercises can be completed by reading alone — every one asks you to produce or revise actual text. That is the point. Keep your responses; several later chapters revisit material you generate here.

A difficulty key appears on each part: ⭐ approachable, ⭐⭐ requires real effort, ⭐⭐⭐ stretches you, ⭐⭐⭐⭐ for the motivated/Deep Dive reader.

How to use these. Do Parts A and B for sure — they're short and they install the chapter's core reflex. Pick at least two tasks from Part C. Part M (interleaved) will feel thin now because there are no prior chapters to mix in; it points forward instead, planting habits you'll reuse. Where a task is open-ended, a self-assessment rubric follows it in place of an answer key.

Part A — Analyze This ⭐

Read the writing; name what works and what's broken. Don't rewrite yet — diagnosis comes before treatment.

A1. Here are two ways the same person recorded the same idea. Which one represents more finished thinking, and what specifically does the better version force its author to decide that the weaker one doesn't? - (i) "Tests flaky. CI red sometimes. Maybe timing? Look into it." - (ii) "Three integration tests fail intermittently — always the ones that hit the payments service — which suggests a race condition in how we wait for that service, not a problem with the tests themselves."

A2. A teammate says, "I completely understand the caching layer, I just can't seem to write the design doc for it." Based on this chapter, what is the most likely diagnosis, and what's the cheapest way to test whether it's correct?

A3. Read this sentence from a draft report: "The utilization of the new framework resulted in the facilitation of a reduction in build times." Without rewriting it, identify (a) how many actual facts it contains and (b) at least two specific things making it hard to read.

A4. Below is the opening line of a real-looking status update. State, in one sentence, what action (if any) the reader is supposed to take after reading it — or explain why you can't tell.

"Following up on our discussion, there are a few things that are still outstanding and that we wanted to make sure were on everyone's radar going forward."

A5. This chapter claims "the best writing is invisible." Find a paragraph — in this book, a textbook, or documentation you use — where you didn't notice the writing and just understood the content. Then find one where the writing got in your way. For each, name one concrete feature that produced the effect.

A6. Look at the two churn-memo versions in §1.2 (the bullets and the paragraph). The paragraph contains a claim the bullets do not: "the problem is not price but unrealized value." Is that claim in the bullets in any form? What does its absence tell you about what the bullet list was actually doing for its author?

A7. A colleague defends a wall-of-text README by saying, "All the information is in there." Using the distinction from §1.3, explain why "the information is in there" does not mean the README works.

A8. ⭐⭐ Find a piece of your own writing from the past month (an email, a lab report, a comment, a message). Read it as a stranger would. Mark one sentence where you can now tell you hadn't fully decided what you meant. Quote it and explain what you'd have had to figure out to write it clearly.

Part B — Revise This ⭐⭐

Rewrite each passage. The goal in this chapter is the first move only — clarity through plain words and committed claims. (Structure and audience get their own chapters; don't reorganize yet, just clean up.) For each, note your word count before and after.

B1. "It was determined by the team that the performance of the system was not in alignment with the expectations that had been established at the outset of the project."

Self-check: Your version should be roughly half the length, name who did the determining, and use a real verb instead of "was determined / was in alignment."

B2. "There are several factors that contribute to the difficulty associated with the maintenance of legacy code."

Self-check: "There are several factors that contribute to" is almost always deletable. What's the actual subject of this sentence, and what's it actually doing?

B3. "The implementation of the caching mechanism was undertaken in order to achieve an improvement in the response times that are experienced by end users."

Self-check: Find the buried verb (something was implemented to improve something). Put the actor in front of it.

B4. A note-to-self that's too vague to act on later:

"Experiment didn't really work. Numbers were off. Probably the setup. Redo?" Rewrite it as 2–3 full sentences that commit to what specifically went wrong and what specifically you'd change. Notice what you have to decide to do this — that deciding is the exercise.

B5. "In today's rapidly evolving technological landscape, it is becoming increasingly important to note that documentation plays a key role in the overall success of any given software project."

Self-check: Cross out every word that adds no information. (Hint: the first eleven words can go entirely.) What's left?

B6. ⭐⭐⭐ Take this five-line comment and rewrite it so it explains why, not what. Invent a plausible reason if you need to — the skill is producing intent, not echoing the code.

# increment the counter
# check if counter is greater than 3
# if so, reset the counter and log a warning

B7. ⭐⭐⭐ Below is one of your own three-sentence explanations from the §1.1 Productive Struggle exercise (if you didn't do it, do it now). Revise it once. Did revising change the sentence, the idea, or both? Write one line describing what changed — this is the chapter's whole thesis, observed in your own hands.

Part C — Write This ⭐⭐–⭐⭐⭐

Produce a real document. Each task names a scenario, an audience, and a length. Write the actual thing, not a description of it.

C1. ⭐⭐ The one-sentence conclusion. Think of a project, experiment, or piece of work you've done. Write the single sentence that states your main conclusion plainly — the sentence a busy reader most needs. If you can't write it in one clean sentence, that's the finding: write a paragraph instead working out what your conclusion actually is, then compress it back to one sentence.

C2. ⭐⭐ Explain it to a smart outsider. Pick a concept from your field and write one short paragraph (4–6 sentences) explaining it to an intelligent person who doesn't share your background — no diagrams, a strict budget of one technical term. Then write one sentence naming where you had to slow down or look something up. That spot is a gap the writing found.

C3. ⭐⭐⭐ The same fact, two readers. Take one true fact about your work (a result, a status, a decision). Write it twice: once for a fellow expert who knows the domain, once for a manager who funds the work but doesn't know the details. Make them genuinely different documents, not the same text with words swapped. (You're previewing Chapter 2; do your best with instinct for now.) In one line, name the biggest difference between them and why the reader caused it.

C4. ⭐⭐⭐ The portfolio charter. Complete the Project Checkpoint from §1.7's checkpoint if you haven't: choose your portfolio subject and write the half-page charter (subject in one sentence; why you chose it; two or three audiences; one thing you'd like to be able to do with it that you can't yet). This is a graded portfolio artifact, not a throwaway — write it as if a reviewer will read it, because in later chapters one will (you).

Self-assessment rubric (score each 0–2): Subject is specific enough that you can picture seven different documents about it (0 = a vague field, 2 = a concrete system/project/question). Reason is honest and personal, not generic. Audiences are distinct — at least one expert and one non-expert. The "can't yet" line names a real gap, not a humblebrag. 7–8 = strong start; 5–6 = sharpen the subject; below 5 = your subject is probably too narrow or too vague — pick again.

C5. ⭐⭐⭐ The case for caring. Write a 150-word note to a skeptical younger version of yourself — someone who thinks writing is a soft skill that won't matter for a "real" technical career — making the case that it does. Use at least one idea from §1.3 (stakes) or §1.4 (career). Persuasion is writing too; notice how much you have to decide you believe to write it convincingly.

Part D — Synthesis & Critical Thinking ⭐⭐⭐

D1. This chapter argues that "I understand it, I just can't explain it" is usually false. Construct the strongest honest counterexample you can — a case where someone genuinely understands something but truly cannot put it in words. (Think: physical skills, intuition, tacit knowledge.) Then argue where the boundary lies: for what kinds of knowledge does "writing is thinking" hold strongly, and where does it weaken? Two paragraphs.

D2. The chapter claims clear writing makes you more precise, not less — that the seven-word fix is more precise than the twenty-three-word original. Defend or attack this. Can clarity ever cost precision? Give a concrete example of a sentence where cutting words would genuinely lose meaning, and explain how to tell that case apart from mere bloat.

D3. Translate for three audiences. Take this single technical sentence and rewrite the idea for three readers: a domain expert, a non-technical manager, and a curious 14-year-old.

"Our model's AUC improved from 0.74 to 0.81 after we added the user-tenure feature, reducing false negatives on the churn-prediction task." For each version, the test is: would that reader both understand it and care? Afterward, write one sentence on what stayed constant across all three (it's the actual finding) versus what changed (everything about how it's said).

D4. ⭐⭐⭐⭐ Re-read the four "Common Misconceptions" in §1.8. Pick the one you most resisted or most believed yourself. Steelman it (make the strongest case for the misconception), then dismantle it. The discipline of arguing the side you reject is itself a thinking-on-paper exercise — notice whether writing the steelman changed how strongly you hold the view.

Part M — Mixed Practice (Interleaved) ⭐⭐

Interleaving — mixing problem types so you must choose the right approach — is one of the most effective study techniques, and most chapters' Part M draws on earlier chapters. This is Chapter 1, so there are none yet. These tasks instead install reflexes you'll interleave from here on.

M1. For each of the following, decide in one word whether the writer's main problem is most likely thinking (they haven't figured it out) or expression (they know it but said it badly) — then give the one-line reason. This is a judgment you'll make constantly. - (a) A paragraph that's grammatically perfect but you finish it unsure what the author actually concluded. - (b) A sentence whose meaning is obvious once you fight through three nominalizations and a double hedge. - (c) A design doc that contradicts itself between page 2 and page 5. - (d) An email that says everything correctly but buries the one request at the very bottom.

M2. Keep a "stall log" for the next three days. Every time your own writing stalls — the sentence won't come, the transition won't land — jot one line: what you were trying to say and whether, on reflection, the block was a word problem or a not-yet-decided problem. Bring the log to Chapter 2. (Most people are startled how often it's the second.)

M3. Start a one-line habit you'll carry through the book: before you send/submit anything this week, write its conclusion or its ask in a single sentence at the very top, then check whether the rest of the document actually delivers it. Note one case where doing this exposed that your draft didn't.

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

E1. This chapter asserts the "explanation effect" / "self-explanation effect" as a real, attributed phenomenon (Tier 2). Without inventing citations, write a 200-word explanation of why, mechanistically, explaining something to a rubber duck (or a page) would help you understand it better — grounded in the §1.2 distinction between recognizing and producing. Then write one honest sentence on the limits of what you can claim here without the primary research in front of you. (You're practicing citation honesty, the subject of Chapter 11, before you're taught it.)

E2. Find a genuinely well-written technical document in your field — a paper, a README, a spec, an essay — that you admire. Write a one-page analysis of what specifically makes it invisible: not "it's clear" but the concrete moves (where it puts the point, how it handles jargon, sentence length, what it omits). You're learning to read as a writer, the practice Chapter 39 argues sustains a writing life. Keep this analysis; you'll add to it as the book gives you more vocabulary for what you're seeing.

Solutions and rubrics. Selected solutions live in appendices/answers-to-selected.md. For the open-ended tasks (most of Parts C and D), use the self-assessment rubrics provided inline. The honest test for every writing exercise in this book is not "does it match an answer key" but "does it work for the reader you wrote it for" — read your output aloud, hand it to someone if you can, and revise once.