Does This Sentence Help the Passage Do Its Job?

Chapter 5

Imagine you're editing a friend's draft for a school newspaper or a club newsletter. A sentence can be true, well-written, and genuinely interesting, and still not belong in this specific paragraph, because it's not doing the job that paragraph needs done right now. A paragraph can contain three perfectly good facts and still feel disorganized, because they're in the wrong order. And an argument can offer real evidence that still doesn't actually answer the objection someone just raised, even though the evidence itself is accurate.

Every one of those is a judgment you already know how to make when you're reading something a friend wrote and trying to help them improve it: does this sentence earn its place here, does this order make sense, does this actually answer the point being raised. You're not checking grammar in any of these cases. You're checking whether the writing is doing its job.

ACT tests this exact editorial judgment, at the level of the sentence, the paragraph, and the whole passage, including a specific family of questions about how a writer builds and defends an argument: what counts as real evidence, how an objection like that gets acknowledged and answered (what ACT calls a counterclaim), and what makes a conclusion actually earned rather than just asserted. ACT groups all of this under Production of Writing: relevance, purpose, organization, and argument development.

Keep only details that serve the paragraph's job.

Production of Writing questions come in a small number of recognizable types: whether a detail belongs, what a sentence is doing there, whether a stated goal has been met, where a sentence should sit, whether a new paragraph is needed, what should open or close the passage, and how one sentence or paragraph should connect to the next. Naming which of these jobs a question is actually asking before doing anything else determines what kind of evidence to look for and how far into the passage to read.

Treating every rhetoric question the same way, as a vague "does this sound right" judgment, is what makes this category feel scattered. It is not scattered. Each question type has its own decision procedure, and this chapter works through them one at a time.

Read only as far as the decision requires. Different question types need different amounts of context, and reading more than a question requires wastes time without improving accuracy, while reading less than it requires produces a confident, wrong answer. A local sentence-purpose question can often be answered from the sentence and its immediate neighbors. An add/delete or relevance question needs the whole paragraph, since the paragraph's focus is the standard the detail is judged against. A placement question needs the whole paragraph too, since the correct spot has to fit both what comes before it and what comes after. A paragraph-transition question needs the end of the paragraph before the break and enough of the new paragraph to name its job. A whole-essay goal question needs the whole passage; a title or a single supporting paragraph is never enough on its own.

Matching the reading distance to the task is a skill in itself, separate from knowing the underlying rhetorical rule, and it is worth deciding on purpose rather than defaulting to the same habit for every question.

Find the paragraph's focus before judging a detail. Before deciding whether a sentence belongs, state what the paragraph is actually about and what job it is doing in the essay, not just its general subject. A paragraph's focus is usually narrower than the essay's overall topic, and a detail has to serve that narrower focus, not merely relate to the topic in a general way.

A paragraph about a scientist's early education, sitting inside an essay about her later career breakthrough, is not simply "about her life." It is about the training that shaped what she was later able to discover. A true, interesting sentence about her childhood hobbies might relate to her life broadly and still not serve this paragraph's narrower, functional focus.

Keep a detail when it helps the paragraph do its work. A detail earns its place by doing something for the paragraph: illustrating a claim, supplying evidence, giving necessary background, clarifying a term, qualifying an overstatement, or providing a contrast the paragraph needs. If a sentence cannot be shown to do one of these jobs for the paragraph's actual focus, keeping it because it is true or interesting is not a good enough reason.

A related detail can still be the wrong detail. Sharing a broad topic with the rest of the paragraph is not the same as being relevant to what the paragraph is doing. This is one of the most persistent traps in this category, because a sentence that is true, well written, and clearly connected to the essay's general subject can still be the wrong sentence for this specific paragraph, because it shifts the paragraph's focus rather than developing it.

A true detail can still be wrong if it serves the wrong paragraph focus.

A paragraph describing the specific engineering challenges of building a bridge across a particular river includes a sentence about how bridges in general have shaped the history of trade and transportation.

The sentence about bridges in general is true and clearly related to the topic of bridges. But the paragraph's actual focus is the specific engineering problem at hand, and a sentence about the general historical importance of bridges shifts attention away from that focus rather than developing it. This sentence would likely serve a different paragraph, one about why bridges matter broadly, better than it serves this one.

Separate relevance from repetition. Relevance and redundancy are answering two different questions, and treating them as the same thing produces wrong answers on both. Redundancy asks whether a piece of information has already been stated; relevance asks whether the information serves the paragraph's focus and rhetorical job. A sentence can repeat information the paragraph has already given and still be relevant in a general sense, and a sentence can state something entirely new and still be irrelevant to what this paragraph needs. Deleting a sentence because it repeats earlier content is a redundancy decision, covered fully elsewhere in this book; deleting it because it does not serve the paragraph's job is a relevance decision, and mixing the two justifications together produces an answer that may be right for the wrong reason.

Decide add or delete before judging the reason. Add/delete questions typically separate the decision from the reason: choices support keeping or removing the material and pair that action with a justification. Decide the action first, based on whether the material actually serves the paragraph, and only then check which of the two reasons attached to that action accurately describes why.

Deciding the action and the reason at the same time, or picking a reason first because it sounds persuasive, is what causes students with otherwise strong grammar to lose points here. The two decisions are separate, and skipping the first one to jump straight to comparing reasons is a common process error in this question type.

Make the reason prove the decision. Once the action is decided, the paired reason has to accurately describe the sentence and the specific job it is or is not doing, not just point in the general direction of the correct action. A reason can correctly match a yes or a no answer and still be the wrong choice, because the explanation it gives is not actually true of the sentence or the paragraph.

Make the keep/delete decision first, then choose the reason that proves it.

Should the writer add a sentence explaining the technical specifications of the bridge's suspension cables to a paragraph that otherwise focuses on the emotional experience of the town's residents watching the bridge's construction?

The correct action here is almost certainly no, don't add it, since the paragraph's focus is emotional experience, not engineering detail. But the reason matters just as much as the action: a reason claiming the sentence is factually inaccurate would be wrong, since nothing suggests the specifications are incorrect; the accurate reason is that the technical detail does not serve this paragraph's focus on the residents' experience, however accurate the detail itself might be. A student who agrees with the correct no but selects the reason about inaccuracy has answered the question wrong despite getting the action right.

Choose the detail that gives the paragraph what it still needs. When a question asks which detail would best support a paragraph, the answer is not simply the most vivid, specific, or memorable option among the choices. It is the option that supplies exactly what the paragraph's argument or description is missing at that point, whether that is a concrete example, a piece of evidence, a clarifying definition, or a necessary qualification.

Choose the detail that most directly supports the exact claim.

A paragraph claims that the new policy reduced processing delays significantly, but offers no specific figures.

Among several true, well-written sentence options, the one that supplies an actual number or comparison, such as average processing time fell from eleven days to four, does more for this paragraph than a vivid but vague sentence describing how relieved staff felt about the change. The paragraph's claim needs support in the form of evidence, and a detail that is engaging but does not supply that specific kind of support is not the best choice, even if it is well written.

Use background when the reader needs it. Some sentences that seem tangential when checked only against the local sentence are actually necessary because a later part of the paragraph or essay depends on the reader already knowing that information. Deleting background material because it appears unconnected to the immediate context, without checking whether something later relies on it, is a common and avoidable error.

Keep background only when the later passage depends on it.

A paragraph mentions in passing that the research team had worked together for over a decade before describing, several sentences later, how their long-established trust let them challenge each other's conclusions openly during the project's most difficult phase.

Checked only against the sentence immediately around it, the detail about the team's decade of collaboration might look like an unnecessary aside. But the later sentence about their openness in challenging each other's conclusions depends on the reader already knowing they had worked together for years; without that background, the later point about trust would come out of nowhere. Deleting the earlier sentence because it seemed locally unnecessary would remove something the paragraph needs to make its later point land.

Keep an example when it actually supports the point. An example belongs in a paragraph only when it actually illustrates the specific claim the paragraph is making, not merely when it is a true and interesting fact placed near that claim. An example that is engaging on its own but does not clearly connect to the claim it is supposed to support is not doing its job, regardless of how well written it is.

Build the claim with evidence that actually supports it.

Argument questions in English are editing questions. The task is not merely to understand what a writer believes. It is to decide what the developing argument still needs, what a sentence is doing to that argument, and whether a proposed revision makes the reasoning stronger, weaker, clearer, or less defensible. That requires tracking the argument as a structure rather than treating each sentence as an isolated statement.

Map the argument before judging a sentence. An argumentative passage usually contains several different kinds of statements, and treating all of them as equal is the first source of error. Before deciding whether a sentence belongs, identify which role it plays.

Do not ask only, "Is this sentence relevant?" Ask the more precise question: "Relevant to what part of the argument, and for what job?" A sentence about cost may be relevant to the topic of a transportation proposal but still fail to support a claim about safety. A statistic may be impressive but useless if it measures a different outcome from the one the claim asserts.

Separate a claim from the evidence that would actually support it. The best supporting sentence is controlled by the exact claim, not by the passage's broad topic. Reduce the claim to its measurable or provable core. Then ask what kind of evidence would directly bear on that core.

Claim: the redesigned bus route reduced commute times. Strong support: before-and-after commute-time data for riders on the redesigned route. Weak support: a survey showing riders like the new bus shelters.

Claim: the school's later start time improved attendance. Strong support: attendance or tardiness data before and after the schedule change. Weak support: students reporting that they feel less tired.

Claim: the community garden increased neighborhood food access. Strong support: data showing more residents obtained fresh produce, especially where access had been limited. Weak support: the number of volunteers who helped paint the garden fence.

A related sentence is not automatically supporting evidence. The most dangerous distractors are often true details from the same topic that support a different claim.

Choose the strongest support, not merely acceptable support. When two choices both help a claim, compare how directly each one connects to the exact assertion. Strong support usually reduces the number of extra assumptions between evidence and claim. It measures the outcome the claim names, uses a comparison the claim depends on, or addresses the specific population, period, or condition the claim covers.

Minimal-contrast example. A writer claims that a new tutoring program improved algebra performance. One choice says that 82 percent of participants said they enjoyed the program. Another says that participants' median algebra scores rose from 71 to 83 while scores in comparable nonparticipating classes remained nearly unchanged. Both are favorable to the program. Only the second directly supports the performance claim.

Match the scope of the claim to the scope of the evidence. Evidence can be relevant and still be too narrow to support the claim as written. If a study involves one school, one neighborhood, one season, or one small group, a sentence that turns that result into a universal claim overreaches. The editor's job may be to narrow the claim rather than replace the evidence.

Example. A pilot in two city blocks found that shaded bus stops were used more often during July. The evidence can support a claim about those locations during that period. It cannot, by itself, support "shaded bus stops always increase public-transit use in every climate." A revision such as "the pilot suggests that shade can increase use at exposed summer stops" preserves what the evidence earns without pretending the study proved more.

Identify the counterclaim before evaluating the response. A counterclaim is not simply any negative fact. It is a competing reason to reject, limit, or question the writer's position. First state the counterclaim in plain language. Then identify what the writer does with it.

A concession does not automatically weaken the argument. Strong arguments often become more precise after conceding a real limitation. Suppose a writer argues that a city should convert one downtown lane into a protected bicycle lane. Opponents note that removing a traffic lane may slow cars during the evening peak. The writer may concede that point and still argue that the change is justified because the corridor moves more people overall once bicycle and bus traffic are counted. The concession changes the shape of the argument, but it does not erase the central claim. An answer choice that treats every concession as surrender is too crude; ask what the writer gives up and what the writer continues to defend.

The best response addresses the counterclaim at its strongest point. A weak response changes the subject. A stronger response meets the objection on the dimension that makes it threatening. If the counterclaim concerns cost, a response about popularity does not answer it. If the counterclaim concerns whether an effect will last, a response about the size of the first-month effect does not answer it.

Close-choice example. A school argues for replacing disposable cafeteria trays with washable trays. Critics object that washing trays will increase water use. Choice A says students strongly support reducing cafeteria waste. Choice B says a life-cycle analysis found that the washable system used more water on campus but reduced total resource use after manufacturing and disposal were included. A is favorable but sidesteps the objection. B answers the objection on its own terms.

Distinguish evidence that supports the claim from evidence that only sounds impressive. Numbers create an illusion of rigor. A statistic is useful only if the quantity measured bears on the claim. A passage arguing that a tutoring program improves retention is not strengthened by the fact that the program has 37 tutors unless the argument depends on staffing. Likewise, a large percentage is not automatically stronger than a smaller one if it measures the wrong outcome.

Check the reasoning bridge, not only the facts. An argument can contain accurate facts and still make an unsupported leap. The editor should be able to state the bridge between evidence and claim: "If this evidence is true, why does it make the claim more likely?" If that sentence cannot be completed without adding an assumption the passage never establishes, the support is weak.

Example. A city library extended weekend hours, and weekend attendance rose 28 percent. That supports the claim that more people used the library on weekends after hours were extended. It does not by itself prove that the change improved literacy, reduced social isolation, or increased total yearly library use. Those conclusions require additional evidence.

Use qualification when the evidence is real but limited. Sometimes the correct revision is neither to delete the claim nor to find stronger evidence. It is to make the claim fit the evidence. Words such as may, can, in this pilot, among surveyed participants, and during the study period can convert an overstatement into a defensible conclusion. Qualification is not weakness when it accurately represents what the evidence can establish.

Build the argument in an order the reader can evaluate. Argument organization is controlled by dependency. A claim should generally appear before evidence whose purpose cannot be understood without it. A counterclaim should be recognizable before the response to it. A concession must come before the writer's pivot away from it. A concluding claim should not depend on evidence that has not yet appeared. A useful default sequence is central claim, supporting reason, evidence, limitation or counterclaim, response, refined conclusion, but this is not a formula every passage must follow. When a sentence is moved, ask what information it requires the reader to already know and what later sentence depends on it.

Five checks for any argument-development question. What exact claim is being supported, challenged, or revised? What does the proposed evidence actually establish, without interpretation? What reasoning step connects that evidence to the claim? Does the claim's scope and strength match the evidence? If a counterclaim appears, does the response address that counterclaim or merely move to a different favorable point?

Progressive practice: from recognition to discrimination.

Level 1: Identify the role each sentence plays in the argument. Paragraph: The city argues that replacing diesel buses with electric buses will reduce street-level air pollution. A sentence reports that diesel exhaust is a major source of nitrogen oxides along the route. Question: what is the sentence doing? Answer: supplying evidence relevant to the pollution claim, not merely background about buses.

Level 2: Separate relevant evidence from directly supportive evidence. Claim: the redesigned park increased use among older adults. Choice A: the redesign added 42 new trees. Choice B: counts of visitors age 65 and older rose from an average of 38 per day before the redesign to 71 per day afterward. Answer: B. A is relevant to the redesign; B directly measures the population and outcome named in the claim.

Level 3: Choose the stronger of two plausible support choices. Claim: a school's device-lending program reduced homework-access disparities. Choice A: ninety-one percent of borrowers said the devices were easy to use. Choice B: the gap in assignment-completion rates between students with and without a home computer fell from 18 percentage points to 5 after the lending program began. Answer: B. Both are favorable; only B measures the disparity named in the claim.

Level 4: Evaluate the counterclaim and the response. Claim: the city should preserve a downtown loading zone despite pressure to convert it to parking. Counterclaim: the loading zone removes six spaces from a block where parking is scarce. Response A: downtown stores contribute substantial sales-tax revenue. Response B: delivery trucks that lose the loading zone would have to stop in the travel lane, and a two-week pilot of that arrangement increased average bus delay on the block by 19 percent. Answer: B. A changes the subject to store importance; B answers the cost of keeping the loading zone by showing a transportation cost of removing it.

Identify what the sentence is doing and what the passage needs.

A sentence purpose question is asking for the sentence's function in the passage, not a restatement of what the sentence says. Confusing content with function, describing what a sentence talks about rather than what job it is performing for the paragraph, is a common way students miss this question type even when they understand the sentence's meaning perfectly well.

Name the sentence's immediate job in the paragraph.

A sentence in the middle of a paragraph reads: "This shift did not happen overnight; it took nearly a decade of gradual policy changes."

A student describing this sentence's content might say it is about how long the shift took. Describing its function instead means naming what the sentence is doing for the paragraph: it is qualifying a claim made in the previous sentence, correcting a possible misimpression that the change was sudden, and setting up the paragraph's move into describing that gradual process in more detail.

Distinguish what a sentence says from why the writer put it there. This mechanism is worth repeating because it is easy to nod along to in principle and still fail to apply under time pressure. When a question asks what a sentence "accomplishes" or "does," an answer choice that merely paraphrases the sentence's content, restating its meaning in different words, is not answering the question, even when that paraphrase is accurate. The correct answer names a rhetorical job: providing context, anticipating an objection, emphasizing a consequence, setting up a contrast, or a similar function.

Match the sentence to the purpose named in the question. When a question specifies a purpose the writer is trying to achieve and asks which sentence or wording accomplishes it, every answer choice may be grammatically correct and reasonably related to the paragraph, but only one actually performs the specific job the question names. A well-written, relevant sentence that does a different job than the one requested is still the wrong answer.

Choose the sentence that actually accomplishes the stated purpose.

A question asks which sentence would best help the writer "acknowledge a limitation of the study before presenting its main finding."

A sentence that simply restates the study's main finding, however well written, does not accomplish the requested job at all; it needs to be a sentence that names a limitation, such as a small sample size or a narrow testing window, positioned before the finding is presented. Among several plausible, grammatical options, only the one that actually names a limitation performs the specific function the question asked for.

Check the scope before deciding whether the goal is met. A whole-essay goal question states a specific target, and before judging whether the essay meets it, the scope of that target has to be defined precisely: is the goal about one example or an entire class of things, a description or an explanation, how something happens or why it happens, a historical account or a current process. An essay can satisfy part of a stated goal's scope while missing another part entirely, and judging the goal correctly requires checking every dimension the stated goal actually specifies, not just the general direction.

Judge a whole-passage goal with whole-passage evidence. A whole-essay goal question cannot be answered from a title, an introduction, or a single supporting paragraph alone. The evidence for whether the essay accomplishes its stated goal has to come from across the passage, because a goal can be partially addressed in one place and left unaddressed elsewhere, and only checking the whole passage reveals that gap.

Reject an answer that fulfills the goal at the wrong scope.

A question asks whether an essay "explains both the causes and the long-term consequences of the policy change."

If the essay's body paragraphs describe the causes of the policy change in detail but only mention, in a single closing sentence, that the policy "affected many communities" without describing any specific long-term consequence, the essay has satisfied one half of the stated goal's scope, causes, but not the other, consequences. A student who sees the essay talk about the policy at length and assumes the goal is met because the essay is clearly on topic has skipped the actual scope check the question requires. The correct answer here is no, and the correct reason names the specific missing half of the scope, not a general sense that the essay does not fully address the goal.

Put each sentence where its references and logic belong.

A sentence often carries a signal pointing backward to something already established: a pronoun, a demonstrative like this or such, a repeated noun, or a synonym for something named earlier. Finding what that signal refers to identifies the sentences it could plausibly follow, narrowing the field of possible positions before checking anything else.

Then check what must come after it. Backward reference alone is not enough to fix a sentence's position, because more than one location in a paragraph can sometimes satisfy the same backward link. The sentence's own content also has to set up whatever comes immediately after it, and checking that forward fit is what breaks a tie between two positions that both seem to satisfy the backward reference.

Place the sentence where both its backward reference and forward connection work.

A paragraph describes a company's early struggles, and a candidate sentence begins with "Despite these setbacks, the team refused to abandon the project."

Two earlier sentences in the paragraph could both plausibly count as "these setbacks": one describing a failed product launch, another describing a funding shortfall a few sentences later. Backward reference alone does not resolve which position is correct. The deciding factor is what comes next: if the sentence immediately following the candidate sentence describes the team's response specifically to the funding shortfall, the candidate sentence belongs right after the funding-shortfall sentence, not the earlier one, because only that position lets the pronoun's reference and the following sentence's content both make sense together.

Use reference words to find the sentence's anchor. Demonstratives and pronouns like this, these, such, they, the result, and the change are some of the most reliable placement signals in this category, because they require a specific, identifiable antecedent nearby. A reference word with two plausible antecedents nearby is exactly the kind of hard case this section is built to prepare for, and resolving it always comes back to checking which antecedent also makes the surrounding sentences work together, not just which one is closer.

Follow chronology when time controls the order. When a paragraph describes a sequence of events, the correct order usually follows the timeline the events actually happened in, even when the paragraph does not state explicit dates or use words like first and then. Chronological order can be implied entirely by what logically had to happen before something else, and recognizing that implied sequence is part of this skill, not just recognizing an explicitly dated timeline.

Put events in the order the passage establishes.

A paragraph describes a building's history: its original construction, a fire that damaged part of it, and a renovation that restored the damaged section.

Even without explicit dates, the renovation logically had to happen after the fire, and the fire logically had to happen after the original construction; a sentence about the renovation cannot precede the sentence describing the fire, because the renovation only makes sense as a response to damage that has already occurred.

Follow cause and effect when one idea depends on another. When a paragraph's sentences describe a cause and its result, the cause usually needs to appear before the effect, since a reader generally needs to understand what happened before making sense of what it led to. This is a strong default, not an unconditional law: a paragraph can open by stating a striking effect and then explain the cause behind it, a structure writers use deliberately for emphasis. What actually matters is information dependency, whether the reader has what they need to interpret each sentence when they reach it, not a fixed rule about which idea has to come first. A sentence describing a consequence placed before any explanation of what caused it disrupts a paragraph's logic only when the reader is left without the context needed to make sense of it.

Move from a general point to the detail that develops it. A common and reliable paragraph pattern states a general claim first and then develops it with more specific supporting material. A specific supporting detail placed before the general claim it supports can leave a reader without the context needed to understand why that detail matters, even though the same information appears somewhere in the paragraph either way.

Keep an example beside the claim it supports. An example needs to sit next to the claim it illustrates, not scattered elsewhere in the paragraph where its connection to that specific claim becomes unclear. This is the placement version of a mechanism this chapter has already covered from the relevance side: the example's rhetorical job, supporting a specific claim, only works if it is positioned close enough to that claim for the connection to be visible.

Keep an example beside the claim it illustrates.

A paragraph states that the new curriculum improved student engagement, and elsewhere includes a sentence describing a specific classroom where attendance rose sharply after the curriculum was introduced.

If the classroom example sits several sentences away from the engagement claim, separated by unrelated material, its function as supporting evidence for that specific claim becomes muddled. Moving the example to sit immediately after the claim it illustrates makes the paragraph's supporting logic clear rather than requiring the reader to reconstruct the connection across a gap.

Keep a response beside the problem or question it answers. The same logic applies to a paragraph that poses a problem, a question, or an objection and then offers a response: the response usually belongs close to the problem, question, or objection it answers. The key is dependency: the reader should still be able to see exactly what the response addresses without having to reconstruct that connection across unrelated material.

Use contrast and concession to control order. When a paragraph's structure depends on a contrast or a concession, an idea that is granted before it is challenged, the sentence that grants the point has to come before the sentence that challenges or complicates it. Reversing that order changes the paragraph's actual argument, since a concession made after the challenge it is supposed to precede no longer functions as a concession at all.

Choose the order that builds the paragraph's idea. Every placement decision in this section reduces to the same underlying test: does this order let the paragraph's idea develop, with each sentence able to build on what the reader already knows from the sentences before it, rather than requiring the reader to hold an idea in suspension until a later sentence explains it. When several placements sound smooth read aloud, the reliable check is not which one sounds best, but which one preserves the paragraph's chain of evidence: what each sentence needs to have already been established, and what it sets up for the sentence that follows.

Start a new paragraph when the writing changes jobs. A new paragraph is needed when the writing's rhetorical job changes, not simply when a certain amount of text has accumulated or when the topic shifts slightly. Two sentences can discuss the same general topic and still belong in different paragraphs if one is doing a fundamentally different job than the other, such as moving from describing a problem to proposing a solution.

A new paragraph can begin even when the topic stays the same if the job changes.

A paragraph describes the symptoms of a medical condition in detail, and the next sentences begin discussing the treatments available for that same condition.

Both stretches of text are about the same medical condition, so a student checking only for topic continuity might not see a reason for a paragraph break. But describing symptoms and describing treatments are different rhetorical jobs, diagnosis versus response, and separating them into their own paragraphs lets each one develop its own point clearly rather than blending two different jobs into one undifferentiated block of text.

Keep one paragraph centered on one controlling purpose. A single paragraph should be organized around one controlling job, and every sentence inside it should serve that job. When a paragraph starts drifting into a second, different purpose, that drift is itself the signal that a paragraph break is needed, independent of how long the paragraph has become.

Open, connect, and close the passage without changing its direction.

An introduction's job is to orient the reader: establishing the subject, providing necessary context, framing a central question or problem, or signaling the direction the passage is about to take. An introduction is not required to follow a single fixed formula, but whatever it does has to genuinely prepare the reader for what the passage actually goes on to do.

Choose an opening that fits the passage's actual direction. An engaging, well-written opening sentence can still be the wrong choice if it does not fit the direction the passage actually develops. A dramatic hook that promises one kind of essay, when the passage that follows turns out to be about something else entirely, misleads the reader about what is coming, regardless of how well crafted that opening sentence is on its own.

Choose an opening that prepares the reader for what they are about to read.

An essay's body focuses entirely on the technical process of restoring a historic clock tower, described in a measured, procedural tone throughout.

An opening sentence that begins with a dramatic, suspenseful account of the clock tower's near-total collapse during a storm decades earlier might be engaging, but if the essay itself never returns to that dramatic history and instead stays focused on the calm, procedural details of the restoration work, that opening promises a different essay than the one the passage delivers. An opening that instead orients the reader to the restoration project itself fits the passage's actual direction, even if it is less immediately dramatic.

Let the conclusion close the work the passage has done. A conclusion's job is to provide closure appropriate to what the passage has actually been doing, not simply to end on a true and reasonably related final statement. A sentence can be accurate and connected to the passage's topic and still fail as a conclusion if it does not close the specific rhetorical work the passage has done.

Choose the kind of closure the passage needs. Closure can take several different forms depending on what the passage needs: returning to an idea raised at the beginning, resolving a question the passage posed, synthesizing points made across several paragraphs, implying a broader significance without stating it outright, or leaving the reader with a final emphasis. Not every passage needs the same kind of closure, and assuming every conclusion must restate the passage's central claim is exactly the kind of formula this section is built to move past.

Do not introduce a new major idea in the conclusion.

An essay describing a town's decades-long effort to restore a historic theater ends with a sentence introducing an entirely new fact about the theater's original architect that was never mentioned anywhere earlier in the passage.

Even if the fact is true and mildly interesting, introducing brand-new information in the passage's final sentence does not provide closure; it opens a new thread the essay has no room left to develop. A stronger conclusion would return to the restoration effort's significance, the theme the essay has actually been building, rather than ending on an unrelated new detail.

Choose the ending that closes the passage's actual line of thought.

An essay traces a scientist's years of failed experiments before her eventual breakthrough discovery, without ever directly stating a "moral" or explicit takeaway.

A conclusion that summarizes the sequence of experiments the essay already described would be accurate but redundant, restating information the reader has just finished reading rather than closing the essay's arc. A conclusion that instead implies, without stating outright, what the years of failure ultimately made the breakthrough mean to her, without introducing new factual claims, can close the passage's emotional and thematic arc more effectively than either a new fact or a plain summary would.

Name the relationship before choosing a transition. Before selecting transition wording between sentences or paragraphs, state the actual relationship between the ideas on both sides: is the second idea adding to the first, contrasting with it, resulting from it, or something else. This is the same discipline this book has already taught for clause-level connectors, applied here at the level of a full sentence or a full paragraph rather than a single clause.

Connect sentences by the ideas on both sides. A sentence-level transition has to fit the specific idea that comes before it and the specific idea that comes after it, not just sound smooth in isolation. Two transition words can belong to the same general category, such as two different words that both signal contrast, and still not be equally correct, because the exact relationship between these two particular ideas may match one of them more precisely than the other.

Name the relationship between the two sentences before choosing the transition.

A sentence describes a company's aggressive expansion into new markets; the next sentence describes how that expansion ultimately strained the company's existing operations beyond what it could support.

A transition word signaling straightforward addition would understate the relationship; a transition signaling consequence, showing that the strain followed as a result of the expansion, fits the relationship between these two specific ideas more precisely than a same-category alternative that signals mere contrast without capturing the cause built into the sequence.

Bridge paragraphs by the movement of the passage. A transition between two paragraphs has to bridge the movement of the passage as a whole, not just link the final sentence of one paragraph to the opening sentence of the next in isolation. Reading only the last line of the previous paragraph, without considering what the new paragraph is actually about to do, produces a transition that may connect smoothly to one sentence while failing to represent the actual shift the passage is making.

Use the new paragraph's job to choose the bridge. The new paragraph's own purpose often controls what kind of bridge is needed more than the previous paragraph's final sentence does. A paragraph that is about to shift from describing a problem to proposing a solution needs a bridge that signals that shift, even if the previous paragraph's last sentence does not obviously call for one.

Read the end of one paragraph and the start of the next before choosing the bridge.

A paragraph ends by describing the extent of a city's traffic congestion problem in detail; the next paragraph begins describing a proposed light-rail system.

A bridge sentence needs to do more than smoothly follow the previous paragraph's final detail about congestion; it needs to signal the shift the new paragraph is actually making, from problem to proposed solution. A transition that merely continues describing congestion, however smoothly it reads next to the prior paragraph's last sentence, fails to prepare the reader for the new paragraph's actual job, while a transition explicitly framing what follows as a response to the problem just described bridges both sides correctly.

Know when no transition is needed. Sometimes the clearest and most effective choice is no explicit transition at all, when the structure and content of the sentences already make the relationship clear without one. Assuming a transition word is statistically likely to be the right answer, rather than checking whether the sentences already cohere without it, is exactly the kind of probability-based shortcut this section rejects.

Leave the sentence alone when no transition is needed.

Two adjacent sentences describe, in sequence, the first and second steps of a straightforward process: mixing the ingredients, then baking the mixture at a specified temperature.

The sequence between these two steps is already clear from their content and their order; inserting an explicit transition like next or after that adds nothing the reader does not already understand from the structure alone. Leaving the sentences to stand on their own, or using only a simple period between them, is often the cleanest choice precisely because the relationship needs no additional signal.

When several placements sound smooth, follow the evidence chain. Across every kind of placement and organization decision in this chapter, from where a single sentence belongs to where a paragraph break should fall to what should open or close the passage, the reliable method is the same: trace the chain of evidence a position depends on and creates, rather than testing each option by how naturally it reads out loud. A position, an order, or a transition can sound smooth and still be wrong, because smoothness is a property of surface rhythm, while the actual test this whole chapter has been building toward is whether the passage's logic, reference, and development genuinely hold together at that point.

Judge whether a whole passage accomplishes a stated goal.

One current, recurring Production of Writing item type states a specific writing goal, sometimes matching the essay's actual purpose, sometimes not, and asks whether the essay as a whole accomplishes it. This is a different task from any single add/delete or purpose question: it requires holding the entire passage in view at once and checking it against one precise, stated standard, rather than judging one sentence's local fit.

Break the stated goal into its exact parts before judging the essay against it. A goal phrased as a single sentence often contains two or more separate conditions, and an essay can satisfy one while missing the other entirely. Treat every clause of the stated goal as its own checklist item, and only call the goal accomplished once every part of it, not just the most obvious part, is actually satisfied by the passage as a whole.

Goal-accomplishment item.

Essay summary: The essay opens by describing how the telegraph was built, including the specific wiring and signaling method its inventor developed. Its second half traces a direct line from that signaling method to the error-correction codes used in modern data transmission, naming the specific principle both systems share.

Suppose the writer's goal had been to describe a specific historical invention and explain its lasting influence on a modern technology. Would the essay accomplish this goal?

A) Yes, because it describes the invention in detail.

B) Yes, because it explains both the invention's original design and its specific influence on a named modern technology.

C) No, because it only explains the invention's original design without connecting it to anything modern.

D) No, because the essay focuses on modern technology rather than history.

A goal question can also correctly conclude that the essay fails, and the reason has to be specific. Just as often, the correct answer is "no," and the wrong answers on a "no" version of this question typically supply a reason that sounds plausible but does not match what the passage actually omits or gets wrong. Confirm the specific missing or mismatched element directly against the passage before accepting a "no" answer's stated reason, the same way you would confirm a "yes" answer's reason.

Practice set: goal accomplishment

A. Essay summary: The essay explains, in plain language, how vaccines train the immune system to recognize a pathogen without causing the disease itself. Its closing paragraph directly addresses the common belief that vaccines "give you a mild version of the illness," explaining specifically why that belief is inaccurate and what actually happens instead.

Suppose the writer's goal had been to explain a scientific process and address a common misconception about it. Would the essay accomplish this goal?

A) Yes, because it explains the process in clear, accessible language.

B) Yes, because it explains the process and directly identifies and corrects a specific misconception readers commonly hold.

C) No, because the essay does not explain the process in enough technical detail for an expert reader.

D) No, because the essay's tone is too informal for a scientific topic.

B. Essay summary: The essay traces a jazz musician's forty-year career, from an unpaid house-band position through international recognition. Its final two paragraphs explicitly argue that this musician's innovations in rhythm section arrangement influenced better-known contemporaries who received more credit at the time.

Suppose the writer's goal had been to profile a person's career and argue for their overlooked importance in their field. Would the essay accomplish this goal?

A) No, because the essay spends most of its length on biographical detail rather than argument.

B) Yes, because the essay traces the person's career and explicitly argues that their contributions have been undervalued relative to better-known contemporaries.

C) Yes, because the essay is entirely biographical.

D) No, because the essay never mentions the person's contemporaries.

C. Essay summary: The essay describes a town's annual lantern festival in vivid sensory detail, including its origin as a harvest ritual roughly a century ago. It does not address how or why the tradition has continued into the present, despite the town's economy and population changing substantially since the festival began.

Suppose the writer's goal had been to describe a community tradition and explain why it has persisted despite changing circumstances. Would the essay accomplish this goal?

A) Yes, because the essay describes the tradition in vivid detail.

B) No, because the essay describes the tradition and explains its origin, but never addresses why it has continued despite change.

C) Yes, because the essay explains why the tradition continues today.

D) No, because the essay is too short to cover both parts of the goal.

Apply relevance, purpose, order, and argument together.

A real ACT English passage mixes every mechanism in this chapter within the same set of questions, and the first move on any of them is still the one this chapter opened with: name the question type before doing anything else, because that decision determines how far to read and what kind of evidence actually answers it.

Apply the chapter skills without being told which rule is tested.

A paragraph describes a nonprofit's early struggles to secure funding. A candidate sentence reads: "Still, the organization's founders refused to scale back their original mission." Elsewhere in the essay, a question asks whether the writer should add a sentence describing the founders' personal financial sacrifices during this period, and whether the essay's stated goal, to "explain both the founding challenges and the organization's long-term impact," has been met.

Three separate mechanisms are at work. Placing the candidate sentence about refusing to scale back requires finding what "still" points back to, likely the funding struggles just described, and checking that the sentence after it continues logically from a refusal to compromise, not from some other idea. The add/delete question about the founders' financial sacrifices requires deciding first whether that detail serves this paragraph's focus on funding struggles, which it plausibly does, and only then checking whether the paired reason accurately describes that role, rather than a less accurate justification like emotional appeal alone. The whole-essay goal question requires checking both halves of the stated scope separately: if the essay thoroughly covers the founding challenges but only briefly gestures at the organization's impact in a single closing line, the goal is not fully met, and the correct answer has to name the specific missing half rather than treating the essay's clear coverage of the first half as sufficient for the whole.

Name the question type, read only as far as it requires, and test conclusions against the paragraph's actual evidence chain rather than against how smoothly an option reads. Those habits carry across every mechanism this chapter has covered.

Practice Set: Production of Writing

Each question below tests one mechanism from this chapter. Name the question type first, decide what evidence actually answers it, and only then compare the choices.

Edit a complete argument from claim through conclusion.

The questions below all refer to one continuous argumentative passage. Read the whole passage once before answering, and return to the specific paragraph each question references rather than relying on memory.

The curb is part of the transit system

Cities often treat the curb as leftover space: a narrow strip to be divided among parking meters, delivery trucks, ride-hail pickups, bus stops, bike-share docks, and whatever use happens to have arrived first. That approach worked tolerably when private parking dominated the competition. It works poorly now. On busy commercial streets, cities should manage curb space according to how many people and essential activities each use can serve, not according to how long a particular use has occupied the block.

Consider the block outside Riverton's Central Market. Until last year, eleven of its fourteen curb spaces were reserved for private parking. The remaining three were shared by deliveries and passenger pickup. During a six-week pilot, the city converted four parking spaces into two loading zones, one short-term passenger zone, and a longer bus boarding area. The pilot did not make every trip easier. Drivers searching for parking spent an average of 42 seconds longer finding a space within two blocks of the market.

That cost was real, but it was not the only measurable effect. Double-parking by delivery trucks fell 63 percent. Buses spent less time waiting for blocked curb access, cutting average delay on the block by 24 seconds per trip. The number of passengers boarding or exiting ride-hail vehicles in the travel lane fell sharply as well. City staff estimated that the four converted spaces directly served more than five times as many passenger or delivery movements per hour as they had when used for private parking.

Critics of the pilot object that movement counts favor short transactions by design. A parked car may occupy one space for an hour while its driver shops, they argue, but the economic activity associated with that visit can matter more than the number of times the curb itself turns over. That criticism identifies a real limitation in the city's initial analysis. Counting curb movements alone cannot establish which arrangement produces the greatest economic benefit.

The city therefore added sales-receipt and foot-traffic data from participating businesses during the final three weeks of the pilot. Those data did not show a decline in customer visits or average sales compared with the same weeks a year earlier. They also did not prove that the curb changes increased sales; too many other factors differed between the two periods. What they did show is narrower but useful: the most immediate economic fear raised by opponents, that reducing four parking spaces would sharply depress customer activity on the block, was not supported during the pilot.

Riverton should not treat six weeks on one commercial block as a universal formula for every street. Residential blocks, entertainment districts, schools, and hospitals create different curb demands. But the pilot demonstrates why cities should stop treating existing parking allocations as neutral. A curb policy is already a transportation policy. The relevant question is not whether a proposed change inconveniences anyone; almost every reallocation will. The better question is which combination of uses moves people, supports deliveries, and serves the block's actual function with the least avoidable interference.

Test the same argument skills in a different passage.

This passage uses a different structure from the curb-space passage: it opens with a common objection, reframes the issue, offers two supporting lines of evidence, concedes a real limitation late, and ends with a narrowed recommendation. Read the whole passage once before answering.

What a Dying Language Still Knows

(Original practice passage; the language, community, and details described are fictionalized for test-preparation use.)

Why spend public money recording a language that fewer than sixty people still speak fluently? It is a fair question, and anyone working to document an endangered language hears some version of it constantly. The honest answer is not that the speaker count is misleading. Sixty is sixty. The honest answer has nothing to do with the number itself.

A language is not simply a different set of words for ideas every language already has. It is often a different way of organizing which ideas get named at all. Reframed this way, the question is not how many people speak a language, but what disappears from human knowledge when nobody does.

Consider Kelmani, a language spoken by a small community in the highlands of a single valley. Kelmani grammar has several features that make it distinctive. A English sentence would need an entire paragraph to approximate, and even then imperfectly. When linguists began recording Kelmani speakers describing seasonal weather patterns, they documented forecasting knowledge, developed over centuries of close observation, that had never been written down in any form. That knowledge now exists nowhere else.

A second line of evidence comes from an unrelated documentation project in a different region entirely. Researchers involved in the project noticed some unexpected results partway through. None of this was the original goal. The documentation effort was funded to study grammar. What it preserved also included something worth remembering that the original project had not set out to document.

Documentation alone, however, does not keep a language alive. A recorded archive is not a living language, and no outside linguist can decide, on a community's behalf, whether continuing to speak a language matters enough to justify the real effort revitalization requires. This raises a difficult question that the essay must address carefully. Recording is not a substitute for that choice. It is, at most, a way of ensuring the choice remains possible.

The right policy, then, is not documentation instead of revitalization, and not documentation as a consolation prize for languages assumed to be beyond saving. It is public support for both: funding archival documentation wherever a language is at risk, and funding community-led revitalization wherever a community chooses to pursue it, without one effort being treated as a replacement for the other.

Narrative passage practice: personal narrative

English passages are not only informational or argumentative. A genuine share of real ACT English passages are personal or biographical narratives, told in first or third person, with characters, a timeline, and small concrete details doing the rhetorical work. The underlying skills tested, sentence boundaries, punctuation, word choice, and organization, are identical to what you have already practiced; only the genre changes. Work this passage the same way you would any other.

A Season at My Grandfather's Bench

The summer I turned twelve, my grandfather decided I was old enough to learn how to fix things. [1] Not big things, not cars or roofs, but small things: a sticking drawer, a wobbling chair leg, a clock that had stopped keeping honest time. His workshop smelled like sawdust and machine oil, and every tool [2] hung on the wall in an outline he had traced himself, so that a missing hammer announced itself instantly.

The first afternoon, he handed me the stopped clock and said nothing. I turned it over in my hands, [3] looking for a door and finding a screw instead. [4] Removing the back panel I found a maze of tiny gears, one of them clearly loose. "Most people," he said, watching me squint at it, "throw this away and buy a new one." He didn't say whether that was foolish. He let the sentence sit there.

By the third week, I could take apart a doorknob without losing a spring across the floor. [5] My grandfather rarely praised me directly, he would simply hand me something harder the next day, a toaster, a rusted bicycle bell, a music box whose song had slowed to a groan. [6] Fixing the music box took me four tries, and on the fourth my grandfather was not in the room to see it work.

I found him on the back porch, and when I told him, [7] he nodded slow, like a man confirming something he had suspected for a while. "You didn't need me for that one," he said. I understood, eventually, [8] that this was not a small thing to say. He had spent the summer handing me tools and stepping back a little further each time, [9] and the music box was the first repair that was entirely mine.

[10] Some of what he taught me left with the workshop when we sold the house after he died. The names of certain screws. The trick for a drawer that swells in humid weather. But the rest of it, [11] patience with something that resists you, care with tools that can be damaged easily and the willingness to open a thing up before deciding its broken beyond saving, has stayed with me [12] longer than any single repair ever could.

Part II: Reading