Craft & Structure + Cross-Text
Central ideas · purpose & structure · paired-passage reasoning
6 min read
Read for what a sentence does, not just what it says — function beats content on these questions. This module covers the remaining Information & Ideas and Craft & Structure types: Central Ideas & Details, Text Structure & Purpose, and Cross-Text Connections (the paired-passage type, usually the hardest single reading item on the test).
Sources: Meltzer ("read for function"); UWorld. Examples MERIDIAN-original. Point convention: each = 1 of ~54 R&W questions; Module-1 misses are ceiling-risk.
TYPE 1 — CENTRAL IDEAS & DETAILS
What it tests
Identify the main idea of a short passage, or a specific detail it supports.
Method
1. Summarise the passage's MAIN POINT in your own ≤8-word phrase before the choices.
2. Main idea = the claim the whole passage serves (not one supporting detail).
3. Detail questions: locate the exact sentence; the answer paraphrases it faithfully.
4. Eliminate choices that are too narrow (a detail posing as the main idea) or too broad.
Why (mechanism): the trap for main-idea is a true detail that is too narrow; the trap for detail is a distorted paraphrase. Pre-summarising forces you to separate the central claim from its supports. [A wrong main-idea pick is usually a real detail — check scope.]
Trap taxonomy
- Detail-as-main-idea: true but covers only part of the passage.
- Too-broad: a generic statement the passage doesn't fully establish.
- Distorted paraphrase: changes a word that changes the meaning.
- Opposite: reverses the author's stance.
Central Ideas — two levels (MERIDIAN-original)
Passage: "Although the city's new bus lanes drew complaints during construction, ridership rose 30% within a year and car traffic fell — and the author notes several such projects where early backlash gave way to measurable gains." Q: main idea? (A) The bus lanes drew complaints during construction. (B) Transit projects often face early opposition that later yields measurable benefits. (C) Ridership rose 30% in a year. (D) Car traffic is a serious urban problem.
- ~700 (trap): (C) — a true, specific figure. Trap: detail-as-main-idea (too narrow). [−1]
- 780+ (precise): pre-summarise the whole passage — "early backlash → later gains, across projects" → (B). Eliminate (A)/(C) too narrow, (D) too broad. Move: summarised the whole-passage claim before reading the choices. [secures 1 question; Module-1 safe]
TYPE 2 — TEXT STRUCTURE & PURPOSE
What it tests
The rhetorical function of a sentence/section ("what does the underlined sentence do?") or the overall structure/purpose of the passage.
Method
1. Ask "what JOB does this sentence do?" — introduce a counterargument? give an example?
qualify a claim? present then complicate? define a term?
2. Describe the function in a verb phrase BEFORE the choices.
3. Match the choice whose VERB matches your function. Ignore choices that merely restate content.
Why (mechanism): wrong answers describe what the sentence says or a function it doesn't perform. The right answer names the move. Training yourself to read structurally ("this sentence concedes, then the next rebuts") is the 800-level habit. [Function words — "introduces," "qualifies," "illustrates," "contrasts" — are the signal.]
Trap taxonomy
- Content-restatement: summarises meaning instead of naming the function.
- Wrong function: a real rhetorical move the sentence doesn't actually make.
- Scope error: describes the whole passage when asked about one sentence (or vice versa).
Text Structure — two levels (MERIDIAN-original)
Passage: "Most historians attribute the method to Newton. Recent archival work, however, shows Hooke described the core idea years earlier." Q: what does the second sentence do? (A) restates the first sentence. (B) gives an example of Newton's work. (C) complicates the conventional attribution with new evidence. (D) praises Hooke's career.
- ~700 (trap): (A)/(B) — describes the content. Trap: content-restatement / wrong function. [−1]
- 800 (precise): name the function first — "introduces counter-evidence that complicates the standard view" → (C). Move: named the FUNCTION (a verb) before the choices. [secures 1 question; Module-1 safe]
TYPE 3 — CROSS-TEXT CONNECTIONS (paired passages)
What it tests
Two short passages on one topic. You determine the relationship — how Author 2 would respond to Author 1, where they agree/disagree, how their emphases differ.
Method
1. Summarise PASSAGE 1's position in a phrase.
2. Summarise PASSAGE 2's position in a phrase.
3. Define the RELATIONSHIP: agree · disagree · qualify · extend · explain differently.
4. Answer in those relationship terms — predict, then match.
Why (mechanism): the trap is a choice that is true of one passage but not of the relationship asked about. Holding both positions as short phrases — and naming the link — prevents you from answering about a single passage. [This is the hardest reading type because it taxes working memory; reduce each passage to one phrase to offload it.]
Trap taxonomy
- One-passage answer: true of P1 or P2 alone, not the relationship.
- Overstated disagreement: says they conflict when P2 only qualifies P1.
- Invented agreement: claims shared ground the texts don't establish.
- Direction flip: swaps which author holds which view.
Cross-Text — two levels (MERIDIAN-original)
P1: "Remote work raises productivity by cutting commute fatigue." P2: "Reported productivity gains from remote work fade once managers account for unmeasured overtime." Q: How does P2 relate to P1? (A) P2 offers independent support for P1. (B) P2 questions whether P1's gains are real once a confound is considered. (C) P2 and P1 discuss unrelated topics. (D) P2 agrees the commute is the key factor.
- ~700 (trap): (D) — "both mention productivity/work." Trap: one-passage/invented-agreement; P2 never endorses the commute point. [−1; and this is often a hard-module item — costly]
- 780+ (precise): P1 = remote work ↑ productivity; P2 = those gains may be an artefact (unmeasured overtime). Relationship = P2 challenges/qualifies P1 → (B). Eliminate (A) reversed, (C) false, (D) invented. Move: one phrase per passage, then named "challenges." [secures 1 question; Module-1 safe]
DIAGNOSE YOUR ANSWER
- Main idea wrong → picked a true detail? Scope error. Fix: pre-summarise the whole-passage claim.
- Purpose wrong → described content? Fix: name the function with a verb first.
- Cross-text wrong → answered about one passage? Fix: phrase each passage, then state the relationship before choosing.
- Overstated a disagreement? Fix: distinguish qualify/complicate from contradict.
DRILLS (timed · self-marked · MERIDIAN-original)
Drill 1 — Function-naming (target 6/6, ≤2 min): for 6 underlined sentences, write the function verb ("concedes," "illustrates," "defines," "contrasts," "qualifies," "introduces") before choices. Drill 2 — Two-phrase cross-text (target 4/4, ≤4 min): for 4 passage pairs, write P1 phrase, P2 phrase, and the relationship word; then match. PASS: relationship correctly named each time. Repetition: fail → repeat next day.
TRANSFER
Abstract rules: main idea = the claim the whole text serves; purpose = the function a part performs; cross-text = the relationship between two positions. All topic-independent. Second context: a poetry excerpt's "purpose" question and a geology passage's "purpose" question both reduce to "what move does this sentence make?" Constant: pre-summarise / name-function / phrase-both-then-relate. Changes: subject only.
REFERENCE CARD — CRAFT & STRUCTURE + CROSS-TEXT
CENTRAL IDEA: pre-summarise whole-passage claim (≤8 words); beware detail-as-main-idea
PURPOSE/STRUCTURE: name the FUNCTION (verb) before choices; ignore content-restatements
CROSS-TEXT: P1 phrase + P2 phrase → name RELATIONSHIP (agree/disagree/qualify/extend) → match
TRAPS: one-passage answer · overstated conflict · invented agreement · direction flip · scope error
Each item = 1 R&W question; Module-1 miss = ceiling risk.
MERIDIAN™ · © 2026 · For Neo's personal study use. Not affiliated with or endorsed by the College Board. SAT® is a trademark of the College Board.
MERIDIAN · FORGE · CRAFT & STRUCTURE + CROSS-TEXT | v1.0
Up next
Vocabulary-in-Context Vault
Vocabulary is back — but it's a reasoning skill, not a memorised list
3 min