Section I 1 — what the prose is doing

Twelve questions on Section I — 45% of the exam, and the half a question bank can honestly train.

Seven reading questions. Each gives a short non-fiction excerpt and asks how the prose works: what a named sentence accomplishes in the argument, what a metaphor achieves that a plain statement would not, what a rhetorical question is doing, whom the writer appears to be addressing.

That last one is the most useful question in the set, because audience is the thing students most often fail to notice. An excerpt here opens "Anyone who has stood at a lambing pen at three in the morning knows what I am about to say" and follows it with "The rest of you may need convincing" — two audiences, named in the open, addressed differently. Its wrong options all make the same mistake in different ways: they confuse who is being talked about with who is being talked to. The pronouns are almost always the giveaway.

The options throughout name rhetorical functions — concedes, qualifies, anticipates an objection, reframes a familiar thing as an unfamiliar one — rather than restating the content. That is the whole distinction the course rests on: what a text says against what a text does. A rhetorical analysis essay that summarises beautifully scores badly, and the habit that prevents it is trained here, one sentence at a time.

Five writing questions. A short draft, a stated purpose, and four revisions of which exactly one accomplishes it. The other three are grammatical, accurate and beside the point — which is the format's whole idea. A revision that reads better but does not do the named job is wrong.

One of them works differently and is worth singling out: a draft promises that "the councils that have actually tried it tell a different story", and you are asked which sentence would supply the evidence that promise creates. Three of the four options are true statements about the situation — that shopkeepers feel strongly, that the debate has run a long time, that opponents have published no figures of their own — and none of them is the evidence that was promised. Noticing that a sentence has made a promise, and that the next sentence has to keep it, is a skill that transfers directly into the argument and synthesis essays.

What this does not do. Section II is 55% of the score: three essays in two hours and a quarter, scored out of six each on a thesis point, four points of evidence and commentary, and one point for sophistication. Nothing that is not an essay prepares you for that.

But the perception the essays depend on is trainable, and it is what these questions train. A student who can say what a concession is doing in someone else's paragraph writes a rhetorical analysis; a student who cannot writes summary. Then go and write one under time, and mark it against the published rubric, counting how many of your sentences are commentary rather than quotation. Most people are surprised.

Nothing here is quoted from anywhere. Every excerpt was written for this material, and no College Board content is reproduced.

• Say what a sentence DOES in an argument, rather than what it says
• Name rhetorical functions precisely: concession, qualification, anticipating an objection, reframing
• Identify a writer's audience from the pronouns and the concessions, and tell it from the subject matter
• Explain what a metaphor or an image achieves that a plain statement would not
• Recognise a promise made by one sentence and what the next one has to deliver
• Choose the revision that accomplishes a stated purpose, among four that are all grammatical and true
• Tell summary apart from analysis — the distinction the whole course rests on

"Anyone who has stood at a lambing pen at three in the morning knows what I am about to say, and will forgive me for saying it plainly. The rest of you may need convincing." — Whom does the writer appear to be addressing? The split is made in the open: one sentence names readers with direct experience, the next turns to those without it. Policymakers are the subject of the criticism and landlords are its object, and neither of those is the same as being its audience.