Thirteen questions in the three shapes this domain uses, and they demand three different kinds of attention.
Seven words in context. A short passage with a blank and four words that all mean roughly the same thing. Nuanced, systematic, cursory, tenuous, refute, undermine, exacerbate — ordinary academic vocabulary, not obscure words, and the question is never "do you know this word" but "does this word do what this sentence needs done". The sentence always contains the constraint: a contrast with "random" forces systematic; a description of skimming rather than deep work forces cursory; a plan that had the opposite effect forces exacerbate.
The method that works is to predict before you look. Decide from the sentence alone what the blank has to mean, in your own words, and only then read the options. Reading the four first makes all four sound possible, which is exactly what they were chosen to do.
Three on what a sentence is doing. These ask for the function of a piece of text, and the answers are verbs: this sentence concedes, illustrates, qualifies, introduces a complication, blocks an interpretation the reader was about to form. One of them turns on a single word — the passage says "Still", and "still" answers an objection, which means the sentence before it must have raised one. Learning to read connectives as structural signals is most of this question type.
The habit to build: after reading any sentence, be able to say in one verb what it did, not what it was about. That is a skill you can practise on anything you read, and almost nobody does.
Three cross-text pairs. Two short passages on the same topic — standardised testing, electric vehicles, remote work — that agree about most of it and part company on one specific step. The question asks how the second author would respond to a particular claim in the first.
The trap is always the same and it is worth naming: the two texts usually agree about more than they disagree about, so a wrong choice will either take something both authors accept and present it as the dispute, or will have the second author reject the first wholesale when in fact they accept the claim and object to what is drawn from it. Find the sentence in Text 2 that the author of Text 1 could not have written. The answer is about that sentence and nothing else.
Fifteen flashcards carry the vocabulary the way it is actually tested — each word against the near-synonym it is most easily confused with, because that pairing is what a question turns on.
"The claim that the Antonine Wall was abandoned within a generation of its building rests almost entirely on coins. Coins do not prove occupation, only supply. Still, the pattern is hard to account for any other way." — What is the function of the middle sentence? The word after it is "Still", which answers an objection; so the sentence before must have raised one. The writer concedes that coin finds show what was delivered rather than who lived there, and then defends the conclusion anyway.
In the study of historical linguistics, the theory that language change is entirely random has been largely abandoned. Most researchers now argue that while some shifts are unpredictable, the overall trajectory of a language is ________, shaped by social pressures and communicative efficiency rather than mere chance.
systematic
Systematic is correct because it contrasts with the idea of randomness, implying an orderly process. Arbitrary means random, which contradicts the context. Cursory means hasty or superficial, which does not fit the description of language change. Tenuous means weak or flimsy, which is not the intended meaning here.