V1 · Anatomy of an Argument
There are eight kinds of Critical Reasoning question and each has its own techniques. That is the next chapter. This one is about the thing all eight have in common, and it is a single skill — one you can learn in an afternoon and will use on every Critical Reasoning question you ever see.
Here is the pattern that makes this chapter necessary. A candidate reads a short argument, understands every sentence, forms an opinion about whether it is convincing, and picks the wrong answer. Not because they misread anything. Because they were evaluating the topic rather than the argument.
Those are different activities, and the difference is worth seeing before it is explained.
Read this and decide, before going further, whether you find it convincing.
The museum abolished its admission charge in January. Attendance over the following year rose by 18 percent. Free admission has clearly widened the museum’s audience.
Most readers find it reasonable. Free entry, more visitors — the connection is obvious.
Now answer a different question. Not is this true? but: what would have to be the case for this reasoning to work?
Try it before reading on.
Here is what the argument actually gives you. Two facts: the charge was abolished, and attendance rose. And one claim built on them: the museum’s audience is now wider.
Between those two things sits a gap, and it has two parts.
First, the argument assumes nothing else caused the rise. A major exhibition, a new transport link, a change in opening hours, a competitor closing — any of these would produce 18 percent with free admission contributing nothing.
Second — and this one is easier to miss — more attendance is not the same thing as a wider audience. If the same people now visit more often because it costs nothing, attendance rises and the audience has not widened at all. The argument has quietly swapped one claim for another.
Neither of those is stated anywhere. Both have to be true for the conclusion to follow. And an exam question about this argument will almost certainly be about one of them — because that is where arguments break.
You did three things in sequence:
- Found what the argument concludes
- Found what it offers in support
- Named what has to be true to get from the second to the first
That is the whole skill. It has a vocabulary — conclusion, premises, assumption — and the rest of this chapter teaches you to do it reliably, on arguments where the conclusion is not the last sentence and the gap is less visible.
But notice what you did not do. You did not decide whether free admission is good policy. You did not consider what you know about museums. You did not question whether attendance really rose. The facts were given, you accepted them, and you examined only the step from them to the conclusion.
That shift — from is this right? to does this follow? — is what the chapter builds, and it is the single highest-return habit in the Verbal section.
By the end of this chapter you will be able to take any short argument and say, in your own words and in about fifteen seconds: this concludes X, on the basis of Y, and it only works if Z.
That sentence is the answer to most Critical Reasoning questions before you have read a single option. Strengthen questions ask you to confirm Z. Weaken questions ask you to attack it. Assumption questions ask you to state it. Evaluate questions ask what would tell us whether Z holds.
Chapter V2 is those eight varieties. This chapter is the sentence.
Every Critical Reasoning argument has at most three parts. Two are on the page. One is not.
- Premises are the facts you are given. You accept them. This is not optional, and it trips up experienced professionals more than anyone, because in your working life the first thing you do with a claim is interrogate it. Here, if the argument says sales rose 40 percent, sales rose 40 percent. Your job is not to audit the data.
- The conclusion is what the argument wants you to accept on the strength of those premises. It is the destination.
- The assumption is what has to be true for the premises to actually get you to the conclusion — and it is not written anywhere. It is the gap between what you were told and what you were asked to believe.
The number of cyclists commuting into the city centre has risen 40 percent since the protected bike lanes opened three years ago. The lanes have clearly made cycling more attractive.
Premise: cyclist numbers rose 40 percent since the lanes opened.
Conclusion: the lanes made cycling more attractive.
Now find the gap. The premise establishes that two things happened in sequence. The conclusion says one caused the other. For that step to hold, something unstated must be true: that nothing else explains the rise.
Notice how much work that assumption is doing. Fuel prices might have doubled. The city might have grown. There might have been a two-year transit strike. Any of those would produce the same 40 percent with the lanes contributing nothing.
The assumption is where the argument is vulnerable, and that is why every question type in the next chapter is, underneath, a question about the assumption. Strengthen questions confirm it. Weaken questions attack it. Assumption questions name it. Evaluate questions ask what would test it.
Before you look at the answer choices — always before — say to yourself, in your own words: "This argument concludes X, based on Y, and it only works if Z."
Fifteen seconds. It is the highest-return habit in the entire Verbal section, and candidates who skip it end up choosing whichever option feels most relevant to the topic.
In a meeting, a number you distrust is the first thing you challenge, and being the person who challenges it is part of your value. That reflex is correct everywhere except here.
Critical Reasoning hands you the premises as fixed. The only live question is whether they reach the conclusion. Time spent doubting the 40 percent is time not spent on the step that actually breaks — and every option built around "but is that figure reliable?" is a distractor written for exactly your instinct.
Arguments in this format do not follow a fixed shape. The conclusion can be the last sentence, the first, or buried in the middle. Assuming it is last is a reliable way to answer the wrong question.
The lanes have made cycling more attractive. Cyclist numbers have risen 40 percent since they opened, and riders report feeling safer.
Test it: "The lanes have made cycling more attractive, because numbers rose and riders feel safer." That works — the first sentence is the conclusion.
Reverse it: "Numbers rose, because the lanes made cycling more attractive." That is a different claim, and not what the passage argues.
- Words that introduce a conclusion: therefore, thus, so, consequently, it follows that, clearly, hence.
- Words that introduce a premise: because, since, given that, as, for, in light of.
"But" and "however" do not mark conclusions — they mark a turn. Often the conclusion follows shortly after, but the turn word itself is not the signal.
"Clearly" and "obviously" are frequently a warning. In this format, when an argument tells you something is obvious, that is usually the step where the reasoning is weakest. Writers use those words to hurry you past a gap.
The one where the argument states a position it disagrees with first:
Supporters of the wetland restoration argue that it would save the county the six hundred thousand a year it now spends on water treatment. But the treatment plant would still be needed for industrial discharge, so those costs would not disappear — at best they would fall by a fraction. The restoration cannot be justified on treatment savings alone.
The first sentence is a position the argument rejects. Candidates who grab it as the conclusion get every subsequent step backwards. The actual conclusion is last here — but the trap is at the start.
When you see a passage open with "Supporters argue…", "It is often said…", "Critics maintain…" — that is almost always the position being attacked, not the argument's own.
A step that genuinely follows does not need to be announced as obvious — the premises carry it. Writers reach for "clearly" exactly where the reader might otherwise stop and object, which makes the word a reliable pointer to the join the argument would rather you did not inspect. Treat it as a highlighter the author did not mean to leave on.
Once you have the premises and the conclusion, the whole game is finding the space between them. The most reliable technique: look for a term that appears in the conclusion but not in the premises, or vice versa. Arguments break where the language shifts.
Employees who take the stairs instead of the elevator report fewer sick days than those who do not. The company should remove its elevators in order to reduce absenteeism.
Compare the vocabulary. The premise is about employees who choose the stairs. The conclusion is about removing the choice. That shift is the gap — and it contains two separate problems.
First, direction. Perhaps taking the stairs makes people healthier. Or perhaps healthier people take the stairs. The premise is equally consistent with both; the conclusion requires the first.
Second, transferability. Even if stairs improve health, that was measured on people who chose them. Forcing everyone up the stairs — including those who avoided them for a reason — need not produce the same result.
Neither problem is about whether stairs are healthy. Both are about the step from what was observed to what is proposed.
| Gap | The shape | The tell |
|---|---|---|
| Correlation to causation | Two things happened together, so one caused the other | The premise reports an association; the conclusion asserts a mechanism |
| Sample to population | It was true here, so it is true generally | The premise describes a specific group; the conclusion is about everyone |
| Part to whole | Each piece has a property, so the whole does | Percentages, averages or components in the premise; a total in the conclusion |
| Prediction from past | It worked before, so it will work again | Past tense in the premise, future tense in the conclusion |
Any argument that concludes something will happen, based on something that did happen, assumes the conditions have not changed. That assumption is almost never stated, and is almost always where the answer lives.
Here is the thing nobody tells you about Critical Reasoning answer choices: most of them are not wrong because they are false. They are wrong because they are about something slightly different. An answer choice can be entirely true, entirely reasonable, and entirely wrong, because it addresses a claim the argument did not make.
Too wide. The argument was about a specific case; the option makes a general claim.
Argument: this hospital's triage nurse reduced waiting times. · Option: hospitals should always employ triage nurses.
The option goes far beyond what the argument supports. If you found yourself thinking "well, that seems sensible" — that is the pull. Sensible is not the test.
Too narrow. The argument was general; the option addresses one instance.
Argument: the LED lighting programme reduced collisions city-wide. · Option: one intersection saw no change after the lights were installed.
A single exception rarely damages a claim about an aggregate. It feels like a counterexample and is not.
Words that signal a scope problem: all, every, always, never, none, only, must on the wide side. Any specific instance, single case or named example on the narrow side.
That said, do not turn this into a rule that eliminates any option containing "all". Sometimes the argument itself is universal and the correct answer must be too. The test is always match to the argument, never avoid strong words.
You are reading fast, and truth is far easier to check than relevance. An option that is plainly true gets a tick from the part of your mind that is scanning for errors, and the scan stops there.
The exam writes distractors that pass the truth check and fail the scope check, because that ordering is the one candidates actually use. Reversing it — scope first, truth second — costs nothing and disqualifies most of them before you have to weigh anything.
The exam is precise about how strongly a claim is made, and it expects you to be. Two options can be identical except for one word, and that word decides.
| Word | Means | Does not mean |
|---|---|---|
| some | at least one — possibly all | a few |
| most | more than half | nearly all |
| many | an unspecified number | a majority |
| all / every | without exception | almost always |
| none / no | zero cases | very few |
The one that catches people is "some". In everyday speech, "some employees" suggests a minority. In this format, "some" means at least one, and it is entirely compatible with "all". An argument that establishes "some" has not established "most", and an option claiming "most" goes beyond it.
| Word | Strength |
|---|---|
| must, will, always | Absolute — requires the strongest support |
| is likely to, probably | Probable — a lower bar |
| may, might, could, can | Possible — a very low bar |
That last row matters more than it looks. An option saying something may happen is claiming almost nothing, which makes it very hard to be wrong — and also, usually, very hard to be the answer, because it does not do enough work. Except in Inference questions, where the correct answer is often the weakest option on the page, precisely because it is the only one that must be true.
For strengthen and weaken questions, the opposite. Weak language does little. An option saying a policy "might have some effect" barely moves an argument either way.
The two question types ask opposite things of an option. Inference asks "is this guaranteed?", and a modest claim is far easier to guarantee — strength is a liability. Strengthen and weaken ask "does this move the argument?", and a modest claim moves nothing — strength is the whole point. Same words, inverted value, decided entirely by what the stem is asking for. Which is why reading the stem before the options is not a nicety.
Every error this chapter has described, in one place. When a Critical Reasoning item goes wrong before you have even reached the question type, it is almost always one of these seven. Open each to see the shape it takes and the defence against it.
Two parts to the practice. This one trains the anatomy directly — no answer choices, just structure. Part B is four full items that reward it. Do Part A first; it is the one that builds the habit.
Conclusion: the extended hours were a success.
Premise: attendance rose 15 percent in the year after they were extended.
Assumption: the rise is attributable to the extended hours rather than to something else — a major exhibition, a change in admission price, a tourism increase.
Correlation to causation — and note that the attendance figure is annual while the change was made in spring, so part of the rise predates it.
Conclusion: fencing should be required along the entire coastline.
Premise: towns with fencing lost less dune material.
Assumption: what worked in the towns that installed it will work along the rest of the coast — same exposure, same dune structure, same storm conditions.
Sample to population, with a second gap hiding underneath: the towns that installed fencing chose to, and that choice may correlate with other protective measures.
Conclusion: the policy has not lowered standards.
Premise: the proportion admitted to selective programmes held steady.
Assumption: admission to selective programmes is a reliable indicator of standards.
Note the structure — the first sentence is the position being rejected. This is the shape from V1.2, and the reason to run the "because" test rather than trusting position.
Conclusion: the software reduces delays.
Premise: every warehouse that switched reported fewer delays.
Assumption: switching caused the improvement — and not, say, that the warehouses that switched were the better-run ones, or that delays fell everywhere that quarter.
Correlation to causation, plus a self-selection problem: nobody assigned the software randomly.
Conclusion: the workshop should be mandatory for all patients.
Premise: attendees had lower average blood pressure than non-attendees.
Assumption: the workshop caused the difference, and the effect would carry over to people who chose not to attend.
The stairs argument from V1.3 in different clothes: the premise is about people who selected themselves in, the conclusion is about removing the choice.
Conclusion: the reservoir will be below capacity next year.
Premise: rainfall has declined for six consecutive years.
Assumption: the trend continues, and rainfall is the determining factor — no change in extraction, no new supply, no reversion to the mean.
Prediction from past, the most common gap in forecasting arguments.
Four items in the real format. Run the habit from V1.1 on each one: name the conclusion, the premise and the gap in your own words before you read a single option.
The board's conclusion depends on which of the following assumptions?
The board observed a correlation and is planning on the basis of causation. If something else drove the 22 percent — a staffing increase, a seasonal drop in admissions, a nearby clinic opening — then replicating the nurse elsewhere achieves nothing. Test it by negation: if the reduction was caused by an unrelated factor, the conclusion collapses. That is the definition of an assumption.
A is tempting and not required: the plan could still work at facilities with different baseline waiting times. C is a cost comparison the argument never makes — being cheapest is not needed for the method to work. D concerns satisfaction, not whether times will fall. E is a detail about one individual with no bearing on the general claim.
Which of the following, if true, most seriously undermines the authority's explanation?
The authority's explanation requires that the lighting is what changed. B supplies a control group: districts with the old lighting saw essentially the same decline. Whatever caused an 18 percent drop also caused a 17 percent drop where nothing was installed, so the lighting is not doing the work. This is the single most powerful way to attack a causal claim — show the effect occurring without the cause.
A is about budget, not causation. C is about comfort. D concerns daytime, when lighting is irrelevant. E is a real benefit of LEDs and says nothing about collisions. All four are on-topic and none touches the link — the trap from V1.4.
Which of the following most accurately expresses the main conclusion of the argument?
The conclusion is the first sentence, restated. Apply the "because" test: "Treatment savings are not sufficient grounds, because the facility would still be needed and costs would only fall by a fraction." That works.
A is a premise — and specifically a point made by the supporters, the position being answered. B is also a premise, the one that does the argumentative work. D misstates it: the argument does not claim the costs were overstated, it claims they would not disappear. E overreaches — the argument rejects one justification, not the restoration itself, and that distinction is exactly the scope problem from V1.4.
If the statements above are true, which of the following must also be true?
The first statement says: germinated → hard coat. Its contrapositive is no hard coat → did not germinate. That is B, and a contrapositive is always true when the original is.
A reverses the first statement and is directly contradicted by the second. C asserts causation where only a shared property was reported — the V1.3 gap. D introduces a quantity ("most") that neither statement supports; "some" tells you nothing about proportion, per V1.5. E claims the archive contains no soft-coated varieties, which nothing establishes — and note that B does not require any to exist. It is a conditional about any that might.
Find the conclusion, find the premises, and name what has to be true to get from one to the other — in your own words, before you read a single option.
V2 takes the eight Critical Reasoning question types one at a time. Every one of them is, underneath, a question about the assumption you learned to name here — which is why this chapter came first.