The 12 logical reasoning argument structures

Every logical reasoning question on the LSAT asks you to do one of twelve things to an argument, and once you can name which one, the right answer has a shape you can predict before you read the choices. This guide is the full list, in the order we teach it, with what the correct answer must do for each and an original example you can check yourself against. If you already know the names and still read answer choices cold, the prephrase lines are the part to steal.

The structures matter more now than they did two years ago. Since the August 2024 administration, the test has two scored logical reasoning sections of roughly 24 to 26 questions each, one scored reading comprehension section, and one unscored section, all 35 minutes, per LSAC. Roughly two thirds of your scored questions are these twelve structures. Learn them as a closed set, not as a stack of flashcards.

Group one: describe the argument

The first three structures ask you to say what the argument is, not whether it works. Students skip them in study because they feel easy, then drop one or two per section to a trap that restates a premise as the conclusion.

1. Main conclusion

Prephrase: the one sentence everything else is trying to prove, in my own words, before I look. Reword it rather than quoting it, because the trap answers quote.

2. Role of a statement

Prephrase: is the quoted sentence the conclusion, a premise, a counter-premise the author argues against, or background? Say which.

Take the library stimulus. If the question quotes “the branch was never meant to hold a large collection,” its role is a premise offered to support the main conclusion that the critics are using the wrong standard. The trap will call it the main conclusion or an intermediate conclusion. It supports; it is not supported. That makes it a premise.

3. Method of reasoning

Prephrase: in abstract terms, what did the author do? Attack a standard, give a counterexample, draw an analogy, rule out alternatives, appeal to a general principle?

The library author’s method is to reject the critics’ standard of evaluation and substitute a different one. The right answer will say something like “argues that an evaluation rests on an inappropriate criterion.” The wrong answers will describe methods that would also have worked but were not used, such as “cites statistics about library usage.” Match what happened, not what could have.

Group two: find the flaw

The next three structures all live on the same thing: the gap between what the premises establish and what the conclusion claims. Find the gap and you have prephrased all of them. The guide on the five misses at 161 covers the gaps that students miss most.

4. Flaw

Prephrase: name the gap in plain words first (“assumes the only cause is X,” “shifts from residents to attendees”), then translate it into test language. Plain words come first because the test language is where the wrong answers hide.

5. Necessary assumption

Prephrase: what is the smallest thing that must be true for this to work? Then confirm with the negation test: negate the answer, and the argument should collapse.

6. Sufficient assumption

Prephrase: which statement, added to the premises, would make the conclusion follow with certainty? Usually a bridge between a term in the premise and a term in the conclusion.

Group three: change the argument

Four structures ask you to add a fact and watch what happens to the conclusion. The common error is evaluating whether the answer is true. It is given as true. The only question is what it does.

7. Strengthen

Prephrase: the gap, then the fact that would close it, or rule out the most obvious alternative explanation. Say the rival cause out loud.

8. Weaken

Prephrase: the gap, then the fact that widens it, usually an alternative cause, a counterexample, or evidence that the sample does not represent the population. Match the conclusion’s timing and degree exactly.

For the website argument, weaken with “The quarter in question included the company’s annual clearance sale, which has produced a 15 to 25% jump every year.” Note the degree and timing match the conclusion exactly, which is what makes it a weakener and not background. Weaken answers do not have to destroy the argument. They have to make the conclusion less likely than it was.

9. Evaluate

Prephrase: what single fact, if I knew it, would tell me whether the conclusion holds? Phrase it as a question, then check the answer by imagining both yes and no.

Website argument: “Did the company change anything else about its sales process during the quarter?” If yes, the argument is weakened; if no, strengthened. An answer that produces the same verdict either way is not the evaluate answer, even if it sounds relevant.

10. Resolve the paradox

Prephrase: state the two facts that seem to clash, then the kind of fact that would let both be true at once. Both facts must survive.

Group four: extract and match

The last two structures treat the stimulus as a set of facts or as a pattern, and both reward diagramming when conditionals are present. They are also where the converse does most of its damage, so the contrapositive check from the plateau guide applies on every one.

11. Inference: must be true, and most strongly supported

Prephrase: combine two statements, or take a contrapositive, and say what falls out. For must-be-true, demand certainty. For most-strongly-supported, accept the conclusion the facts make very likely.

The point-at-issue variant gives two speakers and asks what they disagree about. Prephrase it the same way: find a statement one speaker must affirm and the other must deny. If either speaker’s view on the answer is unknown, it is not the point at issue.

12. Parallel reasoning, and parallel flaw

Prephrase: abstract the stimulus into a skeleton (what kind of conclusion, how many premises, what logical move), then match skeletons, ignoring subject matter entirely. Topic is noise.

How to use the taxonomy

Knowing the twelve is step one. Making them automatic is the work, and it has an order. First, tag before answering: for three full sections, write the structure number next to each question before you solve it. Your tagging should be near-perfect within a week; if it is not, the question stems themselves are costing you time.

Second, write the prephrase before the choices, every time, for two weeks. It will feel slow. It is slow for about forty questions, and then it becomes faster than reading the choices cold, because you stop evaluating five answers and start looking for one.

Third, keep a tally of misses by structure in blind review. Students in the low 160s typically find that three structures carry most of their misses, most often necessary assumption, flaw and the inference pair. That is where the prephrasing drills go. The rest of the list is maintenance.

Rick’s note: I teach these as a closed set on purpose. When a student believes there are a hundred kinds of question, every stimulus is new and frightening. When they believe there are twelve, every stimulus is one they have already met. The second belief happens to be true.

What to do this week

1. Print this list of twelve and tape it where you study. Say each structure’s job out loud until you can do it without looking.
2. Take two timed logical reasoning sections from recent PrepTests. Before answering each question, write its structure number in the margin.
3. Check your tags against the question stems afterward. Fix any structure you mis-tagged more than once by rereading its section above.
4. For the questions you missed, write which of the twelve each belonged to and whether your prephrase existed, was wrong, or was skipped.
5. If two structures carry most of your misses, drill those two untimed for the rest of the week. If the misses are spread flat or you can’t tag confidently, that is the sign to request a diagnostic debrief.