Verbal reasoning on the GMAT is built around a core framework: what kind of reasoning is being used, and what kind of text is being analyzed. Understanding this framework makes CR and RC dramatically more predictable.

A) What Kind of Text Is It? Four Types

Every GMAT verbal stimulus — CR or RC — falls into one of four categories. Identifying the type immediately tells you what to look for.

1. Arguments

An argument gives one or more reasons (premises) for accepting a conclusion. The premises are offered as evidence or support for the claim. Often some of these ideas are implied but not stated — the argument assumes a connection the author doesn't spell out. These unstated ideas are the assumptions CR questions test.

• Key signal: conclusion words (therefore, thus, hence, so, it follows that) and premise words (because, since, given that).
• CR questions are almost always arguments. For RC, look for passages where the author advances a claim and backs it up.
• Example: "The sidewalk is dry, so it must not have rained last night." The argument implies but doesn't state that rain typically leaves sidewalks wet.
• Example: "Sales of electric vehicles rose 40% last year. Consumer interest in sustainability is clearly increasing." Premise = sales data. Conclusion = consumer interest claim.

Core definitions every argument has:

• Premise: an idea given as a reason to accept another idea. An argument can have any number of premises. See D) for the full OG marker word list.
• Conclusion: the idea the argument is trying to get you to accept. The main conclusion is the final claim — it doesn't support anything else in the argument. An intermediate conclusion is one that acts as both a conclusion (supported by premises) and a premise (supporting the main conclusion). See D) for the full OG marker word list.
• Assumption: an idea the argument takes for granted but doesn't state. A necessary assumption must be true for the conclusion to hold — negating it destroys the argument. A sufficient assumption guarantees the conclusion if true. See E) for full definitions and examples.

2. Explanations

An explanation tries to account for something that has already been established as fact. Unlike an argument, the conclusion (the phenomenon) is not in dispute — what needs explaining is why or how it happened.

• Key signal: "explains why," "accounts for," "is responsible for," "resulted from."
• Paradox/Discrepancy CR questions are almost always explanations — you find what explains two puzzling facts.
• Explanations are NOT arguments: the "conclusion" is already accepted. The question is what caused it.
• Example: "Despite increased advertising, sales fell 20%. An explanation: a competitor launched a superior product that same month." The fact (sales fell) is given. The explanation is the cause.
• Trap: confusing an argument with an explanation. If the event is already accepted as fact, you are in explanation territory — weaken/strengthen logic does not apply the same way.

3. Plans and Proposals

A plan or proposal argues that a specific course of action should be taken (or will achieve a goal). Premises describe the problem or desired outcome; the conclusion recommends the action.

• Key signal: "should," "recommend," "in order to," "the best approach is," "propose," "strategy."
• Common in CR Weaken and Flaw questions: "A company proposes X to achieve Y. Which of the following, if true, would most undermine this plan?"
• To weaken a plan: show it won't achieve the goal, has an unacceptable side effect, or that a better alternative exists.
• To strengthen: show it achieves the goal, eliminates objections, or is more effective than alternatives.
• Example: "To reduce employee turnover, the company should raise salaries by 10%." Weaken: "Research shows most employees leave for reasons unrelated to compensation." Strengthen: "Exit surveys show salary dissatisfaction is cited in 80% of departures."

4. Narratives and Descriptions

A narrative or description presents facts or events without making an argument or offering an explanation. It reports what happened or what something is like.

• Key signal: past tense storytelling, factual inventory, "X is characterized by," "the study found that."
• These appear most in RC. A science passage describing an ecosystem, or a history passage recounting events, may be primarily descriptive.
• For RC Main Purpose questions: if the passage is descriptive, the correct answer will say "describes," "examines," or "surveys" — NOT "argues" or "advocates."
• Do not import your own conclusions: a descriptive passage does not take sides. The author is presenting information, not arguing for a position. Inference questions from descriptive passages require sticking closely to what is stated.

B) Inductive Reasoning: Arguments from Evidence

Inductive reasoning draws general conclusions from specific observations. The premises make the conclusion probable — but not certain. No matter how much evidence you gather, an inductive conclusion can always be wrong (new evidence could reverse it). This inherent uncertainty is what CR questions exploit.

Inductive vs. Deductive: The Core Distinction

• Deductive: if the premises are true, the conclusion must be true. Valid deductive arguments guarantee their conclusions.
  Example: "All humans are mortal. Socrates is a human. Therefore Socrates is mortal." — guaranteed.
• Inductive: premises make the conclusion likely, not certain. Evidence supports but does not guarantee.
  Example: "The last 10 times I took this route, there was traffic. Therefore there will be traffic today." — probably true, not guaranteed.
• GMAT implication: almost all CR arguments are inductive. This is why they can be weakened — inductive conclusions are not logically guaranteed, so they can always be made less likely.

Generalizations

A generalization draws a conclusion about all (or most) members of a group based on observations of a sample.

• Structure: [Sample observation] → [Conclusion about the whole group]
• Valid when: the sample is large, random, and representative of the target group.
• Vulnerable when: the sample is small, self-selected, or systematically different from the broader group.
• Example: "We surveyed 200 customers at our flagship store. 80% were satisfied. Therefore 80% of all our customers are satisfied." Weakness: flagship customers may differ from customers at other stores.
• Weaken a generalization: show the sample is unrepresentative or too small. Strengthen: show the sample closely matches the target population.

Predictions

A prediction concludes that something will happen in the future based on past patterns or current trends.

• Structure: [Past trend or current condition] → [Future outcome]
• Vulnerable when: the future context differs from the past, or the trend reverses for reasons not accounted for.
• Example: "Online retail has grown 15% each year for five years. Therefore online retail will grow 15% next year." Weakness: saturation effects, new competitors, or economic downturns could break the trend.
• Weaken a prediction: show a factor that will change conditions going forward. Strengthen: show the conditions producing the trend will continue.

Analogical Arguments

An analogy argues that because two situations are similar in relevant ways, what is true of one is likely true of the other.

• Structure: [A and B are similar in key respects] + [X is true of A] → [X is likely true of B]
• Strength depends on: how many relevant similarities exist, and whether the similarities that matter for the conclusion are present.
• Vulnerability: the two situations may differ in a way that is crucial to the outcome.
• Example: "This marketing strategy worked for a tech company in California. We should use it for our retail chain in Texas." Weakness: tech and retail companies face different customer bases, product types, and market dynamics.
• Weaken: identify a relevant difference between the two situations. Strengthen: show the difference doesn't affect the outcome, or that the relevant similarities hold.
• Parallel Reasoning CR questions are built on analogical structure — find the answer with the same logical pattern.

C) Deductive Reasoning: Logical Precision

Deductive reasoning produces conclusions that are guaranteed given true premises. GMAT Must Be True questions are almost entirely deductive — you apply the rules below to extract what the stimulus guarantees.

Logical Operators

• AND: "A and B" — both must be true. Negate: "not A or not B" (De Morgan).
• OR: "A or B" — at least one is true (inclusive). Negate: "not A and not B."
• IF–THEN: "If A then B" — A is sufficient for B; B is necessary for A. P→Q does NOT mean Q→P.
• NOT: negation flips a statement. Negating "all" → "not all (some are not)". Negating "none" → "at least one."

Valid Deductive Inference Patterns

• Modus Ponens: If P→Q and P is true, then Q must be true.
  Example: "All doctors studied medicine. Ana is a doctor → Ana studied medicine." ✓
• Modus Tollens (Contrapositive): If P→Q and Q is false, then P must be false.
  Example: "All doctors studied medicine. Carlos did not study medicine → Carlos is not a doctor." ✓
• Chain reasoning: P→Q and Q→R → P→R. Valid.
  Example: "All A are B. All B are C → All A are C." ✓
• Affirming the Consequent (INVALID): P→Q and Q is true does NOT mean P is true.
  Example: "All doctors studied medicine. Maria studied medicine → Maria is a doctor." ✗ — she could be a nurse who also studied medicine.
• Denying the Antecedent (INVALID): P→Q and P is false does NOT mean Q is false.
  Example: "If it rains, the ground is wet. It did not rain → the ground is not wet." ✗ — sprinklers could have run.

Necessity, Possibility, and Probability

These three modal concepts determine how strongly a conclusion follows from premises — and how strong GMAT answer choices can be.

• Necessity (must): the conclusion is guaranteed by the premises. Deductive inference yields necessary conclusions. Key words: must, necessarily, cannot but, is certain to, always, in all cases.
• Possibility (could/might): the conclusion is consistent with the premises but not guaranteed. Key words: could, might, may, is possible that, in some cases.
• Probability (likely): the conclusion is probable but not certain given the premises. Inductive arguments produce probable conclusions. Key words: likely, probably, generally, tends to, in most cases.

GMAT Application: Must Be True questions require necessity — the correct answer must be guaranteed, not merely possible or likely. An answer that "could be true" or "is probably true" is NOT correct for a Must Be True question. The bar is logical certainty given the premises.

Example: Premises: "Some economists are politicians. All politicians are public figures." What must be true? → "Some economists are public figures." ✓ (guaranteed by chain + quantifier logic). "All economists are public figures" — NOT guaranteed (only some economists are politicians). "All public figures are economists" — NOT guaranteed (reverse is invalid).

Quantifiers and Reasoning Rules

• All A are B: any A is also B. Does NOT mean all B are A. Contrapositive: not-B → not-A.
• No A are B: zero overlap. Reversible: "No B are A" is also true.
• Some A are B: at least one A is B. "Some" = one or more — not most, not all. Reversible: "Some B are A."
• Most A are B: more than half of A are B. NOT reversible (most B are not necessarily A).
• Chain with "some": "All A are B" + "Some B are C" does NOT guarantee "Some A are C." The "some B" may or may not be in the A subset.
• Chain with "all": "All A are B" + "All B are C" → "All A are C." Valid chain. ✓

Example trap: "All dogs are mammals. Some mammals are dangerous. Are some dogs dangerous?" — NOT guaranteed. The "some dangerous mammals" might all be cats. This is one of the most common Must Be True traps on the GMAT.

D) OG Premise & Conclusion Markers

The Official Guide provides specific word lists for recognizing argument structure. Memorizing these markers accelerates CR parsing speed.

Premise Markers (OG List)

These words introduce premises — the reasons offered to support a conclusion. What follows is evidence, not the main claim.

Conclusion Markers (OG List)

These words introduce the conclusion — the main claim being supported. What follows is what the author is trying to prove.

Key insight: marker words are the fastest way to identify argument structure. However, markers can be absent — always verify by asking "Is this a reason for something else, or is this the main claim?"

Intermediate Conclusions

A conclusion can itself serve as a premise for another, higher-level conclusion. This creates a chain:

• Premise → supports → Intermediate conclusion → supports → Main conclusion
• The intermediate conclusion is both a conclusion (relative to what supports it) and a premise (relative to the main conclusion).
• Example: "The program increased graduation rates (P₁). Programs that increase graduation rates benefit communities (P₂). Therefore, the program benefited the community (IC). This benefit justifies continued funding (MC)."

The "Why Test" for Finding Conclusions

Ask: "Why does the author believe [statement X]?" If Y answers that question, then Y is a premise supporting X, and X is a conclusion. The main conclusion is the statement that is supported but does not itself support anything else in the argument.

Example: "Expanding the park will improve air quality (C). Expanding the park will also provide recreational space (P₂). These benefits justify the cost (MC)." The main conclusion is the last statement — the other two are premises that feed into it.

E) Valid vs. Sound Arguments & Assumption Types

• Valid argument: the conclusion follows from the premises IF they are true. Validity is about logical structure, not truth of premises. A valid argument can have false premises.
• Sound argument: valid AND the premises are actually true. A sound argument guarantees a true conclusion.
• GMAT implication: CR questions tell you to accept premises as true ("if the above is true"). You are always evaluating validity — whether the conclusion follows. Soundness is assumed if premises are given as true.
• Flawed argument: an argument whose conclusion does not follow from its premises even if the premises are true. This is what Flaw questions test.

Valid vs. Sound — OG Examples

Example (i): "Everyone who tries fried eggplant is guaranteed to love the taste. So, if you try it, you'll love the taste too."
This is valid but not sound. The logical structure holds (if everyone loves it, you will too), but the premise is false — not everyone is guaranteed to love fried eggplant.

Example (ii): "Some people who try fried eggplant dislike the taste. So, if you try it, you'll probably dislike the taste too."
This is not valid (and therefore not sound). The premise is true — some people do dislike it. But "some dislike" does not mean you personally will probably dislike it. The logical gap makes it invalid.

Assumption Types

Every argument with a gap between premises and conclusion relies on at least one unstated assumption. An assumption is an idea taken for granted — it may be a premise, a causal claim, a condition, or any idea the argument needs but doesn't state. A passage may have implicit assumptions the author considers too obvious to state. These fill logical gaps between statements. An argument with implausible assumptions is weak.

• Necessary assumption: MUST be true for the conclusion to follow. Negating it destroys the argument. Verified with the Negation Test.
  Example: "Mario's flight arrives at 5:00 pm, so by 7:00 pm we'll be going out to dinner with him." Necessary assumptions: (1) the flight will arrive not much later than scheduled, and (2) Mario caught the flight. Without either, the conclusion fails.
• Sufficient assumption: IF it is true, the conclusion is guaranteed. Adding it makes the argument valid. May be stronger than strictly needed.
  Example: "The study of poetry is entirely without value since poetry has no practical use." One sufficient assumption: "Studying anything with no practical use is entirely without value." If that assumption is true, the conclusion must follow — but this is not a necessary assumption (there could be other routes to validity).
• GMAT Assumption questions usually ask for necessary assumptions — verified with the Negation Test.

Argument Types by Conclusion Kind

• Prescriptive: concludes what should or shouldn't be done. Advocates for or against policies, procedures, or actions. Key signal: "should," "recommend," "the best approach."
  Example: "Our company's staff is too small to handle this project. So the company should hire more employees."
• Evaluative: concludes that something is good or bad, desirable or undesirable — without advocating any specific action.
  Example: "This early novel is clearly one of the greatest of all time, pioneering brilliant narrative techniques with exceptional grace."
• Interpretive: concludes about something's underlying significance, meaning, importance, or implications — often applied to observations, theories, art, or historical events.
  Example: An argument that a historical policy shift reflects a deeper change in societal values, not just a practical decision.

F) Logically Equivalent Forms of Conditionals

Multiple phrasings express the same IF–THEN relationship. Recognizing them prevents mistaking an equivalent form for a new claim.

NOT equivalent (common traps): "If A then B" does NOT mean "If B then A" (converse) or "If not-A then not-B" (inverse). Only the contrapositive is equivalent.

Conjunction markers (and, although, even though, but, however) signal that both parts are asserted. Disjunction: "A or B" + "not A" → "B must be true" is a valid deductive inference.

G) Official Modal Word Table (Necessity / Probability / Possibility)

The GMAT distinguishes three modal strengths. Knowing which words signal which modal type is critical for Must Be True questions. probably means "more likely than not" (>50%), not "virtually certain."

H) Quantifier Synonyms Table

The GMAT uses all synonyms below interchangeably. Failing to recognize a synonym is a common Must Be True error.

Special case — "any": "any" can mean either "all" or "some" depending on context. In "Do any students qualify?" it means "some." In "Any student who qualifies may register," it means "all who qualify." Read carefully.

I) Syllogisms: Valid and Invalid Forms

A syllogism has two premises and a conclusion. The GMAT tests whether you can distinguish valid from invalid syllogistic reasoning.

Valid Syllogism Forms

Invalid Syllogism Forms (Common Traps)

Reasoning with Quantifiers: Practice Framework

1. Identify the quantifier for each premise (all, most, some, no).
2. Draw a Venn-style mental picture: do the groups overlap or not?
3. Ask: does the stated conclusion follow necessarily, or only possibly?
4. Test the contrapositive: if "All A are B," then "not-B → not-A" must also be true.
5. For "some" conclusions: ask "could the 'some' that satisfies premise 1 be a different subset than the 'some' in premise 2?"