CSAT Part 2: Reasoning — Syllogisms, Assumptions and Course of Action, Exam-Ready Notes

Category: Civil Exams · Series: UPSC CSAT Paper II · Read time: ~9 minutes

The reasoning block of CSAT is formula-driven: syllogisms follow Venn-diagram rules, assumption questions follow the negation test, and course-of-action questions follow an administrative logic identical to the decision-making block. Learn the formulas once and this becomes the most mechanical scoring section of Paper II.

Table of Contents

1. The Reasoning Question Card
2. Syllogisms: The Venn-Diagram Method
3. The “Some/All/No” Conversion Rules
4. Assumptions and Statements: The Negation Test
5. Course of Action: The Administrative Filter
6. Inference, Conclusion and Punch-Line Types
7. How Exams Probe This Topic
8. Quick Revision: One-Glance Facts

1. The Reasoning Question Card

• The block. CSAT’s general mental ability carries roughly 12–20 marks across logical reasoning, analytical ability and data-based questions; syllogisms, assumption/conclusion sets, and course-of-action items are the recurring formats.
• The marking. +2.5 per correct, −0.83 per wrong in the standard format — reasoning rewards elimination discipline because options are engineered.
• The good news. Unlike RC, these questions have algorithms. Two algorithms (Venn diagrams, negation) cover the majority.

2. Syllogisms: The Venn-Diagram Method

• The format. Two or three statements (“All A are B; Some B are C”) followed by conclusions (“Some A are C — true/false/uncertain?”).
• The method. Draw the minimal diagram consistent with the statements, then test each conclusion: it is true only if it holds in every possible diagram, false if it fails in every diagram, and uncertain/follows-probably otherwise.
• The either/or option. When two conclusions are complements (“Some A are C” vs “No A is C”) and neither is definite but together they exhaust possibilities, the answer is “either I or II follows” — the most-skipped trick in syllogism sets.
• The definite/possible split. Recent papers phrase conclusions as “definitely true” vs “possibly true” — a conclusion is definitely true only if no diagram can break it.

3. The “Some/All/No” Conversion Rules

Memorise the immediate-inference table:

• “All A are B” → “Some B are A” (valid conversion). It does not give “All B are A” (converse fallacy — the classic trap).
• “No A is B” → “No B is A” (symmetric, valid).
• “Some A are B” → “Some B are A” (symmetric, valid).
• “All A are B” + “All B are C” → “All A are C” (chain).
• “Some A are B” + “All B are C” → “Some A are C” (valid).
• The invalid pair to flag: “Some A are B” + “Some B are C” gives nothing definite about A–C. This single rule eliminates half of all designed wrong conclusions.

4. Assumptions and Statements: The Negation Test

• The format. A statement (usually a policy announcement or advertisement: “The government will distribute free solar lamps in villages”) and candidate assumptions (“Solar lamps reduce kerosene use…”).
• The test. Negate the assumption. If the statement makes no sense or collapses without it, it is an implicit assumption; if the statement survives the negation, it is not.
• The filters. An assumption must be (a) unstated, (b) necessary to bridge statement → intent, (c) reasonable — not extreme (“all villagers want lamps” fails; “lamps would be used by villagers” passes).
• Assumption vs inference. Assumptions come before the statement (unstated premises); inferences come after (what follows from it). Phrasing tricks flip this direction constantly.

5. Course of Action: The Administrative Filter

• The format. A problem situation (“rising cases of food adulteration reported”) with candidate actions (“increase laboratory testing capacity”, “ban all packaged food”).
• The filter (three tests).
  1. Addresses the cause, not the symptom.
  2. Feasible and proportionate — administrative measures survive; bans on entire sectors, mass dismissals, instant criminalisation fail.
  3. Immediate or long-term tagged correctly — pair the question’s asked horizon (“immediate course”) with the option’s horizon.
• The logic source. Identical to CSAT decision-making (Part 7 of this series): procedural, balanced, solution-first. Prepare the two blocks together and one skill serves both.

6. Inference, Conclusion and Punch-Line Types

• Inference from data statements. Two statements with numbers (“literacy rose 8%; female literacy rose 12%”) — the valid inference is modest (“female literacy grew faster”), never causal (“female literacy caused the rise”).
• Cause-effect questions. If both events are described as simultaneous without a stated link, “A caused B” fails; “the events may be related” passes.
• Logical conclusion (“what best follows”). The correct option restates the argument’s endpoint with controlled scope; extreme or out-of-scope conclusions are the engineered wrong answers.

7. How Exams Probe This Topic

• The practice bank: CSAT 2014–2025 reasoning questions plus CAPF and CDS papers (same style, larger volume) — build 200+ solved items with the method named beside each (Venn / negation / filter).
• The drill standard: 20 questions in 18 minutes at 90% accuracy; any question over 90 seconds is a “name the algorithm faster” signal, not a difficulty signal.
• The error log: tag every miss by algorithm and by trap (converse fallacy, either/or missed, extreme inference) — the trap profile, not the count, is what improves.

8. Quick Revision: One-Glance Facts

• Syllogisms. Minimal Venn; true only if it survives every diagram; “either/or” for exhaustive complements; Some+Some ⇒ nothing.
• Conversions. All A are B ⇒ Some B are A; No and Some are symmetric; All is not.
• Assumptions. Negate-and-collapse test; must be necessary, unstated, reasonable.
• Course of action. Cause-addressing, feasible, horizon-matched.
• Drill bar. 20 in 18 at 90%; log traps by name.

Conclusion. Reasoning marks are algorithm marks. The Venn method, the conversion table, the negation test and the administrative filter convert this block from “feels tricky” to “runs on rails” — and combined with Part 7’s shared logic, they form the most reliably trainable 30-odd marks in the CSAT paper.

Fifteen Rapid-Drill Items (The Whole Section’s Grammar)

Syllogisms (1-6). All P are Q; some Q are R → “some P are R”? — no definite conclusion. No P is Q; all Q are R → “no P is R”? — no (the Q-and-R overlap is the unknown). All A are B; all B are C → “all A are C” — yes (the chain). Some M are N; all N are O → “some M are O” — yes. “No cats are dogs; some dogs are black” → “some black things are not cats”? — yes, definite (the black-dogs are outside-cats). Either-or-check: “some P are R” and “no P is R” with-no-definite — either-or follows (the complements-exhaust).
Assumptions (7-10). “Drink milk for strong bones” — assumes milk-builds-bones (negate: the statement-collapses — yes); assumes people-want-strong-bones (yes — the persuasion’s-premise); assumes milk-is-cheap (no — the statement-survives-its-negation). “Book-train-tickets-online to save time” — assumes online-is-faster (yes) and users-have-access (no — the access-gap-is-not-the-statement’s-bridge).
Course-of-action (11-13). “Rising road-accidents at night” → “install-reflectors-and-audit-lighting” (cause-addressed, feasible — follows); “ban night-driving” (disproportionate — doesn’t); “Rising onion-prices” → “release-buffer-stock-and-crack-hoarding” (follows); “import-onions-immediately-and-permanently” (the symptom-and-the overreaction — doesn’t).
Cause-effect (14-15). “Sales-rose after the rebrand” → the rebrand-caused-the-rise? — no (the concurrent-unexcluded); “Literacy-rose as electricity-reached the villages” → causation? — probably-related, unproven (the exam’s-honest-answer).
Fifteen-items, the four-algorithms — run-this-bank weekly; the section’s-questions-are-these-fifteen-in-new-clothes.

The Venn-Drawing Discipline (The Two-Minute-Investment)

For any 3-statement-syllogism, draw-both-minimal-and-maximal-diagrams before the conclusions: the minimal (the least-overlap-consistent) and the maximal (the most). A-conclusion-holding in-both is-definite; holding-in-one-only is-uncertain; failing-in-both is-false. The-discipline’s-payoff: the mind’s-eye-Venn is where syllogism-errors-live — the paper-drawn-pair-eliminates-them, and the 60-seconds-spent-drawing-saves-the 90-seconds-of-doubting.