Inference vs Stated Fact: The Boundary
A line is drawn in every Reading Comprehension passage — invisible, but very strict — between what the passage says and what the passage allows you to conclude. SBI PO setters know aspirants stumble on that line, so they design inference traps that look reasonable but cross it. Learning exactly where the line sits is one of the highest-return skills for the English section.
Definition: A stated fact is information explicitly written in the passage. You can underline it and quote it.
Definition: An inference is a logical conclusion the passage supports but does not state outright — "reading between the lines" within strict limits, using only the passage's own logic and no outside knowledge.
The two big traps: over-inference and under-inference
SBI typically attacks from two opposite directions in the same question.
Over-inference is the option that goes too far. The passage hints; the option asserts. The passage says "many" or "often"; the option says "always" or "must". Whenever you see absolute words — all, none, never, always, must, only, every, certainly — your alarm should ring. A correct inference is almost never absolute, because passages are almost never absolute.
Under-inference is the option that doesn't go far enough — it merely rephrases a sentence from the passage. That is a detail, not an inference. If the question explicitly asks "Which of the following can be inferred?", an option that simply paraphrases a line is wrong even if it is true, because it isn't an inference.
The correct inference sits in a narrow middle band: it is defensible from passage logic, it is not stated outright, and it is the most modest statement that still answers the question.
The golden test — "Could the author disagree with this?"
This is the single most powerful filter in your toolkit.
Take an option. Ask yourself: based on what the passage actually said, could the author of the passage push back on this statement? If the answer is "yes, plausibly," the option is wrong. If the answer is "no, the author is logically committed to this," the option is a defensible inference.
This test works because inference questions are not about your opinion or general knowledge — they are about what the author has, perhaps without saying so, already endorsed. If the author could reasonably disagree, the inference is not defensible.
Three more elimination rules
Beyond the golden test, four operational rules will kill most wrong options quickly:
- Reject extreme/absolute language unless the passage itself is equally absolute. If the passage says "researchers tend to find," an option saying "researchers always find" is over-inference.
- Reject options that introduce new topics not discussed in the passage. If a passage discusses inflation in India and an option talks about Indian exports, exports are out of scope — eliminate.
- Reject options that require outside knowledge. Even if you happen to know the option is true in the real world, if the passage doesn't support it, it is wrong on the test.
- Reject options that contradict the tone or stance of the author. A pro-author option in an anti-passage is wrong by default.
After applying these rules, you will typically be left with just one or two options. The safer of the two — the one that says less — is usually the correct inference.
A worked example
Question: Consider this passage extract:
"While digital banking has expanded financial access in urban India, rural penetration remains limited by patchy connectivity and low financial literacy. Banks have begun deploying business correspondents, but the model has not yet been profitable at scale."
Which of the following can be inferred?
(A) Digital banking has completely solved India's financial inclusion problem.
(B) Rural India will never adopt digital banking.
(C) Expanding rural digital banking faces structural challenges that banks have yet to fully overcome.
(D) Business correspondents are a failed model.
(E) Indian banks should stop investing in rural areas.
Solution:
Step 1: Apply the absolute-words filter. (A) says "completely solved" — eliminate. (B) says "never adopt" — eliminate. (D) says "failed model" — too strong; the passage says only "not yet profitable at scale" — eliminate. (E) prescribes action and contradicts the passage's neutral tone — eliminate.
Step 2: Test the remaining option (C) with the golden test. The passage explicitly says rural penetration is limited by connectivity and literacy (structural challenges) and that the business-correspondent model is not yet profitable (challenge not yet overcome). The author is logically committed to (C). The author cannot plausibly disagree with it.
Step 3: Confirm (C) is modest, defensible, and a true inference (not a restatement of any single line).
Conclusion: (C) is the correct inference.
Why it matters
Why it matters: SBI PO's English section gives 6–10 marks per RC passage, and 2–3 of those are inference questions. If you reliably distinguish inference from over-inference, you convert these from "50–50 guesses" into near-certain marks. Cracking inference also lifts your accuracy in Tone, Assumption, and Conclusion questions, since they all rely on the same logic.
Real-world example: News editorials rely on inference all the time. When The Hindu writes that "the government's growth projections remain optimistic compared with consensus estimates," it does not directly say the projections are unrealistic — but the inference is clearly available. Inference questions are designed to test this everyday skill of reading carefully under pressure.
Common misconception: "If an option is true in real life, it must be a correct inference." Wrong. An option is correct only if the passage supports it. A factually accurate statement that the passage doesn't argue for, or actively avoids arguing for, is a trap. SBI explicitly designs distractors that are real-world true but passage-unsupported.
| Type of option | What it does | Verdict |
|---|---|---|
| Direct restatement of a line | Just paraphrases a fact | Wrong — it's a detail, not an inference |
| Modest, passage-supported claim | Reads between the lines using author's logic | Correct inference |
| Absolute claim (always/never/all) | Goes beyond what passage allows | Over-inference — wrong |
| New-topic claim | Introduces ideas the passage didn't discuss | Out of scope — wrong |
| Real-world-true claim | True outside the passage but not argued | Trap — wrong |
| Author-contradicting claim | Clashes with the author's stance | Wrong |
- ✓- A stated fact is in the passage; an inference is supported by it but not stated.
- ✓- Apply the "Could the author disagree?" test — if yes, eliminate.
- ✓- Reject absolute words: always, never, all, none, only, must, certainly — unless the passage is equally absolute.
- ✓- Reject options that introduce new topics or require outside knowledge.
- ✓- A direct restatement is a detail, not an inference.
- ✓- The correct inference is the most modest option that still answers the question.
- ✓- Real-world truth is not enough — only passage support matters.
- ✓- Tone of the answer should match the tone of the passage.
The CADE filter — when you see an option, check:
- Could the author disagree? (If yes, drop it.)
- Absolute words? (Drop unless passage is absolute.)
- Different topic? (Drop if out of scope.)
- External knowledge needed? (Drop.)
One-liner: "The safest inference is the smallest step beyond the text."
- ✓- Inference = supported, not stated. Detail = stated. Over-inference = stated too strongly.
- ✓- Use the "Could the author disagree?" test as your primary filter.
- ✓- Eliminate absolutes, new topics, outside-knowledge claims, and tone-contradictions.
- ✓- When in doubt, pick the most modest option that still answers the question.
Tone Vocabulary Toolkit
Tone = author's attitude. SBI options use precise tone words you must know: CRITICAL/CENSORIOUS (disapproving), LAUDATORY/EULOGISTIC (high praise), SARDONIC/SARCASTIC/IRONIC (mocking), OPTIMISTIC vs PESSIMISTIC, OBJECTIVE/NEUTRAL/DISPASSIONATE (no opinion), CAUTIOUS/CIRCUMSPECT, ANALYTICAL (examining), CYNICAL (distrustful of motives), AMBIVALENT (mixed), CONDESCENDING (patronising), NOSTALGIC (longing for past), POLEMICAL (strongly argumentative). Speed trick: scan for EVALUATIVE adjectives/adverbs and verbs ('regrettably', 'cleverly', 'ignores', 'fails to'). A factual data-heavy passage with no opinion words = OBJECTIVE. Beware: 'ambivalent' fits only when BOTH positive and negative signals appear.
Worked Example: Spotting Over-Inference
Passage: 'India's solar capacity grew fivefold in five years, though it still meets under 5% of total demand.' Question: What can be inferred? Trap option: 'India will soon meet all its energy needs through solar.' This is OVER-inference—'fivefold growth' plus 'under 5%' does NOT support 'all needs soon'; the author even flags the small base. Correct inference: 'Solar growth has been rapid but remains a minor share of total energy supply.' This stays within the data. Lesson: when a passage gives BOTH a positive (fivefold) and a limiting fact (under 5%), the correct inference usually balances both. Reject any option ignoring the qualifier 'though'.
Inference & Tone Questions — Flashcards
Cover the answer, recall, then check. 11 cards on inference, tone and attitude questions in SBI PO reading comprehension.
Q1. What is an inference question really asking?
A1. What must be TRUE based on the passage but is NOT stated directly — one logical step beyond the text, never a leap.
Q2. How far can a valid inference go?
A2. Only as far as the passage guarantees; if you need an extra assumption the passage does not support, it is not a valid inference.
Q3. How do you detect the author's tone?
A3. From word choice (connotation), adjectives, and any praise or criticism; neutral reporting versus loaded words reveals the attitude.
Q4. Name common tone options and their cues.
A4. Optimistic (hopeful words), critical (fault-finding), objective/neutral (factual, no bias), sarcastic/ironic (mocking praise), cautious (hedged "may/might").
Q5. Difference between tone and the author's attitude?
A5. The question usually wants the author's attitude to the subject, so focus on how the author judges the topic, not on the mood it evokes in you.
Q6. Why is "objective/neutral" often correct in banking passages?
A6. Because report-style economic passages usually present facts without emotion — pick it when there are no clearly loaded words.
Q7. What eliminates an emotional tone option?
A7. The absence of emotional or loaded language; if the passage merely states data, "euphoric", "angry" or "pessimistic" are too strong.
Q8. How do you answer "the passage implies that…"?
A8. Find the sentence the implication rests on and choose the option that follows from it directly, without adding new information.
Q9. What is the classic trap in inference options?
A9. Over-reach — an option that is a bigger claim than the passage licenses, or that brings in outside cause and effect.
Q10. How do you judge "the author's primary purpose"?
A10. Ask why the passage was written — to inform, argue, criticise, describe or persuade — and match the verb in the option to that intent.
Q11. Quick test for a candidate inference?
A11. Ask "Does the passage force this to be true?" If you can imagine the passage true while the option is false, reject it.