How to read a judgment: ratio vs obiter in 10 minutes

Ask why a junior's research note went wrong and the answer is usually not a missed case - it is a misread one: a quotable sentence lifted from paragraph 40 of a judgment whose actual decision turned on something else entirely. Judgments are not statutes; they are not law from first page to last. Only the ratio decidendi - the principle the case actually decides - binds later courts, and everything else, however eloquent, is obiter dicta. This guide gives you a working ten-minute method for telling them apart, anchored in the Supreme Court authorities that define the exercise. It is written for Indian practice - where Article 141 gives the distinction constitutional stakes - but the method is common-law universal. (Legal information for professional skill-building, not legal advice.)

Why most people misread judgments

The classic warning is Lord Halsbury's, in Quinn v Leathem [1901] AC 495: every judgment "must be read as applicable to the particular facts proved", and "a case is only an authority for what it actually decides" - not "for a proposition that may seem to follow logically from it". The Supreme Court of India adopted that passage almost verbatim in State of Orissa v Sudhansu Sekhar Misra, AIR 1968 SC 647: what matters in a decision "is its ratio and not every observation found therein nor what logically follows from the various observations".

The reason misreading persists is that judgments are written as continuous prose: the binding principle, the supporting reasoning, the responses to counsel's wider arguments, and the judge's reflections sit in the same register, unlabelled. Extracting the ratio is genuinely an act of analysis - which is why it rewards a method rather than a highlighter.

What the ratio actually is - the three postulates

The most operational Indian definition is in Union of India v Dhanwanti Devi, (1996) 6 SCC 44. Every decision, the Court said, contains three postulates: (i) findings of material facts, direct and inferential; (ii) statements of the principles of law applicable to the legal problems disclosed by the facts; and (iii) the judgment based on the combined effect of the two. The ratio is the second element as connected to the third: the principle of law actually necessary for the result reached on those facts. "It is not everything said by a Judge while giving judgment that constitutes a precedent."

Article 141 of the Constitution - "the law declared by the Supreme Court shall be binding on all courts within the territory of India" - attaches to exactly that. In Director of Settlements, A.P. v M.R. Apparao, (2002) 4 SCC 638, the Court spelled out the consequences: what binds is "the ratio of the decision and not any finding of facts"; the principle emerges from "a reading of a judgment as a whole"; and a point disposed of on concession is not law declared at all. A Constitution Bench applied the same discipline in Krishena Kumar v Union of India, (1990) 4 SCC 207, refusing to extend D.S. Nakara beyond what it necessarily decided: the doctrine of precedent "is limited to the decision itself and as to what is necessarily involved in it".

What obiter is - and what it is worth

Obiter dicta are everything in category (ii) that the result did not require: commentary on questions not raised, hypothetical variations, views on arguments the court expressly left open. M.R. Apparao gives the fair valuation - Supreme Court obiter is "not binding" but carries "considerable weight". In practice, obiter is tomorrow's argument: minority views and asides have a long history of becoming later ratios. The professional discipline is not to ignore obiter but to price it honestly - cite it as persuasive, never dress it as binding, because opposing counsel (or the bench) will re-price it for you if you do.

The ten-minute method

A repeatable sequence for a first serious read of any judgment:

1. Minute 1 - the metadata: court, bench strength, date, and what order was actually passed (allowed? dismissed? remanded?). Bench strength decides conflicts: larger benches bind smaller ones.
2. Minutes 2-3 - the dispute: identify the concrete question the court had to answer to dispose of the case. Not the themes - the question. Write it as one sentence.
3. Minutes 4-5 - the material facts: list the facts the court treated as decision-driving (Dhanwanti Devi's first postulate). A fact recited but never used is not material.
4. Minutes 6-8 - the principle: find the passage where the court states the rule it applies to those facts to produce the order. That candidate passage is your provisional ratio.
5. Minute 9 - the inversion test: ask whether the decision would have gone the other way if the candidate principle were reversed. If the result survives the reversal, the passage was not necessary - it is obiter, not ratio.
6. Minute 10 - the boundary note: record what the case does not decide - concessions, questions left open, facts distinguished. This is the note that saves you in court when the other side over-reads the case.

The inversion test in step 5 is the workhorse. It operationalises "necessarily involved" from Krishena Kumar: a proposition whose negation changes nothing was never load-bearing.

The traps: headnotes, wide propositions, and facts ignored

Three failure modes account for most citation embarrassments. Headnote-reading: headnotes are an editor's summary, not the court's holding - they compress away the facts that confine the ratio, and they are where "a case is only an authority for what it actually decides" goes to die. Wide propositions: judgments often state principles more broadly than the dispute required; Krishena Kumar is the standing authority that the wide statement does not travel - only the principle as applied does. Fact-blindness: citing a ratio to a court without checking whether your material facts match is not citation, it is hope. The comparison of material facts is the whole game of distinguishing - and it is why step 3 of the method exists.

When a precedent does not bind: per incuriam and sub silentio

Two doctrines let a later court decline to follow an otherwise binding decision - both narrow, both requiring express justification. Per incuriam: a decision rendered in ignorance of a governing statute or binding precedent loses precedential force. The dramatic authority is A.R. Antulay v R.S. Nayak, (1988) 2 SCC 602, where a seven-judge bench recalled the Court's own earlier directions as per incuriam - given in ignorance of the exclusive-jurisdiction provision of the 1952 Act. Sub silentio: a point that passed through a judgment without argument or consideration decides nothing. Municipal Corporation of Delhi v Gurnam Kaur, (1989) 1 SCC 101, is the standard citation: "precedents sub silentio and without argument are of no moment", and a consent-based direction in an earlier case created no binding rule.

Use both doctrines the way courts do - as scalpels, not shields. Arguing per incuriam requires showing the specific provision or authority the earlier bench never saw; it is not a synonym for "wrongly decided".

The authorities to keep at hand

Every one of these is freely readable on Indian Kanoon or the Supreme Court's Digital SCR - our companion guide on citing Indian case law correctly covers the citation systems themselves.

Reading at scale: where AI helps and where it cannot

The method above is per-judgment; a live matter involves dozens. This is where AI research assistance changes the economics without changing the analysis: citation-grounded research surfaces the candidate authorities with pinpoint passages and links to the full text, and treatment signals (followed, distinguished, overruled) flag where a case's authority has moved - the mechanical layer our guide to researching Indian case law with AI walks through. What the tools do not do is the inversion test: deciding what was necessary to a decision is legal reasoning on the material facts, and it stays with the reader. The honest division of labour: AI finds, links, and verifies; the lawyer reads for ratio. Ten minutes per judgment, spent where they count - and if you want the finding-and-verifying half compressed, Judicio's free 7-day trial (500 credits, no card) is the fastest way to test it on your current matter.