Reading law

English law is made in two places: Parliament, which passes statutes, and the courts, which decide cases. To read either kind of law you first need the map of the courts, because a court's place in the hierarchy determines whose decisions it must follow and who must follow its own.

At the bottom sit the magistrates' courts, which handle the great bulk of criminal work — minor offences, first appearances, sending serious cases upwards — and the County Court, which handles most civil disputes: debts, contracts, personal injury below certain values. Neither creates precedent. Nobody is bound by what a magistrates' court thinks.

Serious criminal trials happen in the Crown Court, before a judge and, if the defendant pleads not guilty, a jury. Serious or complex civil cases start in the High Court, which is divided into three divisions: King's Bench (most general civil claims, and judicial review through its Administrative Court), Chancery (property, trusts, companies), and Family. High Court decisions bind the courts below, and while one High Court judge is not strictly bound by another, they will normally follow each other for consistency.

Appeals go to the Court of Appeal, split into a Civil Division and a Criminal Division. It binds every court below it and, with narrow exceptions, itself. Most of the cases on your reading lists were decided here, because it decides far more appeals than the court above it.

At the top is the Supreme Court, which in October 2009 took over from the Appellate Committee of the House of Lords — so when an older case says "the House of Lords held", read "the top court held". It binds every other court. It does not bind itself: since a practice statement of 1966 the top court has been free to depart from its own past decisions when it appears right to do so, a power it uses sparingly. You will meet cases this year where it did.

Two more courts confuse everyone at first. The Judicial Committee of the Privy Council is staffed largely by the same Justices as the Supreme Court but is not part of the domestic hierarchy: it is the final court of appeal for some Commonwealth states and British overseas territories. Its decisions do not, as a rule, bind English courts, but because of who sits on it they are treated as highly persuasive.

Then there are two European courts, and they belong to two entirely different systems. The European Court of Human Rights in Strasbourg is the court of the Council of Europe — a body of forty-six states that has nothing to do with the European Union — and it rules on the European Convention on Human Rights. The United Kingdom is still a party to the Convention; leaving the EU changed nothing about that. Under the Human Rights Act 1998, UK courts must take Strasbourg case law into account, but they are not strictly bound by it. The Court of Justice of the European Union in Luxembourg is the EU's court. Since Brexit its new decisions no longer bind UK courts, though its older case law still matters in ways your European Law module will unpick. For now, hold on to the one distinction that matters: Strasbourg is human rights; Luxembourg is the EU. Mixing them up is the single most reliable way to signal that you have not understood the system.

The common law works on a simple bargain: like cases are decided alike. A court must follow the legal rules laid down by the courts above it, and usually by its own past decisions. That is the doctrine of precedent, and the whole skill of reading cases is working out exactly what a past decision laid down — because only part of a judgment binds anyone.

The binding part is the ratio decidendi — the rule of law the court needed in order to decide the case as it did on its material facts. Everything else a judge says is obiter dicta — remarks by the way. A judge musing about how a different set of facts might come out, a wider observation about where the law ought to go, an entire dissenting judgment: all obiter. Obiter can be enormously influential, particularly from senior courts, but no court is ever obliged to follow it.

Textbooks make identifying the ratio sound mechanical. It is not, and it is worth understanding early why not. First, appellate courts sit in panels, and each judge may give a separate judgment. When five judges agree on the result but reason differently, what exactly did the court decide? Brown, which you will meet in Criminal Law, upheld convictions by three votes to two with the majority reasoning in different ways; lawyers have argued about its precise ratio ever since. Second, the ratio depends on which facts you treat as material, and facts can be described at different levels of generality — is the rule about "contracts made by post", or "acceptances communicated through a third party", or "all acceptances"? The later court choosing how widely or narrowly to read the earlier case is doing real legal work. That is why reasonable people — including judges — genuinely disagree about what old cases decided, and why "what is the ratio of X?" is a live question and not a memory test.

Because precedent binds, the law has developed pressure valves:

• Distinguishing. A court persuaded that the present facts differ in some material way from the precedent's facts may decline to apply it. This is the everyday tool of advocacy: most legal argument is a fight about whether the awkward case is "really" like this one.
• Overruling. A higher court may declare a lower court's precedent — or, in the Supreme Court's case, its own — wrongly decided, killing it for the future. Do not confuse this with reversing, which is what an appeal court does to the decision in the very case before it. A case is reversed on appeal; a different, older case is overruled.
• Per incuriam. A precedent decided in ignorance of some binding statute or authority that would have compelled a different result need not be followed — but only by the court that made it or one at the same level. It is, classically, one of the Court of Appeal's narrow escapes from its own past decisions, not a licence for a lower court to ignore a higher one. The escape hatch is deliberately narrow — it is for decisions that are demonstrably broken, not merely disliked.

Finally, keep the vocabulary of weight straight: decisions from above — and, for most appellate courts, their own past decisions — are binding; obiter remarks, Privy Council decisions, courts in Australia or Canada, dissents are at most persuasive. A persuasive authority is a reason; a binding authority is an instruction.

A case reaches you as a law report, and the first thing to decode is its name and citation. The name gives the parties: Smith v Jones. The "v" is written but never spoken — in civil cases lawyers say "Smith and Jones", in criminal cases "the Crown against Smith". One more label shifted under your feet: the person suing was "the plaintiff" until 1999, and is "the claimant" now — the reports on your reading list use whichever word their decade did. Criminal cases are brought by the Crown, abbreviated R for Rex or Regina depending on who reigns — the older reports on your lists were decided under a queen, so they say R v Cunningham with R meaning Regina.

Citations come in two forms, and telling them apart matters. Since the early 2000s every judgment of the senior courts gets a neutral citation issued by the court itself: Jogee is [2016] UKSC 8, meaning the eighth judgment of the Supreme Court in 2016. Neutral citations point to numbered paragraphs — you cite [42], not a page — and belong to the judgment itself, not to any publisher. A report series citation, by contrast, points into a printed series: [1893] 1 QB 256 means volume 1 of the Queen's Bench reports for 1893, page 256. Not all series are equal. The Law Reports proper — AC (Appeal Cases), QB or KB, Ch, Fam, published by the Incorporated Council of Law Reporting — are the ones courts require to be cited where available, because the judges themselves revise the text before publication. The ICLR's Weekly Law Reports and the All England Law Reports sit on the next rung. One judgment can therefore carry several citations pointing at the same words in different places.

Open a report and before the judgment proper you meet the headnote: catchwords, a summary of the facts, and a statement of what the court "held". Understand exactly what this is. A law reporter wrote it, not the court. It is usually accurate, it is genuinely useful, and it is not the law — where the headnote and the judgment diverge, the judgment wins, and examiners enjoy the difference. Below it you will find the judges' names, counsel, the cases cited, and then the judgment or judgments.

Plural, because each member of an appellate panel may deliver one. Some concur in a sentence; some concur through separate reasoning; some dissent. A dissent binds nobody, but great dissents have a habit of becoming the law a generation later, which is one reason they are worth reading. At the very end comes the disposal — appeal allowed or dismissed, the orders made. Always find it. It is surprisingly easy to read thirty pages of elegant reasoning and lose track of who actually won.

A Supreme Court judgment can run past a hundred pages. Nobody — not your lecturers, not the barristers arguing the next case — reads every word of every judgment at the same depth, and trying to is how first-years drown in week three. Read with a search pattern instead:

1. Find the issue first. What question did the court actually have to answer? Judgments usually state it within the first few paragraphs ("the question in this appeal is whether…"). Until you can say the issue in one sentence, nothing else in the case will organise itself.
2. Find the ratio second. Where does the court answer that question, and what rule does the answer rest on? Work from the issue to the reasoning that resolves it. In multi-judgment cases, start with the lead judgment — the one the others say they agree with.
3. Read the headnote last — as a check, not a shortcut. If your one-sentence issue and your statement of the rule broadly match the headnote, you have read the case correctly. If they do not, one of you is wrong, and finding out which is where the real learning happens. Read the headnote first instead and it will quietly decide for you what the case says, which is precisely the skill you are supposed to be building yourself.

Around that spine, skim. The recitation of facts matters at the start (facts drive materiality, and materiality drives the ratio); long passages surveying the authorities can be skimmed on first pass; concurrences and dissents matter when the case is contested enough for your module to care. You can always go back. The aim on first reading is a working grasp of issue, rule and result — the case note in section 6 gives that a shape.

The other thing you will read constantly is legislation. An Act of Parliament arrives divided into numbered sections, subdivided into subsections, then paragraphs: "s 2(1)(a)" is section 2, subsection 1, paragraph a — and you write "s" for section, "ss" for sections. Long Acts group sections into Parts. At the back sit the Schedules, which carry detail, lists, and transitional machinery; a schedule only has force because some section gives it force. Somewhere near the end you will usually find an interpretation section, where the Act defines its own terms — always check, because an Act is entitled to make a word mean something narrower or stranger than English does.

Reading the words is one thing; deciding what they cover at the edges is another, and the courts have developed recognisable approaches. The traditional trio, which you should know as history and vocabulary: the literal rule takes the words in their ordinary meaning even when the outcome looks odd — famously, a flick knife displayed in a shop window was held in Fisher v Bell not to be "offered for sale", because in contract law a window display is an invitation to treat, not an offer. The golden rule lets a court bend an ordinary meaning to avoid absurdity — in Adler v George, "in the vicinity of" a prohibited place was read to cover someone actually inside it, since Parliament can hardly have meant to catch the man at the fence but not the man past it. The mischief rule, as old as Heydon's Case, asks what defect in the old law the Act was passed to cure and reads it to cure that mischief.

The modern reality is that courts read statutes purposively: words in their context, in the light of the purpose the Act discloses. The old rules survive as habits of argument within that approach rather than as competing masters. One refinement worth knowing precisely: under Pepper v Hart a court construing an ambiguous or absurd provision may look at Hansard — the record of parliamentary debates — but only for clear statements by the minister or promoter of the Bill, and only through that narrow gateway. It is an exception with conditions, not a licence to mine speeches for helpful quotes.

Public Law will complicate this picture productively: section 3 of the Human Rights Act requires courts to read legislation compatibly with Convention rights "so far as it is possible to do so", which is interpretation under instruction and stronger medicine than anything above. Park that for now; know that it exists.

Two practical habits will carry you through the year. The first is the case note. For every case that matters in a module, write five short entries, in your own words, on one card or half a screen:

• Facts — two sentences, only the material ones.
• Issue — the question the court had to answer, as a question.
• Holding — the answer, with the disposal.
• Reasoning — why: the rule and the argument that carried it.
• Why it matters — what this case is for in your module: what it settled, changed, or left open.

The fifth entry is the one that separates useful notes from transcription. A hundred case names with "why it matters" attached beats three hundred summaries you never reread. The subject maps on this site model the format: every case in them carries a one-line significance, which is the fifth entry done for you — the other four are your job.

The second habit is knowing where law lives. Westlaw and LexisNexis are the subscription databases King's gives you — full reports, current versions of statutes, journals; learn one of them properly in the first fortnight. BAILII is free and public, with judgments back decades — invaluable, though without the editorial apparatus. The ICLR publishes the official Law Reports and its site tells you the authoritative citation for a case. When a problem question cites a statute, read the actual current text on Westlaw or Lexis rather than trusting a textbook's paraphrase from two years ago — amendment is constant, and "the section says" is only ever true of the section as it stands.