Where laws come from
You obeyed about forty laws before breakfast and didn't notice one of them.
Which side of the road you drove on. The fact that the coffee you bought was a contract, complete with implied terms you never read. That the building you're sitting in had to satisfy a fire code, that the money in your pocket is legal tender because a statute says so, that nobody could enter your flat this morning without a specific legal reason. None of it felt like law, because law mostly doesn't feel like anything. It feels like weather.
So where does the weather come from? There are four honest answers — judges, legislatures, constitutions and custom — and they're mixed in different proportions depending on which country you happen to be standing in. This is a tour of the machinery, not an argument about what any particular rule should say.
There are two big family trees
Almost every legal system on earth is descended from one of two traditions, or is a blend of both.
Civil law starts with Rome. The emperor Justinian had the accumulated mess of Roman legal writing compiled into an ordered body of texts in the 530s, and that compilation — later called the Corpus Juris Civilis — was rediscovered and studied intensively at Bologna from the late eleventh century onwards. The idea it carries is that law should be written down in a comprehensive, systematic code, arranged by subject, from which a judge reasons downwards to the case in front of them. France produced the model modern version in 1804 with the Napoleonic Code; Germany produced a more academic one, the Bürgerliches Gesetzbuch, in force from 1900. Between them those two texts were copied, adapted and imposed across most of Europe, Latin America, and large parts of Asia and Africa. Japan built its modern system largely on the German model in the late nineteenth century. Turkey adopted a version of the Swiss civil code in 1926.
Common law starts with English kings wanting control. After 1066 the Crown pushed royal justice out across a country full of local and manorial courts, sending judges on circuit and gradually building a body of decisions that was "common" to the whole realm rather than varying village by village. There was no code. There was a growing pile of decided cases, and lawyers argued by analogy from them. Because Britain later ran an empire, that method ended up in the United States, Canada, Australia, New Zealand, India, Pakistan, Nigeria, Kenya, the Caribbean and elsewhere.
Then there are the hybrids, and there are more of them than the neat two-family story suggests. Scotland has always been mixed, with a strong Roman-law inheritance sitting alongside English-style precedent. South Africa runs Roman-Dutch law with a common-law overlay. Louisiana and Quebec are civil-law islands in common-law countries. Plenty of states run religious or customary law alongside a state system, often confined to particular subjects such as family or inheritance, and often with disputes about where the boundary sits.
The difference between the two families is easy to overstate, and most people do. Modern common-law countries are drowning in statutes, and modern civil-law countries pay close attention to what their courts have already decided. But the instinct differs, and it shows up in what a lawyer reaches for first: the code, or the case.
What precedent actually means
People say "judges make law" and then argue about whether they should. The mechanics are less dramatic than the slogan.
In a common-law system, a decided case binds later courts — but only partly. What binds is the ratio decidendi, the reasoning strictly necessary to the outcome. Everything else the judge said, however interesting, is obiter dicta: persuasive, quotable, not binding. Working out which is which in a long judgment is a genuine professional skill, and lawyers don't agree about it even slightly.
Precedent also runs downhill. A higher court binds the courts below it; a decision from a court in another country binds nobody but might be cited as persuasive. And there are two escape hatches. A judge can distinguish a precedent, saying the facts here differ in a way that matters, which is how the law bends without breaking. Or a sufficiently senior court can overrule it outright.
That second one was startlingly recent in England. The House of Lords, then the highest court, held itself absolutely bound by its own past decisions until it issued a Practice Statement in 1966 saying it would depart from them when it seemed right to do so. Its successor, the UK Supreme Court, has held that power since 2009.
Civil-law systems formally have no doctrine of binding precedent at all. A judge applies the code to the case. In practice, a consistent line of decisions from the senior courts carries enormous weight — French lawyers call it jurisprudence constante — and no sensible advocate ignores it. You won't win by pretending it isn't there. So the two systems converge from opposite directions: common law admits it's building rules out of cases and tries to constrain the process, civil law denies it's doing so and does something similar anyway.
Statutes, and the much bigger pile underneath them
Legislation is the part everyone can picture, and the picture is usually of the wrong stage.
The general sequence is much the same across parliamentary systems. Somebody proposes — usually the government, occasionally an individual member, sometimes a citizens' initiative. Specialists turn the policy into legal text, which is a craft in itself; get the drafting wrong and you'll get twenty years of litigation over a comma. The text goes through readings and debates, a committee stage where it's picked over line by line and amended, votes at each stage, and then the same again in a second chamber if there is one. At the end the head of state signs or promulgates it. And then, often overlooked, it has to be commenced — brought into force, sometimes in pieces, sometimes years later, sometimes never.
Presidential systems differ mainly at the end and at the start. The executive isn't drawn from the legislature and can't rely on a majority, so bills originate more often with individual legislators, and the president typically has a veto that the legislature can override with a supermajority. Bicameral systems give a second chamber some power to delay or block; unicameral ones don't have the question.
Here's the part that surprises people. Measured by volume, most legal rules aren't made by legislatures at all. Statutes routinely grant a minister, department or agency the power to fill in the details by regulation, and those details are where the operative law usually lives — the actual permitted limit, the actual form you must file, the actual list of controlled substances. This is called delegated or secondary legislation, and countries produce it at several times the rate of primary statutes.
The reason is capacity. No legislature has the time or the expertise to set every technical threshold, and updating a number shouldn't require a fresh act of parliament. That isn't a conspiracy; it's arithmetic about hours in the day. The trade-off is scrutiny: delegated rules typically get far less debate, and how much oversight they should receive is a live constitutional argument in a great many countries, on which people across the political spectrum have landed in different places.
Constitutions, and who gets to say what they mean
A constitution is the law about how law gets made. It sets out the institutions, allocates power between them, and usually protects some rights against ordinary majorities.
Most are codified in a single document. A few famously aren't: the United Kingdom, New Zealand and Israel operate from a mixture of statutes, court decisions, conventions and practice, which is uncodified rather than unwritten — most of it is written down, just not in one place, and it can be changed by ordinary legislation.
Codified constitutions are normally entrenched, meaning they're harder to amend than an ordinary law. The American one is at the demanding end: an amendment needs two-thirds of both chambers of Congress and ratification by three-quarters of the states, which is why there have been twenty-seven in more than two centuries, ten of them arriving together in 1791. Other countries set the bar lower and amend far more often. Neither approach is obviously right. A rigid constitution is stable, and it's also stuck. A flexible one adapts, and it can also be rewritten by whoever happens to be in charge that year.
Then the hard question: if the constitution is supreme, who decides what it says? Two models dominate. In the diffuse or American model, any court can rule that a law conflicts with the constitution and decline to apply it — a power the US Supreme Court asserted for itself in Marbury v Madison in 1803. In the concentrated or European model, designed largely by the Austrian jurist Hans Kelsen for the constitutional court established in 1920, ordinary courts can't do this and a single specialised constitutional court handles it. Germany, Italy, Spain, South Korea and many others followed that pattern.
Whether unelected judges should be able to strike down laws passed by elected legislatures is one of the oldest arguments in the subject, and it doesn't map neatly onto left and right — the answer people give tends to depend on what their courts and their legislatures have recently been doing.
The law nobody voted for
Plenty of binding rules never pass through a parliament.
Custom is the oldest source, and it isn't extinct. International law still recognises customary rules that arise from consistent state practice combined with a belief that the practice is legally required. Many countries recognise customary land tenure. Merchants developed their own body of trade practice long before states codified it, and modern commercial arbitration still leans on standard-form contracts, trade terms and industry usage that no legislature wrote.
Treaties add another layer, and how they land inside a country varies. In monist systems a ratified treaty becomes part of domestic law more or less automatically. In dualist systems — the UK is the standard example — it binds the state internationally but does nothing to your rights at home until legislation is passed to implement it.
And a large amount of the rule-making that shapes daily life is private. Technical standards bodies decide what a plug or a shipping container or a file format is. Sports federations run disciplinary systems with their own tribunals. Platform terms of service govern the speech of billions of people. Professional bodies license and discipline their members. None of that is law in the strict sense, and all of it is enforced. That's why the boundary between law and regulation-by-other-means is such a busy area of argument, and it isn't going to settle soon.
Where legitimacy is claimed to come from
Every system has a story about why anyone should obey it. The stories are worth knowing because they're still being used, in updated wording.
The oldest is divine or cosmic authority. The stone stele carrying Hammurabi's laws, cut in Babylon around the eighteenth century BC, shows the king standing before a god at the top and the provisions below — the layout is the argument. Scholars still debate whether the text functioned as a working statute or as a royal statement of what a just king was, but the legitimating move is unmistakable.
Then natural law: the claim that there are standards of right conduct independent of any ruler, discoverable by reason, and that human law is only properly law when it conforms to them. Aquinas is the canonical exposition; Grotius carried the idea into international relations. Its practical bite is that it licenses saying an enacted rule isn't really law at all.
Social contract theories relocate authority into agreement. Hobbes argued people would rationally submit to a sovereign to escape a violent state of nature; Locke made the submission conditional and retained a right of resistance; Rousseau grounded it in a general will. Nobody signed anything, which is the standing objection, and you can't produce the document. The reply has always been about tacit consent and about what the alternative looks like.
Legal positivism refuses the whole question of moral merit. On this view law is whatever the system's own rules identify as law, and an unjust law is still a law — you can then argue separately about whether to obey it. John Austin's nineteenth-century version defined law as the command of a sovereign backed by sanctions. H.L.A. Hart, in The Concept of Law in 1961, replaced that with something subtler: a legal system is a union of primary rules telling people what to do and secondary rules about how rules are made, changed and identified, resting on a "rule of recognition" that officials in fact accept and use. Kelsen posited a basic norm at the top of the pyramid that has to be presupposed rather than proved.
Hart's exchange with Lon Fuller in 1958 is the classic collision, and it's still the argument. Fuller replied that law has an internal morality of its own: rules must be public, intelligible, prospective rather than retroactive, not contradictory, not impossible to comply with, reasonably stable, and actually applied by officials as written. A system that fails those tests badly enough, he argued, isn't a bad legal system — it isn't a legal system.
Modern democracies mostly run on a mixture: consent expressed through elections, plus constitutional limits derived from something like rights, plus a positivist account of how you tell whether a given rule is valid. The mixture isn't tidy, and nobody has made it tidy.
Law travels, usually by being copied
Systems rarely invent from scratch. They import.
The comparative lawyer Alan Watson used the phrase "legal transplants" for this, and once you're looking for it the pattern is everywhere: colonies inheriting the metropole's system wholesale, newly independent states keeping most of it, Japan and Turkey deliberately shopping for a code, post-communist states drafting constitutions with foreign advisers in the room, commercial law converging because international business demands predictability.
Transplants don't always take. A rule that works in one country can behave completely differently in another because the courts, the professions, the enforcement capacity and the surrounding customs are different. Text is portable. Institutions aren't, and you can't post them.
The rules are only as real as the enforcement
The last thing to hold on to is that a rule and a practice are different objects.
Statutes sit on the books unenforced for decades. Prosecutors exercise discretion. Police decide what to stop. Regulators pick their battles with the resources they've got. Civil rights that require an expensive lawsuit to vindicate are worth less to someone who can't afford one, which is why access to justice is treated as a component of the rule of law rather than an optional extra.
So "where do laws come from" has a formal answer and a practical one, and they aren't identical. Formally: legislatures, courts, constitutions, treaties, custom. Practically: also from whoever drafts, whoever enforces, whoever can afford to litigate, and whoever the system quietly leaves alone.
The forty rules you obeyed this morning came from all of it at once.
