Jurisdictional Discipline, Carried Out of Islamabad

Legalise was built for one jurisdiction first. The argument at the time was that a drafting tool which knows the Islamabad Capital Territory precisely is worth more to an Islamabad advocate than a national tool that is accurate about nothing in particular, and that argument was right. It is also, as it turns out, the reason the platform could expand at all.

Legalise now drafts, reviews and compiles for five jurisdictions: the Islamabad Capital Territory, Punjab, Sindh, Khyber Pakhtunkhwa and Balochistan. This post is about what that actually meant to build, and about the rule that governs it, which is that five jurisdictions are five jurisdictions and never an average of them.

The failure mode we were avoiding

The tempting way to make a legal tool national is to pour every jurisdiction's law into one collection and let a model find whatever seems relevant. It produces something that demonstrates well and fails in a filing.

The failure is not usually a wrong statute. It is a plausible one. A Punjab eviction petition that cites the Islamabad Rent Restriction Ordinance is obviously wrong to the advocate reading it. A Punjab eviction petition that pleads bona fide personal need, a ground the Islamabad Ordinance provides and the Punjab Rented Premises Act 2009 does not, looks entirely normal until a Rent Tribunal asks which provision it is founded on. Nothing in the document announces the error. The cause title is right, the forum is right, the register is right, and the ground does not exist.

That is what a national soup produces: documents that are wrong in exactly the places where being wrong costs the most, and wrong in a way that reads as competent.

Partition first, then expand

So the rule is strict partitioning at every level, and expansion means building a jurisdiction rather than admitting one.

A document's jurisdiction is fixed when you submit it, and it then governs the whole chain rather than a paragraph of instructions at the top. It decides the court header and the forum the matter is routed to. It decides which statutes the drafting engine is permitted to cite, so a provincial matter is not offered the Capital Territory's list. It decides which body of provincial clause material is available at all. It reaches the reviewer, so a review of a Punjab agreement is conducted against Punjab law rather than graded against Islamabad's. It reaches the compiler, so a briefing describes the forum the matter is actually going to.

Each of the four provinces was built the same way and separately. The rent forum was established from that province's own rent legislation, which is different legislation in each of them and confers different grounds. The civil route was established from that province's own courts legislation. The family forum, the writ jurisdiction and its principal seat, the governing statutes for each template: all of it province by province, from that province's own code portal and gazette rather than from a federal source that says nothing about provincial law.

What is deliberately not there

Two consequences follow that are easy to read as gaps and are in fact the discipline working.

The first is that a jurisdiction the platform does not have does not fall back to one it does. An unrecognised jurisdiction resolves to an explicit statement that the material is unavailable. It does not quietly resolve to the Capital Territory, which was the default the codebase had inherited, and it does not resolve to federal law either, because eviction grounds are provincial law with no federal counterpart and falling back to federal would produce a document founded on nothing.

The second is that where a province genuinely has no confirmed figure, no figure appears. Some provinces publish court fee and stamp schedules that have been established from primary sources, and those are used. Others have not been established yet, and rather than borrow Islamabad's numbers and print them under a provincial cause title, the platform prints nothing and says so. A missing number is a research gap. A borrowed number is a false statement about that province's law, and it is indistinguishable from a correct one until somebody pays a wrong duty.

The tools that stayed in Islamabad, and why

Drafting, review, compilation, the statute library and the case law database all work across the country. A specific set of reference tools does not, and it says so on every one of them.

The stamp duty and court fee calculators, the courts and police station directories, the jurisdiction decoder, the practice rules, the bench composition and the process guides remain scoped to the Islamabad Capital Territory. These are not statute text that can be collected; they are administrative facts about a particular place, and each one needs establishing from that province's own authorities before it can be published. Selecting a province greys those tools and tells you which jurisdiction they cover rather than showing you Islamabad's figures under another province's name.

One tool was in that set and should not have been. Limitation periods come from the Limitation Act 1908, a federal statute with no provincial variation, so restricting it was refusing a provincial advocate a tool that applies perfectly well to them. It is open to all five jurisdictions now.

What it means in practice

You select a jurisdiction, and below the High Court a district, and the platform resolves the court, the forum and the governing law from that choice rather than from a default. The companion post on choosing your jurisdiction sets out line by line what changes in the document when you switch.

The claim is not that Legalise covers Pakistan. It is that it covers five jurisdictions, separately, and tells you where each one ends.

This post replaces two earlier ones written when the platform was scoped to a single jurisdiction, Why Legalise Covers Only Islamabad: The Case for Jurisdictional Specificity and What Jurisdiction-Specific Legal Tools Mean for Pakistani Advocates. Both remain available and are marked as archive, and the argument in the first of them is the reason this one exists.