How to conduct legal research in Zambia

Step 1 — Reduce the problem to legal issues

Before searching anything, write the problem out as a short set of legal issues. A client's account of events is rarely framed in legal terms, and searching the client's words tends to produce irrelevant results. Convert the facts into questions a court could answer: is there a binding contract, was the termination lawful, does the court have jurisdiction, is the claim out of time?

Each issue should be narrow enough to be answered by a rule. "Was my dismissal unfair?" is a topic; "does an employer's failure to hold a disciplinary hearing render a summary dismissal wrongful under the Employment Code Act?" is an issue. Narrow issues produce precise searches, and precise searches are what make the difference between an afternoon and a week of work.

Step 2 — Start with legislation, not with cases

Zambia is a statute-led jurisdiction for most everyday disputes. Employment, company, tax, land, criminal and family matters are all governed in the first instance by an Act of Parliament, often supplemented by statutory instruments. Begin by identifying the governing Act by name, then read the specific sections that create the right, the obligation or the procedure you are interested in.

Read the definitions section before the operative sections. A very large number of arguments in Zambian courts turn on whether a person or thing falls within a statutory definition, and definitions are frequently narrower than ordinary usage. Read the commencement provisions too: an Act may have been passed but only brought into force in part, or on a later date fixed by statutory instrument.

Step 3 — Find the judgments that interpret the provision

Once the provision is identified, look for judgments that have interpreted it. Search by the section number and the Act name together, and separately by the legal concept. Give priority to the superior courts: the Supreme Court and the Constitutional Court within their respective jurisdictions, then the Court of Appeal, then the High Court.

Read the judgment itself rather than a headnote or an AI summary. A summary tells you the outcome; only the judgment tells you the ratio decidendi — the reasoning necessary to the decision — which is the part that binds later courts. Note the composition of the coram, the date, and whether any judge dissented.

Step 4 — Verify that the law is still good law

This is the step most often skipped, and it is the one that costs cases. Two checks are required. First, statutory: has the section been amended or repealed, and has a consolidating Act replaced the one you read? Second, judicial: has the judgment been overruled, distinguished, or displaced by legislation enacted after it?

A judgment interpreting a repealed provision is not worthless — it may still illuminate an identically worded successor provision — but it must be cited for what it is. Never cite a case for a proposition without knowing the current statutory context in which that proposition now sits.

Step 5 — Record the research so it can be re-run

Keep a research note that records, for each issue: the provisions relied on, the judgments relied on with full citations, the date on which you verified currency, and the searches that produced nothing. The negative results matter — they are what allow a supervising advocate or an examiner to see that the field was covered.

Where AI tools are used, the research note should record which parts were AI-assisted and which authorities were opened and read in full. Anything not personally read should not appear in a submission to a court.

Where LEX ZED fits

LEX ZED holds a searchable library of Zambian judgments and legislation and layers an AI research assistant over it. The assistant is designed to point you to the governing provision and the authorities that interpret it; the verification steps above remain yours. Users can browse the case library and legislation library without asking the assistant anything at all.