Legal opinions: six questions to ask before you rely on one.

A legal opinion is the one deliverable an adviser produces that is bought, in part, for a day the buyer hopes never comes. On the day it is commissioned, it guides a decision. On the later day (the review application, the audit, the inquiry, the disaffected shareholder), it becomes the first document pulled from the file: the proof that the decision-maker asked the right question and acted on considered advice. Most opinions are judged on the first day only. They should be commissioned for both.

The law treats the quality of an opinion as load-bearing. Section 76(4) and (5) of the Companies Act 71 of 2008 permit a director to rely on the advice of legal counsel and other professionals, but only on matters the director reasonably believes fall within that person’s professional or expert competence, and only where the person merits confidence. Reliance on weak advice is not a defence; it is exhibit two. For organs of state the stakes are sharper still: a decision must be lawful and rational, and when it is taken on review, the opinion in the record is read as closely as the decision it supported. In both worlds, the protection an opinion offers is exactly as strong as the opinion itself.

Six questions decide whether it will hold. Any board member, registrar or executive can ask them. None requires legal training.

1. Does it answer a precisely stated question?

A proper opinion begins by restating its instructions in the writer’s own words: the question presented, and nothing wider. Read that restatement first. If it is not the question you meant to ask, stop there and send the opinion back; everything that follows answers something else. A precise question is half the answer. A vague one guarantees a conclusion that sounds considered and decides nothing, which is discovered only when the decision is tested and the opinion turns out not to cover it.

2. Does it declare its facts and assumptions?

An opinion stands on facts, and it is only as strong as the facts it stands on. The writer must set out the facts relied on and label what is assumed rather than established, on the face of the document, because if the facts later prove different the opinion falls away to exactly that extent. The board is entitled to see where the ground is firm and where it is taken on trust. Distrust an opinion that silently converts a contested fact into a convenient assumption: it has decided the case in its own favour before the analysis begins.

3. Does it face the strongest contrary argument?

Advice that only builds the case for the answer the client wanted is advocacy wearing advice’s clothing. A reliable opinion states the best argument against its own conclusion and explains why that argument does not prevail, or concedes the point and adjusts the conclusion. The absence of a contrary view in an opinion does not mean there is none. It usually means nobody went looking, and the first person to go looking will be acting for the other side.

4. Can the reasoning be verified?

Every step in the chain (from instrument, to interpretation, to conclusion) should be visible, and every authority identified precisely enough for a second reader to find it and check it. “It is settled law that” followed by nothing is not a method; it is an appeal to trust. The point has become sharper in the age of machine-drafted text and invented citations: ask whether the authorities were verified at first hand, and whether a research record exists that shows they were. An opinion whose reasoning cannot be checked cannot honestly be relied on; it can only be believed.

5. Does it commit to an answer, and calibrate it?

An opinion hedged into meaninglessness protects its writer, not its reader. A proper opinion answers the question asked, in terms; states the degree of confidence in that answer; and identifies what would change it: a fact proving otherwise, an amendment pending, a point on which authority is divided. And it keeps two findings apart that are too often blended: whether a course is lawful, and whether it is wise. A board is entitled to both, stated separately, because it may lawfully do an imprudent thing and must know that it is choosing to.

6. Does it state its limits?

Scope, date and reliance. What the opinion covers and what it expressly does not. The date at which it speaks, because an opinion does not update itself when the statute is amended or the authorities move, and a two-year-old opinion re-used for a new decision is a new risk wearing old comfort. And who may rely on it, for what purpose. These clauses read as boilerplate until the day they are the whole argument. An opinion that claims to cover everything covers nothing.

Commissioning for reliance

The buyer’s side of the discipline is shorter. Brief the question, not the answer you are hoping for. Hand over the documents, including the inconvenient ones, because an opinion built on a curated file is built on sand. Ask for a plain-language summary that the board, council or committee can actually table and debate, with the full opinion behind it. And when the body acts on the advice, minute that it did: reliance that was never recorded is remarkably difficult to prove years later, when it matters most.

The test of an opinion arrives late, and it arrives without warning. When it does, the opinion should be the strongest document in the file: proof that the organisation asked the right question, confronted the strongest answer against it, and acted on advice that committed itself. That is what it was bought for.

This piece is general commentary on advisory practice. It is not legal advice on any specific matter, and no professional relationship arises from reading it.