Bar questions are more repetitive in form than in content. Across subjects and years, the same handful of question architectures recur — and each one wants a different arrangement of the same raw material: the issue, the rule, the facts.
Learning the shapes is the fastest available improvement in exam technique, because it converts the first minute of every question from improvisation into recognition.
1. The distinction
"What is the difference between a void and a voidable contract?"
What it wants: two definitions, the axis of difference, and the consequence.
Shape: state A and B in one clause each, name what actually distinguishes them (not a list of every difference), then state what turns on it — who may invoke it, whether it can be ratified, what follows for third parties.
The trap: writing two essays side by side and never stating the distinction. A marker who has to infer the difference from your two paragraphs has not been answered.
2. The requisites question
"What are the requisites of a valid novation?"
What it wants: a numbered list, and — if facts are given — a sentence applying it.
Shape: "Four requisites: (1)…; (2)…; (3)…; (4)…." Then, if there are facts: "Here, the third is present because…; the fourth is not, because…."
The trap: prose that contains all four requisites but hides them. Numbered elements are marked as elements; the same content in a paragraph is marked as prose, and usually less generously.
3. The multi-part hypothetical
"X did this. (a) May Y sue? (b) What is Y's remedy? (c) May Z intervene?"
What it wants: each sub-question answered separately, in order.
Shape: label the answers (a), (b), (c) — even when the question does not. Each gets its own conclusion, rule and application. If one is a consequence of another, say so in one clause and move on.
The trap: answering the most interesting part at length and leaving the others to inference. The most expensive error in a bar paper is an unanswered sub-question, and it is also the most common.
4. The status question
"The contract lacks X. What is its status, and what are the rights of the parties?"
What it wants: the classification first (void, voidable, unenforceable, rescissible, inexistent), then the consequences that follow from that classification.
Shape: "It is [status], because [reason]. It follows that [who may do what], and that [effect on third parties / on the price / on the period]."
The trap: starting with the consequences. The classification governs everything else; if you get it wrong, the rest of the answer is wrong in a well-organised way. Commit early and defend it.
5. The liability question
"Is X criminally liable? Civilly liable? For what?"
What it wants: three separate questions, often, disguised as one — criminal liability, the specific crime or its absence, and civil liability (which can survive an acquittal, or exist without criminal liability).
Shape: "X is liable for [crime] under [provision]: the elements are (1)…(2)…(3)…, and each is present — [apply]. On civil liability, [ex delicto / separate], because…"
The trap: treating the three as one. An answer that convicts elegantly and says nothing about civil liability has answered two-thirds of the question.
6. The procedural scenario
"X filed a motion in the wrong court, three days late, without notice. What happens?"
What it wants: a sequence through jurisdiction → remedy → period → effect of the defect.
Shape: "The court has no jurisdiction because [test]. The proper remedy is [motion/appeal/petition] under [rule], filed within [period] from [date]. Filed as it was, the defect is [curable/not], so the filing [produces/does not produce] the effect sought."
The trap: jumping to the court's discretion. Procedural questions are mostly determinate: a period is a period, and a remedy that does not exist cannot be granted in the interest of substantial justice.
How to classify in the first minute
Four signals, read off the question before you write anything:
- The verb. "Define", "distinguish", "enumerate", "may he", "what is the remedy", "is it valid" — each maps to a shape.
- The number of question marks and the connective words. "And", "also", "if not" — each usually marks another sub-question.
- The presence of facts. Facts mean application; a question with no facts wants the rule.
- The last sentence. The question is usually there, and it is usually narrower than the story before it.
Write the shape in the margin before you write the answer: distinction / 3 parts / status → rights. Five seconds, and the next twelve minutes have a structure.
Where shapes combine
Most real questions are two shapes stitched together — a status question whose third part is a distinction, or a liability question with a procedural tail. That is why classifying is worth practising: the composite is obvious once you name the parts, and invisible if you treat the question as one undifferentiated story.
Practising shapes without writing twenty answers
Take one year of past questions in one subject and do nothing but label each question with its shape. An hour's work, and it produces a map of how the subject is asked — the exercise described in how to use past bar questions.
Then write skeletons: three lines per question, one per shape element. A year's questions in two hours. Then take the five questions whose skeletons were hardest and write those at full length against the clock.
The one failure that shape cannot fix
Shape organises knowledge; it does not supply it. A perfectly shaped answer around a rule you have misremembered scores for the shape and loses the marks for the law — and, because it reads fluently, it hides the problem from you until someone marks it.
So the shapes are the second layer. The first is the review method: retrieve, apply, write. The shapes are what make the last of those reliable.