Revision Hub / Module 4 · Skills & Defences

Problem-Solving Skills

Lecture L10 Skills topic — applies to every problem question Reading: Horsey & Rackley ch 1; Nolan & Oliphant casebook

Snapshot

A tort problem question is a story at the end of which you are asked to advise the parties. It tests not your knowledge of the law in the abstract but your capacity to apply it. The default scaffold is IRAC — Issue, Rule, Application, Conclusion — and within negligence the rule layer is the four-element chain: duty, breach, causation/remoteness, defences. You are solving this problem, not showing off; you are writing for a marker who has thirty scripts to read and rewards clarity, structural signalling and authority cited per proposition.

1. Reading the problem

Allow yourself 5–10 minutes to read before you write a single sentence. The temptation under exam pressure is to start drafting the moment you spot a familiar fact pattern. Resist it. Candidates who get Firsts plan first; those who run out of time started writing before they understood the question.

On the first pass, read the facts straight through without pen on paper. On the second pass, annotate. You are looking for four things:

The area of law. Is this negligence? Is there an occupier, a manufacturer, a public authority, a statement causing economic loss, a witness suffering psychiatric harm? The labels you attach now control the rules you will deploy.
The parties. Who is the claimant? Who is the victim of the negligence (not always the same person — think secondary victims)? Who is the defendant — and is there more than one? List them.
The harm. Personal injury, property damage, pure economic loss, psychiatric injury, distress falling short of injury? Different heads of damage attract different duty rules; misclassifying here will derail every subsequent step.
What is obvious vs what is contentious. An established duty (driver–road user; manufacturer–consumer) is taken as read in two lines. Contentious points — was there a duty of affirmative action? was the harm too remote? — deserve the bulk of your ink.

Ask yourself the lecturer's three questions on re-reading: "Where have I seen this before?" (which seminar fact pattern does this echo?), "What is obvious?" (which elements are conceded?), "What is contentious?" (where is the marker waiting for you to argue both ways?).

2. Structuring the answer

Strong scripts signal their structure; weak scripts bury analysis in undifferentiated prose. The recommended skeleton, drawn from the lecturer's worked example, is:

Outline claims. A short opening paragraph identifying who is suing whom, in respect of what harm, on what cause of action. One sentence per claim is enough. e.g. "Joe v Supersmooth (negligence — personal injury); Joe v local authority (negligence/breach of statutory duty); Sam's estate v Fred (negligence — fatal injury) and v Supersmooth and Joe."
One heading per claim, per claimant. Treat each claimant–defendant pair as its own mini-answer. Do not lump them together; a marker awarding marks per claim cannot give you credit for analysis they cannot find.
Within each claim, the four elements in order. Duty → Breach → Causation → Remoteness → Defences. Even where an element is uncontested, name it and dispose of it in a line ("Duty is plainly owed: Donoghue v Stevenson"). Skipping elements signals to the marker that you do not know they are required.
Conclusion per claim, then an overall conclusion. Each claim ends with a one-line advice ("Supersmooth is likely to be liable to Joe in negligence"). The overall conclusion ties the threads together: who is liable to whom, and on what basis.

Use bold or underlined sub-headings ("Joe v Supersmooth — Duty") in coursework; in exams, even a short underlined phrase signals the shift. Markers reward a script they can navigate.

3. The IRAC method, and why it earns marks

IRAC is the engine inside each element — a discipline, not a formula: every analytical paragraph should do four things in order.

Issue. State the legal question precisely. Not "is there a duty?" but "does a manufacturer owe a duty in negligence to the ultimate consumer of its product?". The narrower the issue, the easier the rule and application.
Rule. State the rule of law with authority. A proposition without a case is an assertion; a proposition with a case is a legal argument. "A manufacturer owes a duty of care to the ultimate consumer: Donoghue v Stevenson [1932] AC 562; Grant v Australian Knitting Mills [1936] AC 85."
Application. Apply the rule meticulously to the facts. This is where most marks live. Do not paraphrase the facts back; map the facts onto each ingredient of the rule. If the rule has three limbs, walk through three limbs. Where the facts are ambiguous, say so and argue both ways.
Conclusion. Reach a conclusion with caveats. "On balance, Supersmooth is likely to be in breach, although a court might take a more lenient view if the defect was not reasonably discoverable on testing." Hedged conclusions are not weak — they are realistic, and they are what practitioners actually write.

A worked micro-example

Facts: D crashes into C while texting; C suffers a fractured leg.

Issue. Whether D, a motorist who was texting at the wheel, is liable in negligence to C, a pedestrian, for personal injury (a fractured leg) sustained when D's vehicle struck C.

Rule. Road users owe one another a duty of care: Nettleship v Weston [1971] 2 QB 691. The standard is that of the reasonably competent driver, judged objectively (Nettleship), and inattention to the road in order to use a phone falls below it. The but-for test of factual causation applies (Barnett v Chelsea & Kensington HMC [1969] 1 QB 428); personal injury is a foreseeable kind of damage in a road traffic collision (The Wagon Mound (No 1) [1961] AC 388).

Application. D plainly owes C a duty as a fellow road user. By texting while driving D failed to keep a proper look-out, falling below the standard of the reasonably competent motorist; the breach is straightforward. But for D's distraction the collision would not have occurred — factual causation is satisfied. A fractured leg is a foreseeable kind of harm in a vehicle/pedestrian impact, so the damage is not too remote.

Conclusion. D is liable to C in negligence for the fractured leg. Subject to any contributory negligence (e.g. if C stepped out without looking), no defence appears on the facts.

That paragraph is roughly 200 words and earns marks across all four elements. Notice what it does not do: narrate the facts back, list every road traffic case in the textbook, or equivocate where the law is settled.

4. Time management

For a three-hour exam answering three questions, the arithmetic is brutal: roughly 55–60 minutes per question including planning. A workable allocation:

0–8 minutes: read the question twice, identify parties, claims and harm, draft a skeleton plan in the margin or on a separate sheet.
8–50 minutes: write. Stick to the plan. If you spot a missed point, slot it in; do not redesign the structure.
50–55 minutes: finish the conclusion and re-read for sense.
Stop at the cap. A finished but imperfect answer outscores a brilliant unfinished one because the unwritten paragraphs score zero.

When an answer is running long, the temptation is to keep writing. Don't. Marginal returns inside a single answer fall sharply after the second hour; marginal returns from starting the next answer are huge, because the first ten marks of any question are the easiest to win.

5. The most common examiner traps

Examiners construct facts to test specific failure modes. Recognising the trap is half the battle:

Multiple claimants and defendants. A single set of facts often supports four or five separate claims. Miss one and you cap your mark. The fix is the disciplined "who can sue whom?" opening list.
PEL and psychiatric subtypes hiding in plain sight. A statement causing financial loss is Hedley Byrne territory, not ordinary negligence. A bystander watching a horror unfold is a secondary victim with the Alcock control mechanisms. Misclassify and the wrong duty rules apply.
Consent and illegality. When the claimant participated in the activity, ask whether volenti non fit injuria or the illegality defence (Patel v Mirza) is in play before you conclude liability.
Vicarious liability lurking. If the careless actor is an employee, the deep-pocketed defendant is the employer. Do not stop at the individual's negligence; ask whether the employer is vicariously liable on the close-connection test.
Omissions and public authorities. A failure to act is generally not actionable (Stovin v Wise; Gorringe v Calderdale). Spot the omission framing before you assert a duty.
Novus actus interveniens. Where a third party intervenes between breach and damage, ask whether the chain is broken (Wright v Lodge) or whether the second event was a foreseeable consequence of the first (Rouse v Squires).

Wright v Lodge [1993] 4 All ER 299

Holding: A second driver's reckless speed was held to be the sole legal cause of a subsequent collision; the original negligent obstruction did not extend to that loss. Illustrates how a sufficiently extreme intervening act can sever causation.
Why it matters: Pair with Rouse v Squires [1973] QB 889 (foreseeable subsequent collisions remain attributable to the original wrongdoer) when arguing both sides of novus actus in road-traffic problems.

Mansfield v Weetabix [1998] 1 WLR 1263 (CA)

Holding: A lorry driver who lost control because of an undiagnosed hypoglycaemic condition was not in breach: the standard is that of the reasonably competent driver unaware of any incapacitating condition.
Why it matters: A textbook word-association case (the Weetabix lorry) — useful for testing whether you remember the qualified objective standard, and the kind of fact pattern that turns on awareness of incapacity.

6. Coursework vs exam

The 2,000-word coursework problem and the time-pressured exam problem are the same exercise at different intensities. The differences:

Depth of authority. In coursework you are expected to cite primary authority precisely — full citations, paragraph numbers, key dicta. In the exam, "Donoghue v Stevenson" is enough; year and report are a bonus, not a requirement.
Footnotes and OSCOLA. Coursework is footnoted in OSCOLA. Exam scripts are not.
Counter-argument. Coursework rewards genuinely two-sided argument and engagement with academic commentary (Nolan, Stevens, Stapleton). Exam answers reward concise both-ways analysis without the academic apparatus.
Polish. Coursework is drafted, redrafted and proof-read. Exam prose is allowed to be functional. Markers calibrate accordingly.
Word discipline. Coursework has a hard cap; over-running is penalised. Exam answers are capped only by time. In coursework, every paragraph must earn its place.

The structural skeleton — claims list, four elements per claim, IRAC within each, hedged conclusion — is the same in both. Learn one method; dial depth up or down.

7. Reading list and memorising cases

The two indispensable books are Horsey and Rackley, Tort Law (9th edn, OUP 2025) and Nolan and Oliphant, Tort Law: Text and Materials (7th edn, OUP 2023). They serve different purposes:

Horsey & Rackley is the textbook: it gives you the doctrinal narrative, the structure of each tort, the policy debates and worked illustrations. Read it for understanding, not for case detail. After each chapter, close the book and write the four-element scaffold for that tort from memory.
Nolan & Oliphant is the casebook: extracts of judgments, edited for what matters. Use it to read the words of the judges themselves on the cases you most need to know (Donoghue, Caparo, Robinson, Bolam, Bolton v Stone, Wagon Mound, Alcock, Fairchild). Quoting a single sharp phrase from a judgment in an answer is worth more than a paragraph of paraphrase.

Memorising cases

You cannot bluff a tort exam without case names. The slides recommend a small number of techniques that work best in combination.

Word association. Tie each case to a vivid mnemonic anchor: Mansfield v Weetabix — the lorry driver who blacked out behind the wheel of a cereal lorry; Barr v Biffa Waste Services Ltd — Biffa, the bins, the smell, the nuisance. The sillier the image, the better it sticks.
Flash / cue cards. One side: case name and citation. Other side: one-line facts, one-line ratio, one-line significance. Carry a deck. Test yourself in dead time — bus, queue, kettle.
Flowcharts. Draw the negligence chain as a flowchart with branch points (established duty? novel duty → Robinson/Caparo; factual causation → but-for, then exceptions; remoteness → kind of damage foreseeable?). Each branch point is where a case lives. Drawing the chart from blank paper is a daily exercise; redrawing it under exam conditions is a five-minute warm-up before you write.
Reading aloud and quizzing. Cases read silently slip away; cases read aloud to a study partner who fires questions ("what's the ratio? what's the standard? what's the limb that failed?") embed. Reciprocal quizzing in pairs is the single highest-yield revision technique on the slides.
The four elements as a default scaffold. If you remember nothing else under exam stress, write down "Duty / Breach / Causation / Remoteness / Defences" before you read the question. That scaffold alone will catch eighty per cent of the marks on any negligence problem.

8. Common pitfalls

Treating the problem as an opportunity to show what you know. A problem question asks you to solve the dispute presented, not the one you wish had been set. Every paragraph that does not advance the analysis of these facts is dead weight.

Stating rules without authority. "There is a duty of care" is a marker waiting for a citation. Each rule needs at least one case. Each contested rule needs the leading case and a counter-authority where one exists.

Skipping uncontested elements. Markers award credit per element. If duty is obvious, dispose of it in a line; do not leave it out altogether. An answer that jumps straight to breach signals incomplete understanding.

Conclusions that refuse to conclude. "It is unclear whether D is liable" earns nothing. Reach a view, with caveats: "On balance D is likely to be liable, although a court might take a different view if X." Hedged is fine; agnostic is not.

9. Pre-flight checklist

Read the question twice before writing anything; spent 5–10 minutes planning.
Identified every claimant and every defendant, including the non-obvious ones (employers, public authorities, secondary victims).
Classified the harm in each claim (PI / property / PEL / psychiatric).
Drafted a "who sues whom" opening list before the body of the answer.
One heading per claim per claimant; four elements ordered within each.
Each rule cited with at least one case authority; key dicta quoted briefly where they help.
IRAC discipline applied within each element: issue, rule, application, conclusion.
Examiner traps actively checked: multiple parties, PEL/psychiatric subtypes, consent/illegality, vicarious liability, omissions, novus actus.
Hedged conclusion per claim, then an overall conclusion tying the threads together.
Stopped at the time cap and moved to the next question — finished beats perfect.