The substantive topics in Foundation Paper F form four distinct legal inquiries: whether the claimed subject-matter is excluded, whether a single prior-art disclosure destroys novelty, whether the invention is obvious, and whether the disclosure is sufficient. Because each gate has its own test and vocabulary, an argument phrased under the wrong gate does not answer the objection in front of it, even when the facts support it. Work through the scenarios below to see how that happens, then use the rubric and preparation sequence to test your own claim maps.
Why exclusion, novelty and inventive step must stay separate
Every substantive topic on Paper F sits inside one of a few sequential legal gates. Before analysing anything, decide which gate a question targets, because each gate has its own test, its own vocabulary, and its own characteristic failure modes.
The gates run in a fixed order. First, is the claimed subject-matter an invention within a field of technology, or does it fall into an excluded category as such? Second, is it already disclosed, in an enabling way, by one single prior-art document? Third, would it have been obvious to the skilled person starting from that prior art? Only after these come the requirements about the application itself: industrial applicability and sufficiency of disclosure. The requirements are cumulative, so passing one gate never excuses failing another, and each has a different evidentiary standard.
Argument routing is where the analytical discipline shows. An argument that a prior-art document would have been obvious does not answer a novelty objection, because novelty asks only whether everything in the claim is already disclosed once, explicitly or inherently. An argument that the claim is excluded as a business method does not respond to a sufficiency attack. When you draft practice answers, open each gate in the same order, dispose of it with its own test, and refuse to borrow reasoning from the neighbouring gate. The table below is a routing aid: find the row whose question matches the objection in front of you, and use the argument shape listed there.
| Gate | Question it answers | Argument shape that fits | Analytic trap |
|---|---|---|---|
| Excluded subject-matter | Is the claim an invention in a field of technology, or an excluded category as such? | Classify which claim features contribute technical character | Dismissing a mixed technical/non-technical claim wholesale |
| Novelty | Is everything claimed disclosed in one enabling prior-art document? | Map each claim feature to a single disclosure, explicit or inherent | Combining two documents, or treating a broad disclosed range as disclosing every value inside it |
| Inventive step | Would the claimed solution have been obvious from the closest prior art? | A problem-solution chain from distinguishing features to an objective technical problem | Rebuilding the problem using the claim itself as the checklist |
| Sufficiency and industrial application | Does the application enable the claimed scope, and can it be used in industry? | Compare claim breadth with what the description teaches the skilled person | Answering with prior-art arguments instead of disclosure arguments |
Excluded subject-matter: judging mixed inventions feature by feature
Exclusion is assessed on the claim as a whole, but the decisive question is which features contribute technical character. Claims mixing technical and non-technical features must be decomposed, not dismissed wholesale because their label sounds commercial or mathematical.
European patent law excludes certain categories as such, including schemes and business methods, mental acts, presentations of information and computer programs, but it excludes only those categories as such, not every claim that touches them. A mixed-type claim therefore survives the gate if enough of its features contribute technical character, for example by changing how a computer, a network or a machine operates. Train yourself to label every feature technical, non-technical, or ambiguous before concluding anything about the claim as a whole.
Worked scenario one. A claim recites a computer-implemented method for optimising delivery routes, processing customer preference data and applying a routing algorithm to minimise travel time. A tempting shortcut is to classify the whole claim as a business method as such and stop there. The better decision is to decompose: the routing algorithm determines how a technical system, the delivery operation supported by servers and vehicles, actually runs, so features affecting that internal functioning can contribute technical character, while the customer-preference aggregation may not. The distinction matters because a wholesale dismissal ends the analysis prematurely, whereas the gate itself demands feature-level classification before any conclusion. Practise this on claims whose titles mislead: a claim called an advertising method may contain network-management features, and the label never decides the gate.
Novelty: the single-document rule and inherent disclosure
Novelty compares the claim with one enabling prior-art disclosure, not with an averaged picture of the field. A feature destroys novelty only when it is explicitly disclosed or inevitably results from what the document teaches the skilled person.
At the novelty gate you may use exactly one prior-art disclosure as the basis, and it must be enabling: the document must teach the skilled person how to arrive at what it discloses. You may not combine a feature from a second document, however close it is, because combination belongs to the inventive-step gate, not to novelty. When mapping a claim, go feature by feature and ask where each one appears in that single disclosure, either in so many words or as an inevitable consequence of what is taught.
Worked scenario two. A prior-art document discloses a protective coating and states a thickness range of five to twenty micrometres. A later claim requires a coating of twelve micrometres. A plausible mistake is to conclude that novelty fails because the claimed value falls inside the disclosed range. Under European practice a value merely located within a broader disclosed range is normally novel: novelty fails only where the prior art explicitly discloses that value, or where its teaching, for example a preparation method as taught and exemplified, inevitably results in precisely that value. If neither holds, twelve micrometres is your distinguishing feature, and the real fight moves to inventive step, where a technical effect tied to the chosen thickness can carry the day. The distinction matters because conflating a range with every value inside it hands the case away at the wrong gate, while the inherence check must target the specific claimed value itself, not the span around it.
- One document only: if you find yourself citing a second document at the novelty gate, move the whole analysis to inventive step
- After listing explicit disclosures, write one sentence per claim feature stating what the skilled reader would inevitably obtain
- For range claims, test inherence against the claimed value itself, not the disclosed span
- Any feature that survives both lists becomes your distinguishing-feature inventory for the next gate
Inventive step: building the problem-solution chain without hindsight
The problem-solution approach runs in named steps: identify the closest prior art, determine distinguishing features, establish their objective technical effect, formulate the objective technical problem, then ask whether the skilled person would have arrived at the claimed solution.
Each step constrains the next. The closest prior art is the single disclosure pointing most promisingly toward the invention, judged before you know the claim's wording in detail. Distinguishing features come from the novelty-style mapping you already performed. Their technical effect must be one the application itself makes credible from the teaching as filed, not an advantage invented after reading the prior art. Features that do not contribute technical character do not become technical problem elements; they enter the objective technical problem as aims or constraints the skilled person is free to meet.
Hindsight control is the practical discipline here. A reliable check is to redraft your objective technical problem so that it no longer restates any claim language, then ask whether a skilled person, starting from the closest prior art and facing exactly that problem, would have had a reasonable expectation of reaching the claimed solution. In a paper exercise this may show that a problem phrased as how to display data more clearly is really the claim in disguise, while how to reduce processing load on a constrained device is a genuinely technical formulation. The outcome always depends on the facts of the case, which is why the chain, not a fixed answer, is what you practise. Keep a one-page template with the five steps as labelled slots; filling them in order, with the effect-credibility check between steps three and four, makes inventive-step writing fast and internally consistent.
- Closest prior art: chosen from a single disclosure, before consulting the claim as a checklist
- Distinguishing features: the features your novelty map could not find in that disclosure
- Objective technical effect: an advantage the application as filed teaches or makes credible
- Objective technical problem: technical features become problem elements; non-technical features become aims
- Obviousness assessment: would the skilled person, facing that problem, arrive at the claimed solution?
Industrial application and sufficiency: gates about your own disclosure
Unlike the prior-art gates, industrial application and sufficiency test the application itself: whether the invention can be carried out and used in industry, and whether the disclosure enables the skilled person across the full scope claimed.
Sufficiency is claim-scope relative. The same description can enable a narrow claim completely while leaving a broad claim with embodiments the skilled person cannot achieve without inventive effort or new research. That is why a sufficiency argument is built as a gap analysis: pick the widest elements of the claim, often a numerical range or an open class of compounds or methods, and ask what the skilled person would actually have to do to make the invention work at the edges of that scope. Prior-art documents play no role at this gate; they cannot cure an inadequate disclosure.
To practise, take a sample claim containing a range and write, for each end of the range, what the skilled person must be able to do. Then read the description and mark whether it teaches, suggests or is silent about each requirement. The expected observation is that silence at an extreme, combined with a claim that demands that extreme, is the shape of a sufficiency argument, whereas a description that enables the whole range leaves you with nothing to argue at this gate and sends you back to the prior-art gates instead.
Procedure and formalities: a separate knowledge block, not an afterthought
Paper F also covers application procedure and formalities, which reward precise checklist knowledge rather than legal argument. Keep this material in its own revision stream so it does not blur with the substantive gates or their case reasoning.
Procedural questions test whether you know the sequence of a European application, what act is required at each stage, and who must perform it. This is recall material, and recall material decays differently from reasoning skill. Build a one-page procedural map showing the stages of the granting process, each node labelled with the act required and the actor responsible, and redraw it from memory weekly. A blank node is a precise revision target in a way that a vague feeling of unfamiliarity never is.
Resist the urge to fold procedural learning into your substantive practice sessions. Reasoning under the gates improves through slow, argumentative writing, while procedure improves through short, frequent, closed-book reconstruction. Time limits, fees and administrative requirements change and are not restated in this guide; for anything of that kind, rely on the EPO website at epo.org, which is the issuer's official source for such details. Treating the two streams separately prevents the drift where procedural facts get memorised loosely and substantive tests get memorised rigidly, exactly opposite to what each type of knowledge needs.
A claim-mapping exercise with a self-check rubric
Take a granted European patent in a technology you know, read one independent claim, and build a gate map: classify every feature, name the closest prior art you can find in the patent's own citations, and phrase an objective technical problem. Score yourself against the rubric.
Run the exercise three times over one week using different technology areas, for example a mechanical device, a software-implemented method and a chemical composition. Expected observations: at least some features resist clean technical or non-technical classification, which is normal and is precisely what the decomposition skill is for; your first objective technical problem will probably borrow wording from the claim, which is the hindsight signal to rewrite it; and the patent's cited prior art will usually contain one document that maps to most claim features, making it a realistic stand-in for closest prior art in the exercise.
A suggested six-week sequence you can adapt: weeks one and two, the gate framework and mixed-invention decomposition drills; weeks three and four, novelty mappings with inherence checks plus problem-solution chains from those mappings; week five, sufficiency gap analyses on ranged claims; week six, the procedural map and two full claim maps against the clock. Readiness checks before any exam sitting: you can state each gate's test in one sentence; you can map a mixed claim without relying on its title; you can phrase a problem statement free of claim language; you can redraw the procedural map with no blank nodes. Treat your rubric scores as learning milestones only, not as a prediction of any exam result.
- Rubric, scored 0 to 2 per line, with 10 to 12 total indicating a solid working grasp: every claim feature labelled technical, non-technical or ambiguous with a reason
- Novelty map uses one document only, notes inherent disclosures, and applies the range-versus-value inherence check to the specific claimed value
- Objective technical effect is grounded in the application's own teaching
- Objective technical problem contains no claim wording
- The obviousness step addresses the skilled person's path, not the result
- Sufficiency check identifies the claim's widest elements and what the description teaches about them
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.