From Problem to Effect: Navigating EP Patent Communication with Phrases for the Problem–Solution Approach

Understanding the EP Problem–Solution Approach (PSA) and Its Communication Expectations

The European Patent Office (EPO) uses the Problem–Solution Approach (PSA) as a structured method to assess inventive step. Because PSA structures the examiner’s reasoning, it should also structure how you communicate—in interviews and in written submissions. Your goal is to speak the examiner’s language: identify the closest prior art, isolate the distinguishing features, explain the resulting technical effect, and formulate an objective technical problem that is free from hindsight. Each component has linguistic expectations, and the words you choose can either strengthen or undermine your case.

Under PSA, the first step is the selection of the closest prior art, not the best art overall, but the most promising starting point from which the claimed invention could realistically be developed. Examiners expect you to justify why an art piece is the closest starting point, or why the one they propose is not “closer” in terms of purpose and structural overlap. Communicatively, this calls for calm, purpose-driven language. Avoid broad attacks on references; instead, use measured statements that connect technical purpose, context, and claim scope.

Next, you must delineate the distinguishing features over that closest prior art. The examiner will likely have listed these in the communication. Your task is to confirm or carefully refine that list, ensuring each difference is explicitly tied to claim language. Clarity of reference is paramount: cite claim numbers or elements, and avoid ambiguous phrases that look like claim drafting. The EPO wants factual contrasts, not promotional gloss.

The heart of the PSA is the technical effect arising from the distinguishing features. Here, many applicants falter by overstating or by relying on implied knowledge. The EPO scrutinizes whether an asserted effect is (a) credibly achieved across the claimed scope, and (b) supported by the application as filed. If your language drifts into unsubstantiated performance claims or introduces new concepts not traceable to the specification, you risk both credibility and added-matter objections. The register should be precise and restraint-based: anchor to data, mechanisms, or consistent principles disclosed on filing.

Finally, you must formulate the objective technical problem, which is a neutral statement that flows from the technical effect, without importing pieces of the solution. The EPO disfavours problems that are reverse-engineered to make the invention appear obvious; equally, it rejects problems that smuggle in your claim’s solution. Clear phrasing uses conservative, effect-driven nouns and verbs, not solution-laced specifics. Your aim is to state what the skilled person would genuinely try to achieve, given the closest prior art and the accepted effect, without hindsight.

Throughout, you should adopt jurisdiction-appropriate, effect-driven language that avoids claim-like generalities, marketing superlatives, and unanchored causality. The EPO is particularly sensitive to added matter: anything you state must be directly and unambiguously derivable from the application as filed. At the same time, careful phrasing can remain compatible with US enablement and written description by emphasising mechanisms, ranges, and representative embodiments disclosed from the outset.

Phrase Toolkit: Precise, Jurisdiction-Appropriate Stems for Interviews and Responses

Below are reusable stems that fit each PSA stage. They aim for concision, traceability to the specification, and credibility.

• Closest prior art

  • “The skilled person would select [Reference X] as the closest prior art because it pursues the same overall purpose of [purpose], addresses [context], and shares [key structural/functional overlap] with claim [n].”
  • “While [Reference Y] is technically relevant, it is directed to [different purpose/context], which diverges from the problem space addressed in the present application.”
  • “On balance, [Reference X] is the more promising starting point given its alignment with [specific objective] as evidenced at [paragraph/figure].”

• Distinguishing features

  • “Relative to [Reference X], claim [n] requires [feature A], [feature B], which are not disclosed or suggested at [citations].”
  • “The configuration at claim [n], lines [x–y], mandates [parameter/range/sequence] that is absent from [Reference X] and cannot be inferred without hindsight.”
  • “The cited passage at [X:col:line] references [general concept], but does not teach [specific structural constraint/ordering] as recited in claim [n].”

• Technical effect

  • “As disclosed at [spec paragraph/figure], [feature A] in combination with [feature B] results in [effect], evidenced by [data/trend/mechanistic rationale].”
  • “The effect is plausibly achieved across the claimed scope because [underlying principle] is not limited to a specific [embodiment] but arises from [structural/functional relation] recited in the claims.”
  • “Where quantitative improvement is at issue, the application reports [performance indicator], which supports the asserted effect without extrapolating beyond the disclosed ranges.”

• Objective technical problem

  • “In light of the closest prior art and the technical effect above, the objective technical problem can be formulated as: ‘to [achieve/improve/avoid] [effect-driven outcome] in [context]’.”
  • “A neutral statement of the problem avoids presupposing [feature A/B] and expresses only the effect observed.”
  • “Formulated accordingly, the problem is: ‘to enhance [parameter] under [operational constraints], while maintaining [compatibility/efficiency]’.”

• Non-obviousness logic

  • “The prior art provides no teaching or hint toward combining [feature A] with [feature B] to achieve [effect], particularly under [conditions] where [art teaches away/contradicts].”
  • “The proposed combination relies on hindsight: [Reference X] does not recognize [problem/effect], and [Reference Y] addresses a different objective, so the skilled person lacks a motivation to combine them in the claimed manner.”
  • “Even if combined, the cited teachings would not arrive at the claimed configuration because [missing constraint/order/parameter window].”

• Scope credibility and sufficiency

  • “The specification provides a reproducible pathway via [methods/parameters], enabling the skilled person to achieve the effect without undue burden across the claimed range [cite support].”
  • “Where the effect depends on [parameter], the application discloses representative values and control strategies that render the effect plausible within the claim’s boundaries.”

• Added-matter avoidance

  • “The present submission adheres to the application as filed by citing [paragraphs/figures] for each asserted feature–effect link; no new properties or mechanisms are introduced.”
  • “Where we refine terminology, it is a paraphrase of [original wording] at [citation], not an extension of scope.”

• Interview diplomacy and narrowing

  • “If helpful to advance prosecution, Applicant is prepared to clarify claim [n] by specifying [parameter/range/source] as supported at [citation], to reflect the effect-driving feature explicitly.”
  • “To focus the discussion, we propose to center on [feature A], which is the locus of the effect and the point of departure from [Reference X].”

These stems are intentionally conservative. They reduce the risk of overreach, integrate citations naturally, and reflect the EP expectation that assertions are traceable and effect-driven.

Applying the Toolkit: Mini-Scenarios and Contrasts Between Good and Risky Phrasing

Communication quality is often the difference between a persuasive PSA argument and an objection that snowballs. The following contrasts illustrate how to keep language precise, avoid hindsight, and prevent added matter.

• Justifying the closest prior art

  • Prefer: “Reference X targets the same operational context (on-device inference) and shares the pipeline structure recited in claim 1 (preprocess–quantize–execute), making it the most promising starting point.”
  • Avoid: “X is basically the same system as ours.” This is too broad, invites dispute, and lacks the purpose linkage that PSA expects.

• Stating distinguishing features

  • Prefer: “Claim 1 requires quantization after normalization with a clipped dynamic range (−a, a), which X neither discloses nor suggests at col. 3:15–25.”
  • Avoid: “Our quantization is more nuanced.” This is non-technical and untraceable. It risks reading like marketing copy.

• Articulating technical effect credibly

  • Prefer: “The clipped range reduces saturation error, yielding a lower mean absolute deviation as shown in Table 2, which is consistent with the disclosed mechanism at para. [0048].”
  • Avoid: “Performance is dramatically superior.” Without numbers or mechanism, this overclaims and can be challenged as unsupported.

• Formulating the objective technical problem

  • Prefer: “To reduce quantization-induced error in normalized signals under fixed compute budgets.” Neutral, effect-driven, no solution leakage.
  • Avoid: “To implement clipped normalization after which quantization is applied.” This bakes in the solution structure and invites an obviousness rejection.

• Addressing combinations and hindsight

  • Prefer: “Y addresses server-side training stability, not on-device inference accuracy; its teaching does not motivate introducing a clipped normalization stage into X’s inference pipeline.”
  • Avoid: “A person of skill would never combine X and Y.” Absolute statements are fragile and can be rebutted by any plausible motivation.

• Handling sufficiency and scope

  • Prefer: “The effect is tied to ‘a’ within 0.6–1.0 as disclosed; claims are commensurate because the mechanism depends on saturation avoidance, achieved within the cited range.”
  • Avoid: “Any clipping will work.” Overbroad claims without support strain plausibility and sufficiency.

• Preventing added matter

  • Prefer: “We refer to ‘saturation avoidance’ as a summary of the mechanism described at [0048], where the same concept is expressed as ‘limiting dynamic range to avoid outliers dominating quantization.’”
  • Avoid: “We discovered a novel noise-shaping effect.” If “noise-shaping” is not in the as-filed text, this risks added matter.

• Interview tone and concessions

  • Prefer: “If the Office considers the range for ‘a’ material to the effect, Applicant can amend claim 1 to recite 0.6–1.0 as supported at Fig. 3 and para. [0052].”
  • Avoid: “We can always add numbers later.” This suggests post-filing supplementation and raises added-matter concerns.

These contrasts demonstrate a consistent theme: good phrasing ties claims to disclosures and effects, uses neutral problem statements, and avoids conclusory or embellished language. Risky phrasing either imports hindsight, lacks factual grounding, or introduces concepts not present in the as-filed specification.

Aligning with US–EP Strategy While Preserving EP Sufficiency and Avoiding Added Matter

Applicants often pursue protection in both the US and Europe. While the legal tests differ, you can craft language that supports both jurisdictions without compromising EP standards. The EPO’s PSA expects effect-driven, evidence-based reasoning; the USPTO’s enablement and written description requirements expect that the specification teach how to make and use the invention across its scope and demonstrate possession of claimed subject matter. The alignment tactic is to focus on disclosed mechanisms, parameterizations, and representative embodiments, using phrasing that remains constant across jurisdictions.

First, emphasize mechanism-linked effects. In EP, this shows plausibility for the technical effect; in the US, it supports written description by demonstrating that you possessed a causal understanding or at least a consistent correlation at filing. Phrases like “The effect follows from [mechanism] as disclosed at [citation]” work on both sides. Avoid introducing new terminologies or speculative mechanisms post-filing; if a shorthand is helpful, explicitly tie it to the original wording.

Second, keep claims commensurate with support. In EP, plausibility must span the claim scope; in the US, enablement must avoid undue experimentation. Use language that tether claims to disclosed ranges, control variables, and boundary conditions. For example: “The claimed range of [parameter] reflects the intervals where the specification evidences [effect], and the skilled person can reproduce the result using the provided control method.” This satisfies EP sufficiency and bolsters US enablement.

Third, maintain neutral, effect-based problem statements that remain jurisdiction-agnostic. An EP objective technical problem like “to improve [metric] under [constraint]” dovetails with US narratives about technical improvement and utility without embedding the solution. If the US file history benefits from highlighting advantages, ensure any asserted advantage in EP is traceable to the filing so that later US arguments do not appear to expand the disclosure retroactively.

Fourth, manage combinations and teaching-away consistently. In EP, you argue non-obviousness by lack of motivation and incompatibility with the objective technical problem. In the US, similar themes appear as lack of motivation to combine and reasonable expectation of success. Harmonize phrasing that cites concrete incompatibilities: “Reference Y seeks [goal], which conflicts with [constraint] recognized in X; thus, the skilled person would not adopt Y’s teaching in X’s context without hindsight.” This line is persuasive in both forums.

Fifth, plan for amendments that reflect the effect-driving feature. In EP, amendments must be directly and unambiguously derivable. In the US, narrowing amendments can secure allowance but must remain enabled. Pre-identify effect-critical parameters and ranges in the specification; during prosecution, phrase potential amendments as clarifications rather than conceptual shifts: “To reflect what is already disclosed and practiced, Applicant can clarify that [parameter] is within [range], as supported at [citation].” This avoids EP added matter and preserves US enablement.

Finally, be vigilant about terminology stability. Choose technical terms at drafting that you can use verbatim in prosecution. If you need an umbrella term later, declare it as a synonym of language present in the application: “For brevity, ‘[Term]’ refers to the arrangement described at [paragraphs], i.e., [original wording].” This prevents EP added matter and supports the US record by showing consistent possession of the concept.

When executed together, these tactics let you move fluidly between EP’s PSA logic and US enablement narratives. Your phrasing remains anchored to disclosure, your effects are credible, and your objective problems are neutral and hindsight-free. The result is a coherent global prosecution posture: EP communications that pass PSA scrutiny and US arguments that rest on the same disclosed foundations. By mastering the phrase toolkit and its discipline—purpose-first closest prior art, explicit distinguishing features, credible effects, and neutral objective problems—you ensure that every interview and response builds a record aligned with both jurisdictions without compromising the integrity of either.