How courts read 'comprising': a verify-before-commit claim-construction guide

Premium Deals
Mighty Travels Premium
Travel in style,
save up to 90%

On flights and hotels worldwide by booking the best deals when they appear.

See Deals

Sponsored

TakeawayDetail
Verify the live, complete option before committing to a Munich or Düsseldorf infringement theory.Reader rule: compare like-for-like totals and terms before relying on a single “comprising” limitation.
A single “comprising” limitation can shift infringement analysis between UPC divisions.Thesis condition: the 2025 UPC Court of Appeal claim-construction standard changes the division-specific infringement route.
Claim construction remains central to infringement analysis.Grounding identifies “Claim Construction and Infringement Contentions: A Feedback Loop” and stresses the importance of claim construction.
Interpretation (III) spans 31 articles covering claim construction and infringement analysis.Grounding states the draft comprises 31 articles, including jurisdictional rules, standing to sue, claim construction, and infringement analysis.

Courts read “comprising” through claim construction, and the working thesis here — which should be verified against the actual 2025 UPC Court of Appeal record before reliance — is that a single “comprising” limitation can shift infringement analysis between divisions such as Munich and Düsseldorf.

The answer changes when the live, complete option is not verified against like-for-like totals and terms before commitment.

Sunlit library reading room with tall shelves pale
Sunlit library reading room with tall shelves pale

How It Works

Under the claim-construction approach this guide tests — and which you should confirm against the live 2025 UPC Court of Appeal record before relying on it — the mechanism is a verify-before-you-commit loop: you read the live, complete claim language first, then test the accused embodiment against each construed limitation before choosing where and how to plead.

The single “comprising” limitation this section uses is: “A writing instrument, comprising a barrel and a cap,” drawn from the claim discussed in the Patent Infringement and Validity Analysis materials attributed to U.S. Patent No. 5,123,456 to Green et al. That one transitional phrase sets the floor for the comparison; it does not, by itself, close the analysis.

“Comprising” is an open transition. In ordinary patent practice, it means the named elements are required but additional elements may also be present. “Consisting of” would be closed, and “consisting essentially of” sits between the two. The practical check is to confirm the exact transitional word in the live claim text before you map any product feature to it, because swapping one transition for another changes the infringement question without changing the rest of the sentence.

Claim construction is the step that fixes what those words mean for the infringement analysis. Lumenci’s 2026 guide on infringement contentions describes claim construction and infringement contentions as a feedback loop, which is the useful way to run this: the construed meaning tells you what to allege, and the detail of the allegation tells you whether your construction is doing the work you assumed. You verify the complete option—full claim, full construction, full accused feature set—before you commit to a division-specific theory.

That matters because the same “comprising” limitation can read differently depending on how a tribunal treats the open transition in context. One division’s analysis may treat the listed barrel and cap as the required minimum and let an extra component sit outside the infringement question; another may focus on how the accused instrument’s additional structure interacts with the construed limitation before reaching the same result. Whether Munich and Düsseldorf actually diverge on the 2025 appeal is a fact to verify against the UPC record, not assume.

The check is like-for-like: compare the identical claim term, the identical construed scope, and the identical accused feature set on both sides rather than comparing a broad summary from one with a narrow reading from the other.

Direct infringement, as framed in the Patent Infringement Demand Letters material under 35 U.S.C. § 271(a), covers making, using, selling, offering to sell, or importing the patented invention without authority. Use that as a separate verification layer: once the construction is fixed, confirm which act is actually at issue and whether the accused instrument meets every construed element of the claim as written. The open “comprising” transition lowers the bar for extra features, but it does not remove the need to account for the barrel and the cap.

The working rule is simple to state and strict to apply: verify the live claim, confirm the transition, lock the construction, then run the accused product through the same element-by-element check in each forum’s framing. If the totals or terms differ between the two records, you are not comparing like for like, and the analysis is not ready to commit.

How It Works — How courts read 'comprising'

Key Factors to Consider

Before you commit to a Munich or Düsseldorf theory, lock down three decision criteria in the order a reviewing panel would test them. First, capture the live, complete claim text rather than a summary or an earlier prosecution version. Second, confirm that every limitation in that text has a construed meaning you can apply to the accused embodiment. Third, check that the same construction holds when you move the analysis across divisions, because a single “comprising” limitation can change which infringement theory survives. This is the verify-before-you-commit discipline: compare like-for-like language and like-for-like element mapping before you pick a forum theory.

The first criterion is source integrity. Slideshare’s “Patent Infringement And Validity Analysis” deck uses U.S. Patent No. 5,123,456 to Green et al. with a claim beginning “A writing instrument, comprising...” as its teaching example. That matters here only as a reminder of the check: pull the operative claim, not a paraphrase. If your working copy drops a comma, a dependency, or a “wherein” clause, your element map is no longer like-for-like. Record the exact claim identifier and the exact claim text you tested, then carry that same text into the division comparison.

The second criterion is construction coverage. Lumenci’s “Patent Infringement Contentions: Guide for 2026” describes infringement contentions as formal written disclosures that set out, in specific detail, how the accused product, process, or technology allegedly infringes one or more asserted claims, and frames claim construction and contentions as a feedback loop. Use that loop as a checklist: for each construed term, write the accused counterpart beside it and mark whether the mapping is supported. If one “comprising” limitation remains unconstrued, you do not yet have a complete option to commit to.

The third criterion is divisional consistency. Lexology’s discussion of the draft Interpretation notes that it spans 31 articles covering areas including jurisdictional rules, standing to sue, claim construction, and infringement analysis. Treat those as separate gates. A construction that clears the claim-language gate still has to clear the infringement-analysis gate in the division you choose. When the open-ended “comprising” limitation is read one way, your element set may be satisfied; read another way, the same accused embodiment may fall short. Verify which reading the division has adopted before you commit.

The numbers that matter are few and should be used as anchors, not shortcuts. The 31-article scope reported by Lexology tells you how broad the interpretive framework is; it does not tell you which article controls your limitation. The 2026 contentions guidance from Lumenci tells you the disclosure standard is specific detail, not a label. The Green et al. patent number, U.S. Patent No. 5,123,456, identifies the teaching example in the Slideshare deck; it is not a substitute for your own asserted claim. Where a source gives no figure for a fee, deadline, or threshold, leave that cell blank in your check rather than filling it with an estimate.

Your final pre-commit check is a one-page matrix with three columns: exact claim language, construed limitation, accused counterpart. Add a fourth row only if you need to note the division whose construction you verified. Compare the matrix for Munich against the matrix for Düsseldorf using the same claim text and the same accused embodiment. If either matrix relies on a different claim version, a missing limitation, or an unverified construction of “comprising,” stop and repair the record before committing.

Common Mistakes

Most missteps here are not about the word “comprising” in isolation. They come from treating a single open-ended limitation as if it carries the same weight in every division. The practical check is to read the live, complete claim text, then ask which construed limitation actually does the infringement work. As the Slideshare deck on patent infringement and validity analysis puts it, infringement turns on claim construction; a writing instrument claim framed as “comprising” still has to be tested element by element, not as a general concept.

Pitfall 1 is substituting a summary for the operative text. A team will work from an earlier prosecution version, a patent-family abstract, or a contentions summary and assume the open transition covers an extra component in the accused embodiment. Lumenci’s guide to infringement contentions describes those contentions as formal written disclosures that set out, in specific detail, how the accused product, process, or technology allegedly infringes. That detail is the point: if the live claim says “comprising a barrel and a cap,” the check is whether the accused article has those construed limitations, not whether it also has a grip, clip, or sensor. Adding an unrecited feature does not repair a missing recited one.

Here is the concrete example this section uses. Assume the live limitation reads: “A writing instrument, comprising a barrel and a cap configured to seal the barrel.” One accused pen has a barrel, a cap that seals it, and an added grip. A second has a barrel and a textured grip but no cap that performs the sealing function. Under an open “comprising” construction, the first embodiment still has the recited barrel and sealing cap; the extra grip is additional matter. The second fails the cap limitation even though the grip may look commercially important. The verify step is to map each limitation to a feature on the accused article before choosing a forum theory.

Pitfall 2 is letting the label of the division drive the construction. Divisional theories can diverge when one side reads the open transition broadly and the other treats a functional phrase as a tight constraint; whether Munich and Düsseldorf in fact diverge on the 2025 appeal should be confirmed against the UPC record before the label is used as an argument.

The Lexology discussion of the draft Interpretation (III) notes that claim construction sits alongside jurisdictional rules and infringement analysis as a distinct critical area. Keep those buckets separate: first construe, then apply, then select the theory. If the sealing-cap phrase is construed functionally, the question is whether the accused cap actually seals the barrel, not whether the parties call the part a cap.

A third recurring error is mixing validity and infringement records. Prior-art comparisons belong to patentability, while infringement asks whether the accused embodiment meets the construed claim; the Slideshare material keeps those analyses distinct. A reference that makes a claim look crowded may still leave the infringement mapping straightforward. Conversely, a strong infringement map does not cure a construction problem. Keep separate notes for each, and carry forward only the live claim language and the construed limitations into the verify-before-you-commit check.

The working rule is simple: compare like for like. Use the same claim text, the same construed limitation, and the same accused feature set on both sides of the analysis. If one theory relies on a summary, an omitted component, or a different construction, it is not a like-for-like comparison. Re-run the mapping against the complete option before committing to a Munich or Düsseldorf infringement theory.

Insider Tactics

The core tactical move is to treat the claim-construction exchange as a live feedback loop, not a one-way filing.

Lumenci’s Patent Infringement Contentions: Guide for 2026 describes claim construction and infringement contentions as a feedback loop, and that framing is useful here: before you commit to a Munich or Düsseldorf theory, pressure-test the construed limitation against the accused embodiment and then revisit the contention language while the record is still flexible.

A non-obvious strategy is to draft your infringement contention around the narrowest plausible reading of the disputed “comprising” limitation, even if your broader theory is more aggressive. The slideshare deck Patent Infringement And Validity Analysis uses a “writing instrument, comprising…” claim to underscore why construction drives the infringement comparison. If your proof survives the narrower reading, you preserve a cleaner path when a panel adopts a stricter construction; if it does not, you learn that the broader theory carries construction risk you may not want to carry into a divisional choice.

Another tactic is to keep a parallel “construction-dependent” and “construction-neutral” version of the same contention. The construction-dependent version maps the accused feature to the disputed term as your side construes it. The construction-neutral version lists only the undisputed elements and the technical facts that remain true under either reading. That lets you see, before commitment, whether the case turns on the single limitation or whether the infringement story is robust enough to travel across divisions without rewriting the factual record.

Timing matters because contentions are formal written disclosures that set out, in specific detail, how the accused product, process, or technology allegedly infringes asserted claims, as Lumenci explains. Use that detail requirement as a timing check: serve or lock the detailed mapping only after you have verified the live, complete claim text and tested the accused embodiment against the construed limitation. If you wait until after the detailed mapping is fixed, a later construction shift can force you to explain why the same facts now support a different theory.

A practical timing tip is to run the verification step at two points: once when the claim language is first captured, and again immediately before the infringement theory is committed. The second pass should compare like-for-like terms, not just like-for-like products. If the construed limitation changes, reread the accused feature against the new wording rather than carrying forward the earlier conclusion. That discipline is especially useful when a single open-ended limitation is the hinge between theories.

Finally, treat the divisional choice as the last step, not the first. Munich and Düsseldorf should be evaluated only after the live claim, the construed limitation, and the accused embodiment have been aligned in writing. Lexology’s discussion of the draft Interpretation (III) notes that claim construction and infringement analysis sit among the core areas of patent enforcement, which reinforces the practical point: construction is not housekeeping. It is the gate you pass through before the infringement theory, and the division, become commitments.

Comparison

This section compares the two divisional options side by side, on the working assumption — to be verified against the UPC Court of Appeal’s 2025 record — that a single “comprising” limitation can drive different infringement readings.

The comparison below assumes you have already isolated the live claim language and are now choosing which divisional theory to commit to on like-for-like terms: the same claim text, the same accused embodiment, and the same disputed limitation.

Option Like-for-like comparison point When this option wins
Munich-leaning infringement theory Map 1 accused feature set against 1 “comprising” limitation and document whether additional unclaimed elements are present. Wins when the accused embodiment clearly satisfies every required element of that limitation and any extra features do not negate the match.
Düsseldorf-leaning infringement theory Map the same 1 accused feature set against the same 1 limitation, focusing on whether the full context narrows how that element must be embodied. Wins when the surrounding claim context or specification support shows the disputed limitation is met only under a narrower reading that the accused product still satisfies.

The Munich-leaning option is the stronger commitment when your side-by-side review shows a direct, element-by-element correspondence. As general infringement analysis principles emphasize, the accused product or process must be compared against the construed claim language (Slideshare, “Patent Infringement and Validity Analysis”). If the contested “comprising” limitation is plainly present and the remaining differences are immaterial additions, this path offers the cleaner alignment between claim scope and proof.

The Düsseldorf-leaning option wins when the same comparison reveals that the dispute turns on the boundaries of that single limitation rather than its basic presence. In that situation, the prevailing theory is the one that ties the accused embodiment to a defensible reading of the limitation in context, consistent with the feedback loop between claim construction and infringement contentions described in practice guidance (Lumenci, “Patent Infringement Contentions: Guide for 2026”).

To declare a winner before committing, place both theories in one matrix and test them against identical evidence. The option that maintains a consistent reading of the “comprising” limitation across the claim, the accused embodiment, and your supporting materials is the one to advance; if neither does, the comparison signals that further verification is required before filing or forum selection.

What to do next

StepActionWhy it matters
1Verify the live, complete Munich or Düsseldorf infringement option before committing to either division-specific theory.A single “comprising” limitation can shift infringement analysis between UPC divisions.
2Compare like-for-like totals and terms before relying on one “comprising” limitation.It prevents treating one division’s reading as the complete option.
3Re-check claim construction against infringement contentions through the “Claim Construction and Infringement Contentions: A Feedback Loop” framing.Claim construction remains central to infringement analysis.
4Review Interpretation (III) across its 31 articles on claim construction and infringement analysis.It covers the construction-infringement interface in the draft.
5Confirm the jurisdictional coverage among those 31 articles before applying the theory.The draft includes jurisdictional material that may affect the division route.
6Test the theory against the 2025 UPC Court of Appeal claim-construction standard.The thesis condition is that this standard changes the division-specific infringement route.

Frequently Asked Questions

What must be verified before committing to an infringement theory in Munich or Düsseldorf?

The live, complete option must be verified before committing to a Munich or Düsseldorf infringement theory.

How should terms and totals be evaluated when relying on a 'comprising' limitation?

Readers should compare like-for-like totals and terms before relying on a single “comprising” limitation.

Can a single 'comprising' limitation affect which UPC division's infringement analysis applies?

A single “comprising” limitation can shift infringement analysis between UPC divisions such as Munich and Düsseldorf.

What condition must be met before relying on the thesis about division-specific infringement routes?

The working thesis should be verified against the actual 2025 UPC Court of Appeal record before reliance.

How many articles does Interpretation (III) span, and what topics does it cover?

Interpretation (III) spans 31 articles covering claim construction and infringement analysis.

What specific topics are included in the draft that comprises 31 articles?

The draft comprises 31 articles, including jurisdictional rules, standing to sue, claim construction, and infringement analysis.

Quick answers

What should be verified before committing to a Munich or Düsseldorf infringement theory?Verify the live, complete option before committing to a Munich or Düsseldorf infringement theory.
What is the reader rule before relying on a single “comprising” limitation?Compare like-for-like totals and terms before relying on a single “comprising” limitation.
What condition is stated in the thesis?The 2025 UPC Court of Appeal claim-construction standard changes the division-specific infringement route.
How do courts read “comprising”?Courts read “comprising” through claim construction.
When does the answer change?The answer changes when the live, complete option is not verified against like-for-like totals and terms before commitment.

Also worth reading: Johnson and Johnson faces first CAR T patent battle at the UPC: Johnson and Johnson faces first · UPC vs EPO: 2026 Software Patent Validity & Claim Divergence: UPC vs EPO: 2026 Software · Artificial intelligence streamlines patent analysis workflow: Artificial intelligence streamlines patent analysis

Premium Deals
Mighty Travels Premium
Travel in style,
save up to 90%

On flights and hotels worldwide by booking the best deals when they appear.

See Deals

Sponsored

Research Methodology & Editorial Standards

We begin by defining the specific objectives the reader needs to accomplish. Primary product documentation and authoritative secondary sources are assembled into a verified research corpus; drafting occurs only after this foundation is in place.

Every quantitative claim is subjected to dual-source verification. Any figure that cannot be independently corroborated is either qualified or omitted.

Published · Last reviewed · Owned by the Patentreviewpro editorial desk (About, Contact, Privacy).

Related answers