Checklist · Intellectual Property

Claim Construction and the Markman Hearing: A Practical Checklist

Type
Checklist
Jurisdiction
National
Date
2024-05-08
Topic
Intellectual Property

Claim construction is the hinge on which most patent cases turn: the court's constructions of a handful of disputed terms routinely decide infringement, validity, or both, and they do so long before trial. This checklist walks the full workflow — starting from the case theory rather than the claim language, selecting the small number of terms that actually matter, building the intrinsic record of claims, specification, and prosecution history under Phillips v. AWH Corp., and applying the evidentiary hierarchy that keeps extrinsic evidence in its subordinate place. It covers the two escape hatches from ordinary meaning (lexicography and disavowal), the cardinal sin of importing limitations from the specification, prosecution disclaimer, means-plus-function terms under 35 U.S.C. § 112(f) after Williamson v. Citrix, and indefiniteness under Nautilus. It then turns to procedure: patent local rule exchanges, the joint claim construction statement, the Markman brief, the technology tutorial, and the hearing itself. Finally, it addresses what to do with the ruling — stipulated judgment, summary judgment, and preserving the issue for de novo appellate review under Teva v. Sandoz. A running hypothetical involving a disputed term that may or may not require a physical connection shows how the pieces fit together.

Claim Construction and the Markman Hearing: A Practical Checklist

Intellectual Property > Patent Litigation | Published 8 May 2024 - Updated 30 September 2025 | UpLaw - uplaw.us

Patent cases are won and lost on a few words. Claim construction is where a judge decides what those words mean — and once the meaning is fixed, the infringement and validity arguments that seemed so promising in the complaint either survive or quietly die.

Keywords: claim construction, Markman hearing, Phillips v. AWH, intrinsic evidence, prosecution disclaimer, lexicography, disavowal, means-plus-function, Section 112(f), indefiniteness, Nautilus, Teva v. Sandoz, joint claim construction statement, technology tutorial, de novo review

The one thing to understand before you start

Claim construction is a question of law for the court. Markman v. Westview Instruments, Inc., 517 U.S. 370 (1996), took claim interpretation away from the jury and gave it to the judge, and the Federal Circuit's en banc decision in Phillips v. AWH Corp., 415 F.3d 1303 (Fed. Cir. 2005), supplies the methodology. The practical consequence is enormous: the single most outcome-determinative event in a patent case happens in a courtroom with no jury, no witnesses (usually), and a record that you largely built years earlier during prosecution.

Everything in this checklist follows from that. If you want a broader orientation before diving in, start with The Utility Patent Toolkit: From Invention to Enforcement and Litigating Patent Infringement: A Roadmap and Research Guide.

Our running example. Thornbury Dynamics owns a patent on an industrial vibration-damping mount. Claim 1 recites a "sensor coupled to the damping element." Vantail Systems sells a competing mount whose sensor sits three inches away and communicates with the damping element wirelessly. Thornbury sues. The entire case reduces to one question: does "coupled to" require a physical connection? If yes, Vantail wins on non-infringement. If no, Thornbury may win on infringement — but Vantail's prior art, which discloses wirelessly linked sensors, suddenly becomes anticipatory. That symmetry is the essence of claim construction, and it is why you cannot construe terms in isolation from the case theory.

Phase 1 — Start from the case theory: which terms actually decide infringement or validity?

  • [ ] Draft (or update) preliminary infringement contentions and invalidity contentions before you pick terms.
  • [ ] Build a claim chart mapping every limitation of every asserted claim to the accused product and to each primary prior-art reference.
  • [ ] For each limitation, ask: Is there a genuine dispute about what this limitation covers, and does resolving it change the outcome?
  • [ ] Identify the "case-dispositive" terms — usually one to three — and rank the rest.
  • [ ] Run the validity-side test on every proposed construction: if I win this construction, does the claim read on the prior art or become indefinite?
  • [ ] Confirm the claim scope you need is consistent with your damages theory and your willfulness/notice theory.

Why it matters, and the traps. Claim construction is not an academic exercise in lexicography; it is advocacy toward a result. Every construction you propose is simultaneously an infringement argument and a validity argument, and the two pull in opposite directions. Broaden "coupled to" to capture Vantail's wireless sensor and you invite Vantail's prior art. Narrow it to exclude the prior art and you lose infringement. This is the classic patentee's vise, and the Federal Circuit has been unsympathetic to litigants who try to have it both ways. The governing principle is that a claim means the same thing for infringement as it does for validity — a patent cannot be twisted like a "nose of wax," read broadly to capture the accused product and narrowly to dodge the prior art. Amazon.com, Inc. v. Barnesandnoble.com, Inc., 239 F.3d 1343 (Fed. Cir. 2001), applies that principle in the preliminary-injunction posture, and Phillips assumes it throughout.

A related point of orientation: claim construction is a matter of federal patent law, and the Phillips methodology applies uniformly in every district and on appeal to the Federal Circuit. That uniformity is unusual — it is the opposite of the state-by-state variation that governs, say, insurance coverage or corporate fiduciary duty (where Delaware is the customary reference point, not a national rule). What does vary in claim construction is procedure: patent local rules, term caps, tutorial formats, and hearing practice differ materially from district to district and judge to judge. Learn the doctrine once; re-learn the local rules every case.

The trap is sequencing. Lawyers who read the claims first and the accused product second end up with elegant constructions that decide nothing. Read the accused product first. For the substantive framework you are serving, see What Counts as Patent Infringement: An Anatomy of the Section 271 Analysis and Patent Infringement, Claims, and Defenses: A Practical Guide for Businesses and Counsel.

Phase 2 — Select terms strategically (construing everything is a losing strategy)

  • [ ] Check the court's local patent rules and standing order for a numerical cap on disputed terms (many districts limit the parties to ten, and judges frequently cut further).
  • [ ] Rank your terms: (1) case-dispositive, (2) materially narrowing, (3) housekeeping, (4) vanity.
  • [ ] Cut categories (3) and (4). Then cut half of (2).
  • [ ] For each surviving term, write the one-sentence reason a construction is necessary — i.e., what the parties actually dispute and why the jury cannot be left to sort it out.
  • [ ] Decide, for each non-selected term, whether "plain and ordinary meaning" is genuinely safe or whether you are ducking a fight you will lose later at trial.
  • [ ] Preserve indefiniteness challenges separately — do not bury a § 112(b) argument inside a construction dispute.

Why it matters, and the traps. Judges have limited patience and limited technical bandwidth. A party that shows up with fourteen disputed terms tells the court two things: it does not know what its case is about, and it is hoping something sticks. A party that shows up with two tells the court exactly where to look. Credibility in a Markman proceeding is a finite resource; spend it on the terms that matter.

The counter-trap is under-designation. Under O2 Micro International Ltd. v. Beyond Innovation Technology Co., 521 F.3d 1351 (Fed. Cir. 2008), when the parties present a fundamental dispute about claim scope, the court must resolve it — a bare instruction to apply "plain and ordinary meaning" is legal error if it leaves the scope question to the jury. But O2 Micro is a sword, not a shield: if you fail to raise the dispute at the Markman stage, you may find at trial that the court refuses to hear it, and your opponent will argue waiver. Raise the dispute, or lose it.

Worked example. Thornbury's team initially listed eight terms, including "damping element," "housing," and "substantially planar." Only "coupled to" changes any outcome. They drop the rest and open their brief with a single sentence: "This case turns on one word, and the specification defines it." Vantail, by contrast, lists all eight — and spends its page limit defending "substantially planar" instead of winning on "coupled to."

Phase 3 — Build the intrinsic record: claims, specification, prosecution history

  • [ ] Assemble the complete intrinsic record: the patent as issued, every claim (asserted and unasserted), the specification and drawings, the full prosecution history including all office actions, responses, interviews, and notices of allowance.
  • [ ] Pull the prosecution histories of every related application: parents, continuations, divisionals, and foreign counterparts where the same term was discussed.
  • [ ] Pull any post-grant record: reexamination, ex parte or inter partes review filings, and patent owner responses. Statements made in an IPR preliminary response can create disclaimer. See Aylus Networks, Inc. v. Apple Inc., 856 F.3d 1353 (Fed. Cir. 2017).
  • [ ] Collect every prior-art reference cited on the face of the patent — cited art is intrinsic-adjacent and often decisive on what a term meant to a person of ordinary skill.
  • [ ] Chart every appearance of each disputed term across the claims, the abstract, the summary, the detailed description, and the file wrapper.
  • [ ] Note claim differentiation: does a dependent claim add the very limitation your opponent wants read into the independent claim?

Why it matters, and the traps. Phillips is emphatic that the intrinsic record is where claim construction begins and, in the overwhelming majority of cases, ends. Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576 (Fed. Cir. 1996), had already established that if the intrinsic evidence resolves the ambiguity, it is improper to rely on extrinsic evidence.

The most common failure here is incompleteness. Counsel reads the patent and the office actions in the litigated file and stops. The killer admission is almost always in the parent's file, or in a foreign prosecution, or in a patent owner's preliminary response before the PTAB. If your client has a co-pending inter partes review, coordinate your positions across the two forums with obsessive care — the arguments you make to survive institution can be quoted back at you in district court. See Filing an Inter Partes Review Petition: A Practical Checklist.

Phase 4 — Apply the Phillips hierarchy and the ordinary-meaning presumption

  • [ ] Start with the ordinary and customary meaning the term would have to a person of ordinary skill in the art (POSITA) at the time of the invention, reading the claim in the context of the entire patent.
  • [ ] Define the POSITA — education, experience, and the relevant date (usually the effective filing date). This is a factual predicate and it matters.
  • [ ] Work the hierarchy in order: claim language → specification → prosecution history → extrinsic evidence.
  • [ ] Read the term in context of the claim itself: surrounding words, the preamble, other claims (claim differentiation), and the antecedent-basis chain.
  • [ ] Consult the specification — Phillips calls it "the single best guide to the meaning of a disputed term."
  • [ ] Consult the prosecution history, remembering it is an ongoing negotiation and therefore "often lacks the clarity of the specification."
  • [ ] Only then, and only to the extent needed, consider extrinsic evidence.

Why it matters, and the traps. Phillips was written to bury a methodology, not just to state one. The court expressly disapproved the dictionary-first approach of Texas Digital Systems, Inc. v. Telegenix, Inc., 308 F.3d 1193 (Fed. Cir. 2002), which had elevated abstract dictionary definitions over the patent's own teaching. The Phillips insight is that a patent is a document written by a person of skill for people of skill, and the specification is where the inventor tells you what the words mean.

The trap is treating the hierarchy as a checklist to be marched through rather than a set of nested contexts. "Coupled to" has an ordinary meaning in electrical engineering that expressly includes indirect and even wireless coupling. That is the starting presumption for Thornbury. Vantail's job is to find something in the intrinsic record that overcomes it — and only two things can. That is Phase 5.

Phase 5 — Lexicography and disavowal: the two ways to escape ordinary meaning

  • [ ] Search the specification for explicit definitional language: "as used herein," "means," "is defined as," "refers to," "i.e.," parenthetical glosses.
  • [ ] Test lexicography: did the patentee (a) clearly set forth a definition, and (b) clearly express an intent to redefine the term?
  • [ ] Search for disavowal: statements that the invention "is," "must," "always," "never," "requires," or that distinguish the prior art as lacking the feature; statements that the invention overcomes a problem by the very feature at issue.
  • [ ] Distinguish disparagement of an embodiment (usually not disavowal) from disparagement of the approach (often disavowal).
  • [ ] Apply the exacting standard: both lexicography and disavowal require a clear and unmistakable expression. Thorner v. Sony Computer Entertainment America LLC, 669 F.3d 1362 (Fed. Cir. 2012).
  • [ ] If neither applies, say so loudly — the ordinary meaning controls, full stop.

Why it matters, and the traps. Thorner is the cleanest statement of the rule: there are exactly two exceptions to ordinary meaning, and both demand clarity. This is the doctrine that most often saves a patentee from a narrowing construction — and the doctrine most often abused by defendants who mistake a preferred embodiment for a definition.

Worked example. Vantail finds this sentence in Thornbury's specification: "In the invention, the sensor is coupled to the damping element by a rigid mechanical linkage, which is essential to transmitting vibration data without latency." That is a strong disavowal argument — "the invention," "essential," and a stated technical reason. Compare SciMed Life Systems, Inc. v. Advanced Cardiovascular Systems, Inc., 242 F.3d 1337 (Fed. Cir. 2001), where the specification's statement that all embodiments of "the present invention" used a coaxial lumen structure limited the claims accordingly. Thornbury's best answer is that the sentence describes a preferred embodiment and that the specification elsewhere states that "coupling may be direct or indirect." Whoever wins that fight wins the case.

Phase 6 — The cardinal sin: importing limitations from the specification

  • [ ] Ask, for every proposed narrowing: am I using the specification to understand the term, or importing a limitation from an embodiment into the claim?
  • [ ] Check whether the specification describes only one embodiment. That alone does not limit the claims.
  • [ ] Check whether the proposed construction would exclude a disclosed embodiment. That is "rarely, if ever, correct."
  • [ ] Check whether the applicant could have claimed the limitation and did not — and whether a dependent claim already recites it.
  • [ ] Ask whether the narrowing is really an unstated attempt to rescue the claim from invalidity. Courts may construe claims to preserve validity only where the term remains genuinely ambiguous after the full analysis.

Why it matters, and the traps. Phillips itself frames this as the central difficulty: "the line between construing terms and importing limitations from the specification into the claims can be discerned with reasonable certainty and predictability if the court's focus remains on understanding how a person of ordinary skill in the art would understand the claim terms."

The doctrinal anchor for patentees is Liebel-Flarsheim Co. v. Medrad, Inc., 358 F.3d 898 (Fed. Cir. 2004): a claim is not limited to the preferred embodiment even if that is the only embodiment disclosed. But there is a sting in the tail — on remand, the broadly construed claims were held invalid for lack of enablement. Liebel-Flarsheim Co. v. Medrad, Inc., 481 F.3d 1371 (Fed. Cir. 2007). "You get the claim scope you asked for, and the invalidity that comes with it" is the whole lesson of that litigation. Argue for breadth with your eyes open.

Phase 7 — Prosecution disclaimer and its effect on scope

  • [ ] Read every argument the applicant made to overcome a rejection. What did the applicant say the prior art lacked?
  • [ ] Read every claim amendment. Was it made to overcome art (narrowing amendment), or for a formal reason?
  • [ ] Apply the standard: prosecution disclaimer requires a clear and unmistakable surrender of subject matter. Omega Engineering, Inc. v. Raytek Corp., 334 F.3d 1314 (Fed. Cir. 2003).
  • [ ] Distinguish prosecution disclaimer (a claim-construction doctrine that narrows literal scope) from prosecution history estoppel (an equitable doctrine that bars the doctrine of equivalents for surrendered territory). See Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 535 U.S. 722 (2002).
  • [ ] Sweep post-grant statements: reexamination, reissue, and PTAB filings all count.
  • [ ] If you are the patentee, audit your own prosecution history before your opponent does, and build the counter-narrative.

Why it matters, and the traps. Disclaimer is where cases die quietly. In our hypothetical, Thornbury's prosecution history contains a response distinguishing a prior-art reference on the ground that "the cited reference discloses a sensor in wireless communication with, but not coupled to, the damping element." That single clause may be a clear and unmistakable disclaimer of wireless coupling — and it is worth more to Vantail than any expert report. This is why Phase 3 insists on the complete file, and why prosecution counsel should read Turning Patent Rejections Into Granted Claims: A Prosecution Response Guide and Responding to a Patent Office Action: A Practical Checklist with litigation in mind. Every sentence written to a patent examiner is a sentence a district judge may one day read aloud. For the prosecution side of that coin, see Prosecuting a Patent Application: A Roadmap and Research Guide.

Phase 8 — Extrinsic evidence: dictionaries, treatises, and experts (and their limits)

  • [ ] Identify what extrinsic evidence, if any, you actually need — usually only to establish the POSITA's technical background or the meaning of a genuine term of art.
  • [ ] Prefer technical dictionaries and treatises contemporaneous with the filing date over general-purpose dictionaries.
  • [ ] If you use an expert declaration, tie it to the intrinsic record; conclusory declarations that a term "means what my client says" carry no weight.
  • [ ] Confirm the court's local rules on whether expert declarations are permitted at the Markman stage and whether Markman-phase experts may be deposed.
  • [ ] Be prepared to defend your expert's qualifications and methodology — see Fed. R. Evid. 702 and Challenging an Expert Under Daubert and Rule 702: A Practical Checklist.
  • [ ] Remember the ceiling: extrinsic evidence can never contradict an unambiguous intrinsic record.

Why it matters, and the traps. Phillips catalogued why extrinsic evidence is "less significant": it is not part of the patent, dictionaries capture the full universe of a word's meanings rather than the one the patentee used, experts are hired, and there is an unbounded universe of it. Yet extrinsic evidence has one genuinely important role after Teva — it is the only route to a factual finding that gets deference on appeal (Phase 13).

The trap is over-reliance. A brief that leads with a Merriam-Webster definition of "coupled" signals that the intrinsic record does not help you. Lead with the specification. And note the evidentiary posture: the Markman record is not a trial record, and the ordinary rules of admissibility apply loosely, but sloppy foundation still costs credibility. See The Federal Rules of Evidence: A Comprehensive Guide.

Phase 9 — Means-plus-function terms under Section 112(f) and the corresponding-structure requirement

Why it matters, and the traps. § 112(f) is a bargain: the applicant gets to claim by function, and in exchange the claim covers only the disclosed structure and its equivalents. Williamson dismantled the previously "strong" presumption against § 112(f) for claims lacking the word "means," and in doing so turned thousands of software claims — full of "modules" and "units" — into indefiniteness targets. A defendant facing a software patent should run the Williamson screen on every claim as a matter of routine — and should understand how the patent claim interacts with the other regimes protecting the same code. See How the Law Protects Software: Copyright, Patents, Trade Secrets, and Contract Rights.

The trap for patentees is thinking § 112(f) is a mere construction issue. It is often a validity issue in disguise: identify the function, ask what algorithm the specification discloses, and if the answer is "a flowchart box labeled 'compute result,'" the claim is likely dead.

Phase 10 — Indefiniteness under Nautilus: reasonable certainty

  • [ ] Apply the standard: a claim is indefinite if, read in light of the specification and the prosecution history, it fails to inform those skilled in the art about the scope of the invention with reasonable certainty. Nautilus, Inc. v. Biosig Instruments, Inc., 572 U.S. 898 (2014); 35 U.S.C. § 112(b).
  • [ ] Do not apply the old "insolubly ambiguous" or "amenable to construction" test — Nautilus rejected it.
  • [ ] Target the usual suspects: terms of degree ("substantially," "about," "unobtrusive"), subjective terms ("aesthetically pleasing," "user-friendly"), relative terms without a reference point, and means-plus-function terms lacking structure.
  • [ ] For a term of degree, ask whether the specification supplies an objective baseline for measurement. Terms of degree are not per se indefinite; unanchored ones are. Interval Licensing LLC v. AOL, Inc., 766 F.3d 1364 (Fed. Cir. 2014).
  • [ ] Remember the burden: invalidity, including indefiniteness, must be proved by clear and convincing evidence. 35 U.S.C. § 282; Microsoft Corp. v. i4i Ltd. Partnership, 564 U.S. 91 (2011).
  • [ ] Decide when to raise indefiniteness — in the Markman briefing (most districts), or on summary judgment, or both. Check the local rules.

Why it matters, and the traps. Indefiniteness is claim construction's twin. It is decided on the same intrinsic record, usually at the same hearing, and it is the only claim-construction-adjacent outcome that kills the claim outright rather than merely narrowing it. The trap is treating it as an afterthought: a party that argues "this term is indefinite, but if the Court construes it, it should mean X" must be careful that its alternative construction does not prove the term is, in fact, reasonably certain.

Phase 11 — The exchange of terms and constructions; the joint claim construction statement

  • [ ] Read the patent local rules for your district (N.D. Cal., E.D. Tex., D. Del., and W.D. Tex. all have their own, and judges have standing orders on top). Calendar every date.
  • [ ] Exchange proposed terms for construction by the deadline. Terms not disclosed are typically waived.
  • [ ] Exchange preliminary proposed constructions together with the intrinsic and extrinsic evidence supporting each — including identification of any expert and a summary of the opinions.
  • [ ] Meet and confer in good faith to narrow the disputes. Actually narrow them; a judge can tell the difference between a real meet-and-confer and a ritual one.
  • [ ] File the joint claim construction and prehearing statement: agreed constructions, disputed terms with each side's proposal, supporting evidence, the anticipated length of the hearing, and whether live testimony is requested.
  • [ ] Coordinate the Markman schedule with fact discovery — you will need the accused product's technical documents. See Serving Written Discovery Requests: A Practical Checklist.
  • [ ] Confirm whether your court decides claim construction before or with summary judgment; the answer drives your entire pretrial strategy.

Why it matters, and the traps. Patent local rules are the closest thing patent litigation has to a procedural constitution, and they are enforced. The most expensive mistake in this phase is a late-disclosed construction or a late-disclosed piece of intrinsic evidence: courts routinely strike them. The second most expensive is stipulating to a construction that sounds harmless and turns out to concede the case — read every agreed construction against your claim chart before you sign the joint statement.

Phase 12 — The Markman brief, the technology tutorial, and the hearing

  • [ ] Structure the brief term by term, not argument by argument. For each term: the claim language in context → the specification → the prosecution history → (only if needed) extrinsic evidence.
  • [ ] Quote the intrinsic record. Do not paraphrase it. Judges reading a Markman brief want to see the words.
  • [ ] Give the court a clean, adoptable construction — a phrase the judge can paste into an order and a jury can understand. "A sensor joined to the damping element by a physical structure that transmits vibration" beats a three-clause construction with two provisos.
  • [ ] Include a claim-term appendix and, where the technology allows, an annotated figure.
  • [ ] Prepare the technology tutorial: neutral, jargon-free, roughly 20–30 minutes, focused on how the technology works — not on why you win. Confirm the court's format (live, video, joint, or separate) and whether it is evidence or merely background.
  • [ ] Prepare the hearing itself: know your two best record cites for each term cold, prepare for a hot bench, and be ready to be asked "what is your fallback construction?"
  • [ ] Have an answer to the O2 Micro question: if the court adopts plain and ordinary meaning, what exactly does the jury do with the dispute?
  • [ ] If live testimony is allowed, prepare the expert for cross on the intrinsic record, not just on the technology.

Why it matters, and the traps. The tutorial is the most underrated document in patent litigation. It is your first and best chance to make the judge comfortable with the technology, and comfort produces confidence, and confidence produces adoption of your construction. A tutorial that is transparently argumentative squanders it.

The other trap is over-engineering the construction. Judges reject constructions that read like claim amendments. Ask yourself: could a juror apply this sentence to the accused product without a lawyer's help? If not, simplify.

Worked example. Thornbury proposes "connected, directly or indirectly, such that vibration data is communicated." Vantail proposes "physically joined." The court, after the tutorial, adopts a middle construction: "joined by a structure that transmits mechanical vibration, directly or through intervening components." Vantail's wireless sensor does not infringe. The case is over — and it was over at the hearing, not at trial.

Phase 13 — After the ruling: stipulated judgment, summary judgment, and preserving appeal (Teva and de novo review)

  • [ ] Re-run the claim charts against the actual constructions the same week the order issues. Do not wait.
  • [ ] Decide the endgame: (a) stipulated judgment of non-infringement (preserving the appeal), (b) summary judgment, (c) trial, or (d) settlement.
  • [ ] If you lost dispositively, negotiate a stipulated judgment that expressly preserves the right to appeal the construction and that does not concede facts you need on appeal.
  • [ ] If the construction helps you, move for summary judgment promptly. See Moving for Summary Judgment Under Rule 56: A Practical Checklist and The Civil Pretrial Toolkit: From Pleadings to Summary Judgment.
  • [ ] Move for reconsideration or clarification only if the order is genuinely ambiguous or omits a term the parties disputed (an O2 Micro problem).
  • [ ] Preserve the issue: object to jury instructions that embed the construction, and renew objections in your Rule 50 motions.
  • [ ] Understand the standard of review. Under Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., 574 U.S. 318 (2015), the ultimate construction is reviewed de novo, but subsidiary findings of fact based on extrinsic evidence are reviewed for clear error under Fed. R. Civ. P. 52(a)(6). Constructions resting solely on the intrinsic record remain entirely de novo.
  • [ ] Plan appellate strategy accordingly — and see Federal Civil Appeals: A Comprehensive Guide.

Why it matters, and the traps. Teva changed the appellate calculus. Before it, Cybor Corp. v. FAS Technologies, Inc., 138 F.3d 1448 (Fed. Cir. 1998) (en banc), gave the Federal Circuit plenary review of everything. Now, a party that wins a factual finding below — say, what "coupled" meant to a POSITA in 2016, based on credited expert testimony — has armored that finding with clear-error review. The strategic implication is real: if you expect to win the construction and then defend it on appeal, building an extrinsic-evidence record and asking the district court for express factual findings can be worth more than a cleaner intrinsic argument. Conversely, if you expect to lose below, keep the dispute on the intrinsic record, where the Federal Circuit will look at it fresh.

The trap is a stipulated judgment that is too generous. Stipulate to non-infringement under the court's construction; never stipulate to non-infringement generally, and never stipulate away the facts you will need if the construction is reversed and the case comes back.

Common Mistakes to Avoid

  1. Constructing terms before understanding the accused product. Constructions are tools, not truths. Know what you need the term to mean before you argue what it means.
  2. Bringing fourteen terms to a hearing that can decide one. Judges read the number of disputed terms as a proxy for the quality of your case.
  3. Ignoring the parent's prosecution history. The admission that kills your client is rarely in the file you read first.
  4. Confusing a preferred embodiment with a definition. Thorner and Liebel-Flarsheim exist because this mistake is constant. Disavowal must be clear and unmistakable.
  5. Leading with a general dictionary. Phillips buried the dictionary-first approach. If your best evidence is Webster's, your best evidence is weak.
  6. Missing a Williamson trigger. "Module," "unit," and "mechanism" are not safe harbors from § 112(f) — and if there is no corresponding algorithm, the claim may be indefinite.
  7. Taking "plain and ordinary meaning" as a resolution. Under O2 Micro, an unresolved scope dispute handed to the jury is reversible error. Force the court to decide.
  8. Proposing a construction you cannot survive on validity. Every broadening argument is an invitation to the prior art. Chart both sides before you file.
  9. Treating the tutorial as argument. It is a teaching opportunity. Waste it and you will spend the hearing fighting the judge's confusion instead of your opponent.
  10. Blowing the appeal at the stipulation stage. A carelessly drafted stipulated judgment can waive the very issue you took the judgment to preserve.
  11. Forgetting design patents play by different rules. Detailed verbal claim construction is disfavored for design patents. Egyptian Goddess, Inc. v. Swisa, Inc., 543 F.3d 665 (Fed. Cir. 2008) (en banc); see Litigating Design Patents: From Claim Construction to Total-Profit Damages.
  12. Failing to think about the jury. The construction the court adopts will be read to a jury. If it cannot be applied without a lawyer, it will not be applied.

The consolidated claim-construction checklist

Strategy

  • [ ] Infringement and invalidity contentions charted before term selection
  • [ ] Case-dispositive terms identified; every proposed construction stress-tested on both infringement and validity
  • [ ] Term list ruthlessly pruned to the local-rule cap or below

Record

  • [ ] Complete intrinsic record assembled: patent, full file wrapper, parents/continuations, foreign counterparts, post-grant filings
  • [ ] Every appearance of each disputed term charted
  • [ ] Claim differentiation analyzed
  • [ ] Cited prior art reviewed

Doctrine

  • [ ] POSITA defined as of the correct date
  • [ ] Phillips hierarchy applied in order; ordinary meaning presumed
  • [ ] Lexicography and disavowal tested against the Thorner "clear and unmistakable" standard
  • [ ] No limitations imported from embodiments; no construction excludes a disclosed embodiment
  • [ ] Prosecution disclaimer audited (including PTAB statements)
  • [ ] Extrinsic evidence used only to fill genuine gaps, never to contradict the intrinsic record
  • [ ] Every claim screened for § 112(f) triggers under Williamson; corresponding structure (or algorithm) identified
  • [ ] Indefiniteness assessed under Nautilus's reasonable-certainty standard

Procedure

  • [ ] Local patent rules and standing orders read; all exchange deadlines calendared
  • [ ] Proposed terms and preliminary constructions exchanged with supporting evidence
  • [ ] Meet-and-confer completed; disputes genuinely narrowed
  • [ ] Joint claim construction statement filed
  • [ ] Markman brief organized term-by-term, quoting the intrinsic record
  • [ ] Technology tutorial prepared — neutral, clear, on time
  • [ ] Fallback constructions prepared; O2 Micro answer ready

After the order

  • [ ] Claim charts re-run against the actual constructions
  • [ ] Endgame chosen: stipulated judgment, summary judgment, trial, or settlement
  • [ ] Objections preserved through jury instructions and Rule 50
  • [ ] Appellate posture set with Teva's bifurcated standard of review in mind

Primary Authority

Statutes

Rules

Supreme Court

Federal Circuit

Secondary sources

  • Chisum on Patents (the standard treatise; see the claim-construction volumes)
  • Manual of Patent Examining Procedure (MPEP) § 2181 (means-plus-function) and § 2173 (definiteness)
  • Federal Judicial Center, Patent Case Management Judicial Guide

Related UpLaw Resources

Related Guides & Articles

Related Checklists

Related Toolkits

This checklist is general legal information from UpLaw (uplaw.us), not legal advice, and does not create an attorney-client relationship. Laws vary by jurisdiction and change over time; patent local rules differ materially from district to district. Verify current requirements and consult qualified counsel before acting.