A question of law about the meaning of words, decided by a judge, that settles most of what follows.
Start a conversation with the IP Concierge, already scoped to claim construction. Pick a starting point, or describe the matter directly.
Claim construction is the point at which a patent case stops being about technology in general and becomes about specific language. The court determines what disputed terms mean, viewed through the eyes of a person of ordinary skill in the art at the time of the invention, and the sources it consults have a settled hierarchy: the claims themselves, the specification, and the prosecution history first, with outside material such as treatises and expert testimony playing a supporting role. What makes this the highest-leverage stage of the case is that the parties are usually arguing about the accused product while pretending to argue about the invention. Both sides know which construction wins, and the constructions that win on infringement are frequently the ones that lose on validity.
The sources, in the order they carry weight, and what each contributes.
Including how other claims use the same term, and what a dependent claim implies the independent one does not require.
The single best guide to meaning — and the place where an inventor may have defined a term, or disclaimed scope, without meaning to.
What was argued and amended to get the patent granted, and what that surrendered permanently.
The hypothetical person whose understanding governs. Defining their skill level is a technical question with large downstream consequences.
Functional language that can be limited to the structure actually disclosed, which sometimes narrows a claim dramatically.
Dictionaries, treatises and expert testimony — admissible, useful, and subordinate to everything above.
How a technical expert supports construction without overstepping into it.
Almost everything, which is why it is litigated so hard so early.
A term read broadly enough to capture the accused product is frequently read broadly enough to read on what came before. Proposing constructions without testing them against the prior art in the same breath is the most common self-inflicted wound in patent litigation, and it is entirely avoidable.
No. Claim construction is a question of law reserved to the court. An expert can explain how a person of ordinary skill in the art would have understood a term at the relevant time, describe how the field used the language, and explain the technology well enough for the judge to reach a construction — but an opinion that simply asserts the correct construction is both unhelpful and vulnerable.
Because meaning is fixed as of then, not as of trial. Fields move quickly, and a term that has acquired a precise industry meaning in the intervening years may have been used loosely, or differently, at the time of filing. Anchoring the analysis to contemporaneous sources — textbooks, standards, papers and product literature from the period — is a substantial part of the technical work, and it is the part most often skipped.
If an applicant narrowed a claim during prosecution to get the patent allowed, they generally cannot recapture the surrendered ground later by arguing equivalence. It is a fairness rule: what you gave up to obtain the patent, you gave up. It is also why the file wrapper is read closely by both sides, and why arguments made to an examiner years earlier can decide a case nobody anticipated.
Generally not on its own. It is usually reviewed after final judgment, which is a large part of why the stage carries so much settlement pressure — parties may live with a construction through an entire trial before an appellate court examines it. Practically, that means the effort spent getting it right at the outset is the cheapest effort in the case.
Describe the claims and the accused product. The Institute will help you see what the analysis needs.