What is claim construction actually deciding?
Claim construction determines what the disputed words in a patent claim mean, as a person of ordinary skill in the art would have understood them at the time of the invention. It is decided by the judge as a matter of law, usually in a proceeding called a Markman hearing, and it is not a factual question handed to a jury.
What makes it consequential is that a patent claim is a single sentence doing an enormous amount of work. It was drafted years before the accused product existed, by someone trying to describe an invention broadly enough to be valuable and narrowly enough to be granted. That language is genuinely capable of bearing more than one reading, and the reading the court adopts becomes the fixed reference point for every other question in the case.
What evidence does a court rely on to construe a claim?
Intrinsic evidence first, in a settled order: the claims themselves, the patent specification, and the prosecution history. The specification is treated as the single best guide to what the terms mean, because it is where the inventor described the invention — and, sometimes without intending to, defined a term or disclaimed scope.
Extrinsic evidence — dictionaries, treatises, technical literature and expert testimony — is admissible and genuinely useful, and it is subordinate to the intrinsic record. An expert who explains how the field used a term in the relevant period is helping. An expert who simply asserts the correct construction is both unhelpful and vulnerable, because that conclusion belongs to the court.
Why does the priority date matter so much in claim construction?
Because the meaning of a claim term is fixed as of the priority date rather than as of the trial, and technical fields move fast enough that this frequently changes the answer. A term that has since acquired a precise industry definition may have been used loosely, or differently, when the application was filed.
The practical consequence is that the persuasive sources are contemporaneous ones: textbooks, standards documents, conference papers and product literature from the period. Assembling that record is a substantial part of the technical work in claim construction, and it is the part most often skipped in favor of arguing from present-day understanding — which is exactly the argument the other side will be glad to meet.
Why can winning a claim construction lose you the case?
Because claim scope cuts in both directions simultaneously. A patent owner wants terms construed broadly enough to cover the accused product; a broad construction also reaches further back into what already existed, which is the territory an invalidity case occupies. Practitioners describe this as the vice, and it closes on parties who argue construction with only infringement in view.
The discipline it demands is to test every proposed construction against the prior art in the same breath as against the accused product. A construction that captures the defendant and also captures a 1998 product manual is not a victory. This is the main reason the infringement and validity analyses are best run together rather than in sequence, and why an expert asked to opine on one should usually be asked about the other.
Can a claim construction ruling be appealed straight away?
Generally not on its own. A Markman ruling is usually reviewed only after final judgment, which means parties can litigate an entire case — through discovery, expert reports and trial — on a construction an appellate court has never examined.
That structural fact is why the stage carries so much settlement pressure, and why effort spent getting construction right at the outset is the cheapest effort available in a patent case. It is also why a party that loses badly at Markman frequently reassesses the whole matter immediately rather than pressing on and hoping to fix it later.