(Im)pure Questions About Claim Construction In Patent Cases After Phillips

Patent claim construction issues raised in the heralded Phillips case
( Phillips v. AWH Corp. , 412 F.3d 1303 (Fed. Cir. 2005) ( en
)) offered the promise of a decision calculated to reduce a 50%
reversal rate in the Federal Circuit that has left litigants and lawyers longing
for predictability in patent proceedings.

Many commentators expected the Federal Circuit to clarify the methodology we
use to give meaning to the claims that define patented inventions. Savvy
observers also hoped that the Court would correct the course charted in Cybor
Corp. v. FAS Technologies, Inc., 138 F.3d 1448 (Fed. Cir.
1998) (en banc), by turning away from the fiction that the construction of a
patent claim is a pure question of law subject to review de novo , with
no deference given to the trial court's determination of underlying matters of

Phillips answered some questions about how we should determine what
patent claims mean, but failed to address and finally abandon the false notion
that claim construction is a pure question of law. That fatal flaw has
implications for appellate review that prevent Phillips from providing
patent owners and accused infringers with the predictability they need and
deserve. Circuit Judge Mayer in his dissenting opinion got it right: the Federal
Circuit is "rearranging the deck chairs" and asking the orchestra to play on as
if nothing is wrong as the ship heads straight for Davey Jones' locker.

In its 1995 Markman decision, the Federal Circuit worked a sea change
in the law when it held that the judge - not the jury - must determine what the
patent claims mean. The Federal Circuit concluded that "a judge, trained in the
law, will analyze the text of the patent and its associated public record and
apply the established rules of construction, and in that way arrive at the true
and consistent scope of the patent owner's rights." In Cybor Corp.
, the Federal Circuit exalted Markman by holding that claim
construction is a pure question of law subject to review de novo , with
no deference at all given to the trial court's underlying determinations about
"allegedly fact-based questions relating to claim construction."

In the decade since Markman , the Federal Circuit has struggled to
develop a principled approach for construing patent claims. In Vitronics
v. Conceptronic, Inc. , 90 F.3d 1576 (Fed. Cir. 1996), it
directed the courts to construe claims in light of the "intrinsic" record: (1)
the language of the claims; (2) the patent specification; (3) the prosecution
history; and (4) prior art cited in the prosecution history. Under Vitronics
, consideration of the "extrinsic" record (other prior art documents,
dictionaries and expert testimony), is appropriate only if the claim meaning
remains ambiguous after review of the intrinsic evidence.

The hierarchical, rules-based Vitronics approach shared the stage with
another line of cases that culminated in Texas Digital Systems, Inc. v.
Telegenix, Inc., 308 F.3d 1193 (Fed. Cir. 2002), in which the Federal
Circuit instructed the courts to first rely on extrinsic dictionaries to give
claims their "plain meaning" and avoid improperly importing definitions from the
specification. Under Texas Digital, the intrinsic record was examined
only to determine whether the patentee had rebutted the dictionary definition,
either by specifically defining the meaning of a claim term, or by disclaiming a
particular definition through words of manifest exclusion.

We expected Phillips to provide some certainty about the appropriate
methodology for claim construction, and that much the Court did. The Federal
Circuit acknowledged that Texas Digital "placed too much reliance on
extrinsic sources such as dictionaries, treatises, and encyclopedias," and then
expressly reaffirmed the Vitronics approach.

The problem is that the Federal Circuit stopped short after overturning
Texas Digital and declined to take advantage of an opportunity to correct
mistakes made in Cybor.

Recent studies report that the Federal Circuit reverses 25 to 50 percent of
the claim constructions that are appealed. The Federal Circuit's reversal rate
is unusually high because shifting precedent directed to claim construction
methodology, coupled with de novo review where no deference is given to
the trial court, is a recipe for reversal. The Federal Circuit, however, seems
content to deal in formulaic claim construction methodologies, while ignoring a
fundamental problem and refusing to acknowledge that claim construction depends
on underlying factual determinations that should be reversed on appeal only if
clearly erroneous.

Markman hearings now are akin to mini-trials: full evidentiary
hearings that run for days or weeks, replete with testimony from inventors and
experts about the science, inventions, prior art, state of the art and
hypothetical person of ordinary skill in the art. The district courts assess the
credibility of experts, inventors and other witnesses proffered to prove
underlying facts, and they determine facts relevant to arguments about the
meaning of the claims. Our jurisprudence demands that these kinds of factual
determinations should not be disturbed on appeal unless clearly erroneous, but
Cybor has stood the law on its head. Ignoring reality to perpetuate the
fiction that claim construction is a pure question of law allows the Federal
Circuit to freely substitute its judgment for that of the district courts, but
that does violence to the Court's role as an appellate court and to
predictability. Trial means little and nothing is predictable except that the
loser will roll the claim construction dice in the Federal Circuit.

Judge Mayer is right. The Court needs to put Cybor to

Published June 1, 2006.