Guest commentary by Peter S. Menell. Professor Menell is Koret Professor of Law and at UC Berkeley School of Law and Director of the Berkeley Center for Law & Technology. Professor Menell filed an amicus brief along with Professors Jonas Anderson and Arti Rai in Teva Pharmaceuticals USA Inc. v. Sandoz Inc., 13-854. That brief can be accessed at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id= 2457958.
The U.S. Supreme Court heard argument six weeks ago in Teva Pharmaceuticals USA v. Sandoz, a case addressing the scope of appellate review of claim construction rulings. While less headline-grabbing than recent cases addressing the patentability of business methods, computer software, and DNA, Teva nonetheless confronts a critical issue that has long plagued the patent adjudication system.
Nearly two decades ago, the Supreme Court sought to promote more effective, transparent patent litigation through its ruling in Markman v. Westview Instruments that “the construction of a patent, including terms of art within its claim, is exclusively within the province of the court.” While carefully avoiding characterizing claim construction as a pure question of law, the court nonetheless removed interpretation of patent claims from the black box of jury deliberations by holding that the Seventh Amendment right of trial by jury did not extend to patent claim construction and that trial judges were better equipped than juries to resolve the mixed fact/law controversies inherent in construing disputed patent claim terms. The decision ushered in a new patent adjudication era in which judges construe patent claims before trial through a process that has come to be known as a “Markman” hearing.
Notwithstanding this shift, the transparency that the Supreme Court sought has only been partially achieved and the quality of claim constructions falls short on evidentiary and analytical rigor. The U.S. Court of Appeals for the Federal Circuit’s sharply divided 1998 en banc decision in Cybor Corp. v. FAS Technologies held that claim construction is purely a legal issue subject to de novo appellate review, downplaying the Supreme Court’s more nuanced description of claim construction as a “mongrel practice.”
The Cybor decision dissuaded district judges from holding evidentiary hearings regarding the meaning of disputed claims and discouraged judges from explaining their reasoning. As several district judges have candidly acknowledged, why go through all of that trouble if the Federal Circuit is going to construe the terms de novo on appeal? Claim construction reversal rates and consternation among district (and some Federal Circuit) judges rose precipitously following Cybor. And although the Federal Circuit has twice revisited this issue en banc since Cybor and reversal rates have subsided (but not as a result of more transparent records), the controversy has festered, leading to the Supreme Court’s review.
Why the Fuss?
Patent law requires that claim terms be construed from the standpoint of persons of ordinary skill in the technical art using the patent specification and prosecution history (intrinsic evidence) as well as pertinent scientific and technical sources (extrinsic evidence) where necessary as of the time of the patent filing.
Federal district judges are thus thrust into challenging, but not unfamiliar, territory. As in other areas of their docket — such as products liability, toxic torts or even criminal cases turning on forensics — in which they are called upon to resolve factual disputes about scientific and technological matters, judges have at their disposal a familiar toolbox — presentation of evidence and expert testimony. And when district courts assemble a record and explain their reasoning, Federal Rule of Civil Procedure 52(a)(6) provides that their determinations “must not be set aside unless clearly erroneous, and the reviewing court must give due regard to the trial court’s opportunity to judge the witnesses’ credibility.”
Patent claim constructions add a few additional twists to this well-established general judicial framework. Even if a district judge found that a patent claim term meant “X” to skilled artisans as of the effective patent filing date, the court would still have to construe the term to mean “Y” if the intrinsic evidence established that the patentee intended such a specialized meaning. Furthermore, many disputes about patent claim meaning turn not on meaning within the technical art but rather on claim drafting conventions. Thus, courts must pay special attention to the intrinsic evidence. But the understandings of skilled artisans will be pertinent in many claim constructions, and such subsidiary factual determinations should be subject to deference under general judicial rules as well as trial judges’ proximity to the extrinsic evidence.
For these reasons, the Supreme Court should adopt a hybrid appellate standard for claim construction: Factual determinations underlying claim construction rulings should be subject to the “clearly erroneous” standard of review, while the Federal Circuit should retain de novo review over the ultimate claim construction decision. In this manner, district court judges, in their capacity as fact-finders, could better surmount the distinctive challenges posed by the technical, mixed fact/law controversies inherent in patent claim construction. The Federal Circuit would ensure that trial judges’ constructions comport with the intrinsic evidence supporting the patents.
Parsing the Supreme Court Argument
The Teva argument began with a colloquy regarding what constitutes a “subsidiary fact” and how it relates to the ultimate claim construction determination. Midway through Teva’s opening argument, Justice Samuel Alito expressed doubt that departing from the de novo standard was worth the candle. He noted that a recent empirical study could not find any case in which a shift from the de novo standard to clear error review of factual determinations would have affected the appellate review outcome.
In an unintended manner, Justice Alito’s query highlights the reason de novo review of factual determinations raises serious concerns for a justice system. The study to which he refers does not and cannot establish the “fact” that it is purported to prove. Its analysis — which is not revealed or scrutinized in the Supreme Court colloquy — boils down to the following logic: (1) three-judge panels of the Federal Circuit will be better than a single generalist district court judge at applying the appellate court’s rules of claim construction; and (2) because most claim construction reversals (77.1 percent) are unanimous, a shift to a clear error standard would not change the outcome of any cases. The first point ignores the distinctive role of judges in fact-finding, a pillar on which our court system is based. The second point does not prove that no cases would come out differently. It also overlooks the norm that appellate judges rarely dissent unless they disagree on legal questions.
More fundamentally, according greater deference to trial courts through clear error review would encourage district judges to use focused evidentiary hearings (in conjunction with the technical tutorials that are common) to substantiate the basis for their claim constructions, thereby promoting more systematic, well-founded claim construction analysis. Contrary to the study to which Alito referred, we have reason to expect clearer substantive analysis, more settlements following claim construction and trial, more effective appellate review, and fewer reversals and remands in a parallel universe in which Rule 52(a)(6) applies to claim constructions — not business as usual in the Cybor regime.
The Court also probed how deference on subsidiary facts would affect uniformity in patent interpretation. At a basic structural level, it is difficult to see how to achieve such uniformity to the extent that claim construction has a factual component. Under Blonder-Tongue Labs., Inc. v. Univ. of Ill. Found., 402 U.S. 313 (1971), the Federal Circuit’s interpretation of a patent in one case cannot be asserted offensively by the patentee in a later infringement action against other defendants to the extent that there are disputed factual underpinnings. At most, that prior adjudication can bar the patentee from seeking an alternative interpretation under judicial estoppel principles. Because the patent system has no mechanism for conclusively establishing patent scope with regard to all potential infringers where subsidiary factual disputes exist, the certainty that flows from appellate interpretations is not ironclad, as subsequent defendants can potentially bring new evidence or more effective advocacy to bear on claim meaning. Nonetheless, district courts routinely consider prior interpretations of patent claim terms in conducting claim construction. Moreover, concerns about the clarity of patent claims are more appropriately addressed through improvements to the patent prosecution process, post-grant review and reexamination, consolidation of claim construction through multi-district litigation, and adjustments to substantive claim construction jurisprudence.
Back to Basics
The Teva case poses a fundamental question going to the heart of the American justice system. Should factual determinations underlying patent claim construction be subject to substantially less deference (or more distrust) than the factual determinations made in every other area of federal adjudication? In view of the Supreme Court’s efforts over the past decade to bring patent adjudication more in line with general civil litigation principles, it is difficult to see why the Court would make an exception here. Combining deferential review of factual findings with de novo review of the overarching claim construction determination — focused on ensuring that claim construction comports with the intrinsic evidence and that the trial court followed generally applicable rules of claim construction — would enhance the quality of claim constructions, produce a more capacious and transparent appellate record, and provide the appropriate appellate safeguard.
Did Jason Rantanen write this… or did Peter Menell?
Is the piece a “summary” of the linked to article or is it just useful “context” for one interested in reading the article (for which access is forbidden or 404-Not found)?
By appearances of the blog layout, text formatting etc. the “author” seems to be Jason Rantanen, while the “title” self-referentially implies that the piece is commentary by Mr. Menell.
So confusing.
It’s by Professor Menell. I’ve clarified that in the text at the top and fixed the link.
Meh, giving the factual findings underlying a claim construction deference might help a tad but what is really missing in claim construction these days is the courts bowing to what claims actually are. They are mere attempts to put into words what the patentee regards as his invention. Treating claims as such ends 99.99999% of claim construction squabbles (when 112 2nd is also at hand to help out).
So, 6, do I read you right?
The UK Supreme Court declared some years ago, that the way to do claim construction was for the court to ask itself one simple question, namely, what was the writer of the claim “using the language of the claim to mean” (mean to its notional PHOSITA reader, of course).
There is only that one way, both in the PTO and the courts. in the UK, a claim has only one meaning, regardless of forum.
Many patent practitioners find that a useful, real world, technologist- and business-friendly and efficient approach? How is it for you?
Professor, true, Markman held that claim construction is for the court. But did that case hold or even address whether the issued of definiteness was for the court or the trier of fact, which could be a jury?
IIRC, the present case involves differences between the Federal Circuit and the District Court on whether a claim term was indefinite. Is this really a claim construction issue given the new Nautilus standard of reasonably clear to one of ordinary skill in the art?
Good question there at the end Ned.
The entire question is a farce. The meaning of a claim should be clear on its face, not some Schrodinger’s Patent sort of thing where you don’t know what it means until it gets litigated.
This.
It doesn’t matter much what the standard for review of the process is when the Federal Circuit’s various panels are applying different processes or applying the same process inconsistently or arbitrarily, or have different views about the goal of the process.
More importantly, heavy reliance on compromised, paid-to-deliver, biased “experts” to tell you what a word or a claim “means” is a recipe for disaster for any patent system. Claims that require such “expertise” — and hundreds of thousands of dollars — merely in order to be understood by scientifically literate members of the public should never be granted and, if they are granted, they should generally be construed against the patentee’s interest or found indefinite.
If a fact-finder is incapable of figuring out the meaning of a claim term from the claims themselves, the specification, the prosecution history and contemporary treatises, then instead of listening to two dueling “experts” recite their scripts, how about a court-appointed polling expert takes an objective poll (with the methodology agreed on by both parties) that gathers the opinions of (wait for it) actual skilled artisans “who were there.” If there’s a tie, then the patentee loses on that claim construction issue. Fee for the court-appointed expert can be shifted towards one party in any amount as appropriate given the outcome and the merits of the arguments in the briefs.
In addition, a set of unambiguous and consistently applied rules for interpreting certain common terms and resolving issues that arise repeatedly (e.g., singular versus plural) seems foundational.
MM, I somewhat agree.
However, experts can be useful in explaining background technology. I just don’t like it when they testify on the ultimate legal question and give an opinion. I say, keep your opinions to yourself.
I have always thought that courts should appoint experts.
This is true, but some of these ambiguities are as much about how the issue arises in litigation as how the claim is drafted.
For many claim terms, it’s very easy to read the patent and figure out whether device X infringes, but very difficult to figure out whether device Y infringes.
The manufacturer of device X either takes a license or bets on a good invalidity defense. The manufacturer of device Y goes to court with a paid “expert,” arguments that the patentee gave up everything in litigation, and a pile of dictionaries, and fights it out with the patentee in this absurd Schrodinger’s patent ritual.
There’s a definite problem of incentives to draft vague claims. But that’s not the only issue, and some ambiguities in patent claims are unavoidable.
Why do we continue to conflate broad claims with lazy prosecutorial lawyering.
In my experience: What can feel like a painfully narrow claim when drafting (near the time of conception) can feel nakedly broad years later at the time of enforcement. Technology moves fast these days. What was bleeding edge soon becomes conventional wisdom.
>>a fundamental question going to the heart of the American justice system.
Except with Alice we know that the SCOTUS doesn’t care about the justice system. They view themselves as legislators (and writers of the Constitution.)
There is moral or ethical boundary or component to that lot. (Let’s hope Ginsburg retires.)
Comments are closed.