Saga of a Baffled Dry Cleaner
A couple of weeks ago, the Federal Circuit issued its en banc ruling in Phillips v. AWH Corporation. Like most folks with a deep interest in patent law (and virtually no one else), I had been awaiting this decision for its promised clarification of the tools of claim construction. When it finally came down, I was pleasantly surprised. Though the law still does not stand where I think it should, I thought that the court arrived at the most sensible among the possible positions.
I promised a colleague that I would digest the case for an internal IP newsletter. I did that--and in keeping with the digest’s tone, ended up with a very vanilla and uninteresting little essay (note that I am not knocking vanilla, which is in truth one of my favorite flavors). Nevertheless, I chose to replicate it below. But I also decided that maybe, just maybe, I could use this posting as an excuse to promote a little patent education. So I’ve also done something gutsy. I also included a draft of an essay that I prepared as part of my second book. Of course, like all of my ealry drafts, it’s quite overwritten; it will be much condensed in later edits. Nevertheless, if you look below, you will find my introductory essay on the state of claim construction.
(WARNING! This is an unacceptably long post. Proceed at your peril).
(WARNING! I haven’t quite figured out how to include footnotes in a convenient matter, so I’ve just dropped them. Everything that I claim is a quote is, in fact, a quote. The state of the technology, or my aptitude, make it difficult for me to source quotes appropriately).
The Phillips Digest
In Phillips v. AWH Corp., 2005 U.S. App. LEXIS 13954 (Fed. Cir., en banc, July 12, 2005), the Federal Circuit resolved a subtle yet important question lingering over the legal determination of a patent’s meaning: How should a court resolve seeming differences between the “common” meaning of a word and a more specific meaning that the patent drafter may have intended? As Judge Bryson, writing for the court, noted,
This is hardly a new question. The role of the specification in claim construction has been an issue in patent law decisions in this country for nearly two centuries. We addressed the relationship between the specification and the claims at some length in our en banc opinion in Markman v. Westview Instruments, Inc., 52 F.3d 967, 979- 81 (Fed. Cir. 1995) (en banc), aff’d, 517 U.S. 370 (1996). We again summarized the applicable principles in Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576 (Fed. Cir. 1996), and more recently in Innova/Pur e Water, Inc. v. Safari Water Filtration Systems, Inc.,381 F.3d 1111 (Fed. Cir. 2004). What we said in those cases bears restating, for the basic principles of claim construction outlined there are still applicable, and we reaffirm them today. We have also previously considered the use of dictionaries in claim construction. What we have said in that regard requires clarification.
This characterization of the question alludes to the need for the court’s en banc clarification. Ten years ago, Markman announced that a patent is a legal document that rests upon a factual foundation, rather than a factual description of an invention. From a litigation perspective, this distinction means that judges—and judges alone—must construe patent claims as matters of law. Juries, often the triers of fact in patent cases, play no role in resolving linguistic disputes in claim construction. As a practical matter, the legal rulings about claim construction (frequently in pre-trial motions known as “Markman hearings”) frequently determine the outcome of the entire dispute, and lead to rapid summary judgment for the side whose linguistic argument prevailed.
In the years following Markman, the Federal Circuit began to develop and to promulgate rules for claim construction. Two unremarkable principles of legal construction proved to be quite useful. First, words in legal documents usually mean what they mean outside those legal documents; in cases of doubt or dispute, dictionaries and other contemporary writings are useful references. Second, drafters of legal documents often affix technical meanings to words that may differ from their colloquial meanings; in cases of doubt or dispute, the legal document, its accompanying correspondence, and the testimony of experts versed in technical language are better determinants of meaning than are more public and popular resources.
These two rules led to two distinct lines of cases, which though not explicitly contradictory, did create tension and confusion with regard to claim construction. The “Texas Digital” line, Texas Digital Systems, Inc. v. Telegenix, Inc, 308 F.3d 1193 (Fed. Cir. 2002), emphasized the importance of dictionaries in determining “plain meaning,” and then turned to the patent specification itself to see if a clear veto could override that meaning. The “Vitronics” line focused on the specification, and turned to dictionaries only when the specification left some doubt. The Federal Circuit agreed to hear Phillips en banc to resolve this tension. The court explained that Texas Digital
placed too much reliance on extrinsic sources such as dictionaries, treatises, and encyclopedias and too little on intrinsic sources, in particular the specification and prosecution history. . . . In effect, [this] approach limits the role of the specification in claim construction to serving as a check on the dictionary meaning of a claim term if the specification requires the court to conclude that fewer than all the dictionary definitions apply, or if the specification contains a sufficiently specific alternative definition or disavowal. . . . That approach, in our view, improperly restricts the role of the specification in claim construction.
The court therefore reaffirmed the guidelines that it had set forward in Vitronics:
The underlying goal of our decision in Vitronics was to increase the likelihood that a court will comprehend how a person of ordinary skill in the art would understand the claim terms. In that process, we recognized that there is no magic formula or catechism for conducting claim construction. Nor is the court barred from considering any particular sources or required to analyze sources in any specific sequence, as long as those sources are not used to contradict claim meaning that is unambiguous in light of the intrinsic evidence. For example, a judge who encounters a claim term while reading a patent might consult a general purpose or specialized dictionary to begin to understand the meaning of the term, before reviewing the remainder of the patent to determine how the patentee has used the term. The sequence of steps used by the judge in consulting various sources is not important; what matters is for the court to attach the appropriate weight to be assigned to those sources in light of the statutes and policies that inform patent law. In Vitronics, we did not attempt to provide a rigid algorithm for claim construction, but simply attempted to explain why, in general, certain types of evidence are more valuable than others. Today, we adhere to that approach.
The court then turned to apply these principles to the specific patent dispute at issue, namely the meaning of the term “baffle” in Phillips’ patented approach to pre-fabricated prison construction, rejected the restrictive definition of the term that AWH had advocated and the district court accepted, and remanded the matter for trial.
The significance of this case to anyone other than the parties, however, lies not in its specific application but rather in the general principles that it establishes. The court’s acceptance of the Vitronics approach rather than the Texas Digital approach elevates the importance of technical expertise and testimony above that of colloquial understanding. It attempts to inject as much substantive, technical knowledge into the claim construction process as is possible within what remains a legal ruling. It is likely to create both new opportunities and new needs for expert assistance in preparing patent cases, both in the early stages and at trial.
A Draft Essay Introducing Claim Construction
Most patent lawsuits begin with disagreements about the precise meaning of at least some words in the claims. The court must begin by determining what those disputed words mean. This exercise is trickier than it sounds like it should be.
Why? Well, before we can think about construing disputed terms in any particular patent, we’ve got to nail down the rules. Where should the court look to resolve linguistic disputes? For that matter, who should do the looking—the judge or the jury? The answer to this latter, seemingly straightforward, question leads straight into metaphysics: What is a patent? Is it a technical document that one skilled artisan wrote for another, or is it a legal document that lawyers wrote for other lawyers? Should technical witnesses testify about words’ meanings in their narrow fields of specialization, or should the court just look them up in a standard dictionary? Is claim construction a factual question or a legal question?
Litigation and metaphysics are not comfortable bedfellows. Patent scholars may devote their time and energy to the metaphysical aspects of these questions, but litigants, practicing lawyers, and judges need concrete answers. Hundreds of thousands of active patents, billions of dollars, and the respective roles of judges and juries all hang in the balance.
If claim construction is a factual question about the meaning of a technical document, the obvious path towards resolution would be for the court to hire a technically trained “special master,” well versed in the relevant technology and beholden to neither party, and assign that expert the task of disambiguating disputed terms. Unfortunately—at least for present purposes—that’s not the way our justice system works. Remember, you’re in the midst of a jury trial. That means that the jury, and the jury alone, is the “trier of fact.” Certainly, if you and the defendant agree to let the judge appoint a technical expert, and you agree to drop all terminological disputes in favor of the expert’s construction, no one would object. But such agreements are rare, and though the trial judge may try to cajole the parties into one, even such cajolery succeeds only rarely. No, in a jury trial, the parties present all disputed facts to the jury, present evidence supporting their interpretation of the facts, and then let the jury resolve all factual disputes. Unless the jury reaches a conclusion that is entirely inconsistent with the evidence, neither the trial judge nor an appellate court will reopen its factual findings.
In short, if the claims section of a patent is a technical document and disputes over the meaning of its words are factual disputes, only the jury can decide what they mean. But jurors are rarely expert in either the technology underlying a patent or the legal niceties involved in patent drafting. No one, regardless of his training, has ever read his first patent and found it comfortable and familiar. Patents are quite simply unique documents, similar in some ways to technical articles and similar in others to legal documents, but also quite different from both. The idea that an untrained jury could resolve disputes over specific technical terms embedded in a patent is a “legal fiction,” at best. If we want juries to construe claims, we’re just going to have to pretend that they know what they’re doing.
Such a pretense, however, is not strictly necessary. We could go the other way, and insist instead that a patent is nothing more than a legal document, like a contract or a deed, that lawyers write for other lawyers to read. If that’s the case, judges shouldn’t really need technical testimony or expert help. Federal judges are all qualified attorneys, more than capable of differentiating between the arguments that the parties’ equally capable attorneys put forward. And if they fail, appellate courts—in this case, the Federal Circuit—need show them no deference. Recall that appellate courts review legal questions de novo, or from scratch. And claim construction appears to be precisely the sort of issue that motivated Congress to found the Federal Circuit. Though most district court judges, who encounter patents infrequently, may be unfamiliar with the nuances of patent language and patent law, Federal Circuit judges have no such excuse. Not only are they immersed in those nuances, they wrote them. Uniformity in patent law may be the court’s primary raison d’être. If we can’t trust the Federal Circuit to impose uniformity on claim construction, who can we trust?
This approach, namely classifying patents as legal documents so that courts can resolve disputed terms as matters of law, seems to make a lot more sense. Except for one tiny little item: It’s absurd! The only reason that we have patents is to promote progress. Motivating innovation is an important part of that promotion, but it’s only part. If motivation were the only goal, we could let patentees keep their inventions secret. But we don’t. We insist that they buy their claims by explaining their inventions in terms comprehensible to their colleagues. How in the world can we then turn around and insist that patents are not technical documents? Either inventors write them so that other inventors can understand them, or lawyers write them so that other lawyers can understand them. Which is it? Patent policy seems to imply that patents are technical documents. But patent law suggests that there is no meaningful way to resolve linguistic disputes unless patents are legal documents.
Congress was silent on this question; the Patent Act doesn’t provide any useful insights. That leaves the matter to the courts. For 205 years, the courts rose to the challenge by keeping the issue open. Sometimes judges construed claims as a matter of law, and sometimes juries construed them as findings of fact. Sometimes appellate courts were highly deferential, and sometimes they weren’t. And then one day in 1995—more than two centuries after Congress passed our first Patent Act—a man named Herbert Markman decided to force the issue. The Federal Circuit responded with an en banc session. Here is his story:
Markman was a pioneer in the field of dry cleaning, the owner of a patented technology for tracking “inventory” in a dry cleaning establishment. His patent, like all patents, contained two parts. Its written description taught the public how to implement an inventory tracking system. Its claims section specified what the public gave Markman in return for the written description. Like most patentees, Markman had claimed a number of variations on the basic theme of a computerized inventory control. Each claim specified a slightly different setup involving an input device, a data processing device “including memory operable to record [inventory] information and means to maintain an inventory total,” a dot matrix printer (after all, Markman filed his patent application in 1984), and a scanner capable of reading bar codes. The system, operating as a whole, must be able to “detect and localize spurious additions to inventory as well as spurious deletions therefrom.”
In other words, the bargain that the PTO—acting on behalf of the public—had struck with Markman was simple. For the life of his patent, no member of the public could use any of the claimed setups without Markman’s permission—but Markman could not prevent members of the public from using similar but unclaimed devices or setups. One day, Markman noticed that Althon Enterprises, the owner of two dry cleaning establishments, was using Westview Instruments’ DATAMARK and DATASCAN devices. Markman sued both Westview (for making and selling an infringing device) and Althon (for using an infringing device). Westview’s machines combined a data processor, a printer, and a scanner to track dry cleaning invoices and receipts.
Now comes the part where our sudden interest in dry-cleaning inventory control systems pays off. Markman requested and received a jury trial. Three witnesses testified on Markman’s behalf, the first two to explain the claim language to the jury and the third to convince the jury that Westview’s devices infringed the patent: Markman testified that he knew what his patent and its claims meant. He hired a practicing patent lawyer to testify about the way that a trained patent expert would read the claims. And he hired an expert on bar-code technology to explain Westview’s devices to the jury. All of this testimony, of course, only makes sense if claim construction is a factual question that juries decide.
Westview took a different approach. It called only a single witness, its president, who explained to the judge and the jury how his company’s devices worked. In particular, he pointed out that the devices tracked only invoices and receipts, not clothing. As he saw it, a dry cleaner’s “inventory” is clothing; a tracking device incapable of tracking clothing may be many things, but it cannot be an infringing inventory tracking system. Though this testimony makes sense whether claim construction is factual or legal, Westview’s next move did not. Westview asked the judge to rule, as a matter of law, that its products could not possibly infringe Markman’s patent.
The judge probably should have decided then and there whether the question was factual or legal. However, since this case occurred in the pre-1995 muddle, he decided not to decide. He told the jury to “determine the meaning of the claims using the relevant patent documents including the specifications, the drawings and the file histories. . . . also relevant are other considerations that show how the terms of a claim would normally be understood by those of ordinary skill in the art.” He also told the jury “to compare the claims with the Westview device to determine if it infringes.” In other words, the district court judge decided to treat all of the relevant questions as factual, and to let the jury, as trier of fact, decide them. The jury did, and concluded that Westview’s devices infringed some but not all of Markman’s claims.
So far so good—except for the denouement. Having told the jury—and the parties—to treat patents as technical documents and claim construction as a factual question, the judge then turned around and announced that claim construction is a matter of law, for the judge alone to decide. He adopted Westview’s view of the patent, announced that as a matter of law, the DATAMARK and DATASCAN could not infringe, and ruled for Westview. Not surprisingly, Markman appealed. According to Markman, the trial judge was just wrong when he announced that claim construction was a legal question; claim construction is, he asserted, a factual question, and factual questions in jury trials belong to the jury.
The Federal Circuit took the unusual step of bypassing a three-judge panel and agreeing to hear the case en banc. Why? Because, as Chief Judge Archer explained,
The opinions of this court have contained some inconsistent statements as to whether and to what extent claim construction is a legal or factual issue, or a mixed issue. . . . At its inception, the Federal Circuit held that claim construction was a matter of law. Our first opinion deciding a question of claim construction, [decided July 15, 1983] said so explicitly. . . . The first Federal Circuit case to deviate from this precedent and state that claim construction may have underlying factual inquiries that must be submitted to a jury was [decided on April 27, 1984]. . . Notwithstanding the apparent inconsistencies in our opinions, the Supreme Court has repeatedly held that the construction of a patent claim is a matter of law exclusively for the court. . . . We therefore settle inconsistencies in our precedent and hold that in a case tried to a jury, the court has the power and obligation to construe as a matter of law the meaning of language used in the patent claim.
Though Archer stated this inconsistency in rather matter-of-fact terms, it was quite an admission of failure. Numerous committees and commissions had spent almost fifteen years debating the wisdom of a single court for patent appeals. Finally, after both the Carter and Reagan administrations signed onto the deal, Congress passed an enabling statute and told the Federal Circuit to clean up patent law. The court held its first session in October 1982. Seven-and-a-half months later, it announced unambiguously that claim construction is a legal question—implying that jury input is inappropriate and that judges must resolve linguistic disputes themselves. Nine-and-a-half months after that, it announced unambiguously that claim construction is a factual question—implying that a ruling from the bench is inappropriate and that juries must resolve all linguistic disputes. According to Archer, that second ruling contradicted explicit and consistent instructions from the Supreme Court. Yet, he conceded that for over a decade, the Federal Circuit had generated numerous decisions following each of these contradictory unambiguous announcements, thoroughly confusing patentees, accused infringers, and district court judges. In his view, thirteen years after its birth and eleven years after sowing (or at the very least perpetuating) confusion about this basic metaphysical question of patent law, the Federal Circuit was finally going to do the job that Congress had assigned it.
Archer announced a ruling providing clear instructions to anyone—litigant, judge, or jury—involved in patent litigation. A patent is a legal document that lawyers write for other lawyers to read. If litigants disagree about the meaning of a word in a claim, the trial judge must resolve this legal dispute before the trial. Juries must never hear erroneous interpretations of a patent—any more than they should hear erroneous interpretations of a contract, a law, or the constitution. After all, “it is emphatically the province and duty of the judicial department to say what the law is.” Juries exist as finders of fact. Who cares what a jury thinks about the law?
How should we feel about this conclusion as a matter of policy? Archer had some views about that, as well.
There is much wisdom to the rule that the construction of a patent should be a legal matter for a court. A patent is a government grant of rights to the patentee. By this grant, the patentee owns the rights for a limited time to exclude others from making, using, or selling the invention as claimed. Infringement of the patentee’s right to exclude carries with it the potential for serious consequences: The infringer may be enjoined and required to pay increased damages, costs and attorney fees. . . . To treat the nature of the patented invention as a matter of fact, to be inquired of and determined by a jury, would at once deprive the inventor of the opportunity to obtain a permanent and universal definition of his rights under the patent, and in each case of infringement it would subject him to the danger of false interpretation, from the consequences of which he could not escape. By confiding this duty to the court, however, its decision as to the nature of the patented invention becomes reviewable to the same extent as any other legal question.
Archer focused on the meaning of a patent in litigation. When people get involved in a lawsuit, all sorts of documents assume a strange new status as “evidence.” In different sorts of trials, evidence can include contracts, invoices, newspaper and magazine articles, handwritten notes, etc. We have some pretty clear rules about evidence. Those categorized as “legal documents” have legally fixed meanings. All others have factual meanings that might be subject to differing interpretations. As a result, if litigants dispute the meaning of a contract, the judge will resolve the dispute as a matter of law. If they dispute the implication of a newspaper article or a love letter, the jury will resolve the dispute as a question of fact. Which is a patent more like? Archer is almost certainly correct in answering “contract.” A patent is a legal document, in many ways a bill of sale between the government and an inventor; the inventor owns the claims and the public the underlying knowledge as explained in the written description.
So far, so good. A patent is a legal document, its meaning is a legal issue, and legally-trained judges must resolve disputes about that meaning as a matter of law. But Archer also provided another explanation. “Further,” he argued,
it is only fair (and statutorily required) that competitors be able to ascertain to a reasonable degree the scope of the patentee’s right to exclude. They may understand what is the scope of the patent owner’s rights by obtaining the patent and prosecution history—the undisputed public record—and applying established rules of construction to the language of the patent claim in the context of the patent. Moreover, competitors should be able to rest assured, if infringement litigation occurs, that a judge, trained in the law, will similarly 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 to be given legal effect.
This second argument is hardly compelling; if anything, it negates the point that Archer was trying to make. The patentee’s “competitors” reading the patent to gain the knowledge that the government secured on their behalf are technologists, not lawyers. They want to keep up with the state-of-the-art of their fields, and incorporate it into their products. Is it fair to insist that every scientist who reads a patent must also consult an attorney to see how a legal eye might view this technical document? If so, what does that say about the patent system and the “little guy?” What protection does the tinkerer-in-his-garage have? And just how large do our corporate patent-law departments have to be anyway? The whole purpose of the patent system was supposed to be to promote progress. How can requiring a legal interpretation of these important new technical documents possibly promote progress?
Archer, who wrote on behalf of an eight-judge majority, didn’t answer any of these questions. Two of his colleagues, though, saw the problems quite clearly, and disagreed with him quite vehemently. Judge Newman was the calmer of the two. She simply explained that
Patent infringement is a factual question. Its resolution often requires finding the factual meaning and scope of the terms of scientific art and technology and usage by which the patentee described and claimed the invention. These findings usually require testamentary and documentary evidence and occasionally experiments or demonstrations, as illustrated in many of our previous decisions that are now overruled. Deciding the meaning of the words used in the patent is often dispositive of the question of infringement. Thus in the case at bar the infringement controversy is decided by finding the meaning and scope of the term “inventory” in Markman’s patent, in light of the accused Westview system: if “inventory” is limited to clothing, the patent is not infringed; if “inventory” includes invoices, it is. The majority holds that this is a matter of law, devoid of any factual component; and subject to de novo appellate determination. The jury is eliminated, and new and uncertain procedures are imposed on trial judges. . . The court creates a litigation system that is unique to patent cases, unworkable, and ultimately unjust.
Judge Mayer verged on the apoplectic. In his view, the Federal Circuit had been doing its job all along. Just because a patent is a legal document hardly means that its interpretation is devoid of factual inquiries. The seeming contradictions that Archer had cited were more apparent than real. District court judges had been holding trials under the current incarnation of the Patent Act since 1952, and under the Federal Circuit’s supposedly contradictory guidance for more than a decade. Somehow, despite the confusion, they had all muddled through their caseloads and decided which issues belonged to them and which belonged to the jury. Now, in one fell swoop, the Federal Circuit was ejecting juries from an important category of civil lawsuits, an affront to, if not a violation of, the Seventh Amendment. He complained:
Today the court jettisons more than two hundred years of jurisprudence and eviscerates the role of the jury preserved by the Seventh Amendment of the Constitution of the United States; it marks a sea change in the course of patent law that is nothing short of bizarre. Sadly, this decision represents a secession from the mainstream of the law. It portends turbulence and cynicism in patent litigation. For this is not just about claim language, it is about ejecting juries from infringement cases. All these pages and all these words cannot camouflage what the court well knows: to decide what the claims mean is nearly always to decide the case. But today’s action is of a piece with a broader bid afoot to essentially banish juries from patent cases altogether. If it succeeds juries will be relegated, in those few cases where they have any presence at all, to rubber stamps. . .
The question is whether the interpretation of patent claims is a purely legal exercise—always decided by the judge as a matter of law and never raising a question of fact—or rather a mixed question of law and fact, in which some factual matters might need to be resolved by the factfinder on the way to construing the claims as a matter of law. . . . Contrary to what it says today, this court (including the judges in the majority) has always held that claim interpretation is a matter of law depending on underlying factual inquiries. . . . Patents are directed to those skilled in the art. . . . By broadly proclaiming all aspects of claim interpretation to be legal, the court today usurps a major part of the functions of both trial judge and jury in patent cases, obliterating the traditional, defined differences between the roles of judge and jury, and trial and appellate courts. . . . There is simply no reason to believe that judges are any more qualified than juries to resolve the complex technical issues often present in patent cases. Indeed, the effect of this case is to make of the judicial process a charade, for notwithstanding any trial level activity, this court will do pretty much what it wants under its de novo retrial. We have consistently stressed that the same rules apply to patent cases as apply to all other civil disputes. The court subverts this principle and the demands of the Seventh Amendment by the ruse of reclassifying factual questions as legal ones.
Mayer’s passion was to no avail. He was unable to convince even a single one of his Federal Circuit colleagues to join his opinion, and he fared no better when the Supreme Court entered the fray. There, Justice Souter wrote for a unanimous court in terms strikingly similar to Archer’s. As he saw things, the Seventh Amendment only guaranteed jury trials for the sorts of disputes that existed when the framers wrote it. Because both patents and infringement suits existed in the Eighteenth Century, patentees and accused infringers possess the constitutional right to demand a jury trial. Because Eighteenth Century patents lacked formal claims sections, however, the Seventh Amendment couldn’t apply to claim construction. That left our contemporary courts free to do pretty much whatever they wanted.
As a matter of the sound administration of justice. . . judges, not juries, are the better suited to find the acquired meaning of patent terms. The construction of written instruments is one of those things that judges often do and are likely to do better than jurors unburdened by training in exegesis. Patent construction in particular is a special occupation, requiring, like all others, special training and practice. The judge, from his training and discipline, is more likely to give a proper interpretation to such instruments than a jury; and he is, therefore, more likely to be right, in performing such a duty, than a jury can be expected to be.
Souter, like Archer, elevated a patent’s status as a legal document above its status as a technical document, all in the name of “the sound administration of justice.” But does his ruling really improve the administration of justice? Or does it open more questions than it resolves? A leading casebook that went to press shortly after the Supreme Court’s ruling (and thus before much if any empirical evidence of its effect existed), saw at least one set of further issues that the Federal Circuit would now have to address:
On the practical side, Markman raises a number of potentially challenging issues for the patent system. Most importantly, it threatens to change a number of routine practices in patent cases. For example, patent cases are usually appealed only after a full trial. The trial will very often include a claim interpretation issues but will of course also cover a host of other issues. Markman raises the threat that a full trial on many subordinate issues may become moot depending on how the Federal Circuit interprets a claim. For example, if the claim is interpreted as excluding all of the accused infringer’s products, there is no need to consider defenses. . . . Indeed, to assure that time and money are not spent on pointless issues, the parties in patent cases may well seek to obtain a trial court’s legal interpretation of claims and then ask the trial court to certify that finding for interlocutory appeal to the Federal Circuit.
This prediction was half right. Many litigants and trial judges are eager to gain early input from the Federal Circuit. No one wants to go through a full-blown trial for nothing, only to have an appeals court overturn an early legal ruling and make everyone—litigants and trial judges—do it all over again. As a result, numerous trial judges have held pre-trial claim construction “Markman hearings” and then asked the Federal Circuit to review their work.
The Federal Circuit, however, has been adamant in its refusal to help. One of the basic rules of our appellate system is that only “final judgments” are appealable. Intermediate rulings, even important ones, are subject to “interlocutory” appeal only under extraordinary circumstances. There are mechanisms, however, that let district court judges reach out for appellate help. A trial judge who reaches a particularly critical ruling may “certify” that ruling for an appeal. The appeals court can then review the ruling, consider the judge’s reasoning, and decide to help out—but it is under no obligation to do so. Appellate courts retain sole discretion over which certified questions they accept and which they send back. When it comes to claim construction, many trial judges have sought the Federal Circuit’s input. The Federal Circuit has rarely if ever acceded to such a request. This refusal has caused great consternation among patentees and trial judges alike. Parties, judges, and jurors pour valuable time, effort, and money into a trial, the losing party appeals, and the Federal Circuit issues a ruling overturning the claim construction from the Markman hearing. And such an overruling is hardly a rare event. Though the often cited statistic claiming that the Federal Circuit overturns forty percent of all claim constructions on appeal is almost certainly a vast exaggeration, there is little doubt that the court’s refusal to review constructions between Markman hearings and trials has led to an enormous waste of both public and private resources.
The Federal Circuit’s ruling in Markman thus leads to at least one clear failure of public policy—an inefficient allocation of litigation resources. But it hardly stops there. Policy failures abound in this brave new world of purely legal claim construction.
Few trial judges seem to know precisely what set of legal tools they’re supposed to use to construe these legal instruments called patents. And believe it or not, the point of greatest confusion involves the role of dictionaries! Now, there’s no way for anyone to understand that particular problem without a bit of background. So let’s back up to that great skill that Archer and Souter attributed to our judiciary: the ability to review linguistic disputes in light of “the evidence.”
Judges group evidence into a series of concentric circles, and work their way outward until they has resolved the disputed over claim terms. They begin with “intrinsic” evidence, or items that relate specifically to the document (here, a patent) that they are analyzing. Suppose, for example, that Markman had written a definition of “inventory” into his patent. He could have chosen any definition, either a narrow one recognizing that a dry cleaner’s inventory usually includes only dirty clothing, or a broad one explaining that for the purposes of his patent, term inventory also includes invoices, receipts, and cash on hand. Had he chosen that route, the dispute would have been over before it started. Regardless of any standard understanding of dry-cleaning inventory, the patent itself would have told anyone reading it—dry cleaner, patent lawyer, or judge—what the term meant in the claim.
But Markman included no such definition in the patent itself. Under such circumstances, judges turn to other documents specific to the patent, such as the prosecution history. If ambiguity remains after the judge has reviewed all of the intrinsic evidence, he must turn to “extrinsic” evidence, or generally available reference materials that might explain the patent’s terms. And that’s where things get confusing. What constitutes usable extrinsic evidence? How should a judge choose between two extrinsic sources that compound the ambiguity rather than clarify it? These are the sorts of questions that only the Federal Circuit can answer. Federal Circuit panels must review the various approaches that district court judges take, critique them, and craft a “claim construction toolkit” telling trial judges how to conduct valid Markman hearings.
Unfortunately, the Federal Circuit’s two most significant attempts to build this toolkit pointed judges—not to mention litigants—in subtly different directions. And these directions were different enough to compound the confusion, and to lead, once again, to arguably conflicting lines of cases—much as Archer contended the court had once done regarding the factual or legal status of patents.
Judge (now Chief Judge) Michel provided the first useful post-Markman guidelines:
In determining the proper construction of a claim, the court has numerous sources that it may properly utilize for guidance. . . . The court should look first to the intrinsic evidence of record, . . . [as] the most significant source of the legally operative meaning of disputed claim language. In most situations, an analysis of the intrinsic evidence alone will resolve any ambiguity in a disputed claim term. In such circumstances, it is improper to rely on extrinsic evidence. . . . Extrinsic evidence is that evidence which is external to the patent and file history, such as expert testimony, inventor testimony, dictionaries, and technical treatises and articles. (Although technical treatises and dictionaries fall within the category of extrinsic evidence, as they do not form a part of an integrated patent document, they are worthy of special note. Judges are free to consult such resources at any time in order to better understand the underlying technology and may also rely on dictionary definitions when construing claim terms, so long as the dictionary definition does not contradict any definition found in or ascertained by a reading of the patent documents.) However, . . . extrinsic evidence in general, and expert testimony in particular, may be used only to help the court come to the proper understanding of the claims; it may not be used to vary or contradict the claim language. Nor may it contradict the import of other parts of the specification.
Judge Linn offered an expanded—and confusingly different through probably not contradictory—set of advice:
Dictionaries, encyclopedias and treatises are particularly useful resources to assist the court in determining the ordinary and customary meanings of claim terms. Dictionaries are always available to the court to aid in the task of determining meanings that would have been attributed by those of skill in the relevant art to any disputed terms used by the inventor in the claims. When a patent is granted, prosecution is concluded, the intrinsic record is fixed, and the public is placed on notice of its allowed claims. Dictionaries, encyclopedias and treatises, publicly available at the time the patent is issued, are objective resources that serve as reliable sources of information on the established meanings that would have been attributed to the terms of the claims by those of skill in the art. Such references are unbiased reflections of common understanding . . . Indeed, these materials may be the most meaningful sources of information to aid judges in better understanding both the technology and the terminology used by those skilled in the art to describe the technology. . . It is entirely proper for both trial and appellate judges to consult these materials at any stage of a litigation, regardless of whether they have been offered by a party in evidence or not. Thus, categorizing them as “extrinsic evidence” or even a “special form of extrinsic evidence” is misplaced and does not inform the analysis. . . .Consulting the written description and prosecution history as a threshold step in the claim construction process, before any effort is made to discern the ordinary and customary meanings attributed to the words themselves, invites a violation of our precedent . . .
Most readers would look at these two sets of instructions and ask: “so what?” And it’s certainly true that to anyone without a legal training, Michel and Linn seem to be in broad agreement: If the patent is explicit about the meaning of its words, the inquiry is over. Otherwise, dictionaries and other forms of extrinsic evidence are useful ways to determine what they would mean to most readers—and thus to affix a concrete legal meaning to the disputed terms. But to the legally trained eye, the difference between these two toolkits is glaring. Michel noted that judges, like most other people, occasionally look words up in a dictionary. Linn launched into a panegyric of the genre. He elevated dictionaries to some rarified state above either intrinsic or extrinsic evidence. Where does that leave them? He didn’t tell us—for to do so would apparently not inform the analysis.
More to the point, though, this seemingly minor dispute led to two distinct schools of thought on the Federal Circuit about the role of dictionaries in claim construction. Polk Wagner and Lee Petherbridge, two scholars who have served as law clerks to Judge Clevenger, conducted an empirical study of every Federal Circuit ruling on claim construction since Markman. They reached a startling conclusion: different judges apply different methodologies. In particular, they identified two distinct approaches to claim construction; they called the first “procedural” and the second “holistic.” The procedural approach, as they describe it, “starts with a general presumption in favor of the ordinarily understood meaning of claim language, typically drawn from a relevant—often technical—dictionary, reference works, or common usage. . . . [A]ny suggested alteration from the ordinary meaning must be accompanied by significant proof that such an alteration is required under the circumstances.” The holistic method, on the other hand, “adopts a distinctly more free-form approach, seeking the correct meaning according to the particular circumstances presented. . . The holistic approach is significantly more relaxed than the procedural method in moving away from the abstracted ‘ordinary meaning’ of a term in favor of a more localized understanding.” Not surprisingly (given their characterizations of the two approaches), they emerge as strong advocates of the procedural approach.
Their analysis of the differences, however, is almost certainly more striking than their ultimate recommendations. For as they describe it, three of the courts judges—Dyk, Clevenger, and Linn—are strong and consistent proceduralists. These judges employ Linn’s claim construction toolkit, relying heavily as it does on dictionaries to provide an abstract notion of “common meaning.” Three other judges—Lourie, Bryson, and Newman—are strong and consistent holistics. These judges employ Michel’s claim construction toolkit, focus closely on the specifics (and often the peculiarities) of the intrinsic evidence, and use dictionaries only as necessary. The court’s remaining six judges, including Michel himself, fall somewhere in the middle; they are considerably more deferential to the intrinsic evidence than are the proceduralists, but more likely to cite favorably to dictionary definitions than are the holistics.
Now once again, this debate may not sound like much, but when there’s as much riding on an outcome as there often is in a patent case, seemingly subtle distinctions in claim construction methodology can have enormous effects. According to Wagner and Petherbridge, the court’s claim construction methodology hinges upon the makeup of the panel hearing a case. Though they do not say so explicitly, their analysis implies that the proceduralist and holistic methodologies lead to different outcomes often enough to cause problems and promulgate uncertainty. If that implication is true, Federal Circuit opinions must exhibit significant panel dependence; the Clerk of Court’s roulette wheel may determine the case’s outcome when he assigns it to a panel. Litigants, who have little or no advance warning of the makeup of their panel, cannot possibly prepare accordingly. For a court chartered specifically to remove randomness and impose predictability upon the patent law, this criticism is quite damning.
Wagner and Petherbridge see an easy solution—and one already in the making. They “recommend that the court recognize the importance of methodology and move to standardize the procedural methodological approach—the evidence suggests that the procedural approach is inherently more consistent than holistic analyses. Individual judges also have an important role to play in enforcement and evangelism.” I am less enamored of the procedural approach. Though I would never argue against methodological consistency, I see the reliance upon dictionary definitions, and at times their elevation of dictionary definitions above evidence intrinsic to the patent itself, to be problematic. And I’m hardly alone. Consider what happened—shortly after Wagner and Petherbridge’s article appeared in print—to Novartis.
Novartis, of course, is a major pharmaceutical company. One of its patents described a way to administer cyclosporin—a drug that reduces the risk of a patient’s rejecting a transplanted organ—as a “hydrosol.” What is a hydrosol? According to Novartis, any time that cyclosporin dissolves in water, it forms a hydrosol—even if it doesn’t dissolve until it’s inside the patient’s stomach. So Novartis sued Eon Labs for selling capsules in which the cyclosporin is dissolved in ethanol; when a patient ingests the capsule, the ethanol frees the cyclosporin, which then combines with stomach water to form a hydrosol. Eon predictably disagreed, arguing that Novartis’ patent covered only hydrosols formed outside the body.
Now, believe it or not, this sort of argument is actually quite common among pharmaceutical companies, and in many ways it’s something of a garden variety patent dispute. Which of these interpretations is correct? To a layman (or at least to this layman), Eon’s argument sounds more reasonable. But no one writes patents for laymen. Depending on where you stand on the fact/law divide, the intended audience for Novartis’s patent was either pharmacologists, patent lawyers and judges, or both. One member of that audience, the district court judge, agreed with Eon’s construction, resolved the linguistic dispute at the Markman hearing, and granted a summary judgment in Eon’s favor. Novartis appealed to the Federal Circuit.
The Court of Clerk spun his wheel, and the case landed in front of Judges Clevenger, Dyk, and Prost. With two of the court’s strong proceduralists on the panel, Wagner and Petherbridge would have had little doubt as to how the analysis would proceed. The court would look to a dictionary for the plain, commonly understood, meaning of “hydrosol,” and then turn to the intrinsic evidence to see if anything necessarily contradicted that dictionary definition. Clevenger and Dyk would both be on board, and Prost—one of the court’s swing methodologists—could either join them or dissent. But that’s not quite what happened. Dyk did pursue a proceduralist analysis that led him to agree with the district court. Prost joined him to form a panel majority affirming Eon’s victory. But Clevenger dissented in a manner that revealed not only a stark division within the proceduralist camp, but also an inherent limitation of the entire methodology:
I differ not with the majority over the tools of claim interpretation to be applied; my difference is in how the majority has used those tools to arrive at its perceived ambiguity. Let me explain.
The majority starts with its dictionaries in hand, and goes first, as makes sense, to the word “hydrosol.” It uses Webster’s Third New International Dictionary to learn that “hydrosol” means “a sol in which the liquid is water.” Other dictionaries could be used to arrive at the same definition . . . [The majority] pursues the search in the dictionary, relying on Webster’s Third New International Dictionary (2002). There, it finds as a definition for “sol” the following: a “dispersion of solid particles in a liquid colloidal solution.” For the majority, this definition from this particular chosen dictionary is crucial, for it permits the court to move to a further degree of separation away from the word “hydrosol” to investigate the meaning of “solution.” Had the majority used the Merriam Webster Medical Dictionary (2003) definition of “sol,” it would have been deprived of the opportunity to pursue the meaning of “solution,” because the more pertinent medical dictionary defines “sol” simply as “a fluid colloidal system; especially: one in which the dispersion medium is a liquid.” The pursuit of the meaning of “solution” is also blocked if one had turned to Webster’s Ninth New Collegiate Dictionary (1984) (the copy provided to my chambers by the court), which defines “sol” as “a fluid colloidal system; esp: one in which the continuous phase is a liquid.”
Having found a need to define “solution”—remember, the need is in order to know the meaning of “hydrosol”—the majority again relies on its dictionary in the next degree of separation to produce two meanings for “solution.” One meaning, “a liquid containing a dissolved substance” is deemed by the majority, correctly, to be broad enough to relate back to “hydrosol” without imposing any restrictions on the manufacturing site of the hydrosol. The other meaning, however, is pay dirt for the majority. That second definition of “solution” is “a liquid and usu. aqueous medicinal preparation with the solid ingredients soluble.” . . . In the next degree of separation, the majority embraces the term “medicinal preparation,” and forthwith abandons Webster’s in favor of The Oxford English Dictionary 548-49 (2d ed. 1989), to plumb the depths of the meaning of the word “medicine.” (The reader should not forget that the word we are trying to define is “hydrosol,” not “medicine.”). . . .
At the end of the long search in various dictionaries, the majority concludes that “hydrosol” is ambiguous because of “multiple possible” conflicting dictionary definitions. I am at a loss to understand why this dictionary search creates such an ambiguity. . . . “Hydrosol” simply means “a sol in which the liquid is water.” . . . Dictionaries are fine tools to assist in the exercise of claim interpretation, for sure, but in this case the majority has simply overworked the dictionaries to a point of error.
The dance of the dictionaries had gone too far even for one of the court’s proceduralists; his critique served as something of a wakeup call. The court realized that it had reached a critical juncture. It needed to reconcile the differing implications of Michel’s and Linn’s approaches to claim construction. It needed to decide just how central dictionaries were to the analysis deemed so inherently legal that judges had to conduct it before even impaneling a jury. And it needed to do so quickly.
As luck would have it, about the same time that Novartis and Eon were slugging it out over “hydrosol,” a man named Edward Phillips sued the AWH Corporation for infringing his patented invention of vandal-resistant building modules. The key word in this dispute was “baffle.” Once again the Clerk of Court spun his wheel, and once again it fell upon Dyk. This time, however, he found himself sitting with Judges Newman and Lourie, two of the court’s strongest holistics. To no one’s surprise, these two holistics formed the court’s majority. Lourie looked at both the intrinsic evidence and a dictionary. He concluded that though Webster’s defines a “baffle” as “something for deflecting, checking, or otherwise regulating flow,” Edwards himself had defined the term somewhat more specifically in his patent. Dyk dissented, arguing that the dictionary definition should have prevailed.
Phillips, like most losing parties, asked the court to reconsider its decision en banc. The court saw this dispute over a single word governing a patent in an easily understandable area of technology as a golden opportunity to clean up its claim construction jurisprudence. It agreed to hear the case, framed specific questions that it wanted Phillips and AWH to address, and invited all interested parties to submit answers in the form of amicus briefs. The court’s questions addressed: The relative roles of technical and general purpose dictionaries; the best ways to reconcile conflicts between dictionaries; the proper role for dictionaries among judges who turn first to intrinsic evidence; the proper way to view the relationship between the proceduralist and holistic methods; and a few related technical legal points. By and large, though, it was clear that the proceduralists had stacked the deck. The majority of the questions, and the way that the court phrased them, elevated dictionaries to a hallowed position, and requested feedback primarily to determine just how sacred these books are.
Only two judges even took the time to call their colleagues to task about the biased nature of the questions. “For the sake of completeness,” Judge Rader would have at least balanced the inquiry by adding a complementary line of questions. He would have asked about the relative merits of an “algorithmic” approach that imposed uniformity upon claim construction and an approach that selected patent-appropriate tools from a claim construction toolkit (i.e., Wagner and Petherbridge’s holistic approach). Mayer (by that time, Chief Judge), went back to the heart of the matter: “Nearly a decade of confusion has resulted from the fiction that claim construction is a matter of law, when it is obvious that it depends on underlying factual determinations which, like all factual questions if disputed, are the province of the trial court, reviewable on appeal for clear error. To pretend otherwise inspires cynicism.”
Once again, we seem to be verging upon a major failure of public policy. Ambiguous patents serve no one. They motivate insufficient innovation, complicate the diffusion of knowledge, chill competitive entry, engender needless litigation, frustrate trial judges, and aggravate appellate judges. Yet after more than a decade of concerted effort, claim construction jurisprudence has devolved to the level at which an appellate court must ask the patent bar how to best use dictionaries. Could that possibly be what Justice Souter had in mind when he agreed that judges were better equipped than jurors to engage in the complex evidentiary task of claim construction?
Is this problem endemic? Is there any way to resolve this issue? Obviously, it is impossible to remove all ambiguity; language is a finite precision instrument. But perhaps there are ways to minimize it. Perhaps the reason that this method has been so elusive is that the courts have always approached the question from the perspective of law, rather than from the perspective of policy. The legal perspective begins with a debate over the nature of a patent—is it a legal document or a technical document—and then slides inexorably into discussions of evidentiary analysis. We’ve already seen where that approach leads.
A policy perspective begins with a very different set of questions. Rather than asking what a patent is, it asks what a patent does. A patent promotes progress in two ways: First, it motivates innovation by granting the inventor exclusive rights in the claims. Second, it advances public knowledge by explaining the invention in the written description. These two parts of the patent specification also provide “public notice” of the claims’ scope. A skilled artisan—the group from which both competitors and infringers are likely to spring—reading the patent should emerge with a reasonable understanding of what the patentee owns and what remains in the public domain. The key claim construction challenge from a policy perspective is thus how to ensure that the public has reasonably accurate notice of the claims’ scope.
The Federal Circuit encountered a similar challenge early in its history. Yujiro Yamamoto, inventor of a voice activated dictaphone, was embroiled in patent litigation against Dictaphone when the PTO decided to reexamine his patent. Though the PTO reexamines relatively few patents, such proceedings are hardly rare. They usually occur because either the inventor discovered a flaw in his claim that would cost him an infringement suit, or as in Yamamoto’s case, because some competitor who had no opportunity to participate in the initial negotiation between the inventor and the PTO convinces the PTO that the patent may have issued in error. In a reexamination proceeding, the examiner reopens the prosecution history and engages in new rounds of correspondence with the inventor. Once again, this correspondence becomes a public document when the patent “reissues.”
In Yamamoto’s case, as in most cases, these two proceedings pushed in opposite directions. The PTO insisted that if Yamamoto were to read his patent broadly, it would have to rule his patent invalid, and cancel it rather than reissue it. The examiner asked Yamamoto to adopt a narrow definition of a particular claim term by officially “disclaiming” a reasonable broader meaning that would have encompassed not only Yamamoto’s invention, but also important prior art recording devices. The PTO canceled the patent. Yamamoto appealed. He argued—accurately—that the courts like to protect patents. When a trial judge sees two reasonable claim constructions, a broad one that would render the patent invalid and a narrow one that would maintain its validity, the judge is required to adopt the narrow definition—allowing the patentee to keep her patent despite possibly losing her infringement case. Yamamoto wanted the PTO to apply the narrower construction without any of his own written input, and reissue the patent so that he could then propose the broader construction to the trial judge, who likely lacked the examiner’s training and might not invalidate the broader patent.
At the Federal Circuit, Judge Baldwin not only shot him down, but actually took the time to explain the public policy rationale behind applying different rules in reexamination proceedings and infringement trials:
The PTO broadly interprets claims during examination of a patent application since the applicant may amend his claims to obtain protection commensurate with his actual contribution to the art. This approach serves the public interest by reducing the possibility that claims, finally allowed, will be given broader scope than is justified. Applicants’ interests are not impaired since they are not foreclosed from obtaining appropriate coverage for their invention with express claim language. . . . When an application is pending in the PTO, the applicant has the ability to correct errors in claim language and adjust the scope of claim protection as needed.
In Yamamoto’s case, the PTO did exactly what it was supposed to do: it interpreted his claims broadly and invalidated his patent. The Federal Circuit adopted the entirely sensible rule that inventors who do not avail themselves of the opportunity to narrow their claims to exclude prior art do not deserve patents.
Now flip the public interest argument. The examiner has given all potentially ambiguous terms their broadest reasonable interpretation. The inventor has incorporated explicit narrowing terms where necessary. The examiner okays the new language and the PTO issues the patent. Some unsuspecting member of the public, of ordinary skill in the prior art, comes along and reads it. She may or may not notice potential ambiguities, but being a competent reader, she assigns reasonable definitions to at least enough of the patent’s words to understand its teachings. As a potential competitor, she also pays particularly close attention to the claim language to ensure that she does not infringe. She does not, however, have the opportunity to discuss the language with either the patentee or the examiner. What sorts of interpretations might she apply to potentially ambiguous terms? Presumably, if she is reading the patent in good faith and exerting reasonable efforts to avoid infringement, she will interpret all claim language “reasonably.” Sound policy requires rules that encourage her to do what we want her to do: learn the disclosed science, incorporate it into her own knowledge of the state-of-the-art of her field, and if possible, develop competing, noninfringing products. The best way to motivate that behavior is to bind her only to the narrowest reasonable interpretation of potentially ambiguous terms.
But wait! Wouldn’t that reduce the inventor’s motivation? Perhaps somewhat. After all, a patent is a bargain between the inventor and the public. Anything that decreases the price that the public pays also decreases the price that the inventor receives. But the imposition of a “narrowest reasonable interpretation” requirement does more than that: It places the burden of disambiguation on the shoulders of the party best able to bear it. Patentees are free to draft any language that they desire. If a patentee chooses words with multiple reasonable definitions, it is the patentee’s responsibility to make sure that readers know which one he intended. Before the patent issues, the PTO works to ensure that he can’t claim things that existed prior to his invention. After the patent issues, district court judges must ensure that the public doesn’t relinquish anything that the inventor may have failed to request.
So anything goes? Any accused infringer who enters a courtroom asserting an interpretation necessarily wins? Not at all. That’s where we remember that we’re in a court of law, not a policy seminar. The accused infringer will undoubtedly arrive armed with an interpretation that removes her product from the patent’s scope. The court must then answer a key question: Is her interpretation reasonable? Courts answer questions about reasonableness all the time. In fact, juries answer questions about reasonable behavior all the time. Defendants who trot out single obscure sources favoring offbeat interpretations are unlikely to prevail very often. We can finally agree with Judge Mayer about the factual nature of the interpretive task. And we can restore patents to their proper roles as technical documents rather than as legal ones.
Is this approach too radical? Perhaps. It is certainly not along any glide path foreseeable from recent Federal Circuit jurisprudence. It completely upends the Markman ruling, and represents a significant change in patent law—the very sort of occurrence that we founded the Federal Circuit to avoid. It is, at heart, an argument grounded in the economics of transaction costs rather than in the legal tradition of patent law. Such arguments are common in the “law and economics” school of legal analysis, and draw heavily upon the work of Ronald Coase, winner of the 1991 Nobel Prize in economics. According to Coase:
[With zero transaction costs,] the allocation of resources remains the same whatever the legal position . . . . [W]ith positive transaction costs, the law plays a crucial role in determining how resources are used. But it does more than this. With zero transaction costs. . . contractual arrangements will be made to modify the rights and duties of the parties so as to make it in their interest to undertake those actions which maximize [value.] With positive transaction costs, some or all of these contractual arrangements become too costly to carry out. The incentives to take some of the actions which would have maximized [value] disappear. What incentives will be lacking depend on what the law is. . .
Ambiguity in a patent creates transaction costs. It generates uncertainty among members of the public, makes inadvertent infringement more likely, and allows a patentee to file infringement suits even when he knows full well that the accused device falls outside the scope of the invention that he patented. On the flip side, it can also allow intentional infringers to escape liability. Ambiguous claim language therefore creates two types of publicly detrimental transaction costs: tangible litigation expenses, and less-tangible improper pricing that arises from wrongly deterred products and from undeterred infringing products. The public as a whole loses while some patentees and some infringers gain.
There are also positive transaction costs in disambiguation. Someone must bear the cost of ensuring that every word in every patent is clear enough for competitors who behave reasonably and in good faith to avoid inadvertently infringing. The law can assign this burden in one of two places. Either the inventor and the PTO can bear the cost during examination, or litigants and the court system can bear the cost during infringement trials. Though the latter assignment avoids some costs on the majority of patents that never arise in litigation, it also increases the costs of prosecuting our most valuable inventions—precisely those that give rise to commercial competition and infringement suits. Our current system assigns most of the costs to litigants and courts. A “narrowest reasonable interpretation” standard would assign the bulk of the cost to inventors, and in particular to those few inventors who believe that their inventions are important enough to warrant the added expense. Overall costs would almost certainly decline, and many intangible transaction costs of inappropriate pricing would evaporate. Progress would be promoted and public policy well served.
All told, claim construction poses a cautionary tale. The Federal Circuit identified it—correctly—as a source of uncertainty impeding the smooth flow of the patent system. In keeping with its charge, it tried to remove this uncertainty by concentrating the task in judicial hands, rather than in the hands of untrained and often unpredictable juries. A decade after the effort began, reports of rampant uncertainty and reversals of district court rulings persist. Trial judges know that they lack the tools to derive confident constructions, and repeatedly beg the Federal Circuit for help. The Federal Circuit refuses virtually all such pleas. Federal Circuit judges try to explain the procedure, but even subtle differences among their explanations can cause more confusion than they resolve. And even these attempts have degenerated to a new code for dictionary usage. The court is likely to stay baffled until it steps off the path it has taken and considers a new one. Contemporary claim construction jurisprudence creates confusion and motivates expensive litigation. It is not moving us towards a predictable, reliable patent system—the basic characteristics of any system truly capable of promoting progress. The Federal Circuit may be trying, but effort alone is not enough. Markman was a failure. The policy we have is not the policy we want.
The trackback URL for this entry is: http://www.theinformationist.com/index/trackback/97/YY6oV2L9/
Trackbacks:
No trackbacks yet.Comments:
No comments yet.Next entry: Darknet on Hold
Previous entry: Through the Looking Glass