Cybor Corporation v. Fas Technologies, Inc., and Fastar Ltd., Defendants-Cross

*1473 RADER, Circuit Judge,

dissenting from the pronouncements on claim interpretation in the én banc opinion, concurring in the judgment, and joining part IV of the en banc opinion..

The en banc court chooses this case, resolvable under the prosecution history estop-pel doctrine, to state that it accords no deference to the district court’s central role in patent litigation, namely claim interpretation. Yet, based on this court’s unanimous concurrence in the judgment, the standard of review. does not affect the outcome in this case. Accordingly, this ease obscures what really is at stake when claim construction is subject to de novo review and appellate revision. 1

In Markman I, this court en banc declared that claim interpretation resides solely with the judge. The Supreme Court agreed. See Markman v. Westview Instruments, Inc., 517 U.S. 370, 371-73 , 116 S.Ct. 1384, 1387 , 134 L.Ed.2d 577 (1996) (.Markman II). By removing lay juries from complex technological decisions, these decisions promised to improve the predictability and uniformity of patent law. See Markman II, 517 U.S. at 390-91 , 116 S.Ct. at 1395-96 .

To evade the strictures of the Seventh Amendment, Markman I necessarily reasoned that claim construction is purely a matter of law. As a pure question of law, claim construction became subject to plenary appellate review. Because jury involvement remained the focus of Markman I, the Supreme Court did not address appellate review of claim construction. Instead the Supreme Court repeatedly intimated that claim construction was not a purely legal matter. See, e.g., Markman II, 517 U.S. at 377-79 , 116 S.Ct. at 1389-91 (calling the construction of a term of art following receipt of evidence a “mongrel practice”); id. at 388 , 116 S.Ct. at 1395 (suggesting that claim interpretation “‘falls somewhere between a pristine legal standard and a single historical fact’ ” (quoting Miller v. Fenton, 474 U.S. 104, 114 , 106 S.Ct. 445, 451-52 , 88 L.Ed.2d 405 (1985))); id at 390, 116 S.Ct. at 1396 (“notwithstanding [claim construction’s] evidentiary underpinnings”). Nonetheless, in the time since Markman II, this court has repeatedly extolled its own authority to “review the issue of claim interpretation independently without deference to the trial judge.” Exxon Chem. Patents, Inc. v. Lubrizol Corp., 64 F.3d 1553, 1556 , 35 USPQ2d 1801, 1803 (Fed.Cir.1995) (Lubrizol) (emphasis added); see also Hoechst Celanese Corp. v. BP Chem. Ltd, 78 F.3d 1575, 1578 , 38 USPQ2d 1126, 1128 (Fed. Cir.), cert. denied — U.S. -, 117 S.Ct. 275 , 136 L.Ed.2d 198 (1996). Indeed, at one 'point, this court noted that its law requires “independent determination of the construction of the claims, as a matter of law, unencumbered by the trial process.” Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., 72 F.3d 857, 863 , 37 USPQ2d 1161, 1165 (Fed.Cir.1995), vacated and remanded on other grounds, — U.S.-, 117 S.Ct. 1240 , 137 L.Ed.2d 323 (1997). Unencumbered by the trial process? Far from an encumbrance, the Supreme Court suggests that the trial should be the “main event.” Wainwright v. Sykes, 433 U.S. 72, 90 , 97 S.Ct. 2497, 2508 , 53 L.Ed.2d 594 (1977).

Because jury issues dominated Markman I, this court has yet even to receive briefing and oral argument on the proper standard of review for a trial court’s claim construction. Nonetheless this court relies on its earlier en banc decision, Markman I, 52 F.3d at 979, to make claim construction purely a question of *1474 law, subject to independent appellate review without deference to dr encumbrance by the trial process. To my eyes, this rejection of the trial process as the “main event” will undermine, if not destroy, the values of certainty and predictability sought by Markman 1 .

I.

The question of the proper standard of review seems an esoteric legal topic of interest only to law professors and appellate judges. In most cases, however, the review standard influences greatly both the trial judges who preside over the trial process and patent practitioners who must advise clients to accommodate their business plans to an uncertain legal regime.

From the vantage point of trial judges, Markman I dictates many deviations from the normal procedural course for litigation. 2 Perhaps the central deviation, however, affects the trial court’s discretion to use expert testimony. When confronted with sophisticated technology, district court judges often seek testimony from experts to help them-understand and interpret the claim. Under the guise of setting standards for claim construction, this court instructs experienced trial judges that they may use experts to understand, but not to interpret, the claim terms. As a matter of logic, this instruction is difficult to grasp. What is the distinction between a trial judge’s understanding of the claims and a trial judge’s interpretation of the claims to the jury? Don’t judges instruct the jury in accordance with their understanding of the claims? In practice, how does this court’s lofty appellate logic work? As this court acknowledges, a trial court must often resort to experts to learn complex new technologies. See, e.g., Markman I, 52 F.3d at 986. What happens when that learning influences a trial judge’s interpretation of the *1475 claim terms? Are trial judges supposed to disguise the real reasons for their interpretation? How will this perverse incentive to “hide the ball” improve appellate review?

As a matter of legal analysis, the en bane court’s direction to trial judges is equally hard to justify. The objective of claim interpretation is to discern the meaning of the claim terms to one of ordinary skill in the art at the time of invention. See Multiform Desiccants, Inc. v. Medzam, Ltd., 133 F.3d 1473, 1477 , 45 USPQ2d 1429, 1432 (Fed.Cir. 1998) (“It is the person of ordinary skill in the field of the invention through whose eyes the claims are construed.”). What then defeats the relevance of the testimony'of one of skill in the art at the time of invention? Of course this relevant testimony must hot conflict with or attempt to trump contemporaneous intrinsic evidence from the patent document itself, see Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582-83 , 39 USPQ2d 1573, 1576-77 (Fed.Cir.1996), but both trial and appellate judges are poised to halt that abuse. Moreover, by assigning claim interpretation to the judge, Markman II has already corrected the major source of the problem with experts, namely their ability to influence lay jurors with the strength of their resumes rather than the strength of their reasoning. In any event, it seems a contradiction to bar those of skill in the art at the time of invention from a search for the meaning of terms to one of skill in the art at the time of the invention. In effect, the en banc opinion has sub silentio redefined the claim construction inquiry as “how a lawyer or judge would interpret the term.” District courts have already expressed their frustration with the strictures of Mark-man I:

When two experts testify differently as to the meaning of a technical term, and the court embraces the view of one, the other, or neither while construing a patent claim as a matter of law, the court has engaged in weighing evidence and making credibility determinations____ But when the Federal Circuit Court of Appeals states that the trial court does not do something that the trial court does and must do to perform the judicial function, the court knowingly enters a land of sophistry and fiction.

Lucas Aerospace, Ltd. v. Unison Indus., LP, 890 F.Supp. 329 , 333-34 n. 7, 36 USPQ2d 1235, 1239 n. 7 (D.Del.1995); see also Elf Atochem North Am., Inc. v. Libbey-Owens-Ford Co., 894 F.Supp. 844, 857 , 37 USPQ2d 1065, 1075 (D.Del.1995); In re Mahurkar Double Lumen Hemodialysis Catheter Patent Litigation, 831 F.Supp. 1354, 1359 , 28 USPQ2d 1801, 1805 (N.D.Ill.1993) (“[Jjudges should not pretend that all nominally ‘legal’ issues may be resolved without reference to facts____ What seems clear to a judge may read otherwise to [one skilled in the art].”).

II.

From the patent practitioner’s standpoint, this court’s enthusiastic assertion of its unfettered review authority has the potential to undercut the benefits of Markman I. Mark-man I potentially promised to supply early certainty about the meaning of a patent claim. This certainty, in turn, would prompt early settlement of many, if not most, patent suits. Once the parties know the meaning of the claims, they can predict with some reliability the likelihood of a favorable judgment, factor in the economies of the infringement, and arrive at a settlement to save the costs of litigation. 3 Markman I promised to provide this benefit early in the trial court process. To provide fairness under the Mark-man I regime, trial judges would provide claim interpretations before the expense of trial. Patent practitioners would then be armed with knowledge of the probable out *1476 come of the litigation and could facilitate settlement.

The problem with this plan was in its implementation because as a question of law, claim interpretation is subject to free review by the appellate court. The Federal Circuit, according to its own official 1997 statistics, reversed in whole or in part 53% of the cases from district courts (27% fully reversed; 26% reversed-in-part). Granted this figure deals with all issues in cases with many issues. Nonetheless, one study shows that the plenary standard of review has produced reversal, in whole or in part, of almost 40% of all claim constructions since Markman 7 . 4 A reversal rate in this range reverses more than the work of numerous trial courts; it also reverses the benefits of Markman I. In fact, this reversal rate, hovering near 50%, is the worst possible. Even a rate that was much higher would provide greater certainty.

Instead, the current Markman I regime means that the trial court’s early claim interpretation provides no early certainty at all, but only opens the bidding. The meaning of a claim is not certain (and the parties are not prepared to settle) until nearly the last step in the process — decision by the Court of Appeals for the Federal Circuit. To get a certain claim interpretation, parties must go past the district court’s Markman I proceeding, past the entirety of discovery, past the entire. trial on the merits, past post trial motions, past briefing and argument to the Federal Circuit — indeed past every step in the entire course of federal litigation, except Supreme Court review. In implementation, a de novo review of claim interpretations has postponed the point of certainty to the end of the litigation process, at which point, of course, every outcome is certain anyway.

In practical terms, this implementation record has other perverse effects. Trial attorneys must devote much of their trial strategy to positioning themselves for the “endgame” — claim construction on appeal.. As the focus shifts from litigating for the correct claim construction to preserving ways to compel reversal on appeal, the uncertainty, cost, and duration of patent litigation only increase. Thus, the en banc court’s de novo regime belies the purpose and promise of Markman I.

Several high profile appeals have illustrated the problem created by the Federal Circuit’s high reversal rate. In Lubrizol, 64 F.3d 1553 , and J.T. Eaton & Co. v. Atlantic Paste & Glue Co., 106 F.3d 1563 , 41 USPQ2d 1641 (Fed.Cir.1997), this appellate court rejected not only the trial judge’s claim reading, but also the readings advocated by both parties at trial and the readings advocated by all experts in the trial. In fact, in Eaton, the Federal Circuit did not base its interpretation on the specification or the totality of the prosecution history, but instead found its meaning in a brief excerpt from the affidavit of a single expert witness at the end of a lengthy prosecution. See 106 F.3d at 1568, 1570 . If the parties might succeed in convincing the Federal Circuit to reverse an entire trial result with an argument never presented to the trial court or with a brief excerpt from hundreds of pages of prosecution (again not presented to the trial court), would they be wise to settle after a trial court’s reading of the claims?

One other case makes the point even more persuasively. In CVI/Beta Ventures, Inc. v. Tura LP, 112 F.3d 1146 , 42 USPQ2d 1577 (Fed.Cir.1997) (Tura), the Federal Circuit reversed its own earlier claim interpretation as a question of law when the defendant and stage of the proceedings changed. On a motion for a preliminary injunction in an infringement action, a Maryland district court interpreted the 3% elasticity limitation in a claim for flexible eyeglass frames. On appeal the Federal Circuit reviewed this claim interpretation as a question of law and affirmed in a nonprecedential opinion. See *1477 CVI/Beta Ventures, Inc. v. Custom Optical Frames, Inc., 1996 WL 338388 (Fed.Cir.1996) (nonpreeedential); see also Tura, 112 F.3d at 1160 n. 7. Understanding that the law must not change from case to case or from circumstance to circumstance, a New York district court applied the Federal Circuit’s claim interpretation in a separate infringement action involving the same patent. After trial, the defendant appealed. This time the Federal Circuit reversed the district court’s claim interpretation. See Tura, 112 F.3d at 1149 (“We conclude that the district court did err in its claim construction.”) (emphasis added). Although this court referred to the New York district court’s “error,” the Federal Circuit had in fact reversed itself.

One potential trial court response to the CVI/Beta problem might be the issuance of “tentative” claim constructions as a matter of law. See International Communication Materials, Inc. v. Ricoh Co., Ltd., 108 F.3d 316 , 41 USPQ2d 1957 (Fed.Cir.1997). This response, of course, would further frustrate the parties’ desire to receive a certain claim construction as early as possible. ;

Regardless, if the Federál Circuit’s reading of the very same claim can vary from one appeal to the next, every patent litigant has an incentive to appeal every action to the Federal Circuit in hopes, that the statistics will hold up and eventually the appellate court will reverse. Even the Federal Circuit’s claim interpretations as questions of law are not certain. Is this the “uniformity” outlined by the Supreme' Court in Markman II?

III.

Because patent trial practitioners understand the distinct prospect of overturning trial court results on appeal, the trial arena loses some of its luster as the center stage of the dispute resolution drama. Instead the trial court becomes a ticket to the real center stage, the Court of Appeals for the Federal Circuit. Taking a cue from the Supreme Court, this court would more wisely take a functional approach to setting a standard of review for claim construction.

In Markman II, the Supreme Court noted that neither history nor precedent provided “clear answers” about the role of the jury and the factual or legal nature of claim construction. 517 U.S. at 388-90 , 116 S.Ct. at 1395 . Therefore, the Court pursued a functional inquiry to determine whether the judge or jury could best balance the complexities of claim construction. See id. A similar functional inquiry might best clarify the roles of the trial and appellate benches during claim interpretation.

The Supreme Court has provided some guidelines for such a functional approach. The Court counsels appellate courts to defer “when it appears that the district court is ‘better positioned’ than the appellate court to decide the issue in question or that probing appellate scrutiny will not contribute to the clarity of legal doctrine.” Salve Regina College v. Russell, 499 U.S. 225, 233 , 111 S.Ct. 1217, 1222 , 113 L.Ed.2d 190 (1991). At another point, the Court cautions: “[T]he reviewing attitude that a court of appeals takes toward a district court decision should depend upon ‘the respective institutional advantages of trial and appellate courts.’ ” First Options of Chicago, Inc. v. Kaplan, 514 U.S. 938, 948 , 115 S.Ct. 1920, 1926 , 131 L.Ed.2d 985 (1995) (quoting Salve Regina, 499 U.S. at 233 , 111 S.Ct. at 1222 ).

Applying this general counsel, the trial judge enjoys a potentially superior position to engage in claim interpretation. For the complex case where the claim language and specification do not summarily dispose of claim construction issues, the trial court has tools to acquire and evaluate evidence that this court lacks. Trial judges can spend hundreds of hours reading and rereading all kinds of source material, receiving tutorials on technology from leading scientists, formally questioning technical experts and testing their understanding against that of various experts, examining on site the operation of the principles of the claimed invention, and deliberating over the meaning of the claim language. If district judges are not satisfied with the proofs proffered by the parties, they are not bound to a prepared record but may compel additional presentations or even employ their own court-appointed expert.

An appellate court has none of these advantages. It cannot depart from the record of the trial proceedings. To properly marshal its resources, the appellate bench must *1478 enforce strict time and page limits in oral and written presentations. 5 Moreover a sterile written record can never convey all the nuances and intangibles of the decisional process. Indeed a careful consideration of the institutional advantages of the district court would counsel deference. This court’s categorical response that claim interpretation involves no factual assessments does not advance a functional analysis of trial and appellate roles in claim construction. As a matter of fact (so to speak), claim construction requires assessment of custom and usage in the relevant art, assessment of events during prosecution, assessment of the level of ordinary skill in the art, assessment of the understanding of skilled artisans at the time of invention — to name just a few factual components of the complex process of claim interpretation. A careful . functional analysis counsels deference for district court claim interpretations.

IV.

The Supreme Court may have offered a path out of this predicament. At least three times in Markman II, as noted earlier, the Court alluded to the factual component of claim interpretation. At no point did Mark-man II address the appropriate standard of review. Nonetheless this court misses the opportunity to improve certainty in patent practice by giving appropriate deference to trial court claim interpretations, particularly in complex cases. If this court accords more deference to trial court interpretations in the complex cases, soon the district courts will provide the desired certainty early in the process. At that point, Markman I will fulfill its promise. Administration of patent law will move toward less costly disputes and earlier settlements.

Markman I set out to improve patent law administration by removing uncertainties from the dispute resolution process (the chief uncertainty, of course, being jury results). Inadvertently the reasoning in Markman I has postponed the point of certainty to the extreme end of the judicial process. This delay both disrupts the orderliness of trials and practitioners’ hopes for more efficient and earlier claim constructions. The Supreme Court removed the jury uncertainty with its decision and, by recognizing a factual component in claim interpretation, provided a way for this court to accomplish much of its early goals. By according some deference where appropriate, this court can restore the trial court’s prominence in the claim interpretation function and bring again more certainty at an earlier stage of the judicial process. Applying the Supreme Court’s reasoning for habeas proceedings to claim construction, “adoption of [a rule of appropriate deference] ... will have the salutary effect of making the [district court’s] trial on the merits the ‘main event,’ so to speak, rather than a ‘tryout on the road’ for what will later be the determinative [appeal to the Federal Circuit].” Wainwright v. Sykes, 433 U.S. at 90 , 97 S.Ct. at 2508 . This certainty, in turn, would stimulate more settlements and efficient decision making — the promise of Markman I. With this reasoning in mind, I respectfully dissent from the claim interpretation pronouncements of the en banc opinion, concur in the judgment, and join only part IV .of the en banc opinion.

PAULINE NEWMAN, Circuit Judge,

with whom Chief Judge MAYER joins; additional views.

As this ease illustrates, perfection is elusive in the aftermath of the Federal Circuit’s decision in Markman v. Westview Instruments, Inc., 52 F.3d 967 , 34 USPQ2d 1321 (Fed.Cir.1995) (en banc), aff'd, 517 U.S. 370 , 116 S.Ct. 1384 , 134 L.Ed.2d 577 , 38 USPQ2d 1461 (1996). The expectation of greater stability in the application of patent law — thus enhancing consistency in result, reducing the cost of litigation, and indeed reducing litigation by diminishing the uncertainties of jury trials — has not been well achieved..

*1479 Most of the shortfalls between expectation and reality arise from the manner of implementation of our de novo authority for claim interpretation. I cite three principal areas. The first area relates to the treatment of certified questions. Although the district courts have extended themselves, and so-called “Markman hearings” are common, this has not been accompanied by interlocutory review of the trial judge’s claim interpretation. The Federal Circuit has thus far declined all such certified questions. Indeed, the certified question issue was an early warning of the difficulties that could flow from premature claim interpretation, for it was often apparent from the petition that the claims could not be finally and correctly interpreted without evidence beyond the patent documents. The absence of extrinsic evidence, of resolution of conflicting positions, and of detailed analysis and findings by the trial judge, inhibited claim interpretation by certified question. Thus, instead of conducting the expected dispositive de novo review, we simply declined the question. The possibility of early finality to claim interpretation has not materialized, with two untoward consequences; first, the district court has had to conduct a perhaps unnecessary trial; and second, the eventual issuance of a new claim interpretation by the Federal Circuit, on appeal after final judgment, has sometimes required a second trial of the issue of infringement. None of these consequences comports with our insistence that claim interpretation is purely a matter of law, that it needs no findings at trial, and that it will be decided de novo by the Federal Circuit.

The second area of disappointed expectations has flowed from the unexpectedly creative de novo claim interpretations that the Federal Circuit has issued in a few cases. This unpredictability in administration of the law of patent claiming has added a sporting element to our bench. It has not only released appellants’ imaginations on appeal, but it will surely add complexity to future trials, as lawyers attempt to guard against the judicial imagination.

A third concern, although rare in occurrence, is of great significance, for even one case wherein the Federal Circuit has deemed itself unconstrained by its own prior interpretation of the same patent removes finality and encourages relitigation of every patent. The promise of uniformity and finality, flowing from decisions of national effect, is a failed promise if we are not bound by stare decisis in our own claim interpretation.

These flaws are of serious concern, no less because they are of our own making. They are not irremediable, although remedy may require a larger vision..than we possess. However, today’s en banc opinion adds another encumbrance to the procedures for interpretation. of claims, further inhibiting fruition of the Markman expectation. When the issues in litigation involve complex questions of science and technology, a special effort is required of the judicial process. 1 In litigation of patent disputes the nature of the evidence that is received and considered, the balance between the trial and appellate functions, and the evidentiary rules governing opinion testimony and experts, are of particular importance. Today’s decision falls short in each of these areas.

Procedural and evidentiary rules should weigh toward facilitating judicial understanding of the issues and thereby reaching the correct result. Yet the Federal Circuit rules today that it will not consider factual findings of the trial court, expressly disavowing such actions by prior panels. The court continues to deny the need to make findings of disputed facts when interpreting claims: “[By] using certain extrinsic evidence that the court finds helpful and rejecting other evidence as unhelpful, and resolving disputes en route to pronouncing the meaning of claim language ... the court is not crediting certain evidence over other evidence or making factual evidentiary findings.” Maj. Op. at 1454 (quoting Markman, 52 F.3d at 981 , 34 USPQ2d at 1331). The court states that it neither accepts the trial judge’s findings of fact, nor accepts that there are factual issues *1480 in claim interpretation. With these strictures on evidence, witnesses, and findings, it is far from clear how the Federal Circuit proposes to reach the correct claim interpretation.

■ By continuing the fiction that there are no facts to be found in claim interpretation, we confound rather than ease the litigation process. Without doubt, factual disputes arise and must be .resolved in order to interpret the claims. Such facts are normally resolved at trial — yet we now deny ourselves the opportunity even to consider the findings of the trial court. See maj. op. at 1456 (disavowing the procedure in a Federal Circuit opinion whereby the court recognized the trial court’s ‘“trained ability to evaluate 1 [expert] testimony in relation to the overall structure of the patent’ and the trial court’s ‘better position to ascertain whether an expert’s proposed definition fully comports with the specification and.claims’ ”). The court today not only rejects the opportunity to give normal appellate deference to the proceedings and findings of trial, but also rejects the opportunity to consider them at all.

In Markman the en banc court took the position that in patent cases, unlike any other area of law, a disputed question of the meaning, scope, and usage of terms of technologic art is not a question of fact, or even of law based on underlying fact, but is pure law. However, the Supreme Court has relieved us of adherence to this fiction, by its recognition of the factual component of claim interpretation. Further, the Court’s affirmation that claim interpretation “is exclusively within the province of the court,” 517 U.S. at 371-73 , 116 S.Ct. at 1387 , 38 USPQ2d at 1463, did not shut out the trial judge along with the jury. In declining to affirm the Federal Circuit’s fact/law theory, the Court opened the door for retreat from this artificial construct. I urge us to do so, for experience shows that unforeseen and undesired consequences are flowing from its rigidity. Now that this fact/law rigor is no longer necessary for constitutional combat, let us review these consequences and accept this invitation to advance our procedures.

For example, in Markman the Federal Circuit stated that it is not a finding when one side’s evidence is accepted and the other side’s evidence is rejected.- - 52 F.3d at 981 , 34 USPQ2d at 1331. Since accepting one side’s evidence and rejecting the other’s based on an assessment of validity would be a finding, some other criterion must be the basis for the choice. The court has never stated what this criterion could be, except that accepted evidence should be “helpful,” an answer that begs the question.

A fresh view of fact and law would also ameliorate the constraints we have placed on the presentation of extrinsic evidence on the issue of claim interpretation.' Patent litigation now often starts with a preliminary hearing to interpret the disputed claim terms, and often produces an early summary judgment, a path perhaps fostered by our foreclosure of the certified question. This preliminary ruling can be dispositive of the dispute, for the scope of the claim often decides whether there can be literal infringement. Thus I add to my concerns the position of the Federal Circuit, here reaffirmed, that extrinsic evidence is of strictly limited availability in claim interpretation. Such evidence should be encouraged, not restrained, if summary disposition is at hand.

The value of extrinsic evidence in claim interpretation is not surprising, because patent documents are written by and for persons in the field of the invention, not for judges. Judges not only need a larger understanding of the science or technology, but we also need help with understanding how the particular terms as used in the patent are viewed by persons in the field of the invention. As Judge Sehwarzer observed:

The context in which [issues of science and technology] arise varies widely, but generally they share one characteristic. They challenge the ability of judges and juries to comprehend the issues — -and the evidence — and to deal with them in informed and effective, ways. As a result, they tend to complicate the litigation, increase expense and delay, and jeopardize the quality of judicial and jury decision making.

Federal Judicial Center, Reference Manual on Scientific Evidence 1 (1994). The Federal Circuit’s ruling that extrinsic evidence must be restricted unless there is a facial ambiguity in the meaning of the claim is an unnecessary restraint on potentially useful evidence. See Daubert v. Merrell Dow Phar *1481 maceuticals, Inc., 509 U.S. 579, 588 , 113 S.Ct. 2786, 2794 , 125 L.Ed.2d 469 (1993) (“ ‘If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert ... may testify thereto in the form of an opinion or otherwise.’ ”) (quoting Fed.R.Evid. 702).

Of course the primary source of information concerning the claimed invention is the patent documents. But such documents are directed to persons knowledgeable in the field; additional evidence and expert testimony as to their meaning should be the rule, not the exception. So-called “extrinsic” evidence — the evidence of expert witnesses and of experimentation, exhibits, demonstrations, and explanation — should be treated like any other evidence, and received and given weight and value as appropriate. Our broad constraint on resort to such evidence is an unnecessary bar to enlightenment. It is also an incursion into the authority of the trial court. See General Electric Co. v. Joiner, — U.S.-,-, 118 S.Ct. 512, 515 , 139 L.Ed.2d 508 (1997) (decision of trial court to admit or exclude expert testimony is reviewed on abuse of discretion standard).

The real issue with respect to such evidence is not the threshold question of admissibility, as this court appears to hold, but of weight, upon examination of the evidence and in conjunction with the other evidence. These too are matters for the trial judge, and when the threshold criteria of relevance and reliability are met, see Daubert, the evidence should be received and considered as appropriate to its credibility and weight. It follows that the trial court’s factual findings with respect to evidence relevant to claim interpretation should be treated, on appeal, like any other finding of the trial court.

Thus it is quite discouraging to observe this en banc ruling wherein the Federal Circuit prohibits itself from considering the findings of fact made by the trial court. The majority opinion disavows this court’s prior statement that “[t]he district court’s findings of scientific/technological fact were material to the issue of construction of the term.” This disavowal deprives the court, and the parties, of the accumulated progress and experience of the trial, including the findings of the trial judge, and leaves us on appeal with an expurgated record and generally inferior basis of decision. Recognizing that our appellate role is to decide whether the claims were correctly interpreted in light of all of the evidence, it is mysterious why we choose to self-censor what we will consider on appeal. It is equally obscure why this court would prohibit itself from relying on a trial court’s findings, or from choosing between disputed expert positions. I strongly disagree with the majority’s view of the role of extrinsic evidence, at trial and as considered on appeal.

In contrast, a return to the traditional trial/appellate relationship would achieve several important results. It would rationalize the admissibility of relevant extrinsic evidence, appellate deference would be restored instead of disavowed, and appellate review would be in accordance with the rules. The processes of both trial and appeal would benefit. This is particularly important because the evidence involved in claim interpretation, whether intrinsic or extrinsic, is often scientific or technologic. The evidence of what the invention is, how it works, what the technical words meant to persons in the field at some past time, can be of extreme complexity. When there is a dispute as to what a term of technical art or usage means or encompasses, such evidence is relevant and often is indispensable. Why would our court foreclose, or place obstacles in the path of, adducing and considering such evidence?

Surely the better view is to encourage judicial access to scientific evidence and findings based thereon. The ultimate beneficiary would be the parties, for the courts would be less restricted in the search for the correct and just result in patent cases. Thus I must, respectfully, dissent from the court’s rulings on these issues.

. In fact, the district court submitted this case to a jury before this court decided Markman v. Westview Instruments, Inc., 52 F.3d 967, 979 , 34 USPQ2d 1321, 1329 (Fed.Cir. 1995) (Markman I). Thus, the district court’s jury instructions only partially construed the claims, and the en banc opinion relies, as it must, upon what it presumes must have been the jury’s claim construction. Nor did the district court deliberately accept and rely upon expert testimony to understand and interpret the claim terms, a central issue in Markman I. Heedless of these problems, this court has selected this case for en banc review. Fromson v. Anitec Printing Plates, Inc., 132 F.3d 1437 , 45 USPQ2d 1269 (Fed.Cir.1997), may have been a more suitable choice for meaningful en banc consideration of the issue of appellate deference in claim construction. The district court decided Fromson by applying the Markman I regime. The Fromson district court also relied upon extrinsic evidence and findings of scientific/technologic fact to interpret the meaning of claim terms to one of ordinary skill in the art at the time of the invention (over twenty years ago). Thus, the facts of Fromson might have better illustrated how different standards of review can direct the outcome on appeal.

. The following is an incomplete list of procedural deviations required by Markman I:

1. Multiple trials, problem I: If hearings are necessary to interpret complex claims, the trial court must set aside time in its crowded docket for one proceeding to interpret claims and a second (potentially with a jury) to determine infringement and other issues.

2. Claim interpretation, problem I: Fearing that it may not receive the opportunity to supplement expert reports or reopen discovery after the judge’s interpretation, a party often argues alternative, claim construction theories from the outset of litigation. This extends the time and expense of the claim interpretation proceedings.

3. Bias toward summary judgments: In pracr tical terms, Markman I directs the proceedings toward summary judgment on 1 the central issue of the litigation at a potentially premature stage of issue development. Prematurely addressing issues, even at the appellate level, can result in expensive repetition of effort. See CVI/Beta Ventures, Inc. v. Tura LP, 112 F.3d 1146, 1157-58 , 1160 n. 7, 42 USPQ2d 1577, 1585, 1587 n. 7 (Fed.Cir.1997) (finding error in a claim construction that had been affirmed in an earlier appeal), cert. denied, -U.S. -, 118 S.Ct. 1039 , - L.Ed.2d-(1998).

4. Claim interpretation, problem II: As soon as the trial court issúes a claim interpretation, both sides often seek to shift their original claim interpretations to accommodate the judge's views. Thus, the parties seek to revise expert reports or reopen discovery to account for the judge’s interpretation. This maneuvering leads to procedural battles over surprise- and motions for additional time to prepare for trial. See Loral Fairchild Corp. v. Victor Co., 906 F.Supp. 798 (E.D.N.Y. 1995) (interpreting claims); Loral Fairchild Corp. v. Victor Co., 911 F.Supp. 76, 80-81 (E.D.N.Y. 1996) (preventing plaintiff from changing theory of infringement in response to claim interpretation).

5. The new evidence dilemma: As a result of the new and perhaps somewhat unexpected interpretation, the parties scramble to create and acquire new evidence for their infringement arguments.

6. The learning curve problem: Like all human éndeavors, claim interpretation is a learning process. The trial judge makes every effort to state the precise scope of the claims at the close of the initial proceeding, but often, with the additional learning during the infringement trial, realizes that the initial interpretation was'too broad or too narrow in some respects. The judge then faces the dilemma of changing the rules in the middle of the game.

7. The judge as a trial issue: With the judge's claim interpretation central to the issues of infringement,’trial counsel'will try to exploit the judge’s stature with the jury to show that the court is on their side.

8. Multiple trials, problem II: In the words of United States District Court Judge Roderick McKelvie: "[I]n spite of a trial judge’s ruling on' the meaning of disputed words in a claim, should a three-judge panel of -the Federal Circuit disagree, the entire case could be remanded for retrial on [a] different [claim interpretation].” Elf Atochem North Am., Inc. v. Libbey-Owens-Ford Co., 894 F.Supp. 844, 857 , 37 USPQ2d 1065, 1075 (D.Del.1995).

Trial judges can often address each of the above with careful case management, but at the cost of expending scarce trial court resources.

. Three variables affect the settlement calculus of each party to litigation: p, the probability of the plaintiff obtaining damages; J, the expected value of a judgment for the plaintiff; and c, the cost of litigation. See Richard A. Posner, The Federal Courts: Challenge and Reform 89-94 (1996). If p x J (pj) exceeds c, then plaintiff will sue. The plaintiff values the case at pj - c. If the defendant agrees on the values assigned to the variables, the suit will cost him pj + c. This rough model poses an interesting question. Because the costs of litigation invariably exceed the costs of settlement, why do not all cases settle? Chief Judge Posner answers: "[Ujncertainty as to outcome is the key to the settlement rate----” Id. at 90. This uncertainty leads each party to overestimate its chance of prevailing. Accordingly, each party will assign different values to the variables, most notably p, thereby diminishing the likelihood of settlement.

. This figure is based on a survey of every patent decision rendered by the Court of Appeals for the Federal Circuit between 5 April 1995 (the date Markman I was decided) and 24 November 1997. A total of -246 patent cases, originating in the Board of Patent Appeals and Interferences (BPAI), the district courts, and the Court of Federal Claims, were evaluated. Of the 246 cases, 141 cases expressly reviewed claim construction issues. Among these 141 decisions, this court reversed, in whole or in part, 54 or 38.3% of all claim constructions. With respect to the district court and Court of Federal Claims cases, the rate of reversal of claim constructions is 47 out of 126 or 37.3%.

. These necessary strictures contribute to the Federal Circuit’s perception of claim" construction. Pressured by these necessary rules, parties before the Federal Circuit strive mightily to reduce their cases to a few issues controlled by a few passages from the specification or the prosecution history. For these reasons, the appeal process does not present a fair picture of the complex task of untangling the knot of legal and factual issues presented for trial. A seemingly simple issue of claim construction on appeal takes on an entirely different complexion in its proper context as one tiny facet of a massive corpus of litigation.

. In 1993 the Carnegie Commission on Science, Technology, and Government wrote:

The courts’ ability to handle complex science-rich cases has recently been called into question, with widespread allegations that the judicial system is increasingly unable to manage and adjudicate science and technology issues.

Report, March 1993, p. 11.