OPINION AND ORDER ON MOTION TO PIERCE ATTORNEY-CLIENT PRIVILEGE AND ATTORNEY WORK PRODUCT PRIVILEGE BASED ON CRIME/FRAUD EXCEPTION

I. INTRODUCTION

Plaintiffs Laser Industries and Sharplan Lasers (“Laser”) sued defendant Reliant *419 Technologies (“Reliant”), alleging infringement of U.S. Patent No. 5,411,502 (“the ’502 patent”). Reliant sought certain discovery relating to the prosecution of the ’502 patent and the initiation of Laser’s lawsuit. Laser resisted some of this discovery, asserting that specific documents sought are protected by the attorney-client privilege and/or the work-product doctrine.

Through its moving and reply papers, Reliant contends that there are two independent grounds on which the court should order Laser to disclose these documents. First, Reliant argues that it has made a prima facie showing that the communications were made in order to further the commission of a fraud on the Patent Office (PTO), thus making the documents discoverable under the crime/fraud exception to the privilege. 1 Second, Reliant argues (in its reply brief) that Laser has waived whatever protections these documents might once have enjoyed by having its patent prosecution counsel submit declarations (in these proceedings) which purport to describe counsel’s ignorance of the prior art that Reliant says Laser was required to disclose to the PTO during the prosecution of the patent in issue. We address these contentions in separate sections, below.

II. RELIANT’S EFFORT TO INVOKE THE CRIME/FRAUD EXCEPTION

Reliant contends that Laser perpetrated a fraud on the U.S. Patent Office by failing to disclose to the Patent Office, during the prosecution of the ’502 patent, one of its own previous patents, U.S. Patent No. 4,587,396 (“the ’396 patent”), as well as existing technology and published materials relating to that patent.

A. Law

The parties’ papers raise several legal issues, some of which lend themselves to relatively straightforward resolution and some of which do not. It will be useful, at the outset, to identify the major legal issues that we will address in the paragraphs that follow.

(1) Whose knowledge, conduct, and intentions are in issue when courts determine whether the crime/fraud exception applies? Should courts limit their inquiry to the lawyers who prosecuted the patent, or should courts look beyond the lawyers and also consider the persons (or entities) on whose behalf the prosecution was undertaken?

(2) May a party penetrate the attorney-client privilege under the crime/fraud exception by making a sufficient showing of “inequitable conduct” (as that phrase is defined in substantive patent law) in the prosecution of the patent, or must a party’s showing satisfy the more demanding standards of common-law fraud?

(3) When determining whether a party has made a showing sufficient to establish the crime/fraud exception, must courts consider only the evidence, inferences, and arguments advanced by the party attempting to penetrate the privilege (i.e., that support the suggestion that the communications were made in furtherance of a crime or fraud), or should courts also consider evidence, inferences, and arguments advanced by the holder of the privilege (i.e., that tend to rebut the suggestion that the communications were made in furtherance of a crime or fraud)?

(4) Should courts view the evidence they consider in this setting in the light most favorable to the party trying to establish the crime/fraud exception? Or should the courts view the evidence with complete neutrality, i.e., neither in the light most favorable to the challenger nor the light most favorable to the holder of the privilege? Or, given the location and magnitude of the burden of proof at trial of a claim of fraud, 2 and the location of the burden of showing that otherwise privileged communications should be disclosed, should the courts resolve evidentiary and inferential doubts against the party trying to establish the crime/fraud exception?

*420 (5) What level of certainty must the court feel before it may conclude that a challenger has established that the crime/fraud exception applies? What level of certainty satisfies the requirement of making a “prima facie” showing in this setting? Does the phrase “prima facie” mean the same thing in' this setting as it does in others? Should the level of certainty vary with the specific nature of the basis for the assertion that a “fraud” (or crime) was committed?

As we shall see, the last three of these issues are so closely intertwined in the pertinent authorities that it is impossible to address them as if they were fully separable. Some courts, for example, without much independent analysis, have accepted a definition of the phrase “prima facie case” that has dictated the answers to the third and fourth questions posed above. For reasons I hope to make clear, I am not persuaded that such semantically tidy but ultimately circular approaches do justice to the policies and interests that are in tension in these settings. We are likely to think more reliably about these matters if we consider the third and fourth issues at least in some measure independently of the fifth, i.e., before we accept a definition of “prima facie case.” There is some tension between approaching this task in the way I propose and at least some parts of some of the precedents, but I hope that exploring that tension will lead us to answers whose underpinnings we more clearly understand.

Before addressing the five issues listed above, it is important to emphasize that Reliant has not asked the court to conduct an in camera examination of the communications in issue here. Because Reliant has not asked us to conduct an in camera inspection of the communications in issue, we are not going through the two-stage analytical process and applying the two different standards that are set forth in United States v. Zolin, 491 U.S. 554, 572 , 109 S.Ct. 2619, 2630-31 , 105 L.Ed.2d 469 (1989), and discussed in In re Grand Jury Investigation, 974 F.2d 1068, 1072-73 (9th Cir.1992). 3 Thus we are not applying to the showing that Reliant has made in the case at bar the less stringent standard that the Zolin court makes applicable when a judge is trying to decide, at the threshold, whether she must proceed further to consider whether to exercise her discretion to conduct an in camera inspection. 4 Nor are we applying the criteria that Zolin instructs judges to consider, after a party has made a sufficient threshold showing, when the court is deciding whether or not, even given satisfaction of the threshold showing requirement, the court should exercise its discretion in the direction of conducting the in camera review. 5

For the parties’ guidance, however, we make two observations: (1) we are not sure whether we would hold that Reliant’s showing satisfies even the relatively less stringent requirements for a threshold showing under Zolin , but (2) we are sure, on the record before us, that Reliant’s showing is not sufficient to persuade us that it would be appropriate to conduct an in camera inspection of the privileged documents that Reliant is trying to discover through this motion. The discussion in § 11(B), below, will expose some *421 of the principal reasons for the latter conclusion. 6

We address now the issues identified in the preceding paragraphs.

1. In Determining Whether the Crime! Fraud Exception Has Been Established, Should Courts Consider Only the Knowledge, Conduct, and Intent of the Lawyers or Agents Who Prosecuted the Patent, or Should Courts Also Consider the Knowledge, Conduct, and Intent of the Persons on Whose Behalf the Prosecution Was Undertaken?

Laser contends, citing no authority, that in determining whether a showing sufficient to penetrate the privilege has been made, the court should examine only the knowledge, conduct, and intent of the lawyers who prosecuted the application for the ’502 patent. We decline to accept this view.

As Professor Paul Rice emphasizes in his treatise on attorney-client privilege, “[t]he rationale behind the erime/fraud/tort exception to the attorney-client privilege is to deprive those clients of the protection of the privilege when they seek to abuse the attorney-client relationship. Consequently, its focus has been on the intent of the client.” Paul R. Rice, Attorney-Client Privilege in the United States § 8:4, at 572 (1993). Elaborating on this point, Professor Rice notes:

Id. at 576.

In the case at bar, Laser is the client and the holder of the privilege. Since Laser is the holder of the privilege, Laser has the capacity to waive it. Since, as a matter of law, Laser must have the capacity to waive, it would make no sense for the court to ignore Laser’s knowledge, conduct, and intent when the court tries to determine whether the privilege was lost (or never attached in the first place) by virtue of the crime/fraud exception.

Nor are there any obvious fairness concerns that would support a contrary conclusion. Laser appears to have been a sophisticated user of the patent system, and Laser clearly understood and controlled 7 the information that Reliant contends should have been disclosed to the PTO. It would frustrate the policy that informs the crime/fraud exception to the attorney-client privilege if the court were to permit a client thus situated to retain the protections of the privilege while self-consciously orchestrating a fraud on the PTO by intentionally concealing obviously material prior art from the lawyers whom the client retained to prosecute the patent application. The law cannot permit a client to insulate itself from all efforts to penetrate its privilege simply by keeping its counsel in the dark. As the Court of Appeals for the Ninth Circuit has stated, “[t]he crime or fraud exception applies even where the attorney is completely unaware that his advice is sought in furtherance of such an improper purpose.” United States v. Hodge & Zweig, 548 F.2d 1347, 1354 (9th Cir.1977); see also Clark v. United States, 289 U.S. 1, 15 , 53 S.Ct. 465, 469-70 , 77 L.Ed. 993 (1933) (in dictum, the high court declared that “[t]he attorney may be innocent, and still the guilty must let the truth come out”).

We hasten to add that we are not persuaded that Laser adopted a strategy under which it would hide information from the lawyers it retained to prosecute the patent and thus use those lawyers, unwittingly, to pursue any unlawful or improper objective.. The issue here, however, is not whether Reliant has made a showing sufficient to establish the crime/fraud exception; rather, the question at this juncture is whether the court should consider information known during the patent’s prosecution by Laser but arguably not known by its patent counsel. The answer is yes.

2. Under the CrimelFraud Exception, Can a Party Penetrate the Privilege on a Sufficient Showing of “Inequitable Conduct,” or Must the Showing Satisfy the Elements of Common-Law Fraud?

Statutory and case law impose a duty on applicants for patents to proceed before the PTO with a high degree of candor and good faith. See, e.g., Research Corp. v. Gourmet’s Delight Mushroom Co., 560 F.Supp. 811, 819 (E.D.Pa.1983). A party who breaches that duty by engaging in “inequitable conduct” can forfeit the right to enforce the patent. And “inequitable conduct” can be established by proving that (1) the party prosecuting the patent application failed to disclose “material” information (information is considered “material” if there is a substantial likelihood that the examiner would have considered it important in deciding whether or not to grant the patent) and that (2) the omission of the information was either intentional or the product of gross negligence. See, e.g., Union Carbide Corp. v. Dow Chemical Co., 619 F.Supp. 1036 , 1051-52 & n. 14 (D.Del.1985).

It can be appreciably easier to prove the kind of “inequitable conduct” that is sufficient to prevent enforcement of a patent than to prove common-law fraud. See id. In the context of a challenged patent prosecution, to prove common-law fraud it is necessary to establish three elements: (1) that a knowing, willful, and intentional misrepresen *423 tation or omission took place before the Patent Office; (2) that the misrepresentation or omission was material; and (3) that the Patent Office relied upon the misrepresentation or omission in issuing the patent. Union Carbide, 619 F.Supp. at 1052 ; American Optical Corp. v. United States, 179 U.S.P.Q. 682 , 684, 1973 WL 20128 (Ct.Cl.1973). The reliance requirement is by far the most significant difference 8 between what is necessary to prove common-law fraud and what is necessary to establish “inequitable conduct.” In order to establish the reliance element, in this kind of setting, the challenger must demonstrate that having the information that the applicant omitted, or knowing the truth about what the applicant misrepresented, would have been sufficient, without more, to cause the Patent Office to reject the application for the patent. See Union Carbide, 619 F.Supp. at 1052 ; American Optical, 179 U.S.P.Q. at 684 (to demonstrate reliance, the challenger must show that “but for the omission or misrepresentation the claims would not have been allowed”).

For the reasons set forth below, we conclude that, under the most persuasively reasoned authorities, the alleged “wrong” on which courts should focus when deciding whether to pierce the attorney-client privilege in settings like these is not simply “inequitable conduct,” but, instead, the more demanding norm of common-law fraud. See, e.g., Union Carbide, 619 F.Supp. at 1052 ; American Optical, 179 U.S.P.Q. at 684.

Research Corp., 560 F.Supp. at 819-20 , contains the most detailed discussion we found of why it is appropriate to require a showing that addresses the elements of common-law fraud rather than the less demanding elements of “inequitable conduct.” As explained in more detail in Research Corp., using “inequitable conduct” as the norm would permit application of the crime/fraud exception to conduct and states of mind less reprehensible than the “serious unlawful activity” that serves as the justification for loss of this important privilege. See 560 F.Supp. at 820 . 9 And using that lesser standard for matters related to patent prosecutions might imply that communications between lawyers and clients in the patent context merit less respect and privacy than communications between lawyers and clients in other areas. See id. at 819 .

While we agree with the reasoning in the Research Corp. opinion, we note some additional considerations that support the conclusion that the appropriate norm is common-law fraud rather than inequitable conduct. Patent cases can be quite complicated, and determining whether “inequitable conduct” occurred as a result of omissions or misrepresentations before the Patent Office will often require detailed and difficult technical analysis. Moreover, because of the relative elasticity of the concept of “materiality” as it is used in this context, and because of the fact that often what separates sophisticated products and processes are fine and technologically esoteric (but not necessarily insignificant) differences, many times there is an element of unpredictability and (at least at the last analytical stages) arbitrariness in decisions about whether or not some piece of information or prior art should be categorized as “material.” In a phrase, these determinations often seem to be “close judgment calls.” And courts should be slow to deprive parties of important rights (here the *424 right to communicate in confidence with counsel) on the basis of close judgment calls.

In part because of the difficulty inherent in these kinds of judgment calls, courts are likely to make them with appreciably greater reliability (accuracy, and, therefore, “justice”) in the context of a full-blown trial or Mark-man hearing, 10 with the benefit of expert testimony and opportunities for opposing attorneys and the court to question witnesses, than in the context of a pre-trial discovery motion. Simply put, there is a greater risk (if we are to be realistic about these matters) that courts will err in making judgments like these when they are addressing a discovery motion than when they are resolving issues at trial or in a motion for summary judgment.

Moreover, if they were to use the appreciably less demanding test for “inequitable conduct,” courts would intensify lawyers’ incentives to attempt to pierce the privilege by “pre-litigating,” in the discovery stage of the case, the “inequitable conduct” issue. The incentives lawyers already have for undertaking such efforts seem quite sufficient. Lawyers love to know the secret communications between opposing parties and their counsel. And some lawyers love to strike fear into the hearts of opponents by attempting to discover those communications. Motions to penetrate clearly applicable privileges can be tools for harassment — and can chill the communications that the privileges are designed to encourage. The looser the standard for the crime/fraud exception, the greater the risk that it will be abused for such tactical purposes. And, unfortunately, intellectual property cases often seem to be litigated with -unusual levels of adversarial intensity and inventiveness.

In sum, if we were to adopt standards that were too light, we would invite substantial harm both to the administration of justice and to justice itself. We would increase the risk of unjustified consumption of party and judicial resources, and of duplicative litigation of the same issues. Whether there had been “inequitable conduct” would more often be litigated twice, first in the discovery stage (under the lesser demands of a “prima facie” standard — discussed below) and then, later, at trial and/or through a motion for summary judgment. More importantly, we would increase the risk to justice itself by increasing the risk of erroneous pretrial decisions, made without application of full judicial resources and on the basis of a less than fully developed record.

These substantial costs clearly outweigh the benefits that might flow from using the less demanding standard. Use of that standard would result in more frequent disclosure of attorney-client communications, but it is not at all clear that it would improve significantly the likelihood that judges or juries would make correct determinations about whether inequitable conduct occurred. That follows because the parties very often seem to know (without access to any privileged communications) what the prior art or other undisclosed information was. What they dispute is whether that information was “material” and whether it was at least grossly negligent not to have disclosed it. Access to privileged communications will shed little reliable light on whether the prior art was “material” — that issue is resolved on the basis of independent, objective standards.

There is a greater likelihood that access to privileged communications could shed light on the state of mind of the party prosecuting the patent application (at least when that party was not very clever), but since a showing of intent is not required (to prove “inequitable conduct”), and since gross negligence (or intent) can be inferred from circumstantial evidence, including evidence about how obviously material the prior art was, 11 it is not clear that access to privileged communications will very often be necessary in order to resolve the inequitable conduct dispute fairly. So the “benefit to justice” interests of penetrating the privilege on a lesser showing usually would be modest, at best, while the risk of serious harm that would be created by *425 using the less demanding standard would be considerable. For all these reasons, we agree that courts should use the more demanding common-law fraud standard when they determine whether conduct in prosecuting a patent application justifies piercing the attorney-client privilege under the crime/ fraud exception.

3. Whose Evidence May the Courts Consider In Determining Whether a “Prima Facie Showing” Has Been Made?

In support of its motion, Reliant relies in part on a passage from Duplan Corp. v. Deering Milliken, Inc., 397 F.Supp. 1146, 1196 (D.fe.C.1974), where the district court, in determining whether the crime/fraud exception had been established, refused to consider arguments (actually factual allegations) that the holder of the privilege contended would tend to disprove that it had engaged in any fraudulent or otherwise tortious conduct. The district court explained this refusal as follows:

Id.

While the Duplan trial court cited no authority for this proposition, there is support for it in some other eases. See Paul R. Rice, Attorney-Client Privilege in the United States § 8:6, at 588-89 & nn. 65-70, Supp. 119 (1993 & Supp.1996), and cases cited therein. 12 Some of that support seems to be based on assumptions about what the phrase “prima facie showing” means, 13 and some seems to be based on announcements of law in criminal cases (or in connection with ongoing grand jury investigations) that cannot be reliably imported, wholesale and without analysis that is sensitive to context, into all civil litigation settings. As we discuss below, the phrase “prima facie showing” or “prima facie case” can mean different things in different environments — and courts must take care not to answer the difficult question we address in this section on the basis of circular reasoning that is driven by a definition of “prima facie” that was developed in other settings to serve other purposes. See, e.g., In re Feldberg, 862 F.2d 622, 625-26 (7th Cir.1988) (where the court discussed briefly the separate meaning of the phrase “prima facie case” in the context of federal civil rights claims based on alleged race or age discrimination). In other words, we would reason tmreliably if we were to conclude that courts in the setting we confront here should be permitted to consider only the evidence favorable to the party trying to penetrate the privilege simply because the phrase “prima facie case” means, in some other setting, introduction of sufficient evidence to rationally support the inference suggested by the proponent of the evidence, without regard to the persuasive power of evidence that might subsequently be presented by the opposing party.

We also must bear in mind that there can and should be a relationship between how we answer this question and what the consequence of the answer is. If the only consequence of making a prima facie case is creating an obligation in the holder of the privilege to come forward with some evidence tending to rebut the suggestion of crime or fraud, determining whether a prima facie showing has been made without first considering evidence and argument offered by the holder of the privilege is not likely to give rise to serious concerns. Cf. Feldberg, 862 F.2d at 625-26 . But if the consequence of making a prima facie case is *426 not limited to shifting a procedural or even substantive burden, or is not limited to triggering access to an in camera inspection of documents by a judge (which constitutes a relatively less severe invasion of the privilege), but if, instead, the making of a prima facie showing results directly in disclosure of the privileged communications to an opponent in litigation, then the potential harm caused by refusing to consider evidence and argument proffered by the holder of the privilege could be severe.

Most of the judicial opinions that can be cited in favor of the view that a trial court need consider only the evidence offered by the challenger when determining whether there has been a showing sufficient to establish the crime/fraud exception have been generated in criminal cases and, more specifically, in cases where a grand jury has subpoenaed testimony or documents from the lawyer of a person who is the target of an ongoing investigation. In these settings, many courts have concluded that the trial court need consider only the evidence of crime or fraud that is presented by the government, and that the court may consider some or all of that evidence in camera and ex parte, meaning that the holder of the privilege and his lawyers don’t even have an opportunity to learn what that evidence is, let alone to challenge it through argument or through competing evidentiary submissions. See, e.g., In re Grand Jury Proceedings, 867 F.2d 539, 540-41 (9th Cir.1989); In re Vargas, 723 F.2d 1461, 1467 (10th Cir.1983); see also Rice, supra, § 8:5, at Supp. 116-19 (Supp.1996), and cases (some of which are unpublished) cited therein.

The interests that are in tension in these settings, however, are fundamentally different from the interests that are in tension in the case at bar. Most obviously, there is a substantial governmental interest in maintaining the secrecy and sustaining the effectiveness of grand jury investigations. These important societal interests in the role of the grand jury, as well as in the effective detection, punishment, and deterrence of criminal acts, have no analog in the case at bar. And it is largely on the basis of the weight of these interests that courts have concluded that it offends no constitutional (due process) or other norms' for trial judges to determine whether the crime/fraud exception has been established by considering only the challenger’s evidence.

It is significant that the only published 14 opinion from the Court of Appeals for this circuit that at least arguably has implications for this specific issue involved a criminal prosecution. See United States v. Laurins, 857 F.2d 529, 541 (9th Cir.1988), cert. denied, 492 U.S. 906 , 109 S.Ct. 3215 , 106 L.Ed.2d 565 (1989). In Laurins , a defendant was attacking his criminal conviction on appeal. One of the bases for his attack was a contention that the trial court had erred in concluding that the crime/fraud exception applied to certain communications (which were admitted at trial). In rejecting this contention and upholding the district judge’s finding that the government had made “a prima facie case for obstruction of justice,” the Court of Appeals offered the following one-sentence pronouncement: “All the government needed to show was evidence that if believed by the jury would establish the element of an ongoing violation.” Id. (citing In re Sealed Case, 754 F.2d 395, 399 (D.C.Cir.1985)).

It is significant that this pronouncement, which the Court never explained or elaborated, was made in the context of a criminal case. It is even more significant that the Court borrowed this passage from an opinion in which the Court of Appeals for the District of Columbia was addressing the requirements for the crime/fraud exception in the context of an ongoing grand jury proceeding. See In re Sealed Case, 754 F.2d 395, 399 (D.C.Cir.1985) (“The government satisfies its burden of proof if it offers evidence that if believed by the trier of fact would establish the elements of an ongoing or imminent crime or fraud”). These facts about the source of the key passage, and the context in which it was adopted, make me reluctant to *427 conclude that it was intended to apply to the situation at bar.

Moreover, even if the contexts were sufficiently similar to support an inference that the Court of Appeals intended this passage to apply in the circumstances presented in the case at bar, it would not be clear, from this passage or the case on which it relied, that the appellate courts intended to foreclose consideration of evidence and argument offered by the holder of the privilege. The ambiguity in the key passage resides in a potentially critical qualification that the Court imposed on the evidentiary showing it required from the government. The Laurina formulation of the test implied, at least, that the trial court would be called upon to decide whether the evidence presented by the government could be “believed by the jury.” See Laurins, 857 F.2d at 541 (emphasis added). What is unclear, of course, is what the trial court could consider in deciding whether a rational jury could believe the government’s evidence. There certainly is insufficient clarity in this passage to foreclose the possibility that a district court could at least have discretion to decide, in making this key determination, whether to take into account competent, admissible evidence (and the inferences that evidence could support) that was offered by the holder of the privilege.

My inference that the phraseology adopted by the Laurins court does not foreclose the possibility that trial courts could be permitted to consider evidence offered by the holder of the privilege is supported dramatically by a carefully reasoned opinion from the Court of Appeals for the Third Circuit, Haines v. Liggett Group Inc., 975 F.2d 81, 96-97 (3d Cir.1992). In that ease, Judge Aldisert, writing for a unanimous panel, endorsed a definition of “prima facie showing” that is virtually identical to the definition the Ninth Circuit endorsed in Laurins . As Judge Aldisert phrased it, to make a sufficient showing “the party seeking discovery must present evidence which, if believed by the fact-finder, would be sufficient to support a finding that the elements of the crime-fraud exception were met.” Haines, 975 F.2d at 95-96 . After having endorsed this definition of a prima facie showing, the Haines court proceeded to hold, after clearly deliberate consideration, that in civil cases the court must consider, before deciding whether the challenger’s showing satisfied the prima facie standard, evidence and argument that are offered by the party resisting the challenge (the holder of the privilege). See id. at 96-97 . Thus it is clear that at least the Haines court, which has thought about these matters more fully than most others, finds no tension between (1) the definition of a prima facie case adopted by the Ninth Circuit in Laurins and (2) the trial judge taking into account, before ruling on the crime/fraud exception, evidence and argument offered by the party trying to rebut the prima facie showing of crime or fraud.

Given the fact that most of the pronouncements that clearly support the view.that trial courts need not consider evidence offered by the holder of the privilege are found in criminal cases involving powers and procedures before grand juries, i.e., cases where the interests in tension are very different from the interests in the matter at bar, and given the ambiguity in the only arguably pertinent pronouncement that I could find in a published Ninth Circuit opinion, I conclude that at least in civil cases where the alleged wrong is fraud on the Patent Office the question of whether a trial court may consider evidence and argument offered by the holder of the privilege remains open in this circuit. Because the question remains open, it is important to describe the reasoning on the basis of which I resolve it.

The opinion that has addressed this question most carefully and persuasively, in my view, is Haines, 975 F.2d at 96-97 . Notably, Haines did not involve a criminal prosecution or grand jury proceedings. It was a civil case — and that was a fact of some moment to the Court of Appeals, which expressly limited its “discussion regarding the appropriate procedures to be followed in ascertaining the applicability of the crime-fraud exception to the civil context.” Id. at 97 n. 8. Clearly signalling that it appreciated that different interests would be implicated in non-civil settings, the Court declined even to “intimate” any “view as to whether the same proce *428 dures should be used in the grand jury context.” Id. One of the reasons that I am inclined to find Haines persuasive is that it is one of the few opinions to acknowledge that reasoning about these matters in the civil context is not fungible with reasoning about them in the context of grand jury investigations and criminal prosecutions.

The Haines court devoted several pages of discussion to the crime/fraud exception. Writing three years after the Supreme Court published Zolin, 491 U.S. 554 , 109 S.Ct. 2619 , 105 L.Ed.2d 469 , Judge Aldisert carefully distinguished the kinds of showings and procedures that were appropriate to support an in camera inspection of documents by a judge, on the one hand, from, on the other hand, the kinds of showings and procedures that should be required for the much greater invasion of the privilege that would result from ordering disclosure of confidential materials to a litigation opponent. See Haines, 975 F.2d at 96-97 . Judge Aldisert recognized that “the objectives of the two proceedings are completely different. One merely seeks in camera examination of documents by the court; this is a comparatively nondispositive procedural waystation. The other seeks to break the seal of a highly protected privilege.” Id. at 96 . It is because of this considerable difference in the consequences of the two proceedings, according to the Haines court, that standards appropriate for the first might be wholly inadequate for the second:

Id. at 96-97.

This reasoning acquires additional persuasive power in the setting I confront here, where the alleged wrongdoing on the basis of which the privilege would be lost consists of fraud on the Patent Office with respect to a very technologically sophisticated set of claims. As I discussed in the preceding section, the question of whether particular items of prior art are “material” is hardly self-answering in this kind of setting. The meanings of key terms in the patent, the file history, and the prior art often are not obvious, and differences in phraseology, technology, or process that might appear insignificant *429 to the uninformed eye can be critical to one versed in the pertinent art. Given these facts of technological life, and the relative imprecision in the concept of “materiality,” there is a very substantial risk that generalist judges will not make reliable determinations about the strength of fraud showings if they are permitted to proceed ex parte and are required to limit their consideration to evidence and argument presented by the party challenging the privilege. 16 As the magnitude of that risk of error increases, so does the threat to due process values, to say nothing of the threat to the values underlying the privilege.

We should not underestimate the threat posed to privilege values here. As suggested above, if we were to endorse procedures or standards that created significant risks of error by the trial judge and that increased substantially the prospects that challenges to assertions of privilege would succeed, we would increase the incentives, considerably, for launching such challenges. And if the incidence of challenges were to increase, and if the outcomes of such challenges were to become appreciably less predictable, parties would have much less confidence that they could communicate freely with their counsel. Eroding that confidence would reduce the flow of information from clients to lawyers, and would compromise the reliability of the advice that lawyers would provide in return. Such a spectre would strike at the heart of the goals that underlie the law’s creation of the attorney-client privilege in the first place.

The competing considerations cannot justify taking these high risks. A challenger might argue that requiring trial courts to consider evidence offered by the holder of the privilege will reduce the number of occasions when courts will order documents disclosed — and thus will reduce the number of occasions when courts will uncover efforts to use lawyers for unlawful purposes. This argument indicts fundamental notions of due process and, at least in civil cases, would threaten our entire adversary system. It also ignores the distinct possibility that, without the aid of evidence from the holder of the privilege, the occasions on which courts erroneously deprived parties of the protections of the privilege would exceed, perhaps by a considerable margin, the occasions when ordering disclosures confirmed or more fully exposed unlawful activity. That prospect is especially real in cases like the one at bar, where the basis for the claim of fraud (on the PTO) is so fraught with analytical uncertainty. 17

Nor is it clear that requiring courts to consider evidence and argument offered by the holder of the privilege would cause a net increase in burdens on the judiciary. The prospect of facing a fairer, more open, and more demanding hearing is likely to discourage some parties from launching thinly premised or largely tactical efforts to penetrate the privilege. This is as it should be— and would result in courts having to hear fewer motions. The upshot of the approach adopted here is likely to be more-labor-intensive proceedings — but fewer of them. So we certainly cannot say with confidence that there will be any net increase in demands on judicial resources. And even if there were some increase in the burdens on us, we would be required, by our commitment to fairness, to bear them.

For all the reasons set forth above, I conclude that I am required, given the civil character of this case and the specific basis *430 for the claim that the fraud exception applies, to permit the holder of the privilege to submit evidence and argument that tends to rebut an inference of any of the necessary elements of the crime/fraud exception.

4. Should the Court Consider the Evidence Offered by Either One of the Parties to This Kind of Dispute in a Light More Favorable than the Evidence Offered by the Other Party?

At trial, a defendant who is challenging the enforceability of a patent must bear the burden of proof on the issue of whether the plaintiff was guilty of inequitable conduct in the prosecution of the patent. Moreover, to satisfy that burden, it is not enough to prove the inequitable conduct by a preponderance of the evidence. Rather, the defendant must satisfy the more demanding requirements of proof by clear and convincing evidence. See LaBounty Mfg., Inc. v. United States Int’l Trade Comm’n, 958 F.2d 1066, 1070 (Fed.Cir.1992). Given the fact that it is the challenger on whom the law places these substantial burdens at the liability stage, one might conclude that it would be appropriate to give the evidence offered by the holder of the privilege the benefit of any real inferential doubts when resolving a pretrial discovery motion that is based on alleged fraud on the Patent Office. While such a conclusion clearly has some logical appeal, at least some of the underpinnings for it seem to have been rejected by the Supreme Court.

In Bourjaily v. United States, 483 U.S. 171, 175-76 , 107 S.Ct. 2775, 2778-79 , 97 L.Ed.2d 144 (1987), the high court, speaking through Chief Justice Rehnquist, rejected the suggestion that there should be a connection between the standard a party must meet to satisfy the requirements of Fed.R.Evid. 104(a), on the one hand, and, on the other, the magnitude of the burden of proof that the law imposes at the liability stage. The Court held that because different legal and policy considerations inform the Rules of Evidence than inform the substantive issues at trial, “the evidentiary standard is unrelated to the burden of proof on the substantive issues, be it a criminal case or a civil case.” Id. at 175 , 107 S.Ct. at 2778-79 (citation omitted). Because Fed.R.Evid. 104(a) applies to determinations by courts of “[p]reliminary questions concerning ... the existence of a privilege,” it appears that we cannot use the location and magnitude of the burden on the challenger at trial to justify a rule under which we would view the evidence on a discovery motion like this in the light most favorable to the holder of the privilege.

While the Bourjaily court held that the facts necessary to support a ruling under Fed.R.Evid. 104(a) must be established by a preponderance of the evidence, see id. at 176 , 107 S.Ct. at 2779 , it would be dangerous to infer from the Court’s adoption of that standard that a court should consider evidence on a motion challenging a privilege in anything but an even-handed light. The preponderance standard applies in most civil settings— and in most of those settings the courts initially receive the evidence offered by the opposing parties equally, without an inferencing filter that would bring more skepticism to one party’s evidence than to his opponent’s. This equality of initial treatment of evidence, of course, is to be distinguished from ultimate placement of the burden of proof on the liability issues. If, after considering both parties’ evidence with an equally open mind, the trier of fact concludes that the evidence that supports the contentions of the party who bears the ultimate burden of proof does not have at least a little more persuasive power, the party on whom the law places the burden must lose. The issue we address here is not how much persuasive power the challenger’s evidence ultimately must carry (we turn to that question in the next section); rather, we ask here whether courts, when considering that evidence, should view it with a level of skepticism that they do not apply to the evidence presented by the holder of the privilege. We see no reason for doing so.

Nor, however, do we see any reason for viewing the evidence presented by the challenger “in the most unfavorable light against the [holder of the privilege].” Cf. Duplan Corp. v. Deering Milliken, Inc., 397 F.Supp. 1146, 1196 (D.S.C.1974). Basic notions of fairness, reinforced by the values that under *431 lie the privilege, 18 compel the conclusion that it would be inappropriate to draw inferences more generously or more loosely from the challenger’s evidence than from the evidence offered by the holder of the privilege. I cannot understand how it would comport with due process to apply to evidence offered by one party (the party who already has shown that the privilege presumptively applies) a filter of skepticism that I would not apply to the evidence offered by the other party (the party on whom the law places the burden of making a sufficient crime/fraud showing).

I hasten to add that I appreciate that how one thinks about this issue could be affected substantially by the standard one adopts for defining a “prima facie ease.” It appears, for example, that the Duplan trial court accepted a definition of “prima facie showing” (imported expressly from the grand jury context) that required the trial judge not only to confine her vision to the evidence offered by the party challenging the privilege, but also to consider that limited universe of evidence “in the most unfavorable light against the accused.” See Duplan, 397 F.Supp. at 1196 .

As the discussions in other sections of this opinion make clear, however, I am not persuaded that that definition of “prima facie showing” represents, at least in the setting I confront in this case, an appropriate accommodation of the competing interests and policies. Rather, as explained below, I have concluded that the standard for or definition of “prima facie showing” that is appropriate for this kind of case not only permits, but requires, even-handed initial consideration of the evidence offered by the challenger and by the holder of the privilege. Thus, I hold that I must consider the evidence presented by the holder of the privilege in the same light as I consider the evidence offered by the party who is attempting to penetrate the privilege.

5. What Level of Certainty Satisfies the Requirement of a “Prima Facie Showing” in This Kind of Case?

The notion that a party who attempts to penetrate the attorney-client privilege under the crime/fraud exception must make a showing sufficient to earn the label “prima facie” has venerable but curious roots. The seminal federal court authority is Clark v. United States, 289 U.S. 1 , 53 S.Ct. 465 , 77 L.Ed. 993 (1933), penned by Justice Cardozo. While characteristically scholarly, the portions of the Clark opinion that are relevant to the attorney-client privilege suffer from one obvious limitation: they are all dicta. The issues that the Court actually decided in Clark , which involved a criminal prosecution, related to whether and under what circumstances the government could penetrate the privilege of jurors not to be examined about the nature of their deliberations. See id. at 12-19 , 53 S.Ct. at 468-71 . It was with respect to that different privilege that Justice Cardozo wrote that “we do not mean that a mere charge of wrongdoing will avail without more to put the privilege to flight. There must be a showing of a prima facie case sufficient to satisfy the judge that the light should be let in.” Id. at 14, 53 S.Ct. at 469 (emphasis added).

It was when he sought support for this rule from other contexts that Justice Cardozo described what he suggested was a similar rule for the attorney-client privilege. Relying in large (but by no means exclusive) measure on authorities from England,- the Clark court suggested that, under the law of attorney-client privilege, for the protection to be driven away “there must be ‘something to give colour to the charge’; there must be ‘prima facie evidence that it has some foundation in fact.’ ” Id. at 15 , 53 S.Ct. at 469-70 (citations omitted). Later Justice Cardozo added that a “privilege surviving until the relation is abused and vanishing when abuse is shown to the satisfaction of the judge has been found to be a workable technique for the protection of the confidences of client and attorney.” Id. at 16 , 53 S.Ct. at 470 (emphasis added).

A great deal of law, much of it not painstakingly considered, has been built on these *432 modest passages, which, while supported by substantial citations, were not explained. Only a few things seem clear from the Clark formulations: (1) the standard under which the privilege can be penetrated is appreciably less demanding than “beyond a reasonable doubt” and appreciably more demanding than mere suspicion, (2) the showing made must be sufficient to satisfy the trial judge that it is appropriate to expose the otherwise privileged communications, and (3) the trial judge is vested with some discretion in making that determination. 19

Almost sixty years after Clark was published, the Supreme Court conceded, in a wonderful example of studied understatement, that

United States v. Zolin, 491 U.S. 554 , 563 n. 7, 109 S.Ct. 2619 , 2626 n. 7, 105 L.Ed.2d 469 (citations omitted) (emphasis added). Thus, at the highest judicial level, the question with which we grapple here remains open.

Moreover, as suggested in Zolin , over the years the courts of appeals and the district courts have adopted a fairly wide range of standards for “prima facie” showings in this context. On the less demanding end of the spectrum are approaches like those endorsed by the trial court in Duplan Corp. v. Dee-ring Milliken, Inc., 397 F.Supp. 1146, 1196 (D.S.C.1974), 20 and by the Court of Appeals for the Seventh Circuit in In re Feldberg, 862 F.2d 622, 625-26 (7th Cir.1988). See also supra § 11(A)(3).

The Feldberg court, speaking through Judge Easterbrook, did not define the level of conviction the trial court would need to reach under the standard announced in that opinion — but insisted that it was less than the level of conviction required in some other circuits. According to Judge Easterbrook, some other courts of appeals required presentation of evidence by the challenger sufficient to support a verdict (on the crime/fraud claim) in favor of the challenger. See 862 F.2d at 625 . Judge Easterbrook apparently believed that this standard required the challenger to present “evidence that by itself satisfies a more-likely-than-not standard:” See id. at 626 . The Feldberg court rejected that standard, however, claiming that the test should “not [be] whether the evidence supports a verdict but whether it calls for inquiry.” Id. Unfortunately, the Feldberg opinion did not describe the quantum of proof that would be sufficient to “call[ ] for inquiry.” See id. In fact, this approach seems to leave considerable discretion in the trial judge.

Moreover, it is not clear, from Feldberg , what the nature of the burden is that shifts after the challenger presents enough evidence to “call[ ] for inquiry.” See id. Does only the burden of going forward shift to the holder of the privilege? Or does the chal *433 lenger’s initial showing shift the burden of proof (persuasion)?

There is support in Feldberg for either answer. On the one hand, Judge Easter-brook explicitly drew analogical guidance from procedures used in litigating claims of race or age discrimination: “Here, as in the law of discrimination, a prima facie case must be defined with regard to its function: to require the adverse party, the one with superior access to the evidence and in the best position to explain things, to come forward with that explanation.” Id. at 626 (citing Texas Dep’t of Community Affairs v. Burdine, 450 U.S. 248 , 254 n. 7, 101 S.Ct. 1089 , 1094 n. 7, 67 L.Ed.2d 207 (1981)). The burden that shifts in race and age discrimination eases, of course, is only the burden of going forward, not the burden of proof (the latter, more substantial burden remains where it was at the outset, on the plaintiff, as long as the defendant satisfies the more modest burden of going forward). See, e.g., Burdine, 450 U.S. at 256 , 101 S.Ct. at 1095 .

On the other hand, the next sentence of Judge Easterbrook’s opinion might imply that he would shift the burden of persuasion. He wrote, “If the court finds the explanation satisfactory, the privilege remains.” Feldberg, 862 F.2d at 626 . Under this approach, the holder of the privilege bears the risk of not providing an explanation that is “satisfactory.” The key question, of course, is what is the standard for the phrase “satisfactory,” which is hardly a well-defined term of legal art. If, under the Feldberg approach, doubts and close questions are to be resolved against the holder of the privilege, who clearly is the person who must offer an explanation that the court finds “satisfactory,” then the burden that has shifted is the burden of persuasion.

If construed as shifting only the burden of going forward, the Feldberg approach looks a lot like the test for determining whether to consider conducting in camera inspections that the Supreme Court announced nine months later in Zolin, 491 U.S. at 572 , 109 S.Ct. at 2630-31 . But if construed as shifting the burden of persuasion, I believe the Feldberg approach would impose unjustifiable burdens on courts and would jeopardize, without sufficient grounds, the values that underlie the privilege. Challengers clearly could satisfy Feldberg’s undemanding prima facie test pretty easily. And if such relatively soft showings were sufficient both to generate substantial adversarial proceedings and to shift the risk of non-persuasion to the holder of the privilege, the temptation to initiate such challenges would be considerable. The proceedings triggered by such challenges would create burdens and cause distractions — diverting party and court resources from proceedings that should focus on the primary evidence and law. Moreover, the proceedings themselves would create risks to the confidentiality of the privileged communications — as counsel might feel pressured to disclose considerable information about those communications and the circumstances in which they were made in order to reduce the likelihood that the court would not find their explanations “satisfactory.”

As Judge Easterbrook pointed out in Feldberg, 862 F.2d at 625 , more demanding standards for a “prima facie case” have been adopted by other courts. While it is clear that some of the other standards are intended to be more demanding, many things about them remain unclear. For example, I cannot tell whether there is a meaningful difference between the standard adopted in the Second Circuit, which is described as closely akin to “probable cause,” see In re Grand Jury Subpoena Duces Tecum Dated Sept. 15, 1983, 731 F.2d 1032, 1039 (2d Cir.1984), and the standard adopted in the courts of appeal for the Fifth Circuit and for the District of Columbia Circuit, which Judge Easterbrook apparently understood as requiring presentation of “evidence that by itself satisfies a more-likely-than-not standard.” See Feldberg, 862 F.2d at 625 -26 (citing In re International Systems & Controls Corp. Sec. Litig., 693 F.2d 1235 , 1242 (5th Cir.1982); In re Sealed Case, 676 F.2d 793 , 814-15 & n. 88 (D.C.Cir.1982)).

Judge Easterbrook appears to believe that the “probable cause” standard is less demanding than the “evidence sufficient to support a verdict” standard, but not all courts agree. After substantial analysis, for example, the district court in Haines v. Liggett *434 Group, Inc., 140 F.R.D. 681, 692 (D.N.J.), vacated on other grounds, 975 F.2d 81 (3d Cir.1992), concluded that “all of these proposed standards amount to the same basic proposition — has the party seeking discovery presented evidence which, if believed by the fact-finder, supports plaintiffs theory of fraud?” The Court of Appeals for the Third Circuit quoted this view with apparent approval in Haines v. Liggett Group, Inc., 975 F.2d 81, 95 (3d Cir.1992). Similarly, both the Court of Appeals for the Second Circuit and the Court of Appeals for the District of Columbia Circuit have suggested that, especially given the discretion vested in trial courts to make judgment calls with respect to these matters, “there is little practical difference between the two tests.” See In re Sealed Case, 754 F.2d 395 , 399 n. 3 (D.C.Cir.1985) (quoting with approval a similar sentiment voiced by the Second Circuit in Grand Jury Subpoena, 731 F.2d at 1039 ).

While, as the Zolin court confirmed, these matters remain confused, see 491 U.S. at 563 n. 7, 109 S.Ct. at 2626 n. 7, we are relatively confident that the Court of Appeals for the Ninth Circuit has endorsed a standard for defining “prima facie” that is on the more demanding end of the doctrinal spectrum. While the Court of Appeals for this Circuit has not offered detañed analysis or guidance on this point, and has been more active in this area of the law in unpublished dispositions than in published opinions, 21 in United States v. Laurins, 857 F.2d 529, 541 (9th Cir.1988), cert. denied, 492 U.S. 906 , 109 S.Ct. 3215 , 106 L.Ed.2d 565 (1989), the Court explicitly drew the applicable standard from the opinion of the District of Columbia Circuit in Sealed Case, 754 F.2d at 399 . Under that standard, the challenger must present evidence which, “if believed by the jury would establish the elements of [the alleged crime or fraud].” Laurins, 857 F.2d at 541 . At least according to the apparent thinking of the Court of Appeals for the Seventh Circuit, this approach would align the Ninth Circuit with those more demanding courts that seemingly require evidence which “by itself satisfies a more-likely-than-not standard” (or, stated differently, evidence that would be sufficient “to support a verdict in favor of the [chañenger]”). See Feldberg, 862 F.2d at 625-26 .

To shed some additional light on the meaning of the Ninth Circuit’s standard, we turn to examine the opinion of the Court of Appeals for the Fourth Circuit in Duplan Corp. v. Deering Milliken, Inc., 540 F.2d 1215, 1220 (4th Cir.1976). That court adopted a standard that appears to be identical to the standard later adopted by the Ninth Circuit, thus implicitly but necessarily rejecting the looser standard that the Duplan trial court had used. 22 The Duplan appeñate court phrased the test in these words: “[W]hñe a prima facie showing need not be such as to actuaUy prove the disputed fact, it must be such as to subject the opposing party to the risk of non-persuasion if the evidence as to the disputed fact is left unrebutted.” Id. (footnote omitted). This formulation has become relatively common — and sheds little additional light on our subject.

There is more light avañable through the Duplan opinion, however — in fact, through the next sentence in that opinion, a sentence that has not received as much attention as it deserves. In that next sentence, the Court of Appeals for the Fourth Circuit added the foUowing important qualifier to the standard it had just announced: “Whñe such a showing [sufficient to subject an opposing party who offers no response to the risk of nonpersuasion] may justify a finding in favor of the offering party, it does not necessarily compel such a finding.” Id. (emphasis added) (citing Wright v. Rockefeller, 376 U.S. 52, 57 , 84 S.Ct. 603, 605 , 11 L.Ed.2d 512 (1964)).

*435 To understand fully the implications of this important qualifier, one must look carefully at Wright, 376 U.S. 52 , 84 S.Ct. at 603 , the source cited for the pertinent concept. In that case the Supreme Court wrestled with alleged racial discrimination in shaping congressional districts. Despite this substantial difference in settings, the Duplan court of appeals determined that the Wright court’s thinking about the meaning of the phrase “prima facie” should become a part of the standard for determining whether a party had made a prima facie showing of the crime/ fraud exception.

Specifically, the Wright court confronted the following situation. A plaintiff with the burden of proof had made a prima facie showing in support of a particular inference of fact (that the shaping of the districts had been racially motivated). Moreover, even after considering all the contrary evidence adduced by the defendant, the evidence presented by the plaintiff was sufficient to permit a trier of fact, rationally, to draw the inference that the plaintiff wanted drawn. But, significantly, the Supreme Court also agreed that, taking the entire record into account, a rational trier of fact could have concluded that the evidence supporting the contrary inference was equally or more persuasive. In other words, on the state of the evidentiary record as a whole, it would have been rational and unassailable for the trier of fact to draw either inference (so the evidence was not so strong, in either direction, that it admitted of only one rational inference). See Wright, 376 U.S. at 55-57 , 84 S.Ct. at 604-06 . In this setting, the Supreme Court made the following statement:

Id. at 57 , 84 S.Ct. at 605-06 .

By insisting that this concept be imported into the law of the crime/fraud exception, the Duplan, 540 F.2d at 1220 , appellate court empowers a trial court, even in the face of a showing sufficient to make rational a finding of the disputed crime or fraud, to decline to make such a finding as long as, taking into account “the entire record,” 23 the judge could rationally conclude that the persuasive support for the contrary inference was at least equal to the persuasive support for the crime/fraud inference. Stated differently, in these evidentiary settings, the making of a “prima facie case” would not compel a court to find that the challenger has established the crime/fraud exception to the privilege.

The Court of Appeals for the Ninth Circuit has not said whether it would accept this view. It is significant, however, that the opinions from that court, as well as the opinions from other appellate courts, whose tone seems least receptive to this position have been written in cases involving criminal prosecutions and/or grand jury investigations. In contrast, Duplan, 540 F.2d 1215 , and Wright, 376 U.S. 52 , 84 S.Ct. 603 , 11 L.Ed.2d 512 , are civil cases — and, as such, are less likely to be affected by considerations irrelevant to the case at bar.

It also is significant that the Duplan appellate court is one of the most sensitive to the stage of the proceedings at which courts are pressed to make rulings on these kinds of issues. As that court pointed out, a determination about whether to pierce a privilege under the crime/iraud exception that is made during the discovery stage of litigation does not represent the challenger’s only or last opportunity to mount evidence on this issue. The challenger can try again during the trial — and at that stage, the challenger may be able to present additional evidence to support the inference. See Duplan, 540 F.2d at 1222 . As significant, at that stage “[t]he district *436 judge will have the advantage of seeing the witnesses and hearing them testify.” Id. In fact, the judge will have more direct and systematic exposure to all the pertinent evidence — and therefore will be in a better position to rule reliably on the important issues presented by such a challenge. These considerations further support the view that judges should exercise considerable caution when they are pressed during the discovery stage of complex litigation to find that a showing of crime or fraud that is sufficient to justify penetrating the privilege has been made.

Before answering the question about the level of certainty that should be required in a setting like this we must acknowledge one additional unresolved issue about the test. When the judge, in the pretrial period, is assessing the sufficiency of the showing, is she making judgments about how her own judicial mind responds to the evidence, or is she making judgments about what the zone of responses could be by a rational jury? The way the Court of Appeals for the Ninth Circuit articulated (without explanation) the test in Laurins, 857 F.2d 529 , suggests the latter. In its one sentence about this issue the Laurins court stated that the challenger needed to show “evidence that if believed by the jury would establish the elements of an ongoing violation.” 857 F.2d at 541 (emphasis added).

The opinion from which Laurins took this statement, however, Sealed Case, 754 F.2d at 399 , did not refer to a jury, but simply to “the trier of fact.” The Court of Appeal for the District of Columbia Circuit declared in Sealed Case that the challenger (the government) “satisfies its burden of proof if it offers evidence that if believed by the trier of fact would establish the elements of an ongoing crime or fraud.” 754 F.2d at 399 (emphasis added). 24 Moreover, the court in Sealed Case went on, in the next sentence, to emphasize that the “determination that a prima facie showing has been made lies within the sound discretion of the district court.” Id. These passages create some ambiguity about whose mind is the appropriate target of judicial inquiry here.

It seems to me that that ambiguity is resolved by Fed.R.Evid. 104(a), which is almost never discussed by the appellate courts who have made pronouncements on these matters. 25 According to the unequivocal language of the Rule, “[preliminary questions concerning ... the existence of a privilege ... shall be determined by the court____” In making a determination under 104(a), the judge generally employs a preponderance of the evidence standard. 26 This approach stands in sharp contrast to the process contemplated in paragraph (b) of Rule 104. 27 *437 That paragraph applies to determinations of different kinds of facts, facts on whose existence the relevance of other evidence turns. When making judgments under paragraph (b), the trial judge is not determining how her own judicial mind, as fact-finder, would respond to the evidence; instead, she is making a judgment about how a rational jury could respond to the evidence (under paragraph (b), she must rule in favor of admission if a rational jury could draw the necessary fact inference). 28

Given the plain language of Rule 104(a), and the absence from the cases of any articulated reason for reaching any other conclusion, I can see no basis for not concluding that a determination of the kind made necessary by Reliant’s motion to establish the erime/fraud exception falls squarely within the ambit of paragraph (a) of Rule 104, not paragraph (b). The issue here is not relevance that is conditioned on the possible fact of fraud. Rather, the issue is whether a privilege exists. The privilege does not exist, in the case at bar, only if a sufficient showing of crime or fraud is made. Moreover, the direct consequence of the court’s ruling on this kind of motion is immediate loss of the privilege — which stands in sharp contrast to the much less dramatic consequence of a ruling under paragraph (b) of the Rule (under that paragraph a ruling in favor of the moving party puts off the day of ultimate decision — which is made, after additional opportunities for presentation of evidence and argument, by a jury).

It may be significant, moreover, that the Supreme Court opinion, Clark, 289 U.S. at 14 , 53 S.Ct. at 469 , that is cited as the source of the phrase “prima fade case” — for purposes of the erime/fraud exception — was published more than forty years before the Federal Rules of Evidence clearly divided the universe of showings preliminary to the admissibility of evidence into the spheres reflected in paragraphs (a) and (b) of Rule 104. Given the fact that the phrase “prima facie case” was introduced into this doctrinal arena well before Rule 104 was adopted, and well before the courts (even later) clearly separated the standards that were to be applied under paragraphs (a) and (b) of that Rule, 29 and given the admission in Zolin, 491 U.S. at 563 n. 7, 109 S.Ct. at 2626-27 , that the Court’s use of the phrase “prima facie ease” in Clark “has caused some confusion,” it is in some measure understandable that appellate courts have used the “prima facie” language without thoroughly exploring its meaning and without trying to reconcile the meaning they assumed the phrase had with the structure and requirements of Fed.R.Evid. 104. Whatever the source of past confusion, however, I am duty bound to apply the law as it has developed over the last twenty years under the Federal Rules of Evidence. Because I can see no principled basis for declining to do so, I am not free to ignore the requirements of Rule 104(a).

Thus, even though one sentence in the Laurins opinion at least obliquely suggests otherwise, see 857 F.2d at 541 , I conclude that in ruling on a pretrial motion to pierce the privilege under the erime/fraud exception the court should proceed under Fed.R.Evid. 104(a), not 104(b), and should not make a judgment about the breadth of the zone of inferences a jury could rationally make, but, instead, should make a judgment *438 about what inferences the judge herself makes from the evidence and argument. 30

With the assistance of the considerations exposed by this review of the pertinent cases, we return to face squarely the question we posed at the beginning of this section: what level of certainty must judges reach, in the discovery stage of civil litigation, before concluding that a challenger has made a showing sufficient under the crime/ fraud exception to justify turning over to the challenger communications between an opponent and her counsel that otherwise would be protected by the attorney-client privilege? I conclude that Fed.R.Evid. 104(a), as well as a constellation of other considerations that point in the same direction, require me, at least in the circumstances here presented, to refuse to order disclosure of the otherwise privileged communications if, after taking into account all the evidence and argument presented by both parties, I conclude that the likelihood that Laser sought or used legal advice to commit a fraud is no greater than the likelihood that it did not. In other words, if a judge in the pretrial period concludes that the evidence and argument are essentially in equipoise, i.e., that the evidence commands exculpatory and inculpatory inferences equally, the judge must deny the motion to penetrate the privilege under the crime/fraud exception.

I hasten to point out that use of this standard is not tantamount to requiring Reliant to “actually prove” the alleged fraud. 31 At the trial stage, a defendant who is trying to block enforcement of a patent must prove “inequitable conduct” by clear and convincing evidence 32 — but the standard I adopt here would permit penetration of the attorney-client privilege without proof by clear and convincing evidence. Rather, the moving party could penetrate the privilege by satisfying the less demanding “preponderance of the evidence” standard. 33

At this juncture it may be useful to summarize the considerations that would support the adoption of this standard even if it were not mandated by the authorities that have construed Fed.R.Evid. 104(a). First, it is important to emphasize that the rulings I announce here apply only in civil cases. Different procedures and different standards may well be appropriate in criminal eases, or in connection with grand jury investigations, where different and arguably weightier interests sometimes support efforts to discover communications that would otherwise be protected by privilege.

Second, we must bear in mind that the standard I adopt applies only to the ultimate determination about whether the privilege has been lost, 34 not to a preliminary judgment about whether to conduct an in camera inspection to help determine whether the privilege has been lost. Since the standard for deciding whether to inspect documents in *439 camera is lower (easier for the challenger to satisfy), see Zolin, 491 U.S. at 572 , 109 S.Ct. at 2630, 2630-31 , and since in camera inspections can serve as effective tools for uncovering abuses of the privilege and thus for discouraging unlawful uses of legal services, there is no great risk that the courts will create too much room for corrupt use of lawyers by adopting a somewhat more demanding standard for actually disclosing the communications. 35

Use of the somewhat more demanding standard for determining whether the crime/ fraud exception has been established also is supported by considering the consequences of that determination. The decision to conduct an in camera inspection of documents “does not have the legal effect of terminating the privilege,” id. at 568 , 109 S.Ct. at 2628-29 , and represents “a smaller intrusion upon the confidentiality of the attorney-client relationship than ... public disclosure.” Id. at 572 , 109 S.Ct. at 2630-31 . In sharp contrast, a determination that the crime/fraud exception has been established results in a legal conclusion that the privilege does not attach to the communications and in their direct disclosure to a litigation opponent. Thus the confidentiality of the communications is likely to be immediately and permanently lost. 36 *440 This is not to say, of course, that the case is lost, 37 or even that the communications will ultimately be admitted into evidence (there could be other evidentiary barriers, e.g., Fed. R.Evid. 403). But the confidentiality itself is destroyed by the disclosure that is directly ordered — and that confidentiality can never be restored.

This is a matter of some consequence — in part because this loss strikes at the heart of the law of attorney-client privilege. After all, the underlying goal of the law of privilege is to encourage a feeling of confidence in clients that will lead them to disclose to their lawyers everything that is arguably pertinent to providing them with reliable legal advice. See, e.g., United States v. Hernandez, 937 F.2d 1490, 1493 (9th Cir.1991). And the promise of confidentiality (which is not just that the clients’ words won’t be used as evidence, but also that their words won’t be known) is presumed to be crucial to encouraging that necessary feeling of confidence. A showing of real substance should be required before we sacrifice something this essential to enabling the law of privilege to achieve its primary goal.

While a determination that the erime/fraud exception has been established has immediate and substantial consequences for the party asserting the privilege, the consequences for the challenger of an adverse ruling, especially in the pretrial period, are not permanent and often may not be of great moment. As noted above, the challenger can press his motion again during trial. At that juncture the judge will have seen all the pertinent evidence and will be in a much better position to assess its relative persuasive power (e.g., by more reliably gauging the credibility of competing witnesses). Armed with more information, the judge might change her mind about the challenger’s showing and order disclosure of the communications. At a minimum, her decision will be more reliable.

Even if the challenger loses again in that setting, however, he has only lost access to confidential communications — not to underlying evidence or facts. While this loss could damage his ability to prove by direct evidence the mental element of the alleged fraud, he would still have full access to the evidence that could prove that element circumstantially. And when the underlying allegation is of fraud on the Patent Office by failing to disclose material prior art, the challenger can show the trier of fact the items of prior art themselves and argue that they were so clearly material that the mental element can be inferred. See Glaxo Inc. v. Novopharm Ltd., 52 F.3d 1043, 1048 (Fed. Cir.), cert. denied, — U.S.-, 116 S.Ct. 516 , 133 L.Ed.2d 424 (1995).

Use of the somewhat more demanding standard for establishing the erime/fraud exception also reduces the likelihood that these issues will be wrongly decided. If it is important to preserve the vitality of the privilege, it is important to make sure that decisions that strip parties of privilege protections that otherwise would attach are reliably made. Moreover, as noted earlier, the risk of error grows when the fraud allegation is based on technically dense or esoteric evidence (matters well beyond the common experience of generalist judges). Given that risk, the magnitude of the harms that can be caused by erroneous rulings on these *441 kinds of motions, and the availability of other tools (most notably, in camera inspections) to discourage abuse of legal advice, there is little reason to use any of the relatively less substantial standards that might be appropriate in other settings (e.g., when a court needs to protect the secrecy of an ongoing grand jury investigation).

It also is important to bear in mind that the underlying justification for penetrating the privilege in these settings is at least substantial concern that the client has used or attempted to use legal services to facilitate the commission of a serious wrong. It simply doesn’t seem consistent with our basic notions of fairness to take away a valuable right (or to invade an otherwise protected and valued interest) if a judge, even after examining (in camera) the very communications that are alleged to have been made in furtherance of the unlawful activity, as well as all the independent evidence the challenger can muster, cannot say that it is at least marginally more likely than not that the client was using the lawyer to commit the alleged wrong. 38

Using a less substantial, poorly defined, and only vaguely understood standard also makes rulings on challenges to the privilege less predictable. When the rulings are less predictable, the incentives to launch challenges increase — as do the burdens on the courts. More importantly, less predictable outcomes undermine, fundamentally, clients’ confidence in the permanence of the confidentiality of their private communications with their counsel. As noted above, without that confidence, clients are less likely to share all pertinent information with then-lawyers. And the core purpose of the law of privilege is defeated when clients withhold pertinent information from their counsel.

For all these reasons, I hold that a judge sitting in a civil case must deny a motion to penetrate the attorney-client privilege under the crime/fraud exception if, after considering all the evidence and argument offered by both the challenger and the party who is invoking the privilege, the judge cannot say that it is more likely than not that the party resisting the disclosures sought or used legal advice to commit or to try to commit a crime or fraud.

In the next section we explain why, when we apply the standards described above, we conclude that Reliant has failed to establish the crime/fraud exception to the attorney-client privilege.

B. Application of the Legal Standards to the Pertinent Evidence

1. Introduction.

Given the specific predicates for its motion, Reliant may not pierce the attorney-client privilege under the crime/fraud exception unless it can show that it is more likely than not that Laser perpetrated fraud — traditional, common-law fraud — upon the Patent Office. In this context, Reliant must make sufficient showings that (1) Laser or its agents made a *442 “knowing, willful [39] and intentional” misrepresentation or omission to the Patent Office during the prosecution of the application for the ’502 patent, (2) the misrepresentation or omission was material; and (3) the Patent Office relied upon the misrepresentation or omission in issuing the patent. See Union Carbide Corp. v. Dow Chemical Co., 619 F.Supp. 1036, 1052 (D.Del.1985); American Optical Corp. v. United States, 179 U.S.P.Q. 682 , 684, 1973 WL 20128 (Ct.Cl.1973). To satisfy the third element, Reliant must show that the Patent Office would not have issued the ’502 patent if it had been given the information that Laser allegedly omitted or if it had not been misled in the manner that Reliant alleges. See Union Carbide, 619 F.Supp. at 1052 ; American Optical, 179 U.S.P.Q. at 684.

In explaining why Reliant has not made the kind of showing necessary to pierce the privilege, we proceed in the following manner. First, we describe the prior art that Reliant accuses Laser of failing to disclose to the Patent Office. Then we describe the bases for our conclusion that Laser in fact disclosed to the Patent Office much of the prior technology that Reliant inaccurately claims Laser failed to disclose. In the next section we identify what appear to us to be important differences between what was claimed in the ’502 patent and the limited prior art that Laser arguably failed to disclose. It is in this last section that we conclude that because of these significant differences, Reliant cannot show that it is more likely than not that the Patent Office would have denied the application for the ’502 patent if Laser had disclosed the limited prior art about whose existence we cannot be confident the patent examiner was aware.

2. The ’502 Patent, the ’396 Patent, and Related Devices.

a. The ’502 patent.

The ’502 patent is a patent for a surgical laser system. The ’502 patent is described in its abstract as a “system for causing uniform ablation [40] of irradiated material of living tissue while not causing damage below a predetermined depth where laser radiation is provided sequentially and continuously in a predetermined pattern.” Reliant Ex. 1.

The system provided for by the ’502 patent works as follows. The surgeon uses a manual laser (which does not damage tissue) to optically map the borders of the area of the tissue which he wishes to remove. Then, a surgical laser automatically goes over the area which was mapped and removes (“ablates”) the undesirable tissue. The surgical laser goes over the tissue area in a curved, continuous pattern. The ’502 patent claims that the combination of elements that make up the system it discloses enables physicians to remove entire sections of target tissue while limiting, to a pre-determined, uniform depth, the damage to tissue that is beneath the tissue that is to be removed. See Reliant Ex. I. 41

The ’502 patent states that a carbon dioxide laser is used in “the preferred embodiment of this invention,” but the patent apparently does not require a carbon dioxide laser to be used. See Reliant Ex. 1.

b. The ’396 patent.

Laser did not disclose patent number 4,587,396, which had been issued in 1986 to Rubin (“the ’396 patent”). The ’396 patent is not, on its face, limited to surgical lasers. Succinctly, the ’396 patent discloses a laser device for ablating tissue whose boundary has been mapped by another, manually-controlled laser. After the manually-controlled laser maps the boundary of the target area, the automatic laser traverses the tissue in a “raster” pattern, first row-by-row, then column-by-column. See Reliant Ex. 4.

c. Devices based on the ’396 patent.

Laser manufactured at least three surgical laser devices based on the ’396 patent — the Sharplan 771 Microscan, the Sharplan 771B *443 Mierosean, and the Sharplan 775. During the prosecution of the application for the ’502 patent, Laser disclosed neither these devices nor certain brochures and a published article about them. Reliant contends that Laser’s failure to submit these materials to the patent examiner supports a finding that Laser committed the alleged fraud.

The published article to which Reliant points was written by Dr. Jacob Dagan, and was submitted by Reliant as an exhibit to its moving papers. See Reliant Ex. 5. The Dagan article focuses on the “Model 771 Mierosean,” which apparently is the same device elsewhere called the “Sharplan 771 Mierosean.” The article states that the device’s “microprocessor-controlled scanner introduces the capability of swift accurate scanning to the extent that damage in adjacent tissues is minimized even with the use of a continuous beam.” Id. (emphasis added). The article states that the 771 “increas[es] the speed of dissection by allowing the use of a continuous beam and minimizing the damage to surrounding tissues.” Id. (emphasis added). Elsewhere, the Dagan article states that the surgical laser goes over the tissue to be excised “first in the horizontal direction, from left to right, and then in the vertical direction, from top to bottom.” Id. The article also states that the surgery department where the article’s author worked used a carbon dioxide laser with the 771 device. Id.

Reliant also attached the brochures in issue to its moving papers. See Reliant Ex. 6-8. Exhibit 6 is a brochure about the Sharplan 771 Mierosean. The brochure states that the 771 “[mjinimizes damage to surrounding healthy tissue.” Reliant Ex. 6.

Exhibit 7 is an operating manual for the Sharplan 771B Mierosean. The manual explains that the laser goes over the unwanted tissue row-by-row first, and then eolumn-bycolumn. See Reliant Ex. 7.

Exhibit 8 is a brochure about the Sharplan 775. The brochure states that it enables a surgical laser to excise tissue “with utmost uniformity and reproducibility.” Reliant Ex. 8.

3. Prior Art Laser Submitted to the Patent Office.

a. Elements which Reliant claims would have been disclosed by the ’396 patent and related devices and which Reliant alleges were withheld from the Patent Office.

Reliant claims that the ’396 patent and related devices taught or used several technological elements which also are components of the ’502 patent. While Reliant’s arguments are somewhat confusing and repetitive, Reliant appears to contend that four main technological elements would have been revealed to the Patent Office through disclosure of the ’396 patent and the related devices and materials. First, Reliant argues that the devices based on the ’396 patent were able to keep “necrosis” (damage to tissue underlying excised tissue) from going beyond a predetermined depth, which is what the ’502 patent claims to do. Second, Reliant contends that the devices used a carbon dioxide laser, which is the type of laser that the ’502 patent recommends. Third, Reliant argues that the ’396 patent and related devices disclosed a system for continuously moving a laser over the area of tissue which the surgeon desires to excise, something which the ’502 patent also claims. Fourth, Reliant argues that the ’396 patent taught a system for uniform “ablation” (removal) of an area of tissue, which is another claim of the ’502 patent.

Before discussing whether the ’396 patent and related devices actually contained or taught the four technological elements that Reliant claims they did, we describe how Laser in fact disclosed some of these four technological elements to the Patent Office through other prior art that Laser cited.

b. The Itzkan patent.

Laser disclosed U.S. Patent No. 4,733,660, granted to Itzkan (“the Itzkan patent”), to the Patent Office. The Itzkan patent disclosed the use of an argon or dye laser for selective ablation of tissue, where the laser was moved continuously to prevent necrosis (damage) of tissue outside the specifically selected target areas. See Laser Ex. E. While the Itzkan patent did not recommend *444 the use of a carbon dioxide laser, it did discuss other prior art which had used carbon dioxide lasers. See Laser Ex. E. The patent examiner, in rejecting one of Laser’s earlier applications for what eventually became the ’502 patent, 42 stated that it would have been obvious to provide the system described in the Itzkan patent with a carbon dioxide laser. See Reliant Ex. 3 at 84. Significantly, the examiner also took the position that “Itzkan provides a system for causing ablation of a target material of living tissue while not causing necrosis below a certain depth, comprised of a scanning mechanism for moving laser radiation in a predetermined path, as claimed.” Id. at 83 .

Thus, in his discussion of the Itzkan patent, the patent examiner had concluded that he had been exposed to three of the four technological elements that Reliant claims were not disclosed by Laser — prevention of necrosis (damage) below a certain depth, continuous movement of a laser, and use of a carbon dioxide laser.

c. The Davi patent.

Laser also disclosed to the patent examiner U.S. Patent No. 4,469,098, granted to Davi (“the Davi patent”). Like the Itzkan patent, Davi taught a system for removing tissue while limiting unwanted necrosis. See Laser Ex. J. In fact, the examiner asserted, as part of his justification for one of the rejections of Laser’s application, that “Davi discloses a system for causing ablation of living tissue without causing necrosis below a predetermined depth.” Reliant Ex. 2 at 36. Unlike the Itzkan and ’502 patents, however, the Davi patent used a pulsed laser beam to limit necrosis, instead of a continuous laser beam. See Laser Ex. J.

It is clear from the file history that the examiner believed that the Itzkan and Davi patents disclosed three of the four technological components of the ’502 patent that Reliant claims Laser failed to disclose. Of the components as characterized by Reliant, only one appears not to have been disclosed, at least in the examiner’s mind, by either the Itzkan or the Davi patents: the uniform ablation of an entire pre-marked area of tissue. We do not know whether this technological component was disclosed to the Patent Office through any other prior art, whether it was generally known to the Patent Office, and/or whether it represented any kind of innovation at the time the ’502 patent application was submitted.

4. Key Differences Between the ’502 Patent and the ’396 Patent.

Despite the fact that Laser has not demonstrated that it informed the Patent Office of the component of uniform ablation of a tissue area, we conclude that Reliant cannot show that the patent examiner would not have issued the ’502 patent if the ’396 patent and related devices had been disclosed. In reaching this conclusion, we focus on the justification that the examiner ultimately offered for approving the application for the ’502 patent. The “Examiner’s Statement of Reasons for Allowance” reads as follows: “The prior art of record fails to provide a scanner means for moving laser radiation in a predetermined pattern sequentially and continuously as well as providing means for controlling the scanning means so as to cause uniform irradiation to a predetermined depth.” Reliant Ex. 2 at 53.

This statement indicates that what the patent examiner considered innovative about the ’502 patent were the elements (in combination) of (1) providing means for continuous (wholly uninterrupted) movement of the laser beam and of (2) preventing necrosis below a uniform depth by providing means for controlling the scanning process, not the element of uniform ablation of an area of tissue or the element of using a carbon dioxide laser.

The significance of this basis for the examiner’s decision is that Laser has submitted «¿«.contradicted evidence that the ’396 patent and the devices and literature related to it did not disclose a means for continuous irradiation of tissue through continuous movement of a laser beam that was neither shuttered nor turned off during the surgery. Nor did the earlier, undisclosed material pro *445 vide for using that continuous movement and irradiation (in combination with other elements) to limit necrosis to a predetermined depth. Reliant has failed to adduce any evidence that would support a finding that the ’396 patent (or related materials) disclosed a beam that moved and irradiated continuously in a system that limited unwanted necrosis to a pre-determined and fixed depth.

Laser submitted a declaration from Joshua Raif, who served as Laser’s vice-president of research and development and then as Laser’s chief scientist during the time frames in which the ’396 patent and the ’502 patents were developed. Laser also submitted a declaration from the inventor of the subject matter of the ’502 patent, Eliezer Zair, which was similar to Raifs declaration but much less detailed. See Laser Ex. D. Raif declared that the three devices based on the ’396 patent (the 771, 771B, and 775) did not continuously irradiate tissue, but turned off the laser at the end of each scanned line or column, in order to move to the next line or column. Raif Decl. at 2. Thus, even though the 771, 771B, and 775 used a continuous (as opposed to a pulsed) beam, the laser was periodically turned off, so irradiation was not continuous. Raif Decl. at 4; see also Zair Decl. at 6. Unlike the devices based on the ’396 patent, which used a row-by-row and column-by-column “raster” scanning pattern, the ’502 patent provides for continuous, uninterrupted irradiation of tissue through the use of curved scanning patterns. Raif Decl. at 5; see also Zair Decl. at 6.

In addition, Raif declared that the ’502 patent has several other features which enable it to minimize damage to underlying tissue and keep such damage from going below a predetermined depth, features which Reliant has apparently not claimed were present in the ’396 patent or the devices based on it. First, the ’502 patent claims to be able to keep the laser from dwelling on any one spot for more than one millisecond. By contrast, the devices based on the ’396 patent had a minimum dwell time of about seven milliseconds. Raif Decl. at 5-6; see also Zair Decl. at 5-6. Second, the ’502 patent provides for a laser spot size (the diameter of the laser beam as it hits the tissue) of .2 millimeters to .35 millimeters. On the other hand, the ’396 patent devices were unable to use a spot size smaller than .45 millimeters. Raif Decl. at 6. Finally, the ’502 patent prescribed specific power levels for the purpose of limiting the depth of necrosis; the ’396 patent and related devices did not do so. Raif Decl. at 6-7. For these reasons, the devices based on the ’396 patent, according to Raif, were “not designed to, did not and could not produce tissue ablation with either minimal thermal necrosis or avoiding necrosis below a predetermined depth.” Raif Decl. at 2; see also Zair Decl. at 5.

Reliant has not adduced evidence that contradicts these aspects of Raifs and Zair’s declarations, or that otherwise would support a finding that the ’396 patent and related devices claimed to irradiate tissue continuously or to be capable of preventing damage to underlying tissue below a predetermined, specified depth. Reliant has pointed to statements in an article and a brochure about the Sharplan 771 Microscan that claimed that the devices were able to minimize damage to tissue described as “adjacent” in some passages and “surrounding” in other passages. See Reliant Ex. 5, 6. However, none of the literature about the devices based on the ’396 patent specifically claimed that those devices were designed to or able to keep damage to tissue below the tissue a surgeon desired to excise from going beyond a certain, minimal depth. 43

*446 Since the undisclosed prior art did not contain or teach the technological elements which were the basis of the patent examiner’s decision to grant the patent, we conclude that Reliant has not shown that it is more likely than not that the Patent Office would have denied the application for the ’502 patent if the prior art had been disclosed. Reliant’s failure to make this showing requires us to DENY its motion to pierce the attorney-client privilege based on the crime/iraud exception.

Because Reliant has failed to make the necessary showing with respect to the reliance element of its fraud claim, we need not address the other two elements that Reliant also would be required to satisfy in order to prevail in this motion.

III. WAIVER

Reliant contends that Laser waived the attorney-client privilege through declarations that Laser submitted as part of its opposition papers. In a lengthy declaration, 44 attorney Richard Samuel, who prosecuted the ’502 patent for Laser, stated that he never engaged in fraud or inequitable conduct before the U.S. Patent Office and that he had no knowledge (at the time the patent application was prosecuted) of the ’396 patent or the items relating to it which were not disclosed to the Patent Office. Samuel Deck at 2. Samuel also declared that he never communicated about the prosecution of the ’502 patent with Benjamin Barish, a patent attorney who prosecuted an application for the equivalent of the ’502 patent in Israel. Samuel Deck at 5. Eileen Ebel, another attorney who worked on the prosecution of the ’502 patent for Laser, submitted a brief declaration stating that she also did not know about the ’396 patent and related materials at the time the ’502 patent was prosecuted and that she did not engage in inequitable conduct before the Patent Office.

Waiver of the attorney-client privilege can be either express or implied. An implied waiver of the privilege occurs if “(1) [t]he party asserting the privilege acts affirmatively (2) to place the privileged com-,munications in issue between the party seeking discovery and itself (3) such that denying access to the communication becomes manifestly unfair to the party seeking discovery.” Principle Business Enterprises, Inc. v. United States, 210 U.S.P.Q. 26 , 27 (Ct.Cl.1980). Courts are less likely to find waiver when the statements alleged to constitute waiver do not purport to disclose the contents of a specific communication, and when the statements are “pre-trial matters, not evidence.” See American Optical Corp. v. Medtronic Inc., 56 F.R.D. 426, 432 (D.Mass.1972). Courts also are less prone to conclude a waiver occurred where the statements at issue do not go beyond “a mere denial of intent.” Cf. General Elec. Co. v. Hoechst Celanese Corp., 15 U.S.P.Q.2d 1673 , 1679, 1990 WL 154218 (D.Del.1990).

. In this case, Samuel’s and Ebel’s denials of knowledge of the undisclosed prior art did not purport to disclose the contents of a specific communication, cf. American Optical, 56 F.R.D. at 432 , and do not appear to go beyond “mere denial[s].” Cf. General Elec., 15 U.S.P.Q.2d at 1679. More important, their denials have not placed the contents of any privileged communication “in issue,” cf. Principle Business Enterprises, 210 U.S.P.Q. at 27, because our ruling on Reliant’s crime/iraud motion is not affected at all by assumptions or evidence about the states of mind, knowledge, or intentions of Laser’s patent counsel. 45 As the assertions by Laser’s attorneys about their alleged lack of knowledge of prior art, as well as their denials of inequitable conduct, have brought Laser no advantage in the battle over this motion, it would not be “manifestly unfair” to Reliant to deny it access to Laser’s communications. Cf. id.

*447 The portions of Samuel’s declaration which essentially consist of argument about the materiality of both disclosed and undisclosed prior art, the similarities between the ’502 patent and disclosed and undisclosed prior art, and the quasi-legal issue of whether Laser’s conduct before the Patent Office amounted to “inequitable conduct,” likewise cannot constitute a basis for waiver of the privilege. They do not put Samuel’s state of mind or the contents of any confidential communication in issue.

While we reject Reliant’s waiver argument, we agree that it would be unfair for Laser’s attorneys to offer testimony at trial about their knowledge (before the ’502 patent issued) of the ’396 patent and related materials without giving Reliant an opportunity to examine the confidential communications which may confirm or refute the accuracy of such testimony. Thus, if Laser intends to have its attorneys make statements at trial about the contents of any confidential communications, or about the attorneys’ knowledge about pri- or art at the time the ’502 patent application was being prosecuted, Laser will have to disclose any confidential communications which may verify or disprove the truth of such statements.

IY. SUMMARY AND ORDER

It is ordered that:

1. Reliant’s motion to pierce the attorney-client privilege based on the erime/fraud exception is DENIED. Reliant has failed to make a sufficient showing that the Patent Office would have denied Laser’s patent application if Laser or its counsel had presented the prior art that was not disclosed.

2. Reliant’s waiver argument also is DENIED at this time. In ruling on this motion, the court has not relied in any measure on the statements by Laser’s patent prosecution attorneys about their knowledge of the prior art and their good faith. Moreover, those statements do not disclose the contents of any specific communication.

IT IS SO ORDERED.

. Reliant does not argue here that the attorney-client or work-product privileges never attached to the discovery in issue in the first place.

. Some jurisdictions require fraud to be proven by clear and convincing evidence, while others require only a preponderance of the evidence. See, e.g., General Signal Corp. v. MCI Telecommunications Corp., 66 F.3d 1500, 1506 (9th Cir. 1995), cert. denied, - U.S. -, 116 S.Ct. 1017 , 134 L.Ed.2d 97 (1996).

. After considering several pertinent policy considerations, the U.S. Supreme Court in Zolin concluded:

491 U.S. at 572 , 109 S.Ct. at 2630-31 (citations omitted).

. See supra note 3.

. See supra note 3.

. The one Zolin factor that the discussion below does not even implicitly reach is "the volume of materials the district court has been asked to review.” See Zolin, 491 U.S. at 572 , 109 S.Ct. at 2630-31 . While we do not have sufficient information from counsel to determine the full extent of the documents that are being sought here, we note that Reliant's motion asks the court to order Laser to:

The breadth of these demands suggests that Reliant would be asking the court to examine a vety large volume of documents that would qualify for protection under the privilege unless Reliant were to establish the crime/fraud exception.

The larger the volume of documents reviewed in camera, the greater the potential harm to the privilege and the greater the burden on the courts. These kinds of burdens consume finite judicial resources, resources that judges could otherwise be committing to other pressing obligations, like presiding at trials, ruling on substantive motions, and helping parties negotiate settlements. In other words, large-scale inspections of the kind that would be involved here impose real and substantial costs on the administration of justice.

. See Raif Decl.; Zair Decl.

. There may be another, less significant difference. It may be somewhat more difficult to prove the requisite mental element under a common-law fraud theory than under an inequitable conduct claim. The former appears to require clear proof (at least in some jurisdictions, and perhaps through circumstantial evidence) of intent to deceive, while the latter could he satisfied on a showing that the cause of the material omission was less culpable “gross negligence.” Compare Grand Union Mount Kisco Employees Fed. Credit Union v. Kanaryk, 848 F.Supp. 446, 455 (S.D.N.Y.1994) with Union Carbide, 619 F.Supp. at 1052 n. 14.

. Our conclusion also is supported by the fact that the Court of Appeals for the Ninth Circuit seems to have accepted the notion that the privilege cannot be penetrated on a showing of offense to just any legal norm; rather, the showing must be of “an offense serious enough to defeat the privilege.” See United States v. Laurins, 857 F.2d 529, 540 (9th Cir.1988) (citing In re Sealed Case, 754 F.2d 395, 399 (D.C.Cir.1985)), cert. denied, 492 U.S. 906 , 109 S.Ct. 3215 , 106 L.Ed.2d 565 (1989).

. See Markman v. Westview Instruments, Inc., 52 F.3d 967 (Fed.Cir.1995), aff'd, -U.S.-, 116 S.Ct. 1384 , 134 L.Ed.2d 577 (1996).

. See Glaxo Inc. v. Novopharm Ltd., 52 F.3d 1043, 1048 (Fed.Cir.), cert. denied , — U.S.-, 116 S.Ct. 516 , 133 L.Ed.2d 424 (1995).

. Note, however, that some of the cases discussed by Professor Rice in this section are unpublished opinions which, in the Ninth Circuit at least, have no precedential value and cannot be cited. Compare Rice, supra, § 8:6 with 9th Cir.R. 36-3.

. In a footnote of its opinion in Zolin the high court acknowledged that the meaning of the phrase "prima facie case" remained unclear in this setting, but declined to shed light on this subject. See United States v. Zolin, 491 U.S. 554 , 563 n. 7, 109 S.Ct. 2619, 2626, n. 7 , 105 L.Ed.2d 469 (1989).

. Professor Rice's treatise discusses some unpublished Ninth Circuit opinions that also arguably have implications for the issue discussed in this section. See Rice, supra, § 8:5, at Supp. 112-19 (Supp.1996).

15. Feldberg, 862 F.2d 622 , was decided before Zolin, 491 U.S. 554 , 109 S.Ct. 2619 , 105 L.Ed.2d 469 . Feldberg endorsed a less demanding standard for determining whether a "prima facie case” had been made because the Feldberg court took the position that the function of making such a case was somewhat limited: rather than resulting directly in loss of the privilege, the prima facie showing created an obligation in the holder of the privilege to come forward with an explanation that satisfied the trial court that the inference of crime or fraud should not be drawn.

See 862 F.2d at 626 . Thus the Feldberg court clearly contemplated giving the holder of the privilege an opportunity to present evidence and argument whose purpose was to rebut the prima facie showing.

In a subsequent section of the text, I discuss whether, under the authorities from the Court of Appeals for the Ninth Circuit, it would be appropriate to adopt the Seventh Circuit's less demanding definition of a prima facie case. See infra § 11(A)(5).

. In addition, the more we have worked on this case, the more anxious we have become that a rule requiring only the moving side’s papers to be considered would create incentives for parties in technologically complex cases to adopt strategies designed to confuse and overwhelm the judge in order to increase their chances of piercing the privilege. It seems probable, especially if the standard for making a prima facie case is fairly soft and the court is required to view evidence in the light most favorable to the opponent of the privilege, as in Duplan, 397 F.Supp. at 1196 , that presenting a confusing case could lead at least some judges to be more likely to pierce the privilege.

. Another reason to consider evidence presented by both sides in ruling on motions to pierce the privilege in patent cases is that there will normally be a Patent Office record of the proceedings before the PTO and the prior art presented to the PTO. If the opponent of the privilege presents only part of that record to the court, it would not seem to make sense or be just for the court to refuse to consider the rest of that record.

. These values clearly could be threatened here because it has not been disputed that the privilege applies to the communications in issue unless Reliant can establish the crime/fraud exception.

. The existence of that discretion has been confirmed in subsequent opinions by at least some courts of appeals. See, e.g., In re Sealed Case, 754 F.2d 395, 399-400 (D.C.Cir.1985) (stating that "[t]he determination that a prima facie showing has been made lies within the sound discretion of the district court and may be disturbed on appeal only if the district court abused its discretion”).

As we discuss in the text below, it is not clear that the Court of Appeals for the Ninth Circuit would apply the "abuse of discretion” standard when reviewing at least some kinds of rulings on the crime/fraud exception. See infra note 36. It is noteworthy, however, that the Court of Appeals for this Circuit explicitly adopted the standard for (definition of) a "prima facie case" in this setting from the Sealed Case, 754 F.2d at 399 , opinion. See United States v. Laurins, 857 F.2d 529, 541 (9th Cir.1988), cert. denied, 492 U.S. 906 , 109 S.Ct. 3215 , 106 L.Ed.2d 565 (1989).

. As we will make clear later in the text, the Court of Appeals for the Fourth Circuit did not endorse the relatively undemanding standard for defining a prima facie showing that the trial court used in that case. See Duplan Corp. v. Deering Milliken, Inc., 540 F.2d 1215, 1220 (4th Cir.1976).

. The Ninth Circuit has devoted relatively little analytical attention to these issues in published opinions, but has ventured into this arena on several occasions in unpublished (and uncitable, see 9th Cir.R. 36-3) dispositions. See Paul Rice, Attorney-Client Privilege in the United States §§ 8:5 — 8:6, at Supp. 113-19 (Supp.1996).

. That standard, which Reliant urges us to adopt, was articulated by the trial court as follows: “|TIhe court’s function in making a determination as to whether there has been a prima facie showing is akin to that of a grand jury, i.e., only the evidence of the prosecution is considered and in the most unfavorable light against the accused.” Duplan, 397 F.Supp. at 1196 .

. This phrase is quoted by the Supreme Court from Judge Feinberg's concurrence with the majority opinion of the three-judge district court which initially evaluated the evidence in Wright. See Wright, 376 U.S. at 56 , 84 S.Ct. at 605 (quoting Wright v. Rockefeller, 211 F.Supp. 460, 471 (S.D.N.Y. 1962) (Feinberg, D.J., concurring), aff'd, Wright, 376 U.S. 52 , 84 S.Ct. 603 , 11 L.Ed.2d 512 ).

. See also Haines, 975 F.2d at 95-96 (referring to "evidence which, if believed by the fact-finder, would be sufficient to support a finding that the elements of the crime-fraud exception were met”) (emphasis added).

. The one notable exception is the Supreme Court's opinion in Zolin, 491 U.S. at 565-68 , 109 S.Ct. at 2627 -29 — but there most of the discussion of Rule 104 was directed not toward which paragraph of the Rule applies in this setting, but to whether paragraph (a) of the Rule prevents courts from conducting an in camera review of the otherwise privileged communications to help determine whether the crime/fraud exception has been established. As noted earlier, the Zolin court expressly left open the issue with which we wrestle in the case at bar, i.e., what "quantum of proof [is] necessary to establish the applicability of the crime-fraud exception.” See id. at 563 n. 7, 109 S.Ct. at 2626 n. 7.

. See, e.g., Bourjaily v. United States, 483 U.S. 171, 175-176 , 107 S.Ct. 2775, 2778-79 , 97 L.Ed.2d 144 (1987) (ruling on the level of certainty required under Rule 104(a) when a district court determines whether the preliminary facts necessary to admit evidence under Fed.R.Evid. 801(d)(2)(E) have been established); In re Paoli R.R. Yard PCB Litigation, 35 F.3d 717 , 743 n. 9 (3d Cir.1994) (while discussing the standard a trial judge should apply when determining the admissibility of scientific evidence under Rules 702 and 104(a), the Court of Appeals declared: "By holding that the admissibility of scientific testimony is governed by Rule 104(a), Daubert[v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 591-93 , 113 S.Ct. 2786, 2796 , 125 L.Ed.2d 469 (1993),] clearly holds that the parly seeking admissibility must make out more than a prima facie case of reliability.”), cert. denied, -U.S. -, 115 S.Ct. 1253 , 131 L.Ed.2d 134 (1995).

. Fed.R.Evid. 104(b), entitled "Relevancy conditioned on fact,” provides: “[W]hen the relevancy of evidence depends upon the fulfillment of a condition of fact, the court shall admit it upon, or subject to, the introduction of evidence sufficient to support a finding of the fulfillment of the condition."

. In United States v. Enright, 579 F.2d 980, 983-985 (6th Cir.1978), the court (as in Bourjaily, 483 U.S. at 175-76 , 107 S.Ct. at 2778-79 ) discussed the standard a trial judge should apply under 104(a) when determining whether the requirements of Fed.R.Evid. 801(d)(2)(E) had been satisfied. In that setting, and en route to concluding that the appropriate standard is “more likely than not,’’ the Sixth Circuit Court of Appeals declared:

Enright, 579 F.2d at 984-985 .

. That clarification arguably was not completed until at least 1987, when the Supreme Court published its decision in Bourjaily, 483 U.S. at 175-76 , 107 S.Ct. at 2778-79 .

. Thus, in ruling on a motion like Reliant's, the question for the judge should not be whether a rational jury that took into account all the evidence could conclude that the challenger had proved fraud by the holder of the privilege.

. Some courts have stated that it is not necessary to "actually prove" the alleged wrong in order to make the kind of prima facie showing that is sufficient to penetrate the privilege under the crime/fraud exception. See, e.g., Duplan, 540 F.2d at 1220 & n. 5.

. See, e.g., LaBounty Mfg., Inc. v. United States Int'l Trade Comm’n, 958 F.2d 1066, 1070 (Fed. Cir.1992). In order to prove Common-law fraud at the liability stage at trial, some jurisdictions, such as New York, require clear and convincing evidence, while other jurisdictions, such as California, require only a preponderance of the evidence. See, e.g., General Signal Corp. v. MCI Telecommunications Corp., 66 F.3d 1500, 1506 (9th Cir.1995), cert. denied, - U.S. -, 116 S.Ct. 1017 , 134 L.Ed.2d 97 (1996).

. What the moving party must establish by a preponderance of the evidence, however, is common-law fraud, not just "inequitable conduct.”

. In theory, establishing the crime/fraud exception does not cause the "loss” of the privilege; rather, after the requisite showing, the law concludes that the privilege never attached in the first place. In the real world, however, a party who can prove that all the other requirements for attachment of the privilege are satisfied would be able to maintain the confidentiality of the communications until the challenger established the crime/fraud exception. Thus it is accurate to say that the party who invokes the privilege "loses” the protection that that party would otherwise enjoy when the challenger makes a showing that is sufficient to establish the crime/fraud exception.

. Under the approach I adopt here, a judge who has conducted an in camera inspection of otherwise privileged conununications would apply the "more likely than not" standard when she decides whether the challenger has established the crime/fraud exception. Before making that determination, the trial judge will have carefully considered not only the evidence that is independent of the subject communications, but also evidence that she discovers in those communications themselves during her in camera inspection. It seems entirely appropriate to deny the motion if, even after the judge has studied the challenged communications themselves, she cannot say that it is more likely than not that the communications were made in furtherance of a fraudulent or criminal purpose.

. A holder of the privilege usually would face substantial obstacles in any effort to persuade a court of appeals to intervene, before entry of final judgment in the district court, to "correct” an erroneous ruling that the crime/fraud exception had been established. Appellate courts generally require extraordinary showings before they will hear challenges to discovery orders on an interlocutory basis. See, e.g., Chase Manhattan Bank, N.A. v. Turner & Newall, PLC, 964 F.2d 159, 162-64 (2d Cir. 1992). And even though there are a number of reported appellate decisions that respond to attacks on orders by trial judges to disclose arguably privileged materials, it appears that appellate courts are not likely to give prompt attention to such matters unless the district judge.has made an error of law that is both clear and important. See, e.g., United States v. Billmyer, 57 F.3d 31, 34-35 (1st Cir.1995); Admiral Ins. Co. v. United States Dist. Court for Dist. of Ariz., 881 F.2d 1486, 1491-92 (9th Cir. 1989). Where the district judge’s ruling appears to turn on judgment calls about facts (or competing fact inferences from evidence), the deference that appellate courts would normally give trial court decisions reinforces the already considerable disinclination to accept interlocutoiy appeals. Cf. Ralls v. United States, 52 F.3d 223, 225 (9th Cir.1995).

It is worth noting, here, that the opinion on which the Court of Appeals for the Ninth Circuit relied in Laurins, 857 F.2d at 541 , when it described “prima facie case” for purposes of the crime/fraud exception, declared unequivocally that the “determination that a prima facie showing has been made lies within the sound discretion of the district court and may be disturbed on appeal only if the district court abused its discretion.” See Sealed Case, 754 F.2d at 399 -400 (citing In re Berkley & Co., 629 F.2d 548 , 553 (8th Cir.1980); In re September 1975 Grand Jury Term, 532 F.2d 734, 737 (10th Cir.1976)).

Despite its reliance in Laurins on Sealed Case, it appears that the Court of Appeals for this circuit would not always give this level of deference to rulings on the crime/fraud exception. In a good many settings, Ninth Circuit panels have insisted that many privilege issues often present mixed questions of fact and law that are to be reviewed de novo. See, e.g., Tennenbaum v. Deloitte & Touche, 77 F.3d 337, 340 (9th Cir. 1996). On the other hand, the Court also has stated repeatedly that a determination of fact that is pertinent to a privilege ruling will be reviewed under the more deferential "clear error” standard. See, e.g., Ralls, 52 F.3d at 225 ; In re Grand Jury Subpoenas Dated Dec. 10, 1987, 926 F.2d 847, 858-59 (9th Cir.1991). So how much intellectual independence the appellate court brings to its review of any given ruling on the crime/fraud exception should turn on whether that ruling is based principally on inferences from the evidence about what the facts are or principally on at least quasi-legal reasoning about the legal implications or significance of those facts. But even when the ruling (as in the case at bar) arguably involves the latter kind of reasoning, appellate courts will not be quick to intervene on an interlocutory basis unless the case involves matters of unusual visibility and consequence or a clearly legal principle of more general applicability is at stake (as opposed to more case-specific and limited disagreement about what the proper outcome should be, on a specific record, of a quasi-legal reasoning process). *440 The upshot is that in a good many cases, and especially in those where the ruling on the crime/ fraud challenge turns principally on inferences from the evidence about what the holder of the privilege did and what his motives were, a determination by the trial judge that the exception has been established is likely to result in a loss of confidentiality that is permanent — even if there is some modest chance of reversing, eventually, the harm to the merits of the litigation.

. As several courts have pointed out, a ruling that a showing is sufficient to establish the crime/ fraud exception is made under different standards and is not equatable with a ruling on the ultimate liability questions. Cf, e.g., Bourjaily, 483 U.S. at 175-76 , 107 S.Ct. at 2778-79 . In the case at bar, for example, a ruling that Reliant had earned the right to discover the communications in issue by establishing the fraud exception would leave open for resolution at trial the issue of whether Laser's patent was unenforceable as a result of fraud on the Patent Office. This would be true even if Reliant had to establish the same elements in each of these two settings — because the standard of proof that Reliant would be required to meet in order to establish unenforceability at the trial stage is more demanding (clear and convincing evidence) than the standard of proof for penetrating the privilege (no more than a preponderance of the evidence).

. In making this point I am not unmindful of the courts’ insistence that the standard for determining ultimate liability for the alleged wrong and the standard for establishing facts necessary to satisfy the rules of evidence are independent. For example, in Bourjaily , while discussing the interplay between and the requirements of Fed. R.Evid. 104(a) and 801(d)(2)(E), Chief Justice Rehnquist wrote the following for the majority:

483 U.S. at 175 , 107 S.Ct. at 2778-79 (citation omitted).

The test that I adopt in the text does not require judges who are making rulings under Fed.R.Evid. 104(a) to apply the standards that must be met in order to convict a defendant of a crime (beyond a reasonable doubt) or to impose liability for civil fraud in some jurisdictions (clear and convincing evidence). What drives my point in the text, above, is not deference to the appreciably more demanding burdens of proof for those other purposes, but concern about fundamental fairness and a desire to achieve an appropriate accommodation between the competing interests and values that motions like Reliant’s implicate.

39. It is not clear what the word "willful” adds to the requirements imposed by the words "knowing" and "intentional."

40. "Ablation” means surgical removal of tissue.

. Laser uses the term "necrosis" to refer to damage to tissue underlying the tissue that the surgeon wishes to remove.

. Laser’s applications for what eventually became the '502 patent were rejected by the Patent Office at least twice before the '502 patent was issued.

. At the hearing on its motion, Reliant played a videotape of a demonstration of one of the devices manufactured by Laser based on the '396 patent. This videotape was never introduced into the record as an exhibit, and no expert testimony was presented to explain it. Objecting to the court’s consideration of the videotape, Laser’s attorney stated that the tape could not show whether the surgical laser was actually turned on and off at the end of each row or column, because that laser was not visible to the eye. Reliant’s attorneys did not appear to dispute that statement. Thus, even if it had been admitted into evidence, the videotape could not support a finding that the devices based on the '396 patent were able to continuously irradiate tissue.

. Most of Samuel’s declaration consists of arguments responding to positions that Reliant advanced in its opening papers. Arguments like these belong in briefs, not in declarations. Declarations should be limited to statements of relevant facts and identification of attached exhibits.

. Our ruling about the crime/fraud exception would be no different even if it were clear that the lawyers who prosecuted the '502 patent on Laser’s behalf had had complete knowledge of all the materials to which Reliant points.