Recently, the Supreme Court sent Dennison Mfg. v. Panduit Corp. back to the Court of Appeals for the Federal Circuit (CAFC).[1] It remanded with explicit directions that the lower court consider the extent to which Fed. R. Civ. P. 52(a)[2] governs appellate review of nonobviousness determinations. Law and Fact in Patent Litigation:
Thomas G. Field, Jr.
Form versus Function*Introduction
On remand, the CAFC should attempt to relate the issue to the scope of review for other issues that arise in patent appeals. Neither the narrow nor the broad problem has ever received the attention it deserves -- particularly from the standpoint of the fundamental law/fact dichotomy.
It is by no means certain that nonobviousness determinations should be treated as questions of law. Nevertheless, there is ample evidence that courts seek to review these findings more intensely than would be appropriate for questions of fact under "clearly erroneous" or "substantial evidence" standards. This paper argues that, if the courts are inclined to persist in more intense review of nonobviousness, two other matters need to be considered: First, whether more liberal review should be extended to all questions concerning patent validity, and, second, whether such review should be conducted under a "constitutional fact" doctrine. The former would address apparent inconsistencies in the current law, and the latter would allow appellate courts an expanded role (function) without unduly confusing terminology (form).
The Law/Fact Dichotomy The most directly relevant precedent governing appellate review of patent litigation is cryptic. Quoted in the per curiam decision that remands Dennison,[3] It consists of one sentence from the Court's decision in Graham v. John Deere Co.:[4] "While the ultimate question of patent validity is one of law,... the § 103 condition... lends itself to several basic factual inquiries."However, what that sentence means is unclear; Graham indicated neither reasons nor authority for denominating questions of patent validity, in general, or nonobviousness, specifically, as ones of "law" (as contrasted with "fact"). As shown below, without an analysis of those matters, it is very difficult to determine either whether validity issues other than nonobviousness should be treated as ones of "law" or, in any case, what sort of treatment should be afforded. It is hoped that, following the remand of Dennison, some progress will be made toward resolving those important problems.
While much has been written on the law/fact dichotomy, a brief overview is in order.[5] The terms, "law" and "fact" are used several ways,[6] and it is necessary to define them in the present context. Approached that way, the term, "fact," is more precisely called "adjudicative fact,"[7] and the most accurate definition is operational. In short, a question of adjudicative fact is one of the kind traditionally regarded as appropriate for juries. Conversely, a question of "law" is anything that remains -- including questions of "legislative" fact, the latter being used, even by courts, in conjunction with values, to evolve general rules.[8]
Unfortunately, the water is muddied when courts occasionally resolve fact issues as a "matter of law."[9] Yet, even then, the operational definition stands: If a matter is sufficiently well established in the record that reasonable minds could not differ, it is within the ultimate control of judges, not a jury. Hence, in sorting out "law" and "fact" issues, roles of judge and jury are paramount.
Indeed, this is inherent in rights to jury trials; without the distinction, such rights would be hollow,[10] and appellate judges are as bound as trial judges. While appellate courts may be the ultimate arbiters of issues of law, issues of fact properly entrusted to juries are entitled to exceptional deference under the "substantial evidence" standard.[11]
For somewhat different reasons, deference to fact finding is also extended, as in Dennison, to bench trials. There "clearly erroneous" review applies.[12] That standard is less deferential, but the Supreme Court demonstrated in Dennison, and more so in Inwood Laboratories[13] more than token deference is required.
Until 1985, the depth of review sometimes varied depending on whether cases turned on demeanor[14] or documentary evidence. The latter had sometimes been accorded less deference,[15] but in explaining the change made that year, the Advisory Committee said:
The principal argument advanced in favor of a more searching appellate review of findings... based solely on documentary evidence is that the rationale of Rule 52(a) does not apply when the[y]... rest on... an evaluation of documentary proof and the drawing of inferences from it, thus eliminating the need for any special deference.... These considerations are outweighed by the public interest in the stability and judicial economy that would be promoted by recognizing that the trial court, not the appellate tribunal, should be the finder of the facts. To permit courts of appeals to share more actively in the fact-finding function would tend to undermine the legitimacy of the district courts in the eyes of litigants, multiply appeals by encouraging appellate retrial of some factual issues, and needlessly reallocate judicial authority."Constitutional Facts" Despite that, and regardless of whether facts were initially before a jury, an argument can be made that courts should give above-average scrutiny to cases where constitutional interests are at stake. Indeed, the Supreme Court held exactly that in Bose Corp v. Consumers Union U.S.[16] While the decision concerned freedom of speech and product disparagement, not patents, a modest extension would have substantial impact on the latter.In Bose, it was held that appellate courts are obligated to exercise independent judgment in determining when one should be liable to another for misstatements of fact (in this situation, as contrasted with "opinion")[17] concerning the other's products.[18] One commentator has argued that the Court's reasoning makes it difficult to restrict that rule to first amendment cases.[19] He also urges that deep review of constitutional facts is compelling when they have been found by administrative agencies.[20]
Consider whether findings that deal with patent validity do not fit in either instance. At least when no new art is introduced in the litigation, a challenge to validity involves not only issues of constitutional interest (Art. I § 8 cl. 8) but also a collateral challenge to findings of the Patent and Trademark Office. It is doubtful that the Supreme Court had occasion to consider the implications tof his. Yet, it may well have had Bose (perhaps unconsciously) in mind when it remanded Dennison.
If the CAFC, or ultimately the Supreme Court, believes that extraordinary review of facts in patent cases is warranted, consideration should be given to whether the "constitutional fact" doctrine might be usefully extended. Patent attorneys are fond of referring to the constitutional underpinnings of the patent system. Yet rarely will a case turn on them,[21] and it is doubtful that any Supreme Court validity decision ever did.
Even an 1850 case often regarded as the genesis of the nonobviousness requirement was not based on the Constitution.[22] On the contrary, notwithstanding a dissent claiming otherwise,[23] Justice Nelson did not admit to adding a new requirement -- much less one mandated by the Constitution. Indeed, he insisted that mere novelty, even coupled with market success, had long been inadequate to support patent validity.[24]
The first Supreme Court allusion to constitutional interests being involved in standards of patentability does not seem to have occurred until the 1950[25] -- almost exactly a century later. Moreover, it arose in the context of the Court's addressing the scope of appellate review in patent cases. There, the Court was skeptical of the merits of a patent found valid by two lower courts. While it had a rule against re-examining findings consistently made by two lower courts, the Court nevertheless found the patent invalid. The majority maintained, however, that it was reversing because the wrong legal standard had been applied.[26]
The most noteworthy aspect of that case, for present purposes, is a concurring opinion by Justice Douglas.[27] There, it was argued that the majority would have been justified in reversing, notwithstanding the two-court rule. He reasoned that validity is a question of law because "The standard of patentability is a constitutional standard; and the question of validity is a question of law."[28] For authority he cited Mahn v.Harwood,[29] an 1884 case in which the Court reviewed a decision of the Patent Commissioner. Yet, Mahn merely points out that, while the Commissioner might be entitled to some deference on questions of fact, courts need not defer on questions of law[30] -- hardly an earth-shaking proposition.[31]
Had Graham cited Justice Douglas' opinion in A&P, it would have been helpful in understanding what was meant by calling validity a question of law.[32] Nevertheless, the explicit constitutional origins of the patent system, perhaps coupled with the fact that validity may involve collateral review of administrative action, could serve as a basis for de novo review of questions of fact. Whether the constitutional fact doctrine or anything else should be used to that end, however, is another question. Before attending to it, a closer examination of nonobviousness may be useful.
Is Nonobviousness a Special Case? Whatever the CAFC decides about the appropriate standard of review for nonobviousness, it ought also to address why that standard should differ from that applied to other validity requirements. While nonobviousness was the primary concern in Grahamand companion cases, that decision calls validity a question of law. If the CAFC chooses to single out nonobviousness for special treatment, it should justify it.[33]It might do so based on the origins of the nonobviousness requirement. Yet, even if nonobviousness had been found in Hotchkiss[34] to be constitutionally mandated, an implicit requirement would hardly be more important than the utility requirement, which is about as explicit as the Constitution gets.[35] Thus, even if the "constitutional fact" doctrine were used to justify more appellate review, nonobviousness would not stand out.
The CAFC might also attempt to distinguish nonobviousness on the basis of its technical difficulty and the tendency of fact finders to see inventions as obvious through hindsight.[36] Nonobviousness is clearly the most technically difficult issue in patent litigation. Still, neither that nor the tendency to Monday-morning quarterback should any more justify extra-ordinary appellate review than it would justify withholding nonobviousness (or, for that matter, medical malpractice) from the jury altogether.[37]
The Bottom Line The CAFC has full control of issues of patent law, subject only to legislative and Supreme Court oversight. When reviewing infringement litigation, whether there is a jury or not, it has the capacity, for example, to correct for the application of erroneous legal standards, to expand the application of principles, or to carve out exceptions to rules.[38]It can also correct clear errors of fact. Its power is even greater where no jury is involved.[39] Yet, regardless of the standard of review, it cannot second-guess the resolution of disputes that turn on the relative credibility of witnesses. If a case turns on credibility, it is difficult to justify reversal.[40]
If closer review of certain issues of fact seems justified notwithstanding, e.g., problems with demeanor evidence, perhaps on the basis of (often correct) judicial intuition, it behooves the CAFC to confront the issue squarely. It could turn out, on closer inspection, that such review is unwarranted or, if warranted, falls within a recognized exception. It is possible, too, that a new theory may need to be evolved to explain a new exception. The worst that could happen would be to fail to deal with the issue head on. That would leave trial courts and the bar guessing. It would also seem to encourage unduly the inefficient practice of relitigating issues in a setting which would not have been appropriate for dealing with them in the first instance.[41]
Should review more intensive than permitted under Rule 52(a) be somehow justified for fact disputes unlerlying one or more validity requirements, calling them issues of "law" changes nothing and merely confuses matters. Hence, the "constitutional fact" doctrine deserves close attention. It would highlight the need for intense appellate review without concealing the essential nature of the issue; in short, it would serve the needs of both form and function.[42]
Regardless of the outcome, everyone should appreciate an honest attempt to grapple with an issue which has been in limbo far too long. In Dennison, the CAFC has an opportunity to shape the very foundations of patent litigation. It is to be hoped that the court will rise to the occasion.[43]
Notes * Based on 27 IDEA 153 (1987).[1] 106 S.Ct. 1578.
[2]Rule 52(a) provides that "[f]indings of fact, whether based on oral or documentary evidence, shall not be set aside unless clearly erroneous, and due regard shall be given to the opportunity of the trial court to judge the credibility of the witnesses." (Emphasised language added in 1985.)
[3] Id., at 1579.
[4] 383 U.S. l, 17-18 (1966).
[5] The earliest paper on the narrow topic seems to be Note, Nonobviousness in Patent Law: A Question of Law or Fact? 18 WM.&M.L.Rev. 612 (1977); in it the author cites several more comprehensive articles. More recently, 12(4) AIPLA Q.J. (1985) was devoted entirely to the topic, both broadly and narrowly.
[6] E.g., "law" versus "equity" or "fact" versus "opinion."
[7] See F.R.Evid. 201, advisory committee note.
[8] Indeed, the distinction rests on the same operational definition: questions of adjudicative fact are the kind traditionally entrusted to juries whereas legislative facts are not. The closest one can come to avoiding circularity is to look at whether a fact is critical only to the outcome of the specific dispute or goes to establishing the rights and duties of the classes of which the parties are merely members. See generally, e.g., B. Schwartz, Administrative Law, 213-16 (2d Ed. 1984).
[9] See Railroad Dynamics v. A. Stucki Co., 727 F.2d 1506 (CAFC 1984); Judge Markey, at 1513, appears to use the terms in the alternative. Compare their more traditional use, e.g., in Garst v. General Motors Co., 484 P.2d 47 (Kan. 1971); at 63, both the majority and dissent are of interest in that regard
[10] Thus, the Seventh Amendment commands that "no fact tried by a jury, shall be otherwise re-examined... than according to the common law."
Whether one is entitled to a jury trial in patent litigation, by virtue of the Constitution as contrasted with the statute, is another matter altogether. See e.g., NLRB v. Jones & Laughlin Steel Corp., 301 U.S. 1, 48-49, holding that one is not entitled to a jury trial in an action to award back pay under the National Labor Relations Act. See also, Constantino and Master, The Seventh Amendment Right to a Jury Trial in Complex Civil Litigation..., 12 AIPLA L.Q. 279 (1985).
[11] See, e.g., Coolley, Patent Jury Issues . . ., 67 J.P.O.S. 3, 4-6 (1985).
[12] See supra note 2.
[13] Inwood Laboratories v. Ives Laboratories, 456 U.S. 844, 855-56 (1982).
[14] Demeanor evidence is specifically flagged in the Rule. See supra note 2.
Indeed, the logic of deferring to decisions based on such evidence extends to situations where review would otherwise be de novo; see, e.g., First Fed. Savings & Loan Ass'n. v. Fed. Home Loan Bank Bd., 426 F.Supp. 454, 475 (W.D. Ark. 1974), aff'd 570 F.2d 693 (8th Cir. 1978).
[15] See, e.g., Fur Information Fashion Council v. Timme, 501 F.2d 1048, 1050-51 (2d Cir. 1974).
[16]466 U.S. 485, 500 (1984), rehear den., 467 U.S. 1267.
[17] 466 U.S., at 490; see particularly n.4.
[18] The question of whether the New York Times standard ought to be applied to a product was not before the court; see 466 U.S., at 492
[19] Monaghan, Constitutional Fact Review, 85 Colum.L.Rev. 229, 238 (1985).
[20] Id., at 276.
[21] See e.g., Sears, Roebuck & Co. v. Stiffel, 376 U.S. 225, 228-31 (1964). See also Kewanee Oil Co. v. Bicron Corp., 416 U.S. 470, 479-93 (1974). Yet neither case is based on the patent clause; both rest on the supremacy clause. Indeed, in Sears, at 231, "limited times" for patents is discussed as reflecting Congressional policy!
[22] Hotchkiss v. Greenwood, 52 U.S. 248.
[23] 52 U.S., at 267; there, Justice Woodbury argued that the proper test was whether the invention was "new, and better and cheaper" not whether "an ordinary mechanic could have devised it."
[24] He even discussed, without citing, an earlier case in which he had been involved. As Justice Nelson describes it, that case was, indeed, very similar; in it, a patent on a button using a wooden rather than a more expensive bone foundation had been found invalid.
[25] Great Atlantic and Pacific Tea Co. v. Supermarket Equipment Corp., 340 U.S 147.
[26] Id., at 153-54.
[27] Id., at 154.
[28] Id., at 155; note omitted.
[29] 112 U.S. 354, 358.
[30] Id.
[31] However, one will occasionally confront circumstances where there is deference to administrative decisions on both facts and law. See, e.g., Pittston Stevedoring Corp. v. Dellaventura, 554 F.2d 35, 49-50 (2d Cir. 1976); therein Judge Friendly discusses and attempts to rationalize two conflicting lines of authority. See also, Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984).
[32] Supra note 4.
[33] For a discussion of which issues have been called "law" and which "fact," see Hofer, The CAFC and Fact/Law Questions in Patent Cases..., 12 AIPLA L.Q. 295 (1985). Indeed, everything except issues of obviousness and claim construction appear to be treated as questions of fact.
[34] Supra notes 22-24 and discussion.
[35] The reference is to the word, "useful" in the term, "useful arts." Yet, even Brenner v. Manson was not decided on this basis; see 383 U.S. 519, 529, 535 (1966).
[36] See, e.g., the CAFC decision in Dennison, 744 F.2d 1082, 1091-93 (1985).
[37] However, the answer to the question is by no means clear. See, e.g., Constantino and Master, supra note 10.
[38] See generally, L. Carter, Reason in Law (2d Ed. 1984).
[39] Bose, ssupra note 14.
[40] See, e.g., the 1st Circuit decision in Bose, 692 F.2d 189,195. See also supranote 13. Nevertheless, credibility appears to have played a minor role in Dennison; see 774 F.2d, at 1090, n. 14.
[41] Again, this more than anything else seems to underlie Rule 52(a). See also, Preemption Devices v. Minn. Mining and Mfg. Co., 732 F.2d 903 (CAFC, 1984).
[42] However, it would seem to open the record as much to Supreme Court as to CAFC review. This may well have been what Justice Douglas had in mind in A&P; see supra notes 27-29 and discussion.
See also, Sakraida v. Ag Pro, Inc., 425 U.S. 273, 280 (1976) ("Our independent examination of that evidence persuades us of its sufficiency to support the District Court's finding…."). That and the following comparison of the invention and the prior art seems to signal the Court's inclination to review facts closely indeed.
[43] It would be fairly easy to evade the issue if the court were so inclined. Not only did credibility play a minor role in the bench trial (supra note 40), but the CAFC could also ascribe reversal to purely legal errors. See, e.g., 774 F.2d, at 1097, where it was held that the trial court had given too little weight to the presumption of validity.
Modified Sept. 4, 2004. tgf
© Franklin Pierce Law Center
Franklin Pierce Law Center
Two White Street • Concord, NH 03301 • 603.228.1541
© 2002-2007 Franklin Pierce Law Center. All rights reserved.
