In this case the trial court has applied F.R.Civ.P. 56 to enter a summary judgment holding that Patent No. 4,283,455 ('455 patent) is unenforceable for inequitable conduct by the patent attorney or agent who applied for it. The court, however, denied summary judgment, as also sought on grounds of obviousness and noninfringement, but its decision on these issues may have been influenced by its interpretation of the patent respecting the dispositive issue. Burlington Industries, Inc., assignee of the patent, appeals and there is no cross-appeal as to the issues on which summary judgment was denied. The case is somewhat unusual in that the alleged misinformation supplied the examiner related to the invention itself, a supposed discrepancy being discovered, as to which no triable issue of fact existed, between the invention itself as it really was, and the invention as described in the application, an amendment thereto, and the patent as issued. Had the invention been truly described, the claims it is said would have read on the prior art. We vacate and remand because, reading the evidence as we must in the manner most favorable to the party opposing the motion, we do not perceive any showing of intent to deceive, such as our cases require, sufficient to eliminate any triable issue of fact. The evidence on the motion is, in our view, consistent with the innocent use of ambiguous language, such as perhaps a super lawyer would never be guilty of, but which all of us, being what we are, lapse into now and then.

The inventor, James N. McGee, an employee of the assignee Burlington, conceived and reduced to practice an improved V-belt cover fabric. He ran a bias-cut fabric through a thin solution of rubber polymer and conductive carbon black. He said he "impregnated the individual fibers in the yarn bundle." Disclosure of April 23, 1979. The patent attorney who prosecuted the patent application says* he understood the invention covered by the patent was to "impregnate the fiber bundles and encapsulate the individual fibers," and it never occurred to him the patent he applied for would or could be read as claiming an ability to impregnate the individual fibers. To a lay reader it might appear, as it certainly did to the trial judge, that the two phrases mean something different, but the attorney says he and others unconsciously inserted the word "bundles" after the word "fiber." The patent itself, summary, specifications, and claims, use both phrases with a fine impartiality. Moreover, no one asserts there was more than one invention: thus the correct full description of it must be one or the other and cannot be both, taken literally. The importance of the issue derives from the trial court's belief that by claiming impregnation of individual fibers, the patent attorney avoided a prior art rejection which he at first confronted and which caused a Patent and Trademark Office (PTO) rejection of the first application.

The continuous fabric strip 12 formed during slitting is then treated with a fluid mix of elastomeric polymeric material to impregnate in the yarn bundles thereof and to encapsulate the individual fibers of such bundles with the polymeric material. [Col. 2, line 61.]

The numeral 12 refers to Fig. 1 of the patent which shows means to coat the belt by running it through a shallow dish containing the fluid mix 13. Here the inventor carefully avoids saying he impregnates the fibers. Yet he goes on:

The fluid mix 13 according to the present invention preferably has a low viscosity so that impregnation of the fibers is insured. [Col. 3, line 13.]

The purpose being to provide a solvent mix of suitable viscosity permitting effective yarn bundle impregnation to coat the individual fibers. [Col. 3, line 43.]

(* * * the present invention deliberately avoids dyeing in favor of coloring only the fiber surface by pigmenting.) Since each of the fibers of the present cover fabric is encapsulated with the impregnating polymeric material * * *. [Col. 5, line 19.]

Thus, the writer continuously shifts his phraseology while apparently under the impression he is saying the same thing. A like phenomenon occurs in the patent claims:

Claim 1 adopts once more the phraseology of the more careful formulation:

* * * [t]he individual fibers of said yarns being encapsulated in a dried pigmented elastomeric polymer, * * *.

Claim 2 is dependent on Claim 1. Yet claim 3 declares:

[t]he fabric cover being produced from a continuous fabric strip by impregnating the individual fibers.

Claims 4 and 5 again impregnate "the individual fibers." There follow some dependent claims, and Claims 12 and 13 again impregnate the fabric, while 16 impregnates the individual fibers once more.

In the original application, with 20 claims, a similar confusion is found--we say confusion if one assumes the inventor had made but one invention and that he knew, as he says, of no way a mix such as he described would impregnate the individual fibers if they were man made; also, if the mix did impregnate the individual fibers, the asserted advantage of white showing through, revealing the extent of wear, when the outer encapsulation wore off, would be lost. [Col. 1, line 62 to Col. 2, line 14.] The confusion could not possibly have been the product of any rejection by the examiner as it obtained in the file papers from the beginning.

Mr. Robert A. Vanderhye, the patent attorney who prosecuted the application, asserted in an affidavit that he used the shorter formulation, as illustrated in Claim 3 as "shorthand" for a "complicated concept." He "unconsciously" inserted the word "bundles" after the word "fiber." We do not have any explanation by the examiner of his mental attitude, but it is difficult to believe he did not share in the patent attorney's mistake. He must be charged with notice of the actual contents of the patent, as issued with his approval, and it is difficult to believe he would not have noticed the confusion, and demanded a clarification, if he had ever consciously adverted to there being a difference in meaning between impregnating an individual fiber and impregnating a bundle. The trial court, after the controversy had arisen, and with the "20-20 vision of hindsight," read the patent entirely differently than did those concerned with its issue, and we wonder how anybody skilled in patent law and practice could have generated such confusion without a sinister purpose. That is the fascinating puzzle this case presents.

We start with the proposition that to sustain a motion for summary judgment the court must determine that there is no triable issue of material fact, because on the version most favorable to the party opposing the motion, that party still cannot prevail. Howes v. Medical Components, Inc., 814 F.2d 638, 643, 2 USPQ2d 1271, 1273 (Fed.Cir.1987); KangaROOS U.S.A., Inc. v. Caldor, Inc., 778 F.2d 1571, 1573-74, 228 U.S.P.Q. (BNA) 32, 33 (Fed.Cir.1985). In reviewing a trial court's grant of summary judgment, an appellate court is not bound by the "clearly erroneous" rule, because proper summary judgment is not based on debatable findings of fact. Cable Electric Products, Inc. v. Genmark, Inc., 770 F.2d 1015, 1020, 226 U.S.P.Q. (BNA) 881, 885 (Fed.Cir.1985); Lemelson v. TRW, Inc., 760 F.2d 1254, 1260, 225 U.S.P.Q. (BNA) 697, 700 (Fed.Cir.1985).

By 37 C.F.R. Sec. 1.56(d), the PTO will now not grant a patent when fraud on the office was attempted, or the duty of disclosure was violated through bad faith or gross negligence. This court will not hold unenforceable a patent once granted in the absence of an intent to mislead, although the nondisclosure of facts of which the applicant should have known the materiality may justify an inference of intent to mislead in appropriate cases. FMC Corp. v. Manitowoc Co., 835 F.2d 1411, 1415, 5 USPQ2d 1112, 1115 (Fed.Cir.1987). Construction of disputed claims requires reference to the specifications, the prosecution history, and the other claims. Fonar Corp. v. Johnson & Johnson, 821 F.2d 627, 631, 3 USPQ2d 1109, 1112 (Fed.Cir.1987), cert. denied, --- U.S. ----, 108 S. Ct. 751, 98 L. Ed. 2d 764 (1988). Expert testimony as to the understanding of those skilled in the art may be necessary. Moeller v. Ionetics, Inc., 794 F.2d 653, 656, 229 U.S.P.Q. (BNA) 992, 994 (Fed.Cir.1986). In a classic statement, it is said the inventor may be his own lexicographer. Autogiro Company of America v. United States, 384 F.2d 391, 397, 155 U.S.P.Q. (BNA) 697, 702 (Ct.Cl.1967). But he must use his words consistently in the claims and in the specifications. Fonar Corp., 821 F.2d at 632, 3 USPQ2d at 1113.

Here the trial court's theory required it to suppose that the examiner rejected the original application because it read on two prior art patents, Brown, Patent No. 2,518,220 ('220) and Foti, Patent No. 3,962,511 ('511). In order to refute that, McGee's counsel, Vanderhye, persuaded the examiner of a falsehood, i.e., that the McGee invention involved the impregnation of individual fibers, while "in truth and in fact" it involved impregnation of fiber bundles or yarn and only encapsulated the individual fibers.

If the examiner really ever believed the McGee invention involved the impregnation of individual fibers, he ought to have been given pause by some of the claim language. He might have been thereby disabused of his false notion, or he might have thought to look at the specifications. There he would have found much to refute the idea that McGee impregnated the individual fibers, and much to support the notion that he, or his attorney, was using language in a manner peculiar to himself, i.e., was being his own lexicographer. It is hard to read the specifications and not conclude that the writer of them, his own lexicographer, perceived no clear difference between impregnating a fiber and impregnating a bundle. He used both sets of words, referring to the same diagram, Fig. 1, as depicting both. Rather than anyone being a crook, it is surely a permissible inference that both the applicant's attorney and the examiner were alike confused, not being possessed of the analytical vision of hindsight. This may be a mistaken inference, but on summary judgment it cannot be rejected out of hand in favor of a less plausible sinister interpretation. If Mr. Vanderhye really obtained the patent by a willful false practice upon the Patent Office, or negligence so gross as to imply bad faith, it will require not summary judgment to establish the fact, but a careful and thorough examination into the understanding of Mr. McGee, his counsel, and the examiner, so far as permissible, what the words chosen meant to each, and what they would have meant to persons skilled in the art.

We are not, therefore, necessarily endorsing Mr. Vanderhye's explanation of his own conduct. He says he indulged in "shorthand." Shorthand is a method of transcribing words by which the skilled stenographer can keep up with most persons' spoken words, but at a cost of having a script, readable only by himself, whereas the slower conventional method of arabic script is supposed to be readable by anyone not illiterate. In a patent, language readable only by the author is inappropriate. Besides arabic script, we all learn in school a penchant for "elegant variation," i.e., a reluctance to repeat even a single word more than once in a paragraph. If San Francisco is named once, on the next reference it becomes "the Pacific Coast port above mentioned," or even more elegantly, "the City on the Golden Gate." This is how we learn to write. Mr. Vanderhye naturally found it tiresome to repeat so many times what he calls "complicated concepts" and so preferred what he calls "shorthand." A patent, like any other legal document, is likely to have its intentions defeated by "elegant variation," which should be reserved for less mundane documents. Those admissions by Mr. Vanderhye may confess errors, but they do not sink to the level of inequitable conduct. While not compelling, they do represent the version of the facts most favorable to the nonmoving party, and thus should have precluded the summary judgment if consistent with other established facts, as here they are.

The trial court denied summary judgment on obviousness and infringement issues, and this part of its decision is not before us for review. However, our remand on the inequitable conduct issue may lead the trial court to change its opinions as to what the patent means, which might, in turn, bear on obviousness and infringement. The trial court will be able to reconsider its conclusions on all the issues it previously passed upon.

We add one final word: the habit of charging inequitable conduct in almost every major patent case has become an absolute plague. Reputable lawyers seem to feel compelled to make the charge against other reputable lawyers on the slenderest grounds, to represent their client's interests adequately, perhaps. They get anywhere with the accusation in but a small percentage of the cases, but such charges are not inconsequential on that account. They destroy the respect for one another's integrity, for being fellow members of an honorable profession, that used to make the bar a valuable help to the courts in making a sound disposition of their cases, and to sustain the good name of the bar itself. A patent litigant should be made to feel, therefore, that an unsupported charge of "inequitable conduct in the Patent Office" is a negative contribution to the rightful administration of justice. The charge was formerly known as "fraud on the Patent Office," a more pejorative term, but the change of name does not make the thing itself smell any sweeter. Even after complete testimony the court should find inequitable conduct only if shown by clear and convincing evidence. A summary judgment that a reputable attorney has been guilty of inequitable conduct, over his denials, ought to be, and can properly be, rare indeed. We cannot improve on the discussion of this in KangaROOS, Inc., supra, and incorporate what is there said by reference herein.

The summary judgment entered in this case is vacated, and the cause is remanded for further proceedings consistent with this opinion.

Costs to be borne by the appellee.

VACATED AND REMANDED.