F. Anthony Mooney (argued), F. Anthony Mooney and Associates, Wellesley, MA, for Appellant.

Lisa D. Stern (argued), Miller, Turetsky, Rule, McLennan & Stern, Norristown, PA, for Appellee.

Before BECKER, ROTH, Circuit Judges, and SHADUR,1 District Judge.

SHADUR, District Judge.

Compass Technology, Inc. ("Compass") appeals the district court judgment, following a bench trial, that accepted the position of defendant Tseng Laboratories, Inc. ("Tseng") in this contract dispute. Compass contends that the district court erred (1) by admitting extrinsic evidence to interpret the contract between Compass and Tseng and (2) by refusing to reopen the evidence after a key witness had first been located within a few days after the close of the 1- 1/2 day bench trial.

Jurisdiction in the district court was invoked on diversity-of-citizenship grounds under 28 U.S.C. Sec. 1332 (originally-named codefendant Wang Laboratories, Inc. ("Wang") was dismissed by the district court for lack of jurisdiction). We have jurisdiction over this appeal from the district court's final judgment under 28 U.S.C. Sec. 1291.

We hold that under any view of the evidentiary issues the district court erred in refusing to hear the newly-located witness. And because that alone requires us to reverse the district court's judgment and remand for a new trial, we then address the related evidentiary issues as a guide to the handling of that second trial.

This dispute arises out of a "Manufacturer's Rep Agreement" (the "Agreement") entered into between Tseng and Compass effective February 19, 1988. Under the Agreement Compass was to serve as the exclusive selling representative for Tseng, a manufacturer of computer graphics chips, in six New England states. In return Compass was to receive a commission on the Tseng products sold by Compass within the six-state region.

Tseng's then Director of Sales and Marketing John Ciarlante ("Ciarlante") prepared the form of Manufacturer's Rep Agreement based on his experience with a previous employer (Tseng not having previously used such agreements). It took the form of a standard printed document, with blanks left to be filled in as appropriate. While the Agreement is quite straightforward in most respects, its Paragraphs 3 and 4 give rise to the present controversy:

3. PRODUCTS--The Representative shall sell the "products" of the manufacturer set forth herein which may be changed by the Manufacturer upon sixty (60) days prior notice, subject further to Addendum # 1, attached hereto:

4. AMOUNT OF COMPENSATION--Representative's compensation for services performed hereunder shall be 5%2 of the "net invoice price" defined herein below, of the Manufacturer's product for which an order is taken by Representative. However, when engineering, execution of the order, or shipment involve different territories the Manufacturer will split the full commission among the Representatives whose territories are involved. The Manufacturer will make this determination and advise the interested Representatives at the time the order is submitted to the Manufacturer.

No Addendum # 1 is in the record, and that has proved to be the focal point of the dispute between the parties. Even though Tseng was unable to produce a copy of any such addendum or to provide any witness who could testify to its claimed contents or could even recall seeing one, it nonetheless says that there was such an animal and that the addendum specified that Compass was to receive no commission whatever on any sales of Tseng products to Wang. For its part, Compass claims that no Addendum # 1 ever existed and that the Agreement is clear that Compass was to receive a 5% commission on all sales within its territory, including sales to Wang. What is at stake, if Compass is indeed entitled to such a commission on sales to Wang during the time that the Agreement was in effect, is an amount close to $200,000 exclusive of prejudgment interest.

Like most such catch phrases, the Chinese proverb that "One picture is worth ten thousand words" is obviously not intended to be taken literally as a universal rule. In this instance, however, the relevant picture is of the words themselves--the Agreement's pages showing its standardized form, the placement of the blanks and the filling in of the blanks (or perhaps more importantly, the failure to fill in the blanks)--and that picture is worth a good many words in the context of this case. We have therefore annexed a photocopy of the Agreement's most relevant and most illustrative page, its page 1.

Ciarlante had been Tseng's sole participant in negotiating and signing the Agreement on its behalf, while Compass was represented by its President Donald Rheault ("Rheault"). At trial Rheault testified that he could not recall whether or not there was an Addendum # 1 attached to the Agreement when he signed it. Because Ciarlante could not be located by either party before the trial, Tseng's only witness who spoke to the issue at all was John Gibbons, a founder, director and business consultant for Tseng, who testified that he had instructed Ciarlante to exclude sales to Wang from the Agreement. But Gibbons admitted on cross-examination that he did not actually see the Agreement until March 1989--more than a year after it was executed and delivered--and that he has never seen any Addendum # 1.

Early in the trial the district court determined as a matter of law that the Agreement's reference to the missing Addendum # 1 created an ambiguity (1994 WL 446853, at * 1). Over Compass' objections the district court then allowed testimony about the parties' intent as to what commission was to be paid on sales to Wang.

Although the district court's comments during the short bench trial had reflected a healthy skepticism as to whether there had ever been an Addendum # 1 (let alone what its terms were if it actually existed), the court ultimately reached the following conclusion (id. at * 2):

On the basis of the credible evidence presented by the parties the Court finds that Addendum # 1 provided that Compass would receive no commission on the sale of Tseng Products to house accounts and that, at the time the Agreement was executed, Tseng's only house account was Wang. Compass knew and was aware that no commission was to be paid on house accounts and that Wang was Tseng's most substantial account and its only house account. Compass knew and was aware of the fact that a commission would not be paid on the Wang account.

That holding was based, according to the district court, on three strands of evidence presented at trial (id. at * 2-3):

1. In addition to the stated awareness on Compass' part reflected in the last two quoted sentences, Compass was also aware that Tseng's previous manufacturer's representative had been dismissed for requesting commissions on the Wang account.3

2. Compass received two small commission statements for periods during which substantial sales were made to Wang, yet did not question Tseng or complain when those statements did not cover those sales to Wang.

3. In November 1988 (six months after the contract became effective) representatives of Compass and Tseng met to discuss a commission for servicing the Wang account and agreed that Compass would receive a 1% commission on sales to Wang after December 1, 1988.

Based on those findings the district court included in the total damages it awarded to Compass only 1% of the sales to Wang between December 1, 1988 and May 24, 1989 (that figure came to $18,402) plus prejudgment interest (id. at * 3).4

On August 16, 1994--after the trial had ended on August 11 but before the judge's August 15 decision was docketed and sent to the parties on August 18--Compass' efforts to locate Ciarlante, launched some months earlier, bore fruit. Its counsel immediately filed a Motion To Stay Issuance of Decision, or Vacate Decision, and Re-Open Evidence to Permit the Testimony of Previously Missing Witness.5 Then on August 22, after it had received the district court's decision, Compass filed a second motion--its Motion To Alter or Amend Judgment, or, Alternatively for New Trial, or Re-Opening of the Evidence--under Fed.R.Civ.P ("Rule") 59(a). Compass then supported that motion with an affidavit from Ciarlante in which he stated what his testimony, if it were credited at all, would be if the Court reopened the evidence and allowed him to testify.

Suffice it to say at this point that Ciarlante's testimony, if it were credited at all, would have been devastating to Tseng's position. Among other sworn assertions that supported Compass' position in its entirety, Ciarlante flatly controverted Gibbons' testimony as to any discussions having taken place about Wang sales being noncommissionable (or being commissionable at a reduced rate) and just as flatly negated the existence of the mysterious Addendum # 1:

29. It was in this context that the manufacturers representative agreement with Compass Technology, Inc., a copy of which is attached hereto as Exhibit C, was entered into in February of 1989.

30. I prepared that contract and I dealt exclusively with Mr. Rheault in connection with its execution.

31. There was never an Addendum # 1 attached to the contract, nor was it ever intended that there would be an Addendum # 1 attached to the contract.

32. It was never the intent on the part of either Tseng or Compass that commissions on sales to Wang would be excluded from the contract or would be treated as subject to a reduced commission.

33. One of Compass' responsibilities was to salvage the Wang account and develop the Wang business. Under those circumstances, it would have made no sense for Tseng to require that there be no commission paid on Wang sales through Compass Technology, Inc. or that there be a reduction of any commission earned on those sales.

34. There were never any discussions within Tseng Laboratories, Inc. that there would be no commissions paid to Compass with respect to Wang sales and I was not instructed to include such a provision, or a provision calling for reduced commissions, in any contract, addendum or other arrangement entered into with Compass Technology, Inc. on behalf of Wang.

35. I was never told that any manufacturers rep agreement I entered into on behalf of Tseng would have to be approved, initialed or countersigned by anyone, including John Gibbons or Jack Tseng.

36. The form of agreement used in connection with the Compass Technology, Inc. contract was one I used while employed at Princeton Graphics and that form of agreement provided for an Addendum # 1, which was occasionally used by Princeton Graphics to exclude certain products, not to list house accounts.

37. I have never seen an Addendum # 1 in connection with the Compass agreement or any other manufacturers rep agreement entered into during my tenure at Tseng.

38. I was present at the breakfast meeting which took place on November 2, 1988 at which John Gibbons and Donald Rheault were present.

39. I did not meet with Donald Rheault the night before that meeting to advise him that Compass Technology, Inc. would now be receiving a 1% commission on Wang sales, and I was not instructed by anyone at Tseng to have any such conversation with Mr. Rheault.

40. Had such an instruction been given to me, I would remember it, because it would have been directly contrary to the arrangement which Tseng had with Compass Technology, Inc. for the payment of a 5% commission on sales in Compass' New England territory, and it would have represented a drastic change in what had been agreed upon, namely, that Compass would receive a full commission on all sales in its territory.

41. There was no conversation at the breakfast meeting on November 2 in which Donald Rheault thanked John Gibbons for allowing Compass Technology, Inc. a 1% commission. Again, I would have a definite recollection of any such comment because it would have been inconsistent with what I knew to be the arrangement between Compass Technology, Inc. and Tseng, namely, that Compass was to receive a 5% commission on all sales in its territory, including sales to Wang. The breakfast meeting was to discuss protocol and the agenda of a meeting scheduled to take place at Wang immediately following breakfast.

On December 20, 1994 the district court denied both of Compass' motions, finding that Ciarlante could have been located earlier had Compass exercised "reasonable diligence," that Ciarlante's testimony about Addendum # 1 would be "merely cumulative" and that at any rate his testimony would not change the outcome (1994 WL 719616). That decision tied up the last loose ends before the district court and set the stage for this appeal.

We begin where the factual account has just ended--with the district court's rejection of Compass' motion, brought under this portion of Rule 59(a):

A new trial may be granted to all or any of the parties and on all or part of the issues ... in an action tried without a jury, for any of the reasons for which rehearings have heretofore been granted in suits in equity in the courts of the United States.

Although that provision does not in terms speak of such relief being based on new evidence, Rule 60(b)(2) provides that "newly discovered evidence which by due diligence could not have been discovered in time to move for a new trial" can give rise to relief from a judgment or order. Rule 59 and Rule 60(b)(2) share the same standard for granting relief on the basis of newly discovered evidence (11 Charles Wright, Arthur Miller and Mary Kay Kane, Federal Practice and Procedure: Civil 2d Sec. 2808, at 86 (2d ed. 1995)).

That standard requires that the new evidence (1) be material and not merely cumulative, (2) could not have been discovered before trial through the exercise of reasonable diligence and (3) would probably have changed the outcome of the trial (Bohus v. Beloff, 950 F.2d 919, 930 (3d Cir.1991)). Any party requesting such relief "bears a heavy burden" (id., quoting Plisco v. Union R. Co., 379 F.2d 15, 17 (3d Cir.1967)). Though we consequently review a district court's decision in that respect for an abuse of discretion (cf. Olefins Trading, Inc. v. Han Yang Chem Corp., 9 F.3d 282, 290 (3d Cir.1993), dealing with the other side of the coin--the erroneous admission of evidence), in this instance we hold that the district court did indeed abuse its discretion by refusing to reopen the evidence to allow Ciarlante to testify.

First, the district court held that Compass could have located Ciarlante had it exercised "reasonable diligence." But that determination gives insufficient credence to the affidavit of Compass' counsel F. Anthony Mooney ("Mooney") that accompanied its motion. Mooney there chronicled Compass' continuing efforts to locate Ciarlante, which began as soon as Mooney learned through discovery in the case about the vital position of Ciarlante as Tseng's only participant in the negotiations with Compass that resulted in the Agreement (as well as Ciarlante's having been Tseng's signatory to the Agreement).

Indeed, it is plain that both sides were searching for Ciarlante as a possible witness: When Tseng's counsel was asked by the district court about his whereabouts, she replied, "Heaven only knows. Neither party has found him." Based on Compass' counsel's pursuit of several possible avenues to trace Ciarlante's whereabouts, on counsel's ultimate success in locating him through the lawyer who represented Ciarlante in his lawsuit against Tseng (with whom Ciarlante had a falling out), and on counsel's immediate request to the district court to reopen proofs (before counsel knew that the judge had issued his written ruling a day earlier), it must be concluded that Compass made the appropriate showing of diligence.6

More critically for Rule 59 purposes, the district court's characterization of Ciarlante's proposed testimony as "merely cumulative" is unacceptable. That surprising label was ascribed to Ciarlante's affidavit because Rheault had testified that there was no Addendum # 1 that excluded products sold to Wang. But it is plainly different in both degree and kind to have such testimony emanating from the sole representative of Tseng in the negotiation and signing of the Agreement. Moreover, Ciarlante's affidavit expressly contradicts the testimony of Tseng's witness (Gibbons), who had said that such an exclusion of Wang sales had been intended, and other previously-quoted portions of the affidavit (its Paragraphs 38-41) are also wholly at odds with Gibbons' testimony as to the asserted post-contract negotiation of a lower commission rate on Wang sales.

We do not of course rule on which witness or witnesses are ultimately to be credited--that is a matter for the district court to decide on remand. But the point at this stage of the proceedings is that if Ciarlante's testimony were to be believed in any material respect, it could not fairly be viewed as "merely cumulative." One thing should be added in that regard, occasioned by the district court's comment (just after another use of the term "merely cumulative") that Tseng would challenge Ciarlante's credibility because he had been terminated by Tseng and had later sued it for wrongful discharge. Just as this court does not make credibility determinations as to witnesses whom we have not seen and heard (else we might review district court credibility findings de novo, rather than according them the respect that we do), so it would be improper for the district court to consider discrediting Ciarlante sight unseen.

Lastly in Rule 59(a) terms, the district court ruled that Ciarlante's testimony would not change the outcome of the trial. But that is closely linked to the point we have just made. At a minimum such a determination cannot be made without the prior--and impermissible--determination that Ciarlante would be disbelieved without ever seeing and hearing him testify. Indeed, even on that premise it cannot safely be said that the prior trial outcome should stand in any event--and that is so for other reasons to be discussed in the next section.

Our conclusion is additionally fortified by the posture of the case before the district court. What we have said would call for the granting even of a conventional Rule 59(a) (or Rule 60(b)(2)) motion for a new trial. But that result obtains a fortiori where what had taken place here was a short bench trial, with the district court facing only the need to reassemble counsel (and their clients, if they desired) to hear Ciarlante out through direct and cross-examination, and with no need to recall the other witnesses to testify anew (except perhaps to amplify their testimony in light of Ciarlante's)--let alone any need to reconvene with a new jury for a full-blown rerun, as a new trial most often requires.

In sum, a new trial must be ordered because of the district court's erroneous ruling on Compass' Rule 59(a) motion. That conclusion logically leads to a brief discussion of some of the evidentiary ground rules for that new trial, as called into question by this appeal.

Both before the district court and before us, the litigants have focused their principal fire on issues of parol evidence: whether under Pennsylvania law (which is specified by Agreement p 12 to provide the rules of decision, a designation that both parties to this diversity-of-citizenship action have honored) it is proper to resort to matters outside of the Agreement itself that assertedly bear on the intent of the contracting parties. Analysis demonstrates that another aspect of Pennsylvania law calls for those issues to be scrutinized in conjunction with considering what constitutes the appropriate level of proof on the evidentiary issues.

It is of course familiar and noncontroversial doctrine that the fundamental object in interpreting a contract is to ascertain the intent of the parties (Z & L Lumber Co. of Atlasburg v. Nordquist, 348 Pa.Super. 580, 585, 502 A.2d 697, 699 (1985)). And if their intent can be cleanly extracted from the clear and unambiguous words that the parties have used, it is equally conventional wisdom that they are held to those words contained in the contract (Mellon Bank, N.A. v. Aetna Business Credit, 619 F.2d 1001, 1013 (3d Cir.1980)). Those clear waters become murkier when an issue is raised as to a lack of clarity or a claimed ambiguity in the contractual language--in that event the court "should hear the evidence presented by both parties and then decide whether 'there is objective indicia that, from the linguistic reference point of the parties, the terms of the contract are susceptible of different meanings' " (Z & L Lumber, 348 Pa.Super. at 586, 502 A.2d at 700, quoting Mellon Bank, 619 F.2d at 1011). In doing so the court must consider the words of the contract, the alternative meaning proffered by the challenging party, and the nature of the evidence that party could provide (Mellon Bank, 619 F.2d at 1011). It is up to the party claiming that an ambiguity exists to show that a contract (Metzger v. Clifford Realty Corp., 327 Pa.Super. 377, 388, 476 A.2d 1 (1984)):

is reasonably or fairly susceptible of different constructions and is capable of being understood in more senses than one and is obscure in meaning through indefiniteness of expression or has a double meaning.

In this instance, Agreement p 11 contains the conventional integration clause prohibiting resort to other discussions and agreements between the parties:

GENERAL--This Agreement contains the entire understanding of the parties, shall supersede any other oral or written agreements, and shall be binding upon and inure to the benefit of the parties' successors and assigns. It may not be modified in any way without the written consent of both parties.

Nonetheless the Agreement itself poses an obvious question on its face: Does the reference in Paragraph 3 of the standard form Agreement to "Addendum # 1, attached hereto" match up with an actual document, or is it simply a part of that standard form that has no real significance (in somewhat the same way that, for example, a printed standard form of real estate contract may often contain a boilerplate reference "See also rider(s) attached" whether or not a particular contract has any riders at all)? And that basic question subsumes a whole set of such subsidiary questions as:

Is it likely that if an Addendum # 1--containing a special provision as to Tseng's single largest account (Wang)--had in fact existed and had been "attached," neither contracting party would have anywhere in its possession any Agreement other than the version that was wholly lacking in any Addendum # 1 or any other attachment?

Is it likely that if an Addendum # 1 had in fact existed, it would have contained a provision dealing, not with some exception to Tseng's "products" that were to be sold by Compass ("products" are, after all, the subject to which the reference to any such addendum in Agreement p 3 relates), but rather with a claimed exception to the totality of Tseng's customers to whom commissionable sales were to be made?

In the same respect, is it likely that if an Addendum # 1 had in fact existed, no further explanatory reference to the subject matter of that addendum would have been inserted in the space following the colon in Agreement # 3, rather than the parties having left that space totally blank?

As a variant on those last two questions, is it likely that if an Addendum # 1 had in fact existed, and had dealt with an exception to the flat 5% commission rate prescribed in Agreement p 4, no exception to that rate (or at least no cross-reference to Addendum # 1) would have been referred to in the Agreement's blank space that immediately followed that printed Agreement p 4?

There may be other questions that bear on the issue, and we do not suggest the answers to any of them (again that is something for the trier of fact)--but clearly all of those matters call for the district court to do more than to limit itself in terms of the sometime arcane questions of what separates "patent" from "latent" ambiguities (see Metzger, 327 Penn.Super. at 386, 476 A.2d at 5) or of what constitutes an "internal" as contrasted with an "external" ambiguity. Certainly the potential tyranny of such labels made it appropriate for the district court to have considered the testimony of Gibbons as to the alleged content of a meeting between the parties that took place well after the Agreement was entered into. And with that testimony before the trier of fact, it was just as plainly necessary for the district court to consider the testimony of Ciarlante that directly challenged Gibbons' statement as to the holding of any meeting having the content to which Gibbons testified.

But to return to the point of beginning, the central focus of the parties' dispute is clearly on the question whether or not Addendum # 1 truly existed--a question that bears not only on the meaning of the Agreement to begin with but also on whether the testimony of Gibbons or Ciarlante as to the claimed post-Agreement negotiation is to be credited. Because Tseng bears the burden of proving the existence of an ambiguity (under Metzger and like cases), and because it hangs its ambiguity argument on an addendum that the parties do not agree even exists and that Tseng cannot produce, this case bears a substantial resemblance to the "lost instrument" issue that arises from time to time in contract cases. And in that context the Pennsylvania courts have imposed a stringent standard of proof on the party that seeks to rely on a document that it cannot produce.

Thus the plaintiff in Hacker v. Price, 166 Pa.Super. 404, 407, 71 A.2d 851, 853 (1950) claimed that a written agreement gave him the right to purchase three shares of stock. Although the original Agreement was lost, plaintiff was allowed to introduce secondary evidence to show the content of the agreement and was ultimately successful in forcing specific performance. In affirming that decision, the Pennsylvania Superior Court set out the following requirements for recovering on a lost instrument (id., citations omitted):

To recover on an instrument, the original of which has been lost, the burden of proving the loss of the original and that a diligent, bona fide and thorough search was made without success is upon the one offering secondary evidence. He is also required to prove its former existence, execution, delivery and contents. The evidence to sustain these averments must be clear and convincing. Whether the party offering secondary evidence has met this burden of proof successfully is a matter to be determined by the trial Court and rests largely in the Court's discretion which will not be disturbed on appeal unless there is a manifest abuse thereof.

That "clear and convincing evidence" requirement echoes established Pennsylvania doctrine. See, e.g., Mahoney v. Collman, 293 Pa. 478, 482, 143 A. 186, 187 (1928) (imposing a "very heavy burden" on the party seeking to rely on a lost instrument, as well as announcing the "clear and convincing" evidentiary standard); In re Greggerson's Estate, 344 Pa. 498, 500-01, 25 A.2d 711, 713 (1942) (following Mahoney).

We recognize of course that those Pennsylvania cases have dealt with the "lost instrument" approach in a somewhat different context, applying it to plaintiffs who attempt to recover on such a missing document, rather than to a defendant who needs to prove such a document to satisfy its burden of establishing an ambiguity (in that respect only Haagen v. Patton, 193 Pa.Super. 186, 190, 164 A.2d 33, 34-35 (1960) is factually parallel to this case, and that opinion had no occasion to discuss the required standard of proof). Accordingly we do not opine on the level of proof to be applied by the district court on remand. Instead we leave it to that court to determine in the first instance whether the same considerations that have called for the higher "clear and convincing" standard of proof in the Pennsylvania "lost document" cases apply with like force to the situation involved in this case.

On remand, then, the district court is required to hear and consider the testimony of Ciarlante as to all of the matters dealt with in his affidavit. And in the district court's reconsideration of all of the evidence in light of that testimony, it should give consideration to whether Tseng's burden of proving:

1. that Addendum # 1 existed;

2. that the claimed addendum could not be found after a bona fide search; and

3. that Addendum # 1 excluded sales to Wang from the across-the-board 5% commission provision set out in Agreement p 4;

should be scrutinized through a more demanding "clear and convincing evidence" lens. We reverse and remand the case for proceedings consistent with this opinion.