FLYING J INC., a Utah corporation;
CFJ PROPERTIES, a Utah
partnership; TON SERVICES, a Utah
corporation; TFJ, a Utah partnership;
NCR; TCH, a Utah corporation,
No. 03-4262
Plaintiffs-Appellees,

COMDATA NETWORK, INC., a
Maryland corporation,

Defendant-Appellant.

APPEAL FROM THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF UTAH
(D.C. No. 1:96-CV-66-DAK)

Gregory J. Kerwin, Gibson, Dunn & Crutcher LLP, Denver, Colorado (F. Joseph
Warin and Melanie L. Katsur, Gibson, Dunn & Crutcher, LLP, Washington, D.C.,
and Casey K. McGarvey, Berman, Tomsic & Savage, Salt Lake City, Utah, with
him on the brief), for Plaintiffs-Appellees Flying J, Inc., et al.

J. Gordon Cooney, Jr., Morgan, Lewis & Bockius LLP, Philadelphia,
Pennsylvania (Gary F. Bendinger and John H. Bogart, Bendinger, Crockett,
Peterson & Casey, Salt Lake City, Utah, with him on the briefs), for Defendant-
Appellant Comdata Network, Inc.
Before McCONNELL and McKAY , Circuit Judges, and FRIOT , District Judge. *

McCONNELL , Circuit Judge.

This appeal is the latest episode in a lengthy antitrust suit brought by

Plaintiffs Flying J, Inc., TCH, L.L.C., CFJ Properties, TON Services, Inc., and

CFJ (collectively “Flying J”) against Defendant Comdata Network, Inc.

(“Comdata”). The underlying lawsuit, filed in 1996, arose out of Comdata’s

alleged monopolization of two product markets related to the U.S. truck stop

industry. In May 2001, after nearly five years of discovery and pretrial motions,

the parties entered into a settlement agreement. Comdata agreed to pay $49

million in damages and grant Flying J two licenses intended to open the markets

at issue. Conflict soon resumed, however, when the parties could not agree about

the meaning of one of the two licenses. Aggrieved by what it interpreted as

Comdata’s refusal to honor one of the licenses, Flying J filed a motion to enforce

the settlement agreement in May 2002. The district court determined that

Comdata had breached the settlement agreement and ordered Comdata to

implement the license as interpreted by Flying J. Exercising jurisdiction under 28

U.S.C. § 1292(a), we REVERSE. In so doing, we emphasize that it is not our task

*
The Honorable Stephen P. Friot, of the United States District Court for
the Western District of Oklahoma, sitting by designation.

-2-
to determine what would best remedy the underlying antitrust violation, but solely

to interpret the agreement reached between the parties, in light of its plain

language and the intent of the parties.

Plaintiff Flying J, Inc. owns and operates a nationwide chain of

approximately 160 truck stops. Flying J estimates that there are approximately

4000 truck stops in the United States, roughly 1000 of which offer services

comparable to its own truck stops. This appeal involves two markets closely

related to the trucking industry: point-of-sale systems and trucker fuel cards.

Comdata provides transaction-processing services to the trucking industry.

Among other things, it sells the Trendar System, a PC-based point-of-sale system

that allows merchants to process customers’ card transactions. When the

customer’s payment card is swiped through the Trendar card reader, Trendar

software sends the transaction data to a third-party financial institution for

authorization and settlement. Flying J presented evidence in the underlying

litigation that Comdata had secured approximately 90% of the point-of-sale

systems market by 2001. Virtually every major U.S. truck stop other than Flying J

used the Trendar System.

Comdata also issues fuel cards to trucking companies, including a

proprietary payment card (“the Comdata Card”) and a Comdata MasterCard,

-3-
referred to as the Comdata Fleet Card. According to Flying J, Comdata cards

have dominated the trucker fuel card market for many years, securing a 70% share

by 2001.

Proprietary payment cards offer two features relevant to this litigation. The

first, called data capture, requires the truck driver to enter certain data at the time

of the fuel purchase, typically including driver identification and odometer

readings. The data is relayed in real time to the trucking company, allowing it to

monitor the driver’s location and activities. The second feature, called purchase

control, allows trucking companies to restrict the type and quantity of items that

the driver can purchase with the fuel card. A trucking company might, for

example, allow its drivers to purchase motor oil and a certain amount of diesel

fuel each day but prevent them from purchasing alcoholic beverages, a truck

stereo, or enormous bags of pork rinds. Each trucking company can tailor

purchase controls to suit its preferences. According to Flying J, data capture and

purchase controls are essential to effective management of a long-haul trucking

fleet, and they provide the principal incentive for trucking companies to give

trucker fuel cards to their drivers rather than ordinary credit cards. Ordinary

MasterCards do not have these features.

Comdata and Flying J (through its subsidiaries, TAB and TCH) issue

proprietary cards with no logo other than their own, as well as MasterCards. The

-4-
Flying J propriety card is called the TCH Fuel Card; its MasterCard is called the

TCH Mastercard. Comdata MasterCards are called the Comdata MasterCard or

Comdata Fleet Card, which mean the same thing. The Comdata and TCH

MasterCards can be used either as ordinary MasterCards or as proprietary cards,

depending on the arrangements reached with the particular merchant. These are

called “dual function” cards.

As a condition of participating in the MasterCard network, MasterCard

requires merchants to accept all MasterCards, regardless of the issuer. It is up to

the particular merchant to decide whether to accept proprietary cards. In the

discretion of the merchant, MasterCard regulations permit use of a “dual

function” MasterCard as a proprietary card for a “proprietary account,” defined as

“an account . . . that is separate from a MasterCard account and is maintained by a

company or organization other than the MasterCard member.” MasterCard

Bylaws and Rules, October 2002, Rule 6.6, JA 546. The proprietary account rules

require a preexisting account, established according to a separate agreement, with

each merchant.

Comdata MasterCards and TCH MasterCards therefore must be accepted by

all merchants in the MasterCard network. But only if the merchants have an

agreement with the issuer do the merchants treat these cards as proprietary cards

(meaning that they effectuate the data capture and purchase control features of

-5-
these cards). If the particular merchant is part of the MasterCard network but has

not made any agreement with Comdata or THC, Trendar must process the

transaction through the MasterCard network as an ordinary MasterCard

transaction, without the data capture or purchase control features. If the merchant

has agreed to accept the issuer’s proprietary card, the merchant processes the

issuer’s MasterCard transaction as a proprietary card transaction, meaning that the

transaction will be authorized and settled over a private network as opposed to the

MasterCard network.

In an ordinary MasterCard transaction, Trendar sends transaction data to

the merchant’s chosen third-party financial institution, known as the acquirer.

The acquirer, through its own system, sends the transaction data through the

MasterCard network for authorization by the card issuer’s bank and settlement

through the MasterCard network. When a dual function card, such as the TCH

MasterCard or the Comdata Fleet Card, is processed as an ordinary MasterCard, it

does not provide trucking companies with real-time data capture or purchase

control because the MasterCard network does not relay data quickly enough.

When Flying J entered the trucker fuel card market in the mid-1990s, it

faced two significant barriers to entry. First, the Trendar System did not accept

the TCH Fuel Card. Merchants using Trendar could not process TCH Fuel Card

transactions; therefore, Flying J customers could not use the TCH Fuel Card at

-6-
Trendar locations. Given the ubiquity of the Trendar System, this was a big

problem for Flying J. 1 Second, the record shows that Comdata engaged in a

campaign to pressure truck stops not to accept the TCH Fuel Card. Prior to the

2001 settlement, Comdata placed over four hundred telephone calls to truck stops

and threatened to raise transaction fees on Comdata cards if they agreed to accept

the TCH Fuel Card.

In 1996, Flying J filed suit, charging Comdata with violation of the

Sherman Act, 15 U.S.C. §§ 1–7, through monopolization of the trucker fuel card

market and the point-of-sale systems market. The parties settled in 2001.

Comdata agreed to pay $49 million in damages and grant Flying J two licenses:

the Comdata License and the Trendar License. In accordance with the parties’

request, the district court reserved jurisdiction to enforce the terms of the

settlement agreement. See Kokkonen v. Guardian Life Ins. Co., 511 U.S. 375

(1994). This appeal concerns only the Trendar License as it pertains to

processing the TCH MasterCard at unaffiliated merchants; that is, merchants who

have not agreed to accept the proprietary TCH Fuel Card.

1
The Federal Trade Commission has found that trucker fuel cards “exhibit
strong network effects” so that “[d]emand for a fleet card rises with the number
of truck stops that accept the card, which in turn depends on the number of fuel
purchase automation systems that accept the card.” The FTC’s analysis identifies
access to the point-of-sale system as the primary factor in the demand for a fuel
card. Prior to the settlement, this factor inhibited demand for TCH cards because
Trendar did not accept them.

-7-
The Trendar License grants to TCH a “nonexclusive license . . . to access

and use the Trendar System solely for the Permitted Uses.” Trendar License Art.

4.1 Permitted uses include “access and use of the Trendar System by TCH and its

cardholder customers.” Id. Art. 1. In short, the Trendar License provides TCH

access to the Trendar System, and it enables TCH to effect data capture and

purchase controls in transactions involving the TCH MasterCard. Pursuant to the

Trendar License, Comdata and Flying J configured the Trendar System to accept

the TCH Fuel Card. They also configured Trendar to process TCH MasterCards

as proprietary transactions if the merchant had agreed to accept the TCH Fuel

Card.

This cooperative endeavor soon reached an impasse when Comdata refused

to configure Trendar to process TCH MasterCards as proprietary transactions at

retail outlets where the merchants had not agreed to accept the TCH Fuel Card.

The dispute centers on Article 4.2 of the Trendar License, which provides:

All TCH Card Transactions processed through the Trendar
System, including without limitation TCH Cards bearing a
MasterCard or Visa brand, shall be cleared directly through TCH as
opposed to any third party network, such as and without limitation
the MasterCard network or Visa network, to the fullest extent
permitted by the policies, rules, or regulations, as amended from time
to time (including their interpretations thereof) by third party
network providers.

Trendar License Art. 4.2. Many Flying J competitors, such as Pilot, Petro, and

Travel Centers of America, do not accept the TCH Fuel Card. Because truck

-8-
stops operated by these competitors accept MasterCard, they must accept the TCH

MasterCard, but they do so as an ordinary MasterCard transaction, not as a

proprietary transaction. Thus, if Trendar processes the TCH MasterCard through

the MasterCard network at these locations, the transaction does not provide the

trucking company with data capture and purchase control. This restriction on

TCH MasterCard processing hinders Flying J’s ability to compete in the trucker

fuel card market. Flying J argues that in order for the TCH MasterCard to be

fully effective, and therefore attractive to trucking companies, all TCH

MasterCard transactions must provide data capture and purchase control.

As discussed more fully below, Comdata interprets the Trendar License as

requiring Comdata to configure the Trendar System to treat TCH cards the same

way that it treats its own. Accordingly, Comdata was willing to route TCH

MasterCard transactions directly through TCH only if TCH secured MasterCard’s

approval. MasterCard, in turn, refused to agree to financial settlement through

private networks except where the merchant entered into a separate agreement

with the issuer. Although most merchants have entered into such agreements with

Comdata, many have declined to enter such agreements with TCH. According to

Comdata, the reason for this discrepancy is that Flying J is a competitor and

Comdata (which does not operate truck stops) is not.

In an effort to provide data capture and purchase control in all TCH

-9-
MasterCard transactions, Flying J filed a motion to enforce the settlement

agreement. Flying J argues that at the time of the settlement, Trendar processed

the Comdata Fleet Card as a proprietary card; therefore, the Trendar License

obligated Comdata to provide universal private processing for TCH MasterCard

transactions, regardless of whether the merchant accepts the TCH Fuel Card.

In support of its motion to enforce, Flying J presented two models for TCH

MasterCard transactions. The primary model replicated the proprietary

processing method used for Comdata MasterCards, which processed the entire

transaction over a private network. In the alternative, Flying J proposed a split-

transaction or dual-processing model, in which TCH would authorize the

transaction over a private network, providing data capture and purchase control,

but financial settlement would take place on the MasterCard network. Under the

dual-processing model, Trendar would send data to TCH and MasterCard in two

different steps, and financial settlement would occur in the same manner as in an

ordinary MasterCard transaction.

Flying J’s primary model did not conform to MasterCard regulations for

proprietary accounts. Joan Hennessey, MasterCard’s designated corporate

representative, confirmed that Flying J’s primary model depicted a proprietary

transaction subject to MasterCard’s proprietary account rules, which required

prior agreement by the merchant. Addressing the dual-processing model, Ms.

-10-
Hennessey testified that it was not prohibited by MasterCard rules; rather,

transmission of authorization data to TCH was “outside of the MasterCard

transaction and so it would be outside of our rules.” Deposition of Joan

Hennessey, JA 1898.

After the Hennessey deposition, Comdata moved for summary judgment,

arguing that (1) Flying J failed to secure MasterCard’s approval for the primary

proposal, and (2) the split-transaction model did not satisfy the requirements of

the Trendar License, specifically the requirement that clearing occur “directly

through TCH as opposed to . . . the MasterCard network.” The court continued

the hearing to permit Flying J to ensure compliance with MasterCard procedures.

Shortly before the evidentiary hearing resumed, MasterCard confirmed in writing

that Flying J’s primary model did not meet its proprietary account rules. The dual-

processing model fell outside of MasterCard rules. MasterCard noted, however,

that it could not require that a merchant participate in the dual-processing

arrangement, nor could it require a merchant to provide transaction data to Flying

J. MasterCard suggested that the merchant and its acquirer authorize its

participation in the dual-processing arrangement.

Because MasterCard rejected its original proposal, Flying J relied exclusively on

the dual-processing model at the evidentiary hearing.

The district court held that Comdata had breached its obligations under the

-11-
Trendar License, including Article 4.2, by failing to implement Flying J’s dual-

processing model for TCH MasterCard transactions. Order 27, JA 109 (Comdata

breached its obligations “by failing promptly to implement a method for direct

processing of TCH MasterCard transactions directly through TCH that is

acceptable to TCH and does not violate MasterCard rules.”). The court found the

dual-processing method to be consistent with Comdata’s own method of

processing the Comdata Fleet Card, rejecting Comdata’s argument that proprietary

processing of the TCH MasterCard required merchant consent. In its conclusions

of law, the court held that Article 4.2’s requirement that all transactions be

“cleared directly through TCH” included authorization and approval in addition to

financial settlement, thereby permitting dual processing. The court granted Flying

J’s motion to enforce the settlement agreement and ordered Comdata to

implement the dual-processing method for TCH MasterCards at all Trendar

locations. Comdata appeals.

II.

We review the district court’s determination of state law de novo. See

Salve Regina College v. Russell, 499 U.S. 225, 238–39 (1991); Grant v.

Pharmacia & Upjohn Co., 314 F.3d 488, 491 (10th Cir. 2002). We review the

district court’s findings of fact for clear error. “A finding is clearly erroneous

-12-
when, although there is evidence to support it, the reviewing court on the entire

evidence is left with the definite and firm conviction that a mistake has been

committed.” United States v. United States Gypsum Co., 333 U.S. 364, 394–395

(1948) (internal quotation marks omitted). “Whether the district court failed to

consider or accord proper weight or significance to relevant evidence are

questions of law we review de novo.” Harvey ex rel. Blankenbaker v. United

Transport Union, 878 F.2d 1235, 1244 (10th Cir. 1989) (citing Pullman-Standard

v. Swint, 456 U.S. 273, 291, 292 (1982)).

As a preliminary matter, Comdata complains that the district court adopted

Flying J’s proposed findings of fact and conclusions of law almost verbatim.

Regrettably, this appears to be the case. 2 The Supreme Court has criticized the

verbatim adoption of findings of fact, “particularly when those findings have

taken the form of conclusory statements unsupported by citation to the record.”

Anderson v. City of Bessemer City, 470 U.S. 564, 572 (1985). The court’s

wholesale adoption of one party’s proposed findings of fact and conclusions of

law provides little aid on appellate review, see Everaard v. Hartford Accident &

2
Comdata attached to its opening brief a computer-generated comparison
of the district court’s order and Flying J’s proposed findings of fact and
conclusions of law. This document shows that, aside from minor, formal
alterations (e.g., changing “I find” to “the court finds”), the court’s order is an
exact copy of Flying J’s proposed findings of facts and conclusions of law. See
Aplt. Br. Exh. B.

-13-
Indem. Co., 842 F.2d 1186, 1193 (10th Cir. 1986), particularly in the likely event

that the adopted submission takes an adversarial stance. 3

The district court’s adoption of a party’s proposed findings does not change

the standard of review, however. Though not made by the district judge himself,

the findings “are formally his; they are not to be rejected out-of-hand, and they

will stand if supported by evidence.” United States v. El Paso Nat. Gas Co., 376

U.S. 651, 656 (1964) (citing United States v. Crescent Amusement Co., 323 U.S.

173, 184–85 (1944)); see also Sanpete Water Conservancy Dist. v. Carbon Water

Conservancy Dist., 226 F.3d 1170, 1177 n.7 (10th Cir. 2000) (“[E]ven if we

believed the district court improperly adopted [the] proposed findings without

reasoned consideration, we would still review the district court’s decision under

the clearly erroneous standard.”). We therefore review the court’s findings of fact

for clear error.

For its part, Flying J suggests that the district court’s familiarity with this

3
Judge J. Skelly Wright warned district judges that

lawyers, and properly so, in their zeal and advocacy and their
enthusiasm are going to state the case for their side in these findings
as strongly as they possibly can. When these findings get to the
courts of appeals they won’t be worth the paper they are written on
as far as assisting the court of appeals in determining why the judge
decided the case.

Seminars for Newly Appointed United States District Judges 166 (1963), quoted
in El Paso Natural Gas, 376 U.S. 651, 656 n.4 (1964).

-14-
case entitles its order to greater deference on appeal. See, e.g., Stone v. City and

County of San Francisco, 968 F.2d 850, 856 (9th Cir. 1992); Ferrell v. Pierce,

743 F.2d 454, 461 (7th Cir. 1984). We reject this suggestion as well. It is true

that the district court hosted this litigation for almost five years, and we have no

doubt that it became excruciatingly well-acquainted with the long-distance

trucking industry and the details of Flying J’s antitrust claim. The record shows,

however, that the question central to this appeal—whether Article 4.2 was

intended to authorize a dual-processing model for TCH MasterCard

transactions—did not arise until shortly before the evidentiary hearing, well after

the parties settled their claims in the underlying suit. Even assuming that the

district court’s familiarity could increase the already substantial deference owed

to findings of fact under the clear error standard, the rationale for such deference

would not apply to the issue before us.

This case requires us to reconcile two competing interpretations of the

mandate that all TCH Card transactions “shall be cleared directly through TCH as

opposed to any third party network . . . to the fullest extent permitted by . . . third

party network providers.” Trendar License Article 4.2. Comdata contends that

the district court erred by failing to determine whether the contract was

ambiguous, by refusing to give the term “cleared” its prevailing meaning in the

-15-
financial industry, and by disregarding Comdata’s evidence of intent. Comdata

claims further that the district court’s ultimate interpretation of the contract was

clearly erroneous. Flying J responds that Comdata’s interpretation of the contract

is unreasonable and, in the alternative, that the court’s factual findings are

supported by the evidence. We hold that the court interpreted the term “cleared”

under an incorrect legal standard, that differences in meaning of the term render

Article 4.2 ambiguous, that accordingly the court should have considered

Comdata’s proffered evidence, and that in light of that evidence the court’s

factual findings regarding the meaning and intent of the contract are clearly

erroneous.

III.

We begin with an argument about whether the district court found the

contract ambiguous. Comdata contends that the district court failed to make this

threshold determination. Flying J responds that the court explicitly found

Comdata’s interpretation of the contract unreasonable, and that it follows that, if

one of the two possible interpretations is unreasonable, Article 4.2 is not

ambiguous. Fortunately, de novo review provides a narrow escape from this

tangle of compound ambiguity. Because Flying J and Comdata each advance

plausible interpretations of Article 4.2, we find the disputed language to be

ambiguous.

-16-
A.

The general rules of contract interpretation under state law apply to

settlement agreements. See Sackler v. Savin, 897 P.2d 1217, 1220 (Utah 1995). 4

When a court interprets a contract, it must make a threshold determination

whether the contract is ambiguous. See Peterson v. Sunrider Corp., 48 P.3d 918,

925 (Utah 2002). A contract is ambiguous if it supports more than one reasonable

interpretation. Id; see also Nielsen v. Gold’s Gym, 78 P.3d 600, 601 (Utah 2003).

Utah law is unsettled on the issue whether the court may go beyond the four

corners of the contract to determine whether the contract is ambiguous. Compare

WebBank v. Am. Gen. Annuity Serv. Corp., 54 P.3d 1139, 1145 (Utah 2002) (“If

the language within the four corners of the contract is unambiguous, the parties’

intentions are determined from the plain meaning of the contractual language, and

the contract may be interpreted as a matter of law.”), and Central Fla.

4
The Trendar License calls for the application of Tennessee law. The
Settlement Agreement calls for the application of Utah law. Both parties agree
with the district court’s determination that there is no material conflict between
Tennessee and Utah contract law. Because the parties do not raise a choice-of-
law issue, we construe and interpret the Trendar License under Utah law. See In
re Korean Airlines Disaster, 932 F.2d 1475, 1495 (D.C. Cir. 1991) (“Unlike
jurisdictional issues, courts need not address choice of law issues sua sponte.”),
cited in GBJ Corp. v. E. Ohio Paving Co., 139 F.3d 1080, 1085 (6th Cir. 1998)
(declining to undertake a full most-significant-contacts inquiry where the parties
acquiesced in the application of New York law); Railway Express Agency, Inc. v.
Super Scale Models, Ltd., 934 F.2d 135, 139 (7th Cir. 1991) (“Where the law of
the two states is essentially the same, we apply the law of the forum state.”).

-17-
Investments, Inc. v. ParkWest Associates, 40 P.3d 599, 605 (Utah 2002) (“We first

look to the four corners of the agreement to determine the intentions of the

parties.”), with Nielsen, 78 P.3d at 601 (in determining ambiguity, the court may

consider “[r]elevant, extrinsic evidence of the facts known to the parties at the

time they entered the [contract])”(brackets in original) (quoting Yeargin, Inc. v.

Auditing Div. of Utah State Tax Comm'n, 20 P.3d 287 (Utah 2001)), Peterson, 48

P.3d at 918 (“In determining whether a contract is ambiguous the court is not

bound to consider only the language of the contract.”), and Ward v. Intermountain

Farmers’ Ass’n, 907 P.2d 264, 268 (Utah 1995) (“When determining whether a

contract is ambiguous, any relevant evidence must be considered.”). Because

both parties follow the expansive view, we assume that the court may look beyond

the four corners of the contract to determine ambiguity.

Ambiguity is a question of law. Id. If the contract is ambiguous, the court

can admit extrinsic evidence of intent to clarify the meaning of ambiguous terms.

Id. If the court determines the contract to be ambiguous and admits extrinsic

evidence, the court’s interpretation of a disputed term is a finding of fact

reviewed for clear error. Nielsen, 78 P.3d at 601 (citing Kimball v. Campbell,

699 P.2d 714, 716 (Utah 1985)).

Flying J contends that consideration of extrinsic evidence transforms the

threshold determination of ambiguity into a finding of fact. That is not an

-18-
accurate statement of Utah contract law. The cases cited by Flying J involve the

admission of extrinsic evidence to determine the meaning of ambiguous terms, a

question of fact. See Valley Improvement Ass’n, Inc. v. United States Fidelity &

Guaranty Corp., 129 F.3d 1108, 1115 (10th Cir. 1997); Cavic v. Pioneer Astro

Indus., Inc., 825 F.2d 1421, 1424 (10th Cir. 1987) (“[W]hen the trial court resorts

to extrinsic testimony to ascertain the meaning of the contractual terms, the

interpretation is factual.”). In its threshold determination of ambiguity, however,

the court decides whether there is anything for the trier of fact to find. Even if

the court refers to extrinsic evidence to make that determination, contractual

ambiguity presents a question of law that we review de novo. See Peterson, 48

P.3d at 925. We now explain our conclusion that Article 4.2 is ambiguous.

Comdata maintains that the language of Article 4.2 does not permit dual or

split transactions through Trendar, but requires that financial settlement,

authorization, and approval occur over the same network. It argues that the

technical meaning of the term “cleared” for purposes of financial transactions,

and in particular as reflected in MasterCard glossary, refers exclusively to

financial settlement. The MasterCard glossary defines “clearing” as “[t]he

process of exchanging financial transaction details between an acquirer and an

issuer to facilitate posting of a cardholder’s account and reconciliation of a

-19-
customer’s settlement position.” MasterCard, Bankcard Glossary, JA 1151.

Relying on this definition, Comdata argues that split transactions violate the

requirement that all transactions be cleared directly through TCH as opposed to a

third-party network. Comdata thus interprets “to the extent permitted” to mean in

as many transactions as permitted by the third-party network provider.

Flying J urges reliance on a more general definition of “clear” found in the

Merriam-Webster online dictionary: “5a: to submit for approval b: AUTHORIZE, APPROVE : as (1): to

certify as trustworthy (2): to permit

(an aircraft) to proceed usually with a specified action.” Merriam-Webster Online

Dictionary, at http://www.m-w.com/cgi-bin/dictionary, cited in Order 22, JA104

(“These meanings include the functions of authorizing and approving, in addition

to financial settlement.”). If clearing is defined in this way, Trendar can transfer

authorization and approval data “directly through TCH,” while financial

settlement takes place in the MasterCard network. Thus, even if a merchant has

not agreed to accept the TCH Card, Trendar could send authorization and

approval data through a private network, thereby permitting data capture and

purchase control on all TCH MasterCard transactions. Thus, Flying J reads “to

the extent permitted” to mean as much of the transaction as permitted by

MasterCard.

-20-
The language of Article 4.2 will bear either party’s interpretation. Comdata

interprets the provision consistently with its own arrangement with MasterCard.

Though Flying J’s interpretation is not based on any pre-existing processing

model, it too reflects a plausible reading of Article 4.2, it does not violate any

specific MasterCard regulation, and it is technically feasible. 5

The district court agreed with Flying J. The court considered itself bound

to “construe contract language in a way that gives effect to the ordinary meaning

of words unless the evidence indicates the parties intended to adopt a special

meaning.” Order 22, JA 104. Finding no evidence demonstrating that the parties

intended to define “clear” according to its technical meaning in financial

transactions, as illustrated by the MasterCard glossary, the court considered itself

bound by the “plain meaning.” The district court looked to the dictionary as

evidence of plain meaning, concluding that “clear” “include[s] the function of

authorizing and approving, in addition to financial settlement.” Id. The court

reasoned that if clearing involves multiple functions, then Article 4.2 permits

different aspects of clearing to occur over different networks. The court rejected

Comdata’s argument that Article 4.2 did not authorize split transactions as “an

5
In a letter submitted after briefing, see Fed. R. App. P. 28(j), Comdata
informs the Court that “in the early-to-mid 1990’s, Trendar was configured to
process a narrow category of non-MasterCard transactions in two steps.”
Comdata denies, however, that Trendar has ever split MasterCard transactions.

-21-
unreasonable interpretation of the agreement’s language that is inconsistent with

the parties’ intentions for the Trendar License.” Order 24, JA 106.

We do not agree with the district court’s application of the plain meaning

rule. Under Utah law, if “the contract is unambiguous, the parties’ intentions are

determined from the plain meaning of the contractual language, and the contract

may be interpreted as a matter of law.” Green R. Canal Co. v. Thayn, 84 P.3d

1134, 1141 (Utah 2003); see also Grynberg v. Questar Pipeline Co., 70 P.3d 1, 10

(Utah 2003) (“This court interprets unambiguous contracts as a matter of law.”).

Where there are two plausible meanings—one from ordinary speech and one from

the specialized terminology of the relevant industry—the contract is not

unambiguous and the plain meaning rule does not apply. See R&R Energies v.

Mother Earth Indus., 936 P.2d 1068, 1074 (Utah 1997) (“A specialized meaning

would create ambiguity because at least two plausible meanings, one ‘plain’ and

one ‘specialized,’ would exist.”). Giving “cleared” its ordinary dictionary

meaning therefore requires a preliminary determination that Article 4.2 is

unambiguous.

Even if the plain meaning rule controlled, the definition adopted by the

district court is less than plain in the context of Article 4.2. The Restatement of

Contracts provides that unless the parties manifest a different intention,

“technical terms and words of art are given their technical meaning when used in

-22-
a transaction within their technical field.” Restatement (Second) of Contracts §

201(3)(b) 6; see also Holland v. Brown, 394 P.2d 77, 78–79 (Utah 1964) (“When

terms used in a contract appear to have a specialized meaning, they must be

understood in accordance with the particular connotation they may have acquired

in such transactions.”). Considering that this lawsuit focuses specifically on

MasterCard transactions, MasterCard’s own definition might reasonably be

consulted to determine the plain meaning of the term in its immediate context. At

the very least, Comdata’s reliance on the MasterCard glossary should have

insulated its interpretation from the charge of unreasonableness.

Flying J does not dispute that the verb “clear” has a specialized meaning in

the financial industry; rather, it contends that such usage would be unreasonable

in the context of an antitrust settlement. This is an unduly narrow

characterization of the issue. It is true that this dispute arises out of an antitrust

claim, but the antitrust claim involved financial transactions in the trucker fuel

card market. Both parties litigate, at least in part, in their capacity as providers of

6
The Utah Supreme Court has not specifically adopted Restatement
(Second) of Contracts § 201; however, it has consistently adopted other
Restatement provisions. See, e.g., Andreini v. Hultgren, 860 P.2d 916, 921 (Utah
1993) (adopting the standards of duress provided in Restatement Second §§ 175,
176); Southeastern Equip. Co. v. Mauss, 696 P.2d 1187, 1188 (Utah 1985) (noting
that Utah recognizes promissory estoppel as defined by § 90); Bitzes v. Sunset
Oaks, Inc., 649 P.2d 66, 68–70 (Utah 1982) (explaining that Utah followed the
doctrine of frustration of purpose set forth in § 265).

-23-
financial services. The fact that this is an antitrust case therefore does not render

irrelevant the specialized meaning of “cleared” in the financial industry.

Interpreting “cleared” in light of its industry-specific meaning does not,

however, entirely resolve the ambiguity in Article 4.2. Flying J argues that the

Trendar License sanctions dual processing even if “cleared” refers exclusively to

financial settlement. Substituting “financial settlement” for “cleared,” Article 4.2

requires that all transactions “be [financially settled] directly through TCH as

opposed to any third party network . . . to the fullest extent permitted . . . by third

party network providers.” If MasterCard does not permit financial settlement

over the TCH network, the transaction can be settled over the MasterCard

network, but nothing in MasterCard’s rules, and therefore nothing in Article 4.2,

prevents the transfer of other data over the TCH network. Flying J argues, in

other words, that if “cleared” refers only to financial settlement, Article 4.2 does

not address dual processing; if it does not address it, it does not prohibit it; and if

it does not prohibit it, the Trendar License permits dual processing.

We turn, then, to extrinsic evidence of the intended meaning of Article 4.2.

The district court found that Flying J intended the Trendar License to allow

TCH cards “effectively to compete with Comdata’s proprietary trucker fuel card,

-24-
including the Comdata MasterCard Fleet Card.” Order 5, JA 87. The court found

that Comdata failed to produce contradictory evidence:

Comdata did not present evidence at the hearing indicating that the
parties did not intend to accomplish this result through the Trendar
License. Mr. Sheridan participated in negotiation of the Trendar
License but did not offer testimony specifically contradicting that
explanation of the parties’ intent.

Order 5–6, JA 87–88. Based on what it perceived as a lopsided evidentiary

record, the court concluded that

enforcing the Trendar License in a way that requires TCH
MasterCard Fleet Card transactions to be cleared directly through
TCH, as opposed to the MasterCard I-Net network, at truck stops that
do not currently accept the TCH proprietary card as a method of
payment . . . is consistent with the reasonable expectations of the
parties at the time they entered into their Settlement Agreement and
the Trendar License in May 2001.

Order 6, JA 88. The court rejected Comdata’s position, which it characterized as

“an unreasonable interpretation of the agreement’s language that is inconsistent

with the parties’ intentions for the Trendar License.” Order 24, JA 106. We do

not agree with this characterization of the record.

In our review of the record, Comdata produced substantial evidence

indicating that it did not contemplate or intend that the Trendar License would

lead to proprietary processing of TCH MasterCards at unaffiliated merchants,

much less that this would be accomplished through a dual-processing model.

Comdata’s chief counsel, Michael Sheridan, testified that Comdata intended to

-25-
replicate its own processing model for Flying J: “I thought, Your Honor, that we

were allowing them, if MasterCard approved, to send the entire transaction to

them and have a proprietary network just like we have.” Sheridan Testimony, JA

1632. He specifically disavowed any intention to authorize split transactions:

“Section 4.2 was not intended to, and does not, require Comdata to reconfigure

the Trendar system to route part of the transaction through TCH and clear other

parts of the transaction through the MasterCard network.” Sheridan Declaration,

JA 1141–42. See also id. at 1141 (“At no time did I contemplate that Comdata

would be required to configure Trendar to route part of a transaction through TCH

and another part of the transaction through the MasterCard network.”). This

constitutes substantial evidence that the parties did not intend to implement the

dual processing model, which did not exist at the time the settlement was

negotiated and is different from the model used for Comdata’s own proprietary

cards. The district court’s failure to consider Comdata’s evidence was a legal

error. See Blankenberger, 878 F.2d at 1244.

The district court found that Flying J’s interpretation of the contract would

promote the pro-competitive goals of the Sherman Act by creating a network of

truck stops where TCH cards provided data capture and purchase control. In the

district court’s view, this “pro-competitive intent” encompassed data capture and

-26-
purchase control for all TCH MasterCard transactions at all U.S. truck stops using

the Trendar System. It thus concluded that Comdata was bound to implement

dual-processing unless it could show that the parties did not intend to restore

competition to the trucker fuel card market.

We have no occasion to question the district court’s conclusion that Flying

J’s interpretation of Article 4.2 would better promote the competitive objectives

of the antitrust laws. But the issue at this stage is not what would be the most

effectual remedy for a proven antitrust violation; the issue is what the parties

agreed to in settling the litigation. The parties did not frame their settlement in

terms of maximizing competition, but in terms of a specific procedure for

processing credit card transactions. The record is unequivocal on one point: both

Flying J and Comdata intended the Trendar License to provide Flying J with data

capture and purchase control similar to the so-called “Comdata model.” JA

1119–20, 1690 (Adams); JA 1632 (Sheridan); Appellee’s Br. 33; Appellant’s Br.

34. Flying J makes no attempt to dispute this. Appellee’s Br. 33. Comdata’s

own cards were, and are, processed as single transactions. The most

straightforward interpretation of Article 4.2, therefore, is that TCH MasterCard

transactions would be processed in the same way.

At least at the time of the evidentiary hearing, Comdata and Flying J had

different ideas of what the Comdata model entailed. Comdata thought that it

-27-
referred to a particular procedure: access to the Trendar System, plus data capture

and purchase control at affiliated merchants. The main goal of the license, in

Comdata’s view, was to provide TCH with access to the Trendar System:

Comdata . . . understood that the Plaintiff’s goal in litigation and
settlement was to have the TCH card accepted over the Trendar
System and cleared directly through TCH, if permitted by
MasterCard. Indeed, the very title to the operative agreement in this
matter is: “Trendar License (TCH Card Access to Trendar).”

Sheridan Declaration, JA 1134. Comdata did not view the license as intended

“solely to provide TCH with full data capture, regardless of how transactions

were to be cleared or regardless of whether TCH obtained MasterCard’s consent.”

Id. Neither did Comdata anticipate that the Trendar License would require

Trendar to split transactions.

Flying J argues that the Comdata model was associated solely with a result:

data capture and purchase control at all Trendar locations that accepted

MasterCard. Flying J apparently believed that this result was “comparable to the

direct processing of the Comdata MasterCard Fleet Card.” Order 17, JA 99.

Flying J’s understanding depended on the assumption that Comdata obtained data

capture and purchase control in Fleet Card transactions without merchant consent.

See Declaration of J. Phillip Adams para. 18, JA1118 (“My understanding both in

May 2001 and now, is that Comdata did not then, and does not now, have specific

agreements with truck stops authorizing such private processing of Comdata

-28-
MasterCard Fleet Card transactions.”). Flying J defends this assumption by

reference to the testimony of Comdata’s General Counsel, who indicated that

Comdata’s agreements with truck stops do not specifically notify the merchant

that Trendar processes the Comdata Fleet Card over Comdata’s private network.

Based on this testimony, the district court found that

Comdata does not have written agreements in place with the major
U.S. truck stops that accept its card (including Pilot, Petro, and
Travel Centers of America) by which Comdata explicitly informs
those truck stops that it will be privately processing Comdata
MasterCard Fleet Card transactions outside the MasterCard I-Net
network or explicitly obtains their permission for such private
processing. Instead, the truck stop merchant agreements on which
Comdata relies to argue that it has contracts for private card
processing in place with truck stops, are the standard form ‘service
center’ agreements by which truck stops agree to accept Comdata
products including Comdata’s proprietary (non-MasterCard) card.

Order 18, JA100. The district court found “no evidence that Comdata has such

written authorizations in place for its own direct processing through Trendar of

MasterCard Fleet Card transactions.” Order 17, JA 99. If Comdata did not have

specific agreements authorizing proprietary Comdata Fleet Card transactions,

Flying J argues it also need not obtain merchant consent to proprietary TCH

MasterCard transactions.

Flying J’s reliance on the lack of an explicit agreement with the merchant

attempts to sidestep the evidence that Comdata Fleet Cards provided data capture

and purchase control only when the merchant had agreed to accept its proprietary

-29-
cards. MasterCard’s representative testified that Comdata processed transactions

outside the MasterCard network based on the exception for proprietary accounts.

MasterCard Bylaws and Rules, October 2002, Rule 6.6, JA 546. The proprietary

account exception required that the merchant agree to accept the card issuer’s

proprietary card before the issuer’s fleet card could be processed privately. See

id. Rule 6.6.1, JA 546; Rule 8.05, Dec. 2001, JA 539; Hennessy Deposition, JA

1902. There was no contrary evidence. It follows that Comdata did not enter into

specific agreements to process Comdata Fleet Cards over its own network because

it would have been redundant to do so. The merchant had already agreed to

accept Comdata’s proprietary card; therefore, it had consented to proprietary

processing. The district court’s finding that Comdata provides data capture and

purchase control in MasterCard transactions without notice to merchants is thus

without support in the record.

Flying J’s response—that Comdata did not and cannot show a single

merchant that does not accept the Comdata Card—misses the point. Under the

circumstances, widespread, even universal, acceptance of Comdata’s proprietary

card does not show that Comdata did not have to obtain the merchant’s consent to

get proprietary processing; it proves only that Comdata always got it. Comdata

offers an explanation for the Comdata Fuel Card’s popularity: unlike Flying J,

Comdata is not a competitor in the truck stop industry, and unaffiliated merchants

-30-
therefore are more willing to accept its card. Far from proving Flying J’s point,

this supports the inference that merchant consent was central to Comdata’s

commercial arrangements and therefore to the Comdata model. Flying J does not

produce any evidence to the contrary. The court therefore clearly erred in finding

that the parties intended the Trendar License to have the meaning urged by Flying

If Comdata’s processing model involved unitary processing and merchant

consent, the meaning of Article 4.2 depends in part on whether Flying J knew or

had reason to know of these characteristics at the time of the settlement

agreement. Flying J and Comdata attribute different meanings to Article 4.2,

creating an apparent misunderstanding regarding a key term of the Trendar

License. Where the parties assign different meanings to a term,

it is interpreted in accordance with the meaning attached by one of
them if at the time the agreement was made

(a) that party did not know of any different meaning attached by the
other, and the other knew the meaning attached by the first party; or

(b) that party had no reason to know of any different meaning
attached by the other, and the other had reason to know the meaning
attached by the first party.

Restatement (Second) of Contracts § 201(2). If Flying J knew or had reason to

know that Comdata understood Article 4.2 to mean that Trendar would process

-31-
TCH MasterCards in the same manner as Comdata Fleet Cards—through unitary

transactions—then Article 4.2 must be interpreted accordingly.

At the time of the settlement, neither party had reason to expect that Article

4.2 would result in a dual-processing arrangement. Flying J concedes that it did

not conceive of the dual-processing model until after the parties entered the

Trendar License. Kevin Farnsworth, a TCH employee who was involved in

programming Trendar to accept TCH cards, testified that Flying J originally

requested that the TCH MasterCard be processed exclusively through the TCH

network. Testimony of Kevin W. Farnsworth 167:11–25, JA 1588. Ted Jones,

the President of TCH, testified that he was not aware of any split transactions

over the Trendar System. Testimony of Ted David Jones 149:9–21, JA 1570. At

the time of the settlement, therefore, Flying J expected that TCH MasterCards

would be processed exclusively through TCH rather than split between the TCH

and MasterCard networks.

Comdata likewise had reason to expect that the Trendar License would be

limited to unitary processing based on its own arrangement with MasterCard. As

previously noted, MasterCard’s representative testified that Comdata processed

transactions outside the MasterCard network based on the exception for

proprietary accounts. MasterCard Bylaws and Rules, October 2002, Rule 6.6, JA

546. The proprietary account exception required that the merchant agree to

-32-
accept the card issuer’s proprietary card before the issuer’s fleet card could be

processed privately. See id. Rule 6.6.1, JA 546; Rule 8.05, Dec. 2001, JA 539;

Hennessy Deposition, JA 1902. Thus, according to MasterCard, Comdata’s own

procedures are consistent with the unitary processing model. Comdata’s

arrangement clarifies the ambiguous language of Article 4.2.

Flying J counters with the argument that the parties intended to achieve a

particular result; therefore, the particulars of Comdata’s own processing model do

not matter. Flying J draws a distinction between the actual Comdata model and

an arrangement similar to the Comdata model, which would provide Flying J with

“the same real-time data capture and purchase control functionality for its TCH

MasterCard at truck stop locations that use the Trendar System that Comdata was

providing to the Comdata MasterCard Fleet Card through the Trendar System.”

Adams Declaration, JA 1119–20; see also JA 1690 (“[W]hat we contemplated in

the settlement was that we were going to be able to process transactions the way

Comdata was processing Comdata MasterCard transactions.”).

There is evidence in the record, however, that suggests Flying J knew or

had reason to know that Comdata intended Article 4.2 transactions to replicate the

procedure used in Comdata Fleet Card transactions. In his declaration submitted

to the district court, Flying J’s president described his knowledge of Comdata’s

arrangement:

-33-
As of May 2001, I knew that the Trendar system was
programmed to process Comdata MasterCard Fleet Card transactions
outside the MasterCard network. Thus, . . . the Trendar system
would automatically route the transaction directly to
Comdata—instead of through MasterCard’s network—so it could be
cleared directly through Comdata just like transactions involving the
Comdata Fuel Card.

Declaration of J. Phillip Adams, April 4, 2003, at 3, JA 842. Mr. Adams’s

statement supports the inference that Flying J actually understood “direct”

clearing—the term used in Article 4.2—to be equivalent to exclusive

clearing—the meaning urged by Comdata. In other words, it understood “direct”

clearing to mean that the entire transaction would be routed through Comdata’s

network, not split between different networks. At the very least, Flying J had

reason to know Comdata’s intended meaning. Even if Flying J thought that

Article 4.2, as written, entitled it to data capture and purchase control in all TCH

MasterCard transactions, regardless of merchant consent, there is evidence that it

had reason to know Comdata believed the contrary.

Both parties anticipated that Article 4.2 would follow the Comdata model

of MasterCard processing. The record shows that Comdata’s arrangement with

MasterCard followed a specific MasterCard policy for proprietary accounts,

which involved unitary transactions and merchant consent. There is evidence that

Flying J had reason to know that Comdata intended these limits to apply to Article

4.2. Accordingly, we hold that the district court clearly erred in finding that the

-34-
parties intended Article 4.2 to require proprietary TCH MasterCard transactions at

all Trendar locations, regardless of merchant consent, by any feasible means.

IV.

For the reasons stated above, we REVERSE the district court’s order and

REMAND the case for further proceedings consistent with this opinion.

-35-
No. 03-4262, Flying J Inc., et al. v. Comdata Network, Inc.

McKAY, Circuit Judge, dissenting:

I have a much simpler view of this case than the majority. This case is a

dispute over a claim that Comdata violated antitrust laws by restricting Flying J’s

access to the market. On the eve of trial, because Comdata realized it was in

jeopardy of an adverse judgment, it agreed to pay forty-nine million dollars to

Flying J to compensate for past violations. In addition, to rectify future

restrictions on market entry, it entered into the license that is the subject of

dispute on appeal. As the majority points out, certain sections of the settlement

agreement are ambiguous. Whatever their private post-hoc perception of the

license, the language the parties chose does not itself tell the court which

perceptions they intended.

What we do know is that the overall purpose of the license was to open

access to the market. After hearing extensive evidence about the undisclosed

understanding of each of the parties, the district court opted to find Flying J’s

understanding to be the more reasonable one. It then ordered a remedy that most

effectively implemented that perception. Our job on appeal is to determine

whether that finding is clearly erroneous. In my view, the parties’ evidence lends

itself to either party’s interpretation. The district court’s selection of the

interpretation that favors optimal opening of the competitive market seems to me
to be eminently reasonable and supported by the record. Resolving any doubt in

favor of the purpose of the antitrust statutes strengthens this conclusion. I see no

reason to criticize the trial court’s judgment. See DUCivR 54-1(c); Blankenship

v. Herzfeld, 721 F.2d 306, 310 (10th Cir. 1983); Ramey Constr. Co. Inc. v.

Apache Tribe of Mescalero Reservation, 616 F.2d 464, 467 (10th Cir. 1980). I

would affirm the trial court’s decision.

-2-

Flying J Inc. v. Comdata Network, Inc.

Boost your productivity today