Brooks, Tarlton, Gilbert, Douglas & Kressler, Etc., Cross-Appellants v. United States Fire Insurance Company, Cross-Appellee

RANDALL, Circuit Judge: Defendant, an insurance company, appeals the district court’s decision — based on competing cross-motions — to grant summary judgment in favor of plaintiff, the insured, on a breach of contract claim. In a related argument, defendant also challenges the district court’s award of damages on the contract claim. Next, defendant appeals the district court’s determination that defendant’s conduct toward its insured violated the Texas Deceptive Trade Practices Act. Plaintiff cross-appeals the Deceptive Trade Practices Act claim, arguing that the district court incorrectly applied the Act’s 1979 amendments to plaintiff’s claim and, as a result, awarded plaintiff only a fraction of the punitive damages which it should have received. Plaintiff also asks this court to correct a mathematical mistake in the judgment entered below. We agree with the district court that defendant breached the terms of the insurance policy between plaintiff and defendant. However, we reject the court’s conclusion that this breach of contract was also a breach of an express warranty and, therefore, a Deceptive Trade Practices Act violation. Moreover, the unorthodox nature of the proceedings below prohibits us from deciding whether either the alternative ground asserted by the district court in support of its judgment or the additional arguments pled by plaintiff demonstrate a violation of the Act. Consequently, we affirm the district court’s entry of summary judgment against defendant, and its subsequent order on damages, as they relate to the breach of contract claim. We reverse, however, the district court’s decision on the Deceptive Trade Practices Act claim and return the claim to that court for further proceedings. I. FACTS AND PROCEDURE In April, 1975, the law partnership of Brooks, Tarlton, Gilbert, Douglas & Kres-sler and its individual members (“the Lawyers”) entered into a contract for insurance with United States Fire Insurance Company (“USFIC”). The insurance policy (“the policy”) purchased by the Lawyers was a professional liability policy — it recited that USFIC would pay, on the Lawyers’ behalf, all sums which the Lawyers became legally obligated to pay as damages arising out of the Lawyers’ performance of professional services for others. In addition, the policy contained a “Defense, Settlement, Supplementary Payments” provision by which USFIC agreed that, with respect to the insurance afforded by the terms of the policy, it would defend the Lawyers in any suit brought against them, regardless of whether the suit was “groundless, false or fraudulent.” The policy also, however, contained an exclusionary clause (“exclusion provision”) which recited that the policy did not apply “to any dishonest, fraudulent, criminal or malicious act or omission of the insured.” The policy coverage was for one year, but the Lawyers and USFIC renewed it in 1976, 1977, 1978, and 1979 on essentially the same terms. Consequently, the Lawyers were continuously covered by the policy from April, 1975 to April, 1980. In April, 1975, Eva D. Clifton (“Clifton”) hired the Lawyers to represent her in a divorce action filed against her by her husband. In early 1976, Clifton retained Bel-vin R. Harris (“Harris”) of the firm of Sullivant, Meurer, Harris & Sullivant to assist the Lawyers in the divorce proceeding. Although initially Clifton agreed to pay the Lawyers on an hourly basis, sometime early in 1976, Clifton, the Lawyers, and Harris entered into a contingent fee contract which obligated Clifton to pay for her representation based on the amount of property that she was awarded in the divorce action. In June, 1976, Clifton and her husband were granted a divorce according to the terms of a negotiated settlement agreement. Immediately prior to the filing of the decree, and in fulfillment of the contingent fee agreement, Clifton signed a deed conveying to the Lawyers and Harris 10% of one-half of all of her undivided *1360 interest in the properties she retained under the settlement agreement. Clifton, however, subsequently became dissatisfied with the terms of the contingent fee arrangement. Consequently, in April, 1979, Clifton brought suit against the Lawyers and Harris to have the deed conveying the 10% interest set aside. 1 As her basis for setting aside the deed, Clifton argued in her petition that the deed involved a “breach of confidential relationship between the attorneys and their client, and therefore involved a breach of trust and [was] illegal, void and unenforceable.” Moreover, Clifton alleged that the deed was procured “under fraud, duress, undue influence, and involved the breach of an attorney-client privilege, which was a fiduciary relationship.” Finally, Clifton alleged that the contingent fee agreement was “unconscionable and therefore in violation of the Deceptive Trade Practice Act of The State of Texas” because it took advantage of her lack of knowledge and resulted in a “gross disparity between the services received and the consideration paid therefor.” After being served with Clifton’s petition, the Lawyers informed USFIC of the suit and demanded that USFIC provide them with a defense, as required by the terms of the policy. USFIC, however, evaluated Clifton’s allegations in the petition and concluded that the petition alleged only “dishonest, fraudulent, criminal or malicious” acts; therefore, on April 10, 1979, USFIC informed the Lawyers by letter that the exclusion provision removed its obligation to provide the Lawyers with a defense. Subsequently, Clifton amended her petition five times. As the amended petitions were served on the Lawyers, the Lawyers made demand on USFIC for a defense under the policy. USFIC, however, adhered to its original position and continued to refuse the Lawyers’ demand. 2 Because of USF-IC’s position, the Lawyers hired their own counsel to defend them against Clifton. Beginning August 22,1983, Clifton’s claims were tried to a jury; on September 6, 1983, the case ended in a mistrial when the jury “hung.” On January 24, 1984, the Lawyers brought this suit against USFIC based on USFIC’s refusal to defend the Lawyers in the Clifton suit. In their complaint, the Lawyers asserted two separate causes of action. First, the Lawyers claimed USFIC breached its insurance contract with them by failing to defend. Second, the Lawyers charged that in refusing a defense, USFIC violated several provisions of the Texas Deceptive Trade Practices Act — including sections 17.46(a), 17.46(b)(5), 17.46(b)(12), and 17.50(a)(1), (2), (3), and (4). Under the first cause of action, the Lawyers sought actual and punitive damages and attorneys’ fees; under the second, they sought their attorneys’ fees and three times their actual damages. After USFIC answered, both *1361 parties filed motions for summary judgment with the court on the breach of contract claim. 3 Each party argued, in its motion, that the unambiguous terms of the insurance contract required the court to support its position as a matter of law. On November 9, 1984, the district court denied USFIC’s motion, granted the Lawyers’, and entered summary judgment for the Lawyers on the issue of liability under the contract. However, contrary to the position taken by both parties in their competing motions, the court reached its conclusion by determining that the terms of the insurance policy were ambiguous as a matter of law. Consequently, the court construed the policy in favor of the insured and, therefore, in favor of USFIC’s liability. In its order memorializing its decision, the court directed that “[a] hearing on the issue of damages will be set as soon as practicable.” For more than a year, no action was taken. Finally, on January 14, 1986, the district court entered an order setting the case for “a hearing on damages.” After reluctantly granting USFIC a one-month continuance, the district court convened the parties on February 20, 1986. At that time, the court announced that, because of the partial summary judgment which it previously granted the Lawyers, it was ready to take evidence on the issue of damages. The Lawyers, therefore, called several witnesses to testify to the attorneys’ fees the Lawyers incurred in defending the action brought by Clifton — the action for which the court found USFIC owed the Lawyers a defense — and the fees incurred in bringing suit against USFIC for its failure to defend. Along the way, the Lawyers also placed into evidence copies of the insurance policy between USFIC and the Lawyers, and correspondence between the two concerning USFIC’s duty to defend the suit brought by Clifton. Although USFIC cross-examined the Lawyers’ witnesses, it did not present any evidence or witnesses of its own. At the conclusion of the hearing, both parties presented “summation” arguments to the court. Neither argument, however, focused on the evidence which had been presented during the hearing; instead, both parties directed their attention to the Deceptive Trade Practices Act claim which the Lawyers alleged in their complaint. In their summation, the Lawyers focused on the version of the Act under which the court should award damages. They explained that before 1979, the Act required an automatic trebling of all actual damages once a violation of the Act has been found. Since 1979, however, the Act only permits an award of damages beyond actual damages where a defendant is shown to have committed a “knowing” violation. In this case, because USFIC’s initial decision not to defend occurred while the automatic trebling provision was in effect, the Lawyers concluded, any actual damages the court found should automatically be trebled. USFIC’s summation also addressed the question of which version of the Act applied. The court, USFIC argued, should apply the 1979 Act because Clifton’s final amended petition was not filed until after the automatic trebling provision was deleted from the Act. USFIC also, however, argued that no Deceptive Trade Practices Act violation existed at all; instead, USFIC informed the court, the Lawyers’ claim was only a breach of contract claim and, therefore, was not actionable under the Act. When the summations were concluded, the court invited both parties to furnish the court by letter additional legal precedent to support their positions. Both parties subsequently accepted the court’s invitation, and continued in writing their debate over the Deceptive Trade Practices Act claim. In addition, both parties filed “Requested *1362 Conclusions of Law” with the court for its consideration. On October 6, 1986, the district court filed its memorandum opinion concerning the hearing on damages. In its opinion, the court addressed two discrete subjects. First, it awarded the Lawyers actual damages in the amount of $171,127.29 as “reasonable and necessary fees and expenses in connection with the Clifton action.” These fees were recoverable because of the summary judgment the court previously entered in favor of the Lawyers, and included sums the Lawyers spent on attorneys’ fees in the Clifton suit and sums spent in bringing this action against USFIC. Second, the court addressed the Lawyers’ claims under the Deceptive Trade Practices Act. With respect to liability under the policy, the court explained that, “The Court previously held that at least two of the claims in the Clifton suit were not excluded from defense under the policy, yet [USFIC] repeatedly refused to defend [the Lawyers].” Therefore, the court continued, “defendant breached an express or implied warranty under the terms of the professional liability policy when they failed to defend plaintiffs in the Clifton Suit [sic].” However, the court concluded that the Lawyers could only recover damages under the post-1979 version of the Act because Clifton’s Fifth Amended Original Petition was not filed until 1981. According to the court, therefore, the only remaining question was whether USFIC violated the Act in a manner which permitted the Lawyers to recover more than their actual damages. Consequently, the court considered whether USF-IC “knowingly” violated the Deceptive Trade Practices Act when it denied the Lawyers a defense. With respect to this question, the court explained: An insurer is held to a higher standard of care in examinig [sic] its policy to determine coverage because an insurance company’s business expertise is in providing insurance coverage and interpreting such policies. It is clear that a thorough review of Clifton’s petitions would have evidenced the responsibility of [USFIC] to provide a defense under [the Lawyers’] liability policy. From the foregoing facts, the Court can only conclude that [USFIC’s] failure to defend was done knowingly. Because USFIC knowingly violated the Act, the court concluded, USFIC was liable to the Lawyers for the additional amount of $57,042.46, a one-third increase of the actual damage award. Consequently, on October 6, 1986, the district court entered judgment in favor of the Lawyers in the amount of $221,232.35. 4 Both USFIC and the Lawyers appeal from the district court’s rulings. The questions the parties raise involve issues decided by summary judgment and by order from the damages hearing. They involve arguments on the contract claim, the Deceptive Trade Practices Act claim, and the damages relating to each. Consequently, for the sake of clarity, we will wait to set forth in detail each argument as we address it. After evaluating all the arguments made by the parties, however, we conclude as the district court did that USF-IC breached its contract by refusing to provide a defense for the Lawyers when they were sued by Clifton. Consequently, we affirm the summary judgment against USFIC on the contract claim, and we affirm that part of the district court’s order from the damages hearing which determines the actual damages the Lawyers suffered as a result of the breach. But we disagree with the court’s conclusion that USFIC’s promise to provide a defense constitutes an express warranty under the policy. In addition, because the parties never were given an opportunity to present evidence on the issue of liability under the Deceptive Trade Practices Act claim, we cannot evaluate the propriety of the court’s possible conclusion that the failure to defend constituted a breach of an implied *1363 warranty. Nor are we able to assess, at this level, the additional violations the Lawyers asserted under the Act. Like the implied warranty claim, the validity of these assertions depends on proof of facts. Consequently, having no evidence from which we can conclude that the district court’s decision on the Deceptive Trade Practices Act claim was correct, we must reverse the court’s holding and remand for further consideration of those claims which have yet to be developed by evidence. II. ANALYSIS A. The Breach of Contract Claim The most contested issue, and the first one USFIC appeals, is whether USFIC had a responsibility under the insurance policy to provide the Lawyers with a defense in the action brought against them by Clifton. To understand the parties’ arguments, the district court’s holding, and our conclusions, it is necessary to first understand which clauses in the policy decide the issue. In the order they appear in the policy, the relevant clauses read: 5 INSURING AGREEMENTS I. COVERAGE. To pay on behalf of the Insured all sums which the Insured shall become legally obligated to pay as damages arising out of the performance of professional services for others in the Insured’s capacity as a lawyer and caused by the Insured or any other person for whose acts the Insured is legally liable. II. DEFENSE, SETTLEMENT, SUPPLEMENTARY PAYMENTS. As respects such insurance as is afforded by the other terms of this policy the Company shall (a) defend in his name and behalf any suit against the Insured alleging damages, even if such suit is groundless, false or fraudulent; but the Company shall have the right to make such investigation and negotiation of any claim or suit as may be deemed expedient by the Company. The Company, however, shall not make settlement or compromise any claim or suit without the written consent of the Insured. EXCLUSIONS This Policy Does Not Apply: (a) to any dishonest, fraudulent, criminal or malicious act or omission of the Insured, any partner or employee. The district court read these clauses and determined, on summary judgment, that they imposed on USFIC the responsibility for providing the Lawyers with a defense when Clifton sued them. We cannot examine this conclusion, however, until we decide the focus in which we must view it; consequently, we first note the standard of review on an appeal from summary judgment and the law which will control this issue. *1364 When an appeal is taken from a summary judgment, we review the record under the same standards the district court applied to determine that summary judgment was appropriate. Reid v. State Farm Mut. Auto. Ins. Co., 784 F.2d 577, 578 (5th Cir.1986). Therefore, we will affirm a summary judgment only when we are convinced, after an independent review of the record, that “there is no genuine issue as to any material fact” and that the movant is “entitled to a judgment as a matter of law.” Fed.R.Civ.P. 56(c); see Galindo v. Precision Am. Corp., 754 F.2d 1212, 1216 (5th Cir.1985). When fact questions are raised, we must “review the facts drawing all inferences most favorable to the party opposing the motion.” Reid, 784 F.2d at 578. However, when questions of law control the disposition on summary judgment, we will subject to full appellate review the questions raised by the parties. See id.; Turbo Trucking Co. v. Those Underwriters at Lloyd’s London, 776 F.2d 527, 529 (5th Cir.1985). Moreover, because the parties chose our federal forum in which to litigate their diversity suit, we are bound by the principles of Erie R.R. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188 (1938), to settle this dispute as would the Texas courts — by applying Texas substantive law. In this case, summary judgment raised only the question of how properly to construe the exclusion provision of the insurance policy. In Texas, “[i]t is a fundamental rule of law that insurance policies are contracts and as such are controlled by rules of construction which are applicable to contracts generally.” Barnett v. Aetna Life Ins. Co., 723 S.W.2d 663, 665 (Tex.1987). Usually, the first question faced in contract interpretation is whether the contract terms in dispute are ambiguous. And the determination of whether terms are ambiguous is, in Texas, a question of law; it is only the resolution of ambiguities through the resort to extrinsic evidence which creates a question of fact. J.B. Watkins v. Petro-Search, Inc., 689 F.2d 537, 538 (5th Cir.1982); see also Ideal Mut. Ins. Co. v. Last Days Evangelical Ass’n, 783 F.2d 1234, 1238 (5th Cir.1986). The reason for the focus on ambiguity is simple; “[wjhen terms of an insurance policy are unambiguous, they are to be given their plain, ordinary and generally accepted meaning unless the instrument itself shows that the terms have been used in a technical or different sense.” Ramsay v. Maryland Am. Gen. Ins. Co., 533 S.W.2d 344, 346 (Tex.1976). Moreover, when an unambiguous contract exists, the contract alone will be deemed to express the intention of the parties, “for it is objective, not subjective, intent that controls.” Watkins, 689 F.2d at 538 (applying Texas law). By contrast, in Texas an ambiguous contract— when the contract is an insurance policy— is tested by special rules of construction. For example, when an ambiguity exists, exceptions and limitations in insurance policies are construed against the insurer. Ranger Ins. Co. v. Bowie, 574 S.W.2d 540, 542 (Tex.1978). Moreover, the existence of an ambiguity requires a court to resolve the ambiguity in a way which favors the insured and permits recovery, as long as the construction which the court must adopt to do so is not itself unreasonable. Ideal Mut. Ins. Co., 783 F.2d at 1238 (quoting Ramsay, 533 S.W.2d at 346 and Continental Casualty Co. v. Warren, 152 Tex. 164, 167, 254 S.W.2d 762, 763 (1953)). In other words, “[T]he insurer may not escape liability merely because his or its interpretation should appear to us a more likely reflection of the intent of the parties than the interpretation urged by the insured.” Continental Casualty Co., 152 Tex. at 167, 254 S.W.2d at 763. The test for determining whether a contract is ambiguous is whether, after applying established rules of construction, the contract is reasonably susceptible to more than one meaning. Richland Plantation Co. v. Justiss-Mears Oil Co., 671 F.2d 154, 156 (5th Cir.1982) (citing Universal C.I.T. Credit Corp. v. Daniel, 150 Tex. 513, 243 S.W.2d 154, 157 (1951)); see also Barnett v. Aetna Life Ins. Co., 723 S.W.2d 663, 665 (Tex.1987). As far as insurance contracts are concerned, therefore, a decision that the policy is ambiguous appears necessarily to predetermine that the insured will pre *1365 vail; it takes two reasonable interpretations to make the policy ambiguous, but given two reasonable interpretations in an ambiguous policy, the insured’s interpretation must be accepted. An allegation of ambiguity, therefore, is of pivotal importance. Moreover, “[a]s necessity is the mother of invention, so is ambiguity the father of multiple reasonable constructions, and where lawyers are involved, one never lacks an eager parent of either gender.” Ideal Mut. Ins. Co., 783 F.2d at 1238. It is interesting, therefore, that the allegation of ambiguity in this case came, not from the parties, but from the district court. In ruling on the competing motions for summary judgment with which it was faced, the district court concluded that “the policy in question is susceptible of more than one reasonable interpretation” and, therefore, ruled in favor of the Lawyers “that the exclusionary language of the policy does not pertain to the duty to defend but only to the coverage of the policy.” It is this conclusion which USFIC first appeals. It appears from the district court’s opinion that it felt compelled to hold as it did because of the Fourth Circuit decision of Donnelly v. Transp. Ins. Co., 589 F.2d 761 (4th Cir.1978). In Donnelly, the Fourth Circuit was faced with a question, under the law of the District of Columbia, very similar to the one which is before us now: Did an insurance company’s refusal to defend its insured in a particular action constitute a breach of the insurance policy’s “duty to defend” clause? Even more familiar to us, however, are the terms of the policy which the Fourth Circuit evaluated to answer the question. With no significant deviation, the insurance policy before the court in Donnelly was the same one we find before us. 6 Finally, the argument made by the insurer in Donnelly is the same one USFIC argues today — that the exclusion provision precludes it from being responsible for providing a defense. In deciding the scope of the policy before it, the Fourth Circuit preliminarily explained that, “Although basically a duty to defend and a duty to pay are to some extent coextensive, the insurance company’s duty to defend its assured is usually said to be broader than its duty to pay a judgment against him.” Id. at 765. From this rule of law, the Fourth Circuit extrapolated to the conclusion that, in this policy, the insurer is not relieved from providing its insured a defense even though the allegations against the insured are such that, were the insured found liable in the suit, the terms of the policy would relieve the insurer from paying the judgment. As the Fourth Circuit noted: [T]he policy makes no express reference to the insurance company being excused from defending suits in which dishonest, fraudulent, criminal or malicious acts are alleged. While, as noted, there may be sound policy reasons why an insured should not be indemnified from the consequences of acts such as those enumerated, there are no similar reasons why an innocent person, charged with such acts, should not have the benefit of a defense. Policy language, susceptible of more than one interpretation, is construed, if reasonably possible, to provide coverage. Id. at 768 (emphasis added). The Donnelly court, therefore, held that the exclusion *1366 provision in the insurance policy before it— which is the same as the insurance policy before us — applied only to the general coverage clause and not the duty to defend clause. When ruling in this case, therefore, the district court was faced with two possible constructions of the insurance contract before it: first, that the exclusion provision does not modify the duty to defend and second, that it does modify the duty to defend. Moreover, it was also faced with the Fourth Circuit’s opinion that the former construction is reasonable. Consequently, since the former construction also results in coverage for the insured, the district court was compelled under Texas law to accept that construction. The dilemma the district court faced, therefore, was created by the Fourth Circuit’s determination, under the law of the District of Columbia, that it is reasonable to construe the duty to defend clause independent of the exclusion clause. Whatever the validity of the Fourth Circuit’s opinion as it construes the law of the District of Columbia, we do not believe that its result can ultimately be squared with either the express terms of the policy or Texas law. Consequently, we decline to adopt the Fourth Circuit’s reasoning in Donnelly, and conclude that the insurance policy in issue is unambiguous. Our decision is compelled by Texas law. As a general rule of contract construction, a contract should not be construed in a way which renders any of its terms meaningless. Ideal Mut. Ins. Co., 783 F.2d at 1238; Blaylock v. Am. Guar. Bank Liab. Ins. Co., 632 S.W.2d 719, 722 (Tex.1982). The position advanced by the Fourth Circuit, however, does just that. Specifically, it ignores the fact that by the policy’s terms, the duty to defend only arises “[a]s respects such insurance as is afforded by the other terms of this policy.” If the exclusion provision is read, as even the Fourth Circuit agrees it must be, to modify the general coverage provision of the policy, it is incongruous to say that the coverage provision modifies the duty to defend provision without reference to the exclusion provision. Moreover, the Fourth Circuit’s position ignores the structure of the insurance policy. For example, both the general coverage provision and the duty to defend provision are located under the section of the policy titled “Insuring Agreements.” The exclusion provision, however, appears in a separate section appropriately titled “Exclusions.” The exclusions which follow the title are introduced by the disclaimer that “This Policy Does Not Apply” to the situations detailed by the exclusions. No logic supports applying the disclaimer language to one of the several separate provisions in the “Insuring Agreements” section of the policy at the expense of the others. It is clear to us that the only reasonable construction of the policy is that none of the insurer’s duties under the policy arises in situations covered by the exclusion provisions. Perhaps more importantly, to hold otherwise would be to run afoul of the express dictates of the Texas Supreme Court, which has held that, “An insurer is required to defend only those cases within the policy coverage.” Fidelity & Guar. Ins. Underwriters, Inc. v. McManus, 633 S.W.2d 787 (Tex.1982); see also Holmes v. Employers Casualty Co., 699 S.W.2d 339, 340 (Tex.App.—Houston [1st Dist.] 1985, writ ref’d n.r.e.); International Serv. Ins. Co. v. Boll, 392 S.W.2d 158, 160 (Tex.Civ.App.—Houston 1965, writ ref’d n.r.e.). Determining that the district court erred in concluding that the exclusion provision did not modify USFIC’s duty to defend does not, however, end USFIC’s problems. USFIC was only relieved of its duty to defend if Clifton’s suit actually did not fall within the policy coverage. In Texas, the duty to defend “is determined solely from the face of the pleadings and without reference to facts outside the pleadings.” 7 *1367 Continental Sav. Ass’n v. United States Fidelity & Guar. Co., 762 F.2d 1239, 1243 (quoting Rhodes v. Chicago Ins. Co., 719 F.2d 116, 119 (5th Cir.1983)). “If the petition only alleges facts excluded by the policy, the insurer is not required to defend.” Fidelity & Guar. Ins. Underwriters, Inc., 633 S.W.2d at 788. This is because “the insurer is entitled to rely on the plaintiffs allegations in determining whether the facts are within the coverage.” Id.; see also Heyden Newport Chemical Corp. v. Southern Gen. Ins. Co., 387 S.W.2d 22, 24-25 (Tex.1965). Although these principles seem to permit an insurance company great latitude in deciding whether to defend, the reality is that as they have been applied by the Texas courts, the principles are extremely confining. The Texas Supreme Court has directed that “in considering [the] allegations [in the pleadings] a liberal construction of their meaning should be indulged.” Hey-den, 387 S.W.2d at 26. This means that if the allegations made against the insured, when taken as true, even potentially state a cause of action within the terms of the policy, the insurer must defend. Continental Sav. Ass’n, 762 F.2d at 1243; Rhodes, 719 F.2d at 119; Heyden, 387 S.W.2d at 26. More ominous still is the supreme court’s direction that, “in applying the pleadings only rule, the court may indulge the most liberal interpretation of the allegations of which they are susceptible and doubts as to the import of the allegations are to be resolved in favor of the insured and coverage.” Continental Sav. Ass’n, 762 F.2d at 1243 (paraphrasing Heyden, 387 S.W.2d at 26). In addition, as long as the complaint alleges at least one cause of action within the policy coverage, the insurer must defend. Rhodes, 719 F.2d at 119. And finally, we have held that the responsibility for defending is an ongoing one; therefore, “[a] complaint which does not initially state a cause of action under the policy, and so does not create a duty to defend, may be amended so as to give rise to such a duty.” Id. In addition to concluding that summary judgment for the Lawyers was appropriate because the duty to defend existed independent of the exclusion clause, the district court evaluated the allegations in the six petitions Clifton filed against the Lawyers. According to the court, “all of the petitions alleged the same basic claims, except that the later petitions contained more detail.” These claims, the court explained, encompassed (1) fraud; (2) duress; (3) unconscionable fee arrangements; (4) breach of the attorney-client relationship; and (5) violation of certain sections of the Code of Professional Responsibility. The exclusion provision on which USFIC relied, however, excluded only “dishonest, fraudulent, criminal or malicious acts.” As the court read the allegations, the last two claims were not “clearly beyond the coverage of [the Lawyers’] policy.” Consequently, the court held that “even if the exclusionary language did pertain to [USFIC’s] duty to defend, not all of the claims of the Clifton petition fall under the listed exclusions.” Therefore, USFIC had a duty to defend and summary judgment for the Lawyers was appropriate. This conclusion by the district court is the second issue USFIC appeals. As USFIC explains it, the district court’s conclusion that the allegations in Clifton’s numerous petitions imposed a duty to defend is wrong for two reasons. First, USFIC asserts, because of the timing *1368 of Clifton’s suit, she could not have alleged a cause of action within the policy’s terms. Clifton’s lawsuit was brought well over two years after her cause of action accrued; consequently, the statute of limitations precluded her from asserting a valid cause of action against the Lawyers for either their negligence or legal malpractice. We dismiss this argument outright. Never have the Texas courts focused on the validity of allegations to determine the duty to defend. In fact, the Texas Supreme Court expressly rejected such a standard when it concluded that: We think that in determining the duty of a liability insurance company to defend a lawsuit the allegations of the complainant should be considered in the light of the policy provisions without reference to the truth or falsity of such allegations and without reference to what the parties know or believe the true facts to be, or without reference to a legal determination thereof. Heyden, 387 S.W.2d at 24; accord Argonaut Southwest Ins. Co. v. Maupin, 500 S.W.2d 633, 635 (Tex.1973); St. Paul Ins. Co. v. Rahn, 641 S.W.2d 276, 279 (Tex.App. —Corpus Christi 1982, no writ). Moreover, the express terms of the Lawyers’ insurance policy refute USFIC’s argument; the duty to defend provision explains that USF-IC’s duty — once it arises — exists regardless of whether the suit is “groundless, false or fraudulent.” Consequently, the proper question is not what could Clifton successfully have pled, but what did Clifton in fact plead. USFIC’s second argument opposing the district court’s holding focuses on the petitions as pled. Significantly, USFIC does not argue that the district court erred in characterizing the claims in Clifton’s various petitions. Moreover, it adopts the court’s conclusions that the claims remained the same from petition to petition and that only the amount of information concerning the claims changed. USFIC does, however, focus its attention on the nature of the claims identified by the court. According to USFIC, when the common, ordinarily accepted meanings of the phrases used in the exclusion provision are reviewed, it is clear that the exclusion provision encompasses the claims which concerned the district court. More precisely, USFIC argues, both the Texas Supreme Court and the Fifth Circuit recognize that an attorney’s breach of his ethical duties in an attorney-client relationship constitutes a form of fraud; since the exclusion for fraudulent actions is express, USFIC concludes, Clifton’s claims of breach of the attorney-client relationship and violations of the Code of Professional Responsibility are expressly excluded from coverage. We begin our analysis of USFIC’s argument by adopting, as both parties have, the district court’s characterization of the claims in Clifton’s various petitions. We add only our observation that Clifton’s un-conscionability claim, as the Deceptive Trade Practices Act permits, took the form of alleging both that the Lawyers took advantage of her lack of knowledge and that the contingent fee arrangement resulted in a gross disparity between services received and consideration paid. 8 See Tex. Bus. & Com.Code Ann. § 17.45(5) (Vernon Supp.1987). In addition, we agree with USFIC’s assertion that it is the common, ordinarily accepted meaning of the phrases in the exclusion provision on which we should focus our attention. See Ramsay, 533 S.W.2d at 346 (holding that when the terms of an insurance policy are unambiguous, they are to be given their plain, ordinary and generally accepted meanings). We turn, therefore, to the Texas Supreme Court decision upon which USFIC’s argument hinges. *1369 In Archer v. Griffith, 390 S.W.2d 735 (Tex.1964), a former client brought an action to set aside a deed against the attorney who represented her in her divorce. She had executed the deed in favor of the attorney pursuant to a contingent fee contract between them. In deciding the client’s claim, the supreme court found it necessary to examine the nature of her complaint. As the court saw it: The relation between an attorney and his client is highly fiduciary in nature, and their dealings with each other are subject to the same scrutiny, intend-ments and imputations as a transaction between an ordinary trustee and his ces-tui que trust. ... “Although an attorney is not incapacitated from contracting with his client for compensation during the existence of the relation of attorney and client, the courts, because of the confidential relationship, scrutinize with jealousy all contracts between them for compensation which are made while the relation exists. There is a presumption of unfairness or invalidity attaching to the contract, and the burden of showing its fairness and reasonableness is on the attorney.” Id. at 739 (quoting Pomeroy, Equity Jurisprudence, (§ 960d 5th ed. 1941). Since Archer was decided, it has been generally understood by the courts — including this one — to set the standard for constructive, or legal, fraud. See, e.g., H.L. Peterson Co. v. Applewhite, 383 F.2d 430, 435 (5th Cir.1967); Castleberry v. Branscum, 721 S.W.2d 270, 273 (Tex.1986); Tuttlebee v. Tuttlebee, 702 S.W.2d 253, 257 (Tex.App—Corpus Christi 1985, no writ). It also, however, explains the distinction between actual and constructive fraud: Actual fraud usually involves dishonesty of purpose or intent to deceive, whereas constructive fraud is the breach of some legal or equitable duty which, irrespective of moral guilt, the law declares fraudulent because of its tendency to deceive others, to violate confidence, or to injure public interests. Archer, 390 S.W.2d at 740. Based on Archer, we agree with USFIC that Clifton’s allegations against the Lawyers concerning the breach of their ethical duties are properly characterized as charging the Lawyers with constructive fraud. The benchmark of constructive fraud, as defined by Archer, is the existence of a fiduciary relationship. It is the existence of the relationship, apart from any state of mind of the attorney, which casts doubt on a transaction between the attorney and client and imposes the presumption of “fraud.” In order to defeat the presumption, the attorney bears the burden of establishing that the transaction was “fair, honest, and equitable.” Archer, 390 S.W. 2d at 740. Significantly, the Texas courts refuse to find that, by itself, the attorney’s good faith in entering into the transaction establishes fairness. See Miller v. Miller, 700 S.W.2d 941, 947 (Tex.App. — Dallas 1985, writ ref’d n.r.e.) (listing Texas cases). Consequently, it is possible for an attorney to be liable for constructive fraud even though the attorney operated completely in good faith during the entire transaction. Under the liberal interpretation of pleadings standard which the Texas Supreme Court has directed us to apply in determining an insurer’s duty to defend, we must read Clifton’s constructive fraud allegations against the Lawyers in light of this possibility. The question which we must decide, therefore, is whether the common, ordinary meaning of the term fraud — as it appears in the exclusion provision of the insurance policy — includes constructive as well as actual fraud. After carefully considering the question, we conclude that it does not. For help in determing the common, ordinary meaning of the term “fraud,” we have consulted several dictionaries. A sampling of the definitions we have uncovered leads us to the conclusion that the term “fraud” is generally understood to carry with it the connotation of a wrongful intent. Webster’s defines “fraud” as “the intentional perversion of truth for the purpose of inducing another in reliance upon it to part with some valuable thing belonging to him, or to surrender a legal right.” Webster’s New International Dictionary 1003 (2d ed. unabridged 1956). Similarly, “fraud” is de *1370 fined elsewhere as “[a] deception deliberately practiced in order to secure unfair or unlawful gain.” American Heritage Dictionary 531 (2d college ed. 1982). Even Black’s Law Dictionary, when defining “fraud” generally, gives the definition of actual fraud — which includes the element of wrongful intent. See Black’s Law Dictionary 594 (5th ed. 1979); see also Philipp Bros., Inc. v. Oil Country Specialists, Ltd., 709 S.W.2d 262, 264 (Tex.App.—Houston [1st Dist.] 1986, writ ref’d n.r.e.) (defining “fraud” as a false representation of material fact made with intent to induce the listener to act upon it). It is only when explaining the distinction between actual and constructive fraud that the dictionary defines the latter. See Black’s Law Dictionary 595 (5th ed. 1979). We are also aided in our search for the common meaning of the term by the language of the insurance policy. The exclusion provision specifies that coverage is not afforded for “dishonest, fraudulent, criminal or malicious acts.” We find it significant that the terms which both precede and follow the one on which we focus are also commonly understood as including an element of wrongful intent. See Webster’s New International Dictionary 626, 748, 1489 (2d ed. unabridged 1956); American Heritage Dictionary 340, 405, 759 (2d college ed. 1982). In interpreting the contract between the parties, we must construe the terms of the exclusion provision in harmony with each other. Therefore, we conclude that Clifton’s allegations of constructive fraud did not clearly fall within the language of the exclusion provision. Consequently, as a matter of law USFIC had a duty to defend the Lawyers in the Clifton suit. We affirm the district court’s decision to grant summary judgment for the Lawyers on their breach of contract claim against USFIC. 9 *1371 B. The Deceptive Trade Practices Act Claim Leaving the worlds of contract claims and summary judgments, we now turn our attention to the Texas Deceptive Trade Practices Act (“the Act” or “the DTPA”) and the parties’ final arguments on appeal. As discussed earlier, in its order entered after the damages hearing, the district court specifically found that USFIC “breached an express or implied warranty under the terms of the professional liability policy when [it] failed to defend [the Lawyers] in the Clifton suit.” Consequently, the court determined that USFIC’s failure to defend violated section 17.50(a)(2) of the Act; in addition, the court concluded that the violation was “knowing” as that term is defined within the 1979 amended version of the Act, and awarded the Lawyers additional damages based on that finding. USFIC appeals the court’s determination that its failure to defend constituted a violation; the Lawyers cross-appeal the court’s conclusion that the amended version of the Act controlled the award of damages on its claim. To address either argument, we must first examine the posture in which they reach us. In their complaint, the Lawyers alleged the USFIC committed five separate violations of the Act: (1) they represented that goods or services had characteristics, ingredients, uses, benefits, or qualities which they did not have; (2) they represented that an agreement conferred or involved rights, remedies, or obligations which it did not have or involve; (3) they breached on an express or implied warranty; (4) they engaged in an unconscionable action or course of action; and (5) they used an act or practice in violation of Article 21.21 of the Texas Insurance Code, or the rules or regulations issued by the State Board of Insurance under Article 21.21. 10 Once the complaint was filed, no subsequent pleadings addressed the DTPA claim. Instead, the parties filed competing motions for summary judgment on the contract claim. After these motions were resolved in favor of the Lawyers, a hearing on damages was held. At the hearing, the testimony concerned the amount of attorneys’ fees incurred in the Clifton suit; additional evidence was limited to the Lawyers’ letters requesting a defense and USFIC’s responsive letters denying the duty to defend. See note 2 supra. As a result of the hearing, the district court made an award of contract damages; it also, however, ruled on the Lawyers’ DTPA claim. This claim reaches us, therefore, having never been the subject of a trial or other eviden-tiary hearing. Significantly, evidence of how USFIC decided whether to defend was not put before the court; the only fact not related to the issue of damages for which evidence was introduced was that USFIC did not defend — and this fact was not contested. Consequently, the district court’s ruling is based solely on the Lawyers’ broad allegations in the complaint and the court’s legal conclusion that USFIC was wrong to interpret the insurance policy not to require a defense. Since the district court did not engage in any “fact finding,” therefore, we conclude that the district court must have ruled on the DTPA claim as a matter of law. 11 Consequently, as we *1372 do with other questions of law, we will review the court’s ruling de novo. The district court found that “[u]nder the circumstances of this case, ... [USFIC] breached an express or implied warranty-under the terms of the professional liability policy when [it] failed to defend [the Lawyers] in the Clifton suit.” We first turn, therefore, to the question of express warranty. The Texas courts treat the question of whether a promise constitutes an express warranty as a question of law. See, e.g., S.I. Property Owners’ Assoc. v. Pabst Corp., 714 S.W.2d 358, 361 (Tex.App.—Corpus Christi 1986, writ ref d n.r.e.) (holding that representations did not rise to the level of a warranty); Blackwood v. Tom Benson Chevrolet Co., 702 S.W.2d 732, 734 (Tex.App.—San Antonio 1985, no writ) (construing a promise in a contract as an express warranty); Cravens v. Skinner, 626 S.W.2d 173, 176 (Tex.App—Fort Worth 1981, no writ) (holding that a promise to deliver an appraisal was not an express warranty). This is consistent with the view that “[a]n express warranty is entirely a matter of contract, wherein the seller may define or limit his obligation respecting the subject of the sale, and provide as to the manner of fulfilling the warranty or the measure of damages for its breach.” Pabst Corp., 714 S.W.2d at 361 (quoting Donelson v. Fairmont Food Co., 252 S.W.2d 796, 799 (Tex.Civ.App.—Waco 1952, writ ref’d n.r.e.); accord Cravens, 626 S.W.2d at 176. We must determine, therefore, whether USFIC’s promise that it would “defend in [the Insured’s] name and behalf any suit against the Insured alleging damages, even if such suit is groundless, false or fraudulent” is, under Texas law, an express warranty. In deciding this, we are mindful both of our role as an Erie court and of the import of the question before us. It is “well-established that a breach of a contract occurs when a party fails or refuses to perform an act or thing that he has expressly or impliedly promised to do.” Fidelity & Deposit Co. v. Stool, 607 S.W.2d 17, 24 (Tex.Civ.App.—Tyler 1980, no writ) (citing cases). Because this case reaches us without evi *1373 dence of other conduct, however, we are faced with another question: Does a seller’s failure to deliver a service promised under a contract, without more, also constitute a breach of warranty? If it does, we are holding that USFIC’s breach of the contract for failure to deliver a promised service constitutes a violation of the DTPA. Such a ruling, however, would seem to place us in direct conflict with the Texas Supreme Court’s repeated admonition that “[a] mere breach of contract is not a violation of the DTPA.” La Sara Grain v. First Nat’l Bank, 673 S.W.2d 558, 565 (Tex.1984); accord Helms v. Southwestern Bell Telephone Co., 794 F.2d 188, 191 (5th Cir.1986); Dura-Wood Treating Co. v. Century Forest Indus., Inc., 675 F.2d 745, 756 (5th Cir.1982). We have noted before that “[t]he ‘more’ that is required to change a breach of contract action into a DTPA claim is not settled by Texas case law.” Helms, 794 F.2d at 191. In this case, therefore, we must endeavor to determine whether the requirement of “more” can properly be satisfied by designating the promise to perform as an express warranty. Cf. id. at 191 n. 3 (indicating that because the “mere breach of contract” rule arose in claims charging violations of the DTPA other than breach of warranty, the rule’s applicability in a warranty case might be questioned); but see Ashford Dev., Inc. v. USLife Real Estate Serv. Corp., 661 S.W.2d 933, 935 (Tex.1983) (applying “mere breach of contract” rule to DTPA claim without examining the section of the Act defendant was claimed to have violated). While the DTPA makes the breach of an express warranty actionable under the Act, “[t]he DTPA does not define the term ‘warranty.’ Furthermore, the act does not create any warranties; therefore any warranty must be established independently of the act.” La Sara Grain, 673 S.W.2d at 565; accord Melody Home Mfg. Co. v. Barnes, S.W.2d, 30 Tex.Sup.Ct.J. 489, 492 (June 17, 1987). The most accessible definition of “express warranty” appears in Chapter 2 of the Texas Uniform Commercial Code (“the Code”). See Tex.Bus. & Com. Code Ann. § 2.313 (Tex. UCC) (Vernon 1968). Chapter 2 of the Code, however, governs only the sale of goods. Id. § 2.102. Under the DTPA, an insurance policy is a service. McCrann v. Klaneckey, 667 S.W.2d 924, 926-27 (Tex.App.—Corpus Christi 1984, no writ); Dairyland County Mut. Ins. Co. v. Harrison, 578 S.W.2d 186, 190 (Tex.Civ.App.—Houston [1st Dist.] 1979, no writ); McNeill v. McDavid Ins. Agency, 594 S.W. 2d 198, 202 (Tex.Civ.App.—Fort Worth 1980, no writ). Unfortunately, we have been able to find no source which instructs us when, if ever, a seller “warrants” the service it has agreed to furnish. 12 We are not prepared to hold, however, that the dearth of law on express warranties in service contracts means that the law only recognizes express warranties when they arise in contracts for the sale of goods. See Melody Home, 30 Tex.Sup.Ct.J. at 492 (holding that section 17.45(2) of the Act expresses the legislative intent that providers of services should not escape the requirements of the Act). But the absence of law does, when coupled with the fact that we are faced with an insurance policy, complicate our analysis. 13 After reflection, *1374 however, we have concluded that for our purposes, the Code’s provisions on express warranties are instructive, even if they are not controlling. Therefore, we turn to these provisions. Section 2.313 of the Code details how express warranties are created in contracts for the sale of goods. Most significantly, the Code explains that “[a]ny affirmation of fact or promise made by the seller to the buyer which relates to the goods and becomes part of the basis of the bargain creates an express warranty that the goods shall conform to the affirmation or promise.” Tex.Bus. & Com. Code Ann. § 2.313(a)(1) (Tex. UCC) (Vernon 1968). Moreover, the Code admonishes that “[i]t is not necessary to the creation of an express warranty that the seller use formal words such as ‘warrant’ or ‘guarantee’ or that he have a specific intention to make a warranty.” Id. § 2.313(b). These broad descriptions of which promises create warranties appear, at first glance, to encompass almost every statement about the goods a seller makes to a buyer in connection with the sale. For example, taken alone, section 2.313 would find that a seller who promises to sell a buyer a new widget is liable for breach of warranty when he, instead, sells the buyer an old one. More important for our purposes, however, section 2.313 seems to hold that a seller who promises to deliver a new widget to a buyer is liable for breach of warranty when he fails to deliver anything at all. Consequently, it could be argued by analogy that a service provider breaches a warranty when it completely fails to provide the service it has promised. Section 2.313, however, is not the only Code section which defines the scope of express warranties. The scope of a warranty is also measured by how the warranty is breached and what remedies are available once a breach is found. Sections 2.711, 2.712, 2.713, and 2.714 of the Code determine a buyer’s remedies when the seller either fails to make delivery or delivers goods which do not conform to the promises the seller gave the buyer. James J. White and Robert S. Summers, in their textbook on the Uniform Commercial Code, explain the interrelationship between these provisions: [The] buyer’s rights under 2-712 (cover) and 2-713 (contract-market damages) [arise] when the seller fail[s] to deliver or when the buyer rightly reject[s] or revokes] his acceptance of delivered goods.... The principal Code section that governs the rights of a buyer who has accepted nonconforming goods is 2-714. We believe that only buyers who have accepted and neither rightfully rejected nor effectively revoked can use 2-714. Thus, in our view, sections 2-713 and 2-712 on the one hand and 2-714 on the other are mutually exclusive. J. White and R. Summers, Handbook of the Law Under the Uniform Commercial Code 375 (2d ed. 1980); see also [3A Sales & Bulk Transfers] R. Duesenberg & L. King, Bender’s U.C.C. Serv. (MB) ¶ 14.01[1] (May 5, 1985). According to White and Summers, therefore, the remedy sections of the Code distinguish between when the buyer has nothing at all and when the buyer has something, but it is not what *1375 was promised. In other words, the Code’s remedies distinguish based on delivery and acceptance of the goods. This distinction is important because, under the system the Code sets up, a buyer’s ability to sue for breach of warranty arises only under section 2.714, where acceptance of nonconforming goods is a prerequisite. See Tex. Bus. & Com. Code Ann. § 2.714 (Tex. UCC) (Vernon 1968) (titled “Buyer’s Damages for Breach in Regard to Accepted Goods’’) (emphasis added). Consequently, under the Code a breach of warranty does not occur when the seller completely fails to deliver. When that happens, the buyer’s remedy is found under sections 2.712 and 2.713 for the seller’s breach of contract. Moreover, the. Code retains this distinction between breach of contract and breach of warranty when dealing with installment contracts: when delivery is in separate lots, the failure to deliver one or more installments due under a contract, when the failure is significant enough to constitute a breach, gives rise to damages for breach of contract, not breach of warranty. Id. §§ 2.612, 2.711, 2.712, 2.713. When the broad definitional language of section 2.313 is read in light of the limitation on remedy found in section 2.714, it seems to us that a clearer understanding of what the Code envisions “express warranties” to be emerges. We believe that the Code’s focus is on the nature of the goods — their quality and description — and not on the fact of delivery; it would be incongruous for the Code’s authors to intend that the promise of delivery, by itself, be considered an express warranty and then to write the Code in a way which would never permit warranty damages for the seller’s failure to keep its promise. Moreover, since contracts for the sale of goods commonly contain promises that the goods will be delivered to the buyer, we cannot label as an oversight the inconsistency which would result from reading the definition of “express warranty” more broadly than its remedy. To us, by fashioning the remedy provisions as they did, the drafters of the Code demonstrated an intent to distinguish between promises which are merely terms of the contract and promises which rise to the level of warranties. We believe that the distinction between contract promises and express warranties established by the Code survives the transfer to the DTPA intact. We have not found any authority for the proposition that a buyer of goods has a cause of action under the DTPA for breach of express warranty when the buyer cannot bring the same cause of action under the Code. Moreover, we think that such a proposition would conflict with Texas Supreme Court decisions holding that the DTPA does not create warranties and that a breach of contract, without more, does not give rise to a cause of action under the DTPA. Consequently, we conclude that with respect to goods, a buyer is limited to bringing actions under section 17.50(a)(2) of the DTPA to situations where the buyer’s complaint is focused on the qualities of the goods received rather than on the fact that no goods were ever delivered. Moreover, because we see no reason why the sale of a service should receive more protection under the DTPA than the sale of a good, we hold that the same limitation applies to service contracts — the buyer has a cause of action under the DTPA for breach of an express warranty when the focus of the express promise is on the way in which the service is to be performed, but not when the focus is only on the fact that service is promised. 14 *1376 We make this ruling cognizant of the fact that the DTPA carries with it a mandate that the Act be liberally construed to promote its underlying purpose of “protecting consumers against breaches of warranty and providing efficient economical procedures to secure such protection.” Melody Home, 30 Tex.Sup.Ct.J. at 492; see Tex.Bus. & Com.Code Ann. § 17.44 (Vernon Supp.1987). We are, however, equally cognizant of our limited role as an Erie court in fashioning the parameters of that protection. As we have explained previously: In matters of Texas substantive law, our relationship to the Texas Supreme Court is all but identical to that of a Texas intermediate appellate court. Indeed, if it differs at all, as regards substantive innovation it is weaker instead of stronger than that of such a court. Even in the rare case where a course of Texas decisions permits us to extrapolate or predict with assurance where that law would be had it been declared, we should perhaps — being out of Texas jurisprudential development — be more chary of doing so than should an inferior state tribunal. Rhynes v. Branick Mfg. Corp., 629 F.2d 409, 410 (5th Cir.1980); accord Galindo v. Precision Am. Corp., 754 F.2d 1212, 1217 n. 8 (5th Cir.1985). In this case, we find ourselves face-to-face with the Texas Supreme Court’s admonition that we should not permit simple breach of contract cases to escalate into DTPA violations. We do not see how we can hold that a promise to perform constitutes an express warranty of performance without doing violence, in this and every other breach of contract case, to the supreme court’s directive. Consequently, we hold that the mere promise by an insurance company that it will defend is not an express warranty of performance. Instead, the promise of performance is simply a term of the contract which, when breached, gives rise to an action for contract damages. The district court, therefore, erred in finding that in addition to breaching its contract with the Lawyers, USFIC breached an express warranty by failing to perform. This holding does not, however, end our review of the matter before us. As explained earlier, the district court found that when USFIC refused to defend the Lawyers, it breached an “express or implied warranty.” In view of the disjunctive nature of the court’s finding, the court may or may not have found that USFIC breached an implied warranty. If it did, however, it did not articulate the nature of the implied warranty which it determined USFIC breached. 15 Moreover, the district court never conducted a trial or other evidentiary hearing where the evidence needed to establish this cause of action could be placed before the court. We will not imply, from a silent record, answers to all the questions which must be asked before the breach of a specific implied warranty can be found, especially when the unanswered questions in- *1377 elude both whether a warranty has been breached at all and, if one was, which warranty the focus is on. Therefore, we must remand this case to the district court for development of the implied warranty cause of action. 16 In addition, because the Lawyers have had no opportunity to develop the other aspects of their DTPA claim, we direct the district court to address, as necessary, the other violations of the Act which the Lawyers previously pleaded. Finally, because it may arise again on remand, it is appropriate for us to discuss briefly the question of which version of the DTPA governs the Lawyers’ claim against USFIC. 17 See Stallworth v. Monsanto Co., 558 F.2d 257, 268 (5th Cir.1977); United States v. Myers, 550 F.2d 1036, 1048 (5th Cir.1977), cert. denied, 439 U.S. 847, 99 S.Ct. 147, 58 L.Ed.2d 149 (1978). The district court concluded that the 1979 version of the DTPA applies to the Lawyers’ claims because, even though her first petition was filed before the effective date of the 1979 amendments, Clifton’s last amended petition was filed after the effective date. Consequently, the court concluded, the Lawyers’ cause of action against USFIC for its failure to defend them in the Clifton suit developed on the date USFIC last refused to defend the Lawyers—January 5, 1982. On cross-appeal, the Lawyers argue that the court erred in concluding that the 1982 date was controlling and, therefore, that the 1979 amendments to the Act apply to their claim. We agree. The 1979 amendments to the DTPA contain a “savings clause,” which provides that, “This act shall be applied prospectively only. Nothing in this act affects either procedurally or substantively a cause of action that arose either in whole or in part prior to the effective date of this act.” Act of June 13, 1979, ch. 603, § 9, 66th Leg., 1979 Tex.Gen.Laws 1327, 1332. The effective date of the 1979 amendments was August 27, 1979. Consequently, to decide which version of the DTPA applies to the Lawyers’ claim we must determine whether the Lawyers’ cause of action under the DTPA first accrued before or after August 27, 1979. Generally, a cause of action accrues on the date of the deceptive act or practice which gives rise to the complaint. La Sara Grain, 673 S.W.2d at 565; Padre Island Inv. Corp. v. Sorbera, 677 S.W.2d 90, 93 (Tex.App.—San Antonio 1984, writ dism’d w.o.j.). When the cause of action is for breach of warranty, the cause of action accrues when the warranty is breached. See La Sara Grain, 673 S.W.2d at 565; *1378 McAllen State Bank v. Linbeck Constr. Corp., 695 S.W.2d 10, 23-24 (Tex.App.—Corpus Christi 1985, writ ref'd n.r.e.). Because we do not know what the evidence on remand will uncover, we cannot determine exactly the date upon which the court should focus when deciding which version of the Act to apply. We can, however, point out that USFIC had a duty to defend from the time demand was made and received following filing of the first petition by Clifton on April 2, 1979. Moreover, it is undisputed that on April 10, 1979, USFIC unequivocally denied its duty to defend for the first time. We believe that since USF-IC’s duty to defend was ongoing, and since it is conceded that the changes in the subsequent petitions were immaterial to the duty, on remand the district court should not consider the dates on which the latter petitions were filed as relevant to resolving this question. Instead, if the district court finds that USFIC violated the DTPA, it should focus on the earliest point in time that USFIC took the action that the court finds violated the Act. III. For the foregoing reasons, the district court’s order granting the Lawyers’ summary judgment on their breach of contract claim is AFFIRMED. The court's judgment of October 6, 1986 as it relates to the damages which the Lawyers are entitled to recover for USFIC’s breach of contract is AFFIRMED. The remainder of the October 6, 1986 judgment, which awards the Lawyers damages under their Deceptive Trade Practices Act claim, is REVERSED and that claim is REMANDED to the district court for further proceedings consistent with this opinion. Each party shall bear its own costs associated with this appeal. Footnotes [1] At the same time the divorce action was proceeding, Harris and the Lawyers also agreed to help Clifton recover a gift tax refund from the Internal Revenue Service. To pay for these services, Clifton signed a contract agreeing to turn over to Harris and the Lawyers 25% of $369,000.00, the amount recovered from the IRS. In her subsequent suit to set aside the contingent fee contract and deed in the divorce proceedings, Clifton also sought to set aside the contractual agreement relating to the gift tax fee. However, Clifton’s basis for setting aside the gift tax contract was the same as the basis for setting aside the deed in the divorce proceeding; therefore, for the sake of convenience, this separate count will be treated as subsumed within Clifton’s complaint about the contingent fee agreement. [2] Clifton’s First Amended Original Petition was filed on June 24, 1980. Her Second Amended Original Petition was filed February 26, 1981, followed closely by her Third Amended Original Petition on March 13, 1981. Although it is not clear what became of the first amended petition, on April 10, 1981, the Lawyers forwarded both the second and third amended petitions to USF-IC and demanded a defense. However, they received no response to their demand. Clifton filed her Fourth Amended Original Petition on October 29, 1981, and the Lawyers forwarded it to USFIC, along with a demand for defense, on November 2, 1981. Again, USFIC failed to reply. Finally, on December 4, 1981, Clifton filed her Fifth Amended Original Petition, which the Lawyers sent to USFIC, along with their usual demand, on December 7, 1981. To this demand USFIC chose to reply; by letter dated January 5, 1982, USFIC reasserted its position that the exclusion provision relieved it of the duty to defend against Clifton’s allegations. As discussed later, the parties and the district court all agreed that Clifton's allegations remained the same from the first petition filed to the last. [3] USFIC sought summary judgment on all claims against it. However, the only argument USFIC asserted in support of its position was that the insurance policy, as a matter of law, did not require USFIC to provide the Lawyers with a defense. USFIC’s position in its motion, therefore, was that it did not breach the insur-anee contract. The Lawyers’ motion was for partial summary judgment only, and argued that, as a matter of law, the insurance policy required USFIC to provide the defense. Therefore, the Lawyers requested summary judgment on contract liability and trial on the question of damages. [4] As the Lawyers point out in their brief, the sum of $171,127.39 and $57,042.46 is not, as the judgment holds, equal to $221,232.35. It appears, therefore, that the district court added incorrectly in concluding that the Lawyers total recovery was $221,232.35 instead of $228,-169.85. However, because we are reversing the Lawyers’ recovery under the Deceptive Trade Practices Act, we find it unnecessary to correct the district court’s math. [5] As discussed earlier, the Lawyers’ coverage actually involved five separate insurance policies, each of which was in effect for a one-year period. The first four of the policies contained the language on which we focus for this opinion. However, in the fifth policy — which was effective from April, 1979 to April, 1980 — both the wording and form of the provisions are slightly modified in ways which generally do not affect the meaning of the provisions. One of the relevant clauses, however, is modified in a way which deserves comment. In the fifth policy, provision II relating to USFIC’s duty to defend reads: With respect to such insurance as is afforded by this policy, the Company shall defend any suit against the insured alleging such act or omission and seeking damages which are payable under the terms of this policy, even if any of the allegations of the suit are groundless, false or fraudulent. The italicized portion of this provision in the fifth policy does not appear in the provision’s counterpart in the first four policies. As we develop the basis upon which the district court found that USFIC had a duty to defend the Lawyers, USFIC’s reason for adding the new language will become clear. However, as we resolve the question, this change in language does not denote a change in USFIC’s responsibility under the policy; as will be discussed shortly, we hold that even under the first four policies, USFIC only had a duty to defend in situations where it would be responsible for paying a judgment, should one be returned against the insured. [6] The policy provision at issue in Donnelly read: The Company ... Agrees ... To pay on behalf of the insured all sums which the insured shall become legally obligated to pay as damages because of any act or omission of the insured ... and arising out of the performance of professional services for others in the insured’s capacity as a lawyer. ... With respect to such insurance as is afforded by this policy, the company shall: (a) defend any suit against the insured alleging such act or omission and seeking damages which are payable under the terms of this policy, even if any of the allegations of the suit are groundless, false or fraudulent.... This policy does not apply: (a) to any dishonest, fraudulent, criminal or malicious act or omission of any insured, partner or employee. Donnelly, 589 F.2d at 763 (emphasis added). The italicized portions above are the only differences between the Donnelly policy and the one before us. However, none of these differences are significant for purposes of our analysis of the policy. See supra note 5 and accompanying text. [7] In their briefs, the parties debate USFIC’s contention that several Texas cases have now recognized that extrinsic evidence may be addressed in determining the scope of the allegations in the petition. The argument was raised because USFIC attached to its motion for summary judgment below the affidavit of John B. Kyle (“Kyle"), the attorney who represented Clifton in her suit against the Lawyers. According to Kyle’s affidavit, “None of the six petitions concerned any allegations of negligence or error or omission as to professional work on the part of *1367 the Defendant lawyers." Instead, the causes of action Kyle intended to plead included "(a) fraud (b) unconscionable fee arrangements (c) breach of fiduciary attorney/client relationship (d) duress, and (e) concurrent Deceptive Trade Practice Act violations.” Record at 96. The cases USFIC cites in support of its position that we should consider Kyle’s affidavit are not Texas Supreme Court cases. See Gonzales v. Am. States Ins. Co., 628 S.W.2d 184 (Tex.App.—Corpus Christi 1982, no writ); Cook v. Ohio Casualty Ins. Co., 418 S.W.2d 712 (Tex.Civ.App.—Texar-kana 1967, no writ). However, we see no need to determine whether these cases adequately distinguish Heyden Newport Chemical Corp. v. Southern Gen. Ins. Co., 387 S.W.2d 22 (Tex.1965), the Texas Supreme Court decision from which we have drawn our rule that extrinsic evidence cannot be considered. As we explain above, we believe that even the allegations which Kyle admits to pleading were sufficient to impose on USFIC the duty to provide the Lawyers with a defense. [8] We add this observation only because it seems to us that the same analysis we employ to conclude that Clifton's allegations of constructive fraud were not clearly disclaimed by the exclusion provision, see text infra, supports a finding that Clifton’s allegations of unconscionability are also not clearly outside the protections which the policy affords. Unconscionability based on the finding of gross disparity between consideration paid and services received does not require a finding of wrongful intent. Franks v. Associated Air Center, Inc., 663 F.2d 583, 592-93 (5th Cir. Dec. 1981). [9] With respect to the breach of contract claim, USFIC also argues on appeal that the district court erred in calculating the award of damages which were recoverable because of the breach. As discussed earlier, along with the Lawyers, an attorney named Harris represented Clifton in her divorce. When Clifton sued to avoid the contingent fee contract, she also sued Harris and his lawfirm. After USFIC refused to defend the Lawyers, the Lawyers and Harris were both represented in the Clifton suit by the same counsel. USFIC’s argument is that the district court improperly permitted recovery for attorneys’ fees which were incurred readying Harris’ case for trial in addition to those fees spent on the Lawyers’ defense. This allegation has its basis solely in the following colloquy which took place during the hearing on damages: Q. (by counsel for USFIC): Judge Street [the Lawyers’ counsel during the Clifton suit], I just wanted to ask briefly about your mention of your representation of gentlemen by the name of Harris and Sullivan at the same time? A. Yes. Q. They were co-defendants, is that correct, with Mr. Tarlton and his partners [the Lawyers]? A. Yes. Q. And did you begin to represent Mr. Harris and Mr. Sullivan at the outset of the 17th District Court case? A. Yes. Q. Were they also obligated to pay your attorney’s fees? A. They were obligated to pay me attorney’s fees, yes. Q. Did they pay attorney's fees also? A. No, Sir. I have not been paid any attorney’s fees by anybody so far. Q. During the time that you were representing Mr. Tarlton and his partners, is there any way to differentiate between the work that you did for Mr. Harris and Mr. Sullivan as opposed to the work that you were doing for Mr. Tarlton and his partners? A. Well, when people are sued as co-defendants and they are trying to get a joint and severed judgment against all of them, well, you would hope that anything you did in defense of the lawsuit would benefit everybody on the defense side. That's your hope. This exchange takes up approximately one and one-half pages of testimony from a seventy-six page hearing record. Prior to this exchange, USFIC filed no pleadings asserting the argument they have fashioned from this testimony. During the extensive summations on the issue of damages in which both parties engaged after the witnesses finished testifying, USFIC made no mention of the testimony’s relevancy. In its “Requested Conclusions of Law" which were filed after the hearing, USFIC submitted no conclusion raising the point. In the letter briefs submitted to the district court during its deliberation of the damages claim, USFIC made no mention of the argument. And finally, after learning of the judgment entered by the court concerning damages, USFIC made no move to have the court consider the issue. Even viewing the evidence in the light most favorable to USF-IC, as our sister circuit remarked under similar circumstances, "we find it difficult to see how the district court could have considered this argument.” Laketon Asphalt Refining, Inc. v. *1371 United States Dep’t of the Interior, 624 F.2d 784, 788 (7th Cir.1980). Therefore, "[a]bsent plain error we [will] not reverse the district court on an issue not clearly presented to it for decision.” Nathan Rodger Constr. & Realty Corp. v. City of Saraland, 676 F.2d 162, 163 (Former Sth Cir.1982). After examining the briefs, we find no plain error. Consequently, we decline to address USFIC’s argument. [10] The Lawyers actually pleaded only that USF-IC’s refusal to defend constituted a violation of "Sections 17.46(a), Section 17.46(b)(5) and (12), as well as Section 17.50(a)(1) through (4)’’ of the Act. Our list of their claims is the result of synthesizing the sections cited in the manner contemplated by the Act. Consequently, claim (1) is the combination of sections 17.46(a), (b)(5) and 17.50(a)(1); claim (2) is the combination of sections 17.46(a), (b)(12) and 17.50(a)(1); claim (3) is derived from section 17.50(a)(2); claim (4) is derived from section 17.50(a)(3); and claim (5) is derived from section 17.-50(a)(4). See generally Tex.Bus. & Com. Code Ann. §§ 17.46, 17.50 (Vernon Supp.1987). [11] This conclusion comports with the apparent view of the parties and the court, as evidenced by their actions from the time of the damages hearing. During the summation arguments, the court engaged counsel for the Lawyers in the following conversation: *1372 Mr. Barlow: I am going to briefly address the Court because the Court has caught the gut issue in this case as of November 9, 1984 in which the Court has already held that the defendant wrongfully refused the defense. Now, the question before the Court is what damages do I now assess, and what am I mandated to assess under the Texas Deceptive Trade Practice Act, and has the plaintiff proved that the cause of action under the Texas Deceptive Trade Practice Act because you already held in your ruling of November 9, 1984 that we have a breach of contract issue there already. The Court: But that doesn’t necessarily mean it’s deceptive trade practice— Mr. Barlow: That’s what I have to argue to you. I agree with you, Your Honor. The Court: In other words, I haven’t [sic] found that they had a duty to represent. I have not ruled that the Deceptive Trade Practice Act is applicable. Mr. Barlow: No, you have not, and that’s what I am about to persuade you, Your Hon- or. That’s what I was saying. We had a breach of contract already established and the attorney’s fees that flow from any breach of contract under Texas law, and now it’s my burden to show you that this is a Deceptive Trade Practice Act. After this discussion, Mr. Barlow argued to the court that the Lawyers had "proved” every element of the DTPA claim. He then summarized the "proof' he had presented. His summary included: (1) that the DTPA is to be liberally construed by the court; (2) that USFIC, as evidenced by the correspondence between the parties, contended that it did not have the obligation to defend; and (3) that the Lawyers incurred damages as a result of USFIC’s actions. After Mr. Barlow finished his summation, counsel for USFIC began his summation by explaining: I think it is well established that this is a breach of contract case. There is an insurance policy that has two parties, an insured and a carrier. It has a contract that provides certain things be done, one of which is to furnish a defense. Defense was not furnished. The insured lawyers then had an option to sue for specific performance or to sue for breach of contract and seek damages.... [A] breach of contract case is not one to which the Deceptive Trade Practice Act applies. Mr. Barlow responded to this argument by telling the court, "I can, likewise, cite for the Court with a few minutes’ research cases in which breach of insurance contracts have been held to be deceptive trade practices." He also explained that ”[w]e know that the Deceptive Trade Practice Act refers to the purchase and sale of goods and services. By [USFIC's] definition, all of those are contracts, and the fool’s definition would be that they are all excluded.” [12] For example, in his treatise on contracts, Professor Williston does not mention the possibility that express warranties can be found in service contracts. See generally S. Williston, A Treatise on the Law of Contracts (3d ed. 1957 and Supp. 1987). In his own treatise, neither does Professor Corbin. See generally A. Corbin, Corbin on Contracts (one volume ed. 1952). Moreover, a review of other jurisdictions demonstrates that courts may go to great lengths to avoid recognizing the existence of an express warranty in a service contract. See, e.g., McCarty v. E.J. Korvette, Inc., 28 Md.App. 421, 347 A.2d 253, 18 UCC Rep.Serv. 14 (1975). In those cases we have found where courts are apparently willing to recognize the idea of express warranties in service contracts, no definition of "express warranty" is given and the warranty which is recognized always relates to promises about the manner of performance. See, e.g., Noel v. Proud, 189 Kan. 6, 367 P.2d 61 (1961); King v. Ohio Valley Terminix Co., 309 Ky. 35, 214 S.W.2d 993 (1948); Sullivan v. O'Conner, 363 Mass. 579, 296 N.E.2d 183 (1973); Sidney Stevens Implement Co. v. Hintze, 92 Utah 264, 67 P.2d 632 (1937). [13] There is a line of cases which discuss what constitutes an express warranty by the insured, as opposed to the insurer, in an insurance policy. These cases hold that "for a statement in a *1374 policy to constitute a warranty, the parties must have intended that the policy stand or fall on the literal truth or the falsity of the statement.” Lane v. Travelers Indemnity Co., 391 S.W.2d 399, 402 (Tex.1965); accord Old. Reliable Fire Ins. Co. v. Alduro-Raynes Arabians, Inc., 717 S.W.2d 124, 127 (Tex.App.—Houston [14th Dist.] 1986, writ ref'd n.r.e.); Cartusciello v. Allied Life Ins. Co., 661 S.W.2d 285, 286 (Tex.App.—Houston [1st Dist.] 1983, no writ). The question before the court in these cases was whether false statements which the insured made in connection with obtaining insurance are to be construed as misrepresentations or warranties. The distinction between the two is significant. If the statement is found to be a warranty, the policy is treated as never having been in effect. However, if the statement is only a misrepresentation, the insurer must show that the misrepresentation was material before it can deny policy coverage. See Lane, 391 S.W.2d at 400-01. The reason for the restrictive definition of "warranty,” therefore, is to avoid forfeitures, which are disfavored in law. Id. at 402; Cartusciello, 661 S.W.2d at 286. These concerns are inapposite when it is the insurer who is making the promise, since a forfeiture will not occur regardless of what the promise is called. Therefore, we find it necessary to examine other definitions of "warranty” which are more relevant to the issue before us. [14] We recognize that under our holding a seller would seem to be better off not even trying to perform than he would be if he tried to perform and did so inadequately. This is because by attempting performance, the seller ostensibly risks DTPA damages; however, by not performing at all, the seller is shielded from DTPA liability. To the extent that this anomaly exists, we believe that it was created by the combined statutory schemes of the Code and the DTPA, not by us. However, we do not believe that the reasoning which leads to this anomalous conclusion is sound. Our holding today is simply that the mere promise to perform does not constitute an express warranty. Therefore, nonperformance by the seller does not result in the breach of an express warranty. If all that the seller promises is performance, however, the result is not changed just because the seller attempts to perform. Unless the seller made *1376 additional promises concerning the manner of performance — as opposed to the fact of performance — it does not matter what action the seller takes. Not performing at all, or performing incorrectly — neither leads to liability under the DTPA for breach of an express warranty because the seller has made no express warranty which can be breached. This is not to say that either conduct cannot, when coupled with other evidence, otherwise prove to support DTPA liability. However, because we are not factfinders, we cannot determine either the amount or nature of additional evidence which, when coupled with the fact of nonperformance, will prove a violation. Therefore, we make no comment on the use of the fact of nonperformance in proving a violation of the DTPA other than by breach of a warranty. [15] In their DTPA demand letter to USFIC, the Lawyers charged USFIC with violating an implied warranty that it would use "good faith and fair dealing in connection with the duties which [USFIC] owed [the Lawyers]” under the insurance policy. The Texas Supreme Court has recently recognized that, because of the fiduciary nature of the relationship between an insured and its insurance company, "there is a duty on the part of insurers to deal fairly and in good faith with their insureds.” Arnold v. National County Mut. Fire Ins. Co., 725 S.W.2d 165, 167 (Tex.1987). The Texas Supreme Court gave no indication, however, that this duty of "good faith and fair dealing” is, in fact, an implied warranty for the breach of which recovery can be had under the DTPA. Instead, the supreme court appears to have characterized the cause of action as a simple common law tort. [16] We note that under the Code, section 2.714 limits recovery for both express and implied warranties to those situations where the buyer has received the goods. Consequently, the analysis we used to determine the Code’s concept of express warranties is equally applicable to understanding implied warranties which arise under the Code. The implied warranties which the Code specifically recognizes bear out the distinction we have made between promises which are simply terms of the contract and promises which rise to the level of warranties. Both the warranty of fitness for a particular purpose and merchantability focus on the nature of the goods received by the buyer. See Tex.Bus. & Com. Code Ann. § 2.314, 2.315 (Tex. UCC) (Vernon 1968). Recently, the Texas Supreme Court found that an implied warranty exists in service contracts, with respect to “services such as repairs,” that the service provider will perform its service in a good and workmanlike manner. Melody Home, 350 Tex.Sup.Ct.J. at 491. Even this new implied warranty, however, arose in a case where the service provider had clearly performed under the contract and the focus of the complaint was on the manner of performance. See id. Because there is no indication from the record before us what implied warranty, if any, the district court was focused on, we make no comment on whether the implied warranty to perform in a good and workmanlike manner may extend farther than the facts of Melody Home suggest. [17] This issue is relevant because deciding which version of the Act applies also decides whether the Lawyers must be permitted to recover treble damages if the court finds that USFIC violated the Act. Prior to the 1979 amendments to the DTPA, trebling of any actual damages was mandatory. Jim Walter Homes, Inc. v. Valencia, 690 S.W.2d 239, 241 (Tex.1985); Woods v. Littleton, 554 S.W.2d 662, 669 (Tex.1977). Under the 1979 amendments, however, only the first $1000 of an actual damage award is automatically trebled. Any additional award of damages is at the trial court’s discretion. However, the trial court cannot award these additional damages at all unless the deceptive act or practice which is the subject of the suit is found to have been committed knowingly. Tex.Bus. & Com.Code Ann. § 17.50(b)(1) (Vernon Supp.1987); Jim Walter Homes, 690 S.W.2d at 241.