OPINION

RANSOM, Justice.

Brenda Hansen was injured in January 1990 when the car she was driving collided with a car driven by Della Irene Pease. In settlement of her claims against Pease, Hansen executed a general release in April 1991. Two years later Hansen sued Ford Motor Company, Al Allred Ford Lincoln-Mercury, Inc., and TRW, Inc. (collectively, "Ford"), claiming that the air bag in her automobile malfunctioned during the collision, causing her injury. Ford moved for summary judgment, claiming that the general release signed by Hansen released Ford from all liability arising from the accident.

Ruling that the release unambiguously includes Ford, the trial court entered summary judgment. Hansen appeals to this Court pursuant to SCRA 1986, Section 12-102(A)(1) (Repl.Pamp. 1992) (count sounding in contract). We question whether under principles of traditional contract law the terms of a general release facially and unambiguously include third parties who are not specifically identified as beneficiaries. The better policy, in any event, is to recognize an inherent circumstantial ambiguity as to the intentions of the parties to a general release. Because boilerplate language purporting to discharge all persons whose conduct may have contributed to Hansen's injuries is not sufficiently specific to identify Ford as a third-party beneficiary of the release, we reverse and remand for a determination of the parties' intentions under a rebuttable presumption that the release benefits only those persons specifically designated.

Facts and proceedings. The release in question — a standard form which includes blanks for entering the amount of consideration paid, the names of releasees, and the date and location of the accident — provides:

For the Sole Consideration of [$29,000] to be paid the undersigned hereby releases and forever discharges Paul M. Pease, Della Irene Pease, and American National Property and Casualty Company[,] their heirs, executors, administrators, agents and assigns, and all other persons, firms or corporations liable or, who might be claimed to be liable . . . from any and all claims, demands, damages, actions, causes of action or suits . . . on account of all injuries . . . which have resulted or may in the future develop from an accident which occurred on or about the 31st day of January. . . .

. . . .

Undersigned hereby declares that the terms of this settlement have been completely read and are fully understood and voluntarily accepted . . . for the express purpose of precluding forever any further or additional claims arising out of the aforesaid accident. (Emphasis added.)

Hansen has conceded that she had seen releases of this type before and routinely used them as part of her job as an insurance claims adjuster for State Farm. She argues that the release is ambiguous because it is "reasonably and fairly susceptible of different constructions" and that she in fact had intended to release only the persons specifically named in the release — Paul Pease, Della Pease, and the Peases' insurance company, American National Property and Casualty Company. Hansen proffered her deposition testimony as to her intent and also requested the court to hold an evidentiary hearing to determine the intent of the parties to the release. Ford countered that Hansen's products-liability claims arose out of the accident and that it was thus included within the language of the general release. Relying on Hendren v. Allstate Insurance Co., 100 N.M. 506, 672 P.2d 1137 (Ct.App. 1983), the trial court stated that it was "convinced that the law in New Mexico declares that the general release signed by Plaintiff . . . necessarily includes all Defendants in the case at bar."

Hansen appeals the entry of summary judgment, citing four grounds for reversal: the trial court erroneously applied the "four-corners" standard of contract interpretation; the terms of the release are ambiguous, and the trial court should have held an evidentiary hearing to determine the intent of the parties; the release should be rescinded or reformed because there was a mutual mistake; and the release should be rescinded because there was no "meeting of the minds."

Burden is on third party who neither negotiated nor gave consideration for release to prove that it was an intended beneficiary of the release. Ford seeks to be discharged as a third-party beneficiary of the general release. It is undisputed Ford neither took part in the settlement negotiations that culminated in the execution of this release nor contributed any money to the consideration paid for the release. Hansen contends that Ford had the burden to prove that she and the Peases intended by the release to discharge Ford from liability.

In Hoge v. Farmers Market Supply Co., 61 N.M. 138, 143, 296 P.2d 476, 479 (1956), this Court agreed that "one claiming to be a third party beneficiary of an agreement made by others has the burden of proving that he was intended by the makers of the agreement to be such beneficiary." A prospective third-party beneficiary may prove the intent of the parties to an agreement by relying on the unambiguous language of the agreement itself, Valdez v. Cillessen Son, Inc., 105 N.M. 575, 581, 734 P.2d 1258, 1264 (1987), or, in the absence of such language, on extrinsic evidence such as the circumstances surrounding the execution of the agreement, Permian Basin Inv. Corp. v. Lloyd, 63 N.M. 1, 7, 312 P.2d 533, 537 (1957).

A number of courts have applied these general principles of contract law in cases involving the scope of a general release. See, e.g., Neves v. Potter, 769 P.2d 1047, 1053-54 (Colo. 1989) (en banc) (holding that parol evidence is admissible to determine intent of parties); Harris v. Grizzle, 599 P.2d 580, 586 (Wyo. 1979) (allocating burden of proof to prospective third-party beneficiaries). We agree that "alleged wrongdoers who were not parties to the release and made no payment toward satisfaction can fairly be called on to show that . . . the release which they rely on was intended to discharge them." Harris, 599 P.2d at 586. Hence, consistent with well-established principles applicable to contracts in general, and releases in particular, Ford had the burden to prove that it was an intended beneficiary of the release executed by Hansen and the Peases.

An inherent ambiguity in any general release is recognized as a matter of policy. Hansen asserts that the trial court committed error by refusing to consider parol evidence to determine whether the terms of the release were ambiguous. Ford counters that the trial court did consider parol evidence but that the evidence adduced was insufficient to show an ambiguity, grounds to void the release, or grounds to reform the release. Hence, argues Ford, the court properly entered summary judgment.

Under previous New Mexico decisions, releases have been interpreted using general contract law principles. See, e.g., Ratzlaff v. Seven Bar Flying Serv., Inc., 98 N.M. 159, 162, 646 P.2d 586, 589 (Ct.App.), cert. denied, 98 N.M. 336, 648 P.2d 794 (1982). Hansen thus correctly asserts that under this Court's decisions in Mark V, Inc. v. Mellekas, 114 N.M. 778, 781, 845 P.2d 1232, 1235 (1993), and C.R. Anthony Co. v. Loretto Mall Partners, 112 N.M. 504, 508-09, 817 P.2d 238, 242-43 (1991), the trial court could consider extrinsic evidence to determine whether the facially unambiguous terms of the release are in fact ambiguous. Hansen incorrectly asserts, however, that her deposition testimony established an ambiguity and that it was thus error for the court not to consider it.

On the third-party-beneficiary issue — whether the general release executed by Hansen and the Peases inures to the benefit of Ford — the intent of the parties controls. Neves, 769 P.2d at 1053-55. Hansen presented evidence that discloses her subjective understanding that the release did not include Ford. Nothing indicates that the Peases thought so or knew this to be Hansen's understanding. A "unilateral, subjective intent not to include" third parties as beneficiaries of the agreement is not determinative. Jaramillo v. Providence Washington Ins. Co., 117 N.M. 337, 342, 871 P.2d 1343, 1348 (1994). From the deposition testimony presented by Hansen, we are able to discern nothing new about the meaning attributed to the language of the release by the Peases and their insurance company.

Because Hansen's deposition discloses only her unilateral, subjective intent, it was insufficient in itself to establish an ambiguity in the terms of the release. For similar reasons, Hansen's testimony was insufficient to establish a mutual mistake. Finally, Hansen did not adduce any evidence of fraud or overreaching that would establish grounds to void the release. Thus, even if the trial court refused to consider Hansen's deposition testimony, we would not reverse the entry of summary judgment on this ground. Tsosie v. Foundation Reserve Ins. Co, 77 N.M. 671, 676, 427 P.2d 29, 32 (1967) (stating that court will not be reversed when it reaches right result for wrong reason).

Although Hansen's deposition testimony standing alone was insufficient to overcome the apparent import of the "arising out of the accident" language, for reasons of policy discussed below we conclude that boilerplate universal release language such as that used here is circumstantially ambiguous. The best way of determining and enforcing the actual intent of the parties expressed in boilerplate language is to adopt a rebuttable presumption that a general release benefits only those persons specifically designated in the release document.

There is a rebuttable presumption that a general release benefits only those persons specifically designated. As noted recently by our Court of Appeals, there are three approaches to determining the scope of a general release. See Perea v. Snyder, 117 N.M. 774, 778, 877 P.2d 580, 584 (Ct.App.), cert. denied, 118 N.M. 90, 879 P.2d 91 (1994). These three approaches are known as the absolute or flat bar rule, the intent rule, and the specific identity rule. See id. at 778-79, 877 P.2d at 584-85 (discussing the three approaches). In Perea the Court of Appeals reasoned that our decisions in Mark V and C.R. Anthony indicated New Mexico should follow the intent rule. The Court went on to hold that "an absolute bar will be presumed to be the intent of the parties to a general release unless an ambiguity is shown to exist by extrinsic evidence." Id. at 779, 877 P.2d at 585.

Ford relies extensively on the Court of Appeals' analysis in Perea to support the trial court's entry of summary judgment. Hansen replies that the district court's decision is erroneous under Perea. Because the arguments in support of and against the Court of Appeals' reasoning and conclusions have been fully briefed, we take this opportunity to review the Perea decision.

Amicus New Mexico Trial Lawyers Association argues that the formulation of the intent rule in Perea is closer to the "flat bar" rule than the "intent of the parties" rule formulated in other jurisdictions. The Trial Lawyers ask this Court to adopt a "specific identity" rule. In particular, they urge this Court to adopt as a matter of public policy a rule of release interpretation under which every general release will be deemed to contain a term implied in law limiting the effect of general language. Under this rule, a releasee could obtain the release of other possible tortfeasors only by specifically naming those tortfeasors or by using some other specific identifying terminology. Because no such language was included in this release, under the specific identity rule Hansen's release does not bar her claims against Ford.

Ford counters that adoption of a specific identity rule ignores the unambiguous language of the release and ignores the intentions of the parties. Ford also argues that adoption of the specific identity rule would deprive the settling defendant of important rights, including contribution rights and the right to be free from further litigation. Finally, Ford argues that Perea correctly concluded that releases should be interpreted according to general contract law principles and thus urges us to affirm the trial court's entry of summary judgment.

Those courts adopting the flat bar rule hold that language such as "all other persons, firms or corporations liable" is unambiguous and discharges all potential tort-feasors from liability. See, e.g., Battle v. Clanton, 27 N.C. App. 616, 220 S.E.2d 97, 99 (1975), cert. denied, 289 N.C. 613, 223 S.E.2d 391 (1976); Hasselrode v. Gnagey, 404 Pa. 549, 172 A.2d 764, 765 (1961). The parties to a general release containing "all other persons" language are deemed as a matter of law to have expressed their intent to discharge all potential tortfeasors. Flat bar courts thus look only to the four corners of the release document and do not allow consideration of extrinsic evidence.

Jurisdictions adopting the intent rule have developed two formulations with similar purposes. Some jurisdictions hold that the parol evidence rule is inapplicable in an action by a party to a release and a stranger to that agreement. See, e.g., Neves, 769 P.2d at 1054; Sims v. Honda Motor Co., 225 Conn. 401, 623 A.2d 995, 1003 (1993). Under this formulation of the intent rule, parol evidence of the parties' intentions is admissible even when the terms of the release are facially unambiguous. See, e.g., Sims, 623 A.2d at 1004 n. 12. Other jurisdictions hold that extrinsic evidence of the parties intent is admissible only when the court determines as a matter of law that the terms of the release agreement are ambiguous. See, e.g., Wells v. Shearson Lehman/Am. Express, Inc., 72 N.Y.2d 11, 530 N.Y.S.2d 517, 523-24, 526 N.E.2d 8, 14-15 (1988); Krauss v. Utah State Dep't of Transp., 852 P.2d 1014, 1019-20 (Utah Ct.App.), cert. denied, 862 P.2d 1356 (Utah 1993).

Jurisdictions adopting the specific identity rule conclusively presume that the liability of a party not named or otherwise specifically identified by the terms of the release is not discharged. The specific identity rule compels the settling parties either to name nonsettling tortfeasors, see, e.g., Young v. State, 455 P.2d 889, 893 (Alaska 1969), or to include such terms or descriptions as make the identity of the unnamed beneficiaries or the class thereof reasonably apparent, see, e.g., Aid Ins. Co. v. Davis County, 426 N.W.2d 631, 633 (Iowa 1988) (holding that terms identifying persons "in a manner that the parties to the release would know who was to be benefitted" sufficient under specific identity rule); Pakulski v. Garber, 6 Ohio St.3d 252, 452 N.E.2d 1300, 1303 (1983) (holding that release which discharged agents and employees sufficiently identified law firm and individual member thereof who had represented specifically named releasees).

A distinct minority of jurisdictions continues to adhere to the flat bar rule. As between the remaining two approaches — the intent rule and the specific identity rule — the intent rule has been adopted by a bare majority of courts. See, e.g., McInnis v. Harley-Davidson Motor Co., 625 F. Supp. 943, 956-57 (D.R.I. 1986) (predicting Rhode Island would adopt intent rule); General Motors Corp. v. Superior Court, 12 Cal.App.4th 435, 15 Cal.Rptr.2d 622, 625-27 (1993) (applying traditional rules of contract construction and determining that terms of release were facially unambiguous); Neves, 769 P.2d at 1053 (holding that extrinsic evidence admissible to determine intent of parties); Sims, 623 A.2d at 1001-02 (holding that extrinsic evidence of intent must be considered); Hurt v. Leather-by Ins. Co., 380 So.2d 432, 433-34 (Fla. 1980) (adopting intent rule and holding that release containing handwritten and preprinted terms was inherently ambiguous); Cram v. Town of Northbridge, 410 Mass. 800, 575 N.E.2d 747, 749 (1991) (reversing summary judgment because accident victim filed an affidavit in which she stated she intended to discharge only the driver who injured her); Wells, 530 N.Y.S.2d at 521, 526 N.E.2d at 12 (holding that court must determine intent of parties by reference to language of release, turning to extrinsic evidence only when it determines as a matter of law that terms are ambiguous); Krauss, 852 P.2d at 1019-20 (same).

A significant number of courts, however, have opted for the specific identity rule. See, e.g., Young, 455 P.2d at 893 (holding that release discharges only those tortfeasors specifically named in the release agreement); Moore v. Missouri Pac. R.R., 299 Ark. 232, 773 S.W.2d 78, 82 (1989) (holding that release must specifically name or otherwise specifically identify the persons to be discharged); Lackey v. McDowell, 262 Ga. 185, 415 S.E.2d 902, 903 n. 3 (1992) (noting that rule that only specifically identified tortfeasors are released eliminates need to inquire into intent); Saranillio v. Silva, 78 Haw. 1, 889 P.2d 685, 700 (1995) (holding that release discharges only persons named in or sufficiently described by terms of release); Alsup v. Firestone Tire Rubber Co., 101 Ill.2d 196, 77 Ill.Dec. 738, 740-41, 461 N.E.2d 361, 363-64 (1984) (holding that in order to fulfill legislative intent to abolish common-law release rule, general release must specifically identify parties to be released); Aid Ins. Co., 426 N.W.2d at 633-34 (holding that release must identify persons in such a manner that parties to release would know who was to be benefitted); Beck v. Cianchetti, I Ohio St.3d 231, 439 N.E.2d 417, 420 (1982) (holding that general release must name or specifically describe parties to be discharged); McMillen v. Klingensmith, 467 S.W.2d 193, 196 (Tex. 1971) (holding that parties not specifically identified are not released); Bjork v. Chrysler Corp., 702 P.2d 146, 162-63 (Wyo. 1985) (holding that release must specifically identify persons to be discharged in order to ensure that intent of parties is fulfilled).

In deciding which of these three approaches we should adopt, we are not obligated to choose the position that has garnered the support of more courts. Rather, we will adopt the position that reason and logic counsel is the better choice under the law of New Mexico. Griego v. New York Life Ins. Co., 44 N.M. 330, 338, 102 P.2d 31, 36 (1940) (noting that goal of court is to make choice compatible with justice, reason, and logic even if the choice represents a minority view).

Support for a flat bar has its roots in the common-law rule that the release of one joint tortfeasor discharges other joint tortfeasors as a matter of law. See, e.g., Young, 455 P.2d at 891; McMillen, 467 S.W.2d at 195. The "unity of discharge" rule, as it is known, was based in turn on the notion that in cases involving joint tortfeasors, the plaintiff had suffered a single and indivisible wrong, the damages for which could not be apportioned among defendants. Missouri, K. T. Ry. v. McWherter, 59 Kan. 345, 53 P. 135, 137 (1898); Muse v. De Vito, 243 Mass. 384, 137 N.E. 730, 731 (1923); McBride v. Scott, 132 Mich. 176, 93 N.W. 243, 245 (1903). Under early American rules of joinder, a plaintiff could join all claims against concurrent tort-feasors in a single suit. See W. Page Keeton et al., Prosser and Keeton on the Law of Torts § 47, at 326 (5th ed. 1984). Because concurrent tortfeasors could be joined in the same suit they came to be called joint tort-feasors. William L. Prosser, Joint Torts and Several Liability, 25 Cal.L.Rev. 413, 419-21 (1937). In cases involving such tortfeasors, because the plaintiff was considered to have suffered a single wrong, and because defendants were considered joint tortfeasors entirely liable for a single harm, once plaintiff had surrendered his or her cause of action in favor of one of the joint tortfeasors, he or she was conclusively presumed to have surrendered it as against all others. See, e.g., Whitt v. Hutchison, 72 O.O.2d 30, 43 Ohio St.2d 53, 330 N.E.2d 678, 681 (1975) (discussing bases for common-law release rule).

In Bartlett v. New Mexico Welding Supply, Inc., 98 N.M. 152, 158, 646 P.2d 579, 585 (Ct.App.), cert. denied, 98 N.M. 336, 648 P.2d 794 (1982), our Court of Appeals held that "[j]oint and several liability is not to be retained in our pure comparative negligence system on a theory of one indivisible wrong," reasoning "[t]he concept of one indivisible wrong . . . is obsolete." Id. As recognized in Bartlett, with the adoption of comparative fault came the ability to divide a previously indivisible injury. Id. at 159, 646 P.2d at 586. Damages for a singular wrong to the plaintiff could be apportioned among the wrongdoers according to fault.

The notion that a person injured by concurrent tortfeasors has suffered an indivisible wrong that gives rise to a single cause of action is erroneous for the reasons stated in Bartlett and its progeny. Further, giving effect only to the abstractly unambiguous words of an agreement without regard to the actual intent of the parties elevates form over substance. A tortfeasor who has taken no part in the satisfaction of a plaintiff's claim should not gratuitously benefit from settlement arrangements undertaken at the time and expense of others. Only when the parties to a release intend to discharge the liability of others not parties thereto should the release be given such effect. Interpreting a release by looking only to the four corners of the document poses too great a risk that a plaintiff will be trapped into surrendering a separate cause of action when this was not his or her intent. We thus join the majority of jurisdictions in rejecting the flat bar rule.

Most courts that require "specific identity" have relied as justification for the rule, at least partially, on legislative adoption in their state of some form of the Uniform Contribution Among Tortfeasors Act, 12 U.L.A. 57 (1975). See, e.g., Moore, 773 S.W.2d at 81-82; Saranillio, 889 P.2d at 700; Alsup, 461 N.E.2d at 364; Aid, 426 N.W.2d at 634; Beck, 439 N.E.2d at 420; Bjork, 702 P.2d at 161-62. Section 4 of the Uniform Act governs the effect of a release in cases involving joint tortfeasors. See, e.g., NMSA 1978, § 41-3-4 (Repl.Pamp. 1989). Under that Section, "[a] release by the injured person of one joint tortfeasor . . . does not discharge the other tortfeasors unless the release so provides." Id. Those courts adopting the specific identity rule have noted initially that the purpose behind the Uniform Act was to abrogate the common-law release rule. See, e.g., Alsup, 461 N.E.2d at 363; Beck, 439 N.E.2d at 420. This Court has noted a similar purpose. See Herrera v. Uhl, 80 N.M. 140, 141, 452 P.2d 474, 475 (1969). In order to ensure that this legislative purpose is satisfied and to eliminate the harsh, unwarranted effects of the common-law release rule, some courts have concluded that it is necessary to construe narrowly the statutory language "unless the release so provides." See, e.g., Beck, 439 N.E.2d at 420 ("The thrust of [Section 4] is to retain the liability of tortfeasors and, thus, the phrase 'unless its terms otherwise provide' should be narrowly construed and require a degree of specificity."); Bjork, 702 P.2d at 162-63 ("To permit discharge of noncontributing tortfeasors . . . based upon broad, general language which does not identify the tortfeasors effectively perpetuates the [common-law rule] in contravention of the legislature's intention.").

Courts rejecting the specific identity rule in favor of the intent rule have reasoned that the touchstone in interpreting any agreement is that the intent of the parties should prevail. See, e.g., Neves, 769 P.2d at 1053-55; Sims, 623 A.2d at 1001-03. Many such courts have noted that a specific identity rule swings the pendulum too far in favor of plaintiffs; while the common-law release rule often unfairly deprives a plaintiff of a separate cause of action by artificially presuming that the release of one joint tortfeasor releases all others, the specific identity rule artificially presumes that parties who have not included specific names or identifying terminology did not intend to discharge others. McInnis, 625 F. Supp. at 955; Neves, 769 P.2d at 1052-53; Sims, 623 A.2d at 1000-01. These courts have reasoned that while a plaintiff should not be unwittingly duped into surrendering a cause of action, defendants also have rights — contribution rights among them — that would be sacrificed if a specific identity rule were adopted. McInnis, 625 F. Supp. at 955.

For general applicability, it would be inappropriate to decide whether to adopt the specific identity rule or the intent rule based on a statutory construction of the New Mexico Uniform Contribution Among Tortfeasors Act, NMSA 1978, §§ 41-3-1 to -8 (Repl.Pamp. 1989). That Act applies only to cases involving joint tortfeasors as that term is defined in Section 41-3-1. Under that Section joint tortfeasors "means two or more persons jointly or severally liable in tort for the same injury." Id. (emphasis added). For purposes of general discussion, we assume as did the parties that the respective liabilities of Pease and Ford would be determined according to principles of proportionate fault. See Bartlett, 98 N.M. at 158-59, 646 P.2d at 585-86. As we observed in Sanchez v. Clayton, 117 N.M. 761, 765, 877 P.2d 567, 571 (1994), "[l]iability for proportionate fault is a liability for a distinct part of the damages and not for the same damages that may be apportioned to others."

If the Uniform Act is not applicable to this action, the rationale for a narrow construction of the statutory provision that other tortfeasors are not discharged "unless the release so provides" is nonetheless generally applicable to New Mexico's abrogation of contribution among concurrent tortfeasors. The Trial Lawyers argue that adopting a specific identity rule is the better-reasoned and more appropriate approach to interpreting the scope of general releases in a jurisdiction that has adopted comparative negligence and abolished joint and several liability for concurrent tortfeasors. The fact that each concurrent tortfeasor is liable only for a distinct portion of the damages lies at the heart of the Court of Appeals' conclusion in Wilson v. Galt, 100 N.M. 227, 231, 668 P.2d 1104, 1108 (Ct.App.), cert. quashed, 100 N.M. 192, 668 P.2d 308 (1983), that " Bartlett effectively eliminates any basis for contribution among concurrent tortfeasors." The pre- Bartlett/Wilson need for concurrent tortfeasors to secure the release of other tortfeasors in order to secure the former's right to contribution and in order to insulate themselves from a later action for contribution by unreleased tortfeasors explains the prevalence of release language purporting to discharge the entire world from liability. The Trial Lawyers argue that courts in the past felt constrained to give far-reaching effect to such broad language in order to facilitate the policy favoring settlement. Because concurrent tortfeasors no longer have a right to contribution, argue the Trial Lawyers, courts should construe general release language narrowly to avoid perpetuating the harsh, unwarranted effects of the common-law rule.

We agree that with the adoption of comparative fault the need to obtain the release of other concurrent tortfeasors has disappeared except for circumstances in which joint and several liability may yet obtain or in which potential indemnitees exist. Even without the release of other concurrent tort-feasors, the settling tortfeasor remains free of further liability both to the plaintiff for damages and to the nonsettling tortfeasors for contribution. See Wilson, 100 N.M. at 231, 668 P.2d at 1108. Relying on Kansas' adoption of comparative fault and abrogation of joint and several liability, the Kansas Court of Appeals held that "the type of release given will have no effect on any party not specifically named in the instrument." Geier v. Wikel, 4 Kan. App. 2d 188, 603 P.2d 1028, 1030 (1979). The current state of the law has led to our conclusion that boilerplate universal release language is circumstantially ambiguous and serves only to trap unwary plaintiffs into surrendering claims when it is not necessary to do so in order to protect the party with whom they are settling.

A release is contractual in nature and as such our primary objective in construing its terms is to give effect to the intent of the parties. Shaeffer v. Kelton, 95 N.M. 182, 185, 619 P.2d 1226, 1229 (1980). We agree with those courts which have observed that adopting a pure specific identity rule entirely ignores the actual intent of the parties to a general release. As we indicated when rejecting the flat bar rule, we should not elevate form over substance. We conclude that sound policy requires more evidence of intent to release potential tortfeasors not named in a release agreement than boilerplate language that objectively purports to release the entire world from liability. See Hurt, 380 So.2d at 434 (stating that "manifestation of intent must be more explicit than signing a printed form . . . providing spaces for the specifically discharged parties"). As a matter of policy, those persons whose wrongful acts cause injury to another should be called upon to answer in damages for those wrongful acts. Having contributed nothing to the settlement of plaintiff's claims, a nonsettling tortfeasor should not be allowed to claim gratuitously a benefit for which he or she did not bargain and for which other parties did not bargain on his or her behalf. Cf. Duncan v. Cessna Aircraft Co., 665 S.W.2d 414, 422 (Tex. 1984) ("A nonsettling tortfeasor should not fortuitously escape compensating his . . . victim simply because of settlement arrangements that did not encompass him or his conduct and to which he contributed nothing.").

It is therefore because release agreements are most often drafted by the party seeking to be released, because the attorneys who draft these agreements and the insurance companies they represent are laboring under the influence of a bygone era, and because the victim's interest in obtaining recompense frequently overshadows concerns with boilerplate language that is often introduced for the first time in a release prepared after the specific parties have agreed to a settlement, that we have concluded boilerplate release language like that used here is inherently ambiguous. We hold that a general release raises a rebuttable presumption that only those persons specifically designated by name or by some other specific identifying terminology are discharged. See Stueve v. American Honda Motors Co., 457 F. Supp. 740, 747 (D.Kan. 1978) (applying Kansas law and adopting presumption that third parties not discharged by general release silent as to other tortfeasors). In the absence of such specific terminology, the person seeking to be discharged must prove by extrinsic evidence that the parties to the agreement actually intended to discharge him or her from liability. Such a rule best mitigates the harsh effects of the common-law release rule, insures that injured parties have in fact intentionally released claims against third parties, and works no unfair hardship on the party negotiating and receiving the release or the prospective third-party beneficiary.

The Court of Appeals was persuaded in Perea that any specific identity requirement must be rejected in order to preserve the freedom of parties to contract as they wish. See 117 N.M. at 778-79, 877 P.2d at 584-85. We do not agree. Even in those jurisdictions that have adopted a specific identity rule, a settling party remains free to obtain the release of nonsettling parties, such as potential indemnitees, by including specific identifying terminology in the release. In this case, for example, Hansen and Pease could have discharged Ford by including language like "and the manufacturers and suppliers of the Releasor's car and its component parts." In an ordinary automobile accident involving three cars, the owner and/or operator of one of which is unknown to the other two, the settling parties could release the phantom driver by including terminology such as "and the driver and persons responsible for the operation and maintenance of the blue car." Finally, in the case of an automobile accident that appears partially attributable to the presence of foliage obscuring a stop sign, the settling parties could release the persons responsible for the foliage by including terminology like "and the owners, occupiers, and any other persons responsible for the upkeep or maintenance of the premises on which the shrubs that obscured Releasee's view of the stop sign were growing."

We recognize that general release clauses without specific identifying terminology have been used extensively and have been relied on as full and final settlement of all claims. Our ruling today shall, therefore, apply only prospectively, except that we will also apply it to this case and to all other cases in which the issue is preserved. Cf. Alsup, 461 N.E.2d at 364-65 (holding that because of widespread use of and reliance on general language, specific identity rule would apply prospectively only).

Conclusion. Because of the circumstantially ambiguous terms of any general release such as the one executed here by Hansen, we today adopt a rebuttable presumption that a general release benefits only those persons specifically designated. Applying that rule to this case, we conclude that the boilerplate language purporting to release all persons from liability for injuries resulting from the January 1990 automobile accident is not sufficiently specific to identify Defendants as third-party beneficiaries of the general release. The summary judgment entered by the trial court is therefore reversed and this cause is remanded for further proceedings.

IT IS SO ORDERED.

BACA, C.J., and FRANCHINI, J., concur.