This case involves an appeal from a jury's determination that a woman was entitled to recover various elements of damage because her fiance broke off their engagement. The plaintiff, Sharon Wildey, brought suit under Illinois's Breach of Promise Act, 740 ILCS 15/1 et seq. (1993). She alleged that the broken engagement harmed her in several ways, and consequently sought recovery for medical expenses, lost business profits and pain and suffering. The jury returned a $178,000 verdict in her favor. The district judge remitted that amount by $60,000, concluding that the lost business profits were not fairly attributable to the broken engagement. The district judge left the remainder of the verdict intact, however. The parties now appeal, alleging various points of error. We agree with the suggestion that the case should have been dismissed because Sharon failed to provide her fiance with notice of each element specified by the statute. We therefore reverse.

Sharon Wildey and Richard Springs enjoyed a relationship early in 1992. When Richard broke off the engagement at the end of April, Sharon became quite distressed. She eventually decided to sue under Illinois's Breach of Promise Act, 740 ILCS 15/1 et seq. (1993).

Sharon is an attorney in Chicago, and Richard is an Oregon cattle rancher. Introduced by a mutual friend, the two began their long-distance relationship over the telephone. After a number of telephone calls, Richard decided to visit Sharon in Chicago in January of 1992. The relationship apparently went well because Richard visited Chicago on two more occasions during January and February. The parties decided that a Florida vacation was in order.

At the end of the five-day Florida trip, the parties agreed to be married. While waiting for a plane in the Orlando airport, Sharon suggested that she and Richard consider marriage. Although Richard agreed to her suggestion, he expressed doubts about the couple's ability to overcome the practical difficulties of a long-distance relationship. Sharon wanted to stay in Chicago until her children got out of school; Richard did not intend to leave his Oregon ranch. The parties eventually decided on a "commuter-type" marriage for a five-year time period, leading to an eventual relocation for both in Florida or the Caribbean. They then flew back to Chicago, and Richard returned to Oregon.

Richard again visited Sharon later in March. Once in Chicago, he began shopping for engagement rings and eventually purchased one. When the ring was ready, Richard got down on one knee and formally proposed. The couple set a wedding date and planned a Chicago ceremony. All of these activities occurred in Chicago.

The couple spent more time together during April. Early in the month, Sharon visited Richard's ranch in Oregon. She and Richard later attended her son's wedding in Los Angeles. Afterwards, Sharon and her daughters flew to Oregon to spend several days on Richard's ranch. In another week or so, everyone flew to Chicago to spend time there.1

After the last weekend in Chicago, however, Richard began having second thoughts about the marriage. On a flight to Florida on April 27, 1992, Richard decided to break the engagement. He composed a letter to Sharon explaining his doubts about the marriage. He also suggested that she keep the engagement ring and some money he had placed in a Chicago bank account. Pl.Ex. 1.

Sharon responded with a letter. She stated that the broken engagement had caused her to become "extremely depressed and anxious," and she asked for financial help. Pl.Ex. 3. Richard wrote back, defending himself and writing that he felt that he had fulfilled any obligations he may have had. Pl.Ex. 2. Sharon drafted a second letter in which she apprised Richard of her intent to sue. Pl.Ex. 4.

Sharon's second letter contains the date that the parties had planned to be married. It also references several elements of damage, such as medical bills from counseling and lost income, that Sharon believed she suffered as a result of the broken engagement. Sharon further wrote that she had spoken with an attorney and intended to file suit. She also stated that she had been willing to marry Richard and gave him the option of discussing the matter with her. Sharon did not, however, include any of the dates upon which the parties had exchanged marital promises.2

Sharon filed suit soon afterwards. She contended, among other things, that the broken engagement had occasioned large expenditures on medical care and had also caused her to lose business income. She additionally sought recovery for pain and suffering. The jury returned a verdict totaling $178,000 in Sharon's favor. Of this amount, $25,000 was awarded for past and future medical costs; $60,000 was awarded for lost business profits; and $93,000 was awarded for pain and suffering.

At Richard's request, the district court remitted the award to $118,000, concluding that lost business profits were not attributable to the broken engagement. The court affirmed the verdict as to the other elements of damage, however. The district court also refused to reverse the verdict on various other grounds that Richard suggested. It concluded that Illinois law, rather than Florida law, had properly governed the issues at trial. It also determined that the notice Sharon provided in her second letter had been adequate; "substantial compliance" with the notice provision, in the district court's view, was all that Illinois law required.

Richard now appeals. We agree with the district court that Illinois law properly governs this case. We do not agree, however, with that court's notice analysis. We believe instead that the omission of the date that the parties became engaged is fatal to Sharon's claim. The plain language of the Breach of Promise Act requires that a potential plaintiff include several items in the notice that must be sent to a defendant. The date that the parties became engaged is clearly one of these items. Although the strict construction that we rely on may seem unwarranted, several factors render it appropriate in this case. First, the history behind actions for the breach of a promise to marry argues in favor of a strict construction; the strictures of the Promise Act exist because these actions were unpopular and subject to abuse. Further, the date of the engagement, though plausibly insignificant when the fact of the engagement is not disputed, was potentially dispositive here. An admission concerning the date the parties became engaged determined the place of that engagement; for reasons that will become clear in the choice-of-law analysis, this fact threatened Sharon's ability to maintain her suit. In addition, we note that Sharon is a seasoned attorney. She had consulted with counsel when she drafted the second letter. In light of these considerations, and of the clarity of the Statute's requirements, we reverse.3

A. Breach of a Promise to Marry: An Historical Perspective

The action for the breach of a promise to marry is of antique vintage. First conceived as a creature of the English ecclesiastical courts, the action was originally used to pressure a reluctant lover into fulfilling a marital promise. W.J. Brockelbank, The Nature of the Promise to Marry--A Study in Comparative Law, 41 Ill.L.Rev. 1, 3 (1946); Michael Grossberg, Governing the Hearth: Law and the Family in Nineteenth-Century America 34 (1985). The common law eventually adopted the action, however, and permitted the recovery of monetary damages. Although developed by English courts, the action found its way into the American colonies and was later used by post-revolutionary American lawyers.

The action served dual ideals in colonial America. On the one hand, a breach of promise action continued to appeal to vestiges of the older notion that marriage was a property transaction completed after complex family negotiations. Grossberg, supra, at 35. But on the other hand, the action began to pay tribute to the emerging ideal that marriage was a sacred contract premised upon affection and emotional commitment. Id. The suits soon became utilized almost universally by women, and were justified by lawmakers largely on these grounds. Id. at 37. Marriage was considered necessary to secure both a woman's social and financial security. Margaret F. Brinig, Rings and Promises, 6 J.Law, Economics, and Organization 203, 204-05 (1990). But more importantly, the actions, and the judges who were willing to enforce them, recognized that promises to marry sometimes occasioned a loss of virginity. Id. at 205. Because of the importance the society of that day placed on premarital chastity, the economic and social harm suffered by a jilted woman were often reflected in large damage awards. Id.

The actions were characterized by a lack of legal formality peculiar for the law of that time. Foreign observers noted that, unlike wills or commercial contracts, little was needed to support an allegation that the parties had become engaged. Grossberg, supra, at 51. Consequently, appellate courts deferred widely to jury determinations in credibility contests. Id. at 49. Traditional rules relating to damages, too, were relaxed. Despite the originally contractual nature of the action, judges refused to confine the damage measure to immediate loss. Instead, they permitted recovery for elements such as mental anguish and a loss of social position. Id. at 43.

Largely because of the perceived vagaries of the suits, the actions had fallen into disrepute by the early twentieth century. Three principal reasons are given for their decline in popularity. The first is the unfounded use of the suit, given the lax standards of proof, to extort out-of-court settlements. Brockelbank, supra, at 13; Note, Heartbalm Statutes and Deceit Actions, 83 Mich.L.Rev. 1770, 1776 (1985).4 Second, the excessive damages awarded prompted disdain for the actions. Id. at 1774. Finally, the ideals that the action served came to be viewed as anachronistic. The greater social and economic freedom incident to women's entry into the workforce meant that the loss of an initial suitor posed a lower threat to future prospects than it might have in the nineteenth century. Grossberg, supra, at 55.5

As concerns grew, legislatures began to act.6 Florida and Illinois were among these. Florida abrogated the cause of action by enacting a statute providing that agreements to marry made in Florida could not be enforced "within or without this state." Fla.Stat.Ann. Sec. 771.04 (1986). Illinois, too, outlawed the common law cause of action. In 1935, the Illinois legislature passed the "Heart Balm Act," making it "unlawful" to file an action based on the breach of a promise to marry. Ill.Rev.Stat., ch. 38, Secs. 246.1 & 246.2 (1943). Although the Florida statute eliminating the cause of action remains in force, the Illinois statute soon encountered troubles.

In Heck v. Schupp, 394 Ill. 296, 68 N.E.2d 464, 466 (1946), the Illinois Supreme Court held the statute unconstitutional under provisions of the Illinois Constitution providing a legal remedy for all injuries.7 The Illinois legislature soon responded in kind. Faced with the court's determination that the State was required to leave plaintiffs a remedy for the breach of a promise to marry, the legislature enacted Illinois's Breach of Promise Act in 1947. See generally 740 ILCS 15/1 et seq. (1993) (the Promise Act). This second attempt at circumscribing the perceived abuses of the common law action leaves the plaintiff's ability to recover intact. The statute does, however, suggest that actions for breach of promise were subject to both abuse and the danger of an excessive damage award. The legislature clearly states its misgivings in the Act's Preamble, determining that "the remedy heretofore provided by law for the enforcement of actions based upon breaches of promises or agreements to marry has been subject to grave abuses and has been used as an instrument for blackmail by unscrupulous persons for their own unjust enrichment...." 740 ILCS 15/1. It therefore uses an elaborate notice provision and limits the types of damages that a plaintiff may recover. 740 ILCS 15/2, 15/4 & 15/5. This constricted version of the original breach of promise action passed muster under the Illinois Constitution, Smith v. Hill, 12 Ill. 2d 588, 147 N.E.2d 321, 326-27 (1958), and has survived until the present.

In light of the fact that Florida will not recognize a cause of action for breach of promise to marry, and Illinois will, the parties hotly contest which state's law governs the claim. Richard contends that Sharon's claim is barred because Florida does not recognize the cause of action, and the parties became engaged in Florida. Sharon, however, claims that the place of making the contract is not dispositive. Instead, she suggests that we apply the Second Restatement's "most significant contacts" test because the place of contracting (Florida) was different from the place where the parties planned to marry (Illinois). In light of Illinois's contacts with the facts of the dispute, she claims that Illinois law controls the fate of her claim. The district court found that Illinois law applied to Sharon's claim. We agree with that conclusion.8

As a federal court sitting in diversity, we must look to the conflict-of-laws rules in the jurisdiction in which we sit to choose the substantive law applicable to the case. Florida Risk Planning Consultants, Inc. v. Transport Life Ins. Co., 732 F.2d 593, 595 (7th Cir.1984) (quoting Klaxon Co. v. Stentor Electric Mfg. Co., Inc., 313 U.S. 487, 61 S. Ct. 1020, 85 L. Ed. 1477 (1941)). Under the law of Illinois, Richard relies on an appeal to traditional conflict-of-laws principles governing contracts. These traditional principles suggest that a contract is construed according to the law of the state in which it was entered, unless performance was to be in another state--in which case that state's law governs. William J. Lemp Brewing Co. v. Ems Brewing Co., 164 F.2d 290, 293 (7th Cir.1947), cert. denied, 333 U.S. 863, 68 S. Ct. 745, 92 L. Ed. 1142 (1948). If the state in which the contract was formed does not recognize a cause of action for its breach, however, comity demands that a court refuse to enforce the contract at all. Frankel v. Allied Mills, Inc., 369 Ill. 578, 17 N.E.2d 570, 572 (1938). A straightforward application of the traditional rules would result in the application of Florida law because the agreement to marry was first formed in Florida, and Florida refuses to recognize a cause of action for breach of promise to marry.

We do not believe that traditional choice-of-law principles should control here, however. Illinois courts have long since abandoned a formalistic resort to the traditional conflicts rules. Palmer v. Beverly Enterprises, 823 F.2d 1105, 1107 (7th Cir.1987). The Illinois Supreme Court refuses to appeal to the more traditional principles when a "wooden application" of them would result in "unjust and anomalous results." Ingersoll v. Klein, 46 Ill. 2d 42, 262 N.E.2d 593, 596 (1970). Therefore, that court expressed considerable doubt about the efficacy of the vested rights doctrine because it ascribes undue significance to the place at which the rights of the parties "vest." Id., 262 N.E.2d at 595. In Ingersoll, the vested rights doctrine would have required the court to look to the law of the jurisdiction where the injury occurred to conclude that the plaintiff possessed a cause of action not recognized by the state where the parties had the most significant contacts. 262 N.E.2d at 596. Instead, the Ingersoll court adopted the Second Restatement's "most significant relationship" test. Id. See generally Restatement (Second) of Conflicts Sec. 145 (1971).

We find Ingersoll's reasoning persuasive here. The cause of action for a breach of a promise to marry has a hybrid character. Although it is premised on the existence of a "contract," or an exchange of promises between the parties, White v. Prenzler, 7 Ill. 2d 624, 131 N.E.2d 540, 543 (1956), it nevertheless recognizes tort-like elements of damage. Stanard v. Bolin, 88 Wash.2d 614, 565 P.2d 94, 96 (1977). This hybrid character does not itself compel the adoption of conflicts principles governing torts. It does, though, render a strict application of the traditional conflict-of-laws principles for contracts somewhat less appealing. Our holding does not, however, rest on these grounds.

Instead, we believe that Ingersoll is indicative of the approach the Illinois Supreme Court would take if confronted with the question that we decide today. As we have previously noted, Ingersoll's reasoning extends to contract cases. Palmer, 823 F.2d at 1107. "The criticism of choice of law rules grounded on the vested rights doctrine applies with full force to [the] case where [a state's] only interest in [the] dispute is by virtue of the alleged making of the contract there ... and the significance traditionally bestowed upon that fact by the vested rights doctrine." Id. at 1108. Although we deal with comity here, the same considerations are relevant. The appeal to Florida law is premised entirely on the fact that the parties became engaged in the Orlando airport while waiting for a return flight to Chicago. Florida's connection to the case is therefore tenuous, and the belief that Florida law should govern can only be defended with reference to the earliest conflict-of-law principles.

We continue to believe, however, that the Illinois Supreme Court would apply the Second Restatement's "most significant contacts" test. See generally, Restatement (Second) of Conflicts Sec. 188 (1971). This is the test that we have relied upon in the past to govern contract disputes. Palmer, 823 F.2d at 1107-08; Florida Risk Planning, 732 F.2d at 595. And this is the test that a number of Illinois appellate courts have endorsed. See generally Champagnie v. W.E. O'Neil Construction Co., 77 Ill.App.3d 136, 32 Ill. Dec. 609, 614-16, 395 N.E.2d 990, 995-97 (1979). That State's appellate courts have continued to apply this test in both substantive contract cases, see, e.g., Purcell and Wardrope v. Hertz Corp., 175 Ill.App.3d 1069, 125 Ill. Dec. 585, 592, 530 N.E.2d 994, 1001 (1988); Boise Cascade Home & Land Corp. v. Utilities, Inc., 127 Ill.App.3d 4, 82 Ill. Dec. 180, 186-87, 468 N.E.2d 442, 448-49 (1984), and in forum non conveniens cases involving contracts. See Certain Underwriters at Lloyd's London v. Bertrand Goldberg Ass'n, 238 Ill.App.3d 692, 179 Ill. Dec. 709, 714, 606 N.E.2d 541, 546 (1992); Illinois Tool Works v. Sierracin, 134 Ill.App.3d 63, 89 Ill. Dec. 40, 44-45, 479 N.E.2d 1046, 1050-51 (1985). In fact, we have no indication that the Illinois Supreme Court would be dissatisfied with the most significant contacts approach. Quite to the contrary, at least where insurance contracts are concerned, the Illinois Supreme Court seems to approve of the factors considered by the Second Restatement test. See Lee v. Interstate Fire & Casualty, 826 F. Supp. 1156, 1159 (N.D.Ill.1993) (discussing dicta in Hofeld v. Nationwide Life Ins. Co., 59 Ill. 2d 522, 322 N.E.2d 454 (1975)).

The fact that Illinois's appellate courts might occasionally depart from the otherwise general application of the Second Restatement test does not alter our analysis. See, e.g., Olsen v. Celano, 234 Ill.App.3d 1045, 175 Ill. Dec. 799, 803-04, 600 N.E.2d 1257, 1261-62 (1992) (applying the traditional principle that a brokerage contract that is invalid under the law of the place of contracting is unenforceable everywhere). We are not required to follow the holding of a state appellate court if we do not believe it reliably predicts the state supreme court's stance on an issue. Indiana Harbor Belt R. Co. v. American Cyanamid, 916 F.2d 1174, 1176 (7th Cir.1990). Courts have previously recognized that the choice-of-law rule in Illinois for contracts cases "appears to be in flux." Barry Gilberg, Ltd. v. Craftex Corp., Inc., 665 F. Supp. 585, 590 (N.D.Ill.1987); see also Lee, 826 F. Supp. at 1159-60 (finding the fact that the appellate court has occasionally stated a choice-of-law rule different from the most significant contacts test insignificant). But this hardly alters our responsibility to adopt the approach by which we believe the Illinois Supreme Court would abide. Absent indication that the Illinois Supreme Court is moving away from the common-sense conflicts approach that it embraced in Ingersoll, 262 N.E.2d 593, we will continue to apply the Second Restatement's "most significant contacts" test.

The appellate court's holding in Olsen, contrary to the defendant's contention, does not constitute such an indication. Olsen involved a contract for the sale of real estate. In it, the appellate court affirmed the dismissal of a complaint because the contract involved was void under the law of the state where it was made. In resorting to this older rule, the appellate court relied on Frankel v. Allied Mills, Inc., 369 Ill. 578, 17 N.E.2d 570 (1938) (holding that the validity of a contract for a commission for the sale of real estate is determined by the law of the state where the contract is made), and did not apply the significant contacts test. 175 Ill. Dec. at 802-03, 600 N.E.2d at 1260-61. While this holding might have some predictive validity for Illinois's approach to cases involving real estate (a subject on which we offer no opinion), it hardly governs here.

The present case involves a broken engagement. For reasons noted earlier, this cause of action is not purely contractual in nature. Neither is it in any way "tied" to Florida. Instead, the fact that the agreement to marry was first discussed in Orlando was something of a fortuity. A glance at the record reveals a fair amount of wrangling about what exactly counts as a complete and formal "engagement." Sharon apparently argued to the district court, for instance, that despite the original Florida discussion, the contract to marry was not "consummated" until the parties later purchased an engagement ring and set a wedding date in Illinois. Mem.Op. at 10-11. Whatever logical difficulties one might have with such an argument, we note that a more formal engagement--in the sense that Richard got down on one knee and proffered a diamond ring--later occurred in Chicago (according to Richard, once, and according to Sharon, twice). The amount of confusion surrounding the finality of the parties' agreement to marry demonstrates the transitory nature of an engagement. Unlike land, it is not invariably tied to a particular place.

In contracts involving the transfer of an interest in land, on the other hand, traditional rules typically require a court to look to the situs of the land to determine the governing law. See Eugene F. Scoles & Peter Hay, Conflict of Laws Sec. 18.23 (1992); see also Restatement (Second) of Conflicts Secs. 188 & 189. Although Olsen stated a somewhat different rule, the court in fact applied the law of the situs in that situation. 175 Ill. Dec. at 803, 600 N.E.2d at 1261. We therefore find Olsen unpersuasive in the present case. Whatever its import for conflicts cases involving real estate, it has no application here.

In light of the stated considerations, we believe the Second Restatement's most significant contacts test should govern choice of law analysis. See generally, Restatement (Second) of Conflicts Sec. 188. Under that test, the contacts relevant to a choice-of-law decision include "the place of contracting, negotiation, performance, location of the subject matter of the contract, and the domicil[e], residen[ce], place of incorporation, and business of the parties." Palmer, 823 F.2d at 1109-10 (quoting Champagnie, 32 Ill. Dec. at 615, 395 N.E.2d at 996).

Illinois law clearly governs under the most significant contacts test. To the extent that the parties ironed out the details of the proposed marriage, they did so in Illinois (or on a plane back to Chicago), where Sharon lived and worked. Richard purchased Sharon's wedding ring in Chicago, and he got down on one knee and restated the marital promise there. The parties eventually planned to be married in Chicago, and had begun searching for appropriate wedding locations in that area. These aspects of the agreement to marry are sufficient to convince us that Illinois law should govern the parties' transaction.9

Because Illinois law governs the parties' agreement, the strictures of Illinois's Promise Act guide our analysis. See generally 740 ILCS 15/1 et seq. The Promise Act requires a plaintiff to provide a defendant with notice of her intent to sue:

If the notice provided for by Section 4 is not given as provided in that section, any such civil action for breach of promise or agreement to marry shall be dismissed and the person to whom any such cause of action accrued shall be forever barred from further suing.

740 ILCS 15/5. The required notice must be written, signed and sent within three months of the date of breach. In addition, it must contain the following information:

1) the date upon which the promise or agreement to marry was made;

2) the date upon which the marriage ceremony was to have been performed;

3) the damages suffered by the person signing the notice; and

4) a statement of whether the person signing the notice is or is not still willing to marry the person to whom the statement is given.

740 ILCS 15/4.

It is not disputed that Sharon provided Richard with a signed, written notice within three months of the broken engagement. Nor is it disputed that the notice contained only three of the four required items. Sharon's second letter to Richard twice stated the date upon which the marriage ceremony was to have been performed (September 5, 1992). That letter also noted the damages, such as medical expenses, depression and business problems, that Sharon felt Richard had caused by breaking the engagement. Finally, the letter stated that Sharon had been willing to marry Richard, and was still willing to discuss the matter with him if he so desired.10 The letter did not, however, contain the date upon which the promise or agreement to marry had been made.

On appeal, the parties dispute the significance of this omitted element. Richard contends that the omission of this element was fatal to Sharon's claim.11 Sharon advances several arguments favoring a liberal construction of the statute, and suggests that substantial compliance with the Promise Act's notice provision suffices in light of Richard's knowledge of their engagement. The district court agreed with Sharon, holding notice sufficient under the Promise Act and submitting the case to the jury. We disagree and reverse.

The Illinois Promise Act plainly requires that the specified notice include "the date upon which the promise or agreement to marry was made." 740 ILCS 15/4. We believe that Sharon's failure to include this item in the written notice is fatal to her claim. In addition to the clear premium the Promise Act places on the date of the agreement, the Act states that claims shall be dismissed "[i]f the notice provided for by Section 4 is not given as provided in that section." 740 ILCS 15/5. This plain language reflects the Illinois legislature's priorities. Faced with the Illinois Supreme Court's mandate that the cause of action for the breach of a promise to marry could not be completely abrogated, Heck, 68 N.E.2d at 466, the legislature restricted the action's particulars. It limited the recovery of certain elements of damage. 740 ILCS 15/2. It instituted the elaborate notice provision at issue here. 740 ILCS 15/4. And it determined that the failure of notice would bar a claim. 740 ILCS 15/5. This Act eventually withstood constitutional challenge. Smith, 147 N.E.2d at 326.

As a federal court sitting in diversity, we hesitate to disturb the balance that Illinois's legislature and its Supreme Court eventually struck on this controversial subject matter. By ignoring any of the required statutory items we risk running afoul of the principle that we should be guided by the express statutory limits of state law. Butler v. Sentry Ins. A Mut. Co., 640 F. Supp. 806, 812 (N.D.Ill.1986) (noting that "it is certainly impermissible for a federal court exercising diversity jurisdiction to attempt to expand in common-law fashion the express statutory limits of state law"); see also Afram Export Corp. v. Metallurgiki Halyps, S.A., 772 F.2d 1358, 1370 (7th Cir.1985) (noting that "[f]ederal judges are disinclined to make bold departures in areas of law that we have no responsibility for developing"). Expanding the perimeters of a state's considered cause of action is not our function. Rubinstein v. Collins, 20 F.3d 160, 172 (5th Cir.1994); Burris Chemical, Inc. v. USX Corp., 10 F.3d 243, 247 (4th Cir.1993). Lacking contrary indication from the Illinois Supreme Court, we therefore choose to enforce the Promise Act as it is written.

Sharon argues, however, that the Illinois Supreme Court has issued a contrary indication. She points our attention to White v. Prenzler, 7 Ill. 2d 624, 131 N.E.2d 540 (1956), in which the Illinois Supreme Court stated that:

In construing notice requirements, as well as statutes of limitations, courts must look to substance, rather than merely to form, in seeking the true intention of the legislature.

131 N.E.2d at 543. Unfortunately, the import of this passage is not entirely clear in the present case.

White itself dealt with a case of concealment. There, the defendant breached his promise to marry the plaintiff when he married another woman. Because the plaintiff was not aware of his marriage to another, she did not send a notice of her intent to sue within three months of the breach. She did, however, send the notice within three months after she learned of the breach. The court therefore needed to determine whether the three-month time period began running as of the date of the breach (in which case the plaintiff's claim was barred), or whether the time period began running as of the date that the plaintiff knew about the breach (in which case her claim was timely). The court first examined the statutory language, which required notice from one "who is about to commence" a breach of promise action, 740 ILCS 15/4. On this basis, it concluded that "[o]ne cannot be a person about to sue until he knows of a reason to sue." 131 N.E.2d at 542. The court also noted that it would be unfair to allow a defendant to use the notice requirement to "escape a duty as a result of his own failure to disclose an incapacity to perform his part of the contract." Id., 131 N.E.2d at 543. The court therefore held that knowledge that a claim may exist is a necessary prerequisite to beginning the three-month time period during which notice must be given. This rule governs the timing of notices under the statute. The court did not, however, make any determinations regarding the sufficiency of the notice actually given.

White is therefore of limited guidance here, where we are concerned with the sufficiency of the notice that Sharon mailed to Richard. We can certainly heed White's general command to look to substance rather than form, but even that does not dispose of the matter before us. Richard complains that the notice was substantively insufficient; Sharon suggests that it was not. Unfortunately, the parties have pointed to no further authority construing the sufficiency of notice under Illinois's Promise Act. Instead, Sharon suggests that the Act's requirements should be relaxed for several reasons. First, she asserts that statutes in derogation of the common law should be liberally construed in favor of the plaintiff. This, she suggests, renders the notice valid in light of its purported substantial compliance with the statutory mandate. Finally, she claims that the date of the engagement was not important given the fact that the existence of the engagement was uncontested. Although we are sympathetic to Sharon's plight under an adverse ruling on this issue, we do not find any of these arguments persuasive.

Liberally construing the statute in favor of a plaintiff runs contrary to express legislative intent. The Promise Act was instituted to eliminate the perceived common law abuses of the actions. The Preamble to the Promise Act states that:

... the remedy heretofore provided by law for the enforcement of actions based upon breaches of promises or agreements to marry has been subject to grave abuses and has been used as an instrument for blackmail by unscrupulous persons for their unjust enrichment, due to the indefiniteness of the damages recoverable in such actions and the consequent fear of persons threatened with such actions that exorbitant damages might be assessed against them.... Consequently, in the public interest, the necessity for the enactment of this chapter is hereby declared as a matter of legislative determination.

740 ILCS Sec. 15/1. In light of this concern, the legislature engrafted various limitations, such as the notice provision, onto the common law cause of action. In interpreting these provisions, the legislature instructed that the "act shall be liberally construed to effectuate the objects and the purposes thereof and the public policy as herein declared." 740 ILCS 15/9 (emphasis added). Those objects and purposes were the avoidance of exorbitant damage awards and the winnowing of the multitude of claims. The Illinois legislature's wisdom in choosing these objects is none of our business. As a federal court sitting in diversity, we certainly have no mandate to assert any policy preferences we might have with respect to either the breach of promise lawsuits or the purported efforts to restrict them. We are bound by Illinois' statement of intent--a statement only recently endorsed by that state's courts. See Vann v. Vehrs, 260 Ill.App.3d 648, 198 Ill. Dec. 640, 642, 633 N.E.2d 102, 104 (1994) (relying on the Act's stated legislative purpose to note that "persons bringing actions for the actual damages incurred from the breach of a promise to marry must comply with the notice and limitations period set forth in the Act"). The legislature intended to make it more difficult for plaintiffs to bring these suits, not less. A liberal construction favoring plaintiffs is not, therefore, in order--despite the fact that the statute may place limitations on the common law.12

Neither do we believe that relaxing the statutory requirements is warranted under the doctrine of substantial compliance. We can assume for the purposes of argument that Illinois courts would use the doctrine of substantial compliance to evaluate notices under the Promise Act. This, perhaps, is the import of White's statement that courts should value substance over form. 131 N.E.2d at 543. But we note that Illinois courts have largely developed the doctrine of substantial compliance under the Tort Immunity Act's (TIA) notice provision. Compare Ill.Rev.Stat., ch. 24, Sec. 1-12 (1947) with Ill.Rev.Stat., ch. 89, Sec. 29 (1947). Although cases discussing the TIA's notice provision might give some indication of how a court should treat the Promise Act's notice provision, see White, 131 N.E.2d at 543, we can fairly question the value of the more liberal interpretations of the TIA's notice provision here.

The Illinois legislature has repealed the TIA, along with its restrictive notice provision. See 745 ILCS 10/8-102 (1993). This repeal was the culmination of a dialogue between Illinois's courts and its legislature concerning, among other things, the manner in which the TIA's notice provision should be interpreted. The more liberal interpretations of the TIA's notice provision were a result of the legislature's relaxations of the TIA's strictures. See Reese, 149 Ill. Dec. at 462-63, 561 N.E.2d at 1158-59 (noting that the "continual trend" in Illinois courts toward a "liberal interpretation of the Tort Immunity Act" was supported by different legislative relaxations of the Act). We benefit from no such exchange indicating that a relaxed interpretation of the Promise Act is in order. Instead, the latest statement of intent that we have under the Promise Act is that of the legislature's original hostility to breach of promise claims.

Despite this, we will assume that White's statement is sufficient to indicate that Illinois courts would use the doctrine of substantial compliance to evaluate notices under the Promise Act. Even doing so, however, we find that the doctrine fails to save the present claim. Illinois courts, even under the most liberal interpretations of the TIA's notice provision, have not relied on the doctrine of substantial compliance when confronted with the complete omission of one of the items mandated by statute.

The doctrine of substantial compliance can be applied to preclude the dismissal of certain complaints despite the fact that a plaintiff's notice was in some way inadequate. The doctrine generally requires a court to engage in a common sense assessment with respect to whether a notice, despite its inadequacy, fulfilled its statutory purpose and did not prejudice a particular defendant. See, e.g., McClintock, 169 Ill. Dec. at 467-68, 591 N.E.2d at 971-72. Here, Sharon relies on the doctrine to argue that the omission of the date of the engagement was insignificant in light of the fact that the existence of the engagement was not contested.

Illinois courts have not, however, stretched the doctrine of substantial compliance to lengths that would excuse the complete omission of a required item of information in a notice. In Carroll v. Chicago Housing Authority, 155 Ill.App.3d 710, 108 Ill. Dec. 124, 125, 508 N.E.2d 285, 286 (1987), the court stated:

Illinois courts construe the Tort Immunity Act to require, in the first instance, that all elements listed in [the notice provision] be present in the written notice; the omission of even one element is fatal. Nevertheless, once a particular element is furnished, the courts take a liberal approach as to the sufficiency of that element, provided that it is 'reasonably sufficient to fulfill the requirements of the statute ... and the public entity has not been misled or prejudiced thereby.'

108 Ill.Dec. at 125, 508 N.E.2d at 286. See also Lando v. Chicago, 128 Ill.App.3d 597, 83 Ill. Dec. 752, 755, 470 N.E.2d 1172, 1175 (1984) (holding that "[w]here, as here, an element has been completely omitted from the formal notice ... a plaintiff will not be found to have complied with the Act"); Reese, 149 Ill. Dec. at 463, 561 N.E.2d at 1159 (holding that "each element required under the Act must be included in the written notice and each element must be sufficiently clear and accurate ..."). These cases suggest that the doctrine of substantial compliance only applies to save notices in which all the elements are provided, but some are vague or inaccurate. See, e.g., H. Winter Metal Co. v. Chicago, 143 Ill.App.3d 542, 97 Ill. Dec. 611, 615, 493 N.E.2d 93, 97 (1986) (inaccurate description of accident location plus actual knowledge of that location was sufficient); McClintock, 169 Ill. Dec. at 467, 591 N.E.2d at 971 (substitution of attorney's address instead of plaintiff's address was sufficient because plaintiff could be reached through attorney).

The doctrine of substantial compliance therefore fails to save the notice in the present case. We simply have no indication that Illinois courts would excuse the complete omission of one of the statute's required elements; the parties have not come forward with evidence that Illinois would use the doctrine to save notices that completely overlook a required element. We therefore find that the doctrine of substantial compliance would not be used to save Sharon's claim. Given Illinois's handling of the doctrine, the trial court improperly submitted the issue to the jury.13

Even if we believed the doctrine of substantial compliance governed a case such as this, however, we would nevertheless have difficulty accepting the argument that the date of the engagement was immaterial. Sharon's substantial compliance argument relies on the claim that the purpose of the Promise Act's notice provision is to provide the defendant with notice of the alleged date of engagement so that he may prepare to rebut the allegation that the parties were in fact engaged (much in the manner that the prosecution needs to prepare to rebut an alibi defense). Because the fact of the engagement was not in issue in this case, Sharon insists that her omission of the engagement date was insignificant.

This argument overlooks the fact that the date of the engagement may be important for reasons other than determining whether or not a defendant proposed. Here, for instance, the date of the engagement was potentially dispositive. An admission concerning the date of the engagement would have unequivocally established the place of the engagement--a fact that was quite important for reasons noted in our choice-of-law analysis.

The record reveals a substantial amount of confusion concerning the date of the engagement. Sharon's original complaint (and Richard's answer) suggested that the date of the engagement was February 20, 1992 (a date that did not place the parties in Florida). The parties later stipulated, however, that the date of the engagement was March 9, 1992 (in the Orlando airport at the end of the parties' five-day Florida trip). The record fails to reveal either the reason for the original confusion or for the later stipulation. Despite this, we do not find the eventual stipulation as to the date of the engagement significant in light of the confusion that has surrounded the term "engagement" since the proceedings between the parties began.

The stipulation as to the date of the engagement was entered as part of the final Pretrial Order. Appellant's Br.Ex. F. This eventual stipulation, which fixed the place of the engagement as Florida, was only entered after the time for dispositive pretrial motions had apparently passed. The district judge later used the failure to raise the choice-of-law issue prior to trial as a basis for ruling against Richard. When denying his post-trial motion for relief, the district judge stated:

The crux of Springs' argument is that he and Wildey entered a contract to marry at the moment that they became engaged. At this late stage of the case, the court cannot rule on the narrow legal question that Springs' argument raises. Springs properly should have raised this legal point via a motion to dismiss or a motion for summary judgment. Instead, Springs waited until the middle of trial to raise the issue of the legal effect of the engagement.... Because the court was not asked before trial to rule on the effect of the engagement, it is impossible to rule as a matter of law that the agreement to marry took place in Florida when the parties became engaged.

Mem.Op. at 10. Richard's Motion to Dismiss did not in fact raise the choice-of-law issue. Def.Reply Br.Supp.Mo.Dismiss. Instead, it focused on dismissal for lack of notice. Id. Although we can only hazard a guess at the reasons for this, we find it noteworthy that at the earlier stages of the proceedings the date of the agreement to marry was not yet fixed. The parties' depositions similarly reflect this confusion over the precise date of the formal agreement to marry.14

Although the facts of the March 9th exchange at the Orlando airport were never in dispute, whether this event should be characterized as a "formal engagement" or "agreement to marry" most assuredly was. Even when responding to Richard's post-trial motions, Sharon argued that there was a difference between an engagement and an agreement to marry:

... there were repeated and mutual promises to marry made in Illinois. The engagement ring was purchased and given in Illinois. The contract was, therefore, executed in Illinois. Assuming arguendo that it is not conclusive that the contract was made in Illinois, there is still not conclusive evidence that it was made in Florida....

Pl. Post-Trial Res. at 10. The eventual stipulation thus decided very little. The continual arguments attempting to fix the date of formal agreement to marry demonstrate that it was in Sharon's interest not to have that date precisely fixed. It is impossible to say precisely how including the date of the agreement to marry in the notice would have altered the course of litigation. But we note that the date was quite significant.

In light of these factors, we believe the omission of the date of the agreement to marry was problematic--despite the fact that the parties's engagement was never itself disputed. One purpose of the Promise Act's notice provision, as Sharon suggests, was certainly that of apprising a defendant of the date of a possibly contested engagement. But we have no reason to believe that this was the only purpose. The legislature, in recognizing the indefinite nature of affairs of the heart, undoubtedly wished potential plaintiffs to spell out all facts relevant to the cause of action at an early point in time. By admitting certain facts in the required notice, a plaintiff would later be prevented from changing postures as litigation needs arose.

Here, the letter notifying Richard of the impending suit did not include an admission as to when (and consequently, where) the engagement occurred. Yet the express language of the Promise Act requires a plaintiff to put this item in the prescribed notice. Sharon's failure to do so therefore bars her cause of action. Sharon is a seasoned attorney. She consulted with an attorney before sending the required notice. Under these circumstances, we simply cannot ascribe the omission to ignorance.15 The history of Illinois's Promise Act, its pointed Preamble, and the controversy that has surrounded it, practically dictate this strict construction of the Act's provisions. In any event, the doctrine of substantial compliance is of no avail as Illinois courts have interpreted it; it does not apply to save notices that completely omit an item required by statute. Even if the doctrine applied, however, the fact that the agreement to marry was quite important to the resolution of this case would preclude relief.

We agree with the district court's conclusion that the parties' dispute was governed by Illinois law. But under the circumstances of this case, we cannot conclude that the notice provided was adequate.

For this reason, the judgment of the district court is REVERSED.