Doe v. Etihad Airways, P.J.S.C. (870 F.3d 406, 2017)

United States Court of Appeals for the Sixth Circuit 2017-08-30

Judiciary and Judicial Procedure Transportation

Citations

Full opinion text

OPINION

BOGGS, Circuit Judge.

Plaintiff Jane Doe and her eleven-year-old daughter flew aboard Etihad Airways from Abu Dhabi to Chicago. For the duration of the fourteen-hour journey, Doe’s tray table remained open in her lap because a knob that was meant to hold it in place had fallen to the floor. During the flight, Doe’s daughter found the knob on the floor and gave it to Doe, who placed it in a seatback pocket. When it came time to descend, an Etihad flight. attendant (unaware of the detached knob) gave Doe the familiar reminder to .place her tray table in the upright and locked position for landing. Doe, of course, could not comply. To aid in explaining her problem, she reached into the seatback pocket to retrieve the fallen knob. But when she stuck her hand into the pocket, she was unexpectedly pricked by a hypodermic' needle that lay hidden within. She gasped, and the needle drew blood from her finger.

Doe claims damages from Etihad for both her physical injury and her “mental distress, shock, mortification, sickness and illness, outrage and embarrassment from natural sequela of possible exposure to” various diseases. Her husband claim's loss of consortium. The Montreal Convention of 1999, an international treaty under which these claims arise, imposes strict liability (up to a monetary cap) upon Etihad “for damage sustained in case of death or bodily injury of a passenger upon condition only that the accident which caused the death or injury took place on board the aircraft.” Etihad concedes that an accident onboard its aircraft caused Doe to suffer a bodily injury. But. Etihad argues that “damage sustained in case of ... bodily injury” means only “damage caused by bodily injury,” and thus does not include Doe’s fear of contagion and other emotional-distress and mental-anguish damages— damages that Etihad claims were caused not by Doe’s bodily injury (the small hole in her finger) but by the nature of the instrumentality of that injury (the needle). The district court agreed and granted partial summary judgment for Etihad; But the district court erred both in reading the additional “caused by” requirement into the treaty and in concluding that Doe’s bodily injury didn’t cause her emotional and mental injuries. The plain text of the Montreal Convention allows Doe to recover all her “damage sustained” from the incident, which includes damages for both physical injury and accompanying,emotional or mental harm. So, for the, reasons that follow, we reverse .and remand. s

I

. When Doe was pricked by the needle, the passenger seated in the aisle seat to her right heard Doe exclaim, “ouch,” and saw her finger bleeding. The Etihad flight attendant who had come to Doe’s seat picked up the needle and what was later determined to be its accompanying insulin syringe, both of which Doe had placed on her tray table. But the flight attendant then returned the items to the tray table and left to summon the assistance of her supervisor. Because the airplane had begun its descent, the flight attendants did not have access to the flight deck, which was where the only onboard sharps box was located, nor were the flight attendants permitted to call the flight deck absent a more pressing emergency.

The flight attendant returned with her supervisor. The flight attendant took the needle and syringe, placed them in. an empty water bottle, capped the bottle, and later turned the bottle over to her cabin manager. The supervisor, meanwhile, gave Doe an antiseptic wipe, which Doe used to wipe her finger, arid a Band-Aid, which the supervisor himself wrapped around her finger. The cabin manager wrote a report of the incident and told Doe that Etihad would contact her. A flight attendant recommended that Doe see a doctor, but Eti-had provided no medical assistance other than the antiseptic wipe and Band-Aid.

The next day, Doe saw a family physician, who noted a “small needle poke” on Doe’s finger. Doe was prescribed medication for possible exposure to hepatitis, tetanus, and HIV, and she underwent several rounds of testing over the following year. Thankfully for Doe, all the tests came back negative. Nevertheless, Doe claims that she refrained from sexual intercourse with her husband and from sharing food with her daughter until one year after the incident, when her doctor told her that she could be certain that she had not contracted a disease from the needles-tick.

Two days after the flight, Doe sent an email to Etihad to follow up because Eti-had had neither sent her a copy of the incident report nor offered her any further assistance. One week later, Etihad replied by email to offer a “purely goodwill gesture” of “possible reimbursement” of Doe’s medical expenses, “without any admission of liability.” This litigation followed.

II

Plaintiffs filed suit against Etihad in the United States District Court for the Eastern District of Michigan. Etihad, an entity wholly owned by the Government of Abu Dhabi, United Arab Emirates, is a “foreign state” within the meaning of the Foreign Sovereign Immunities Act, 28 U.S.C. § 1603(a). But as a condition of Etihad’s Foreign Air Carrier Permit—issued by the United States Department of Transportation to permit Etihad to fly to United States airports—Etihad waived sovereign immunity from suit in United States courts and could thus be sued “in any judicial district in which [Etihad] is licensed to do business or is doing business,” which includes the Eastern District of Michigan because of Etihad’s codeshare and other business agreements with airlines operating from points within that district. 28 U.S.C. § 1391(f); see 49 U.S.C. § 41301.

Following discovery, Etihad moved for, and the district court granted, partial summary judgment in favor of Etihad as to Doe’s claims for mental-anguish and emotional-distress damages, including fear of contagion. (For simplicity, we will refer to these various claims collectively as Doe’s claims for mental anguish.) The partial-summary-judgment order also dismissed Doe’s husband’s derivative claim for loss of consortium. Doe declined to pursue a lost-earnings claim that she had pleaded in her complaint, leaving only her claim for the physical pain, suffering, and medical expenses caused by the needles-tick, which the parties stipulated to be de minimis relative to the dismissed claims. (These de minimis damages include the physical pain and suffering from being pricked by the needle: the small hole in Doe’s finger and the “ouch,” so to speak. But they do not include any mental anguish arising from the fact that it was a stray needle and not, for example, a sterilized toothpick, that pricked Doe’s finger. The logic behind this distinction is that if something like a sterilized toothpick had caused Doe’s bodily injury, then Doe would not have had any reasonable fear of contagion, so Doe’s fear of contagion must arise from the fact that it was a needle that caused her injury, rather than arising from the injury itself, and Doe’s fear of contagion is therefore not recoverable as “damage sustained in case of bodily injury” under the Montreal Convention. This logic is faulty, of course, because Doe’s injury was an injury caused by a needle and was not the same as the injury that a sterilized toothpick would have caused, even if arguably similar. We will discuss this more fully in Section III.A, infra.) The parties reached a settlement as to these de min-imis damages, and the parties agreed to a “Stipulation and Order of Dismissal with Prejudice,” so that Plaintiffs could immediately appeal the district court’s partial-summary-judgment order.

We first discuss, in Section III, whether the district court erred in holding that Doe’s mental-anguish damages were not recoverable under Article 17(1) of the Montreal Convention, and—after analyzing both the plain text of the treaty and relevant persuasive authorities—we conclude that the district court did so err. Then, in Section IV, because the Montreal Convention provides rules for liability but looks to local law for the measure of damages, we conduct a choice-of-law analysis and hold that Michigan damages law governs both the amount of any damages Etihad comes to owe Doe and the ability of Doe’s husband to recover loss-of-consortium damages.

Ill

The parties agree that Article 17(1) of the Montreal Convention, a multilateral treaty to which the United States is a signatory, provides Plaintiffs’ only avenue for recovery against Etihad. See Convention for the Unification of Certain Rules for International Carriage by Air, art. 17, May 28, 1999, S. Treaty Doc. 106-45, ICAO Doc. No. 9740, 1999 WL 33292734 (entered into force Nov. 4, 2003) (Montreal Convention). More than 125 countries, including the United Arab Emirates, have signed, ratified, or acceded to the Montreal Convention since 1999.

The interpretation of a treaty is a question of law that we review de novo. United States v. Page, 232 F.3d 536, 540 (6th Cir. 2000). Under the Supremacy Clause, treaties are “the supreme Law of the Land.” U.S. Const. art. VI, cl. 2. Neither our court nor the Supreme Court has yet interpreted any provision of the Montreal Convention. The Warsaw Convention (the Montreal Convention’s longstanding predecessor treaty), however, has been the subject of much litigation over the past eighty years, and interpretations of the Warsaw Convention have at least some persuasive value in interpreting parallel provisions of the Montreal Convention. See Convention for the Unification of Certain Rules Relating to International Transportation by Air, Oct. 12, 1929, 49 Stat. 3000, T.S. 876, 137 L.N.T.S. 11 (Warsaw Convention); In re Air Crash at Lexington, Ky., 501 F.Supp.2d 902, 907-08 (E.D. Ky. 2007) (noting that “the ‘common law’ of the Warsaw jurisprudence is vitally important to understanding the meaning of the Montreal Convention”) (quoting Baah v. Virgin Atl. Airways, 473 F.Supp.2d 591, 596 n.7 (S.D.N.Y. 2007)); see also, e.g., Zicherman v. Korean Air Lines Co., 516 U.S. 217, 116 S.Ct. 629, 133 L.Ed.2d 596 (1996) (interpreting Warsaw Convention Article 17), Eastern Airlines, Inc. v. Floyd, 499 U.S. 530, 111 S.Ct. 1489, 113 L.Ed.2d 569 (1991) (same), Air France v. Saks, 470 U.S. 392, 105 S.Ct. 1338, 84 L.Ed.2d 289 (1985) (same). As with the Montreal Convention, the Warsaw Convention provided international air passengers’ exclusive, remedy for claims governed by that treaty. See, e.g., El Al Isr. Airlines, Ltd. v. Tsui Yuan Tseng, 525 U.S. 155, 168-69, 119 S.Ct. 662, 142 L.Ed.2d 576 (1999) (holding that the Warsaw Convention provided the sole remedy for personal-injury claims arising from injuries sustained during international air travel, even if the injured party could not state a claim for relief under the Warsaw Convention, in which case no remédy was available at all).

A. Textual Analysis

Our analysis of Article 17(1) of the Montreal Convention “must begin ... with the text of the treaty and the context in which [its] written words are used.” Saks, 470 U.S. at 397, 105 S.Ct. 1338 (citing Maximov v. United States, 373 U.S. 49, 53-54, 83 S.Ct. 1054, 10 L.Ed.2d 184 (1963)), The text of Article 17(1) provides:

The carrier is liable for damage sustained in case of death or bodily injury of a passenger upon condition only that the accident which caused the death or injury took place- on board the aircraft or in the course of- any of the operations of embarking or-disembarking.

Montreal Convention art. 17(1).

1. Etihad’s Argument

The contested language here is “in case of.” Etihad’s argument has two components: its understanding' of what “in case of’ means, and its application of that understanding to the facts of this case.

First, Etihad argues that “in case of’ means “caused by,” Appellee’s Br. 4, or perhaps “caused directly by,” see id. at 21. If we impose Etihad’s reading of Article 17(1) back onto the text of the treaty, Etihad is then “liable for damage sustained [caused directly by] death or bodily injury of a passenger upon condition only that the accident which caused the death or injury took place on board the aircraft....” Thus, according to Etihad, in order for Doe to recover for her mental anguish under Article 17(1), Doe would have to prove that (1) an “accident” caused her “bodily injury” on board an aircraft and (2) her “bodily injury” (i.e. the small hole in her finger) directly caused her “damage sustained” (i.e., her mental anguish).

Second, Etihad concedes that an accident caused Doe to suffer a bodily injury on board its aircraft, but Etihad argues that Doe’s bodily injury did not directly cause her mental anguish: according to Etihad,, Doe’s anguish was caused not by her “bodily injury” (i.e., the needlestick, the physical puncture wound) but rather by the “accident” that caused the injury (i.e., being stuck by a needle, as opposed to being stuck by. something else). Order Granting Def.’s Mot. for Partial Summ. J. 4 (emphasis added) (citations- omitted) (“Plaintiffs mental distress damages were not caused by her physical injury. It is not the physical needle prick itself that caused Plaintiffs distress, but the possibility that she may have been exposed to an infectious disease.”); see Appellee’s Br.. 17 (“[Doe’s] mental anguish damages arise from the nature of the accident itself and were not caused by the bodily injury”), id. at 20 (“because the plaintiffs’ mental injuries were caused- by the ‘accident’ itself and not the- ‘bodily injuries’ sustained in the accident, there could be no recovery under the Convention”).

A simple diagram helps to illustrate Eti-had’s curious understanding:

As this diagram indicates, according to Etihad, mental anguish caused directly by the bodily injury is recoverable, but mental anguish that merely accompanies the bodily injury, and which is instead caused more generally by the accident, is not recoverable.

2. Plain Meaning of the Text

But “in case of’ does not mean “caused by.”

Rather, the plain meaning of “in case of’ is “if there is” .or “in the event, of’ or “during a case in which there is.” The Oxford English Dictionary, for example, defines “in case” (as a conjunction)' as, “In the event that; if it should happen that; if,” and defines “in ease of’ (as an adverb) as “in the event of (esp. something untoward). Now frequently in in case of emergency.” In case, Oxford English Dictionary, http:// www.oed.com/view/Entry/426263 (last visited Aug. 29, 2017). The Canadian Oxford Dictionary has similar definitions and is a seemingly apt dictionary for identifying the contemporaneous meaning of terms in the Montreal Convention, given that the dictionary was first published in 1998 and then updated in 2004, while the treaty was signed in 1999 (in Canada) and entered into force in 2003. See Case, The Canadian Oxford Dictionary (2d ed. 2004), http:// www.oxfordreference.com/view/10.1093/ acref/9780195418163.001.0001/m_en_ca 0011030?rskey=8fa6U0&result=11001 (defining “in case” as “in the event that; if,” and defining- “in case of’ as “in the event of’).

Clearly, the plain meaning of “in case of’ is conditional, not causal. To say in case of X, do Y is to say “if X happens, then do Y”—none of which means that there is a causal relationship between X and Y—just as to say in case of a compen- sable bodily injury, the passenger may recover damage sustained is to say “if there is a compensable bodily injury, the passenger may recover damage sustained.” But to adopt Etihad’s meaning of “in case of,” we would impose an additional causal restriction onto the text of Article 17(1) that the plain text does not contemplate. Indeed, imposing such an additional causal restriction would contradict the plain text, which states that '“[t]he carrier is liable for damage sustained in case of ... bodily injury ... upon condition only that the accident which caused the death or bodily injury took place on board the aircraft or [while] embarking or disembarking.” Montreal Convention art. 17(1) (emphasis added).

The phrase “upon condition only” is new to the Montreal Convention—it is not found in the Warsaw Convention (either in English or in the official French version)—and it makes clear that the passenger’s recovery is conditioned only on the occurrence of an accident that causes death or bodily injury either on board the aircraft or during boarding or deplaning. Surely, the drafters of the Montreal Convention could have used a word or phrase with causal meaning instead of “in case of’-if they wanted to impose such a causal restriction on the kinds of “damage sustained” that are recoverable when an accident on board an aircraft causes a passenger to incur a bodily injury, Indeed, the drafters did impose such a causal requirement in stating that the accident must have “caused” the death or bodily injury. The drafters’ use of “caused” to express that an accident must have caused the bodily injury thus provides additional support for our conclusion that the drafters did not, in the very same sentence, use “in case of’ also to mean “caused by.”

3. The Underpinnings of Etihad’s Argument

Admittedly, in light of the foregoing discussion, Etihad’s position—that “in case of’ does mean “caused by”—may seem absurd. But it is not, and that is because Etihad’s argument is rooted in a Warsaw Convention decision of the Second Circuit Court of Appeals in which that court held that American Airlines was not liable under the Warsaw Convention “for mental injuries that were not caused by physical injuries.” Ehrlich v. Am. Airlines, Inc., 360 F.3d 366, 368 (2d Cir. 2004) (emphasis added). Etihad asks us to adopt the Second Circuit’s Warsaw Convention decision in Ehrlich to decide the Montreal Convention case before us. But the Montreal Convention is a new treaty that we interpret as a matter of first impression, and there is no legal authority that would require us to import Ehrlich’s Warsaw Convention determination to govern this Montreal Convention claim.

In Ehrlich, an American Eagle aircraft overshot its designated runway upon landing at New York’s JFK International Airport. An arrestor bed—a bed of material made of water, foam, and cement that crushes under the weight of an airplane, increasing drag and helping bring the airplane to a stop—saved the plane from plunging into the waters of Thurston Bay, which lay 200 feet beyond where the plane came to a halt. To evacuate the aircraft, passengers had to jump six to eight feet from its doorway. Ibid.

Gary and Maryanne Ehrlich were passengers on the flight. They contended that they suffered bodily injuries (neck, back, shoulder, hip, and knee injuries; hypertension; and a heart problem) during the abnormal landing and subsequent evacuation. They also alleged mental injuries including a fear of flying, nightmares, and trouble sleeping. The district court granted partial summary judgment for the airline defendant as to the mental injuries on the basis that “a plaintiff may only recover for emotional damages caused by physical injuries.” Id. at 369 (quoting Ehrlich v. Am. Airlines, 99-CV-6013, 2002 U.S. Dist. LEXIS 21419, at *10 (E.D.N.Y. June 21, 2002) (emphasis added)). The Second Circuit affirmed, noting that “the Ehrlichs had offered no evidence demonstrating a causal connection between their mental and physical injuries.” Ehrlich, 360 F.3d at 369.

Ehrlich reached its conclusion only after grappling at length with the original French text of the Warsaw Convention, finding it ambiguous as to whether it held airlines liable for mental injuries that are-, not caused by a compensable bodily injury, and inquiring into the original purpose of the Warsaw Convention when it was signed in 1929. Indeed, Ehrlich discussed the Montreal Convention as well: the Montreal Convention was signed just weeks .after the Ehrlichs’ emergency landing, and the Montreal Convention entered into force after the Second Circuit heard argument in Ehrlich but before it issued its opinion. See id. at 372. But Ehrlich expressly rejected the argument that .the Montreal Convention had any retroactive applicability to the Ehrlichs’ claim, and the Second Circuit based its decision entirely on its interpretation of the Warsaw Convention. See id. at 373 (“neither the Montreal Convention nor the intentions of .its drafters govern this appeal”).

In reaching its conclusion, Ehrlich followed the lead of Jack v. Trans World Airlines, 854 F.Supp. 654, 663-68 (N.D. Cal. 1994), a district-court decision that also concluded that “only emotional distress flomng from the bodily injury is recoverable” under Article 17 of the Warsaw Convention. Id. at 665 (emphasis added). Jack expressly acknowledged (after rejecting other possible interpretations of the Warsaw Convention) that its interpretation “does read a causal component into the phrase ‘damage sustained in the event of,’” but nevertheless went ahead with such an interpretation because that interpretation was “not prohibited” by the United States Supreme Court’s Warsaw Convention precedents. Id. at 668.

But “to alter, amend, or add to any treaty, by inserting' any clause, whether small or great, important or trivial, [is] an usurpation of power, and not an exercise of judicial functions.” The Amiable Isabella, 19 U.S. (6 Wheat.) 1, 71, 5 L.Ed. 191 (1821) (holding that the 1795 U.S.-Spain Treaty for safe passage of ships did not protect a Spanish claimant from United States condemnation of a schooner during the War of 1812 when the requisite passport mandated by the, treaty was not affixed to the vessel). Both Ehrlich and Jack interpolated a causal component into the Warsaw Convention that was not required by the text, and both did so expressly to serve the Warsaw Convention’s purpose of “limiting the liability of air carriers in order to foster the growth-of the fledgling commercial aviation industry.” Ehrlich, 360 F.3d at 385 (quoting Floyd, 499 U.S. at 546, 111 S.Ct. 1489); see also Jack, 854 F.Supp. at 662, 665 (“such an approach furthers the pro-airline industry goals of the Warsaw Convention because it is so restrictive of passengers’ rights”). To be sure, both Ehrlich and Jack found ambiguity ip the original French text of the Warsaw Convention before inquiring into the purpose of that treaty and seeking to give' effect to that purpose. But what that should mean for us is not, as Etihad would have it, that we should blindly adopt Ehrlich as the law of our circuit for claims under Article 17(1) of the Montreal Convention, but rather that we should grapple with the text of the Montreal Convention'itself, and then, to the extent that we find any ambiguity therein, look to relevant persuasive authority—which may include evidence of the purpose of the Montreal Convention, but almost certainly not the nearly century-old purpose of the Warsaw Convention—to assist us in resolving that ambiguity.

Ehrlich recognized that “the Montreal Convention is an entirely new treaty that unifies and replaces the system of liability that derives from the Warsaw Convention.” Ehrlich, 360 F.3d at 371 n.4. So do we. The Montreal Convention was signed in 1999, in six languages including English, and we are charged with interpreting that English text in the first instance rather than clinging to the Second Circuit’s pur-posivist interpretation' of a French-language predecessor treaty signed in 1929. In Sections III.B through III.E, infra, to fortify our textual analysis of Article 17(1)', we will discuss more fully the relative purposes of the Warsaw' and' the Montreal Conventions, 'and we will address relevant decisions of the United Statés Supreme Court and other cóurts, which provide useful context for both Ehrlich-and our decision here. But for now, it suffices to say that Ehrlich and Jack do hot provide insight into meaning of' the plain text of Article 17(1) of the Montreal Convention.

4. Our Textual Interpretation

Here, then, is a fairer illustration of what damages are recoverable under Article 17(1) according to the plain text of the Montreal Convention:

As this diagram makes clear, because an accident onboard Etihad’s aircraft caused Doe to suffer a bodily injury (a fact that Etihad concedes), Doe may therefore recover damages for her mental anguish, regardless of whether that anguish was caused directly by her bodily injufy or more generally by the accident that caused the bodily injury. That is because, either way, Doe’s mental anguish is “damage sustained in case of’—i.e,, “in the event of’ a compensable bodily injury.

What the plain text of Article 17(1) also makes clear is that a passenger cannot recover damages for mental anguish if there is no requisite accident or if the accident does not cause a bodily injury. For example, if ordinary turbulence causes a passenger to suffer an anxiety attack, the Montreal Convention would not allow the passenger to recover damages for the anxiety attack because ordinary turbulence is not an “accident.” Likewise, if there is an accident, such as an emergency landing, and a passenger escapes physically unscathed but mentally harmed, the passenger is barred from recovering mental-anguish damages for want of the required bodily injury. This understanding is supported by the plain text of Article 17(1) of the Montreal Convention—and it also happens to have the advantage of being simpler than Ehrlich’s approach.

Admittedly, however, the text of Article 17(1) is still not entirely clear as to what connection must exist between the required bodily injury and claimed mental anguish. The plain text of Article 17(1) is sufficient on its own to reject Etihad’s interpretation of it. And the plain text of Article 17(1) allows our conclusion that when a single “accident” causes both bodily injury and mental anguish, that mental anguish is sustained “in case of’ the bodily injury. But the plain text on its own does not necessarily require that a single accident cause both the required bodily injury and the claimed mental anguish in order for that mental anguish to be “sustained in case of’ the bodily injury, as our conclusion suggests.

What if, for example, there are two accidents: first, unusually rough turbulence (which causes a passenger mental anguish but no bodily injury), and second, an unrelated emergency landing, during which every passenger sustains at least some bodily injury. Does the bodily injury sustained in the emergency landing allow the passenger who had previously suffered severe emotional distress to recover for that distress? That is, is mental anguish from the first accident considered “damage sustained in case of bodily injury” because it was sustained during the same flight as the second ■ accident, which caused bodily injury?

On the one hand, it seems reasonable to read the “in case of’ language as precluding recovery of damages for mental anguish in the example presented in the preceding paragraph, and our interpretation of Article 17(1) implicitly supports such a conclusion; but on the other hand, the text of the treaty does not explicitly prohibit such recovery. So, both to bolster our conclusion that mental anguish is “sustained in case of’ a bodily injury when it arises from the same accident that caused that bodily injury, and to reinforce the proposition that Ehrlich does not control this case, we review relevant persuasive authorities that provide insight into the meaning of Article 17(1) in the context of its ratification by its signatories. See, e.g., Saks, 470 U.S. at 396, 105 S.Ct. 1338 (“[Tjreaties are construed more liberally than private agreements, and to ascertain their meaning we may look beyond the written words to the history of the treaty [and] the negotiations” that produced the treaty, (alteration in original) (quoting Choctaw Nation of Indians v. United States, 318 U.S. 423, 431-32, 63 S.Ct. 672, 87 L.Ed. 877 (1943))). We therefore turn next to the history of the negotiations that culminated in the signing of the Montreal Convention and to evidence of the signatories’ purpose in ratifying the Montreal Convention.

This historical inquiry is important because the question before us is important. And the question before us is important for several reasons. First, Article 17(1) governs not only claims for need-lesticks, snakebites, and the like, but also claims for injuries and fatalities sustained in plane crashes. Second, “uniformity is an important goal of treaty interpretation,” Sanchez-Llamas v. Oregon, 548 U.S. 331, 383, 126 S.Ct. 2669, 165 L.Ed.2d 557 (2006), so we look to the history of the Montreal Convention to ensure that the conclusion we draw today is consistent with how our sister signatories would understand the text of Article 17(1). Third, the Warsaw Convention’s analogue to the question before us was expressly left unanswered by the United States Supreme Court in Floyd when it ruled that mental injury standing alone was not recoverable under the Warsaw Convention because of the absence of the required death or bodily injury:

We conclude that an air carrier cannot be held liable under Article 17 when an accident has not caused a passenger to suffer death, physical injury, or physical manifestation of injury. Although Article 17 renders air carriers liable for “damage sustained in the event of’ ... such injuries, we express no view as to whether passengers can recover for mental injuries that are accompanied by physical injuries. That issue is not presented here because respondents do not allege physical injury or physical manifestation of injury,

Floyd, 499 U.S. at 552-53, 111 S.Ct. 1489.

Fourth, although we have expended considerable effort explaining that the Montreal Convention is a new treaty that we should interpret independently of the Warsaw Convention, such that Ehrlich does not inform our decision here, there is nonetheless evidence that the drafters of the Montreal Convention intended Article 17(1) to be construed consistently with well-settled Warsaw Convention precedents of the United States Supreme Court. See Explanatory Note to Montreal Convention, art. 17, S. Treaty Doc. 106-45, 1999 WL 33292734, at *16 (2000). These precedents, as we will discuss, include the United States Supreme Court’s decision defining “accident” (in Saks, a decision from 1985 that has gained global currency). But these precedents do not include Ehrlich, which was decided well after the Montreal Convention was signed (and which was not a decision of a signatory’s highest court, in any event, see, e.g., Husain, 540 U.S. at 655 n.9, 124 S.Ct. 1221).

And finally, while Ehrlich was a Warsaw Convention decision, we recognize that our conclusion today is directly contrary to Ehrlich’s conclusion as to a similarly worded provision. The history behind the Montreal Convention will make clear why the conclusion we reach today is correct, and why we cannot use the same lines of reasoning that Ehrlich and Jack used in reaching their holdings that denied recovery for mental injuries that accompanied but did not directly flow from a bodily injury.

B. History and Purpose of the Montreal Convention

The Warsaw Convention was opened for signature in 1929, just two years after Charles Lindbergh famously flew his Spirit of St. Louis solo from New York to Paris, and eight years before Amelia Earhart disappeared over the Pacific Ocean, The original parties to the Warsaw Convention had the “primary purpose of ... limiting the liability of air carriers in order to foster the growth of the fledgling commercial aviation industry.” Floyd, 499 U.S. at 546, 111 S.Ct. 1489 (citing Trans World Airlines, Inc. v. Franklin Mint Corp., 466 U.S. 243, 256, 104 S.Ct. 1776, 80 L.Ed.2d 273 (1984); Minutes, Second Int’l Conf. on Private Aeronautical Law, October 4-12, 1929, Warsaw 37 (R. Horner & D. Legrez trans. 1975) (“Warsaw Conference Minutes”); and Lowenfeld & Mendelsohn, The United States and the Warsaw Convention, 80 Harv. L. Rev. 497, 498-99 (1967)).

The Warsaw Convention itself was the product of four years of work by a committee of experts that was appointed in 1925 at an international conference in Paris at which an early draft protocol was circulated. That draft protocol included an expansive liability provision, holding the carrier “liable for accidents, losses, breakdowns, and delays” without imposing any requirement of death or bodily injury. Ministére des Affaires Étrangéres, Conference Internationale de Droit Privé Aérien (27 Octobre-6 Novembre 1925), 79 (1926), as translated in Floyd, 499 U.S. at 542, 111 S.Ct. 1489.

By the time the conference in Warsaw began in 1929, the committee had divided the protocol on liability into three separate provisions (one for injury to passengers, one for damage to goods, and one for losses from delays). This text was then further developed in Warsaw until the final version of the Warsaw Convention was agreed upon—with much narrower language in Article 17 for air carriers’ liability to injured passengers. See Warsaw Conference Minutes at 205-06; Floyd, 499 U.S. at 543, 111 S.Ct. 1489. Moreover, unlike the Montreal Convention’s strict-liability scheme, the Warsaw Convention imposed a cap on damages at 125,000 gold French francs (at the time, approximately $8,300) per passenger, which carriers could reduce to zero upon showing that they had exercised due care by taking “all necessary measures to avoid the damage or that it was impossible” to do so. The cap on damages was lifted (so as to allow potentially unlimited liability) only if the carrier’s “■willful misconduct” caused the injury or death. Warsaw Convention arts. 17, 20, 22.

The United States Supreme Court has stated that “it is reasonable to infer that the Conference adopted the narrower language [in Article 17] to limit the types of recoverable injuries.” Floyd, 499 U.S. at 543, 111 S.Ct. 1489. “Whatever may be the current view among Convention signatories, in 1929 the parties were more concerned with protecting air carriers and fostering a new industry rather than providing a fuli recovery to injured passengers.” Id. at 546, 111 S.Ct. 1489.

The Warsaw Convention, entered into force in 1933, and the United States became a party to it in 1934. Paul S. Dempsey & Michael Milde, International Air Carrier Liability: The Montreal Convention of 1999, 13 (McGill Univ. Centre for Research in Air & Space Law) (2005). The United States subsequently led various efforts to modernize it and raise its liability limits. See Montreal Convention, 1999 WL 33292734, at *3-5 (Letter of Submittal from President Clinton to United States Senate) (“Letter of Submittal”) (detailing history of Warsaw Convention and proposed modifications). In the early 1950s, the newly created International Civil Aviation Organization (ICAO) began evaluating a potential increase to the liability limits at international conferences in Rio. de Janeiro and The Hague. At The Hague, the United States proposed raising the personal-liability limits to approximately $25,000, but the majority of other participants resisted; the United States countered with a reduced proposal of approximately $20,000, which was also met with disapproval. “It.was not until the United States began to threaten denunciation” that any agreement to increase the personal-liability limits was reached, and even then, the United States “succeeded only in doubling the original Warsaw Convention liability limit to $16,600,” in a proposed amendment to the Warsaw Convention known as the Hague Protocol. Dempsey & Milde, supra, at 19 n.51; Letter of Submittal, 1999 WL 33292734, at *3; see Protocol to Amend the Convention for the Unification of Certain Rules Relating to International Carriage by Air, done at The Hague September 28, 1955. The United States, dissatisfied with the low liability limits, refused to ratify the Hague Protocol.

In 1965, in response to what some courts have described as the “unconscionably low” liability limits under the Warsaw Convention, Dunn v. Trans World Airlines, Inc., 589 F.2d 408, 411 (9th Cir. 1978), United States Secretary of State Dean Rusk gave Poland six months’ notice that the United States intended to denounce the Warsaw Convention. Dempsey & Milde, supra, at 29 n.87 (citing Dep’t of State Press Release No. 268, 50 Dep’t of State Bull. 923-24 (1965)). The notice included a proviso that the United States would retract its notice of denunciation if personal-liability limits were raised to $75,000 to $100,000 per passenger.

As a result of this notice, the ICAO held a conference in Montreal in 1966 at which the United States unsuccessfully sought to increase the. personal-liability limits. The airlines themselves, however—including all major air carriers that served the United States—entered into a private intercarrier agreement (the Montreal Agreement) that made two broad changes to the Warsaw Convention’s limitations. First, the Montreal Agreement increased the personal-liability limit to $75,000 per passenger. Second, the Montreal Agreement imposed strict liability up to the $75,000 limit (while retaining the preexisting provision that allowed liability beyond that limit upon a showing of willful misconduct by the airline). CAB Order E-23680 (May 13, 1966), 31 Fed. Reg. 7,302 (May 19, 1966), reprinted at 49 U.S.C.A. § 1502 (1970). The United States retracted its notice of denunciation. Dempsey & Milde, supra, at 30. The Montreal Agreement remained in force among its signatories for approximately thirty years, and was applicable to all carriage to, from, or through the United States.

In the wake of the Montreal Agreement of 1966, various other international agreements were also reached to increase liability. In 1974, various European and Japanese carriers agreed to increase passenger liability in an informal “Malta Agreement.” Id. at 31. In 1992, Japanese carriers agreed to strict liability for personal injury up to 100,000 Special Drawing Rights per passenger. And in 1995, a dozen airlines signed a “Washington Intercarrier Agreement,” endorsed by the International Air Transport Association, to which the United States Department of Transportation had given antitrust immunity to facilitate discussion of the modernization of international air-carrier liability. Id. at 33-34. This Washington Intercarrier Agreement, signed in Kuala Lumpur, imposed strict liability up to 100,000 SDRs per passenger and removed the “willful misconduct” provision for liability beyond the cap, replacing that provision with something more like a negligence standard that imposes unlimited liability above the 100,000-SDR cap if the airline cannot prove that it took “all necessary measures” to avoid the injury. The major United States-based airlines joined the Washington Intercarrier Agreement within a week of its initial signing in Kuala Lumpur. Ibid.

Also in the wake of the Montreal Agreement of 1966, aside from the private inter-carrier agreements that were negotiated, the United States continued to seek amendments to the Warsaw Convention that would impose higher personal-liability limits. In 1971, the Guatemala City Protocol came close to achieving a limit of 1,500,000 gold francs (then equivalent to approximately $100,000) per passenger, but that Protocol would have imposed an absolute limitation on liability, even in cases of willful misconduct. See id. at 22-26. The United States Senate refused to ratify the Guatemala City Protocol in part because it used the gold standard for liability limits and because it would have imposed an absolute, unbreakable limitation on liability. In 1975, various “Montreal Protocols” were proposed at a diplomatic conference as part of an initiative to replace the Warsaw Convention’s gold standard with the SDR. But the only protocol that entered into force worldwide was Protocol No. 4, which affected only cargo liability and not personal-injury liability. See id. at 26-29.

Against that backdrop, the Montreal Convention of 1999 was revolutionary: it replaced not only the Warsaw Convention but also “all of its related instruments and ... eliminate[d] the need for the patchwork of regulation and private voluntary agreements” that then dominated the world’s air-carrier liability regime. Letter of Submittal, 1999 WL 33292734, at *7. The Montreal Convention imposes strict liability for injuries that are compensable under Article 17(1), up to 100,000 SDRs per passenger, with a decennial adjustment for inflation. (The first official adjustment came in 2009, increasing the strict-liability limit to 113,100 SDRs—or approximately $160,000—per passenger. Montreal Convention arts. 21, 24; see Inflation Adjustments to Liability Limits Governed by the Montreal Convention Effective Dec. 30, 2009, 74 Fed. Reg. 59,017 (Nov. 16, 2009).) Above that strict-liability limit, a carrier remains liable for all damage sustained, with no limit, unless the carrier can prove either that “such damage was not due to the negligence or other wrongful act or omission of the carrier or its servants or agents,” or that “such damage was solely due to the negligence or other wrongful act or omission of a third party.” Montreal Convention art. 21. Finally, an exoneration provision allows a reduction in compensation for injuries caused by or contributed to by the plaintiff, in the same manner as a pure-comparative-negligence or pure-comparative-fault scheme; this exoneration provision applies to all claimed damages including those falling under the strict-liability limit. Montreal Convention art. 20. In short, the Montreal Convention replaced a “restrictive,” “pro-airline industry” regime, Jack, 854 F.Supp. at 662, 665, with “a treaty that favors passengers rather than airlines.” Lexington, 501 F.Supp.2d at 908 (quoting Ehrlich, 360 F.3d at 371 n.4 (describing Montreal Convention)). And it did so on terms that reflected decades of effort by the United States to abolish the outdated limitations of the Warsaw Convention.

Moreover, by 1999, when the Montreal Convention was opened for signature, the aviation industry was anything but “fledgling,” and the purpose of the Montreal Convention was not to protect the aviation industry, but rather to provide a “modernized uniform liability regime for international air transportation.” Letter of Sub-mittal, 1999 WL 33292734, at *6.

In light of the great difference between the purpose of the Warsaw Convention and the purpose of the Montreal Convention, then, it hardly seems appropriate for us to look to the purpose of the Warsaw Convention, as Etihad would have us do in relying on Ehrlich, in order to arrive at a different conclusion from one compelled by the plain text of the Montreal Convention. Our Supreme Court’s Warsaw Convention jurisprudence has relied consistently on analysis of the purpose of that treaty as it was implemented in 1929. See, e.g., Zicherman, 516 U.S. at 221-23, 116 S.Ct. 629; Saks, 470 U.S. at 400-05, 105 S.Ct. 1338; Floyd, 499 U.S. at 546, 111 S.Ct. 1489. What the historical record makes clear is that the considerations favoring a close textual reading of the Montreal Convention—a product of at least five decades of international negotiations—far outweigh whatever considerations would weigh in favor of rewriting the text of the Montreal Convention in order to accommodate Ehrlich or effectuate the purpose of the Warsaw Convention, as Etihad would have us do.

C. Relevant Warsaw Convention Litigation

We turn next to (1) relevant decisions of our Supreme Court under the Warsaw Convention; (2) a brief summary of our reasons for rejecting Ehrlich in light of the foregoing discussion of the history and purpose of the Montreal Convention; and (3) a brief discussion of relevant district-court cases.

1. United States Supreme Court Decisions Under the Warsaw Convention

From 1984 to 2004, the United States Supreme Court handed down a series of seven opinions clarifying various aspects of the Warsaw Convention, most of which involved Article 17. The first of these, Trans World Airlines, Inc. v. Franklin Mint Corp., 466 U.S. 243, 104 S.Ct. 1776, 80 L.Ed.2d 273 (1984), upheld the Convention’s liability limit for cargo and is not particularly relevant to our case. More relevant is the Court’s 1986 decision in Air France v. Saks, in which it held that “accident” in Article 17 of the Warsaw Convention means “an unexpected or unusual event or happening that is external to the passenger.” Saks, 470 U.S. at 405, 105 S.Ct. 1338 (denying passenger’s recovery for deafness caused by cabin depressurization where the depressurization was ordinary and the plaintiff was the only passenger on the flight who was affected),

A year later, the Court decided Chan v. Korean Air Lines, Ltd., 490 U.S. 122, 109 S.Ct. 1676, 104 L.Ed.2d 113 (1989), holding that the Warsaw Convention’s limitations applied even if an airline defendant failed to provide notice of the Convention in at least 10-point type as the airline defendant agreed to in the Montreal Agreement of 1966. Chan, 490 U.S. at 135, 109 S.Ct. 1676 (holding that although Korean Air Lines had joined the Montreal Agreement in 1969 and had violated that agreement by providing notice of the Convention only in 8-point, type, the Warsaw Convention’s limitations still applied because the Montreal Agreement did not impose any sanction at all for failure to provide notice in the required typeface, let alone the sanction of forfeiting liability limitations).

In 1991, the Court decided Eastern Airlines v. Floyd, which as we noted earlier held that “bodily injury” in Article 17 of the Warsaw Convention does not allow for the recovery of mental injuries on their own (that is, with no physical injury incurred whatsoever), but which “expressed] no view as to whether passengers can recover for mental injuries that are accompanied by physical injuries.” Floyd, 499 U.S. at 552, 111 S.Ct. 1489. Next came Zicherman v. Korean Air Lines in 1996, in which the Court held that although the Warsaw Convention provided rules for liability and limitations of liability, it did not govern the measure (or calculation,'so to speak) of damages, which was instead a matter to be determined in each case by applicable domestic law. Zicherman, 516 U.S. at 225, 116 S.Ct. 629 (“ [Questions of who may recover, and what compensatory damages they may receive, ... were unresolved by the Convention and left to ‘private international law’— i,e., to the area of jurisprudence we call ‘conflict of laws,’ dealing with the application of varying domestic laws to disputes that have an interstate or' international component.”).

In 1999, the Court decided El Al Israel Airlines, Ltd. v. Tsui Yuan Tseng, in which it held that the Warsaw Convention provided the sole remedy for personal-injury claims arising from injuries sustained during international air travel, even if the injured party could not state a claim for relief under the Warsaw Convention, in which case no remedy was available at all. Tseng, 525 U.S. at 161, 119 S.Ct. 662 (alleged assault by El Al agents during pref-light security search that did not result in bodily injury was- not an “accident” and was not compensable under the Warsaw Convention; the Warsaw Convention , nevertheless continued to preempt local claims for damages from the assault).

Finally, in 2004, the Court decided Olympic Airways v. Husain, in which it clarified that finding an “accident” to have occurred for the purpose of applying Article 17 does not require identifying'a single “injury producing event” -but may rather involve a chain of causation that results in death or bodily injury, so long as there is an unexpected or unusual happening external to the passenger in that chain, following Saks. Husain, 540 U.S. at 651-54, 124 S.Ct. 1221 (when asthmatic passenger died after flight attendant repeatedly but wrongly refused to reseat him away from the smoking section, the failure to reseat counted as an “accident,” even though the presence of ambient smoke in the cabin— not itself an unexpected or unusual happening on a flight that allowed smoking— could be viewed as an “injury producing event”).

Saks, Chan, Floyd, and Zicherman were all decided unanimously, and Franklin Mint and Tseng were both decided eight to one over the dissent of Justice Stevens. Husain was decided six to two, with Justice Scalia dissenting, arguing that because two other Warsaw Convention signatories (England and Australia) had rejected the proposition that an airline’s inaction could constitute an “accident” under Article 17, and because the text of Article 17 did not clearly resolve that issue, the Court should instead have followed the English and Australian decisions. Husain, 540 U.S. at 659-64, 124 S.Ct. 1221 (Scalia, J., dissenting).

These opinions have enjoyed wide acceptance among our sister signatories, which have given them (especially Saks, Tseng, and Floyd) at least some deference and have developed théir own jurisprudence using these opinions as guideposts. See, e.g., Plourde c. Service aérien F.B.O. inc., 2007 QCCA 739, para. 29 (Court of Appeal of Québec) (applying Floyd to deny recovery for purely psychological injury in a Montreal Convention case); Povey v. Qantas Airways Ltd. (2005) 223 CLR 189, 190 (High Court" of Australia) (applying Saks and considering Tseng and Husain in Warsaw Convention case); King v. Bristow Helicopters Ltd [2002] UKHL 7 (House of Lords) (applying Saks, Tseng, and Floyd in Warsaw Convention case).

Because these Supreme Court cases analyzed aspects of the Warsaw Convention that we have no reason to believe have changed following "the ratification of the Montreal Convention (and that neither party has argued have changed following the ratification of the Montreal Convention), it is reasonable to conclude that these cases form part of the “precedent” consistent with which, according to the Explanatory Note (see n.11, supra), the drafters expected signatories to construe Article 17(1) of the Montreal Convention. Accordingly, we have adopted Saks’s definition of “accident,” and our discussion of damages in Section IV will be guided by Zicherman’s deference to the forum jurisdiction’s choice-of-law rules.

2. Why the Second Circuit’s Ehrlich Decision Does Not Govern Montreal Convention Claims

In light of the discussion in Sections III.A and III.B, there are several reasons why we decline to adopt Ehrlich to govern Doe’s claims. First, Etihad’s argument that we should adopt Ehrlich is unconvincing in part because of how thorough Ehrlich itself is: Ehrlich reaches its conclusion only after plumbing the depths of the original French meaning of the Warsaw Convention, Ehrlich, 360 F.3d at 376-78 (analyzing whether “dommage survenu en cas de ... lésion corporelle” incorporates a requirement that the bodily injury [lésion corporelle] cause the damage [dommage]), French legal materials, id. at 380, the purpose of the Warsaw Convention, id. at 385, and the “negotiating history” of the Convention, ibid. Indeed, if Ehrlich is persuasive, it is persuasive not for the conclusion it reached but for how it got there, and our similarly searching analysis leads us to a conclusion opposite Ehrlich’s.

Second, Ehrlich interpreted the authoritative French text of the Warsaw Convention, and found ambiguity in that text (in the original French) that Ehrlich thought could accommodate a causal meaning. Specifically, Ehrlich examined French-language dictionaries and found that the word “cas” in “en cas de” (the French phrase that was the Warsaw Convention analogue to the Montreal Convention’s “in case of’) could actually mean “cause.” Ehrlich, 360 F.3d at 377-78 (“If ‘cas’ means ‘cause,’ then the phrase ‘dommage survenu en cas de ... lésion corporelle,’ as those words are used in Article 17, would hold carriers liable for any ‘damages sustained in the cause of ... bodily injury.’ Such a translation is amenable to an interpretation that would allow passengers to recover for mental injuries only where they were caused by a bodily injury.”). Setting aside the fact that the French word cas does not actually mean “cause” except perhaps in the same way that we might say a “hopeless case” is a “lost cause,” the range of ambiguity in the English “in case of’ is far, far narrower than the range of ambiguity that Ehrlich found in the French “en cas de” and, as we concluded in Section III.A, notwithstanding any ambiguity in the English “in case of,” the plain text of the English “in case of’ does not contain a requirement that “damages sustained” be “caused by” bodily injury.

Third, as we discussed in Section III.B, the purpose of the Montreal Convention vastly differs from the purpose of the Warsaw Convention, such that we have no reason to interpret Article 17(1) of the Montreal Convention in such a way as to serve the purposes of the Warsaw Convention, as Ehrlich did.

Fourth, although the Ehrlich court stated that its ruling was necessary to avoid anomalous results, it appears that under the Montreal Convention, following Ehrlich would be more rather than less likely to lead to anomalous results. Ehrlich explained its reasoning as follows:

The interpretation of Article 17 favored by the [plaintiffs] would give rise to anomalous and illogical consequences because “similarly situated passengers [would be] treated differently from one another on the basis of an arbitrary and insignificant difference in their experience.” For example, a passenger who sustained a mental injury but no bodily injury would be unable to look to Article 17 for relief whereas a co-passenger who suffered the same mental injury yet fortuitously pinched his little finger in his tray table while evacuating and thereby suffered an unrelated bodily injury would be able to hold the carrier liable under the Warsaw Convention.

Ehrlich, 360 F.3d at 386.

But our interpretation of Article 17(1) of the Montreal Convention does not necessarily imply this result. Ehrlich’s hypothetical here presumably involves some sort of crash or emergency landing (no context is provided in Ehrlich itself for the portion quoted above). Under our interpretation of the Montreal Convention, if an airplane crash-landed, then any passenger who sustained a bodily injury caused by that crash-landing would also be permitted to recover for mental anguish sustained in that crash-landing—i.e., anguish sustained “in case of’ a compensable bodily injury. If a passenger sustained a broken leg, that passenger would be able to recover for the broken leg, for mental anguish caused by the broken leg, and for mental anguish arising from the crash-landing that accompanied the broken leg—all of that would be “damage sustained in case of’ the broken leg. True, another passenger escaping the same crash-landing physically unscathed would be barred from recovering damages for mental anguish alone, but that’s not an “anomalous” result. Rather, it is a result that is fully consistent with (and compelled by) the text of the Montreal Convention.

Returning to Ehrlich’s hypothetical of the passenger who escapes the same crash-landing entirely unscathed except for a pinched pinky finger: what result? As we read the Montreal Convention, so long as that passenger can prove that the accident (i.e. the crash-landing) caused the injury to the pinky finger, that passenger would be able to recover both for the physical injury to the finger and for mental anguish sustained—and that passenger would be able to recover mental anguish sustained on the same terms as the passenger who suffered the broken leg.

None of these outcomes produce an “illogical or unreasonable result” that might caution against our ruling today. Int’l Union, United Auto., Aerospace & Agric. Implement Workers of Am. v. Brock, 816 F.2d 761, 766 (D.C. Cir. 1987). Thus, we are not persuaded by Ehrlich that we must read in an additional causation requirement to avoid interpreting the Montreal Convention in a manner that would produce absurd results.

To the contrary, it would be odd to require the passenger to prove which mental injuries in fact were caused by the physical injury as opposed to being caused more generally by the accident. In our crash-landing hypothetical, a passenger might, for example, be conscious for the duration of the crash-landing and then realize that he has suffered a grievous injury to his leg. Perhaps the passenger fears losing the leg for hours or days while he is in the hospital. And perhaps the passenger, who has a compensable bodily injury (bruised ribs and a broken leg), suffers mental anguish and other emotional damages—some as a result of the fear of losing the leg (which surely would be caused by the bodily injury, even if the leg was ultimately not lost), and some as a result of having experienced the crash-landing.

It would not “favor[ ] passengers,” Ehrlich, 360 F.3d at 371 n.4, to require the passenger to prove which mental harms were caused directly by the broken leg as opposed to being caused more generally by the accident. After- all, causation is difficult. Surely, for example, harm such as insomnia, fear of flying (or other fears), or emotional distress might initially result from the crash-landing but then be exacerbated by the bodily injury. Would only the portion of harm traceable and subsequent to the bodily injury be recoverable? Or, what if some of the harm resulted from the realization of an imminent crash-landing— should that harm be excluded from recovery because its cause preceded the accident, while harm occurring together with or flowing from the crash-landing would be recoverable?

Thus, at the end of the day, adopting Ehrlich would mean requiring Doe and other Montreal Convention plaintiffs to prove causation in a way that burdens the injured passenger far more than the text requires; that would be an anomalous result.

For all these reasons, we decline to adopt Ehrlich.

S. Relevant District Court Cases

We now turn to Jack, the district-court opinion that Ehrlich followed, and which was the first district-court opinion to analyze the text and history of the Warsaw Convention at any serious length. Jack followed on the heels of the Supreme Court’s decision in Floyd. In the Jack case, where fire consumed a plane following an aborted takeoff and crash but all the passengers survived, the court addressed the question whether Floyd’s bar to recovering purely mental damages under the Warsaw Convention also meant that mental anguish was recoverable only if caused by bodily injury. Jack, 854 F.Supp. at 657. The court posited four theories for the recovery of emotional-distress damages under Article 17:

1. Emotional-distress damages are never recoverable. Id. at 665.

2. Emotional-distress damages are always recoverable as long as the plaintiff has a bodily injury, even if the bodily injury is wholly unrelated to the emotional distress. Id. at 665-66 (notably, the court found that this approach “would read emotional distress as damages resulting from the accident (as opposed to the injury), which is difficult to do under the wording of Article 17,” although the court did not explain the difficulty). This theory is broader than our interpretation, in that it allows recovery for mental injuries that are wholly unrelated to a compensable bodily injury, which would seem to capture more than just those mental injuries “sustained in case of’ a com-pensable bodily injury.

3. Emotional-distress damages are recoverable as “an element of the damages for bodily injury,” but “need not be about the injury,” so long as the distress occurs “at the same time or later than the bodily injury.” Id. at 666-67 (noting that in a plane crash that caused an injury, distress about the plane crash would be recoverable so long as it occurred after the injury, just as federal common law would allow the victim of a racially motivated false arrest to recover for emotional distress subsequent to physical injuries sustained, and not only for the minor physical injuries). This theory attempts to limit the scope of recovery to something narrower than what-t