Austrian Airlines (Premiers soins a bord d'un aeronef) (Air transport - Montreal Convention - Liability of air carriers for death or bodily injuries sustained by passengers - Opinion) [2023] EUECJ C-510/21

BAILII is celebrating 24 years of free online access to the law! Would you consider making a contribution?
No donation is too small. If every visitor before 31 December gives just £1, it will have a significant impact on BAILII's ability to continue providing free access to the law.
Thank you very much for your support!

	[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	Court of Justice of the European Communities (including Court of First Instance Decisions)
You are here: BAILII >> Databases >> Court of Justice of the European Communities (including Court of First Instance Decisions) >> Austrian Airlines (Premiers soins a bord d'un aeronef) (Air transport - Montreal Convention - Liability of air carriers for death or bodily injuries sustained by passengers - Opinion) [2023] EUECJ C-510/21_O (12 January 2023) URL: http://www.bailii.org/eu/cases/EUECJ/2023/C51021_O.html Cite as: [2023] EUECJ C-510/21_O, EU:C:2023:19, ECLI:EU:C:2023:19

(Request for a preliminary ruling from the Oberster Gerichtshof (Supreme Court, Austria))

(Reference for a preliminary ruling – Air transport – Montreal Convention – Liability of air carriers for death or bodily injuries sustained by passengers – Exclusivity of the Convention – Article 29 – Scope – Claims related to bodily injuries sustained by passengers as result of ‘accidents’ within the meaning of Article 17(1) – Claim for compensation, based on domestic rules of civil liability, relating to injuries allegedly caused by the inadequate first aid administered by the flight attendants after an accident – Sufficient causal nexus between the injuries and the accident – Claim exclusively governed by Article 17(1) – Claim pre-empted by the Convention)

1. The Convention for the Unification of Certain Rules for International Carriage by Air (2) (‘the Montreal Convention’) and its precursor, the Warsaw Convention, (3) have, throughout their many years of application, been the subject of considerable judicial controversy. One issue, among those at the very centre of debate, is the extent to which those conventions, which notably deal with the liability of air carriers in case of death or bodily injury sustained by passengers, prevent – or, in other words, ‘pre-empt’ – claims for compensation based not on their provisions, but on domestic law.

2. That controversy resurfaces in the present case, which concerns an action for compensation brought by DB against Austrian Airlines AG (‘Austrian Airlines’). DB claims, on the basis of the Austrian rules of civil liability, damages for injuries he sustained, as a passenger on an international flight operated by that carrier, allegedly as a result of the flight attendants’ failure to administer proper first aid after a jug of scalding coffee spilled on him during the flight. As he did not lodge his claim under the Montreal Convention and since, in any event, he initiated proceedings after the expiration of the limitation period laid down therein – but in accordance with the longer one set out in Austrian law for civil liability actions – the pivotal issue at hand, which lies at the heart of the two questions that the Oberster Gerichtshof (Supreme Court, Austria) referred to the Court of Justice, is whether that convention pre-empts such a claim. In this Opinion, I will explain why that is indeed the case.

3. The third recital of the Montreal Convention states that the States Parties ‘[recognise] the importance of ensuring protection of the interests of consumers in international carriage by air and the need for equitable compensation based on the principle of restitution’.

4. The fifth recital of that convention states that ‘collective State action for further harmonisation and codification of certain rules governing international carriage by air through a new Convention is the most adequate means of achieving an equitable balance of interests’.

5. Article 17 of the Montreal Convention, entitled ‘Death and injury of passengers – damage to baggage’, provides in paragraph 1 that ‘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’.

6. Pursuant to Article 29 of that convention, entitled ‘Basis of Claims’, ‘in the carriage of passengers, baggage and cargo, any action for damages, however founded, whether under this Convention or in contract or in tort or otherwise, can only be brought subject to the conditions and such limits of liability as are set out in this Convention without prejudice to the question as to who are the persons who have the right to bring suit and what are their respective rights. …’.

7. Article 35 of the same convention, entitled ‘Limitation of Actions’, provides, in paragraph 1, that ‘the right to damages shall be extinguished if an action is not brought within a period of two years, reckoned from the date of arrival at the destination …’.

8. Article 3(1) of Council Regulation (EC) No 2027/97 on air carrier liability in respect of the carriage of passengers and their baggage by air, (4) as amended by Regulation (EC) No 889/2002 (5) (‘Regulation No 2027/97’), states that ‘the liability of a Community air carrier in respect of passengers and their baggage shall be governed by all provisions of the Montreal Convention relevant to such liability’.

9. On 18 December 2016, DB travelled from Tel Aviv (Israel) to Vienna (Austria) on a flight operated by Austrian Airlines, pursuant to a contract of carriage by air concluded with that company.

10. During the flight, a jug of coffee fell from a serving trolley being manoeuvred by the attendants through the rows of seats. Hot coffee spilled and scalded DB. The attendants subsequently administered some form (6) of first aid on the injuries sustained.

11. On 31 May 2019, DB brought, on the basis of the Austrian rules of civil liability, and within the three-year limitation period set out therein, (7) an action against Austrian Airlines before the Handelsgericht Wien (Commercial Court, Vienna, Austria), seeking damages to the amount of EUR 10 196 and a declaration establishing the carrier’s liability for all future injury arising from that incident. In essence, DB claimed that he suffered severe burns and that, under Austrian law, Austrian Airlines ought to be liable not only for the carelessness of its employees which led to the fall of the jug, but also for the allegedly insufficient and inadequate first aid that was administered on his injuries thereafter. (8)

12. By judgment of 17 June 2020, the Handelsgericht Wien (Commercial Court, Vienna) dismissed the action in its entirety. In essence, that court considered that the claim at issue was governed exclusively by the Warsaw Convention and that, in accordance with that instrument, it was time-barred. Indeed, pursuant to Article 29 thereof, any claim for compensation against an air carrier is subject to a two-year limitation period, and DB brought proceedings after that period had elapsed.

13. By judgment of 28 October 2020, the Oberlandesgericht Wien (Higher Regional Court, Vienna, Austria) upheld that decision. Considering the Montreal Convention to be applicable to the facts at hand, (9) it ruled that DB’s claim fell within the scope of Article 17(1) of that convention. Indeed, DB’s injuries were to be regarded as caused by an ‘accident’ that took place on board the aircraft, within the meaning of that provision, namely the fall of the jug of scalding coffee, even if those injuries could possibly have been mitigated or avoided by the proper administration of first aid. Accordingly, that claim was time-barred since it was brought outside of the two-year limitation period laid down in Article 35 of that convention.

14. DB lodged an appeal on a point of law against that judgment before the Oberster Gerichtshof (Supreme Court). While admitting that the fall of the jug of coffee constituted an ‘accident’ within the meaning of Article 17(1) of the Montreal Convention, DB submitted that the allegedly inadequate first aid administered on his injuries thereafter had been a separate and autonomous cause of damage that did not fall within the scope of that provision. Therefore, in so far as his claim rested on that particular cause, that claim was governed not by that convention, but by Austrian law, in accordance with which it would not time-barred.

15. It is in those circumstances that the Oberster Gerichtshof (Supreme Court) decided to stay the proceedings and to refer the following questions to the Court of Justice for a preliminary ruling:

‘(1) Is first aid which is administered on board an aircraft following an accident within the meaning of Article 17(1) of the [Montreal Convention] and which leads to further bodily injury to the passenger which can be distinguished from the actual consequences of the accident to be regarded, together with the triggering event, as a single accident?

Does Article 29 of [the Montreal Convention] preclude a claim for compensation for damage caused by the administration of first aid where that claim is brought within the limitation period under national law but outside the period for bringing actions which is laid down in Article 35 of [that] convention?’

16. The request for a preliminary ruling, dated 5 August 2021, was lodged on 19 August 2021. Written observations have been submitted by DB, Austrian Airlines, the German Government, as well as the European Commission. No hearing was held in the present case.

17. The Montreal Convention is a treaty establishing certain uniform rules for international carriage by air. Since that convention was concluded by, inter alia, the European Union, (10) it has been an integral part of the EU legal order from the date on which it entered into force as far as that organisation is concerned, namely 28 June 2004. Accordingly, as of that date, the Court has had jurisdiction to deliver preliminary rulings concerning its interpretation. (11)

18. As the referring court observes, the flight at issue in the main proceedings falls within the general scope of the Montreal Convention. Indeed, DB concluded a contract of carriage with Austrian Airlines and that contract concerned an ‘international carriage’, within the meaning of Article 1 of that convention, as the place of departure and the place of destination of that flight are situated within the territories of two States Parties, namely the State of Israel and the Republic of Austria. (12)

19. Chapter III of the same convention contains several provisions relating to the liability of air carriers. In particular, Article 17(1) deals with their liability in case of death or ‘bodily injury’ sustained by passengers, such as DB, on board an aircraft (13) during an international flight.

20. Actions for damages, under that provision, are subject to a two-year limitation period, set out in Article 35(1) of that convention, which commences from the date of the arrival of the flight in question at its destination. (14) Here, it is not disputed that DB brought proceedings against Austrian Airlines almost three years after the flight during which he sustained injuries landed in Vienna. (15) Accordingly, under the law of the Convention, irrespective of the merits of his claim, DB would have no remedy, since that claim would simply be time-barred.

21. That said, as I indicated in the introduction to this Opinion, the pivotal issue is whether the Montreal Convention also prevents DB from bringing a civil liability claim against Austrian Airlines on the basis of his domestic law.

22. The referring court explains, in that respect, that, under Austrian law, the limitation period generally applicable to a civil liability claims is three years. (16) Under that law, DB’s claim would not be time-barred and could proceed on the merits, which would then be assessed in the light of the conditions of liability set out therein. Thus, under domestic law, the applicant may have a remedy against the defendant.

23. In view of those circumstances, and for the purposes of providing as much guidance as possible, I shall begin my analysis with the second question of the referring court concerning, in essence, the scope of the so-called ‘exclusive’ and ‘pre-emptive’ effect of the Montreal Convention (A). In my view, clarifications on that issue should be made from the outset for it will help the reader understand the relevance of the first question – which relates, in substance, to whether DB’s injuries must be regarded as caused by an ‘accident’ within the meaning of Article 17(1) of that convention. That first question will accordingly be discussed subsequently (B).

A. The scope of the pre-emptive effect of the Montreal Convention (second question)

24. As follows from the previous section, when passengers sustain injuries during an international flight, the Montreal Convention provides them, in certain circumstances, with a cause of action – that is, a legal basis for the carrier’s liability. Article 17(1) (17) of that convention refers to, more specifically, a situation where a passenger dies or suffers a ‘bodily injury’ because of an ‘accident’ (18) which took place on board the aircraft. In such circumstances, a claim may be brought under that provision against the air carrier responsible for the flight at issue, the success of which depends on the conditions set out in that convention – including the limitation period laid down in Article 35(1) thereof.

25. In addition, injuries sustained by passengers on international flights may potentially give rise to causes of action under domestic law. Indeed, various harmful events occurring on board an aircraft may theoretically qualify as breaches of the contract of carriage, torts under the general rules of civil liability, and so on – each of those causes, of course, coming with its own set of conditions, including limitation periods to bring proceedings.

26. Those different causes of action sometimes relate to distinct matters. If a passenger purchases, in advance, a meal to be served during the flight and the carrier does not provide it, that usually constitutes an actionable breach of contract under domestic law. However, in that scenario, there is neither ‘bodily injury’ nor ‘accident’ and, thus, no claim under Article 17(1) of the Montreal Convention. By contrast, causes of action ‘accumulate’ in certain circumstances. The same harmful event, such as the inadvertent spilling by a flight attendant of a scalding hot drink on a passenger, resulting in the latter’s bodily injury, may simultaneously be regarded, for example, as (i) negligence under domestic tort law, (ii) a breach of the obligation of safety arising from the contract of carriage, and (iii) an ‘accident’ under that provision. (19) In such a scenario, the applicant would theoretically have a choice of legal basis to bring a claim against the carrier, and his or her skilful legal counsel would naturally pick the one most favourable to his or her interest – a choice most obvious where, for instance, one of the causes of action is time-barred but the other(s) are not.

27. The issue of the different and sometimes competing rules of liability potentially applicable to air carriers with respect to passenger injuries was addressed by the drafters of the Montreal Convention. In order to tackle it, they intended to give a certain ‘exclusive’ effect to the ones set out therein, and notably Article 17(1). To make that clear, a specific provision was included in that convention, namely Article 29, which states, in relevant parts, that ‘in the carriage of passengers … any action for damages, however founded, whether under this Convention or in contract or in tort or otherwise, can only be brought subject to the conditions and such limits of liability as are set out in this Convention … ’.

28. That said, clarity on the matter was not fully achieved, to say the least. In fact, that last provision, like Article 24 of the Warsaw Convention before it, (20) has been the subject of intense controversy, as emphasised by the referring court in the present case.

29. A first interpretative controversy, concerning the method of exclusivity, can be addressed briefly. On the one hand, Article 29 can be read as providing that, where the Montreal Convention exclusively applies, a claim can only be brought under that convention, to the outright exclusion of domestic causes of action. On the other hand, it can also be understood in the sense that, in such a situation, an applicant may bring a claim under a domestic cause of action, but when he or she does so, the conditions and limits of liability laid down in the Convention must still be respected. While the second interpretation is, in my view, the most natural reading of that provision, (21) the first interpretation seems to be the dominant view. (22) That being said, that controversy bears little, if any, practical consequences. Both methods give, in fact, the same imperative effect to the Convention: when it applies exclusively, liability of the carrier can exist only where, and to the extent that, that instrument provides so, and an applicant cannot circumvent it by pleading his or her case under domestic law.

30. Far more significant is the controversy surrounding the scope of exclusivity of the Montreal Convention – that is, the issue of which liability claims against air carriers fall within its exclusive reach. This calls for a more detailed examination.

31. As the referring court observes, there are two opposing views on the matter. According to a first view, which I shall describe as ‘broad’, the Montreal Convention governs exclusively all potential claims against air carriers, however pleaded, for any injury sustained by passengers during an international flight falling within the general scope of that convention, (23) irrespective of its cause. Under that view, carriers may be held liable only in the circumstances envisioned in Article 17(1), that is to say, in case of a passenger’s death or ‘bodily injury’ caused by an ‘accident’. In other circumstances – where, for instance, a passenger sustained pecuniary loss, or where the cause of the injury was something other than an ‘accident’ – not only is there no remedy available under that provision, but any possibility of relief under domestic law is also excluded under Article 29 of that convention.

32. Pursuant to a second view, which I shall describe as ‘narrow’, the Montreal Convention does not govern all claims against carriers that may arise from international travel by air, but only the ones relating to passenger deaths or ‘bodily injuries’ caused by ‘accidents’, as envisioned in Article 17(1). When a claim, however pleaded, objectively satisfies that definition, Article 29 of that convention prevents the applicant from relying on more favourable conditions and limits of liability set out in domestic law. By contrast, other types of injuries sustained by passengers on board an aircraft are beyond the scope of the Convention: while no remedy is available under that instrument, an applicant is free to bring a claim against the carrier on the basis of domestic law.

33. In the present case, the issue of whether the first or the second view is the correct one has a direct impact on the relevance, as regards the outcome of the dispute in the main proceedings, of the referring court’s first question.

34. Indeed, if one adopted the broad view of the exclusivity of the Montreal Convention, then it would not matter, for the purposes of deciding whether DB’s claim could proceed on the basis of Austrian law, whether his injuries were caused by an ‘accident’ within the meaning of Article 17(1) or by something else. Under that view, that claim would be regarded as governed exclusively by that convention – and pre-empted by the latter as time-barred – for the simple reason that it concerns injuries sustained by a passenger during an international flight falling within the general scope of that convention, irrespective of their actual cause. (24)

35. Conversely, if one were to follow the narrow view, the question of the cause of DB’s injuries would be decisive for the fate of his claim. Under that view, that claim would be regarded as governed and pre-empted by the Montreal Convention only if his injuries were ‘accident’-related. If not, it could proceed on the basis of national law.

36. Thus, from a logical standpoint, the second question, regarding the scope of exclusivity of the Montreal Convention, could have been asked first. That said, the order in which the referring court chose to refer its questions makes perfect sense from a pragmatic point of view. Indeed, that complex and sensitive issue (1) does not need to be entirely decided in the present case for, as I will explain, a claim such as DB’s is pre-empted anyway (2).

37. One may be forgiven for considering that the issue of the scope of exclusivity of the Montreal Convention has been settled. Although this Court has never taken a position on it, (25) the courts of other States Parties have. The broad view was endorsed by no less than the Supreme Court of the United Kingdom in its judgment in Sidhu (26) and the Supreme Court of the United States in its judgment in Tseng, soon followed by numerous other supreme and high-level courts throughout the world. (27) While Sidhu and Tseng concerned the Warsaw Convention, the solution has already been transposed (28) to the Montreal Convention. In addition, many experts in academic literature have come out in its support. (29) Understandably, Austrian Airlines also suggests, before the Court, that it is the correct and established view on the matter.

38. I have already indicated, in my Opinion in Austrian Airlines (Exoneration of air carrier from liability), (30) that, since the Court is but one of many jurisdictions throughout the world that are competent to interpret the Montreal Convention, and since the uniform application of that convention in all States Parties is an aim to be pursued, it is appropriate for the Court to duly take into account, and give the required weight, to the decisions handed down by the courts of those States Parties.

39. That being said, obviously, the Court should not follow such national precedents without question. (31) A cautious review of the ratio decidendi underpinning the solutions endorsed in other State Parties and of the practical consequences that stem from them is always called for.

40. In that regard, the trend set by Sidhu and Tseng has not gone uncontested. Back when it was decided, Tseng was seen by some as a significant overturn of a good number of decisions going in the opposite direction, delivered by lower courts in the United States. (32) Furthermore, as DB indicates, those judgments are still criticised by a part of the academic literature. (33) In fact, DB, the German Government and the Commission suggest – expressly or impliedly – that, should the Court address the issue in the present case, it should depart from those national precedents and endorse the narrow view instead.

41. No doubt, that controversy stems, in the first place, from the fact that the scope of exclusivity of the Montreal Convention is, from a theoretical point of view, a complex question. The applicable rules of interpretation, as codified in the Vienna Convention on the Law of Treaties of 23 May 1969, (34) do not lead to an unequivocal answer. On the one hand, the broad view, as endorsed in Sidhu and Tseng, rests on a solid good-faith interpretation of the terms of the Montreal Convention in their context and in the light of its object and purpose. Its travaux préparatoires were also used by those two Supreme Courts to confirm that interpretation. On the other hand, the very same elements can, in my view, reasonably support the opposite construction.

42. As regards the terms of the Montreal Convention, and of Article 29 more specifically, proponents of the broad view point to the fact that its language is not limited to injuries caused by ‘accidents’, but refers more generally to actions for damages arising from ‘the carriage of persons’ – an expression from which one should imply that all claims for passenger injuries are concerned. (35) By contrast, supporters of the narrow view emphasise, as DB does before the Court, that nothing in the text of Article 29 expressly mandates the exclusion, to the detriment of passengers, of any domestic cause of action even when no ‘accident’ is involved. Such a drastic outcome could only result from clear and unambiguous language, (36) not from implication.

43. The object and the purpose of the Montreal Convention are equally equivocal. It is not disputed, that, pursuant to its very title and preamble, (37) that instrument aimed at unifying only ‘certain rules’ for international carriage by air – it deals with some of the issues arising from air carriage, not all of them. However, there is fundamental disagreement as to what its purpose exactly is when it comes to the liability of air carriers.

44. Proponents of the broad view submit that liability of air carriers, generally, is an issue that the Montreal Convention was intended to address. In their opinion, the rules set out in its Chapter III constitute a comprehensive scheme for claims arising from international carriage by air. The main purpose of that convention is to achieve uniformity of law in that respect. Those rules lay down the circumstances – that is, the only circumstances – where carriers ought to be liable for injuries caused to passengers. By restricting the types of claims that may be brought against them, the Convention, and notably Article 29, provides certainty to air carriers. It ensures that they can determine and calculate in advance the compensation burden imposed on them – something essential for, inter alia, insurance purposes. That objective of uniformity and certainty would be frustrated, were passengers able to bring other claims against them. (38)

45. Like DB and the German Government, supporters of the narrow view reply that, in their opinion, the Montreal Convention was actually intended to unify the carriers’ liability in certain situations only, notably in the case of aircraft-related accidents. Indeed, the risk of carriers facing crippling liability in the event of airplane crashes was one of the concerns that led to the adoption of the Warsaw Convention in 1929. (39) It was not supposed to shield carriers from liability in other circumstances. Read in that light, the sole aim of Article 29 of the Montreal Convention would be to prevent the applicant, where he or she brings a claim that is objectively ‘accident’-related, from circumventing the conditions and limits of liability set out therein by pleading his or her case under domestic law. Uniform application of the Convention would require no more than that. (40)

46. Supporters of the narrow view also emphasise a difference in purpose between the Warsaw and Montreal Conventions. As DB and the Commission point out, whereas the former was adopted to foster the development of the nascent airline industry, the latter was intended to further consumer protection in international carriage by air. (41) Regardless of what the correct approach under the Warsaw Convention should have been, the broad view on exclusivity cannot, at least for that reason, be transplanted to the Montreal Convention. (42) Proponents of the broad view reply that, beyond merely consumer protection, the Montreal Convention was intended to achieve ‘an equitable balance’ between the interests of passengers and those of carriers. (43) That balance is inherent in the liability regime of the Convention. Against carriers, passengers have limited remedies. However, when such remedies are available, they are obtained easily and swiftly, in large part due to the strict liability regime set out therein. (44)

47. The preparatory works of both conventions do not mitigate the controversy since they are also ambiguous as regards the intention of their drafters vis-à-vis exclusivity. Indeed, the travaux préparatoires of the Warsaw Convention do not clearly point to a definitive intention on the part of the delegates on the matter. (45) Those of the Montreal Convention are not helpful either. The only direct and substantial statement in that respect, from the conference Chairman, is open to interpretation. (46)

48. In the second place, the controversy surrounding the scope of exclusivity of the Montreal Convention arises from the significant – and often dire – practical consequences that the broad view entails. In Sidhu, passengers who were detained for three weeks by Iraqi forces after their plane landed at Kuwait International Airport in August 1990, in the first hours of the invasion of Kuwait by the Iraqi forces, were denied a remedy. So was, in Tseng, a passenger who was subjected to a possibly abusive security search by the airline’s personal at John F. Kennedy International Airport in New York before she boarded a flight to Tel Aviv. In other decisions, passengers whose handicaps were not accommodated properly by the airline, in direct breach of EU law requirements, (47) or who were allegedly discriminated against by carriers on prohibited grounds, such as their race, were also denied a remedy. (48) Clear breaches of contract, such as the failure of airlines to deliver pre-ordered meals, mentioned in point 26 above, also remained uncompensated. (49) Indeed, in all of those cases, the claims directed against the responsible carriers were deemed to be exclusively governed by the Warsaw or Montreal Convention because they arose from international carriage by air. However, since there was neither ‘bodily injury’ nor ‘accident’, no relief could be obtained under the relevant convention. Yet, applicants were not allowed to proceed on the basis of domestic law.

49. Supporters of the narrow view point to the unfairness of those outcomes. The broad view amounts, in many cases, to denying justice to passengers. Beyond the few hypotheses contemplated in the Montreal Convention itself, carriers are shielded from any kind of liability, regardless of the source and purpose, even liability that would otherwise stem from disregarding statutory duties and/or passenger’s fundamental rights. In effect, those duties and rights cannot be – privately – enforced against them. (50) Proponents of the broad view maintain that those outcomes are but a necessary implication of the uniformity and certainty that that convention was intended to achieve. While sometimes acknowledging the unfairness for passengers, particularly in discrimination cases, they emphasise, not without merit, that it is not for the courts to redraft an international treaty to make it fairer. (51)

50. All in all, the issue of the scope of exclusivity of the Montreal Convention calls for some serious reflexion on the part of the Court. Granted, as the Commission submits, in its judgments in IATA and ELFAA (52) and Nelson and Others, (53) the Court already slightly touched upon the subject and adopted, in a way, a cautious view. However, when the Court ruled, in those judgments, that there was no conflict between the Montreal Convention and Regulation (EC) No 261/2004, (54) it mostly avoided the issue, by ruling that the first instrument covered individual actions for damages only, whereas the second one contains obligations to assist passengers and standardised compensation. That said, the hostile reactions in response to those two judgments, as found in a significant part of the academic literature, further emphasise the sensitive nature of the debate. (55)

2. It is not necessary for the Court to take a position in the present case

51. Amidst the obscurity of all that controversy, there is one thing that shines through nonetheless: as I indicated before, it is not necessary for the Court to take a complete position on the scope of exclusivity of the Montreal Convention in the present case.

52. Indeed, as the broad view encompasses the narrow one, they coincide on one aspect. At the very least, pursuant to Article 29 of the Montreal Convention, a claim against an air carrier, however pleaded, which objectively relate to death or bodily injuries sustained by a passenger during an international flight falling within the general scope of that convention, because of an accident that took place on board the aircraft, as envisioned in Article 17(1) of that convention, is, without any doubt, governed exclusively by that instrument. In such a scenario, it is generally accepted that the applicant is not allowed to circumvent the conditions and limits of liability laid down in that instrument by pleading his or her case under domestic law. (56) It is only with respect to claims that involve neither passenger death or bodily injury, nor accident, that the two views diverge: under the narrow view, the applicant may proceed on the basis of domestic law; under the broad view, he or she is denied any potential remedy. (57)

53. In the present case, as I will explain in the section below, in my view, a claim such as the one brought by DB against Austrian Airlines objectively relates to a passenger’s bodily injury caused by an accident within the meaning of Article 17(1). Thus, the Court does not need to choose between the broad and narrow views. Irrespective of the correct view, that claim is, in any case, governed exclusively by that convention and pre-empted by the latter as time-barred. (58)

54. In my opinion, it would be a laudable display of judicial restraint if the Court were to limit itself here to that obvious finding, and refrain from taking a definitive position on the wider issue of the scope of exclusivity of the Montreal Convention. There may be, in the future, cases in which the Court will have to swallow that pill. The example of discrimination claims brought by passengers against air carriers may come before it. If and when such a case arrives, it would be best decided by the Grand Chamber, after careful consideration of all the aspects laid out above.

B. Injuries such as the ones sustained by DB are to be regarded as having been caused by an ‘accident’ within the meaning of Article 17(1) of the Montreal Convention (first question)

55. I have already indicated, in the previous section, that, in my view, a claim such as the one brought by DB against Austrian Airlines is governed exclusively by the Montreal Convention, since it objectively relates to the situation envisioned by Article 17(1) of that convention. Consequently, the applicant is unable to circumvent the two-year limitation period set out therein by pleading his case under domestic law. I shall now explain my view on that matter.

56. In the main proceedings, it is not disputed that DB’s claim relates to ‘bodily injuries’ – namely, severe burns – which were sustained by a passenger – that is, himself – during an international flight falling within the general scope of the Montreal Convention – as explained in point 18 above – and that the event(s) that led to those injuries took place on board the aircraft. The sole contested issue is whether or not those injuries must, in law, be regarded as having been caused by an ‘accident’ within the meaning of Article 17(1).

57. In that regard, I recall that, during the flight transporting DB to Vienna, a jug inadvertently fell from a serving trolley that was being manoeuvred by the attendants through the rows of seats. Scalding coffee was spilled, leaving DB with severe burns. (59)

58. As the referring court observes, such an unfortunate incident undoubtedly qualifies as an ‘accident’ within the meaning of Article 17(1) of the Montreal Convention. That is so, irrespective of whether one applies the definition of that concept stated many years ago by the United States Supreme Court in Air France v. Saks (60) – ‘an unexpected or unusual event or happening that is external to the passenger’ – or the novel definition adopted by the Court in the judgment in Niki Luftfahrt – ‘an unforeseen, harmful and involuntary event’. (61) Obviously, when boarding an aircraft a person does not expect/foresee that a jug of hot coffee will be spilled over himself or herself during the flight. (62)

59. That finding is not contested by DB. (63) However, before the court of appeal and the referring court, the applicant framed his claim as not being about that ‘accident’, but rather about what took place afterwards. Indeed, pursuant to the theory of recovery upon which DB relies, his claim stems from another cause, that is, the allegedly (64) insufficient and inadequate first aid that the flight attendants administered on his wounds, (65) in breach of the duty of care that carriers owe to their passengers. That cause is distinct from the initial ‘accident’ and has resulted in a separate injury, namely, the aggravation of his burns. It is only for that particular injury that he seeks compensation.

60. Therefore, in DB’s view, his claim – as it was pleaded – solely concerns bodily injuries which have been caused not by the ‘accidental’ fall of the jug of scalding coffee, but by the subsequent response of the cabin crew. Accordingly, that claim is not governed by Article 17(1) of the Montreal Convention and, thus, may proceed on the basis of Austrian law, under the conditions – notably the limitation period – set out therein.

61. In the light of that argument, the referring court wonders, by its first question, whether the fall of the jug of scalding coffee, and the first aid which was administered by the crew on the burns thereafter, should be treated as separate causes of injuries or as forming part of a single ‘accident’ within the meaning of Article 17(1) of the Montreal Convention.

62. In my view, that first question should be slightly reformulated. Indeed, the case in the main proceedings, seen in light of DB’s argument, raises, quite plainly, an issue of causation. In essence, whether or not the applicant’s claim falls within Article 17(1) of the Montreal Convention and, as such, is governed exclusively by that instrument, depends on whether or not the initial accident that took place on board the aircraft, that is to say the fall of the jug, can be regarded as having ‘caused’, within the meaning of that provision, the injuries in respect of which he is seeking compensation, given that the administration of first aid by the crew intervened in between. That issue should be addressed as such. (66) By contrast, it would be inappropriate to deal with it somewhat indirectly, under the concept of ‘accident’, by wondering whether those two distinct factors should be treated as one event for the purposes of the same provision. (67)

63. Accordingly, I will explain in the passages below why the ‘accidental’ fall of the jug must, for the purpose of Article 17(1) of the Montreal Convention, be regarded, in law, as having ‘caused’ the bodily injuries which are the object of DB’s claim, irrespective of the fact that the administration of first aid thereafter may have also contributed to those injuries (1). (68) Moreover, as Austrian Airlines has raised before the Court the issue of whether, in any case, that second event could also, in itself, qualify as an ‘accident’ within the meaning of that provision, I will address it briefly for the sake of completeness (2).

1. The ‘accidental’ fall of the jug must be regarded, in law, as having ‘caused’ the bodily injuries sustained for the purpose of Article 17(1) of the Montreal Convention

64. As a starting point on the matter, it is worth emphasising that, although Article 17(1) of the Montreal Convention does not specify in which situations a given accident may be regarded as having ‘caused’ a passenger’s death or bodily injury, and although that Convention generally does not define that term, it should not be interpreted by reference to the domestic law applicable to the contract of carriage at issue, (69) as DB and Austrian Airlines suggest. In fact, considering the objective of uniformity pursued by that convention, (70) and since the latter does not say otherwise, (71) that term should instead be given an independent construction, in the light of the rules of interpretation laid down in the Vienna Convention. (72) Independence from domestic law, however, should not mean complete disregard thereof. When it comes to a fundamental concept of liability law such as causation – the particulars of which can hardly be deduced, with intellectual honesty, from the ‘ordinary meaning’ of that term, the ‘context’ in which it is used and the ‘object and purpose’ of the Montreal Convention alone – one must also draw inspiration from the general principles which are common to the laws of the States Parties. (73)

65. In accordance with those rules of interpretation, and in the light of those principles, I am of the view, together with all the interveners before the Court, that, in order to determine, in each case, whether the relevant ‘accident’ can be regarded as having ‘caused’ the passenger injury which is the object of the claim, two complementary tests should be applied consecutively.

66. As Austrian Airlines, the German Government and the Commission submit, the first test stems directly from the ordinary meaning of the verb ‘to cause’ used in Article 17(1), namely ‘to make something happen’. That test is a factual one. It corresponds to what is known in the laws of the State Parties as ‘sine qua non’, ‘but for’, or ‘equivalence of conditions’ causation. Under that test, any conduct or event which, in fact, constitutes a necessary condition of a given injury – that is, but for which it would not have been sustained – is regarded as having caused the latter. Since any injury is the result of a combination of factors, each contributing to its occurrence, (74) it is viewed, under the ‘but for’ test, as having not one, but multiple causes. Indeed, all of these factors are considered to be factual causes of that injury – or ‘links’ in the ‘chain of causes’ that lead to it. (75)

67. Thus, for the purposes of Article 17(1) of the Montreal Convention, the first test is satisfied when the relevant ‘accident’ is, in fact, a necessary condition of the passenger injury which is the object of the claim – that is, but for that event, the injury would not have been sustained. It also means that it suffices that the ‘accident’ is one of the factors that contributed to the bringing about of that injury – a ‘link’ in the ‘chain of causes’ that led to it. As the referring court noted, that approach has been expressly endorsed, notably, by the United States Supreme Court in Air France v. Saks. (76)

68. In my view, that first test is consistent both with the objective of consumer protection (77) and with the objective of uniformity pursued by the Montreal Convention. In that regard, let us consider the hypothetical example of a passenger’s bodily injury, produced by the combination of – at least – two factors, namely (i) the prior health condition of the passenger in question, which made him or her prone to heart attacks, and (ii) the extreme stress provoked by the ‘accidental’ emergency landing of the aircraft, which triggered such an attack. In that scenario, on the one hand, the first test described above contributes to the victim being able to be compensated easily (78) under Article 17(1) of the Montreal Convention, since the mere fact that the ‘accident’ is one ‘link’ in the ‘chain of causes’ that led to the passenger’s death or injury is sufficient. If, by contrast, the ‘accident’ had to be the ‘only/sufficient cause’, or even the ‘main cause’ of the passenger’s injury, that would seriously impede his or her claim. The first alternative test could exclude relief altogether in the example given. (79) The second one would lead to uncertainty. Indeed, the choice between the two factors set out above would be arguable and, as such, likely to generate considerable debate between the litigants, with the decision ultimately taken by the judge being somewhat arbitrary. (80) On the other hand, the ‘but for’ test also contributes to the uniform application of that convention. Indeed, whether or not a claim is governed by Article 17(1) depends on some objective relationship between a passenger’s injury and an ‘accident’, not on how the cause of the former has been pleaded by the applicant or defended by the carrier – a point on which I shall return to later.

69. That being said, as all the parties and the interveners univocally emphasised before the Court, for the purposes of Article 17(1) of the Montreal Convention, the ‘but for’ test discussed above cannot be applied in an unfettered manner. Otherwise, the scope of that provision and, ultimately, that of the carrier’s liability thereunder would be excessive.

70. In that regard, consider the following hypothetical scenario: the aircraft transporting a national football team to the World Cup sustains ‘accidental’ damage to one of its engines and makes an emergency landing. The players are not injured during that incident, but are understandably shocked by it. The trauma of that ‘accident’ increases over a week until, during a game, one of them becomes distracted, misses the ball, loses his balance, falls and twists his ankle – an injury which could give rise to substantial damages, given the detrimental impact it would certainly have on that player’s ability to participate further in that sporting event.

71. It may be that, but for the initial ‘accident’, the player would not have sustained the bodily injury in question. Nevertheless, it would be stretching unreasonably the provision of Article 17(1) of the Montreal Convention to consider that injuries such as that one, which relate only very remotely to an ‘accident’, are governed by that provision. Moreover, if air carriers were liable thereunder for such far-off consequences of an accident that occurred on board of their aircrafts, it could impose on them a very heavy compensation burden, which would be difficult to determine and to calculate. The ‘equitable balance of interests’ between passengers and carriers sought by the drafters of the Convention would not be maintained. (81) By analogy, in the laws of the States Parties, as DB pointed out, it is generally recognised that the ‘but for’ test is not sufficient to keep liability within reasonable limits. (82)

72. For that reason, a second test should, in every case, be applied. That test is a legal one and, as such, involves a policy choice. It is a matter of determining whether the causal relationship between an ‘accident’ and a passenger injury, beyond mere satisfaction of the ‘but for’ test, is sufficiently significant, so that it appears justified and reasonable, in light of the object and purpose of the Montreal Convention, to apply Article 17(1) and to hold the carrier liable under that provision. That test corresponds to what is known as ‘adequate cause’ in civil law countries and ‘proximate cause’ in common law countries.

73. Generally speaking, in the laws of the State Parties, (83) among all the factors which contributed to a particular injury, a given conduct or event will be regarded as an (84) ‘adequate’ or ‘proximate’ (85) and, thus, actionable cause of that injury if the latter is a natural result of the former. The classic sub-test is to verify whether the injury at issue was a foreseeable consequence of that conduct or event – that is, whether, in retrospect, a hypothetical bystander could reasonably have foreseen, in light of all the circumstances and past experience, that it would cause such an injury. Another closely related sub-test, mentioned by the German Government, consists in verifying whether the injury may be regarded as the materialisation of a risk inherent to the conduct or event in question. (86) Neither sub-test is fulfilled when it would appear improbable that the relevant conduct or event would have resulted in that injury, and that it happened only because of a particularly atypical or highly extraordinary chain of events. That legal test should be applied not in an abstract manner, but rather with its clear and sound policy intent in mind: whereas people should generally answer for their harmful actions, civil liability cannot reasonably extend to far-fetched consequences of the latter.

74. In the context of Article 17(1) of the Montreal Convention, as all the parties and interveners univocally suggested, (87) it means that the reach of that provision, and the carrier’s scope of liability thereunder, are limited to injuries that are the foreseeable consequences of the relevant ‘accident’ – or, put differently, injuries which constitute the materialisation of a risk inherent to that event. Those should be regarded, in law, as having been ‘caused’ by the ‘accident’ in question, for the purposes of that provision. By contrast, injuries which would appear, in retrospect, improbable consequences of the ‘accident’, and only arose from it because of a particularly atypical or highly extraordinary chain of events, should not. Here again, notably, courts of the United States follow that approach. In their view, to satisfy the causation requirement of Article 17(1), an applicant must not only prove that an ‘accident’ is part of the ‘chain of causes’ that leads to his or her injury, but also that the former has ‘proximately’ caused the latter. (88)

75. In the case in the main proceedings, it is beyond dispute that the first test described above is satisfied. It has already been established by the referring court. Obviously, from a factual point of view, but for the ‘accidental’ fall of the jug of coffee, DB would not have sustained the injuries for which he seeks compensation – and that is so, even if one accepts DB’s reasoning that the initial burns he sustained should, in that regard, be distinguished from their subsequent ‘aggravation’. (89) Indeed, as that court and all the interveners observed, if one removes the fall of the jug of coffee from the equation, DB would not have been burned and, obviously, his – then – non-existent burns could not possibly have been ‘aggravated’ thereafter.

76. Certainly, the alleged failure of the crew to administer proper first aid could – if proven – be regarded as another ‘link’ in the ‘chain of causes’ that lead to that so-called ‘aggravation’. Assuming that (i) the crew was legally required to respond to DB’s scalding in a certain way, because of a duty of care towards passengers (90) and/or industry standards, (91) (ii) they failed to do so, and that (iii) if they had not so failed, it could have mitigated DB’s burns, one could arguably consider that, but for that failure, these injuries would not have been sustained in the form they ultimately were – they would not have been ‘aggravated’ in the sense that DB claims. That consideration is, however, immaterial. As stated in point 67 above, to satisfy the first test, it suffices that the ‘accident’ is one of the factors that contributed to the production of the injury at issue: it need not be the only cause.

77. The second test is also clearly satisfied here, despite DB’s efforts to convince the Court of the contrary. In essence, the applicant argues that the ‘accidental’ fall of the jug of coffee could not be regarded as having ‘adequately’ or ‘proximately’ caused the injuries claimed – the ‘aggravation’ of his burns – since they were not foreseeable consequences of that ‘accident’. In DB’s view, that ‘aggravation’ happened only because of an atypical chain of events. Indeed, what typically happens after person is scalded on board an aircraft, is that he or she receives sufficient and adequate first aid, preventing his or her wounds from aggravating. The extraordinary failure of the crew to do so here would function as a ‘superseding cause’, breaking the ‘chain’ leading back to the initial ‘accident’.

78. That argument, however, does not withstand close scrutiny. In the first place, here again, even if DB’s initial burns were to be distinguished from their ‘aggravation’, the latter would obviously be a foreseeable consequence of the scalding, given that that ‘aggravation’ is what would have resulted – or here, possibly, has resulted – from the natural evolution of the wounds. In other words, as the German Government puts it, the risk of ‘aggravation’ of the burns was already inherent from the ‘accidental’ fall of the jug. The first aid administered from the crew was, precisely, intended to prevent the materialisation of that risk. Thus, there is a clear and close ‘cause and effect’ relationship between DB’s injuries – again, even if they could/should be distinguished – and the relevant ‘accident’.

79. In that regard, let us imagine a scenario, removed from the carriage of passengers but, in my view, quite illustrative. A person negligently drops a lit candle on the curtain of his or her neighbour’s house. A fire ensues. The firefighters called to intervene negligently lack the necessary equipment and fail to tackle the flames properly. In the end, the house burns down. Following DB’s reasoning, that would be regarded as an improbable consequence of the dropping of the candle, because, if the firefighters had fulfilled their duty of care and managed to extinguish the fire, the damage to the house would have been mitigated. The logical fallacy is obvious. Clearly, it was foreseeable that dropping a candle on a curtain could eventually result in the house burning down. Put differently, that end result was clearly ‘within the risk’ inherent to that action.

80. In the second place, contrary to what DB claims, in the present case, while the response of the crew to the fall of the jug, if proven to have been subpar, may certainly, as I have already indicated, be regarded as another cause of the ‘aggravation’ of his injuries, it could not, however, go as far as ‘breaking the chain’ leading back to the triggering ‘accident’. That would be the case, as is obvious from points 73 and 74 above, only if the crew’s conduct had been particularly atypical or highly extraordinary, to the point that that conduct – and the injuries eventually sustained – would appear, in retrospect, improbable to a hypothetical bystander. (92) Only then would it be, in terms of policy, unreasonable to impute those injuries to that ‘accident’. That is a high bar, which is simply not reached in a situation such as the one in the main proceedings.

81. Indeed, as Austrian Airlines, the German Government and the Commission astutely observe, the fact that a flight attendant, who may have had limited medical training only, and who has other duties and passengers to attend to, fails to provide, in the potentially stressful aftermath of an ‘accident’, the required medical care and attention to an injured passenger could hardly appear, in retrospect, improbable to such a bystander. (93) It is not that ‘atypical’ or ‘extraordinary’. It may plausibly happen for, in the light of experience, people do make mistakes in such circumstances. Similarly, a negligent failure to refill an emergency kit before a flight could also hardly be regarded as such. (94)

82. A similar example would be that of a person being hit by a car driven negligently by someone else, resulting in a broken arm requiring surgery. During the operation, the surgeon fails to apply sufficient care and attention and does not manage to mitigate the injury, or aggravates it, or even causes a different injury. Usually, in the laws of the State Parties, such medical mistakes do not ‘break the chain’ leading back to the initial negligence of the driver, since they are not particularly atypical or highly exceptional, to the point of appearing, in retrospect, improbable because, unfortunately, they do happen. (95)

83. In my view, that interpretation of the causation requirement set out in Article 17(1) of the Montreal Convention, in a situation such as the one in the main proceedings, is perfectly in line with the system, object, and purpose of that convention.

84. First, in the liability regime set out in the Montreal Convention in case of ‘bodily injury’ sustained by a passenger, the issue of whether the carrier’s personnel, in the event of an ‘accident’, has taken the required steps to avert the injury is intended to play a part only at the defense stage, under Article 21(2)(a) of that convention, for the purpose of limiting the carrier’s liability under Article 17(1). (96) Therefore, in all logic, that aspect should not be taken into account, upstream, when assessing whether or not that ‘accident’ can be regarded as having ‘caused’ that injury and, thus, whether or not the claim at issue is governed by the latter provision.

85. Second, that interpretation contributes to the ‘equitable balance of interests’ between passengers and carriers sought by the drafters of the Montreal Convention. It ensures that, where a claim for compensation has been brought, pursuant to Article 17(1) of that convention, within the two-year limitation period laid down therein, an applicant may be compensated, under that provision, for the entirety of the harmful consequences ‘proximately’ arising from the relevant ‘accident’, even those that may have occurred, in part, due to the subsequent administration of inadequate or insufficient first aid. (97) At the same time, such a liability would not be excessive, but rather justified and reasonable, given the importance of the ‘accident’ in the picture. (98)

86. Third, I believe, together with the referring court and the Commission, that the same interpretation also contributes to the uniform application of the Montreal Convention. In that regard, as I have indicated in the first section of my analysis, that instrument lays down an imperative regime of liability. Accordingly, as mentioned in point 68 above, whether a claim is governed by Article 17(1) thereof should be determined, not in the light of how it was pleaded, but objectively, in view of the actual facts. My interpretation guarantees, precisely, that every time a claim concerns injuries which are objectively and closely related to an ‘accident’, a skilful counsel will not be able to circumvent the Convention, and notably the two-year limitation period to bring proceedings, by craftily separating, as DB attempted to do, (99) these injuries from that event. (100)

2. Whether the failure of the crew to administer adequate and sufficient first aid could constitute, in itself, an ‘accident’

87. In its submissions, Austrian Airlines argued that, even if the fall of the jug of coffee could not be regarded, in law, as having ‘caused’ the injuries claimed by DB, it would not matter. Indeed, in the carrier’s view, the factor presented by the applicant as the real ‘cause’ of those injuries, namely the first aid administered by the flight attendants, qualifies, in itself, as an ‘accident’ within the meaning of Article 17(1) of the Montreal Convention. Thus, in any event, DB’s claim would, even under a narrow view of the exclusivity of that instrument, (101) fall within its scope and, accordingly, be pre-empted as time-barred.

88. That question is substantially different from the one being asked by the referring court. It is no longer an issue of causation per se, but of legal qualification of a given conduct of the crew. Moreover, that question was not substantively debated before the Court. In fact, it was only superficially addressed by DB and Austrian Airlines in their submissions. Above all, as I will explain in the next point, it is unnecessary, for the purposes of the present case, to address it. For all those reasons, I would respectfully urge the Court not to enter into that discussion. Should it nevertheless choose to do so, I shall address it briefly, and in the alternative.

89. Generally speaking, where – as is the case in the main proceedings – an ‘accident’ occurs during an international flight, such as the inadvertent fall of a jug of scalding coffee on a passenger, resulting in bodily injuries, one need not wonder whether the subsequent failure of the crew to administer proper first aid on the wounds also qualifies as such. As explained in the previous section, the fact that the initial ‘accident’ ‘adequately’ or ‘proximately’ contributed to those injuries is enough for the purposes of Article 17(1) – whether it be to order the carrier to compensate the victim or, as is the case in the main proceedings, to dismiss the latter’s claim as pre-empted by the Convention. Scrutinising further the ‘chain of causes’ that lead to the injuries in search of (an)other ‘accident(s)’ would be superfluous. (102)

90. That question becomes relevant only in situations – different from the one in the main proceedings – where passengers suffer medical issues during international flights, such as strokes or heart attacks, not brought about by any abnormal event, but solely as a result of prior health conditions which happen to manifest themselves on board. Indeed, these medical issues are generally not regarded as ‘accidents’ under Article 17(1) of the Montreal Convention, since they are not ‘external’ to the passengers concerned, (103) but indeed purely ‘internal’ to them. (104) In that context, the argument has been raised – either by passengers or by airlines, depending on the context (105) – that the failure of the crew to respond adequately to the medical issue in question – because they ignored it, or did not administer sufficient first aid, or did not have on board the required equipment to do so, wrongfully decided not to divert the aircraft to a nearby airport for immediate treatment, and so on – constituted, in itself, such an ‘accident’, which contributed to the passenger’s eventual death or bodily injury.

91. There are many decisions from, notably, the courts of the United States on that particular question. That case-law, however, is not entirely settled. Indeed, DB and Austrian Airlines each put forward decisions supporting their opposite points of view.

92. In a first line of decisions, presented by DB as the ‘majority view’, these courts rejected the argument that the failures of the crew to provide proper medical care, or to have adequate medical equipment on board, or to divert the aircraft to a nearby airport could constitute, in themselves, ‘accidents’ within the meaning of Article 17(1). (106)

93. Nonetheless, a second line of decisions, presented by Austrian Airlines as the ‘true’ majority view, and starting with the seminal judgment of the United States Supreme Court in Husain, (107) states, conversely, that the response of the crew to a passenger’s medical issue may, in certain situations, qualify as an ‘accident’. That is the case where the carriers’ personnel deviates from industry standards of care and/or airline policies and procedures, which require them to adopt a particular conduct in the circumstances – providing oxygen in the event of a heart attack, and so on – to such a significant extent that their response can be considered ‘unusual’ and/or ‘unexpected’. (108)

94. As I indicated above, I do not believe that the Court should take a position on that question in the present case, notably because it is, I repeat, irrelevant for the purposes of the case in the main proceedings. However, if it were still to do so, I would suggest a certain degree of caution. The reasoning of the United States Supreme Court in Husain stretches, in my view, the concept of ‘accident’ quite significantly. That the conduct of the crew could, in itself, be regarded as an ‘event’ already raises some conceptual difficulties. (109) However, the most challenging aspect of that construction is the underlying idea pursuant to which the ‘unusual or unexpected’ (or ‘unforeseen’) nature of such an ‘event’ depends on whether the crew deviated from a given conduct required by law – in other words, whether they have been negligent.

95. Indeed, that approach has the potential to turn the examination of whether an ‘accident’ happened – something which should be rather straightforward – into a lengthy dispute involving complex assessments of facts and law. In that regard, while it is recognised, in many jurisdictions, that carriers have a duty of care towards passengers, and although industry standards exist on medical issues, (110) what exactly is required in a given situation, save for the rare obvious cases like Husain, is often arguable. (111) Above all, I recall that the concept of ‘accident’ within the meaning of Article 17(1) of the Montreal Convention is intended to be an objective one. Negligence on the part of the carrier is, normally, not relevant for deciding whether a given event qualifies as such. (112) As stated in point 84 above, it should play a part only at the defense stage, pursuant to Article 21(2)(a) of that convention. In that respect, Husain could be criticised for shifting the focus of the ‘accident’ inquiry from the nature of the event which caused the injury to the alleged failure of the air carrier to avert it.

96. That said, all things considered, what should, in my view, ultimately guide the Court as to the interpretation of the concept of ‘accident’ in such situations (113) – and another reason why it should reserve a decision on that question to a future case – is the position it will eventually take, one day, on the scope of exclusivity of the Montreal Convention.

97. On the one hand, if the Court were to endorse the broad view of that exclusivity, it should rule that a failure of the crew to respond adequately to a passenger’s medical issue does constitute an ‘accident’, despite the conceptual difficulties raised by that interpretation. Indeed, the opposite would deprive the passenger victims of any remedy – for no claim could be brought under Article 17(1), and negligence claims under domestic law would be pre-empted by the Convention – whereas they sustained an injury partially attributable to the conduct of the carrier’s personnel. Such a result could hardly reflect the ‘equitable balance of interests’ between air carriers and passengers. In addition, it would also remove a serious incentive for airlines to comply with their duty of care and the relevant industry standards, since they would not be liable towards the victims in the case of breach.

98. In fact, Husain should very much be seen in that light. It was decided after the same Supreme Court ruled, in Tseng, that a passenger has either a remedy under the Montreal Convention or no remedy at all. By including cases of negligent response of the crew to medical issues within the concept of ‘accident’, that court ensured that the victims of such conduct may obtain relief. (114)

99. On the other hand, if the Court were to adopt the narrow view of the scope of exclusivity of the Montreal Convention, then I would suggest maintaining the ‘traditional’ construction of the concept of ‘accident’ used in Article 17(1). Then, finding that the carelessness of the crew in dealing with sick passengers does not, in itself, qualify as such would, certainly, exclude a remedy under that instrument. However, it would, at the same time, open the way for negligence claims under domestic law. As such, these situations would not remain unaddressed.

100. In the light of all the foregoing considerations, I suggest that the Court answer the questions referred for a preliminary ruling by the Oberster Gerichtshof (Supreme Court, Austria) as follows:

(1) Article 29 of the Convention for the unification of certain rules for international carriage by air concluded on 28 May 1999 in Montreal, signed on 9 December 1999 by the European Community, and approved on its behalf by Council Decision 2001/539/EC of 5 April 2001 (‘the Montreal Convention’),

must be interpreted as meaning that a claim against an air carrier, however pleaded, which is objectively related to bodily injuries sustained by a passenger, during an international flight falling within the general scope of that convention, because of an accident that took place on board the aircraft, within the meaning of Article 17(1) of that convention, is governed exclusively by that instrument. Accordingly, the conditions and limits of liability set out therein, including the two-year limitation period laid down in Article 35(1) of that convention, imperatively apply to such a claim. That answer is without prejudice to the issue of whether or not other types of injuries sustained by passengers are also exclusively governed by the same convention.

must be interpreted as meaning that, for the purposes of Article 17(1) of that convention, an ‘accident’ is to be regarded as having ‘caused’ a passenger’s bodily injuries when (i) but for that event, those injuries would not have been sustained and (ii) the latter was a foreseeable consequence of the former, irrespective of the fact that another factor, such as the subsequent administration of improper first aid by the flight attendants, may also have contributed to the injuries in question.

2 That convention, concluded in Montreal on 28 May 1999, was signed by the European Community on 9 December 1999 and approved on its behalf by Council Decision 2001/539/EC of 5 April 2001 (OJ 2001 L 194, p. 38).

3 Convention for the unification of certain rules for international carriage by air, signed in Warsaw on 12 October 1929 (‘the Warsaw Convention’).

5 Regulation of the European Parliament and of the Council of 13 May 2002 (OJ 2002, L 140, p. 2).

6 Whether what was done was sufficient and adequate is a point in dispute between the parties in the main proceedings (see footnote 8 below).

7 See paragraph 1489 of the Allgemeines Bürgerliches Gesetzbuch (Austrian General Civil Code).

8 DB alleges that the attendants did not cool his burns with cold water and only gave him a small tube of ointment to apply onto them. Furthermore, the first aid kit did not contain ointment to treat the burns or bandages in sufficient quantities. Austrian Airlines appears to dispute these facts.

9 The first-instance court assumed that the Warsaw Convention, and not the Montreal Convention, was applicable because it believed that Israel was only a party to the former. However, as the court of appeal noticed, in fact, Israel deposited an instrument of accession to the Montreal Convention on 19 January 2011, and that convention entered into force, with respect to that State, on 20 March of the same year (see https://www.icao.int/secretariat/legal/List%20of%20Parties/Mtl99_EN.pdf).

11 See, notably, judgment of 19 December 2019, Niki Luftfahrt (C‑532/18, EU:C:2019:1127, paragraph 30 and the case-law cited; ‘the judgment in Niki Luftfahrt’).

12 See point 9 above. Moreover, since Austrian Airlines is, as it would appear, a ‘Community air carrier’ within the meaning of Article 2(1)(b) of Regulation No 2027/97, that regulation also applies. However, when it comes to the liability of such a carrier in respect of injuries suffered by passengers, Article 3(1) of the same regulation simply refers to the provisions of the Montreal Convention.

13 Or ‘in the course of any of the operations of embarking or disembarking’. In the present Opinion, however, I shall focus on injuries sustained on board the aircraft.

14 Or, alternatively, from the date on which the aircraft ought to have arrived, or from the date on which the carriage stopped.

17 I will leave Articles 18 and 19 of the Montreal Convention aside in the rest of this Opinion, as those are not relevant in the present case.

20 Despite some textual differences, those two provisions are generally equivalent. Therefore, judicial decisions and academic literature relating to Article 24 of the Warsaw Convention is relevant for the interpretation of Article 29 of the Montreal Convention. The same applies with respect to Article 17 of the first Convention and Article 17(1) of the second (see Opinion of Advocate General Saugmandsgaard Øe in Niki Luftfahrt (C‑532/18, EU:C:2019:788, points 26, 27 and 43). I will therefore refer without distinction to decisions relating to one or the other convention.

21 See, for the same view, Chapman, M., Prager, S., Harding, J., Saggerson on Travel Law and Litigation, 5th edition, Wildy, Simmonds & Hill Publishing, London, 2013, p. 513, § 10.103, and Giemulla, E., ‘Article 29 – Principles for claims’, in Montreal Convention, Kluwer, the Netherlands, 2006, p. 5-7, §§ 10-13.

22 See, notably, Supreme Court of the United States, 12 January 1999, El Al Israel Airlines, Ltd. v. Tsui Yuan Tseng, 525 US 155 (‘Tseng’). See also the view of Tompkins, G.N., expressed notably in Hobe, S., Ruckteschell (von), N., Heffernan, D. (eds.), Cologne compendium on air law in Europe, Carl Heymanns Verlag KG, 2013, pp. 1004-1005, §§ 143-148.

24 Under the broad view, had DB filed his claim within the two-year limitation period laid down in Article 35(1) of the Montreal Convention, the issue of whether his injuries were caused by an ‘accident’ should have been addressed, but only for the purpose of deciding whether a remedy were available at all.

25 See, nevertheless, the judgments discussed in point 50 below. That question was also raised in the case which gave rise to the judgment of 20 October 2022, Laudamotion (C‑111/21, EU:C:2022:808). However, the Court held that it did not need answering in that case (see paragraph 34 of that judgment).

26 House of Lords (Scotland), 12 December 1996, Abnett v. British Airways Plc, [1997] A.C. 430 (‘Sidhu’).

27 Sidhu and/or Tseng have been followed by, notably, the Court of Appeal of New Zealand in its judgment in Emery Air Freight Corp. v Nerine Nurseries Ltd ([1997] 3 NZLR 723); the Federal Court of Australia in its judgment of 9 September 1998, South Pacific Air Motive Pty Ltd v Magnus (157 ALR 443 (1998)); and the Supreme Court of Canada in its judgment of 28 October 2014, Thibodeau v. Air Canada ([2014] 3 S.C.R. 340). The broad view was also adopted by the Bundesgerichtshof (Federal Court of Justice, Germany), in its judgment of 15 March 2011 (Az X ZR 99/10).

28 See, notably, Supreme Court of the United Kingdom, 5 March 2014, Hook v British Airways Plc [2014] WL 795206.

30 Opinion in case C‑589/20, ‘my Opinion in Austrian Airlines (Exoneration of air carrier from liability)’, EU:C:2022:47, point 29. See by analogy, judgment of 6 October 2020, Commission v Hungary (Higher education) (C‑66/18, EU:C:2020:792, paragraph 92). Accordingly, I will refer to relevant national decisions in this Opinion.

31 The Court has already made clear that it does not intend to do so. See, notably, the different definitions of the concept of ‘accident’ restated in point 58 below.

32 See, amongst many decisions, United States Court of Appeals, Third Circuit, 19 July 1984, Stanley Abramson v. Japan Airlines Co., Ltd, 739 F.2d 130; US District Court, S.D. New York, 24 September 1991, Walker v. Eastern Air Lines, Inc., 775 F.Supp. 111 (stating that the narrow view on exclusivity corresponded, before Tseng, to ‘the overwhelming consensus of the courts that have addressed the issue’); and United States Court of Appeals, Eleventh Circuit, 25 August 1997, Krys v. Lufthansa German Airlines, 119 F.3d 1515.

33 See, notably, Giemulla, E., op. cit., p. 5, § 9, p. 8, §15, and the academic references; McDonald, M., ‘The Montreal Convention and the Preemption of Air Passenger harm Claims’, The Irish Jurist, Vol. XLIV, 2010, pp. 203 to 238, and Bernard, N., ‘Taking Air Passenger Rights Seriously: the Case Against the Exclusivity of the Montreal Convention’, International Community Law Review, Vol. 23, issue 4, 2021, pp. 313 to 343.

34 United Nations Treaty Series, Vol. 1155, p. 331 (‘the Vienna Convention’). See Article 31 and Article 32 of that convention.

35 The evolution of that provision through time reinforces that argument. The original version of the (then) Article 24 of the Warsaw Convention provided, in substance that ‘in the cases covered by Article 17 …’, actions for damages could only be brought subject to the conditions set out in that convention. However, the language of Article 24 was later altered by the Montreal Protocol No. 4 to Amend the Convention for the Unification of Certain Rules Relating to International Carriage by Air. There, the reference to the ‘cases covered by Article 17’ was replaced by the expression – arguably more inclusive – ‘in the carriage of persons …’.

36 The drafters of the Convention could, for instance, have stated that ‘no cause of action of any kind …, other than ones provided under this Convention, may be taken in respect of any loss or injury of any type, howsoever caused, arising out of international carriage by aircraft’. See McDonald, M., op. cit., pp. 205 and 227.

40 See, notably, United States Court of Appeals, Third Circuit, 19 July 1984, Stanley Abramson v. Japan Airlines Co., Ltd, 739 F.2d 130, and the case-law referred. See also Giemulla, E., op. cit., p. 5, § 9, p. 8, §15, and the various academic references.

45 Something recognised with much honesty by Lord Hope in Sidhu: ‘There are various passages in [the Minutes of the Second International Conference on Private Aeronautical Law, 4 to 12 October 1929 at Warsaw] where the delegates are recorded as expressing views about the object of the Convention. … But these are mere straws in the wind which emerged during several days of detailed discussion to which many delegates contributed. I do not find a sufficiently clear and consistent expression of views here … on the point which is at issue …’.

46 ‘The purpose behind Article [29] was to ensure that, in circumstances in which the Convention applied, it was not possible to circumvent its provisions by bringing an action for damages … in contract or in tort or otherwise. Once the Convention applies, its conditions and limits of liability were applicable’ (see International Conference on Air Law, minutes of the seventeenth meeting, p. 3, § 10). Unfortunately, the chairman did not elaborate on the ‘circumstances in which the Convention applie[s]’.

47 See Supreme Court of the United Kingdom, 5 March 2014, Stott v Thomas Cook Tour Operators Limited, [2014] UKSC 15, relating to Regulation (EC) No 1107/2006 of the European Parliament and of the Council of 5 July 2006 concerning the rights of disabled persons and persons with reduced mobility when travelling by air (OJ 2006 L 204, p. 1).

48 See, notably, United States Court of Appeals, Second Circuit, 22 March 2002, King v. American Airlines, Inc., 284 F.3d 352.

49 See, notably, United States District Court, D. Maryland, 31 January 2007, Knowlton v. American Airlines, Inc., 31 Avi 18,486). See, for other examples of passengers deprived of remedies, McDonald, M., op. cit., pp. 220 to 223.

51 See, notably, Supreme Court of the United Kingdom, 5 March 2014, Stott v Thomas Cook Tour Operators Limited, [2014] WL 795206, § 63 to 65.

53 Judgment of 23 October 2012 (C‑581/10 and C‑629/10, EU:C:2012:657, paragraphs 41 to 60).

54 Regulation of the European Parliament and of the Council of 11 February 2004 establishing common rules on compensation and assistance to passengers in the event of denied boarding and of cancellation or long delay of flights and repealing Regulation (EEC) No 295/91 (OJ 2004 L 46, p. 1).

55 See, notably, Wegter, J.J., ‘The ECJ Decision of 10 January 2006 on the Validity of Regulation 261/2004: Ignoring the Exclusivity of the Montreal Convention’, Air & Space Law, Vol 31, No 2, 2006, pp. 133 to 148; Radošević, S., ‘CJEU’s Decision in Nelson and Others in Light of the Exclusivity of the Montreal Convention’, Air & Space Law, Vol. 38, No 2, 2013, pp. 95 to 110; and Tompkins, G.N., ‘Are the Objectives of the 1999 Montreal Convention in Danger of Failure?’, Air & Space Law, Vol. 39, No 3 (2014), pp. 203 to 214.

56 That is, in the event of ‘accumulation’ of causes of action (see point 26).

58 As I just explained, only if injuries such as the ones sustained by DB could not have been regarded as resulting from an ‘accident’ would it have been necessary for the Court to take a position on the scope of exclusivity. That is why the referring court pragmatically raised the issue in a subsidiary manner.

60 Supreme Court of the United States, 4 March 1985, 470 U.S. 392 (1985) (‘Air France v. Saks’).

64 I underline that, at the currently stage of the main proceedings, this aspect has not been established. Indeed, Austrian Airlines contests it (see footnote 8), and the referring court indicated, if it were relevant for the fate of DB’s claim, further evidence and findings would be necessary in that respect.

66 At this juncture, one may legitimately wonder how the Court could, in the context of the preliminary ruling procedure laid down in Article 267 TFEU, limited to questions of interpretation of EU law, address the issue of whether a given accident should be regarded as having ‘caused’ a certain injury for the purpose of Article 17(1) of the Montreal Convention. In my view, it can do so, but only to a certain extent. Indeed, as I will explain in the next section, causation boils down to two questions. First, one must decide whether the ‘accident’ contributed or not to the production of the injury (see points 66 to 68). That is, plainly, a factual issue, to be clarified by the national judge in each case. In the present case, that issue is not in dispute (see point 75). A second aspect of causation is whether the ‘accident’ is sufficiently related to the injury at issue to justify, policy-wise, the application of Article 17(1) (see points 69 to 74). That is a question of (EU) law, and the only one at stake here (see points 77 et seq.). One could also wonder whether the Court should do so. I have already detailed, in my Opinion in Austrian Airlines (Exoneration of air carrier from liability) in point 77, the reservations I have when it comes to the Court answering, by means of the preliminary ruling procedure, questions tailored to very detailed and specific sets of facts. That said, given that the present case is the first one regarding the causation requirement set out in Article 17(1), and given its illustrative nature for future cases, I consider it useful and appropriate to do so here.

67 The artificiality of that approach is evidenced, in my view, by the fact that Austrian Airlines and the Commission answer the question of whether the fall of the jar and the subsequent administration of first aid should be treated as a single ‘accident’ by bringing forward what is, by nature, a causation reasoning – which I will refer to where relevant. Even the German Government, which addresses, first, whether the two events could be seen as a single ‘accident’ and second, whether DB’s injuries could be regarded as ‘caused’ by the fall of the jar, brings forward essentially the same reasoning in both parts. In fact, the referring court itself discusses the issue from a causation angle in the order for reference.

68 I share, in that respect, the view of the court of appeal in the main proceedings (see point 13 above)

69 Here, that would be Austrian law, by means of Article 5(2) of Regulation (EC) No 593/2008 of the European Parliament and of the Council of 17 June 2008 on the law applicable to contractual obligations (Rome I) (OJ 2008 L 177, p. 6).

70 The conditions of liability of the carrier, at least in cases involving accidents, are certainly among the rules for international carriage by air that the Montreal Convention was intended to unify (see points 44 and 45 above). To assess one of those conditions by reference to the lex contractus would undermine that objective of uniformity, for liability could then vary depending on the forum before which a claim is lodged – as the various choice-of-law rules of the States Parties may designate different lex contractus – and, ultimately, on the content of that law.

71 The Montreal Convention refers certain issues expressly – see, notably, Article 33(4) – or impliedly – see Article 29, first sentence, last line (‘… without prejudice …’) – to domestic law. It does not do so when it comes to the causation requirement set out in Article 17(1) thereof.

72 See, by analogy, judgment of 20 October 2022, Laudamotion (C‑111/21, EU:C:2022:808, paragraph 21 and the case-law cited).

73 See my Opinion in Austrian Airlines (Exoneration of air carrier from liability) (point 72).

74 A typical example would be a cyclist hitting a pedestrian. That happened because of a combination of the speed of the bike, the defect of its breaks, the inattentiveness of the pedestrian, and so on.

75 See, notably, for a comparative analysis of the laws of the Member States of the European Union and of the United Kingdom, Von Bar, C. et al. (eds), Principles, Definitions and Model Rules of European Private Law. Draft Common Frame of Reference (DCFR); prepared by the Study Group on a European Civil Code and the Research Group on EC Private Law (Acquis Group), Sellier, European Law Publishers, Munich, 2008, Vol. IV, book VI (‘Non contractual liability arising out of damage caused to another’), Chapter 4: Causation, pp. 3566 to 3608.

76 ‘Any injury is the product of a chain of causes, and we require only that the passenger be able to prove that some link in the chain was [an “accident” within the meaning of Article 17(1)].’

78 That said, there may be situations where it could be difficult to establish that, ‘but for’ a given ‘accident’, the injury at issue would not have been sustained. See, on that topic, notably, Defossez, D., ‘Contaminated Air: Is the “But For” Test Saving Air Carriers?’, Air & Space Law, Vol. 44, No 2, 2019, pp. 185 to 202.

79 Since the ‘accidental’ landing would not, in isolation, have caused the passenger’s heart attack and bodily injury, but only did in conjunction with the prior health condition of the latter.

80 See, to that effect, Supreme Court of the United States, 24 February 2004, Olympic Airways v. Husain, 124 S.Ct. 1221.

82 See, notably, Von Bar, C. et al. (eds), op. cit., p. 3570, and, for a comparative analysis of the law of the Member States on that issue, pp. 3574 to 3585.

83 See, notably, for a comparative analysis of the law of the Member States on that issue, Von Bar, C. et al. (eds), op. cit., pp. 3574 to 3585.

84 There may be more than one ‘adequate’ or ‘proximate’ cause for an injury.

85 For the purposes of the present case, ‘adequate’ and ‘proximate’ will be used synonymously with all the nuances from national law set aside.

86 See, for US law, Restatement (Third) of Torts: liability for physical harm (Basic Principles), ‘Scope of liability – Proximate cause’, § 29. Consider the example of a student who left university late because his or her class finished with a delay, was hit by a car while crossing the street, and dies as a result. Driving a car, in itself, generates a risk of accident. The time at which the class finishes does not. The first factor is an ‘adequate’/‘proximate’ cause of the injury; the second one, although it is a necessary condition of the same (factual cause), is not.

88 See, among many decisions, United States District Court, S.D., New York, 6 September 2007, Zarlin v. Air France, 2007 WL 2585061, and the case-law cited.

89 See point 59 above. While the referring court seemingly endorses that aspect of DB’s reasoning – for it is a premiss of its first question – I have some difficulty with it. Indeed, what DB framed as two distinct injuries are, in fact, but one and the same: the burns he sustained. The alleged failure of the crew to administer first aid did not result in ‘separate’ injuries, in the proper sense – that would have been the case, for instance, if the attendant, while caring for DB’s burns, had stepped on his foot and broken his toes. Nor were the burns ‘aggravated’ by the conduct of the crew, in the proper sense of ‘made more severe than they were’. Indeed, as I explain in point 76 below, what is at issue, here, is whether the crew negligently failed to mitigate his burns, that is, to make them less severe than they initial were, or better than they would have been following their natural evolution. That being said, as hinted above, even if the administration of first aid had resulted in a genuinely separate injury the reasoning suggested in this Opinion would be valid nonetheless, mutatis mutandis.

90 In many jurisdictions, it is recognised that carriers have a duty of care towards passengers. Notably, under Austrian law, which is applicable to the contract of carriage at issue (see footnote 69), the conclusion of such a contract of carriage brings about, on the part of the carrier, an obligation to ensure their safety (see my Opinion in Austrian Airlines (Exoneration of air carrier from liability), footnote 5).

91 In the European Union, industry standards can be found in Council Regulation (EEC) No 3922/91 of 16 December 1991 on the harmonisation of technical requirements and administrative procedures in the field of civil aviation (OJ 1991 L 373, p. 4), as amended by Regulation No 1899/2006 of the European Parliament and of the Council of 12 December 2006 (OJ 2006 L 377, p. 1), Annex III. See, notably, OPS 1.745 (obligation to be equipped with first-aid kits, readily accessible for use); OPS 1.755 (obligation to be equipped with an emergency medical kit); OPS 1.760 (obligation to be equipped with first-aid oxygen); OPS 1.1005 and OPS 1.1010 (initial and continuous training of the crew concerning, notably, the content and the use of first-aid kits, first-aid oxygen, emergency medical equipment).

92 By analogy, in the domestic law of the State Parties, when the conduct of a third party intervenes between the initial wrongful actions of the defendant and the injury eventually sustained by the victim, that conduct is regarded as a ‘superseding cause’ only where that conduct would have appeared, in hindsight, unforeseeable or improbable. Where that is not the case, it does not break the ‘chain of causes’ leading to the actions of the defendant. A classic example is a person negligently leaving a loaded gun in a school courtyard and a child subsequently picking it up and firing it at his friend. In such a situation, the conduct of the child does not break the ‘chain of causes’ leading back to the initial person’s negligence. Indeed, if one leaves a gun in a school courtyard, plainly and simply, one could reasonably foresee that a child would pick it up and use it. See, notably, Von Bar, C. et al. (eds), op. cit., pp. 3571 to 3572 and 3578 to 3581.

93 I wish to underline, to avoid any confusion, that the question of whether the alleged failure of the crew to administer proper first aid was ‘foreseeable’, for the purpose of deciding whether DB’s injuries could be regarded as having been adequately ‘caused’ by the ‘accidental’ fall of the jug of coffee, is sensibly different from the issue of whether that failure could constitute, in itself, an ‘accident’ – that is, under the definition of that concept adopted by the Court (see point 58), an ‘unforeseen’ event (see Section B.2. of this Opinion). Imagine, if you will, a flight attendant serving a cup of hot coffee to a passenger. That cup ends up slipping from the tray table of the passenger’s seat, resulting in scalding. On the one hand, that is a ‘foreseeable’ course of events, for the purpose of causation. A reasonable observer could have anticipated that it would occur. Risks of scalding are inherent in the act of serving hot drinks. A cup falling from a tray, if one takes into account usual circumstances like the movement of the aircraft, can plausibly happen and does indeed happen. That is why cups of coffee are sometimes served with a lid. On the other hand, the cup falling, when it actually happens, can still be regarded as an ‘unforeseen event’ and, thus, as an ‘accident’, because the injured passenger (unless omniscient) did not know, before it happened, that it would happen at the moment it did. In sum, many ‘accidents’, although ‘unforeseen’, were ‘foreseeable’.

94 By contrast, as the German Government hints in its submissions, if, after the fall of the jug of coffee, the attendant who cared for DB’s injuries had, for some inexplicable reason, had a fit of rage and intentionally beaten the passenger, that conduct would, no doubt, have appeared, in retrospect, unforeseeable/improbable.

95 In that scenario, the driver could be held liable for all the injuries (see, notably, Cour de cassation (Court of cassation, France), 27 January 2000, No 97-20.889), irrespective of the fact that the victim would also potentially have a claim against the doctor, with respect to the second injury only. See, notably, for a comparative analysis of the law of the Member States on that issue, Von Bar, C. et al. (eds), op. cit., pp. 3574 to 3585.

96 Pursuant to that provision, ‘the carrier shall not be liable for damages arising under [Article 17(1)] to the extent that they exceed for each passenger 100 000 Special Drawing Rights if the carrier proves that … such damage was not due to the negligence … of … its servants or agents’. Indeed, whether the carriers’ personnel complied with their duty of care and industry standards, after an ‘accident’ happened, is traditionally assessed thereunder. See, for instance, United States District Court, S.D. Florida, 10 March 2018, Quevedo v. Iberia Lineas Aereas de España, Sociedad Anónima Operadora Co., 2018 WL 776754).

97 Concretely, it means that, for instance, if DB had brought his claim within the limitation period set out in the Montreal Convention, he could have obtain compensation, under Article 17(1), for the full extent of his injuries, given that even the allegedly separate consequences of the first aid were nonetheless ‘adequately’ or ‘proximately’ related to the ‘accidental’ fall of the jug of coffee. By contrast, the opposite interpretation would be more burdensome for passengers. It would mean that, in order to be fully compensated in a similar scenario, a passenger would need to plead – and establish – not just one, but two causes of action. Instead of simply having to show that a hot beverage spilled, he or she would also need to prove that the response of the crew was negligent – a task which would not always be easy, as explained in point 95 below.

98 That seems all the more reasonable where the inadequate first aid has been administered by the carrier’s personnel. That said, even in a scenario where first aid would have been administered by, for instance, a nurse that happened to be on board, the carrier should still be liable under Article 17(1) for the entire passenger injury, since, as I have demonstrated in this Opinion, the ‘accident’ can be regarded as having ‘caused’ all of it, within the meaning of that provision. Admittedly, in that scenario, the passenger could also have a claim against the nurse, since his or her conduct also caused the injury. However, that would be a question of domestic law, since the Montreal Convention only governs liability claims against air carriers.

99 Such counsel’s games are made obvious, here, by the fact that DB seems to have modified his particulars of claim between the first and subsequent instance of the trial, potentially to avoid the Convention (compare points 11 and 14 above).

100 See, for a very similar case, United States Court of Appeals, Second Circuit, 5 January 1998, Fishman by Fishman v. Delta Air Lines, Inc., 132 F.3d 138. In that case, a child was scalded during a flight after an attendant inadvertently spilled boiling water on her. That was clearly an ‘accident’ within the meaning of Article 17 of the (then) Warsaw Convention. The passenger and her mother filed negligence claims against the carrier, but did so more than two years after the event. To avoid a finding of pre-emption by that convention, the applicants, much like DB in the main proceedings, invoked as the ‘cause’ of the injury, not the ‘accident’ itself, but the allegedly negligent refusal of the crew to administer first aid thereafter. The national court rejected that argument, observing that ‘the underpinning of the claims … [was] the scalding of [the child] … The precipitating cause, the “accident,” cannot be artificially separated from its results … in order to avoid the Warsaw Convention’.

101 I recall that, if one were to adopt a ‘broad’ conception of the exclusivity of the Convention, DB’s claim would be pre-empted simply because it concerns an injury sustained on board the aircraft, whether or not it was caused by an ‘accident’ within the meaning of Article 17(1).

102 Notably, recognising that not just one, but two ‘accidents’ caused a passenger’s injury would not influence the amount of damages that the latter could obtain under Article 17(1) of the Montreal Convention. Indeed, such damages are meant to compensate the injury suffered. Their amount accordingly depends on the extent of that injury, not on how many ‘accidents’ contributed to it.

103 A relevant criterion, at least in the definition of the concept of ‘accident’ provided in Air France v. Saks (see point 58 above).

104 See, notably, United States Court of Appeals, Eleventh Circuit, Krys v. Lufthansa German Airlines, Aug 25, 1997, 119 F.3d 1515.

105 Indeed, before Sidhu and Tseng, when a narrow view of the exclusivity of the Warsaw/Montreal Convention was endorsed, it was in the interest of the applicants to plead that there was no ‘accident’ in such circumstances: that finding meant that the Convention was considered not to be applicable, which enabled claims under national law, often more favourable to them. By contrast, airlines had an interest in arguing that an ‘accident’ did occur because that could bring the case back within the exclusive realm of that convention. After Sidhu and Tseng, when a broad view of the exclusivity became the norm, it also became in the interest of the passenger to claim that an ‘accident’ happened – since that was the only way to obtain a remedy. By contrast, it became the interest of airlines to argue that there was no ‘accident’: such a finding would shield them from any liability.

106 See, notably, and respectively, United States Court of Appeals, Third Circuit, 19 July 1984, Stanley Abramson v. Japan Airlines Co., Ltd, 739 F.2d 130; United States District Court, S.D. New York, 22 February 1996, Tandon v. United Air Lines, 926 F.Supp. 366, and United States Court of Appeals, Eleventh Circuit, 25 August 1997, Krys v. Lufthansa German Airlines, 119 F.3d 1515.

107 In the case that lead to that judgment, a passenger was wrongfully seated in the smoking section of the aircraft while he suffered from a prior asthmatic condition which made him very sensitive to second-hand smoke. The passenger’s wife repeatedly asked a flight attendant to move her husband to a different area of the plane, which was met with refusal to do so. As the smoking increased during the flight, he felt ill, received some medical assistance from a doctor travelling with him, but died thereafter. The Supreme Court ruled that the attendant’s refusal to reseat the passenger qualified as an ‘accident’ since it was an ‘event’ clearly ‘external’ to the passenger and also ‘unexpected and unusual’ as it deviated from industry standards and the airline’s policy.

108 See, for instance, United States District Court, S.D. Indiana, Indianapolis Division, 10 October 2007, Watts v. American Airlines, Inc., 2007 WL 3019344.

109 One could very well consider that acts or omissions of the crew cannot, in themselves, qualify as ‘accidents’ as they are not, strictly speaking, ‘events’, although they can lead to such accidental events. For instance, in the present case, the ‘accident’ is not the failure by the attendant to manoeuvre carefully the cart on which the jug of coffee was resting: it is the resulting fall of that jug.

111 Here, for example, that point seems very much to be disputed between the parties in the main proceedings, (see footnote 8 above). Notably, the adequate response to DB’s burns could depend on how severe they appeared on the spot, which could be arguable. See also, notably, United States District Court, S.D. Florida, 15 June 2011, Cardoza v. Spirit Airlines, Inc., 2011 WL 2447523. In that case, whether the decision of the pilot not to deviate the flight to a nearby airport, given the medical issue of a passenger, was appropriate in the circumstance and, as such, ‘usual’ and ‘expected’ for the purpose of Article 17(1), was debated at length between the parties, involving arguments about fact-intensive issues such as (i) the actual severity of the passenger’s condition and to what extent was the pilot actually informed of it, (ii) what the industry standards and procedures actually require and, (iii) what the pilot actually did.

112 See, notably, judgment of 2 June 2022, Austrian Airlines (Exoneration of air carrier from liability) (C‑589/20, EU:C:2022:424, paragraphs 22 and 23).

113 I repeat, for the sake of clarity, that this discussion is relevant only for cases where the medical issue of the passenger resulted solely from a prior health condition that happened to manifest itself during an international flight. Conversely, where that issue results, like in the present case, from an ‘accident’, then the claim is already, for that very reason, governed by Article 17(1) of the Montreal Convention, and the carrier liable thereunder (see points 89 and 90 above).

114 By contrast, the cases mentioned in footnote 106 were decided before Tseng. There, findings of absence of ‘accident’ also meant that claims for negligence under domestic law were allowed to proceed, and often succeed. In cases decided after Tseng, the same finding resulted in the claimant being deprived of any remedy (see, notably, United States District Court, E.D. New York, 13 March 2000, Rajcooar v. Air India Ltd., 89 F.Supp. 2d 324). Husain can be seen as a response to that unsatisfying situation.

Court of Justice of the European Communities (including Court of First Instance Decisions)