2.1 Lotus’ legacy

The Lotus case is unique in international jurisprudence for at least four reasons.

First, Lotus is often cited as the only major – but certainly the most well-known – instance when the PCIJ’s (or ICJ’s) core holding in a dispute was subsequently overturned by treaty law.Footnote ⁸

As the story is reproduced in legal textbooksFootnote ⁹ and both commentaries to the United Nations Convention on the Law of the Sea (UNCLOS),Footnote ¹⁰ Lotus’s holding on the legality of the exercise of concurrent jurisdiction by France and Turkey ‘produced alarm among seafarers’, and a long campaign ensued against the decision to avoid multiple prosecutions for the same offence.Footnote ¹¹

This led to the codification of rules providing for the exclusive jurisdiction of the flag state over criminal matters in cases of ‘collision or any other incident of navigation’. The rules were codified first in the 1952 Brussels Convention, then later in the 1958 High Seas Convention, and took their final shape as codified by UNCLOS Article 97(1), providing for the exclusive jurisdiction of the flag state for penal or disciplinary proceedings ‘[i]n the event of a collision or any other incident of navigation concerning a ship on the high seas’.

While this provision (and its precursors) are sometimes described as the ‘overruling’Footnote ¹² of a key holding, this article proposes that Article 97(1) is better understood as a carve-out rule. Accordingly, in incidents involving vessels or high seas that are not collisions or incidents of navigation, Lotus is still good law.Footnote ¹³ What is more – as confirmed by the Enrica Lexie award – Lotus is the applicable law for these issues.Footnote ¹⁴

Second, the Lotus case, despite its fame, is barely ever referenced in the ICJ’s decisions. Handeyside, in a recent article, identified only three occasions when the ICJ referenced the Lotus case in a merits decision, and even in these decisions, the Court refused to meaningfully engage with the Lotus ruling.Footnote ¹⁵ Commentators identified further cases, where the spectre of Lotus seemed to be lingering in the background – the Corfu Channel case,Footnote ¹⁶ the Kosovo opinion,Footnote ¹⁷ and the Arrest Warrant Footnote ¹⁸ case are such examples – but no explicit references were made to the judgment itself. It is almost as if Lotus was considered as taboo or, at the very least, an overly sensitive topic by the Court. (Against this background, the comparative relevance of the Norstar and Enrica Lexie tribunals’ engagement with Lotus is even more significant.)

Such aversion on the ICJ’s part is not so surprising, considering that some judges of the Court heavily criticized Lotus in their separate and dissenting opinions: the Separate Opinion of Judges Higgins, Kooijmans, and Buergenthal to the Arrest Warrant case characterized Lotus’s position on the law of jurisdiction as ‘the high water mark of laissez-faire in international relations, and an era that has been significantly overtaken by other tendencies’.Footnote ¹⁹ Higgins criticized Lotus elsewhere too, noting that it should be looked at ‘with great caution’.Footnote ²⁰

The third reason why Lotus is special is that discourse on the judgment has been largely diverted from what the case was actually about. Granted, the PCIJ was, at least partially, responsible for this diversion through the formulation of an obiter dictum, whose controversy grew so large that it overshadows the narrower legal point of the case in its ratio decidendi. ^{Footnote 21} According to the infamous adage – the so-called Lotus principle – ‘[r]estrictions upon the independence of States cannot therefore be presumed’. ^{Footnote 22} It is unnecessary to engage here either with the holding itself or the criticisms levelled against it. I will limit myself to making a small observation on how the PCIJ’s holding grew to be – at least partially – misconstrued.

The root of the problem was probably Judge Loder’s dissenting opinion to the Lotus majority judgment, in which he wrote: ‘under international law everything which is not prohibited is permitted … in other words, on the contention that, under international law, every door is open unless it is closed by treaty or by established custom’. ^{Footnote 23} This recharacterization has become all but synonymous with the original judgment’s holding.Footnote ²⁴ There are two problems with this. First, Judge Loder’s rephrasing was not identical in meaning to the original. ^{Footnote 25} Second, Loder used this quote to summarize Turkey’s position in the proceedings – although, admittedly, according to Loder, the majority adopted essentially the same position.Footnote ²⁶

Thus, it would seem that an obiter dictum,Footnote ²⁷ taken out of context and mischaracterized by a dissenting judge, was largely influential in casting a negative light on the judgment in toto. This is all the more astonishing, considering the conflict between the obiter and the decision’s narrower ratio decidendi. Guilfoyle notes that ‘if the Lotus principle means what it literally says, then the PCIJ in Lotus failed to apply it’. ^{Footnote 28} This is correct. What is more, the Lotus principle was not only irrelevant to the Court’s final holding, it was at odds with it, as will be shown below.

Finally, no case of the PCIJ or the ICJ – save for the highly problematic South West Africa judgmentFootnote ²⁹ – has been subject to the amount of scholarly criticism Lotus has. Already in the wake of the judgment, many commentators voiced their concerns about it.Footnote ³⁰ Criticism remained the norm for the decades to follow: a representative example is Mann’s 1964 Hague Academy Lecture, where he made the point that ‘nothing was decided in the Lotus case’.Footnote ³¹

Thus, it is all the more surprising that recent years have seen a slight shift in scholars’ attitudes towards Lotus. By offering a critique of the criticism, Hertogen defended Lotus from a policy perspective, arguing that what we understand under the ‘Lotus principle’ is based on a mistaken reading of the majority judgment and that it is the prevailing reading of Lotus that should be ‘discarded as inaccurate’ and not the judgment itself.Footnote ³²

Others have also followed course, calling for a more nuanced reading of the judgment, considering the Lotus principle ‘in its immediate textual context or framed against the wider circumstances of the case’.Footnote ³³ According to another recent article, it is only the ‘conceptual origins’ of Lotus that are problematic and not the ‘core content’ of the case.Footnote ³⁴ The same article talks about a ‘scholarly ambiguity about the current standing of the Lotus rule in international law’.Footnote ³⁵ Considering the original approach of outright hostility towards the case, ‘ambiguity’ is already a great step forward for Lotus and its legacy.

2.2 Jurisdiction in Lotus

An undeniable privilege one enjoys when contributing to a century of commentary on perhaps the most famous inter-state dispute is that one can remain cursory in lining out the facts of the case. Lotus, in the PCIJ’s very own words, was about:

Following the incident, Turkish authorities arrested the Lotus’s first officer, Lieutenant Demons, and a Turkish court sentenced him to 80 days of imprisonment (and a fine of £20).Footnote ³⁷

The question before the Court was whether Turkey had breached international law by instituting criminal proceedings against Lieutenant Demons. The Court found in favour of Turkey. The Court’s decision was supported by the narrowest possible margin of majority: six judges for and six judges against, with President Huber’s casting vote.Footnote ³⁸

A brief summary of the PCIJ’s reasoning is reproduced below. I will only quote segments of the award that will be directly relevant in the context of the article’s discussion on Norstar and Enrica Lexie as well.

The Court first spelled out the general rule of prohibition on the extra-territorial exercise of jurisdiction. Notably, this paragraph followed directly after the Court’s pronouncement of the Lotus principle and is in stark contrast to it (or, alternatively, it is a carve-out of massive proportions):

The PCIJ then introduced another carve-out to the above rule (i.e., an exception to the exception):

The distinction between these two paragraphs is among Lotus’s most valuable contributions to international law. What the Court did here is it introduced a distinction between extraterritorial enforcement jurisdiction on the one hand (the exercise of which, as a general rule, is prohibited), and extraterritorial prescriptive (legislative) and adjudicative jurisdiction on the other (the exercise of which, as a general rule, is permitted).Footnote ⁴¹ In case of the latter categories (prescriptive and adjudicative jurisdiction), the judgment makes this pronouncement explicit, while in case of the first category (enforcement jurisdiction), the reference is implicit (and also follows from the process of elimination, i.e., if there is a special rule governing prescriptive and adjudicative jurisdiction, the only type of jurisdiction ‘remaining’ is enforcement). This trichotomy still provides, to this day, the basic framework of how we think of jurisdiction in international law.Footnote ⁴²

As the facts of the Lotus case concerned the exercise of Turkey’s adjudicative jurisdiction, the case could have been decided (with the same result) based on the above holding alone. Granted, the rest of the reasoning still added some welcome nuance. A couple of pages later, the Court applied the above general considerations to the specific – law of the sea – context of the case with the expected results. It is useful to quote the relevant section in full, as it played an important role in Norstar and Enrica Lexie as well:

Thus, the Court applied the same dichotomy between enforcement and prescriptive–adjudicative type of jurisdiction in the context of the law of the sea and found no lex specialis rule to necessitate a diversion from its original approach.

Notably, both the Norstar and the Enrica Lexie tribunals relied on the italicized first sentence of the first paragraph in the quote.Footnote ⁴⁴ However, they did so without also quoting the second italicized segment of the text, thus breaking the internal logic of the reasoning and distorting its meaning.Footnote ⁴⁵ This will be discussed in more detail below.

The final Lotus holding that is relevant for our purposes concerned the exercise of concurrent jurisdiction (in case of incidents on high seas):

The finding on concurrent jurisdiction is not particularly controversial in itself. However, the Court, in reaching this conclusion, also held that vessels should be regarded as the ‘floating territory’ of the state.Footnote ⁴⁷ This legal fiction has attracted much criticism from the dissenting judges, and has since been deemed to have ‘fallen into disrepute’.Footnote ⁴⁸ The Danish Company Tax Liability case offers a helpful explanation as to why this fiction is ‘misleading’. Just one reason is that vessels would have their own territorial sea around them if state territory was interpreted verbatim.Footnote ⁴⁹ This is understandably problematic.

This, however, does not mean that Lotus’s holding, in this context, is erroneous. Rather, it seems to be the case that it is permissible to treat vessels as the floating territory of the state for the (narrow) purposes of exercising adjudicative jurisdiction as established in Lotus,Footnote ⁵⁰ but not otherwise.

2.3 Jurisdiction after Lotus

The Lotus case concerned the narrow issue of states’ right to exercise their adjudicative (criminal) jurisdiction in a law of the sea context. In spite of this, the judgment has contributed greatly to the discipline of jurisdiction in international law in general. Lotus is the locus classicus for rules on concurrent jurisdiction, the objective territoriality principle, the ‘effects’ doctrine, and collision on the high seas (in this last regard, with the caveat of UNCLOS’s carve-out on collisions and incidents of navigation).Footnote ⁵¹

The trichotomy that was never spelled out, only implied in Lotus, has come to govern our understanding of international law rules on exercising jurisdiction.Footnote ⁵² The Third Restatement provides a concise taxonomy of the three types of jurisdiction:

Most authorities follow this trichotomy.Footnote ⁵⁴ There are good reasons for this, as ‘[e]ach form of state jurisdiction is distinct. They are subject to different tests and differ in scope …’.Footnote ⁵⁵ Some scholars, however, only recognize the two major categories of ‘legislative/prescriptive’ and ‘enforcement’ jurisdiction, making only passing mention of adjudicative jurisdiction.Footnote ⁵⁶

This article proposes that it is advisable to follow the three-pronged approach of the Restatement, as adjudicative jurisdiction possesses sufficiently distinct characteristics. Thus, treating it as analogous to either enforcement or prescriptive jurisdiction could prove to be problematic. The Norstar dispute is a prime example why.

3.1 Summary

The M/V ‘Norstar’ was an oil tanker owned by a Norwegian company, chartered to a Maltese company, flying the flag of Panama. From this flurry of jurisdictions, it is the Panamanian one that is going to be relevant for our purposes, as it was Panama, the ‘Norstar’s’ flag state, that initiated proceedings against Italy before the ITLOS. The dispute was about a Decree of Seizure issued by the Italian authorities that eventually led to the arrest and seizure of the ship by Spanish authorities in Spanish territorial waters.Footnote ⁵⁷

From 1994 to 1998, the ‘Norstar’ engaged in supplying gas oil to mega yachts in international waters ‘off the coasts of France, Italy and Spain’.Footnote ⁵⁸ The sale of gas oil to vessels at sea is known as ‘bunkering’ or ‘offshore bunkering’.Footnote ⁵⁹ In 1997, the Italian fiscal police initiated an investigation against the ‘Norstar’ and its affiliated broker company registered in Italy. The investigation revealed that the ‘Norstar’s’ activities were in breach of the applicable tax laws:

Italian authorities initiated criminal proceedings against a number of individuals involved, including the ‘Norstar’s’ captain.Footnote ⁶¹

On 11 August 1998, the Public Prosecutor at the Court of Savona, Italy, issued a Decree of Seizure against the ‘Norstar’, proclaiming that the vessel ‘as well as the oil product transported therein must be acquired as corpus delicti’.Footnote ⁶² The same day, the Italian prosecutor’s office also sent a request for judicial assistance to Spanish authorities to enforce the Decree of Seizure. The request was based on the European Convention on Mutual Assistance in Criminal Matters (European Convention). On 25 September 1998, Spanish authorities in Palma de Mallorca seized the M/V ‘Norstar’.Footnote ⁶³

In 2003, the competent Italian court acquitted the persons that had been charged and also revoked the ‘Norstar’s’ seizure, ordering that the ship be released to its owner.Footnote ⁶⁴ According to Panama, the owner was never made aware of this.Footnote ⁶⁵ Finally, the ‘Norstar’ reached a sad fate when a waste management company bought it at a public auction in 2015 to convert it into scrap.Footnote ⁶⁶

The case marked the first preliminary objections judgment in the ITLOS’ history.Footnote ⁶⁷ The tribunal delivered the judgment on 4 November 2016, ruling that it had jurisdiction to adjudicate the dispute and Panama’s claims were admissible. Although the dispositif did not make this clear, based on the judgment’s reasoning, the tribunal in effect only upheld its jurisdiction to Panama’s claims under UNCLOS Articles 87 and 300, and found that the rest of the provisions that Panama had relied on ‘[could not] be invoked in the present case’.Footnote ⁶⁸ The tribunal’s decision on upholding its jurisdiction enjoyed a 21–1 majority – with Judge Treves, the ad hoc judge appointed by Italy, dissenting – while the decision on admissibility was adopted with a 20–2 ratio.Footnote ⁶⁹

One controversial aspect of the preliminary objections judgment was the tribunal’s decision that Italy (and not Spain) was the proper respondent. Italy argued that the tribunal had no jurisdiction ratione personae as Spain was the proper respondent. Italy’s defence rested on the question: how exactly could the arrest and seizure of the ‘Norstar’ – both carried out by Spanish authorities – be attributed to Italy? The tribunal was unable to offer a satisfying answer to this conundrum,Footnote ⁷⁰ and held that ‘[as the Tribunal] found that Italy is the proper respondent, the question whether the conduct of Spain is attributable to Italy is irrelevant for the purpose of determining the proper respondent’.Footnote ⁷¹ Other than being slightly tautological, this holding is probably the most baffling of all in both Norstar judgments, and one that had conceptual and legal repercussions for the merits judgment as well.

Admittedly, the tribunal’s analysis is on point about the integral part that Italy played in the proceedings. Nevertheless, this only proves that there would have been no case against Spain as a sole respondent. It does not prove that there was a case against Italy as a sole respondent.

The solution would have been for Panama to sue both Italy and Spain together. There seems to be no indication on the fact record as to why Panama did not involve Spain in the proceedings or whether it had attempted to do so. According to Panama, ‘Spain has not been mentioned, summoned, cited, or even referred to in this case either as defendant or as a third party, nor has it shown any interest in participating through any of the possible methods accepted by the Convention’.Footnote ⁷² One can only wonder why Spain had not lined up to be on the receiving end of a claim that was in the hundreds of EUR millions.

The final leg of the tribunal’s analysis was the rejection of Italy’s admissibility objection on the basis that Spain was an indispensable third party in the proceedings.Footnote ⁷³ This outcome followed logically from the tribunal’s holding that Italy was the proper respondent. On the whole, the holding is aptly described as ‘a decision that while technically consistent with prior authority, was markedly inconsistent with reality’.Footnote ⁷⁴

On the merits, the tribunal also found in Panama’s favour, the victory was, however, a pyrrhic one. In terms of the compensation to be paid, the tribunal found that the Italian court’s 2003 judgment ordering the ‘Norstar’s’ unconditional release interrupted ‘the causal link between the wrongful act of Italy and damage suffered by Panama’. Therefore, Panama was not allowed to claim any damages for the loss of the ‘Norstar’ sustained after this date. In addition to this, the tribunal rejected the rest of Panama’s monetary claims (for the shipowner’s loss of profits, the payment of wages, payment of fees and taxes, damage to the charterer, material and non-material damage for natural persons). Thus, the final amount awarded was US$ 285,000 plus pre-award interest compounded since 1998, amounting altogether to approximately US$ 500,000. This was less than one percent of Panama’s original claim.Footnote ⁷⁵

3.2 Freedom of navigation and exclusive flag state jurisdiction in the judgment

The tribunal’s (re-)framing of Panama’s claims in its judgment on jurisdiction created problems for the merits analysis as well. Italy viewed the tribunal’s judgment as ‘curtailing’ the dispute and limiting it to the lawfulness of Italy’s Decree of Seizure.Footnote ⁷⁶ According to the tribunal, however, the dispute involved ‘not only the Decree of Seizure and the Request for its execution’ – undertaken by Italy, the respondent – but ‘also the arrest and detention of the Norstar’ – undertaken by Spain.Footnote ⁷⁷

Next, the tribunal turned to the facts of the case, and found that ‘the Decree of Seizure and its execution concern both alleged crimes committed in the territory of Italy and bunkering activities conducted by the “Norstar” on the high seas’.Footnote ⁷⁸ Naturally, for the purposes of the dispute, only the activities conducted on the high seas were relevant.Footnote ⁷⁹

The tribunal then undertook a detailed analysis under UNCLOS Article 87(1), which is, without doubt, the most exciting and controversial part of the merits judgment. Article 87(1) lays down the fundamental rule on states’ right to freedom of navigation. In its reasoning under Article 87(1), the tribunal extensively relied on Article 92(1) as interpretative guidance. Article 92(1) provides for the general rule of exclusive flag state jurisdiction on the high seas:

The most recent – 2017 – commentary to the UNCLOS limits the scope of Article 92(1) to enforcement jurisdiction. According to the commentary, the provision:

In addition to being ‘canonical’ by virtue of its inclusion in the commentary, this is also the mainstream viewpoint shared by the majority of scholars.Footnote ⁸¹ While the UNCLOS commentary helps limit the scope of Article 92(1), it, unfortunately, does not adhere to the three-pronged approach in qualifying the different types of jurisdiction. Accordingly, it does not make an explicit reference to adjudicative jurisdiction. Nonetheless, for the purposes of Article 92(1), adjudicative jurisdiction should logically be grouped with prescriptive jurisdiction, and thus remains unaffected by the provision.

Turning back to the analysis in Norstar, the tribunal’s linking of Article 87(1) to Article 92(1) was not entirely without textual support: Article 87(1) itself provides that ‘[f]reedom of the high seas is exercised under the conditions laid down by this Convention and by other rules of international law’. The tribunal made the point that ‘[f]reedom of navigation would be illusory if a ship – a principal means for the exercise of the freedom of navigation – could be subject to the jurisdiction of other States on the high seas’.Footnote ⁸² In support of this statement, the tribunal relied on the holding of the Lotus judgment already quoted above:

This is the majority judgment’s only reference to Lotus, however, it is strategically placed. The tribunal followed up the reference with a confirmation that Article 92(1) UNCLOS was the reflection of the above Lotus quote. Hence, Lotus helped provide a crucial link between Articles 87(1) and 92(1). It is somewhat unfortunate that the PCIJ’s own caveatFootnote ⁸⁴ to the quote did not make it into the Norstar judgment, as it would have offered valuable guidance for the analysis that followed.

Next, the tribunal confirmed – relying on consistent ITLOS case law – that bunkering activities fell within the scope of freedom of navigation.Footnote ⁸⁵

The tribunal then continued its legal analysis of Article 87(1), holding that, in addition to physical or material interference, ‘even acts which do not involve physical interference or enforcement on the high seas may constitute a breach of the freedom of navigation’.Footnote ⁸⁶ The tribunal further noted that ‘Italy recognizes the possibility that acts falling short of enforcement action on the high seas could be relevant in terms of a breach of article 87 of the Convention, if such acts produce some “chilling effect”’.Footnote ⁸⁷ Thus, it seems that it was common ground between the parties – and the tribunal – that acts that did not amount to physical interference could breach Article 87(1).

In the next paragraph, however, the tribunal discarded Italy’s qualification requiring a ‘chilling effect’ and stated instead that:

For the analysis of the actual breach, the tribunal relied on Article 92(1) (or, in the tribunal’s formulation, ‘the principle of exclusive flag state jurisdiction’) to hold that the provision:

Quite bafflingly, the tribunal offered no authority or further justification on this point, which the whole case turned upon. Actually, in the 121-page long award, this was the first mention of the term ‘prescriptive jurisdiction’ other than in the summaries of the parties’ positions. The tribunal concluded by stating that:

One can only wonder why the ITLOS decided to adopt such a controversial position and why it did so in such a cursory manner. Leaving the legal correctness of the decision aside for the moment, the tribunal could certainly have made more of an effort to come up with justifications to support a position that ‘run[s] against the grain of prevailing scholarly opinion’.Footnote ⁹¹

Other commentators have made compelling points on why such an expansive reading of the rule on exclusive flag state jurisdiction is problematic from a policy perspective.Footnote ⁹² I do not wish to repeat these arguments here. I would instead like to make the point that other than erroneous; the tribunal’s key holding was superfluous.

This is because (i) the issuance of the Decree of Seizure did not constitute an exercise of prescriptive jurisdiction, rather it qualified as an exercise of adjudicative jurisdiction; (ii) accordingly, on the facts of the case, it was unnecessary for the tribunal to attribute such an extensive scope to Article 92(1); and therefore, (iii) the tribunal’s holdings are essentially obiter dicta, at least to the extent that they concern prescriptive jurisdiction.

According to the ITLOS, Italy’s breach of Article 87(1) entailed the following acts: (i) the adoption of the Decree of Seizure; (ii) the request for judicial assistance addressed to Spain; and (iii) the arrest and detention of the ship.Footnote ⁹³ The question is: which of these actions qualified as an exercise of Italy’s prescriptive jurisdiction? The arrest and detention of the ship most certainly did not: they are clear examples of the exercise of enforcement jurisdiction (leaving aside, for the moment, the fact that the arrest was carried out by Spain and not Italy). That leaves the issuance of the Decree and the request for judicial assistance.

The Decree was adopted by the Italian Prosecutor’s office in the course of proceedings before a criminal court.Footnote ⁹⁴ Italy’s aim with the adoption of the Decree was to ‘subject persons or things to the process of its courts’.Footnote ⁹⁵ Thus, the adoption of the Decree was an exercise of Italy’s adjudicative jurisdiction. Italy’s prescriptive jurisdiction was exercised through the adoption of its laws that served as the basis of the adoption of the Decree of Seizure. However, the Norstar judgment did not discuss this regulatory framework as part of the factual background and provisions of Italian law were only referenced in passing. Indeed, the tribunal limited its focus and analysis to the Decree, which, however, is not a legislative act.Footnote ⁹⁶ And this is unsurprising in light of the mainstream view, according to which ‘… the mere exercise of prescriptive jurisdiction, without any attempt at enforcement, will not normally have to pass the test of international law’.Footnote ⁹⁷

Italy’s request for execution under the European Convention is harder to classify. As the request was issued by an Italian court in the course of the criminal proceedings, it should also belong under the category of adjudicative jurisdiction. However, it could be argued that – as the request entailed the arrest and detention of the ship – it is better characterized as an exercise of enforcement jurisdiction. Either way, the request was not an exercise of prescriptive jurisdiction.

In light of this, there seems to be no good reason why the tribunal chose to attribute an unnecessarily broad reading to a legal standard in a case, where the factual background did not even warrant the application of that legal standard. Even the narrower grounds suggested by this article would still leave room for criticism: it is not self-evident that the issuance of the Decree of Seizure as an exercise of adjudicative jurisdiction impaired the ‘Norstar’s’ freedom of navigation. But at least the judgment would be conceptually and legally more consistent, and its holdings would not have the far-reaching policy implications that are referenced by its critics.

3.3 Freedom of navigation and exclusive flag state jurisdiction in the joint dissenting opinion

Seven judges of the ITLOS penned a joined dissenting opinion, heavily criticizing the award.Footnote ⁹⁸ The dissent’s core point was that Article 87(1) did not apply to the case and, even if it had been, Italy had not violated it. Notably – and somewhat surprisingly – the dissenting judges accepted and adopted the majority’s classification of Italy’s acts as an exercise of its prescriptive jurisdiction, and formulated their arguments accordingly.

Where the dissenters did disagree with the majority judgment was the scope of Article 87(1). The dissenting judges attributed a narrower reading to Article 87(1), holding that it protected the free movement of vessels ‘primarily from the exercise of enforcement jurisdiction by non-flag States on the high seas’.Footnote ⁹⁹ Notably, just a couple of paragraphs later, the opinion seemed to concede that Article 87(1) ‘may also protect vessels on the high seas from the prescriptive jurisdiction of non-flag States’ under certain conditions (not met in Norstar).Footnote ¹⁰⁰

In another two paragraphs, the dissenters returned to their main argument against the majority’s broad interpretation of exclusive flag state jurisdiction.Footnote ¹⁰¹ Thus, while the dissenting judges were willing to concede that Articles 87(1) and 92(1) might, in certain cases, be breached through the exercise of prescriptive jurisdiction, they disagreed with the ‘wholesale’ prohibition approach of the majority. To support this point, they relied on a string of citations, referencing Guilfoyle’s contribution to the 2017 UNCLOS commentary, who, in turn, had referenced a French treatise from the 1930s and – as it should be no surprise to the reader by this point – the Lotus case. As stated above, Guilfoyle’s views on prescriptive jurisdiction represent the majority view on this issue.Footnote ¹⁰² Also, the position seems to enjoy some post hoc support as well by subsequent commentary and criticism of the majority decision.Footnote ¹⁰³

The dissent also picked up on the majority judgment’s perplexing phrasing of its main holding: ‘[exclusive flag state jurisdiction] prohibits … the extension of their prescriptive jurisdiction to lawful activities’.Footnote ¹⁰⁴ According to the dissenters, ‘[t]his would seem to suggest that a non-flag state is not excluded from extending, in conformity with international law, its prescriptive jurisdiction to the unlawful activities of foreign vessels or of persons on the high seas’.Footnote ¹⁰⁵ Indeed, the majority’s qualification here is unhelpful. ‘Lawful’ – under which set of laws? International law or domestic law? In vacuo, every activity is ‘lawful’ until it becomes subject to regulation. There is also a temporal side to the question: as it were the case in Norstar, the Italian courts ended up finding that the ‘Norstar’s’ activities had been lawful all along. This, however, could not have been foreseen at the time of the issuance of the Decree of Seizure (or at the time of adopting the relevant Italian legislation, for that matter).

The opinion closed with the argument that ‘a State may exercise its prescriptive criminal jurisdiction with respect to conduct on the high seas where such conduct is integral to an alleged crime committed in the State’s territory, not when it is justified or allowed by international law to do so, but when it is not prohibited by international law to do so’.Footnote ¹⁰⁶ The statement is supported by a reference to the Lotus case, elegantly tying together Lotus’ narrower point on the limits of exclusive flag state jurisdiction, with its meta-points on the structure of international law. In legal terms, the argument is somewhat superfluous in an otherwise mostly convincing dissenting opinion, also for the reason that it engages with the more controversial aspects of Lotus’ legacy.

Nevertheless, it shows the considerable might of the Lotus decision that it guided the majority as well as the minority judges in their considerations. At the same time, it should caution us that it was also the Lotus case – together, of course, with the applicable treaty provisions – that allowed the two groups of judges to arrive at polar opposite conclusions.

In the following section, the article will discuss the Enrica Lexie award. As will be shown, the Enrica Lexie tribunal was more successful in utilizing the Lotus judgment than the ITLOS judges, particularly in its interpretation of concurrent jurisdiction over the incident. Alas, the arbitrators ran into some of the same pitfalls when discussing the scope of UNCLOS Article 92(1).

4.1 Summary

The facts of the Enrica Lexie case revolved around the near-collision between the ‘Enrica Lexie’, an Italian-flagged vessel, and the ‘St. Antony’, an Indian vessel,Footnote ¹⁰⁷ and a shooting incident costing the lives of two Indian fishermen. The ‘Enrica Lexie’s’ crew included a six-member vessel protection detachment of Italian marines.Footnote ¹⁰⁸ The marines’ task was to protect the ship from threats of piracy.

On 15 February 2012, as the ‘Enrica Lexie’ was passing through India’s contiguous zone, the ship’s crew noted that the ‘Enrica Lexie’ was on a collision course with the ‘St. Antony’. Given that the attempts of the ‘Enrica Lexie’s’ crew to contact the ‘St. Antony’ through radio and visual signals were unsuccessful, the crew had been under the impression – mistakenly, as it later turned out – that the ‘St. Antony’ was a pirate ship targeting the ‘Enrica Lexie’. In reaction to this – and also in an attempt to avoid a collision – the marines aboard the ‘Enrica Lexie’ fired three consecutive rounds of warning shots. After the third round of shots, when the two vessels were 30 metres apart, the ‘St. Antony’ changed course, avoiding the two vessels’ collision.Footnote ¹⁰⁹

The ‘St. Antony’ was an Indian fishing boat, and the shots fired by the marines killed two Indian fishermen. The reason why the ‘St. Antony’ did not react to the ‘Enrica Lexie’s’ attempts to communicate was that the boat’s captain and most of the crew were asleep after a long night of fishing.Footnote ¹¹⁰

Following the incident and upon the Indian coast guard’s request, the ‘Enrica Lexie’ changed its course towards India’s territorial sea. Once the ship arrived at the Kochi anchorage, the Indian authorities commenced an investigation and later arrested the two Italian marines.Footnote ¹¹¹

While the crux of the case was whether India’s exercise of jurisdiction (prescriptive, adjudicative, and enforcement) was lawful and whether the marines enjoyed immunity from jurisdiction, Italy brought claims under several other UNCLOS provisions as well. India also brought counterclaims against Italy, arguing, inter alia, that the marines’ actions violated India’s right to freedom of navigation.

Already on these facts, the similarities between Enrica Lexie and the Lotus case are apparent. The factual background of Lotus (‘a collision on the high seas between two vessels flying different flags, on one of which was one of the persons alleged to be guilty of the offense, whilst the victims were on board the other’Footnote ¹¹² ) is almost identical to facts in Enrica Lexie. One important difference is that there was no collision in Enrica Lexie, only a near-collision. The fact that the events of the Enrica Lexie case took place in India’s contiguous zone is, however, not a real distinguishing feature: by virtue of UNCLOS Article 58(2), the Convention’s provisions on the high seas are also applicable to the contiguous zone, and the case was decided under these provisions.

Enrica Lexie provides an excellent opportunity to pick up the threads of Norstar on the importance of distinguishing between the distinct types of extraterritorial jurisdiction on the high seas: Italy alleged that India had breached international law through (i) the exercise of its prescriptive jurisdiction (by adopting its Maritime Zones Act); (ii) the exercise of its adjudicative jurisdiction (by bringing criminal proceedings against the marines for acts committed on the high seas aboard an Italian vessel); and (iii) the exercise of its enforcement jurisdiction (through ‘interdicting’ and ‘escorting’ the ‘Enrica Lexie’ to an Indian port.)

Similarly to the Norstar case, Enrica Lexie was an UNCLOS-based dispute, however, it was adjudicated by an ad hoc Annex VII arbitral tribunal, and administered by the Permanent Court of Arbitration. Prior to the rendering of the award in 2020, two provisional measures orders were delivered: in the first order, the ITLOS, acting under Article 292 of the UNCLOS, ruled on the suspension of all criminal proceedings in India and Italy in 2015;Footnote ¹¹³ and in the second order, the Annex VII arbitral tribunal ruled in 2016 that India release one of the two marines held in its custody, due to his health conditions.Footnote ¹¹⁴

In the merits award, delivered in May 2020, the tribunal rejected most of Italy’s claims on India’s violations of various UNCLOS provisions but upheld the most foundational one: India’s exercise of jurisdiction was found to have violated the marines’ immunity, and the tribunal ordered the marines to be released to Italy to be tried. The tribunal also upheld one of India’s counterclaims, finding that the shooting incident amounted to a violation of India’s freedom of navigation.

Most of the award’s key issues were decided based on a narrow majority of three arbitrators against two. In an unusual matrix of dissenting opinions, the two dissenting arbitrators penned individual dissents and a joint dissent as well.Footnote ¹¹⁵ While these dissenting opinions offer valuable criticism and insights on the points they raised, this article will not subject them to closer scrutiny, as the legal issues highlighted here all happened to enjoy unanimous support among the five arbitrators.

The Enrica Lexie award was better received than Norstar, and most commentators limited their observations to smaller criticisms on more nuanced points. These criticisms were focused mostly on the tribunal’s ruling on the marines’ functional immunity, and whether the tribunal had jurisdiction to adjudicate upon this incidental issue.Footnote ¹¹⁶

While these are certainly the ‘central’ issues of the case and they break new legal ground, the singular attention devoted to them is somewhat regrettable. This is the case all the more so as the award is exceptionally rich in legal detail, addressing questions which have rarely been in the focus of international adjudication: leaving aside the issues of jurisdiction and applicable law, the case engaged – at least – 14 different provisions of the UNCLOS.Footnote ¹¹⁷ This article aims to remedy this, at least to some extent, by focusing on two less-explored issues in Enrica Lexie, also guided by this article’s underlying Lotus theme.

4.2 Concurrent adjudicative jurisdiction in the trichotomy of jurisdictions

From the many issues discussed in the award, there is one that directly engages with the Lotus case: the question of whether Italy had exclusive jurisdiction to initiate criminal proceedings against the marines, or, alternatively, Italy and India both had concurrent – adjudicative – jurisdiction over the incident (the same way as France and Turkey in Lotus).

The issue bore great importance as had the tribunal sided with Italy, the dispute would have been over, and there would have been no need for the tribunal to discuss the thorny issue of the marines’ immunity (and the equally complex question of its own jurisdiction over the question of the marines’ immunity).

The tribunal’s analysis on this point is murky because it is split into two (or, potentially, three)Footnote ¹¹⁸ different parts throughout the award. The first part of the analysis is located under Italy’s first claim on how the adoption of certain maritime acts – i.e., the exercise of India’s prescriptive (legislative) jurisdiction – allegedly breached numerous UNCLOS provisions. It is in this context that the arbitral tribunal first analysed whether India had a potentially valid basis under international law for exercising its jurisdiction.Footnote ¹¹⁹

However, even before commencing the analysis, the arbitral tribunal already discarded Italy’s claim, seeing ‘no need to address that issue in the context of the present dispute’.Footnote ¹²⁰ Thus, the Enrica Lexie award said nothing – neither in the reasoningFootnote ¹²¹ nor in the ratio decidendi Footnote ¹²² – on the lawfulness of India’s exercise of its prescriptive jurisdiction.

In the very next paragraph, the tribunal jumped to the question of the legal basis of India’s exercise of its jurisdiction ‘over the “Enrica Lexie” incident and the Marines’.Footnote ¹²³ This was a stark shift to the questions of adjudicatory and enforcement jurisdiction, and one that was not made explicit by the tribunal.Footnote ¹²⁴ The award found that the application of the objective territoriality principle to ships and aircraft was well-founded in international law, and was not the same as treating the ship as the floating territory of the state. The key reference here was, once more, the Lotus case.Footnote ¹²⁵

Having found that the objective territoriality principle was a valid basis for India’s exercise of jurisdiction, the tribunal had to test this basis against (i) the rule of exclusive flag state jurisdiction under Article 92(1); and (ii) Article 97(1)’s carve-out rule on collisions and incidents of navigation of the UNCLOS.Footnote ¹²⁶

Regarding UNCLOS Article 92(1), the tribunal held that India’s exercise of jurisdiction was:

While it is not spelled out explicitly, it is relatively clear that the tribunal was referring to the exercise of India’s adjudicative jurisdiction. Thus, according to the tribunal, at least in the context of adjudicative jurisdiction, there was no conflict between Article 92(1)’s requirement of exclusive flag state jurisdiction and the Lotus-based exercise of concurrent jurisdiction based on the simultaneous application of the objective and subjective territoriality principles. The two could be reconciled.

Whether the same considerations applied to enforcement and prescriptive jurisdiction, the award did not say. The tribunal later returned to the interpretation of Article 92(1) in the context of India’s exercise of its enforcement jurisdiction, which is something I will address in more detail in Section 4.3.

The tribunal’s analysis of the exercise of India’s adjudicative jurisdiction was, however, incomplete at this point. The award picked up the thread approximately 80 pages later.Footnote ¹²⁸ A final hurdle was the question of whether Article 97(1)’s carve-out rule on collisions and incidents of navigation applied to the case.

While the Enrica Lexie incident was decidedly not a collision, it was possible to raise the argument (and Italy did so) that it constituted an incident of navigation, and, under Article 97(1), only Italy – the flag state – had jurisdiction to initiate criminal proceedings against the ‘Enrica Lexie’s’ crew members.

In this segment, the tribunal also elaborated upon the facts and the afterlife of the Lotus judgment, then continued with an in-depth analysis of the interpretation of the term ‘incident of navigation’.Footnote ¹²⁹ Finally, based on a fact-driven analysis, the tribunal found that the incident did not qualify as an ‘incident of navigation’, so there was no exception to the applicable rule of concurrent jurisdiction.Footnote ¹³⁰ Accordingly, India would have won the case, had it not been for the question of the marines’ immunity, which finally decided the case in Italy’s favour.

Figure 1 depicts the tribunal’s thought process in the framework of the UNCLOS’s provisions.

Figure 1.

Flowchart modelling the reasoning of the arbiral tribunal in the Enrica Lexie case

In light of the above analysis, some of the criticism levelled against the tribunal’s treatment of Articles 92 and 97 can be, at least partially, rebutted. Devaney and Tams, for example, argue the following in connection with the two articles:

Contrary to the above observation, there is not necessarily a ‘tension’ between the tribunal’s reading of Articles 92 and 97, or rather: any tension is reconciled by the tribunal’s application of the Lotus rule. The tribunal held that (i) the Lotus rule on concurrent jurisdiction and Article 92(1) can be reconciled (at least in the case of adjudicative jurisdiction); and that (ii) Article 97 creates an exception to the Lotus rule (and not (directly) to Article 92(1)). So, for the purposes of adjudicative jurisdiction, the analysis is coherent. That being said, Devaney and Tams were otherwise correct to criticize the tribunal’s expansive reading of Article 92(1), which issue is further explored in the next section.

4.3 Enrica Lexie’s endorsement of Norstar

The tribunal analysed Article 92(1) in more detail under a separate claim: Italy’s claim for a breach of the provision through India’s exercise of its enforcement jurisdiction by directing, interdicting, and escorting the ‘Enrica Lexie’ to an Indian port. Notably, Italy brought no claims under Article 92(1) concerning India’s exercise of its prescriptive or adjudicative jurisdiction. But this did not stop the arbitral tribunal from attributing the same expansive interpretation to Article 92(1) that the Norstar tribunal did, finding that the provision prohibited the exercise of all kinds of jurisdiction, including both prescriptive and adjudicative jurisdiction.Footnote ¹³²

The tribunal supported its conclusion by recalling that the concept of jurisdiction under international law is generally understood to entail these three sub-categories, therefore, Article 92(1)’s prohibition must encompass all three of these domains.Footnote ¹³³ While the explanation is far from satisfactory, it is already more detailed than what the ITLOS provided in Norstar.

As it was stated in the Norstar Dissenting Opinion, the 2017 UNCLOS commentary adheres to a different interpretation, limiting Article 92(1)’s scope to a prohibition on non-flag-state enforcement.Footnote ¹³⁴ Unfortunately, the Enrica Lexie tribunal did not use the opportunity to reflect upon this, all the more so, as the tribunal otherwise heavily relied on the 2017 UNCLOS commentary, citing it on 17 different counts.

Another issue is the apparent contradiction between this holding and the tribunal’s holding on how Article 92(1) can be reconciled with the exercise of concurrent jurisdiction. How can Article 92(1) be at the same time an all-encompassing prohibition and the flexible rule that allows for Lotus-based carve-outs for the exercise of concurrent adjudicative jurisdiction? The tribunal offered no answer.

Next, the tribunal turned to a fact-driven analysis of India’s conduct and found that India’s actions – in interdicting and escorting the ‘Enrica Lexie’ – did not amount to an exercise of enforcement jurisdiction, so there was no breach of Article 92(1). While, on the facts of the case, this holding is not entirely unproblematic either, in a strictly legal sense, the analysis is sound.

The tribunal’s endorsement of Norstar’s expansive interpretation of Article 92(1) can be criticized for the same policy reasons as the Norstar decision, and so it has been.Footnote ¹³⁵ Indeed, as Honniball notes, ‘[f]or readers of the M/V Norstar Judgment, this oversight is a moment of déjà vu’.Footnote ¹³⁶ Less evident, however, is that Enrica Lexie’s holding on Article 92(1) – similarly to Norstar’s – is also an obiter dictum, although in a way that is somewhat different from Norstar.

First, as argued above, the tribunal made no pronouncements on India’s exercise of prescriptive jurisdiction. According to the award’s dispositif, there was ‘no need to address the question of the compatibility with UNCLOS of India’s 1976 Maritime Zone Act and its 1981 Notification’.Footnote ¹³⁷ The tribunal made this pronouncement unanimously. None of Italy’s additional claims engaged with India’s exercise of prescriptive jurisdiction. Thus, the award’s findings on the application of Article 92(1)’s prohibition of prescriptive jurisdiction were not integral to any of the findings in the dispositif; as such, they are obiter dicta.

Second, the article already reflected upon the tribunal’s analysis of India’s adjudicative jurisdiction. In essence, the tribunal held that Article 92(1) can be reconciled with the Lotus-type exercise of concurrent adjudicative jurisdiction. Thus, the tribunal either (i) held (or rather implied) that Article 92(1) did not encompass the exercise of adjudicative jurisdiction and later contradicted itself when it endorsed Norstar’s broad scope or, more probably, (ii) treated the Lotus rule as a carve-out to Article 92(1)’s prohibition. Either way, the tribunal’s holding on the broad prohibition under Article 92(1) was, once more, not integral to the decision on the exercise of adjudicative jurisdiction.

The most recent edition of Akehurst’s Modern Introduction to International Law makes the same observation: ‘[i]n Enrica Lexie … [t]he Tribunal seemed to be speaking generally about jurisdiction, including prescriptive jurisdiction, but India’s activities involved in that case were essentially about enforcement jurisdiction’.Footnote ¹³⁸

While it is challenging to piece together the Enrica Lexie tribunal’s reasoning, most of the award is legally sound and offers valuable guidance to less-discussed issues of the law of the sea. The Enrica Lexie tribunal’s utilization of Lotus is also commendable as it confirms the important role Lotus continues to play in the law of concurrent jurisdiction. In addition to Lotus, the arbitral tribunal also made good use of other timeworn cases, such as the Muscat Dhows caseFootnote ¹³⁹ and the Owners of the Jessie case.Footnote ¹⁴⁰ Finally, the tribunal relied heavily on the available commentaries to the UNCLOS and other scholarly publications. This is a strong (and welcome) departure from what the international community is used to in terms of the ICJ’s – or the ITLOS’s – practice of referencing secondary sources outside of their own output.Footnote ¹⁴¹

It is regrettable and, indeed, peculiar that, following the ITLOS’s footsteps, the Enrica Lexie tribunal also chose to adopt a controversial position on UNCLOS Article 92(1), especially as the position was not in any way needed for the tribunal to arrive at its final ruling.