Maritime Rule of Law: Some Preliminaries

Can there be rule of law at sea? In extending the traditionally terrene ideal seaward, there are a range of conceptual difficulties. These difficulties are outlined, and a recurring thought pattern is set out that is found in the traditions of thought about the rule of law as protecting members of a community from the abuse of power. Drawing on Jeremy Bentham’s scattered remarks about maritime governance, three assumptions underlying this thought pattern, regarding territoriality, community, and protective function, are identified as requiring modification in the maritime context. Achievement of the rule of law at sea is possible, but reflects a number of limitations as compared to its terrene counterpart, limitations related to these three traditional assumptions.

the oldest forms of transnational law, developed in coastal communities around the Mediterranean, Indian Ocean, and Southeast Asia well before the emergence of the modern state, and even before some ancient forms of political community (Anand 1982;Tetley 2002, 3-32). These pre-modern origins may suggest that there could be law at sea in some sense where there is not even in principle the possibility of the rule of law-if, for example, that ideal required something akin to a stable, statebased legal system to flourish. If that were so, maritime law would be a degenerate form of law, somehow unable to realize its potential as law. To put the point a different way, given the very persistence of maritime law over the centuries, despite the historically tenuous control of land-based legal institutions over the seas, it may be that that control has amounted to only rule by law, a far cry from the fully realized ideal (Ginsburg and Moustafa 2008;Krygier 2012). In either case, the prospects of the rule of law at sea would be dimmer than those of the worst failed states. By identifying, and then modifying, the implicit assumptions of the concept of the rule of law as a terrene ideal, I will argue, however, that we are not driven to such a conclusion. The rule of law, as traditionally conceived in the philosophical tradition, can and does extend to the sea, but understanding how does not come without its difficulties.
One way into the difficulties is to compare the rule of law at sea to the modern maritime extension of another venerable political ideal, one with which the rule of law is often paired: human rights. The proposed Geneva Declaration on Human Rights at Sea asserts as two of its fundamental principles that "human rights are universal; they apply at sea, as they do on land" and that "all persons at sea, without any distinction, are entitled to their human rights." 2 These principles affirm the universality of human rights standards, but do not assume that observance or violation of them at sea will take the same forms as on land. To the contrary, it is precisely because the relations between masters, crews, and passengers on ships are of distinctive sorts, and because those relations make possible actions and practices which violate human rights in distinctive ways, that the Declaration is important. The present point is that these principles rely on the idea that the rights themselves somehow travel with persons when they leave the terra firma of states, out into the nearby 'territorial' waters, and even into the high seas, where it has long been held that no claim of sovereignty by any state will be tolerated or accepted. 3 The reason why these principles do not pose any inherent conceptual difficulties is clear enough: the Footnote 1 (continued) present purposes, maritime law includes, on the one hand, the body of public international law known as the Law of the Sea, the now many international treaties governing navigation, the marine environment, shipping, and labour practices at sea, and, on the other, the body of mostly domestic private law referred to in common law countries as admiralty law, encompassing a patchwork of distinctive substantive and procedural rules governing maritime personal injury, contracts, and other transactions relating to maritime commerce.
bearers of human rights are persons, and persons do not change in any obviously relevant way when they set sail from port. And regardless of one's preferred accounts of the justification and nature of such rights, it is easily acknowledge that human rights apply to humans, wherever they go and whatever they do, including at sea.
But it is just here that the difficulty around the idea of the rule of law at sea emerges. For the rule of law does not seem as portable as such rights, or at least not portable in the same way (Krygier 2011). That is, the rule of law does not seem to travel out to sea with people as human rights do, and for a similar reason: the rule of law does not attach to, nor is directly had by, persons. The rule of law is an ideal of law, and perhaps only of legal systems displaying a certain institutional complexity. So it would seem that law would similarly have to leave port with people, in some sense to be explained, in order for the rule of law to be possible at sea. One traditional thought is that the rule of law requires judges in standard adjudicative institutions, who consistently and impartially apply a system of rules on a case-by-case basis, which in part protects both the rights of persons under the law and the integrity of the system as a whole (Raz 2009). But of course law, at least in this sense, is notably absent on the seas. These officials, and their institutional functions, are landbound; there are no floating courthouses or seafaring judges. The relatively recent establishment of the International Tribunal of the Law of the Sea is not to the point, for even it cannot provide the kind of ongoing accessibility to justice that courts display in regard to land-bound disputes, and to land-bound parties. Tribunals and other legal institutions initially seem unable to perform the same functions at sea for seafarers as they do when applying terrene law to disputes between landlubbers, if only for reasons of geography and circumstance. 4 It may be thought, however, that the rule of law extends to the sea just as human rights do because the rule of law travels, not with people, but with the ships on which they sail. For centuries, the ship registration (or 'flagging') system has been the primary mechanism by which law and the coercive powers of the state extend to peripatetic ships and their crews and passengers (Coles and Watt 2019). One route to our destination is to consider how that system actually brings law to the sea, and its limitations in doing so, to begin to determine how the ideal of the rule of law at sea could then be realized. The philosopher Jeremy Bentham once called a ship an 'ambulatory province ' (1802, 332) and thereby encouraged the juridical view that ships that fly the flag of a state are floating territories of that state. A ship that flies the French flag is a floating piece of France, and therefore French law applies naturally to the conduct of those aboard, including the actions of the ship itself (e.g., in cases of a collision or allision), wherever in the world the ship may be when the conduct occurs. 5 The rule of law for oceans, then, would not be an exercise in extra-territoriality (Kopela 2016), but rather a straightforward expansion of the territory of states. This view also brings about an elegant symmetry between human rights and the rule of law at sea: just as human rights attach to people, so the rule of law applies to territory, and extends to the sea by way of ships as mobile extensions of a state's territory.
The 'floating territory' theory flourished in the nineteenth century among international lawyers as a foundational principle of the ship registration system, as a way to provide order in an otherwise lawless and chaotic space (Clapham 2012, 227). 6 It later came to be regarded by some courts as a dangerous legal fiction, 7 and has allegedly "long fallen into disrepute" (Crawford 2012, 464). It has not. 8 And while the idea of a ship as a floating territory as a jurisdictional principle of international law may lead to unpalatable consequences (in which the borders of states would be in chaotic, constant flux), it is nonetheless a useful heuristic in clarifying how maritime rule of law is possible but distinctively complex.
I suggest that, despite its flaws, Bentham's theory helps us identify three traditional but rarely articulated philosophical assumptions about the rule of law, ones which require reassessment and refinement in the maritime context. The first is that the rule of law is achieved only over a territory: that legal governance itself allegedly requires extension over "a definite portion of the earth's surface" (Kelsen 1945, 217). Bentham attempts to extend this territorial assumption to ships, but I will argue that the maritime rule of law requires us to jettison any such assumption and to embrace the idea of a genuinely non-territorial legal regime. 9 Second, the rule of law is thought to require some form of political community, indeed, that it is impossible without it. Bentham seems to assume there can be political community at sea; I argue this is a view we should accept, but that he misidentifies the bounds of that community. Third, the rule of law, whatever else it may be, is thought to yield a distinctive sort of benefit to members of the community whose rule of law it is. The benefit has classically been expressed in terms of protection from or prevention against something undesirable, for example, that the rule of law is a bulwark against tyranny, or requires a government of laws and not of people who seek to rule according to self-interest, at the expense of the powerless. I suggest that the direct beneficiary of the rule of law at sea is the seafaring community itself, in that it protects its members from various forms of arbitrariness at sea. But it often has far more indirect beneficiaries, including the broader maritime community (e.g., shipowners and marine insurers), and indeed the rest of us in land-based communities, for example, in bringing about more effective protections of the marine environment and facilitating maritime commerce and international trade.
But I first excavate the concept of the rule of law itself, and identify what I call a thought pattern running through the Western tradition, regarding what it is, what it is for, and why it is valuable. The pattern provides a point of reference in modifying the rule of law seaward with respect to the background assumptions of territoriality, political community, and protective function. The adaptation of these assumptions provides a basis for understanding how the traditionally terrene ideal remains afloat in its new environs. 10

The Rule of Law: A Thought Pattern
One recent trend is to view the rule of law with a certain deflationary suspicion, either with outright hostility and dismissiveness toward the very idea (Shklar 1998), or else with qualifications that still carry a whiff of scepticism-for example, that it is an 'essentially contested concept', and related claims. 11 But in light of those traditions which have brought us the idea, from the ancient Greeks onwards, the trend is puzzling. For we can discern in those traditions what I call a recurring thought pattern about its conceptual core. The pattern is not exhibited perfectly across every conception in the Western tradition; for example, it is less common in some of those developed from the mid-twentieth century onwards. However, it does capture a line of thought present in most conceptions developed before, and some during, the twentieth century. The pattern is intended as precisely that, a kind of schema, through which to understand the considerable historical variations.
The most abstract discernible pattern among claims about the rule of law is that, whatever else it is, it avoids or prevents something from happening. Aristotle thought the rule of law keeps the "wild beast" in human nature from wielding political power (2017, 79). Locke argued the rule of law requires legislators to not make law arbitrarily and that "known authorized judges" are to decide cases "by promulgated, standing laws" or else a community will end up in a situation even worse than lawlessness (2003,160). And Mill, echoing Magna Carta, wrote that "the aim…of patriots was to set limits to the power which the ruler should be suffered to exercise over the community; and this limitation was what they meant by liberty" (1997,4). In each case, the rule of law circumvents a certain sort of catastrophe.
But even from this small set of examples, a more complex pattern emerges. These quite abstract formulations obscure how the rule of law mediates certain social relations within a political community. The relations have variously been characterized as between ruler and ruled, the powerful and the powerless, those vested with legal powers and those who are vulnerable to their abuse, among others. An enriched pattern references an essential relation between the rule of law and the two groups, 10 It may be wondered whether the international or domestic rule of law is under discussion. This raises important questions about their relations, which cannot be adequately addressed here. But as indicated (above, note 1), a clarifying philosophical account of the maritime rule of law is likely impeded by prioritizing such distinctions. 11 Cf. Waldron 2002, which was as much a cause as an effect of this recent trend. It is worth emphasizing that the notion of an essentially contest concept, as first introduced by W.B. Gallie (1956), is not at all a sceptical one, though many writers have taken it as such.
such that the law mediates relations between them by protecting one group from the other. How the rule of law protects, and what it protects people from, continues to be intensely debated. But so as not to prejudice that issue, we may say it helps avoid a distinctive sort of harm or undesirable state of affairs brought about by the dominant or powerful group, at the expense of the vulnerable one. Thus, as a rough approximation, the recurring thought pattern about the rule of law is that when it prevails in a political community, a. that community has law with certain characteristics that is in force b. those characteristics enable it (the law) to protect members of that community c. from them suffering a distinctive sort or harm or undesirable state of affairs.
I briefly elaborate on each aspect of this pattern.

(a) The community has law with certain characteristics that is in force
There are two initial points to emphasize here. The first is that the rule of law requires law in a community may seem a trivial truth, hardly worth mentioning. 12 On its face, the idea is descriptive, regarding the necessity of law in a community for the ideal even to be possible. But it has sometimes been taken as implying a moral judgement that there ought to be law in a community: that law is a superior means of resolving disputes and maintaining social order, over any alternatives (Marmor 2003, 2-5). This may smack of the old imperialist assumption that distinctively legal forms and institutions are marks of "civilization", as is the case in some British conceptions of the rule of law. The case for these connections between theoretical and political assumptions has been convincingly made (Lino 2018). But even in the postcolonial world, the moral claim is treated as a platitude; this may explain why explicit arguments in support of it are indeed few and far between. 13 And yet an argument is needed, if only to answer now familiar objections that the law itself is an instrument of oppression (Pashukanis 1978), or that, in any case, disputes can be reliably and equitably resolved without traditional legal institutions (Ellickson 1991).
The second point is in regard to the last clause: that law must be "in force." This is also not the tautology it appears; rather, it refers to the need for efficacious law. It is perhaps part of the everyday concept of law that where it is proper to even speak of valid law, there must be law that is "in force." Some writers take that to mean that there are law-applying institutions able to issue credible threats of coercion to compel compliance with legal requirements, 14 while others go further and require that there be widespread obedience to the law among members of a political community. 15 The efficacy of law from a rule-of-law perspective is said to involve recognition and use of the law and its institutions, perhaps also accompanied by some belief in their legitimacy. These latter theorists very often do not have a nuanced view of what obeying the law involves, so the claim is difficult to assess (Sevel 2018). An efficacious legal system is nonetheless a common background assumption of accounts of the rule of law.
With regard to the characteristics law must have to exhibit the rule of law, much of the theoretical debates over the centuries has been about which features enable it to discharge its protective function. And it is primarily the jarring diversity of such views that encourages talk of the rule of law as "essentially contested." While lists of the relevant features have varied widely in both length and content, there are several recurring themes. For example, many since Aristotle have held that the ideal requires that laws are sufficiently general, with respect to regulating both kinds of conduct (rather than singular acts) and groups of people (rather than individuals). The overriding importance of generality has been called into question, given that certain sorts of law that fails to be general in these two respects are well-entrenched across otherwise diverse legal systems, e.g., executive and judicial orders, acts of attainder, and some standard forms of common law adjudication (Raz 2009;Endicott 2013). However, the concerns of those who emphasize the generality of law often lie elsewhere: in the supposed connection between general law and there being equality under the law. A common assumption among those who include generality as a perennial feature of the rule of law is that general rules themselves bring about, or in any case facilitate, their consistent application across cases, thereby discouraging ad hoc applications, or the disregarding of general rules altogether. Some writers have even identified this 'legal equality' with the abstract principle 'treat like cases alike' (Hart 2012, 157-167 andRawls 1999, 206-213); others have doubted the validity of that principle while still recognizing the importance of generality (Lyons 1973).
Another common theme is that law must be knowable by those to whom it applies: that it is sufficiently 'public', 'announced', 'promulgated', 'declared', and so on. The idea has sometimes been construed as an imperative on lawmakers to make and declare law such that the rights, duties, or permissions contained in it may be ascertained with reasonable effort by anyone of normal capacities, and especially by the people who are the bearers of those rights, duties, and permissions (Hobbes 2003, 183-200). The requirement is sometimes said to extend downstream from legislative activity: that there must be sufficient distribution of legal information among the community whose law it is, that is, that the law must be actually known. However, some theorists have understood the demands of promulgation more modestly; for example, if a law coincides with a moral or social norm that is already widely known, its successful promulgation as law does little to further the rule of law (Fuller 1965, 49-51).
Beyond these common themes, there is great variety among supposed rule-of-law features. A trend over the last quarter century has been to rely on an alleged distinction between 'formal' or 'thin' features that "focus on the proper sources and form of legality" and 'substantive' or 'thick' features that "include requirements about the content of the law (usually that it must comport with justice or moral principle)" (Tamanaha 2004, 92). However, as in many domains, form and substance here are often hard to keep separate; generality and promulgation, to take the examples just mentioned, both relate to the form and content of law in different ways, and are often included in rule of law features for moral reasons, e.g., respect for and fair notice to legal subjects. The utility of the distinction, both in theory and practice, therefore remains an open question.

(b) The system protects members of that community
Whatever the list of characteristics which law must have to achieve the rule of law, theorists agree that a legal system that has them can then perform a distinctive protective function. The diversity of views here is regarding how, for whom, and by whose actions, the rule of law discharges its function. We can discern two general approaches. One is a 'top-down', institutional approach in which officials with law-applying and enforcement powers (including judges, police, and administrators) exercise them to protect the rights and freedoms of those vulnerable to the abuse of those powers. This is the common approach of common law-oriented conceptions of the rule of law, as they often entrust judges with the responsibility of protecting members of the community both from each other and from legal officials who may exercise power over them to do harm. 16 A second, 'bottom-up' approach is that the rule of law empowers members of a community to protect themselves by using such institutions to protect their rights and freedoms. In the modern world of rule of law promotion and international development policy, this 'bottom-up' approach prioritizes 'access to justice' in a community, an umbrella term which refers not only to access to efficient and competent courts, but also adjacent matters of, e.g., adequate judicial salaries and keeping court costs such as filing fees and the hiring of competent legal representation reasonably low. 17 These two approaches are clearly compatible; indeed, they may be complementary within the same theory, and the same community. Judges are well-situated to ensure that citizens are empowered both to argue their rights in court and to challenge court decisions if necessary. And judges can of course exercise review powers over administrative and executive action, as well as judicial actions in lower courts. And yet different variations on the pattern make one or the other approach more salient in achieving the rule of law.
On the question of for whom the protective function is discharged, I have used the terms 'citizens' and 'members of a community' interchangeably, though they are clearly not equivalent. There has been, in fact, precious little discussion about whom the rule of law is meant to protect: whether citizens (wherever they may be, inside or outside a territory), all persons within a given territory at a given time, or a broader class which may include persons who live under other legal systems-for example, those living in immediately adjacent states, or persons from near or far seeking refuge or political asylum within, or from outside, a state. Who is considered to be within the 'community' for purposes of explaining how the rule of law protects its members will have consequences for how, and how well, it discharges that function, as well as for what accountability relations there are that determine who has moral standing to protest when it does not. The dramatic increase in recent decades of refugees, displaced and stateless persons, and other migrants attempting to move across international borders, and often undertaking perilous journeys at sea to do so, has made the question of who is meant to be protected by achieving the rule of law in a given state an increasingly urgent moral and political issue. The general issue requires independent discussion, though I briefly return to it below.
(c) From suffering a distinctive harm or undesirable state of affairs.
When the rule of law performs its protective function, what it protects members of a community from, according to the pattern, is presumed to be something morally and politically undesirable. It first appears there is broad scholarly consensus about what the undesirable thing is. Writers commonly refer to the arbitrary exercise of power as the problem for which the rule of law is a, indeed the, solution. 18 But the apparent consensus falls away once it is asked what is meant by 'arbitrariness'. For example, does such abuse of power involve only the exercise of power created or recognized by law (for example, a legal power to legislate, or the use of executive power)? 19 Or is it also about the abuse of political power, whether or not that power is regulated or even countenanced by the law? Are the standards by which to determine whether the exercise of a power is arbitrary provided by the law or by an independent political morality?
On these questions (and others) there has been remarkably little systematic reflection, and only very recently have two distinct approaches emerged, one narrow and the other broad. 20 The narrow approach holds that only a valid power, recognized or created by law, can be exercised arbitrarily. To act arbitrarily is to act contrary to or with indifference towards the reasons which justify its existence (Endicott 2014). Sometimes these reasons are explicit, for example, in judicial interpretations of the scope of the power based in statutory or constitutional texts that establishes it. But 18 Cf Aristotle 2017, Locke 2003, Hayek 1960, Raz 2009, Fuller 1965, Krygier 2016, and Gowder 2016 If answered in the affirmative, it is an open question whether arbitrariness, at least the sort most relevant to the rule of law, concerns only the exercise of public powers rather than also private ones (for example, the making of contracts). See Austin and Klimchuk 2014, which begins to challenge the "public law presumption" in influential accounts of the rule of law. The rule of law in employment relationships is discussed in Sempill 2017, but more general questions around the rule of law and private powers have yet to be addressed. 20 For a useful critical overview, see Krygier 2016. often the reasons exist only as implicit policy or political morality. Arbitrariness of this sort is possible even if there is no explicit legal norm forbidding it; it is enough that the law has, as it were, provided the opportunity for it by establishing powers which are liable to abuse.
A second, broader approach holds that not only legal powers, but any political power, can be exercised arbitrarily, and that this should be explained from the point of view of those who are vulnerable to it. For example, a power is arbitrarily exercised if it does not allow those vulnerable to it to reliably predict how it will be exercised in advance, thus rendering them unable to plan their lives in light of it (Krygier 2016). An even more capacious variation is to view arbitrariness as exercise of any power which is disrespectful of a person's rights, interests, or expectations, in that the powerholder fails to attribute the proper deliberative weight and significance to them (Sempill 2016, 366-374). These two approaches tend to yield very different treatments of putative examples of abuses of power. And it is relatively early in the scholarly assessment of their respective virtues and vices in relation to the evergrowing set of examples of rule of law failures by governments around the world; that debate will no doubt continue and bear fruit in the coming years.

Territory
Though not always explicit, the many conceptions of the rule of law that exemplify the thought pattern have relied on the assumption that the legal authority of the state (or its analogue) necessarily extends over a territory. The consequent dim prospects for rule of law at sea were noted long ago. Hugo Grotius famously argued in the seventeenth century that, in effect, the rule of law at sea was not possible because the sea, like a wild animal, was not susceptible to continuous occupation and control, which was for him the sort of sovereign control that made legal governance of any sort possible (Grotius 2004, 25-27). And while the history of maritime law in the twentieth century would appear to have proven him wrong-by the early 1970s, one delegate to the Third UN Conference of the Law of the Sea quipped that its purpose was "to bury Grotius" (Oxman 2006, 832)-the extent to which coastal states now actually control and regulate their 'territorial' seas and beyond is very much an open question, and one of great practical and political importance. While there have been a few attempts in recent years at literal continuous occupation of parts of the sea-for example, China's attempts to militarily occupy some parts of the South China Sea, or the United States' vast oil and gas industry in the northern Gulf of Mexico and elsewhere-the idea that the sea is itself part of the territory of a coastal state is at least partly a fiction, if only due to the scarce resources of coastal states needed to achieve the requisite sort of control and occupation over it. In order to meet Grotius' standard, a coastal state would have to continuously police, or at least to surveil, vast expanses of the seas. To be sure, that is now done at some maritime boundaries for various legal and political reasons, but Grotius would require a far more comprehensive effort over a far greater area. 21 We are lead back at this point to Bentham's idea that the territorial assumption is not in regard to the sea itself, but the ships that sail upon it. The practice of ships flying the flag of a territorial state thus extends the territory of the state, and with it the possibility of law, and the rule of law, wherever the ship may go, including to the distant high seas. The suggestion has been taken seriously by both international law scholars and judges over the last two centuries, and has become part of the standard history of the ship registration system, though some courts have regarded it as itself a suspicious legal fiction. It seems to have gone unnoticed that Bentham, in making the suggestion, was inspired by an actual fiction. He says a ship is a little wandering province "like the one of Laputa." The reference is to Swift's Gulliver's Travels, in which Gulliver comes upon the flying island of Laputa, one large enough to support a small society, and which roams the seas as ships do (Swift 2009, 141ff). Bentham thinks that the storybook fiction is not as far-fetched as it once was, and adds that, as he writes, some warships can transport as many citizens as then lived in the French Caribbean settlement of Saint Martin (Bentham 1802, 333).
These and other of Bentham's scattered remarks on maritime matters reflect his appreciation of the unusual fluidity and versatility of the ship as an object of legal governance, depending on purpose and context: that the ship can be regarded as a chattel, a place of residence, a workplace and factory, the site of recreation, or of warfare, among many others, and that many of these roles can be played by a ship simultaneously, and often while it moves across many jurisdictional boundaries. Under these circumstances, regarding the ship as the extension of a state's territory would bring some order to this chaos. And yet, Bentham's examples of Laputa and Saint Martin emphasize, not the territoriality of ships, but the fact that they can be the situs of a form of human community. The territorial assumption is an imperfect proxy explanation for this. But we can acknowledge the point about community without accepting the dubious territorial metaphor.
For it is by now clear in international practice that maritime law, in all its forms, composes a non-territorial legal regime. This is the sense in which the modern Law of the Sea "buried Grotius": in rejecting the assumption that law and successful governance requires continuous occupation of a territory. It remains the case that ships and their zones of navigation present difficulties as objects of direct state regulation, as compared to, for example, roads, cars, and their drivers within land territories. For example, both domestic and transboundary overfishing continues to be a threat to biodiversity worldwide. 22 This is true of other objects in other domains of law historically inspired by maritime governance, such as aviation and space law, regarding internationally transient planes and spacecraft. And yet in each case a non-territorial conception of the rule of law is possible and, indeed, necessary. And there is little difficulty in the idea, once a territorial conception of law is given up; theorists have begun to develop a variety of non-territorial concepts of law, within a diversity of conceptual frameworks. 23 And if the rule of law is at its core about law having such characteristics that allow it to discharge a protective function across all members of a community, as we observed in the thought pattern, then dispensing with the territorial assumption does not prima facie present any intractable difficulties in extending the rule of law to the sea. But the rejection of territoriality leads our attention to two further aspects of the rule of law, community and protection, that take different forms at sea.

Community
Does the rule of law as such require political community? The notions of community and territoriality have often been run together as standard conditions for the possibility of law. And the recent revival of the rule of law in the broadly liberal tradition has brought with it the idea that the rule of law assumes a political community whose identity is bound up with a shared set of political ideals and principles, and perhaps a range of other social and cultural norms (Kymlicka 1986). And from a more practical, law-and-development perspective, it has long been taken for granted that the rule of law does require a sufficient orientation of the "hearts and minds" of a community around the law and its institutions (Faundez 2011), for example, a widely shared disposition among members of a community to see themselves and each other as legal persons, with standing to engage legal institutions for the exercise of public or private powers, or to protect themselves from the exercise of them by others, what Dicey long ago called having a "legal turn of mind" (Dicey 1960, 187). It is less clear, however, what the bounds of the relevant community is or should be, i.e., whether it is limited to citizens of a state, to its legal residents, or extends to visitors, immigrants, refugees, and asylum seekers. As a result, and as we noted in part (b) of our thought pattern, it remains unclear who is the intended beneficiaries of the rule of law. But this debate rests on the more basic assumption that when the rule of law prevails, it does so over a determinate and distinct political community.
This may suggest an insurmountable obstacle for achieving the rule of law at sea; while the oceans are continuously, albeit sparsely, populated with human beings, it may nonetheless seem utterly bereft of human community. Much has been written, particularly since the nineteenth century, about the unavoidable solitude of being at sea. Melville wrote of the "wild watery loneliness" of seafarers, and portrayed them as each living on a continent of their own while at sea. 24 A generation earlier, this image of the life of mariners was incorporated into the law across the common law world, and has since become part of the rationale for how admiralty courts have afforded protections to seafarers. For example, United States Supreme Court Justice Joseph Story famously characterized seafarers as "generally poor and friendless," and for that reason the courts should treat them as their "wards." 25 As we have seen, Bentham thought there are such things as maritime communities, but identified them narrowly, only as the determinate, but transient, groups of people on-board particular ships at a given time: masters, crews, and passengers en route from port to port. But even this view overlooks the persistent and vast community dispersed around the world, and indeed over the centuries, who navigate the seas for commercial and recreational purposes, that is, the seafaring community. It is composed of distinct but overlapping smaller communities of those in e.g. maritime shipping, commercial and recreational fishing, scientific research, offshore seabed and subsoil resource exploitation, and many others. What unifies the seafaring community are the common dangers its members face in navigation. These include both the perennial 'perils of the sea' (drowning, shipwreck, and dangerous marine species, among others) and the various harms they are liable to suffer at the hands of their fellow seafarers. The seafaring community, then, may seem of the right sort to underwrite and enable the maritime rule of law. 26 And yet, it may be objected that this is still an implausibly narrow conception of the maritime rule-of-law community because it rests on a conceptual confusion about the character of the required community. 27 The confusion rests on the idea that the most important community supporting the rule of law is the one whose protection is ensured when the rule of law prevails. But this is mistaken, so the objection goes: the community necessary for realizing the rule of law is, rather, the one which 24 Melville described the crew of the Pequod as "nearly all Islanders…Isolatoes too, I call such, not acknowledging the common continent of men, but each Isolato living on a separate continent of his own" (2008, 106). 25 "Seamen are by the peculiarity of their lives liable to sudden sickness from change of climate, exposure to perils, and exhausting labour. They are generally poor and friendless sand acquire habits of gross indulgence, carelessness and improvidence…Every court should watch with jealousy an encroachment upon the rights of seamen, because they are unprotected and need counsel; because they are thoughtless and require indulgence; because they are credulous and complying; and are easily overreached." Harden v. Gordon (C.C. Me. 1823) 11 Fed. Cas. 480, No. 6,047 (US). 26 I acknowledge, only to set aside, the objection that even this broad suggestion is inadequate, on grounds that it is speciesist, since it includes only human beings, and excludes all non-human marine species. But all of these species, so the objection goes, merit and require rule-of-law protections. This argument raises several important issues, such as the nature of legal and moral personhood, which cannot be adequately dealt with here. However, the anthropocentric conception of community appealed to here, for better or worse, is continuous with those used in the philosophical traditions of the rule of law which are the primary focus here. But as I explain below, assuming the bounds of the relevant community to track those of humanity is compatible with the fact that achieving the rule of law in that community has collateral, and perhaps even necessary, salutary effects on other human and non-human populations. 27 I thank an anonymous reviewer for pressing this objection.
constitutes the law, that is, the law-making community. The ruling, rather than the protected, community is where the ideal is properly cultivated. These two communities may overlap, but often are not identical. For example, a democratic law-making community may legislate to protect a powerless minority within a society, such that there is virtually no overlap in membership across the two groups. The maritime law-making and seafaring communities are similarly distinct. A great variety of legal institutions, across many jurisdictions, contribute to the legal regimes which govern seaward, including national and local state-based legislative institutions, as well as the work of states internationally, including the many maritime conventions and the creation and use of international courts such as the International Tribunal of the Law of the Sea. The objection under consideration suggests that the community supporting the rule of law at sea is the collection of transnational and international communities that make law for the protection of the seafaring community and marine environment. The stronger the rule of law is within this diverse set of legal and political institutions, the stronger the rule of law will be at sea, and the resultant protective function will thereby be enhanced.
The suggestion that the rule-of-law community is simply the law-making community fails, however, to account for the immanence of the rule of law, an extension of the aspect in our thought pattern of what it is for law to be in force within a political community. The rule of law is immanent in the sense that it.
appears as a sort of social glue, a connective tissue holding society together… [A] generalised obedience to the law combines with a pervasive legalism in both public and private spheres. The law itself functions in the background, largely internalized and functionally independent of the coercive and administrative power that guarantees its efficacy…It is already present in the everyday interaction of law-abiding citizens. Its reappearance in official and legal processes is the concrete expression of profoundly held principles and habits of thought-but not their source (Humphreys 2010: 29).
The reference to 'citizens' here is evidence of the terrene tendencies in thought about the rule of law already noted, which the maritime rule of law challenges. The core point is nonetheless clear: it is the seafaring community which exemplifies and concretizes the rule of law at sea, regardless of the nature and source of the law that governs the interaction of its members. This aspect of the maritime rule of law should profoundly shape how we view interactions at sea that have recently caused great injustices. For example, Frontex, the border guard agency of the European Union, has been accused of serious human rights violations of migrants and asylum seekers on the Mediterranean sea in recent years (Marinai 2016;Pallister-Wilkins 2015). An implication of the view of rule-of-law community I have suggested is that while Frontex agents and officials may be part of different land-based political communities than the migrants or asylum seekers, they all are equal members of the rule-of-law seafaring community. And when some members of that community, empowered as agents of the European Union, commit patent injustices against more vulnerable seafarers, the "social glue" of the rule of law at sea thereby begins to erode. It can of course be true that any unjust and illegal actions of Frontex agents at sea weakens the rule of law in traditional land-based legal systems (e.g., in Italy or the European Union itself), but I have argued that a further rule-of-law failure is distinguishable, one suffered by the seafaring community, as the primary locus and agent of the maritime rule of law.
While the seafaring community is the natural and direct beneficiary of the protective function discharged by the rule of law at sea, there are many other, indirect beneficiary communities which extends landward. One is the maritime community more broadly construed, to include land-based workers who in some way facilitate ships in navigation, for example, marine insurers, ship builders, ship owners, stevedores, and so on. The rule of law also indirectly protects traditional land-based political communities, in facilitating solutions to marine problems that have direct effect on those communities, e.g., transboundary marine environmental problems like vessel-sourced oil pollution, sustainable fish stocks, and of course sea-level rise and climate change. In that respect, the value of the rule of law at sea may match or exceed that of any state, in potentially benefitting everyone in both coastal and landlocked communities, as well as marine ecosystems worldwide.

Protection
If the seafaring community is the direct beneficiary of the rule of law at sea, how does it achieve its classic protective function within that community? Since, as we noted, there are no onsite legal institutions to discharge, on an ongoing basis, that protective function, it would seem to require what I called a 'bottom-up' vigilance among seafarers worldwide, to build and sustain the rule of law themselves. This is of course greatly facilitated by the work of traditional terrene legal institutions in coastal states and internationally, but ultimately must be realized by seafarers who regard themselves as both subject to and empowered by law to enforce their rights, and in some cases to protect the rights of others, or other declared objects of legal protection, such as the marine environment or particular marine species. The flagging states play a key role in regulating as far as possible ships while in port, but it is seafarers who must display a "legal turn of mind" while at sea for all to be adequately protected. 28 What does the rule of law at sea protect its many beneficiaries from? As noted in our thought pattern, the classic catastrophe avoided is the arbitrary exercise of power, an already difficult notion and perhaps more so in a maritime setting. There are at least two sorts of arbitrariness at sea. One does not center on the arbitrariness of legal officials as such, but the power that may be exercised by seafarers at sea against each other, among masters and crews of ships and their passengers. Friedrich Hayek thought that the direct result of achieving the rule of law is a "sphere of unimpeded activity" which protects individuals from the coercive forces of the arbitrary whim and will of others. The value of this sphere is particularly salient aboard a ship, where the finite limits, and often remote location, of one's physical environment create risks of such coercion occurring, both in the form of physical coercion among masters and crew of ships, 29 as well as the unchecked degradation by seafarers of the marine environment, e.g., by discharge of oil and other toxins, dumping of waste, and abandonment of fishing nets. The other is arbitrariness done by states themselves, by means of their ships and other floating structures, both in the territorial sea and beyond. This may include mismanagement of fish stocks, destruction of habitat in exploring the seabed and subsoil for minerals and other natural resources, and many other harms to the marine environment. It may also include a failure to enforce rules and regulations with regard to customs, human and drug trafficking, and other crimes at sea. Again, the human rights violations of asylum seekers at sea by Frontex is illustrative. By Frontex agents physically abusing or coercing migrants and asylum seekers at sea, as has been alleged at the Greece-Turkey maritime border (Fallon 2022), they exhibit the first sort of arbitrariness. In engaging in such abuses as agents of the member states of the European Union, they exhibit the second. Each of these sorts of arbitrariness are minimized when the rule of law at sea is achieved, to the benefit of all seafarers. And while the particular means of achieving it in each case will vary, the result will be protection against these harms for members of those distinct but overlapping maritime communities.

Conclusion
The truth in Bentham's 'floating territory' theory is that a ship is the primary means by which the rule of law extends to the sea, and the tethering of the ship to a state by registration extends the reach of the law of the registering state to the ship and those aboard. The flagging system is an imperfect proxy for land-based legal institutional structures, but nonetheless serves a similar purpose. Indeed, maritime law scholars characterize this system as grounding the rule of law to function for precisely the same reason that Hobbes recognized in the seventeenth century in regard to terrene law: to prevent anarchy and disorder, which would be to the detriment of everyone (Coles and Watt, 10). I have argued against the view that the 'maritime rule of law' is an empty slogan, or a misleading characterization of what is possible. A persistent but mistaken view is that the rule of law at sea is possible only in the attenuated sense of states and their institutions performing their usual functions over maritime matters as they arise for review and adjudication. There is therefore no rule of law at sea. Seafarers, their employers, and many others who find themselves at sea take refuge in legal institutions only if they avail themselves of their protections while in port. And legal solutions to marine environmental problems must be land-bound in the same way.
I have rejected this land-based interpretation by appealing to the rough idea of a seafaring community, its members bound by their shared life at sea, and the protections it can afford to them in a 'bottom-up' manner, by practicing the rules applied and created not only by land-based legal institutions, but by its own members over the centuries, for example, in adhering to those ancient rules often called the 'general maritime law.' The ways in which the rule of law at sea protects the powerless will vary, relative to its effects on its direct beneficiary of the seafaring community, or its many indirect beneficiaries, which, often enough, includes all of us. While the maritime rule of law adheres to the classical 'thought pattern' for the most part, some of its background assumptions must be either rejected or adapted to maritime circumstances. The extent to which maritime communities will be successful in maintaining the rule of law will depend on many factors, including the diligence of both state governments and the many international legal and non-governmental actors, who work to entrench rule-of-law relevant legal and cultural norms worldwide. I have tried nonetheless to characterize, in outline, the sort of thing they should aim to bring about.
Funding Open Access funding enabled and organized by CAUL and its Member Institutions. No funding was received to assist with the preparation of this article.

Conflict of interest
The author has no relevant financial or non-financial interests to disclose.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.