This article examines Israel's enforcement of a maritime blockade against the Gaza Strip implemented in the course of an ‘armed conflict’ with Hamas. The first question is the legal characterisation of this conflict and whether it is one to which the laws of naval warfare apply. The conclusion of this article is that, irrespective of the status of the Gaza Strip as an occupied territory, at the relevant time Israel was at best involved in a non-international armed conflict (NIAC) with Hamas. There is only limited support for the proposition that blockade is available in NIACs, and then only in conflicts reaching a high level of intensity. On this basis, Israel had no applicable right of blockade.
In the alternative, the article considers the requirements of lawful blockade and concludes they were not met in the present case. The central issue is proportionality. The maritime blockade was part of a comprehensive closure regime that had disproportionate effects on the civilian population of Gaza. A maritime blockade in support of other measures causing disproportionate damage must itself be disproportionate. In the further alternative, the article assesses whether Israel could have justified its actions on the basis of other belligerent rights.
Finally, the article considers the law governing the use of force during maritime interdiction operations under the laws of naval warfare. It concludes that a ‘policing’ paradigm of force is applicable. The law of individual self-defence and war crimes is also considered.
I. I ntroduction
A. The Mavi Marmara incident
At approximately 4.30 am on 31 May 2010, the Israeli Defence Force boarded a flotilla of six vessels on the high seas bound for the Gaza Strip including the largest ship the Mavi Marmara . 1 The Mavi Marmara was flagged to Comoros 2 and carried activists from the Free Gaza Movement and the Turkish charity Foundation for Human Rights and Freedoms and Humanitarian Relief (IHH) as well as a cargo of humanitarian supplies. The stated aim of the flotilla was both to deliver aid to Gaza and to breach Israel's naval blockade of Gaza. 3 Israel published details of its naval blockade in January 2009, 4 and has justified the blockade as a measure permitted by the laws of armed conflict (LOAC) and necessary to prevent weapons reaching Hamas. 5 Hamas is the political/armed group that is presently the elected government in the Gaza Strip, which claims responsibility for rocket attacks on Israel and which does not recognise Israel's right to exist. Israel made it clear through ‘public and diplomatic channels’ that unless the vessels diverted to the port of Ashdod and submitted their cargoes to inspection and distribution through Israeli channels they would be boarded. 6
The boarding of the Mavi Marmara involved from the outset the use of stun grenades and subsequently live fire. The other five flotilla vessels were intercepted without loss of life; aboard the Mavi Marmara clashes between Israel Defence Forces (IDF) soldiers and activists resulted in nine civilian deaths. The principal difference between the boarding of the Mavi Marmara and other flotilla vessels is that the others were boarded by boat. When two efforts at boarding the Mavi Marmara by boat failed due to resistance from those aboard, a helicopter was used. 7 While the precise facts are contested, it would appear that the IDF inserted soldiers one at a time by a single fast rope from helicopter onto the Mavi Marmara upper deck where they met resistance from approximately 15 protestors and encountered an increasingly volatile situation. 8 Military experts have suggested that the aim of such a boarding is to place a sufficient number of troops on deck sufficiently quickly to assert and maintain immediate control. 9 Several of the first few soldiers inserted were captured by protestors and appear to have been injured. Within minutes two groups of soldiers had been deployed by helicopter and assembled on this ship's roof before descending to take control of the vessel. 10 Subsequent live fire used by the IDF killed nine civilians and left many others wounded. Both sides have publicly claimed they acted in self-defence after first being attacked by the other. 11 Despite some initial claims to the contrary, there is no evidence that IDF soldiers in the boarding party were armed with anything other than modified paint-ball guns and other riot control weapons and personal protection sidearms in the form of 9 mm pistols (the calibre of the latter is significant in that the lower exit velocity of such bullets makes them less lethal). 12 IDF claims that they were met immediately with active and violent resistance on boarding the Mavi Marmara have eyewitness support 13 and the IDF has released video recordings showing protestors dragging on the helicopter fast rope and beating the first soldiers inserted.
The Turkish autopsy reports on the nine protestors killed concluded ‘five of the victims were shot either in the back of the head or in the back’. 14 These reports and other medical records were examined by an independent pathology expert retained by the UN Human Rights Council inquiry into the incident (discussed below), who concluded in respect of several of the killings that persons were fatally shot in the head at point-blank range or from above and at a distance; in some cases the injuries are consistent with the deceased having already been shot once and then being shot again in the head once already prone on the deck. 15 The IDF has made an internal inquiry into the incident and the individual circumstances of each fatality but has not made the results of that inquiry public: instead it has suggested that such force would only be used in self-defence where an attacking protestor was ‘bending over a soldier with an ax[e] or with an iron bar’ and alleged that all those killed were violent activists who were willing to die as martyrs. 16 Issues of self-defence will be explored further below.
The Mavi Marmara , once captured, was diverted to Ashdod where those aboard were detained. Detainees, including journalists, variously report having been kept incommunicado and in some cases denied consular access, being photographed by Israeli soldiers, being beaten, being denied adequate food or medical attention and being placed under pressure to sign untranslated documents in Hebrew. 17 It appears that all computing, communications and photography equipment aboard the flotilla was confiscated by Israel and that none has been returned. 18 Most personal possessions also appear not to have been returned.
The civilian deaths sparked an angry response with Turkish official statements describing the interception as a ‘gross violation of international law’. 19 The Turkish foreign minister Ahmet Davutoğlu called the episode ‘tantamount to banditry and piracy’ and ‘murder conducted by a state’. 20 The Security Council also issued a presidential statement ‘condemn[ing] those acts which resulted in the loss’ of civilian life and calling for a ‘prompt, impartial, credible and transparent investigation’ into the incident. 21 The UN Human Rights Council also voted ‘to dispatch an independent international fact-finding mission [FFM] to investigate violations of international law, including international humanitarian law and human rights law, resulting from the Israeli attacks on the flotilla of ships carrying humanitarian assistance’. 22 The FFM itself noted that this language could be taken to suggest the issue of ‘violations’ and ‘attacks’ had been pre-judged and interpreted its task to be investigating the facts of the incident and whether violations of applicable law occurred. 23 As noted above there has already been an Israeli internal military inquiry into the incident, which found that there were failures of intelligence and planning but no grounds for disciplinary action or prosecution. 24 An Israeli public inquiry including two international observers (commonly called the Turkel Commission, after its chair) has reported in part. Its mandate is to evaluate whether the maritime closure of the Gaza Strip and the measures taken to enforce it on 31 May 2010 were ‘compatible with international law’ and to assess ‘the actions undertaken by the organizers of the flotilla and those who participated in it and their identities’. 25 The first part of the Turkel Commission Report was released on 23 January 2011 and its approach to the law of blockade is addressed principally in a post-script. In August 2010, Turkey announced it would conduct an internal inquiry: 26 this was released on 11 February 2011 and is also briefly addressed in the postscript. Finally, there will be an independent review of the two national inquiries by a high-level committee formed under the auspices of the UN Secretary-General. 27 So far this committee has met twice formally to adopt a set of working methods. 28 A key difficulty in assessing any inquiry's conclusions will be that no inquiry will have had extensive first-hand access to both IDF and protestor witnesses. Nonetheless, the episode, on any assessment of the contested facts, raises a significant number of international law issues under both LOAC and the law of the sea.
B. Outline of legal issues
As a preliminary matter, a vessel on the high seas is ordinarily subject to the exclusive jurisdiction of its state of nationality (the flag state) unless an exception applies. 29 Generally, such exceptions are codified in the High Seas Convention and the UN Convention on the Law of the Sea and are limited to cases involving suspicion of piracy, the slave trade, unauthorised broadcasting and vessels though displaying other markings suspected of having the same nationality as the interdicting warship. 30 Other possible justifications for interfering with a vessel's freedom of navigation include statelessness, specific treaty arrangements, or the ad hoc consent of the flag state. Israel's action fits within none of these categories. The question arises whether LOAC may also provide justification for taking measures against foreign vessels at sea. 31 It is sometimes argued that UNCLOS Articles 88, 141 and/or 301 have demilitarised the high seas, thus effectively abolishing the prior law applicable to naval operations during armed conflicts. These articles reserve ocean areas beyond state jurisdiction for ‘peaceful purposes’ and re-state the Article 2(4) UN Charter prohibition on the use of force against ‘the territorial integrity or political independence of any State’. However, it is now generally accepted that UNCLOS does not ban any uses of the ocean consistent with:
self-defence under Article 51 of the UN Charter (for example, intercepting a vessel designed to act as a floating bomb);
the rules of LOAC (which may be invoked by belligerent states irrespective of whether they are later found to have been acting in Article 51 self-defence); or
operations authorized by a Security Council Resolution. 32
At most the relevant UNCLOS provisions can be said to outlaw the use of the sea for acts of aggression 33 but this is only to say that such acts, obviously illegal in themselves, will also violate UNCLOS. While there was from time to time some suggestion that the UN Charter had effectively abolished the laws of naval warfare, this position was never accepted by major naval powers during UNCLOS negotiations and military manuals continue to include chapters on naval warfare. 34
There remains, however, some doubt in the general public international law literature as to the scope of application of the laws of war under the UN Charter. As Churchill and Lowe put it, ‘[o]ne view … is that force may be lawfully used only with the authorisation of the United Nations Security Council or alternatively in the exercise of the inherent right of self defence preserved by Article 51 of the … Charter.’ 35 On this approach, each individual use of force must be justified under one head or the other. The alternative view is: ‘regardless of the legality of the use of force in international relations, when force is used on a large scale it is regulated by the Laws of War and the law of neutrality.’ 36 The latter view is preferred by almost all international humanitarian law scholars, as well as in military manuals and the case-law of international criminal tribunals; indeed, the majority of these authorities would find all of LOAC applicable the moment there is any recourse to force between states. 37 Churchill and Lowe seem to prefer the first view as
under the traditional laws of war and neutrality belligerents would have a right systematically to stop and search all neutral vessels and to seize contraband, whereas under Article 51 combatants would have only the right to stop and search particular ships suspected on reasonable grounds of taking arms to the other side for use in the conflict. 38
The other possibility, of course, is that the law of the UN Charter, or subsequent state practice, has modified the traditional rules. The San Remo Manual (discussed below) generally posits not absolute rights, but rights that can be exercised only on reasonable suspicion. Indeed, the UK military manual goes further and imports UN Charter concepts of necessity and proportionality into the strategic choice of methods and means of warfare in contexts where the traditional LOAC conception of ‘proportionality’ would allow a belligerent considerably more latitude. 39 Some controversy remains over the correct approach on such questions, but the continued applicability of LOAC rules during the UN Charter era is not generally doubted.
A legal naval blockade may be enforced against the vessels of flag states not parties to the conflict (neutrals). There are some uncertainties, however, in the applicability of the law of blockade to Israel's conflict with Hamas and whether the blockade was legally established and implemented. There may be alternative rights of interception in times of armed conflict exercisable against neutral vessels which will also be explored. If, however, there is no legal basis to interdict a vessel on the high seas under LOAC, then we must consider the law applicable to an unlawful boarding under LOAC and perhaps the general international law of the sea.
The present discussion will generally not touch on human rights law or the question of its application during armed conflict.
II. T he A pplicable L aw of A rmed C onflict
A. Characterising the Israel-Hamas conflict
An important question in the debate surrounding the legality of the Mavi Marmara interdiction is the classification of the conflict between Israel and Hamas. The usual assumption made by commentators is that there is a continuous armed conflict in the LOAC sense which must either be classified as an international armed conflict (IAC) or as a non-international armed conflict (NIAC). Israel has acknowledged that it ‘is currently in a state of armed conflict with the Hamas regime that controls Gaza’ but does not expressly classify that conflict as either an IAC or NIAC. 40 For practical purposes, it may often be the case that the precise characterisation of a conflict is not important. The Goldstone Report, in the context of assessing events in the Gaza Strip in December 2008 – January 2009 (‘Operation Cast Lead’), appeared to endorse the view that the legal characterisation of hostilities between Israel and Hamas was not crucial to determining the applicable law, as a fairly wide body of humanitarian law is common to IACs and NIACs. 41 However, in the context of naval blockade the distinction is important. This is because the most widely accepted restatement of the rules of naval warfare, the San Remo Manual, ‘does not expressly deal with non-international armed conflicts’ and is ‘primarily meant to apply to international armed conflicts at sea’. 42 Further, with the possible exception of the 2006 Israel-Lebanon conflict, the only twentieth century conflicts clearly involving recognised blockades were all IACs (state practice in NIACs is explored further below). 43 The content of the rules of naval warfare applicable to NIACs is at best uncertain and the availability of blockade in such cases open to doubt. There is a third possibility also: that while the Gaza Strip may be occupied territory under the Fourth Geneva Convention (GC IV), 44 there was no armed conflict governed by LOAC at the time of the Mavi Marmara interception on 31 May 2010. If there was no armed conflict of any type, the law of blockade could not apply. In the alternative, it may be possible to conclude on the basis of state practice that a right to blockade is acknowledged only in conflicts reaching a certain scale and intensity. It would then be open to question whether the violence in the Israel-Hamas conflict passes that threshold. The logic of each of these positions is explored below.
The starting point remains identifying the existence of an armed conflict. The generally accepted test was stated in Prosecutor v Tadić as:
an armed conflict exists whenever there is a resort to armed force between States or protracted armed violence between governmental authorities and organized armed groups or between such groups within a State. 45
On this test, an IAC occurs whenever there is recourse to any degree of violence between states; a NIAC requires ‘protracted armed violence’—which may be a question of intensity more so than duration 46 —involving armed groups organised along military lines. Prima facie , as Hamas forces do not represent a state this may lead to the conclusion that the conflict should be classified as a NIAC, always provided that the conflict meets the ‘protracted armed violence’ criterion. 47 (The position that the Palestinian Territory is already a state 48 is not widely accepted by commentators. The occupied Palestinian Territory does not satisfy the legal precondition to statehood of independent government. 49 While the International Court of Justice has held that the occupants of the Palestinian Territory have a right to self-determination, that finding is as consistent with a right to future statehood as any other conclusion. 50 In any event, even if there was a single Palestinian state, it is an open question whether Hamas could be considered a state organ capable of engaging international responsibility).
However, three arguments may be put that, despite the basic test, this conflict is an IAC on the basis of some exceptional rule. First, it may be argued that Israel remains the occupying power in Gaza, and as the laws of occupation apply only in an IAC, the conflict with Hamas is an IAC. Secondly, it could be suggested that any conflict crossing an international border is an IAC. Thirdly, one could invoke the concept of ‘recognition of belligerency’, meaning that when a state acknowledges an opponent as having full belligerent status, the laws applicable to an IAC govern their conflict. 51 Such recognition may occur implicitly through declaration of blockade against a non-state actor: a blockade constitutes a serious measure of war 52 and therefore a declaration of blockade is tantamount to recognising that an adversary cannot be overcome by measures short of IAC. Thus one could claim that Israel, by the very act of blockade, has recognised Hamas as a belligerent and the conflict thereafter became an IAC. If any of these arguments is accepted, then it would appear to follow as a matter of principle that blockade is an available means of waging war, provided it is legally implemented and enforced. In fact, none of these arguments is compelling and state practice does not unequivocally support the conclusion that blockade is available in a conflict of limited intensity irrespective of whether it might technically qualify as an IAC.
B. Arguments based on occupation
The first step in the argument based on occupation, that Israel remains in occupation of the Gaza Strip under the terms of GC IV, has widespread support. In 2004 the International Court of Justice in the Wall Advisory Opinion unequivocally found that the entirety of the Palestinian Territory was occupied territory subject to GC IV based both on an assessment of the facts and the attitude of states parties to the Geneva Conventions. 53 The question is whether that assessment remains correct following Israel's 2005 unilateral ‘disengagement plan’ and the associated withdrawal of its troops from the Gaza Strip. On 10 December 2009, General Assembly Resolution 64/92 affirmed the applicability of GC IV to the Occupied Palestinian Territory; on the same day General Assembly Resolution 64/94 directly referred to Gaza as part of the Occupied Palestinian Territory. 54 Both resolutions were passed by an overwhelming majority of member states. While these may be thought of as politically motivated statements made within one of the United Nation's political organs, the attitude of states to such resolutions remains relevant evidence of opinio juris . 55 In the present context, such acts cannot be presumed to have no legal significance in assessing what states believe to be the correct interpretation and application of GC IV. 56 The Security Council practice on point is less instructive. In the period 1969-1994 the Council repeatedly emphasised that Israel was the occupying power in the Palestinian territories. 57 However, since the 2005 ‘disengagement plan’, Security Council resolutions have not directly dealt with the legal status of the Gaza Strip. While Security Council Resolution 1860 of 2009 stressed ‘that the Gaza Strip constitutes an integral part of the territory occupied in 1967’, this is not an unequivocal statement that the Gaza Strip is occupied at present; in debates the next day several Security Council members did, however, expressly refer to Israel's responsibilities in the Gaza Strip as the occupying power. 58
Even if the situation in the Gaza Strip is to be classified as an occupation, it is sui generis : the usual test of occupation is ‘effective control’ over an area. 59 Israel conducts at best limited acts of civil administration within Gaza (an area governed by a hostile political entity) and since 2005 it does not have the troop presence within the territory usually associated with occupation. However, Israel maintains exclusive control of the airspace over Gaza, the maritime approaches and all or most land borders. As the Goldstone Report put it:
the powers that Israel exercises from the borders enable it to determine the conditions of life within the Gaza Strip. Israel controls the border crossings (including to a significant degree the Rafah crossing to Egypt, under the terms of the Agreement on Movement and Access) and decides what and who gets in or out … It also controls the territorial sea adjacent to the Gaza Strip … thereby regulating economic activity in that zone. It also keeps complete control of the airspace … It makes military incursions and from time to time hit targets within the Gaza Strip. No-go areas are declared within the Gaza Strip near the border … and enforced by the Israeli armed forces. Furthermore, Israel regulates the local monetary market based on the Israeli currency (the new sheqel) and controls taxes and custom duties. 60
Israel also continues to control some functions of civil administration (such as the population registry) and has not only shown the ability to conduct incursions into Gaza at will 61 but has expressly reserved the right to do so, as military imperatives require. 62 Case-law supports the proposition that an occupation does not necessarily require a continuous troop presence throughout the relevant territory, so long as the occupier has at least ‘the capacity to send troops within a reasonable time to make the authority of the occupying power felt’. 63 Shany argues that such an assertion of authority in the case of Gaza would be more ‘lengthy and costly’ than the case-law anticipates. 64 On such a basis there is an arguable (if minority) view that the situation between Israel and Gaza is now better described as a siege. 65 Alternatively, the Israeli Supreme Court in Al Bassiouni has held that following the 2005 withdrawal of Israeli troops there is no longer an occupation in Gaza, but that nonetheless ‘certain obligations [to supply Gaza with fuel and electricity] derive from the degree of control exercised by the State of Israel over the border crossings between it and the Gaza Strip’ along with the historic relationship of dependency by Gaza on Israel for electricity. 66 It is hard to reconcile this conclusion with the Supreme Court's premise that there is a continuing armed conflict in effect but not an occupation. Dinstein has observed of this logic that ‘[t]he notion that a Belligerent Party in wartime is in duty bound to supply electricity and fuel to its enemy is plainly absurd’ and that the ‘sole reason’ such an obligation might exist during a conflict is that ‘the occupation is not over’. 67
An assessment based on the present facts alone could perhaps conclude that Gaza is not occupied. However, the overwhelming majority opinion of states 68 appears based on the view that Israel remains the occupying power. This involves at least an implicit assessment that the change in the character and intensity of Israel's control over the Gaza Strip since 2005 is not sufficient to relieve it of all of the duties of an occupying power which had previously been established. 69 Indeed, article 6 of GC IV acknowledges that an occupation can extend beyond ‘the general close of military operations’ and that the occupier retains certain core obligations ‘for the duration of the occupation, to the extent that … [it] exercises the functions of government’ even in the absence of hostilities. ‘Belligerent occupation goes on as long as the Occupying Power continues to exercise effective control in an occupied territory’. 70 Of course, the degree of control exercised and the government functions retained will affect which responsibilities are engaged: Israel cannot be held liable for the acts of the Hamas administration in the Gaza Strip.
There are three possible views of the consequences of a continued occupation. The first, based on older authorities, is that any blockade of occupied territory by the occupier is a nullity. The second is that any level of hostilities that occurs in an occupied territory—even between a state and non-state actor—is governed by the laws of IAC (which include the laws of blockade). A third follows from the proposition that, although an occupation must commence in an IAC, it can continue after ‘the general close of military operations’. 71 On this third view there can be a continuing occupation without an armed conflict occurring in fact or even being presumed in law: any subsequent hostilities within that occupied territory are therefore separate matters to be judged by the Tadić test. The question becomes one of parallel application: either an IAC or NIAC could logically occur in occupied territory irrespective of the underlying application of GC IV. 72 Notably, the Gaza Strip was seized in 1967 in an IAC with Egypt and a peace treaty was concluded between Israel and Egypt in 1979.
The first view, that any blockade of occupied territory by the occupier is inherently unlawful, can be briefly stated. The proposition is that blockade is a means of subduing a territory, and once that territory has been occupied the justification for blockade has passed. Article 1 of the London Declaration of 1909 thus required that a ‘blockade must not extend beyond the ports and coasts belonging to or occupied by the enemy’. 73 The Harvard codification project similarly concluded in 1939, after reviewing relevant case-law: ‘[i]t is doubtful whether a belligerent may institute a regular blockade against ports under his own occupation.’ 74 However, while the London Declaration is widely accepted as stating the customary international law applicable at the time it cannot be taken to be ‘a reliable guide to contemporary law’ a century later. 75 The San Remo Manual and modern military manuals (discussed further below) do not refer to any prohibition on blockading occupied territory; its continued status as part of the laws of naval warfare is therefore open to doubt. Further, as discussed above, article 6 of GC IV acknowledges that an occupier may exercise control in occupied territory to a variable extent. If an occupation can continue despite less than complete control (and impose variable duties on the occupier), then an occupied territory need not be by definition completely subdued. In this context the all-or-nothing rule of the London Declaration may be inappropriate. Nonetheless, if it remains correct to say both that an occupier may not blockade territory under its control and that the Gaza Strip is occupied by Israel, then the Israeli blockade was illegal from the outset and the high-seas interception of the Mavi Marmara was unlawful. Ultimately, given the silence of modern restatements of the law on point, it remains relevant to assess arguments based on the law of IAC and NIAC.
The second view, that all hostilities in an occupied territory must be governed by the laws of IAC, is supported by some authority, but little sustained argument. The only case-law on point is the Israeli Supreme Court judgment in the Targeted Killings case as reaffirmed without further elaboration in A and B . The relevant passage from Targeted Killings is worth quoting in its entirety:
Professor Cassese [has] discussed the international character of an armed conflict between the occupying state in an area subject to belligerent occupation and the terrorists who come from the same area, including the armed conflict between Israel and the terrorist organizations in the area, stating:This law includes the laws of belligerent occupation. However, it is not restricted only to them. This law applies in any case of an armed conflict of international character—in other words, one that crosses the borders of the state—whether or not the place in which the armed conflict occurs is subject to belligerent occupation. 76
An armed conflict which takes place between an Occupying Power and rebel or insurgent groups—whether or not they are terrorist in character—in an occupied territory, amounts to an international armed conflict’ (A. CASSESE, INTERNATIONAL LAW 420 (2nd ed. 2005) … ).
The Israeli Supreme Court relied on this passage without more to state in A and B ‘that an international armed conflict prevails between the State of Israel and the terrorist organizations that operate outside Israel’. 77 The two propositions put forward by the Supreme Court then are: (1) ‘[a]n armed conflict which takes place between an Occupying Power and rebel or insurgent groups … in an occupied territory, amounts to an international armed conflict’; and (2) the law of IAC also ‘applies in any case of an armed conflict of international character—in other words, one that crosses the borders of the state’. It is convenient to deal with the second proposition first. The idea that a conflict which crosses a state border should be considered an IAC might be thought intuitive. However, the fact that a conflict is ‘external’ is not enough to make it ‘international’ as a matter of law. As will be explained more fully below in the analysis offered of the Israel-Lebanon war of 2006, the question is not one of the geography of a conflict but of the identity of the parties to it. There is little in the treaty law of LOAC to suggest that the fact an armed conflict crosses a border results in an IAC. The fundamental definition of an IAC under the four 1949 Geneva Conventions is that it is a conflict involving ‘High Contracting Parties’: states. The Tadić case expands this definition to include as an IAC a conflict involving state-sponsored forces. 78 This, however, remains a test of identity, not geography. This focus on the identity of parties is sound. As Lubell has pointed out, the ‘primary reason for states creating … two separate categories of [armed] conflict has been to avoid giving any recognised status’ to non-state forces and ‘it cannot be the case that by stepping across a border a rebel is suddenly entitled to claim elevation to prisoner of war status and subsequent immunity [from criminal prosecution] for attacks against state forces.’ 79 As the US Supreme Court has put it: ‘[t]he term “conflict not of an international character” is used … in contradistinction to a conflict between nations.’ 80 One should also note that non-state organised armed groups are less likely than state forces to be able to comply fully with the higher standards under the law of IAC, especially those regarding the treatment of prisoners of war: the logic of holding such groups to less onerous NIAC standards unless and until their conflict with state forces crosses a border, at which point higher standards obtain, is questionable. 81 Such a finding would also seem unlikely to promote compliance with the law.
Returning to the Supreme Court's first proposition, it is quoted directly from Cassese's International Law , 82 where it is stated as a proposition derived from principle only. Cassese argues that:
a conflict with a non-state actor in an occupied territory lacks the character of an internal conflict and so should be treated as an IAC;
the object and purpose of international humanitarian law is to provide humanitarian protection on a basis that is ‘ as extensive as possible ’ and this justifies the application of IAC; and
it would be contradictory to hold the occupation is governed by the law of IAC and hostilities within that occupied territory by the law of NIAC. 83
None of these arguments is convincing. On the first, one might as easily say such a conflict lacks the essential character of an IAC—being a conflict between states—and therefore should be classed as a NIAC. Further, common Article 3 of the Geneva Conventions does not refer to ‘internal’ armed conflicts, only ‘conflicts not of an international character’; and as noted above, the international character of a conflict is not a question of geography but of the identity and status of the parties involved. 84 Cassese's second argument is an argument for the abolition of the law of NIACs in toto , thus making the law of IAC applicable to all armed conflicts. While perhaps an admirable goal de lege ferenda , the general view is that we are not there yet. As to the third argument, as will be explained more fully below, there is no inherent contradiction in holding that, within an occupied territory, the laws of NIAC may apply. However, one could couple the thrust of Cassese's third argument with Dinstein's observation that occupation has an inherently international character, as it can commence only in an IAC and is unknown in the law of NIAC. 85 One might therefore argue that the occupation's international character confers a derivative international status on events within occupied territory, even after the close of the original IAC. This conclusion may be intuitive, but it is not supported by clear authority nor is it logically inevitable. It is difficult to accept, therefore, the Israeli Supreme Court's position that cross-border violence between Israel and organised armed groups must be governed by the law of IAC.
As noted, under Article 6 of GC IV the international law of occupation can continue to govern an occupied territory long after the close of an IAC. GC IV deals with the occupying state's rights and duties as an occupier . If protracted armed conflict breaks out in that territory between a state actor (the occupier) and a non-state actor, there is no reason in principle preventing that conflict being governed by the laws of NIAC. 86 There is nothing contradictory about saying a NIAC may break out in a space where certain rights and duties derive from the laws of IAC. It is simply a case of parallel regimes applying in the same territory but in respect of different obligations: the occupying state's duties towards the civilian population may be governed by GC IV; its rights and duties in the conflict with the non-state actor may be governed by the law of NIAC. So long as the obligations owed are owed to different parties, there is no inherent contradiction. By way of analogy, one might consider the following analysis of the Israeli incursion into Lebanon in 2006 during its conflict with Hizbollah. Israel advanced into Lebanon's territory without its consent. Prima facie this is an act of aggression, which could be considered as triggering an IAC irrespective of whether Lebanon resisted. On this basis the laws of IAC are the applicable jus in bello as between Israel and Lebanon, and in particular any civilians in Lebanese territory falling under Israeli control would be subject to the protections of GC IV. One might argue that Israel lacked a belligerent animus towards Lebanon as it was (on its own account) acting only in self-defence. This question of the jus ad bellum is irrelevant: it is well established that the applicable jus in bello is not determined by jus ad bellum concerns. However, Israel was engaged only in active hostilities with a non-state actor: again, prima facie , this conflict with Hizbollah was a NIAC. 87 There is no reason in principle it would not be possible to say that as between Hizbollah and Israel the law of NIAC applied, but that in respect of Israel's rights and duties towards the local population (and any Lebanese forces sent to oppose Israel) the law of IAC would apply. 88 There is no intrinsic problem in this parallel application so long as the different bodies of law apply to different legal relationships.
On such an analysis it would have to be shown under the Tadić test that there was a protracted armed conflict between Israel and relevant non-state actors in the Palestinian Territory before the law of NIAC applied. The Israeli Supreme Court has said that ‘[t]he general, principled starting point is that between Israel and the various terrorist organizations active in Judea, Samaria, and the Gaza Strip … a continuous situation of armed conflict has existed since the first intifada [of 1987-1992]’. 89 It is not a proposition which is easy to sustain. Within the context of an occupation it is generally accepted that there must be some ‘resumption’ or ‘outbreak’ of hostilities before LOAC applies, it does not apply continuously and automatically to every act of violence that occurs. 90 Various periods since 1987 have seen the virtual cessation of Israeli-Palestinian hostilities. In 1999, for example, Israeli-Palestinian violence occasioned only 4 Israeli deaths in total. 91 Subsequent to the conclusion of Operation Cast Lead on 18 January 2009, the present author's review of items posted to the Israeli Foreign Ministry website would appear to indicate that six Israeli citizens (only three of whom were state officials) in total were killed in the period to 31 December 2009. 92 A further three soldiers were wounded during this period. In the period 1 January to 31 May 2010, the time of the flotilla interception, four Israel military officers were killed along with one foreign civilian and a further three police officers were wounded. Figures for Palestinian casualties in this period are difficult to come by, but Israel claims to have killed 13 Palestinian ‘terrorists’ in Gaza in the relevant 16 month period and wounded more. 93 Despite the obvious fact that there is not a numerical threshold of deaths required for an armed conflict to come into existence, 94 a total of 24 deaths and six other confirmed casualties in 16 months does not appear consistent with a finding an armed conflict was underway. 95 However, according to Israel there were 692 rocket and mortar attacks on Israel from Gaza in 2009 and in the period January-May 2010 there were 44 rocket attacks and 2 mortar attacks. 96 On this basis one might argue that the continued attacks on Israel in 2009 met the intensity criterion. The figures for the first five months of 2010, however, provide a much weaker justification for such a conclusion. While clearly criminal, 97 whether such a relatively low level of violence could satisfy the ‘intensity’ criterion for a NIAC under the case-law of the International Tribunal for the former Yugoslavia (ICTY) is open to doubt. In Limaj , for example, an ICTY Trial Chamber found an armed conflict to exist in Kosovo only by the end of May 1998 despite, two months earlier, an attack on a series of villages killing 83 people and police operations causing ‘great devastation to a limited number of buildings’. 98 While rocket attacks from Gaza and other cross-border exchanges of fire have continued, they did not in early 2010 clearly constitute of themselves ‘protracted, large-scale violence’ or ‘substantial clashes’ of organised forces and have not resulted in significant casualties. 99 Acts of violence 100 or terrorism 101 alone do not ordinarily engage LOAC. There is some (albeit limited) support for the proposition that a NIAC must continue to meet the Tadić criterion of intensity and if it drops below this threshold any armed conflict ceases to exist; on this approach the situation in Gaza may be considered an armed conflict only ‘on an irregular and sporadic basis’ when, for example, there are ‘frequent exchanges of gunfire between the Israel Defense Forces and Palestinian gunmen’. 102
Of course acts of terrorism on a sufficient scale may give rise to a right of self-defence as a matter of the law on the use of force under the UN Charter. As Gray notes, UN Security Council Resolutions 1368 and 1373, dealing with the 11 September 2001 terrorist attacks on New York, ‘implicitly recognised the right to use force in self-defence against terrorist action’. 103 The International Court of Justice held in the Wall Advisory Opinion , in response to Israel's claim that the construction of the security barrier in occupied territory was a justified act of self-defence against terrorism, that:
Article 51 of the Charter … recognizes the existence of an inherent right of self-defence in the case of armed attack by one State against another State. However, Israel does not claim that the attacks against it are imputable to a foreign State. The Court also notes that Israel exercises control in the Occupied Palestinian Territory and that … the threat which it regards as justifying the construction of the wall originates within, and not outside, that territory. The situation is thus different from that contemplated by Security Council resolutions 1368 (2001) and 1373 (2001) … Consequently, the Court concludes that Article 51 of the Charter has no relevance in this case. 104
Gray notes that, strictly construed, the first sentence quoted ‘does not say there is only a right of self-defence only in the case of an armed attack by one state against another’. 105 The rest of the paragraph may be taken as ‘leaving open the possibility of self-defence against non-state actors in situations like those contemplated in Security Council Resolutions 1368 and 1373’. 106 It is unlikely, however, that sporadic rocket attacks inflicting few civilian casualties rise to this level, although heavy and sustained rocket attacks could plausibly do so. 107 UN Charter issues relating to the jus ad bellum are, however, not directly relevant to the present discussion. The point to note is that the fact that a state may be entitled to claim they are acting in self-defence against terrorism in certain cases under the UN Charter does not of itself justify the conclusion that the applicable law in those cases is that of an IAC. 108
In summary, there is no convincing argument that the conflict between Israel and Hamas is an IAC. It is, at best, a NIAC. It is further open to question whether the level of violence as at 31 May 2010 could realistically be characterised as an ‘armed conflict’ at all in the strict LOAC sense. The best counter-argument would obviously be that even if the violence involved was sporadic and of a relatively low intensity, as a matter of principle armed conflicts do not end with a lull in hostilities or ceasefire but continue ‘until a general conclusion of peace is reached; or … a peaceful settlement is achieved’. 109 The facts, however, might be thought to place this principle under some strain.
C. Belligerent recognition and other cases of blockade in NIACs
Having reached the conclusion that any armed conflict between Israel and Hamas must be classified as a NIAC, one must then consider whether the law of blockade can apply in conflicts with non-state actors. The US Civil War (1861-1865) is often cited as a (somewhat anachronistic) example on point as: 110
During the American Civil War the state of belligerency was considered to exist [between the United States and the Confederate states] as a result of the presidential proclamation of blockade, accompanied by the British proclamation of neutrality. 111
Thus there was support in the nineteenth century for the idea that the declaration of blockade against an insurgency automatically resulted in recognition of belligerency and the laws of IAC applying between the parties to a conflict which might otherwise be classified as a NIAC. 112 Certainly, reviving recognition of belligerency as a means of resolving the law applicable to conflicts with non-state actors outside civil conflicts has an attractive explanatory power. As noted above, one could suggest under the doctrine that by the very act of declaring a blockade Israel has recognised the belligerency of Hamas; one could also suggest that some recognition of belligerency is inherent in pronouncements by the Israeli Supreme Court that the conflict between Israel and ‘terrorist groups’ is governed by the law of IAC. However, it is unclear, even doubtful, that the doctrine survived the nineteenth century. 113 Even if the law of belligerency remains applicable, there would still be a real question whether measures against neutral shipping are permissible in such cases. Much of the law suggests that before a neutral vessel could be subject to any measures of interdiction at sea its flag state must have acknowledged the international status of the conflict. 114 Insurgents (as non-state actors) had no right to take action against neutral vessels and were liable to be treated as pirates unless the flag state had recognised their belligerency. 115
Given the uncertainties surrounding recognition of belligerency it is appropriate to enquire whether blockade (or other belligerent rights against neutrals such as contraband) are available in a NIAC. Some would vigorously deny that participants in a NIAC have any rights against the vessels of third states. 116 As noted, in the nineteenth century it was widely thought that such powers rested on the recognition of belligerency by the affected neutrals themselves; the point, however, was fiercely debated. 117 In the twentieth century, the most important example would appear to be the Spanish Civil War (1936-1939). Despite the scale of hostilities involved and the degree of international intervention on both sides (in violation of non-intervention agreements), no European state conceded to any party to the conflict any right to interfere with neutral shipping. 118 During the Algerian revolution (1954-1962) France instituted a high-seas maritime surveillance zone up to 50 nautical miles off the coast of Algeria and asserted a right to visit, inspect, and in some few cases divert merchant shipping. As a consequence 4,775 ships were searched on the high seas in the first year of operations alone, though in total only 10 vessels carrying arms were diverted during the conflict. 119 No formal blockade was declared and the measures, justified on the basis of self-defence, attracted considerable protest. 120 A more recent civil conflict, that in Sri Lanka (1983-2009), certainly involved maritime operations. References, however, to a government naval ‘blockade’ during that conflict are strictly a misnomer. In 1984 Sri Lanka implemented ‘a special naval surveillance zone within Sri Lankan waters … with the … purpose of preventing illegal entry and exit. This move followed mounting concern … that Tamil rebels were receiving both training and equipment from the southern Indian state of Tamil Nadu.’ 121 Most reported maritime interceptions appear to have occurred with Sri Lanka's territorial sea or contiguous zone, ostensibly on suspicion the vessels were engaged in smuggling weapons or supplies to the Tamil Tigers (LTTE). 122 Such interdictions could be justified under ordinary customs and policing powers available within 24 nautical miles of Sri Lanka's baselines and do not require the invocation of blockade. The practice certainly involved no assertion of rights against neutral vessels on the high seas. Finally, Israel implemented a blockade of Lebanon during its 2006 operation in Lebanon against Hezbollah, which was generally not protested. 123 As noted above, however, it is not clear that the Lebanon conflict can be classified as a straight-forward NIAC.
The right of states to implement measures against neutral vessels in NIACs is thus at best an unsettled question. The most one can say is that in higher-intensity conflicts states have sometimes acknowledged or acquiesced in blockades targeting non-state actors. The key examples would be the US Civil War and the Lebanon war of 2006. However, in equally violent conflicts such a right has sometimes not been recognised and attempts to assert rights of blockade or similar measures have been protested (for instance, the Spanish Civil War and the Algerian rebellion). Where such measures are protested as contrary to international law those protests must weigh against the conclusion that there is opinio juris supporting the rule of custom invoked. 124 On the basis of relevant state practice one can at most hazard a suggestion that irrespective of the precise classification of a conflict, states are likely to tolerate the assertion of a blockade only in cases of higher-intensity conflicts on a par with the traditional understanding of war. 125
In the final analysis, the applicability of the law of blockade in a NIAC is uncertain and the level of hostilities between Israel and Hamas in May 2010 were not comparable to the types of conflicts where neutrals have tolerated blockades in recent history.
D. The law of naval warfare and measures against neutral shipping: blockade, contraband and belligerent rights of visit and search
The conclusion to this point is that the conflict in Gaza at the relevant time was most likely not an armed conflict in the strict sense. Alternatively, even if it must be treated as a continuing armed conflict on the basis that no successful and lasting peace has been concluded, there is no sound basis to classify it as an IAC. The conflict should thus be treated as a NIAC. In a NIAC the right to blockade is at best uncertain and its exercise has been tolerated by neutral states only in much higher intensity conflicts than that occurring in Gaza outside of discreet operations such as Operation Cast Lead.
In the event that I am wrong in any of these conclusions it is worth considering the potentially applicable law of naval warfare. The following discussion presumes that either the conflict in Gaza is an IAC or that it is a NIAC and that the laws of naval warfare are applicable in toto in such a conflict. In such cases a belligerent could take a range of measures against neutral shipping. It may forbid all commerce with specified ports or coastlines (blockade). It may prohibit certain listed goods from being shipped to the enemy (contraband). Or it may have a general power of visit and search against neutral commerce (belligerent interdiction). Contraband will not be considered in significant detail here, but belligerent interdiction will be considered at greater length as the most plausible alternative to blockade.
The debate surrounding the lawfulness of Israel's blockade of Gaza has generally invoked the 1994 San Remo Manual on naval warfare. 126 This was an attempt to codify customary law, carried out by a round-table group of naval law experts of diverse nationalities and both military and academic backgrounds. 127 While participants acted only in their personal capacity, the wide participation and inclusion of operational experts gives it a unique status and makes it a necessary starting point for any sensible discussion of the modern law of naval warfare. It is extensively drawn upon in national military manuals including the UK Manual, 128 the Canadian Manual 129 and to a lesser extent the German Manual. 130 It has been expressly invoked in legal analyses offered by Israel 131 and the US has submitted before the International Court of Justice that ‘most of [the San Remo Manual's] provisions reflect customary international law’. 132 I will therefore generally refer to the San Remo Manual alongside national manuals, principally the UK Manual, the German Manual and the US Manual (the latter being used by 25 other countries as a ‘principal legal guide’). 133 As I have noted elsewhere: ‘[w]hile these works lack codifying power, they take a relatively common and expert view of the rules governing interdictions during a [UN] Charter-era armed conflict, and represent (in the military manuals’ case) statements of opinio juris .’ 134 The general approach taken here will be to start with the San Remo Manual and then note points of difference and controversy.
As a final preliminary point, one should note the argument that the flotilla interception could be justified on the alternative basis that if Israel is considered the occupying power in the Gaza Strip it has a right to control maritime access to that territory. 135 While occupation and effective control of land territory results in occupation and control of ‘any abutting maritime areas’ this could not in the present case justify action beyond the territorial sea or contiguous zone. 136
Where there is lawful blockade the blockading power may, under the San Remo Manual (paragraph 98):
capture and divert to port merchant vessels ‘believed on reasonable grounds to be breaching a blockade’; and
attack merchant vessels ‘which, after prior warning, clearly resist capture’.
The right of capture also applies to vessels suspected on reasonable grounds of breaching a blockade if they resist visit and search. 137 Paragraph 146(f) specifies that ‘[n]eutral merchant vessels are subject to capture outside neutral waters if they … are breaching or attempting to breach a blockade’. Neutral waters in the San Remo Manual are defined as the territorial seas and internal or archipelagic waters of non-parties to the conflict (paragraph 14), thus the provision clearly contemplates capture and diversion of blockade-runners on the high seas. As the Mavi Marmara had as its stated aim the breach of the blockade, Israel clearly had ‘reasonable grounds’ to capture it if its blockade was lawful. The reference in paragraph 98 to ‘attack’ where capture is resisted is not, however, a unilateral right to sink the vessel: such action remains subject to the duty to ‘take all feasible precautions … to avoid or minimize collateral [civilian] casualties or damage’. 138
There are some suggestions that a blockade should be enforced only in a defined area of operations 139 in accordance with Article 17 of the London Declaration of 1909. 140 Historically, there was a difference of view between Anglo-American approaches, holding that the entire voyage to a blockaded port was a breach of blockade (rendering the blockade-runner liable to capture at any point) and continental approaches ‘based upon an analogue of the right of hot pursuit’ that would allow capture only after breach of a cordon. 141 In particular Heintschel von Heinegg holds that:
neutrals have only in rare cases been willing to tolerate interference with their merchant shipping in areas distant from the blockaded coasts or ports. … [A]n obligation of States not participating in an international armed conflict to tolerate belligerent measures can be justified only under strict conditions. … If a neutral merchant vessel is captured outside the range of operation of the blockade forces because it–in fact or presumably–was destined to a blockaded port, that violation of the law of neutrality results in a duty to return the vessel and its cargo and to compensate any damage. 142
His position appears to follow from the presumption that the principle of effectiveness ‘would be rendered meaningless if belligerents were entitled to enforce a blockade at a far distance from the [blockaded] area in question’. 143 The argument appears to be that ‘long-range’ enforcement of a blockade would mean belligerents were not required to station an adequate number of vessels near the blockaded territory. This is less than compelling: it rests on an assumption that effectiveness requires close stationing. It is not apparent why long-range enforcement inherently cannot be effective. Such a restrictive view is not generally accepted in military manuals 144 or, as noted above, by the San Remo Manual. The generally accepted opinio juris would appear to be, then, that blockade may be enforced on the high seas even at a distance from the area of naval operations and prior to breach of any cordon.
The larger question is whether the blockade of the Gaza Strip, presuming the doctrine to be available, was legally established. A blockade must satisfy requirements of:
non-discrimination or impartiality (it ‘must be applied impartially to the vessels of all States’); 147 and
As to proportionality, a blockade is prohibited under San Remo Manual paragraph 102 if it either: (a) ‘has the sole purpose of starving the civilian population’ or (b) results in, or may be expected to result in, excessive ‘damage to the civilian population … in relation to the concrete and direct military advantage anticipated’. 148 It is important to distinguish the two criteria set out here: starvation or hunger as a weapon of war and excessive civilian damage. Starvation is not the only basis for judging proportionality. The more specific prohibition on starvation derives from the wider prohibition on excessive damage, 149 which is obviously a cardinal principle of LOAC. 150 In addition to the paragraph 102 proportionality standards, under paragraph 103 of the San Remo Manual:
If the civilian population of the blockaded territory is inadequately provided with food and other objects essential for its survival, the blockading party must provide for free passage of such foodstuffs and other essential supplies, subject to:
the right to prescribe the technical arrangements, including search, under which such passage is permitted; and
the condition that the distribution of such supplies shall be made under the local supervision of … a humanitarian organization which offers guarantees of impartiality[.]
Under paragraph 103 a blockader thus has the right to search inbound supplies and prescribe the mode of their delivery, precisely the rights Israel claimed in respect of the Mavi Marmara .
The provisions of paragraph 103 of the San Remo Manual draw on the text of Article 70 of the First Additional Protocol to the Geneva Conventions of 1977 (‘AP I’). 151 Though Article 70 is applicable only to ‘territory under the control of a Party to the conflict, other than occupied territory’ near identical language governs the position of occupied territories under Article 59 of GC IV. 152 Indeed, the duty under Article 59 is arguably a stronger one, as it does not allow the occupier the right to prescribe technical arrangements, such as search. 153 Nonetheless, paragraph 103 will be presumed to be a correct statement of the applicable law in the present case for several reasons: first, its status as a consensus text that attempted to codify customary rules; and secondly, the acceptance of article 70, AP I as the relevant standard in other military manuals. 154
Several arguments may be made concerning the legality of the blockade either on the basis of rules against starvation or collective penalties (prohibited methods of warfare), or on the basis of excessive damage to the civilian population (the principle of proportionality). The argument here will be that, in this context, all analyses based on prohibited methods of warfare ultimately fall to be assessed on the basis of proportionality.
First, a strict textual literalism might suggest a blockade can be legal so long as the civilian population is not starved to death and is maintained above a minimum level needed for survival. Indeed the Israeli government's stated position is that it is ‘not require[d] … to allow the passage of unessential goods or amounts of goods that exceed what is required for basic humanitarian needs’ into Gaza 155 or that it may restrict the passage of goods to a ‘humanitarian minimum’. 156 The problem with this approach is that there is no right to prevent or limit the passage of food. The free passage of food, where the civilian population is ‘inadequately provided’ with it (a point returned to below), is an unqualified and absolute obligation subject only to technical implementation measures which may not result in ‘relief consignments … [being] diverted or delayed unnecessarily’. 157 The drafters of the San Remo Manual intended this provision to be an ‘unequivocal statement that the blockading power is obliged to allow transit of relief shipments through the blockade’. 158 At the time of the flotilla interception Israel's policy on which foodstuffs were allowed into the Gaza Strip was notoriously restrictive and, indeed, arbitrary as anything not expressly permitted was prohibited; Israel also carefully regulated the availability of basic supplies in Gaza in an effort to deny such supplies to Hamas. 159 The promptness with which aid consignments were allowed through land check-points is disputed, though since the flotilla incident the volume of goods passing into Gaza has markedly and suddenly increased. 160 While this analysis may support a finding that paragraph 103 obligations were not fully complied with, this would not support a claim that the blockade became void for illegality. That can only occur if interference with humanitarian aid in turn leads to a proportionality violation under paragraph 102.
As to the question of the prohibition on ‘starvation’ as a method of warfare, the prohibition is not one on ‘starvation to death’. The ordinary definition of ‘starvation’ is simply to cause hunger. 161 However, it might be thought that a significant difficulty is introduced by the San Remo Manual's further requirement that this be the sole purpose of the blockade. 162 While the San Remo manual's drafters rejected the view that the prohibition on starvation rendered blockade per se unlawful, they did take the view that:
a ‘blockade, in order to be of itself illegal must have the sole purpose of starving the population or have a disproportionate effect as indicated in [paragraph 102](b)’; and
‘[w]henever … blockade has starvation as one of its effects’ this triggers the obligation to allow passage of relief shipments.’ 163
It is therefore possible that if the civilian population is inadequately supplied with food and that lack is not being met by relief shipments, that situation may itself constitute a ‘disproportionate effect’ irrespective of intent. Thus, ‘[i]f a blockade has the effect of starving the civilian population it becomes illegal’. 164 This approach appears more consistent with AP I, article 54(1) which prohibits ‘[s]tarvation of civilians as a method of warfare’ without reference to purpose or motive. Article 54 clearly applies to blockades as AP I expressly modifies the rules of ‘sea warfare which may affect the civilian population … on land’. 165
It may also be argued that the Gaza Strip is simply not ‘inadequately provided’ with food. It is common knowledge that there is an extensive network of smuggling tunnels along the Egyptian border, some large enough to drive vehicles through; as a consequence it is sometimes claimed that food is readily available in the shops and markets of Gaza. 166 The claim is contestable, but even if true it is clear that many Gazans, due to the economic collapse engendered by the closure regime and blockade, lack the means to buy (or buy consistently) the food that is available. 167 The final result, starvation in the ordinary sense of causing hunger, is the same.
In the alternative, the blockade might be thought disproportionate not in the paragraph 102 sense but on the basis that other, better-calibrated means of achieving Israel's stated security goal of keeping weapons and weapon-making supplies out of Gaza are available. The argument would run that this objective could be better and more proportionately met by issuing a list of goods which will not be allowed into Gaza thus permitting all commerce not expressly prohibited. The latter approach was adopted as Israeli policy on 4 July 2010 when a list of prohibited weapons and dual-use items was issued, with all other goods presumed clear for entry. 168 Such an approach is also permissible in naval warfare under the doctrine of contraband. 169 Nonetheless, cement and concrete remain restricted ‘dual use’ items under the new Israeli policy and no provision is made for the export of goods from the Gaza Strip. This view would accommodate arguments that food is not the central issue in the Gaza Strip; the central problem is the collapse of an economy dependent on exports and foreign earnings (typically through employment in Israel). 170 While it was open to Israel to argue that the blockade measures have successfully reduced the number and intensity of rocket attacks on its territory, other and more proportionate measures (such as those now adopted) would perhaps have been sufficient throughout. Whether this argument is sustainable in law is a difficult question. There is some limited support for the proposition that principles of ‘necessity and proportionality … restrict the choice of targets, the use of methods and means of warfare and measures taken against neutral shipping’; this is said to follow from the requirement that such actions must not only be ‘in accordance with the rules of international humanitarian law’ but must also be ‘necessary and proportionate to the needs of self-defence of the parties.’ 171 Indeed, the San Remo Manual includes as paragraph 5 the principle that: ‘[h]ow far a State is justified in its military actions against the enemy will depend upon the intensity and scale of the armed attack for which the enemy is responsible and the gravity of the threat posed.’ 172 While the view that proportionality may constrain the choice of means and methods of warfare is not without controversy, 173 the UK Manual holds:
even when resort to force is justified, it should not exceed what is necessary and proportionate to the achievement of a goal for which force may be used. In a conflict of limited scope, this may mean a belligerent state is constrained … in the action it may lawfully take against the shipping or aircraft of states not involved in the conflict. 174
This does not appear to be a widely-adopted view. There is significantly more support for the proposition that the duty to take precautions to minimise collateral damage may limit the choice of means of carrying out a particular operation at the tactical level, but this does not assist the present argument. 175 It would at best be a minority view that the availability of other more precisely calibrated means of achieving a relevant military advantage will per se render a blockade disproportionate.
In the further alternative, the closure of the Gaza Strip (including the blockade) has also been denounced by the International Committee of the Red Cross (ICRC) as a ‘collective punishment’ of the civilian population and a ‘clear violation’ of humanitarian law. 176 If one takes the purpose of the closure/blockade as being to punish Hamas for rocket attacks and the continued detention of Gilad Shalit, 177 then it might well be thought a ‘general penalty, … inflicted upon the population on account of the acts of individuals for which they cannot be regarded as jointly and severally responsible’ and thus a collective punishment contrary to Article 50 of the Hague Regulations and the law applicable in both IACs and NIACs. 178 However, if one accepts the Israeli case that the principal aim of these measures is to prevent the smuggling of weapons and enemy personnel into the Gaza Strip, 179 both legitimate military objectives, then the question becomes one of proportionality not collective penalties.
On the narrower view of paragraph 103 proportionality, there is evidence that the civilian damage involved is excessive. The UN Food and Agriculture Organization estimates unemployment in the Gaza Strip at more than 30% and considers 61% of households to be ‘food insecure’. 180 The point may be taken that this situation is brought about not principally by the naval blockade, but by a complex of measures including restricting the passage of goods over the land border. 181 However, in the absence of the blockade, if relief ships could arrive and exports depart by sea, one might expect the situation to be ameliorated. While Gaza lacks a port, supplies may still (if much less efficiently) be disembarked at sea to smaller craft to take ashore. 182 The claim that the impact of the maritime blockade is ‘zero’ because this was not a traditional route for goods into or out of Gaza is less than compelling. 183 Proportionality is a contextual assessment: if the excessive civilian damage is actually being caused by the disproportionate effects of land border closures, it is hard to see how implementing a maritime blockade in support of those other disproportionate measures, can itself be proportionate. 184 The blockade alone does not of itself cause civilian suffering (positive causation), but it does present an obstacle to its amelioration (negative causation).
The position adopted here is that the right of blockade was not available ab initio as the conflict was neither of an international or sufficiently intense character to give rise to the right. Even if it were available, the preferable conclusion is that it was disproportionate as it was implemented as part of a comprehensive denial of commerce and supplies above a level needed to meet basic humanitarian needs. These measures inflict disproportionate civilian damage. A maritime blockade implemented as part of a package of disproportionate measures should not itself be judged proportionate.
If, however, one presumes LOAC applies to the conflict, but presumes the blockade to be illegal, this would raise the question of whether there were alternative grounds on which the flotilla interdiction could have been justified. This raises the question of the application of the law of contraband and whether there is a general right of belligerent interdiction.
Under the law of contraband a belligerent may prohibit certain goods from being shipped to the enemy and may exercise rights of visit, search and seizure over neutral vessels on the high seas to this end. Contraband goods liable to seizure are those ‘susceptible to use in armed conflict’ and ‘ultimately destined for [enemy] territory’: this category will obviously include weaponry, but may also include materiel susceptible to both military and civilian use (‘dual-use materiel’). 185 As I have noted elsewhere the essential features of the doctrine of contraband are:
belligerent warships may visit and search any neutral ship outside neutral waters if there is a reasonable suspicion that the vessel is carrying contraband ultimately destined for territory under enemy control;
this right cannot be exercised against neutral vessels escorted under convoy, though the convoy's leader may be asked to give particulars of, and destinations for, their cargoes; and
neutral nations should be forewarned what goods will constitute contraband [through the publication of contraband lists] — though weapons shipments will clearly always constitute contraband. 186
The San Remo Manual, the UK Manual and the Canadian Manual all expressly precondition the exercise of rights of visit and search upon both reasonable suspicion (or ‘grounds’ or ‘belief’) that the neutral vessel is carrying contraband and the prior publication of lists of prohibited goods (‘contraband lists’). 187 The US Manual posits an absolute right of visit and search of neutral commerce, irrespective of any particular suspicion that it is carrying contraband. 188 The language of the German Manual suggests that, in practice, exercise of a right of search and seizure will be preceded by a suspicion that a vessel is carrying contraband. 189 However, as discussed below, the German Manual subscribes to an absolute right of belligerent visit and search which would appear inconsistent with any strict requirement of reasonable suspicion prior to interdiction. Earlier editions of the US Manual once suggested that publishing lists of ‘free goods’ (all other trade thus being implicitly contraband) might fulfil any requirement to publish lists; however, that view is not retained in the current edition. 190 The German Manual similarly does not require that contraband lists be published for the right to be enforced, but does limit contraband strictly to materiel either ‘essential for war’ or ‘destined for the administration or armed forces’ of the enemy. 191 The general view, however, appears to be that measures of visit and search under the doctrine of contraband are only permissible on the basis of reasonable suspicion and where contraband lists have been published in advance. Historically, contraband lists were often overly broad 192 and to prevent their abuse most Manuals now require that ‘[c]ontraband lists shall be reasonably specific’. 193
Even if the Israel-Hamas conflict was one which permitted recourse to high-seas measures against neutral shipping, the requirements of the doctrine of contraband were not met. Israeli military legal advisors have been quite clear there was no positive basis for belief the cargo of the Mavi Marmara consisted of weapons or was destined for Hamas. 194 Further, no list of prohibited goods had been published. Such a list was first published on 4 July 2010. 195 Israel's policy at 31 May 2010 of prohibiting all goods not permitted had also not resulted in the publication of a list of ‘free goods’: ‘there [was] no official list for traders to observe’ as decisions were made solely ‘on a case-by-case basis’. 196 This approach self-evidently cannot provide sufficient certainty to allow neutral merchant vessels to know in advance whether they risk interception for breach of contraband by carrying particular goods. 197
Israel deliberately chose not to invoke contraband in relation to the Mavi Marmara episode. Israel's Chief Military Advocate General, Avichai Mandelblit, made essentially two points on the topic to the Turkel Commission: first, that in his view intercepting a vessel on the basis of contraband would require ‘a plausible basis for suspicion’ that there was contraband aboard; secondly, that invoking contraband rather than blockade involves complex judgements and some uncertainty as to what dual-use materiel may be declared contraband (for example, concrete may be used for civilian construction or military purposes). 198
In conclusion, Israel made no case for the application of contraband and any attempt to invoke it must fail in any event. No contraband lists were published and the Israeli evidence is that there was no suspicion the flotilla was carrying weapons or other goods destined for Hamas.
4. Belligerent interdiction
Could Israel have argued that it had an absolute belligerent right to visit and inspect neutral vessels in the course of an armed conflict irrespective of any declared blockade or system of contraband? The German Manual supports such a position, stating:
Warships of a party to a conflict are entitled to stop, visit, and search merchant ships flying the flag of a neutral state on the high seas and control the contents and destination of their cargo. 199
The US Handbook asserts a similarly wide right of interdiction. 200 The commentary to the German Manual notes, however, ‘[t]he state of international customary law is controversial in many details concerning the extent of this [power of] control’. 201
The San Remo Manual takes the narrower view that any such right of visit and search can only be exercised upon ‘reasonable grounds’ for suspecting the vessel is subject to capture; 202 and grounds for capture include ‘breaching a blockade’. 203 The UK and Canadian Manuals takes the same approach. 204 Further, the San Remo Manual provides that a neutral vessel subjected to visit and search may only be diverted from its destination for the purpose of conducting such a search either with its consent or where search at sea is impossible or unsafe. 205 In the Mavi Marmara context invoking a separate right of visit and search may not have greatly assisted Israel. Such a right would not necessarily have allowed diversion or capture, and may have required a reasonable case that this particular vessel was smuggling weapons or that the aid aboard would be diverted to Hamas. 206 Indeed, the view of the Israeli Chief Military Advocate General was that absent a blockade belligerent rights could only be exercised against neutral shipping where there was a ‘plausible basis for suspicion’. 207 Israel's stated position is that all vessels approaching the Gaza Strip by sea must be searched, a position that is best sustained under the general law of blockade which involves an absolute prohibition on commerce and allows for the diversion of aid supplies for search and their supervised distribution. 208
The conclusion of this section is that, even presuming this was a conflict to which the laws of naval warfare applied, the blockade imposed would have been illegal and any alternative right of interdiction would have to have been based on, at the least, a plausible suspicion that the flotilla was taking war materiel or combatants into Gaza. It has not been the position of the Israeli military witnesses before the Turkel Commission that any such suspicion existed. Where an underlying blockade is not lawful, any boarding is obviously an act engaging state responsibility and compensation would have to be paid. Certainly under the laws of peace, where a vessel is unlawfully interdicted there is an obligation to compensate the vessel (including persons aboard) for loss, injury or delay. 209 Such liability is strict, even in cases where a reasonable suspicion proves unfounded, in order to deter abuse of exceptional powers to interfere with freedom of navigation. 210 Similarly, under LOAC where there is a breach of the law of neutrality there is a duty to compensate the vessel. 211 (There may also be a right of compensation in any case where a neutral vessel is sunk, whether lawfully or not. 212 ) The question remains, however, as to what law was applicable during the enforcement action against and aboard the Mavi Marmara .
E. Enforcement action against the Mavi Marmara
Despite the conclusion that there was no legal basis for the blockade, this section will commence with an examination of the rules applicable to enforcement measures against merchant vessels under LOAC, if only for the sake of completeness. Consideration will then be given to the applicable legal framework if there was no lawful blockade and if the interception occurred outside the context of an armed conflict. The final position reached is that irrespective of the legal basis for the action aboard the Mavi Marmara , the applicable law is likely the same.
Where a lawful blockade has been implemented it must be enforced effectively and against all vessels. A blockade that is ineffective or selectively enforced is, or becomes, unlawful. Stopping a blockade-runner is thus in all cases a legitimate military objective, irrespective of that vessel's cargo, passengers or mission. The question of whether intercepting any given vessel is ‘proportionate’ in the ordinary LOAC sense therefore does not arise. Under the law of blockade a vessel which resists visit and search or capture becomes subject to attack: 213 resistance is not expressly prohibited, 214 but resistance is very much at the vessel's risk and historically led to automatic condemnation before a prize court. 215 As previously noted, a right of attack against a blockade-runner does not imply a license to sink a vessel carrying hundreds of civilians. 216 An attempt at capture should normally precede attack 217 and attacking a neutral civilian vessel is only permitted exceptionally. 218 Even in cases where attack is permissible, the belligerent must still ‘take all feasible precautions … to avoid or minimize collateral [civilian] casualties or damage’. 219 In any tactical-level action where the safety of civilians is in question this obligation to take precautions requires, in the words of the US Manual, that:
the commander must decide, in light of all the facts known or reasonably available to him, including the need to conserve resources and complete the mission successfully, whether to adopt an alternative method of attack, if reasonably available, to reduce civilian casualties and damage. 220
Other national military manuals similarly interpret the requirement of taking ‘feasible precautions’ as potentially limiting the means and methods of attack. 221 As one commentator has put it:
As a last resort, the right of [belligerent interdiction] … may be exercised by using force. However, this is limited by a strict principle of necessity. Only such force is permissible which is indispensible to enforce the right … , in particular to prevent a merchant ship from evading such control. 222
Force cannot be used against a vessel which has clearly signalled its intention to surrender by means such as ‘hauling down the flag, hoisting a white flag, stopping the engines and responding to the attacker's signals’; continuing an attack after such surrender is a war crime. 223 However, no such questions arise here as the IDF quite properly moved to capture rather than attack the flotilla vessels. The question is one of the applicable legal standards governing the use of force against persons aboard a vessel during such an interdiction.
There are essentially two possibilities: the application of a LOAC paradigm or a law-enforcement paradigm governing the use of force. The first approach could be justified on one of two bases: either that irrespective of the legality of the blockade there was an underlying armed conflict (a conclusion this article considers dubious) and this was a belligerent act connected to that conflict; or that as an act involving military force against another state's flag vessel, the interdiction should in principle be governed by LOAC. The latter position is hard to sustain: while I have contended elsewhere that the use of force against a foreign flagged vessel on the high seas may violate the prohibition on the threat or use of force in international relations, 224 there is only limited and ambiguous authority suggesting a military attack on a single merchant vessel constitutes an act giving rise to a right of self-defence under the jus ad bellum . 225 It seems unlikely that the lesser act of capturing (as opposed to attacking or sinking) a single flag vessel could be said to be in some sense an ‘act of war’ giving rise either to an IAC or the application of LOAC on some other basis. In any event, even during an IAC it is acknowledged that there are situations where military forces may be restricted to a ‘policing’ use of force paradigm. The duty to maintain order in occupied territory is an obvious example. 226 Occupation forces may, for example, have to quell or disperse rioting civilians. LOAC provides no guidance on dealing with riots, and in such cases armed forces ‘must respond … according to the law enforcement model’ under which they are not entitled to target individuals with lethal force except in self-defence. 227 Startlingly, the laws of naval warfare provide no positive guidance on the use of force in capturing neutral merchant vessels. One must suggest that the appropriate analogy then is the law-enforcement paradigm as it applies in other LOAC situations such as maintaining order during an occupation.
The only exception to this approach might be where a civilian aboard a captured vessel became a direct participant in hostilities (always presuming there to be an armed conflict). The ICRC guidance on what constitutes direct participation in hostilities emphasises three criteria: (1) a ‘threshold of harm’ must be crossed, such as violence that threatens a member of enemy forces with death or injury; (2) there must be a causal link between the civilian's act and that harm; and (3) this must be done ‘in support of a party to the conflict … (belligerent nexus)’. 228 Spontaneous rioters aboard the Mavi Marmara armed with metal rods and kitchen knives might satisfy the first two criteria but are unlikely to satisfy the third. Violent pro-Hamas activists might satisfy all three criteria—and the possibility that some protestors aboard the Mavi Marmara could have crossed the line into being direct participants in hostilities is more than theoretical. However, several caveats must be entered: first, as the conclusion of this article is that there was not an armed conflict at the relevant time, the concept of participation in hostilities has no application; secondly, even if there were an armed conflict, in the absence of clear evidence of belligerent nexus and to the extent there was doubt as to the status of any rioter on the part of the military boarding party that person must still be treated as a civilian; and thirdly, the issue of how to respond to life-threatening violence from rioters is amply covered by the law of self-defence.
If a law-enforcement paradigm is applicable under LOAC, what is its content? The only available analogy is supplied by the general international law of the sea applicable to maritime policing operations. The International Tribunal for the Law of the Sea restated the law on the use of force to carry out interdictions in MV ‘Saiga’ (No. 2) where it found that general international law:
requires that the use of force must be avoided as far as possible and, where … unavoidable, it must not go beyond what is reasonable and necessary in the circumstances. Considerations of humanity must apply … It is only after the appropriate actions fail that the pursuing may, as a last resort, use force … Even then … all efforts should be made to ensure that life is not endangered. 229
On the specific question of the use of force by government officials once aboard a civilian vessel ITLOS endorsed the view that government agents must:
avoid the use of force except when and to the degree necessary to ensure the[ir] safety … and where … obstructed in the execution of their duties. The degree of force used shall not exceed that reasonably required in the circumstances. 230
Obviously, these same standards would be applicable if the operation was judged a maritime enforcement action outside of LOAC. The conclusion then is that under the laws of peace or war much the same standards should be considered to apply. Indeed, even when the capture of an unarmed civilian vessel becomes a legitimate military objective under LOAC, targeting it with more than a ‘policing’ level of force is unlikely to be consistent with the fundamental duty to take all feasible precautions to minimize civilian casualties.
Whether these standards were breached is obviously a question of fact. Inserting soldiers or other armed government agents by helicopter is not an obviously disproportionate use of force in relation to an effort to stop and seize control of a vessel otherwise refusing to stop. Similar action was taken, albeit with flag state consent and without the use of stun grenades or live fire, by Spain in the So San incident where a vessel was boarded on suspicion of being engaged in the proliferation of weapons of mass destruction. 231 The real question is whether the level of force used in conducting the interdiction was necessary.
The actual use of force once aboard the Mavi Marmara was condemned in terms in the UN Security Council by Turkey, Mexico, Brazil, Austria, Russia, Uganda, France, Bosnia and Herzegovina, and Lebanon. 232 France made the clearest statements that the level of civilian casualties was at least prima facie evidence that disproportionate force had been used. 233 The United Kingdom took the view:
Given the number of casualties in this incident, Israel now bears a responsibility to provide a full account of what occurred, what efforts were made to minimize the loss of life and why the death toll was so high. It will be particularly important to establish whether enough was done to prevent deaths and unnecessary injuries. 234
This is surely the absolute minimum required under any view of the law. Israel's account of events is that live fire was not used from the outset as a means of taking control of the vessel, but in self-defence by soldiers attacked by protestors. This implicates the law of personal self-defence or defence of others, available to soldiers and civilians alike in peace or war. Under LOAC reasonable and proportionate self-defence or defence of others remains an available defence to soldiers charged with unlawful killing. 235 The FFM Report does not consider the possible claims of self-defence (on the part of IDF officers or civilians) in terms. Instead it addresses the question of deaths aboard the Mavi Marmara solely in terms of human rights violations (not considered in this article) and the law of war crimes. The issue of war crimes will be returned to below but the question of self-defence remains an important one.
The FFM Report raises several disturbing categories of deaths: (1) civilians on the top deck attempting to obstruct the boarding who were either shot once in the chest or lower limbs and then shot again in the head or who were shot from above and not at close range (the inference being that at least one live round was fired from the boarding helicopter); and (2) civilians on lower decks who were shot and killed from above with live fire (i.e. from the top deck where the boarding party landed). 236 It is hard to see how persons in the latter category could have posed any imminent threat to Israeli soldiers justifying action in self-defence. The case of the protestors on the top deck raises the most acute factual questions of self-defence on either side. It is acknowledged that protestors were armed with, at the least, iron bars. If used against the head these could clearly pose a risk to life justifying recourse to firearms. In such a case it is not necessarily inconsistent with self-defence to shoot a person once in an attempt to disable them and then a second time in the head if they cannot be overcome by other means. Nor, obviously, is using an iron bar against a person armed with a 9 mm pistol who shoots first. One should also note that once several Israeli soldiers had been captured this could have justified recourse to further force to liberate them if there was an apprehension that their lives were at risk. Finally, there is, of course, the tragic possibility that the use of lethal force may have commenced with an act of mistaken self-defence on one side or the other. Whether a mistake of fact (i.e. a genuine but erroneous belief one was being attacked) can found a claim of self-defence under international criminal law is contentious. 237 It is otherwise a question of the applicable criminal law of the flag state or the state of nationality of the perpetrator or victim depending on where any resultant criminal case is heard. It remains to be seen what, if any, detail from the Israeli internal military inquiry into the circumstances of each particular death is made public.
The FFM Report characterises the circumstances of some or all of the deaths on board as providing prima facie evidence of the war crime of wilful killing, 238 invoking article 147 of GC IV as applicable law. This raises an underdeveloped point in the law of naval warfare. Presuming, contrary to the argument here, that there was an IAC, were those aboard the Mavi Marmara protected persons under the Geneva Conventions who could be the victims of war crimes? They were certainly civilians protected from direct targeting under the LOAC principle of distinction, but the class of civilians protected by GC IV (and thus the law of war crimes) is textually much narrower. Article 4 of GC IV, first sentence, reads: ‘Persons protected by the Convention are those who, at a given moment and in any manner whatsoever, find themselves, in case of a conflict or occupation, in the hands of a Party to the conflict or Occupying Power of which they are not nationals’. Taken literally this would appear to apply to citizens of neutral states aboard a vessel interdicted at sea during an IAC, at least once the vessel is under the control of the boarding party. Such persons are ‘in the hands of a Party to the conflict … of which they are not nationals’. However, the intention behind Article 4 as a whole, according to the ICRC commentary, was only to protect ‘(1) “enemy nationals” within the national territory of each of the Parties to the conflict and (2) “the whole population” of occupied territories (excluding nationals of the Occupying Power)’ (emphasis added). 239 This does not appear to contemplate the nationals of neutral states captured at sea. In addition, the moment these neutral citizens were removed to the territory of the interdicting belligerent state, some or all of them could lose GC IV protected status as under Article 4, third sentence:
Nationals of a neutral State who find themselves in the territory of a belligerent State … shall not be regarded as protected persons while the State of which they are nationals has normal diplomatic representation in the State in whose hands they are.
The question of such persons’ treatment becomes solely a matter for diplomatic relations. The alternative is perhaps to argue that to the extent that naval warfare is not expressly contemplated by GC IV the concept of protected person under the customary law of war crimes must be interpreted in light of the Convention's object and purpose. As was said in the Tadić Appeal Chamber judgment:
Article 4 of Geneva Convention IV, if interpreted in the light of its object and purpose, is directed to the protection of civilians to the maximum extent possible. … Its primary purpose is to ensure the safeguards afforded by the Convention to those civilians who do not enjoy the diplomatic protection, and correlatively are not subject to the allegiance and control, of the State in whose hands they may find themselves. 240
This passage emphasises a literal and broad reading of the first sentence of Article 4. Such a broad and progressive interpretation of the treaty law is controversial. There was no consensus that such an approach should be adopted in the Rome Statute of the International Criminal Court: the question of the scope of persons protected by GC IV was left to the Court to decide in future cases. 241
It would, perhaps, have been a more secure approach to argue that the applicable law of war crimes protecting neutral civilians intercepted at sea stemmed from article 3 of each the four Geneva Conventions (Common Article 3), which protects inter alia persons taking ‘no active part in hostilities’. Common Article 3 lists certain prohibitions, including murder, applicable during conflicts ‘not of an international character’. The International Court of Justice has held that the content of Common Article 3 is equally applicable in IACs 242 and the ICTY held in Tadić that breaches of Common Article 3 ‘entail individual criminal responsibility, regardless of whether they are committed in internal or international armed conflicts’. 243 The offence of murder under Common Article 3 has the same content as the crime of wilful killing under GC IV in any event. 244 As a residual category of minimum protections, the most appropriate source of applicable law of war crimes during naval operations would thus appear to be Common Article 3.
Also worth consideration under either a policing or LOAC paradigm is the extensive appropriation of private property from the flotilla. It is worth noting that the only property offence agreed to exist in both IACs and NIACs is pillage, the offence of appropriating property for private use. 245 (There has been some reported evidence of misappropriation of private property by IDF personnel for personal use. 246 ) The broader crimes of extensive and wanton appropriation or destruction of property not justified by military necessity apply only in IACs. Obviously, such crimes can be committed only against civilians protected by one of the four Geneva Conventions, as discussed above. Otherwise, the appropriation of property without legal cause from foreign citizens and corporations would appear principally an issue of human rights or the law of expropriation and diplomatic protection.
A final point to note is that if one concludes there was an armed conflict and evidence of war crimes, as the FFM Report has done, a question arises as to whether such crimes might be prosecuted before the International Criminal Court (ICC). Comoros, the Mavi Marmara 's flag state, is a party to the ICC Statute. Any offences committed aboard therefore occurred within the ICC's spatial jurisdiction, which extends to crimes ‘committed on board a vessel or aircraft’ registered in a state party. 247 As these events occurred on the high seas, it is not presently relevant to examine the legal effect of the 2009 declaration by the Palestinian National Authority's accepting ICC jurisdiction over events occurring on its territory. 248 A significant obstacle to the ICC's taking jurisdiction in this case, however, would be the requirement that a case be of ‘sufficient gravity to justify’ the involvement of the Court. 249 There is also a relevant consideration (or discretion) implicit in the ICC's having jurisdiction over war crimes ‘ in particular when committed as a part of a plan or policy or as part of a large-scale commission of such crimes’ (emphasis added). 250 No matter how deplorable, the loss of nine lives seems unlikely to warrant the attention of a Court with jurisdiction ‘limited to the most serious crimes of concern to the international community as a whole’. 251
III. C onclusions
The present article has started from the premise that the widely-shared view that Israel remains the occupying power in the Gaza Strip under GC IV, part of the laws of IAC, has no necessary implications for the characterisation of any conflict in that territory. The possibility that the laws of IAC and NIAC may apply in the one territory at the same time is generally overlooked. There is no contradiction in having two legal regimes of armed conflict applicable in the same space where they bind different sets of parties. As the conflict between Israel and Hamas is not a conflict between states, it can be only a NIAC. The contention that the conflict should be considered an IAC on the basis of the applicability of GC IV to the Gaza Strip is at best a position of principle supported by little authority. Arguments that because the conflict is ‘external’ to Israel it must be an IAC misapprehend the concept of ‘non-international’ armed conflict. The distinction is one of identity (between states and non-states), not of geography. Equally, arguments based on belligerent recognition face significant obstacles given that the doctrine is widely presumed to have fallen into desuetude and state practice since the nineteenth century is effectively non-existent.
However, for the conflict between Israel and Hamas to qualify as a NIAC it must reach a certain level of intensity. It has certainly reached that threshold in the course of particular engagements, such as Operation Cast Lead, but may be generally regarded as falling far below it. The situation may thus be characterised as only sporadically an armed conflict in the LOAC sense and at such time the laws of NIAC would apply. It must be conceded, however, that there is authority supporting the proposition that even lengthy cease-fires or lulls in hostilities do not necessarily terminate an armed conflict. One might thus have an underlying NIAC that continues (in law if not in fact) until a final peace is concluded.
The question of a state's ability to invoke blockade or other belligerent rights in a NIAC or against a non-state actor turns on an assessment of state practice. With the exception of the US Civil War it is hard to point to examples of NIACs where neutral states have tolerated the assertion of a blockade. Even if one suggested that any rule arising from the US Civil War was limited to comparable fact situations (a high intensity conflict recognised as such by neutral states), those facts are not present here. Indeed, even when comparable facts were present in the Spanish Civil War they were not seen as giving rise to a right of blockade. Other examples of state practice are also less than conclusive: France's actions against neutral shipping in the Algerian conflict were widely protested and cannot be seen as rule-creating; properly understood, the Sri Lankan civil conflict involved no assertion of rights against neutral shipping on the high seas; and Israel's generally unprotested blockade of Lebanon in 2006 is inconclusive as that conflict may be considered to have involved an IAC between Israel and Lebanon and a NIAC between Israel and Hizbollah. There is thus no consistent state practice and opinio juris suggesting blockade is available outside an IAC. It follows from this that Israel had no right to impose a blockade on the Gaza Strip and its enforcement of that unlawful blockade against the flotilla including the Mavi Marmara was an act incurring state responsibility.
In the alternative, this article has suggested that even if there was a right to blockade the Gaza Strip that blockade had become unlawful by virtue of inflicting disproportionate civilian damage. The argument is put that the closure regime imposed by Israel on the Gaza Strip by land, sea and air has a single military objective: denial of military supplies and personnel to Hamas. The maritime blockade does not fall to be assessed separately because in its absence aid shipments could enter by sea. Proportionality must be a contextual assessment; where an objectively related package of measures with a single military aim creates disproportionate damage in toto , it should not be judged legal through the device of considering its components piecemeal. The alternative argument that a blockade is disproportionate where less-damaging means of achieving the same military aim are available is, at best, a minority view supported by limited authority. Arguments that, presuming the law of naval warfare to apply, Israel would have had a right to intercept the flotilla based on a general right of belligerent interdiction also fail as there was no reasonable basis for suspicion that the flotilla intended to convey materiel to Hamas (a consideration that would also be fatal to arguments based on the law of contraband). The wrongfulness of the interdiction, however, is a separate question from the standards governing conduct of Israeli forces once aboard. The argument here has been that under LOAC or the international law of the sea the applicable law would be a policing paradigm where force should be limited to the minimum necessary to achieve legitimate objectives. However, in peace or war both soldiers and civilians have a right to self-defence: a right invoked by all parties to the Mavi Marmara incident. If soldiers were attacked first with iron bars, they had a clear right to defend themselves. Once several of the boarding party were captured, force was also justified in rescuing them. To the extent that there is evidence that some civilians aboard the Mavi Marmara were shot from above and at a distance, however, this raises real questions about whether a plausible case of necessity or self-defence could be made. If the interdiction is considered to have occurred in the course of an armed conflict a further question arises about the protections afforded to citizens of neutral powers captured at sea. The argument here has been that in such cases civilians would be covered by Common Article 3 to the Geneva Conventions, applicable in IACs and NIACs, and could thus be victims of, inter alia , the war crimes of murder and pillage.
P ostscript : T he T urkel C ommission and the T urkish N ational C ommission of I nquiry R eports
After the present article was readied for publication, the Turkel Commission released the first part of its report on 23 January 2011. At 295 pages it is a wide-ranging examination of the facts and applicable law relevant to the legality of the blockade, the actions of the IDF and the actions, identity and motivations of the protestors. This postscript outlines and discusses principal features of the Commission's legal argument on blockade: space precludes discussion of other relevant issues. 252 As to blockade, the Commission:
refers to state practice both during and outside IACs to reach the conclusion that the conduct of military operations to deny passage to and from enemy territory by sea is an ordinary function of navies;
asserts that blockade is applicable in NIACs;
classifies the conflict between Israel and Hamas as an IAC, and concludes therefore that the law of blockade is applicable;
finds that the blockade did not cause starvation and was not otherwise disproportionate given Israel's legitimate military objectives; and
on this basis concludes that the blockade was in conformity with international law.
There are problems with each of these arguments. On the first point it is understandable that the Commission wished to establish that naval interdiction operations are not inherently inconsistent with the freedoms of the high seas. However, the Commission itself notes that not all its cited practice would necessarily satisfy the legal requirements of blockade, 253 the 1962 Cuban Missile Crisis being a possible case in point. 254 To the extent it cites state practice during IACs, this supports the uncontroversial proposition that blockade is legal in IACs.
With respect to its conclusion that the law of blockade is also applicable in NIACs, the Commission cites only two episodes of supporting state practice. 255 The first is the US Civil War. The Commission acknowledges that this episode rested on the traditional requirements of recognition of belligerency, including that the conflict reach a certain threshold of intensity ‘similar to that of a war between states’. 256 However, the Commission concludes that these requirements are now ‘almost irrelevant’. 257 It is uncertain what the reader is meant to infer from this. Recognition of belligerency was a permissive rule extending the laws of war between states to situations they otherwise did not cover. It is not the logical result of the desuetude of recognition of belligerency as a doctrine 258 that blockade becomes automatically applicable in NIACs. The second episode, and the only item of post-second world war practice, referred to is the ‘conflict of the “nationalists” with the People's Republic of China’. 259 In 1949, the Nationalist Chinese government purported to close certain ports held by Communist forces. Correspondence with the US, however, made it clear the Nationalists were not invoking belligerent blockade. 260 Neither case appears to support the Commission's conclusion. The Commission ultimately makes arguments de lege ferenda in support of its position. It refers in particular to the ‘increasing acceptance’ or ‘consensus that is developing’ as to the common body of law applicable to IACs and NIACs in order to conclude ‘it is likely there will be a willingness on the part of courts and other bodies to recognise’ that the law of blockade is applicable in NIACs. 261 Although protective humanitarian standards are clearly converging between the two bodies of law, it is far from obvious this same dynamic justifies the extension of belligerent rights against third parties to NIACs. Further, one must always prove on some reasoned basis that such a cross-over of rules from the law of IACs to NIACs has already occurred. 262 The Commission's choice of language makes it clear that it has no firm basis to conclude this has already happened; it simply considers that such development is likely or desirable.
This lack of sustained analysis is most readily explicable on the basis of the Commission's conclusion that the conflict with Hamas is an IAC. Here it relies entirely on the arguments in Targeted Killings outlined above: that an IAC is one which crosses state borders (which appears to be the Commission's view) or one that occurs within occupied territory (an argument it rehearses without accepting that Gaza remains occupied). As before, neither argument is convincing. A conflict with a non-state actor is by definition a NIAC. The Commission acknowledges that this is established as matter of ‘traditional-formal’ law but nonetheless asserts ‘the complexities of modern warfare pose a significant challenge when classifying an armed conflict’. 263 There is, in fact, no difficulty at the level of classification. Confusion arises only if extra-territorial operations against non-state actors are assumed automatically to implicate the laws of IAC or to require some tertium quid that is neither NIAC nor IAC. 264 As discussed above there is no reason for thinking this is necessarily the case. Of course, if the Commission is correct and the conflict is properly considered an IAC, then blockade is available.
On the question of the blockade's compliance with the principle of proportionality the Commission seems in part to base its conclusion on two assumptions: first, that ‘starvation’ can mean only the intentional infliction of hunger as a weapon of war; second, the mere existence of hunger or food insecurity in the Gaza strip does not constitute ‘starvation’ in this sense. 265 Strictly, both propositions are correct, but they are not necessarily to the point. As noted above, the intentional infliction of starvation is not the only test of a blockade's legality. It is also possible that where the civilian population is inadequately supplied with food so as to cause hunger (‘starvation’ in an ordinary sense) this may constitute a disproportionate effect rendering a blockade illegal irrespective of whether this effect was intentional (‘starvation’ as a prohibited means of war). Food insecurity is defined as ‘people lack[ing] secure access to sufficient amounts of safe and nutritious food for normal growth and development and an active and healthy life’; only the strictest interpretation of these words could sustain a view that even in a situation where 61% of households are food insecure no one is necessarily suffering hunger in a manner relevant to proportionality. 266 Not only does the Commission apparently adopt such an approach, it further notes that food insecurity does not mean there was not enough food in the Gaza Strip but rather ‘an inability of the population to purchase this food’. 267 It seems difficult to conclude that the central problem for the population of Gaza is not access to food but the means to buy it without addressing this question conjointly with the impact of the blockade on Gaza's economy. Instead the Commission appears to conclude that so long as Israel ensured there was sufficient food in Gaza no question of ‘starvation’ or hunger as a matter of proportionality can arise. 268 While one must accept that the assessment of proportionality can never be an exact science, this is a narrow view. The Commission does, however, take in principle quite a broad approach to the issue of damage to the civilian economy. It notes that the longer the blockade and land closure regime lasts:
there is a real danger … systematic damage to the economy will result. Therefore it may be assumed that the ability of the civilian population to recover from the blockade after it is removed will be adversely affected the longer it lasts. 269
Further, it concludes ‘there is a danger that comprehensive restrictions may not be regarded as proportionate indefinitely’. 270 Nonetheless, it concludes that any such tipping point has not yet been reached. This view would not appear to be shared by the UN Department of Political Affairs which has stated that, given ‘the scale of the unmet needs’ in Gaza, ‘the blockade is unacceptable and counterproductive and must end’. 271
The Commission also makes a strange argument that individuals have no right to challenge a blockade on the grounds they consider it illegal:
[T]he claim that neutral shipping is free to ignore the existence of a naval blockade for the reason that it prima facie breaches the provisions of article 102 of the San Remo Manual amounts to a claim that the blockade—an act of a sovereign state—is null and void. 272
In the Commission's view there are few cases where state acts ‘are regarded as void—as opposed to illegal or unjust’. 273 The approach taken here is odd at best. Blockade is a practical doctrine based on factual effectiveness: if the blockading state cannot or does not enforce it, it becomes illegal. This recognises the fact that vessels may, at their own risk, attempt to run a blockade. To ask whether LOAC grants a right to individuals to run a blockade (legal or illegal) is unusual: insofar as LOAC contemplates individual civilians its rules are negative and/or protective. The Commission seems really to be making an argument about breach of neutrality. That is, there was some debate historically about whether a ‘right’ to blockade imposed a duty upon neutral states to respect it (and to ensure that their nationals did) or whether blockade involved a ‘clash’ of rights between the belligerent right of blockade and the neutral right to trade (implying no duty of respect). 274 The San Remo Manual addresses the issue in only a limited and practical way, stipulating essentially that certain acts by private persons (amounting to breach of neutrality) render a vessel subject to capture. 275 However, even on the view most favourable to belligerents, the law never directed individuals not to breach a blockade: at most it required flag states or states of nationality to prevent acts breaching neutrality ‘according to domestic laws and regulations already in effect’. 276
The Turkish National Commission of Inquiry Report followed on 11 February 2011 (‘the Turkish Report’). Several points are noteworthy, though they can be treated here only briefly. First, the Turkish Report highlights the fact that Israel ‘does not accept the characterization of its armed conflict with Hamas and other groups in Gaza as international’ but contends that Israel relies upon the ‘international law applicable to international armed conflicts to justify its naval blockade on the Gaza Strip’. 277 Israel's stated position is not quite so stark: it contends it is engaged in an ‘armed conflict’ with Hamas to which many rules of IAC will apply. Nonetheless, the Turkish Report's contention that Israel's legal position allows it selectively to choose the applicable law is sound; equally sound is its contention that blockade is only applicable in IACs. The Turkish Report draws a further conclusion, however, arguing that Israel's refusal to characterise the conflict as an IAC means it is estopped from relying on the law of blockade. 278 The Report would have been on more secure ground to argue that the conflict is not objectively an IAC. The estoppel argument attributes a decisive weight to Israel's characterisation of the conflict that squares poorly with any doctrine other than recognition of belligerency. Second, it argues that the notification of the blockade was defective for not specifying a finite duration and for not clearly stipulating which goods would be allowed entry. 279 The latter point was certainly true of the land ‘blockade’ but is irrelevant in the maritime context where the objective of blockade is to stop the passage of all goods other than humanitarian supplies. Third, it argues that ‘the general lack of access to food and nutrition should be considered to constitute a level of unacceptable starvation’. 280 The thrust of the analysis seems to be that this is a question about disproportionate damage (with which the present writer would agree), not deliberate starvation as a weapon of war. 281 Fourth, the Turkish Report attempts to establish the maritime blockade was not effectively enforced by reference to events occurring prior to its formal declaration (on the basis that there was a de facto blockade in progress) and the arbitrary enforcement of the land ‘blockade’; it also criticises enforcement of the blockade at a significant distance from the blockaded territory. 282 All three arguments fail. The Report cannot logically emphasise the importance of a formal declaration of blockade and then treat the legal regimes before and after that date as one entity. While the enforcement of the land ‘blockade’ might—in context—have a bearing on the proportionality of maritime measures, there is no authority for the idea land-based measures are relevant to assessing the effectiveness of a maritime blockade. The legality of ‘long range’ blockades has been discussed above. Fifth, the Report argues that because the Mavi Marmara was engaged in a humanitarian mission it was absolutely immune from attack and capture. To this end it relies upon the San Remo Manual requirement that deliveries of humanitarian aid to blockaded territory must be allowed where the civilian population is inadequately provided for but fails to note the qualifying belligerent right ‘to prescribe … technical arrangements, including search’ under which this happens. 283 Boarding and diversion for search are thus not necessarily incompatible with the Manual. Sixth, the Report also asserts (without fully explaining either argument) that the blockade constitutes collective punishment and that blockade of occupied territory is a nullity. Both points have been explored above. Finally, the report makes its own argument based on individuals having positive rights under LOAC: that as vessels carrying humanitarian supplies are exempt from capture, those aboard have a lawful right to ‘resist the Israeli attempts to stop, seize and search the ship’. 284 The premise, that humanitarian vessels have an absolute immunity from capture, is obviously flawed. 285 The further argument suggesting an individual positive right under LOAC to resist boarding is flawed for the same reasons outlined in relation to the Turkel Commission's argument that there is no individual right to breach an illegal blockade. Further, it ignores the right of a belligerent to prescribe ‘technical arrangements’ for the delivery of humanitarian shipments which could clearly entail boarding. 286 As explained above, however, none of this affects the individual right of self-defence available in principle to soldiers and civilians alike in times of peace and war.