1. Introduction
At the heart of the Nuclear Non-Proliferation Treaty (NPT or Treaty)Footnote 1 is the idea that only the five nuclear-weapon States—China, France, Russia, the United Kingdom (UK) and the United States (US)—can possess nuclear weapons.Footnote 2 These States are not entitled to transfer their weapons, or control over them, to any other entities and all other States Parties (the non-nuclear-weapon States) must forgo receiving, manufacturing or acquiring them. These propositions are embedded in the first two Articles of the Treaty.Footnote 3 Article I provides:
Each nuclear-weapon State Party to the Treaty undertakes not to transfer to any recipient whatsoever nuclear weapons or other nuclear explosive devices or control over such weapons or explosive devices directly, or indirectly; and not in any way to assist, encourage, or induce any non-nuclear-weapon State to manufacture or otherwise acquire nuclear weapons or other nuclear explosive devices, or control over such weapons or explosive devices.Footnote 4
Article II states:
Each non-nuclear-weapon State Party to the Treaty undertakes not to receive the transfer from any transferor whatsoever of nuclear weapons or other nuclear explosive devices or of control over such weapons or explosive devices directly, or indirectly; not to manufacture or otherwise acquire nuclear weapons or other nuclear explosive devices; and not to seek or receive any assistance in the manufacture of nuclear weapons or other nuclear explosive devices.Footnote 5
While the basic premise of who can and cannot possess nuclear weapons appears clear cut, one of the most divisive issues in the nuclear realm is whether two sets of nuclear sharing practices—the nuclear sharing between the US and its North Atlantic Treaty Organization (NATO) allies, and the nuclear sharing between Russia and Belarus—contravene Articles I and II NPT. Nuclear sharing by a nuclear-weapon State can involve a range of practices, including: stationing its nuclear weapons in a non-nuclear-weapon State; providing a non-nuclear-weapon State with access to nuclear weapons; allowing a non-nuclear-weapon State to participate in decision-making, policy-making and strategising around nuclear weapons; training military personnel in a non-nuclear-weapon State in how to deploy and use nuclear weapons; and allowing non-nuclear-weapon State military personnel to deploy and use nuclear weapons.
The US has engaged in forms of nuclear sharing with its NATO allies since 1954.Footnote 6 At present, it stations nuclear weapons in Belgium, Germany, Italy, the Netherlands and Turkey.Footnote 7 During peacetime, these weapons remain under the custody and control of the US, but NATO policy provides that, during times of conflict, military personnel from the non-nuclear-weapon States can be authorised to deploy and use the nuclear weapons.Footnote 8 To ensure this policy can be actioned effectively, non-nuclear-weapon States’ military personnel are trained in how to handle and use nuclear weapons.Footnote 9 A further component of NATO’s nuclear sharing arrangements is that NATO’s nuclear policies are developed and implemented by the Nuclear Planning Group,Footnote 10 a body made up of members from all NATO States bar France.Footnote 11
Russia’s nuclear sharing arrangements with Belarus are much more recent than those between the US and its NATO allies.Footnote 12 The possibility of Russia entering into nuclear sharing arrangements with Belarus was first raised in early 2022 and, on 27 February of that year, Belarussians voted in favour of adopting a new constitution which paved the way for nuclear sharing by removing a ban on nuclear weapons in Belarussian territory.Footnote 13 Then, on 25 March 2023, Russian President Putin announced that he would deploy tactical nuclear weapons to BelarusFootnote 14 and, the following month, Russian military personnel began training their Belarussian counterparts in nuclear matters.Footnote 15 On 16 June 2023, Putin declared that ‘the first nuclear warheads have been delivered to Belarus, but only the first batch. There will be more. By the end of the summer, by the end of this year, we will complete this work.’Footnote 16 The President of Belarus, President Lukashenko, confirmed in late December 2023 that the transfers of nuclear weapons from Russia had been completed.Footnote 17
Many States Parties to the NPT object to all forms of nuclear sharing. They contend that the stationing of nuclear weapons in the territory of a non-nuclear-weapon State and allowing non-nuclear-weapon State military personnel to undertake nuclear training amounts to a ‘transfer’ of nuclear weapons under Articles I and II NPT and that allowing non-nuclear-weapon States to participate in decision-making involves them exercising control over the weapons, as does allowing military personnel to deploy and use nuclear weapons in times of war.Footnote 18 In response, the NATO States assert that a ‘transfer’ of nuclear weapons only occurs when the weapons are physically in another State and control over the weapons has been transferred.Footnote 19 Consequently, the stationing of nuclear weapons in other States during peacetime—without the transfer of control over the weapons—does not violate the NPT. This position raises questions about the legality of nuclear-weapon States transferring control over nuclear weapons to non-nuclear-weapon States’ military personnel during times of conflict. However, the US has sought to circumvent this issue by contending that the NPT does not apply in times of war.Footnote 20 As for the involvement of nuclear training and decision-making, it holds that the Treaty does not address these matters and, consequently, there is nothing preventing NATO States from collaborating on them.Footnote 21 Russia and Belarus have not released lengthy defences of their nuclear sharing arrangements; rather, they have relied on the fact that the US and NATO engage in comparable practices. For example, in 2023, when justifying the new nuclear sharing practices with Belarus, President Putin stated:
the United States have been doing this for decades. They have long ago deployed their tactical nuclear weapons on the territory of their allied countries, NATO countries … the Americans do this with their allies, deploy on their territory, teach, by the way, their crews, their pilots to use this type of weapons if necessary. We agreed that we will do the same—without violating our obligations.Footnote 22
Discerning which interpretation of Articles I and II is correct is not straightforward. As explored in Section 2, the ordinary meaning of the words in Articles I and II can be interpreted in different ways; there are copious—and at times conflicting—historical records to sort through to try to determine what the parties believed the terms to mean when the NPT was being drafted; and there is a significant amount of subsequent practice to consider.
Over the years, a number of non-lawyers have put forward thoughts as to what Articles I and II mean by drawing on parts of the travaux préparatoires and other historical records. For example, Martin Butcher, Otfried Nassauer, Tanya Padberg and Dan Plesch have conducted extensive research into the history of the relationship between the NPT and NATO’s nuclear sharing programme;Footnote 23 Daniel Khalessi has provided detailed analysis of conversations that took place about the drafting of the NPT within the US Government, amongst NATO allies and amongst delegates to the NPT negotiations;Footnote 24 and William Alberque has delved into the bilateral negotiations concerning the drafting of Articles I and II NPT that took place between the US and the Union of Soviet Socialist Republics (USSR, as Russia then was) in the 1960s.Footnote 25 While their research contains important information about different understandings of Articles I and II during the drafting of the NPT, the authors have (understandably) not applied the rules of treaty interpretation to their factual findings. Consequently, their research does not identify which facts are legally significant or draw any legal conclusions. In addition, there are some parts of the historical record—in particular, aspects of the NPT negotiations which took place in the Eighteen Nation Committee on Disarmament (ENDC)Footnote 26 —that were not considered in these accounts.
Surprisingly few international lawyers have grappled with whether nuclear sharing is permitted under Articles I and II NPT and the two most significant studies have some limitations. Mohamed Shaker penned an extraordinary three-part treatise in 1980—The Nuclear Non-Proliferation Treaty: Origins and Implementation 1959–1979—which provides detailed information about the development of the NPT and its first few years of operation.Footnote 27 While the treatise sets out significant information about the formation of Articles I and II,Footnote 28 it does not evaluate that information in light of the principles of treaty interpretation. Its age is also a limitation, given that there have been four decades of activity since Shaker’s analysis which potentially give rise to subsequent practice for treaty interpretation purposes. In 2021, Mika Hayashi published an article that examined the legality of NATO’s nuclear sharing practices.Footnote 29 While the article applied some of the rules of treaty interpretation to NATO’s nuclear sharing arrangements, it did not look at the ordinary meaning of the Treaty’s terms. Further, and very significantly, it did not engage with any of the primary travaux préparatoires documents, relying instead on secondary accounts of the negotiations. The secondary accounts focused on deliberations between the US and USSR, leaving out important perspectives and nuances put forward by the other negotiating States Parties. Additionally, Hayashi looked only at subsequent practice since 1995, rather than from the Treaty’s first Review Conference in 1975. All of this means that Hayashi’s main conclusion—that the NPT permits nuclear sharing—is questionable.Footnote 30
In order to provide greater insight into the legality of nuclear sharing under Articles I and II NPT, in Section 2 this article engages with the extensive historical records surrounding the two Articles, the secondary literature and the principles of treaty interpretation set out in Articles 31 and 32 of the Vienna Convention on the Law of Treaties (VCLT).Footnote 31 The key issues explored are whether the prohibition on a nuclear-weapon State transferring, and a non-nuclear-weapon State receiving, nuclear weapons—directly or indirectly—and the prohibition on a nuclear-weapon State transferring control over nuclear weapons—directly or indirectly—permit or prohibit the following nuclear sharing activities:
i. a nuclear-weapon State stationing its nuclear weapons in a non-nuclear-weapon State;
ii. a nuclear-weapon State training military personnel from a non-nuclear-weapon State in how to deploy and use nuclear weapons;
iii. nuclear- and non-nuclear-weapon States jointly making decisions and devising strategies around nuclear weapons; and
iv. nuclear-weapon States transferring control over nuclear weapons to military personnel in non-nuclear-weapon States in times of conflict.
Section 2.1 begins by examining the ordinary meaning of the words in Articles I and II in light of their context and the Treaty’s object and purpose. Section 2.2 explores what can be discerned about the scope of Articles I and II in the context of nuclear sharing from the Treaty’s travaux préparatoires. Section 2.3 considers whether there have been any subsequent agreements or subsequent practice that have affected the meaning of Articles I and II since the Treaty was concluded. It is argued that, of the four aspects of nuclear sharing under investigation, only one can be conclusively settled by the principles of treaty interpretation. The ordinary meaning of the NPT’s terms reveals that the Treaty does not cease to exist in times of war, and it is therefore not permissible for nuclear-weapon States to transfer nuclear weapons and/or control over them to non-nuclear-weapon States in times of conflict. This understanding is supported by the travaux préparatoires and subsequent practice. In contrast, it is not possible to determine whether the Treaty forbids or permits the other nuclear sharing practices: the text of the Treaty permits competing interpretations of Articles I and II on these points; the preparatory materials expose the fact that there were differing views about the meaning of the terms at the time the Treaty was concluded; and there has been no subsequent agreement or practice in the last 57 years that has settled these matters. This results in a Gordian knot:Footnote 32 ambiguity is so deeply embedded in Articles I and II that deciphering whether they permit nuclear sharing or not is impossible.
Unfortunately, unlike in the Greek legend, there is no way to slice through the NPT’s Gordian knot.Footnote 33 However, Section 3 of this article considers a different solution, analysing whether the nuclear disarmament obligation in Article VI NPT can provide an alternative way to resolve the question of whether nuclear sharing is permitted under the NPT. It argues that while Article VI did not prohibit nuclear sharing when the Treaty was adopted in 1968, there are compelling reasons to believe that today it does. It is thus concluded that those States that wish to see an end to nuclear sharing would be best placed to focus their argument on the fact that the practice is a violation of Article VI and therefore must be brought to an end to ensure compliance with the Treaty. Section 4 concludes the article with some brief overarching reflections.
2. Is nuclear sharing permitted under Articles I and II NPT?
2.1. Under their ordinary meaning
The starting place for interpreting Articles I and II NPT are the rules of treaty interpretation, as reflected in Article 31(1) VCLT: ‘[a] treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose’.Footnote 34 Thus it must be determined whether the ordinary meaning of the terms in Articles I and II NPT, in their context and in light of the Treaty’s object and purpose, prohibit nuclear sharing. More specifically, it must be determined, first, whether, if the principles reflected in Article 31(1) VCLT are applied, the prohibition on ‘transfer[ing]’ nuclear weapons in the NPT includes a prohibition on stationing nuclear weapons in non-nuclear-weapon States or allowing individuals from non-nuclear-weapon States to access nuclear weapons during training exercises. Second, it is necessary to ascertain whether the prohibition on transferring ‘control’ over nuclear weapons prevents non-nuclear-weapon States from being involved in the decision-making about nuclear weapons. Third, it must be ascertained whether the text of the NPT provides guidance as to whether the treaty applies in times of conflict or not. This section considers these questions and argues that the application of the principles reflected in Article 31(1) VCLT to the third question reveals that the NPT was intended to apply in times of conflict, but that it provides no clear answer in relation to the other two questions.
2.1.1. Stationing of weapons and training of military personnel
Article I NPT prohibits nuclear-weapon States from ‘transfer[ing]’ nuclear weapons to any recipient directly or indirectly and Article II prohibits non-nuclear-weapon States from ‘receiv[ing] the transfer’ of nuclear weapons. The question here is whether stationing nuclear weapons in a non-nuclear-weapon State—that is, the physical placement of nuclear weapons on the territory of a non-nuclear-weapon State without the ownership of or control over the weapons being transferred—amounts to a ‘transfer’ and whether the non-nuclear-weapon State’s military personnel accessing nuclear weapons during training exercises constitutes a ‘transfer’. The Merriam Webster Dictionary offers multiple definitions of the term ‘transfer’, each of which leads to a different conclusion on this point.
One of the definitions provided is ‘to convey from one person, place or situation to another’.Footnote 35 Following the idea that transfer means conveying from one place to another, stationing would fall foul of the NPT as it involves conveying nuclear weapons from the territory of a nuclear-weapon State to the territory of a non-nuclear-weapon State. Further, applying the idea that transfer means conveying from one person to another, allowing a non-nuclear-weapon State’s military personnel access to nuclear weapons during training might fall within this definition. However, a further definition states that transfer means ‘to make over the possession or control of’ something.Footnote 36 Under this understanding of the term, the stationing of nuclear weapons in non-nuclear-weapon States would not be prohibited because the arrangement does not allow the non-nuclear-weapon State to possess or control the nuclear weapons. It is possible that training exercises might be prohibited under this definition, although this would turn on whether the military personnel are allowed to ‘possess’ nuclear weapons and how much control they are granted over the weapons during training.
One piece of information that could potentially sway the interpretation of the word ‘transfer’ in favour of the restrictive position is the fact that one object and purpose of the NPT is to prevent the proliferation of nuclear weapons.Footnote 37 It makes little sense to take a permissive approach to the word ‘transfer’ if the Treaty’s goal is to prevent nuclear weapons from spreading. However, States that engage in nuclear sharing would insist that proliferation only occurs when control or ownership is transferred and, thus, that a permissive interpretation of ‘transfer’ is not contrary to the goal of the Treaty.Footnote 38
Another possible argument in favour of the restrictive approach is the fact that Article I refers to the concepts of ‘transfer’ and ‘control’ separately. This could suggest that the term ‘transfer’ is meant to have a definition that does not include the concept of ‘control’. The difficulty here, however, is that an understanding of the term ‘transfer’ that contains the word ‘control’ does not make a separate reference to ‘control’ superfluous. It is possible that there was a desire to emphasise that there was an intention to prohibit physical transfers of nuclear weapons where control was also transferred as well as situations where control alone (without physical transfer) was handed over. It is thus difficult to conclude from the ordinary meaning of the word ‘transfer’ in the Articles whether nuclear stationing and training are permitted.
2.1.2. Participating in nuclear policy decision-making
Articles I and II NPT prohibit nuclear-weapon States from transferring control over nuclear weapons and non-nuclear-weapon States receiving control over nuclear weapons. What must be determined at this point is whether non-nuclear-weapon States contributing to decision-making about nuclear weapons amounts to exercising ‘control’ over nuclear weapons. The Merriam Webster Dictionary defines control as ‘to exercise restraining or directing influence over’Footnote 39 or ‘to have power over’.Footnote 40
Similarly to the definitions of ‘transfer’, these definitions can be interpreted and applied in different ways. It is possible to argue that a non-nuclear-weapon State that is involved in nuclear decision-making may have a restraining influence over what happens to nuclear weapons because they can advocate for or against particular nuclear policies being implemented. Equally, however, it could be argued that because non-nuclear-weapon States do not have complete influence over what happens to nuclear weapons, they cannot be said to have a ‘directing influence’; their voice is only one among many and will not necessarily determine any particular policy or strategy discussion. There is also uncertainty with the second definition—‘to have power over’—as it is not clear-cut whether this requires complete power over something or whether some shared level of power would be sufficient.
Once again, turning to the object and purpose of the Treaty provides limited assistance as it depends on how one understands the scope of the term ‘proliferation’: does it occur when some level of control over nuclear weapons is transferred or only in the event that complete control is vested in another entity?
2.1.3. Application in times of war
The final issue to consider is whether the NPT applies during times of war: if it does, then NATO’s argument that it would be permissible for non-nuclear-weapon States to assume control of nuclear weapons during times of conflict would fail. On its face, there is nothing in the NPT to suggest that it does not apply in wartime. The only mention of war is in the preamble and the reference there does not suggest that the Treaty would cease in the event of conflict breaking out.Footnote 41 There is no mention of war in the body of the Treaty, let alone any suggestion that war would result in the Treaty no longer applying.
There is, however, a provision which does not mention war but helps to answer this question. Article X(1), which sets out when individual States can withdraw from the Treaty and the process they must follow, states:
Each Party shall in exercising its national sovereignty have the right to withdraw from the Treaty if it decides that extraordinary events, related to the subject matter of this Treaty, have jeopardized the supreme interests of its country. It shall give notice of such withdrawal to all other parties to the Treaty and to the United Nations Security Council three months in advance. Such notice shall include a statement of the extraordinary events it regards as having jeopardized its supreme interests.Footnote 42
The fact that the Treaty envisages individual States withdrawing when ‘extraordinary events’ occur reinforces the idea that is implicit in the rest of the Treaty: that it does not automatically cease to apply in times of conflict. Instead, it is for each State Party to determine if it wants to stop being bound by the Treaty in a time of war.
The conclusion that the NPT does not automatically cease during times of war is supported by the International Law Commission’s (ILC) 2011 Draft Articles on the Effects of Armed Conflicts on Treaties (Draft Articles).Footnote 43 Article 3 provides that ‘[t]he existence of an armed conflict does not ipso facto terminate or suspend the operation of treaties’, and Article 5 states that the ‘rules of international law on treaty interpretation shall be applied to establish whether a treaty is susceptible to termination, withdrawal or suspension in the event of an armed conflict’.Footnote 44 While the Draft Articles are not a binding source of law, the ILC is a highly respected body, and it is therefore noteworthy that the application of these principles confirms the idea that the NPT would not automatically cease to apply in times of war.
2.1.4. Conclusion on the ordinary meaning of the text
From the above discussion it is possible to conclude that the claim by NATO States that the NPT does not operate in times of conflict is not supported by the ordinary meaning of the Treaty’s text. It is less clear, however, whether the Treaty supports or prohibits the stationing of nuclear weapons, the training of non-nuclear-weapon State military personnel in nuclear weapon matters or the involvement of non-nuclear-weapon States in nuclear decision-making. When it is not possible to discern the meaning of a treaty term from the methods of interpretation in Article 31 VCLT, regard may be had to the treaty’s travaux préparatoires, as reflected in Article 32 VCLT.Footnote 45 The following section considers the NPT’s preparatory materials to see what light they shed on the matter.
2.2. Under the travaux préparatoires
As mentioned in Section 1, the NPT negotiations in the 1960s produced a large number of documents. Over the years, academics, civil society leaders and diplomats have dipped into these documents and selected excerpts to support one side or the other in the nuclear sharing debate. There has not, however, been a comprehensive attempt to consider all the material in the negotiating documents in light of the rules concerning what constitutes travaux préparatoires. This section begins by outlining what can be considered under Article 32 VCLT before turning to examine the NPT’s preparatory material in this light. It argues that the materials reveal that there was no agreed upon position amongst the negotiating parties as to whether stationing nuclear weapons in non-nuclear-weapon States, training military personnel from non-nuclear-weapon States and non-nuclear-weapon States’ participation in nuclear decision-making and strategising were permitted or excluded under Articles I and II.Footnote 46
2.2.1. Defining travaux préparatoires
There is no precise definition of what constitutes travaux préparatoiresFootnote 47 but it is well established that only material from the negotiating process ‘during the preparation of the treaty up to its conclusion’ that is ‘apt to illuminate a common understanding of the negotiating parties to treaty provisions’ can be used.Footnote 48 This means that material prepared by just one or two negotiating parties can be used only if it was made available to all negotiators and elicited a common understanding between all negotiators.Footnote 49 Documents that represent the ‘unilateral hopes, inclinations or opinions’ of just one negotiating party are not considered travaux préparatoires.Footnote 50
2.2.2. Meaning of Articles I and II in the travaux préparatoires
In light of the legal test above, it is necessary to determine whether there are materials from the NPT’s travaux préparatoires that clarify what the negotiating parties to the NPT understood as being permitted by Articles I and II in relation to the stationing of nuclear weapons in a non-nuclear-weapon State; the training of non-nuclear-weapon States’ military personnel in nuclear weapons matters; and the involvement of non-nuclear-weapon States in nuclear decision-making.
The materials that are relevant to these issues come from three different stages of the NPT negotiations:
i. Early discussions (between 1965 and early 1967) in the ENDC—the main international forum where the NPT was negotiated.Footnote 51
ii. Bilateral discussions between the US and USSR over the Articles, which took place in 1966 and 1967, and which resulted in the text of the Articles being agreed.
iii. Discussions in the ENDC after the conclusion of the bilateral negotiations between the US and USSR and materials that emerged around the time the NPT was signed and ratified by States Parties, which shed light on how States understood the scope of the provisions to which the US and USSR had agreed.
2.2.2.1. Divergent views in early discussions in the ENDC
In the early stages of the NPT negotiations, it was apparent that there were very different conceptions of what Articles I and II should contain among the States in the ENDC. The approaches can be grouped broadly into three categories: the permissive approach; the restrictive approach; and the highly restrictive approach.
The permissive approach was adopted by the NATO States who advocated for very few limits to be embedded in Articles I and II. They wanted to ensure that these Articles were formulated in a way that allowed them to maintain their existing nuclear sharing practices and preserved the right for them to develop an arrangement—termed a multilateral force (MLF)—that would enable all NATO Member States to participate in the joint control, manning and command of nuclear weapons.Footnote 52 To this end, they advocated for language in Articles I and II that would have prohibited non-nuclear-weapon States from gaining independent control to fire nuclear weapons but would have permitted them to have nuclear weapons stationed on their territory and to participate with other NATO States in the control, command and use of those weapons. This approach was expressed by the UK’s delegate in the following way:
[t]he western delegations have explained what proliferation means in their view – and it is a straightforward, common-sense definition; proliferation occurs when a non-nuclear weapon state or group of states acquires its own independent capability to fire nuclear weapons, without the explicit and concurrent decision of a nuclear weapon state.Footnote 53
The restrictive approach was advanced by the USSR and some of its allies. The Soviets were vehemently opposed to the idea that the NPT would permit non-nuclear-weapon States to jointly own or control nuclear weapons via some form of multilateral body such as the MLF.Footnote 54 They objected to the language put forward by the US on the basis that it would afford non-nuclear-weapon States ‘the possibility of participating in the control of nuclear weapons and in taking decisions concerning these weapons’Footnote 55 and they made it clear that they did not support ‘the right [of non-nuclear-weapon States] to participate in the ownership, disposal and use of nuclear weapons’.Footnote 56 The USSR also opposed the idea that military personnel from non-nuclear-weapon States could have any form of access to nuclear weapons.Footnote 57 While the USSR’s views were much more restrictive than those of NATO, it did not insist on a complete ban on moving nuclear weapons to non-nuclear-weapon States. Instead, it put forward proposals that allowed for the possibility of nuclear weapons being stationed in the territory of non-nuclear-weapon States, provided that the weapons remained under the command and control of a nuclear-weapon State.Footnote 58
The highly restrictive approach was adopted by States from the Non-Aligned MovementFootnote 59 and some of the USSR’s Eastern Bloc allies. These States were opposed to nuclear weapons being moved to non-nuclear-weapon States under any circumstances (including in the form of stationing) and non-nuclear-weapon States having any involvement in nuclear weapon policy or decision-making. For example, Burma argued for ‘watertight non-proliferation’, asking ‘should not the text be favoured that tried to close all possible avenues, to prevent proliferation in the very sense of the word?’Footnote 60 India similarly insisted on a comprehensive ban on the spread of nuclear weapons, asserting that:
the treaty must prohibit all aspects of proliferation … in any form or shape. As a non-aligned nation we are unable to understand why members of military alliances should receive a special dispensation in the context of non-proliferation. There cannot be three categories of nations: nuclear nations, non-nuclear nations in alliance with nuclear nations, and non-nuclear nations.Footnote 61
Similar sentiments were expressed by the Latin American States, which advocated for Article I NPT to say:
the Contracting Parties undertake ‘to use exclusively for peaceful purposes the nuclear material and facilities which are under their jurisdiction, and to prohibit and prevent in their respective territories’ both ‘the testing, use manufacture, production or acquisition by any means whatsoever of any nuclear weapons’ and ‘the receipt, storage, installation, deployment and any form of possession of any nuclear weapons, directly or indirectly, by the Parties themselves, by anyone on their behalf of in any other way’.Footnote 62
Mexico declared this formulation to be ‘undoubtedly one of the most comprehensive ever to have been drafted at world or regional level and [it] certainly does not appear to contain any loophole’.Footnote 63 Further, Poland argued that all forms of disseminating nuclear weapons, including ‘physical’ transfer without necessarily transferring ‘ownership, possession or control’ over them, should be banned to ensure the Treaty’s effectiveness.Footnote 64 Finally, throughout the ENDC negotiations, numerous States referred to the fact that the United Nations General Assembly (UNGA) Resolution imbuing the ENDC with the power to negotiate the NPT noted that one of the main principles which should be respected was that ‘the treaty should be void of any loop-holes which might permit nuclear or non-nuclear Powers to proliferate, directly or indirectly, nuclear weapons in any form’.Footnote 65
The approach which should be taken to Articles I and II proved one of the most contentious issues in the early years of the NPT negotiations and consumed extraordinary amounts of time. By April 1966, it was apparent that there was little hope of agreement being reached in the full committee of the ENDC and it was thus agreed that the US and the USSR should enter bilateral discussions in a bid to reach agreement on the matter.
2.2.2.2. US/USSR bilateral discussions
The US and the USSR conducted a series of bilateral negotiations about the scope of Articles I and II across the second half of 1966 and into 1967.Footnote 66 There are significant questions concerning what was agreed between the two States and whether the understanding they appeared to come to was then communicated to the other negotiating States Parties. To appreciate the points of contention it is helpful to provide an overview of the negotiations and the communications that unfolded.
When the two States entered the bilateral talks, the key sticking points were the interest of the US in developing an MLF and its determination to protect all of its existing NATO nuclear sharing arrangements. Across the course of the negotiations, the US agreed to give up its hopes for an MLFFootnote 67 but it remained steadfast in its commitment to preserve NATO’s nuclear sharing arrangements. Conscious that the Soviets were not prepared to adopt provisions that explicitly allowed for all of its nuclear sharing practices to persist, the US worked on drafting ambiguously worded versions of Articles I and II in the hopes that the right phrasing might placate Soviet concerns while also allowing for NATO’s arrangements to be accommodated.Footnote 68 The text that they settled on and that was incorporated into the final version of the NPT appeared to do precisely this. In particular, as discussed in Section 2.1, the words ‘transfer’ and ‘control’ have multiple meanings that can be used both to support and oppose nuclear sharing practices.
The Americans made it clear in a document sent to the USSR on 28 April 1967 that they believed the agreed-upon drafting of Articles I and II protected their nuclear sharing arrangements. It detailed the understanding that the US was permitted to station nuclear weapons in NATO States and that the treaty would cease to operate in times of war in the following way:
[The Treaty] does not deal with arrangements for deployment of nuclear weapons within allied territory as these do not involve any transfer of nuclear weapons or control over them unless and until a decision were made to go to war, at which time the treaty would no longer be controlling.Footnote 69
It also made it clear that the US did not consider the Articles to prevent NATO States from engaging in nuclear discussions and policy-making, stating that ‘[the Treaty] does not deal with allied consultations and planning on nuclear defense so long as no transfer of nuclear weapons or control over them results’.Footnote 70
While the US interpretation of Articles I and II was clearly set out to the Soviets, it is evident from US Government documents that the USSR never explicitly agreed to these understandings. Indeed, when briefing the US Secretary of Defense about the state of Articles I and II NPT on 10 April 1968, the Under Secretary of State, Nicholas Katzenbach, explained that the US ‘nuclear defense deployment arrangements … are not explicitly sanctioned by Articles I and II, since the USSR was not prepared to provide such an endorsement of NATO arrangements’.Footnote 71 However, Katzenbach went on to say that ‘the Soviets were informed that if they took an official position in opposition to these interpretations, a very serious problem would arise’.Footnote 72
All of this suggests that the USSR can perhaps be said to have acquiesced to the US interpretation of Articles I and II. In international law, a State acquiesces to a position (and is thus bound by it) when it remains silent in circumstances where ‘expressing disagreement or objection in relation to the conduct of another State would be called for’.Footnote 73 Arguably, if the Soviets did not want the NPT to permit NATO’s nuclear sharing arrangements and they were aware the US took the position that it did, it was incumbent upon them to make their objection known. Interestingly, however, even the US was sceptical at the time that the Soviets were bound by their own interpretation, with Katzenbach writing in his letter to the US Secretary of Defense in 1968 that ‘the USSR could not be expected to be bound by unilateral interpretations’ and ‘[w]e have not heard from the Soviets any indication that they will contradict the US interpretations when they are made public … This does not mean that they will necessarily agree with them.’Footnote 74 There is thus some ambiguity as to whether the Soviets can be said to have acquiesced to the US interpretation of Articles I and II.
Even if it is determined that the Soviets acquiesced to the US understanding of Articles I and II and the negotiations between the US and USSR ‘illuminate a common understanding’ between them as to the scope of the provisions, it is not possible to conclude that all other negotiating States were aware of this understanding, let alone supportive of it, as will become apparent in the next section.
2.2.2.3. The ENDC and signing the Treaty
The US and USSR introduced the text of Articles I and II that they had agreed upon to the rest of the ENDC on 24 August 1967. It is clear from archival records that in addition to informing the Soviets of their understanding of Articles I and II, the Americans did share their understanding of the provisions with their NATO allies. Indeed, Katzenbach wrote in his 1968 letter that ‘[w]e worked out interpretations on … Articles I and II with our allies (and in particular the FRG)’.Footnote 75 There is, however, no evidence that all the non-NATO States in the ENDC were aware of how the US understood the Articles, though Shaker states that the US interpretation of Articles I and II was shared with ‘key delegations’ in the ENDC.Footnote 76 However, as set out above, the rules on travaux préparatoires are clear that it is not sufficient for ‘key’ delegations to be aware of a particular textual interpretation; there is a need for all negotiating States to be involved. There is no evidence in the ENDC’s NPT negotiation records that the Americans’ understanding of the provisions was discussed by the full ENDC at any stage.Footnote 77
Furthermore, there are a number of pieces of evidence that suggest that at least some of the ENDC States were not aware of the US interpretation. For example, when expressing support for the content of Articles I and II in August 1967, the Czechoslovak delegate stated that ‘[t]hey leave no loop-holes allowing for a spread of nuclear weapons in any way’,Footnote 78 a position that they would have struggled to adopt if they were aware of the American understanding of the provisions. In 1999, a Swedish diplomat confirmed that Sweden signed the NPT in 1968 without being aware of the US position.Footnote 79 He stated that the idea that non-nuclear-weapon States would have access to and be able to use nuclear weapons during times of war went against Sweden’s reasons for signing the NPT.Footnote 80
Finally, US Government records reveal that the US deliberately decided not to release their interpretation of Articles I and II prior to the NPT being opened for signature on 1 July 1968.Footnote 81 It was only eight days after the Treaty opened for signature (and 56 States had already signed it) that the US publicly released its understanding of Articles I and II during the US Senate’s debate as to whether to ratify the treaty.Footnote 82 The literature on travaux préparatoires makes it clear that understandings of a treaty term that are set out in the domestic legislative documents of an individual State can only be considered when they ‘illuminate a common understanding of the agreement’Footnote 83 amongst the negotiating parties. This is a very difficult test to pass.Footnote 84 A rare example is found in the Oil Platforms case when the International Court of Justice (ICJ) found that, for interpretation purposes, regard could be had to documents conveyed to the US Senate during a treaty ratification process.Footnote 85 However, the documents related to a bilateral treaty and both States Parties to that treaty introduced and relied on the material before the ICJ, suggesting that their contents did indeed represent a common understanding of the relevant treaty provision.Footnote 86 This scenario is thus very different to the Senate debate over the NPT where the provisions in issue were embedded in a multilateral treaty and there was no indication that the American understanding of the terms amounted to a ‘common understanding’ amongst all negotiating States.
In light of the fact that some States in the ENDC were not informed of the US interpretation during the negotiating process and that the US Senate records cannot be read as elucidating a common understanding of the States that negotiated the NPT, it is not possible to conclude that the US understanding of Articles I and II represents the meaning of the terms in 1968.
It must then be asked what the common understanding of Articles I and II amongst the ENDC States was at the time that the NPT was concluded. This question is difficult to answer as the ENDC records do not clearly explain the scope of the two Articles. It is apparent from the records of the ENDC debate following the release of the joint US–USSR agreed text in August 1967 that there was no discussion of whether nuclear-weapon States could transfer nuclear weapons to their non-nuclear allies in times of war, and certainly no agreement on the matter amongst all negotiating States. It seems likely that the non-NATO States in the ENDC assumed that the prohibition on non-nuclear-weapon States controlling nuclear weapons precluded non-nuclear-weapon States from participating in nuclear decision-making or policy development. This is because they had objected vociferously to the idea that any non-nuclear-weapon State could be involved in decisions over nuclear weapons in the early stages of the NPT negotiations,Footnote 87 and if they suspected this was permitted by the final version of Articles I and II, there is little doubt they would have spoken out in protest. It appears, however, that no mention was made of this issue following the introduction of the US–USSR agreed text of the provisions to the ENDC.
What is less clear is whether the non-NATO States believed that the Articles permitted the stationing of nuclear weapons and the training of military personnel from non-nuclear-weapon States. The fact that, as noted above, Czechoslovakia stated that Articles I and II ‘leave no loop-holes allowing for a spread of nuclear weapons in any way’Footnote 88 suggests that it believed the Articles captured the highly restrictive position that many Non-Aligned Members had supported. However, other ENDC States believed (much to their consternation) that the Articles in their final form were not capable of banning stationing or training. For example, when commenting on the final text of the Articles, India stated that ‘[n]o attempt appears to have been made, however, to deal with the question of the transfer of nuclear weapons to and their stationing in the territories of other countries, or with … the training of the armed personnel of non-nuclear nations in the use of nuclear weapons’.Footnote 89 Similarly, in 1968, soon after the Treaty had been opened for signature, Mauritius noted that ‘[t]he Non-Proliferation Treaty had other weaknesses. It did not prohibit the storing or transporting of nuclear weapons within the territories of the parties’.Footnote 90
In light of the discrepancies in State responses to the Treaty, it is difficult to reach a definitive conclusion as to what the common understanding was. Instead, it is clear that different understandings of key provisions of the Treaty were present from the outset. The NATO States believed that their existing nuclear sharing practices were permissible and that the Treaty would cease to operate if a war were to break out. Some non-NATO States believed that nuclear stationing and training were permitted by the NPT, but not other forms of nuclear sharing, and some States believed that all forms of nuclear stationing were forbidden under Articles I and II. The seeds of division and disagreement were thus planted at the Treaty’s inception.
2.3. Subsequent agreements and subsequent practice
The meaning of a treaty’s terms at the time of its conclusion is not the end of the story. Instead, pursuant to the principle reflected in Article 31(1)(3) VCLT, the meanings of terms can evolve over time through subsequent agreements and subsequent practice. This section begins by setting out the rules on subsequent agreements and subsequent practice before considering whether any agreements or practice have arisen over the last five decades that have clarified or altered the understanding of Articles I and II in relation to nuclear sharing practices. It argues that there is a subsequent agreement that has confirmed the understanding that the NPT continues to operate in times of war but that there has been no subsequent agreement or practice that elucidates an understanding amongst all States Parties as to whether other nuclear sharing practices are permitted under Articles I and II. Instead, the divisions that emerged during the NPT negotiations have persisted.
2.3.1. VCLT definition
Article 31(3) VCLT provides that the interpretation of a treaty shall take into account ‘any subsequent agreement between the parties regarding the interpretation of the treaty or the application of its provisions’Footnote 91 and ‘any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation’.Footnote 92 Subsequent agreements encompass any ‘firm agreement’ amongst the States Parties as to the meaning of provisions in a treaty.Footnote 93 Such agreements can be formal treaties but do not have to be; the key requirement is that the States Parties have reached agreement as to the meaning of a term.Footnote 94
Subsequent practice for the purposes of Article 31(3)(b) VCLT comprises conduct of the States Parties in the fulfilment of a treaty’s obligations that reveals a common understanding about the meaning of its terms.Footnote 95 The conduct needs to be ‘concordant, common and consistent’ over time so that a discernible pattern is established.Footnote 96 Isolated acts are not sufficient to amount to subsequent practiceFootnote 97 but it is possible that subsequent practice exhibiting agreement amongst States Parties can be identified when some States Parties to a treaty remain silent or inactive.Footnote 98
Finally, the ILC has determined that a decision adopted by a Conference of States Parties to a treaty may amount to either a subsequent agreement or subsequent practice for the purposes of Article 31(3) VCLT ‘in so far as it expressed agreement in substance between the parties regarding the interpretation of a treaty, regardless of the form and the procedure by which the decision was adopted’.Footnote 99
2.3.2. Relevant subsequent agreements and subsequent practice
Under Article VIII(3) NPT, the States Parties to the Treaty are required to hold a Review Conference every five years to examine the operation of the Treaty. The records of these Review Conferences are helpful in determining whether there have been any subsequent agreements or subsequent practice that affects the interpretation of the Treaty’s terms. Some Review Conferences have resulted in a Final Document being agreed by consensus. Where this has occurred, it can be relevant to have regard to the content of the Final Document to see whether there is language that demonstrates a firm agreement amongst the parties as to the meaning of particular terms in the Treaty. Regardless of whether a Final Document has been agreed, the debates from each Review Conference are useful in examining whether any ‘concordant, common and consistent’ interpretations of terms appear over time that could constitute subsequent practice.
A review of the records of the ten NPT Review Conferences that have taken place to date reveals that the issue of nuclear sharing has not been an issue that has attracted a large amount of time. It has, however, persistently been raised by a handful of States across the Review Conferences and the records reveal that, with the exception of an agreement in 1985 which affects whether the Treaty applies in times of war, there has been no agreement as to whether nuclear sharing practices fall within the scope of Articles I and II. To understand this state of affairs, it is helpful to provide an overview of what occurred at each Review Conference.
The very first NPT Review Conference, in 1975, resulted in a Final Document which stated that ‘obligations undertaken under Articles I and II of the Treaty have been faithfully observed by all Parties’.Footnote 100 This statement, combined with the fact that the US and USSR were openly stationing nuclear weapons in non-nuclear-weapon States at the time, suggests that States Parties to the NPT did not believe that stationing was a violation of Articles I and II. However, there is a need for caution in assuming that all States were of this view. Yugoslavia, for example, stated that the ‘Conference had failed to reach consensus on any substantive issues’ and that this ‘reflected the existence of profound divergencies on fundamental issues’.Footnote 101 It went on to say that ‘the draft Final Document did not faithfully reflect the deliberations of or the positions adopted at the Conference, nor did it contain all the pertinent elements of the proposal made’ and that it would have voted against the Final Document if a vote had been held, but it had decided that it would ‘not stand in the way of consensus provided that its statement was fully recorded’.Footnote 102
In 1980, there was no consensus on a final text and the divergent views as to whether Articles I and II permitted any form of nuclear sharing came to the fore more explicitly in the debates. While some declared that the NPT States Parties were in compliance with Articles I and II (again suggesting that nuclear sharing practices were not contrary to the provisions embedded in those Articles),Footnote 103 others raised doubts and questions. Morocco, for example, said that ‘[i]t was difficult not to consider the transfer of nuclear technology and equipment and the deployment of nuclear weapons on the territory of third States or in international waters as contrary to article I’.Footnote 104 Other States expressed qualms at the existence of nuclear sharing without directly stating that they violated Articles I and II. For example, the People’s Republic of the Congo noted that the safeguarding of the security of non-nuclear-weapon States ‘should lead to the dismantling by [the nuclear-weapon States] … of their military bases on foreign territory’.Footnote 105 It added that this ‘was felt by most States to be a legitimate demand’.Footnote 106
Unlike in 1980, the 1985 Review Conference did reach a consensus agreement. With respect to Articles I and II, the Final Declaration provided that:
The Conference acknowledged the declarations by nuclear-weapon States Party to the Treaty that they had fulfilled their obligations under Article I. The Conference further acknowledged the declarations that the non-nuclear weapons States Party to the Treaty had fulfilled their obligations under Article II. The Conference was of the view therefore that one of the primary objectives of the Treaty had been achieved in the period under Review.Footnote 107
At first glance, these clauses appear to endorse the idea that nuclear sharing practices at the time were not in violation of Articles I and II. However, the language is somewhat equivocal. It only ‘acknowledged’ the States Parties’ declarations that they had complied with their obligations. It did not accept those declarations. Further, the phrase ‘of the view’ is not particularly strong, definitive language.
Looking behind the Final Declaration at the debates that preceded it, it becomes apparent that the pre-existing divisions over whether nuclear sharing is consistent with Articles I and II persisted. As with previous Review Conferences, numerous States believed that the Articles permitted nuclear sharing.Footnote 108 However, Peru stated that it was unclear whether Article I had been implemented by the nuclear-weapon States and highlighted the fact that nuclear-weapon States ‘had stationed nuclear weapons in the territory of other States parties, with a possibility, and in some cases the declared intention, of transferring control over those weapons to the receiving States in the event of an armed conflict’.Footnote 109 Further, Sri Lanka determined that it ‘was concerned at the stationing of nuclear weapons on the soil of non-nuclear-weapon States in pursuance of military alliance obligations, for the act of transfer itself was precluded by article I and receipt of the transfer was a breach of article II’.Footnote 110 It stated that it ‘would also like to see its understanding that article I precluded transfers between nuclear-weapon States embodied in a final document of the Conference’.Footnote 111
Yugoslavia also spoke out against nuclear sharing. It said that ‘the horizontal proliferation of [nuclear] weapons through their deployment in the territories, waters or airspace of some non-nuclear weapon States as well as in international waters … could only harm the credibility of the Treaty’.Footnote 112 Additionally, it submitted a Working Paper designed to ‘reaffirm the main obligations of the States parties in respect of articles I, II and VI of the Treaty and to propose measures for fulfilling those obligations’.Footnote 113 A key part of that Working Paper was a request for nuclear-weapon States to conclude a treaty to halt the future deployment of nuclear weapons in the territories of non-nuclear-weapon States and secure the withdrawal of nuclear weapons that were currently in such territories.Footnote 114
One further point of note from the 1985 Final Declaration is that it stated:
The Conference agreed that the strict observance of the terms of Articles I and II remains central to achieving the shared objectives of preventing under any circumstances the further proliferation of nuclear weapons and preserving the Treaty’s vital contribution to peace and security, including to the peace and security of non-Parties.Footnote 115
The use of the words ‘under any circumstances’ suggests that the States Parties were affirming that the obligations in Articles I and II apply at all times, including in times of war. This underscores the idea that was agreed during the drafting of the Treaty that, contrary to NATO’s nuclear doctrine, it would be impermissible for the US to transfer command and control of nuclear weapons to non-nuclear-weapon States in the event that war broke out.
The Review Conference that took place at the end of the Cold War, in 1990, failed to achieve a final agreement. As with previous debates, there was a divergence of opinion as to whether nuclear sharing was permitted under the Treaty. Venezuela, for example, stated that ‘articles I and II of the Treaty, read in conjunction with each other, did not explicitly prohibit a nuclear weapon State from establishing in the territory of a non-nuclear weapon state nuclear weapons which remained under its jurisdiction and control’,Footnote 116 while the Democratic Republic of Korea believed that the objectives of the Treaty required States to give up nuclear sharing arrangements. It urged States to make the most of the period of ‘détente and co-operation’ to adopt ‘decisive measures to attain the Treaty’s fundamental objectives’.Footnote 117 At the top of the list of measures it asserted were required for proper fulfilment of the Treaty was that ‘the nuclear-weapon States parties to the NPT must be prohibited from deploying their weapons outside their territory’.Footnote 118
Once again, there was no agreement on a final text in 1995 and tensions between competing viewpoints on Articles I and II continued to simmer. The NATO States insisted that the nuclear sharing they engaged in was permitted under the NPT. In the words of Germany, ‘[t]here had never been any transfer whatsoever of nuclear material or know how for military purposes’Footnote 119 to it as the US had always maintained command and control of the weapons on German territory. A number of non-nuclear-weapon States, however, questioned this position. The Philippines challenged the Netherlands as to whether it was a nuclear- or non-nuclear-weapon State, emphasising that ‘[f]or its part, the Philippines was a non-nuclear weapon State which complied with the provisions of article II’.Footnote 120 Egypt and Tanzania both stated that Article I was unclear,Footnote 121 with Tanzania arguing that ‘the deployment of nuclear warheads in the territory of non-nuclear weapon States would represent a violation of article I if it were interpreted as a transfer of nuclear weapons’.Footnote 122 The confusion surrounding the issue prompted Mexico to state that it ‘would like to clarify some matters relating to articles I and II of the Treaty, in particular the question of the transfer of nuclear weapons or other nuclear explosive devices’.Footnote 123
The division over nuclear sharing’s compatibility with Articles I and II persisted into the twenty-first century. The two Final Documents that were approved by consensus after the turn of the century, in 2000 and 2010, did not directly address the issue one way or another but instead simply noted that the parties reaffirmed their commitments to the principles within the Articles.Footnote 124 However, the debates that took place in the Review Conferences from 2000 onwards reveal the ongoing split in opinion. The NATO States maintained that the non-proliferation provisions permitted nuclear sharing while various other States opposed the idea.Footnote 125 For example, in 2005, Libya argued that the Review Conference ‘should highlight the importance of the full observance of articles I and VI of the Treaty. Nuclear-weapon States should be called upon not to share or export nuclear technology or know-how except for peaceful purposes.’Footnote 126 Further, the Non-Aligned Movement stated that ‘[n]uclear-weapon States must refrain from nuclear sharing for military purposes under any kind of security arrangements’.Footnote 127 This was echoed by China, which said that ‘States should withdraw and repatriate all nuclear weapons deployed outside their own territories, abandon “nuclear umbrella” and “nuclear sharing” policies and practices’.Footnote 128
In 2010, Iran said that ‘[t]he nuclear-weapon States should comply with their obligations under article I by refraining from nuclear-sharing, under any pretext, including security arrangements or military alliance’.Footnote 129 It reiterated its concerns in 2015, arguing that ‘[n]uclear-weapon-sharing … was a clear violation of the concerned States parties’ explicit obligations under articles I and II of the Treaty’.Footnote 130
In 2015, Russia also expressed its view that nuclear sharing contravened the NPT. The Russian delegate argued that his Government considered ‘the nuclear-sharing missions of the North Atlantic Treaty Organization (NATO) … a blatant violation of articles I and II of the Treaty’.Footnote 131
Opposition to nuclear sharing being compatible with Articles I and II continued at the 2022 Review Conference. There, the Non-Aligned Movement stated:
In the view of the Group, any horizontal proliferation of nuclear weapons and nuclear weapon-sharing by States Parties constitutes a clear violation of non-proliferation obligations undertaken by those Nuclear Weapon States (NWS) under Article I and by those Non Nuclear Weapon States (NNWS) under Article II of the Treaty. The Group therefore urges these States parties to put an end to nuclear weapon-sharing with other States under any circumstances and any kind of security arrangements, including in the framework of military alliances.Footnote 132
Russia stated that ‘[t]here are U.S. nuclear weapons on the territory of non-nuclear bloc allies. Its practical use is being exercised with the involvement of non-nuclear members of the bloc. Such actions … are contrary to Articles I and II of the NPT’.Footnote 133
While not explicitly claiming that nuclear sharing was contrary to Articles I and II, China spoke out strongly against it, declaring that nuclear sharing ‘has never been universally and clearly recognized by all the parties’ and that ‘it runs counter to the purposes and principles of the Treaty, and is itself nuclear proliferation’.Footnote 134
It is apparent from the above exposition that there has been very little agreement or ‘concordant, common and consistent’ practice over time from the States Parties to the NPT as to whether nuclear sharing practices are permissible under Articles I and II. The one element of agreement displayed is found in the 1985 Final Declaration, which suggests that States are agreed that the Articles continue to operate in times of war.
2.4. Conclusion on nuclear sharing practices
The application of the principles of treaty interpretation reflected in the VCLT to Articles I and II NPT reveals a deep ambiguity in their scope and uncertainty as to the extent to which they support or prohibit the various types of nuclear sharing practices. Having regard to the ordinary meaning of Articles I and II in their context and in light of the NPT’s object and purpose does not provide a clear answer in relation to the stationing of nuclear weapons in non-nuclear-weapon States, the training of military personnel from non-nuclear-weapon States or the participation of non-nuclear States in nuclear decision-making, strategising and planning exercises. Furthermore, neither the travaux préparatoires nor the records of debates or Final Documents from NPT Review Conferences provide any greater clarity. Instead, they reveal the deep divisions between States Parties as to whether different nuclear sharing practices are permissible under the Articles.
The only area where the application of the principles reflected in Articles 31 and 32 VCLT to the NPT does provide some clarity is with respect to the war issue. The ordinary meaning of the Treaty’s terms, the travaux préparatoires and the subsequent agreement from the 1985 NPT Review Conference all demonstrate that NATO’s assertions that the Treaty ceases to operate in times of conflict, and that the US can transfer command and control over nuclear weapons to non-nuclear-weapon States in the alliance, cannot be supported.
3. The disarmament obligation in Article VI NPT
While most of the legal discussions surrounding nuclear sharing have focused on Articles I and II, there are other provisions that are relevant when considering the legality of these practices. Most significantly, the nuclear disarmament obligation in Article VI should be considered. This section begins by setting out the scope of Article VI and how it has been interpreted before considering its implications for nuclear sharing practices.
3.1. Scope of Article VI NPT
Article VI provides that ‘[e]ach of the Parties to the Treaty undertakes to pursue negotiations in good faith on effective measures relating to … nuclear disarmament’.Footnote 135 Precisely what this obligation entails has been the subject of considerable debate,Footnote 136 and its vague terms have enabled the nuclear-weapon States to justify a lack of progress towards nuclear disarmament.Footnote 137 However, despite efforts by the nuclear-weapon States to undermine Article VI, there is strong support from the ICJ and international law literature for the proposition that it entails concrete obligations to achieve nuclear disarmament.
The ICJ determined in its Advisory Opinion on the Legality of the Threat or Use of Nuclear Weapons that ‘[t]here exists an obligation to pursue in good faith and bring to a conclusion negotiations leading to nuclear disarmament in all its aspects under strict and effective international control’.Footnote 138 In other words, Article VI contains both an obligation of conduct (to take part in negotiations) and an obligation of result (to achieve nuclear disarmament).Footnote 139
Some scholars have questioned whether the ICJ went too far in reading an obligation of result into Article VI,Footnote 140 and have suggested instead that a plain meaning interpretation of the provision simply requires States Parties to focus on engaging in good faith negotiations.Footnote 141 The concept of good faith negotiations is well known in international law and was explored by the ICJ in the North Sea Continental Shelf cases of 1969. The Court held that good faith negotiations mean that:
The parties are under an obligation to enter into negotiations with a view to arriving at an agreement, and not merely to go through a formal process of negotiation … they are under an obligation so to conduct themselves that the negotiations are meaningful, which will not be the case when either of them insists upon its own position without contemplating any modification of it.Footnote 142
Good faith negotiations have further been held to require States to ‘move forward simultaneously’Footnote 143 and to ‘make every effort … to reach a mutually satisfactory compromise, even going so far as to abandon previously inflexibly held positions’.Footnote 144
Interestingly, in many respects there is very little difference in what these two interpretations of Article VI achieve over time. While the ICJ’s approach explicitly requires States to achieve nuclear disarmament at some point, if States are properly engaging in good faith negotiations under the second approach and focusing on reaching agreement, ensuring negotiations are meaningful, compromising with one another and moving forwards, they will also eventually reach a point where nuclear disarmament is achieved. It is thus clear that both interpretations of Article VI require States Parties to be making meaningful progress towards nuclear disarmament until it is achieved.
3.2. Applying Article VI NPT to nuclear sharing practices
The requirement under Article VI for States Parties to the NPT to be making progress towards nuclear disarmament has significant implications for all of the States Parties engaged in nuclear sharing practices. In relation to the NATO States engaged in the practice, it does not mean they were under an immediate obligation upon joining the NPT to give up these practices. However, it does mean that they must be able to show that from the point they joined the Treaty they have been progressively taking steps towards nuclear disarmament. Arguably, if this obligation was being met, the NATO States would have long ago reached the point where they needed to start giving up their nuclear sharing practices to demonstrate their good faith movement towards the goal.Footnote 145 Indeed, a number of States have made this point over the years in Review Conferences. For example, as early as 1975, Yugoslavia said ‘[t]o initiate the process of gradual nuclear disarmament, nuclear weapon States [must] remove tactical nuclear weapons from foreign territory as soon as possible’. More recently, in 2010, China argued that States should ‘abandon the policy and practice “nuclear umbrella” and “nuclear sharing”’ so that they could ‘promote nuclear disarmament’.Footnote 146 And, in 2022, prior to re-engaging in the practice itself, Russia stated that NATO’s nuclear sharing arrangements ‘generally hamper nuclear disarmament efforts’.Footnote 147
Article VI also raises difficulties for the nuclear sharing arrangements between Russia and Belarus. Introducing nuclear sharing practices in 2023 was a step away from nuclear disarmament, as it involved spreading the reach and influence of the weapons rather than diminishing them. Given that Article VI requires States to work progressively towards nuclear disarmament, Russia and Belarus’s actions can be seen as a violation of the provision.
4. Conclusion
The argument that I have put forward in this article about the interaction between Articles I and II NPT and nuclear sharing was not the one that I had expected to make. I had always read Articles I and II as prohibiting the practice. On reflection, however, I should not have been surprised that the wording of the Articles is in fact ambiguous and that the history of the Treaty—from the earliest days of the ENDC negotiations through to the most recent Review Conference—reveals deep contrasts in how the words are understood. This is a result of the fact that the NPT was drafted to accommodate the nuclearism of the nuclear-weapon States (and their allies) while paying lip service to the anti-nuclear ideals of other States.Footnote 148 The ambiguity that is woven into the fabric of the Treaty makes it exceedingly difficult for those States and civil society organisations that want to see progress made towards a nuclear-free world to undo the Gordian knot of the Treaty and gain traction on their agendas. Instead, they are pulled into endless arguments about the meaning of the Treaty’s terms.
In light of the uncertainty that underlies Articles I and II, there is perhaps little to be gained by seeking to resolve the nuclear sharing debate through continuing to litigate the meaning of these provisions. Instead, those States that are opposed to the practice would be advised to challenge it under Article VI NPT. As noted in Section 3, all States Parties are required by Article VI to work progressively towards nuclear disarmament. If the NATO States Parties to the Treaty had been fulfilling this obligation across the decades and continuously taking steps towards this goal, they ought to have at least made a start in dismantling their nuclear sharing practices, if not given them up completely. Furthermore, if Russia and Belarus were complying with their Article VI obligations, they would not have entered into their 2023 nuclear sharing arrangements at all.
In closing, it is important to note that the issues explored in this article are likely to become increasingly significant in the coming years. There is growing pressure for the establishment of new nuclear sharing initiatives. In particular, both Japan and the Republic of Korea have expressed a strong interest in entering nuclear sharing arrangements with the US to protect against threats from North Korea, Russia and China.Footnote 149 If Japan and the Republic of Korea (and for that matter any other non-nuclear-weapon States Parties to the NPT) want to maintain their good standing within the NPT, and not contravene their clear obligations under Article VI, then ideas of nuclear sharing need to be abandoned and alternative means of achieving security pursued.
Acknowledgements
I am incredibly grateful to Arianna Bacic and Katie Marshall for their wonderful research assistance. My thanks too for the very helpful comments and insights provided by Peter Hayes.
Open access
