Russia-Ukraine War

The Black Sea Grain Initiative and the Law of Treaties: A Response

By Gregor Novak
Tuesday, May 9, 2023, 8:31 AM

In a recent piece on Lawfare, Mykhailo Soldatenko offered an insightful analysis of the Black Sea Grain Initiative (BSGI), as well as the separate memorandum of understanding (MoU) between Russia and the United Nations (UN) Secretariat. It is excellent to see legal scholars continue to grapple with this important topic, including in the form of Soldatenko’s critical engagement with a contribution by Helmut Aust and myself, published in Just Security on Nov. 10, 2022. Because Soldatenko explicitly referred to some of the points we made in that piece, I wish to respond by recalling some of our initial arguments. This response should not be read as an update or elaboration of our earlier article, which must be left for another occasion. Further, while I am the sole author of this response, Aust has reviewed and endorsed the points developed in this response.

Soldatenko begins his contribution by referring to a recent difference among the parties to the BSGI about the interpretation and application of its paragraph H (or 5.H.), which lays out the automatic renewal of the agreement for 120 days, “unless one of the Parties notifies the other of the intent to terminate the initiative, or to modify it.” Later on, Soldatenko argues that the BSGI is most likely not a legally binding agreement, while claiming that Aust and I had found that it was “based primarily on [the] specific” use of the “mandatory modal verb ‘shall.’” According to Soldatenko, and in light of the doubts about the BSGI’s status as a treaty, the agreement’s continued existence hinges not on international law, but on other factors, summarized in the final paragraph of the piece. In contrast, Soldatenko suggests that we had claimed that the law of treaties had an “influence on the parties’ loyalty to the deal.” While Soldatenko makes a number of important observations in his piece, several of which I am both in agreement with and believe offer valuable insight, I consider that they merit a few clarifications.

The Role of the Law of Treaties for the Black Sea Grain Initiative

First, the focus of our Just Security piece was on the role of the law of treaties as a “conceptual and argumentative framework” for the management of treaty relations. We did not seek to examine the various underlying factors sustaining the BSGI, something that requires a transdisciplinary prism. Nonetheless, in our piece, we did recognize that the BSGI stands on a number of pillars—humanitarian, economic, and political—that have little or nothing to do with the law of treaties.

Second, we did not come to our conclusion about the BSGI’s likely legal nature “primarily” based on the use of the modal verb “shall,” as Soldatenko claims. Rather, like Soldatenko, we offered various separate arguments rooted in the customary law of treaties, acknowledged the uncertainties surrounding the parties’ positions, pointed to the inherent limits of our analysis, and drew on information openly available at the time of publication.

Third, we did not argue that the law of treaties itself had an influence on the “parties’ loyalty to the deal,” as Soldatenko argues, but made a more nuanced claim. After examining the legal nature of the BSGI, our focus was on the role of the law of treaties at moments when the BSGI’s continued existence was under acute pressure, notably in the context of Russia’s short-lived attempt at claiming a right to “suspend” the initiative and its claim of a unilateral right of “withdrawal,” neither of which is foreseen in the text of the BSGI. At the same time, we emphasized how the lack of a dispute settlement clause in the BSGI shapes the role that international law can play in such scenarios.

Fourth, in our piece, we argued why the BSGI is a treaty and suggested how the law of treaties likely “played a significant background role in stabilizing” it. Specifically, we argued that the law “provided a conceptual and argumentative framework for treaty-making and the management of treaty relations in the most difficult of circumstances.” At the same time, we acknowledged that we did not know “whether constraints reflected in the law of treaties played a role in [the mediators’] efforts” and therefore referenced “publicly available information [that pointed] to the role of formal and legal arguments” in that context. 

The crucial point is that even if one were to accept all of Soldatenko’s reservations about the legally binding nature of the BSGI (several of which we also raised in our piece but did not find to be determinative), the law of treaties may still have played precisely the role we attributed to it. After all, Soldatenko himself engages in interpretations of the BSGI, refers to concepts like the purported “suspension” of the BSGI, invokes the parties’ practice in implementing the BSGI, and mentions Russia’s claim of a unilateral right to “exit” the BSGI (in this context, Russian President Vladimir Putin also used the notion of a “right to withdraw”). These phrases are part and parcel of the language of the law of treaties. As long as there is some uncertainty about the legal nature of the BSGI, and Soldatenko is right to recall the continued uncertainty on that issue, the law of treaties is indeed liable to affect its application and interpretation, if only as the only available set of standards to guide similar scenarios covered by agreements, be they purely political or also legally binding.

Additional Points for Further Reflection and Debate

Soldatenko makes a number of other points that are related to our arguments about the relationship between the BSGI and the law of treaties that merit closer scrutiny, even if they are of a broader ambit.

First, Soldatenko on several occasions points to what he calls the “flexible” framing of the BSGI and speculates that “the parties likely wanted more flexibility than is usually available under rigid treaty law rules.” In other words, clarifying the BSGI’s status as a treaty would have entailed “rigid treaty law rules” and, therefore, the parties chose to endow the BSGI with a legally nonbinding “political” character. 

As an initial matter, relatively little is known in the public domain about the legal aspects of the parties’ discussions of the BSGI, be it during the negotiations about its conclusion, during discussions of its two extensions so far, or during other contacts between the parties. On the face of it, however, ample flexibility seems to be built into the BSGI already. Notably, it foresees a period of effectiveness of only 120 days, which is considerably shorter than the time required to plant and harvest most crops whose export from Ukraine was one of the main motivators for the BSGI. The deal also established the Joint Coordination Center, tasked with conducting “general oversight and coordination” of the BSGI, which allows the parties’ representatives to address day-to-day challenges in the BSGI’s operation flexibly. Moreover, while the various rules and principles of the (customary) law of treaties may indeed be relatively rigid, they have demonstrated their flexibility in many other contexts, for example, by supplying recognized rules of treaty interpretation and general principles of the law of treaties.

More importantly, in the case of the BSGI, little suggests that all parties were seeking more flexibility than is already offered in the text of the agreement. Rather, it appears that it was only Russia that repeatedly sought to challenge the BSGI’s plain text. For example, after we wrote our piece, Russia effectively claimed a right to unilaterally modify the BSGI. This attempt does not appear to have been endorsed either by the other parties or by the UN Secretariat. None of them signaled that the BSGI had been modified, even though the UN Secretariat took no clear position on the precise period of the latest extension, as Soldatenko also notes. The UN Secretariat’s, Turkey’s, and Ukraine’s reactions to Russia’s behavior suggest that the various “flexibilities” invoked by Russia were in tension with the text of the BSGI, even if one were to consider it a purely political agreement. While states do often conclude purely political agreements to maintain a degree of flexibility and conclude treaties with elastic provisions (or provisions that are purely hortatory), the particular instances in which Russia tried to suspend the BSGI or invoke a unilateral right of withdrawal or modification are not apt examples of the type of “flexibility” all parties would have desired. From that perspective, the claim that the BSGI is likely not legally binding should ideally stand on firmer factual ground than currently seems to be offered.

Second, Soldatenko invokes the fact that the parties have “referred to the agreement using a nonbinding vernacular” in support of his assessment of the legal nature of the BSGI. In this context, he also gives the example of a statement by Putin calling the BSGI a “deal” in Russian. The BSGI has indeed been referred to in various ways by all parties, the UN secretary-general, as well as observers. At the same time, Russia’s leadership has regularly invoked the BSGI as part of a “package deal,” consisting of the BSGI and its separate MoU with the UN Secretariat, which would go a long way toward explaining why Russian officials sometimes choose the language of “deal” rather than “treaty.” More importantly, the parties have in fact referred to the BSGI also with “vernaculars” customarily used in the context of the law of treaties, for example, when Russia attempted to suspend its participation in the BSGI or invoked a “right of withdrawal.” While this multiplicity of labels only adds to the uncertainty surrounding the BSGI’s legal status, a party’s post-hoc qualification of an agreement is irrelevant to its legal qualification, especially in a situation in which that party has taken several actions at odds with its commitments under it.

Third, Soldatenko’s argument that the BSGI is likely only a political agreement allows him to take license with the interpretation of the BSGI that would otherwise be constrained by the customary rules of treaty interpretation. As a treaty, the BSGI would need to be interpreted by reference to the rules of treaty interpretation, including those that take into account the subsequent practice of the parties in the application of the BSGI. Interpreting the BSGI without such a framework would call for even greater circumspection. Instead, Soldatenko’s conclusions about the implications of the parties’ ambiguous actions and statements following its most recent extension seem to skew to one side. 

Fourth, even if one were to accept Soldatenko’s argument that the BSGI is likely not a treaty, this would neither imply that international law had no effect on the BSGI and its implementation nor mean the end of the necessary legal inquiry. Rather, it would call for examining whether the BSGI could have had any other effects under international law and what their implications could be—for example, whether a party’s commitments under the BSGI could qualify as a unilateral act or whether a party’s nonadherence to a (legally nonbinding) agreement concluded in the context of the International Convention for the Safety of Life at Sea implicated any obligations under that latter convention, such as the duty to perform it in good faith. 


A discussion of the BSGI’s legal nature may appear like a tangential question in the face of Russia’s ongoing war of aggression against Ukraine, whose effects on global food security form the backdrop to the UN secretary-general’s efforts at supporting the BSGI—be it a “deal” or a treaty. At the same time, the legal dimensions of the BSGI offer valuable insights into the actual and potential roles of international law at times of crisis. Soldatenko’s recent contribution to Lawfare was an insightful addition to the debate over the legal nature of the BSGI and some of its implications, which Helmut Aust and I had discussed in Just Security last year. As outlined above, in my view, Soldatenko’s discussion of our article did not reflect several important nuances of our argument. I hope this response provides the author and other interested readers with helpful context and some clarifications. 

Editor’s Note: The views expressed in this article are personal and in no way reflect the views of any EU institution