Article

Between Rhetoric and Effects: The ICJ Provisional Measures Order in South Africa v. Israel

| Written By:

In its order on provisional measures, the Court appeared to engage in transactional justice. It used harsh rhetoric to describe the catastrophic situation in the Gaza Strip while minimizing the impact of the order by adopting ambiguous language on the contents and scope of the plausibility test.

The provisional measures order in the Genocide Convention case brought by South Africa against Israel, which was issued by the International Court of Justice (ICJ) on January 26, 2024, was accepted by both parties to the case with a degree of satisfaction. South Africa viewed the order as vindicating its decision to legally challenge Israel’s military campaign in Gaza under the terms of the Genocide Convention. Israel, by contrast, was relieved that the Court did not order it to halt the military campaign and noted positively the Court’s call for the immediate and unconditional release of the Israeli hostages held in Gaza. It therefore appears as if the Court succeeded in striking a delicate balance between the competing expectations of the two parties. As we discuss below, however, in order to reach such an outcome the Court adopted a strategy of legal ambiguity in fact-finding and in application of law. The ambiguous nature of the decision might result in the order having a limited impact on the actual conduct of the Israel-Hamas war.

We discuss below three aspects of the order that are especially ambiguous: (1) The plausibility analysis, where the Court did not clearly explain how it reached its decision and the scope of its holding on that issue; (2) the treatment of Israel’s use of force (self-defense) claims, which the Court largely avoided; and (3) the actual measures ordered, which provide Israel with limited guidance as to what aspects of its policy it should change, if at all. That said, there is a residual set of legal expectations that the Court’s order creates, and we identify as best we can what that is.

Plausibility

Given the consent-based jurisdictional structure of the Court, South Africa could bring a claim against Israel in relation to its military campaign in Gaza only under the Genocide Convention – due to the Convention’s compromissory clause – to which both States are parties. Consequently, South Africa had to present its claims as genocide claims, notwithstanding the fact that other legal frameworks – most obviously, international humanitarian law (IHL) – seem more appropriate for evaluating the legality of Israel’s military campaign and the difficult questions it raises about the balance struck between military necessity and the protection of humanitarian interests. (The two dissents by Judge Sabutinde and Judge ad hoc Barak made the point that the real dispute is about the interpretation and application of IHL.)  Further complicating things for South Africa is the exceptionally high evidentiary threshold required by the ICJ for proving genocide claims – establishing specific genocidal intent on the basis of fully conclusive evidence (cf.  Bosnia-Herzegovina v. Serbia, para. 209; Croatia v. Serbia, para. 178). This renders its case on the merits according to some commentators something of a long shot.

Still, for the purposes of issuing provisional measures, it is enough for the Court to establish that the rights asserted in the claim are at least plausible. The Court has not clearly articulated, to date, how plausibility is defined and how is it to be applied. The reasonable prospect of success standard articulated by Judge Christopher Greenwood has not been invoked, at least not explicitly, by other judges; still, the Court did appear to suggest in previous cases that a finding of plausibility should follow “a careful examination of the evidence and arguments presented” (see e.g., Costa Rica v. Nicaragua (2011), para. 58), and that the threshold of plausibility is not affected by the severity of the allegations raised in the case (Gambia v. Myanmar, para. 56). Furthermore, the Court did indicate that the test of plausibility must be applied to key elements of the definition of the right (see Ukraine v, Russia (2017), para. 75: “in order to determine whether the rights for which Ukraine seeks protection are at least plausible, it is necessary to ascertain whether there are sufficient reasons for considering that the other elements set out in Article 2, paragraph 1, such as the elements of intention or knowledge noted above (see paragraph 74), and the element of purpose specified in Article 2, paragraph 1 (b), are present”).

In the Gambia v. Myanmar case, the Court relied on detailed UN Fact Finding reports that explicitly identified reasonable grounds to infer genocidal intent (para. 55), and the judges were therefore able to assert that the claims made by Gambia, including those premised on establishing special genocide intent, were plausible:

“In the Court’s view, all the facts and circumstances mentioned above … are sufficient to conclude that the rights claimed by The Gambia and for which it is seeking protection — namely the right of the Rohingya group in Myanmar and of its members to be protected from acts of genocide and related prohibited acts mentioned in Article III, and the right of The Gambia to seek compliance by Myanmar with its obligations not to commit, and to prevent and punish genocide in accordance with the Convention — are plausible.” (para. 56, emphasis added)

Note that the phrase “the rights claimed” used to describe factual plausibility seems to suggest that the Court found as plausible all of the rights claimed by Gambia. This language is markedly different from the language used in para. 30 of the same order to describe whether the acts alleged by Gambia fall within the scope of its jurisdiction. There the Court used a more careful formulation: “In the Court’s view, at least some of the acts alleged by The Gambia are capable of falling within the provisions of the Convention” (emphasis added; note that similar language is found in para. 30 of the South Africa v. Israel order, which deals with the Court’s prima facie jurisdiction over the case).

By contrast, in the South Africa v. Israel case, the Court did not have before it a detailed fact-finding report establishing possible genocidal intent. Instead, it was presented with  circumstantial evidence submitted by South Africa relating to the catastrophic humanitarian situation in the Gaza Strip and statements by Israeli officials which could be construed as genocidal in nature. Israel presented counter-evidence pointing to the responsibility of Hamas for the humanitarian crisis, elaborating its harm-mitigation efforts, and indicating that some of the public statements cited by South Africa were ambiguous in nature and that all of them did not have any impact on the operational orders actually given to the IDF by the Israeli cabinet.

The plausibility section of the order mentions the claims of both parties, and then moves to discuss a variety of UN sources, which described the dire humanitarian conditions in Gaza and alluded to the “discernibly genocidal and dehumanising rhetoric coming from senior Israeli government officials” (three statements by senior Israeli officials, which arguably fall under this characterization, are mentioned in the order in this connection). Notably, unlike the Myanmar case, none of UN reports or press releases cited involved a serious factual investigation or a detailed analysis of genocidal intent and its relationship to the actual military operation (see also Judge Nolte, para. 13: “The evidence provided by South Africa regarding the Israeli military operation differs fundamentally from that contained in the reports by the United Nations fact-finding mission on Myanmar….).

The relatively weak and contested nature of the evidence for genocidal intent perhaps explains the differences between the language found in the Court’s holding in para. 54 of the order and  the parallel language used in the Myanmar case:

In the Court’s view, the facts and circumstances mentioned above are sufficient to conclude that at least some of the rights claimed by South Africa and for which it is seeking protection are plausible. This is the case with respect to the right of the Palestinians in Gaza to be protected from acts of genocide and related prohibited acts identified in Article III, and the right of South Africa to seek Israel’s compliance with the latter’s obligations under the Convention. (emphasis added)

One clue for the reasons for adding the words “at least some” to the order can be found in the individual opinion of Judge Nolte, who wrote: “I do not find it plausible that the military operation is being conducted with genocidal intent.” Still he did find some other claims made by South Africa to be plausible:

“My decision to vote in favour of the measures indicated rests on the plausible claim by South Africa that certain statements by Israeli State officials, including members of its military, give rise to a real and imminent risk of irreparable prejudice to the rights of Palestinians under the Genocide Convention …. At the present stage of the proceedings, it is not necessary to determine whether such statements should be characterized as acts of ‘[d]irect and public incitement to commit genocide’ within the meaning of Article III (c) of the Genocide Convention… Since they were made by high-ranking officials, who thereby also addressed soldiers involved in hostilities in the Gaza Strip, I cannot plausibly exclude that such statements contribute to a potential failure by Israel to prevent and punish acts of public and direct incitement to genocide.”

The final paragraph of Nolte’s opinion reads quite similar to the language ultimately used by the Court: “South Africa has, in my view, shown that some, but not all, of the rights which it has alleged are plausible at the present preliminary stage of the proceedings.”

It is reasonable to assume that disagreements between the judges on whether the plausibility test has been satisfied with relation to all the different rights claimed by South Africa (corresponding to duties not to commit or attempt to commit genocide, to prevent genocide and to prevent incitement to genocide) led the Court to adopt such ambiguous language, which leaves open the question which claims are plausible and which are not. Whereas Judge Nolte may represent the judges interested in applying a relatively stringent approach to plausibility, Judge Bhandari, who opined that, for plausibility purposes, it would be enough to consider the dire humanitarian situations in Gaza, may represent some of those judges who consider a lower evidentiary threshold more suitable.

Interestingly enough, the difference in wording between the Gambia v. Myanmar and South Africa v. Israel regarding plausibility did not seem to affect the approach of the Court regarding the measures actually applied. This is because in para. 59 of the order, the Court invoked another relatively loose standard often found in ICJ provisional measures orders: whether “at least some of the provisional measures … are aimed at preserving the plausible rights.” The ambiguity surrounding both plausibility and the link between the plausible rights and requested measures might have encouraged the Court to opt for provisional measures of a general nature. This appears to us to be part of the general balancing act struck in the case: the dire humanitarian situation in the Gaza Strip led the Court to acknowledge the seriousness of the crisis and to issue provisional measures, even though it could not formulate a clear position on what makes the case legally plausible and which plausible rights the provisional measures were intended to protect. At the same time, the legal weaknesses of the case and the decision militated in favor of issuing general and not specific provisional measures, yielding a limited effect.

The Order Not Issued – Provisional Measures and the Use of Force

As much as the discussion of plausibility described above is important for ascertaining the legal implications of the order, the more concrete effect of the Court’s order concerns the actual provisional measures issued, and their influence on the real-world conduct of hostilities in the Gaza war. In this regard, it is important to evaluate both the measures ordered and the measures that the court did not order – most significantly, the request to impose an immediate and unilateral ceasefire. 

In its application, South Africa’s requested from the court to order Israel to “suspend its military operations in and against Gaza” (para. 144 ). In fact, one could claim that this was the most significant measure requested by South Africa, and a logical consequence of its position regarding the severe humanitarian harms caused by the military operation. In response, Israel allocated large segments of its oral arguments to justify the military operation as stemming from its right to self-defence and duty to protect its citizens from future attacks by Hamas.

The Court mentioned in its order Israel’s claim that “its right to self-defence is critical to any evaluation of the present situation” (para. 40), but did not address this aspect of the case in its decision (by contrast, the right to self-defence is discussed in the separate opinions of Judge ad hoc Barak and Judge Sebutinde; Judge Nolte mentioned the right to self-defence but opined that the Genocide Convention is not designed to regulate armed conflicts). At the same time, the Court refrained from acceding to the South African request to order a suspension of the military operation – despite the Court being willing to do just so in the Ukraine v. Russia (2022) case, where it was “profoundly concerned about the use of force by the Russian Federation in Ukraine, which raises very serious issues of international law.”

The refusal by the Court to issue in the South Africa v. Israel case a measure similar to the one issued in Russia v. Ukraine could be seen as indicative of the Court’s acceptance of Israel’s self-defence claim as being at least plausible in nature, and one which cannot be prejudiced by the issuance of provisional measures. The interpretation given to order by Judge ad hoc Barak – his writing that “[t]he Court has reaffirmed Israel’s right to defend its citizens” — goes in this direction, even if this might be something of an exaggeration. Indeed, the Court had also refused in the past to order Serbia to cease hostilities in that genocide case, with no implied affirmation of Serbia’s use of force (Bosnia-Herzegovina v. Serbia). What’s more, in contrast with ad hoc Judge Barak, Judge Nolte suggested that the Court rightfully avoided the issue. “The Court need not address many well-known and controversial questions, such as those relating to the right to self-defence,” he wrote.

Still, this outcome appears to us to represent in the aftermath of the Russia v. Ukraine case a relative success for the Israeli defense team. In any event, the silence of the Court about the right to self-defence probably reinforces the perception, which we discuss below, that the actual provisional measures that were ordered by the Court might not make a significant impact on the way in which Israel conducts its military campaign in Gaza.

The Provisional Measures Ordered

Another area of considerable legal ambiguity can be found in the actual provisional measures that were ordered, which tend to be general in nature. This is not so exceptional, however, in genocide cases. The remedies ordered in South Africa v. Israel largely mirror those issued in Gambia v. Myanmar (with the major exception being the order pertaining to enabling basic services and humanitarian assistance). Notably, the more specific requests by South Africa — to cease the military operation (modelled after the Ukraine v. Russia order), to allow the entry of fact-finders into Gaza and to take a series of specific measures, including rescinding forced displacement orders (resembling the ones decreed in 2023 in the Armenia v. Azerbaijan case) — were all rejected by the Court. Instead, the Court ordered Israel to adopt six provisional measures: take all measures within its power to prevent genocide, ensure that its military does not commit genocide, take all measures within its power to prevent an punish incitement to genocide, take immediate and effective measures to enable urgent humanitarian assistance and basic services, preserve evidence relating to allegations of acts under article 2-3 of the Genocide Convention, and report to the Court within one month about measures taken to give effect to the order.

Since Israel has taken a firm position that its military campaign cannot be plausibly seen as genocidal in nature – as Israel maintains that a campaign not aimed at harming Palestinian civilians, and which incorporates multiple harm mitigation measures cannot be factually or logically based on genocidal intent – it is not obvious how the implementation of the first two measures ordered would involve any major change in its approach to the conflict. This is especially so, since the Court has not yet made any concrete finding that Israel has violated any of its obligations under the Genocide Convention (See e.g. para. 30 and 84 to the provisional measures order). One may also question, in this context, the relevance of the second measure ordered by the Court (to prevent the Army from committing genocide). Unlike the Gambia v. Myanmar and Ukraine v. Russia cases, where there were claims that armed groups or units within the armed forces were acting outside the central control of the state party, we know of no such claims regarding the Israeli operation in Gaza.

The third provisional measure ordered by the Court –  to prevent and punish incitement to genocide – might prove to be a policy area where more visible change could take place, since neither the Israeli legal or political system has taken, so far, serious and concrete measures against individuals involved in what appear to be acts of incitement to genocide. The Attorney General and the State Attorney warned publicly on 9 January 2024 against making such statements and indicated that they are examining certain cases. Still, no indictments were brought forward; nor was any information about other investigative steps or sanctions made public. What’s more, the Court specifically addressed those developments in the Attorney General’s office and determined, “While steps such as these are to be encouraged, they are insufficient” (para. 73).

The fourth measure ordered by the Court requires Israel to enable the provision of urgently needed basic services and humanitarian assistance. It is worth noting that in the application for provisional measures, South Africa requested much more concrete measures, including ordering Israel to desist from restricting access to food, water, fuel, clothes, medical supplies, etc. In the face of Israel’s insistence that it does allow for the entry into Gaza of humanitarian assistance in sufficient quantities (See pp. 50-52 to the Verbatim record of 12 January), it seems the Court chose to order the enabling of essential supplies, without making any findings regarding which specific humanitarian provisions are urgently needed at the present time.

Notably, the Court’s statement that certain actions taken by Israel were currently “insufficient to remove the risk that irreparable prejudice will be caused” also related to “steps to address and alleviate the conditions faced by the population” (para.73). Once again, this statement by the Court indicates that Israel must do more to enable basic services and humanitarian aid in order to ameliorate the urgent and catastrophic situation of the Palestinians in Gaza (para. 72).

In fact, the third and fourth provisional measures ordered seem to us to hold the greatest promise for actual impact on the ground. On the Issue of the plausibility of the genocide claim, the Israeli public is mostly unified it its rejection of the claim, viewing it more as a form of “blood libel.” The same also holds true with regard to the very high level of public support for Israel’s military campaign against Hamas. However, on the issues of inciting statements by politicians and officials and the enabling of humanitarian assistance, there exists a heated political debate inside Israel. Even before South Africa’s application was submitted to the Court, significant parts of the public condemned these incendiary statements by public figures. As for the provision of humanitarian assistance, after the initial shock following the October 7 massacre receded, the facilitation of such assistance was accepted by large parts of the Israeli political system as an essential part of the war, despite continued calls by the extreme right to condition it on Hamas providing access to the Israeli hostages it holds. We believe that what political scientist Beth Simmons has claimed regarding compliance with international human rights law holds true also here: The greatest potential of the Court to actually change domestic behavior is where it enjoys the support of powerful local stakeholders and can mobilize them through its decisions.

In this last context, it is the sixth measure – the one month reporting requirement – which may ultimately prove to be the most consequential remedy ordered by the Court. This measure   underscores the close international judicial scrutiny to which Israel’s acts in Gaza are now subject. With the date of the submission of the report approaching, the Attorney General should definitely consider changing course and pursuing more effective measures to prevent and punish inciting statements. Moreover, it is notable that the need to report to the Court about the measures taken has already been cited by IDF sources as a reason to disperse right wing demonstrators who were interfering with the operation of the Kerem Shalom crossing that is used for transporting humanitarian aid into the Gaza Strip.

Finally, the report – a copy of which will be forwarded to South Africa – might trigger efforts by the latter to request the Court to revise the provisional measures. The comparison to the Armenia v. Azerbaijan proceedings might be useful in that regard. These latter proceedings have featured five provisional measures requests brought to the Court within two years. Moreover, the last order in this latter case, which included a one-month reporting obligation, was explicitly motivated by “the undertakings made by the Agent of Azerbaijan at the public hearing.” One can therefore presume that the reporting requirement introduced in the South Africa v. Israel case was similarly linked to certain statements made by Israeli counsel regarding ongoing investigations of acts of incitement, ongoing efforts to facilitate more humanitarian aid and steps taken toward winding down the military campaign. Through the reporting duties, the Court could verify whether Israel makes good on its promises and representations before it.

In any event, one should recall that international scrutiny of Israel’s manner of prosecuting the Gaza war is not limited to the South African ICJ proceedings. In February, the ICJ will begin hearings on a request for an advisory opinion on the legal consequences of the prolonged Israeli occupation of the Palestinian Territories, which would also touch upon the Gaza situation. Moreover, the International Criminal Court prosecutor has already declared that his office is investigating alleged crimes committed by Hamas and Israel during the present war.

Concluding Remarks

In its order on provisional measures, the Court appeared to engage in what Professor Abi-Saab has referred to as transactional justice. It used harsh rhetoric to describe the catastrophic situation in the Gaza Strip and agreed to issue provisional measures (arguably also because of the devastating harm which the Court’s reputation would have suffered had it not done so). Nevertheless, it then minimized the impact of the order by adopting ambiguous language regarding the contents and scope of the plausibility test; ambiguated again with regard to its view of the facts of the case; and ordered provisional measures which, due to their general character and considerable overlap with Israel’s self-perception of the legality of its current actions, are unlikely to result in significantly changing, in and of themselves, the dynamic on the ground.

Still, the continued scrutiny of Israel’s practices in Gaza, facilitated, inter alia, by the reporting obligation introduced by the Court, and the growing pressure by Israel’s allies (who might become be more and more nervous about supporting a country against which plausible claims of committing, at least, some violations of the Genocide Convention have been made), might result in greater willingness by Israel to review its war policies. Such a review may strengthen the hands of those elements within Israeli society and government who are calling for a tougher approach against incitement to genocide, greater efforts to facilitate humanitarian assistance, more care afforded in order to minimize the harm to civilians caused by the military operation and an increased propensity to move toward winding down the war.

This article was published by Just Security.