How Would War Crimes Prosecutors Classify the Syrian Conflict(s)?

Over the past few weeks, a great number of excellent scholars have debated how to classify the contemporary armed conflict(s) in Syria. In particular, Ryan Goodman (here, and here), Adil Haque (see here, here), Oona Hathaway (on Twitter), Deborah Pearlstein (see here and here), Gabor Rona (see here), Terry Gill (see excellent article here, and blog here), and Dapo Akande (see excellent article here and blog here) have all debated the trigger points of non-international armed conflict (“NIAC”) and whether the Syrian conflict(s) are now rendered international armed conflicts (“IAC”) by American, Turkish and Russian military intervention. I join the discussion to ask how war crimes prosecutors are likely to see these issues, then to raise the possibility (born of working in this capacity myself) that an analytically satisfying solution to these debates about internationalized armed conflict might be structurally unavailable. I begin by introducing these perspectives, then plot a set of doctrinal points that arise from ICL’s encounter with the phenomenon of conflict classification. I end by reiterating my earlier normative critique of the international/non-international bifurcation in the laws of armed conflict (see here), which emerged from my own intellectual dissatisfaction as a war crimes prosecutor over a decade ago.

To begin, let me flesh out why inquiring about war crimes prosecutors’ perspectives might be a helpful supplement to the debates about conflict classification in Syria thus far. Most obviously, if the reason for insisting on qualifying the armed conflict in Syria is to promote the prosecution of war crimes, it could be helpful to understand how courts tasked with trying war crimes are likely to undertake that classification process if these trials ever come to pass. But perhaps more importantly, war crimes prosecutors have confronted more or less exactly the same difficulties that animate these debates for over twenty years now, albeit in the context of the multiple, changing and overlapping international/non-international conflicts in the former Yugoslavia and the Democratic Republic of Congo rather than Syria. Much of the resulting caselaw is exceptionally detailed, and as a body of experience, it is useful as a tool to help mediate between competing arguments here. In fact, ICL’s long history of classifying armed conflicts is also of assistance in that it shows these institutions attempting to avoid the classification process wherever possible. As I explain below, revealing their attempts to bypass the classification conundrum is an important aspect of the added value a war crimes prosecutor’s perspective may provide.

This brings us to my second initial perspective. If ICL has done its best to pull away from the riddle of classifying internationalized conflicts like that in Syria, it is somewhat peculiar that when faced with Terry Gill’s harrowing revelation that “[t]here are reportedly hundreds (by some accounts approximately 1,500) of armed groups and militias active in the Syrian conflict” that no one has yet objected that this distinction between IACs and NIACs cannot be meaningfully made. The closest we get thus far is Professor Gill’s admission that “[it] would be well-nigh impossible to draw a coherent picture of the entire mosaic of armed groups and their aims, actions and alignments.” (see here, p. 355). In addition, we have not heard that even where the application of these tests for internationalization is more clear cut, the ability of armed groups to comply with standards we ourselves cannot agree upon in the heat of battle, in the context of changing military relationships, based on closely guarded information seems marginal. I am reminded, for instance, of Marco Sassòli and Laura Olsen’s argument that “there is no reason to think that, during a conflict, one could convince a military commander to respect certain rules by arguing that he is an agent of a foreign country.”[1]

Admittedly, I am reiterating here an argument I first made over a decade ago after first working on these issues at two different war crimes tribunals, in which I pointed out the unavoidable dangers of analytical incoherence in qualifying internationalized armed conflict, then posited the possibility of a unified system of IHL that would be applicable to all types of armed conflict (see here). I made this argument in order to circumvent the otherwise insurmountable analytical difficulty I had experienced in practice. As I explain in that article, the way out I propose was not new even then: the ICRC had advanced this argument at every stage in the codification of major IHL instruments, various luminaries such as George Aldrich had endorsed it in light of the Vietnam experience, and the history of ICL is replete with judicial statements like “it is only natural that the aforementioned dichotomy [between IAC and NIAC] should gradually lose its weight.”[2] As I will explain shortly, ICL not only offers a set of lessons about the classification process that have not fully informed the various online debates about Syria thus far, it also reveals a pattern of deliberate attempts to avoid the problem wherever possible. This reality speaks to an ongoing concern about the practicality of the tests in discussion presently and the availability of third approaches that might be appealing to prosecutors if cases from Syria are ever heard.

I move, then, to my five doctrinal observations about the history of conflict classification in ICL and its salience to these debates:

First, although it is probably technically correct to say that Tadić is the leading judicial decision in this area, to leave matters at that risks undervaluing over two decades of judicial experience classifying armed conflicts post Tadić before a wide variety of courts and tribunals (national and otherwise). In fact, a number of initiatives within these institutions deliberately sought to build upon the initial foundations set by Tadić. In 2007, for instance, I was asked (ironically given my earlier article calling for the abandonment of the distinction) to lead a process for the Office of the Prosecutor of the ICTY that developed a more comprehensive matrix of factors that go to making up a non-international armed conflict. This project led to far more detailed sets of factors that would establish both limbs of the test for a non-international armed conflict—intensity and military organization—in a trilogy of cases involving the Kosovo Liberation Army and the Albanian National Liberation Army in Macedonia.[3] The painstaking depth these and other cases go to in applying the tests may be helpful to friends and colleagues engaged in these debates, beyond just the foundational importance of Tadić itself. They may also act as templates for prosecutors asked to prosecute war crimes arising out of modern-day Syria.

Second, let me add one problem from this history to the current debate about Syria in order to point to a sub-issue that adds further complexity, in ways that might also militate in favour of a simpler unified approach moving forward. As all the experts engaged in this debate will know, Additional Protocol II (“APII”) entails a different test for NIAC that arguably lifts the intensity requirement above that applicable to Common Article 3 conflicts and that also appends the requirement that the armed group enjoy territorial control. Some courts, like the ICC, have rejected the territorial control element for purposes of determining a NIAC under its statute,[4] but the Special Court for Sierra Leone has insisted on it as a requisite element of war crimes that derive from APII.[5] So, in the same way that Ryan Goodman has pointed out that Russia is a signatory to API for the purposes of determining the applicable law in Syria (see here), war crimes prosecutors are also likely to be confronted by the reality that: (a) Russia is party to APII too; (b) that APII requires territorial control in addition to the intensity and military organization elements for armed conflicts in Common Article 3 NIACs; and (c) that the question of classifying the law applicable in the Syrian context may be even thornier than our debates to date have revealed.

Third, once these issues are viewed through the eyes of war crimes prosecutors, it will likely become apparent that the first prong of the test for internationalization in ICL appears to have escaped close scrutiny in the debates about Syria thus far. Before now, my friends and colleagues who have engaged in this debate have largely focused on the absence of consent on the part of the Syrian government to the various manifestations of US military force in Syria, arguing about whether the absence of Syrian consent means that the United States is presently engaged in an IAC with Syria and/or Russia. And yet, the test for internationalization in ICL is appreciably wider in scope, and although controversial, a number of ICL cases will act as precedents for war crimes prosecutors focused on Syria who are eager to establish their jurisdiction over the full panoply of war crimes applicable in IAC. To recall, in the famed Tadić Appeal Judgement, the Appeals Chamber of the ICTY stipulated that:

“It is indisputable that an armed conflict is international if it takes place between two or more States. In addition, in case of an internal armed conflict breaking out on the territory of a State, it may become international (or, depending upon the circumstances, be international in character alongside an internal armed conflict) if (i) another State intervenes in that conflict through its troops, or alternatively if (ii) some of the participants in the internal armed conflict act on behalf of that other State.”[6] (emphasis added)

Importantly, a number of international courts and tribunals have interpreted the reference to “another State interven[ing] in that conflict through its troops” as extending beyond actual armed force between States Parties. To be precise, a number of decisions apply the body of law applicable to IACs to all state and non-state actors within a conflict zone where a foreign military intervention only indirectly affects independent internal conflict(s). I provide several examples of this reality from ICL caselaw in my earlier article (see here, pages 328-333), but to cite just one here, the Kordić & Čerkez Judgement found that the Croatian government’s intervention in the conflict against Serb forces in Bosnia internationalized a separate conflict in which the Croatian government had no direct military involvement, namely the conflict between Bosnian Croats and Bosnian Muslims.[7] According to the Trial Chamber, it did this “by enabling the Bosnian Croats to deploy additional forces in their struggle against the Bosnian Muslims.”[8] Thus, the Tribunal applied the laws applicable to IACs to all actors discussed, including non-state armed groups. Our debates about the absence of state consent—certainly a vitally important issue—played no part in this classification. Depending on the war crimes involved, war crimes prosecutors considering trials arising out of Syria may well take inspiration from these precedents.

Fourth, the debate up until now has only tacitly referenced “Mixed” versus “Global” theories of conflict classification, which polarized scholars and practitioners in the early years of ICL’s encounter with this problem. The mixed approach—what Terry Gill calls “parallel conflicts”—is reflected in the refrain from Tadić that the violence in the former Yugoslavia could be characterized “at different times and places as either internal or international armed conflicts, or as a mixed internal-international conflict”[9] and that “depending upon the circumstances, [the conflict may] be international in character alongside an internal armed conflict.”[10] And yet, an alternative “global” approach pre and post-dates these tests; many experts have acknowledged that distinguishing between IACs and NIACs is practically impossible in many modern armed conflicts given the indecipherably complex and constantly dynamic interaction between state and non-state actors in internationalized warfare. According to this global approach, the full body of IHL applicable to IACs apply to all armed groups, state and non-state, in an entire territory that contains multiple conflicts of international and internal origin. As I show in the article (see here, pp. 334-335), the US government, Theodore Meron, various ICTY judges and even the ICRC appear to have endorsed this approach in certain circumstances.

These ideas have featured only tacitly in the recent online debate about conflict classification in Syria. Terry Gill’s excellent article, for instance, starts off assuming a mixed approach to classification (he calls “parallel”), but when faced with the tremendous complexity of the task in certain scenarios he shifts to the global alternative, before he later shifts back. A variant of the global approach seems evident, for example, where he concedes that “[t]he fact that these parties have different objectives and have clashed with one another on occasion (or in the case of ISIS and the Kurdish YPG on an ongoing basis) does not change the fact that there is one overall conflict of a non-international character within Syria with a number of different parties. The alternative of looking at each conflict as a separate conflict makes no factual or legal sense.”[11] To my mind, this quite understandable oscillation between mixed and global approaches emulates that in earlier ICTY caselaw. Moreover, it is also interesting to see the same oscillation play out at the ICC in a conflict strikingly similar to Syria legally speaking; a conflict Madeline Albright once dubbed “Africa’s First World War.”

In recent years, a Pre-Trial Chamber at the ICC in the case against Thomas Lubanga cited the two prong test from Tadić I quote above, then indicated that, “where a State does not intervene directly on the territory of another State through its own troops, the overall control test will be used to determine whether armed forces are acting on behalf of the first State.” (emphasis in original).[12] By implication, the overall control test is not relevant where there is direct military intervention, which will operate to internationalize all conflict in a globalizing fashion in line with cases like Kordić & Čerkez I referenced above. Again, the important topic of State consent we have debated did not feature in this analysis. Thus, the Pre-Trial Chamber concluded that the armed conflicts between various non-state actors in the region were subject to the law governing IAC because Uganda was an occupying power in the region. In the Pre-Trial Chamber’s own words: “as a result of the presence of the Republic of Uganda as an occupying Power, the armed conflict which occurred in Ituri [between various non-state actors] can be characterised as an armed conflict on an international character from July 2002 to 2 June 2003, the date of the effective withdrawal of the Ugandan army.”[13]

At trial, a differently constituted bench disagreed with the Pre-Trial Chamber’s characterization by effectively adopting a more stringent mixed approach. Instead of citing the Tadić two limbed test, which entailed both direct and indirect intervention as bases for internationalization, the Chamber ignored the first element based on direct foreign intervention and the line of cases like Kordić & Čerkez applying it, then discredited evidence about Rwandan and Ugandan control over various armed groups fighting in the region in assessing the second. As a result of this shift in technique, the ICC Trial Chamber in one and the same case revised the Pre-Trial’s position by concluding that the conflict between these various non-state groups “was non-international in nature.”[14] This mixed approach has gained ascendancy at the ICC ever since so is likely to be especially influential to war crimes prosecutors contemplating the terrible conflagration in Syria. Nevertheless, it is also notable that aside from marking a repetition of the oscillation between mixed and global approaches, in a passage I am tempted to read as conveying regret, the Court remarked that “some academics, practitioners, and a line of jurisprudence from ad hoc tribunals have questioned the usefulness of the distinction between international and non-international armed conflicts… The Chamber does not have the power to reformulate the Court’s statutory framework.”[15]

From an analytical standpoint, the problem is that both of the approaches are difficult to justify, which perhaps explains the seemingly constant oscillation between them. The great difficulty is that the metrics for explaining when to prefer the mixed and when to favour the global are extraneous to the legal tests and go unannounced. More broadly, in my earlier piece, I argued that the relative strengths and weaknesses of the “mixed” and “global” views indicate that reaching any sort of agreement about the classification of complex internationalized conflicts like Syria within the present framework will inevitably involve choosing between a theory that cannot work (the mixed approach) and a practice that is not justified (the global approach). (see here, p. 335). The challenge, therefore, is very much to the structure of IHL generally. For present purposes, I wonder if our debates about classification could benefit from keeping these arguments in mind, although as I hint at throughout, the better way of promoting accountability may lie in convincing war crimes prosecutors that they need not try to resolve issues we cannot.

Fifth, avenues exist that allow us to do just this. Because these issues are so factually complex, analytically unsatisfying, enormously time consuming to prove, and ultimately, often morally irrelevant, courts have attempted legal run-arounds wherever possible. A number of the scholars in this online debate have rightly pointed out the areas of substantive disparity between IAC and NIAC, but if war crimes are the emphasis in general, and murder, rape and torture the central pre-occupations in particular, it might be possible to dispense with the characterization process all together. At a certain point in its maturation, the ICTY and various national criminal courts adopted just this approach as a response to the sorts of classification quandaries we are debating (for examples, see here, p. 864). They did this by relying on the ICJ’s dicta in the Nuclear Weapons case that Common Article 3 is an “elementary consideration of humanity” applicable in both variety of armed conflict (an approach later echoed by the US Supreme Court in Hamdan). Armed with this blanket principle, prosecutors might look at a conflict like that in Syria, banish grave breaches immediately because of the difficult technicalities they entail, then opt for Common Article 3 prosecutions to avoid the intractable complexities we are presently engaged with. I have pointed to the problems this approach can give rise to elsewhere (see here, p. 875), but for present purposes, this strategy may be the best way of incentivizing war crimes prosecutors to take up these cases despite our understandably protracted disagreement.

This brings us to a fork in the road, where I move away from a perspective grounded in ICL doctrine into a purely normative mode, which originally developed as a response to my experiences with these problems as a practitioner. I have four normative points:

First, I believe that the idea of single body of IHL applicable in all types of conflicts deserves far greater intellectual engagement. Admittedly, as my earlier article readily conceded, that project is conceptually challenging and politically unlikely because it has to address the absence of combatant status or a law of occupation in NIACs, two issues not likely to be readily resolved. I will not attempt to broach these conceptual issues here, other than to offer up the notion that it is hard to incentivize compliance with IHL by non-state groups without offering something akin to combatant status and to observe how these difficulties already arise in internationalized non-international armed conflicts (see here, p. 345). In addition, I respect the reality that opening up the Geneva Conventions for renegotiation in a post Sept 11 world would likely lead to a net diminution of humanitarian protections, but regardless of whether a unified body of law that is not regressive can ever be politically realized in even the medium-term, I reiterate my now dated calls for greater engagement with the unification project as a normative agenda, perhaps as part of or an appendage to Columbia’s project on harmonizing standards for armed conflict (see here). One upside is that it stands to make war crimes prosecutions in places like Syria easier.

Second, the engagement with what Adil Haque eloquently calls “triggers and thresholds” (see here) is exceptionally interesting. I was particularly struck by Adil’s very insightful conclusion that “we should accept a unilateral trigger and nominal threshold for both IAC and NIAC.” If this approach is defensible, he has found a solution to a key problem for a unified body of norms that dispenses with the IAC/NIAC classification altogether. To address this problem of disparate trigger mechanisms in my earlier proposal, I borrowed from a proposition by the Brazilian government during the negotiation of APII, which suggested that the application of IHL in all types of conflict could be triggered by armed violence between “organized armed forces or other organized armed groups under a responsible and identifiable authority, and clearly distinguished from the civilian population.” (see here, p. 345). I am less sure that this is much of a solution now, so I am particularly enthusiastic for creative new thinking like that Adil Haque offers as well as the intentionality approach Michael Adams and Ryan Goodman have suggested (see here). In the same breadth, my enthusiasm is strictly conditional on the need for these innovative new standards to avoid watering down pre-existing IHL protections and prevent against a new field of application that makes departures from human rights standards easier, more frequent, or simply more justifiable.

Third, I sense that the law of armed conflict is caught between its aspirations for humanitarian protection and an anxiety about its own complicity in violence. On the one hand, I certainly understand and appreciate the argument by Adil Haque and others that IHL does not authorize anything; it merely restrains. So when Gabor Rona complains (see here) that qualifying the armed conflict as international might trigger “the same targeting and detention rules that would apply between the US and Syria.. wherever US and Russian interests rub up against each other,” the retort is that if the US and Russia do carry out these unthinkable actions, it will be for altogether different political reasons that are entirely seperable from the body of IHL that will apply to them as they do so. Conversely, one does not have to tax one’s memory too hard to recall the Bush Administration’s use of the laws of war to publicly justify important excess. The laws of war rhetoric helped enable indefinite detention of detainees at Guantánamo (without conferred them with corresponding protections) and had a quite terrible trickle-down effect in Uganda, Liberia, Chechnya and beyond, where conflicts were quickly re-imagined as “Wars on Terror.” The reality is that historically speaking, the laws of armed conflict are often used to justify violence.[16] To my mind, thinking through ways of undermining this complicity should also be a first order task for IHL scholars.

Fourth, a unified body of IHL could help do just this by depoliticizing the significance of a conflict’s classification one way or the other. Much of the resistance to Ryan Goodman’s argument (see here) that the US is already in an IAC stems from a concern that this recognition would be politically provocative, thereby entailing a weak variant of the complicity-in-violence-anxiety I reference immediately above. Gabor Rona, for example, mentions his concern about the classification “upping the ante” (see here); Deborah Pearlstein reasonably worries that “Syria and Russia would view such a statement as provocatively signaling a U.S. intention to embark upon a new and different course of hostilities” (see here); and Terry Gill’s very helpful article ends with a series of warnings about “drawing conclusions which open the door to a widening of the conflict.” (see here, p. 380). Ironically, in my earlier article, I used Russian intervention in Afghanistan several decades ago to highlight equivalent concerns for politicization then (see here, pp. 342). I also suggested that a unified body of armed conflict that stripped away the IAC/NIAC distinction might offer a way out that minimizes these tensions, at least partially, by allowing us to insist that all parties are bound by IHL in their military actions without saying more.

The foregoing does not offer obvious solutions for the Syrian classification, but I hope that some of the terrain I traverse is useful for further discussions of these important issues.


[1] M. Sassòli and L. M. Olson, “International decision: Prosecutor v. Tadić (Judgement)”, American Journal of International Law, Vol. 94, July 2000, p.

[2] Tadić Jurisdiction Appeal, para. 97

[3] The principle case was Prosecutor v. Boškoski, but the same principles are also evident in the Prosecutor v. Limaj and Prosecutor v. Haradinaj cases.

[4] Prosecutor v Lubanga Trial Judgment, para. 536.

[5] Prosecutor v Sesay et al, SCSL-04-15-T, Judgement, 2 March 2009, para. 966. (“the Prosecution must also prove the elements of Article 1 of Additional Protocol II, namely that the dissident armed forces or other organised groups participating in the conflict: […] (ii) Were able to exercise such control over a part of their territory as to enable them to carry out sustained and concerted military operations.”)

[6] Prosecutor v. Tadić, T-94-1-A, Judgement, 15 July 1999, para. 84

[7] Prosecutor v. Kordić & Čerkez, IT-95-14/2-T, Judgement, 26 February 2001

[8] Ibid., para. 108(2).

[9] Tadić Appeal Judgement, para. 73

[10] Id.

[11] Gill, p. 375.

[12] Prosecutor v Lubanga, Decision on Confirmation of Charges, 29 Jan 2007, para. 220

[13] Prosecutor v Lubanga, Decision on Confirmation of Charges, 29 Jan 2007, para. 220.

[14] Lubanga Trial Judgment, 14 March 2012, para. 567.

[15] Ibid, para 539.

[16] In my view, one of the best arguments to this effect is still Chris Jochnick & Roger Normand  “The Legitimation of Violence: A Critical History of the Laws of War” (1994) 35 Harvard Int’l LJ  49-95; 387-416.