Ever since its very first articulations, the “unwilling or unable test” has relied heavily in the time-tested legitimacy of the 1837 Caroline Affair, where British forces sunk a vessel manned by Canadian rebels in American territory. Dressing such a visible and well-known case in the cloth of “unwilling or unable” allows its proponents to argue that its underlying principles have always been an accepted rule of international law, and we all just missed it.
However, one specific incident is not enough to prove that a particular legal rule has been firmly established in international law for the past 200 years. This is why, as the test’s articulation became more refined, its adherents have sought to find additional situations that may help confirm the suspicions borne by the Caroline, that the test was always part of international law, particularly if they occurred before the adoption of the UN Charter.
To my knowledge, the most comprehensive list of incidents that would give support to the longstanding (pre-Charter) character of the unwilling or unable test is the appendix to Ashley Deeks famous article on the matter. This list starts with an 1817 United States attack on Seminole Indians in Spanish Florida and finishes with the British attack against the Altmark, in Norway, during World War II. Of the 12 pre-Charter cases listed, 8 deal with non-state actors, and of these, to my surprise, 6 involve Latin American states.
I was particularly surprised to see the 1932 Leticia Incident between Peru and Colombia and the 1916 US Expedition against Francisco “Pancho” Villa make the list; two events that – at least the way I learned about them – did not really match the basic requirements of the test. My surprise continued to grow, however, as these cases kept popping up in subsequent conversations and reports, and I decided to take a more detailed look. This two-part post contains the results of my inquiry, starting today with Leticia, leaving Pancho Villa for another day.
The Leticia Incident is a topic I am well aware of. As any Peruvian, I studied it in high school as part of Peruvian history class, as the triggering event of the 1933 Colombo-Peruvian War. The story behind Deeks’ chart is both interesting and uncommon, and a good way to understand how States viewed sovereignty and territorial integrity in the early 20th century.
The crisis traces its origins to the territorial uncertainties caused by the independence of Spanish America, particularly in the under-explored but resource rich Amazon jungle. As can be seen in the map I prepared below, Peru sought control of all territories south of the Caquetá river, whereas Colombia claimed all territories north of the Napo and Amazon rivers. In 1922, they negotiated a compromise fixing the border at the Putumayo river, but granting Colombia a small piece of land protruding to the south, called the Amazon Trapezoid. The area was of strategic importance, as it included the Amazonian port of Leticia.
This accommodation, known as the Salomón-Lozano treaty, proved highly unpopular in Peru. It was speculated that President Leguía signed it only to accommodate the United States, that was seeking to compensate Colombia for the loss of Panama. In 1930, the unpopular Leguía was brought down by General Luis Miguel Sanchez Cerro, a strong opponent of the treaty.
Two years later, on September 1st, 1932, a group of Peruvian citizens and a few rogue soldiers, decided to storm Leticia, expel its Colombian authorities and raise the Peruvian flag in the town’s main square. Sanchez Cerro was caught by surprise and was undecided as to how to proceed. Considering his position on the treaty, he could not fully denounce the action, but he could not afford a full scale war either. His decision was to try to walk a thin line, seeking to take advantage of the rebels’ actions to secure political gains for himself and territorial conquests for Peru, but without renouncing the Salomón-Lozano border. On September 30th, Peru formally invited Colombia to revise the treaty, under the terms of Article 19 of the Covenant of the League of Nations.
Colombia, of course, reacted as expected, demanding that Peru disavow the rebels and declare its neutrality, while beginning preparations to retake Leticia. The main problem, however, was that Colombia had no fleet with which to achieve this goal. In order to gain time, and while it purchased the ships and weapons it required, it addressed Peru’s concerns through diplomatic means. In October, 1932, Colombia responded that it would re-establish control over Leticia “without recognizing any other state or international organization the prerogative of intervening in its internal affairs” (all translations are my own; see here for the print source of these diplomatic notes: the memoirs of General Alfredo Vásquez Cobo, Commander in Chief of the Colombian Expedition to the Amazon, p. 87).
The ensuing diplomatic exchanges paint a very un-Caroline-like situation. Peru justified its call to negotiations in the self-determination of the inhabitants of Loreto (in the Peruvian side) and Leticia (in the Colombian side), calling the episode “the product of spontaneous expressions of uncontainable national aspirations” (Vásquez Cobo, p. 111), and claiming it had not violated, nor renounced, the Salomón-Lozano treaty. Colombia, in turn, complained about Peru’s attempts to “transform an act that for Colombia has essentially the nature of a matter subject to its exercise of interior sovereignty, that repels all interference by a foreign power, into an international affair” (Vásquez Cobo, p. 89). Colombia insisted in Peru’s complicity in the matter, accusing it of sending supplies, weapons and troops to support the rebels.
On December 21st, the Colombian expeditionary fleet reached the Brazilian port of Belém do Pará, in the Atlantic Ocean, with instructions to sail up the Amazon River, towards the Amazon Trapezoid. The instructions were clear: this was an “unquestionable internal police action”, not an international war (Vásquez Cobo, p. 161-162). In order to ensure this, Colombian authorities warned General Vásquez Cobo, to be extra careful to respect Brazilian sovereignty. The League of Nations also issued similar instructions to the parties (Vásquez Cobo, p. 162-163): Peru had to abstain from intervening in the Leticia affair, limiting its actions to only defending Peruvian territory as expressed in the Salomón-Lozano treaty, whereas Colombia had to avoid violating Peruvian territory and limit its actions to the reestablishment of order in its own territory.
This, however, presented Colombian military leaders with a conundrum: Colombia’s fleet was approaching Leticia from the southeast, in Brazilian territory. Simply attacking the rebels as soon as they were within range of fire, from Brazilian waters, was unthinkable, as it was contrary to international law. The Expedition would need to sail up the Amazon until it was within Colombian waters in the north-eastern bank of the Amazon, very close to the rebel positions in Leticia, and potentially subject to fire from Peruvian regular forces staged in the island directly in front of the port. This not only risked defeat, but also risked having to use Brazilian territory as a staging ground, compromising Brazil’s neutrality (Vasquez Cobo, p. 229).
In order to solve the problem, the President instructed Vasquez Cobo to leave the Amazon River, and sail up the Putumayo, to attack the Peruvian garrisons in Tarapacá. The reasons for this detour were clearly set out:
“First: We will act in a section of the Putumayo where, in accordance with the Delimitation Treaty, Colombia owns both banks of the river. Our freedom of action would therefore be absolute. We can move and attack without fear of violating Brazil’s neutrality, which is a factor that carries immense direct and indirect importance (…). Second: While it seems Peru has staged forces in Tarapacá and organized the defence of said sector of the trapezoid, it has done so in a much lesser degree than in Leticia, both because of [natural] elements and its manpower”. (Vasquez Cobo, p. 230-231).
The plan also included sending an additional force staged in the upper Putumayo to assaults the the Peruvian towns of Güepí and Cabo Pantoja (Vásquez Cobo, p. 202). In sum, this would no longer be the “internal police action” Vasquez Cobo had originally planned for, but a full-fledged war with Peru; something that he did not like. In his memoirs he noted that this “turned a military operation of simple police action, such as the occupation of Tarapacá and Leticia, doable within a few weeks, into something I would not call a war, but at least an international conflict that we should avoid at all cost, of undetermined duration” (Vasquez Cobo, p. 207).
After receiving assurances from the signatories of the Kellogg-Briand Pact that they considered Peru’s support for the rebels a violation of the treaty, Bogotá launched its attack on Tarapacá. The war raged on for two full months, until late April, 1933, when Peru’s President Sánchez Cerro was assassinated, as he supervised troops about to be deployed to war. His successor, General Oscar R. Benavides, sued for peace, and reached an agreement: Peru would hand over Leticia to the League of Nations for administration, while Peru and Colombia negotiated peace. In June 1934, both nations signed a Protocol to the Salomón-Lozano treaty, ratifying its terms, and finalising the Colombo-Peruvian War. As legend has it, Benavides refused to allow the Peruvian flag be lowered from the flagpole in Leticia’s main square; the flagpole was withdrawn whole, with the flag still raised, and returned to Peruvian territory intact, to be received by cheering crowds.
I thus see little resemblance between the Leticia and Caroline incidents, and for that matter little support for the unwilling or unable test in either Peru or Colombia’s positions. Colombia never attacked rebels in foreign soil, nor could it: Leticia was in Colombian territory and the rebels had the full support of Peru. The actual hostilities were, thus, a war between Colombia and Peru, in both Colombian and Peruvian lands. If anything, Colombia’s high level of care not to use force in any portion of foreign soil, even in self defence against rebels occupying one of its own cities, is rather proof of just how relevant territorial integrity was for pre-Charter Latin America, even in dire situations.
In fact, this strong respect for sovereignty and distaste for intervention was markedly present in the contemporary international agreements meant to address the problems arising from non-state actors in time of civil war: the 1911 Agreement on Internal Commotions and Neutrality, and the 1928 Convention on Duties and Rights of States in the Event of Civil Strife (which in 1933 was binding on Brazil, Colombia, Costa Rica, Dominican Republic, Mexico, Nicaragua, Panama, and the United States). In brief, neither of these treaties regulated an open-ended right to attack non-state actors in cases of unwillingness or inability of the host State, but rather requested each State to deal with the threats they encountered inside their own borders. In fact, this understanding of a limited set of options available to deal with rebel forces acting across national lines is precisely the legal discussion at the heart of the Mexican and American positions during the 1916 Punitive Expedition, and exactly the reason why I find its inclusion on Deeks’ chart so surprising.
As evidenced in Part I, Latin American states have not been keen to allow expansive interpretations of the rules for use force in foreign soil. Latin America is a region historically subjected to foreign intervention, and as such, the rules it designed, especially in the pre-Charter era, were always very much thought out from a perspective of protecting “the invaded”, not enabling “the invader”.
This was particularly evident for Mexico, a country very much affected by foreign intervention throughout the 19th century. Mexico’s distaste for foreign intervention would not change after the turn of the century, and it would be evident throughout the entire duration of the 1916 crisis, which is why I have been so surprised to see it listed so frequently as an early example of the unwilling or unable test.
Perhaps unsurprisingly, though, the historical retelling of the 1916 Expedition changes depending on whose sources one reads. While American sources tell the story of the “Punitive” Expedition and America’s campaign to sanction Francisco “Pancho” Villa across the border line, Mexican sources, such as renowned historian Bertha Ulloa, rather speak of “the Resisted Intervention”, and remember it as the time when First Chief of the Revolution, Ventusiano Carranza, stood his ground against an overwhelming American power, in defence of Mexican sovereignty and national dignity. In this retelling, more than an early example of “unwilling or unable”, the 1916 Expedition is rather an example of early pushback against it.
In 1916, Mexico and the US were facing a complex diplomatic situation. After long, careful negotiations, American President Woodrow Wilson had finally decided to recognize First Chief Carranza as the de facto ruler of Mexico. This enraged Pancho Villa, his former star General, who had put forth his own competing claim to the Presidency. The hurt pride of Villa turned into hatred, and on March 9th, 1916, he crossed into US territory, exacting his revenge on the small town of Columbus, New Mexico.
Wilson’s response was immediate, demanding Mexico to pursue, capture, and exterminate Villa (Fabela, p. 162). Carranza, who had no love for Villa either, responded rather willingly. He reminded the US government of the 1880 and 1884 incursions of Native American rebels from the US into Mexican territory and the agreements reached in that time so that US and Mexican armed forces be allowed to pass freely from one territory to the other in order to pursue and punish the “bandits”, offering the same solution for the Villa problem. Through a diplomatic note, Mexico requested “the permission necessary to let Mexican forces cross into American territory in pursuit of those bandits, acknowledging due reciprocity in regard to forces of the United States crossing into Mexican territory if the raid effected at Columbus should unfortunately be repeated at any other point of the border” (original Spanish in Fabela, p. 163, English translation in AJIL Vol. 10 No. 3). Three days later, US Secretary of State Lansing replied accepting the Mexican proposal but adding that the United States “understands that (…) the arrangement is now complete and in force, and the reciprocal privileges thereunder may accordingly be exercised by either government without further exchange of views” (See AJIL Vol. 10 No. 3, and Fabela, p. 172 for Spanish translation).
Clearly, both states understood the note differently. Mexico sought to grant permission to the United States to enter Mexico if another raid occurred. The United States, however, assumed the right had been given outright. Whether this was an honest mistake is subject to controversy, particularly considering the admission of Colonel Frank Tompkins, a member of the Expedition, who called the stated interpretation of his own government “false, not founded on fact, and [in] danger of being challenged by the troops of Mexico”. Whichever the case, Wilson based the legality of his now famous expedition on this note (and his interpretation of it), not the unwillingness or inabbility of the Mexican government.
Carranza responded swiftly to what he considered a violation of Mexican sovereignty, and demanded the US sign an agreement with Mexico to regulate the presence of US troops in Mexican territory. Carranza would limit the number of men to one thousand; the number of days they could stay, to five; and the distance they could travel into Mexican lands, to 60 kilometres, demanding that the US started withdrawing its troops upon signing. The US, however, refused, contesting the proposed limitations and claiming any such agreement should not contemplate the troops already in Mexico, given their exclusively “punitive” nature (Ulloa, p. 69). This dynamic, with Mexico demanding the immediate, unconditional withdrawal of all US forces from Mexico and the US seeking a conditional, progressive one, allowing for suspension in case of further raids, would continue throughout all of 1916, in several different rounds of negotiations. Carranza’s position, however, is famous for its unwillingness to yield. In fact, at several points, Carranza himself specifically refused to refrend the agreements his negotiators reached, insisting that negotiations could only move forward once the United States agreed to the unconditional and immidiate withdrawal of troops (Ulloa, p. 96).
While negotiations were ongoing, on May 5th, 1916, Villa forces launched a new cross-border raid, this time in Glenn Springs, which prompted an additional 400 American troops to cross the border, this time clearly without any possible misunderstanding as to lack of Mexican consent. In a long and aggressive diplomatic note, dated May 22nd, Carranza exploded on Wilson: “[a]t all events, the Mexican Government, having expressed clearly its nonconformity with the crossing of additional troops into Mexico, is forced to consider this as an act of invasion of its territory and, in consequence, will be obliged to defend itself against any body of American troops on its soil” (See Fabela, p. 216, for Spanish original, and see here for English translation).
“[The] attitude of the American Government, which was the one which expected to cross the boundaries at such times as might be necessary, in pursuit of the bandits, is clearly indicating its intention of preparing to penetrate further into Mexican territory than the purposes of defense would seem to warrant. The punitive expedition from Columbus, as it has been called, did not have, according to statements of President Wilson, any further object than to capture and punish the band guilty of the raid, and was organized under the supposition that the Mexican Government had consented thereto. Nevertheless it has shown an attitude of manifest distrust toward the Mexican Government and a spirit of such absolute independence that it can not but justly be considered as an invasion without Mexico’s consent, without its knowledge and without the cooperation of its authorities”.
The United States, evidently, disagreed, and responded on June 20th, through a long (and also aggressive) note, stating that:
“It is in fact protection to American lives and property about which the United States is solicitous and not the methods or ways in which that protection shall be accomplished. If the Mexican Government is unwilling or unable to give this protection by preventing its territory from being the rendezvous and refuge of murderers and plunderers, that does not relieve this Government from its duty to take all the steps necessary to safeguard American citizens on American soil. The United States Government can not and will not allow bands of lawless men to establish themselves upon its borders with liberty to invade and plunder American territory with impunity and, when pursued, to seek safety across the Rio Grande, relying upon the plea of their Government that the integrity of the soil of the Mexican Republic must not be violated” (emphasis added).
These diplomatic exchanges are clear: Mexico argued the US needed Mexican consent to send troops into Mexico. The US, explicitly using the phrase unwilling or unable, felt it had a right to send troops into Mexico. At this point, some might feel inclined to vindicate Deeks’ claim that the Expedition is a 100 year old precedent for the unwilling or unable test. I would argue, however, that things are more complicated than this.
Customary international law evolves through breach. An established rule is challenged enough times by enough non-compliying states to consider the rule altered. This was something that First Chief Carranza was very well aware of. As negotiations were ongoing, he told his negotiator team he would not accept any arrangement that created “[any] precedent that may later be of fatal consequences for our motherland” (Fabela, p. 197).
Carranza’s stubborness proved successful. On early January, 1917, and with both countries at the brink of open war, Wilson yielded. By February 5th, his troops were gone; and in the words of Mexican negotiator Luis Cabrera, they were gone just as they had come: “without conditions, without an agreement that served as precedent, tacitly recognizing with their exit the injustice they had committed with their entry” (Ulloa, p. 97). The 1916 crisis, therefore, rather represents a case of resistance to an unwilling or unable rationale.
If we look at the state of international law in the Americas in the early 20th century, we will see few instances where cross border, non-consented uses of force against non-state actors were carried out without resistance (and this includes the Caroline incident!). In fact, as I briefly mentioned in Part I, the relevant treaties in force took a limited approach to the problem of non-state actors.
The 1928 Convention on Duties and Rights of States in the Event of Civil Strife, ratified by the United States, included four fundamental obligations, none of which related to a unilateral right to use force in the territory of another state in cases of unwillingness or inability. According to this treaty, member states were obligated to (i) prevent their inhabitants from participating in, gathering elements, crossing the boundary or sailing from their territory for the purpose of starting or promoting civil strife; (ii) disarm and intern every rebel force crossing their boundaries; (iii) forbid the traffic in arms and war material, except when intended for the Government; and (iv) prevent the arming or equipping of any vessel seeking to favour the rebellion. Similar obligations were included in the 1911 Agreement on Internal Commotions and Neutrality, in force between Ecuador, Bolivia, Peru, Colombia, and Venezuela.
In essence, no rule akin to the modern “unwilling or unable test” existed in contemporary treaty practice, and exceptions allowing for cross-border hot pursuit of outlaws were specifically agreed to by States through international agreement. This was the procedure followed by the United States throughout its dealings with Mexico. The 1916 Expedition was the exception, not the rule. After all, if any doctrinal debate existed at the time, it was not surrounding a test similar to modern “unwilling or unable”, but, rather, whether there existed a right to non-consented cross-border “hot pursuit” of rebels forces (i.e. a right to continue chasing a bandit across border lines), and even this was contested in contemporary scholarship).
Put in context, therefore, neither the 1932 Leticia Incident nor the 1916 Punitive Expedition demonstrate a consistent pre-Charter practice in Latin America regarding the unwilling or unable test. To the extent the 1916 crisis deals with it, it mostly shows how Mexico resisted the US’ attempts to unilaterally force the standard unto it. The only way the 1916 Expedition can be considered a valid precedent for the unwilling or unable standard is if one completely and utterly ignores both Mexico’s strong opposition to it and the treaties existing at the time.
This is not new. Its a pattern that seems to be repeating in today’s iteration of this same debate, with the positions of some powerful Western states taking preeminence over others. In the words of Jutta Brunnee & Stephen Toope, “[g]reat caution is warranted before accepting the argument that changes in customary international law can be produced through the practice of a small number of lead-States combined with the alleged acquiescence of a largely silent majority”. In any case, it should not be at all shocking that every time a group of invading States argues for the existence of a right to invade another, there will be another group of invaded States that argue against it. This was true in 1916, and is still true in 2018. Chosing one argument over the other based solely on familiarity is not good law. International law requires a more stringent test. Sadly, it seems 102 years has not been enough time for us to learn this lesson.