A matter of time: the role of timing in regulating military weapons

Abstract

This article explores the extent to which, and in what way, the timing of the legal regulation of battlefield weapons and practices matters. Specifically, we consider the challenges and opportunities of ex ante regulation—the imposition of rules and controls before a weapon or practice has been used in battle. We argue that timing shapes both the challenges confronted by and the opportunities afforded to norm entrepreneurs, particularly in terms of their exploitation of normative shock. Timing may also have an instrumental impact, influencing the degree to which perceptions of military utility either facilitate or impede the creation of battlefield regulation. We examine three cases to make this argument: the ban on explosive bullets in the 1868 St Petersburg Declaration; the regulation of bombardment from the air beginning in 1899; and the campaign which culminated in the 1995 Protocol on Blinding Laser Weapons. Examining these cases reveals that early regulation has value, despite some inherent challenges, providing a legal and normative template upon which more successful regulation can be built for the future.

Since formal regulation of the battlefield began, states and civil society actors have grappled with a basic tension: the instrumental drive to develop military technologies seen as essential versus the normative drive to regulate or prohibit morally problematic innovations. The apparent conflict between these two impulses represents a challenge for the regulation of weapons. Compounding this challenge, regulation of military technology is sometimes sought when the weapon in question does not yet exist.

This is the case today regarding ongoing efforts to pre-emptively ban fully autonomous weapons, with opposition stemming, in part, from a general apprehension regarding the practicability of ex ante regulation. Military technology does not always evolve as predicted, creating a risk that anticipatory controls will be overly or insufficiently restrictive. With consensus still lacking over some aspects of autonomous weapons technology, including how to precisely define it, it is increasingly clear that ‘technical progress is outpacing political safeguards and norm setting’.¹ Without understanding a weapon's actual practical effects, the instrumental objections to regulation may be difficult to overcome, and the normative urgency of controls may be difficult to convey. Are these likely difficulties severe enough to negate the benefits of ex ante regulation, or does the anticipated value of acting early justify the effort?

Evaluating the opportunities and challenges of ex ante weapons regulation is especially important today, given the dire state of arms control. States are withdrawing from longstanding weapons treaties² and have shown little interest in creating new ones.³ Detailed, historically informed analysis is needed to clarify what, if anything, preventive efforts at weapons control can achieve in such a climate, whether immediately or in the long term. However, there is a paucity of literature examining this precise question.⁴

In order to understand how ex ante regulation works, we selected three cases from the small universe of examples of pre-emptive weapons control: the 1868 St Petersburg Declaration, which banned explosive bullets; the regulation of bombardment from the air from 1899; and the campaign to ban blinding laser weapons, which culminated in the creation of the 1995 Protocol on Blinding Laser Weapons.⁵ These cases reflect different time periods, different stages in the evolution of arms control, different types of weapon and, most critically, different perceptions of ‘success’.

Examining these three cases reveals a number of important features of ex ante lawmaking. First, even in cases that cross different time periods (as our examples do), our research demonstrates that regulators consistently grapple with the competing pressures of a moral desire to regulate weapons (normative considerations) and a military desire to retain control over a potentially useful weapon (instrumental considerations). These considerations, however, face unique challenges in an ex ante environment. Regulatory efforts will invariably run into unpredictable and contingent events that may strengthen or weaken the case for governance. This is particularly the case when the regulation in question targets a weapon that does not yet exist: the weapon may evolve in a way that makes it especially useful, or especially egregious, or simply in a way that early regulators do not predict.

Second, these cases demonstrate that speaking about weapons regulation in terms of ‘success’ and ‘failure’ is, in most cases, overly simplistic. Our cases reinforce the need to analyse the legal regulation of weapons over the long term,⁶ rather than considering a snapshot of a particular moment in the regulatory journey. This broader analysis reveals that international law is more adaptive to the challenges created by contingency than is sometimes claimed. The creation of international law is an iterative process: the framework has the potential to adapt and readapt to changing needs and conditions. Regulation, in turn, is not a stand-alone process: it is enmeshed in and built upon antecedent laws and normative contestation. Extrapolating from this claim about the importance of antecedence and iteration, we argue that contrary to much of the conventional wisdom, ex ante regulation has significant value, providing a legal and normative template upon which future, more successful regulation can potentially be built. These findings enhance our understanding of both the character and challenges of international law, and provide important insights into whether and how to regulate emerging and evolving weapons.

We proceed in three sections. First, we outline our conception of ex ante regulation and explain how it interacts with the instrumental and normative dimensions of weapons control. We then develop these findings empirically through our three cases. In the concluding section, we demonstrate that analysis of ex ante weapons control helps refine our understanding about contingency and the law, and reinforces the view that early regulation can perform a valuable foundational function for future regulation, even when it appears to have failed. We also build on our arguments about weapons control law to propose new possibilities for research in the field.

International law and the challenges and opportunities of weapons regulation

Ex ante regulation, international law, and weapons control

We define ex ante regulation as efforts to legally control a weapon that does not yet exist or has not yet been used extensively on the battlefield. Examining ex ante lawmaking is a useful way to analyse the International Relations (IR) concept of legalization. As Hurd notes, even in the absence of formal international law, issues do not exist in some kind of legal ‘black hole’. Contrary to much of the IR literature, there is not always a sharp difference between a legalized and an un-legalized issue, because ‘the practices of international law fill these ostensibly prelegal spaces’.⁷ Our cases confirm this, highlighting historical episodes of weapons regulation driven not by a clear understanding of the weapons' characteristics or effects, but rather by a clear expectation of what was and was not permissible in battle.

Our three cases have been selected from the surprisingly small number of historical episodes where actors sought to regulate a weapon ex ante. The regulation of explosive bullets, aerial bombing and blinding lasers cross different time periods. These representative cases showcase different levels of regulatory ‘success’ and reveal different processes of contestation. Consideration of this variance allows us to contribute to important debates in IR scholarship. As van Dijk argues, there is a tendency in western scholarship on international humanitarian law to see laws evolving along a historical trajectory that becomes progressively more just.⁸ Our cases demonstrate that would-be regulators, regardless of the century, continue to face similar pressures stemming from the tension between humanitarian imperative and military necessity, and encounter similar challenges and opportunities shaped by the timing of such efforts.

We argue that international law is inherently iterative. It builds upon itself. New weapons, even technology that is seemingly without precedent, does not enter the legal universe in a vacuum. As Hurd argues, new law builds on previous instances not only of law, but also ‘past state practice on the subject, traveaux preparatoires, analogies and examples from other legal systems, consequentialist claims, [and] state interests’.⁹ Bode and Huelss concur, highlighting the role of ‘complex and largely unnoticed’ norm-making practices—including the technical characteristics of the weapons system itself—that work over time to create and shape standards of appropriateness.¹⁰ We argue that analogies are crucially important to understanding how actors develop law in response to novel weapons, through the construction of ‘shock’ events.

We also argue that law may have enduring value even when it appears to have ‘failed’. Mantilla also argues that examining failed cases of international law can yield insights into what makes law succeed;¹¹ we extend this analysis, arguing that in weapons law, the ultimate success of these regulatory efforts cannot be determined in the short term—the norms, practices and law subsequently built upon and shaped by these foundational efforts also matter, and can only be recognized over the longue durée. To be clear, we do not argue that law invariably develops in a more progressive direction. Justice is never guaranteed and often absent from weapons law. But positive developments are achievable through incremental action that builds upon historical antecedent.

Ex ante regulation provides a sharp instrument through which to consider international law creation because, without definitive demonstrated effects, the debate over the balance between a weapon's utility and its moral impact is particularly challenging. As Kinsella and Mantilla argue, international humanitarian law has always ‘been composed of tense, plastic, and contested layers of “agreement” about how to balance humanitarianism and military necessity, the pursuit of war and the protection of its participants, ensuring both the destruction and the security of the population’.¹²

We argue that examining ex ante law provides a particularly effective window to consider the drivers of regulation that create the ‘layers of agreement’ to which Kinsella and Mantilla refer. Because the weapon does not yet exist and there is not necessarily evidence of its impact, lawmakers must make a compelling case for or against regulation. Accordingly, examining ex ante regulation reveals the relationship between both instrumental and normative drivers of regulation. Instrumental drivers focus on the demonstrated or anticipated military benefit of a weapon, while normative drivers focus on the moral arguments stemming from its humanitarian impact. Both drivers influence whether and how weapons are regulated, and both are influenced in turn by the timing of regulatory efforts.

Instrumental factors

The rules and standards of war have always been heavily influenced by the instrumental considerations of belligerents. As Royse argues, a weapon will ‘be restricted in inverse proportion, more or less, to its effectiveness … the more efficient a weapon or method of warfare the less likelihood there is of it being restricted in action by rules of war’.¹³ Price illustrates this instrumental dimension in relation to chemical weapons. Asphyxiating gases, he argues, were easier to regulate at The Hague because the presumption was that these weapons were not militarily essential. Likewise, biological weapons were banned, as the result not only of a moral imperative but also an instrumental calculation: the United States concluded that biological weapons were of ‘limited tactical utility on the battlefield’.¹⁴

However, the concept of military utility is neither objective nor static. Different states have different warfighting preferences. States develop, procure and use weapons in line with their preferred way of war. States that view a particular weapon as essential to their military needs are more likely to resist burdensome restrictions. States that cannot develop those same weapons, or conclude that they will be militarily disadvantaged by their proliferation, will be less likely to oppose proposed controls. This dynamic can be observed in relation to the regulation of autonomous weapons. Actors such as the US view the technology as indispensable to the effective prosecution of future war¹⁵—a view reflected in their consistent opposition to an international legal ban. It is important to observe here that predictions of military utility are often mistaken. The battlefield effectiveness of a particular innovation may—and often does—exceed or fall short of early estimations. The ambiguity of technological change is also relevant when considering the technical challenges of regulation.¹⁶ It is hard to predict how new technology will evolve; once a technology exists, it is often used in a way that differs markedly from initial expectations, or in conditions very different from the regulators' early imaginations.¹⁷ History is replete with examples, as Maas notes, of ‘jumping the gun’ in relation to the regulation of technologies ‘(erroneously) thought to be imminent’.¹⁸ Similar concerns have been expressed throughout the debate over autonomous weapons, with some favouring a wait-and-see approach over early regulatory intervention.¹⁹

There is some validity to this technical criticism of pre-emptive regulation. Imposing restraints on innovation that has yet to emerge is undoubtedly hard to do, and harder still to do well. Weapons and practices may evolve in an unexpected direction, rendering the newly crafted law over- or under-inclusive. Critically though, this challenge seems less severe when international law, and the regulation that helps constitute it, is conceptualized not as a single episode of success or failure, but rather as an iterative process. Law is adaptive, and more capable of responding to technological changes in the weaponry it is created to govern than is often credited. Before exploring this in the next section of this article, we consider the normative dimensions of weapons regulation. As this section will further confirm, the disadvantages of ex ante regulation are real, but frequently overstated.

Normative factors

There is an extensive literature on the role of norm entrepreneurs in the development and creation of regulation aimed at controlling weapons and practices in war.²⁰ Norm entrepreneurs—state and non-state—are vital actors in the creation of international law, and they are present, in one form or another, in all forms of weapons governance. In ex ante regulation, the nature of timing shapes both the opportunities for norm entrepreneurs and the challenges they face.

One important tactic of norm entrepreneurs is the exploitation of ‘shock’—the outrage generated among the wider public in response to a weapon's humanitarian impact. When properly deployed, ‘shock’ can frame the terms of the debate, draw the attention of a wider audience, legitimate existing concerns and undermine opposition to reform.²¹ A ‘shock’ event might highlight the extent to which a new technology in war is in violation of existing rules, or it might expose the law's failure to adapt to a significant change in the character of war. In either case, shock helps convey the scale and urgency of the challenge, as well as the potential consequences of inaction.

A weapon that does not yet exist, or that exists only in a rudimentary form, makes the successful invocation and exploitation of shock difficult. There will be no practical example of misuse to draw audience attention or to address through stricter controls. Ex ante regulatory campaigns have a hypothetical quality that makes it easier for states, particularly those who stand to benefit from the technology, to dismiss its potential harm (cynically or otherwise). Evidence of this ‘urgency gap’ can be found throughout the history of weapons regulation, including the 1899 Hague Convention proposition to ban ‘the use of projectiles the sole object of which is the diffusion of asphyxiating or deleterious gases’. With the shells not yet developed, US delegate Admiral Alfred Thayer Mahan argued that it could not yet be proved that the technology was ‘inhumane or unnecessarily cruel’.²² Similar arguments have been made more recently in the context of autonomous weapons, with opponents of a pre-emptive ban accusing campaigners of exaggerating the negative, dystopian effects of the technology.²³

How can regulators argue that a proposed innovation is intolerable if no practical example of misuse can be identified? Our cases show that in the absence of demonstrated impact, norm entrepreneurs often use analogy, associating their ex ante campaigns with already existing banned weapons. For example, in the case of blinding laser weapons, campaigners made a strong link to chemical warfare to garner support. While ex ante regulators often lack formal legal structures around which to build their campaign, they can harness legal and normative ‘raw materials’, such as analogy. We find, even in the case of the 1868 St Petersburg Declaration, strong evidence for the fact that lawmakers use various normative and legal ‘construction materials’ to buttress their case for regulation. As time evolves (as evidenced by our cases), the development of weapons law creates more and more legal material that can be used to develop and bolster law for technologies that are unregulated.

Regardless of precisely when the regulation of a particular weapon is attempted, the creation of meaningful controls will be made easier when the instrumental benefits of the weapon in question are perceived to be few and minor and the normative problems to be many and severe. Timing does matter, however, in how both instrumental and normative considerations are formed and navigated. The next section empirically validates these claims, through consideration of three different cases of ex ante regulation.

Ex ante regulation

Explosive bullets

The St Petersburg Declaration,²⁴ which sought to prohibit explosive bullets, was the first attempt to ban a weapon in advance of its widespread use on the battlefield and the first specific attempt to create a formal international agreement prohibiting a weapon's use in war.²⁵ The declaration was created in response to the development of new exploding bullets which detonated on impact with a hard surface. Initially designed to hit ammunition wagons,²⁶ by 1867 the functionality of the bullets had advanced to also explode on contact with softer surfaces. If fired at humans, these weapons caused devastating wounds;²⁷ ‘the expanding gases and the scattering of fragments from the casing’ would cause the victim to die in agony.²⁸ The Russian government was willing to renounce the use of these weapons if other states did so, and in 1868 convened a conference on the topic.²⁹ The resulting St Petersburg Declaration had 19 signatories and only applied to conflicts between them. In its preamble, the declaration also clarified the core international humanitarian law principles of unnecessary suffering and proportionality. The agreement was noteworthy because, as Vagts explains, explosive bullets, ‘although reprehensible from a humanitarian point of view, at the same time [were] certainly not devoid of all military utility’.³⁰ Indeed, the Russian proposal possibly stemmed from a desire to forestall the wide-ranging adoption of the weapon.

The evolution of the legal control of bullets provides a useful demonstration of how law evolves alongside technological development. Advances in technology meant that the creation of other munitions, with a similar impact but falling outside the remit of the St Petersburg Declaration, was inevitable. The expanding—or ‘dum-dum’—bullet posed the first such challenge. These bullets, which expanded on contact rather than exploding, reached the agenda of the Hague Conferences of 1899. During these meetings, the bullets' illegality was confirmed, partly because of their similarity to explosive bullets and partly owing to the belief that they fell under the same pre-existing prohibition.³¹

Because of their status as customary international law, the St Petersburg and Hague declarations have been embedded in military practice, making them particularly robust. Any new bullet adopted by the US military, for example, must pass a formal legal review to determine its compliance with international law (particularly on the questions of necessity and superfluous injury),³² ideas which were first articulated in the preamble of the St Petersburg Declaration.

Aerial bombing

The first attempt to prohibit bombing from the air was directed against military balloons, during the 1899 Hague Convention. Troubled by the inaccuracy of this crude technology, participants agreed to prohibit its use for five years.³³ The imposition of a five-year expiry on this prohibition was grounded in the belief that a more accurate dirigible balloon would soon be available. The commitment of major powers to meaningful regulation was soon tested by advancements in air power, including the invention of the aircraft. The military interest this engendered could be witnessed at the Second Hague Peace Conference of 1907, when states failed to generally ratify a renewal of the five-year prohibition on bombing from the air. Instead, the land warfare convention was amended to include aerial attacks: ‘The attack or bombardment by whatever means of towns, villages, dwellings or buildings which are undefended is prohibited’.³⁴

In April 1911 the Institute of International Law met to discuss the law of aviation. The final conclusion reached was that while aerial warfare should be permitted, its use was to be contingent on ‘not creating greater dangers [than those imposed by land or sea warfare] for persons or property not engaged in war’.³⁵ In November 1911, during the Italo-Turkish War, the first successful bombing from an aircraft took place.³⁶ By the end of the First World War, all major belligerents had conducted bombing raids against enemy cities.

Throughout the interwar period there were renewed attempts both to regulate aerial bombing and to prohibit it outright. The 1922 Washington Conference resolved to appoint a commission of jurists to consider the applicability of international law to methods of warfare that had emerged since 1907. The commission subsequently released the 1923 Hague Rules of Air Warfare. Article 22 of the rules prohibited ‘any air bombardment for the purpose of terrorizing the civil population or destroying or damaging private property without military character or injuring non-combatants’.³⁷ These rules, however, were never adopted by the powers concerned. During the 1932–4 League of Nations World Disarmament Conference the US proposed the ‘Hoover Plan’, which if successful would have required the abolition of all bombing aircraft. The proposal was rejected, with some concerned that the dual-use potential of civil aviation would undermine the value of such a prohibition.³⁸

During the Second World War major powers were capable of massive aerial bombardment while still lacking the technological capacity to meaningfully distinguish between military and non-military targets. Despite the controversy of this bombing, Boothby suggests that ‘it was not until 1977 with the adoption of Additional Protocol I [to the Geneva Conventions] that formal international agreement was achieved on a set of rules that would regulate the engagement from the air of targets on land’.³⁹

Blinding lasers

Laser technology developed rapidly in the late twentieth century. By the 1990s, lasers had several battlefield applications: they could be used to aid targeting, range-finding, detection, communication and target destruction.⁴⁰ It was further theorized that they could be developed to blind enemy military personnel. The governments of Sweden and Switzerland were particularly concerned about the advent of these weapons, and led a campaign, alongside the International Committee of the Red Cross (ICRC), to ban their use. In 1986, at the ICRC's 25th international conference, Sweden and Switzerland produced a draft resolution banning blinding lasers on the grounds that they were contrary to international humanitarian law, as they caused unnecessary suffering and superfluous injury. The two governments also brought the issue to the United Nations General Assembly in its 41st and 42nd sessions.⁴¹ Sweden and Switzerland, in concert with the ICRC, took the view that the best way to ban blinding laser weapons would be to add them to the UN Convention on Certain Conventional Weapons. The ICRC began to investigate the expert literature on blinding weapons and hosted a series of meetings to discuss their potential banning.⁴² The ICRC campaign was soon supported by a range of other non-governmental organizations, including Human Rights Watch and organizations working in support of the blind.⁴³

These lobbying efforts, especially those on the part of Human Rights Watch, had significant purchase in the US,⁴⁴ where the Department of Defense reversed its position on blinding lasers and cancelled the US research programme shortly before the next meeting at which the technology was to be discussed.⁴⁵ The defence secretary, William Perry, announced in 1995 that the Department of Defense had prohibited the use of laser weapons for blinding and supported the negotiation process, although the position was later revised to state that while intentional blinding should be banned, ‘accidental injuries caused by lasers designed for other purposes’ should not.⁴⁶ Also in 1995, at the first Review Conference of the 1980 Convention on Certain Conventional Weapons, blinding laser weapons were banned through the addition of a fourth protocol to the Convention.⁴⁷ Blinding lasers are regarded as the first weapon to be banned in advance of use on the battlefield since the nineteenth century.⁴⁸

Instrumental considerations

As argued above, campaigns to ban or regulate weapons ex ante have the highest likelihood of success when the anticipated negative moral effects outweigh the anticipated military benefits. This was the case with blinding lasers, a controversial military innovation aborted in the 1990s due, in part, to a shared recognition among states of their limited battlefield utility.⁴⁹ Calculations between military utility and negative humanitarian impact are rarely so straightforward. In the context of explosive bullets and manned aerial bombing, we can find evidence of military utility perceptions both aiding and obstructing regulatory efforts.

Beginning with explosive bullets, during this period states were committed to preserving the possibility of technical developments that could be of assistance on the battlefield, and were wary of regulation that might restrict such advancements. The British were especially interested in how technical proficiency could help defend the empire, and saw themselves as ‘dependent on technology to compensate for a smaller army both in European conflicts and in imperial wars’.⁵⁰ Explosive bullets fit this logic, and perceptions of indispensability led Britain to insert a clause into the St Petersburg Declaration that made its restriction only applicable to signatories. The St Petersburg Declaration had 19 parties, all European states, so this clause had the effect of making the declaration applicable only to intra-European war, leaving Britain free to use explosive bullets in imperial conflicts.⁵¹

British notions of military utility also influenced the next stage of bullet evolution—expanding, or dum-dum, bullets. Before the 1899 Hague Conference, Britain forcefully asserted its belief in the battlefield efficacy of the technology.⁵² The British delegate Sir John Ardagh infamously argued that, while such bullets were not necessary in ‘civilized’ wars, where the wounded could be counted on to stop fighting and await rescue:

It is very different with a savage. Even though pierced two or three times, he does not cease to march forward … and before anyone has time to explain to him that he is flagrantly violating the decisions of the Hague Conference, he cuts off your head.⁵³

Perceptions of military utility also had an inverse effect, however, at least partly incentivizing Britain's adversaries to support prohibition. British commentary at the time argued strongly that the other European powers at The Hague were unfairly seeking to deprive Britain of a useful weapon.⁵⁴

Aerial bombing regulation also demonstrates how perceptions of military utility influence both opposition and support for restrictions at the ex ante stage. Kuhn argued in 1910 that a key factor in the acceptance of the 1899 prohibition on aerial bombing was the lack of progress ‘made for the offensive use of aircraft’.⁵⁵ This was no longer the case by 1907, when key states at the Second Hague Peace Conference opposed a continuation of the ban.⁵⁶ According to Kuhn, the opposition of France, Germany and Italy was motivated primarily by the ‘relative advance that each had made in aeronautics’.⁵⁷ These states had concluded that manned aerial bombing had the potential to be a decisive mode of warfare and, crucially, that such a development would advantage them specifically.

The predicted battlefield value of aerial bombing had the opposite effect, at least initially, among those who benefited from the existing military status quo. In 1899 Lord Wolseley encouraged the British armed forces to explore the practice of dropping bombs from balloons, arguing that ‘in war superior armament compensates for lack of numerical strength’.⁵⁸ By 1907, however, the British position had notably shifted. Concerned that air power might threaten its naval supremacy, Britain, having refused to ratify the 1899 prohibition, supported an extension of the ban on aerial bombardment.⁵⁹ This continued beyond the First World War, with Britain leading efforts to prohibit aerial bombing during the League of Nations Disarmament Conference. Imperfect knowledge—specifically, the failure of Britain to recognize the potential of early air power to complement rather than nullify its naval defences—was a resource that regulators could exploit. Just as with explosive bullets, this example complicates the conventional wisdom, which typically frames high utility as an inevitable impediment to regulation. The governance of new weaponry, like all other aspects of international law creation, is a political process. In some instances, the high anticipated utility of a new weapon may aid the creation of ex ante regulation, drawing support from states that seek to suppress military innovation for instrumental reasons.

This does not, however, alter the fact that the optimal grounds for prohibiting a newly emerged weapon in war are a determination of low utility, coupled with a high—preferably inherent—tension with moral and legal norms. One challenge for those seeking regulation at the ex ante stage is the prevalence of techno-optimistic accounts of emerging technology that significantly exaggerate its likely effects. In such cases, the creation of explicit preventive controls is unlikely, especially when stronger state actors become convinced of the need to invest.

Normative considerations

One of the challenges of ex ante regulation is that a weapon that does not yet exist does not have any demonstrated ill effects around which to build a campaign. Those seeking regulation have to construct shock. Construction is easier when regulators can credibly link the weapon under review to something already banned or stigmatized. Price describes this process as ‘grafting’, referring ‘to the mix of genealogical heritage and conscious manipulation involved in such normative rooting and branching’.⁶⁰ Shock construction is a challenge for regulators at the ex ante stage, as they must stress the urgency of the potential challenge while avoiding groundless exaggeration.

In the case of blinding lasers, constructing shock was difficult because the weapons did not exist and their impact was hard to imagine. The early ICRC/Swedish attempts to create a ban were thwarted ‘because the vast majority of States were unaware of developments and thought that such weapons were science-fiction’.⁶¹ The ban had to overcome not just the direct opposition from governments who might be interested in using such weapons, but also the ‘general indifference of many other governments that were not interested in a potential problem the urgency of which had not been proved’.⁶² Campaigners, facing regulatory reluctance, constructed shock through analogy. They linked the issue of blinding lasers to the use of chemical weapons in the First World War. In 1994, the ICRC released a pamphlet entitled Blinding weapons: gas 1918 … lasers 1990s?. The publication made a deliberate connection to the blindness caused by gas weapons during the First World War, linking past horrors of war with the significant, if yet to be realized, danger of lasers.⁶³

Norm entrepreneurs also sought to solidify support for the blinding lasers ban by embedding it within existing law, particularly the principles of unnecessary suffering and superfluous injury. To do so, campaigners had to demonstrate that blinding as an injury was particularly egregious, causing extensive psychological damage, and being impossible to treat medically as well as being out of step with even the most extreme forms of punishment. Anderberg et al. noted in 1992 that while the death penalty was widely accepted internationally, blinding as a punishment was no longer used in any state.⁶⁴

Aiding these efforts were advances in laser technology, which helped push the debate from science fiction into reality. In early 1995 two types of blinding laser weapons with anti-personnel usage possibilities were advertised for sale. This had ‘the beneficial effect of eliminating a certain indifference on the part of many government ministries which still believed that such developments would probably take place only in the distant future—if at all’.⁶⁵ The encroaching reality of these weapons appears to have focused the minds of regulators.

Constructing shock in relation to the ban on explosive bullets occurred in a different context. As Fazal has noted, the early stages of codifying the laws of war were dominated by states and militaries, rather than civil society actors.⁶⁶ States only had to convince each other—a method of campaigning less contingent on the creation and exploitation of normative shock. Nonetheless, the speed and nature of military change in the nineteenth century was shocking, especially in the field of small arms. Until the 1840s, small arms had been effectively unchanged for a century; after the 1840s, the speed of change was rapid and bewildering.⁶⁷ The ‘dizzying revolution in military technology raised numerous questions of law and … led to doubts about the legitimacy of the employment of these weapons’.⁶⁸ Russia's decision to promote the ban may have been prompted by the memory of the Crimean War, and specifically by the 256,000 casualties endured by the state.⁶⁹

As an examination of ex ante cases further reveals, the shock generated by new weaponry is rarely sufficient in itself to catalyse new regulation. Equally important is the wider normative and regulatory landscape within which the shock event occurs. Just as with blinding lasers, early efforts to regulate aircraft bombing benefited from an antecedent objection to the practice. The decision by France in 1799 to disband its corps d'aerostiers was motivated by questions over the practical value of military balloons, combined with Napoleon's perception that their use ‘was not a gentlemanly act’.⁷⁰ This ethical discomfort with bombing from the air, particularly in relation to civilian targets, endured throughout the early decades of the aircraft.⁷¹

The odium of aerial bombing was powerful and lasting, but was insufficient to motivate the creation of meaningful regulatory controls in the first half of the twentieth century. This is partly explained by the normative environment within which the bombing took place. Manned aircraft bombing was first conducted in 1911 during the Italo-Turkish War,⁷² and evolved in an imperial context as a punitive tool to suppress ‘native rebellion’.⁷³ The aerial bombing of cities similarly developed in the context of existential war in Europe. Both contexts—colonial violence and total war—functioned to override the significant moral pressures to restrict the practice, which were intensified by the 1937 bombing of the Basque town of Guernica by German and Italian air forces during the Spanish Civil War.

The construction of shock is important to understanding how weapons can be banned prior to their military use. There needs to be a concerted campaign to raise awareness of the potential horrors of a weapon. In the absence of a contemporary shock, norm entrepreneurs will look to construct it, drawing on historical antecedents where possible. But the wider normative and regulatory environment also matters. The success of an ex ante campaign will be determined just as much by the shockability of the targeted audience as by the nature of the weapon in question.

Law and timing

Examining the regulation of weapons in advance highlights the fluid and contingent character of instrumental and normative considerations. It also provides a useful reminder that international law is an iterative process. We argue that, when weighing the challenges and opportunities of military regulation, our scope of analysis needs to widen, to encompass the historical landscape—specifically, the inertial effect of preceding regulation. International law advances in conversation with legal precedent, along with underlying norms, practices and debates—both settled and unsettled. These antecedents are instructive, helping set the conceptual and practical boundaries for what is possible to achieve with new law. What this means is that efforts to control new weapons should be judged twice—first, as a direct response to an explicit regulatory need, and second, as a building block in the construction of future regulation, whether of the same issue or a related one. Initial failures can evolve into partial successes; partial success can evolve into more comprehensive ones.⁷⁴ Recognition of the iterative dimensions of international law has important implications for the way in which we approach the question of timing in the regulation of military weapons. Most crucially, it serves to highlight the frequently overlooked value of acting early.

The value of early regulation

Our cases reveal that early regulation does have drawbacks, including the difficulties inherent in predicting how technology will evolve, and the challenges that are associated with persuading relevant actors to impose a ban on something which has yet to reveal any practical ill-effects. For some, these difficulties are sufficient evidence of the futility of ex ante regulation. We argue that despite these challenges there is considerable value in early efforts because of the way in which international law functions. Our cases demonstrate that international law is surprisingly adaptable to the challenges created by unexpected technological evolutions.

It is firstly important to clarify that it is possible for early regulation to get the scope and approach of regulation ‘right’, or at least sufficiently right, on the initial attempt. The regulation on explosive bullets themselves has gone mostly unchallenged. The ex ante ban on blinding lasers has significantly reduced the harmful effect of this technology. Prevention, as the saying goes, is always preferable to a cure. This is especially true in the context of war, where the consequences of delayed regulation are often tallied in dead or permanently disfigured bodies.

Even when early regulation fails to achieve its objectives, either fully or partly, it can have enduring value. The tracks of regulation are deeply grooved. Once a particular weapon is set on a regulatory pathway, it is hard to dislodge, and once something becomes regulated, it is rare for it to become unregulated.⁷⁵ Regulatory norms can be modified, and are sometimes replaced, but rarely disappear.⁷⁶ One advantage of early regulation is that regardless of its immediate benefits or limitations, it moves a weapon into the regulatory space. Early regulation, whatever its quality, will then have the potential to exert a ‘gravitational pull’ on future technologies of a similar nature, increasing the likelihood of their regulation. This is especially marked because of the tendency of lawmakers to rely on previous law and analogy as the ‘construction materials’ to create new law in order to control a novel weapon.

This ‘gravitational pull’ is evident in how the explosive bullet ban led to the establishment of a prohibition on a related technology. In the late 1890s the British began manufacturing an expanding bullet at the Dum Dum munitions facility outside Calcutta, eventually known as a dum-dum bullet. The decision was in response to the widespread use in the empire of the Lee–Enfield rifle—a weapon that had been considerably less deadly than anticipated, causing small bullet wounds that rarely led to immediate death.⁷⁷

In response, the dum-dum was developed: a bullet that did not explode, but rather expanded on the point of entry and caused considerable damage. There were early suspicions that ‘the Dum Dum bullet had contravened international law’; these were raised in the UK Parliament in 1898. The British government argued that the bullet did not violate the St Petersburg Declaration because it was not explosive, but found it more difficult to rebut the charge that the bullets ‘ignored the spirit of the St Petersburg Declaration’.⁷⁸ Spiers argues that the government's attempts to justify this contradiction became ‘tortuous to the point of absurdity’.⁷⁹ Eventually, the president of the commission charged with examining the issue of dum-dum bullets argued against British reservations, pointing out that the prohibition ‘was rather akin to the language adopted—and approved by Britain—in the Declaration of St Petersburg’.⁸⁰ Other delegations at The Hague were enthusiastic about the prohibition because they believed it would primarily affect the British, and failed to anticipate that the prohibition would have significant impacts on the general development of international humanitarian law.⁸¹ By 1907, British concerns over the reputational cost of using this technology led them to accept a prohibition, even though some military voices still supported its use.⁸² In more recent times, other types of exploding or fragmenting projectiles have also been banned, continuing the process begun at St Petersburg. There is now a prohibition on ‘the use of fragments made of materials that cannot be detected by X-ray machines’, a rule grounded in the desire to limit unnecessary suffering in war, and descended from the bullet regulations discussed here.⁸³ As this trajectory of bullet regulation reinforces, ex ante efforts—even when partial, flawed or cynical—can develop into something more. The prohibition on explosive bullets was far from perfect. It was narrowly crafted and failed to anticipate other small arms munitions that would generate similar injuries in opposing combatants. Ultimately, however, it did not need to be perfect. It began a regulatory process that successful captured a range of different but sufficiently similar technologies.

Ex ante law can have a value beyond its immediate impact. This may even hold true in cases where the early regulation in question is regarded at the time as a categorical failure. This can be seen in the context of manned aerial bombing. Efforts to restrict the use of military balloons and early aircraft helped establish and preserve a normative and legal framework that could be returned to and built upon when states were more receptive to the imposition of meaningful limits on the use of force.

The First World War marked the beginning of the systematic and indiscriminate bombing of urban areas from the air. The prewar stigmatization of such conduct, however, remained in force. This is evidenced by the pressure faced by states to provide special justification for their violence. When the French Air Force conducted its first large-scale bombing of German factories in 1915, they defended the attack on the basis that the factories were producing asphyxiating gases.⁸⁴ Moral and legal mechanisms for restraining bombing from the air were too weak to meaningfully restrain belligerent forces, but not so weak as to be disregarded entirely.

At the end of the conflict, manned aerial bombing was brought further into the regulatory space by the 1923 Hague Rules of Air Warfare. As Biddle notes, though unratified, these rules built on the frameworks of 1899 and 1907 to ‘erect a moral trellis that could not be entirely ignored and would not be entirely torn down, even in the midst of brutal combat yet to come’.⁸⁵ The League of Nations invoked the 1923 Hague Rules in its denunciation of the aerial bombings of Spain and China in the 1930s.⁸⁶ During the early stages of the Second World War, the influence of the Rules was again reflected in the pledge by states to avoid civilian targeting from the air.⁸⁷ When determining the legitimacy of the use of atomic bombs against Hiroshima and Nagasaki, the District Court of Tokyo stated in 1963 that the Rules constituted international customary law.⁸⁸ For most of the twentieth century the regulation of bombing from the air fell tragically short of its goals, but these efforts were not without value. Even in failure, they ensured that a normative and legal pathway endured in relation to the protection of non-military targets in war.⁸⁹

Our cases also demonstrate that considering the ‘success’ or ‘failure’ of weapons regulation depends on many contingent factors. First, as Percy and Sandholtz have argued, whether or not a law is deemed to be influential will depend greatly on the time period when it is examined, and therefore it is essential to examine the rule over a long period.⁹⁰ Confidence in the success of the ban on expanding bullets at The Hague was low, and yet over time this ban has expanded and strengthened. Second, even when law appears to have the most unprepossessing beginnings—being derived from instrumental or even cynical concerns—it still can become binding and influential. Mantilla has argued that sometimes states create ‘face-saving institutions’ in order to ‘divert attention away from robust cooperation towards weak compromises’.⁹¹ We argue that the history of ex ante regulation shows that even these weak compromises may still end up exerting influence, and may eventually form the raw material for genuine and more substantive cooperation in other contexts.

The time limit imposed in 1899 at The Hague on the ban on projectiles launched from the air duly expired, but the idea that bombing from the air ought to be the subject of legal regulation never went away, even if it has always been difficult to ensure the protection of civilians in this context. Britain argued, with some evidence, that the enthusiasm of other nations for the ban on dum-dum bullets stemmed from a desire to suppress Britain's military advantage—but even if this were true, ultimately the regulation came to pass and has been remarkably successful. Early regulation, no matter how it is achieved, can lead to success, even if it appears to be an initial failure. To be clear, early regulation does not guarantee success over the longer term, but it does present a possibility that should not be ignored for more durable and effective regulation.

Conclusion

In this article we have explored the role of timing in the legal regulation of battlefield weapons and practices. Specifically, we have considered the challenges and opportunities of ex ante regulation—the imposition of rules and controls before a weapon has been used in battle. Examining the historical cases of explosive bullets, aerial bombing and blinding lasers, we have shown that timing matters, shaping both the normative and instrumental dimensions of anticipatory action. Our research into three cases of ex ante regulation demonstrates that considering the timing of regulation in general has considerable promise for further scholarship that wishes to consider how international law is created. For example, examining how ex ante regulation differs from post facto regulation can reveal how states use the building blocks of law in different or similar ways. Working at the nexus of history, IR and international law scholarship on the timing of weapons regulation has the potential to yield new insights about how, when and under what circumstances successful weapons regulation occurs.

Our analysis of timing also provides further confirmation of the iterative nature of international law. The desire to regulate particular weapons, and the likelihood of securing such regulation, is guided to a significant degree by the past—antecedent ideas and principles, formal and informal legal precedent, and prior normative contestation. Military regulation is a longstanding, ongoing and unfinished process. Consequently, early regulatory efforts almost always have some value. Even when such efforts ‘fail’ (i.e. regulation is successfully opposed or deficient in nature) they provide a legal and normative template upon which future, more successful regulation can be built—they are the building blocks that mean that no law appears in a vacuum. A greater focus on temporality is essential as we grapple with emerging battlefield technologies of increasing complexity.

Given the challenges facing the legal controls over weapons and techniques of war, this research agenda has considerable policy value. In the case of autonomous weapons systems, state and non-state actors remain divided over whether and how to control the development and use of the technology. With crude autonomous drones now being fielded in battle, including by Ukraine in its ongoing defensive war against Russia,⁹² the years-long effort by campaigners and civil society actors to pre-emptively ban the technology appears to have unequivocally foundered. Compounding this seeming failure is the radical increase of interest in autonomous weapons specifically, and military artificial intelligence more broadly, by a range of actors, including the United States, China and Russia.⁹³ We do not suggest that pessimism over the unwillingness of states to accept meaningful controls over this technology is unfounded: it is likely that autonomous weapons will generate unjust harm on the battlefield. What our research does contest, however, are fatalistic assumptions of regulatory finality. The failure of states to agree to a pre-emptive ban on this technology does not rule out the possibility of future controls. The regulation of autonomous weapons will be an ongoing process of some uncertainty. Conditions may worsen—problematic precedent may calcify into new and worse battlefield standards. Critically, though, the alternative is also possible—the ‘shocking’ effects of autonomous weapons, changes in the broader normative environment, or a failure of the technology to meet its military potential, may open spaces for regulation that do not currently exist. If so, then the policy and advocacy work done ex ante, to better define the technology and clarify expectations of civilian protection and ‘meaningful human control’, will matter. The failures of today may become the construction material of future success. We cannot lose sight of this possibility in our research, nor in our advocacy. Our findings reinforce this, clarifying what is risked—and potentially gained—in relation to the timing of weapons regulation.

Footnotes

Author notes