Information gathering

Data on human and environmental consequences is often central to debates about the legality or wider appropriateness of weapons. By demanding access to state-held information or compiling field data of their own, groups within civil society can identify problems hitherto ignored or they can develop a deeper understanding of problems already identified. Therefore, information gathering can be vital to initiating consideration of a particular issue, or for the development of arguments around an issue that has already been established.

Even the most basic forms of information relevant to weapons’ effects can be contested and problematic. The topic of deaths from armed violence illustrates the importance of information gathering by members of civil society, and the relatively weak practice of states. For example, deaths resulting from the 2003 Iraq War have been a prominent topic of international public concern. With the absence of efforts by parties to the conflict, including the US, the UK, and others, to produce figures on the numbers of civilians killed (indeed with active efforts to remain ignorant about this matterFootnote ¹³) it has fallen on those in civil society to produce figures. Largely based on the systematic evidence of media accounts, the NGO Iraq Body Count not only produced an accessible listing of direct civilian deaths inflicted since the intervention, but also has been able to break them down by perpetrator and weapons type.Footnote ¹⁴ During 2011, many of these dynamics of accountability were repeated when NATO initially denied deaths from its aerial campaign in Libya.Footnote ¹⁵ NGOs monitoring media reports have been able to offer provisional figures about casualties related to weapon types, although recognizing limitations in the sources they have access to.Footnote ¹⁶

Such data can be very valuable for making further assertions regarding the role played by certain weapons in the production of civilian harm and often stand in contrast to states’ own abilities or willingness to provide such data. Despite decades of public concerns regarding cluster munitions, and repeated assurances that such weapons were acceptable given a ‘careful weighing’ of military benefits and civilian risks, the UK was unable in 2005 to point to any data that it had gathered on their humanitarian impact.Footnote ¹⁷

Given such weaknesses in state practice, a group of NGOs have endorsed a Charter for the Recognition of Every Casualty of Armed Violence, and are initiating an ‘Every Casualty Campaign’ calling on states to recognize that they have a responsibility to record, identify, and acknowledge all casualties of violence.Footnote ¹⁸ This initiative builds on recognition that developing controls on deployed weapons is likely to require data regarding harms caused, but that the parties responsible both for using the weapons and establishing such controls rarely produce such data.

However, information gathering raises many questions. ‘Information about what?’ being one. In relation to civilian harms, the question of which deaths should be counted is of critical significance. Should that include only those killings directly resulting from violence or should the numbers also include indirect deaths stemming from a loss of public infrastructure and access to medical facilities, which may be a major element of the overall harm?Footnote ¹⁹ Much of the public controversy about the real number of civilian deaths stemming from the Iraq War stemmed from alternative assumptions about what should be measured and misconceptions about what was being measured.Footnote ²⁰ Decisions about what information should be gathered are likely to be affected by how a problem is depicted, and those decisions may also serve to shape what arguments can subsequently be made. For new weapons technologies, different types of data may be needed at different stages of a weapon's development.

‘Information with what assurance?’ is another question. Ruge recounted how the definition of humanitarian problems related to arms control and disarmament – such as the previously prominent claim there were ‘110 million mines in the ground’ – resulted from limited data being extrapolated into fact.Footnote ²¹ Elsewhere, in arguments about the perceived acceptability of certain weapons it has not been unusual for states to challenge methodologies and data produced by NGOs while offering no data of their own.Footnote ²² Issues of methodology and rigour may also shape practices and debates that follow.

‘Information when?’ is a further question of particular significance for assessment of new weapon technologies. In so far as prohibitions on anti-personnel mines and cluster munitions were driven by information gathered on the humanitarian impact of these weapons, it is important to note that this information only became effective after substantial international use of the weapons and high levels of resulting civilian harm. With respect to emerging technologies, data on harm may not be available and so other types of information may be required. For example, while there is little data on the civilian harms caused by new ‘sensor fuzed’ weapon systems, NGOs such as Landmine Action (now Action on Armed Violence), Austcare (now ActionAid Australia), and Handicap International have called for technical information regarding these weapons so as to better understand the civilian risk.Footnote ²³ At the other end of the process, civil society organizations also take on substantial information-gathering functions in order to monitor the implementation of agreements adopted to control certain weapons.Footnote ²⁴

Finally, it should be recognized that information does not on its own generate meaning. Whether the effects or technical characteristics being documented show a weapon that causes disproportionate harm or kills and wounds in some unacceptable way is a question that cannot be resolved simply by comparison of data. Some previous efforts to strengthen the international regime for controlling new weapons, such as the ICRC's SIrUS project, have arguably fallen foul of too great an emphasis on the decision-making power of data.Footnote ²⁵

Analysing

Building on the final point above, civil society organizations generally go beyond simply providing data, and seek to forward assessments about the scale and nature of the problems being documented, the links between those problems and the technology of specific weapon types, and what needs to be done in response.Footnote ²⁶ However, there are limits to the role of analysis in developing new standards.

Ideally, it might be imagined that choices about the adoption and deployment of weaponry, as with other technologies, might follow a rational set of stages. Operational objectives would be established first; alternative options to meet those objectives would be scrutinized in detail (including with regard to their humanitarian implications); weapons would be deployed; their performance would be systematically monitored and evaluated; and this experience would feed back into a new cycle of examining objectives, options, and performance. However, political theorists examining how choices are made about technology have long questioned whether such rational models are accurate or even desirable as ideals.Footnote ²⁷ A central problem is that they place a great weight on analysis, and do not adequately recognize the extent to which information can be ambiguous and may produce divergent views when approached with different preconceptions or motivations.

Even if data regarding the effects of weapons are relatively undisputed the legal framework governing armed conflict alone provides ample opportunities for analyses to diverge. The meaning of the principles and rules of IHL are uncertain and subject to disagreement in major respects. Phrases such as ‘incidental loss of life or injury to civilians’ and ‘concrete and direct overall military advantage’, for instance, are subject to significantly different interpretations by government officials and legal scholars.Footnote ²⁸ Given these differences, the notion that analysis alone could resolve disputes about legality – let alone wider questions about acceptability – is questionable.

Further, the identification of humanitarian implications that need redress is not done through an exhaustive process of analysing the objective harms of all weapons and then agreeing priority topics for action. Officials, NGOs, and others work with assumptions about what concerns matter. For example, the idea of a weapon that kills or injures by blasting pieces of flesh off the victim is generally not considered problematic, because blasting pieces of flesh off people is seen as common in armed conflict. By contrast, horrific injuries from an ‘unusual’ type of weapon technology (for instance, biological weapons or white phosphorus) may attract far greater attention even if there is a much more extensive pattern of civilian death and injury associated with technologies considered ‘normal’ or the use of which is somehow seen as ‘inevitable’.Footnote ²⁹

The preceding paragraphs are not meant to suggest analysis has no, or only a highly limited, role to play in setting standards. They are intended to indicate that analysis is most likely to be most meaningful when it contributes to ongoing political processes and dialogues. In such circumstances, the ‘framing’ of the issue in question can perhaps be narrowed down sufficiently to limit divergence of opinion regarding the underpinning terms of the debate. A good example is that of the 2007 report by Norwegian People's Aid in collaboration with the Norwegian Defence Research Establishment regarding the in-field reliability of the M-85 submunition.Footnote ³⁰ By directly challenging one of the proposals being debated at the time as part of the Oslo Process on cluster munitions – namely that submunitions with a self-destruct mechanism could sufficiently address humanitarian concerns – the report helped policymakers resolve a choice that was being posed to them.Footnote ³¹

Framing

Claims about the causes of the problems associated with weapons and what needs to be done about them also speak to issues of framing. Gamson and Modigliani referred to frames as the central ideas for structuring our sense of events and the issues at stake.Footnote ³² Frames shape our understanding of what is going on, why, what (if anything) needs to be done, and who needs to do it. This may involve setting the terms of the argument, for example in relation to existing law or other policy commitments; and it may include stipulating which fora are most appropriate. Without the latter, issues might be widely regarded as problems, but not tackled anywhere. Thus, while framing itself requires communication, it is primarily a process of setting the terms for the communication that is to follow.

It is worth noting the different grounds on which civil society and others have raised concerns about particular weapons technologies in the past. There are various subtleties to how concerns have been framed in different contexts. However, some of the grounds for calling for controls on specific weapons could be summarized as follows:

• • The weapons, due to the way in which they function, have a tendency to kill or injure the wrong people (e.g., biological, chemical, nuclear weapons, cluster munitions, anti-personnel mines, incendiary weapons).

• • The weapons have presented a historical pattern of killing and injuring the wrong people (e.g., anti-personnel mines, cluster munitions).

• • The weapons, due to the way in which they function, have a tendency to kill or injure even the intended people in the wrong way (e.g., blinding laser weapons, ‘dum-dum’ bullets, anti-personnel mines, biological, chemical, and incendiary weapons, cluster munitions in the 1970s).

• • The weapons may have wider negative effects on the environment, infrastructure, economic life, etcetera, that last far beyond the period of conflict (e.g., chemical, biological, and nuclear weapons, uranium weapons, landmines and unexploded ordnance).

• • The weapons end up in the hands of the wrong people (e.g., in relation to the transfer of ‘dual use’ technologies and small arms).

In making these arguments, different individuals and organizations may take different orientations to existing law. Some groups tend to urge that, on the basis of one or other of the arguments above, the weapon in question falls foul of existing law by being indiscriminate, causing unnecessary suffering, etcetera. Others might press that, while not straightforwardly disallowed by existing law, one or other of the arguments above provides grounds for new rules to be put in place. Legal, extra-legal, and non-legal arguments run in tandem with assessments about whether new formal rules and restrictions are required and/or whether the delegitimization and stigmatization of weapons within the international community can address identified problems. Such varied orientations are not necessarily mutually incompatible, and within civil society coalitions those with different orientations may still work effectively together.Footnote ³³

A key element of civil society responses to new weapon technologies will be to frame the concerns associated with a particular technology. It should be noted that how a weapon is portrayed as a problem can change over time. As is illustrated in the list above, diplomatic proposals in the 1970s to control cluster munitions were framed around concerns regarding unnecessary suffering and superfluous injury (due to the fragmentation effects of the cluster munitions),Footnote ³⁴ yet this ‘problem’ barely featured in the development of an international ban on cluster munitions in 2008.Footnote ³⁵ It is not yet clear what framings will predominate with respect to weapon technologies now emerging, but some suggested possibilities are:

• • cyber warfare – due to the type of target that may be attacked (e.g., public infrastructure), is likely to harm the wrong people; may cause unforeseeable longer-term harms; and may end up in the hands of the wrong people;

• • autonomous weapons – due to the way in which they function (e.g., by sensor/algorithm decisions to attack), will be prone to killing and injuring the wrong people; may lack adequate human accountability; and may offend against an assumption of human control over lethal decisions;

• • directed energy weapons – due to the way in which they function (e.g., invisibility of microwaves, unknown longer-term effects, incomprehension of the victims), will cause death and injury in the wrong way, including intended targets.

In all such framings, consideration needs to be given not only to the basic moral problem being attributed to the weapons, but also to the causal link between the technology and the harm. Is the problem that an unacceptable outcome will occur in all circumstances, most circumstances, some circumstances, etcetera? Furthermore, it is possible that new technologies will bring to the fore problem framings that have not been used for weapons previously or that relate back to controls over the methods of warfare rather than weapons as types of technologies. For example, drones have raised concerns about a lack of accountability for attacks with such systems.Footnote ³⁶ Both in cyber warfare and the proliferation of drones, the problem might not be framed so much in the permissibility of the weapon technology itself, but in the types of attacks that this technology now facilitates. In any case, the way in which a problem is framed will have a great bearing on the type of solution that follows.

Redefining

‘Redefining’ means providing an overarching mode of analysis that goes beyond the question of how issues with individual weapon types are framed. Past efforts, spearheaded by international civil society, to shift from traditional national security-indebted arms control approaches to ‘human security’ or ‘humanitarian action’ represent instances of redefinition.Footnote ³⁷ By linking discussion about the rights and wrongs in the conduct of conflict to these overarching notions of human security and humanitarian action, a goal of the redefinition was to open up novel possibilities for collaboration and paths for intervention.Footnote ³⁸ Similarly to ‘framing’ discussed above, the work of redefining modes of analysis is about setting the terms of technical, political, or public arguments that will subsequently be worked through.

Redefinitions might be more or less explicit or acknowledged. By shifting from a negotiation process structured around establishing what should be restricted to instead demanding argument for what should be allowed, Rappert and Moyes argued that the Oslo Process leading to the Convention on Cluster Munitions shared important dimensions with ‘precautionary’ approaches to environmental risk.Footnote ³⁹ Borrie has noted that ‘shifting the burden of proof’ was a key (though not often remarked upon) element in the development of the case against cluster munitions.Footnote ⁴⁰ Yet, while some in this process recognized the significance of this shift, many others may not have done.Footnote ⁴¹ As a result, the precautionary precedents set by the Oslo Process may, or may not, inform a wider redefinition in how future negotiation processes are structured.

As an approach that is wider than armed conflict, conceptualizing violence as a health problem is an approach that can both complement and challenge legal and security-related agendas. Shared starting points among health approaches include conceptualizing violence as a substantial and preventable cause of physical and psychological harm.Footnote ⁴² Public health approaches have been advanced in relation to armed conflict in general,Footnote ⁴³ and small arms in particular.Footnote ⁴⁴ Bound up with such redefinitions has been the expansion of what kinds of expertise are required; specifically, an extension of expertise beyond that associated with military operations and legal rules.

Aligned with health definitions, the argument has been advanced that some classes of weapons need to be scrutinized as if they were drugs. This is most evident today in relation to biochemical agents alternatively known as ‘incapacitating’, ‘non-lethal’, and ‘less lethal’ weapons. The use of a fentanyl derivative during the Moscow theatre siege in October 2002 (with tragic results) is the most high-profile example of such a weapon capability, a capability that may yet be adopted more widely by other governments. In the light of such developments, many have questioned the legality of such options as well as the adequacy of the procedures meant to validate their safety.Footnote ⁴⁵

The contention that the uncertain or unpredictable effects of chemical agents require a wider appraisal of their acceptability goes back some time. In light of widespread use of CS smoke (‘tear gas’) grenades in Northern Ireland in the late 1960s, the UK government-appointed Himsworth Committee concluded, among other things, that in the future such chemical agents should be regarded as being more akin to medical drugs than weapons in relation to their operational approval.Footnote ⁴⁶ While the specific meaning of this recommendation is open to question, arguably any such process would need to consist at least of the pre-deployment testing for possible concerns in a manner open to scrutiny and the post-deployment monitoring of operational use. Making public the evidential basis for decisions. as well as the criteria for assessment, would be vital in ensuring the robustness of decisions and the adequacy of attempts to address uncertainties. As with the monitoring of adverse reactions to drugs, rigorous systems for the post-marketing surveillance of weapons’ effects would also be vital in ensuring that outcomes match expectations.

Thus, treating weapons as akin to medical drugs for the purpose of their assessment and control is one way civil society could redefine current approaches to these technologies. The next section illustrates the gulf between such an aspiration and state practice, and thereby the sweeping changes possible through such a redefinition.

Communicating and representing

A fifth role for civil society is in the ongoing communication and representation of this information, analysis, and problem framings to different audiences. The work of communication is seen in NGO publications, placing of media stories, interventions in meetings, mobilization of parliamentarians, and direct lobbying of diplomats and government officials. Such communications may be setting the agenda, framing arguments, pushing for decisions, supporting negotiations, or monitoring instruments already in place. However, underpinning this representational role there is a wider question about how affected populations have their voices heard in discussions regarding the acceptability or appropriateness of certain weapons.

Weapons of armed conflict are often developed and brought into service with assumption that the population amongst whom they may be used will be foreign rather than domestic. As a result, the links of accountability between those introducing the technology and those likely to experience negative effects are very limited. By representing the experiences of that population, civil society organizations can work to reduce this deficit in accountability. The development of global civil society coalitions, where many NGOs in different countries share resources and coordinate their research and advocacy work under a common banner, can help to increase the space for often-marginalized perspectives to be heard.Footnote ⁴⁷ However, capacity in that area is arguably significantly short of what might be required to sustain a systematic scrutiny of weapon effects and implications across the range of relevant theatres and technologies.

A representational role also brings with it challenges. Civil society often presents itself as ‘speaking on behalf of’ populations, but the basis for such a mandate is often unclear. Cluster munition survivors, as activists against cluster munitions, had a strong and active role in the process of banning these weapons, but in any such process there are dangers that ‘victims’ are used as representational figureheads and are without the authority to manage representation directly. Making assertions about what affected communities need and about where any particular issue stands amongst those communities’ priorities is fraught with difficulties for those in civil society. Civil society organizations often face pressure to synthesize diverse experience into a sense of the problem that can fit into the political debate at hand. However, this might downplay or exclude some of the experiences of those in affected communities. For instance, in the case of cluster munitions much of the focus was on civilian populations. This differs significantly from the attention given to these weapons in the 1970s when effects on military personnel were the centre of attention. Despite these concerns about who is represented, such a representational role will likely remain a key one for civil society as state willingness to bring affected populations directly into discussions regarding the acceptability of certain weapons remains very limited.

The challenges of reviews

Before offering an evaluation of the implementation of Article 36 – and thereby illustrating the scope for civil society engagement – it is worth noting some of the general difficulties associated with the type of reviews that might be envisioned based on an analysis of such processes in relation to technologies more broadly. CollingridgeFootnote ⁵⁵ identified a fundamental dilemma in trying to manage technology. Controls are relatively easy to introduce in the early stages of development, yet at such an early stage they often prove difficult to justify because negative effects have not materialized. However, when the need for controls is apparent because of negative effects they are often more expensive and troublesome to put in place. The way technologies become entrenched within organization practice, the investment costs already committed, the formation of beliefs and career structures, etcetera, can all work against the adoption of control measures.

Collingridge's key recommendation is to maintain flexibility in the adoption of a technology. The customary response to the ‘dilemma of control’ is to focus on finding better ways of forecasting technology's effects. This approach has limitations given the fallibilities of analysis. In the case of weaponry, the fallibility of analysis is particularly acute because of the scope for uncertainty and disagreement about costs and benefits (including how such costs and benefits are characterized) and the substantial and irreversible nature of the harms that might be inflicted. In these circumstances, it is vital that processes are established for learning from experience. This underscores the need for openness to scrutiny, involvement of those with expertise and relevant backgrounds, ongoing review of operational use, and the open sharing of experiences.

The challenges of existing practice

However, current implementation of Article 36 seems to fall far short of the type of regime suggested in the paragraph above. While progressive in its intent, Article 36 is reactionary in its terms, which do not prescribe any particular mode of compliance. Instead it is left to states to determine their own processes without international oversight. Consequently, it is difficult to gain a complete picture of whether, or to what extent, states are abiding by the obligation to review. In the context of national defence and security interests there is a dearth of publicly available information on weapons review programmes and their outputs. As Cassese noted shortly after the adoption of Additional Protocol I, Article 36 does not require states to make public their weapons reviews and, consequently, ‘other contracting States have no possibility of verifying whether the obligation laid down [in Article 36] is complied with’.Footnote ⁵⁶ Such secrecy presents particular challenges for civil society organizations that seek to enhance state accountability and promote transparency.

Further, Article 36 is quite evidently not self-executing. Despite the scope afforded to national authorities in determining the mode of compliance with its terms, more than three decades after Protocol I's adoption the number of states known to have formal review processes remains very small. While a limited number of states appear to be actively abiding by the terms of Article 36, it is clear that a much larger number of states are not undertaking weapons reviews.Footnote ⁵⁷ It also appears that some states rely on the review processes of larger military powers when they acquire or develop new weapons, failing to abide by their independent obligation to review.Footnote ⁵⁸

While states are entitled to calibrate their review processes differently, many interpret the Article 36 obligation very narrowly. This tendency can manifest itself in at least four different ways: first, through a primary focus on ensuring technologies in development will not fall into existing categories of explicitly banned weapons, neglecting, or downplaying the applicability of a broader range of general rules that are more difficult to interpret; second, the phrase ‘weapons, means or methods of warfare’ may be interpreted to refer only to physical weapons and their normal or intended use, with the meaning of ‘means or methods of warfare’ poorly understood and not extending to the ways in which certain weapons are used;Footnote ⁵⁹ third, while some states take on board the multidisciplinary approach urged by the ICRC, involving experts from a variety of disciplines in the evaluation process, others leave the determination to military lawyers or ‘experts’ who need not draw on outside inputs despite their obvious relevance to the question of legality;Footnote ⁶⁰ and, finally, most states fail to review existing technologies on an ongoing basis, in light of actual battlefield use and effects. In addition, this analysis is generally done in secret with little or no public information being produced to facilitate learning lessons from the past or from other contexts. As a result, even where they are followed, such processes will tend to produce narrow legal interpretations and thus are unlikely to provide a substantial barrier to the uptake of new weapon systems that present unknown risks. In some cases, states have asserted reservations under Additional Protocol I that seek to exempt whole categories of weapons from falling under those rules. For example, the UK government's reservation to Additional Protocol I states, inter alia, that ‘the rules so introduced do not have any effect on and do not regulate or prohibit the use of nuclear weapons’.Footnote ⁶¹

Engendering a culture of review

Many of the known review processes appear to be ill-suited to assessing certain new technologies of warfare for compliance with international law. First, the acquisition or development of some new technologies will simply not be subject to review at all: either that technology will fall outside the narrow definition generally accorded to ‘weapons, means or methods of warfare’ (as may be the case with many cyber capabilitiesFootnote ⁶²) or its development will occur outside normal military processes and so fail to come to the attention of the bureaucratic apparatus charged with undertaking review. Second, if review does occur, there may be insufficient expertise or capacity within the review body, or those it seeks input from, to adequately understand the operation and effects of the technology. Third, the frame for assessing legality often incorporates the narrow interpretations described above, where only ‘normal’ or expected uses of the technology are considered (with little by way of a boundary established to prevent use outside of such parameters) and no reassessment is made on the basis of actual use and actual effects post-review. Finally, even if a broad approach is taken, existing international law, with its open terms and focus on ‘balances’, may be reasonably interpreted as not addressing the technology or allowing the technology, despite concerns about its humanitarian or environmental impact. These inadequacies should be cause for significant concern: democratic states have a duty to justify the deployment of new technologies that may inappropriately kill, injure, or cause wider harms in both moral and legal terms – even if some form of harm is part of the designed purpose of the technology.

The military application of nanotechnology provides a concrete example of the inadequacies of the current framework. Nasu and Faunce observe that:

As they note, ‘[t]echnological advancement all too often entails adverse effects on the environment or human health that may not immediately be so obvious after its full import into battlefields is experienced’.Footnote ⁶⁴ Further, where significant investment and time have been devoted to a new technology, the pressure on military lawyers to defend its legality may be very great, even if unstated; for example, in the absence of clear evidence of adverse long-term effects, the confidently claimed military advantages of the technology may allow reviewers to strike the balance between military considerations and possible harm in favour of legality.Footnote ⁶⁵ In such a case the standards of proof required of the different elements being balanced may be quite different.

While the creation of an international body charged with scrutinizing new technologies is politically implausible right now, some mechanism for strengthening international capacity and coordination is required.Footnote ⁶⁶ The role of civil society aside, formalized coordination between states in shaping standards around new technologies could be a significant part of the solution. The responsibility for carefully considering each new technology and its relationship to legality and broader considerations of humanity should be a shared one, including during the conceptualization, design, and manufacture of weapons. However, it is not clear from current practice that states are undertaking their responsibilities in a way that adequately assesses the humanitarian and moral problems that weapons technologies can pose. Without greater transparency and sharing of information it is hard to see that the national-level processes currently in place can provide the basis for more progressive efforts to set standards regarding weapons. There are few if any examples of weapons that have been found to be problematic through a national review mechanism and where the state has then gone on to promote a new international standard with respect to the particular technology. Similarly, there are few if any examples of states revisiting their reviews of weapon legality in the wake of evidence that existing weapons are causing unacceptable humanitarian or environmental harm.