Fighting silence covert warfare and the uphill battle against the unsaid

Silences around drone warfare and similar covert state practices have often been encountered as a hurdle that hinders us from understanding and interrogating government acts. Scholars as well as human rights actors have opposed silences in a struggle for greater transparency and have called on governments to speak. Through the case study of drone warfare, this paper analyses the productive role of silences and the political struggle to oppose it. Analysing 125 non-governmental organisation (NGO) reports, UN documents and policy papers, this article investigates how silences are encountered, interpreted and opposed by Western human rights actors. This shows that silence is not encountered as a discrete unit but as interdependent layers of denial, partial withholding of information, redactions, delays, lack of oversight and so on. Situated within unequal power relations, I show how the battle against the unsaid is itself based on what has (not) been heard in Western constructions of drone warfare and risks further enabling violent practices. Discussing ways of subverting the workings of silence, the paper not only contributes to academic literature on covert warfare and silence but also speaks to the practical dilemmas faced by non-state actors who are advocating for more transparency.


Introduction
The War on Terror is permeated with silences. We often encounter practices, such as armed drone attacks, through classified reports, redactions and denials. Research on these practices has, in turn, often worked around silences. The more tangible aspects of attacks has been focused on, such as the new drone technology (Kindervater, 2016;Manjikian, 2014;Wilcox, 2017) and information has been recovered by drawing on interviews, leaked documents or freedom of information requests. Silence is encountered as a hurdle that creates difficulty in understanding drone warfare because governance 'categorically refuses to reveal the evidence' (Heller, 2013: 104;Sterio, 2012). Nongovernmental organisations (NGOs) and other human rights activists have critiqued the silences around drone warfare and called for more transparency in an 'uphill battle' (WILPF, 2017) against the unsaid. This paper takes a different approach. Rather than working around or against silences, I am interested in the political role silences themselves have come to play in the 'uphill battle' against the unsaid. Investigating 125 NGO reports, UN reports and other documents by human rights actors, the paper contributes with an empirically grounded study to recent interest in silence in international relations (IR) scholarship (Dingli, 2015;Guillaume, 2018). Analysing human rights discourses shows that silences are not monolithic but noted in complex layers of denial, partial withholding of information, redactions, delays, lack of response and so on. I argue that these silences are productive in how drone warfare becomes conceptualised as 'a secretive program of surveillance and killings' (Amnesty International, 2013) and in how the practice is politically enabled. These effects, however, are not independent of the human rights discourses within which silences are noted, interpreted and opposed.
Given the political effects of silences, it is not surprising that human rights actors have tried to break the silence and have called on governments to speak (Amnesty International, 2013;Open Society Foundations, 2019). Yet, there is a risk that this ends up further entrenching the dynamics of silences it opposes. I argue that the battle against the unsaid risks focusing on silence in a way that shifts attention away from opposing the violent practices themselves. Calling for less secrecy can end up reaffirming the power position of state actors who are asked to speak and creates a subject position of exclusion from which it becomes difficult to know or to speak. Interpretations of silence tend to recreate the meaning of silences along the lines of dominant interests, for example, as secrecy and legal uncertainty. The unequal power relations within which the battle against the unsaid is situated thus lead to a range of dilemmas for Western human rights actors.
It is important to interrogate this in the context of wider power relations that not only determine who has not been able to speak -but what has been listened to (Chimni, 2018;De Behar, 1995;Enloe, 2004;Hansen, 2000;Hobson, 2007). Silence is inherently ambiguous and dependent on the relations between speakers and listeners, which means that a particular meaning of silence reflects the 'horizon of expectation' of the interpreter (De Behar, 1995: 7; see also Guillaume, 2018). By tracing the role of silences in Western human rights discourses, I follow Spivak's suggestion to investigate our listening skills. Building on the work of subaltern studies, feminism and postcolonialism, the paper reflects on the complicity of the 'well-meaning liberals' (Maggio, 2007: 431) in further marginalising the subaltern. Engaging with the dilemmas which human rights actors face in the battle against the unsaid, the paper then points to spaces of agency and ways to creatively subvert the power relations at play. This not only contributes to scholarly work on silence but also addresses the practical dilemmas faced by non-state actors who are struggling against the unsaid.
While the paper focuses on drone warfare, similar workings of silence are at play in other political practices when actors struggle against silences. For this paper, I empirically focus on documents on drone warfare, though some of the investigated reports have discussed silences regarding more traditional aerial warfare or interventions by special forces in ground wars. A fetishisation of the drone is itself problematic (see for further discussion Schweiger, 2019) and drone warfare is here not used as a stable category but as a lens through which to examine the political workings of silence. Analysing the uphill battle against the unsaid and the dilemmas it poses, the paper is relevant for other struggles against silences by powerful actors.
After a brief overview of existing literature on silence, the second part of the paper analyses reports and policy briefs by human rights actors to map the silences they have encountered in layers of denials, redactions, delays, partial withholding of information and so on. The subsequent sections focus on the uphill battle against the unsaid, engaging with the dilemmas it presents for human rights actors. Section 'An uphill battle against the unsaid' analyses the risks of framing the battle against the unsaid as the core struggle, thus shifting attention away from the violence itself. Section 'Speaking for the subaltern' focuses on the attempt to represent marginalised communities affected by drone warfare, often facing the dilemma of further essentialising and silencing them. Section 'Speech expectations' discusses the underlying speech expectations of the battle against the unsaid and the subject position of exclusion this creates. The final section suggests ways in which the workings of silence might be subverted.

Studying silence
There is a wealth of literature on the effects of silence in Sociology and Sociolinguistics, investigating how silence can become communicative. While a lot of this literature has focused on abstract, ontological or epistemological aspects of silence (Doucet, 2008;Ephratt, 2008;Kurzon, 2007;Scott, 2018), a number of authors have showed the practical functions of silences in everyday settings from concealment, to delay, consent, disapproval or disregard (De Behar, 1995;Jaworski, 1992;MacLure et al., 2010;Zerubavel, 2006). These everyday workings of silence do not take place outside of political hierarchies and, as feminist and anti-racist work on silence has showed, are fundamentally linked to raced, gendered, classed and other relations of power (De Behar, 1995;Mazzei, 2003;Ryan-Flood and Gill, 2010). The wider, political workings of silence have received quite a bit of scholarly attention in recent years, highlighting the dynamics of who can speak and what is spoken about (Dingli, 2015;Dingli and Cooke, 2019;Ferguson, 2003;Schröter and Taylor, 2017;Thiesmeyer, 2003). Scholars have empirically traced the effects of silencing, for example, the silences on homelessness in US newspaper coverage (Huckin, 2002), or the silencing of migration on Twitter (Alcántara-Plá and Ruiz-Sánchez, 2018).
In International Relations scholarship, these underlying relations of power which lead to the silencing and exclusion of certain topics and voices have also been an important subject (Acharya and Buzan, 2007;Bhambra and Shilliam, 2009;Dingli, 2015;Enloe, 2004;Schweiger, 2018;Suleri, 1992). As Lene Hansen (2000) has argued, not everyone is in a position where they can speak. Postcolonial and feminist studies have examined the silence of the subaltern who in Spivak's famous words 'cannot speak' (Spivak, 1988: 294). Silence has here been studied through webs of exclusion from hegemonic discourses in which both the 'West' and 'its silent Other' are constructed (Said, 1985: 93; see also Hobson, 2007). Silence in these strands of literature is thus often equated with exclusion and marginalisation (see Guillaume and Schweiger, 2018). Revealing the structures of power through which voice becomes silenced, this calls for an emancipatory programme of opening up space for silenced perspectives, to 'give[] voice to the marginalised, silenced "other"' (Agathangelou and Ling, 1997: 9;Said, 1985: 91) and 'demand[. . .] that the unseen and the unacknowledged be made visible and heard' (Dingli, 2015;Ryan-Flood and Gill, 2010: 1). Less attention has been paid to the strategic use of silence.
The use of silence as a strategy has received more attention in Security Studies. Silence is here conceptualised as secrecy, or the deliberate withholding of information by state authorities and other elite actors (Bratich, 2006;Masco, 2010;Walters and Luscombe, 2017). A vibrant field of scholarship has developed on the question of secrecy, ranging from the role of secrecy in media representations of armed drone attacks (Kearns, 2016), to the secret detention and torture programme (Raphael et al., 2016), to secrecy in counterterrorism and finance tracking (Goede and de Wesseling, 2017). Scholars have analysed the performative role of secrecy, particularly regarding the open secret, which is linked to 'the public demonstration that one will not reveal' (Blakely, 2012: 49;Masco, 2010).
There is thus a tendency in IR scholarship to approach silences through a particular mode, for example, as secrecy or as exclusion. Yet, silence is inherently ambiguous. Constructing the meaning of silence will always reflect the 'horizon of expectation' (De Behar, 1995: 7) of the listeners -regarding who has been expected to (be able to) speak, what has been considered relevant and how silences are organised and interpreted (Guillaume and Schweiger, 2018). The existence and meaning that is fixed for silence relies on assumptions which the listeners make and, as I will argue, often ends up further enabling relations of marginalisation and violence. Analysing Western human rights discourses and their struggle against silence shows the fractured and ambiguous nature of silence, where acts of concealment are inherently linked to dynamics of exclusion and forms of not listening.
In the paper, I do not conceptualise silence as a discrete unit in a binary relation to utterances. Instead, the paper builds on human rights discourses to show how silences are noted through a spectrum that includes non-disclosures, part-disclosures, ambiguities, delays, explicit refusals to answer and so on. By critically engaging with these layers of silences, the paper contributes to existing literature, not only on covert warfare but also on silence more generally, showing the contingency of the discursive function of silences rather than already interpreting them through a particular mode, such as secrecy. This embeds the analysis within wider structural relations of power in international politics. It also shows the spaces of agency which listeners are holding and points towards an engagement with silence as a 'radical irruption' (Guillaume, 2018: 13).
The paper traces how silences have been encountered and interpreted in human rights discourses around drone warfare and how silences have gained meaning. Human rights actors have used intricate techniques to trace silences by creating points of reference, such as information gained through lawsuits (European Center for Constitutional and Human Rights (ECCHR), 2017), independent investigation of strikes through satellite images and interviews (Amnesty International, 2013) or relative absence of detail in some publications versus others (Drone Wars, 2018). This is important to note from a methodological perspective, since the practical aspects of researching silence are often neglected in literature on silence (see for a notable exception Schröter and Taylor, 2017). Yet, the point of my analysis is not to prove that silences noted in human rights discourses do empirically exist or that they should be interpreted in a particular way. Instead, I conduct an analysis of the meta-discourses about silences (Schröter, 2013: 48) to analyse what kind of silences are noted and opposed in Western human rights discourses. This is important because even if those silences did not empirically exist, they have come to gain political functions.
I use the term human rights discourses quite loosely in the paper as a heuristic device to refer to a range of documents published in English by actors who have raised human rights concerns about the legal, ethical and political repercussions of armed drone attacks, including NGOs, think tanks, international lawyers and UN representatives. The NGO and think tank reports were accessed through the European Forum on Armed Drones (EFAD), a civil society network of organisations. I analysed all 50 reports published on their website between 2011 and March 2020. In addition, the paper builds on UN reports which were accessed through the UN Human Rights Council search engine through the key word 'drone'. This returned 142 results which were further narrowed down by excluding irrelevant documents (such as reports referring to photos taken by civilian drones), duplicates of the above analysis or translations. The remaining 75 documents were searched for key words around drone warfare (UAV, targeted, drone, unmanned). All 125 documents were coded according to themes of silence (secrecy, covert, denial, ambiguity, absence, lack, omission, exclusion, transparency, accountability, acquiescence).
This methodology is not intended to develop an exhaustive understanding of silences around drone warfare -indeed this would be impossible, since both EFAD and the UN are platforms which favour Eurocentric perspectives and powerful state and non-state actors. As will be discussed further below, which silences are heard in the first place and how they are interpreted, is itself dependent on Eurocentric assumptions that risk further marginalising the subaltern. Instead, the point of this analysis is to understand the layers of silences which are noted within Western human rights discourses on drone warfare, how they are interpreted and opposed and to critically engage with the battle against the unsaid within which they are made sense of. This is not only of scholarly relevance but also has policy implications. It addresses the practical dilemmas faced by those who are opposing state silences in an uphill battle against the unsaid.

Mapping layers of silences
When we speak of silence through a particular mode, for example, as secrecy, there is often an idea of silence as a discrete unit, an absence of speech, that is encountered by the researcher. Analysing reports by NGOs, UN agencies and other human rights actors on drone warfare show how silences are encountered through nuanced layers, ranging from complete denial, to partial withholding of information, to lack of oversight, absence of media reporting, to redactions in court proceedings, unclear policy formulations and so on. These layers are not in absolute opposition to utterances but produced with and through text. The following section will trace how human rights actors have encountered and interpreted silences around drone warfare. These silences constitute a complex matrix which is understood to have political effects.
The top-most layer of silence noted in the battle against the unsaid is a general sense of unclear frameworks for the use of force. These unclear frameworks are understood to raise as many questions as they answer: When does an individual pose a 'continuing and imminent threat' to the United States? How does the government decide when capture of a target is 'feasible'? (. . .) Nothing that the government has released to date provides satisfying answers to those questions. (ECCHR, 2017;see also Amnesty International, 2013) The lack of clear policies is seen to be further compounded by non-disclosure of information on attacks, such as when 'the Central Intelligence Agency (CIA) denied the request [by the American Civil Liberties Union], stating that it could not confirm or deny even an "intelligence interest" in the drone program without compromising national security' (ECCHR, 2017). These silences are seen as intertwined in processes in which records are so secret that even the reasons for not disclosing these records are not disclosed, a problem often encountered by those engaged in transparency litigation: Plaintiffs must decide whether to appeal without knowing why the court rejected their arguments or the ultimate basis for the court's reasoning. Even when the court rules against the government and holds that a particular fact or document may not remain secret, plaintiffs may still not get the information they seek. If the US government appeals, its briefs may be riddled with redacted section headings, redacted topic sentences, redacted lists, and entirely redacted pages. (WILPF, 2017) Individual attacks are often not publicly acknowledged with government officials refusing to comment on leaked reports. When individual attacks are acknowledged, they are often understood to withhold crucial information, such as the location and timing of attacks: Three months later, Defence Command scaled back the level of geographic detail by omitting city names. And from March 2015 a caveat was added noting that strikes took place 'primarily' in for example Anbar province, thus leaving open the possibility of strikes elsewhere. (Airwars, 2016; UN Doc A/HRC/31/57/Add. 1, 2016: 642) Generalised, or 'bulk information' (WILPF, 2017) on strikes is understood as problematic because it creates difficulties for human rights actors to interrogate not only the factual basis but also the ethical and legal implications of attacks.
The complex layers of silence also include the absence of clear description in reporting about attacks, through the use of ambiguous terms, such as 'air strike'. This is understood as an absence because what is actually covered by the phrase can vary hugely. At one end of the spectrum an 'air strike' can mean a single aircraft firing a single missile or bomb at a single target, while at the other it can mean multiple aircraft (sometimes even from different nations) launching multiple bombs and missiles at multiple targets within the same geographical area or time frame. (Drone Wars, 2018) Noted in the reports are also silences that are not direct absences but delays in providing information. It has, for example, been stated that 'the first admitted fatalities in Syria were eight months after the event' (Airwars, 2016).
Silences around covert warfare are thus noted and interpreted through nuanced, interdependent layers, which includes lack of clear policies, outright denial of information requests, redactions, non-confirmation of attacks, as well as absence of more particular information on attacks, such as location and timing, the use of ambiguous terms which do not reveal details about attacks or delay in providing information. They tend to be understood as problematic because of the way in which they are seen to undermine accountability for attacks (UN Doc A/HRC/28/38, 2014: 25). This lack of accountability is grasped through its function of further undermining civilian protection and the clarity of frameworks, leading to a 'vaguely defined licence to kill' (Article 36 and Reaching Critical Will, 2014; UN Doc A/HRC/24/34, 2013: 17).
NGOs have noted difficulties in even establishing which state actors are involved in which attacks 'with so many world powers having carried out an estimated 25,000 airstrikes in Iraq and Syria to summer 2016 -alongside actions by aircraft of the Iraq government and the Assad regime' (Airwars, 2016). This becomes even more pronounced when states are providing military assistance to other states, rather than directly engaging which 'has allowed the government to be involved (and in some cases deeply involved) in overseas conflicts with minimal scrutiny -and in some cases completely denying its involvement'. (Oxford Research Group, 2017).
Most governments in the Global North, as well as certain military elites in the states targeted by armed drones, are tied to US foreign policy practices and their secret services are often involved through a tight network of surveillance services, a transnational 'targeting regime' (Zappala, 2015, 254): Amnesty International showed how the UK, Germany, the Netherlands, and Italy all provide crucial assistance to US drone strikes, including in Somalia. This assistance takes the form of general intelligence-sharing as part of historical alliances such as the Five Eyes alliance, a global surveillance network which includes the UK, as well as specific intelligence support which has been provided by various States, including the Netherlands, and used to locate and identify targets for US drone strikes (Amnesty International, 2019).
The opaque workings of the transnational 'targeting regime', which make it difficult to even know whom to ask for more information, have important repercussions for accountability. In an almost Kafkaesque logic, for example, an appeal by Reprieve to the UK High Court to examine Britain's role in US armed drone attacks was rejected. The High Court did not engage with the substance of the issue at all and refused to grant permission for the case on the basis of the Act of State doctrine: 'The court argued that any attempt to examine Britain's role would necessarily require it to sit in judgment of the legality of US actions and in doing so, potentially "imperil relations between the states"' (ECCHR, 2017).
They are also seen as symbolic of the wider relations of power between the West and the marginalisation of victims of drone strikes, for example, through the 'virtual air brushing of civilians from all strike reports together with the blanket assertion that there is "no credible evidence of civilian death"' (Drone Wars, 2018; UN Doc A/HRC/43/61, 2020: 29). This is grasped as fundamentally unethical: 'what really belittles the concept of human rights is not just the lack of investigative journalism conducted into the thousands of civilian casualties but the empty claims of minimal civilian casualties coming from the White House and the CIA' (WILPF, 2017).
For people living under covert wars, silences are seen to further increase the traumas experienced because 'for victims in particular, there is no one to recognize, apologize for, or explain their sorrow; for communities living under the constant watch of surveillance drones, there is no one to hold accountable for their fear' (Center for Civilians in Conflict and Columbia Law School, 2012; UN Doc A/HRC/28/38, 2014: 24). Silences are seen to be representative of a violent relationship that neglects to even recognise the violence that it inflicts: Ali Ahmed Abdulla Saleh (51 years old), for example, lives in Yemen and has repeatedly tried to present his complaints before the responsible authorities, but to no avail. 'We have tried more than once to make our voices heard; that there are no terrorists here and we asked people to come and check, but no one is willing to listen to us', says Ali. Up to this day no investigation into the death of his son and brother occurred. There is only a deafening silence coming from the US and drone strikes that continue to bomb the area. (PAX, 2018a) These layers of silences are seen to reach deep into the recesses of public political life: 'There are established procedures for Congress to declassify and publicize previously secret information, but these procedures have reportedly never been employed, and certainly not with regard to covert drone strikes' (Center for Civilians in Conflict and Columbia Law School, 2012). Absence of debate in parliaments and other institutions is linked to a lack of information and a lack of public debate on attacks, for example, when it shows all the seriousness of the lack of transparency on the part of Italian institutions, the lack of clarity in the doctrine and the absence of a public debate able to promote a full understanding of the political, legal, social and military implications of armed drones. (Instituto di Ricerche Internazionali (IRIAD), 2017) These complex layers of silence have become productive in how drone warfare has been conceptualised as 'a secretive program of surveillance and killings' (Amnesty International, 2013). Reports often carry titles such as 'The Civilian Impact of Drones: Unexamined Costs, Unanswered Questions' (Center for Civilians in Conflict and Columbia Law School, 2012) or 'All Quiet on the ISIS Front: British secret warfare in an information age' (Oxford Research Group, 2017). While War on Terror measures more generally are often understood through this secrecy, there is a sense that armed drone attacks are an epitomised expression of this since 'drone warfare embodies the fundamental problems that covert warfare entails' (WILPF, 2017). Silences are thus not just encountered as absences and obstacles but are fundamentally productive in how drone warfare is understood.
Given the political salience of these layers of silence, it is not surprising that there has been a concerted effort to call on governments to break the silence, 'to establish clear policies, prevent complicity, provide transparency, and establish accountability regarding the use of armed drones' (Open Society Foundations, 2019) and to 'officially protest and pursue remedies under international law when lethal force is unlawfully used by the USA or other states' (Amnesty International, 2013). Yet, the fight against silence produces itself some problematic effects. The next sections will engage with the 'uphill battle' against the unsaid (WILPF, 2017) and show the dilemma it presents.

An uphill battle against the unsaid
Human rights actors have used a range of methods to engage with the layers of silence discussed in the previous section. In these efforts, there is often a sense of 'an uphill battle and hurdles imposed by excessive executive branch secrecy' (WILPF, 2017). The uphill battle against the unsaid, however, poses its own problems. The fight for increased transparency risks focusing resistance on silence, rather than the violent acts themselves. Furthermore, the assumption that silences are the problem is not as obvious as it first appears. Indeed, more transparency can create more -not less -legal uncertainty and legitimacy for violent state acts.
Whole reports, or sections of reports by human rights actors are dedicated to the need for more transparency (Airwars, 2016;APPG, 2018;PAX, 2018a; UN Doc A/HRC/20/22/ Add.3, 2012: 3; UNIDIR, 2017) and even among state actors it has been noted that 'there was consensus across the political divide in the States concerned that Governments could and should be more transparent' (UN Doc A/HRC/28/38, 2014: 26). Typologies of transparency have been developed by human rights actors. Airwars, for example, has evaluated states through categories of good practice, identifying a cluster of best practice including 'Canada, the UK, the United States and France [who] have consistently been the more transparent and accountable partners. Each has generally published significant information relating to the dates, locations and targets of their airstrikes' (Airwars, 2016). The second cluster (including the UAE, Turkey and Denmark) is considered to be less transparent, while 'the final and most troubling cluster includes Australia, Bahrain, Saudi Arabia and Belgium, along with Jordan and the Netherlands' (Airwars, 2016).
Silence is thus encountered as a problem which has 'posed a significant research challenge' (Amnesty International, 2013), has made it 'impossible to determine legality' (APPG, 2018), has led to a situation where actors are 'unable to verify (. . .) allegations or to understand whether civilians may have been killed' (UN Doc A/HRC/33/38, 2016: 41) and has been perceived as a problem affecting international legal frameworks more generally: 'If widespread proliferation of armed UAVs occurs without common understandings about accountability and oversight based on some measure of transparency with a view to mutual confidence building, then the results may be chaotic and undermine the rule of law' (UNIDIR, 2017).
Yet, in approaching silence as the main problem, there is a risk that scrutiny becomes focused on silence -rather than the violent practices themselves. At times, it almost seems like silence becomes understood as the violent act in itself: 'Concealment of covert activities generates confusion, results in the deaths of countless targeted people and bystanders and undermines international law and cooperation' (WILPF, 2017, emphasis added). By focusing on opposing silence, the battle against the unsaid becomes the main struggle. This has led to a situation in which some of the countries with the most problematic track record in terms of aerial warfare have become celebrated as role models. The British All Parties Parliamentary Group thus asks for the British government to 'publish its policy on the targeted killings of individuals in line with the precedent set by the US and Israel' (APPG, 2018) and the United States and United Kingdom have been lauded by NGOs as 'transparent and accountable partners' (Airwars, 2016).
There is an underlying assumption that transparency is more ethical since silence will benefit powerful actors by opening up more ambiguity in the ethical and legal restrictions of violence. Silence by states using armed drones, for example, has been interpreted through its function of undermining international legal frameworks, arguing that there is 'a troubling lack of clarity as to how international norms apply, in part because existing users have not been uniformly transparent concerning their use of armed UAVs' (UNIDIR, 2017: 1, emphasis added; see similarly UN Doc A/HRC/26/36, 137). Yet, this assumption is not as straight-forward as it first appears.
More transparency can create not less but greater legitimacy for violent acts. This is in fact used as leverage by some NGOs in their uneasy position to persuade Western governments to provide more information: 'Greater transparency can bring significant strategic and tactical benefits -and help to distinguish the US and its allies from other belligerents such as Russia' (Airwars, 2016). It has been recognised that publicly reporting attacks can have the effect of normalising and legitimising violent practices: Strikes in Yemen were said to be easier to track since CENTCOM (United States Central Command) started declaring its drone strikes there (. . .) on the other hand it was pointed out that the press releases describing the strikes risk legitimizing what is not a conventional conflict and exacerbating the blurring of the lines between the traditional battlefield and unconventional airstrikes. (ECCHR, 2017) Indeed, transparency and adherence to International Humanitarian Law have been mobilised by state actors to justify military acts in ways that often builds on colonial discourses of civilised war (Gordon and Perugini, 2020: 70). More transparency on strike procedures and official justification of attacks can end up creating more legitimacy of such violent practices. It can also create more legal uncertainty. The public justification of 'targeted killing' by the Israeli and US government has, for example, produced the concept as a category of international law in the first place and caused uncertainty around the interpretation of the right to self-defence (Grayson, 2016;Krasmann, 2012). The notion that more reporting and information would help to prevent violence is thus not self-evident -sometimes the opposite might be the case.
Concerned about the political effects of silence around drone warfare, human rights actors have struggled against silence, calling on governments to reveal more information. This, however, might have some unintended effects. In this section, I showed how the 'battle against the unsaid' risks to shift attention from opposing violence to opposing silence. I argued that silence is not necessarily the main problem and more information, rather than silence, might actually end up further normalising and legitimising violent state acts. The next section will look more closely at attempts to break the silence by speaking for those subjected to bombing and how there is a danger in essentialising and further excluding the marginalised while staying within dominant epistemes and orientalist assumptions.

Speaking for the subaltern
The condition of the subaltern is one of not being able to speak, as not being listened to, as not 'mattering' (Jong and de Mascat, 2016: 721;Spivak, 1999). Silences are often seen as symbolic of the marginalisation of communities in the Global South. The battle against the unsaid thus involves a making-audible of voices that would not otherwise be heard within Western, liberal discourse while also highlighting the neglect that is represented by the absence of such listening: As Liban, a resident of Darussalam village told Amnesty International about the death of three young farmers in a US air strike in November 2017: 'We have been waiting for someone to come ask us about this. We lost these boys, and no one is talking about it. It seared into our flesh. We came [here] to tell the truth, because of the boys we lost, we have to tell the world' (Amnesty International, 2019) Human rights actors are trying to make audible the suffering of those subjected to bombing and to highlight the neglect that is shown towards this suffering. Yet, this is a difficult task. Attempts to break the silence can themselves end up participating in relations of power which the 'battle against the unsaid' tries to oppose. This is especially the case when human rights actors remain within a dominant discourse that follows the logic of warfare and is based on orientalist assumptions.
In attempts to make marginalised voices audible, there is often a focus on the figure of the civilian. Human rights actors have conducted intricate research to establish evidence for civilian death (see for example Amnesty International, 2019; WILPF, 2017). This evidence has been compared to the absence of reporting by states, exposing the hypocrisy, for example, when US state representatives 'assert that the USA killed no civilians in Somalia in 2018, despite the fact that there have been no credible or onsite investigations into allegations of civilians killed in air strikes' (Amnesty International, 2019). Human rights actors have called for more transparency and critiqued the lack of procedures to establish whether those killed were civilians or combatants and called on governments to publicly disclose the 'methodology employed to assess strike outcomes and to determine an individual's civilian or "combatant" status' (Amnesty International, 2019).
These critiques put pressure on state actors to show more concern towards the identity of those that are killed. They draw on strong legal and ethical norms around the laws of war and thus compel state actors to respond. Yet, focusing on procedures of distinguishing between civilians and combatants buys into the language game of armed conflict. The debate then tends to circle around the adequacy of technical procedures to establish the combatant status of terrorist suspects and the precision of weapon technology (Afxentiou, 2018;Gregory, 2015). Moving within the dominant logic of armed conflict, this way of thinking shifts the conversation away from a question of whether military violence should be used to procedural aspects of how violent attacks should be conducted and how information on them should be published. Attempts to represent the marginalised can end up entrenching the interests of the powerful, thus posing fundamental challenges to the 'well-meaning liberals' (Maggio, 2007: 431) who might end up further marginalising the subaltern.
To be sure, the battle against the unsaid often goes beyond an assessment and representation of civilian death. There have been attempts by human rights actors to speak and create space for the subaltern to speak, for example, when PAX and the British NGO Reprieve invited two Pakistani victims of drone strikes, Kareem Khan and Noor Behram, to the Netherlands where they spoke to Dutch politicians (. . .) and journalists about the lack of accountability and transparency of U.S. drones strikes. (Open Society Foundations, 2019) Yet, these attempts to speak for the subaltern or create space for the subaltern to speak risks reproducing an essentialised notion of the subaltern, whose voices are only included to the extent that they confirm their status as victims.
Postcolonial studies have pointed out that attempts to speak for the subaltern are riddled with problems. The very construction of the subjectivity of the subaltern as the subaltern for whom to speak can become complicit in a discourse which essentialises the subaltern and thus further disempowers -hence Spivak's conclusion that the subaltern cannot speak (Spivak, 1988, 294). As soon as attempts are made to break the silence, speak for the subaltern or create space for the subaltern to speak, the underlying relations of power are recreated (Jong and de Mascat, 2016;Maggio, 2007;Spivak, 1988).
This danger is even further accentuated when the representation of the silent subaltern is linked to orientalist constructions of the areas subjected to bombing, following notions of wilderness, such as the 'remote and lawless nature of the region' (Amnesty International, 2013). Said has called such constructions of localities 'imagined geographies' (Said, 1994) which have been used throughout colonial times to justify violence in 'civilising' missions to bring law and order. Orientalist constructions of areas subjected to bombing as 'lawless', 'remote' (remote from what?) and 'dangerous' have also been mobilised to justify current drone attacks (Satia, 2014). Yet, they are invoked in some human rights discourses as well: 'A related issue is that independent investigation of lethal armed UAV strikes can be difficult due to the danger and remoteness of the places where targets are located' (UNIDIR, 2017).
The battle against the unsaid often remains within dominant epistemes -such as the logic of armed conflict -and risks relying on orientalist constructions of areas and peoples subjected to bombing. Speaking for the subaltern within this logic can end up further disempowering those who are already marginalised. This is further compounded by the expectations of speech often underlying attempts to make voices audible where the silence of the subaltern ends up being linked back to the silence of the core: 'Indeed, this is an international failure to redress the victims of drone strikes. In the words of Reprieve, an organisation supporting drone victims, they received 'No answers. No compensation. No apology. Just silence'' (PAX, 2018a). Marginalisation of affected communities is thus often conceived through the silence not of the communities themselves but of powerful state actors: 'Victims of drone strikes have consistently appealed for justice, transparency, accountability and reparations, yet their calls were largely met with silence' (PAX, 2018b). The next section will look more closely at the speech expectations underlying this understanding of silence.

Speech expectations
The attention which the silence of powerful governments has received reflects an asymmetrical structure of knowledge production in which expectations of speech are focused on certain actors -while others are not expected to say anything, not considered relevant as speaking subjects. During the 2014 UN Human Rights Council, 'interactive panel discussion of experts on the use of remotely piloted aircraft or armed drones', for example, speakers included legal experts such as the UN Deputy High Commissioner for Human Rights, representatives of different states, as well as NGOs, such as Amnesty International or the American Civil Liberties Union. Absent were representatives of communities most affected by drone warfare.
Expectations of speech tend to be directed at powerful Western state actors whose silence becomes the centre of attention: 'This secrecy surrounding who the US government has killed and why has prevented affected communities from obtaining official acknowledgment of deaths and injuries or other forms of public accountability' (WILPF, 2017). Secrecy has been a crucial mode through which silences around drone warfare have been interpreted and through which drone attacks have gained meaning as 'secret, unaccountable, borderless wars' (ECCHR, 2017) that are 'opaque, ambiguous and effectively unaccountable' (APPG, 2018). For good reasons, the battle against the unsaid is consequently directed against a small and powerful core of Western -and particular US -secret agencies. Accountability for attacks has been understood to be difficult to establish, not just when there are no answers but when it is unclear who should be asked in the first place.
The opacity of who is responsible is often linked to the new technology of armed drones in attacks that are monitored and controlled from thousands of miles away by a small team of operators, through a network of satellite, ground and radio communications networks, supported by remote intelligence analysts and databases, operating within command structures that may switch between major military commands, joint force structures, covert and traditional military operations, or even between command hierarchies of coalition and treaty partner nations. (WILPF, 2017) The battle against the unsaid is to a large extent configured through this secrecy and the necessity to oppose it. One of the central demands by the UN HRC resolution on armed drones is thus for states 'to ensure transparency in their records on the use of remotely piloted aircraft or armed drones' (UN Doc A/HRC/ RES/25/22, 2014, 2;see also Amnesty International, 2019;APPG, 2018;Article 36 and Reaching Critical Will, 2014;Open Society Foundations, 2019;WILPF, 2017).
Yet, by focusing on the call for governments to speak, there is a risk of reproducing underlying power relations. The attempt to get more information from state actors in the battle against the unsaid implicitly undermines the value of the vast amount of information that has been collected and co-created by local communities, NGOs and human rights actors. Immense research efforts have been launched which include detailed investigation of published strike data, including time and location of attacks, scrutiny of satellite images of attacked areas before and after strikes and interviews with eye witnesses of attacks 'interviewed separately and by different research teams at different times' (Amnesty International, 2013). In calling for governments to reveal information, this wealth of information is implicitly devalued.
The fight against silence risks reproducing particular speech expectations where state actors are presupposed as the agents who speak and act, while the rest of the world is configured as the passive recipient of information who 'must primarily rely on the US government's selective disclosures' (WILPF, 2017). The interpretation of silence as secrecy in the battle against the unsaid thus creates a subject position of exclusion of those who are placed outside of the corridors of power, unable to speak. This can be counterproductive (see Kearns, 2016). It undermines the vast amount of information which local actors and NGOs already possess and reaffirms the position of the powerful who are expected to speak and whom we must wait for to speak. It is an interpretation of silence which creates an impasse for those fighting against state violence.

Alternative ways of engaging with silences
The previous sections have discussed the dilemmas which human rights actors face. I argued that the fight against silence risks shifting the focus of resistance on silence, rather than the violent acts themselves. Attempts to break the silence of state actors often remain within dominant epistemes -such as the logic of armed conflict -and risks relying on orientalist assumptions of areas and peoples subjected to bombing. Trying to break the silence from within this logic can end up reproducing relations of power. This is further compounded by speech expectations being focused on violent state actors whose silence is interpreted as secrecy, creating a subject position of exclusion which reaffirms the status of the powerful as those who are expected to speak and undermines the vast amount of information which local actors and NGOs possess.
The battle against the unsaid does not take place outside of the structures of power and meaning-making which it critiques. The assumptions and expectations through which we listen directly shape what we hear and how we interpret it. The battle against the unsaid carries the risk of organising and interpreting the layers of silences in ways that end up running along the interests of dominant actors. Fractured and unstable layers of silences are thus, for example, grasped as secrecy by a powerful core which leaves the rest of the world unable to know and to speak. By noting, interpreting and opposing silences, there is then also a risk to (re)produce the function of these silences in a self-fulfilling prophecy.
Silence is fundamentally relational and ambiguous. While silence exists logically whenever a potential speaker decides to be silent, it is relational in that it becomes communicative only when silence is noted by others (Schröter, 2013: 31). Interpreting silence as secrecy and opposing it as a problem that undermines legal and ethical frameworks risks bringing these very effects about. This creates dilemmas for human rights actors, some of which cannot be resolved within the asymmetrical power structures we are embedded in. Yet, the unstable and ambiguous nature of the layers of silences discussed above also points to some spaces of agency which listeners hold. As Xavier Guillaume has argued, silence produces effects which explode a static framework that reinforces the interests of the powerful and can instead present a 'radical irruption' (Guillaume, 2018: 13; see also Montoya, 1999). The layers of construction of silence discussed in this paper hence also offer an opportunity for reappropriation which has been taken up by some actors in creative ways.
For example, I discussed above how the assumption that silence creates legal and moral uncertainty is not only questionable but can itself contribute towards such uncertainty. Rather than reproducing the notion of uncertainty linked to silence, some actors have reaffirmed the solidity of norms and argued that 'the existing legal framework was sufficient and did not need to be adapted to the use of drones, rather, it was the use of armed drones that must comply with international law' (UN Doc A/HRC/28/38, 20104: 56). Reversing the assumption that a lack of information creates uncertainty and thus opens up more leeway for state actors, existing standards are instead taken as the base line unless otherwise stated. This avoids the risk that ambiguity is talked into being. It also rearranges the power position of who is expected to speak and know about these practices.
Similarly, I showed above how the battle against the unsaid can create an impasse where human rights actors understand themselves as reliant on state actors for more information. Rather than interpreting the silence of actors on drone warfare and the controversial 'targeted killing' policies as secrecy which hinders our understanding on the practice and further increases normative and legal leeway for powerful actors, some human rights actors have instead interpreted it as an admission of guilt: There is a stigma attached to targeted killing -especially given its controversial use by the US during the war on terror. It is therefore unsurprising that no politician would like to put their name to such a public 'targeted killing' policy. Add this to the fact that the legality of such actions has been questioned by a number of human rights groups and international actors and the government's reticence to release more information is understandable. (Oxford Research Group, 2017) Rather than grasping the silence by military agencies as secrecy which is used strategically by powerful actors, it might also be interpreted as incompetence and inaptitude, for example, when the lack of clarity about who is involved in armed drone attacks is linked to a lack of understanding by the actors themselves, 'so much so that at times even higher-level policymakers do not know whether drone operations are conducted by CIA or JSOC personnel' (Center for Civilians in Conflict and Columbia Law School, 2012).
Silences around individual attacks and civilian death might be interpreted as incompetence since in practice, intelligence assets are limited, and often rely on a single source. As General Bolduc confirmed (. . .) 'the only assessment we [AFRICOM] have is we continue to fly over the objective post-strike to get a sense of the environment from the air'. (Amnesty International, 2019) Taking the knowledge and expertise of non-state actors seriously, some human rights actors have reversed the positionality of exclusion that is at play when asking governments for more information and have instead offered assistance to government agencies in investigating attacks since, as Airwars have argued 'the MoD's monitoring capabilities may not at present be fit for purpose' (Airwars, 2016).
Such reinterpretations do not solve the wider dilemmas which the battle against the unsaid poses. Yet, they do point towards creative ways of subversion within the discourse. This might help to avoid some of the impasses human rights can find themselves in when confronted with layers of silences around state violence. It also shows reflection on the complicity of 'well-meaning liberals' (Maggio, 2007: 431) in the workings of silence. This is all the more important, given the structures of power in which the battle against the unsaid is embedded.

Conclusion
While there has been a lot of interest in the political functions of silence in IR, research has often remained quite abstract and theoretical. This paper has analysed how silences are encountered, interpreted and opposed in human rights discourses around covert warfare by analysing 125 NGO reports, UN documents and policy papers. Contributing with an empirically grounded analysis, the paper showed that silences are not monolithic but noted in complex layers of denial, partial withholding of information, redactions, delays and so on which work with and through speech. Rather than already approaching silence through a particular mode, such as secrecy, this paper has showed how these fractured layers of silence gain meaning by being organised in ways that often runs along the interests of dominant actors.
The paper showed the ethical dilemmas which silence poses for non-state actors who are opposing covert warfare. I argued that the uphill battle against the unsaid risks focusing on opposing silences rather than opposing the violent practices themselves. Calling for less secrecy can end up reaffirming the power position of Western state actors who are asked to speak while the subaltern become further essentialised and silenced. This is particularly important because of how complicit Western human rights discourses can be in reproducing the relations of power they oppose. Following Spivak's suggestion of investigating the listening skills of Western knowledge production (Spivak, 1988), the paper highlighted the pitfalls of trying to break silence without first reflecting on our own listening.
This is not only of scholarly relevance but has important policy implications. In areas ranging from finance to warfare, transparency is often understood as a value per se and silence is opposed by human rights actors. This paper has showed the ethical dilemmas and impasses this can create for those who are opposing covert practices. Engaging with creative ways of dealing with silence, the paper argued that the fractured nature of silences encountered in human rights discourses points to spaces of agency which the listeners hold. While these cannot resolve the wider dilemmas posed by our embeddedness in asymmetrical power relations, they point towards possible ways of subversion and resistance.