Arms, aviation, and apologies: mapping the Boeing social media response to the 2019 Ethiopian Airlines crash

ABSTRACT

Boeing is famous for aviation but also produces arms, making $29.2 billion from the latter in 2018. The role of the arms trade in facilitating death can be considered a ‘public secret’ - known, but socially unacknowledged. This allows Boeing to represent its role as one of ‘neutral’ technological advancement, obscuring violence engendered by certain products. This paper builds on works on public secrecy, which investigate how (un)acknowledgement obscures everyday security arrangements. How can we know the public secret? We argue that public apology and scandal are boundary-delineating practices, locating certain issues within the public secret and rendering others knowable and sayable. We examine Boeing’s Twitter response to the March 2019 Ethiopian Airlines flight 302 crash. The content: 1) produced the crash as a tragedy, positioning Boeing as ‘sorry’ and capable of grief, 2) allowed Boeing to ‘take responsibility’, positioning safe operation of their products as a moral obligation. Within the wider political contexts of the arms trade and responsibility for safety in commercial aviation, we explain Boeing’s Twitter navigation of apology/scandal not as simply corporate face-saving, but as a practice of (re)confirming the public secret, positioning aviation deaths as knowable/grievable, and those lost to the arms industry as neither.

KEYWORDS:

• Public secret
• public apology
• scandal
• social media
• arms companies

Introduction

In 2019, Ethiopian Airlines flight 302 took off from Addis Ababa Bole International Airport in Ethiopia. The Boeing 737 Max model plane crashed six minutes into its scheduled flight time. The crash killed all 157 passengers on-board and was the second of such crashes to affect Boeing aircrafts in 5 months (the other – beyond the scope of this article – being Lion Air flight 610, which departed from Indonesia and carried 189 people). Boeing representatives had, at various points, suggested that ‘pilot error’ was to blame for both crashes (Robison 2021). Specifically, David Calhoun (Chief Executive) suggested that pilots from Indonesia and Ethiopia ‘don’t have anywhere near the experience that they have here in the US’, and that this inferior training played a key role in the crashes (cited in Kitroeff and Gelles 2020). In fact, the company was perhaps performing ‘strategic ignorance’ (McGoey 2012) of the fact that its Maneuvering Characteristics Augmentation System (MCAS) software was not optimally functioning. However, in 2021, Boeing was fined $2.1 billion for misleading the US Justice Department about the causes of these crashes, and in late 2022, was confirmed guilty of misleading investigators, agreeing to pay $200 million and former Boeing CEO Dennis Muilenburg personally agreeing to pay $1 million (Rushe 2022). This has widely been perceived as a serious failure of responsibility and a ‘scandal’ (e.g. Yglesias 2019). The international scope of this scandal is important because it demonstrates the ways in which technological and military ‘progress’ has resulted in the creation of the ‘global accident’, whereby impacts are felt across the world and are increasingly perceived as normal (Virilio 2007, 7).

In recognition, since 2019, Boeing has offered various apologies – including personal ones from Muilenburg – across a variety of platforms. They did not offer apologies in the wake of the Lion Air Flight 610 crash and rather extended their public apology delivered to Ethiopian Airline crash victims to include those of the previous crash retrospectively, in response to an ongoing investigation (Baker 2019). One platform on which Boeing has offered a number of apologies for the Ethiopian Airlines crash is Twitter,¹ considered a key part of the modern ‘public sphere’ (Habermas 1989) as well as an important space for companies to manage their image (Cusumano 2021).

Aviation is not the only component of Boeing’s business, however; they are also one of the top arms-manufacturing companies, making around 33% of their revenue from this enterprise (SIPRI 2019) and partnering with national militaries (Bousquet 2017). The attention paid to the company following the crash of Flight 302 elides the fact that, on a daily basis, Boeing produces technology that is intended to kill. What is the relationship between arms and aviation in this case? How might the activities within the aviation section of Boeing support the maintenance of the status quo regarding their arms sales? Virilio – in conceptualising ‘the accident’ - makes the case that accidental death and destruction should not be considered as ontologically distinct from the process of technological advancement precisely because ‘failure’ is ‘programmed into the product from the moment of its production or implementation’ (Virilio 2007, 211–2). When conceived in this way, the progress of aviation inherently means the potential for increasing numbers of accidental deaths. We argue that the recognition of accidental deaths engenders a ‘cleansing’ of arms manufacturers by presenting these companies as moral, technically knowledgeable, and crucial to the global politics of security. Owens’ (2003, 607) work, which locates the framing of accidental deaths within the ‘entrepreneurial, wartime spirit that sustains the hegemonic order’, can be seen as in conversation with that of Virilio (2007). Of course, deaths within the context of commercial aviation crashes are distinct from civilian war casualties; however, as we demonstrate, Boeing’s apology statements must be understood in relation to its (and more broadly, US) technological advancements made through war/security. As Robison (2021) demonstrates, not only is arms manufacturers’ relationship to war defined through producing weapons, but further, their success even within aviation can only be understood in the context of the ‘opportunities’ afforded by WWII. This fact speaks to the deep intertwining of the ‘military-scientific and industrial complex’ (Virilio 2007), a conceptual blurring of private and public security and technological advancements facilitated by war which Boeing as a company has benefited from financially.

In this paper, we build on an array of critical security studies literature which has dealt with various aspects of the arms trade (e.g. Stavrianakis 2016; Larsson 2020) and private security actors (e.g. Chisholm 2014; Chisholm and Stachowitsch 2016; Joachim and Schneiker 2012; Leander 2010; Stachowitsch 2013). This scholarship demonstrates that ‘[t]here is a distinct political economy, strategic orientation and – crucially – form of justification based on human rights, humanitarianism and morality that frame … arms transfers as part of broader war-making and war-preparation practices’ (Stavrianakis 2016, 841) trends which may even be exacerbated in the context of private arms companies. We therefore focus on the intersection between brand management representations and the political security practices (Cusumano 2020) which are evident within Boeing’s social media apologies.

By analysing Boeing’s social media response to the Ethiopian Airlines crash, this paper draws on and contributes to a growing body of critical scholarship challenging the notion that the politics of scandal engenders ‘an instantaneous “switch” [of revelatory recognition] rather than a process requiring interpretive work’ (Stampnitzky 2020, 598). Similarly, previous work surrounding the performance of public accountability (including expressions of apology) implicitly assumes that public knowledge renders injustice visible (e.g. Daase et al. 2015). However, as some compellingly demonstrate, the politics variously articulated as that of ‘not knowing’ (Nordstrom 1999, 14), of ‘strategic ignorance’ (McGoey 2012), or ‘known unknowns’ (Davies and McGoey 2012, 65) is one which is affectively maintained, rather than referring to a mere absence of information (Walters 2015, 2021). Such a state of not knowing should also be understood as useful and normal within organisational structures. Indeed, ‘the knowledge of what individuals aspire and struggle not to know’ (McGoey 2012, 571) is socially determined and maintained for various strategic reasons. The assumption, therefore, that organisations strive for transparency and demonstrations of competency and expertise might be misleading in many contexts.

We examine apologies delivered by an arms manufacturer outside of the context of arms in order to explicate how the politics of scandal is delineated and bounded, in this case around an understanding of accidental death. As Page (2014) notes, public apologies on social media have a range of impacts, and ambiguities produced within this space have ‘concrete effects on specific practices’ related to war, security and violence (Best 2008, 356). In this article, we aim not to assess whether Boeing’s actions were rational or dismiss their apologies as cynical measures designed to improve the public image of the company. Instead, we suggest that Boeing’s social media-based apologies express regret for unintentional loss of life whilst consigning the intentional sale of arms to the realm of the ‘public secret’ – knowledge that society possesses but cannot or will not discuss (Taussig 1999). An emphasis upon this contradiction illuminates in new ways how public secret, scandal, and apology reinforce and disrupt one another, and helps us consider what the consequences of this might be.

The article proceeds as follows. First, we pay attention to the concept of the ‘public secret’ and suggest that the dynamic of ‘that which is known but cannot be articulated’ (Taussig 1999, 5) is key to understanding arms companies. We explain that the public secret has friction with the concept of scandal, where blame must be apportioned. Second, we demonstrate that public apology is a means of performing accountability and addressing recognised ‘scandals’. After this, we set out the methodology, a discourse-theoretic approach using the concepts of presupposition, predication, and subject positioning (Doty 1993). The final section is an analysis of Boeing’s Twitter output over the period of the crash, as well as the public statements these tweets share through hyperlink. Overall, we argue that Boeing’s social media apologies are not simply face-saving measures but also ambivalent ‘breaks’ in the frame of the public secret, which operate to acknowledge the loss of some lives but re-inscribe secrecy by obscuring others.

Arms, aviation, and the public secret

Secrecy is often conceptualised as a straightforward binary, obscuring the nuances of its operation (Van Veeren 2019, 386). Despite ongoing differences in disciplinary conceptions, a liberal perspective on secrecy (focusing on governance and interrogating the balance between secrecy and openness) predominates within political science (Walters 2015, 288). This binary way of understanding secrecy regulates our understanding of the world by reinforcing the separate nature of other categories such as inside/outside, masculine/feminine, private/public (Hoijtink 2019, 149). Binary conceptualisations of secrecy have epistemological implications, positioning things either as knowable or not knowable, with little space for nuance or recognition of ‘known unknowns’, which illustrate the importance of organisational context for facilitating cultures of ignorance rather than transparency (Davies and McGoey 2012). Relatedly, as Best (2008, 355) notes, mainstream approaches to social research ‘tend to downplay the interpretive dimensions of the limits of knowledge’. In contrast, this article employs an anthropological concept (Walters 2015, 288) to better understand the work of Boeing’s apology: the public secret, which is perhaps best represented through the work of Michael Taussig.

The public secret is defined as ‘that which is generally known but cannot be articulated’ due to social pressures or an inability to form the ideas necessary to explain it (Taussig 1999, 5). This is a form of antiepistemology, which traces the ‘shadow’ of knowledge and its production to ask how social and political processes might obscure or create ambiguities of knowledge (McGoey 2012, 3). Militarised spaces and subjects are good examples of the structures of public secrecy because they can be difficult to access (Ericson and Wester 2022). The arms trade in particular is marked by what Stavrianakis (2020, 231) calls ‘managed openness’ in order to maintain the status quo of arms sales. It is widely known that the arms trade causes fatalities but the management of information around/by arms manufacturers makes social acknowledgement difficult. As a result, Stavrianakis encourages us to ask what information is publicly available. Whilst the conceptualisations of ‘secret’ described above therefore re/produce other binaries, it is possible to disrupt the binary of the public secret and social order by interrogating the relationship between the two (Young 2011). This gives us cause to think beyond the contents of specific secrets, instead asking what political work is being undertaken by creating ambiguous spaces (McGoey 2012).² New sites of possible ambiguity and contestation are emerging all the time. As Kuntsman and Stein (2015, 15) assert in their work on the militarisation of Israeli society, ‘… public secrecy has taken new forms in the social media age. Today social media and social networking function as crucial sites and tools of secrecy management’. That is, digital spaces represent key sites of analysis for contemporary works on public secrecy.

Epistemological contestation is central to the re/production of ambiguity (Best 2008), reinforcing the public secret. One example of this is the designation of ‘accident’. Owens (2003) argues that what is knowable as an accident is socially produced, acting as a site for epistemological debate (see also: Virilio 2007). It is through the epistemological delineation of ideas such as ‘accident’ that the social conditions for the apportionment of blame (or not) are created. The label of ‘accident’ therefore often works to confirm the public secret because it places accountability – and the societal recognition of blame – out of reach. In the case of the arms trade, ambiguities are also created when technology is labelled ‘dual use’. Through the representation of ‘neutral’ technological advancement, a positioning of activities primarily through the lens of the needs of the client, whoever that might be, can be seen as a means of ‘profit[ting] equally from talking peace and fuelling war’ (Larsson 2020, 44). That is, microchips may be produced for arms or aviation, providing a layer of plausible deniability where deaths may be considered accidental or not. Though Owens (2003) largely discusses accidents in the context of war, she makes several comparisons with aviation and air disasters.

Davies and McGoey (2012, 559) argue that ‘to prove the utility of ignorance is to demonstrate that ignorance has a material concreteness that undermines its own ontological status. Ironically, once ignorance is defined, it loses its very definition’. Similarly, the public secret is unsayable, but it can be rendered present through partial glimpses, allowing us to trace the exercise and maintenance of power. Often this means examining where procedures designed to ensure accountability have gone awry and this is left unaddressed. The example given by Taussig (1999) is that of extra judicial police violence, which people often know about but cannot or will not articulate. There are other means through which the public secret is rendered legible, such as discursive processes of denial which reinforce that no wrongdoing has occurred whilst simultaneously invoking the spectre of wrongdoing itself (Cohen 2013). The residue of violence or suffering also resists secrecy by virtue of its materiality (Austin 2017), demanding recognition and rendering it knowable. It is harder to ‘look away’ from physical evidence of suffering, whether that is shells from a bombing or bullet holes in a wall, though this is sometimes minimised discursively, e.g. through claims of exaggeration (Cohen 2013). Relatedly, grievability as a process works against public secrecy because it forces the acknowledgement of humanity (Stec 2016). In contrast, when victims are not considered fully human – ‘if a life is not grievable, it is not quite a life’ (Butler 2004, 34) – the absence of grievability maintains the public secret. Taussig (1999) asserts that the public secret can be made visible through defacement, by a public act of marking; both grief and the material elements of violence may, then, be conceptualised as a form of defacement, whether intended or not.

In asking how the public secret becomes known, we also need to ask the parallel question: how are the boundaries of the public secret re/inscribed? Below we set out a conceptual framework incorporating public apology and scandal before applying this to the Ethiopian Airlines crash. We demonstrate that, together, representations of public apology and scandal work to re/produce the contours of the public secret with regard to arms and aviation: when some deaths are positioned as worthy of public recognition, this re-confirms other deaths as unspeakable, ungrievable, and ultimately unknowable.

Public apology and responding to scandal

The public apology has become a commonplace means of responding to political violence, scandal and various forms of norm violation, and as such has inspired a wide body of interdisciplinary literature (e.g. Brooks 1999; Gibney et al. 2008; Mihai and Thaler 2014 etc.). Apologising actors are diverse and include, for example, states and governments, religious institutions, public figures, and corporations (Celermajer 2009). Academic work on both corporate and political apologies has focused on highlighting ideal standards against which statements can be held in terms of sincerity, brand management, and restoring public trust (e.g. Hall 2020; Lee and Atkinson 2019). However, because the normative criteria of such an adequate statement of apology are hardly agreed upon in any context, institutions may even suffer from fresh criticism that their attempts to mend relationships are simply ‘cheap, “gestural politics” awash in self-interest and crocodile tears’ (MacLachlan 2010, 374). This makes the public apology a particularly precarious discursive strategy with manifold uncontrollable consequences (Renner 2011). Understanding public apology as a way of responding to normative (dis)order (Johnson, Basham, and Thomas 2022), we explore three particularly important facets of the public apology.

First, and in line with our methodological and epistemological approach, we prioritise the affective content and impacts of the public apology. This means that, further than viewing the public apology as necessarily accompanied by the (un)convincing performance of ‘sincerity’, we build on approaches which ask what the emotional/affective content of public apologies does politically (e.g. Ahmed 2014; Augoustinos, Hastie, and Wright 2011). One example of this is the way in which contradictions – such as the recognition of aviation deaths in a broader context of the arms trade – are ‘smoothed over’ epistemologically through apology. This is consistent with a now wide-ranging body of International Relations scholarship which takes a keen interest in emotion as both sustaining and disrupting politics of violence, security, trauma, and war (e.g. Åhäll et al. 2015; Hutchison and Bleiker 2008; Solomon 2012). We view this process of making ‘sense’ of the scandal alongside the politics of grievability, within which only particular lives are recognised as fully liveable (Butler 2004).

Second, the public apology has often been positioned as a definitional challenge to the politics of secrecy, privacy, or obscurity. As such the public apology has come to be recognised as a performative gesture of transparency in line with democratic norms and priorities (Daase et al. 2015). Much scholarship exploring public apology tends toward understanding its function in two opposing ways: ‘a progressive tool for restorative and transitional justice … [or] as ideological weapons deployed to reinforce the moral authority of elites as they cherry-pick which atrocities they will renounce for the benefit of their image’ (Smith 2022, 1). This interpretation relies upon the impact of the speech being determined by the intentions of the speaker, which poststructuralist approaches challenge by definition. The following analysis builds upon this alongside the idea that institutional taking of responsibility requires a public revelation around previously unknown activities or events. We deconstruct, however, the notion that public acknowledgement through an official statement of apology operates to directly challenge the state of secrecy through binary notions of voice versus maintaining silence (Hoijtink 2019, 149). The importance of contesting such binaries is all the more pertinent in the context of norms of transparency, which do not necessarily result in more open practices, but instead lead to ‘putting a gloss on the truth, setting the public agenda, spin-doctoring, tendentious leaks to the media, selective concern about suitable victims, interpretive denials’ (Cohen 2013, 10). That denial and lack of transparency exists alongside norms and practices of openness makes mapping the epistemological boundaries of these states inherently challenging if not futile. Rather, and alongside a long tradition of feminist research, we trouble the binaries between silence/voice, secrecy/revelation, and private/public to demonstrate that public apologies delineate the boundaries of what is sayable/unsayable.

Finally, we suggest that the concept of scandal occupies a central epistemological space between performative transparency and secrecy. Secrets often come to be known through scandal, as is reflected within literature around scandalogy (Andrews, Connor, and Wadham 2020; Johnson, Basham, and Thomas 2022). In order for a scandal to have social effects, a transformative process must occur, wherein it not only becomes publicly known but also socially recognised as significant/dominant and negative (Andrews, Connor, and Wadham 2020; Stampnitzky 2020). Johnson (2017) draws attention to the workings of scandal in the context of war, which in causing outrage in response to some forms of violence, can also produce others as acceptable and even moral. One such example is Abu Ghraib, whereby those detained were subjected to practices of torture that were considered as atypical of the ‘American’ military, and therefore unacceptable (Enloe 2004; Stampnitzky 2020). Such scandalous events incur a range of actions such as inquiries, sanctions, and apologies (Andrews, Connor, and Wadham 2020). Once the status of scandal has been established there ‘is a moral obligation to respond to it … to not respond to a scandal would raise questions about the moral character of a society that would tolerate the violation of its constitutive norms’ (Johnson, Basham, and Thomas 2022, 13–14).

The concepts of the public secret, apology, and scandal should be seen as relational, their friction being (re)productive of the lines through which ‘the secret’ is drawn. As a result, the discourses examined within this article can usefully be viewed as a range of ‘line-drawing manoeuvres’ (Johnson 2017, 706). In asking what becomes scandal, we can also ask parallel questions: for which events is there not an obligation to respond? How are the boundaries of scandal and apology negotiated in ways that bound the public secret? As Johnson (2017, 217) argues, ‘scandals function to enable, excuse and obscure the complex landscapes of violence that define the spectacular and mundane sites of contemporary war’. We aim to show how the negotiation of scandal (a plane crash) and related apologies might obscure other practices of violence engendered by martial industries (the manufacture and sale of arms).

Methodology

We examine the social media output of Boeing over a six-month period between January and June 2019, covering the period just before and after the crash. The data set for this article is therefore 94 tweets (all tweets sent over this time), including retweets (RT) from the official Boeing account. We also examine images, videos, and linked URLs from these tweets because these are part of the same discursive environment (Campbell 1992, 62–63). Our emphasis is on discourse, defined as ‘practices, representations and interpretations’ (Peoples and Vaughan-Williams 2020, 65). We adopt a discourse theoretic approach (Shepherd 2006), exploring how the discursive constitutes subjects by providing momentary fixity of identity. In addition, we pay attention to the ‘affective-discursive’, observing ‘what the emotional, however defined, do’ (Åhäll 2018, 48). Inspired by Ahmed (2004), we consider how both emotional representations and affective (or felt) impacts help to create and naturalise particular discourses and power relations. To do this, we make use of Roxanne Doty’s 1993 concepts, which have been deployed to good effect in many other articles in a similar space (e.g. Åhäll and Borg 2013). Doty encourages us to ask how particular (in)actions are made possible through discourse and practice.

We first consider what is pre-supposed by the tweets (and related materials), knowledge one needs to understand and contextualise what is being said. Doty (1993, 314) explains, 'The intertextual nature of these texts becomes important here. To borrow a phrase from Barthes (1987,135), these texts were ”plugged in” to each other as well as to other texts; other foreign policy texts, social science texts, and nonacademic texts’. Secondly, we draw upon the concept of subject positioning which aims to understand how different subjects and objects relate to one another. This engenders an understanding of who has power and who is marginalised within a given situation. Thirdly, we make use of predication, which considers how people or organisations might be constructed when particular ideas are attached to them. Taken together, these concepts make it possible to tease out contradictions within a discourse, highlighting the way in which the public secret, apology, and scandal each negotiate and delineate the boundaries of the others. We deploy these concepts below to analyse the aforementioned digital material, mapping the discourse in a way that reveals glimmers of contextual security arrangements, whereby value is placed on some lives but not others.

One challenge of digital data is digital decay, where data is removed from the Internet, either deliberately or through lack of maintenance (Gulotta et al. 2013). During data collection, all text from these tweets was collected, copied and pasted into a separate document for ease of analysis; subsequently, some tweets have been removed. Therefore, whilst we still have the URLs, the tweets themselves no longer exist, and the account @BoeingCEO has also been deactivated. We have provided alternative URLs containing the same materials (from other Boeing accounts or the Internet Archive). Use of API scraping tools requires Twitter’s permission and one of the terms of agreement is not to utilise deleted material. As Littman (2019) notes with frustration, this poses a challenge for archivists and the deletion of tweets can therefore be used as a means of brand reputation management. An API was not used in this case, so the rule does not apply. More broadly, as a brief aside: it is interesting to consider this problem in the context of an article concerned with secrecy, the ‘public’, and what is knowable. Whilst digital decay does complicate this, the reasoning for examining Twitter posts – e.g. rather than simply speeches – is because this is a hyper-visible element of the public sphere (Habermas 1989). As Virilio (2007, 59) notes, there is also an affective dimension to ‘instantaneous communication’, which is especially important in producing ‘accidents’. This is, therefore, a fruitful space for the exploration of the public secret, apology, and scandal.

Analysis

As Milliken (1999, 229) explains, ‘discourses produce as subjects publics (audi­ences) for authorized actors, and their common sense of the existence and qualities of different phenomena and of how public officials should act for them and in their name (e.g. to secure the state, to aid others)’. The representations presuppose that Boeing is able to speak as an ‘authorized actor’ within the domain of international safety, security and politics. It is presented simply as common-sense that Boeing, as a corporation, can apologise and perform remorse/regret/sorrow in the context of significant human suffering. The construction of apology establishes Boeing not only as ‘sorry’ – an emotional state – but also maintains its rational security expertise as set out below. This can be considered something of an ontological contradiction; we argue that unpacking the contradictions in this case highlights the negotiation of the boundaries between public secret, apology, and scandal.

This section of the paper offers a two-pronged analysis of Boeing’s social media response to the crash scandal. Thus, this section focuses on the six-month period January–June 2019, which represents a break in the frame of Boeing’s social media presence, previously used to ‘invisibilise … violence by presenting these companies as drivers of human progress’ (Jester 2023, 1). However, we pay specific attention to how Boeing’s response to fatality delineates the frame of the public secret and subverts opportunities for connecting the scandal to wider practices of arms manufacturing and liberal norms of security. In particular, the following analysis draws attention to two interrelated themes which weave throughout the discourse: 1) representations of tragedy and 2) taking (ambiguous) institutional/personal responsibility.

‘Tragedy’

In January 2019, Boeing stated, ‘What a great year! We delivered 806 commercial and 96 military aircraft in 2018’ (tweet 2) and explained that ‘We’re honored to be ranked again as the No. 1 Aerospace & Defense Most Admired Company of 2019’ (tweet 4), highlighting the relationship between security and aviation within their work. Following the crash of Ethiopian Airlines flight 302, however, their statement includes the line ‘we know lives depend on the work we do’ (tweet 10; see also tweet 14 and tweet 18). The contrast between these tweets is indicative of the common-sense and always-already instated nature of the public secret: whose lives ‘depend’ on their work? Whilst such a statement indicating the life-and-death impacts of Boeing’s technology applies to its production of arms, the social and political context makes this unsayable (Taussig 1999). However, as we show below, representations of ‘lives’ make visible a glimmer of the public secret, and bounds public apology/scandal. As Butler (2004, xvii) explains, ‘the public sphere is constituted in part by what cannot be said and what cannot be shown. The limits of the sayable/what can appear circumscribe the domain in which political speech operates and certain kinds of subjects appear as viable actors’.

‘Tragedy’ is a common theme in Boeing’s tweets and can be aligned to the concept of the ‘accident’ (Virilio 2007). For example, tweet 9: ‘As part of standard practice following any accident…’; and the linked statement in tweet 6 referring to the ‘Ethiopia Accident Investigation Bureau’. The crash is, therefore, positioned as unexpected, chiming with Leese’s (2015, 277) explanation that ‘Failure, in the aviation security and the security industry narrative, [occurs] through the element of “surprise”’, and not, as Virilio (2007) suggests, a structural component of technological advancement. Importantly, this draws upon the idea of the crash as being ‘endogenous to political agency or social and cultural structures and processes’ (Johnson, Basham, and Thomas 2022, 7; see also Owens 2003). Thus, whilst there is significant suffering, it is not presented as foreseeable. Examples are: ‘We at Boeing are sorry for the lives lost in the recent 737 accidents and are relentlessly focused on safety to ensure tragedies like this never happen again’ (tweet 14, RT of @BoeingCEO; see also tweet 10, RT of @Boeing); another describes the crashes as ‘two tragic accidents’ (tweet 8). The crash is constructed as an event that happened to/around Boeing, circumscribing the event’s representation as a ‘scandal’ by locating it beyond Boeing’s knowledge and therefore within the domain of tragedy. This representation exists uneasily alongside the aforementioned public apologies, and with below constructions of a responsible subject.

Those who died are referred to variously as ‘victims’ (tweet 20 RT of @cbseveningnews) and ‘passengers’ (tweet 10, RT of @BoeingCEO; see also the statement attached to tweet 6 RT of @boeingairplanes). The representation of this group as worthy of sorrow, sympathy, and grief functions as the enactment of a ‘death ritual’ (Morse 2018, 246) which locates these deaths within the realm of the sayable. Subdued articulations of ‘sadness’ add another affective layer to the portrayal of tragedy. Sadness at this moment underscores the presupposition that Boeing – a private company – is capable of emotion and therefore sincere apology. On 10th March, Boeing retweeted a statement from the company’s media room: ‘Boeing is deeply saddened to learn of the passing of the passengers and crew on Ethiopian Airlines Flight 302, a 737 MAX 8 airplane’ (tweet 6; ‘deeply saddened’ is also used in tweet 3 and tweet 4). Emotional language in these tweets is mostly passive, and as Kampf (2009, 2268) suggests, the use of ‘passive voice’ whilst ‘omitting the agent all together [operates] to avoid agency’. The tweets acknowledge the deaths but do not state clearly who/what caused the tragedy.

The humanisation of Boeing (re)confirms the nature of the crash as tragedy. The term ‘heartfelt’ is particularly evocative (tweet 4) because it is ultimately derived from a focus on the heart, a living, beating organ: ‘Our hearts are heavy, and we continue to extend our deepest sympathies to the loved ones of the passengers and crew on board’ (RT of Boeing statement, tweet 10). In a different context (autonomous weapons), Fosberg (2019) explores the trend towards granting corporate personhood, treating companies as human beings, and argues that it is difficult to make this case concerning weapons due to their human costs. With respect to the crash of Flight 302, however, Boeing is humanised through this ‘heartfelt’ embodiment. Humanisation is also underscored through the subject positioning of Boeing, the passengers/crew who died and those who care about them (tweet 6). The links between these groups are presented as multi-layered, incorporating not only family but other bonds: ‘Our purpose at Boeing is to bring family, friends and loved ones together with our commercial airplanes – safely’ (RT of Boeing statement, tweet 10; ‘family and loved ones’ is repeated in tweet 8).³ Boeing, therefore, positions themselves as responsible for the lives of people’s family members, a representation haunted by the public secret of arms deaths. The discourse in this case once again reveals a presupposition: that Boeing considers themselves responsible for the loss of some people’s family members, in some circumstances. This is reinforced by Boeing’s statement that the crashes are ‘uniting people and nations in shared grief for all those in mourning’ (RT of Boeing statement, tweet 10), constructing these families’ grief as their own.

Representations of then-CEO Dennis Muilenburg also humanise Boeing. Muilenburg is the human face, expressing his own personal feelings of tragedy: ‘RT @cbseveningnews: WATCH: In his first interview since two fatal 737 MAX crashes, Boeing CEO Dennis Muilenburg tells @NorahODonnell that he “personally” apologizes to victims’ families; ‘of my 34 years at Boeing, this has had the biggest impact on me of anything that’s happened’ (tweet 20). Apologies of this nature conduct discursive ‘work to balance … emotion with reason and rationality’ (Augoustinos, Hastie, and Wright 2011, 518). Thus, despite a lapse in technical expertise, apology resists scandal because Boeing cannot be fully at fault if they themselves are emotionally wounded by the crash (Johnson 2017). Muilenburg is therefore represented as “speaking for’’ the company in performing personal regret. This is further emphasised by the use of Muilenburg’s image in photographs and videos attached to tweets. The videos are either of him talking to a camera in a warehouse or appearing on television to be interviewed (e.g. tweet 12RT; 14RT; 15RT; 20RT; 21RT). However, Muilenburg’s presence creates a jarring distance between the apology and the company through performances of contrition. Muilenburg was the owner of the now-deactivated account @BoeingCEO, with the bio ‘Former Boeing intern. Current Boeing CEO. Lifelong aerospace enthusiast. Proud Iowan’, and was replaced by Dave Calhoun in December 2019. Rather than re-assigning this account to Calhoun, Boeing chose to invisiblise the digital space. This can be read ambiguously (Best 2008): as a CEO leaving following a tragedy; a company taking responsibility and removing someone who has made a serious error; or an attempt to remove evidence of the link between a scandal-plagued CEO and a company with significant legal problems. This operates in parallel to the ambiguous representation of responsibility/agency, explored below.

“Responsibility”

Above we have argued that Boeing represented itself as grieving a ‘tragedy’. In this section of the analysis, we show that Boeing presents itself as ‘responsible’, especially through appeals to safety. Emotional discourse of tragedy is uneasily dispersed between technical information about safety updates planned or made to the Boeing 737 MAX in general and the MCAS software in particular. MCAS software was ultimately found – following investigation – at fault within the accidents (Gates and Baker 2019). It is notable that the term ‘safety’ was only used once in Boeing’s tweets in the period from 4th January until 18th March (in relation to another crash, tweet 3). However, it frequently features in Boeing’s social media discourse in the days following the Ethiopia Airlines crash, stating that ‘safety is Boeing’s number one priority’ (tweet 7), they are ‘embracing the highest standards of safety’ (11RT), and that they ‘will take any and all additional steps necessary to enhance the safety of our aircraft’ (13RT, see also tweets 10RT, 14RT, 16RT, 17RT, etc.).

As a result, the predicate ‘safe’ is affixed to Boeing. Writing about Private Security Companies, Leander (2010, 474) argues that their ‘authority is embedded in and derived from a discourse/practice about security/risk expertise’ and Boeing is represented similarly. The representation of safety discursively minimises risk within Boeing’s operations. Facts and figures reassure us of safety, as the tweet 15 video (of Muilenburg at an airfield) tells us, ‘our talented test pilots have now completed 120 737 Max flights, totalling 203 hours of airtime’ using the software updated after the crash. The concept of safety functions as a central feature of Boeing’s identity, produced as – and presupposed to be – always already safe. For example: ‘safety is at the core of who we are at Boeing, and ensuring safe and reliable travel on our airplanes is an enduring value and our absolute commitment to everyone’ (statement attached to tweet 10). This is despite the fact that safety is most explicitly mentioned in tweets following a failure so catastrophic that it was labelled a scandal (Johnson 2017; Yglesias 2019). Indeed, in an interview post-crash, Muilenberg states that action taken to improve safety ‘reinforces our values as a company’ (tweet 21). The invocation of safety as core to the Boeing identity also reveals a shadow of the public secret: ‘safety’ is not necessarily about preventing loss of life, but rather preventing loss of life when products do not function as intended.

Safety is linked to ideas of technological futurity and advancement (Owens 2003). This positions Boeing within the wider technology industry, marking them out as making a valuable contribution. Tweet 17 (RT @commercial_crew) states: ‘S-A-F-E-T-Y, safety is a top priority as @BoeingSpace and @SpaceX prepare for their crewed missions to the @Space_Station. Safety measures already taking place include testing flight hardware, preparing for emergencies & rehearsing rescue techniques’. Boeing is positioned alongside NASA, a highly regarded institution known for undertaking missions with great technical expertise.⁴ Boeing is also constructed as leading the funding of pilot training: ‘We are cleared for takeoff! #Boeing donates $3 million to fund pilot training and aviation maintenance programs @EmbryRiddle’ (tweet 5). Boeing, therefore, consistently represents itself as being at the forefront of ‘neutral’ technological developments that create the ‘future’. These are therefore taken to be unquestionably ‘good’, despite the implication of increasing potential for even accidental harm/death (Virilio 2007): ‘How will the future of air travel continue to evolve? New aircraft concepts for the Transonic Truss-Braced Wing were revealed at the @aiaa #aiaaSciTech forum’ (tweet 3). This is congruent with other tweets utilising predicates representing Boeing as modern or innovative (‘our excitement to innovate and explore’ (tweet 4)). Narratives of institutional responsibility for safety, therefore, sit alongside de-politicised ideas related to neoliberal technological advancement (Leander 2010). The connections between safety and security – as embedded within liberal entrepreneurial notions of war (Owens 2003; Virilio 2007) – provide the context for Boeing’s production as a necessary provider of rational expertise, even in the face of serious failure. This process ‘upholds the United States’ emerging entrepreneurial approach to war for what many see as a new ‘imperial’ age’ (Owens 2003, 597).

‘Integrity’ is another important predicate within this discourse, situated in relation to ‘safety’, and only appears within our dataset in the context of Boeing’s response to the crash. For example, tweet 10 states: ‘lives depend on the work we do, and that demands the utmost integrity and excellence from our team. Here’s our commitment to airlines, passengers and the aviation community’ and tweet 11: ‘Our Boeing team is proud to stand with our customers and partners in embracing the highest standards of safety, excellence and integrity’ (see also tweet 16RT; tweet 22). There is a crucial relationship between safety and integrity, which are both presented as moral issues. However, the practice of apology performs integrity and shapes the boundaries of what can be considered scandal or not (Johnson 2017). It is commonly understood that the public apology acknowledges acceptance of moral responsibility for norm violation/wrongdoing and can indirectly repair the social image of the speaker (e.g. Harris, Grainger, and Mullany 2006). In recognising a disruption, the politics of apology paradoxically (re)produces the apologising actor and their activities as always already morally virtuous. In this case, the predicates of ‘safety’ and ‘integrity’, presented as key facets of Boeing’s identity, are positioned as possible only because the crash(es) presented a key opportunity for the company to reflect on its identity/values. The use of apology attempts here to ‘reproduce the [company] as an ideal’, to paraphrase Ahmed (2004, 109).

Boeing is positioned as holding power over air passengers and having moral/professional responsibility for their safety. As above, whilst it is presupposed that the crash was an accident, indicating error, Virilio (2007) notes that there is a – socially produced – line between accident and negligence. Indeed, institutional failures can be considered as caused either by intentional wrongdoing or despite the best of intentions (Howlett 2012, 544) and Boeing’s apologies are ambivalent in their invocation of intentionality. Because the production of the accident did not result from a singular moment of revelation to the public but rather an evolving process, the discourse of apology operated to continually (re)draw lines surrounding intentionality and therefore scandal itself. Whilst Boeing has unintentionally harmed passengers on the downed flight, they are simultaneously positioned as having responsibility for future passengers and other lives, which is jarring in light of its arms manufacturing endeavours. This is presupposed through the use of active verbs such as ‘ensure’ (tweet 14) and ‘provide’ (tweet 3), actions to improve safety, as well as the explanation that ‘Boeing Supports Action to Temporarily Ground 737 MAX Operations’ (tweet 8). This discourse contests that of tragedy, which is based on the foundational notion that nothing preventative could have been done. Fluctuating and contradictory articulations of agency like these not only typify corporate ‘blame-shifting’ (Cusumano 2021) but also speak to the ways in which scandals are continually re-drawn as they unfold and become known/knowable.

Important in understanding Boeing’s framing of institutional responsibility is its relationship to the government, and Boeing is represented as cooperating with investigations at both the US and Ethiopian level and therefore as a good international citizen, with a role to play internationally (Wheeler and Dunne 1998). This positioning is multi-layered. Firstly, Boeing is represented as subject to government rules and regulations (‘In accordance with international protocol …’ (linked statement in tweet 9)). They are presented as being subordinate to both national/international governance, providing Boeing some distance from scandal. One tweet says ‘A Boeing technical team will be travelling to the crash site to provide technical assistance under the direction of the Ethiopia Accident Investigation Bureau and U.S. National Transportation Safety Board’ (tweet 6). Another states: ‘all inquiries about the ongoing accident investigation must be directed to the investigating authorities’ (tweet 9, in an image). Boeing thus are ‘following the rules of the game’ (Baudrillard 1994, 15, quoted in Johnson 2017) in this context by participating in investigations. This is further reinforced by terms such as ‘leadership’, e.g. tweet 19 is ‘RT @boeingairplanes: We appreciate the FAA’s leadership in taking this important step in bringing global regulators together to share information and discuss the safe return’. This allows governance structures – in the US/Ethiopia – to assume control over the crash investigation, whilst positioning Boeing as acting responsibly in ceding control over the matter.

Boeing is also constructed as learning from its mistakes. On 29 May (marking the end of Boeing’s social media campaign on the scandal), the company Tweeted, ‘Dennis Muilenburg says the company is learning from the accidents, and vows to make improvements’ (tweet 21RT). These articulations reproduce the notion that there is a singular ‘truth’ which can be discovered through investigations/commissions, which will ultimately allocate blame to the secretive individual, and is reminiscent of Vermeir’s (2012) ‘privative’ view of secrecy, ‘a force which deprives us, as publics, citizenries, persons, of knowledge’. That is, understandings of the contours of scandal directed by a ‘dirty secret’ will always subvert structural understandings of how harm is disproportionately experienced by some and not others within everyday arrangements of security (Thomas 2021).

Conclusion

This article has positioned social media apologies for the scandal of a fatal aviation crash within the wider context of arms manufacturing. It has ultimately argued that the process of ‘knowing’ through scandal and performative transparency also works to conceal the unknown or ‘known unknowns’ (Davies and McGoey 2012). On the part of Boeing, of course, it can be concluded that their social media strategy during an unfolding scandal simply works as ‘rational’ corporate face-saving, and that companies wish to appear as transparent, often in spite of empirical evidence to the contrary in the context of normalised organisational cultures of ‘strategic ignorance’ (McGoey 2012). However, our analysis has demonstrated that in order for Boeing’s Twitter apologies to appear to audiences as ultimately rational and successful in repairing the company’s image, we must accept the implicit affective workings of the public secret. In order for Boeing to avoid any threat of a moral reckoning around its production of weapons, the epistemology of ‘accidental’ death in the wider context of neoliberal security practices must be discerned. The workings of the public secret here are subtle but extremely meaningful. Indeed, as Walters (2021, 14) argues: ‘State and non-state actors certainly do at times conspire, invoke norms and laws of secrecy to cover up mistakes or illegalities and much else … But we should avoid the presumption of too much rationality, coherence or success in these efforts’. Our analysis demonstrates the importance of tracing these inconsistencies – often through emotional performances – because crucial to the epistemology of secrecy, scandal and everyday security practices is how certain representations become possible. We do not suggest that Boeing is unsuccessful in their corporate apologies, but that the context for perceiving ‘success’ in this case must be interrogated alongside conceptions of secrecy/scandal.

Further, as Thomas asserts (2015, 17): ‘Resisting secrecy in the name of publicity does not escape the violence that is committed in the name of security’. We have therefore, throughout this paper, implicitly asked what/who is missing from apology discourses, and whose lives are therefore not knowable in relation to the state of ‘scandal’. As Rappert and Bauchspies (2014, 1) argue, absence ‘can provide a remedy to misperception about the solidity of what is deemed present’. Boeing is represented as having a relationship with its CEO, its customers, and international governance organisations, whilst invisibilising arms manufacturing as known but ultimately secret/unspeakable. It is possible to begin to make this visible, as we have done above, by examining what is present; all parties are represented as concerned with passenger safety in the context of aviation, especially in light of the Flight 302 crash. Boeing makes $29.2 billion per year from the sale of arms, which are deployed intentionally to kill, though the reality of their deployment is not widely known/knowable. There is a sharp contrast between the grievable lives lost as a result of a fatal air crash, where the product malfunctioned – paradoxically positioning Boeing’s responsibility alongside the narrative of accident/tragedy – and the lives lost as a result of bombings, where the products kill by intentional design. Ultimately, Boeing’s apologies, by taking responsibility for lives lost through ‘the work we do’, further reify distinctions between seemingly indispensable and neutral technological advancements embedded within entrepreneurial warfare and the tragic accidents which ‘occasionally’ accompany aviation.

Acknowledgement

We would like to thank two anonymous reviewers for sustained and detailed engagement with our work as well as the editors of Critical Studies on Security for an excellent publishing experience. We would also like to thank Katy Parry for kindly reading an early and very rough draft of this paper, as well as participants of the 2022 BISA panel at which this work was presented: 'Remembrance, War and Peace: Critical Approaches to Violence and Militarism’.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21624887.2023.2267328

Additional information

Notes on contributors

Natalie Jester

Natalie Jester is a Lecturer in Sociology and Criminology at the University of Gloucestershire, UK. Her research focuses on militarisation and security discourses in digital spaces. Common themes include gender, state identity, and secrecy/transparency.

Emma Dolan

Emma Dolan is Associate Professor in Peace and Development Studies at the University of Limerick's Politics and Public Administration Department. She works on topics of Feminist IR, political apologies, gender-based violence, and cultural militarisation.

Notes

1. Twitter has undergone rebranding and is now known as ‘X’. However, we use ‘Twitter’ in this article to reflect the branding at the time of data collection.

2. As Pelkmans and Machold (2011) explain, the process of labelling something – whether secret, not secret, transparent, or conspiracy theory – is not neutral and functions as evidence of the aforementioned political work of knowledge production.

3. Family also appears outside of the context of the crash (tweet 1 ‘It’s like one big 787 #Dreamliner hug. Congrats @united on completing your Dreamliner family!’).

4. NASA have also had catastrophic safety failures, e.g. the Challenger Disaster, where crew have been killed.