Foregrounding threats to disaster investigations with actor-network theory

ABSTRACT

To create a safer world, it is vital that lessons are learned from near-misses, incidents and accidents and that those lessons are, through active learning, translated into practical, safety-oriented actions. Not unnaturally, vested interests – fearful of the consequences of being associated with the event in question – may seek to delay, deflect or halt whatever investigation is called for or authorised. Drawing on actor-network theory (ANT), this paper uses a case study approach to explore the lengths to which those implicated in a near-miss, incident or accident will go to protect their perceived interests. Through the inductive analysis of five investigations, it is shown that disruptive tactics range from the mischievous, for example, manipulating an inquiry’s terms of reference, to the devious and illegal, for example, sanitising or manipulating statements. It is suggested that those charged with investigating near-misses, incidents and accidents use ANT to identify potentially hostile actors’ resourcing, reach, networks and likely tactics.

KEYWORDS:

• Investigations
• evidence
• interference
• manipulation
• fabrication
• remedies

Introduction

Near-misses, incidents and accidents are subject to formal investigations. Serious accidents may be subject to, for instance, judicial review and a Coroner’s Investigation. The 1979 Mount Erebus air disaster, 1989 Dryden air disaster and 1989 Hillsborough football stadium disaster were the subject of judge-led inquiries [1].

This paper presents an inductive study of five inquiries – into the 1943 Bari (Italy) chemical poisoning disaster, the 1958 Munich (West Germany) air disaster, the 1979 Mount Erebus (Antarctica) air disaster, the 1989 Dryden (Canada) air disaster and the 1989 Hillsborough (Sheffield, UK) stadium disaster – for the purpose of answering three questions:

1. What strategies might those implicated in a near-miss, incident or accident use to delay, deflect or halt an investigation?

2. What strategies might investigators use to counter efforts to delay, deflect or halt an investigation into a near-miss, incident or accident to ensure that right is done (fiat justitia)?

3. Is ANT useful for showing how, and in whose interests, power is exercised?

These disasters were selected first, because the investigations were subject to extraordinary pressures and/or interference and secondly, because the investigations hold lessons for those who examine contemporary near-misses, incidents and accidents.

Although the investigations were subject to broadly the same pressures, because the disasters occurred in different eras and in different circumstances, the investigative mechanisms differed. For example, responsibility for investigating the Bari chemical poisoning disaster rested with the military. Responsibility for investigating both the Mount Erebus and Dryden air disasters rested with the judiciary. Therefore, while the investigations were subject to similar pressures, those pressures were negotiated by actors with different powers and freedoms. The judges who led the Mount Erebus and Dryden Commissions of Inquiry were drawn from an independent judiciary. In contrast, the US Army doctor who investigated the Bari disaster worked for an organisation that had the power to dismiss the investigation’s findings.

The paper draws on actor-network theory (ANT) to determine how and to what degree the investigations were compromised by vested interests’ defensive tactics such as threats of legal action. Data are drawn from secondary sources; e.g. reports produced by the Hillsborough Independent Panel [1] and the Right Reverend James Jones, Bishop of Liverpool [2].

Literature review

Although there are studies both of the failure of disaster investigations to secure organisational or sectoral change [3–9] and of underperformance in disaster investigation [10–14], there are only a few studies of interference with disaster investigations.

The case study selection permits an examination of the more extreme forms of organisational defensiveness, including:

• the manipulation of the terms of reference of an investigation to favour one or more vested interests

• the seeding of an investigation team with persons professionally obligated to those likely to be investigated

• threats of, or actual court actions

• evidence-tampering

• espionage

• denigration

• character assassination, and

• intimidation.

The author, subscribing to Toft and Reynolds’s [15] view that historic events hold lessons for practitioners, anticipates that those responsible for conducting inquiries will, through processes of double-loop and active learning [15,16], benefit from the analysis.

Methodology and presentation

This paper presents a collective case study (see Stake [17] for a definition) with inferences drawn from five disasters. As noted by Goldthorpe [18], accounts differ in their level of abstraction. Denzin [19] suggests a positive relationship between triangulation and accuracy. Accounts were compared – triangulated – to create an accurate picture of each disaster.

The five disasters were interpreted through the prism of actor-network theory (ANT) [20–25]. Actor-network theory, described at Appendix 1, Table A1, reveals the complexities of purposive action. It recognises the agency of actors both tangible and intangible, e.g. mission statements, physical infrastructure, employees, tooling and organisational culture, maps the relationships between them and describes the strategies used by a network to colonise and dominate an activity or debate. Actor-network theory holds that, in Bennett’s [24] words: ‘[P]ower is exercised, and work performed, through networks of stories (intangibles) and things (tangibles)’ (Figure 1).

Figure 1. An actor-network is composed of stories (intangibles) and things (tangibles). The former shape the nature, environment and behaviour of the latter [24].

The case studies are presented in broadly the same format. Tables describe the actor-networks that influenced the investigations. Actors, tangible and intangible, are documented for each actor-network. For example, the Allies’ mustard gas actor-network is described in terms of President Roosevelt and Prime Minister Churchill (tangible actors) and the Grand Alliance’s fear that the Allies’ stockpiling of mustard gas might provoke Hitler into launching a chemical warfare attack on Allied formations [26].

A critical reflection on the methodology

Finnegan [27] warns against selectivity and encourages the researcher to identify biases reflexively, both personal and in the sources themselves, to the reader when drawing on secondary sources. The author acknowledges selectivity in his choice of historical accounts. For example, regarding the Bari chemical poisoning incident, data are drawn from sources critical of the authorities’ management of the disaster, for example, Conant [26,28]. Similarly, regarding the Munich Air Disaster, data are drawn from sources critical of the British and West German governments’ investigation – e.g. Bennett [14].

As with any research, there are questions about the generalisability of the findings and utility of the recommendations [19,29,30]. Mindful of Toft and Reynolds’s [15] argument that past events hold lessons for practitioners, the author believes his analysis to be of practical use.

Case studies

The 1943 Bari, Italy, chemical poisoning disaster

From 1943, the Grand Alliance’s fortunes were in the ascendancy. In September 1943, British forces liberated the mediaeval port of Bari on the Adriatic Sea. On 2 December 1943, the Luftwaffe attacked the port. Vessels hit included the Liberty ship SS John Harvey. There followed a spate of deaths amongst service personnel stationed in Bari [26,28,31,32]. The deaths were investigated by physician and chemical warfare expert Lt. Col. S.F. Alexander, a senior member of the US Army Chemical Warfare Service’s (CWS’s) Medical Division [28]. Alexander found servicemen presenting with symptoms untypical of those who had been blown off ships: ‘Instead of being restless or anxious, they were apathetic … their extremities were warm rather than cold … Patients [complained loudly] about the intense heat, [tore] off their clothing and, in their frenzy, [tried] to rip off their bandages’ [28]. Patients complained of nausea. They vomited. Their skin blistered. A nurse recalled: ‘We began to realise that most of our patients had been contaminated by something beyond all imagination’ (Rees cited in Conant [28]). Through observation and interview, Alexander concluded that many servicemen had suffered chemical burns, possibly from mustard gas. Alexander confronted the British authorities who denied the presence of mustard gas at the port. Alexander, aware that the Allies were stockpiling chemical warfare agents in-theatre, began testing tissue samples and seawater. During a survey of the harbour floor a diver discovered an M47A2 mustard gas bomb, one of 2,000 transported to Bari by the SS John Harvey [32–34].

Alexander informed both Roosevelt and Churchill of his discovery. Roosevelt asked to be kept informed. Churchill denied the presence of chemical warfare ordnance at Bari, and suggested Alexander re-examine his patients.

When Alexander presented scientific evidence that personnel had been poisoned by mustard gas [34], the military authorities intervened. ‘Any mention of mustard gas was stricken from the official record … Alexander’s diagnosis of toxic exposure was deleted’ explains Conant [28]. Patients’ charts were altered and their records rewritten.

The Allies’ mustard gas actor-network

To support its stockpiling strategy, the Grand Alliance invested in a mustard gas actor-network, described at Appendix 2, Table A2, that managed the politics of creating a chemical warfare stockpile (stockpiling was politically problematic because the use of such weapons had been outlawed) and ensured the stockpile remained secret (the Allies feared a German pre-emptive strike should there be a security breach).

Alexander practised his medicine in an environment dominated by the Allies’ mustard gas actor-network. He resisted being translated by the network, finding evidence of mustard gas contamination and stockpiling. Nevertheless, the mustard gas actor network’s narrative – that there were no chemical weapons in Italy – prevailed in the short term. The network’s success in advancing this narrative reflected its success in translating, aligning and mobilising actants – with the exception of Alexander himself. Conant [28] claims that the mustard gas actor-network’s tunnel vision (sustained by Roosevelt and Churchill) increased the death toll: ‘By failing to alert hospital staff to the risk of contamination [the military authorities] had greatly added to the number of fatalities’. Almost one thousand Bari residents died [32].

The 1958 Munich air disaster

On 6 February 1958, British European Airways (BEA) Flight 609 crashed in Munich while attempting a departure in poor weather. Twenty-three were killed and nineteen injured. The dead included one of the pilots. The German authorities attributed the disaster to wing ice. Despite harbouring doubts about the quality of the West German investigation [35], the British government, seeking a rapprochement with Germany [36], accepted this finding. Flight 609’s captain, James Thain, was blamed. Blaming Thain provided for the better management of the politics of the Munich Air Disaster [14,37]. Blamism and fundamental attribution – the individuation of responsibility for a near-miss, incident or accident that is systemic in origin – are common post-disaster responses [13,14,38–41]. Morrin (cited in Lee [42]) observes: ‘He [Thain] was hung out to dry … ’.

Thain asked the British and West German authorities to consider the possibility that his aircraft’s performance had been affected by runway contamination, specifically snow, slush and standing water. They refused [35], investing instead in the wing-ice actor-network.

The wing-ice actor-network

The wing-ice actor-network, described at Appendix 3, Table A3, a product of the British and West German governments’ blamism, fundamental attribution and politicking, promoted the wing-ice hypothesis and maintained harmonious relations between the two countries. Like the mustard gas network, the wing-ice network drew its weight from political and institutional actants [14].

The runway contamination actor-network

Thain refused to accept he was to blame for the disaster. With support from his trade union, he invested in a competing actor-network, described at Appendix 4, Table A4, to propagate the narrative that his aircraft’s performance had been compromised by runway contamination. Disagreements produce contradictory narratives that find expression in actor networks that compete for attention and influence.

Outcomes

As evidenced by Prime Minister Harold Wilson’s after-dinner speech [43], by 1967, Thain’s runway contamination actor-network had translated and aligned the Prime Minister. Judging that calls for a new investigation could no longer be resisted, the British government organised a new inquiry. It exonerated Thain. The Minister of State at the Board of Trade expressed admiration for the wronged captain:

The cause of the accident was slush on the runway. It is possible, but unlikely, that wing icing was a contributory cause … Her Majesty’s Government accept this finding and the government of the Federal German Republic have been so informed … I am sure that the House would wish me to congratulate Captain Thain on the successful outcome to his long campaign

(Rodgers cited in Hansard [44]). In 1973, Sir William Hildred said this of the West German investigation:

[There was] a readiness to draw conclusions with insufficient data, a good deal of deliberate lying, suppressio veri [a misrepresentation of the truth by the omission or suppression of pertinent facts] and suggestio falsi [an affirmative misrepresentation – by words, conduct or artifice – as distinguished from a mere suppression of the truth] … a holding back of witnesses and an astonishing ignorance of the retarding effects of slush on the runway [45].

The 1979 Mount Erebus air disaster

On 28 November 1979, an Air New Zealand DC10 collided with Antarctica’s Mount Erebus in a controlled-flight-into-terrain (CFIT) incident, killing all on board. Two inquiries were held. The first, led by New Zealand’s Chief Inspector of Air Accidents (CIAA), found the pilots culpable. The second, a Royal Commission of Inquiry (CoI) led by a High Court judge, found the airline culpable.

According to the CIAA, the pilots were culpable because:

• they had breached the 16,000 feet minimum safe altitude (MSA) rule. The pilots, aware that the service had been advertised as a sight-seeing flight, had (with the permission of local air traffic control) taken the DC10 down to 1,500 feet

• they had failed to confirm their position before descending below the MSA.

According to High Court judge Justice Peter Mahon, Air New Zealand was culpable because:

• prior to departure the airline had changed the aircraft’s routing over Antarctica without informing the pilots. Previous sight-seeing flights had been routed over McMurdo Sound. Air New Zealand routed the accident aircraft over land, the DC10’s automatics having been programmed to fly at Mount Erebus. ‘[T]he single dominant … cause of the disaster was the mistake made by those airline officials who programmed the aircraft to fly directly at Mt. Erebus and omitted to tell the aircrew’, observed Mahon [46]

• the airline’s procedures were not fit for purpose. Mahon [46] attributed the risk-laden routing and failure to brief the flight-crew to ‘incompetent administrative airline procedures’.

For Mahon, the loss was an organisational accident – an adverse event rooted in systemic failings (see Reason [47] for a definition) that included:

• Air New Zealand’s tolerance of flight crews’ violation of the airline’s MSA rule. '[L]ow-level sightseeing was a matter of public knowledge, and had not been challenged by either the operator or the [regulatory] authority … [T]his acquiescence became an error-inducing condition; it would perforce have influenced the accident crew’s belief that they were acting in accordance with acceptable precedent', observed Maurino et al [48].

• Air New Zealand’s suspension of the requirement that the pilot-in-command of Antarctic sight-seeing flights should have experience of Antarctic flying. The accident crew had no experience of Antarctic flying [48].

Mahon rounded on the airline at the conclusion of his investigation, accusing it of falsifying evidence. He observed: ‘The palpably false sections of evidence which I heard could not have been the result of mistake or faulty recollection. They originated … in a pre-determined plan of deception’ [46]. Having so charged the airline, Mahon closed the CoI.

Air New Zealand (ANZ), aggrieved that Justice Mahon had not invited it to respond to his allegations, successfully appealed certain of the CoI’s findings [49]. Justice Mahon sought to restore his reputation through the Privy Council. The Privy Council judged that Mahon, by denying the airline the right of reply, had violated the principle of natural justice. Reviewing the Privy Council’s opinion, The Dominion Post [50] ventured: ‘[This was] about the harshest thing that could be said about a judge’.

Air New Zealand (ANZ), in appealing the CoI’s findings, were buoyed by Prime Minister Sir Robert Muldoon’s public criticism of Justice Mahon: ‘Muldoon, a friend of [ANZ’s Chief Executive Officer], savaged Justice Mahon and his report’ [50]. According to The New Zealand Herald [51]: ‘Muldoon responded with venom – the findings were potentially fatal to the Government-owned carrier’. Muldoon refused to table Mahon’s report. Possible reasons for Muldoon’s reaction include:

• his disapproval of Mahon’s conduct of the CoI

• his rejection of the CoI’s findings

• his friendship with ANZ’s CEO

• a desire to protect the nation’s flag carrier

• a desire to safeguard his government’s investment in ANZ – the New Zealand government owned 100% of the airline [52].

Muldoon’s anti-COI actor-network

Seeking to establish the CIAA’s report as the definitive account of the disaster, the Muldoon government invested in an anti-CoI actor-network, described at Appendix 5, Table A5, that translated, aligned and mobilised actors sympathetic to the government’s position. The Muldoon government’s network included, if a report in The New Zealand Herald [51] is accurate, the airline itself. As evidenced by the Privy Council’s rebuff to Mahon, the judge’s efforts to defend his reputation failed. Mahon’s treatment suggests that speaking truth to power – even if one holds the authority of the judiciary – is not without risk. Right (fiat justitia) may not be done in politically-charged environments.

The 1989 Dryden air disaster

On 10 March 1989, Air Ontario’s Captain George Morwood and First Officer Keith Mills were rostered to fly a four-sector day, shuttling a Fokker 28 jetliner between Winnipeg and Thunder Bay. The pilots were unfamiliar with the F28. Morwood had 62 hours on-type and Mills 66 hours. They operated in icing conditions.

The first two sectors included an intermediate stop at Dryden. The F28’s auxiliary power unit (APU) had been out of commission for five days. An APU provides an aircraft power when parked. Without a functioning APU, the F28 was reliant on ground power units (GPUs). Unfortunately, Dryden had no GPUs. Without a functioning APU and with no GPUs, one of the F28’s engines would have to remain powered-up when the aircraft was at Dryden. Air Ontario pilots were not allowed to de-ice with an engine running.

The day’s first Winnipeg to Thunder Bay sector, with an intermediate stop at Dryden, went reasonably well. But, the Thunder Bay turnaround was problematic. Issues included:

• the miscalculation of passenger numbers

• falling behind schedule

• complaints from passengers.

At Thunder Bay, Morwood requested he be allowed to offload passengers rather than fuel. Air Ontario refused his request. Defueling the F28 put the service further behind schedule. Crew morale slumped: ‘Evidence from eyewitnesses disclosed that these events changed the good-spirited mood which the flight crew had showed (sic) earlier’ [48].

By the time the aircraft commenced the Thunder Bay to Winnipeg leg, with an intermediate stop at Dryden, Morwood’s crew faced several challenges:

• the possibility of icing conditions at Dryden

• schedule slip – the F28 departed Thunder Bay circa one hour behind schedule

• frustrated passengers

• operating in challenging conditions with limited experience on-type

• following Air Ontario’s refusal to grant Captain Morwood permission to offload passengers, the crew’s realisation that it was the junior partner

• Captain Morwood’s elevated stress level [53]

• Morwood's limited experience of flying in the Canadian north with its numerous challenges [53].

As the stressors and threats multiplied, safety margins were squeezed. At Dryden, the crew decided to keep an engine running during turnaround, which precluded de-icing (see explanation above). At just after mid-day, the F28, by now seventy minutes behind schedule, commenced its take-off run. The F28 failed to gain altitude and crashed, killing twenty-four. The accident had immediate and proximate causes (see Appendix 6, Table A6).

Although Captain Morwood must bear responsibility for the acts of commission (such as agreeing to fly an aircraft with an unserviceable APU into an airport with no GPUs in icing conditions) and omission (such as failing to perform a walk-around) that were the immediate cause of the accident, failings in the operating environment contributed to the accident (see Appendix 6, Table A6). Dryden was an organisational accident.

The government appointed a Commission of Inquiry (CoI) [54] in preference to having the accident investigated by the Canadian Air Safety Board (CASB). Despite, or perhaps because of, its pioneering ‘organisational’ approach to accident investigation [48], the CoI was opposed by Canada’s civil aviation establishment which, to protect its interests, organised an actor-network-of-reaction (see Appendix 7, Table A7).

Moshansky, determined that his CoI should be neither hamstrung nor corrupted, successfully challenged the Canadian aviation system’s actor-network-of-reaction. He recalled: 'During the early stages of the Inquiry, counsel for the regulator [Transport Canada] attempted to limit the scope of the Inquiry with threats to limit my mandate by seeking an order in the Federal Court of Canada. When it became clear that intimidation would not succeed, these attempts were abandoned …' (Moshansky cited in Maurino et al [48]). Moshansky dealt decisively with a seconded CASB investigator who suggested it was acceptable to doctor witness statements. He recalled: 'I was taken aback by the counsel of a senior CASB member suggesting that I undoubtedly would wish to massage investigator reports and witness interviews, which he made clear was CASB practice. The term ”massage” in this context meant to me manipulation of accident investigator findings to fit an agenda, which later events suggested was not to unduly disturb the aviation industry’s comfort with its own performance … [C]oming from a judicial background where one seeks to ascertain truth rather than subvert it, I was appalled by this advice and dismissed it out of hand' [55].

Moshansky’s investigation can be characterised as a struggle between progressives and reactionaries, the latter producing, through shared interests, an actor-network-of-reaction (see Appendix 7, Table A7) whose purpose was to neuter the CoI. Moshansky’s indefatigable efforts earned him the Order of Canada [56]. Judge Moshansky, it transpired, was an accomplished politician.

The 1989 Hillsborough football stadium disaster

At an association football match in Sheffield, England in 1989, ninety-six supporters were killed in a crush on one of the terraces. Hillsborough was an organisational accident with immediate and proximate causes [57]. One of the immediate causes was the decision by South Yorkshire Police (SYP) to open a gate to relieve pressure outside the ground. Proximate causes included the design of the terraces and a fence that prevented supporters from finding refuge on the pitch. The disaster polarised opinion. Some saw it as an indictment of the squalid conditions that obtained at grounds. Others considered it a manifestation of the phenomenon of football hooliganism. In the 1980s, hooliganism at football matches was a major issue [57–59]. There was a perception that crowd behaviour was poor and that remedial measures (such as membership schemes and perimeter fences) were needed. The then Prime Minister, Margaret Thatcher, ‘was determined to eradicate football hooliganism’ [57]. Those who attributed the Hillsborough disaster to Liverpool supporters’ misbehaviour could reference the Heysel (Brussels, Belgium) stadium disaster that saw fourteen Liverpool supporters convicted of manslaughter [60].

The disaster was investigated against a backdrop of polarised opinion, high emotion, intense media interest and prejudice. Tabloid newspaper The Sun accused fans of pickpocketing the dead [57]. Melnick [58] notes how newspapers ‘ … in their almost hysterical zeal to “get the story out to the public” … frequently trample over the facts’. To defend its officers and reputation, SYP created an actor-network that advanced the narrative that Liverpool fans’ misbehaviour lay at the heart of the disaster (see Appendix 8, Table A8).

Investigation

There were several investigations into the disaster. The first, a 1989 public inquiry chaired by Lord Justice Taylor, held South Yorkshire Police (SYP) responsible for the disaster. In his report, Taylor criticised SYP for scapegoating fans. The most consequential investigation was that conducted by the Hillsborough Independent Panel (HIP). The HIP report, published in 2012, claimed that 164 police accounts had been altered and that numerous negative comments about the policing operation had been erased. Following publication of the HIP report, South Yorkshire’s Chief Constable admitted that, in 1989, ‘disgraceful lies were told which blamed the Liverpool fans for the disaster’ (Crompton cited in Jones [2]).

In 2014, the Independent Police Complaints Commission (IPCC) said it believed that more police accounts than originally thought had been altered, and that fans’ statements may have been doctored (manipulated through alteration).

Closure?

Following the watershed 2012 HIP report, some politicians drew attention to the negative consequences of institutional defensiveness and mendacity. Former UK Prime Minister Gordon Brown said:

No one should ever have to go through what Hillsborough families have had to live through … . No one should be kept in the dark by bureaucratic indifference and deceitful lies [author’s emphasis]

(Brown cited in Brown [61]). Former UK Prime Minister Theresa May said: ‘I have been struck in the case of Hillsborough, but in other cases too, at the way in which the state in its various forms acts to defend itself from blame. The very bodies that we expect to protect and support the public [for example, the police], seek instead to protect themselves, and this defensive attitude means that families are all too often denied access to the truth, and with that, often denied access to justice [my emphasis]’ (May cited in LiverpoolWorld [62]). Mayor of Greater Manchester Andrew Burnham said: ‘When things go wrong, the authorities close ranks. They blame victims. They sometimes create false narratives that can be very hard to shift [my emphasis]’ (Burnham cited in Brown [61]). The British state’s efforts to redeem itself did not include, however, a legal duty of candour on public bodies ‘ … to tell the truth and proactively co-operate with official investigations and inquiries’ [63].

Conclusions

The case studies demonstrate that to preserve reputations and privileges, parties may manipulate the terms of reference of an inquiry, massage or destroy evidence, manipulate the press, denigrate investigators, refuse to co-operate and lie. Consider, for example, how Churchill denied the presence of chemical warfare ordnance at Bari. Consider how West Germany’s Traffic and Transport Ministry refused to entertain the possibility that runway contamination contributed to the Munich Air Disaster. Regarding Judge Moshansky’s investigation, consider:

• how Transport Canada (Transports Canada) attempted to limit the CoI’s scope

• how Canada’s Privy Council attempted to institutionalise the CoI by locating it within the Ministry of Transport

• how seconded CASB personnel suggested Moshansky massage evidence to cast Canada’s air transportation sector in a more favourable light

• how Transport Canada (Transports Canada) initially refused to co-operate with CoI investigators.

Consider how SYP amended officers’ statements. Consider, also, at least the possibility that fans’ statements were massaged.

Actor-network theory provides investigators with a means of conceptualising how vested interests might organise to achieve an outcome convenient to themselves. As documented in the Appendices, following each of the five disasters, vested interests created actor-networks for the purpose of protecting their reputations and those of their associates. It is axiomatic that understanding how an opposing force has organised itself is key to developing effective countermeasures. Actor-network theory can reveal how vested interests might, in defence of reputations, privileges, wealth and the status quo, organise against an investigation. It reveals the resources – human and non-human – that vested interests might marshal against inquiry. Investigators can only secure the truth by identifying and immobilising these resources. The sine qua non for identifying and countering opposition is mindfulness [64,65]. Investigators must be able to imagine who might obstruct them, and how and when that obstruction will manifest. Once these entities are known, countermeasures can be taken.

Disclosure statement

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

Appendix 1.

Actor-network theory

Actor-network theory documents the origins and functioning of purposive networks, for example, a system of law-enforcement or munitions manufacture. Actor-networks (hybrid-collectifs) reify intentions and produce.

Actor-networks:

• Are composed of technical and social elements. For example, machine-tools and corporate culture

• Are dynamic, responding to new challenges in the wider environment

• Compete with other actor-networks for resources and power. During the 1960s, NASA’s man-on-the-moon actor-network competed with United States President Lyndon Johnson’s War on Poverty actor-network for government funds [66].

The nature and dynamics of actor-networks are described using a lexicon specific to ANT (Table A1).

Table A1. ANT concepts.

Term	Meaning
Actor (‘actant’ in Latour’s actor-network argot)	An actor (actant) can be, in nature, tangible – for example, a worker, manager, pit-prop, conveyor belt, machine-tool, factory building, delivery truck or checklist – or intangible – for example, an organisational culture, corporate vision, product proposition, cronyism, religious belief, lie or Zeitgeist. Actors, tangible and intangible, act on other actors, and on the environment, in furtherance of the network’s interests.
Actor-network (‘hybrid-collectif’ in Latour’s actor-network argot)	An actor-network, consisting of actors, tangible and intangible, is a purposive social artefact. Examples include airframers (aircraft) and police forces (security).
Principle of Generalised Symmetry	The Principle of Generalised Symmetry holds that actors, whether tangible or intangible, are invested with agency. In the same way that a shop-floor worker, machine-tool or newspaper article has agency, a prejudice, aspiration or ideology has agency.
Translation	Translation is the process whereby actors seek to recruit – translate – other actors. Networks accrue actors. Translation proceeds through four stages: Problematisation (the problem to be solved is defined); Interessement (prime movers recruit actors to the cause); Enrolment (roles are defined); Mobilisation (actors are primed for action).
Alignment	Alignment addresses actors’ complementarity. The greater this is, the more aligned are the actors. Alignment and network impact and resilience are positively related.

The application of ANT

Actor-network theory has been used to deconstruct social activities and pathogens, from the production of scientific knowledge [20] to corruption [14,25]. Latour and Woolgar [20] used ANT to explain how scientific knowledge is produced, Law [67] used it to describe the genesis of the British Aircraft Corporation’s (BAC's) 1960s strike aircraft, the TSR2, and Bennett [14,25] used it to describe the nature, extent, functioning and consequences of corruption within the New York Police Department (NYPD) in the 1960s and 1970s [68].

Appendix 2.

The Allies’ mustard gas actor network

Table A2. Key actant-components of the Allies’ mustard gas actor-network.

Actants
Tangible	Intangible
President Roosevelt	A determination that if the Axis powers used chemical weapons, the Allies would respond in kind
Prime Minister Churchill	A fear that if Hitler learned that the Allies were stockpiling chemical weapons on Italian soil, he might order a pre-emptive strike against Allied formations
The Grand Alliance’s military apparatus	A fear that chemical weapons could disrupt, or even defeat, a strategic manoeuvre, such as 1944’s Operation Overlord that saw hundreds of thousands of soldiers delivered to the beaches of Normandy in an effort to open a second front, thereby appeasing a restive Marshal Stalin
The M47A2 mustard gas bomb	A culture of secrecy and denial in politics and the military
The Foggia chemical warfare stockpile
Denial, misinformation and disinformation. Such tactics were employed by the Allies to keep the stockpiling operation secret (regardless of the death and suffering such tactics might cause among troops and residents)

Appendix 3.

The wing-ice actor-network

Table A3. Key actant-components of the wing-ice actor-network.

Actants
Tangible	Intangible
The West German government	The post-war Zeitgeist of bridge-building with Germany [36]
The British Government	The prioritising of geopolitical interests over the interests of individual citizens, whatever the consequences for those citizens
The West German Traffic and Transport Ministry	The social pathogens of blamism and fundamental attribution. Cook [40] notes ‘[a] social [and] cultural need to blame specific, localised forces or events for outcomes’
An act of omission, specifically Thain’s failure to check for wing-ice prior to take-off
A photograph taken before the third take-off run that showed what appeared to be snow on the upper-surfaces of the aircraft’s wings
Wing-ice. Hours after the crash, the West German Traffic and Transport Ministry found ice on the wing upper-surfaces

Appendix 4.

The runway contamination actor-network

Table A4. Key actor-components of Captain James Thain’s runway contamination actor-network.

Actors
Tangible	Intangible
The British Prime Minister. Harold Wilson pointedly remarked at a Manchester United FC dinner: ‘The one thing I regret is that the captain of that aircraft has been unjustly treated’ (Wilson cited in Sunday Mirror [43])	Thain’s belief that he and his deceased colleague were blameless
The Royal Aircraft Establishment (RAE). The RAE researched the effect of runway contamination on aircraft runway performance	Thain’s belief that objective scientific data, dispassionately presented, would exonerate him and his colleague
The British Airline Pilots’ Association (BALPA)	Thain’s desire to protect his family (his daughter was bullied at school)

Appendix 5.

The anti-CoI actor-network

Table A5. Key actor-components of New Zealand Prime Minister Robert Muldoon’s anti-CoI actor-network.

Actants
Tangible	Intangible
The Prime Minister’s advisory group on the CoI	A desire to see the nation’s flag-carrier prosper
Air New Zealand (ANZ). According to The New Zealand Herald [51], ANZ participated in the Prime Minister’s advisory group – essentially a political formation – despite the airline being called to give evidence	Respect for the CIAA’s findings amongst the pilots who reviewed the CoI report
The Prime Minister’s friendship with ANZ’s CEO	Since the dead cannot speak, the knowledge that a reductionist analysis that blamed the pilots was incontestable by those most intimately involved. Fundamental attribution would, in this case, be low-risk. The New Zealand Herald [51] insightfully observed: ‘[The] accident report … blamed pilot error – far less damaging to the Government than systems-failure at the state-owned airline’
Leaks to the advisory group from within the CoI
The CIAA’s report (that found the pilots culpable)

Appendix 6.

Causes of the Dryden accident

Table A6. The immediate and proximate causes of the 1989 Air Ontario Dryden (Canada) accident.

Immediate and proximate causes
Immediate	Proximate
Snow/ice on the F28’s wings	Air Ontario’s limited experience of jet operations
Runway contamination	Tolerance of open maintenance items (for example, the Dryden aircraft’s APU)
Flight crew stress	Inadequate training (for example, in respect of runway contamination and crew resource management)
Failure to perform a walk-around	The Dryden crew’s limited experience on the F28
Poor communication: the Dryden Dispatcher failed to inform the crew of the snow/ice accumulation on the aircraft's wings	A dearth of active foresight. The airline required Captain Morwood to fly an aircraft with an unserviceable APU into an airport with no GPUs in icing conditions
	A national regulatory authority, Transport Canada (Transports Canada), that lacked the resources to discharge its statutory duties satisfactorily [48,53]

Appendix 7.

The Canadian aviation establishment’s actor-network

Table A7. Key actor-components of the Canadian aviation establishment’s actor-network-of-reaction.

Actants
Tangible	Intangible
The Privy Council. The Privy Council demanded that the CoI report to Canada’s Minister of Transport. (Fearing a conflict of interest, Moshansky convinced the Federal Government that the CoI should report to the independent Governor General in Council)	The Canadian aviation system’s desire to neutralise ‘outsider threats’ and maintain the status quo [55]
Transport Canada (Transports Canada). The regulator attempted, in Moshansky’s [55] words: ‘[T]o confine the Dryden inquiry to the facts of the accident itself, as opposed to a system-wide investigation’	A belief – not uncommon amongst investigators – that reductionist analyses provide adequate explanation for near-misses, incidents and accidents
The CASB. Seconded CASB investigators tried to convince Moshansky that it was acceptable to massage data and evidence
Air Ontario. The airline sought to prevent the naming of individuals in the CoI report (the airline was defeated in Federal Court)
The Canadian Air Line Pilots Association. The Association sought to prevent five Air Ontario pilots being called to testify (the pilots were called)
Transport Canada (Transports Canada). The regulator obfuscated and delayed. (Transport Canada co-operated after being threatened with subpoenas)

Appendix 8.

The SYP actor-network

Table A8. Key actor-components of the SYP’s Fans-were-to-blame actor-network.

Actants
Tangible	Intangible
SYP	A sub-culture within SYP that tolerated the doctoring of police and witness statements
Whites Press Agency. Whites put out unevidenced stories from SYP officers complicit in the cover-up	Negative perceptions within government of football supporters
SYP Police Federation Branch. (The Police Federation is a staff association that represents the interests of police officers to employers and other parties)	Negative perceptions of Liverpool supporters post Heysel
Irvine Patnick MP. Local Conservative MP Irvine Patnick made derogatory comments about Liverpool supporters	Negative perceptions of the City of Liverpool. The July, 1981, Toxteth riots lasted eight days. The riots saw seventy buildings damaged and 460 police officers injured
The Sun. The tabloid newspaper printed unevidenced allegations against Liverpool supporters	Perceptions of the City of Liverpool as a repository of superannuated rust-belt industries unworthy of major government investment
A number of tabloid and broadsheet newspapers, for example, the Sheffield Star and Daily Telegraph, that reproduced unevidenced allegations against Liverpool supporters [57]	The authoritarian character of the government of the day
	The pro-police stance of the government of the day [57]