The Girl at the Wall

How the movie Eye in the Sky Tests the Principle on Foreseeable Harm to Innocents

The Girl at the Wall.pdf

Download PDF • 128KB

Spoiler Alert: This paper discusses the plot and ending of Eye in the Sky (2015) in detail

Introduction: The Hardest Case

Gavin Hood’s 2015 film Eye in the Sky is widely regarded as the most realistic depiction of modern drone warfare in popular cinema. Military lawyers and targeting specialists have praised its fidelity to the texture of real targeting decisions. It was the last major role of Alan Rickman, whose quiet, devastating performance as Lieutenant General Frank Benson anchors the film’s moral weight.

It is also a very good film. It raises all the right questions, and it was one of the things going around in my head as I tried to imagine a better legal framework for civilian protection - along with my early education in the law of armed conflict, Daniel Ellsberg’s searing account of targeting decisions in The Doomsday Machine, the real-world case of Mohammad Azam, and a decade of news from conflicts in which the technology to see civilians outran the legal obligation to protect them. The film, whether it intends to be or not, is the single best well-known test case in popular culture for the Principle on Foreseeable Harm to Innocents (FHI) - the framework that emerged from those influences.

Because the film is widely known and genuinely liked, it also offers something rarer: a way to make complex legal arguments about targeting law vivid and immediate for an audience beyond the specialist; hence this paper.

If FHI cannot withstand the scenario presented in Eye in the Sky, it is not a serious legal framework. If it can - and if it would have changed something that mattered - then it may earn its claim to operational realism; and, if adopted, will save lives.

This paper walks through the film’s plot against FHI’s decision architecture, step by step. It assumes the reader has seen the film, though the key facts are summarised where necessary. It concludes that FHI not only survives the test but reveals two things about the scenario that the film itself does not explore.

          *

1. The Scenario

A joint UK-US military operation is tracking senior Al-Shabaab leaders who have gathered in a safe house in the Eastleigh neighbourhood of Nairobi, Kenya. Colonel Katherine Powell (Helen Mirren) commands from the UK. Lieutenant General Frank Benson (Rickman) monitors from a COBRA-style briefing room alongside government ministers and military lawyers. US Air Force pilot Steve Watts (Aaron Paul) operates an MQ-9 Reaper drone from a ground control station in Nevada.

The mission begins as a capture operation. But when a micro-drone is flown inside the safe house, it reveals two operatives being fitted with suicide vests. The operation pivots from capture to kill. A Hellfire strike on the building is authorised in principle.

Then a cheerful girl of about nine - Alia - sets up her father’s bread stall against the compound wall, directly in the blast radius of the proposed strike. She is clearly visible on the Reaper’s full-motion video feed. Everyone in the chain of command can see her: the drone pilot in Nevada, the colonel in the UK, the general and ministers in the briefing room.

Everything that follows - eighty minutes of agonised deliberation involving ministers, lawyers, foreign secretaries, percentage estimates of probable civilian harm, and one undercover agent sent to buy bread — flows from a single irreducible fact: there is a child in the blast zone, and they know she is there.

          *

2. FHI’s Decision Architecture

FHI operates through a structured sequence of gates. Each must be passed before the next becomes relevant. The sequence cannot be skipped, short-circuited, or reversed. It looks like this:

Gate One: Foreseeability

The first question: is there an ascertainable civilian whose harm is foreseeable?

In Eye in the Sky, the answer is unambiguously yes. The Reaper has persistent surveillance. Alia is clearly visible, clearly a child, clearly a civilian. She is within the calculated blast radius. Her death or serious injury is foreseeable with near certainty. The foreseeability gate engages. Under FHI, a presumption against attack now arises.

Note the immediate contrast with the real-world case of Mohammad Azam, the taxi driver killed alongside Taliban leader Mullah Mansour in Pakistan in May 2016. In that case, the foreseeability gate was never engaged - not because Azam was invisible, but because the prevailing counting methodology classified all military-age males in a strike zone as combatants by default. The question “is there a civilian who might be harmed?” was never posed. In Eye in the Sky, it is the only question that matters.

Gate Two: Avoidability

The second question: can the harm be avoided? Can the target be struck at a different time, in a different place, using a different method? Can the civilian be moved?

This is where Eye in the Sky becomes genuinely difficult - and genuinely different from the Azam case. The film shows the operators actively engaging with avoidability:

Can they use a smaller weapon? Yes, and they explore this option. A GBU-12 is substituted for the Hellfire to reduce the blast radius. But Alia is too close to the compound wall. The physics cannot be negotiated away.

Can they change the aim point? Yes, and they spend considerable time on this. Powell orders her risk-assessment officer to recalculate the collateral damage estimate using a different strike point within the building, seeking a geometry that will bring Alia’s predicted casualty risk below the threshold. This is genuine precautionary effort - exactly the kind of operational engagement with civilian protection that the law of armed conflict demands and that real-world targeting too often omits. The film deserves real credit for depicting this process with such fidelity. But the recalculation, like the weapon substitution, cannot overcome proximity. Alia is at the wall.

Can the civilian be moved? They try. This is the most striking sequence in the film. Their undercover agent, Jama Farah, is sent to buy all of Alia’s bread. He does. She leaves. The strike window opens - and then her mother sends her back with more bread. Avoidability was momentarily achieved and then lost.

Can they wait for a different opportunity? This is the question the characters never really explore, and it is the question that matters most.

          *

3. The Third Option

The film presents its dilemma as binary: strike the house now, while the targets are inside, or lose them. Once the bombers leave, the argument goes, they will disperse into the crowded streets of Nairobi’s Eastleigh district and the opportunity to prevent a mass-casualty attack will be lost.

But this is a false binary. Consider what the operators actually know:

The safe house is in a residential compound. The targets arrived there somehow - almost certainly by vehicle, since Eastleigh is not within walking distance of Somalia. They are preparing to conduct a suicide attack on a crowded location, which means they need to reach that location - again, almost certainly by vehicle, since walking through Nairobi’s streets wearing bulky suicide vests would be conspicuous and dangerous for the operation.

The Reaper drone has persistent surveillance capability. It can track vehicles. It has been doing precisely this throughout the operation. The same drone that can watch a micro-drone fly through a window can follow a Toyota on a road.

Which means there is an alternative that nobody in the chain of command seriously considers: let the targets leave, follow them by drone, and strike the vehicles en route, on open road, away from Alia, away from the bread stall, away from the residential neighbourhood.

This is not a speculative or fanciful alternative. It is precisely what the United States did in the Mansour strike - tracked a vehicle for six hours across the Balochistan desert and struck it on an open highway. The operational capability is proven. The precedent exists. The only difference is that in the Mansour case, nobody bothered to spare the taxi driver, whereas in Eye in the Sky, the entire moral architecture of the film depends on the proposition that Alia cannot be spared.

There are counterarguments. What if the targets leave on foot? What if they separate? What if you lose them in traffic? These are real operational risks. But FHI asks whether avoidance is feasible - not whether it is guaranteed. A feasible alternative that carries some risk of failure is still a feasible alternative. The question is whether the certain death of a child is justified when a plausible alternative exists that would avoid her death entirely. Under FHI’s avoidability gate, the answer is no. The presumption against attack is not overcome until feasible alternatives have been genuinely exhausted - and following the targets’ departing vehicles is not merely feasible but routine.

This does not mean the strike could never happen. It means the strike should not happen at this moment, in this place, with this child against the wall. FHI does not prohibit the killing. It prohibits the killing when avoidance is feasible.

          *

4. The Second Hellfire

Even if we accept, for the sake of argument, that the first strike was justified - that the avoidability analysis was exhausted, that the emergency override applied, that every condition was met - the film contains a second decision that is indefensible under FHI. And it is this second decision, not the first, that kills Alia.

Here is what happens. The first Hellfire strikes the safe house. The building is largely destroyed. The suicide vests - the imminent mass-casualty threat that justified the override - are eliminated. Then, on the Reaper’s video feed, one of the targets is seen moving in the rubble. She is alive. A second Hellfire is authorised to finish her.

It is the blast from this second strike, combined with the cumulative effect of both explosions, that kills Alia. She is caught in the second blast, rushed to a field hospital, and dies on the table while her father watches.

Under FHI, the second strike is unlawful. And the analysis is straightforward.

The emergency override exists for one purpose: to permit an otherwise-prohibited attack when a specific mass-casualty threat against civilians is imminent and no alternative means of averting it exists. By the time the second Hellfire is authorised, the mass-casualty threat has been liquidated. The vests are destroyed. The safe house is rubble. The surviving target is wounded in the debris. She is now a conventional military objective - a wounded combatant, possibly an hors de combat - and her elimination, however desirable, is not the prevention of an imminent attack on civilians.

Let us walk through the override conditions for the second strike:

Condition 1: A specific mass-casualty attack is imminent. No. The vests are destroyed. The attack has been prevented. The surviving target is crawling in rubble, not deploying an explosive device. There is no imminent mass-casualty threat. Condition not met.

That single failure is dispositive. The override conditions are cumulative; all must be satisfied. But consider the others:

Condition 2: No feasible alternative exists. A wounded woman in rubble is not going anywhere. Ground forces could be deployed. She could be monitored. She could be captured. Condition not met.

Condition 3: Civilian harm is genuinely unavoidable. Alia is visibly present - now potentially injured - in the blast zone. A second strike will foreseeably harm her further. Her harm is entirely avoidable by not firing the second missile. Condition not met.

Condition 4: Intelligence is current and reliable. The target’s location is confirmed via real-time video. Condition met.

Condition 5: Authorised at the highest practicable level. The authorisation chain remains in place. Condition met.

Condition 6: Post-strike review automatically triggered. Procedural. Condition met.

Three of six conditions fail. The override cannot be invoked. Under FHI, the presumption against attack reasserts in full. There is a visibly injured child in the blast zone. The imminent threat has been eliminated. The second Hellfire is prohibited.

Had it not been fired, Alia would probably have survived.

          *

5. What the Film Gets Right

It would be a mistake to treat Eye in the Sky merely as a foil for FHI’s critique. The film gets several things profoundly right, and FHI is built on the foundations the film so vividly illustrates.

It gets distinction right. Everyone in the chain of command recognises Alia as a civilian. Her status is never in doubt. Nobody suggests she might be a combatant, or that her presence in the blast zone is evidence of affiliation with the targets. The contrast with the Mansour strike - where the taxi driver was simply designated “another combatant” - could not be starker.

It gets the moral weight right. The decision-makers agonise. Ministers squirm. The drone pilot weeps. Nobody treats the probable death of a child as a line item on a spreadsheet. The film takes the proportionality question seriously - painfully seriously - in a way that the real-world targeting process often does not.

It gets the politics right. The film’s most acidly observed scenes involve ministers and officials trying to pass responsibility up the chain, each hoping someone else will make the decision. This is not cynicism; it is an accurate depiction of how targeting authority actually works when political consequences are foreseeable.

It gets the human cost right. Alia dies. The film does not flinch from this. The final scene - Rickman’s General Benson buying a doll for his granddaughter, carrying the weight of what has been authorised - is one of the great performances of quiet grief in modern cinema. The audience is left not with resolution but with cost. Even when the system works, a child is dead.

These are not small achievements. They represent the moral seriousness that the law of armed conflict demands and that real-world drone operations too often lack.

          *

6. What the Characters Don’t Consider

The film’s characters do not lack moral seriousness. What they lack is operational imagination - and this is precisely the gap that FHI is designed to close.

The false binary

Eye in the Sky presents two options: strike now, or lose the targets. The characters never seriously consider a third: follow the targets when they leave and strike them away from the residential area. The drama depends on the proposition that the choice is between Alia’s death and a suicide bombing in a Nairobi shopping centre. But this is not the only choice. The same persistent surveillance capability that found the targets in the safe house can follow their vehicles through Nairobi’s streets and strike them on open road.

FHI’s avoidability gate forces decision-makers to ask this question before they reach for the proportionality calculus. The gate does not ask: “is this strike proportionate?” It asks the prior question: “is this strike necessary at this moment, in this place, with this child at the wall?” Only if the answer is yes - only if alternatives have been genuinely exhausted - does the proportionality question arise.

The unexamined second strike

In the film, the second Hellfire is treated as a continuation of the first decision. It is not. It is a separate targeting decision with different facts, different conditions, and different legal consequences. The threat that justified the first strike no longer exists. The override conditions no longer apply. A wounded child is now in the blast zone. Yet the second strike is authorised without the agonised deliberation that attended the first.

Under FHI, the second strike would have required its own independent assessment. The avoidability gate would have asked: can this wounded combatant be dealt with by other means? The foreseeability gate would have noted: there is a child, possibly already injured, who will foreseeably be harmed by a second explosion. The override would have been assessed - and would have failed, because the imminent threat had been eliminated.

This is not a marginal observation. It is the central operational difference FHI would have made. The first strike is arguable under either framework. The second strike is prohibited under FHI. And it is the second strike that kills Alia.

          *

7. The Two Films

There is a useful way to understand what FHI does, and it involves watching Eye in the Sky as though it were two separate films.

Film One is the story of an agonising decision in which every participant can see a civilian, recognises her as a civilian, and struggles with the moral and legal consequences of her probable death. The system works - slowly, painfully, imperfectly - but it works. The law of armed conflict engages. Distinction is applied. Proportionality is weighed. Precaution is taken. The decision is terrible, but it is a decision, made with eyes open and conscience burdened.

Film Two - the real-world story of Mohammad Azam - is the story of what happens when nobody asks the questions at all. The civilian is not identified; he is reclassified. Alternatives are not explored; the question never arises. The decision is not escalated; there is nothing to escalate, because the system sees no civilian in the car. Nobody agonises. Nobody weeps. A taxi driver is vapourised and described as “another combatant,” and the Pentagon tells the press there were “no reports of civilian casualties.”

FHI exists for both films. For Film One, it provides structure: the avoidability gate that forces consideration of alternatives before proportionality is reached, the override conditions that create a documented framework for the terrible decision, and the mandatory post-strike review that ensures accountability. It would not have prevented the first strike. It would have prevented the second.

For Film Two, it provides the legal architecture that was entirely absent: the foreseeability gate that requires recognition of civilians before they can be defined away, the avoidability analysis that demands consideration of alternative timing and methods, and the presumption that cannot be inverted by administrative fiat.

The gap between the two films is not a gap of technology. Both scenarios involve the same weapons systems, the same surveillance capabilities, the same capacity to see individual human beings from twenty thousand feet. The gap is a gap of legal architecture. In one scenario, the system requires decision-makers to look at the person in the blast zone and determine their status. In the other, the system pre-emptively resolves the question by definitional fiat.

          *

8. The Measure of a Legal Framework

The measure of a legal framework for armed conflict is not whether it produces comfortable answers. It is whether it forces the right questions at the right time. Eye in the Sky is the hardest test FHI can face - a scenario where the stakes are extreme, the intelligence is strong, the alternatives are limited, and the cost of inaction may be measured in dozens of civilian lives.

FHI passes the test. It does not produce a naively absolutist answer. It does not say: never strike when a civilian is present. It says: before you strike, ask whether the civilian’s death can be avoided. If it can - if you can follow the vehicles, if you can wait for a cleaner moment, if you can strike on open road - then you must avoid it. If it genuinely cannot - if the override conditions are met, if the cumulative requirements are satisfied, if the stakes are as extreme as the framework demands - then the terrible decision may proceed, documented, authorised, and subject to review.

But FHI also reveals what the scenario’s own logic does not fully explore. The false binary - strike now or lose the targets - is not an inevitable feature of the situation. It is a gap in the characters’ operational thinking, a failure to ask the avoidability question with sufficient rigour. And the second Hellfire - fired after the threat has been eliminated, against a wounded woman in rubble, with an injured child in the blast zone - is not, under FHI, an agonising moral dilemma. It is a prohibited act.

The girl at the wall deserved better. FHI would have given her a chance.