Reproducing Violence and Oppression through Law: An Analysis of the Trial Judgment in Kalma v African Minerals Ltd


October 13, 2021

The transnational pursuit of redress for corporate human rights violations in Africa has been partly premised on Western courts providing such redress. Their perceived failures to do so are thus susceptible to being understood as judicial and legal inaction, as may be observed in early responses to the UK Court of Appeal decision last year in Kalma v African Minerals Ltd (Kalma).[1] That decision unanimously upheld the judgment of Mr Justice Turner of the High Court, which dismissed a civil action brought by 142 Sierra Leonean claimants for human rights violations in the vicinity of the UK-domiciled defendant’s iron ore mine in Sierra Leone.[2]

Yet these cases may also be understood as courts and law actively working to reproduce the conditions that sustain violence and oppression of the kind involved. To this end, I provide a reading of Turner J’s judgment which seeks to draw out that work and the reasoning on which it is founded. While the Court of Appeal judgment merits similar analysis, focusing on the trial judgment permits fuller consideration of how the case’s ‘factual’ basis was judicially constructed. As one of the few Western court decisions on the merits for alleged corporate human rights violations in the global South, Kalma affords particular insight in this regard.

The reading developed here helps to show how cases like Kalma may result not just in failures to provide justice, but in the reaffirmation of, and development on, rationalities through which oppression and violence have long been secured. I conclude by reflecting on the implications of this understanding for delocalised justice.

Trial Judgment Findings

Kalma centred on two incidents, recounted here by reference to Turner J’s findings. The first, in 2010, followed local residents’ efforts to prevent African Minerals’ (AML) activities on their land, which the company was considering for a dam; the second, in 2012, followed strike action by local workers. During these incidents, Sierra Leonean police officers ‘overreacted’ as local residents, some of whom had been forced to relocate for the mine, were ‘beaten, shot, gassed, robbed, sexually assaulted, squalidly incarcerated’, and, in the case of one young woman, killed (para. 4).[3]

Turner J rejected evidence of torts committed directly by AML personnel, but made several findings of relevance to AML’s liability for acts of the police, pleaded by the claimants under various grounds including accessory and vicarious liability, and negligence. Before and after the first incident, AML considered how to deal with local residents deemed to be in its way. Internal emails from management read:

“The central Government should be asked to put in measures or legislature (sic.) to fast-track the relocation/resettlement process - sort of Stalin like thing (not quite)….” (Para. 107.)

“If these guys don’t stop mucking with my people, I’ll turn their water supply off!!!! Please delete this email.” (Para. 179.)

AML began making payments to the police before the 2010 incident (para. 97); after it, AML increased its support for them (para. 178). During the two incidents, AML provided the police with a site at which to hold detainees, accommodation, drivers, vehicles, and other support, while continuing payments to the police without which their services would have been ‘considerably less robust’, possibly non-existent (paras. 196 and 293). AML made its payments, which were unlawful and referred to internally as ‘incentives’, in cash to maintain control over which officers received them (paras. 284, 286, and 296).

Turner J held that AML’s relationship with the police was ‘far removed from what would be considered appropriate in England and Wales’ (para. 331), but did not, in Sierra Leone, give rise to liability. As I will show, this outcome required significant productive work beyond doctrinal interpretation, including through imposing an ordering narrative; producing (un)reliability; reframing resistance as lawlessness; (dis)locating the source of violence; reifying a public/private divide; and projecting a particular conception of the ‘state’.

This productive work, moreover, relied on logics identified in critical scholarship as underpinning law’s relationship to violence and oppression. These logics reflect and reinforce notions of improvement, mediated by race, that have served to identify those worthy of legal protection by their supposedly advanced, as opposed to ‘primitive’, ways of living and relating to land; racialised tropes about non-Europeans as uncivilised, anarchic and violent; gendered tropes about women as overly emotional and irrational; the distinction between public and private domains and the universalisation of Eurocentric conceptions of the state; and associations of law with order and resistance with deviance, whereby violent repression becomes both justified and inevitable.

Imposing an Ordering Narrative

Turner J opens his judgment by situating the case within a broader narrative frame:

"The deeply unhappy events with which this judgment is concerned unfolded in Tonkolili, a remote and inaccessible district in the north of Sierra Leone in West Africa. For centuries, generation upon generation of local villagers had lived and worked on the land. Most of them depended for their livelihoods upon small scale local trading or the subsistence farming of rice and other crops. Then, ten years ago, beneath the lands which they had farmed so modestly over countless generations was discovered the largest iron ore deposit in Africa." (Para. 1.)

From the outset, then, notions of improvement are invoked, shaping understandings as to what the case is about (an encounter with modernity giving rise, as later held, to local benefits and ‘inevitable’ conflicts (paras. 3 and 279)), what it is not about (violence perpetrated so that a company can pursue its project unimpeded), and who will prove worthy of legal protection. Within this frame, the claimants live in an undeveloped state as they have ‘for countless generations’, their ‘modest’ farming contrasted in the next paragraph with AML’s ‘massive infrastructure project’ (para. 2). Thus established, these themes are carried forward in the judgment.

Producing (un)reliability

The claimants’ case turned largely on the intent attributable to AML and on witness testimony, including of abuses by an AML employee which, AML conceded (para. 19), would establish liability if accepted. The judgment’s outcome thus rested in significant part on Turner J finding on a balance of probabilities that the claimants and their witnesses, little of whose testimony conflicted with what might be considered prima facie more objective evidence, were not credible and AML not ill-intentioned.

For the claimants and their witnesses, uncivilised attributes and motives are continually evoked to find them not credible. Discussions about an AML employee in a local community, for example, represent ‘the perfect environment in which evidence is likely to be fatally contaminated by rumour, reconstruction and communal resentment’ (para. 247). Likewise, contamination by ‘mutually reinforcing rumour, recrimination and resentment’ can, for Turner J, explain why various separate accounts from witnesses of abuses by an AML employee are all unreliable (para. 147); a contemporaneous written account dictated by a witness is similarly rejected as a ‘retaliatory response to all the woes which he attributed to [the employee]’ (para. 173(iii)). Gendered tropes are also apparent in the rejection of evidence by two women, including Ms Kalma, who had been shot and according to witness testimony ‘beaten badly’ (para. 225), of AML personnel’s involvement in abuses. Considered ‘openly hostile’ to AML by Turner J, their ‘complaining’ of additional harms by AML such as pollution of the local water mean that neither ‘could be categorised as a neutral observer’ (para. 223); as if ‘neutral observer’ is the expected category for a litigant.

If their supposed backwardness and experience of harms render the claimants untrustworthy, AML’s profit motives make it presumptively well-intentioned, as, reasons Turner J, using excessive force was ‘against [its] economic interests’, being ‘very likely to breed future resentment, hostility and mistrust’ (para. 177(ii)). Turner J maintains this presumption throughout despite finding, inter alia, that correspondence circulated amongst AML’s senior management confirmed that the police to whom AML was providing transportation and other key support were ‘mobilising for an all-out assault’ during the 2012 incident (paras. 195 and 216). He also makes favourable credibility assessments regarding the testimony and contemporaneous written accounts – deemed ‘the most reliable guide’ (para. 124) and ‘accurate’ (para. 306) – of AML’s expatriate witnesses, their evidence apparently uncontaminated by their environment of widely shared ‘frustration’ and threats against local communities (paras. 107-112 and 306).

And while AML’s Sierra Leonean employees remain untrustworthy, imputed good intentions are redemptive: Turner J finds that two such employees lied about their involvement, but that their lies are explicable by a desire to create distance from ‘deplorable’ conduct and avoid ‘the stigma of perceived complicity’, not because the allegations against them were true (paras. 230 and 258).

That rationalisation may be contrasted with Turner J’s rejection of the evidence of the claimants’ witness, a Sierra Leonean police commander (and not a party to proceedings) testifying under an anonymity order, that he acted under standing orders to open fire issued to him and other officers along with cash and alcohol by AML’s Health, Safety and Security Manager as ‘no more than a confection to camouflage the extent of his own responsibility’ (paras. 260 and 270).

Reframing Resistance as Lawlessness

Both incidents involved acts of resistance. According to a Sierra Leonean NGO’s investigation into the 2010 incident, local residents rejected an offer for land that AML wanted for a dam; it was a burial ground, used for cultivation, and the community’s only source of drinking water. When AML’s bulldozers arrived, residents protested, temporarily preventing company personnel from leaving the area. And according to a Human Rights Watch investigation documenting many destructive impacts of AML’s operations on local communities, upon giving notice to strike in 2012, local workers had cited issues which included poor working conditions, abuse by expatriate staff, and the inability to join their chosen union that they said they had long sought to resolve with management without success.

Although these reports were before Turner J, such background is omitted from his judgment. Instead, he writes that the 2010 incident concerned land with compensable crops that was ‘suitable’ for subsistence farming, in which AML was interested for a dam ‘necessary’ to its future operations (para. 114); the 2012 strikers were simply ‘discontented employees’ (para. 181). Just as the cause for resistance is transfigured, so are the acts. Turner J accepted evidence that AML personnel were released unharmed by local residents in the 2010 incident (para. 126(iv)), and that there was no evidence that protesters instigated the violence in the 2012 incident (para. 182). Nevertheless, their acts are defined in terms of illegality and criminality. Their resistance is said to reveal local people as ‘prone to react to real or perceived injustices…by deploying unlawful means’ (para. 279) and to exemplify ‘criminality’, the consequences of which if tolerated ‘would likely be a complete breakdown in law and order’ (para. 357).

(Dis)locating the Source of Violence

Disengaging AML from the violence entailed rendering it intrinsic to the people of Sierra Leone. The violence may not have been instigated by local residents or workers, but violence and criminality are identified as latent within them, the potentiality justifying AML’s actions. Thus, Turner J cites ‘the risk of repeated outbursts of…potentially violent criminality’ (para. 279, emphasis added) and the ‘ever present risk’ of illegality in concluding that AML acted as to be expected and exercised no improper influence over the police (para. 302(iv)).

As for the violence that did occur, while it was consistent with securing conditions amenable to AML by crushing local resistance, the Sierra Leonean police officers’ actions are instead attributable to a ‘lack of training, natural inclination or a combination’ thereof (para. 282, emphasis added) and ‘a mixture of fear, ill-discipline, anger and testosterone’ (para. 266), their ‘proclivities…an institutional fact long before [AML’s] arrival’ (para. 356, emphasis added). Indeed, had AML personnel given orders to police officers during the 2012 incident, Turner J would not have accepted that such orders ‘would have made any difference’ to their unlawful acts, given that certain officers were ‘so…violently anarchic’ (para. 266).

Constructing the ‘State’ and Reifying a Public/Private Divide

According to Turner J, AML provided the police with what Sierra Leone ‘ought to have provided to maintain an efficient police force’ (para. 359). This finding projects a particular, (Eurocentric) conception of the state, premised on it being the sole security provider, which is not necessarily suited to Sierra Leone. His projection of that conception, and conclusion that the former British colony failed to meet it, enables Turner J to explain AML’s extensive support to the police. However, it raises another problem for AML, as taking the state’s place in providing that support, and thereby securing for its benefit, as Turner J notes, ‘the mantle of state authority’ (para. 279), implies that AML assumed responsibility for police conduct. That result is avoided by reifying through law a public/private divide not apparent on the facts. AML is not vicariously liable for police actions, Turner J reasons in circular fashion, because the state authority that AML secured for its own benefit by taking the state’s place was derived from police officers’ relationship to the (public) state – ‘from…constitutional powers and responsibilities’ (331(i)) – not the (private) company.


The work and logics underpinning Turner J’s judgment shed light on how the violence and oppression attending transnational corporate projects may be reproduced through, not despite, law. Analysis of the case through an ostensibly human rights lens which ignores that work while foregrounding its confirmation that ‘businesses operating in challenging environments are entitled to seek protection from state security forces’ demonstrates how human rights discourse may perform a similar function.

This does not militate against the importance of those in circumstances like the claimants in Kalma being able to pursue transnational claims, which may provide redress, if not genuine justice, and prevent other harms. It can, however, inform our understanding of such claims. If we see human rights not merely as the subject of a business ‘responsibility’, but, following those such as Shivji, as an element in broader social movements against oppression, then we would also see cases like Kalma differently. Appreciating that they are firstly civil claims for redress, we could look to them as helping to expose the exploitation and violence of the dominant socio-economic order and their relationship with law, while providing insights into the limits of law’s socially-transformative potential as well as its possible strategic use; and we could look to them, too, as both illuminating instances of resistance and affording opportunities for counter-narratives. In at least these ways, pursuing delocalised justice may thus also inform and animate movements against the conditions that make such cases necessary.

[1] RAID, where I am a Legal and Policy Researcher, supported the claimants’ application for permission to appeal to the Supreme Court, which was refused, by way of a Rule 15 submission, but was not otherwise involved in the case. The views expressed in this blog are my own and are not attributable to RAID.

[2] Two other corporate entities within the AML group were named as defendants, one of which had inherited the rights and obligations of the others by the time of judgment. For ease of reference, I follow Turner J’s lead in referring to the defendants as the ‘defendant’ or ‘AML’.

[3] Some of the violence was attributed to Sierra Leone’s armed Operational Support Unit, but the trial and Court of Appeal judgments do not consistently differentiate this unit from the police.