Introduction
At least since Alcock v Chief Constable of South Yorkshire Police,Footnote 1 few jurists have dissented from the idea that the law of negligence liability for what was once called ‘nervous shock’ and is now often referred to as ‘pure psychiatric damage’ is in a lamentable state.Footnote 2 Orthodoxy has it that this area of law protects the right not to suffer psychiatric damage, or, phrased positively, the right to one’s ‘mental health’ or ‘ability to reason’.Footnote 3 Subscribers to the orthodox view then frequently conclude that the English courts have botched the protection of this right and deplore its differential treatment relative to the right to one’s physical wellbeing, where a test of reasonable foreseeability generally seems to govern matters.Footnote 4
Such criticisms have been most pronounced in the law concerning ‘secondary victims’: claimants who suffer injuries as a result of witnessing another person being killed, injured or endangered. They are required to show, in addition to the reasonable foreseeability of a person of ordinary fortitude suffering injury, that they had close ties of love and affection with the immediate victim (rebuttably presumed with parent-child and spousal relationships),Footnote 5 were in close proximity to the accident or its immediate aftermath and had sight and hearing of the same.Footnote 6 These ‘control mechanisms’ have been said to be ‘apt to produce arbitrary outcomes and to exclude meritorious claims’.Footnote 7 The UK Supreme Court’s most recent intervention in this area, Paul v Royal Wolverhampton NHS Trust,Footnote 8 not only confirmed their application but also restricted claims by secondary victims of clinical negligence. So far, most commentators have not bestowed praise on the decision.Footnote 9
Nor has the law on ‘primary victims’ – those immediately affected by the defendant’s negligence – escaped attack, with Michael Jones once describing it as making ‘the calculation of the number of angels on the head of a pin seem like an entirely rational exercise’Footnote 10 and Rachael Mulheron believing it to be ‘the more troublesome twin’.Footnote 11 Typically, these cases involve claimants who are, or reasonably believe themselves to be, placed at risk of physical danger.Footnote 12 Yet the courts have struggled to explain why some who fit this description – such as those who, after being exposed to asbestos, worry they will develop a lethal disease and suffer psychiatric injuries as a result – have had their claims rejected.Footnote 13 Little wonder that Lord Steyn once described this branch of negligence as a ‘patchwork quilt of distinctions which are difficult to justify’.Footnote 14
In this paper, I provide an alternative account of these doctrines that hinges on forging an analogy between the intentional tort of assault and the paradigm case of primary victims. My claim is that the relevant right in negligence cases where a claimant avoids physical impact but suffers from (psychiatric) injuries resembles that protected by the tort of assault: one’s interest in not being threatened with immediate physical violence. If one combines this interest with a concern, originally articulated in Hambrook v Stokes Brothers,Footnote 15 and recently revived by the Supreme Court in Paul,Footnote 16 with the harm to the law’s legitimacy that could occur in cases where it is difficult to distinguish primary victims (fearful for themselves) and secondary victims (fearful for others), then the Alcock control mechanisms are rendered less intractable. Unlike the orthodox account, which sees the law relating to secondary victims as restricting the right not to negligently suffer a psychiatric injury,Footnote 17 the ‘negligent assault’ view sees it as a modest extension of the right not to be negligently assaulted.
Connections between nervous shock and assault are nascent in some of the older cases and feature obliquely in several present doctrines but they have been lost in the modern debates. In his dissenting judgment in Hambrook, Sargant LJ first made an analogy between negligence liability for nervous shock and assault:
There seems to me to be no magic in actual personal contact. A threatened contact producing physical results should be an equivalent. The principle on which a threatened battery may justify damages for assault is in my view strictly analogous. In the case of a threat of imminent danger to a plaintiff resulting in illness through nervous shock, there is, in my view, as real and direct an interference with the personality of the plaintiff as if the illness had been caused by actual physical contact with him. And the duty of a defendant to avoid acts or omissions which will result in the illness of the plaintiff seems to me as clear and definite in the one case as in the other, though no doubt the occasions on which illness will result are much less frequent in the first case than in the second.Footnote 18
In the century since Hambrook, when jurists have drawn parallels between negligence liability for psychiatric injury and the intentional torts, they have tended to reach for the rule of Wilkinson v Downton,Footnote 19 which could be seen as protecting one’s right to mental health (or right not to suffer shock-induced harm) against deliberate interferences.Footnote 20 A notable exception was Lord Oliver’s judgment in Alcock, where he recognised, ‘There is … nothing unusual or peculiar in the recognition by the law that compensatable injury may be caused just as much by a direct assault upon the mind or the nervous system as by direct physical contact with the body’.Footnote 21 In the academic literature, MA Millner once linked assault and nervous shock but it was given short treatment as part of a broader discussion of the trespass torts.Footnote 22 And Robert Stevens analyses the law in terms of a right to mental health but acknowledges: ‘It would have been possible to see recovery as limited to psychiatric injury consequent upon the infringement of a right not to be subject to the fear of physical injury, just as it is possible to assault someone without him or her being touched’.Footnote 23
Each of these associations between assault and negligence was made in passing. This paper therefore represents the first sustained assessment of this alternative rationale’s ability to explain the extant law. My analysis is targeted at ‘accident’ cases, which the majority in Paul defined as external events that cause, or have the potential to cause, injury.Footnote 24 Successful claims for pure psychiatric injuries do occur outside of this context, typically where the claimant and defendant have a pre-tort relationship, but fall outside the scope of this paper and are not inconsistent with my examination of the accident category.Footnote 25
My attempt to make sense of the ‘patchwork quilt’ proceeds as follows. The next section identifies difficulties the orthodox view faces in explaining English law. From there, I examine the paradigm case of primary victims and forge an analogy between this category of claims and the tort of assault, establishing its consistency with the historical distinction between trespass and the action on the case for negligence and demonstrating that the ‘negligent assault view’ can better accommodate the damage suffered in nervous shock/accident cases. After testing this rationale against various primary and secondary victim doctrines, I then offer conclusions.
This paper is intended to be explanatory and takes no normative position on whether the law should follow the negligent assault view or the orthodox view. However, it will hopefully change the debates in this area, which have too often become a back-and-forth between judges erecting and maintaining restrictions and academics rebuking them for failing to liberalise the law.
1. The right to mental health and the orthodox view
(a) The orthodox view(s)
What I have called the orthodox view could be characterised as a collection of perspectives. Members of this group do not speak with one voice. For example, some advocate a test of reasonable foreseeability for determining whether a duty of care is owed to claimants, regardless of the type of injury.Footnote 26 Lord Burrows in his dissenting judgment in Paul indicated that the ‘only truly principled solution’ was arguably to adopt a test of reasonable foreseeability of psychiatric injury.Footnote 27 In opposition to this has been the suggestion, mostly identified with Jane Stapleton, that recovery should be restricted to those whose mental illness flows from a physical injury and wiped out for the rest.Footnote 28 Between these two poles, there are variations on the status quo: permit recovery but maintain some controls, beyond a reasonable foreseeability test, in order to limit the number of claimants. Often this is in a more liberal direction;Footnote 29 sometimes it involves a ‘counsel of despair’Footnote 30 and preserving the current categories in the absence of an obvious better alternative. Uniting each of these positions, though, is that they view the relevant interest as psychiatric in nature (the right to mental health), even if they castigate the courts for doing a poor job of protecting it.
(b) The right to mental health
It could be argued that references to the orthodox view protecting a ‘right to mental health’ is a misnomer. The right is actually more limited. In England, claimants who have suffered purely mental harm must establish that they have a ‘recognised psychiatric injury’. ‘Normal human emotions’ such as grief or anxiety, as opposed to prolonged grief disorder or anxiety disorders, are, alone, not actionable damage in negligence.Footnote 31 Where damage – whether physical or a recognised psychiatric injury – has been established, a wide array of consequential losses, such as pain and suffering, are recoverable.Footnote 32 Consequential mental harm is not relevant to the present discussion as it involves the loss flowing from the violation of a right.Footnote 33 My concern is with identifying the relevant right in accident cases.
I will maintain the language of the ‘right to mental health’ when referring to the orthodox view in order to avoid excessive repetition of the unwieldy, but more precise, label of ‘right to be free of negligently-caused pure recognised psychiatric injury’ but it is this to which it refers. If English law followed Canada and removed the necessity for a medical diagnosis, the ‘right to one’s mental health’ would be more expansive and refer to the right to be free of any mental injury that rises above the ‘ordinary emotional disturbances that will occasionally afflict any member of civil society’.Footnote 34 Such broadening would not affect the argument presented in this paper, for neither understanding of the right to mental health gives the orthodox view a decisive explanatory advantage over the assault view: the debate is not over what counts as actionable damage.
(c) The persuasiveness of the orthodox view
Nobody could doubt that the orthodox view has powerful explanatory force in some common law jurisdictions. In Tame v NSW/Annetts v Australian Stations Pty Ltd, the High Court of Australia removed many of the restrictions on liability for secondary victims, such as the requirements that a person of ordinary fortitude would suffer psychiatric injury and the necessity that the claimant perceive an accident or its immediate aftermath.Footnote 35 Instead, the test was one of reasonable foreseeability. Following the Ipp Report and the enactment of the Civil Liability Acts from 2002 onwards, some Australian states and territories reversed this and are now more closely aligned with the English position.Footnote 36 But in jurisdictions, such as Queensland, that retained the Australian common law position, the orthodox view is pretty much irrefutable. There, the law undoubtedly protects one’s interest in one’s mental health, and the negligent assault view will be an explanatory failure.
(d) Problems with the orthodox view
Turning back to English law, the orthodox view faces difficulties in explaining the restrictions imposed on negligence’s protection of one’s mental health. Until recently, judges have invoked a catalogue of policy factors in support of this position. These have included: a fear of fraud; the need to avoid burdening defendants with disproportionate liability; evidentiary difficulties (ie the complexity of distinguishing psychiatric injuries and normal emotions); ideas that reform should be left to the legislature; the belief that litigation can act as an unconscious disincentive to recovery; and the need to limit the number of claims.Footnote 37 Academic critics have met each reason with formidable rejoindersFootnote 38 but, even if some of the policies are convincing in the abstract, none satisfactorily justifies the specific boundaries currently in place.
The recent case of Paul, where the Supreme Court reoriented the law around accidents, takes a different tack.Footnote 39 In this conjoined appeal, the secondary victims suffered psychiatric injuries as a result of witnessing the death of their loved ones, or its immediate aftermath, following negligent medical misdiagnoses by the defendant. The Supreme Court held that as there was no accident, the claims must fail but, in doing so, it reaffirmed the necessity of secondary victims satisfying the Alcock control mechanisms. Accident cases were treated as an exception to the supposed general rule that ‘the law does not grant remedies for the effects – whether psychological, physical or financial – of the death or injury of another person’.Footnote 40 Secondary victims whose claims are built upon their witnessing of ‘non-accident’ incidents, such as the death of family members from medical misdiagnosis, have had the rug pulled from under their feet. The majority in Paul proffered three reasons for differentiating accident and non-accident claims. First, observing an accident, even where nobody is injured, was seen as generally more traumatic than witnessing the death of a loved one caused in another manner.Footnote 41 Secondly, a focus on accidents enables a reasonably clear boundary to be drawn.Footnote 42 Thirdly, it can be difficult or arbitrary to distinguish primary and secondary victims in accident cases.Footnote 43
Paul therefore offers a different reason for restricting claims by secondary victims. Could this rescue the orthodox view? Possibly. One could say that the underlying interest is still one’s mental health, and the duty of care is to avoid causing reasonably foreseeable psychiatric injury (or even ‘personal injury’); but it is overlaid with a restrictive rule applicable only to secondary victims.
Alas, this response does not withstand further scrutiny. First, in recasting the law of secondary victims, the Supreme Court did not clarify the interest protected in primary victim cases. After White v Chief Constable of South Yorkshire, the test for primary victims is no longer whether psychiatric injury was reasonably foreseeable. Nor does ‘personal injury’ establish the required foreseeability link. Instead, the focus is on the foreseeability of physical injury. Footnote 44 If the underlying right is in avoiding psychiatric injury, it is unclear why the duty of care is framed around a physical connection. Secondly, Paul has received a cool reception. Early commentaries have expressed concern that the ‘general rule’ may be drawn too broadly.Footnote 45 If one has a right to one’s mental health, then, as Stevens has noted, it should not matter whether by interfering with this right a defendant also violates the rights of someone else.Footnote 46 One might also question whether accidents are generally more traumatic than other forms of negligence, especially when nobody is injured. Finally, certainty is always an important consideration but, alone, it does not justify the current law: the same goal could be achieved by wiping out recovery for pure psychiatric injuries altogether.Footnote 47 The difficulty of distinguishing primary and secondary victims is, I shall argue later, more promising but only acquires persuasive force from the standpoint of the negligent assault account rather than the orthodox view.
My starting premise is therefore that in this enclave of negligence the orthodox view does not easily conform with doctrine. This point is relatively uncontroversial, given that many in the orthodox school tend to advocate a ‘thorough purification rather than partial remodelling’Footnote 48 of the law.Footnote 49
2. The paradigm case: negligent assault and primary victims
In English law, the foundation for the recognition of negligence liability for nervous shock was Dulieu v White & Sons. Footnote 50 The defendants were responsible for negligently driving a pair-horse van into the pub where the claimant, who was pregnant at the time, was serving behind the bar. She sustained a severe shock, which made her ‘ill in body and suffer bodily pain’Footnote 51 and resulted in her giving birth prematurely to a child described as being an ‘idiot’.Footnote 52 Dulieu established that there could be an action on the case for negligence for physical damage ‘though the medium through which it has been inflicted is the mind’.Footnote 53 Mr Justice Kennedy held that the shock must arise ‘from a reasonable fear of immediate personal injury to oneself’.Footnote 54
An analogy can be made between Dulieu and the tort of assault. Assault is not a fashionable tort,Footnote 55 but it protects an important interest:Footnote 56 avoiding threatened physical harm. Protecting people from assaults might have the indirect effect of also promoting their psychiatric wellbeing because, ‘the anticipation of physical harm and sometimes even the anticipation of mere physical contact may be traumatic for the person who is forced to anticipate it’.Footnote 57 But the interest this tort protects is not one’s mental health simpliciter ; it is framed around threatened physical contact.
Assault is one of the ancient trespass torts. It involves, according to an influential definition from Robert Goff LJ, ‘an act which causes another person to apprehend the infliction of immediate, unlawful force on his person’.Footnote 58 It does not require any physical contact to be made; that is the province of the separate tort of battery. There is no need, for present purposes, to unpack all of the details or resolve all doctrinal debates relating to this tort but a sketch of its key features will be useful.
Dulieu refers to ‘fear’ of ‘injury’, whereas assault uses the terminology of ‘apprehension’ of physical force (battery). This may appear to be a distinction without a difference given that they are often interchangeable in everyday speech, but their legal meaning is not identical. Apprehension means ‘expecting’ or ‘anticipating’ a battery; it does not require, though it may perfectly well include, fear.Footnote 59 The hooligan enthusiastically joining a pub brawl might not feel scared as punches fly towards him, but he will apprehend them. Since the full range of what a claimant apprehends will, necessarily, include those things that the claimant fears, we might fairly say that the Dulieu situation can be portrayed as one involving a negligent assault. In this type of case, the defendant’s lack of care has unintentionally caused the claimant to apprehend the infliction of immediate physical force. Furthermore, both the assault and the Dulieu-type situation require any apprehension to be reasonable.Footnote 60
The most significant differences between the two torts involve damage and intention. Physical force can be inflicted – and battery committed – without any ‘damage’. The ‘least touching’Footnote 61 of another can constitute physical force: the unwanted kiss is as much a battery as the punch in the face.Footnote 62 This is because battery, like all trespass torts, is actionable per se. Negligence is different: the claimant must have suffered some damage. As mere touching, without more, could never ground a claim in negligence, it might follow that anticipating a mere touching would likewise be insufficient and that a claimant must apprehend physical injury.
Assault has an intention element, albeit an ambiguous one, that negligence lacks.Footnote 63 Accordingly, the Dulieu-type situation, being unintentional, must be pleaded in negligence. If unintentional assaults were actionable without any damage being suffered, liability would likely be drawn too broadly.Footnote 64 The lack of an intention requirement in negligence is balanced by a requirement for the claimant to have suffered actionable damage.
When all of the above is considered, it requires no great leap to see the principle in Dulieu as a ‘negligent assault’ with some modifications that are necessary because the action is brought in negligence rather than trespass: the intention requirement is removed and a damage requirement is substituted. Defendants will owe a duty of care when they can reasonably foresee that their acts (or, exceptionally, omissionsFootnote 65) will cause a claimant to apprehend an immediate physical injury. The philosopher Frances Kamm once posited that, even without a theory, analogies may help us to ‘see that case A is more like B than C and use that conclusion to help us find a solution to case A’.Footnote 66 In what follows, I will aim to establish that this analogy can be used to show that the doctrine in accident cases is more consistent with the right not to be assaulted than the right to mental health.
3. Trespass and case
The negligent assault framework is consistent with the historical relationship between the trespass torts and the action on the case for negligence. Traditionally, the two were distinguished by directness: it was possible to plead in trespass for careless conduct provided the interference was direct.Footnote 67 Later, Letang v Cooper held that trespass and negligence were divided on the basis of intention: trespass involves intentional interferences with the person.Footnote 68 The argument here is not that claimants can bring a claim in trespass for a ‘negligent assault’ nor that the mental element (intention) of the tort of assault should be modified in England.Footnote 69 Rather, this duty of care category in the tort of negligence protects a similar interest. It is therefore perfectly consistent with Letang’s confinement of trespass to intentional interferences.
Long before Letang, the 1833 case of Williams v Holland held that it was possible to bring an action on the case in negligence for a direct interference. As Tindal CJ stated, ‘where the injury is occasioned by the carelessness and negligence of the Defendant, the Plaintiff is at liberty to bring an action on the case, notwithstanding the act is immediate, so long as it is not a wilful act’.Footnote 70 It was recognised that where something that would otherwise be a trespass is committed carelessly (in other words, where there is a negligently-caused direct and immediate interference) and the claimant suffers damage as a result, one can bring an action in negligence. This means that Sargant LJ’s assault analogy does not contravene the distinction between trespass and negligence.
4. Duty and damage
Under both the orthodox and the negligent assault views, claimants must still establish that they have suffered damage for their claim to be actionable. The two accounts then diverge on the relationship between duty and damage. The orthodox view perceives the duty of care rules as flowing from the psychiatric nature of the injury.Footnote 71 By contrast, the assault view sees the special duty of care rules applying, not to the nature of the damage but to the mechanism by which the injury was caused. Footnote 72 Historically, this was via shock, but one need not adopt that label. When Page v Smith reached the Court of Appeal, Hoffmann LJ noted that damages can be recovered for ‘mental or physical illness resulting from the effect of an event upon the mind’. The distinguishing feature was ‘its causation rather than its symptoms’, the latter of which can be physical.Footnote 73 He therefore preferred the label ‘damage caused by mental trauma’.Footnote 74 The negligent assault view aligns with this thinking by focusing on damage caused by a certain form of trauma. The duty would be to take reasonable care to avoid causing a claimant to apprehend the infliction of immediate physical danger. This would then become actionable when the claimant suffered damage, which could be psychiatric or physical.Footnote 75
This, admittedly old-fashioned,Footnote 76 way of thinking provides one explanatory advantage of the assault view over the orthodox view. Many of the foundational nervous shock decisions do not involve pure psychiatric injuries but physical damage.Footnote 77 One hopes that readers of this paper would agree that when assisting someone having a haemorrhage,Footnote 78 miscarriageFootnote 79 or with an injured back,Footnote 80 the first port of call should not be the psychiatric ward. Nor are physical injuries confined only to the early landmark cases.Footnote 81 Unless one wants to maintain that judges at the turn of this century believed that a stroke was a pure psychiatric injury,Footnote 82 the orthodox view faces problems in categorising and accounting for accident cases, decided under the nervous shock rules, where the claimant has suffered physical injuries.
It is not only cases involving psychiatric injuries that would fall under the negligent assault duty category. Consider the facts of Jones v Boyce,Footnote 83 an action on the case for negligence. The claimant had jumped from a coach and broken his leg. He alleged that the defendant proprietor had negligently conducted the coach. Lord Ellenborough held that it was not necessary for the claimant to have been thrown off the coach; it was sufficient ‘if he were placed by the misconduct of the defendant in such a situation as obliged him to adopt the alternative of a dangerous leap, or to remain at certain peril’.Footnote 84 The key question was whether a ‘reasonable and prudent mind’ would have adopted the same course.Footnote 85 The jury found for the claimant. This case, which was cited with approval in Dulieu,Footnote 86 can also be recast as an example of a negligent assault, one that resulted in physical damage to the claimant. On the understanding of the law adopted in this paper, precisely the same principles in Dulieu would apply to this claimant: a negligent assault is a category of case that attracts a duty of care, which becomes actionable upon the suffering of either physical or psychiatric injury.
5. Primary victims revisited
(a) Page v Smith
I have argued that Dulieu can be seen as protecting a similar interest to the tort of assault: fear of an immediate physical injury caused the claimant to suffer damage, even though there was no physical impact. Since then, the courts have expanded this category of primary victim. In Page v Smith, the claimant was involved in a car accident of ‘moderate severity’ caused by the defendant’s negligence. Neither party was physically injured by the initial impact, but the claimant suffered recrudescence of myalgic encephalomyelitis. His claim was successful, despite the lack of evidence that he had felt fear.Footnote 87 Put another way, primary victims do not simply include those who actually anticipated physical injury, but also those placed at foreseeable risk of bodily injury.Footnote 88 Claimants need only show they were in the ‘zone of danger’ to be classed as primary victims. Given that a claimant might be in the zone of danger but not apprehend or fear any physical injury, this may undermine the explanatory potential of the assault rationalisation.
Page has attracted no shortage of scathing criticism from subscribers to the orthodox view. In White v Chief Constable of South Yorkshire, Lord Goff (dissenting) maintained that Page ‘constituted a remarkable departure from … generally accepted principles’.Footnote 89 This view has gained support from academics, who believe that identifying the zone of danger might ‘generate litigation over a tenuous divide’.Footnote 90 If one accepts the view that Page was wrongly decided then the case can be disregarded and the problem evaporates.
But one need not discard Page in order for the negligent assault view to hold. There may be practical grounds for slightly widening the pool of primary victims to the Page-type situation. Doing so might save court time and litigation costs by avoiding a forensic process devoted to determining whether a particular claimant actually expected violence or not.Footnote 91 As Lord Walker acknowledged in Corr v IBC, Page ‘provides a much simpler test for judges trying personal injury cases’.Footnote 92 Apprehending some touching is enough for an assault, and so it might be thought that debating the risk of physical injury against the fear of physical injury in this context is not worth the candle. A broader test does not undermine the underlying basis or aims of the rule: protection against injuries arising from situations where someone is exposed to the threat of immediate physical harm. This modest extension of the scope of the right can therefore be justified on practical grounds (to save litigation costs). By contrast, proponents of the orthodox view struggle to accommodate the law’s concern with physical imperilment.
(b) Fear of the future claims
A key explanatory advantage of the negligent assault view relates to ‘fear of the future’ claims. Consider Rothwell v Chemical & Insulating Co Ltd. Footnote 93 This case was brought by claimants who had been exposed to asbestos by the defendants and, as a result, developed pleural plaques. Although this indicated that the claimants were at risk of contracting an asbestos-related disease in the future, the plaques were symptomless and so, without more, did not constitute actionable damage in negligence. In the conjoined appeals before the House of Lords, one of the claimants, Mr Grieves, had, after the discovery of the plaques, suffered from clinical depression, a recognised psychiatric illness. This was actionable damage, but his claim was nonetheless rejected on the basis that his psychiatric reaction was not reasonably foreseeable.
The conclusion that it is not reasonably foreseeable that exposing people to the risk of a terrible disease might cause psychiatric injuries in a person of ordinary fortitude seems hard to credit. In any case, post-Page v Smith and White, the test for primary victims is not whether psychiatric injury was reasonably foreseeable: it is whether there was a reasonably foreseeable risk of physical injury. A duty of care would be owed in such circumstances, even if the claimant did not suffer any physical injury. Furthermore, with primary victims, reasonable foreseeability of injury is assessed at the moment of breach (unlike secondary victims, where it is considered in hindsight).Footnote 94 At the time of the exposure, it would have been reasonably foreseeable that the claimant might suffer a physical injury, such as mesothelioma or asbestosis. It is therefore difficult to see how the claimant does not fit the definition of a primary victim. As Stephen Bailey and Donal Nolan have concluded, ‘there was no principled distinction between the facts of Rothwell and Page’.Footnote 95 Despite this, the House of Lords believed the case was ‘distinguishable’Footnote 96 and that it would be an ‘unwarranted extension of the principle in Page v Smith to apply it to psychiatric illness caused by apprehension of the possibility of an unfavourable event which had not actually happened’.Footnote 97 But this does not provide sufficient grounds for the claim’s failure. Why could exposure to asbestos not be an unfavourable event that has actually happened? And why is exposure to danger sometimes classed as a sufficient event, even if no physical injury occurs?
A principled explanation for this result can be found via the assault rationale. The defendant may have exposed the claimant to a risk of physical injury in Rothwell, but it was not sufficiently immediate to constitute a negligent assault. It is a key requirement of assault that the claimant apprehends an immediate battery.Footnote 98 In addition to being part of the definition articulated by Robert Goff LJ, mentioned above, several cases have turned on this point. A notable example is Stephens v Myers, where the defendant advanced towards the claimant with his fists clenched but was prevented from carrying out any physical threat. Tindal CJ held that the question was whether the defendant was advancing ‘so that his blow would almost immediately have reached the chairman, if he had not been stopt [sic]’.Footnote 99
Given the immediacy requirement, a fear of harm in the distant future is insufficiently analogous to an assault and so is not covered by this duty category.Footnote 100 Although no assault analogy was made in Rothwell, there are inchoate statements in the judgments that support this distinction. For example, Lord Hope opined: ‘The immediacy that is characteristic of the situation that applies to primary victims as contemplated in Page v Smith … is lacking in his case.’Footnote 101 Without the underlying right being identified as analogous with assault, a requirement for immediacy makes little sense. Unlike the orthodox view, the negligent assault view enables us to explain the differential treatment of these landmark cases.
(c) Rescuers
The decision in White v Chief Constable of South Yorkshire, where the House of Lords held that police officers who acted as rescuers in the Hillsborough disaster were not owed a duty of care, is easily accountable under the assault rationale. The House of Lords in White disagreed with Lord Oliver’s view in Alcock that rescuers were a separate category of primary victim.Footnote 102 Instead, they must meet the criteria in Dulieu or Page for primary victims: physically unimperiled rescuers are owed no duty of care.Footnote 103
Although the framing of the duty around a physical nexus in White creates tension with the orthodox view, the decision might be explained by policy considerations or by the general rule in Paul, detailed above. However, the negligent assault view perfectly aligns with the decision in White. There was no negligent assault on the police officers in White as they faced no anticipated immediate physical harm. Notwithstanding the criticism the law relating to rescuers has attracted,Footnote 104 it is explicable under the negligent assault view.
(d) Unwilling participants
Turning to the ‘unwilling participants’ category, it potentially defies explanation under the negligent assault rationale. In Alcock Lord Oliver believed that a claimant who is ‘put in the position of being, or of thinking that he is about to be or has been, the involuntary cause of another’s death or injury’Footnote 105 (an unwilling participant) would also be classed as primary victims. The negligent assault view would not be able to classify these claimants as primary victims if there is no question of their facing the infliction of immediate physical imperilment, though it is worth noting that Lord Oliver’s characterisation of them has not been universally acclaimed.Footnote 106 Furthermore, some respected scholars maintain that this category either no longer exists,Footnote 107 or that it involves a special type of secondary victim.Footnote 108 If one of these positions is adopted, then the difficulties that unwilling participants present for the negligent assault account are reduced.
6. Expanding the analogy: secondary victims
If one wanted to adopt a strict interpretation of the right not to be negligently assaulted, then the assault analysis would not conform with a duty being owed beyond primary victims exposed to a threat of immediate physical injury, perhaps with a minor extension to the Page situation on practical grounds. Secondary victims appear to be at odds with the thesis. In what follows, though, I will argue that if one combines the assault rationale with a concern for avoiding the making of invidious distinctions between primary and secondary victims, then aspects of the Alcock control mechanisms begin to make sense.
Following McLoughlin v O’Brian,Footnote 109 debates on secondary victims have tended to focus on whether or not additional factors beyond reasonable foreseeability should be used to contain the protection of their right to mental health. Following the assault paradigm, though, supplies a different perspective. The relevant rules could be viewed as an extension of the duty of care owed to primary victims.Footnote 110 The first case to permit claims by a secondary victim, even if that label was not in use at the time, was Hambrook. A mother feared for her children’s safety and suffered a nervous shock after witnessing a careening lorry go around a bend in the road towards where her children had been. The shock was followed by a severe haemorrhage and miscarriage which led to her death. A majority of the Court of Appeal held that she was owed a duty of care. In defending this expansion of the law, Bankes LJ used an example of two mothers crossing a street, each holding a child’s hand, when a lorry comes thundering down the road. One, ‘courageous and devoted to her child’, is terrified, but ‘thinks only of the damage to the child, and not at all about herself’ whereas the other, ‘timid and lacking in the motherly instinct’, is also terrified but ‘thinks only of the damage to herself and not at all about her child’.Footnote 111 Where both mothers suffered injury through shock, the law would be favouring the ‘less deserving mother’ if only the latter’s claim could succeed.Footnote 112 Such distinctions, thought Atkin LJ, ‘would be discreditable to any system of jurisprudence in which they formed part’.Footnote 113 The Court of Appeal appeared to be recognising that the obligation to avoid negligently assaulting a claimant can be extended in order to preserve the law’s legitimacy.Footnote 114
Similar reasoning was deployed by the Supreme Court in Paul, where it was acknowledged that in accident cases it was ‘often difficult or arbitrary in such cases to distinguish between primary and secondary victims’.Footnote 115 In a case where ‘both a mother and her child are put in physical peril, it would not only be unjust but practically impossible to distinguish between the mother’s emotions of fear for herself and fear for her child’.Footnote 116 Outside of accident cases, allowing the claimant to recover cannot be based on ‘the practical impossibility and injustice of otherwise having to distinguish between injury caused by fear for the claimant’s own safety and injury caused by fear for the safety of a close family member’.Footnote 117
In this respect, the Supreme Court in Paul came close to identifying the grounds for permitting some claims by secondary victims but its failure to characterise the paradigm case as involving a ‘negligent assault’ occluded this insight. At this stage one might reasonably ask whether subscribers to the orthodox view would not also care about avoiding invidious distinctions between comparably situated claimants. Undoubtedly, they would, but the starting point is different. And if one’s point of departure is that the underlying right is to one’s mental health, it becomes hard to see why any difficulty in distinguishing primary and secondary victims would matter. There is no connection between the primary-secondary distinction and the right to mental health. This reason does little to assist the orthodox view.
From the negligent assault position, some key judicial concerns in the duty of care rules relating to secondary victims can be captured. A duty is owed where distinguishing between primary victims and secondary victims would usually be exceptionally difficult and arbitrary by requiring a debate over whether a person was fearful for themselves or their loved one. If a person witnesses a loved one being killed, injured or endangered, the law could be making invidious distinctions if no duty of care was owed in negligence. However, the further removed a person is from the Hambrook-type situation – in other words, the easier it is to distinguish primary and secondary victims – the weaker the reasons for extending the duty to secondary victims.
This is not to say that the assault rationale can explain all aspects of the law on secondary victims. Restrictive interpretations of the close ties requirement have no inherent connection to the right not to be assaulted – nor for that matter, with the right to one’s mental health – and so neither bolsters nor undermines either account.Footnote 118 With this caveat in place, I am now in a position to assess the other control mechanisms.
(a) Proximity to the accident or its immediate aftermath
Requiring close proximity to the accident or its immediate aftermath has long been a contentious restriction on secondary victim claims. It appears baffling under the orthodox view. As Jane Stapleton put it, ‘The parent who comes across her child’s bloodied corpse just a little too late for the current rules on recovery for nervous shock to her apply to her might wonder why the law rules her child’s blood too dry to found an action.’Footnote 119 However, an expanded negligent assault view might account for it. It can be difficult to distinguish a primary victim and a secondary victim when the latter is at the scene of the accident; it becomes much easier if a claimant witnesses the collapse and death of a loved one later. The law can easily tell whether one was fearful for oneself or another if the claimant is not ‘present at or near the scene of the accident’.Footnote 120 It is only when it becomes exceptionally difficult to distinguish the two that extending liability is warranted.
This analysis might help to explain Taylor v Novo. Footnote 121 The claimant’s mother was injured at work when a stack of racking boards fell on her. She appeared to be recovering when, a few weeks later, she died at home. The claimant, who witnessed the death, suffered psychiatric injuries and so brought a claim in negligence against her mother’s employers. The Court of Appeal held that there was one accident with two consequences: the mother’s initial injuries and her death several weeks later. Witnessing the latter was insufficient to establish a duty of care. This makes sense under an assault analysis. A collapse at home a few weeks after an accident cannot be the relevant event in these circumstances because there is no difficulty in distinguishing the primary and secondary victim: it is easy to determine that the daughter was not fearful for her own safety. This is not to say that had the daughter witnessed the falling of the boards it would necessarily be impossible or difficult to separate fear for herself and fear for her mother. But the law needs to develop rules that can be applied easily in future cases and so the current rules on secondary victims may be drawn widely to capture the cases where it would be difficult to tell a primary victim and a secondary victim apart when there has been a negligent assault. This does not prevent their basis being an extension of the law on primary victims in order to preserve the law’s legitimacy. The same is true of misdiagnosis cases such as Paul. As there had been no negligent assault, there was no trouble in discovering whether the claimants were primary or secondary victims. Extending recovery to the relatives could not be seen as an expansion of the right not to be negligently assaulted as there was no threat of immediate physical impact to the primary victims.
The aftermath doctrine creates greater difficulties. If properly confined, it might just fit the above rationale. Take the Hambrook-type situation of an out-of-control lorry. In such circumstances, allowing claimants who witness the aftermath is intelligible. The mother may not have been in actual danger at the moment the lorry turned the corner, but it is well known that a pile-up can soon follow a car crash. An initial impact can cause further accidents. With some accidents, arriving at the immediate aftermath can place one in danger. This concern with not making invidious distinctions may apply where a secondary victim arrives at the immediate aftermath.
The result in McLoughlin, where a parent saw distressing scenes at the hospital an hour or two after the accident, cannot be reconciled with the extended assault rationale. On those facts, the courts face zero difficulty in determining whether the claimant’s concern was for her own safety or for her family: it is very easy to state that it was the latter. Post-Paul, such cases might also create further arbitrariness: two mothers who witness their children die in a hospital might be subject to differential treatment if one’s damage was a consequence of an earlier accident and the other was a product of medical negligence. If the negligent assault rationale is to be followed, then its realisation would require the immediate aftermath to be reined in. Indeed, the problems with stretching the immediate aftermath were recognised by the majority in Paul, which acknowledged that the ‘clarity and certainty of these tests have been compromised to some extent by the decision in McLoughlin’.Footnote 122 However, any tension with McLoughlin does not do much to diminish the cogency of the assault view given that the orthodox view has so far failed to provide a convincing justification for the requirement of close proximity to the accident, never mind the immediate aftermath.
(b) Sight and hearing of the accident
The negligent assault view can also explain the requirement of direct perception or sight and hearing of the accident. This hurdle, once described by the Law Commission as ‘unduly restrictive’,Footnote 123 does not seem to reflect the potential horror that can be inflicted by television, let alone the smartphone,Footnote 124 and so can appear unfathomable under the orthodox account. There is little question that the experience of family members who witnessed the Hillsborough disaster unfold on television was disturbing, regardless of whether the broadcasting code of ethics prevented close-ups of scenes of individual suffering.Footnote 125 Consider, too, an example supplied by Lord Bridge in McLoughlin:
Take the case of a mother who knows that her husband and children are staying in a certain hotel. She reads in her morning newspaper that it has been the scene of a disastrous fire. She sees in the paper a photograph of unidentifiable victims trapped on the top floor waving for help from the windows. She learns shortly afterwards that all her family have perished. She suffers an acute psychiatric illness. That her illness in these circumstances was a reasonably foreseeable consequence of the events resulting from the fire is undeniable. Yet, is the law to deny her damages as against a defendant whose negligence was responsible for the fire simply on the ground that an important link in the chain of causation of her psychiatric illness was supplied by her imagination of the agonies of mind and body in which her family died, rather than by direct perception of the event?Footnote 126
One could even adapt the situation to modern conditions by having the husband video call the wife from the burning building. This could be more traumatic than being on the street outside. But trauma is not the yardstick. Under an assault analysis, the law can quite easily distinguish primary and secondary victims in such cases. There is little question of people anticipating immediate physical injury to themselves when witnessing dangerous events through a screen or reading newspaper stories about them. In such circumstances, primary and secondary victims can be distinguished easily and the reason for expanding recovery to secondary victims is absent. This restriction is consistent with the assault view.
In sum, the orthodox view struggles to provide a compelling account of why the ‘immediate aftermath’ and ‘sight and hearing’ should restrict someone’s right not to be negligently inflicted with purely psychiatric injuries. As mentioned earlier, it would be in keeping with a strict interpretation of the negligent assault view to remove any duty being owed to secondary victims altogether. Extending liability to secondary victims probably made more sense post-Hambrook but pre-Page’s expansion of the paradigm category of primary victims. Situations where it is difficult to tell whether claimants are fearful for themselves or for the immediate victim are likely to be ones where secondary victims were in the zone of danger and so classed as primary victims anyway, even if they are not fearful of harm to themselves. However, if it is thought that a duty should be owed to secondary victims, the Alcock restrictions of close proximity and sight and hearing of the accident or its immediate aftermath make more sense when construed as rippling out from the assault core. Under the negligent assault view, claimants could be advised that the law allows actions by those who are negligently assaulted or, in a very marginal extension, close relatives who are at the scene of a negligent assault where it would be difficult to determine whether the claimant apprehended injury to themselves or a loved one. The rules are drawn widely enough to encompass this limited class of secondary victims, who could be seen as directly involved in the assault, even if they do not apprehend any physical violence to themselves. One could even call them ‘extended primary victims’ or ‘peripheral primary victims’.
Conclusions and areas for future debate
A widespread consensus in the academic literature perceives English negligence law as attempting to protect a right to mental health, even if it is often seen as doing so in an incomprehensible and arbitrary manner. In this paper, I took up an analogy first drawn by Sargant LJ in Hambrook in order to show that large swathes of the extant law are better explained by a different right: the right not to be negligently assaulted.
Under the assault view, a defendant must take reasonable care not to cause a claimant to apprehend the infliction of immediate physical injury. This becomes actionable if the claimant suffers damage (whether physical or psychiatric) and meets the other requirements for a negligence claim. Although Sargant LJ’s analogy has never, until now, been developed into a full account of the underlying right, it is remarkable how often judges have intuitively landed at the correct destination while essentially flying blind. The features that most align with the assault view relate to primary victims. It can explain why the law allows the paradigm case involving fear of bodily injury represented by Dulieu but excludes actions by unimperilled rescuers and ‘fear of the future’ claims.
If one accepts granting an extension of the right not to be assaulted to scenarios where it would be difficult to tell primary and secondary victims apart without potentially bringing the law into disrepute, then the requirements for close proximity and sight and hearing of the accident or its immediate aftermath also begin to make more sense from a ‘negligent assault’ starting point. Secondary victims who meet the Alcock control mechanisms are directly involved in a negligent assault and might warrant protection. The analogy cannot explain everything, though: unwilling participant claims and the proximity of relationship requirement for secondary victims remain equally difficult under the assault analysis as they do under the orthodox understanding. And, even if a broadening of the negligent assault view is granted, it can only go so far. McLoughlin, with its extension of the ‘immediate aftermath’ to the hospital, would have to be decided differently.
This paper should by no means be seen as the last word on this topic. Suppose one believes that the policy reasons outlined in the introduction or the general rule in Paul can provide a compelling explanation of the law’s boundaries. In that case, one can reject the very first premise of this paper and endorse variations of the orthodox view. This paper’s focus was on accident cases. It may be that if one steps back and looks at the broader duty of care categories, problems of fit at this higher level become more challenging for the assault analysis and less troublesome for the orthodox view. Those committed to the idea that damage determines the duty of care rules, rather than merely being a condition of actionability in negligence, might also resist these arguments. Within the assault analysis, there is room for disagreement about whether the right should be strictly interpreted or whether the expansion to secondary victims, and indeed to other categories of claimants, is justified.
Litigating such debates is outside the scope of this paper. So, too, are normative ones over whether the right to one’s mental health is more important than the right not to be assaulted and so should be accorded greater protection. This paper is explanatory. One would still be free to argue that English law should adopt an approach similar to Queensland.
Although I have taken no normative stand in this paper, I hope I can be forgiven for closing with the observation that the negligent assault view neither mandates wiping out recovery altogether nor permitting all claims whenever pure psychiatric injuries are reasonably foreseeable. It therefore offers a principled rationale for those who wish to maintain something approximating the present restrictions on liability. At the very least, if my analysis is accepted, the current rules can no longer be treated as an Aunt Sally by those who advocate a ‘thorough purification’ of the law.
Open access
