Heterogeneity, Risks, and Torts: Exploring the Worlds of Acts and Activities

I Introduction

This paper focuses on heterogeneity in Gregory C Keating’s Reasonableness and Risk,^[1] appreciating its significance while also seeking to expand it. The demonstration of heterogeneity and its defence as a feature of tort theory, as well as tort, is an important distinctive element in the work itself. It is, indeed, the idea which Keating highlights in the concluding chapter, where he shows that it is key to setting his theory of tort apart from a number of others including both law and economics approaches, and neo-Kantian approaches. Supporters of these approaches have tended to criticise the flexibility and open endedness of heterogeneous thinking as leaving the way open for ‘policy-based’ and ‘non-legal’ factors to be discussed both in legal academic work, and in the courts. The responsibility based account that unfolds with heterogeneity at its core is therefore a significant contribution which deserves our focus.

Substantively, heterogeneity finds its key illustration in Reasonableness and Risk in the distinction between responsibility in the world of acts and responsibility 5in the world of activities.^[2] This distinction has significant implications for the interpretation of strict liability, and its place within a theory of tort law. That distinction therefore is an important idea. It also encapsulates a more general conceptual argument that a phenomenon like tort, which moves with the needs of society, cannot be expected to be anything other than heterogenous if it is to remain relevant. Associated with the substantive observation of a distinction between the world of actions and the world of activities is an argument that tort cannot be seen as wholly distinct from other responses to the social problem of harms caused by activities, and that this is not a problem for tort scholarship. All of that is rich, and significant; and it swims against the rising tide of approaches which seek to ring fence tort from other responses to misfortune and risk.

I suggest that we can take these insights further. The world of activities can be understood in other ways too, so that its significance in tort does not stop with the instances presented in the book. The role played by parties who are not necessarily risk creators could be seen as the key challenge within contemporary negligence law, namely where its duties will stop;^[3] and it can be argued that negligence law is as imbued with issues relating to the world of activities – rather than discrete acts – as is strict liability. Thus, the idea of a world of activities can be expanded to shed further light on other roles of the law of tort in relation to risks, beyond what is ordinarily understood by ‘enterprise liability’, because activities are often multi-party affairs, in which different parties take different roles. Those suffering loss are themselves frequently participants in activities linking them to potential defendants. This emerges in the expansion not of strict liability, but of negligence law, and the central debates that have emerged around the justification for, and limits to, this expansion.

In the course of this argument, I consider whether it is really true to say, at this stage in the development of the law of tort, that it is preoccupied with responding to the causation of accidental physical harm.^[4] I suggest that it has become increasingly concerned with the management of risks and with a range of actors whose role is associated with risk, and therefore with risk structures surrounding harm. In this way, I suggest that the world of activities, and the response of tort to these activities, is more diverse and heterogeneous than is indicated within Reasonableness and 6Risk. But it is with recognition of the world of activities, and with the acceptance of heterogeneity, that these observations start.

After this introduction, Section II introduces the role of heterogeneity in Reasonableness and Risk, and outlines the significance of acts and activities. Section III sketches some of the importance of heterogeneity. In Section IV, I explore whether the true reflection of the world of activities extends beyond enterprise liability. We live in a world preoccupied not only with risk creation but also with risk management, and this has been central to the recent development of the law of tort. Ideas of responsibility related to risk management are pervasive within the law of tort, and are themselves heterogeneous. The new relations between parties could be appropriately captured in terms of relations within an activity.

Finally, in Section V, I illustrate by taking up Keating’s invitation to consider a particular instance, namely the design of liabilities for harm done by autonomous vehicles. In the UK, this is approached in terms of an allocation of liability and this has been the subject of consultation and legislative action – itself an indication that contemporary law takes a proactive approach to responsibility in light of change. A theory of tort which had to disown this sort of adaptation to risk and responsibility questions, because it appears to allow a diversity of approaches to responsibility or because it involves statutory design, would be an impoverished theory of tort.

II Heterogeneity and its value in Reasonableness and Risk

Keating’s theory of tort law is able to account for the full range of forms of liability to be found within the law of tort, and to do so with no embarrassment.^[5] Torts of intention, and strict liabilities, are significant component parts of a heterogeneous practice. There is no single form of liability which is representative of the entire law of torts, and neither is this required to make sense of the practice. That is because tort is understood in terms of the role that it plays. While social values may change over time, and protected interests may therefore wax and wane, the role of tort is stable.^[6] Identification of this role is one of the key contributions of the work. It is a 7role properly understood as fundamental: defining the interests which are protected in our interactions with others, and therefore setting out an essential part of our security as members of civil society.^[7] It might be noted that this is a theory which draws on social norms and values, rather than formal values to be found within the law and its structure. As a consequence, it gives the legal scholar no monopoly on normative questions about the nature of protected interests and the manner of their protection, but vindicates the idea of legal scholarship as drawing on the values and practices of the community expressed through law.

It is a great strength of the account that it does not seek to force the law of tort to fit a single normative theory,^[8] but seeks a normative theory which can accommodate the subject as a whole. The notion that a subject that is based in responsibility can at the same time have a social function is important. Further, the relationship between torts and administrative practices which deal with some of the same issues – particularly with the response to accidental harms – is thereby accepted to be an important aspect of the study of torts, and not a threatening alternative to the practice of tort. Here then is another exercise in acceptance of heterogeneity. Direct regulation of risks, and administrative compensation schemes, are approached as both ‘cooperative with, and alternative to, the private law of torts’.^[9] There has been too quick a conclusion on the part of tort scholars that addressing the role of administrative schemes as alternatives to tort must inevitably lead to a critique of tort, for it amounts to an invitation to assess tort by external standards.^[10] It could even be said that this has scared a generation of tort scholars into adopting highly formal accounts which resist both heterogeneity and any comparison with administrative schemes or regulation. On Keating’s account, the coexistence of tort with such schemes is not a threat. Rather, there are occasions where tort is not well able to construct appropriate ‘communities of risk’ in which responsibility for protection of essential interests can be established. Such schemes are ‘cooperative with, and alternative to’ the law of tort. This acceptance that tort and such schemes play similar roles in different circumstances should not be consigned as a surrender of tort’s distinctiveness. Rather, it is an act of confidence in the social role and significance of the law of tort as a branch of private law, dealing with the effects of interactions 8between persons on those essential interests that are worthy of protection as an essential part of our security.

Rather than a single value therefore, Keating instead identifies a core social purpose to the law of tort. This is in clear contrast to the conceptual accounts of the law of tort to be found in neo-Kantian theory for example, which hold that the value of the subject is derived from its conceptual structure. The core purpose is also quite opposed to the ‘pricing system’ that some examples of law and economics propose, and takes seriously the normative idea of responsibility embraced within the law of tort. Instead, Keating builds an important reading of the law of tort and its role and social function. The function of protecting significant interests within interactions helps to explain the nature of tort remedies and to provide a core purpose which can account for heterogeneity in both protected interests, and standards of liability. It can also accept, and to some extent seek to explain, changes over time, as the value placed on security of different interests and on the interactions which might threaten this security itself shifts. Tort law is distinctive but also open to societal needs and changing values. A range of solutions is available within tort law, as well as outside it.

The focus on security of interests within interactions is compatible with the nature of reasoning in private law, and its scope. There is no need to abandon the sense that tort has a key social function in order to understand its nature as an element of ‘private’ law. Governance or regulation of the relations between persons is such a function.^[11] This strengthens the claims of tort as a significant object of study while not abandoning its core values or importing such values from other ‘external’ normative systems, such as economics.

Some other scholars have captured a similar idea in terms of the ‘security’ of selected interests as key to the law of tort.^[12] It is a potentially useful idea in understanding the distinctive role of a subject which is concerned with the protection of what we have and when we are entitled to claim such protection. Focusing on security of interests in the context of creation of risk through social interactions secures a balance within the understanding of tort which is lacking when considering the parties’ relationship in a single way, for example through the fact of causation of harm, or through the defence of rights, or the maximisation of value.

III 9The world of acts versus the world of activities

Of course, Keating does not stop with the observation of heterogeneity, important though this is, but gives this substantive content. It strikes me that the key substantive aspect of heterogeneity defined in Reasonableness and Risk is the distinction between the world of acts and the world of activities and the observation that both play a part in the contemporary law of tort. It is important to recognise the significance of this distinction in terms of the more general points already made. Keating explains the distinction in historical terms, helping to account for shifts in standards of liability for accidental physical harm in particular, in terms of the scale of such harm and the manner in which it is caused over time. With this shift comes changing notions of what is required to justify responsibility.

In Keating’s account, the distinction between the world of acts and the world of activities is inherently connected to the rise of enterprise liability. The enterprise is linked to activities and their effects rather than discrete faulty acts, and in some circumstances it justifies the imposition of strict liability, in the strong sense of liability for harms done by actions which ‘should not have been avoided’.^[13] Harms done in this way, in other words by justified risks, may be compensated on grounds of fairness, and the perception of activities is crucial to this being an acceptable response. I will explain shortly that there is relatively little enterprise liability of precisely this sort (responsibility for the harms done by expressly justifiable risk) in English tort law, but I will also suggest that an understanding of the world of activities is present nevertheless. Connected to this, I will argue that the power of this idea of a world of activities is broader than the notion of enterprise liability. This is not an argument for expanding strict liability further but a suggestion that we use the notion of the world of activities and distinction from an earlier focus on the world of acts to explain some issues about the reach not just of strict liability but of negligence law, for example. The world of activities is not limited to explaining responsibility in terms of scale of operation.

Keating explains the focus on acts, and the focus on activities, as belonging to different social worlds – in other words, in historical terms. The nature and frequency of accidents changed in the progress of the 19th and 20th centuries, but so too did the organisation of economic and social activities. Thus enterprise liability is identified as an appropriate foundation for strict liability in the 20th century.^[14] Keating draws on Holmes as explaining the difference between discrete risks associated with thinking in terms of acts; and generalised and systemic risks associated 10with the world of ‘activities’.^[15] In the modern world, the typical injury does not arise from diffuse or disorganised acts of individuals, but out of organised activities of firms (not individuals) that are large themselves or ‘component parts of relatively well-organised enterprises’. Actuarially large activities lead to a certainty that there will be accidental harm. The world of activities, as Keating presents it, is connected to large actors and large activities, thus to statistical risks which are certain to eventuate. Strict enterprise liability places responsibility with the activity and disperses losses from those who suffer harm. Keating regards this as an appropriate approach for our times. As he puts it, we live ‘in a world of organised enterprises and systemic risks’.^[16] In this world, strict enterprise liability is, or can be, a reasonable response. Negligence liability leaves those who suffer harm from justified enterprise risks to cope with the consequences.

IV Does English law reflect activities and strict liability without enterprise liability?

The idea of enterprise liability is not alien to English lawyers, yet within the law of tort, reasoning which holds that there could be liability for reasonably imposed risks is in extremely short supply. Some ‘strict’ liability does not turn on the idea of liability for consequences of a risk that should not have been avoided. Rather it may be defined more narrowly in terms of the absence of fault. At the same time, I suggest that the world of activities can be identified as very influential in English law in other ways, though we need to work a little harder to explain it. This is a precursor to the next section where I suggest other ways in which the world of activities could be a potent justifying idea without having recourse to enterprise liability as such, and taking account of other changes in social response to risk.

An example of English law being less than clear about the nature of the strict liability it accepts is in private nuisance, which is defined in terms of the reasonableness or unreasonableness of an interference with the claimant’s rights. While nuisance liability is generally perceived as strict (no fault), it is by no means clear that this meets the definition of interferences that should not have been prevented, since the outcome of the (not necessarily negligent) activities must be judged unreasonable for there to be liability. On the other hand, the way that remedies are at11tached to nuisances can considerably impact on the way that we understand its liability rules. Greater use of damages in lieu of injunction would support the sense that the interference should indeed not have been avoided. And yet, it is only in recent years that this interpretation of nuisance remedies has been clearly stated in the higher courts.^[17] Before this, the orthodoxy appeared to be that a nuisance was a wrong that would ordinarily result in an injunction.

Elsewhere in the English law of tort, liability under Rylands v Fletcher appeared to embrace a notion of enterprise liability in an early, individualised form (not related to the rise of corporate harms). But despite the same social changes outlined by Keating occurring or intensifying since the mid-19th century, English courts have deliberately minimised the impact of Rylands.^[18] The same is true for liability based on ‘ultra-hazardous acts’, which has become a dead end in the law.^[19] In both cases, the ideas have not gained as much traction in their original jurisdiction as they have developed in the United States. The responsibility idea embraced by Rylands v Fletcher – that one who creates a significant risk does so at his own peril, not that of his neighbour – has virtually no direct application in English tort law. It should be emphasised that this has happened despite the growth of the same type of strict liability through statute.^[20] Does this play into the hands of those who argue that this sort of development is for statute, and that tort must simply continue to apply reasoning associated with the world of acts? Must it be left to statute to create relevant ‘communities of risk’? I suggest that this is not the case, and that the world of activities is represented in English tort law nonetheless.

In English law, enterprise liability ideas can be seen via more indirect routes, generally dependent on the presence of a fault-based tort. In particular, English law has extensive liability attaching to activities through vicarious liability, and through 12non-delegable duties (particularly on the part of employers^[21]). The action (tort) depends upon faulty acts but at the same time, it is the gateway to liability attaching to the activity (the enterprise). On the face of it, this continues to give priority to the world of acts: someone has acted unreasonably. But importantly, that act is not considered to be discrete and unpredictable.^[22] It is recognised that such unreasonable individual actions are predictable consequences of an activity. We do not attach all consequences to an enterprise; but in principle we do accept that an enterprise can be linked to predicable torts. These are not ‘risks that should have been taken’ from the point of view of the act in question. But they are risks that are part of the enterprise whether or not there is any defect in the risk management of the enterprise: the law of vicarious liability does not concern itself with the quality of the employer’s risk management. It concerns itself with the risk associated with an activity. Because this is related to the responsibility of enterprises for the risks associated with their activities, and not with the profit that they gain from those activities, it has been possible to extend this to activities which are not predominantly concerned with profit, such as charities and churches, as well as public authorities. Unfortunately, the risks associated with these activities have proved to be very significant.^[23]

I would suggest then that English tort law richly reflects the idea of the world of activities, yet at the same time does not chiefly (other than through statutory variants) concern itself with responding to accidental harm arising from risks which are decisively recognised as ‘rightly taken’ (as Keating puts it). That is, it does not operate by addressing the question of rightly taken risks, but does recognise that fault, including malevolence, is itself a predictable component of activities for which the activity, not just the individual, should take legal responsibility. The risk of wrongdoing expands along with other risks when we enter the world of activities.

Other than through the application of vicarious liability, another way in which English law might be said to respond to this without a wholesale embrace of enterprise liability (for justified risks) is that its forensic examination of wrongfulness itself may in practice be either more or less exacting in different circumstances. 13Whether there was a ‘wrong’, and whose responsibility it was, could be the foreground question in different circumstances. Do we accept that very bad people are the responsibility of the organisations that appoint them and give them access to vulnerable claimants; and do we accept that standards of wrongfulness may themselves be manipulated somewhat to reflect the relative ability of enterprises, or communities of risk? For example, will the ‘date of knowledge’ for a specific industrial disease be generously extended for the benefit of claimants, because of a sense that an enterprise or group of enterprises ought to be thoughtful about the risks to which workers are exposed? The last two questions are different questions, but both reflect the influence of the world of activities and its impact on how we understand responsibility, even when applying fault liability.

We could illustrate these issues by addressing change over time in employers’ duties and liabilities to their employees. In the UK, this area was returned to tort when workmen’s compensation was abolished in 1948. Perhaps oddly (at least through American eyes), this was partly because the trade union movement was wedded to a fault principle and full compensation, so that the partial compensation and insurance style solution of workmen’s compensation was perceived as second best despite the greater ease of claiming.^[24] The rise of health and safety duties in the workplace was eventually accompanied by wide ranging civil liability for breach of regulations many of which were (and still are as regulatory duties) strict, or even absolute. Civil liability was removed from breach of health and safety duties in 2013 (prospectively).^[25] How much difference has it made? In a highly significant move from the point of view of understanding the world of activities and its role even in what appears to be fault liability, the Supreme Court soon ruled that a reasonable employer would conduct risk assessments and act upon them. This conclusion was drawn not from previous common law but from decades of application of the regulatory duties, which contained risk assessment duties and are now absorbed at least to this extent to be part of negligence. There is therefore a duty to anticipate risks systematically and it is negligent not to do this adequately, and not to take reasonable steps to mitigate these risks.^[26] Notably this is not about imposing new liabilities, but finding a way of holding onto what has been accepted as fair to this point, and rationalising it in ways that a fault-based (or at least, reasonableness-based) approach can comprehend.

14In summary, it could be argued that English law accepts the implications of a world of activities in a number of ways without at the same time appearing to accept the idea that in common law, there can be liability for harm done by risks that were rightly taken (perhaps the strongest form of strict liability), unless some fault can be found in the process. Statute, it is widely accepted, does the latter in a range of specific circumstances. So there is no failure to grasp the potential fairness of this style of responsibility, but a distinction in the ways in which acceptance of the world of activities operates within the common law of torts.

To this point, and excluding the point about nuisance, the examples have largely been about employers and others connected to individuals in some relationship ‘akin to employment’. This might indeed be a very good place to focus if resisting the explanatory power of Ripstein’s individualising, act-focused explanation that ‘everyone is in charge of themselves and no one is in charge of anyone else’.^[27] Interpretations of vicarious liability diverge in part depending on whether employment is seen as a special case where the employer and employee are close to being identified, or a genuine case of responsibility within activities. But arguably, there are other places we should look for recognition of the world of activities in tort, and these extend deeper into the reach of fault-based torts. While Holmes certainly had a point in his observations of scale of activities and enterprises, we do not always attach blame to the creators of risk, and certainly not solely to them. There is more to the world of activities and risk than these questions of scale.

V The world of activities beyond enterprise liability

Keating suggests that the focus of contemporary tort law is causation of accidental material harm.^[28] It is this that has been subject to the changes connected to the scale of enterprises and risks that he associates with the world of activities. However, it might be argued that this observation discounts much of what preoccupies modern tort law and which could be described as another feature of activities, which are indeed (to take up Keating’s point) now naturally and pervasively understood in terms of the risks that they involve. Rather than perceiving harms to arise typically from discrete acts (the world of acts), or typically from risk creating activities (the world of the enterprise), risks are seen as subject to avoidance and management. 15While the world of activities and enterprise liability invite us to look at the doing of harm and creation of risk, the law is also centrally concerned with the failures of those whose role is claimed to be to prevent or advise or manage these risks. These parties are seen as fulfilling an essential role in relation to the pursuit of activities and in respect of security of interests. They may advise on the extent of risk and on distribution of risk. Equally, within a group of organisations pursuing an activity together, there may be a clear understanding of where the risk of accidental harm is intended to lie.

This focus on risk management is a further indication of the rising role of risk, but rather than focusing on scale, it focuses on diversification in those who are perceived to be connected to those risks and to the harms that flow from them. We live in a world of risk management, not just a world of risk creation. Thus, the rise of professional liabilities on the parts of surveyors, legal advisors, medics, and local authorities is a vital part of the story of contemporary tort law, and is not directly connected to the creation of risk so much as its management and avoidance: this also has placed economic consequences of risk of various types at the forefront of questions about the reach of tort, and not just physical damage. Many of tort law’s most persistent contemporary questions are around these issues. Included in this are questions about the distinction between acts and omissions, harming and failing to secure from harm, and when this distinction is or is not pertinent – for example, where there is an assumption of responsibility in relation to a risk. Thus, there are other facets to the idea of a shift from discrete acts to ‘activities’. One such facet is not about the increase in scale of enterprises, but the creation of multiple roles in relation to an activity, including roles of advice and management. If we think about ‘communities of risk’ now, we may well think in terms of a multiplicity of roles, rather than in terms of a body of risk creators. Whether and where to place liability is a live question in much modern tort law.

The move to a world of risk management (and diversification of parties and roles) may be as significant as the increase in scale of activities. Parties have expectations about how risks will be managed and by whom, and harm done by the eventuation of risks is not solely understood to be the consequence of risk creators. Those who have more direct dealings with the party suffering harm are sometimes understood to have duties in relation to the security of their fundamental interests, even if a more remote party was the originator of the risk.^[29] Keating does note that separate enterprises may well be involved in an ‘activity’.^[30] The point made in this 16section is that they may have significantly different connections to the harm, and to the risk. In fact, if we expand the world of activities to encompass management of risk, we see clearly that legal actions in relation to accidental harm often involve multiple parties with distinct roles. Not least, the pervasive involvement of insurers ensures that this is the case.

Along with risk management ideas come complementary principles in the law, other than causation and enterprise risk, aimed at determining whether it is appropriate for a particular party to be subject to a duty. This includes concepts such as assumed responsibilities and reasonableness of reliance. It is important to note that although this shifts us away from enterprise liability, these ideas are consistent with Keating’s core focus on the protection and security of interests within relations and interactions. With the expansion to ideas of risk management, tort law has developed heterogenous responses, including the idea that claimants should secure those interests themselves in certain circumstances. Tort reasoning therefore asks the question of who should have anticipated and guarded against the risk, including the party who has suffered loss, and whether there is justification for placing the loss with a party involved in an activity. This takes us beyond harms imposed by activities that are large in scale, to harms suffered within multi-party activities where it is thought that someone connected to that activity should have had the responsibility to provide security, whether through advice, avoidance, or insurance.

Thus, quite a lot of modern tort law is not preoccupied with risks that cause accidental harm, and responsibility based on those risks, but with the question of who had the duty to provide or to pay for security against risks, both within activities, and outside them. It is especially difficult here to distinguish between duties to take care to avoid harm, and duties to make good losses, because the courts themselves have sought to develop notions such as ‘assumption of responsibility’ which in some circumstances, associated with risk taking,^[31] make the test of duty depend on the acceptance of legal consequences should there be a failure. Claimants also may be more or less sophisticated in relation to risk. Are they purchasing property in order to invest? Or are they simply looking for somewhere to live?^[32] Tort must 17adapt its principles to new sorts of heterogeneity, of claimants and their role in relation to risk, as well as more nuanced relationships with other parties.

Tort here is securely connected to private law concerns, since it is concerned with the location of responsibility within interactions between parties. But its scope has been expanded to understand the involvement of numerous parties in these interactions or activities. Securing interests is itself a normal part of party interactions. Whether and against whom people have an entitlement to that security, through the law of tort, is not typically a question of whether liability should be strict, but often a question of how far the law of negligence is prepared to identify a duty to secure interests in different circumstances, where the burden of securing against harm is the very question in play. The role of discrete, perhaps careless ‘acts’ in causing harm can indeed seem to become a background issue in these contexts.

In other words, neither society nor the law of tort is necessarily as preoccupied by responsibility for causing accidental material harm, as is generally accepted in tort theory. Tort is also preoccupied with the limits of responsibility for securing interests in the face of a risk of wrongful acts, reflecting the social significance of those interests in a different way. This expands our notion of heterogeneity in tort. But I suggest it is also tied to the idea of activities and risk, and in an extended way to the idea of communities of risk, now perceived to include parties whose connection to the risk may be varied.

Thus, I suspect that the identification of a world of activities, as well as a world of actions, has broader (and more heterogenous) applications than those explored in Reasonableness and Risk.

VI Designing new liabilities

If we consider the design of liabilities for harm done by autonomous vehicles – an issue rightly mentioned by Keating as illustrating the need for change and adaptation within the domain of tort law and the flexibility that a focus on its role in protection of interests can bring – we will see that issues around the allocation of such liabilities within the class of those in various ways involved in the activities in question, including different ‘legal persons’, sheds additional light on the notion of activities and of communities of risk beyond the enterprise.^[33] I suggest that these developments, and the discussions around them, illustrate within a legislative pro18cess many of the issues confronting the common law in relation to risk just discussed. In the UK, as in many other jurisdictions, the design of liabilities has been subject to consultation and legislative action even as the technology for autonomous vehicles is developed (rather than reactively as with previous technological advances).^[34] This chronology and anticipation of issues is itself an indication of the greater awareness of risk and responsibility, and the possibilities of attributing different risk roles to different parties, in our contemporary legal thinking, compared to the days when Holmes was writing about the world of activities. Activities can be understood as mutually desired or beneficial and as involving a multiplicity of parties with different roles in relation to risk, while still understanding the overwhelming significance of one particular type of potential harm, namely personal injury. Nevertheless, UK legislation to this point resolves only certain questions, leaving others to be tested by future parties. The critical point perhaps is that a first, second and third resort are available to those who are injured by autonomous vehicles, and these are a single insurer for both driver and anyone else; the owner; or the driver if at fault. The potential liability of manufacturers is not excluded, but testing the boundaries of product liability in this way would fall to liable parties, and not to injured parties. Thus, the statute could be seen as an exercise in prioritisation.

In designing the liabilities that will attach to autonomous vehicles, fault is not the dominant idea, but nor is the idea of causation or the idea of enterprise responsibility. All of these however may influence the way that primary liabilities are followed up in the search for contribution or shifting of liability. The avoidance of uninsured loss, the avoidance of potentially crippling liabilities or of non-liabilities, and the insurability of liabilities and losses as well as the potential for reflecting product liability principles, are all component parts of the responses chosen. Notably though, the proposed solution is not a ‘compensation fund’ for harm done by autonomous vehicles, though this could have been a realisable solution. Instead, the method has been to determine a division of responsibilities that reflects the world of activities associated with these risks and the development of the technology that marks their difference from the existing regime. At the same time, the very existence of this process goes a long way to demonstrating, through the exercise in distinguishing the new technology, how far our current law relating to road traffic collisions is indeed a solution to the problem of an activity with many parties, and not simply (as it may superficially appear, and is presented in some more or less formalist tort theory) an instantiation of the fault principle in relation to distinct 19acts. The role of insurers as significant participants in debates about liability design, as well as primary sources of compensation, is itself significant. Undoubtedly, they will also have a key role in attempting further development in the law, including potential product liabilities.

Keating is right to offer the development of a new liability regime for autonomous vehicles as an illustration of the need for change and adaptation in the law of tort, and not as standing outside the law of tort. The response to a new set of issues and realisation that this challenges existing liability structures demonstrates that this is what they are, and shows us the complexity of the world of activities and the flexibility of notions of responsibility, but continues to concern itself with the interactions between parties. Those interactions are far richer – and more heterogeneous – than anything that could be captured purely in terms of the doing and suffering of harm or in terms of incentive and deterrence. And we can also see, I argue, that a theory of tort which had to consign all such adaptations to a different subject that is not tort, because it seems to rest on different theories of responsibility or because it stems from statute or from consultation, would indeed be an impoverished theory of tort.

The discussion of liability regimes for autonomous vehicles in this section is designed to reinforce the central point of agreement here: the role of tort, not the identification of a single form of responsibility within it, has a strong claim to be the foundation of successful tort theory. This is the better and much needed way to reflect the truth of heterogeneity of responsibility which can be understood as operating within private law.