The Law of Negligence, Blameworthy Action and the Relationality Thesis: A Dilemma for Goldberg and Zipursky’s Civil Recourse Theory of Tort Law

In this paper, I discuss Goldberg and Zipursky’s Recognizing Wrongs and argue that there is a tension between their philosophy of action as applied to the law of negligence and the idea that the directive-based relationality thesis is central and, therefore, the action and conduct of the defendant should not be part of the core explanation of the tort of negligence.

practices generate legal directives, whose content are indeterminate rights and duties. The institutionalised courts of law engage in the task of elucidating these rights and duties and issue directives (pp.46, p. 239 and pp. 254-259), and therefore rights and duties are neither general nor have their source in morality, but are rather a legal duty and a corresponding legal right of this specific plaintiff and this specific defendant once the courts have elucidated them.
The core arguments of Goldberg and Zipursky's Recognizing Wrongs, are drawn from ideas and published materials emerging from a lengthy engagement with theorists who advocate either corrective justice as foundational of tort law or the view that law can be reduced to economic analysis. 3 For the former group of theorists 4 the relationship between the plaintiff and defendant as formulated by the pair right/duty, which will be called the relationality thesis, is paramount and the action or conduct of the defendant cannot be severed from the loss or injury suffered by the plaintiff. By contrast, for the latter group of theorists, the defendant's action or conduct is severable from the loss or injury of the plaintiff. Furthermore, there is no intelligible or normative connection between them and, therefore, judgements of responsibility in tort law (for example, responsibility in negligence) lack any rational and principled foundation. Consequently, economic analysis theorists argue, tort law should be reduced to policy decisions made by judges whose main concern is to advance a scheme that guarantees a fair distribution of losses and gains. Like corrective justice theorists, Goldberg and Zipursky (p. 4,p.13,p.26,p.89 and p. 181) 5 argue that the relationality thesis is key to understanding and providing an intelligible unity to the complexity of torts or wrongs that are part of current common laws. Thus, defamation, nuisance, product liability, negligence, and privacy, all find a common core in the idea that a wrong has been committed and, therefore, a breach of a duty has occurred, which corresponds to the right of the defendant not to be injured or suffer loss. Unlike corrective justice theorists, however, Goldberg and Zipursky do not rely on the Kantian framework of rights and duties, conceptions of personhood or rational agency. 6 Goldberg and Zipursky ground rights and duties on the emergence of practice-based directives. This view will be called the 'directive-based relationality thesis'. In addition to the directive-based relationality thesis, Goldberg and Zipursky put forward the idea that theorists should also consider the role that courts have in intervening to solve the rightduty dispute. Goldberg and Zipursky aim to show that the law of torts empowers the plaintiff to redress the wrong suffered through civil recourse (p. 13, p. 15, p. 29, pp. 30-37, p. 42, p. 52, p. 72, p. 80, p. 9l, p. 113, p. 115, pp. 163-165). 7 It is an 'empowerment' as the plaintiff instigates the claim and can decide at any moment to abandon it. This additional feature further elucidates the directivebased relationality thesis within tortious relations. Thus, tort law is not a matter of a moral right that has been violated by the defendant's conduct and the corresponding moral right of remedy. It is a matter of a legal empowerment through a civil recourse conferred on the plaintiff by the law.
In philosophical terms, this theory addresses the question of the normative and justificatory force of tort law as specifically legal and not moral. On the normative and justificatory aspects, it relies on the directive-based relationality thesis; but the normative force of the directive-based relationality thesis emerges as a combination of legal practices and self-understanding of these legal practices (p. 15). There is neither, Goldberg and Zipursky tell us, the need to advance a Kantian framework of political and moral philosophy to give content to the relationality thesis and, consequently, to justify the intervention of the State, nor is there the need to use the Kantian framework to show the immanent formal features that are displayed in the correlation between a duty and a right in tort law. Furthermore, there is no need to resort to an empirical reality, such as vengeance (p. 65 and pp. 121-122), to explain the characteristic normative language used by the courts (p. 6 and p. 12), nor to advance an Archimedean conception of objective morality and moral truths.
Goldberg and Zipursky insist, nonetheless, that there is an immanence and an inner normativity of law (pp. 61-62 and pp.190-191), 8 but that it is always legal in character as provided by the empowerment of the law (p.2, p. 4 and pp. 86-89), and by the fact that it is the courts that elucidate the content of the right and duty.
II. THE DILEMMA My criticism of their rich and complex theory, however, will concentrate on the philosophy of action presupposed by Goldberg and Zipursky, which aims to show that action, or the conduct of the defendant, cannot be severed from the plaintiff's loss or injury. I will argue that there is a tension between their philosophy of action as applied to the law of negligence and the idea that the directive-based relationality thesis is central and, therefore, the action and conduct of the defendant should not be part of the core explanation of the tort of negligence. I infer that as their arguments refer to all aspects of tort law, they also apply to specific torts, i.e. negligence.
The criticism is formulated in terms of the following dilemma: If the directive-based relationality thesis is at the centre of the tort of negligence, then a description or explanation of the defendant's action and conduct should not be part of the core explanation of the tort of negligence. On the other hand, if the sound philosophy of action shows that we cannot sever the defendant's conduct from the plaintiff's injury, then the defendant's action might become part of the core explanation of the law of negligence. Therefore, the directive-based relationality thesis-in terms of the right-duty pair-as emerging legal practices is secondary to an explanation in terms of the defendant's action. Either it is the case that the defendant's actions and conduct should not be part of the core explanation of the tort of negligence or, the defendant's action and conduct is part of the core explanation of the tort of negligence and the directive-based relationality thesis is secondary to the former. Furthermore, the directive-based relationality thesis might be dispensable.
The intuitive puzzling idea that emerges from the dilemma is that within Goldberg and Zipursky's tort theory the role of a sound theory of action is unclear. If they were to advance a correct explanation of action, then this theory will have normative consequences and provide a grounding for tort law. This would require us to investigate carefully what action is, and what the conditions are for identifying and evaluating action. Furthermore, methodologically speaking, Goldberg and Zipursky's would be starting from an explanation of action as the 'core' explanation, and subsequently the directive-based relationality thesis will be derivative and follow as a normative thesis to the action theory. Instead, they start with the idea that actions cannot be severed from their results or consequences and they introduce this idea as the correct conception of action. There is no explanation of why this is so, no further argumentation or grounding is provided. The methodological route that Goldberg and Zipursky seem to follow is the opposite from the one suggested. Their conception of action derives from the directive-based relationality thesis that they defend. However, we have no reason to believe that this is a sound explanation of action.
If the premises of the dilemma are sound, then Goldberg and Zipursky would need to choose between the following two options: (a) embrace the directive-based relationality thesis, in which case they would need to ignore the argument that defends the view that action and injury or loss are not severed and, consequently, an explanation in terms of action becomes unnecessary and plays no central role in the tort of negligence; or (b) embrace a sound theory of action whereby action and injury or loss are not severed, in which case they would need to ignore the directive-based relationality thesis or at least explain how it is derivative of their sound theory of action as it will play no central role in the explanation and justification of negligence law. Both alternatives seem unpalatable. In the latter case, if they ignore the directive-based relationality thesis, the explanation should be carried out in terms of action, conduct, capacity and probably outcome-responsibility, 9 partially ignoring or suspending the question of rights and duties. In a weaker version of this option, they would need to show how the directive-based relationality thesis is derivative of or dependant on their sound theory of action. If they embrace the directive-based relationality thesis, they would need to ignore the question concerning action and conduct as non-severable from injury and loss as central or paradigmatic. This is because it would render mysterious the role played by this theory of action in a coherent explanatory and justificatory theory of negligence law; all the explanatory and justificatory work would be done by the directive-based relationality thesis. Inevitably, however, once we ignore or discard the idea that a defendant's action cannot be severed from the plaintiff's injury in describing negligent actions, problems related to moral luck and consequentialist puzzles will arise. It will not be clear why our legal and social practices, and self-understanding do not embrace 'moral luck' conceptions, consequentialist, or economic analyses of the law of negligence. Furthermore, as a counter-analysis, a more robust formulation of the right/duty pair would need to be provided to rescue the relationality thesis (or at a least a variation of it). Therefore, Goldberg's and Zipursky's theory might collapse into Weinrib's or Ripstein's corrective justice theory.

A. First Premise of the Dilemma
I will now explain in detail the two premises of the dilemma.
Premise 1: If the directive-based relationality thesis is at the centre of the tort of negligence, then the defendant's action and conduct should not be part of the core explanation of the tort of negligence.
In English law, the courts identify a number of key conditions that need to be satisfied for a legal action in negligence to be successful. First, the plaintiff needs to show that the defendant owes a duty of care to the plaintiff. This is determined by two key concepts, foreseeability and proximity. 10 Second, the plaintiff needs to demonstrate that there has been a breach of the standard of care, which is the standard of a reasonable person in the same position and circumstances as the defendant. 11 Third, the plaintiff should prove that the defendant's action caused the injury or loss and, finally, 12 that the type of damage, i.e. loss or injury, was not too remote. 13 The analysis in the US legal system differs slightly, as much focus is placed on the notion of risk. One of the factors that needs to be considered when determining whether the defendant has a duty of care to the plaintiff is whether the defendant engaged in the creation of the risk which resulted in the plaintiff's injury. 14 10 Donoghue However, arguably, in both jurisdictions as in many other common law jurisdictions, the primary element is that there is a duty that the defendant owes to the plaintiff and that the latter has a right not to be injured or suffer a loss. In both the English and US legal systems as in many others, 15 the standard of care is objective and the law of negligence becomes closer to a scheme of quasi-strict liability (p. 191). It does not matter whether the defendant has subjectively done everything that she or he could have done, i.e. whether they took all the precautions or steps to avoid the harm or loss that in her own assessment were necessary. The defendant will be liable if he or she has fallen below the objective standard of care of the reasonable person engaged in a similar action or activity. This 'quasi-strict liability' scheme has been the target of criticism by a number of theorists 16 who argue that there can be no blameworthy action as there is no knowledge that the defendant should have had, and therefore liability for negligent acts cannot be justified. In other words, the defendant was not in control of her actions as she was unaware or oblivious of what she was doing when she was doing it. Arguably, Goldberg and Zipursky's theory provides an answer to this sceptical challenge in the following terms. The conduct or action, including how the defendant conceived her action or the knowledge that she had when she was engaged in the negligent action, play no role in the justification and explanation of the law of negligence (pp. 248-249). 17 Goldberg and Zipursky point out that the correct description that identifies a negligent act is 'x negligently injured y' and not 'x acted carelessly, and y suffered a setback because of x's careless act'. They assert: To ground a claim for negligence, the plaintiff's injury must ordinarily be a realization of the aspects of the defendant's behavior 18 that rendered it careless: the plaintiff's injury must be capable of being cogently described as having an immanent within the defendant's carelessness. 19 Negligence, in other words, contains a directive that enjoins careless injuring, not careless conduct itself, nor even harm caused (in any manner) by careless conduct (pp. 248-249).
The legal directive establishes the nexus between the injury and the careless conduct. But the nexus is not in terms of an evaluation of the conduct, but in virtue of the right-duty pair and the directivebased relationality thesis. This is a paradoxical position because the role of the different criteria to evaluate conduct is now unclear, e.g. in English law foreseeability and proximity, and reasonable person as standard of care; in the US, reasonable risk. Because Goldberg and Zipursky do not provide a detailed and justified theory of action through which we could assess conduct or action in general, it is not clear how negligent conduct, philosophically speaking, can be determined and assessed. The philosophical and conceptual implication of phrases in this paragraph such as, ''the plaintiff's injury must ordinarily be the realization of the aspects of the defendant's behavior '' or, ''the plaintiff's injury must be capable of being cogently described as being immanent within the defendant's carelessness'' are unclear. What do 'being immanent' and 'realization' mean here? We will try to elucidate this in the next section when we engage with Goldberg and Zipursky's theory of action and the second premise of the dilemma.
For Goldberg and Zipursky, the core view that explains and justifies the tort of negligence is the directive-relationality thesis, i.e. the plaintiff's right not to be injured or suffer loss and the defendant's duties that are elucidated by the courts and are the content of practised-based directives of the law of negligence. The special 17 There is a specific conception of 'wrong' advocated by Goldberg and Zipursky which is not related to blameworthiness or conduct (pp.187-199). A conduct is 'wrong' when it is attributed as such by the courts in their interpretation of the legal material (p. 106 and 182). The argument underlying the view that conduct is irrelevant in negligence is that there is an objective standard of care, which is defined according to what a reasonable person would do in the specific circumstances (p. 104 and p. 107). Assessment of liability is in terms of what the defendant ought to have done, not what they actually do (pp.191-192). 18 The emphasis is mine. 19 The emphasis is mine.
relationship between the plaintiff and the defendant is also shaped by the empowerment in the form of a civil recourse that is conferred on the plaintiff by the State. According to this view, consequently, the question related to the defendant's subjective understanding and description of her action, or whether the defendant was a proximate cause of injury or loss suffered by the plaintiff is not at the core of the justificatory and/or explanatory framework of the law of negligence. The knowledge and conduct of the defendant and whether it was foreseeable that the action of the defendant will cause the harm seems irrelevant. The directive-based relationality thesis seems to do all the required work as the right-duty pair define what kind of action is a negligent one.

B. The Second Premise of the Dilemma
Premise 2: if the sound philosophy of action shows that we cannot sever the defendant's conduct from the plaintiff's injury, then the defendant's action might become part of the core explanation of the law of negligence. Therefore, the directive-based relationality thesis in terms of the right-duty pair is secondary to an explanation in terms of the defendant's action.
In spite of what has been said in Premise 1, Goldberg and Zipursky engage in an explanation in terms of a theory of action. Their theory of action is a key argument against moral luck views on action and consequentialists theories in tort law. The puzzle of moral luck in contemporary philosophy was first introduced by Bernard Williams and Thomas Nagel, and subsequently applied to tort law by Jeremy Waldron. In reply to Williams' paper ''Moral Luck'', 20 Nagel 21 asks whether we are responsible for actions that are beyond our control, e.g. circumstances, opportunities, capacities, temperament, contingencies in the world. According to Nagel, the success or failure of our actions depends on what is happening in the world when we act. Williams calls this 'moral luck'. The idea of the agent losing control of her actions due to contingencies in the world, including who we are and how we are constituted, undermines any standard notion of responsibility and moral judgement. As Nagel puts it, ''the self which acts and is the object of moral judgment is threatened with dissolution by the absorption of its acts and impulses into the class of events''. 22 But Nagel insists that a moral judgement is not what happens, it is not about a state of affairs, but about an active self. He merely outlines this active self in his piece on moral luck, but provides a clearer idea in The Possibility of Altruism 23 and The View from Nowhere 24 . The active self is capable of identifying and distinguishing between what belongs to us, when we engage in choosing and rational deliberation, and what is just a mere happening. Thus, we cannot take a merely external evaluative view of ourselves, and understanding our actions is key to shaping the contours and borders of what we have done, as opposed to what is merely happening in the world. We cannot, Nagel tells us, operate and make ourselves intelligible if we operate as causes in the empirical and contingent world. Contrast Nagel's position with Williams' view on the matter. Williams focuses on rebutting the view that we are immune to moral luck and that the agent's reflective assessment and justification of her own actions are not subject to luck. 25 The example of the painter Gauguin illustrates Williams' point. The painter abandons his wife and children to pursue a career in painting. His success as a painter, i.e. giving a unique artistic legacy to the world and defining the way human existence is meaningful, cannot be separated from any justification of his actions. But his success cannot be foreseen by him or anyone at the moment of his actions. If Gauguin fails, he has done the wrong thing; if he succeeds, his actions might be justified. 26 Of course, this justification is not moral, but it is within a life that adheres to certain values, i.e. a life that is meaningful through aesthetic experience. Williams recognises that this justification is not towards others, i.e. the children will still have grounds for reproach, but the justification will operate retrospectively. Williams' point is that at the moment of the action, Gauguin cannot act in light of all the relevant rational considerations that apply to him 27 because much of the justification of his actions will depend on his talents, deliberative capacity 28 and character, and how he successfully develops these talents. This is what Williams calls 'intrinsic luck'. 29 He tells us ''the locus of intrinsic luck, largely lie in him -which is not to say, of course, that they depend on his will, though some way''. 30 However, he recognises that the locus of intrinsic luck can lie outside the agent. Williams seems to hint to the idea that agency continues once we have performed our actions, when we look back and ponder over the consequences of our actions. 31 We can see these consequences as part of our actions, in spite of the fact that we may have had no knowledge of the action and did not intend it. According to Williams, we would expect that the lorry driver, who runs over a child by accident, has particular feelings about what happened, and that these feelings cannot be easily eliminated by arguing that what happened was not his fault. The lorry driver recognises that the consequence, i.e. the death of the child, has resulted from his act, i.e. his driving the lorry. 32 For Williams it would be an insane conception of rationality if we expected people not to feel or own the consequences of their actions, or if we expected them to detach themselves from the unintended aspects of their actions. 33 However, let us recall that for Williams, this evaluation is not in terms of moral blameworthiness as one cannot make clear judgements of liability and moral blameworthiness. The latter, in principle, requires control, but in cases of non-intentional action there is no control. 34 He concludes that agency cannot be purified of contingencies in the world and, therefore, any sound account of rationality and responsibility need to consider this impure conception of the self. Furthermore, he wishes to emphasise that our assessment and justification of actions is not exclusively from the moral point of view, where actions and consequences are severed, as Nagel aims to show. Let us now analyse how these ideas have been examined in tort law. Honoré and Gardner, 35 inspired by Von Wright's philosophy of action and norms, 36 outline an alternative conception that navigates between Nagel's and Williams' positions. Gardner, following Honoré's argument that responsibility in the law of negligence is about outcome-responsibility, establishes a distinction between results and consequences. Thus, results are within the description of the action and therefore are part of what the agent is doing and what we bring to the world as agents. By contrast, consequences are outside the scope of the action, and are merely 'happenings'. 37 Waldron 38 also takes up Williams' points on moral luck to establish a case for the replacement of negligent liability for a scheme where losses and gains are distributed in terms of adequate models of fairness and justice. Waldron aims to show that the law of negligence is basically unjust as we cannot justify liability in negligence for the alleged harm that the plaintiff has suffered as result of the defendant's action given that others may have acted in a similar manner, i.e. were equally morally wrong, but were lucky enough not to injure or cause any loss to anyone. His argument is well illustrated with the following example: Two drivers, named Fate and Fortunate, were on a city street one morning in their automobiles. Both were driving at or near the speed limit, Fortune a little ahead of Fate. As they passed through a shopping district, each took his eyes off the road, turning his head for a moment to look at the bargains advertised in a storefront window. In Fortunate's case, this momentary distraction passed without event. The road was straight, the traffic in front of him was proceeding smoothly, and after few seconds he returned his eyes to his driving and completed his journey without accident. Fate, however, was not so fortunate. Distracted by the bargain advertised in the shoe store, he failed to notice that the traffic ahead of him had slowed down. His car ploughed into a motorcycle ridden by a Mr. Hurt. Hurt was flung from the motorcycle and gravely injured. His back was broken so badly that he would spend the rest of his life in a wheelchair. Fate stopped immediately to summon help, and when the police arrived he readily admitted that he had been driving carelessly. 39 Like in Williams' case of the painter Gauguin, Waldron invites us to reflect on the complexity of action and its connection to blame and responsibility. If world contingencies and happenings are in- evitably merged with our actions, then it is difficult to justify from the moral standpoint that we are responsible in negligence, and moral luck seems to pervade our tortious actions. Similarly, if liability in negligence is grounded in our conduct and blameworthy action, then it is difficult to justify that Fate is liable whereas Fortunate is not. The moral viewpoint as defended by Nagel is challenged by this example. Fate and Fortunate conducted themselves in exactly the same way, and there is no justification for both attributing liability to Fate and justifying the transfer via compensation of 5 million US dollars from Fate to Hurt, which bankrupt him. Neither Fate nor Fortunate choose to act in the way they did.
We took a detour in order to understand the context within which Goldberg and Zipursky propose their conception of action. It aims to undermine Waldron's position and therefore, indirectly, any 'moral luck' type of analysis of action and consequentialist position on action theory, including economic analysis of law. But it also aims to put pressure on the moral standpoint of blame and responsibility as advocated by Nagel, for example, where an action is assessed only by virtue of what the active self engages with and, therefore, the self is isolated from results or consequences. They navigate between the Scylla of the moral standpoint and the Charybdis of the 'moral luck' position on action and consequentialist analyses. They argue that in tort law, including the law of negligence, the plaintiff's injury cannot be severed from the defendant's negligent action. According to Goldberg and Zipursky, Waldron's description of Fortunate and Fate formulate actions as a sequence of events in the world where contingencies occur. For Waldron, the description of the action that corresponds to Fate's performance of action would be ''the driver was driving the vehicle at high speed, the child jumped in front of the vehicle, the vehicle ran over the child and therefore the driver's action caused the injury of the child''. By contrast, the description of Fortunate's action would be ''the driver was driving the vehicle at high speed''. Under these two sets of descriptions of negligent action, we cannot understand why Fate is liable while Fortunate is not. They both performed the same actions, but Fate was unlucky due to contingencies of happenings that were beyond her control. By contrast, Fortunate was lucky in terms of happenings. It is unfair to establish liability in negligence on the basis of happenings, especially in negligence, where actions are non-intentional and caused by inadvertence, lack of knowledge or unawareness. It seems that sceptical positions have the upper hand here.
Goldberg and Zipursky aim to show that Waldron's description of negligent action is mistaken. In tort law and the law of negligence and, arguably, (though they do not put the point according to this exact formulation) from the point of view of the law of negligence and tort law entrenched in legal practices, the correct description of negligent action is as follows: ''the speedy driver carelessly ran over a child and this resulted in injury'' and this is, according to tort law, what Fate did. By contrast, Fortunate only ''drove carelessly at high speed''. Of course, from the point of view of the Road Traffic Act, 1988, both Fortunate and Fate violated the Act and the consequence will be a fine. However, within the perspective of tort law, the duty of non-injury is attached to the description of the act and in the example, only Fate violated the duty of non-injury. It is unclear how Honoré and Gardner's description of action in terms of results and not consequences sheds further light on, or perhaps complements, Goldberg and Zipursky's theory of action. I suspect that the key difference is as follows. Goldberg and Zipursky would like to make legal descriptions parasitic on legal concepts and the way these concepts are interpreted and elucidated by the courts. By contrast, Honoré and Gardner's theory of action might stand independently of interpretative legal practices.
If Goldberg's and Zipursky's view on action is sound, then the role of the directive-based relationality thesis is unclear. The justification can be done within this theory of action. The courts would only need to identify the correct description of negligent action, i.e. the plaintiff's action caused the defendant's injury.

C. The Conclusion of the Dilemma
Let us now analyse the conclusion of the dilemma: Either it is the case that the defendant's actions and conduct should not be part of the core explanation of the tort of negligence or, the defendant's action and conduct are part of the core explanation of the tort of negligence, and the directive-based relationality thesis is secondary to the former. Furthermore, the directive-based relationality thesis might be dispensable.
Goldberg and Zipursky need to make up their minds about the role that their theory of action ought to play within the justificatory and explanatory framework of the tort of negligence. If they embrace the directive-based relationality thesis, then the right-duty pair determines and defines the negligent act. A theory of action, therefore, as construed in terms of a description where the injury is part of the act performed by the plaintiff plays no role in the justification and explanation of the tort of negligence. Furthermore, the description of the action is merely a legal description, which is defined by the rightduty pair.
A theory of the right-duty pair will be sufficient to explain key features of the tort of negligence, and the interconnected idea of the civil recourse that is conferred by the State on the plaintiff would remain intact. Goldberg and Zipursky rely on historical and pragmatic arguments to give further flesh to their theory of rights. The so-called conduct theory of rights aims to defend the view that entrenched practices and the interpretations of courts elucidate the content of our rights and duties in tort law and the law of negligence. It is still open to debate whether this conception based on practices is sufficiently normative and robust to ground a theory of rights, which are typically characterised as non-empirical, practical and normative. By contrast, a conduct theory of rights might be contingent on practices, interpretative elucidations and self-understanding, and subject to historical conditions. 40 Additionally, an objector might raise the issue of the character of action and defend the view that action and injury should be severed. This is precisely Nagel's point. We cannot look at the performance of our action from the empirical perspective, where contingencies might undermine our self-understanding of what we are responsible for in terms of what we can control. However, contra Nagel, Goldberg and Zipursky do not look at the action from the empirical perspective. They try to show that the action ought to be looked from the law of negligence point of view. However, arguably, the legal point of the tort of negligence does not automatically convert this point of view into a correct normative conception of the law. It is, Nagel would argue, in conflict with our moral practices and moral understanding where blameworthy action plays a key role. Fur-thermore, if there is to be any meaningful normative language, e.g. duty, breach of a standard, this normative language would need to have some relationship with the moral perspective or moral point of view.
The opposite position, Waldron's and Williams' conception of action, puts further pressure on Goldberg and Zipursky's theory of action. Waldron and Williams would insist that our actions are inevitably merged with contingencies in the world, including our character, temperament, what we can see, know or be aware of at the moment of an action. The law of negligence cannot carve a conceptual space through which action and results are not severed, and at the same time, explicitly ignore the deeper reasons why they should not be severed. They cannot separate the non-severance of action and injury from its deeper rationality. We cannot justify cases like that of Fate as different from cases like Fortunate and it is unclear why this law of negligence or legal point of view should have a privileged position over the 'moral luck' view, the empirical and/or moral point of view. What is the character of this special normative perspective given by legal concepts and legal practices through which we make this distinction?
The idea of entrenched legal practices does not seem sufficient to ground the distinction, which by now might seem arbitrary. Goldberg and Zipursky would need to ignore the theory of action as it plays no role in the directive-based relationality thesis, though they will not be, therefore, armed with a theory of action that dismantles the pervasiveness of moral luck in negligent actions. They will indirectly open the path for reducing negligence law to consequentialist or economic conceptions of law.
As shown by the second horn of the dilemma, Goldberg and Zipursky could fully embrace their conception of action as a description of injury necessarily connected to the description of the negligent act that is performed by the defendant. This conception will explain why we are responsible for negligent actions, but as the primary explanatory and justificatory role will be in terms of a philosophy of action, we need to engage with an explanation of proximate causation, capacity, etc. The directive-based relationality thesis would subsequently be derivative or dispensable as all the work will be done by these key concepts. 41 Goldberg and Zipursky reject the latter argumentative strategy and insist that the directive-based relationality thesis and the idea of civil recourse conferred on the plaintiff are at the core of their explanatory and justificatory theory of tort law and, therefore, also negligence law. They fail, however, to explain the role of their theory of action in tortious actions.
I will now explore a potential response to the challenge of the role of their theory of action for tortious action and, more specifically, negligent action. Goldberg and Zipursky could argue that their theory of action is derivative of the directive-based relationality thesis. The latter and, therefore, the pair right/duty determines and defines what a negligent action is. The pair right/duty is, Goldberg and Zipursky could argue, a scheme of interpretation that gives intelligibility to the unity of bodily movements and injury/loss suffered by the defendant. They could state that the moral standpoint, 'moral luck' viewpoint, and economic analysis theorists fail to understand that the scheme of interpretation as providing intelligibility of action ought to be in terms of rights/duties. Goldberg and Zipursky's underlying argument would therefore be that the directive-based relationality thesis and the concepts of right and duties are the conditions of possibility that enable us to engage with the practice of negligence law. Ideas of right, duty and empowerment through civil recourse are historically entrenched in legal practices and are also currently practised by the courts of different jurisdictions. The right not to be injured and the duty of non-injurious action is a scheme through which we can identify and give content to the driving, and injury and loss caused by the speedy driver. If there is another speedy driver who does not injure anyone, then the scheme of interpretation does not apply to her.
The right-duty pair and, consequently, the directive-based relationality thesis are crucial to our self-understanding and to the practices of courts and legal practitioners of the law of negligence.
Furthermore, Goldberg and Zipursky could argue that 'moral luck' consequentialists and economic analysis theorists' notion of action are mainly empirical and fail to have any intelligibility from the point of view of the agent who engages in the action. On the other hand, Nagel's proposal is merely moral, but not legal. The speedy driver who collides with another vehicle and as a result injures the passengers in that vehicle does not see the injury as severed from his driving.
Arguably, however, Goldberg and Zipursky's hypothetical response would engage neither with Nagel's moral standpoint, nor with the 'moral luck' or economic analysis viewpoint. 'Moral luck' and economic analysis theorists could argue that it is true that the speedy driver does not refuse to appropriate and be responsible for the injuries that he has caused to the plaintiff by his negligent driving. However, speedy driver's self-understanding is confused and muddled by the fact that another speedy driver was lucky and did not actually cause any injury, even though her conduct was equally negligent. In Waldron's example Fate will ask why she ought to pay for losses that she caused through actions she was unaware of, when Fortunate, who behaved exactly like her, did not cause any injury and therefore is not liable simply because she was lucky. The fact that Fortunate did not cause injury was merely 'good' luck 42 . This is precisely Williams's point; it would be insane to advocate a theory of rationality and responsibility that would ask for a purified conception of agency, and that would isolate us either from the contingencies of the world or from a purely normative legal conceptual scheme though it is the result of practices. We can still question it, and the fact that is entrenched in practices does not give any special status or consolation to Fate's pressing question. Fate needs to understand why her action is truly blameworthy. Judgements of liability against her and not against Fortunate contradict our self-understanding. There ought to be a symmetry between Fortunate's and Fate's attributions of liability. Self-understanding can go both ways and might undermine a scheme of interpretation in terms of the right-duty pair onto negligent action.
Goldberg and Zipursky also fail to see the moral standpoint. The moral standpoint resists the idea that descriptions of actions should merge with contingencies of the world which we cannot control. Legal practices show a continuity with moral normativity at the level of the interpretative tasks of the courts. This position is clearly defended by Goldberg and Zipursky. But their theory of action seems to isolate legal normativity from moral normativity. Moral self-understanding can show its ugly head in our legal self-understanding either through the interpretative engagement of the courts or the practices of legal practitioners. Furthermore, the application of a scheme of interpretation of right-duty comes after the specific negligent action has been performed. It does not help to guide us in our actions, when we are engaged in reasons and tracking good-making characteristics as it is a mere scheme of interpretation.

III. CONCLUSION
In this paper it has been shown that there is a tension between Goldberg and Zipursky's theory of action, which defends the view that negligent action and injury should not be severed, and their directive-based relationality thesis. The role of their theory of action is unclear since all the explanatory and justificatory theoretical work is done by the directive-based relationality thesis. I have also tried to demonstrate that either the moral standpoint of action, or the moral luck position on responsibility might appear in our self-understanding of negligent action. Thus, the possibility of a pure law of negligence with its own normative point of view is problematic as courts in their interpretative tasks engage with both the moral standpoint and moral luck analysis to determine liability in the law of negligence. The moral standpoint or a moral luck type of analysis might show the 'impurity' of action and undermine the law of negligence normative point of view that justifies liability. We need further explanations to purge the impurities of a moral standpoint or moral luck analysis in our legal assessment of liability in negligence. It seems that a political or moral philosophical position beyond the notion of historically entrenched practices might be necessary.

OPEN ACCESS
This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and THE LAW OF NEGLIGENCE, BLAMEWORTHY ACTION reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creativecommons.org/licenses/by/4.0/.

University of Surrey Centre for Law and Philosophy, Guildford, England, UK
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.