A. Legal Evolution

Many judicial decisions are not about what tort law should be but about its application to concrete factual patterns: for instance, whether a touching of a body is intentional, a limitation of liberty is coercive, or a certain surgical method is reasonable. While deciding what is intentional, coercive, or reasonable is not purely technical and involves value judgments, a basic premise underlying tort law and its theory is that the connective tissue linking such norms and their underlying common law principles offers sufficient normative guidance to resolve cases, even when they fall outside settled precedent. Therefore, when judges apply such norms and when theorists explain them, they can focus on legal materials without needing to rely on external normative inputs—for example, from morality or economics.

However, while tort law’s operation takes up much space in its theory and practice, it also constantly evolves in response to society’s needs.Footnote ² As Scott Hershovitz observes, tort law is “one of the ways that we work out what we owe each other”; and because “our views about that are bound to change”—with political, social, cultural, economic, scientific, and other changes—“tort law is bound to change too.”Footnote ³ Indeed, in the past century or so, tort law has been far from static:Footnote ⁴ tort norms have been abolished (e.g., against the enticement of spouses), replaced (e.g., by workers’ compensation schemes), introduced (e.g., against invasions of privacy), and expanded (e.g., negligence law’s coverage has extended from physical harms to body and property to include pure economic and emotional harms).

Negligence law is especially illuminating in this regard: its open-ended nature—specifically, its duty element, which determines who is obligated to whom—exposes the pressures exerted by policy reasons, which other torts often better conceal. Its flexibility and transparency have made it a hotbed of legal evolution and therefore of theoretical and practical disagreement. Its evolution along two normative paths therefore highlights general features of tort law’s evolution.Footnote ⁵ On one path, negligence law recognizes general duties not to unreasonably harm others’ bodies or property. The foreseeability of such physical harms is normally treated as sufficient to justify duties of reasonable care. The landmark cases on this path are MacPherson and Donoghue:Footnote ⁶ they replaced a regime that allocated the risks of harm from defective products through contracts with a mandatory regime, thereby imposing duties of care on manufacturers even toward consumers with whom they lack contractual privity—and crystallizing moral norms into legal baselines that constrain markets.Footnote ⁷ Negligence law’s second evolutionary path focuses on interests that are too complex to justify general duties based solely on the foreseeability of harm: duties not to cause pure economicFootnote ⁸ or emotional harm,Footnote ⁹ or to protect against harm by third parties,Footnote ¹⁰ are therefore special—that is, imposed only when duty-bearers assume extended responsibilities toward particular others who rely on their care, such as on universities to students or on doctors to patients.

B. Claiming Authority

These paths, to which we will return, highlight the fact that tort law’s evolution involves a degree of discretion and therefore an exercise of authority, and not only in legislative arenas. Tort law must resolve normative conflicts that arise within its regulatory domain—its unique place within the legal system—not only effectively but also legitimately. Even if tort norms are not moral norms,Footnote ¹¹ their validity does not depend on their moral justification, and there are often good reasons for them to deviate from morality,Footnote ¹² they must still, to be good law, purport to obligate us not only legally but also morally: to tell us what we really ought to do—which requires sufficient congruence with moral principles.Footnote ¹³

Tort law’s claim to authority is often obscured because its regulatory domain does not center around a single normative phenomenon like crime or contract; it obligates us to treat others in certain ways, yet the forms of treatment and the list of protected interests are diverse and dynamic.Footnote ¹⁴ However, tort law’s domain has distinctive features that shape its claim to authority: unlike public regulation, tort norms are enforced by right-holders who decide when and how to call duty-bearers to account; and unlike forms of private ordering, tort norms are mandatory and duty-imposing rather than power-conferring. These features explain why tort law operates under simultaneous pressures from moral forces concerning both voluntarism and vulnerability. On the one hand, tort law must justify its imposition of duties by highlighting the duty-bearer’s responsibilities toward the right-holders to whom they are accountable.Footnote ¹⁵ It does so by relying on concepts such as intention, reliance, and foreseeability, which give duty-bearers reasonable control over their normative situation.Footnote ¹⁶ On the other hand, tort law cannot be reduced to the mere identification of responsibilities: it not only reveals that agents are responsible in light of how their actions fit under the current legal regime but also changes the legal regime itself to make agents responsible,Footnote ¹⁷ often in response to vulnerabilities or to large-scale policy goals.

In evolutionary moments—including cases like MacPherson and Donoghue—tort law solves social problems in a regulatory and authoritative manner by determining “who is best doing what and seeing to what, and hence answering for what.”Footnote ¹⁸ As negligence law’s expansion from duties not to cause physical harm to duties not to cause emotional or economic harm and to protect demonstrates, tort law can change the concepts and baselines by which our mutual responsibilities are evaluated. While some concepts and baselines (like factual causation or the human body) are more robust, there is often room for normative recalibration: for instance, when it comes to concepts like proximity, foreseeability, or injury; to interests such as dignity, privacy, or property; or to the definition and delineation of risks.Footnote ¹⁹

To borrow H.L.A. Hart’s taxonomy of uses of the concept of responsibility, it operates in tort law not merely as a basic capacity, a causal conclusion, or a liability, but also as a cluster of demands attached to roles (or tasks, undertakings, etc.) that require ongoing attention and care.Footnote ²⁰ This process is less visible in relation to general and negative duties, which apply to all social activities. But even battery, for instance, means one thing at the office and another during football practice; and there are strong pressures to evaluate the foreseeability of physical harms, and thus shape duties of reasonable care, while focusing on roles within activities, such as driver or manufacturer.Footnote ²¹ These pressures to attach responsibilities to roles are even more visible when it comes to special duties in negligence law—for instance, not to cause pure emotional or economic harm or to protect—which attach almost exclusively to roles such as accountant, university, or doctor.Footnote ²² Therefore, as Dilan Esper and Gregory Keating note, duty rules in negligence law must “articulate the more particular standards of care owed by certain well-defined social roles and activities … or incurred by certain undertakings.”Footnote ²³

By targeting roles within activities tort law not only reduces information costs and gives private agents greater control over their normative situationFootnote ²⁴ but also affords itself more moral room to assign mandatory duties, also in response to policy reasons. Importantly, this ability to assign responsibilities by imposing duties while deviating from existing normative trajectories means that even if it were possible to explain tort law’s ongoing operation without recourse to policy reasons, such reasons cannot be ignored when explaining why tort law stands where it does or where it should stand: to be normatively robust, such explanations must address the nature and limits of tort law’s ability to evolve in response to vulnerabilities and policy reasons.Footnote ²⁵ This requires venturing beyond the norms that tort law happens to recognize and the juridical principles that underlie them and even beyond social conventions.Footnote ²⁶ However, while common tort law’s inner principles cannot provide a correct and fully justified answer to every question about its evolution—a Dworkinian position that Hart famously referred to as a “noble dream”—it must also not collapse into Hart’s “nightmare”: the realist vision of law as entirely malleable.Footnote ²⁷ Authority and discretion must be exercised in developing tort law, but within justified moral limits—which tort theory must articulate and explain.

C. Moral Foundations

Normative explanations of the nature and limits of tort law’s evolution (which move beyond mere description) must address both policy and relationality. On the one hand, because tort law is a system of mandatory legal norms, its structural and allocative effects must be justified as an exercise of state authority.Footnote ²⁸ On the other hand, because tort law is rights-based, the state’s authority to use it must also be shaped by relational moral obligations: by what we owe to each other as private agents. Tort law’s foundations must provide it with tools to reconcile these pressures in a principled manner. This section highlights three typical failures in mainstream tort theory to achieve this: agnosticism, monism, and open-ended pluralism.

The first failure expresses the fact that normative explanations of tort law’s evolution cannot be agnostic about the values it pursues. The immediate suspects here are economic theories, which see tort law’s evolution as guided by its ability to incentivize socially desirable behavior. Importantly, they either reduce desirability to morally questionable values (such as social welfare) or treat it as extrinsic to tort law’s foundations. At best, moral principles are assigned instrumental and derivative roles: for instance, rights of action should be recognized if private enforcement is more socially efficient than public enforcement.Footnote ²⁹ This instrumentalism has led to criticisms that economic theories fail to explain tort law’s relational structure and the basic concepts operating within it,Footnote ³⁰ but the problems are also normative: economic theories lack the machinery needed to explain how tort law can and should evolve as a morally legitimate system of rights and duties because they reduce this area of the normative universe to mere prices and sanctions as causal levers.

Importantly, such prescriptive lethargy can also afflict theories that take tort law’s relational structure more seriously. For example, John Goldberg and Ben Zipursky offer an insightful theory of tort law as a system of civil recourse for legal wrongs. However, because they hold that tort norms draw their force solely from their legal validity rather than from any reliance on morality,Footnote ³¹ and voice concerns about policy reasons only in judicial decisions, which they see as incrementally elucidating existing norms,Footnote ³² they place few substantive constraints on the possible content of tort law’s primary norms.Footnote ³³

A second type of failure involves confining tort law to a narrow spectrum of moral reasons, while excluding policy reasons. Corrective justice theories are notorious in this regard. They often seem morally inert (say, when claiming that private law’s only goal is to be private law),Footnote ³⁴ or overly focused on litigation and remedies—to the point of having little to say about the content of tort law’s primary norms and even about its (legislative) abolition.Footnote ³⁵ However, they do establish some limits: for example, by excluding justifications for primary norms that do not fit within the correlative structure of the rights and duties of formally free and equal parties,Footnote ³⁶ or that conflict with the principle that no person is in charge of another in terms of their rights over body and property.Footnote ³⁷ The state can pursue policy goals through institutions regulated by public law, but the distributive baselines set by them allow tort law to hold private agents responsible for their actions in isolated interactions (hence, for example, a poor person may be obligated to compensate a rich person).Footnote ³⁸ Mandatory and privately enforced rights that respond to policy goals—for example, against discrimination—are therefore seen as parts of the background public law regime.Footnote ³⁹

This isolation of tort law’s operation and evolution from policy goals raises not only descriptive concerns (it struggles to explain many areas of tort law, such as class actions or strict liability) but also normative ones.Footnote ⁴⁰ It assumes that introducing policy reasons into tort law is inherently opposed to its relational structure—but as we will see, it is not only that this structure can accommodate such reasons, but that doing so is necessary for it to be morally legitimate: tort law cannot be oblivious to its structural and allocative implications. Thus, some degree of pluralism in terms of tort law’s goals is necessary.Footnote ⁴¹ Importantly, this pluralism cannot be superficial: it is not enough to concede that tort law just happens to produce good policy results by operating as a relational system of corrective justice or civil recourse.Footnote ⁴² Accounts of tort law’s foundations must explain how it can respond to policy reasons in ways that meet basic standards not only of interpersonal morality but also of political morality.

This takes us to the third type of failure: open-ended pluralism. It is not enough to see policy reasons and relational reasons as parts of a fortuitous and decentralized process in which so many reasons are involved that no general theory can explain or guide it.Footnote ⁴³ While contextualism is a virtue of morally pluralistic theories,Footnote ⁴⁴ some structure is necessary to avoid ad hoc, unprincipled, and intuition-driven balancing—which undermines not only rule of law values like publicity or stability but also substantive moral values, which must be engaged and realized through processes reflecting their intrinsic qualitative features. The imperatives of comprehensiveness and depth, while not requiring a complete political theory of tort law, demand conflict resolution principles that reach adequate degrees of reflective equilibrium by engaging a sufficient variety of judgments and principles in the central cases.Footnote ⁴⁵ This rules out some overly generalized claims—for example, suggesting case-related priorities: say, a focus on relational reasons in disputes between individuals and on policy reasons in disputes that involve corporations.Footnote ⁴⁶ Such distinctions are too rigid: policy reasons should often shape small-scale disputes and relational reasons should shape large-scale disputes—and, more importantly, policy reasons and relational reasons must often interact within the same case.Footnote ⁴⁷

Crucially, we must not only take all the relevant reasons into account in all casesFootnote ⁴⁸ but also justify the place and priority of each type of reason in each type of case.Footnote ⁴⁹ This poses problems for accounts of tort law that appreciate both its political and its relational aspects and the need to reconcile them, but are not systematic enough: for example, for claims that policy reasons determine which rights tort law must protect while relational reasons determine how,Footnote ⁵⁰ or that relational principles can be defeated by policy reasons as long as the former are properly engaged with as tort law’s deontic core.Footnote ⁵¹ Such accounts do not offer a reasoning scheme that is comprehensive and structured enough to resolve the ongoing conflicts between relational reasons and policy reasons in a principled manner: that is, a coherent and morally robust system of priority, exclusion, and reconciliation principles that responds to the nature and limits of relational reasons and policy reasons in different regulatory contexts.

A. The Structural Layer

Rawls’s theory requires little introduction. As part of the social contract tradition, it unpacks values inherent in the political relationship in which free and equal citizens justify their collective use of coercive power to each other.Footnote ⁵² Reciprocity and respect require the principles of social justice (for Rawls: basic liberties, equal opportunities, and equality unless inequalities serve the worst-off) to be justified in terms that reasonable citizens can accept: by placing idealized deliberating citizens behind a “veil of ignorance,” Rawls aimed to ensure that moral reasoning reflects social solidarity and is not distorted by historical contingencies.Footnote ⁵³

When thinking about tort law in the Rawlsian scheme, a natural place to start is his focus on the basic structure of society: the main political, social, legal, and economic institutions, as a unified and dynamic system.Footnote ⁵⁴ This focus serves, first, to avoid the libertarian reduction of the political to the relational. The profound and pervasive impact of basic institutions justifies the political collective’s complex responsibilities and obligations, which extend above and beyond those of any individual, to act through this structure to mitigate the unjust accumulated results of uncoordinated private actions: to respond to societal patterns involving relationships, practices, attitudes, habits, and so on, in order to preserve “background justice.”Footnote ⁵⁵

The theoretical debate about tort law’s subjection to social justice principles has often been framed in terms of whether it is part of the basic structure of society: for some theorists, because it is mandatory, it is part of the basic structure and thus fully subordinated to social justice;Footnote ⁵⁶ for others, its relationality locates it outside the basic structure and isolates it from social justice.Footnote ⁵⁷ We should resist the temptation of such binary claims.Footnote ⁵⁸ As Rawls noted, for the basic structure to sustain social justice in response to dynamic social conditions, its boundaries cannot be defined too sharply or rigidly.Footnote ⁵⁹ Thus, while tort law’s mandatory and duty-imposing nature and profound and pervasive impact on social spheres like markets, healthcare, and transportation provide strong reasons to see it as part of the basic structure,Footnote ⁶⁰ the important debate is not about whether social justice should shape or limit tort law but about how it should do so. In fact, answers to the latter question can ameliorate some of the pressure on the former by alleviating anxieties over subordinating tort law to imperialistic social justice imperatives.

This leads to another reason for focusing on the basic structure: avoiding the utilitarian subordination of the interpersonal to the political. Social justice risks entering this terrain if it is viewed as comprehensive enough to dictate tort law’s content and as exclusive in the sense that competing moral principles are irrelevant, excluded, or subsumed under it.Footnote ⁶¹ Although Rawls initially argued that social justice determines both the principles of social justice and all subsequent “criticism and reform of institutions,”Footnote ⁶² after his “political turn” (his retreat from extending his theory interpersonally) he clarified that social justice regulates only constitutional essentials and matters of basic justice—and therefore has little to say about many of the problems that regulatory bodies must routinely address and resolve.Footnote ⁶³

Social justice operates at the distributive rather than allocative level: it addresses the ongoing operation of the basic structure as a unified system rather than concrete allocations of benefits and burdens.Footnote ⁶⁴ It therefore not only evaluates basic institutions systematically—we achieve little, for example, if tort law successes are offset by backlashes in other parts of the basic structure—but also remains agnostic about many of their features:Footnote ⁶⁵ the normative data needed to determine the structure and content of many substantive, procedural, and remedial norms is too fine-grained to figure in general social justice principles.Footnote ⁶⁶ Social justice applies to choices between entire “systems of cooperation,”Footnote ⁶⁷ and thus lacks the tools or disposition to address allocations of goods and burdens through particular actions or norms.

Importantly, these limitations do not invite the development or extension of social justice. Rather, they demonstrate that it is not all there is to justice in society or even within the basic structure. An important claim in Rawls’s later work is that while social justice principles constrain or limit “local justice” principles—which express the nature, aims, and demands of different institutions, associations, practices, or relationships—they do not determine them.Footnote ⁶⁸ Private agents, unlike basic institutions, are not required to pursue social justice directly: basic institutions impose more concrete individual duties, which also respond to local justice values that focus on how private agents treat others in their “moral neighbourhood.”Footnote ⁶⁹

This pluralism is reflected in the four-stage sequence through which social justice is unpacked and applied to institutions: as we proceed from the original position to constitutional, legislative, and application stages we deal with progressively more concrete reasons and more specific standpoints.Footnote ⁷⁰ This sequence also responds to concerns about normative theories or systems that, absent clear methods of resolving conflicts, must resort to ad hoc and intuition-based decisions.Footnote ⁷¹ In this sequence, deeply political issues addressing the political collective’s self-organization—such as the electoral system or control of the means of production—are addressed at the higher levels of generality: as social justice principles or constitutional norms. While tort law is constrained by the products of reasoning at these stages, it clearly raises questions that extend beyond them. It also addresses what we owe to each other more directly—not through the basic structure: for instance, in terms of defamation, nuisance, or negligence. Tort law must reconcile not only claims against the political collective in institutional form, but also interpersonal claims. This requires reasoning at lower levels of generality, represented by the legislation and application stages.

B. The Allocative Layer

Scanlon’s theory is particularly apt for structuring the legislative stage of the sequence. It is more relational and less political than Rawls’s theory in three basic senses: it is about what we owe to each other as individuals; it models moral reasoning on a hypothetical conversation between individuals; and a relationship of mutual recognition and respect is an ideal that structures moral justification and motivates conformity—we follow moral principles not just out of a general sense of duty but because of the importance of standing in this relationship.Footnote ⁷² This relationship need not involve familiarity or history: compliance with moral principles can constitute a valuable way of relating to others in any “context of potential interaction structured by legitimate expectations,”Footnote ⁷³ even with strangers (which is significant for tort law).

Standing with others in this moral relationship entails that following a moral principle that someone can reasonably reject is wrong. Driving speed limits, for example, are enforced by state agents, but are not only about what we owe to people through institutions as part of the political collective—they are also about what drivers owe to others. Limits are justified when drivers cannot reasonably justify more lenient rules to pedestrians or fellow drivers. Similarly, tort duties not to defame, batter, or harm negligently, can express each duty-bearer’s respect for the potential victims of such actions. Thus, choices between regulatory regimes must reconcile both vertical personal claims against society and horizontal personal claims of individuals as the targets or beneficiaries of regulation. This person-by-person comparison of claims allows Scanlon’s scheme to explain and justify deontological principles:Footnote ⁷⁴ for example, even if imposing burdens on an individual would benefit many others, it may be unjustified if each beneficiary must acknowledge that the victim’s personal claim is comparatively stronger.Footnote ⁷⁵

In the aforementioned senses, Scanlon’s scheme is more relational than Rawls’s. However, it is nonetheless general and systematic in basic respects—echoing consequentialist themes—rendering it well-suited for structuring the legislative stage (and making the “allocative” label apt). It does not focus on particular people affected by specific actions but on the general standpoints of everyone affected by the adoption of principles for the regulation of society (including members of future generations):Footnote ⁷⁶ we reason about “what the social world will be like,”Footnote ⁷⁷ or “how there is reason to want people in general to go about deciding what to do.”Footnote ⁷⁸ Objections may concern not only the direct results of actions but also the behavioral social patterns they permit or demand: for instance, the psychological burdens of constant self-monitoring.Footnote ⁷⁹ In this vein, our substantive responsibilities under this scheme are shaped not only by how we choose but also by the justification of the conditions under which choices are made.Footnote ⁸⁰ Aggregation, in turn, is permissible when competing claims are on a par;Footnote ⁸¹ and a failure to aggregate when it is justified wrongs all affected parties—even those who happen to benefit (say, because they belong to the smaller group of claimants).Footnote ⁸² Given these systematic and general features, it is unsurprising that Scanlon’s scheme has been described as a “characterization of morality in public policy choice,”Footnote ⁸³ as “quasi institutional,”Footnote ⁸⁴ and as “legislative.”Footnote ⁸⁵

This normative key is well suited for choices between regulatory regimes: for example, between negligence law and public insurance. It is important for tort law because, as Gregory Keating notes, while tort law operates by recognizing rights and duties,Footnote ⁸⁶ the choice of whether to use it is collective, and political, and must balance competing claims to liberty and security on a social scale.Footnote ⁸⁷ While Keating also relies on a Rawlsian notion of fairness,Footnote ⁸⁸ his approach is more Scanlonian, given its focus on personal claims concerning the allocative effects of different regulatory regimes rather than political claims concerning the basic structure as a whole.Footnote ⁸⁹ This focus on allocations of the risk of harm in society was criticized by John Oberdiek, who claimed that, like consequentialist theories, Keating’s theory cannot account for relational wrongdoing and the bilateral structure of tort suits.Footnote ⁹⁰ Oberdiek builds on Scanlon’s idea of justification to a person to emphasize that tort rights and duties “are a function of what can be justified between individuals, full stop,”Footnote ⁹¹ and that the unjustifiability of tortious wrongs is “explained by reference to the wronged individual’s personal reasons.”Footnote ⁹² These ideas are, however, more relational than Scanlon’s scheme. As Scanlon concedes, his theory does not address rights and directed duties:Footnote ⁹³ its generality lends itself more easily to explanations of wrongness than of wrongdoing,Footnote ⁹⁴ which requires a greater focus on our relationships with and justifications to people directly affected by our actions.Footnote ⁹⁵ In other words, Scanlonian justification is also addressed to the world at large, even if it carves it up differently from how Rawlsian justification does, focusing on personal claims about allocative principles rather than political claims about the basic structure. When it comes to the duties of private agents, the allocative normative key better describes public regulation: for example, rules about the permissibility of pharmaceutical trials or the minimal solvency of banks. Legal norms about such matters constrain private agents for the benefit of others, but private agents are accountable for compliance to state agents, acting on the collective’s behalf, whose enforcement decisions must consider the interests of all affected parties.

Yet Oberdiek rightly claims that because tort law regulates by recognizing rights and duties, its explanation and justification cannot remain at the allocative level of reasoning. While its allocative implications, like its structural ones, cannot be ignored, its evolution must also be shaped and constrained by the relationality of its ongoing operation. In this sense, the more pressing issue (which Keating gestures toward by highlighting the distributive grounds of the interests that tort law protects by recognizing rights) is not determining the correct layer of reasoning, but their integration. In fact, concerns about the insufficient relationality of allocative reasoning can be alleviated by appreciating its continuity with more relational contractualist theories—located at the application stage of the four-stage sequence.

C. The Relational Layer

Relational contractualist theories highlight the ties between interpersonal claims and the force of moral obligations.Footnote ⁹⁶ These ties are important because, as Joel Feinberg explains, the special significance of rights stems from their recognition as valid claims within normative practices.Footnote ⁹⁷ Thus, when practices like tort law recognize rights, they assume the normative baggage of claiming and therefore more interpersonal requirements entailed by the idea of mutual recognition and respect, which more relational contractualist theories explain.

In Darwall’s terms, moral obligations are not third-personal and impartial or first-personal and self-referential but second-personal and other-regarding. They operate within an inherently relational normative framework: notions like standing, authority, accountability, responsibility, and reactive attitudes “comprise an interdefinable circle” into which there is no breaking from the outside—moral obligations may enter this circle only if they address its constitutive concepts and their interrelations coherently and on their own terms.Footnote ⁹⁸

Such relational structures take looser forms when duties are not owed to any particular agent and are tied to the representative authority of the community to call agents to account for doing wrong, as is the case in Scanlon’s theory and public regulation. They take tighter forms when duties are owed to particular individuals and are tied to their authority to call others to account on their own behalf, as is normally the case in private law.Footnote ⁹⁹ This tightness reflects the fact that when we are obligated to people whose interests are central to the justification of our obligations and who stand to be wronged by how we discharge them, disregard for the duty is disregard for these people as sources of valid claims.Footnote ¹⁰⁰ Wallace distinguishes in this regard between Scanlon’s (more systematic) focus on what “makes actions right or wrong” and a focus on “what it is for actions to be right or wrong,” which involves concepts like claims, obligations, and wrongdoing: “elements in the relational conception of moral requirements,”Footnote ¹⁰¹ which fall out of but are not addressed by Scanlon’s theory.Footnote ¹⁰² Crucially, though, what it is for actions to be right or wrong has upstream implications for what makes actions right or wrong—that is, more relational moral factors shape and constrain allocative reasoning.

The important point in this regard is that the relational normative structure of rights-and-duties relationships shapes and constrains the possible content of these norms. Even if their justifications rely on reasons external to such relationships, tied to more general standpoints, for these norms to operate within such relationships their content must adequately tie together the claimant’s standing and authority, the obligor’s responsibility and accountability, and attitudinal and remedial responses to wrongs. Consequentialism is often criticized for having little room for this normative structure: our duties are about others but are not really owed to them and a distance remains between the justifications of duties and the basic components of moral relationships—say, the right-holder’s authority to call the duty-bearer to account on their own behalf.Footnote ¹⁰³ Duties are, at bottom, owed to the collective.

Contractualism not only accommodates such relational structures but also explains their reliance on more abstract relationships. Because contractualism locates rights and duties within interconnected normative practices,Footnote ¹⁰⁴ it highlights the moral pressures on such practices to foster and sustain moral relationships of recognition and respect while focusing not only on direct interactions but also on the basic structure of society and allocative institutions and practices. Thus, tort law’s ability to empower right-holders to hold duty-bearers accountable and to guide and constrain duty-bearer actions that affect right-holders depends on this careful integration of the three normative layers. From this perspective, tort law rights and duties do not operate in a detached and technical manner. Rights empower their holders not only by erecting sanction-backed boundaries but also by structuring processes through which right-holders shape their relationships with duty-bearers; and right-based duties guide not only by requiring or prohibiting specific acts but also by requiring their bearers to internalize their requirements, regularly reflect on them, and monitor and modify their conduct accordingly.Footnote ¹⁰⁵ This relational refinement of the contractualist scheme at its point of application is crucial for tort law, as the next sections demonstrate: because tort norms operate within rights-and-duties relationships in different social spheres, policy reasons must draw upon relational reasons to gain weight in decisions about the content of tort norms and may express their more general and systemic features only as negative reasons not to adopt certain tort regimes.

A. Relational Baselines

Prescriptive tort law reasoning must rely on relational baselines to get off the ground: structural and allocative reasons are not sufficient to justify tort norms (even if they possess veto power). This claim is in tension with a few basic features of tort law, which the contractualist scheme highlights. Tort law’s mandatory nature allows it to assign responsibilities that shape and are shaped by other segments of the societal “responsibility tapestry,” such as professional ethics, religion, or morality.Footnote ¹¹⁹ Thus, tort law can concretize abstract moral norms and resolve moral conflicts—for instance, between our personal responsibility to others and our responsibility to ensure fair compensation for certain harms as participants in allocative practices or as members of the political collective—thereby redrawing the boundaries of our agency.Footnote ¹²⁰ Concern and respect for those affected by our actions are not limited to our direct interactions but require us to consider their interests from more general social standpoints and perspectives.Footnote ¹²¹ We must calibrate our actions according to public standards of conduct, including legal ones—even if they are objective or strict—as long as we retain reasonable control of our normative situation: say, if we can avoid assuming certain roles or purchasing liability insurance.Footnote ¹²² Crucially, despite this political basis, failures to meet these standards can amount to relational wrongs,Footnote ¹²³ and thus justify remedies and reactive attitudes like blame—even if in forms more remote from censure, resentment, or anger,Footnote ¹²⁴ and closer to a public notion of mutual accountability.Footnote ¹²⁵ This ability to ground accountability and remedial mechanisms in public standards is a core aspect of tort law, which greatly increases its ability to respond to structural and allocative concerns.

These features of tort law, and the fact that tort norms often recede into the background of private relationships (especially when liability insurance is relied upon—as it often is), create the temptation to see at least general duties not to harm as merely “loosely relational,” to borrow John Gardner’s term: that is, not justified by values intrinsic to the relationship between the right-holder and the duty-bearer, but by their instrumental value in safeguarding the right-holder’s interests.Footnote ¹²⁶ There is a valuable relationship here: such duties define justified ways of relating to others.Footnote ¹²⁷ However, it operates at the allocative level, focusing on agents as the targets or beneficiaries of regulation and on basic features of daily interactions in modern society, such as the badness and reasonable avoidability of physical harm. While this is a relationship in which we engage with others in an attenuated sense, tort law’s ability to assign responsibilities tempts us to see it as capable of grounding accountability and remedial mechanisms. Therefore, the thought goes, justifications for tort norms might be able to remain external to rights-and-duties relationships, and thereby pay more attention to vulnerabilities and policy goals.

The problem with this line of argument is that the fact that the public standards tort law articulates operate within strictly relational normative structures shapes and limits its ability to promote policy goals and highlights the necessity of strictly relational justificatory baselines. The starting point is that tort law’s procedural structure creates avenues of relational moral address: right-holders call tortfeasors to account not merely for acting wrongly or failing to do good but for wronging them in particular.Footnote ¹²⁸ This procedural structure has important substantive implications touching tort law’s evolution. Not only do public institutions not enforce tort norms on behalf of right-holders, but there are strong pressures against seeing right-holders as enforcing their rights in the name of the collective—and duty bearers as discharging their duties as part of the state’s regulatory apparatus. As noted by Chief Justice Cardozo in the famous Palsgraf case, “the plaintiff sues in her own right for a wrong personal to her”—not for breaches of general duties or duties owed to third parties:Footnote ¹²⁹ such public wrongs are responded to by state agents, which call agents to account for on the collective’s behalf (for example, through criminal law).Footnote ¹³⁰ Thus, for example, even if we act wrongly by driving negligently but harmlessly,Footnote ¹³¹ we do not wrong any tort law right-holder—and the fact that stopping such behaviors is something that the police rather than private right-holders are authorized to do reflects a political morality choice to deal with inchoate wrongs through public mechanisms. Tort law’s procedural structure therefore tends to support rights to be treated in certain ways—not rights that others conduct themselves in certain ways.

While, unlike special duties, general duties do not obligate us to act for particular others (general prudence is appropriate), we must not exaggerate the distance between special and general duties in terms of moral detachment. In both cases, the ongoing requirements of duties and the fact that violations wrong a right-holder shape and constrain tort law’s possible content. For tort law to empower and protect right-holders and guide and constrain duty-bearers in social spheres such as roads, hospitals, universities, or workplaces, it must not only recognize fewer and simpler primary norms but also recognize primary norms that operate effectively and justifiably within the normative structure of private rights-and-duties relationships while being sufficiently congruent with their moral foundations.Footnote ¹³² That is, tort law must tie the right-holder’s interests to the duty-bearer’s agency while addressing the inherently relational and interdependent components of these normative ties—expectations, standing, responsibility, accountability, reactive attitudes, and remedies. Tying these parts together is necessary for tort law’s ability to foster relationships of mutual recognition and respect in regulated spheres (rather than only during or following litigation). However, failures in this regard compromise not only tort law’s legitimacy but also its effectiveness, given the strong pull moral structures often exert on judges and private agents. A salient example in this regard is Michael Selmi’s critique of attempts to reorient discrimination law towards greater attention to distributive or allocative patterns, while pulling it away from relational notions like accountability and blame: this trend was met with judicial resistance, had limited social impact, and in some senses even backfired by painting discrimination law in instrumentalist and political colors.Footnote ¹³³

These relational pressures are even clearer once we add the ties between primary rights to certain forms of treatment, secondary rights to remedies in response to (or in anticipation of) violations, and tertiary rights to call duty-bearers to account, which are not merely contingent: these right-types form constitutive parts of coherent wholes.Footnote ¹³⁴ The duty-bearer’s accountability to the right-holder and their duty to remedy the right-holder’s predicament are intrinsically valuable as forms of relational recognitionFootnote ¹³⁵ but only if tied to the duty-bearer’s responsibility to the right-holder rather than, say, to their ability to spread the costs of liability or be made a deterring example of. Because remedial norms form parts of strictly relational processes of calling to account—in and out of court—they are under strong pressure to address relationships between right-holders and duty-bearers (hence their zero-sum structure). And because tort remedies and processes of calling to account are strictly relational, there are strong pressures on primary rights and duties to take strictly relational forms as well—to be the kind of norms that can ground effective and justified accountability and remedial mechanisms.

While tort law does not necessarily assume the role of preventing or mending ruptures in moral relationships, it is still relational in the sense that it guides and limits (if things go badly, with judicial help) dialogical processes through which private agents shape their relationships in response to morally important factors, such as risks of physical harm.Footnote ¹³⁶ Tort law puts people whose basic interests are vulnerable to our actions on our normative radar, and the fact that it does so by recognizing rights is crucial. In the hustle and bustle of modern society, with which the birth of modern tort law is often associated, it is easy to overlook or trivialize other people’s basic interests. A system of legal norms that is discretionary in terms of enforcement yet mandatory in terms of content responds to this factual predicament in a dynamic and iterative yet morally robust manner that contributes to the cultivation of recognition and respect in society (similar claims could be made, for example, about contract law and promissory morality).

The important point is that relational baselines such as intention to harm, foreseeability of physical harm, the existence of a special relationship, or invitation of reasonable reliance, are foundational in reasoning about this form of regulation—even if they can be modified or negated by policy reasons. Their indispensability results from the fact that the place and weight of justifications for tort norms are shaped by how they embed these norms in rights-and-duties relationships expressing mutual recognition and respect: they must address the ways in which tort norms empower and protect right-holders and guide and constrain duty-bearers, within the unique accountability and remedial structures created by tort-type regulation. Policy reasons—or, for that matter, vulnerabilities to certain harms, in and of themselves—cannot address such issues on their own: they lack the normative resources to shape the details of rights-and-duties relationships; and even if they contain sufficient data, they lack the tools to ensure that they do not overstep their domains by deputizing right-holders as private attorneys general and enlisting duty-bearers as cogs in the state’s institutional apparatus. Thus, as the next section elaborates, even if policy reasons can negate or limit the recognition of tort norms, they must garner support from relational reasons to shape their finer details.

B. Positive Policy Reasons

Tort law rights and duties, despite their unique relational normativity, cannot be isolated from the basic structure of society or from broad allocative patterns: recognition and respect require relationships to conform to moral inputs from all relevant standpoints. While private agents might lack the standing to call other private agents to account as citizens or as participants in allocative practices, the state cannot ignore structural and allocative objections to regulation. This means that moral claims about tort law’s evolution, which are directed against the state, cannot simply reject non-relational facets of the moral universe as irrelevant, even if they do not (or rarely do) form parts of non-legal private relationships. Thus, while the evolution of many torts involves less frequent recourse to policy reasons, it remains the case that all torts must be shaped by such reasons to some extent at certain points in their normative lives, even if only to ensure their continued legitimacy. Contractualism highlights, in this regard, not only the ability of policy reasons to enrich tort law’s relational baselines but also the limits within which they must do so, which are tied to the nature of rights-and-duties relationships.

The starting point, as noted above, is the strong moral pressures to pursue structural and allocative reasons through tort law within the limits entailed by the structure of rights-and-duties relationships.Footnote ¹³⁷ Policy reasons do not gain weight if they cross certain thresholds,Footnote ¹³⁸ after all relational reasons have been satisfied,Footnote ¹³⁹ or as required by complex consequentialist schemes.Footnote ¹⁴⁰ Rather, tort law normally engages policy reasons from a qualitatively different normative perspective. If relational paths are insufficient to meet structural or allocative standards, we may have reasons to forsake or revise our structural, allocative, or relational principles—as part of an ongoing attempt to reach reflective equilibrium. But we have strong reasons to start within the endemic environment of relational reasons. This creates a robust filtering mechanism. Because it is grounded in tort law’s value as a regulatory tool rather than in the structure of litigation or the limited law-making authority of judges, it can also push back against instrumentalist claims in legislative arenas. The moral reasons to create congruence between tort norms and relational structures draw their force from the ways in which tort norms empower and guide private agents in practice.

Now, while this article does not seek to justify specific tort principles or reforms, it must address the general connections between this filtering mechanism and tort law’s topography. This is important because, as Oberdiek notes, one of the advantages of contractualism is that it can explain and justify not only secondary and tertiary but also primary tort norms—and do so without bifurcating these inquiries.Footnote ¹⁴¹ To highlight contractualism’s utility in this regard, we venture back to the two evolutionary paths for the recognition of tort norms.

The first point touches on the relational resistance to policy justifications for the recognition of general tort norms. As noted above, tort law aims to give us reasonable control over our normative situation by recognizing affirmative or complex duties only in special relational contexts: general duties, which we cannot unburden ourselves of, tend to be negative in order to make conformity with them easier.Footnote ¹⁴² The clearer scope and greater salience of physical harms allow tort law to erect general and even rigid boundaries (as is the case with harmless trespass, for example) around our bodies and physical property. Moral expectations are stable enough to ground responsibilities not to transgress such boundaries and reactive attitudes and remedies in response to violations, even in interactions with strangers (think about a tap on your shoulder, mistaking you for someone else, not followed by an apology). Even when it comes to the more complex duty not to negligently cause physical harm, mere foreseeability—a concept deeply tied to the contractualist theme of mutual recognition—can serve as a robust baseline within rights-and-duties relationships with strangers.Footnote ¹⁴³ Things are trickier with regard to intangible harms. While there is growing openness to protections against them, their legal standing remains largely where it was over two decades ago when the Supreme Court of California noted that “reliance on foreseeability of injury alone … is not adequate when the damages sought are for an intangible injury.”Footnote ¹⁴⁴ Thus, general rights against intangible harms take more categorical, concrete, and strictly relational forms that require, for example, intention (rather than negligence) and a higher severity of harm.Footnote ¹⁴⁵

It is against this background that we must understand the difficulty of relying on positive policy justifications for recognizing general tort duties. The problem does not lie with the specific values adopted: even reasons about autonomy, dignity, privacy, or equality, for instance, face challenges in operating as general relational baselines.Footnote ¹⁴⁶ One problem is that figuring out what policy goals require and which means of pursuing them are justified often exceeds the capacities of private agents: these goals touch complex structural and allocative issues and demand not only respect and forbearance but also protection and promotion. The second problem is that because these goals are abstract, duties to pursue them involve a level of discretion and authority that liberal societies tend to place in the hands of the state (think, for example, about obligating social media platforms to reconcile privacy and freedom of expression). Even if such general duties do not raise concerns about complexity and authority, they still raise questions about how ongoing interpersonal scrutiny in a political normative key, related to complex collective goals, can harm not only personal autonomy but also some of the values, attitudes, and predispositions that constitute private relationships.Footnote ¹⁴⁷

One response to these concerns is to translate structural and allocative reasons into more concrete general norms, which include components like intention or a higher severity of harm. This response risks ignoring the systematic factors that make structural and allocative reasons important for enriching tort law. Another approach, more common in practice, takes the second evolutionary path of special norms, whose justification is made easier by narrowing their scope of application. It is easier to justify a physician’s duty not to negligently cause emotional harm to patients, a lawyer’s duty not to negligently cause financial harm to clients, and a university’s duty to protect students from physical harm by third parties. These justifications are not made easier by structural or allocative factors (even if they often push in similar directions). Rather, relational reasons, tied to the assumption of roles on which vulnerable individuals rely, support structural and allocative reasons (even when the role’s assumption is implicit, as is sometimes the case with parental obligations). The relationship’s normativity and value make it easier to embed within it tort law baselines that extend beyond physical harms, and to shape those baselines in light of policy reasons that address broader normative concerns.

I will demonstrate this idea by addressing affirmative duties in several relational contexts. It is quite obvious, for example, that state institutions tasked with public safety, like the police, are obligated to protect individuals from foreseeable harm by third parties.Footnote ¹⁴⁸ Take a slightly less straightforward case: a train operator was held liable to a female passenger who was raped as she walked to the station after she was discharged by a conductor between stops.Footnote ¹⁴⁹ Was the train operator being used to respond to a societal risk primarily because of policy reasons for effective harm prevention? Something along these lines was suggested in a case dealing with the duties of owners of apartment buildings to protect tenants from criminal attacks: “the landlord is in the best position to take the necessary protective measures” and “every segment of society has obligations to aid in law enforcement and to minimize the opportunities for crime.”Footnote ¹⁵⁰ While there are many good reasons to obligate train operators to protect passengers and landlords to protect tenants, seeing them as the long arm of the state is not one of them—and has deeply troubling implications. It is better to say that public transportation implicitly includes the promise of a safe commute, just as apartment building ownership carries the promise of a safe environment. The absorption of certain societal risks—say, of assaults by third parties, but not of invading armies—into these relationships, is reflected in the robustness of relational moral structures of expectations, claims, and accountability. A similar story could be told about university duties to protect students: if in the post-war period judicial emphasis was placed on student autonomy, in recent years courts have been more willing to obligate universities to, for example, protect intoxicated students from foreseeable risks of serious physical harm or to take reasonable measures to prevent the foreseeable suicide attempts of particular students.Footnote ¹⁵¹

Taking a further step, we can see that the path of special duties also covers non-physical harms. As I argue elsewhere, discrimination law, while often labeled as external to tort law, demonstrates this path’s broader utility.Footnote ¹⁵² In fact, in recent years, this domain has confronted challenges posed by its tort law roots. They were vividly exposed in a series of decisions by the Supreme Court of the United States, which subordinated discrimination law to relational tort principles—say, of but-for causation—under the assumption that unless Congress clearly states otherwise, when it “creates a federal tort it adopts the background of general tort law.”Footnote ¹⁵³ This tortification process raises structural and allocative concerns (and so far, it has tended to suppress discrimination law’s progressive potential). However, it holds the promise of better integrating equality-related policy goals within rights-and-duties relationships. Doing so cannot rely solely on strictly relational principles that isolate the content of tort norms from policy concerns, nor on unconstrained policy reasoning that ignores the relational forces that necessarily shape and constrain the legal recognition of all mandatory and privately enforced rights.

Discrimination law already exhibits many features of tort law’s special duties branch: it targets relationships between role-holders such as employers or sellers and vulnerable parties such as employees or consumers who reasonably rely on role-holders to properly carry out the structural and allocative aspects of their roles. Contractualism highlights the fact that while it is difficult to integrate some duties within these relationships—say, to meet hiring quotas that focus solely on allocative patterns—they form rich soil for many others: for instance, adopting proper hiring or dismissal policies, accommodating disability, or protecting against harassment. These norms do not merely concretize the social justice obligation to bring about and support just institutions,Footnote ¹⁵⁴ nor are they merely strictly relational: they expand personal responsibilities and embed social justice within the fabric of private relationships by leveraging their inherent normativity—say, the extended authority of employers over their employees or the seller’s invocation of the consumer’s trust. Consider, for example, the distinction between a bakery’s refusal to provide a cake that says “support gay marriage,”Footnote ¹⁵⁵ and a hotel’s refusal to provide a room with a double bed to a gay couple:Footnote ¹⁵⁶ even if social quality is implicated in both cases, in the first case the refusal can be said to express an objection to the message and not to the customer’s sexual orientation (no customer would be sold that cake and gay customers can buy other cakes), while in the second case the service aligns more closely with the protected trait, thereby establishing a connection between social equality and the service provider’s agency.

Other legal areas that regulate by recognizing mandatory and privately enforced special rights could take similar paths. Concerns arise, for example, about tort law’s ability to address emerging concerns related to privacy, touching, say, social media platforms or remotely monitored products: a rigidity reflected in the U.S. context in the persistent pull of the four privacy torts identified decades ago in the case law by William Prosser (intrusion, public disclosure, appropriation, and false light).Footnote ¹⁵⁷ While there are strong relational reasons not to recognize general privacy rights, there is more room to recognize special rights: for example, to impose duties on social media platforms to protect the privacy of their users from other users who threaten to exploit or extort them. Claims about the measures platforms use to protect users from such threats could fit quite neatly within the structure of rights-and-duties relationships, thereby grounding effective and justified accountability and remedial mechanisms, because it is not only that users are vulnerable to how platforms are designed but also that platforms create the potential for exponential increases in the scope and sophistication of certain risks, such that these risks acquire a qualitatively different character. The same ideas can apply to threats to privacy in other settings, such as workplaces, universities, or sports arenas.

These ideas also make room for economic analysis. Keating claims, for example, that allocative reasons can justify adopting “enterprise liability” regimes, which impose strict liability on agents that can spread the costs of recurring harms created by socially beneficial activities among those benefiting from them.Footnote ¹⁵⁸ When considering such allocative practices—for example, how car manufacturing companies can and should allocate the costs of tort duties to car users—economic considerations, touching incentive structures within complex markets are crucial. In such cases, economic reasons rely on a relational moral foundation, situating them within rights-and-duties relationships: a right-holder who relies on economic reasons to, for example, justify strict liability, does not act as a private attorney general but addresses the nature of their relationship with an agent engaging in an allocative social activity.

This evolutionary path is still limited, for example, by tensions between the intimacy and trust intrinsic to some relationships and legal accountability; by the political nature of some harms (e.g., disinformation on social media platforms before elections), or the diffuse nature of others (e.g., carbon emissions); and by the reduced effectiveness of tort rights when there are significant power disparities (e.g., between banks and clients) or when calling wrongdoers to account requires resources that private agents lack (e.g., regarding workplace safety standards). In such cases, alternative regulatory avenues, such as administrative oversight, might be more justified. However, by examining such regulatory questions from within the unique structure of rights-and-duties relationships, we can not only locate cases in which such relationships can tolerate policy influences, but also rely on their relational structure to shape these policy influences in ways that render them more effective and justified and (even if more gradually) make these relationships more hospitable to structural and allocative principles.

C. Negative Policy Reasons

Importantly, even if there are good relational and policy reasons to recognize tort norms, there could still be negative policy reasons against this recognition. This section addresses the nature of and justifications for this veto power—and explains why, when exercising it, structural and allocative reasons are indirectly influenced yet not constrained by relational reasons. It should be noted that this section does not say whether this veto power should actually be exercised: that is, whether some tort regimes successfully pass through the upstream negative policy net.Footnote ¹⁵⁹

The starting point is tort law’s subordination to social justice. While social justice does not determine the local justice principles that shape tort regimes, it places constraints on their permissibility. Its systematicity allows it to identify deep problems with tort law as part of the basic structure of society that more relational reasons cannot fully capture (this was the gist of Rawls’s criticism of libertarianism): for example, how the recognition of tort norms affects patterns of inequality or the democratic process; or how such recognition reflects on society’s core values—the value of life, for instance, can justify not recognizing a parent’s right against the wrongful birth of their child.Footnote ¹⁶⁰ Social justice principles can also be entrenched in constitutional rights, whose horizontal aspects may constrain tort law’s possible content:Footnote ¹⁶¹ for example, the right to freedom of expression can limit the expansion of rights against defamation if they threaten to stifle civil rights dissent, freedom of the press, and public debate.Footnote ¹⁶²

This does not mean that structural reasons are isolated from relational reasons. For example, the list of primary goods or basic interests underlying principles of social justice may be shaped by relational reasons: say, highlighting the wrongness of certain kinds of harms, such as sexual harassment, or the importance of certain forms of treatment by particular others, such as parents or physicians. Relational reasons can elucidate the political collective’s social justice or constitutional obligations, even if doing so requires their translation into higher levels of generality that focus on the impact of the basic structure on individual interests.

This takes us to negative reasons focusing on allocative institutions and practices, which may limit or negate the recognition of tort norms from a less abstract perspective compared to structural reasons. Such reasons can justify using other regulatory tools when the claims against using tort law—even by parties external to regulated relationships—are stronger than the claims that justify its use. For example, it can be justified to publicly prohibit some forms of risk-creation,Footnote ¹⁶³ or to offer public insurance coverage for certain injuries.Footnote ¹⁶⁴

Like structural reasons, allocative reasons must also rely on relational reasons to uncover the social patterns created by tort law accountability and remedial mechanisms. For example, the imposition of tort duties toward vulnerable individuals can alter formal policies or informal attitudes due to fear of liability, which could harm social practices: concerns about “defensive medicine” can justify not extending duties of care to certain procedures, and the autonomy of house party hosts and their guests can justify not recognizing host duties to protect third parties from their drunk driving guests.Footnote ¹⁶⁵ A second example in this regard is the disparity in liability between agents who performed similar actions with different results, such as two similarly negligent drivers, only one of whom hit a pedestrian. Imposing liability only on the unlucky driver is justified only if their claims are weaker than the claims of their victims and of other people whose basic interests will be adversely affected by alternative regimes. Negative allocative reasons must address, in this regard, the limits of tort law rights and duties: for example, the difficulties of using tort law to regulate mere risk creation.Footnote ¹⁶⁶

Thus, figuring out the nature and weight of negative allocative reasons often requires us to take a few steps into tort law’s domain and borrow relational insights about how it empowers and guides. A few further steps are required if allocative reasoning touches competing tort regimes. Consider objections to a right to enjoy one’s property regardless of the costs to third parties—say, even if it means closing down a nuisance-causing factory, causing hundreds of employees to lose their jobs.Footnote ¹⁶⁷ It could be claimed that the allocative objections to this regime are stronger than the objections to a regime of monetary compensation to property owners.Footnote ¹⁶⁸ Because the alternative regime is a tort regime, allocative objections to the first regime are tied to relational reasons that compare the features of different remedial mechanisms.

One of the important features of the contractualist scheme is the ability of structural and allocative reasoning to absorb relational reasons: even isolated relational wrongdoing, such as discriminatory acts with no structural or allocative implications could be understood as social injustices (say, as parts of a basic rights principle) or as allocative wrongs (because the victim’s claim is stronger than those benefitting from a permission to wrong). However, there is still a sense in which reasons not to adopt tort regimes because of their structural or allocative impact have priority over relational reasons: for example, because the former can highlight harms to third parties, such as the factory’s employees. When such reasons are present, the systematic structure of negative policy reasons is imposed on relational reasons and prevents the materialization of relational tort baselines. It is in this sense that we can define negative allocative and structural reasons as policy reasons in a strong sense of the term (which is by no means binary). True, reasons like harm prevention, defendant burdens, or defensive medicine also have much to say about the internal features of tort regimes. But if they do, thereby entering the relational domain of rights and duties, they lose their categorical priority.

The same type of reason can therefore operate both in a policy register and a relational one. For example, when engaging the physician-patient relationship while focusing on allocations of benefits and burdens to a group with competing claims, defensive medicine can operate as a negative policy reason that must be defeated in its own allocative terms. Because we address a relationship, questions arise about how the absence of a right affects its basic features and value. But we engage these questions from a perspective that does not focus on the internal details of rights-and-duties relationships but on the allocative claims of all affected parties. It is only when we delve into such relationships that positive reasons about them—say, that defensive medicine considerations justify specific tort duties—must be diluted and shed some of their features so that they can properly tap into the normativity of rights and duties.