A Thinking in Networks
But remember that words are signals, counters. They are not immortal. And it can happen … that a civilisation can be imprisoned in a linguistic contour which no longer matches the landscape of … fact.Footnote 1
The new century is already twenty years old, but the old debates of the twentieth century still cast a long shadow over tort theory. This book in many ways is an attempt to go beyond the current academic debate in tort law by widening our vista in theoretical, methodological and conceptual terms. Theoretically, it breaks with mainstream tort law scholarship as it is neither a contribution to the philosophy of tort law nor the economic analysis of law. Instead, it opens a new theoretical line of enquiry by adopting insights from systems theory as its ‘analytical frame of reference’ to advance a theory of tort law beyond the state.Footnote 2 Methodologically, it engages in comparative theoretical investigations examining, first, the law France, the United Kingdom and the United States, before, secondly, moving from a comparative to a transnational orientation when confronting the role of tort law in the twenty-first century. The story that it tells is, I hope, a compelling and insightful one. The internal evolution of tort law from models of individual responsibility to organisational responsibility cannot be understood without reference to the broader social epistemologies in which tort law is embedded. The story is one of mutual feedback between law and society in which different knowledge paradigms – ways of seeing the world – have contributed, and also been influenced by, different models of responsibility.Footnote 3 Today, the society of networks forms the social epistemology of law and, it will be argued, tort law is, in bits and pieces, re-pivoting towards the society of networks. The final and conceptual contribution the book will make is to show how a new and distinctive form of responsibility – ‘network responsibility’ – is the best and most persuasive way to coherently braid these ‘bits and pieces’ into a new model of tort law for a new regime of liability beyond the state.
Making the case for network responsibility as an overall architecture for a new approach to tort law requires a reconstruction of tort law and its relationship with its wider social epistemologies. Therefore, once the theoretical ‘frame of reference’ has been established our attention turns to reconstructing the tort law of the nineteenth and twentieth centuries as giving rise to two models of responsibility – individual and organisational responsibility. The first two chapters are devoted to retracing and in some ways retelling the evolutionFootnote 4 of modern tort law from its classical foundations through the twentieth century to the challenges facing tort law today.Footnote 5 In broad terms, they retrace the evolution of twentieth-century tort law from a paradigm of individual responsibility to a paradigm of vertical vicarious liability or organisational liability.Footnote 6 These models of responsibility, it will argued, cannot be divorced from broader transformations in the knowledge base of society. The common threads in this analysis will be drawn out through a sustained focus on product liability and state liability considered as illustrative of broader reorientations within tort law.Footnote 7 The first two chapters recount the rise of individual responsibility as a form of liability appropriate to a ‘society of individuals’ and how it was supplemented by secondary remodelling of tort law in the course of the twentieth century by organisational responsibility as a second and self-standing model of legal responsibility. This evolution will be retraced by examining public bureaucracy (the state) and private bureaucracy (the firm), in which it will be argued that whatever the precise doctrinal designation, organisational responsibility is the most plausible account of the tort liability regime that developed. On the public side, I will examine English and French law in particular,Footnote 8 with the Anns decision in the United Kingdom and Feutry in France representing key turning points in the march of organisational responsibility. Twentieth-century courts viewed the activities of subcontractors in terms of a delegation of public activities and, as such, that public authorities should be held liable either for negligence or on the doctrinal basis of breach of statutory duty. This is the public face, as it were, of organisational liability. Thus, on an organisational model, the real addressee of liability ought to be the state. This mirrors developments on the private side, so to speak, as like the employer in vicarious liability or the firm in the case of product liability, the state was ultimately imputed responsibility when that risk materialised in harm. Organisational liability might be understood as a ‘secondary normative remodelling’ of tort law that made it more reflexively oriented for a society of (hierarchical) organisations.Footnote 9 None of these changes could have been conceived or realised without reference to changing ideas of risk.Footnote 10 However, while two distinct models of liability have developed, they tend to inadequately grasp the contemporary ‘landscape of fact’.
As we enter the society of networks, the existing models of tortious responsibility are failing to adequately respond to their environment. Networks, and their governance, do not respect national boundaries and so the focus of our investigations turn to the context of the EU legal order. Understood as an avant garde experiment in transnational governance, it is the site for developing a theory of network responsibility.Footnote 11 It will be shown how the inchoate beginnings of an idea of network responsibility can, indeed, already be detected. Today, in the context of EU law, while the hard and fast distinction between public and private is blurring, effacing even, the models of both individual and organisational responsibility are strained. They are strained because neither model of liability adequately reflects the societal knowledge base, which has migrated from organisations to networks. These networks, unlike in the society of organisations, cannot neatly divide into public and private spheres. Secondly, the society of networks is also a society of risk in which classical understandings of causation become increasingly difficult to apply to harms occasioned by ‘network failures’.Footnote 12 Thirdly, networks have broken out of the frame of the territorial state and, it is argued, a form of liability law that takes the concept of network as its ‘frame of reference’ in a multi-level legal order is required. This line of reasoning will be pursued in Chapters 3 and 4 with specific focus on the New Approach to Technical Standards (NATS) which is a characteristic example of ‘network governance’Footnote 13 in a risk society.Footnote 14 Ultimately, it will be argued that what is now required is a concept of network responsibility for network failures, which is normative in substance, and represents an overall cupola for which individual responsibility and organisational liability form constituent elements. In the final chapter, it will be demonstrated how by recasting and reimagining Francovich liability, expanding on how its principles have already been applied to both public and private actors, it can be repurposed as a new form of secondary liability within an overarching concept of network responsibility, which can contribute to rethinking tort liability, more generally, in a transnational society of networks.
These preliminary observations can be presented in the form of a matrix (Table 0.1), which forms the entry-point and foreshadows the argument to come.Footnote 15
| Modernity | Tort law | Model of liability | Case law examples |
|---|---|---|---|
| Society ofindividuals (nineteenth century) | Individual wrongdoing: moralistic | Individual responsibility | Winterbottom v Wright; Geddis v Bann Reservoir |
| Society of organisations (twentieth century) | Strict(er) liability (risk creation) | Organisational responsibility | Anns v Merton LBC; Donoghue v Stevenson |
| Society of networks (twenty-first century) | Liability for risk enhancement | Network responsibility: network liability, organisational liability and network liability as components | Francovich liability Schmitt |
The rest of the chapter will be devoted to explaining how I reached this position on contemporary tort law. It will distinguish a systems theory frame of reference from those more commonly associated with tort theory. Then its basic assumptions will be outlined before returning to the main argument about how European law offers the promise of law of torts for a society of networks.
B Breaking Academic Frames of Reference
Tort law is in many ways a subject revitalised due to the processes of Europeanisation and globalisation of law. The process of Europeanisation has provoked scholars of private law to once more focus on the question of legal evolution and the social embeddedness of private law.Footnote 16 This ‘irritates’ private law theory that for a long time has been in the grip of what White describes as ‘neoconceptualism’ – primarily, the seemingly intractable debate between legal economists and moral philosophers of tort law.Footnote 17 In the debate, stripped back to its most essential assumptions, moral philosophers build their model of tort law around the individual, and their obligations in accordance with a theory of individual right and reciprocity, while legal economists focus on aggregative individual preferences, largely arguing how tort law does or should correspond to economic efficiency. In this respect, they argue for tort law’s function as a proxy for contract when transaction costs prevent ex ante voluntary exchange from taking place.Footnote 18
Our starting point, inspired by systems theory, is different. To be sure, individual and aggregative concerns are germane to tort law, hence the impasse in scholarship. Our point can be summarised quickly: there is no law of torts in abstracto, which requires explanation. Instead, tort law is a dynamic and contingent system. Earlier common law theorists grasped this insight. As Holmes argued:
What has been said will explain the failure of all theories which consider the law only from its formal side; whether they attempt to deduce the corpus from a priori postulates, or fall into the humbler error of supposing the science of the law to reside in the elegantia juris, or logical cohesion of part with part. The truth is, that the law is always approaching, and never reaching, consistency. It is forever adopting new principles from life at one end, and it always retains old ones from history at the other, which have not yet been absorbed or sloughed off. It will become entirely consistent only when it ceases to grow.Footnote 19
It is well-known that while Holmes pointed to the ‘bricolage’ of the common law, he went on to provide his own formal theory dividing tort law neatly into negligence, strict liability and intentional torts.Footnote 20 Nevertheless, he was at least cognisant of the importance of the diachronic dimension to law to furnish a good explanation of current law and its various apparent inconsistencies in objectives not to mention the ‘real’ social problems that legal changes were responding to. What the existing approaches in moral theory of tort law lack, it is submitted, is a focus on this practical dimension, suggesting either that these questions are beyond the province of private law or that a pre-political private law has precedence over them.Footnote 21 Law and economics (or economic analysis of law – EAL), on the other hand, offers an instrumental theory of private law suggesting it is merely a tool in a regulatory arsenal that might serve in a complementary fashion to public regulation.Footnote 22 Whereas public regulation is one way to achieve regulatory objectives, private law is an institutional complement to these ends. In the radical form of this argument, private law, tort law in our case, is public law in disguise.Footnote 23 Yet, as the legal pluralists of tort law have shown, neither theory seems to ‘fit’ tort law on its own terms.Footnote 24 Fit is achieved at the expense of some questionable stretching of existing doctrines,Footnote 25 reformulations that seem to ignore the actual history of their development,Footnote 26 or through marginalising developments in the lawFootnote 27 and, in particular, by ignoring the way courts justify their development.Footnote 28 Indeed, Lord Hoffmann describes this as a tension in tort law between individual wrongdoing and compensation as social justice that is not fully reconciled. This fits into a pluralist understanding of the tort system. It is worthy of extended quotation:Footnote 29
the law of negligence attempts simultaneously to pursue two quite different objectives. The first is to attribute responsibility on the basis of moral fault and the second is to award compensation by way of redistribution of loss. Responsibility on the basis of moral fault requires that one should have regard to the actual physical and mental ability of the defendant as far as it is practically possible to do so …. But redistribution of loss is a matter of social justice which does not require any moral fault at all. It depends upon the proposition that it would be fairer for the loss to be allocated to the insurers or customers of the enterprise which caused it than to be borne by the victim alone. And since no moral fault is required, it does not matter that the defendant was doing his individual best. The question is simply whether his conduct fell below whatever standard the law prescribes for the protection of the plaintiff.Footnote 30
That a celebrated judge should point to the diverse aims of tort law is largely ignored by neo-conceptualist theories of tort law favouring an approach that, like their scholastic antecedents, largely places a gloss on tort law. But, then, what do we propose in its stead? With a reading of tort law inspired by systems theory, are we not culpable of exactly what we criticise, namely, placing yet another gloss on tort law – indeed, it might be argued, one that is considerably more obscure and abstract than the valiant attempts of the systematisers? I argue that the ‘added value’, as it were, of an approach informed by systems theory is to better conceptualise the fact that tort law is a ‘historical machine’, which has developed layers of complexity remodelling itself to changes in the knowledge base of society.Footnote 31 From this perspective, the controversy that Lord Hoffmann touches on in the earlier quote is the evolutionary pressures placed on tort law by the development of the ‘practical knowledge base’ of society from one of individuals to one of organisations. Before examining this evolution more closely in Chapters 1 and 2, the basic features of our theoretical standpoint will be unpacked. Once unpacked, it should be apparent that this approach locates tort law in the more general evolution of society in modernity.Footnote 32 This meta-theory aims to provide an analytical frame of reference through which tort law can be reconstructed and provides, it is submitted, an instructive and original lens through which to view and conceptualise contemporary tort law. These latter theme will be developed in chapters 3 and 4. Thus, by adopting a systems theory approach, which stresses tort law as a system in a dynamic and contingent relation to its environment, an interpretation of tort law not primarily based on the individual is proffered, whether conceived in terms of abstract right or an aggregation of preferences.
C Understanding Systems Theory as an Analytical ‘Frame of Reference’ rather than a Grand Theory
1 Tort Law from Individuals to Systems: Ladeur’s Debt to Luhmann
It is now germane to make the theoretical commitments explicit by situating the approach in a systems theory ‘frame of reference’. Within systems theory, the argument is ‘Ladeurian’ because it is concerned with the relationship between the internal differentiation of the legal system and its relationship with changes in the knowledge base of society.Footnote 33 To unpack these theoretical commitments, however, some central features of Luhmann’s theory require explanation. Once explained, the theoretical commitments of this book will be clearer. At the outset, it should be noted that the systems theory approach does not offer a grand theory of society; although it is clearly a meta-theory, as it is a sociological theory about sociological theories. It is viewed as an analytical frame of reference rather than positing an explanation in the sense of the philosophy of tort law or EAL. We do not attempt to find ‘fit’ between our theory and the actual practices of a legal system as understood by participants, whether by explaining particular doctrines as instantiations of deeper philosophical commitments (deontological or utilitarian) or by drawing out the implicit philosophical foundations of the tort system.Footnote 34
Systems theory is meant to be ‘an observation of observations’, second-order observations of legal ‘communications’ rather than reconstructions of the internal point of view.Footnote 35 It is not an external point of view, properly speaking, but rather relies on existing legal theory but reclassifies these theories according to its second-order observation methodology. Legal theories become, in this light, self-descriptions and are relegated to the level of law’s observations. A crucial methodological step in this approach resituates the individual no longer as a sovereign will shaping laws to their ends, or a group of ‘participants’ within a practice; rather, the law’s individual forms part of law’s self-description. To put this in other terms, Luhmann is not concerned with the individual or groups of individuals (including their purported ‘internal point of view’); instead, his analysis occurs at the level of communicative systems. As King and Thornhill state: ‘Luhmann’s primary unit of analysis is not the individual or groups of people but systems. And these systems consist not of people, but of communications.’Footnote 36 It should be stressed that Luhmann constructs a communicative theory that departs from methodological individualism. Instead of searching for the internal point of view of participants, systems theory examines societal communication.
It is radically different to EAL or philosophy of private law in this respect. ‘The individual’ is not the centre of society and nor should they be the basic unit of social analysis. The individual, according to this approach, is a constructed category, a fiction that can be subdivided into three distinct elements: the living, the psychic and the social system:
In this schema people, as we have seen, become living systems, which exist as bodies and bodily parts, and ‘psychic systems’, which produce meaning through consciousness. Society, on the other hand, consists of interdependent social systems which make sense of their environments through their communications. The two remain always quite separate, although dependent on and ‘structurally coupled’ to one another.Footnote 37
The focus in Luhmann’s theory is on how communicative systems interact with one another, which follows from his strict separation of psychic systems (‘consciousness’) from social communication.Footnote 38 This assumes, of course, that there are distinct systems of communication independent of ‘psychic’ systems. The presence of multiple communicative systems is explained in terms of modernity. The move from a stratificatory to a functionally differentiated society leads to the emergence of several communicative systems to solve coordination problemsFootnote 39 leading to multiple interpretations of the same ‘event’.Footnote 40 The emergence of societal sub-systems is explained on the basis of the double contingency problem transposed from bilateral exchange to the level of societal communication. These communicative systems constitute society, as such, because society is defined in terms of communications. The communicative system defines its environment. I will return to this issue when we discuss the place of the legal system in this scheme. Suffice it to say, in this context ‘functional’ does not refer to purposive or goal-directed rationality, as it more commonly denotes, but in Luhmann’s terminology it refers to the role of each differentiated communicative system, which produces meaning although different communicative systems produce different sub-systems of meaning (Sinnwelten). Function obtaining to law, in this sense, denotes law’s role in providing counterfactual normative expectations over time, upholding these normative expectations in the face of experiential disappointment.Footnote 41 This is important for social ‘learning’ because normative counter-expectations help other systems ‘anticipate’, ‘adjust’ and ‘plan’.Footnote 42 Radical separation of communication discourses assumes, paradoxically, interdependence. His examples of these various systems include law, politics, economy, art and so on. In terms of meaning, law produces normative expectations ‘in the face of actions that contradict such expectations’.Footnote 43 Differentiation is explained by King and Thornhill thus:
Systems are functional, therefore, in so far as they are able to organize communications and disseminate them in ways that they and other communicative systems may make use of them. In very general terms, function systems create order out of chaos: they give meaning to events which otherwise would be meaningless for society. Their functionality relates exclusively to communications and is in no way affected by the quality of their performance assessed on any other basis.Footnote 44
From this account, it is clear that for systems theory the function of the legal system is to provide society with normative expectations that help stabilise or ‘de-complexify’ society. While in common with classical legal positivists, Luhmann regards law as ‘operationally closed’ communication, his overall theoretical approach is different. He does not attempt to furnish an internally coherent system of rules, referable to the will of a legislator or participants’ point of view, but rather sees law as ‘a contingent and infinitely alterable system of communication’.Footnote 45 Law does not exist in isolation from ‘society’ according to this view; rather it reconstructs its environment by internal legal means while maintaining its ‘identity’ qua system:
Closure is not to be confounded with impermeability to influence from outside: the system is permanently ‘irritated’ from outside, i.e. by its natural unstructured environment or by other systems. But systems do not share a common reality – they have to and are only able to observe and operate on certain elements of reality – because their own ‘identity’ does not have a stable objective character, they are a product of their own operations using certain distinctions and neglecting others.Footnote 46
It is at this point that a distinction between code and programme is introduced. The code denotes the differentiation of a system from other systems into a functional system. It is akin to its ‘DNA’ and gives a system its identity. Programmes are more variable depending on ‘contingent choices’ and represent the cognitive openness of the law: ‘One needs to appreciate the role of programmes for system operations and the possibility of programmes becoming highly complex in response to the growing complexity of the system’s environment.’Footnote 47 Programming allows for the environment to re-enter the system but in a way that does not undermine its code.Footnote 48 The code law/non-law still determines how the environment, as constructed by the system itself, will be meaningfully understood.Footnote 49 However, programmes reformulate information from law’s environment and reframe them to be consistent with law’s binary code.Footnote 50 As King and Thornhill explain: ‘Programmes exist as organizers of information which allow the application by the system of its binary code.’Footnote 51 Hence, if in law the binary code depends on the distinction between law/non-law, then what counts as law is organised by the law’s programmes which provide criteria of selectivity for law’s code. Although there is no direct relationship between the legal system, in terms of input–output, and other systems, such as law and the economy, the legal system ‘experiences’ irritation in its environment and this provokes internal reactions at the level of its programmes. Programmes relate to law’s function by making the law a ‘historical machine’, allowing for law’s binary code to be applied.Footnote 52 These programmes create ‘redundancy’, meaning the law can incorporate ‘events’ (information) into its ‘structures’, which allow for normative stabilisation over time.Footnote 53 Luhmann locates the centre of these legal communications in courts; it is here that the programmes of legal communications are primarily unfolded.Footnote 54
2 Ladeur’s Models of Modernity: Programmes and Paradigms
(i) The Individual and Organisational Paradigms
Thus, the law as a distinct communicative system with its own ‘internal complexity’ (internal differentiation) emerges with modernity. This is the level at which Ladeur’s analysis becomes germane. The fundamental break here in modernity is the introduction of the liberal law, which is the law of a ‘society of individuals’.Footnote 55 Liberal law, with its separation of the right and the good, is an internal legal programme that reflects the functional differentiation of society.Footnote 56 This is the case, first, because what is legally valid (law/non-law) is determined entirely by the legal system itself, rather than by appeal to an extra-legal source, meaning the law itself is functionally differentiated from other communicative systems.Footnote 57 In other words, no finis ultimus outside legal communication is a source of law, rather the law simply links operation to operation.Footnote 58 Second, the liberal law’s programmes (subjective right; private will) supports ‘space’ for heterarchical communicative systems and, in addition, relies on these systems to enrich its programmes (internal differentiation). Thus, with Luhmann the basic insight follows the claim that while modern law is operationally closed, it is cognitively open. Ladeur’s paradigms of modernity attempt to explain the law’s internal differentiation going beyond Luhmann’s rather more limited examination of the internal dimension to law, and its relationship to its epistemic knowledge base.Footnote 59 The relationship between cognitive assumptions of normality and legal norms is close and complicated. To understand Ladeur’s approach, this is a key (epistemological) methodological assumption. I will expand on this distinction presently.
Within the legal system, programmes develop patterns as a result of ‘events’ and these programmes include what Ladeur refers to as ‘legal dogmatics’. The liberal law is the law for a heterarchical society, without a stable concept of collective order imposed from above or by tradition.Footnote 60 Instead, central ‘institutions’ such as the subjective right support what Ladeur refers to the relational rationality of the society of individuals in which the ‘play of ideas’ lead to the emergence of innovation.Footnote 61 In other words, the non-instrumental form of the subjective right which refers to ‘relationships in terms of rules’ buttresses decentralised cooperation by allowing for future-oriented planning of relations.Footnote 62 The subjective right provides a normative structure for decentralised cooperation because its formal equality allows for the forming of ‘stable patterns of relationships between legal actors’.Footnote 63 It is a background ‘architecture’ that facilitates cooperation outside the legal system understood as state law.Footnote 64 Another important but subsidiary ‘institution’ in this respect is ‘classical’ tort law, that is, tort law limited in scope by a strict distinction between misfeasance and nonfeasance.Footnote 65 There is a strong focus in classical tort law on individual active conduct; causation of harm alone will not ground legal responsibility. Tortious responsibility requires fault, most clearly in the civil law where it is conceptualised as an excess of the will.Footnote 66 The legal ‘institution’ of the subjective right and the misfeasance/nonfeasance concepts in tort law do not impose thick moralistic duties on parties to care for one another.Footnote 67 Rather, they facilitate horizontal cooperation without impediments based on feudal institutions or on the basis of state-reinforced customs.Footnote 68
Ladeur, in addition, views decentralised cooperation as giving rise to ‘conventions’ (or cognitive assumptions of normality, interpersonal standards) that can then be incorporated into the law and normatively coded. Hence, while liberal law provides ‘space’ for decentralised cooperation, the latter produces social conventions that can then be incorporated into law in a ‘feedback loop’.Footnote 69 In tort law, for example, the standard of care fulfils this function. What is normal and what is reasonable behaviour are closely linked.Footnote 70 In this light, the collective order is constructed ‘bottom-up’ out of ‘cognitive assumptions of normality.’Footnote 71 This architecture, Ladeur argues, limits ‘the centralisation of legal material’, by which he means: allows for the development of legal transactions outside the legal system; for example, private contracting that creates complex structures of cooperation.Footnote 72 This inverts the classical account of law, whereby legal norms are deductively applied to facts.Footnote 73 In Ladeur’s account, by contrast, practical knowledge informs the development of legal rules, not the other way around.Footnote 74 The deductive modelling of the law, classical legal formalism, conceals the legal system’s dependence on its practical, social knowledge base. In this regard, Ladeur refers to a ‘reverse process of norm generation’, with the construction of norms dependent on practices beyond the law rather than by means of the more classical understanding of norms applied to facts that the deductive model supplies.Footnote 75 It should be clear that this is not a simple ‘input’ model, however, because the law reconstructs these ‘events’ or data internally through its programmes consistent with law’s autonomy or operative closure.Footnote 76 ‘Open norms’Footnote 77 (‘bridging concepts’) such as reasonableness or good faithFootnote 78 assist in linking law’s programmes to shared ‘cognitive assumptions of normality’; in other words, the law is linked to the decentralised ‘collective order of modernity’.Footnote 79 The task of Chapter 1 is to give shape to the programmes of the liberal law, which is situated in a society of individuals. The liberal law gives rise to a particular type of individualistic attribution/imputation in tort law. This is explored in Chapter 1, drawing on examples from the United Kingdom, France and the United States in product liability and state liability law.
For as long as this practical knowledge can be derived from distributed ‘experience’, the liberal law’s subjective right – the society of ‘individuals’ – is an adequateFootnote 80 normative modelling of society.Footnote 81 This practical knowledge base is ‘dynamic’, meaning that the law, a normative stabilisation of this knowledge base, will be impacted by changes in this knowledge base. However, as soon as the knowledge base fragments with the rise of the ‘society of organisations’, this stable framework of rules of just (individual) conduct is undermined.Footnote 82 The society of organisations is a development in the knowledge base of society, breaking the symmetry between the liberal legal system’s individualism and factual normality. In Ladeur’s words, a ‘rupture’ occurs, which he defines in these terms:
A distinction has to be made between the continuous flow of information which emerges from the permanent variation within societal knowledge basis, and the momentary suspension of normality by the rise of new factual paradigms which induce a kind of ‘break of symmetry’ in the inter-relationship between technical normal knowledge and the feedback loops which have to be designed within the legal network of networks.Footnote 83
The legal system, then, receiving feedback from society, experiences a ‘rupture’ caused by a change of paradigm in the practical knowledge base of society, develops its internal programmes. This constitutes a second normative remodelling in order to adjust to the new knowledge paradigm such that the legal system can continue to fulfil its function. The basic shift from the society of individuals to the society of organisations is the change in how social ‘reality’ or ‘normality’ is perceived. In the common law, at least, Holmes is an important thinker in this regard. However, as Kennedy has noted, the main intellectual impetus comes from France and, as such, we will examine Duguit, Saleilles and Josserand in Chapter 2 as illustrative. There is a move from a view of reality that is premised on the idea of chance,Footnote 84 in which case the loss lies where it falls, absent positive (human) conduct, to one based on accidents as statistical regularities, probabilities to be calculated and losses to be socialised, most clearly represented in the rise of insurance.Footnote 85 The injuries that occur to individuals remain a question of misfortune but are rendered statistically predictable and therefore the subject of prevention ex ante or deterrence and loss spreading ex post.Footnote 86
‘Society’ comes to be viewed, in other words, as an aggregate and the question becomes one of risks and their fair distribution. Especially with the rise of insurance, the emphasis shifts from fault to risks, their distribution and the justification for such a distribution. This is what Ewald means when he refers to risks as a ‘mal à répartir’.Footnote 87 This impacts profoundly on tort law, because it provides a basis from which to argue beyond individuals about who the best risk avoider is and why. Different factual worlds make for different normative conclusions regarding is and why. The emergence of risk calculation, of course, led to what Weber calls rationalisation in both private firms and public bureaucracy. These emerging structures became subjects of law, although not legal subjects in the earlier, liberal sense. Artificial or legal personality concealed, somewhat, the way in which risk calculations succeeded rules of just conduct with regard to organisations. Thus, it is the hierarchical organisation to which decision-making is attributed in the secondary normative remodelling. But we must remember that it is not the organisation X or Y per se that is the target but the group to which the organisation belongs. In legal doctrine, this is both explicitly the case (product liability) and implicitly achieved (the doctrine of vicarious liability). Both examples will be explored in Chapter 2. Suffice it to say, the aim was internalisation of costs in industries. However, the model of organisational liability that developed was also applied, with modifications, to the state.Footnote 88 This normative model of liability will be denoted by the term organisational responsibility, where the legal model looks beyond immediate actors to the groups and their activities, in which acts are ensconced. To put this otherwise, the emergence of collective or corporative structures irritates the law and how normality is understood. As Ladeur states:
The knowledge base is increasingly processed and channelled by representative groups, organisations and a homogeneous technical infrastructure beyond the open processes of self-coordination among individuals and small organisations as used to be the case in the case of the liberal order.Footnote 89
Linking normativity and facts is more ‘proceduralised’, with the legal system coordinating groups rather than individuals. There is, thus, an emphasis on ‘the harmonisation of competing interests’ in an interdependent society.Footnote 90 If the society of individuals developed rules of just conduct, the society of organisations developed rules of just conflict. Thus, Ladeur argues that in the society of organisations, the legal system developed ‘a set of “rules of conflict” for the management and co-ordination of different types of rules (individualistic v collective)’.Footnote 91 In his words, ‘[n]ew types of rules and judgments evolve which try to balance the liberal individualistic rationality against the new collective logic’.Footnote 92 The society of organisations, and the notion of organisational responsibility, coincide with the apogee of the welfare state as the primary societal institution of ‘positive integration’.Footnote 93
In tort law terms, at the level of the law’s programmes, this meant that individual litigants are understood as representatives of conflicting groups within society. This moulded the law of negligence and strict liability, building on earlier models of corrective justice towards a model of what has been described as weak commutative justice.Footnote 94 Thus, stricter forms of liability emerge that decouple legal responsibility from fault in the classical sense. The language of the law may remain rooted in concepts such as fault, but fault is redefined as running an unreasonable risk as distinct from directly and unreasonably acting in the world.Footnote 95 And faulty conduct or acts are replaced by ‘fault’ in activities, whether production processes or in the imprudent hiring of subordinates who are not sufficiently controlled. In both cases, the internalisation of costs within an organisation is what matters and the task of the courts is to determine which risks are unreasonable and which are reasonable and must fall on the ‘victim’.Footnote 96 This involves courts in the complex task of determining reasonable risk taking relying on an economic and sociological (aggregative) perspective but ultimately with a view to making a normative judgment.Footnote 97 This presupposes an epistemology in which risks are calculable and collective, and injuries can be converted into a monetary value.Footnote 98 The nub is that accidents are considered as ordinary instances of social interdependence, not acts of nature or the individual will.Footnote 99 Ewald traces this changed view back, at least in France, to the rise of factory accidents in mid-nineteenth-century France, culminating in the law of 9 April 1898, which led to the internalisation of ordinary costs of activities within firms.Footnote 100 As we will see, the theorists of this approach were Saleilles and Josserand, who attempted to refashion French law in light of the ‘machine age’. This move to stricter forms of liability, meaning liability on the basis of risk creation which locates responsibility in hierarchies, such as firms, has parallels in other leading jurisdictions.Footnote 101 It is not only by means of legislation, but by means of reinterpreting judge-made law that transformations occurred to reflect this changing understanding of accidents, their causes and the best way to deal with their consequences. Most importantly, it represents a transformation through which the law harnessed the collective and special knowledge of organisations to achieve public ends. This transformation from the first modern tort law to the second is treated in Chapters 1 and 2 respectively.
Kennedy’s concept of globalisation of the law finds neat expression in this transformation. The change from legal formalism to the social, reflecting a different understanding of the causes of accidents in the context of tort law, might be considered to reflect the way these changes were intellectualised by their leading proponents.Footnote 102 However, it should be noted pace Ladeur that the aim is not to supplant the liberal order with a different model but to construct a functional equivalent model to that of the nineteenth century.Footnote 103 The aim remains that of facilitating decentralised coordination. To this end, the remodelling of normative rules to the ‘patterns’ of the practical knowledge base is required. This pattern is now pre-eminently organisational and, as such, legal programmes evolve to facilitate coordination between organisations, and between organisations and ‘individuals’.
That is not to say that such secondary normative remodelling was not achieved without difficulty.
One might consider Luhmann’s central preoccupation concerning the distinction between contingent programming and purposive programming in this light. The former refers to the ‘if/then’ structuring of legal reasoning, whereas the latter replace ‘if/then’ with open-ended objectives – ends to be achieved. The contingent programming of the law is historicised by Ladeur, who considers it to correspond, more or less, to liberal law. In this book, we follow that approach. Luhmann contends that the latter, purposive programming, can potentially undermine law’s operative closure, leading to ‘de-differentiation’ or ‘cognitivisation’. With the emergence of ‘world society’, these tendencies are magnified.Footnote 104 While it is surely correct that once one replaces rules of conduct with end-dependent rules there is a blurring between law and non-law, it is also the case that the organisational paradigm of the twentieth century remained strongly normative. We will examine the way in which factual normality was integrated into legal reasoning in Chapter 2.
(ii) Beyond Individuals and Organisations: The Society of Networks and Its Tort Law
In Chapters 3 and 4, the salience of ‘the society of networks’ for tort law is examined. The society of networks is understood as a third rupture in the knowledge base of society in modernity. The society of networks as the knowledge base for twenty-first-century society is one in which our understanding of risk has become reflexive, heterarchy has replaced hierarchy as the main ordering structure in both public and private governance and, whether public or private governance is at issue, networks have surpassed the confines of the territorial state and its centrality for the regulation of society.Footnote 105 The loss of regulatory centrality of the state requires a methodological shift from the national and comparative to a transnational orientation. Legal communication occurs not only within territorial states but, as Ladeur and Teubner argue, within and between functional systems in global society.Footnote 106 The EU legal space, and EU law in particular, will be used as the testing ground for a theory of network responsibility because, theoretically, it is considered an example of a post-Hegelian ‘state’ in which ‘positive integration’ is no longer feasible.Footnote 107 Ladeur conceptualises the European Union’s multi-level structure as a ‘network of networks’ that attempts to coordinate a heterarchical ‘society’. This understanding is, then, applied mutatis mutandis to its legal order.Footnote 108
The role of court-based law viz other sources of law is radicalised in the network society, with court made law conceived of as peripheral to legal communication outside the state.Footnote 109 However, while courts are a ‘last stop’ peripheral to networks, it is argued that they should retain an important role as a site for strategic public intervention in disrupting ‘network pathologies’. These ‘pathologies’ arise where networks are captured by particular discourses crowding out others. In our context of public and ‘private’ product regulation the danger is that in the definition of acceptable levels of risk in certain discourses, notably the economic discourse, comes to dominate to the detriment of other discourses within networks, ultimately leading to ‘network failures’.Footnote 110 The role of courts, then, should be to disrupt these pathologies by creating normative counter-expectations.
Failing to do so, it will be argued, contributes to the ‘cognitivisation of law’; that is, the law’s programme becomes dependent on knowledge from other social systems, most notably the economic system, as the market-leading discourse in world society,Footnote 111 rather than fulfilling its traditional stabilising function of providing normative counter-expectations. Thus, courts in this perspective are viewed as having a role in interrupting ‘negative’ patterns in networks, unblocking their dependence on particular discourses that have ‘external effects’ for society. This will be explored in Chapters 3 and 4 in the context of the New Approach to Technical Standards by focusing on Court of Justice of the European Union (CJEU) case law. The regulatory regime of the New Approach to Technical Standards combines private self-regulation by firms and certification bodies, and transnational public supervision and is considered an example of network governance beyond the state.Footnote 112 The fundamental argument advanced in this book is that the CJEU case law conceptualised in its best light furnishes a nascent concept of network liability. How is this the case? Taken together, a number of trends in the way the CJEU constitutionalises private law, treats the public/private divide and understands risks can be recast into a new normative model of responsibility.
It will be argued, first, that emerging and promising trends in the constitutionalisation of private law can offer a potential path to ‘re-normativise’ ‘private’ law, tort law in our case, beyond the frame of reference of the nation-state.Footnote 113 Boston Scientific is an important case study in this respect because of how the CJEU approaches the issues of scientific uncertainty and market access through a rights-based discourse that, in essence, imposes liability based on the enhancement of a risk.Footnote 114 It is argued that it provides a path for the for ‘re-entry’ of other functional discourses into EU law.Footnote 115 This provides a route to ‘re-norm’ EU law’s ‘programmes’ – fundamental rights that apply to public bodies, but also indirectly to private parties through legal interpretation. It is argued that constitutionalised rights-based reasoning provides placeholder rights for mediating colliding discourses of economics, politics, safety and law. In a sense, it expands the ‘protected interests’ of tort law, but in the context of different systems discourses instead of a discourse of individual right. By reframing consumer expectations in terms of rights, EU law turns policy arguments into principled collisions of norms. This allows for ‘strategic public interventions’ to disrupt pathologies in other functional systems based on an open, and case-by-case experimentalism.Footnote 116
Notwithstanding the potential that constitutionalisation holds,Footnote 117 by providing strong normative backstops against cognitivisation, it alone furnishes no model of liability. An important second feature of CJEU case law in recent yearsFootnote 118 has been, in the tort law context, the increasing willingness of courts to translate principles of state liability to ‘private’ parties in governance networks. The Schmitt judgment is illustrative and important because it ‘applies’ Francovich liability to ‘private’ actors.Footnote 119 The CJEU indicates that whether the regulator (secondary actor) is a private certifier or public body, a liability regime should be developed by member states for omissions’ liability for failure to exercise discretion.Footnote 120 This ties the private liability of certification bodies to the public (administrative) liability of states in EU law. This is distinct from organisational liability because it takes account of the secondary role of the certifier and the state by calibrating the standard of care – sufficiently serious breach – to a ‘faute lourde’ level in recognition of its secondary or network role.
Therefore, the CJEU is developing principles of network liability for circumstances in which secondary actors enhance, as distinct from create, a risk that eventuates in harms. The main dogmatic trigger for an enhanced liability for secondary actors stems from the constitutionalisation of tort law, but the application of these principles to secondary actors is based on applying principles developed in the context of state liability to both ‘public’ and ‘private’ actors.Footnote 121 ‘Sufficiently serious’ failure to exercise ‘discretion’ – in other words, failure to regulate at all or adequately – in normative terms can be conceptualised as liability for enhancing a risk. In this way, secondary actors in networks – whether ‘public’ or ‘private’ – are subject to a regime of network liability. However, while network liability creates legal responsibility for the specific role played by secondary actors, it is not yet a regime of network responsibility. It needs to be reconciled with individual liability and organisational liability within an overall framework of network responsibility.
When one combines network liability with principles of individual liability (AGM.Cosmet)Footnote 122 and organisational liability (producer’s liability) in the context of network governance, the overall cupola of network responsibility emerges, which combining elements of individual, organisational and ‘network’ liability. The CJEU takes account of the collective dimension of risk in networks, but normatively distinguishes primary and secondary actors within this overall cupola. Because network responsibility is normative – in other words, it draws on the law’s internal programmes of responsibility – it is consistent with the law’s requirement of coherence.Footnote 123
This is the positive argument. By developing a regime of secondary, normative liability the CJEU is providing a template for liability in networks. The attitude towards risk evident in judgments such as PIP, Boston Scientific and, more recently, Sanofi shows a shift from risk creation (organisational liability) to risk contribution or enhancement (society of networks). By broadening the addressees of liability through rights (constitutionalisation), by shifting emphasis to the justification for the exercise of discretion (Francovich, Schmitt) and by relaxing causal rules (Boston Scientific, Sanofi), the CJEU is creating a tort law for a network society. Importantly, liability of secondary actors formerly invisible and based on risk enhancement, as distinct from risk creation, completes a new responsibility regime into which personalised and organisational liability may be subsumed. The sociological literature which reconceptualises governance structures as networks, when translated into EU law, results in a new, emerging normative model of tort for the twenty-first century.Footnote 124
Table 0.1 should now be clearer. The model developed will, of course, require a deeper engagement with both the evolution of knowledge base of society and a positive and persuasive argument as to why systems theory offers a new, distinctive and illuminating way to view tort law in its societal context. Each chapter will contribute to reconstructing tort law and pointing the way towards a new concept of tort liability both beyond the individual and beyond the frame of reference of the state.
Let’s take stock. It has been argued so far that once one moves from the internal point of view to a systems theory approach to tort law, one is confronted with the social embeddedness of tort law and its models of liability. A focus on these models of liability demonstrates a clear evolution from an individualistic model to an organisational model of liability within the frame of the nation-state. In both respects, tort law and the legal system help to coordinate society by providing a normative framework for society. When one moves from the society of organisations to the society of networks the position of the legal system in performing this task is challenged. In the society of networks, regulation cannot be confined within the boundaries of the nation-state, and instead governance networks have emerged that are, strictly speaking, neither public nor private in the traditional sense. The networks are sites of discourse collisions; unregulated, they impose ‘externalities’ upon society. The existing models of legal responsibility are ill-equipped for the task of providing liability rules for structures that go beyond market and hierarchy. A tort law for the twenty-first century requires a distinctive form of secondary liability for network actors that fulfil normatively ‘secondary’ roles in networks as well as, conceptually, an overall regime of network responsibility. This book argues that a regime of network liability is emerging in CJEU case law; however, an overall concept of network responsibility – Teubner’s invisible cupola – is missing. This book, for the first time, shows how the bits and pieces, the fragmentary elements of network responsibility, can be congealed into a third normative model of responsibility fit for the problems of tort law today.
