I Defining fictions

Arguably, legal fictions are present in most, if not all, legal systems in which a professional corps of judges decide disputes on the basis of an authoritative corpus of past materials, either laid-down rules or decisions. This is because the legal fiction is a device that is created out of a need – indeed, often a duty – to render one's resolution of the dispute in a manner that respects coherence (and thus some level of predictability and stability), while at the same time attempting to be responsive and efficient in one's dispensation of justice. Legal fictions, if you like, are created in the coalface of legal change, which serves two masters: the conservative pressure of the system and the call of the injured pleading for a remedy.

It should immediately be noted that legal fictions are understood, in this paper, as devices used in the process of adjudication. Those theorists who have sought to classify different kinds of legal fictions – e.g. Lon Fuller (Reference Fuller1930–31) and Pierre Olivier (Reference Olivier1975) – have also made reference to, for instance, legislative fictions,Footnote ⁴ jurisprudential fictions,Footnote ⁵ fictions of legal technique,Footnote ⁶ and the fictional use of legal instruments by citizens.Footnote ⁷ Both Fuller and Olivier also refer to fictions used in adjudication, and in that respect – as we shall see in more detail in a moment – their work is directly relevant here.Footnote ⁸

When it comes to offering a general definition, Fuller and Olivier differ considerably. Fuller's well-known general definition is that a fiction is ‘1) a statement propounded with a complete or partial consciousness of its falsity; or 2) a false statement recognised as having utility’ (Fuller, Reference Fuller1930–31, p. 369). To call a fiction a ‘statement’Footnote ⁹ seems artificial, especially in an adjudicatory context where it is surely more usefully understood as a device of reasoning.Footnote ¹⁰ Oliver's general definition is arguably more disciplined, and also more in keeping with the approach taken in this paper:

Nevertheless, there are some differences between Olivier's definition and the one adopted here. It seems more accurate to say that a legal fiction is a ‘suspension’ of an operative fact rather than an ‘assumption’ of fact made – if one knows there is evidence to the contrary of what an operative fact requires, one does not ‘assume’ it to be otherwise; instead, one knows it is otherwise, but one proceeds anyway to allow the normative consequences to follow. Further, ‘suspension’ arguably better captures the temporary nature of the fiction. Finally, reference to facts ‘probable’ in a particular case is arguably unnecessary: for a fiction, all one needs is that one either has evidence to the contrary of the operative fact, or that one does not have evidence of the operative fact – one does not need to pass judgment on the standard (though it is true that sometimes when fictions are employed for lack of evidence of an operative fact, this is done because the judge feels that were evidence to be available it would uncover the otherwise hidden fact).

Olivier's general definition – especially the reference to assumptionsFootnote ¹² made contrary to probable facts – brings him dangerously close to not recognising a distinction between fictions on the one hand, and presumptions and deemings on the other hand. This is a little surprising, for Olivier is otherwise careful to distinguish his notion of a fiction from, for instance, ‘analogous application of norms, e.g. if fact A is present apply the rules applicable to fact B’ on the basis that there is ‘no false factual assumption’; or ‘metaphors and symbolic expressions which endeavour to describe legal institutions or concepts but which do not require a false actual assumption, e.g. prescription “runs”’; or ‘abstract notions concerning the law or legal institutions, e.g. the concept of subjective rights’ (Olivier, Reference Olivier1975, pp. 81–82).

Considering the examples Olivier provides, though, we can read him generously and ignore the reference to ‘probable facts’ in the general definition. For instance, looking at Roman-Dutch law, Olivier says that it is a fiction ‘that an illegitimate child has no father and does not succeed on intestacy to his father or father's relations’ (Olivier, Reference Olivier1975, p. 133) – clearly there is no question of it being probable that a child has no father; or that ‘an illegitimate child is legitimated by the subsequent marriage of his parents, based on the fiction that the parents had married before the birth of the child’ (p. 133) – again, the need for the fiction would not arise unless it was known that the parents did not marry before the birth (or it could not be known whether they did);Footnote ¹³ or that ‘that subsequent ratification of the contract of a minor by the guardian validates the contract, based on the fiction of retroactivity’ (p. 133) – once more, there is no question of the probability of validation (for a minor could not legally validate a contract, let alone physically if they are a baby), this also being a good example of where evidence of intention is required by the operative facts and where, in order to achieve what is seen as the just result in the particular case, that requirement is suspended.Footnote ¹⁴

Similarly, when Olivier provides an example from the common law – in the case of Des Salles D'Epinoix v. Royal Borough of Kensington and Chelsea [1970] – it is an example that works in a context where the facts are known to be otherwise. As Olivier summarises it:

This case is a good example of the approach taken to fictions in this paper:Footnote ¹⁶ the normative consequences attaching to the concept of ‘occupation’ (in this case, liability for the rates) require the operative fact of dwelling physically in the particular abode. This operative fact, however, is suspended – and signalled by the inclusion of the qualification ‘beneficial’ – in order to reach what is thought to be a just result in the case. Olivier, it should be added, criticises this case, asserting that it should have been decided on the basis of an explicit principle:

From the perspective of this paper, this is too quick – Olivier may be underestimating the importance – at the time – attaching to limiting the liability of a divorced husband, or, put differently, the importance – in part because of its unpredictable effects – of not creating a general liability to pay taxes on an abode lived in by the estranged wife and children. But to make this point is to leap ahead to issues we will be considering in more detail in the second part.

In the case of Fuller, although his classification of fictions is somewhat mixed in with his typology of motives for deploying them,Footnote ¹⁷ there are aspects of his account that are of great assistance to the approach taken here. In particular, Fuller's discussion of the relationship between fictions, presumptions and estoppel as devices used to enable, but also ‘obscure’ (Fuller, Reference Fuller1930–31, p. 531), the growth of the law, is particularly noteworthy. Although lengthy, the following passage is worth citing in full:

There are a number of observations to make about this important passage. The first is that it need not be the case that Fact B (to use Fuller's terminology) is replaced by Fact C. Instead, Fact B might simply be suspended. Second, it is interesting that the example offered by Fuller, where Fact B is replaced by Fact C, is one that relates to a problem with proving intentionality – negligence, from this perspective, becomes an engine of legal change precisely because it offers a practical way of dealing with the unavailability (or, better, rare availability) of proving the requisite intention. Third, it is not quite accurate to say that the existing rules are ‘untouched’, but nor is it quite correct to assert that the rule has been changed and done so in an ‘obscure’ way. Looked at by tracing the role of the device across time, and not its use in a single, isolated decision, a fiction (1) puts normative pressure on a rule by noting circumstances in which it may be appropriate (depending on future cases arising before future courts) to suspend one of the rule's operative facts; and (2) gives some normative impetus (but one which subsequent courts may not take up) to the eventual introduction of a new rule. The fourth point is more positive: Fuller distinguishes, but also notes the relationship between, three kinds of devices: fictions, presumptions and estoppel. In that respect, he adds that:

Fuller is not very generous to fictions here, but his analysis is nevertheless useful: presumptions are more likely to be used (and more likely to be efficient) in situations where the operative fact is probable (given, perhaps, the other operative facts), in which case it also makes sense to leave room for rebuttal should the other side be able to show that the unlikely is the case; with estoppel, on the other hand, an operative fact may again be probable, but this time the rebuttal is itself further constrained by an operative fact (the absence of a certain kind of conduct, e.g. duress); and then there are fictions when, says Fuller, ‘all else fails’. Fuller's judgment of fictions, however, is clouded by his isolation of a single decision in which a fiction might occur, pitting fictions against presumptions and estoppel on the same time wave, so to speak. If, instead, we look across time, we will see that presumptions and estoppels tend to be much more embedded than fictions, i.e. they are established rules, though a little hidden in the evidentiary structure of the law. Fictions, then, are not so much obscurantist as they are tentative. In other words, by isolating a single use of a fiction, Fuller precisely misses its value (at least in this context of comparing them with presumptions and estoppels).

Although Fuller is careful to distinguish fictions from presumptions, he is not as explicit, nor as careful, in distinguishing fictions from deemings. He says, for instance, when giving examples of what he calls ‘exploratory fictions’ – i.e. fictions in which judges create constructions in which they ‘feel their way’ towards some principle – that the ‘doctrine of vicarious liability for tort began with such notions as that the master should be “deemed negligent” for hiring a careless servant’ (Fuller, Reference Fuller1930–31, p. 528). It is true that the line between fictions and deemings is not clear. Nevertheless, Fuller's example does appear to have more of the flavour of a deeming than a fiction; after all, it creates an explicit principle – vicarious liability – the operative fact of which is that if a servant is in the master's employ, then the master is liable for the servant's negligence. Were it to be thought of as a fiction, one would need to identify a rule the operative facts of which were being suspended, and it is not clear that there is such a rule. There is a general concept – liability – but the extension of it via the notion of vicariousness does not smack of fictionalisation – it smacks, instead, of a policy stipulation or postulation.

Deemings, then, are more stipulative – more arbitrary in a sense – and do not have a clear relationship to difficulties associated with proof.Footnote ¹⁸ The line, however, as noted above, is not always clear. For example, consider the following well-known device: ‘a person missing for seven years or more is dead.’Footnote ¹⁹ Rescher refers to this as a presumption, and in a sense that is true for it can be rebutted (whereas a deeming typically cannot). However, there is also something stipulative and arbitrary about it – why seven, and not six or eight? – and it also need not be thought of as taking a stance on the probability or likelihood of a person being alive or dead. It is, then, perhaps a quasi-presumption (or a quasi-deeming?). Other difficult examples are (also from Rescher, Reference Rescher2006, p. 5): ‘that a wife is incompetent to testify against her husband’ or ‘that a minor is too immature to vote or to enter into a valid contract’. These certainly feel more stipulative, for they cannot be rebutted, but nevertheless they may be thought to be taking a stance as to the reliability of any potential evidence to the contrary, i.e. that no matter what proof of a wife's competence is offered, it may still be the case that the wife would be biased, and no matter how mature a child is shown to be, they are unlikely to have understood the legal consequences of their promise (or acceptance of one).Footnote ²⁰

Looking beyond Fuller and Olivier, definitions of legal fictions that are close to the one adopted in this paper can also be found, for instance, in Peter Birks's important piece (but one often overlooked in the literature on legal fictions) entitled ‘Fictions Ancient and Modern’ (Birks, Reference Birks, MacCormick and Birks1986). Birks works with a form of action he considers exemplary for understanding legal fictions: action on money paid. At first, this action was restricted to circumstances (the operative facts) in which (1) there was ‘a request (to pay a third party)’; and (2) there was ‘a promise to reply’ (Birks, Reference Birks, MacCormick and Birks1986, p. 87). In the course of time, these operative facts were dropped, such that even when there was proof to the contrary (i.e. proof that there was no request and no promise), the same form of action could be used to regain the money. As Birks says:

An example is Exall v. Partridge, a case from 1799:Footnote ²¹

The case is an interesting one if only, in the present context, because it shows how the court's suspension of certain operative facts can be made dependent on the proof of other facts, not explicitly noted in the form of action. What makes Birks's discussion valuable is his generalised analysis of how legal fictions work:

Birks's analysis here is somewhat more dynamic than Fuller's or Olivier's. He speaks of the ‘fictionalisation’ of A and D, which is in part an acknowledgement, or can be read to be so, of an ongoing process, rather than an isolated act. Although it could be clearer, one assumes that by the ‘fictionalisation of A and D’ Birks does not mean that one can no longer bring an action on the basis of the facts A, B, C and D, but just that there is another way to reach the same normative consequences by way of proving facts B and C, or B, C and Z. In any event, Birks's approach is close to the one adopted here, for he implicitly recognises the ambiguous, flexible or even unstable character of a situation in which the same normative consequences may be applied via a series of alternative operative facts, at least until – in some unforeseen future – a court decides to articulate a more general principle (e.g. the principle of unjust enrichment, and associated rules) that will encompass these various alternative routes.

This sense of fictions being an instrument via which, incrementally, the law gropes its way towards a principle is important. It echoes the point noted above by Fuller by reference to ‘exploratory fictions’. It is a term often acknowledged to be in the spirit of the common law. For instance, Hamilton argues that it is through the use of ‘exploratory fictions’ that ‘common law judges often feel their way incrementally toward some new legal principle or theory’, providing the example of the ‘genealogy of cases giving rise to the “attractive nuisance” doctrine, which involves a fiction that the children were allured or enticed upon the land, and which eventually results in child trespasser law, such that a child being a trespasser becomes ‘one fact to be taken into account, with others, in determining the defendant's duty, and the care required from him’ (Hamilton, Reference Hamilton1988–89, pp. 1465–66).Footnote ²² As Hamilton summarises it: ‘the initial fiction which allowed children to recover for injuries sustained while trespassing “pointed the way” towards development of a full-blown theory of premises liability that allowed recovery’ (p. 1466). Once again, we see the benefits of taking a long-lens view of the use of the device, which, interestingly, may be enabled by focusing on common-law epistemology. More will be said about these benefits in the next part of the paper.

It will not, of course, be possible here to exhaustively review the way the term ‘legal fictions’ has been used in the literature.Footnote ²³ However, a few further remarks need to be made concerning the bounds of the definition adopted here.

First, there has not yet been occasion to observe that it is important to distinguish fictions from ‘virtual entities’ in the law, such as the concept of ‘legal personality’ – thinking of them as virtual rather than fictional helps see the distinction.Footnote ²⁴ Second, but closely related to the first point, it is not clear how helpful it is to think of fictions as part of, or as further illustrating, the metaphorical character of legal language, where the roots of certain legal concepts can be traced back to their distant more concrete beginnings (e.g. concepts like possession, conversion, delivery, estate, etc.).Footnote ²⁵ Although it is an interesting exercise to trace the genealogy of terms used in the law, it is not clear how including fictions under the canopy of metaphors in legal language assists us to understand them better. Third, fictions have often been associated with – especially in the common-law tradition – jurisdictional wars, as devices used precisely in order to acquire jurisdiction at the cost of a competing court. On the one hand, these fictions could be seen more generously as attempts to provide remedies to those who sought them, but it is difficult to ignore the political and economic context, e.g. in the fictional enlargement of the action of ejectment by the King's Bench (in competition with the Court of Common Pleas). By a series of fictions – e.g. ‘no actual lease is made, no actual entry by the plaintiff, no actual ouster by the defendant’ – the action of ejectment ends up being a means for trying title (see Holdsworth, Reference Holdsworth1937, Vol. 7, p. 12; see also Vol. 1, p. 200, and pp. 218–19). The need for recourse to fictions was the result, of course, of the reliance on rigid (or only apparently rigid!) forms of action.Footnote ²⁶ Certainly, this kind of jurisdiction extension has the effect of changing the law, but such instruments are not fictions in the definition adopted in this paper – they are, for instance, not adaptations to difficulties associated with proof.Footnote ²⁷ This is not to say that certain ways in which the law was developed by medieval and early modern common-law courts will not be of interest to a student of legal fictions on the definition adopted here – it is just to say that some uses of them have a decidedly impatient tenor about them: they are not an exercise of tentative cognition in relation to future courts; they are examples of cognition exercised most assertively for the purpose of competing with other courts of one's own age.

Fourth, and following on from the last point, the approach to fictions adopted here is one closely related to the ways in which a court manages or adapts to difficulties of proof in daily legal practice.Footnote ²⁸ We must remember that it has not always been the case – in all legal systems – that those in charge of the law were also evaluating submissions on the facts.Footnote ²⁹ Arguably, the character of law and legal reasoning changes greatly when the courts become not only forums of the law, but also forums of fact. Others have associated the use of fictions with issues to do with proof, but have not always made clear that fictions are a means of adapting to difficulties that claimants have with proving certain (often the same kind of) elements of particular actions – especially, as noted above, proving kinds of intentionality (e.g. malice, intention to be bound, intention to hold for beneficiary, etc.) or causation. Thus, Baker says that:

This is a valuable observation, but we have to be careful: we need to see that it is only after a period of time that it becomes (or might become) clear that a certain operative fact can be dispensed with (the initial cases suspend, rather than dispense with, certain operative facts). Baker would probably agree, for he is mindful of the changing character of a legal fiction: it might at first ‘slip into practice without challenge, and if repeated often enough could become so rooted that a future challenge would be unlikely succeed’; then, ‘when they became familiar in practice, they might be extended into more contentious situations, provided that they were seen to have a just objective’ (Baker, Reference Baker2001, pp. 54–55). Baker here recognises that the law changes by creeping in degrees, pursuing remedies for claimants, and where fictions are one of the devices via which the waters are tested, so to speak. If enough normative momentum is accumulated, and thus enough cases come before courts to warrant it, there might come a time when a court will articulate the principle or rule said to cover the factual pattern of that string of cases. Further, we have to be careful not to return to the mere opportunism of jurisdiction extension, i.e. the idea is that there is a genuine adaptation to difficulties of proof encountered by otherwise worthy claimants, and not the granting of actions for the sake of attaining jurisdictional (and of course revenue) benefits.

The way legal fictions are framed in this paper, then, is as part of what can be understood to be the study of common-law epistemology.Footnote ³⁰ This is an epistemology whose style – if one can speak of it in those terms – is resolutely pragmatic. As Postema put it:

This is a style that keeps close to the particulars (to paraphrase Bacon: see Postema, Reference Postema2003, p. 16), but also keeps ‘an eye on how decisions in particular cases would function in future as examples and rules’ (p. 4). It is a style that prides itself on ‘the flexibility and opportunity for judicial testing and re-evaluation’ (p. 13), recognising that ‘past cases do not preclude deliberation and reasoning in subsequent cases, but rather they invite and focus that reasoning’ (p. 17). As Last Stone put it, ‘common law reasoning is … a reflective process, a form of collective legal reasoning, in which rules and concepts emerge from and are tested against real and hypothetical narrative situations’ (Last Stone, Reference Last stone2007, p. 127), and where there is a discipline and art of ceaseless ‘re-reading’, ‘re-writing’, ‘re-describing’ (p. 132), ceaseless transformation and experimentation, with some formulations (of rules or principles, connected to certain constructed factual patterns) taking precedence for a time over others, until they are potentially replaced with the accumulated effect of what is creatively constructed as another string of cases. Both Postema and Last Stone come close to seeing common law legal reasoning as relational in character, and one in which legal fictions play an important role. To further analyse that role, let us turn to Part II of this paper.

II Evaluating fictions

Part I has already found it necessary to mention the need for taking a more diachronic, dynamic view of the use of legal fictions. This part seeks to systematise those brief observations, and to do so (1) by reference to the discussion of legal fictions as enablers of legal change in the literature; and (2) by recourse to a contemporary example in the context of personal injuries litigation.

It should immediately be said that it is not always the case that legal fictions are portrayed negatively.Footnote ³¹ Nevertheless, the bulk of the literature is negative in tone. Certainly, legal fictions are recognised as helpful, if not occasionally necessary, for change to take place, but this is typically qualified in various ways, e.g. that legal fictions are in use only initially, in the infancy of a legal system, and that mature systems do not have recourse to such allegedly dishonest means of reforming the law.

An influential proponent of this attitude was Sir Henry Maine. In his Ancient Law (1931 [1861]), Maine recognises legal fictions as one of the three ways in which law can be changed – the other two being equity and legislation.Footnote ³² Maine's evaluation of that function, however, is from the beginning qualified and ends up in harsh criticism. He says, initially, that ‘at a particular stage of social progress’ – already a qualification – legal fictions ‘are invaluable expedients for overcoming the rigidity of law’ (Maine, Reference Maine1931, p. 22). His example is the ‘fiction of adoption, which permits the family tie to be artificially created’, and Maine argues that without this fiction ‘it is difficult to understand how society would ever have escaped from its swaddling clothes, and taken its first steps towards civilisation’ (p. 22). Maine criticises Bentham for being unhistorical in his evaluation of fictions, thereby neglecting ‘peculiar office in the historical development of law’ (p. 22). A few lines later, however, Maine insists that legal fictions have had their day. Indeed, he asserts that legal fictions ‘are the greatest obstacles to symmetrical classification. The rule of law remains sticking in the system, but it is a mere shell. It has been long ago undermined, and a new rule hides itself under its cover’ and he concludes that ‘[i]f the English law is ever to assume an orderly distribution, it will be necessary to prune away the legal fictions’ (p. 23). It is somewhat ironic that Maine chastises Bentham for being unhistorical, and yet himself fails, in his evaluation of fictions, to see how they operate over time, i.e. how they enable courts – including contemporary courts in so-called mature systems – to experiment with which operative facts to attach to what normative consequences, and when to sediment in an explicit way some such association that has proven useful. The law – it might be said a little colourfully – is never fully ordered, or better, it is only fully ordered when it is dead. For law to be alive – to be used – means that there are parts of it that are more flexible than others, but all are somewhat tentative, for their meaning (and in some instances, their very relevance) in the next case depends precisely on the facts of that case (and of course how they are constructed).Footnote ³³

There is no doubt that Maine influenced many a generation of jurisprudes and legal historians. In The Spirit of the Common Law, for instance, Roscoe Pound recognises fictions as ‘modes of growth’ (1999 [1963], p. 169), but says that they are ‘primitive … closely akin to the “let's play” so and so of our childhood’ (p. 169). Later he is even more explicitly critical:

In his earlier five-volume textbook, Jurisprudence (1959), Pound devoted more space to understanding and evaluating legal fictions. Here, he was also more thorough in his review of different modes of legal growth. Thus, he makes an initial division between the traditional element of growth and the imperative (or enacted) element. The traditional element contains eight modes of growth: ‘1) fictions; 2) interpretation; 3) equity; 4) natural law; 5) juristic science; 6) judicial empiricism; 7) comparative law; and 8) sociological study’ (p. 449). Fictions, in turn, he divides into three ‘classes’: first, ‘particular fictions’, which are chiefly procedural and are ‘employed to meet a particular type of case or to change or avoid a particular rule or effect a particular isolated result’ (like Maine, Pound cites the fiction of adoption); second, ‘general fictions’, which have ‘a more sweeping operation to alter or create whole departments of the law, introducing principles and methods rather than isolated rules’ (examples include equity and natural law); and third, and most importantly, ‘dogmatic fictions’, which are ‘fictions worked out after the event by juristic thinking in order to give or appear to give a rational explanation of existing precepts’, examples of which include ‘representation as an explanation of liability for the tort of a servant and constructive trust’ (p. 450), as well as ‘implied malice, implied condition, constructive notice, constructive trust, constructive service, constructive fraud, constructive possession’ (p. 455).Footnote ³⁴

Having mapped fictions in this way, Pound further compliments them as ‘the solvent of formalism’ (Pound, Reference Pound1959, p. 461). In fact, he is even more subtle than that, for he recognises why, in certain circumstances, the preservation of rules (at the expense of using a fiction) can be important, e.g. ‘the rule may not have passed out of a religious or semi-religious stage. Or it may be that the rule has a specially solemn form, such as a law of the XII Tables, having been incorporated in a permanent memorial of the settled customs of the community … or a rule … may be protected against change by a widespread fear of return to a condition of justice without law still remembered with dread’ (p. 449). Pound therefore recognises the way in which a fiction can play an important role in preserving certain rules that are important for the community – but, and this is crucial, he fails to see how fictions are needed as an incremental, experimental, tentative form of legal change. Like many other theorists and historians, he takes a fiction to be the introduction of a new rule, not the (potentially temporary, potentially more stable) suspension of an operative fact attached to an already existing rule's normative consequences. Failure to see this function of fictions over time, results in seeing them – like Maine – as primitive and ‘clumsy’:

Pound's evaluation here clearly echoes Maine's, which is even more obvious when Pound asserts that ‘[a]fter a certain stage of legal development, on the other hand, fictions retard growth and clog development’ (Pound, Reference Pound1959, p. 466). Fictions are in use initially: ‘after law has been put in writing’; they are, indeed, ‘the first agency through which the traditional element of a legal system is enabled to grow’ (p. 449). A ‘rational age’, however, must do without them, for they ‘conceal the substance’ (p. 466); resort to them reveals a ‘lack of general ideas’ (p. 462).

Similar sentiments are recorded by contemporary commentators on the use of fictions. Jeremiah Smith, for instance, draws explicitly on Maine to come to the following conclusion:

The problem with this view, as already noted above, is that it freezes the frame at which it evaluates fictions: it only considers the utility of a fiction from the perspective of sufficient normative accumulation to warrant the articulation of an explicit rule or principle; it does not go back to the first uses (i.e. to the first few suspensions of the relevant operative fact), and thus does not evaluate the role of fictions across time. Olivier, who we encountered in the first part of the paper, also takes this distinctly temporally insensitive approach.Footnote ³⁵ He acknowledges that fictions have ‘played an extremely significant and beneficial role in the development of the law’, but he characterises the judges' deployment of fictions as a matter of ‘concealing their law-making activities’, and one which, ultimately, can cause more harm than good:

Again, like Smith, Olivier focuses his evaluation on the moment at which the fiction is ripe for articulation, and not the period over which the fiction has been used to explore whether articulation is justified and/or useful. Olivier's full evaluation is worth quoting in its entirety. He distinguishes the following three kinds of case:

1. 1. ‘As far as the adoption by judges of historical fictions is concerned, the judge is obliged to adopt the rule forming part of the common law, but should reject its fictional form or fictional basis. The judge is obliged to formulate the rule in direct, non-fictional terms, and to find and formulate the true principle behind the fiction.’

2. 2. ‘Whenever a judge is compelled to apply a statutory fiction, he should determine and formulate the legal rule envisaged by the legislature and the principle on which the rule is based, and should give effect to the rule and its true basis rather than to the fiction.’

3. 3. ‘Whenever a judge finds it necessary to create a new fiction, it indicates that the major premise he is applying is wrong. The judge should then determine whether the major premise is the correct one to apply and if there does not exist a better major premise. Frequently an alternative and better major premise exists which, if applied, will lead naturally and logically to the desired result.’ (Olivier, Reference Olivier1975, p. 151)

Olivier, here, like many others, overestimates the value of logic and explicit articulation, and underestimates the value of tentative, fact-anchored, exploration of what operative facts should be attached to what normative consequences. In this respect, his third scenario is the most misleading, though one might even wish to argue that, in the first two, Olivier misses the benefits to be gained from continuing to tread carefully, i.e. incrementally and pragmatically.Footnote ³⁶

The first step towards a more positive evaluation of legal fictions will need to pay attention to the practical context in which they are used. Unsurprisingly, perhaps, we need to turn to historians rather than theorists for relevant insights in that respect. Thus, Baker has noted that the use of fictions in the early common law was rather secretive (which of course raises a methodological problem for any historian of the common law), for any judicial approval to relieve ‘the plaintiff from proof of certain facts which had to be alleged’ (in a form of action) was ‘given off the record, at the assizes, and the verdict was returned in common form’ (Baker, Reference Baker2001, p. 53; see also the discussion on pp. 54–55). The granting of remedies – on broadly equitable grounds (though of course not articulated explicitly that way) – preceded any ‘theoretical justification’; the law, in other words, changed incrementally, pragmatically, as a result of remedies creeping in by degrees (pp. 56–57). In time, the normative accumulation of cases in which such remedies have been granted might result in the articulation of certain coherent theoretical principles, though this may be driven by academic rational reconstruction of the cases (as with, arguably, unjust enrichment: see the discussion in Baker, Reference Baker2001, p. 57; see also Ibbetson, Reference Ibbetson1999, Chapter 14).

Another contemporary historian of the common law, who has also paid attention to the practical context in which fictions were deployed, is Milsom. Indeed, Milsom uses the practical context to dispel the myth that fictions were ‘packs of lies’, as Bentham would say, used by judges who were surreptitiously ‘stealing legislative power’ (Milsom, Reference Milsom2003, p. 27). He does this by downplaying the active role of the judges in creating fictions. For Milsom, authorship of fictions should lie more with those ‘countless individual lawyers through the centuries, each concerned not with “the law” as such but with a small immediate predicament of his client’ (p. 27; Milsom also recognises the role of jurors), and not with the judges. The role of the judges, according to Milsom, ‘was essentially passive. They might facilitate the later stages of a fictional development, but in the important early stages they just accepted results reached by others’ (this is perhaps especially so in cases where jurors, ‘shut up by themselves’, returned ‘blank verdicts’ that ‘enabled them to do their own justice’: p. 27). While not wishing to challenge Milsom's thesis (concerning the authorship of fictions in fifteenth-century common law), and thus not wishing to question the vital role of the pursuit of remedies for their clients by their lawyers, nor the critical role of the ‘dialogue between judges and jurors’ (p. 45), it is not inconsistent to add that insofar as judges were involved (even in the act of ‘formalising’ some result reached by others), and thus played their part, they did so not necessarily or exclusively for political reasons (wanting to preserve legitimacy), but also, or more so, out of fear of the unforeseen and unforeseeable consequences of introducing a new rule or principle, preferring the more incremental, pragmatic, slow change of the law. Fictions, one might say, are a way of slowing down change – of treading carefully – creating resources for future courts, but ones which they are not compelled to respect. Again, attention to the practical context is clearly important – but that should include not only relations between claimant's lawyers and judges, and between judges and jurors, in instant cases, but also dialogues between courts across time.

That pragmatic incrementalism is the spirit of change in the common law is itself widely recognised in the theoretical as well as the historical literature. Part I above has already mentioned Postema's and Last Stone's characterisations of common-law reasoning. More specifically on the theme of change in the common law, historians have shown that:

Leaving aside the question of appearances, Ibbetson is recognising here that legal change occurs via the ‘initiative’ of the ‘litigants’ (see Ibbetson, Reference Ibbetson1999, p. 299), and thus, in a bric-a-brac fashion, adding a little bit here, taking away a little bit there – in a sense, precisely playing with the resources one has, reshaping them as the need arises, but also revising them (reconstructing the alleged pattern of past cases) so as to fit with the justification for the remedy in the instant case. Explicit articulation and overall coherence may occasionally be attempted, but even then essentially for pragmatic reasons: where there are, for instance, practical advantages to be gained (e.g. clarity for litigants) in generalising from what is (at that point) perceived to be the overall normative pattern of a certain string of cases.

Fictions are, of course, not the only mode or device of reasoning used by judges to enable the incremental change that is characteristic of the common law.Footnote ³⁷ Hutchinson, in his Evolution and the Common Law (2005), takes note, for instance, of techniques for avoiding or loosening the binding force of precedent, including reversing, overruling, distinguishing, not following, asserting the precedent was given per incuriam, characterising subsequent cases as ones that have been doubted in other cases, arguing the precedent was based on a faulty interpretation of earlier cases, claiming the scope of the precedent is unclear, and arguing the precedent has been criticised by academic commentators (see Hutchinson, Reference Hutchinson2005, p. 7).Footnote ³⁸ Hutchinson's overall view is sometimes very radical – he applies an evolutionary model he understands as ‘blind, contingent, haphazard and entirely opportunistic’ (p. 36), going on to characterise law as a ‘work-in-progress’ that is ‘always situated inside and among … the ideological forces at work in society’ (p. 18).Footnote ³⁹ However, when Hutchinson says, for instance, law is ‘never complete or finished’ (p. 18), and that it is ‘a rutted and rough road that has innumerable twists and turns and no particular destination’, such that ‘any particular route taken has been chosen from among the countless and constantly proliferating possibilities for change’ (p. 15), he offers some helpful metaphors for understanding the common law, as well as discarding unhelpful ones (such as inevitable progress, natural harmony, universal telos, etc.).

The common law, then, can only ever be artificially frozen at any point in time: it is better understood as a dynamic domain of possibilities, with each rule or principle better conceived as requiring an understanding of how it might and might not be stretched by way of a redescriptionFootnote ⁴⁰ of the supporting facts of past cases, and thus a reconstruction of the relevant factual pattern.Footnote ⁴¹ This does not mean, as Hutchinson also recognises (2005, p. 10), that ‘anything goes’ – for there are many pressures for continuity and stability. Indeed, the techniques – fictions and precedential techniques among them, but including also analogical reasoning and perspectival reasoningFootnote ⁴² – are better thought of as enabling the striking of a balance between flexibility and stability. There is, if you like, a dynamics of relevance at the heart of the common law – one driven by solving particular problems,Footnote ⁴³ but always with a view both to the past and the future – and the devices, of which fiction is one, are ways of managing relevance: making sure it is not too unpredictable, but also not too static. As Hutchinson says, ‘the common law is a messy, episodic and experimental effort to respect and adapt to the contingent demands that the political and social milieu places upon it’ (p. 270); its success has been ‘this ability to be flexible, open, experimental and adaptable’ (p. 268).Footnote ⁴⁴ ‘The best that can be hoped for’, adds Hutchinson, is that ‘the common law remains supple, experimental and pragmatic’ (p. 236).

Are fictions – in the broad sense conceived in this paper – a thing of the past? Has the common law reached the kind of maturity that some critics of fictions think is attainable, and indeed, according to some, has been attained? This paper would argue that the contrary is the case: fictions continue to be used, and rightly so. Without them, the common law would lose one of its most treasured instruments for creating potential change; for exploring and experimenting with the extent to which the law might or should be changed; for drawing up dynamic resources for future courts in deciding unforeseen and unforeseeable factual scenarios.

Take, as just one brief contemporary example, the decision of the House of Lords in McGhee v. National Coal Board (1972),Footnote ⁴⁵ and the string of cases since then. A neat overview of the significance of the caseFootnote ⁴⁶ is provided by Lord Hope of Craighead in the 2002 Personal Injuries Bar Association Lecture (Hope, 2003). The facts of the case – described in such loving detail by Lord Hope – were, briefly, that Mr McGhee was employed by the National Coal Board (NCB) to clean out brick kilns. Mr McGhee was exposed to clouds of abrasive brick dust, which, as a result of the NCB not providing adequate washing facilities, he only washed off once he got home after cycling home ‘caked with sweat and grime’ (to use Lord Reid's own description, at McGhee, p. 1010). After five days, Mr McGhee developed dermatitis and stopped work. There was medical evidence that the dermatitis was caused by the working conditions in the brick kilns, but the evidence also showed that cycling back home after work added materially to the risk. In the Court of Session, Mr McGhee failed because he could not show that the breach of duty (i.e. failing to provide adequate washing facilities) caused his injury. There was, in short, no positive evidence that it was more probable than not that he would have avoided contracting dermatitis if those facilities had been provided.

The House of Lords allowed the appeal and found for Mr McGhee. There were five separate opinions and, as Lord Hope notes, ‘an examination of their reasoning will show that it was far from uniform’ (Hope, 2003, p. 588).Footnote ⁴⁷ Mr McGhee's problem, as will be evident from the above description, was causation – he could not discharge his onus that on the balance of probabilities the NCB's breach of duty caused or materially contributed to his injury.Footnote ⁴⁸ The Lords, then, were faced with a rule that attached a certain normative consequence (satisfaction of the requirement of causation for the purposes of establishing liability) to certain operative facts (including the onus), namely material contribution to injury. Counsel for Mr McGhee argued that the NCB's failure to provide adequate washing facilities ‘materially increased the risk’ of his contracting dermatitis (p. 592) – in essence, they argued for suspending, at least in this case, the operative fact of ‘material contribution’.Footnote ⁴⁹ The House, in order to reach a conclusion they had thought justifiable – namely, that Mr McGhee was a worthy recipient of a remedy, and that the NCB ought to be found liable – had to reframe the facts (and who had to prove what) vis-à-vis the law, and it could do so precisely by suspending the required operative fact. In the broad sense, then, one might argue that a fiction was deployed: because of the absence of proof, a required operative fact was suspended, such that, as a result, the NCB could be treated as if they caused the injury (‘as if’ because, on the required operative fact, they would not have ‘caused’ it). Further, one of the Lords – Lord Wilberforce – himself characterised that the inference ‘from the fact that the negligence increased the risk that it had materially contributed to the disease was something of a fiction’ (p. 601). It is useful to quote Lord Wilberforce's own words (the concluding paragraph of his speech):

This quotation is instructive, for there is explicit recognition of the intimate link between the use of fictions and the difficulties encountered concerning proof (causation being one of the commonly recurring contexts in which such difficulties are encountered, intention being the other). There is also explicit reference to the redescription of past cases, via the mode of analogy – indicating, also, the link between the use of the device of fictions and the use of the mode of analogy. The suspension of a required operative fact is assisted by the analogical extension of a string of cases now asserted to be relevant to the present problem. Facts extended by analogy are like the bricks in the bridge of the fiction.

As has been mentioned – no doubt at ad nauseam – in this paper, it is not sufficient – both in order to understand but also in order to properly evaluate the use of fictions – to focus on just one instance, on just the single use of it in a particular case. If we did that in the present case we might think that the Lords took on powers that properly belong to the legislature, and that they tried to hide their law-making by concealing it under the cloak of a fiction.Footnote ⁵⁰ What subsequent cases have brought out – in a way that could not have been foreseen – is that the decision in McGhee has provided a fertile though also controversial resource that has allowed the law to adapt and remain responsive to the claims of plaintiffs in personal injury litigation. Some subsequent judges have refused to treat McGhee as a relevant resource – e.g. Lord Bridge of Harwich in Wilsher v. Essex Area Health Authority (1988)Footnote ⁵¹ said that McGhee:

This effectively ignores the invitation offered by the fiction in McGhee, i.e. to suspend the operative fact, and to potentially systematise and make explicit that suspension, and reintroduces the operative fact of the rule, i.e. it treats material contribution as continuing to be the necessary operative fact (rather than materially enhancing the risk of injury).

Other judges, however, have taken up the invitation, and indeed attempted to formulate a principle so as to unify both the facts of McGhee and the facts of the case being dealt with, e.g. in Fairchild v. Glenhavon Funeral Services Ltd and Others (2002).Footnote ⁵² It is interesting that in that case Lord Rodger observed that (in Lord Hope's summary) ‘the House [in McGhee] could have proclaimed more clearly and more openly that they were stating a new principle’ (Hope, 2003, p. 598; see Fairchild, 2002, para. 144). This already is a taking up of the invitation, for it presents McGhee as a principled decision. However, from the perspective of this paper, it is not a drawback of the decision in McGhee that it was not (allegedly) clearer about the effect of its judgment: that was precisely for later courts to either take up or not. The point is that the House in McGhee was proceeding tentatively, creating potential resources, but keeping principled articulation at a minimum,Footnote ⁵³ for this enables future experimentation and exploration of the benefits – including testing impact on other nearby rules and principles – of adopting a rule or principle. Lord Reid in Fairchild, as Lord Hope notes, says on the back of a reading of McGhee that ‘in a case like this we must take a broader view of causation’ (Hope, 2003, p. 599; see Fairchild, 2002, para. 21). What becomes visible here is how the fiction is taken up, transformed into what might not yet be a principle, but at least articulates the need for a different attitude – in this case, ‘a broader view of causation’. In doing that, it also reads back, establishing a factual pattern – ‘in a case like this’ – which further sediments the relevance of McGhee, but also creates resources for future analogical extension.

The debate over the relevance of McGhee is by no means over. Nor has the adventure of ‘the principle’ ceased – it being the point that there is not one formulation, for it is continually being reformulated in the light of future cases. For instance, in Sienkiewicz v. Grief (UK) Ltd (2011)Footnote ⁵⁴ and Willmore v. Knowsley Metropolitan Borough Council (2011),Footnote ⁵⁵ the Supreme Court gave further normative impetus to the operative fact of materially increasing risk (potentially displacing the operative fact of material contribution),Footnote ⁵⁶ and thus arguably extending the legacy of McGhee, via Fairchild.Footnote ⁵⁷ It remains to be seen whether future courts will, in their turn, continue endorsing that normative impetus, or will instead seek to retrospectively narrow the significance of McGhee and/or Fairchild, turning them into exceptions or sui generis outcasts.Footnote ⁵⁸ For the moment, it certainly looks as if – as a result of subsequent courts largely taking up the invitation offered by McGhee – the fiction first entertained in McGhee will not die a quiet death, and indeed may very well blossom, past a change in attitude to causation in personal injury cases, to a more general principle of causation (or at least the proof of causation) in tort law (or indeed other areas of the law). We do not know – that is part of the magic and utility of fictions, and, simultaneously, the magic and utility of the common law.