Tort Law Defences: a Defence of Conventionalism

‘This is the peer reviewed version of the following article: Eric Descheemaeker, “Tort Law Defences: A Defence of Conventionalism”, (2014) 77(3) Modern Law Review 493, which has been published in final form at This article may be used for non-commercial purposes in accordance With Wiley Terms and Conditions for self-archiving.’

Tort Law Defences: A Defence of Conventionalism

Eric Descheemaeker

(Edinburgh)

James Goudkamp, Tort Law Defences, Oxford: Hart Publishing, 2013, 270 pp, hb £60.00

We live in an age where orthodoxy has become suspect. Obligations lawyers from the entire English-speaking world can be persuadedto gather at a conference dedicated to its being ‘challenged’.[1]When one remembers that the word means ‘right thinking’, this is rather extraordinary: one could be forgiven for assuming that right thinking, far from being attacked, should be encouraged. This must mean that the word is taken in a looser sense, presumably as a synonym of ‘traditional thinking’ or ‘convention’. Now it is absolutely true that conventional thinking may be wrong; it might be heterodox, and indeed it is one of the main tasks of legal academia to constantly examine and re-examine received wisdom. But we must also be careful not to assume, perhaps unconsciously in an environment which institutionally values ‘paradigm shifts’, that new thinking is necessarily better than the old: it may or may not be. While truth is the supreme value, meaning that no one should be afraid to discard previously accepted ideas which are shown to be incorrect, stability – all things being equal – is also a desirable good, for one cannot build safely on ground that is constantly moving.

It is against this background that the present review article seeks to defend a view regarding defences in the law of tort which has been described in an important recent book, James Goudkamp’s Tort Law Defences (Hart 2013), as ‘conventionalism’ – by contrast with the author’s views, described as ‘radical’.[2]Indeed, the conventional view is so conventional that it has not been given much thought at all by tort lawyers, for whom the prospect of conceptualising defences is likely to appear as exciting as crumbs in a tin box. Whether or not we end up agreeing with its main theses, one of the greatest merits of Dr Goudkamp’s book is arguably that it forces everyone with a serious interest in tort (and more widely in private law) to think harder about the topic. While most tort lawyers tend to regard defences as a collection of appendices to various causes of action, there is no denying that they are of considerable significance, both theoretically and practically. It follows that anyone critical of the author’s views should probably start by sharing in a collective act of repentance for the fact that this is the first serious study devoted to the subject in the common-law world.

This review article focuses on two – intertwined but analytically distinct – issues which I regard as foundational to the book,in that jointly they make up the ‘radical’ understanding of defences set forth in the study. These are (i) the idea that a defence is defined as ‘a rule that relieves the defendant of liability even though all the elements of the tort in which the claimant sues are present’;[3]and (ii) the idea that defences are distinct from torts, rather than part of the definition of the causes of action. These views are misguided in my judgement, even though they do raise important questions that we all need to examine carefully,and provide much ammunition for us to do so. The first two partsof this article examine the author’s arguments in respect of both questions, contrasting them with the conventionalapproach, and argue that there are least four good reasons to retain the traditional thinking. The third part changes tack and seeks to highlight how the author’s attempt to consider defences as a whole is, despite the above criticisms, extremely precious – for a reasonon which the author himself does not rely, namely, that it paves the way for a reclassification of defences which would be highly beneficial for the rationality and accountability of the law of torts.

‘RADICALISM’ AND ‘CONVENTIONALISM’

The first two sections of this part are concerned with the author’s professed ‘radicalism’ in respect of tort law defences, and highlight those of his positions which I regard as most significant: namely, as explained, his definition of a defence and the main consequence this definition has on the relationship of defences with torts.[4] This part concludes by articulatingthe conventional understanding of defences on both points.

The Definition of a Defence

The Difficulty of Definitions

What is a defence? This question if of paramount importance in the present context; indeed we cannot hope to go anywhere until it has been settled. But anyone interested in definitions (which ought to be everyone caring about right thinking) will know how difficult an art this is: omnis definitio periculosa est, not simply, as Javolenus tells us, because it is in danger of engraving in marble an understanding that misses the mark;[5] but more fundamentally because we struggle to know where to find the authority on which basis we can pass a judgement.

Who is to decide what a defence is? Despite the ever-present temptation to assert that a word only means what people make it to mean, to the effect that usage is the only criterion, this cannot possibly be true: everyone, even the keenest supporterof unbridled language evolution, will accept that there is such a thing as an incorrect use of a term. A use is incorrect when the person does not intend to convey what is objectively the (or an) appropriate meaning; and while this meaning might ultimately evolve, unless and until it has, any attempt to use the word in a different context should be regarded as inappropriate. The possibility of a wedge existing betweenusage and what we assert to be the (or a) ‘proper’ meaning of the term has therefore to be right.

The question then becomes how wide this wedge can be, for it is equally obvious that the process of normative refinement of a term can only take us this far. There comes a point where we are driven so far from what people commonly acceptthat the bystander will have to take the view that the rationaliser ought to carve out a new concept, rather than twist the existing one out of recognition. More fundamentally, the rationaliser needs some authority for his process of refinement and exclusion of ‘wrong’ usage. This could be an appeal to – for instance – history, logic, the degree of fit with related concepts, or etymology(the latter being especially tempting because it seems to give us ‘the right’ answer, but it is evidently dangerous given the depth of assumptions it makes about underlying rationality). At any rate, the authority has to come from outside usage itself, for out of an ‘is’ no ‘ought’ can come.

Defining Defences

Turning to the subject-matter of defences, Dr Goudkamp starts his definitional quest by observing that writers have been using the word defence in a ‘slapdash’ way.[6]This is observably true and most of us must plead guilty as charged. ‘Defence’ is a word that everyone interested in tort (though not only in tort) is bound to use with great regularity; yet it is remarkable that even those with a strong interest in tort theory and taxonomy are unlikely to have given the underlying concept more than fleeting attention.[7]No-one can seriously disagree with the author when he asserts at the outset of his study that ‘the neglect of defences’ is one of the ‘most striking features of the tort law literature’.[8]

His second stepis to identify clusters of usage: this is the descriptive stage, rather similar to that of a dictionary, content to report its observations and put some order in them. Here, the author identifies five understandings of defences:

(i)defences as ‘denials of elements of the tort in which the claimant sues’;

(ii)defences as ‘liability-defeating rules that are external to the elements of the claimant’s action’;

(iii)defences as ‘principles that diminish the claimant’s relief’;

(iv)defences as ‘rules in respect of which the defendant carries the onus of proof’; and lastly

(v)defences as ‘the final element of the claimant’s cause of action’.[9]

The author then proceeds to rule out four of these possible definitions, essentially on two bases: either thatthose who use the term in such ways cannot possibly mean what they say, becauseit leads to consequences they would regard as absurd–e.g. (iv) would lead to describing as a defence the provision that enables defendants to amend their pleadings after service –or because, even though they might quite plausibly mean what they say, such usage leads to other undesirable results –for example (i)obscures the distinction between defences and denials,[10]and (iii) that between liability and remedies, both sets of distinctions the author rightly regards as being foundational.

The last stage is to select one as the ‘correct’ definition, in this case (ii), i.e. defences as ‘liability-defeating rules that are external to the elements of the claimant’s action’.[11] In other words, for Dr Goudkamp, a defence is a plea in confession and avoidance: the defendant recognises that he has committed a tort (in the sense that all the ‘ingredients’ of liability are made out), yet asserts that he ought not to be held liable because of the existence of a further set of facts making out what the law recognises as a defence. This definition is not reached simply by a process of elimination, for one could not rule out that none of the identified usages would be satisfactory, but also because it ‘facilitates clear thinking’[12] about the law of torts, bringing out in particular the distinction between tort and defences (our next point below), considered by the author to be of paramount importance.

It follows that the adopted definition is based mostly on what Dworkin would call ‘justification’, and Dr Goudkamp recognises that its ‘fit’ is less than perfect.[13] In particular, it means that the one doctrine which would almost certainly be the most cited if we asked a random sample of lawyers to name ‘a tort law defence’, namely contributory negligence, will be discarded as being concerned with remedies rather than liability.[14] While one might object in principle to a line of thinking which rejects a possible definition – (iv) –on the basis of lack of fit, yet goes on to propose one that seems no better and possibly worse in this respect, I will not throw the first stone at the author: in part because, if current use is based on imperfect rationality (as it almost invariably is), something will have to give way in the process of refinement; and in part because I am actually sympathetic, albeit for different reasons,[15] with the view that contributory negligence is not ‘properly’ a defence.

Defences and Burden of Pleading

It is important to highlight a principal consequence of the author’s understanding of defences, which is this: defences are not to be defined, as commonly assumed, by looking at the procedural question of who bears the onus of proof of what. While the author is happy to recognise that the distinction between torts and defences commonly dovetails with the division of pleading between claimant and defendant, this for him is no more than a ‘practice’, which in fact is not ‘universally followed’.[16] Indeed, for Dr Goudkamp, the definition of defences is at heart a substantive definition. For the author, strongly influenced by the work of Kenneth Campbell in the field of criminal law,[17] the offence – here, the tort – is identified by the fact that there exists ‘a prima facie reason not to perform’ the act.[18] It follows that the criteria for identification are extrinsic to the law and find their origin in judgements of a broadly moral nature (the loose term ‘philosophical’ can be used to highlight their extra-legality). However, the argument continues, there can exist stronger counter-reasonswhy the person engaging in the conduct should nonetheless escape liability: these are defences to the offence. These counter-reasons are of two sorts: they can either pertain to the reasonableness of the defendant’s conduct, explaining why it should be permissible (what the author calls ‘justification defences’, of which examples would be self-defence and responsible journalism),[19] or to considerations extrinsic to it, which explain why the defendant should not be liable even though his conduct was irredeemable(‘public policy defences’, such as immunities or limitation bars).[20]In neither case is the question of who bears the burden of pleading those countervailing reasons (or their absence) determinative.

It might be helpful to illustrate the logic of reasons and counter-reasons by using the author’s own example of sexual battery. Everyone agrees that an act of sexual penetration performed by the defendant on the claimant is not actionable if the claimant consented to it. The question is whether consent is a defence to the tort of sexual battery or whether lack of consent is factored into the definition of the tort, to the effect that sexual battery is definitionallynon-consensual intercourse. For Dr Goudkamp, the answer is the latter, there being no reason to believe, ‘according to our sexual morality’, that any sexual intercourse is prima facie wrongful.[21] Because it would be absurd to think that any such intercourse needs to be justified, the wrong has to be defined in such a way as to include lack of consent. Consent is not a defence: without ‘unconsentedness’ there is no wrong to confess and avoid in the first place. To re-emphasise the earlier point, the position would remain the same if for whatever reason the law decided to switch the onus of proof on the defendant, requiring him to prove consent on the claimant’s part.

Dualism

Such a substantive definition poses considerable challenges in terms of identifying defences, to which we shall return.[22]It also has this striking and far-reaching theoretical consequence that, on this understanding, defences will be altogether separate from torts—a view which can be described as ‘dualism’. For the author, this is true not only of what he calls public policy defences but also of justifications based on the defendant’s conduct. Where most people, as explained in the next section, would see the availability of such a defence as an element which prevents a tort from arising, Dr Goudkamp sees it as a way to account for the commission of a tort. As the author puts it, it is ‘crucial to realise that … an act which falls within an offence definition constitutes a wrong even if it attracts a defence’.[23] To use a simple example, the defendant who has hit another in self-defence has, on this reading of the law, committed the tort of battery (‘confession’); yet he escapes liability (‘avoidance’) because he had a good reason for acting as he did, which the legal order recognises by granting him the namesake defence.[24]Again, The Times really did defame Mr Reynolds, even though it escaped liability because its conduct amounted to responsible journalism.[25]In other words,a justification defence does not deny wrongdoing, it explains it.[26]

Conventionalism

From the assertions of the radicals, as fleshed out above, we can deduce what the non-radicals argue, those which Dr Goudkamp calls the ‘conventionalists’.[27] Now it is likely that no-one would want to define themselves as a conventionalist, not simply because the label is unattractive, but because the positions they are supposed to adopt are likely to be regarded as so basic as not to be worthy of any positive engagement. Though there are exceptions,[28] most people are conventionalist by default and might only become aware of it when challenged in their position by ‘radicals’ or others. In spite of this, the author is undoubtedly correct to my mind when he ascribes the following views to prevailing thinking.

Contra Dr Goudkamp’s theses, conventionalists would assert two things. First, that defences are, within the ‘ingredients’of tortious liability (i.e. the elements required for the success of the action, whether they be defined positively – by their presence –or negatively – by their absence), those which fall on the defendant to prove:[29] in other words, the definition is a proceduralone.[30]Second, when it comes at least to justification defences– which stand at the heart of the category of defences – their existence prevents the relevant tort from arising: that it to say, it reverses the provisional finding of liabilitybrought about by the claimant successfully ticking all the boxes on his ‘to-prove list’(thereby making out the prima facie tort). To borrow the terminology used in the book, thesedefences do not excuse wrongs, they deny it has happened: in Arthur Ripstein’s words, ‘[j]ustifications exculpate by showing that an apparently wrongful act was not wrongful’. It seems fair to say that this is how most tort scholars would think about it, if only by default.

FOUR PROBLEMS WITH RADICALISM

We are thus presented with two views of defences. To the sceptic, tempted to dismiss the matter as being a dispute over how many angels can dance on the head of a pin, it must be emphasised that the argument goes to the heart of what constitutes a wrong. It is therefore of considerable importance both theoretically and, in a more indirect way, practically.