1

THINKING ABOUT TORT LAW –

WHERE DO WE GO FROM HERE?

Nick McBride*

I. WHERE WE ARE NOW

In August 1900, the German mathematician David Hilbert addressed the International
Congress of Mathematicians in Paris. In the course of his speech he set out a number
of mathematical problems that were as yet unresolved, with the object of challenging
his audience, and future generations of mathematicians, to solve them. The original
speech mentioned 10 problems, but the published version contained 23. By the year
2000, mathematicians had fully or partially resolved 17 of the 23 Hilbert problems,
and in doing so fundamentally extended humanity’s understanding of mathematics
and its foundations.
The legal academy has no equivalent to the Hilbert problems. In this way, law as
an academic discipline is very much like English or history – everyone pretty much
goes their own way, coming up with their own projects to work on and issues to
explore, some of which prove to be fruitful, others not. I think it’s a pity that legal
academics work in this way, for three reasons:

(1) Academics who work alone tend to avoid addressing really big, major questions
because they are just too daunting to be taken on alone. It would be like trying to
climb Everest on your own. So legal academics tend to focus on narrower, more
manageable topics – ones where one can be confident that six months’ or a year’s
worth of work might yield a decent publishable article. Only the young and foolish try
to focus on anything big; and they soon wise up.1

(2) Where different academics do happen to work on the same questions, all sorts of
institutional pressures – in particular, the need to make a name for yourself –
encourage them to find points of disagreement between themselves and make the
most of those disagreements. They never seem to make any attempt to come together,
and work together, to resolve their differences. Instead, the legal journals are littered
with a series of Mexican stand-offs between: (a) a few academics who seem to take
one view of a question; (b) a few academics who seem to take the opposite view; and
(c) a few academics on the sidelines, arguing that sides (a) and (b) are ‘really’ talking
past each other and addressing different issues, or that sides (a) and (b) are ‘really’
saying the same thing. And no one’s backing down.

*
Fellow, Pembroke College, Cambridge. I am very grateful to Jane Stapleton, Peter Cane and Sandy
Steel for their comments on earlier drafts of this paper.
1
I remember attending a talk when I was a student in Oxford where Mark Kelman – one of the biggest
names in ‘critical legal studies’ – observed that academics tended only to focus on big issues when they
were young, and as they grew older they would focus more on much narrower and less interesting
issues like planning law (his example, not mine).
 2

(3) The lack of any sense among legal academics that they form a community of
scholars who are dedicated to working together on resolving issues and questions that
are important and need to be resolved is dispiriting and destabilising. Legal academics
live a Nietzschean nightmare of having to create their own tables of values, their own
horizons of what is important. It’s a very hard trick to pull off successfully – it
requires that you immerse yourself so deeply in your work that you stop thinking
about whether what you are doing actually matters, and start simply assuming that it
does. But not many people are capable of doing this – most people have more
perspective – and the lucky/unlucky ones who can’t, have to live with constant
uncertainty about what they ‘should’ be doing with their lives as academics, and
whether any of it really matters.

Thus the current state of the legal academy, so far as I see it: a lot of individuals
working on their own on quite narrow issues; some debates on particular issues, going
nowhere fast; and none but the most narrow-minded feeling that great about
themselves or what they are doing.
You may think I’m far too pessimistic, but I’m not the only one. 2 Where I am
not at all pessimistic is in my belief that we can do much better. And we will, if we
come together as a community, determined not only to work together on questions
and issues that actually matter, but also determined to resolve those questions and
issues in our lifetimes and not leave them hanging for future generations of legal
academics to pick over and use as a means of racking up research points.
This paper represents my own immodest attempt to try to foster the sort of sense
of community among tort scholars that I think is desperately needed at this time. To
do this, I will do two things. (1) I will set out a series of questions that I think tort
scholars should be working on, and working on together, to resolve. (2) I will set out
the different answers to those questions that are currently floating about in the
academic literature, so that they are all gathered together in one place. (3) I will make
some suggestions as to how we might determine which of these various answers are
correct.

II. THE FIRST QUESTION

The first question is obvious. It is – What are we talking about, when we talk about
tort law? We can hardly think about tort law without first having some idea about
what we are thinking about when we think about tort law. And it is a sign of some
trouble in tort law scholarship that some of its most distinguished names grow evasive
when pressed on this issue. But we must resolve it – we cannot make any progress in
understanding a subject that isn’t even a subject. Fortunately, there are some answers
to this first question that are available to us.
For a long time, tort scholars have assumed that:

2
See Pierre Schlag, ‘Spam jurisprudence, air law, and the rank anxiety of nothing happening (a report
on the state of the art)’ (2009) 97 Georgetown Law Journal 803, 804: ‘...American legal scholarship
today is dead – totally dead, deader than at any time in the past thirty years. It is more dead, vastly and
exponentially more dead, than critical legal studies was ever dead during its most dead period.
Nothing’s happening.’
 3

(A) Tort law determines when one person – D – will be held liable to pay
another – P – compensation for a loss that D has caused P to suffer.

For those who believe that (A) is true, tort law is centred around the concepts of loss,
causation, and liability to pay compensation.
For the last 15 years or so, (A) has come under sustained attack. In Birksian
terms,3 (A) is criticised for converting tort law – which, as even its name suggests, is
an area of law focussed on an event – into an area of law that is focussed on a
response. Instead, it is argued that:

(B) Tort law determines what basic rights we have against other people, and
what remedies will be available to us when those rights are violated.

Where P will have a basic right against D that D act (or not act) in a particular way if:
(i) the law imposes a duty on D to act (or not act) in that way; (ii) it imposes that duty
on D for P’s benefit and not for anyone else’s benefit; and (iii) P does not have to do
anything special for the law to impose that duty on D, such as paying D money.
Given this definition of what a basic right is, (B) can be restated as:

(C) Tort law determines what basic duties other people will owe us, and what
remedies will be available to us when those duties are breached.

(C) and (B) say exactly the same thing. (C) is probably more illuminating, and less
liable to give rise to confusion. But it must be admitted that (B) has a bit more zing
and appeal about it.4
John Gardner is unhappy with (B) (and, by extension, its alternative formulation
in (C)). He thinks that (B) does not effectively distinguish tort law from the law of
equity – or more accurately, that bit of the law of equity that is concerned with
equitable wrongdoing.5 He would like to say that:

(D) Tort law determines what basic rights we have against other people that
they not cause us loss, and give us a right to sue for reparative damages when
that right is violated.6

John Gardner thinks that this helps to distinguish tort law from the law on equitable
wrongdoing as ‘The main reason why the defendant has duties to the plaintiff in the
law of torts is to protect the plaintiff from losses... This contrasts with the position in

3
Peter Birks, An Introduction to the Law of Restitution (Clarendon Press, 1989), 39-40.
4
The first and second editions of McBride & Bagshaw, Tort Law (2001 and 2005 respectively)
described tort law as being made up of ‘duties owed to others’ and remedies for the breach of those
duties. Under the influence of Robert Stevens’ Torts and Rights (OUP, 2007), the third and fourth
editions (2008 and 2012 respectively) switched into using the language of rights to describe the
fundamental building blocks of tort law. Rob has admitted ‘I could have entitled my book... Torts and
the Duties We Owe One to Another but that wouldn’t have been quite so snappy’ (Robert Stevens,
‘Rights and other things’ in Nolan and Robertson (eds), Rights and Private Law (Hart Publishing,
2012), 117). But snappiness has its dangers: see the extended critique of the way Rob uses the word
‘right’ in McBride, ‘Rights and the basis of tort law’, also in the Nolan and Robertson volume.
5
We have to make this qualification because there are plainly some areas of the law of equity that form
part of tort law – for example, the rules on when and what equitable remedies (such as an injunction)
will be available when someone commits a tort.
6
See John Gardner, ‘Torts and other wrongs’ (available on SSRN).
 4

equity where the main reason for the defendant’s duties is to secure that the
defendant’s dealings are conducted for the plaintiff’s advantage, and not the
defendant’s own.’7
I think there are two major problems with (D).

(1) Many of the basic rights that tort law gives us are not obviously designed to
protect us from suffering losses. I’m thinking here of such things as:
(i) the right that a patient has to be told of the (non-insignificant) risks of an
operation that her doctor is suggesting that she have;
(ii) the right that the beneficiary under a will has that the solicitor who has taken
on the job of drawing up the will take care to ensure that the will is valid, and is
drawn up expeditiously;
(iii) all of the rights that property owners have that other people not interfere
with that property (where such interference can occur without in any way causing the
property owner loss).

(2) For (D) to work as a definition, it must be possible to come up with an account of
when someone will suffer a ‘loss’ as a result of another’s actions. After all, if tort law
is concerned to endow us with rights that others not cause us ‘loss’, we must be able
to determine when someone’s actions will result in us suffering a ‘loss’. I don’t want
to say that it is impossible for someone to come up with such an account, but I do
want to observe that practising tort lawyers have never needed or wanted to come up
with such an account. The reason is that practising tort lawyers never need or want to
ask – Did A suffer a loss as a result of B’s actions? Their only concern is to ask – Did
A suffer a loss as a result of B’s violating A’s rights?8 The baseline for tort lawyers’
inquiries into whether A has suffered a loss is set by the position that A would have
been in had B not violated his rights. So the scope of A’s rights against B determine
whether or not A can claim he has suffered a loss at B’s hands. From the point of
view of a practising tort lawyer, Gardner’s position seems to put the cart before the
horse in that it allows the issue of what losses A might suffer at B’s hands to
determine the scope of what rights A will have against B under the law of tort. This
does not provide a very promising basis for an argument that is supposed to tell us
something about the nature of tort law, as it is currently practised.

The only reason John Gardner insists on (D) is to find some way of distinguishing tort
law from the law on equitable wrongs. (Even then, it might be observed that there are
quite a few equitable wrongs which seem just as much concerned with protecting
people from suffering losses as Gardner claims tort law is.)9 If this is a concern
(which it isn’t to most people), there are other ways of distinguishing the two areas of
law. One could argue that tort law and equity both do (B), but from different starting
points. On this view:

7
Ibid, 12-13.
8
So, for example, Andrew Tettenborn’s essay ‘What is a loss?’ (in Neyers, Chamberlain and Pitel
(eds), Emerging Issues in Tort Law (Hart Publishing, 2007)) is solely concerned with the issue of (at
441) ‘just what do we mean when we say that a plaintiff whose rights have been infringed has
nevertheless suffered no loss? Or, if you prefer to put the question positively, what is the law referring
to when it talks about ‘loss’?’ (emphasis added).
9
For example, handing over trust assets to a third party without authorisation, dishonestly assisting
someone to commit a breach of trust, passing off, breach of confidence.
 5

(E) Tort law determines what basic rights we have against each other, and what
remedies will be available to us when those rights are violated, and it does so by
giving effect to, and elaborating on, rules and principles that were first
developed in the old common law courts.

A similar statement could be made about ‘The law on equitable wrongdoing’,

substituting the words ‘courts of Equity’ from ‘common law courts’. (E) works
perfectly fine to distinguish tort law from the law on equitable wrongs without
committing us to any unduly restrictive – and possibly impossible – views on the
nature of the rights that tort law gives us. As between (B) and (E), I think that (E) is
more accurate, but for working purposes in talking about tort law, (B) is perfectly
adequate.
In his piece on ‘Torts and other wrongs’, John Gardner assumes that the
American academics John Goldberg and Benjamin Zipursky endorse (B). They may
well do.10 However, it is worth observing that there is nothing in their ‘civil recourse’
theory of tort liability that requires that we endorse (B). To see why, it’s easier if we
look at tort law using definition (C), rather than (B). (While remembering that (B) and
(C) are formally equivalent.) (C) presents tort law as forward looking – as imposing
duties on us for other people’s benefit, and telling us what remedies will be available
if we breach those duties. So tort law – as defined in (C) – operates as a guide to
conduct, in much the same way that the Highway Code does.
In contrast, Goldberg and Zipursky’s ‘civil recourse’ theory of tort law is
backward looking. It sees tort law as performing the following functions: (i) looking
back at what a defendant has done; (ii) determining whether that conduct was
‘wrongful’ (in some sense still to be explored); and (iii) if it was, providing the victim
of that wrong with a peaceful means of recourse against the defendant wrongdoer.
Nowhere in this picture do we see tort law: (iv) laying down rules that are meant to be
followed in future.
Of course, one could argue that in doing (ii), tort law is implicitly doing (iv).
And there are some passages in Goldberg and Zipursky’s work that support this idea:
that if the law recognises that your conduct is ‘wrongful’ then it is implicitly telling
the rest of us not to do the same thing in the future. 11 But I don’t see that there is any
necessary connection between these two things, particularly when the rest of
Goldberg and Zipursky’s work makes it clear that (in their opinion) what the law is
doing when it is identifying your conduct as ‘wrongful’ is that it is basically saying
that it thinks12 you have committed a moral wrong that is so serious that in a pre-legal

10
See, for example, John Goldberg and Benjamin Zipursky, ‘Torts as wrongs’ (2010) 88 Texas Law
Review 917, 937: ‘Tortious wrongdoing always involves an interference with one of a set of individual
interests that are significant enough aspects of a person’s well-being to warrant the imposition of a duty
on others not to interfere with the interest in certain ways, notwithstanding the liberty restriction
inherent in such a duty imposition’ (emphasis added).
11
Ibid, at 949: ‘The statement that it is wrong to lie, spoken by a parent to a child or written by an
opinion columnist for newspaper readers, contains injunctive force: it condemns lying and conversely
urges refraining from lying. Making such a statement is, moreover, identifying a way of treating other
people as unacceptable. The same is true when a court holds liable a broker who has misrepresented a
company’s financial condition to an investor who relied on that misrepresentation to his detriment. The
court is articulating a norm of conduct that requires certain actors to refrain from deceiving other to
their detriment and condemns doing so as wrongful. The first would be said to be a duty-imposing rule
of morality, the second a duty-imposing rule of law.’
12
Of course, the law may be wrong to think this, so there is a conceptual space between the notion of a
legal wrong (what the law recognises as being wrongful) and a moral wrong (what is actually
wrongful): ibid, 951: ‘The fact that an act falls under an authoritative legal directive that characterizes
 6

state the victim of your wrong would have been morally entitled to seek some sort of
recourse against you.
In order to make some sort of connection between the law’s recognising that
doing x is morally wrongful and the law’s telling us not to do x, we could take the
tough line that whenever the law makes it clear that it thinks that doing x is morally
wrongful, it is telling us not to do x. But the tough line seems to have unacceptable
consequences. For example, one could read s 4 of the Law Reform (Miscellaneous
Provisions) Act 1970 (which provides that ‘no person shall be entitled
to...claim...damages from any other person on the ground of adultery with the wife of
the first-mentioned person’) as saying that making someone a cuckold is morally
wrongful, but not so seriously wrongful that it is prepared to offer the cuckold any
form of redress for the cuckolding. But if this is correct, and the tough line is also
correct, then s 4 of the 1970 Act would seem to be saying that you have a legal duty
not to sleep with another man’s wife – which can’t be right. So the tough line seems
not to work.
But the line that I think Goldberg and Zipursky would like to take – the G&Z
line – seems equally problematic. The G&Z line (so far as I can predict) would be that
we can only say (a) that the law is telling us not to do x, if (b) the law regards doing x
as so seriously morally wrongful that it ought to provide some form of recourse
against whoever does x. On this view we cannot say that the law tells us not to
commit adultery, because s 4 of the 1970 Act makes it clear that the law does not
regard adultery as so seriously wrongful as to warrant providing some means of
recourse to a man whose wife has sex with someone else. The argument may end up
in the right place – there is, of course, no legal duty not to commit adultery – but it
gets there by a less than convincing route. There seems no logical reason why we can
only say (a) when (b) is true.
It is for Goldberg and Zipursky to sort out these difficulties with their theory.
But reflecting on these difficulties indicates that it might be possible for someone – if
not Goldberg and Zipursky themselves – to argue that:

(F) Tort law determines when someone has been the victim of a moral wrong
that is so serious that the victim would, in a pre-legal state, have been morally
entitled to seek some kind of recourse against the perpetrator of that wrong, and
provides the victim of that wrong with a peaceful means of obtaining redress for
that wrong.

I have been arguing that (B) and (F) are different. For example, if you believe (B),
then what the House of Lords was doing in Donoghue v Stevenson13 was determining
whether consumers automatically have a legal right against manufacturers that those
manufacturers take care to ensure that their goods were not dangerous to use, with the
result that they don’t have to do anything special to obtain that right.14 If you believe
(F), then what the House of Lords was doing was considering whether in a pre-legal
state, carelessly causing someone to fall sick by putting into circulation a ginger beer
bottle with a dead snail in it would have counted as such a serious moral wrong as to

it as a legal wrong does not entail that such an act, in the circumstances it actually occurred, warrants
categorization as morally wrongful.’
13
[1932] AC 562.
14
For such a reading of Donoghue v Stevenson, see Andrew Tettenborn, ‘Professional negligence: free
riders and others’ in Economides et al (eds), Fundamental Values (Hart Publishing, 2000) – for me,
one of the top ten articles ever written on tort law in the UK.
 7

entitle the victim of that wrong to seek some sort of recourse against the person who
put the ginger beer bottle into circulation.
So should we accept (F)? Two reasons are commonly given for rejecting (F).

(1) It may be questioned whether there are any wrongs that are genuinely ‘redress-
entitling’ (my term, not Goldberg and Zipursky’s) in the sense that in a pre-legal state,
the victim of that wrong would have been morally entitled to take direct, violent
action against the perpetrator of that wrong to seek some sort of redress for what was
done to them.15 Of course, people feel urges to seek revenge on those they feel have
wronged them (a fact that I will return to later), but that does not mean that those
urges are legitimate or should be given some peaceful form of expression through the
law.

(2) (F) does not seem to fit with what we know about how tort law works, and how
tort cases are decided. Most importantly, there are plenty of occasions where someone
is held liable in tort for doing things that do not seem to amount to a redress-entitling
wrong: for example, where someone innocently converts someone else’s property, or
innocently trespasses on another’s land, or where someone causes a claimant harm by
inadvertently breaching a statutory duty.16 In many cases where a defendant is held
liable in negligence, we may also doubt whether the defendant’s conduct came
anywhere close to amounting to a redress-entitling wrong; for example, where the
defendant was trying to do his best to avoid harming the claimant, but could have
done better.

Goldberg and Zipursky have provided a good explanation why the second reason for
rejecting (F) does not stand up. They readily acknowledge that the law’s catalogue of
‘wrongs’ is very different from the list of redress-entitling wrongs that you and I
might come up with if we were asked to do so. They explain that it is for institutional
reasons that the law ends up stigmatising as ‘wrongful’, conduct that no one would
describe as redress-entitling. These institutional reasons for the law’s adopting
broader and less flexible accounts of when someone will have ‘wronged’ another than
we would expect it to if (F) were true are: (i) the need to make it relatively easy for
the courts to determine whether the defendant ‘wronged’ the plaintiff, and avoid
protracted and difficult hearings; and (ii) the need to send a strong message about
‘how society expects its members to behave’.17 However – so far as I know –
Goldberg and Zipursky have yet to explain why the first reason for rejecting (F) does
not stand up. For myself, I think it does, and accordingly reject (F).
So, of the six definitions of tort law that I have set out above, I believe we
should reject (A), (D) and (F). For working purposes, I think we can proceed on the
basis that (B) is true, while acknowledging that on occasion we may need to
supplement (B) by reference to (C) (its formal equivalent, except using the more
precise language of duties owed to another, rather than rights that one has against
another) and (E) (if we need to distinguish the law of tort from the law of equitable
wrongs).

15
See John Finnis, ‘Natural law: the classical tradition’ in Coleman and Shapiro (eds), The Oxford
Handbook of Jurisprudence and Philosophy of Law (OUP, 2002), 656.
16
See Jane Stapleton, ‘Evaluating Goldberg and Zipursky’s civil recourse theory’ (2006) 75 Fordham
Law Review 1529.
17
Goldberg and Zipursky, ‘Tort law and moral luck’ (2007) 92 Cornell Law Review 1123, 1158.
 8

Others may disagree. For the remainder of this paper, I will proceed on the basis
that (B) is correct, and formulate the next questions tort lawyers should be looking to
answer on that basis. But I hope I will say enough in my discussion of what those
questions should be to enable those who believe that (A) or (F) is correct to see what
form their own ‘next questions’ should take.

III. THE SECOND QUESTION

Academics who endorse (B) have tended to busy themselves with answering
analytical questions such as:

(1) The hard law question: What basic rights does tort law give us, and what remedies
will be available when those rights are violated? and

(2) The interpretive question: What ideas or principles best explain why tort law gives
us the basic rights that it does, and why it makes available the remedies it does when
those rights are violated?18

Though obviously these are questions that need to be answered, I think that there is a
danger that paying so much attention to these analytical questions is actually getting
in the way of our making any worthwhile progress in thinking about tort law. This is
for two reasons.

(1) It is very likely that there are no completely clear answers either to the hard law
question or the interpretive question.
Looking at the hard law question first, it is obviously clear that tort law gives us
certain basic rights – such as a right that others not touch us except in certain defined
circumstances. But the available case law makes it hard to tell whether we enjoy
certain other basic rights. For example, I would say that there are a huge number of
different duties of care that we might owe other people in different situations. David
Howarth argues that, in fact, we each owe each other a basic duty to take reasonable
care not to harm that other by acting unreasonably, with certain exceptions where we
are left free to harm others carelessly.19 I am not sure how the available case law
would allow us to say who is right. Similarly, I would say that someone driving down
the road owes nearby pedestrians a duty to take care not to drive dangerously. Robert
Stevens would say that he owes those nearby pedestrians a duty not to injure them by
carelessly driving badly.20 Again, it is very difficult to find relevant materials that
allow anyone to say with confidence which of us is correct.
When we turn to the interpretive question, the difficulties in finding clear
answers increase. There are lots of different aspects of the rights and remedies that
tort law gives us that are capable of being explained in a variety of plausible ways.
For example, consider the old actio personalis rule – now abolished in the UK in all

18
Though I separate out these two questions, I readily acknowledge that in a common law system
where there is an institutional commitment to ensuring that the law is principled and coherent, there
will be a lot of overlap between these two questions.
19
David Howarth, ‘Many duties of care – or a duty of care? Notes from the underground’ (2006) 26
Oxford Journal of Legal Studies 449.
20
Stevens, ‘Rights and other things’, above, n 4, 117-118.
 9

but defamation cases21 – that a victim of a tort’s right to sue other people in tort would
die with him. Someone who adheres to the interpretive view that damages in tort
claims are designed to restore to the victim of a tort the means that he has been
deprived of by that tort would explain this rule on the basis that if it is no longer
possible to do this (because the victim of the tort is dead), there is no point in allowing
a claim in tort to be made against the tortfeasor. Someone who endorses Goldberg and
Zipursky’s ‘civil recourse theory’ of tort law would give a quite different explanation
of the rule: that if the victim of a tort is no longer around and wanting to seek some
sort of recourse against the person who did him wrong, then there is really no point in
allowing a claim in tort to be made against that person. It is hard to see how we could
tell which of these views is correct. We could come up with some real or imaginary
cases where the result should be different according to which view is correct, and then
ask ourselves how the courts decided, or might have decided, those cases when the
actio personalis rule existed. But if the case is real, who’s to say that the courts might
not have decided the case another way if they had been made aware of the issue of
principle that rode on it? And if the case is imaginary, how can we guess how the
courts would have decided that case if we cannot tell what principles really underlie
the law?
The fact that there are no completely clear answers to the hard law or the
interpretive question means that these questions are particularly prone to give rise to
protracted debates which take up a lot of time and get nowhere. If we are going to
make some progress in thinking about tort law, we have to avoid these kinds of
fruitless debates, and seek to advance on other fronts.

(2) Homer’s Odyssey tells us the story of Odysseus and his men seeking to make their
way back home to Ithaca after the end of the Trojan War. In the course of their
travels, they came across the island of the Lotus Eaters. Anyone who ate the flowers
and fruits of the lotus plant would lose interest in doing anything else. Odysseus sent
some of his men to explore the island, but they never came back. He sent some more,
and they too never came back. Finally, he went to look for them, found them eating
lotus fruits and flowers, realised what had happened, and dragged them back onto the
ship, and set sail. A lucky escape.
What I want to suggest is that the hard law and interpretive questions are, for
legal academics, the equivalent of what the lotus fruits and flowers were for
Odysseus’ men. The game of trying to answer these kinds of questions (let’s call it
‘the AQ game’, for ‘analytical question game’) is addictive, and consequently makes
you lose interest in doing anything else. This is for three reasons.
First, the AQ game is fun to play. I have a lot of fun writing (with Roderick
Bagshaw) my textbook on tort law, taking a whole load of cases and seeing how to fit
them into some sort of intelligible pattern. And when you find a new case that fits
perfectly with your ‘take’ on the law, you get a little dopamine shot in the brain – a
feeling that you are ‘getting somewhere’, that you ‘really understand’ what’s going on
with this area of the law.
Secondly, the AQ game is not too hard to play. Computer games that are simply
too hard are not addictive because users quickly lose interest in them. But the AQ
game is a perfectly balanced game – not so easy that it’s uninteresting, but not so hard
that it’s off-putting.22 For example, to play the interpretive level of the AQ game, all
21
Law Reform (Miscellaneous Provisions) Act 1934, s 1.
22
Here’s Pierre Schlag, again (see above, n 2, at 809): ‘...legal thought is not rocket science. It’s just
not: Putting cases together (case-crunching), recognizing their patterns (theory), weaving them into this
 10

you need is some knowledge of the law, and some knowledge of philosophy (moral
and political) and you are away, constructing theories of why the law says what it
says, and testing them against the available materials – and always with the option of
chucking any cases that are just too intractable into the sad pile marked ‘wrongly
decided’.
Thirdly, there are a number of external goods available to those deemed to be
high achievers in the AQ game – research funding, prestigious academic posts,
citations in real life cases. The prospect of being able to obtain some or all of these
goods is an added incentive to keep on playing the AQ game.
Given all this, why would anyone want to do anything but play the AQ game?
And so it proves. The amount of literature and effort devoted to the AQ game is vastly
out of proportion to the actual tangible progress that we make in coming up with
answers to the hard law and interpretive questions, but still people spend all their time
playing.
The fact that people spend so much time playing the AQ game would not be so
bad if it deserved to be treated as if it were the only game in town. But it does not. For
example, suppose that it could be shown – as some interpretive theorists of tort law
contend that it can be – that tort law in its current state can be best explained as
requiring us not to violate other people’s independence as persons. 23 Well, so what?
Isn’t it more important to know what tort law should be doing, rather than what it is
doing at the moment? It’s important, of course, that we know what direction we are
travelling in at the moment: but it’s far more important that we know whether or not
we are travelling in the right direction.

For those of us who accept (B) as being correct, we cannot dare to make the next
question we address an analytical one, for fear that it will take us into woods from
which we will never be able, and never want, to emerge. So the second question I
would suggest tort academics (who endorse (B)) should focus on is not analytical at
all, but normative. The question is – What basic rights should tort law give us against
other people? This is a much harder question than either the hard law or the
interpretive question, and it is not surprising that academics should shy away from
addressing it explicitly. But I think if we are to make some real progress in thinking
about tort law, we have to confront it head on.
The available literature suggests four different answers to my second question. I
say ‘suggests’ because the available literature is not really concerned to make a
normative case for saying that we should have certain rights and should not have other
rights. Instead, it is concerned to make interpretive claims about what ideas and
principles best explain why we have the rights that we do have. But when an
interpretive theorist of the law argues that a given area of law can be best explained as
giving effect to certain ideas and principles, those ideas and principles almost always
seem to be – through happy chance – ones that the interpretive theorist himself
endorses.24 So I see nothing wrong with mining interpretive claims in the literature as

or that well-known argument... – all in all, it’s just not the sort of thing that requires or permits the
display of great intellectual prowess.’
23
See text at nn 28-31, below.
24
There are, of course, interpretive theorists who attempt to do a hatchet job on tort law, arguing that it
can be best explained as giving effect to ideas and principles that no sensible person would ever
endorse. Such accounts of tort law are pretty worthless, characterised as they are by a signal
unwillingness to consider any other explanations of tort law that might present tort law in a more
edifying light.
 11

to what ideas and principles best explain what tort law is up to, in order to come up
with some possible answers to the normative question of what ideas and principles
tort law should give effect to in endowing us with basic rights against other people.
So – here are the four different answers to my second question that are
suggested by the available literature. There is, first, the Rawlsian contractualist
answer, according to which:

(G) Tort law should endow each of us with those basic rights that everyone
would agree each of us should have, were we to make that agreement under fair
conditions.

As Peter Gerhart observes in his recent book Tort Law and Social Morality:‘By and
large, scholars have not taken advantage of the [Rawlsian] device of the veil of
ignorance when thinking about tort law.’25 However, his own book can be taken as
suggesting (G).
There is, secondly, the balanced answer:

(H) Tort law should endow A with a basic right against B so long as: (i) the
benefit of giving people like A such a right would outweigh the burden to
people like B of A’s having such a right against her; and (ii) the benefit of
giving people like A such a right is not offset by any net costs to the community
from people like A’s being given such a right.

I call this answer – which is the answer set out at greater length in my piece on
‘Rights and the basis of tort law’26 – the ‘balanced’ answer because it involves, first of
all, balancing the interests of people like A against the interests of people like B in
determining whether there is a case in principle for people like A to be given a right
against people like B; and then balancing the interests of people like A against the
interests of the community as a whole in determining whether all things considered
people like A should be given a right against people like B.
There is, thirdly, the social harmony answer:

(J) Tort law should endow A with a basic right against B that B do (or not do) x,
if B’s not doing (or doing) x would result in A developing understandable
feelings of anger towards B that might result in A seeking some sort of revenge,
or recourse, against B.

This is, of course, Goldberg and Zipursky’s ‘civil recourse’ theory of tort law, re-
presented – as Jane Stapleton called for it to be27 – as a normative theory of what tort
law should say.
There is, fourthly, the Kantian Right answer:

(K) Tort law should, in endowing us with basic rights against other people, give
effect to the right to independence that we each have against everyone else.

This is the normative version of the interpretive claim made by theorists like Ernest
Weinrib28 and Arthur Ripstein29 – among others – that this is what tort law already
25
Peter Gerhart, Tort Law and Social Morality (Cambridge University Press, 2010), 93.
26
McBride, above, n 4.
27
See Stapleton, above, n 16, 1562.
 12

does. Tort law (or so it is claimed) protects our right to independence by preventing
other people depriving us of the means that we are entitled to use to pursue the
purposes we have chosen to pursue. A succinct and very clear statement of this
interpretive view of tort law can be found in Ripstein’s very recent paper ‘Civil
recourse and the separation of wrongs and remedies’, from which the following
passages are taken:30

The means that you have are, from the point of view of the law of tort, just your
own person, that is, your bodily powers and mental capacities (as well as your
reputation), and whatever things outside of your own body which are yours, that
is, your property... To say that [these means] are yours, as against others, is...to
say that you, rather than anyone else, are the one who gets to decide how they
will be used. That is...why you can be wronged if somebody touches you without
your authorization: that person uses your person – your body – for a purpose that
you have not set.

[Tort law] restricts the use of means in two straightforward senses: first of all, the
law of tort prohibits trespasses of all forms. You are not entitled to use or even
touch another person’s body or property without that other’s permission, and you
are not allowed to do so, quite apart from whatever worthy or unworthy purpose
you might be pursuing. Second, you are not entitled to injure other people, either
in their person or their property, by using your own property in ways that
characteristically cause such injuries.

The first set of restrictions generates a wide variety of trespass-based torts; the
second set generates...“harm-based” torts. In a trespass-based wrong, liability is
premised on defendant’s use of plaintiff’s person or property. Defendant need
only intend to use something that is another’s, and a trespass can be committed
even if defendant did not, or even could not know, the title of the thing being
used.

Harm based torts concern the side effects of one person’s use of his or her own
means on the ability of others to continue using their own means. For example,
the tort of nuisance is organized around the thought that a landowner’s use and
enjoyment of his or her own land will inevitably have side effects on his or her
neighbours... Given that some side effects are inevitable, the question of whether
a particular side effect constitutes a nuisance turns into the question of whether it
is excessive in relation to the ability of a plurality of neighbors to each use and
enjoy his or her own land, consistent with the ability of other neighbors to do the
same.

In the tort of negligence, the same general form of reasoning prohibits people
from injuring each other’s means through dangerous use of their own means. As
people use their means, whether their own bodies or chattels, some side effects
on others are inevitable... The law of negligence... [determines] whether
defendant’s conduct was too dangerous, that is, whether it is was of a type that, in
the circumstances, was too likely to interfere with plaintiff’s security of his or her
means.

28
Weinrib, The Idea of Private Law (Harvard University Press, 1995).
29
Ripstein, ‘Beyond the harm principle’ (2006) 34 Philosophy and Public Affairs 215; ‘Tort law in a
liberal state’ (2007) 1 Journal of Tort Law 3; ‘Civil recourse and the separation of wrongs and
remedies’ (2011, available on SSRN) (henceforth, ‘Civil recourse’).
30
Ripstein, ‘Civil recourse’, 9-12.
 13

This austere structure of protecting each person’s means against use by others, or
damage through the excessive side effects of other people’s use of their means,
generates the familiar structure of the law of tort. Most importantly, it generates
the fundamental distinction between nonfeasance and misfeasance, that is,
between wronging someone and failing to confer a benefit on that person. To
interfere with what another person already has is a wrong, but to fail to provide
aid to that person, no matter how badly that person needs it, is not.

Neither Weinrib nor Ripstein would claim to have invented this explanation of what
tort law does. That honour goes to Immanuel Kant, the first part of whose
Metaphysics of Morals (‘The Doctrine of Right’), lay the groundwork for thinking
about tort law in this way. Kant’s assertion that ‘Freedom (independence from being
constrained by another’s choice) insofar as it can coexist with the freedom of every
other in accordance with a universal law, is the only right belonging to every man by
virtue of his humanity’31 is key to this view of tort law.
So – we have four different answers to our second question. How do we
determine which one is correct? I suggest a good starting point would be to establish
whether or not we do have – as Kant contends that we do – a right to independence.
Let’s call this the Independence Thesis. If the Independence Thesis is correct, then
that automatically establishes that (K) is correct, and (G), (H) and (J) are incorrect.
This is because if we do have a right to independence, then the only thing that tort law
can justifiably do is to give effect to this right to independence. If it does anything
else, then it will violate our right to independence.
For example, suppose that it were the law that if A were in peril of his life, and
B could easily save his life, then A would have a right that B save his life, and let’s
say that the law took this stance because A would benefit a great deal from his life
being saved, and subjecting B to a duty to save A’s life would not cost B much, and
imposing such a duty on people like B would not have any unacceptable side effects
for society as a whole. If the Independence Thesis is correct then the law would be
unjustified in granting A a right that B save his life. Doing so would do nothing to
preserve A’s independence as a person (as opposed to his welfare), and deprive B of
her independence to choose what to do with the means at her disposal.32

31
Immanuel Kant, Metaphysics of Morals, Introduction, 6:237 (trans. Mary Gregor). (‘6:237’ means
page 237 of the sixth volume of the Prussian Academy edition of Kant’s works.)
32
The fact that, if the Independence Thesis is correct, the only thing tort law can justifiably do is
protect our independence as persons helps to explain certain otherwise strange and (in some cases)
infuriating features of the writings of Kantian tort theorists who make the interpretive claim that tort
law does work to protect our independence as persons. Namely:
(1) The limited range of rights that Kantian tort theorists are willing to consider a claimant may
have had, in considering whether a defendant could be said to have violated that claimant’s rights. If
the Independence Thesis is correct, only independence-protecting rights can count as they are the only
rights that tort law can justifiably recognise.
(2) The way Kantian tort theorists identify corrective justice (which everyone else normally
associates with a principle that a defendant who has wronged a claimant should repair the harm that the
claimant has suffered as a result of that wrong being committed) with repairing the consequences of a
defendant’s violating a claimant’s right to independence, so that any duties to repair that do not arise
out of a violation of the claimant’s independence do not count as examples of ‘corrective justice’ at
work. If the Independence Thesis is correct, then a defendant can only be said by tort law to have
wronged a claimant and incur a consequential duty to repair if he has violated the claimant’s right to
independence.
(3) The tendency Kantian tort theorists have to assert that tort law has no choice but to live up to
the claims that they make for it. If the Independence Thesis is correct, then – as I have observed – then
 14

So if we want to determine which of (G), (H), (J) or (K) is correct, we must find
out whether the Independence Thesis is correct. Unfortunately, Kant did not attempt
to provide any basis for the Independence Thesis, saying instead that his ‘Universal
Principle of Right’ (which elaborates on the Independence Thesis by asserting that
‘Any action is right if it can coexist with everyone’s freedom in accordance with a
universal law, or if on its maxim the freedom of choice of each can coexist with
everyone’s freedom in accordance with a universal law’) was ‘a postulate that is
incapable of further proof’.33
Despite this, I think there are four arguments in favour of the Independence
Thesis that we could consider, to see whether any of them provide a sufficient basis
for endorsing it.

(1) The logic of ownership. The first argument is that the Independence Thesis simply
follows from the fact that certain means (such as my body, and my property) belong
to me, and not to you.
If my body belongs to me, then what that means is that I am the only one who
gets to decide what use to make of my body. If anyone else could decide what use to
make of my body, then it would no longer truly belong to me. Of course, if I decide to
make use of my body to grab you and wrestle you to the ground, I cannot justify that
on the basis that my body belongs to me and as such I am free to do what I like with
it. If I assert that my body belongs to me, then I must also acknowledge that your
body belongs to you, and that grabbing you and wrestling you to the ground is
wrongful.
So the Independence Thesis – and workings out of that thesis in things like the
Universal Principle of Right – simply gives effect to the idea that different means
belong to different people, and ensures that all means-owners enjoy an equal measure
of freedom to use the means at their disposal to pursue purposes that they have chosen
to pursue.
The difficulty with this argument is that it justifies the Independence Thesis at
the cost of opening up debates over what is implied by saying that a given set of
means belongs to someone. We could argue that the fact that A owns a given set of
means implies that he must be allowed as much freedom as possible to determine how
to use those means, subject only to everyone else being allowed as much freedom as
possible to determine how to use the means that belong to them. As Arthur Ripstein
observes, ‘the sense in which your means are your own is that you alone, as against
others, are entitled to determine the purposes for which they will be used.’ 34 However,
to say that a given set of means belongs to you does not necessarily imply that you
must be allowed a maximal amount of freedom to decide how to dispose of those
means. Other models of what owning something means are possible.

the only thing tort law can justifiably do is protect people’s independence as persons. Point (3) reminds
me of a funny story told by Derek Parfit in the Preface to his On What Matters (Oxford University
Press, 2011) about asking a Kantian philosopher about Kant’s Categorical Imperative: ‘I asked a
Kantian, “...if I don’t give myself Kant’s Imperative as a law, I am not subject to it?” “No,” I was told,
“you have to give yourself a law, and there’s only one law.” This reply was maddening, like the
propaganda of the so-called “People’s Democracies” of the Soviet bloc, in which voting was
compulsory, and there was only one candidate. And when I said, “But I haven’t given myself Kant’s
Imperative as a law”, I was told “Yes you have.” This reply was even worse.’ (Volume 1, xlii-xliii.)
33
Kant, above, n 31, 6:231.
34
See Ripstein, ‘Civil recourse’, 10.
 15

For example, under the social view of ownership,35 being vested with ownership
of some thing does not just confer rights on the owner to decide how to use that
property, but also responsibilities to act wisely in exercising those rights. On this
view, the trust provides a more illuminating picture of what it is means to be an owner
of property than that provided by the insitution of legal beneficial ownership of an
item of property. If this view of what it means to own something is correct, then there
is no reason why an owner of a given set of means must be allowed as much freedom
as possible to determine how those means are used.
Alternatively, under the exclusion view of ownership,36 being vested with
ownership of some thing merely means that you are given the power to veto over
what other people may positively do with that thing. On this view, any limits that the
law places on what you may positively do with that thing are not incompatible with
your being an owner of that thing. So, for example, I would still count as being the
owner of Titian’s Diana and Actaeon even if the UK government prevented me from
selling that painting to a buyer located overseas.
So simply to assert that a given set of means belongs to a particular individual
does not, of and in itself, imply that that individual must be allowed a maximal degree
of freedom to decide how to dispose of those means. To establish this, it would have
to be shown that there is something incoherent or irrational or just plain wrong with
the social and exclusion views of what it means to be the owner of something. And I
am not sure there is.

(2) Ends, not means. The second argument is that the Independence Thesis gives
effect to the moral requirement that we should treat other people as ends, not means.
The idea is that if A is drowning in a lake, and we impose a legal duty on a passerby
B to save A’s life, we are merely using B as a means of securing A’s welfare. The
only set of legal rules that will treat A and B and all of us as ends in themselves, is
one that will guarantee to each of us the maximum possible measure of independence
from each other.
The trouble with this argument is that it is vulnerable to the arguments laid out
in Chapter 9 of Derek Parfit’s On What Matters, which explores what it means to treat
someone merely as a means. Parfit argues that:

we do not treat someone merely as a means, nor are we even close to doing that,
if either

(1) our treatment of this person is governed or guided in sufficiently

important ways by some relevant moral belief or concern,

or

(2) we do or would relevantly choose to bear some great burden for this
person’s sake.

For some moral belief to be relevant in the sense intended in (1), this belief must
require direct concern for the well-being or moral claims of the person whom we
are treating in some way. Suppose that some...slave-owner never whips his slaves
because he believes that such acts would be wrong. But what would make such

35
See, most recently, the symposium on the ‘The social function of property: a comparative
perspective’ in (2011) 80 Fordham Law Review 1000ff.
36
See James Penner, The Idea of Property in Law (OUP, 1997), chapter 4.
 16

acts wrong, he believes, is not the fact that he would be inflicting pain on his
slaves, but the fact that he would be giving himself sadistic pleasure. If that is
why this man never whips his slaves, this fact would not count against the charge
that he treats his slaves merely as a means. Another example is Kant’s view that
cruelty to animals is wrong because it dulls our sympathy, making us more likely
to be cruel to other people. If it is only this moral belief that leads some scientist
to avoid causing her laboratory animals any pain, she would be treating these
animals merely as a means.37

By that yardstick, a legal rule which said, ‘If people’s lives are in danger, save those
you can, if you can do so without endangering your own life’ would not involve
treating anyone merely as a means to an end. The fact that the rule is crafted only to
apply to those who can easily save others shows that we have been influenced, in
creating the rule, by a moral concern for the well-being of those who might be
subjected to this rule.
And even a rule which said, ‘If people’s lives are in danger, save those you can,
even if doing so involves great danger for yourself’ would not involve treating anyone
merely as a means to an end. While such a rule may place a great burden on an
emergency worker who is in a position to save someone’s life but only at considerable
danger to her own, had it been the emergency worker who had been in danger, we
would have been willing to risk our lives to save her (as the rule we have created
shows).

(3) Universal laws. Arthur Ripstein has attempted to explain the ‘Universal Principle
of Right’ as ‘the unique moral principle for rational beings who occupy space’:

if you were prohibited from using your body in any way, or, what comes to the
same thing, you were conditionally prohibited, so that your entitlement to do
anything with your own body was subject to the choice of others, as a material
principle would demand (perhaps everyone, or even someone, had to approve
any action you chose to perform), your capacity to set and pursue your own
purposes would be subject to their choice. No material principle of that sort could
be a universal law...because as a rational being you could not will a universal law
under which you could never set a purpose for yourself, or one under which you
could only do so with leave of another...
If moral persons are individuated spatially, then the only way to have
freedom under universal law is for each embodied rational being to have, in
virtue of its humanity, a right to its own person – that is, to its own body. Such a
right must be innate, because nothing could count as an affirmative act
establishing it – the right applies to any rational being that occupies space,
because its right is nothing more than the right that it has to the space that it
happens to occupy.38

This seems far too strong. While we can concede that no rational person could ‘will a
universal law under which you could never set a purpose for yourself’, there surely
must be laws or maxims that a rational person could will that exist somewhere
between ‘Everything I do will be subject to another’s say-so’ and ‘Nothing I do (that
does not violate another person’s independence as a person) will be subject to
another’s say-so’. For example, there seems no reason why a rational person could not
will to be a universal law that ‘A referee must take care not to write a reference that
37
Parfit, above, n 32, 214-215.
38
Ripstein, Force and Freedom (Harvard University Press, 2009), 371-372.
 17

will not unjustifiably damage the subject of the reference’s prospects of getting the
job he is applying for’39 – which is one of the rules of English tort law to which
Kantian theorists of tort law should object, as carelessly writing someone a damaging
reference does not necessarily violate their independence as a person.

(4) The importance of independence. A final argument in favour of the Independence

Thesis – though not one, I think, that would appeal to any Kantians, who prefer their
rights to have a metaphysical, rather than political, origin – goes as follows.
Each person’s being free to pursue the purposes he has chosen to pursue with
the means at his disposal is more important than anything else except everyone else’s
being free to pursue the purposes they have chosen to pursue with the means at their
disposal. So reason dictates that no one’s independence to pursue their purposes can
be violated except in the cause of protecting other people’s independence to pursue
their purposes from being violated.
But this explanation does not seem to stand up. It is of course important that
people be free to pursue the purposes they have chosen to pursue, with the means that
are available to them. But other things are important as well. Having a job is
important. Peace of mind is important. Friendships are important. Happiness is
important. It is not clear that someone’s being independent as a person is more
important than any of these things and cannot be justifiably – rationally – sacrificed in
order to protect these things.

If this final defence of the Independence Thesis – and, with it, view (K) of what basic
rights tort law should give us – falls foul of what we can call the Many Things Matter
Thesis, then it seems so does (J), which tells us that tort law should endow us with
rights that other people not treat us in ways that will cause us, or have a tendency to
cause us, to develop understandable feelings of anger towards those other people.
Social harmony is important, but if many things matter, then it is not clear why we
should not also be endowed with rights against other people where other values are at
stake.
If (J) is to be saved as a plausible view of what basic rights tort law should give
us against other people, we have to find some way of arguing that – at least so far as
tort law is concerned – the Many Things Matter Thesis is wrong, and that in fact, only
one thing matters: either social harmony, or some value intimately linked with social
harmony. There seem to be two ways of doing this.

(1) Public reason. The first way is to argue that while many things matter, the only
thing that we can agree matters is social harmony. Given this, and given the need for
tort law to be acknowledged as legitimate by the overwhelming majority of the
population if it is going to work effectively, the only thing tort can law can do is foster
and maintain social harmony. If it goes beyond that role, and starts to recognise that
other things matter, it will be condemned as illegitimate by those who disagree that
those other things matter.
The problem with this view is as follows. Either we say that the only thing tort
law can do is give effect to goals and values that we all agree on. In which case,
fostering social harmony is out, as there obviously some people who do not value
social harmony at all. Or we say that the only thing tort law can do is give effect to
goals and values that all reasonable people agree on. But if it is true that many things

39
Spring v Guardian Assurance Co Ltd [1995] 2 AC 296.
 18

matter, we have no reason to think that the only thing all reasonable people will agree
on is that social harmony matters. Why would they realise that social harmony
mattered, but not any of the other things that matter?

(2) Love. The second way of defending (J) is to reject the Many Things Matter Thesis
in favour of the view that only one thing matters: that we love each other. This view,
of course, finds its greatest expression in the 13th chapter of St Paul’s first letter to the
Corinthians. As the examples given by St Paul will be very remote to a 21st century
academic audience, I have taken the liberty of rewriting his words to make it
dramatically obvious what a radical challenge St Paul poses to the Many Things
Matter Thesis:

If I write books that sell in their millions and are universally acclaimed, but I
have no love, I might as well have written nothing at all. If I understand
everything, and can do anything, but have no love, I am worth nothing at all. I
may be awarded Nobel Prizes, be on first name terms with heads of state and
prime ministers, have Professorships in universities on every continent – but if I
have no love, it means nothing at all.

What does this imply for tort law? I think it suggests that in an ideal world, where
everyone loved everyone else, there would be no need for tort law to exist at all. If
someone did something wrong to me, I would instantly forgive them for whatever
they had done – in the same way that I find it very easy to forgive those closest to me
whenever they screw up and don’t treat me as they ought to. There would be no
question of my taking them to court and standing on my rights and demanding some
sort of remedy for what they had done.40
But we don’t live in an ideal world. We live in a world where our starting point
is to feel indifferent to most other people. And our relationship with another person
can often find itself moving in the wrong direction – from indifference to hate, rather
than indifference to love – if we feel mistreated by that person. In the kind of non-
ideal world we live in, tort law has a valuable role to play in preventing indifferent
relationships being poisoned and becoming hateful, by: (i) requiring people not to
treat others in ways that will provoke those others to develop justifiable feelings of
anger and recrimination towards the person that has mistreated them, and (ii) where
such mistreatment has occurred, requiring the person who is guilty of that
mistreatment to make up for what he has done, so as staunch the wound created by his
actions. And as the degree of love each of us has for other people is only thing that
matters, doing (i) and (ii) should be the only thing that matters to tort law, in
endowing us with basic rights against other people, and determining what remedies
will be available for violations of those rights.
On this view of tort law, the basic rights I have against other people would not
reflect my interest in being independent of other people – as the Kantian theorists
would have it – but would instead reflect the stake I have in my not being independent
of other people. In other words, my basic rights would protect my interest in having
relationships with other people that are love-filled and have not becomed poisoned
through their mistreating me. And what basic rights tort law endowed me with against

40
Compare St Paul, in the same letter to the Corinthians in which his famous words on love appear:
‘The very fact that you have lawsuits among you means you have been completely defeated already.
Why not rather be wronged? Why not rather be cheated?’ (I Corinthians 6:7.)
 19

you would be strongly influenced by: (i) who you are; (ii) the nature of our
relationship; and (iii) the type of society we live in.41
(i) Who you are would matter because the degree of anger we feel towards
someone for doing or not doing something may depend a lot on who that someone is.
We would feel a lot more resentful towards a police officer for failing to save us from
being beaten up than we would towards a passing stranger.
(ii) The nature of our relationship would matter because the stronger – more
loving – our relationship, the more we could be expected to put up with from the other
person without developing feelings of anger and hatred towards them. So friends and
family members would have fewer rights against each other than they would against
people who were strangers to them.
(iii) The type of society we live in would matter because what sort of society we
live in would strongly influence the anger-making significance of certain actions. For
example, in Shakespeare’s Verona, ‘biting your thumb’ at someone would be liable to
provoke a murderous fight.42 But not in the societies we live in. In the same way, we
could expect people to react differently to having their character defamed, depending
on what kind of society they lived in.43
The love-based argument being made here in favour of (J) may seem very
attractive. It provides us with a model of what tort law should do that is much more
humane, relational, and culturally specific than the ‘austere’ (Ripstein’s words)44
Kantian model we were previously considering. However, the argument that only love
matters suffers from a logical flaw. Loving someone involves wanting that person to
have whatever is good for them to have, and being willing to sacrifice your own
interests to see that that happens. So my loving someone else presupposes: (i) that
there are things that it would be good for the loved one to have; and (ii) that there are
things that it would be good for me to have, but which I am willing to sacrifice in
order to see that the loved one gets the things that it is good for her to have. I can’t
then be said to love someone else unless I acknowledge that there are things that it is
good for the loved one to have, and there are things that it is good for me to have. But
if there are such things, then those things must matter. So while love may be the most

41
Note that none of these factors would be relevant under a Kantian vision of tort law where tort law
protects the right to independence that each of us has against everyone else. (Though each society
would have some ‘margin of appreciation’ for how it gave effect to that right to independence in
relation to matters such as rights arising out of the ownership of property.)
42
Romeo and Juliet, I.1:
Abraham. Do you bite your thumb at us, sir?
Sampson (aside to Gregory). Is the law of our side, if I say ay?
Gregory (aside to Sampson). No.
Sampson. No, sir. I do not bite my thumb at you, sir; but I bite my thumb, sir.
43
Contrast Iago’s incendiary speech in Othello, III.3: ‘Good name in man and woman, dear my Lord /
Is the immediate jewel of their souls: / Who steals my purse steals trash; ’tis something, nothing; /
’Twas mine, ’tis his, and has been slave to thousands; / But he that filches from me my good name /
Robs me of that which not enriches him / And makes me poor indeed’ with Falstaff’s dismissive shrug
in King Henry IV, Part One, V.1: ‘Can honour set to a leg? No. Or an arm? No. Or take away the grief
of a wound? No. Honour hath no skill in surgery, then? No. What is honour? A word. What is in that
word? Honour. What is that honour? Air. A trim reckoning! Who hath it? He that died o’ Wednesday.
Doth he feel it? No. Doth he hear it? No. ’Tis insensible, then? Yea, to the dead. But will it not live
with the living? No. Why? Detraction will not suffer it. Therefore I’ll none of it. Honour is a mere
scutcheon.’
44
Ripstein, ‘Civil recourse’, 12.
 20

important thing, and something without which nothing else matters, love can’t be the
only thing that matters because love itself acknowledges that other things matter. 45

If the Many Things Matter Thesis causes us, in the end, to reject (J) and (K), that only
leaves (G) and (H) standing as plausible accounts of what basic rights tort law should
give us against other people. Let’s remind ourselves of these accounts again:

(G) Tort law should endow each of us with those basic rights that everyone
would agree each of us should have, were we to make that agreement under fair
conditions.

(H) Tort law should endow A with a basic right against B so long as: (i) the
benefit of giving people like A such a right would outweigh the burden to
people like B of A’s having such a right against her; and (ii) the benefit of
giving people like A such a right is not offset by any net costs to the community
from people like A’s being given such a right.

It may be that we have no need to decide which of these answers to our second
question is correct. The arguments in Derek Parfit’s book On What Matters incline me
to believe that whichever of (G) and (H) we adopt, we will end up coming to the same
conclusions as to what basic rights tort law should give us against other people. 46 That
is, a basic right that satisfies conditions (i) and (ii) in formula (H) will also be one that
everyone would agree each of us should have, were we to make that agreement under
fair conditions. And the same applies vice versa.
My own view is that (G) and (H) are the best answers we have at the moment to
our second question. It may be that this is wrong. But I think it will only be shown to
be wrong if someone comes up with a fifth answer to our second question that is
superior to either (G) or (H). I am fairly confident that neither (J) nor (K) provide
better answers to our second question than (G) or (H) do. But that is not to say that (J)
and (K) are valueless as responses to our second question. On the contrary: reflecting
on the values that could justify (J) and (K) reminds us of facts about what really
matters that we are in danger of overlooking in the legal academy today.

45
It may be that John Gardner was alluding to this point in his cliffhanger ending to ‘Obligations and
outcomes in the law of torts’ where he asserts that Kant’s argument that ‘the only source of
unconditional (a.k.a. moral) value in our actions is the good will’ ‘collapses...spectacularly’ but says
that showing this is ‘a task for another paper’: see Peter Cane and John Gardner (eds), Relating to
Responsibility (Hart Publishing, 2001), 144. In a follow-up piece (‘The wrongdoing that gets results’
(2004) 18 Philosophical Perspectives 53), Gardner argues that Kant’s argument fails because (at 68)
‘Morally virtuous people...do not regard the value in their actions as stemming from their virtues [i.e.
their good will], but...from their actions.’ This may be true but doesn’t quite trigger the promised
‘spectacular’ collapse in Kant’s position.
46
Parfit’s book is concerned to come up with a formula that will allow us to determine when it would
be morally wrong to act in a particular way. He ends up considering three formulae: (1) An act is
wrong if and only if such acts are disallowed by some principle that is one of the principles whose
being universal laws would make things go best (rule consequentialism). (2) An act is wrong if and
only if such acts are disallowed by a principle that is one of the only principles everyone could
rationally will (Kantian contractualism). (3) An act is wrong if and only if such acts are disallowed by
a principle that no one could reasonably reject (Scanlonian contractualism). Parfit ends up concluding
that these three different formulae end up generating the same set of principles and that those who think
that ‘there are...deep disagreements between Kantians, Contractualists and Consequentialists’ are
wrong: ‘These people are climbing the same mountain on different sides’: Parfit, above, n 32, 419.
 21

(K) reminds us that human independence is important – that people must be

allowed the room to feel that the lives they are living are ones that they have chosen
to lead. Legal academics often forget this, in their eagerness to propose rafts of new
duties to deal with real or perceived social problems. But if all these proposals are
given effect to, pretty soon you have created a drone society, where no one is allowed
any room to make up their own mind what’s the best thing to do.47
(J) reminds us that causing another to hate us is perhaps the very worst thing we
can do to someone. As Martin Luther King Jr preached, ‘Like an unchecked cancer,
hate corrodes the personality and eats away its vital unity. Hate destroys a man’s
sense of values and his objectivity. It causes him to describe the beautiful as ugly and
the ugly as beautiful, and to confuse the true with the false and the false with the
true.’48
In reminding us of these facts, (J) and (K) play a vital role in enabling us to
apply the formulae set out in (G) and (H) in a way that is truly attentive to what really
matters.

IV. THE THIRD QUESTION

The third question is also a normative question, and follows on from the second
question. It is: What remedies should be made available to someone whose basic
rights have been violated by someone else? (To save words, I will call someone
whose basic rights have been violated, ‘the victim of a tort’, and someone who
violates someone’s basic rights ‘a tortfeasor’.) There are too many possible answers to
this question for me to set down and consider here. However, as an aid to clear
thinking in answering this third question, I want to consider two answers to this third
question which currently enjoy great popularity, and explain why I think the first
answer is deeply problematic, and the second answer completely wrong. The two
answers I want to consider are as follows:

(L) The victim of a tort should be allowed to sue the tortfeasor for
compensatory damages designed to put him in the position he would have been
in had that tort not been committed, as a ‘next best’ way of giving effect to the
basic right that the tortfeasor violated in committing his tort.

(M) The victim of a tort should be allowed to sue the tortfeasor for any profits
made by the tortfeasor from committing his tort because no one should be
allowed to profit from his wrong.

Let’s take (L) first. Next best theories of why the victim of a tort should be allowed to
sue the tortfeasor for compensatory damages are very popular at the moment among

47
An abiding theme of Philip K Howard’s writings on law is how having too many laws undermines
people’s capacities to think for themselves, and take responsibility for their actions. See, most recently,
Life Without Lawyers (WW Norton & Co, 2009). (On the same day I wrote this footnote, there was a
story on a UK newspaper website about how a fire brigade refused to rescue a drowning man from a
lake on the ground that health and safety regulations did not allow it. The lake was three feet deep, and
regulations do not permit the fire brigade to rescue people from waters more than ‘half a boot’ deep.)
48
Martin Luther King Jr, Strength to Love (Fortress Press, 2010), 48.
 22

theorists who endorse (B), as the following passages from, respectively, Ernest
Weinrib, John Gardner, Robert Stevens, and Arthur Ripstein, show:

When [a] defendant...breaches a duty correlative to the [claimant’s] right, the

[claimant] is entitled to reparation. The remedy reflects the fact that even after
the commission of the tort the defendant remains subject to the duty with respect
to the [claimant’s] right. The defendant’s breach of...duty...does not, of course,
bring the duty to an end... With the materialization of wrongful injury, the only
way the defendant can discharge his or her obligation respecting the [claimant]’s
right is to undo the effects of the breach of duty... Thus tort law places the
defendant under the obligation to restore the [claimant], so far as possible, to the
position the [claimant] would have been in had the wrong not been committed.49

When I fail to perform a duty that I owe to someone, there is something that I
still owe that person afterwards. Strictly speaking, I still owe him performance of
the duty, which continues to bind me. But if it is too late to perform – the dirty
deed is done – I now owe him the next best thing. I owe it to him to put him
back, so far as it can now be done, into the position he would have been in if I
had done my duty in the first place.50

Where [a] wrong has been committed, the secondary obligation to pay money
imposed upon the wrongdoer can be seen as the law’s attempt to reach the ‘next
best’ position to the wrong not having been committed by him in the first place.
Where the defendant is required to make good the claimant’s...loss...this is the
law’s attempting to reach this nearest approximation of the wrong not having
occurred.51

[I]f someone interferes with a right that you have, the right does not thereby
cease to exist. Since your right is to the exclusive use and security of your means,
you remain entitled to the security of your means, even if another interferes with
those means and so with your right. A remedy serves to ‘preserve what is mine
undiminished’... If I negligently destroy your coat, the coat no longer exists. But
your right to your coat – that is, the constraint on my conduct generated by the
fact that you, rather than I, are the one who is entitled to determine how it will be
used – is not changed by the fact that I acted contrary to that right. Destroying the
coat does not extinguish the right understood as a constraint on my conduct. If
the coat no longer exists, that constraint may call for different specific actions.
My duty to repair, to restore to you the means to which you have a right against
me, takes the form of a duty to replace the means of which I have deprived you.52

The difficulty with all such next best theories of the foundation of a victim of a tort’s
right to sue the tortfeasor for compensatory damages is that they conceal a problem
that needs to be addressed.
It is tempting to think that because the law is justified in imposing a duty on
Driver to take care not to crash into Pedestrian (= giving Pedestrian a right that
Driver not carelessly crash into her), the law is also entitled – in the case where
Driver violates that right/duty by carelessly crashing into Pedestrian – to require
Driver to repair his violation of that right/duty by putting Pedestrian back (so far as

49
Weinrib, The Idea of Private Law, 135.
50
Gardner, ‘Punishment and compensation: a comment’ (available on SSRN), text at n 20.
51
Stevens, Torts and Rights, above, n 4, 59.
52
Ripstein, ‘Civil recourse’, 17-18. See also Ripstein, ‘As if it had never happened’ (2007) 48 William
and Mary Law Review 1957.
 23

money can do it) into the position Pedestrian would have been had that duty/right not
been violated. But, like most temptations, this is one that needs to be resisted. The
reason for this is that repairing Driver’s violation of the duty he owed Pedestrian (or
the right that Pedestrian had against Driver) is almost always going to be a lot more
burdensome than observing the primary duty/right that was violated in this case. So it
would be wrong to think that just because the primary duty that Driver owed
Pedestrian (or the right Pedestrian had against Driver) is justified, then requiring
Driver to repair his violation of that duty/right is also justified. What we need is an
argument that it would not be unfair on Driver to require him to repair that violation –
and no one has yet been able (successfully) to make such an argument out.53 So (L)
might be correct, but only if it can be supplemented with an argument as to why it is
not unfair to make a tortfeasor repair the violation of the right/duty that he has
violated in committing his tort, given how burdensome such a duty of repair can be
for a tortfeasor.
Let’s now turn to (M), which says that the victim of a tort should be entitled to
sue the tortfeasor for any profits that the tortfeasor has made from committing his tort
because no one should be allowed to profit from his wrong. This is just a
straightforwardly bad argument that should be consigned to the dustbin, and never be
heard again. The reason why is that it is tautologous. The tautology is concealed,
though – which may be the reason why many academics have fallen for (M). To see
that (M) is tautologous, we just have to change the words used in (M) slightly
(without in any way altering their meaning). So (M) says:

The victim of a tort should be entitled to sue the tortfeasor for any profits that
the tortfeasor has made from committing his tort because no one should be
allowed to profit from their wrong.

The following is identical to (M) in all material respects:

(M1) A tortfeasor should be made to give up to the victim of his tort any profits
that he has made from committing his tort because no one who has committed a
wrong should be allowed to profit from that wrong.

The following is identical to (M1) in all material respects (and is different only in the
immaterial54 respect that we have dropped the reference to it being the victim of the
tort who gets the profits made by the tortfeasor):

(M2) A tortfeasor should not be allowed to retain any profits that he has made
from committing his tort because no one who has committed a wrong should be
allowed to profit from that wrong.

The following is identical to (M2) in all material respects:

53
There is, of course, Tony Honoré’s theory of ‘outcome responsibility’ (see, originally, his article
‘Responsibility and luck: the moral basis of strict liability’ (1988) 104 Law Quarterly Review 530), but
it very doubtful whether ‘outcome responsibility’ is, in fact, as fair as Honoré claims that it is.
54
Immaterial because the ‘principle’ that ‘no one should be allowed to profit from their wrong’ itself
admits that it is immaterial who gets the profits made by a wrongdoer – it is just important that they do
not remain in the hands of the wrongdoer.
 24

(M3) A wrongdoer should not be allowed to profit from his wrong because no
one who has committed a wrong should be allowed to profit from that wrong.

The following is identical to (M3) in all material respects:

(M4) A wrongdoer should not be allowed to profit his wrong because a

wrongdoer should not be allowed to profit from his wrong.

(M4) is obviously tautologous, but – as the transition from (M) to (M4) shows – is no
different from (M) in any material respect. If (M4) is tautologous (which it is
obviously is), then so is (M). (M) is a bad argument, and must be dropped. Those who
would argue that the victim of a tort should be entitled to sue the tortfeasor for any
profits made by him as a result of committing that tort need to explain exactly why it
is that ‘no one should be allowed to profit from their wrong.’

V. THE FOURTH QUESTION

Our second and third questions exhaust all the normative questions that we might
want to ask about tort law. Having worked our way through them, we can return –
perhaps with some relief – to the analytical territory that represents our comfort zone.
But this time we need have no worry that dwelling on analytical questions will
prevent us ever getting onto other questions that need addressing: having already dealt
with those other questions, we can take as much time as we like over analytical issues.
And the fourth question I would like us to address will require a lot of analytical work
to be answered properly. The fourth question is this: How far does tort law in its
current state fall short of our ideals as to what tort law should say?
I have put this question in a negative way, focussing our attention on how far
tort law falls short of our ideals for it, instead of asking how far tort law lives up to
those ideals. This is deliberate. I am desperate that we – as a community of tort
lawyers – should not fall into the trap of writing what Pierre Schlag calls ‘the zeitgeist
is going my way’55 articles, where we argue that tort law, properly understood, in fact
largely does exactly what we think it should be doing. So we take a bunch of cases,
and make an argument that the cases – understood as a whole – can be explained as
giving effect to some ideal that coincides with what we would like the law to be
doing. Some cases fit within the argument with no problems. Others can be massaged
to fit the overall argument, either by focussing on particular features of the case and
arguing that they ‘really’ provided the motivation for the decision, or by focussing on
a few lines in the judgment as revealing the ‘real’ reason for the decision. And then
there are the remaining cases that just don’t fit, and they are categorised as ‘wrongly
decided’.
We are all familiar with these kinds of articles. What is a bit more mysterious is
why we write articles like these. Why don’t we just say – ‘This is what we think the
law should be doing, and sometimes it does it, and sometimes it doesn’t, and where it
doesn’t, it needs to be changed.’ The suggestion made by Roberto Unger, and with
which Jeremy Waldron agrees,56 is that the academic obsession with trying to
55
Pierre Schlag, ‘Normative and nowhere to go’ (1990) 43 Stanford Law Review 167, 172.
56
See Roberto Unger, What Should Legal Analysis Become? (Verso, 1996), and Jeremy Waldron’s
review of Unger’s book in ‘Dirty little secret’ (1998) 98 Columbia Law Review 510, especially at 517-
 25

establish that the cases, viewed as a whole, already give effect to some ideal that the
author wants the law to give effect to is that the author is really addressing him or
herself to the judiciary and is essentially saying to the judges: ‘You have to give effect
to my ideal vision for the law, because that ideal vision is already embedded in the
case law.’
So, for example, restitution scholars who wished to see the law explicitly give
effect to the principle that ‘P will be entitled to sue D for restitution of a gain made by
D if D will be unjustly enriched at P’s expense if he is allowed to keep that gain’
never wrote articles saying ‘Look – here’s a principle that we think the law should
give effect to, and this is why, and if the law falls short of implementing that
principle, it needs to be changed.’ Instead, they wrote articles saying ‘This principle is
already part of the law, and we just need to wake up to that fact, and iron out any
inconsistencies in the law that have arisen because of a failure to recognise that this
principle is already part of the law.’ And that is because, I believe, they were
essentially writing for judges, and attempting to get the judges to sign up to giving
effect to what they would like the law to say.
I think we tort lawyers should try very hard to avoid this model of writing. 57 The
process of massaging and squeezing the cases to fit a particular model of what we
think tort law should be is, in the long run, damaging to the quality of our writing and
thinking. The role of a prophet crying in the wilderness is not without honour, and one
that we should not be afraid to assume if it proves that – honestly examined – tort law
does not in fact live up to the ideals we have for it.
It is to help ensure that we do not fall into the temptation of writing ‘the zeitgeist
is going my way’ articles that I suggest that, in addressing my fourth question, we
approach tort law in a critical spirit, seeking to identify any aspects of the law where it
falls short of the ideals we have for it. In order to shore that critical spirit up still
further, I’d like to point out a few very well-established features of tort law that seem
to be very hard to justify:

(1) Companies. I don’t want to suggest that tort law’s protections should only be
reserved for real people. Companies should obviously have basic rights not to have
their property seized or taken by the State or individuals. But I do want to suggest that
a company’s interest in not having its property destroyed or damaged may not be
weighty enough to justify burdening the rest of us with duties to take care not to
destroy or damage a company’s assets. When it comes down to it, a company’s
interest in not having its property destroyed or damaged is purely economic in nature,
and purely economic interests are not generally regarded in the law as sufficient on
their own to justify giving rise to duties of care in other people. On the same basis, it
could be argued that the only basic rights companies should enjoy in respect of their
reputation is a right that other people not maliciously damage their reputation –
something which would give rise to a claim under the tort of malicious falsehood. So
it may be questioned whether the law of defamation should protect companies at all,
as it unquestionably does at the moment.

519 (agreeing with Unger, that the academic obsession with influencing judges is a reaction both to
political defeat at the legislative level, and the perception that it is easier to effect change through
judges than through democratic processes).
57
Which is not to say that we should give up on attempting to influence judges. But we should seek to
do so not through appeals to authority, but to their reason.
 26

(2) Full compensation. I observed above that no one had yet successfully made an
argument that it would not be unfair to make a tortfeasor pay the victim of his tort
compensatory damages designed to put the victim (so far as money can do it) in the
position he or she would have been in had the tort not been committed. It may be that
so far as human tortfeasors are concerned, no such argument can be made. If this is
the case, we should revisit the idea that the standard response to someone’s
committing a tort is to make him or her compensate the victim of the tort in full for
the non-remote losses resulting from that tort. Moderating the amount of damages
payable to take account of a tortfeasor’s ability to pay – at least in cases where the
tortfeasor is a real life human being, and could not reasonably be expected to carry
liability insurance – might make it easier to justify making a tortfeasor pay
compensatory damages to the victim of his tort. It should also be noted that adopting a
‘measured’ approach to assessing the damages payable to the victim of a tort would
lessen the pressure on ordinary people to obtain liability insurance to cover potentially
catastrophic awards of damages being made against them, which pressure allows
liability insurers a huge amount of completely unregulated power to dictate to people
how to conduct their lives and what sort of risks they are allowed to take.

(3) Vicarious liability. If the institution of vicarious liability did not exist, would
anyone feel any need to invent it? Arguments that the rules on vicarious liability are
necessary, in order to allow us to determine when an artificial legal person has
committed a tort58 do not work: we already have such rules (‘rules of attribution’) and
they look nothing like the rules on vicarious liability. Recent decisions on vicarious
liability seem to have been influenced by a perception that where A employs B to do
work for him and the nature of B’s work means that it is likely that he will commit a
particular kind of tort T, A should take steps to guard against the possibility that B
will commit tort T and if he does not (and B commits tort T), then A should incur
some kind of sanction.59 But the law on vicarious liability is not needed to give effect
to this perception. We could easily (and I think rightly) find that A owes those who
might be victims of B’s committing tort T a duty to take reasonable steps to see that B
does not commit that tort. We could also (and again, I think justifiably) reverse the
burden of proof in cases where it is alleged that A has breached that duty of care, so
that A will be found to have breached this duty unless he can prove that he did take
reasonable steps to see that B would not commit tort T.

(4) Strict duties. There are two kinds of strict duties that tort law imposes on us. First,
strict duties of care, where the amount of care we are required to take with regard to
another’s interests is assessed objectively and not according to how much care we are
capable of taking. Secondly, duties to ensure that something does not happen, such as
that we do not defame someone unjustly, or that we do not treat property belonging to
another as our own to dispose of. I think the first type of strict duty can be easily
justified in most situations, but the second type of strict duty is much harder to justify.
However, the matter is so complex, I have deferred discussion of it to an Appendix to
this paper.60

58
See Stevens, Torts and Rights, above, n 4, 260-262.
59
See Gravil v Carroll [2008] EWCA Civ 689 (vicarious liability for rugby player throwing punch on
field), Maga v Trustees of the Birmingham Archdiocese of the Roman Catholic Church [2010] 1 WLR
1441 (vicarious liability for priest sexually abusing child he had met, but not through his being a
priest).
60
See below, p 28.
 27

VI. CONCLUSION

I began with the great German mathematician David Hilbert, and I would like to
finish by going back to him. On Hilbert’s tombstone in the German city of Göttingen,
there are inscribed some words, not from his 1900 speech to the International
Congress of Mathematicians, but from a speech that he gave 30 years later on his
retirement, to the Society of German Scientists and Physicians:

Wir műssen wissen

Wir werden wissen

In English – ‘We must know, we will know’. I would suggest these words as a good
slogan for tort lawyers to adopt. The questions I have been discussing in this paper are
questions to which we must know the answers. We cannot carry on saying that it is
impossible to say what we are talking about, when we are talking about tort law; or
that we don’t really know whether tort law serves any rational or justifiable purpose.
The people whose freedoms are limited every year in the name of tort law are owed
more than that.
As for whether we will ever know the answers to the questions identified in this
paper, I am very doubtful that we will unless we come together, and work together, on
answering these questions. I look forward to a day when instead of firing papers at
each other, people who take different views on the questions I have been discussing
will write papers together, explaining and working through their differences. No more
papers written just by Ripstein or just by Goldberg and Zipurky, but papers written by
Ripstein and Goldberg/Zipursky, together. Not just Stevens and just Stapleton, but
Stapleton and Stevens, together. Can any of these amazingly distinguished academics
really tell us anything more on their own than they have told us already? But together,
there is a real hope that they can say something new, and progress can be made. Even
if a joint paper does not result in unanimity, the process of working together and
achieving a very deep understanding of where the other person is coming from must
be beneficial, and holds out the hope of future breakthroughs.61 So let’s resolve to
come together to work out the answers to the questions discussed in this paper, and
arrive at some determinate answers to them before we reach retirement age, thereby
allowing the next generation of tort academics to grapple with fresh challenges and
fresh questions, as yet unimagined.

61
In this respect, the recent blog debate between Alan Calnan and John Goldberg & Ben Zipursky on
the basis of tort law at http://lawprofessors.typepad.com/tortsprof/2012/01/quandaries-of-civil-
recourse-calnan-replies-to-goldberg-zipursky.html#more provides a model for future collaboration
between seeming opponents on tort law issues.
 28

APPENDIX: STRICT DUTIES62

As has already been observed, tort law imposes two kinds of strict duties on us:

(1) Strict duties of care. These require us to take an objective degree of care with
regard to other people’s interests.

(2) Duties to ensure that something does not happen. These require us to ensure that a
particular event does not occur. Examples of such duties are our duties to ensure that
we do not defame another unjustly, or that we do not convert another’s property by
treating it as our own to dispose of.

Kantian tort theorists have no problem accounting for the existence of strict duties in
tort law. For them, the legal duties that tort law imposes on us serve as boundary
markers, telling us how far we can go before we will start violating someone’s
independence as a person. If we go over those boundaries, then we have done
something wrong – and how and why we went over those boundaries will be
irrelevant.
For me, strict duties in tort law are more of a problem. That is because I take the
view – as thinkers like HLA Hart and Lon Fuller have done before me – that legal
duties exist to provide us with a guide to conduct. 63 If one takes this view, then strict
duties are a problem because their strictness means we are not always capable of
complying with them. Defects in knowledge, skill and resources will lead even the
most law-abiding individual into violating a strict duty at some point. So why would
tort law ever impose such duties on us?
One response to this problem might be to say that it’s a problem that doesn’t
actually exist, except in the minds of people who believe in (B) and (C). The idea is
that people who believe in (B) and (C) see a defendant being held strictly liable in
defamation or conversion to pay a claimant damages and they think (because they
believe (B) and (C)), ‘Oh, the fact that the defendant is being held liable even though
his conduct was blameless must mean that he has breached a strict duty owed to the
claimant to ensure that he not defame the claimant unjustly/not convert the claimant’s
property. How very strange – why would the law impose such strict duties on people
like the defendant?’ But (so the accusation goes) the ‘must’ in the above sentence
only follows if (B) and (C) are correct. For someone who believes in (A), the fact that
the defendant is being held strictly liable in defamation or conversion to pay a
claimant damages does not necessarily indicate that he has breached a duty owed to
the claimant, but may just indicate that it is fair, just and reasonable (for one reason or
another) for the defendant to pay up, despite the lack of fault on his part.
As this paper has been written on the basis that (B) and (C) are correct, I think I
can put this objection aside and instead assume that tort law does impose strict legal
duties on us, and ask how that might be justified. But a beneficial side effect of
62
I am grateful to Jane Stapleton for pressing me to address the issue of strict duties in this paper. The
issue is, of course, discussed in Gardner’s essay on ‘Obligations and outcomes in the law of torts’,
above, n 45, where he distinguishes between ‘duties to succeed’ and ‘duties to try’. I haven’t followed
Gardner’s terminology in what follows because it seems to me that both of the strict duties I am talking
about are ‘duties to succeed’, if they are anything.
63
I think John Gardner, in addressing this issue, is writing more within the Kantian tradition, where
legal duties provide us with an authoritative guide as to what we are morally required to do. So if there
are moral duties to succeed (as he argues that there are), then there is no problem with the law
imposing on us legal duties to succeed.
 29

proceeding in this way may be that if I can show that the law might justifiably impose
strict legal duties on us, that may make it more plausible for those who currently
believe (A) to switch sides and start thinking that (B) and (C) are correct.

1. Strict duties of care

Let’s begin with strict duties of care. To keep the discussion simple, I will consider
the case where Learner and Experienced are each driving on the same area of road. In
principle, we might just want to impose on Learner a measured duty to take as much
care as he is capable of not to drive dangerously, as that is the only duty that Learner
could be expected to comply with. However, a number of different considerations will
lead us instead to impose a strict duty of care on Learner that requires him to take as
much care as a normal driver could be expected to take not to drive dangerously:

(1) In this case, Learner and Experienced each owe each other duties of care not to
drive dangerously. If those duties of care were measured, then Learner would end up
owing Experienced a less demanding duty of care than Experienced would end up
owing Learner. This might be thought unfair: it would seem to punish Experienced
for her expertise, and reward Learner for his inexperience.

(2) Similarly, Other Drivers will owe Experienced duties of care not to drive
dangerously. If those duties of care were measured, then they may end up owing
Experienced more demanding duties of care than Learner would owe Experienced.
This, again, might be thought to be unfair as it punishes the Other Drivers for not
being like Learner and reward Learner for not being like Other Drivers.

(3) Learner could have avoided exposing Experienced to danger by simply doing
nothing. Given that Learner chose to put Experienced at risk of harm by going out
onto the road, it is not unfair to require Learner to take the same sort of care as
normal drivers to ensure that his choice to drive does not result in Experienced
suffering harm, even if Learner is incapable of meeting that standard of care. If
Learner had not wanted to be subjected to such a demanding standard of care, he
could have avoided it by simply not driving.

In contrast, in the case where Landowner has a fire start on his land which threatens
Neighbour’s property, then the case for imposing a strict duty of care on Landowner
to look out for Neighbour’s interests grows considerably weaker. As there is no
danger on Neighbour’s land that threatens Landowner’s property, the duty of care that
Landowner owes Neighbour is not matched by a corresponding duty of care that
Neighbour owes Landowner. And the duty of care that Landowner owes Neighbour
does not correspond with a duty of care that anyone else owes Neighbour: as the
danger to Neighbour’s property has arisen on Landowner’s land, only Landowner is
under a duty of care to abate it. And Landowner did not do anything to incur that duty
of care that might make it not unfair if that duty of care were too demanding for
Landowner to live up to. It may not be surprising, then, that the duty of care that
Landowner owes Neighbour is measured, rather than strict.64

64
Goldman v Hargrave [1967] 1 AC 465.
 30

2. Duties to ensure

Let’s now turn to duties to ensure that something does not happen. Again, to keep the
discussion simple, I’ll consider the case where Customer goes to Jeweller’s store,
offering to sell Jeweller a watch. Customer offers Jeweller proof that the watch
belongs to him, and talks knowledgeably about its history and provenance. Jeweller
buys the watch and then later gives it away as a birthday present. The watch is
subsequently lost or destroyed. It turns out that the watch was stolen, and the true
Owner successfully sues Jeweller for converting his watch, both in taking possession
of it, and then in giving it away.
Again, in principle, we might just want to impose a measured duty of care on
Jeweller not to treat Owner’s watch as his own to dispose of, as that is the only duty
that Jeweller could be expected to comply with. So how can the law be justified in
imposing on Jeweller a duty to ensure that he does not convert Owner’s watch? It
seems to me that the most promising way of justifying imposing such a duty on
Jeweller would go as follows.
(1) We want Jeweller to do what he can to take care to see that he does not get
involved with buying or selling stolen property.
(2) But if we simply impose a duty of care on Jeweller to this effect (what we
can call ‘the target duty’), Jeweller will consistently underestimate how much care he
needs to take to avoid getting involved with dealing in stolen property. In other words,
he will consistently think that he has taken adequate precautions against this risk,
when he has not.
(3) In order to help ensure that Jeweller complies more effectively with the
target duty, we need to impose on Jeweller a more demanding duty to ensure that he
does not get involved with dealing in stolen property. Imposing such a duty on
Jeweller will put him ‘on edge’ and encourage him to take special measures to ensure
that he does not deal in stolen property, and thereby help ensure that he hits the ‘target
duty’ more often than he would do if he were simply subject to that ‘target duty’.
The crucial step in this argument is (2). In this context, I think (2) is quite
plausible. The law on fiduciary duties owes its existence to the courts’
acknowledgment that self-interest can distort people’s judgments as to whether they
are doing the right thing: fiduciaries duties exist in order to eliminate that distorting
effect. I see no reason why a duty to ensure that he does not trade in stolen goods
could not be imposed on a jeweller with the same object.
However, I think this argument only works effectively in this context, where
Jeweller is operating a business. I find it hard to see how this argument could work to
justify the existence of duties to ensure that you do not unjustly defame me, or
trespass on my land. (Though it might in particular contexts, for example where it is a
newspaper doing the defaming, or the State intruding onto my land.)
Instead of assuming that there must exist some justification for the existence of
these duties to ensure, we should consider the possibility that they owe their existence
to a simple confusion. It is noticeable that these duties to ensure that something does
not happen tend to crop up in relation to torts that can be committed in ways that
allow plaintiffs to seek a remedy against someone who is currently committing the
tort. For example, if Jeweller were still in possession of Owner’s watch, Owner would
simply go to court for an order that Jeweller hand over the watch or its value, and he
would get it. It would be no defence for Jeweller to argue that he had not acted
 31

carelessly in buying the watch. It seems a plausible suggestion that the courts have
made the incorrect inference that if the Jeweller’s degree of care (or lack of it) is
irrelevant to his liability in the case where he still has Owner’s watch, it cannot be
relevant either in the case where he no longer has Owner’s watch. After all, the courts
have made the same elision between the case where Jeweller still has Owner’s watch
and the case where Jeweller no longer has Owner’s watch in connection with the
ground of Jeweller’s liability. In the case where Jeweller still has Owner’s watch,
Owner could not simply go to court and say, ‘He has got my watch – make him give it
back, or its value!’ Owner had to go to court and say, ‘He is acting inconsistently with
my rights in hanging on to my watch – make him give it back, or its value!’ But in
allowing Owner’s claim on that basis in the case where Jeweller still had Owner’s
watch, the idea arose that doing anything with the watch that was inconsistent with
Owner’s rights over the watch (such as taking possession of it, or handing it over to
someone else) would make someone liable to Owner to pay Owner the value of the
watch.
So the courts say that Jeweller is under a duty to ensure that he does not convert
Owner’s watch because his liability for converting Owner’s watch is never dependent
on its being established that he was at fault for converting Owner’s watch. But
Jeweller’s liability is never dependent on his being at fault because in the case where
Jeweller is still in possession of Owner’s watch, his fault is irrelevant to his liability
and the courts wrongly assume (in, frankly, a typically and exasperatingly English
way) that what is true in that case must be true in all cases where Owner sues Jeweller
in conversion.
A rational tort law that is built on formulae such as (G) and (H) will rid itself of
confusions like these, if confusions they be. Under such a tort law, we could expect
the existence of duties to ensure that something does not happen to be limited to cases
where arguments analogous to the one made out in steps (1) – (3) above can be made
out.