University of Edinburgh

School of Law

Research Paper Series

No 2014/45

Three Errors in the Defamation Act 2013

Eric Descheemaeker
Lecturer in European Private Law
University of Edinburgh, School of Law
[email protected]

6 Journal of European Tort Law (2015), forthcoming

This text may be downloaded for personal research purposes only. Any additional
reproduction for other purposes, whether in hard copy or electronically, requires the consent
of the author(s). If cited or quoted, reference should be made to the name(s) of the author(s),
the title, the number, and the working paper series
© 2014 Eric Descheemaeker

Edinburgh School of Law Research Paper Series

University of Edinburgh

Electronic copy available at: http://ssrn.com/abstract=2524338

Abstract
This article considers three aspects of a recent British statute on the law of defamation, the
Defamation Act 2013, disputing in each case their opportuneness. First, it argues that the new
requirement of ‘serious harm’ under sec 1 runs against basic tenets of the law. Second, it
expresses concern about the new drafting of the defence of responsible journalism (sec 4),
which is in danger of losing touch with its original rationale. Third, it examines the revamped
version of the defence of fair comment, now known as ‘honest opinion’ (sec 3), and suggests
that comment should never be actionable because it cannot be defamatory in the first place.

Keywords
Defamation, libel, Defamation Act 2013, serious harm, responsible journalism, publication
on matter of public interest, fair comment, honest opinion

Electronic copy available at: http://ssrn.com/abstract=2524338

 University of Edinburgh School of Law Research Paper 2014/45

I Introduction
After a long-winded parliamentary process, which started in 2010 with the introduction of a
private bill by Lord Lester of Herne Hill and nearly derailed towards the end of 2012, short-
circuited by the Leveson Inquiry into the practices of the British media – a new Defamation
Act was passed in England and Wales: the third one since the Second World War, after those
of 1952 and 1996. Its main ambition was transparently to continue making the English law of
defamation less restrictive on the press (and others who might be impacted by it), either by
putting on a statutory footing common-law developments that had occurred since the Human
Rights Act 1998 or by taking further steps in this process of ‘liberalisation’.
To anyone interested in that area of the law, two excellent resources have already appeared: a
new edition of the leading practitioners’ text, Gatley on Libel and Slander, which has been
rewritten in some depth in light of the Act,1 and the Blackstone’s Guide to the Defamation
Act 2013, which helpfully incorporates key passages from the preparatory works.2 With at
least one excellent commentary on the whole Act having also already appeared,3 it seemed
more worthwhile in the present context to select a few issues and offer an in-depth
examination. Accordingly this commentary will focus on three dimensions of the Act where
it will be argued that Parliament has either taken a wrong turn or failed to correct existing
problems. They concern: (i) the new requirement of ‘serious harm’ (which runs against basic
tenets of the law of defamation); (ii) the updated version of the defence of responsible
journalism (which is in danger of losing touch with its rationale); and (iii) the revamped
defence of ‘honest opinion’ (which cosmetically alters the previous version known as ‘fair
comment’ when a wholesale re-examination would have been in order). Some historical and
comparative context will be offered in respect of all these points.

II The requirement of ‘serious harm’

Section 1 of the Act provides that:
1 Serious harm
(1) A statement is not defamatory unless its publication has caused or is likely to cause serious
harm to the reputation of the claimant.
(2) … [H]arm to the reputation of a body that trades for profit is not ‘serious harm’ unless it has
caused or is likely to cause the body serious financial loss.
The intention behind the section is unmistakably to deter claims regarded as trivial or
spurious and, by so doing, further to entrench the protection of freedom of expression in
English law: this is part of a much broader pattern, so broad in fact that it has been rightly
doubted whether the English law of defamation really is as skewed towards claimants as
common opinion makes it out to be, or even at all.4 But, whether the intention is right or
wrong, the way it works itself out in this section causes significant difficulties, which go to
the heart of what defamation exists for as a tort, which need to be examined closely.

1 A Mullis/R Parkes (eds), Gatley on Libel and Slander (12th edn 2013) (hereinafter: Gatley).
2 J Price/F McMahon (eds), Blackstone’s Guide to the Defamation Act 2013 (2013).
3 A Mullis/A Scott, Tilting at Windmills: the Defamation Act 2013 (2014) 77 Modern Law
Review 87.
4 A Mullis/A Scott, The Swing of the Pendulum: Reputation, Expression and the Recentering of
English Libel Law (2012) 63 Northern Ireland Legal Quarterly (NILQ) 27.

Electronic copy available at: http://ssrn.com/abstract=2524338

 University of Edinburgh School of Law Research Paper 2014/45

There are at least three issues that one needs to take argument with here: the requirement that
the injury to reputation be ‘serious’; the need on the part of companies to prove (actual or
likely) financial loss before they can sue in defamation; and finally the question of the causal
link between wrong and injury (‘has caused or is likely to cause’). For the sake of clarity we
need to keep them as separate as possible.

A The ‘seriousness’ of the injury to reputation

First, we turn to the seriousness of the injury to reputation, which forces us to return to first
principles. What interest, or interests, the wrong of defamation protects is not as
straightforward a question as it might seem: indeed, it can quickly descend into considerable
difficulties. In a system like French law, the – criminal – wrong of diffamation is defined as
an injury (atteinte, encroachment) to the victim’s honour (honneur) or estimation
(considération, the modern rendering of Donellus’ existimatio).5 Both concepts are difficult
and it is not obvious how they relate to one another. Honour is clearly reminiscent of the
Roman roots of the wrong in the delict of iniuria, which protected dignitas: dignity in the
sense of rank or status.6 English defamation, since at least the 19th century, has consistently
understood itself in terms of an injury to reputation, that is, – like existimatio – the
‘reckoning’ of the victim in the eyes of others (Latin reputatio, the act of thinking about).7 Of
course the concept of reputation is a difficult legal construct in itself8 and, on closer analysis,
the equiparation between defamation and reputation is imperfect at best; 9 nonetheless the
basic stance of the law was settled on a position that was both intelligible and, in the present
writer’s opinion, entirely correct: defamation as a wrong protects the claimant’s interest in
the way third parties think of him.
Against this background, the requirement of the Defamation Act 2013 that ‘a statement is not
defamatory unless its publication has caused … serious harm to the reputation of the
claimant’ is puzzling – indeed puzzling twice over. For one thing, it is very difficult to see
how the definition of defamatoriness itself could depend on the magnitude of the injury
caused in the mind of others. By contrast, no-one would want to say that a physical injury is
not a physical injury unless it is serious enough to warrant a trip to the hospital: this would be
an attempt to change the meaning of words belonging to ordinary language, which is beyond
the mandate of the law. What the law might plausibly say is that no redress will be given
unless a trip to the hospital was warranted: this would be a policy decision within the power
of the law to make. But the two issues are distinct. Similarly, it seems that what the Act is
trying to say is that a defamatory statement will not be actionable unless its publication has
caused not simply harm, but serious harm, to the claimant’s reputation.10
This, however, appears misguided too. If we take it at face value it would fly in the face of
the long-accepted proposition that the tort of defamation protects the interest in reputation: if
‘non-serious’ injuries can never be redressed through the operation of the wrong then

5 Law of 29 July 1881 on the Freedom of the Press, art 29.

6 E Descheemaeker/H Scott, Iniuria and the Common Law, in: E Descheemaeker/H Scott, Iniuria
and the Common Law (2013) 1, 13, 16 f, 19.
7 L McNamara, Reputation and Defamation (2007) 1.
8 Ibid, part I.
9 E Descheemaeker, Protecting Reputation: Defamation and Negligence (2009) 29 Oxford Journal
of Legal Studies (OJLS) 603, 610-619.
10 This confusion was noted during the debates: see Price/McMahon (fn 2) § 2.40 (Sir Edward
Garnier).

2
 University of Edinburgh School of Law Research Paper 2014/45

defamation cannot be said to protect the claimant against encroachments to his reputation.11
We could not say that it now protects ‘serious reputation’ because this makes no sense
linguistically, but a proposition to that effect will need to be coined. More fundamentally, as
a matter of principle, why would the law want to operate such a restriction? That trivial
injuries ought not to get their day in court is naturally an ancient idea that is not necessarily
objectionable (de minimis non curat praetor or lex); but this is normally dealt with through
the operation of rules that are external to the cause of action being found, for instance, in the
rules of civil procedure. To pursue the parallel with physical injuries, no-one would want to
say that a physical injury is not actionable in negligence or in trespass unless it reaches a
certain degree of gravity. The law relies instead on rules pertaining to abuse of process, or
indeed on the common-sense principle that hardly anyone would want to sue if they stand to
lose more than they gain. Why introduce a different logic into the law of defamation, taking it
yet further away from the mainstream of tort? The fact that, in the pursuit of a personal feud,
some affluent claimants might want to bring a defamation claim over a trivial matter, even if
it costs them, is not good enough a reason (assuming it is in fact an issue): they could just as
well sue over a bruised thumb out of sheer spite. If the law decides that such claims do not
deserve a hearing – which incidentally is far from self-evident – it should deal with them in a
principled way, not with ad hoc rules distorting the definition of individual causes of action.

B Companies and financial loss

The second issue concerns the proposition that, for a body that trades for profit, the ‘serious
harm’ requirement is translated into a requirement of serious financial loss.12 Again, the issue
is twofold: terminological and substantive. First, the wording is infelicitous. The ‘serious
harm’ looks back to the first subsection, where it explicitly means harm to reputation –
rightly so, this being defamation – and it makes no sense to say that harm to reputation is not
serious unless it causes financial loss. It is observably false that an egregiously defamatory
statement will always cause economic loss, even in the case of a company, and so for the law
to assert this is again an impermissible attempt to pervert ordinary language. What the
drafters meant to say is that ‘a body that trades for profit cannot sue in defamation unless it
establishes [serious] financial loss caused by the [seriously?] defamatory statement’.
The main issue is however not presentational; it is substantive – indeed, foundational. It
relates to the relationship between defamation and economic losses and, intertwined with
this, the treatment of juridical persons under that cause of action. Non-physical persons
present, of course, a perennial problem for private law, which reached its modern form (in
Roman and English law alike) long before anything remotely like our juridical persons came
into existence. Rules designed with human beings in mind had to be transposed to profoundly
different entities. Within tort, it is clear that the law struggles much more, to use a division
which already underpinned the Roman distinction between damnum iniuria and iniuria, on its
‘being’ side (protecting the claimant’s personality) than on its ‘having’ side (protecting their
patrimony). It is not difficult to see why: that a company, or association, or charity, or local
authority has suffered a wrong if its property is damaged or taken away is almost impossible
to deny (once we have accepted the principle of a juridical person); but whether or not they
have suffered a wrong if they are, for example, spoken ill of – and if so what the
consequences should be – is much less clear.
In part, this is woven in with the question of injured feelings: it is a common argument that
the reason why persons are granted a right to reputation is because being defamed wounds

11 This is a very different restriction from saying eg that only culpable encroachments are
redressable, or that the operation of the tort is subject to defences.
12 Sec 1(2), extracted above in the text.

3
 University of Edinburgh School of Law Research Paper 2014/45

their feelings, or perhaps their sense of self-esteem. On that basis it is easy enough to find
that non-sentient beings cannot suffer this ulterior injury and, thus, ought not to be granted a
right to reputation. To which it will invariably be a reply that a defamatory statement can still
injure a juridical person in their pocket, thereby warranting protection of their reputation.
Torn between these two conflicting ideas, the common law as a whole has been unable to
settle on a principled position. While it is happy to recognise in principle that juridical
persons have a reputation and can be defamed,13 it is constantly trying to restrict their right of
action some way or another: to take but one recent example, the uniform Defamation Acts
passed in Australia in 2005 removed the right to sue in defamation for trading corporations
with fewer than ten employees.14
The present writer has argued elsewhere that both issues – the question of wounded feelings
(if any) and that of possible consequential economic loss (which affects all persons, not just
companies or juridical persons) – are distractions.15 In particular, that companies might be
injured in their pocket by a defamatory statement is self-evident, but this does not explain
why they should not be able to sue in defamation unless and until they suffer economic loss.
While the underlying objective is clear and easy to sympathise with – to protect individuals
against the threat of lawsuits from corporations, especially large ones instrumentalising the
law of defamation to effectively censor the expression of critical views about them (as in, for
example, the infamous ‘McLibel’ case 16 ) – the ‘solution’ adopted by the Act only brings
confusion into the very foundations of defamation law.17
Two issues of principle arise. The first is: why should the rule only apply to ‘bodies trading
for profit’? Loss of goodwill is not the only way in which a defamed party might suffer in
their pocket, and all persons – physical persons, for-profit legal persons and not-for-profit
legal persons – will potentially be affected in this way by a defamatory statement (indeed, the
economic consequences might be far more significant for physical persons who might, for
instance, lose their job in circumstances where they are unlikely to find another one). The
preparatory works disclose no good reason, besides a rhetorical appeal to the controversial
cases of ‘McLibel’ and Singh,18 to draw a distinction between different classes of persons.19

13 For some history, see F Patfield, The Origins of a Company’s Right to Sue for Defamation
(1994) 45 NILQ 233.
14 D Rolph, A Critique of the National, Uniform Defamation Laws (2008) 16 Torts Law Journal
(2008) 207, 216 and references cited therein.
15 E Descheemaeker, Three Keys to Defamation: Media 24 in a Comparative Perspective (2013) 130
South African Law Journal (SALJ) 435.
16 McDonald’s Corporation v Steel & Morris [1997] England and Wales High Court, Queen’s Bench
Division (EWHC QB) 366, where McDonald’s won a £40,000 libel suit against two environmental
activists over claims contained in a critical pamphlet (though the firm never collected the damages).
17 It is not even clear how the provision prevents the occurrence of future McLibels, as it does
not shield the defendant who really has caused (serious) financial harm by his words – the very
situation where he might need to be protected from claims for very large amounts of consequential
economic loss.
18 British Chiropractic Association v Singh [2009] EWHC QB 1101; [2010] England and Wales Court
of Appeal, Civil Division (EWCA Civ) 350, concerned a case where a journalist had been sued by a
professional body for having written in a column in The Guardian that the association of chiropractors
‘happily promotes bogus treatments’. They lost on appeal but their victory at first instance, which
caused a lot of ink to flow, is widely regarded as one of the main drivers behind the Defamation Bill
introduced in 2010.
19 Defamation Bill Committee Stage Report: Bill No 51 2012-13 (Research Paper 12/49) (2012)
§ 3.13.

4
 University of Edinburgh School of Law Research Paper 2014/45

The second issue is how this stance can be reconciled with the proposition that the wrong of
defamation protects the claimant’s reputation? Here the law, after paying lip service to the
proposition, goes on to assert that no action lies unless a different injury has been or is likely
to be suffered, namely, financial loss.20 The logical implication is that defamation does not
protect reputation but the ‘temporal’ consequences of its violation, that is, those which can be
valued in money: in that sense, defamation would have to be regarded as an economic tort. If
nothing else, this is incompatible with the rhetoric of the law; one might even go further and
wonder if it is meaningful to say that persons on whom such restrictions are imposed have a
right to reputation in the first place.21
Here, the position of English law can be helpfully contrasted with that of South African law.
In the recent case of Media 24,22 the Supreme Court of Appeal in South Africa faced the very
situation which has prompted the hesitancy of the modern law in such jurisdictions as
England or Australia: a company that had been defamed, in the ordinary sense, by a
newspaper claimed not only – modest – compensation for the injury to their reputation per se
(‘general damages’) but also a – very large – sum to redress their alleged loss of profits
(‘special damages’), prompting concerns about the excessive restriction of free expression.
The answer of the court was that the company could indeed receive compensation for its
injured reputation by suing in defamation but, if it wanted to recover for its economic loss, it
would need to use a different cause of action and sue in injurious falsehood (regarded as part
of the actio legis Aquiliae, the action on the Aquilian Law, in a South African context).23
Crucially, what this entails is that demonstrating that the injury to reputation was wrongful
does not in itself prove that the economic loss was also wrongfully caused: in order to
establish this latter proposition, the claimant would need to clear the usual – high – hurdles
standing between him and recovery for his (now pure) economic loss. Although the approach
of the Supreme Court of Appeal has momentous consequences in terms of our understanding
of consequential loss, which are not pursued here, it is to this writer’s mind entirely correct as
a matter of principle; and in practice it would also eliminate any concerns regarding the
‘chilling effect’ of corporations claiming vast sums to compensate for their allegedly lost
profits.

C ‘[L]ikely to cause harm’

One further issue, intertwined with both of the above points, calls for separate attention: in
both subsections, the Act tells us that in order to be recoverable the relevant injury should
either have been caused or be ‘likely’ to be caused by the prior event (the publication of the
statement or injury to reputation). This alternative requirement is a difficult one, in which
Parliament seems to have conflated – indeed, confused – different ideas.
Tort, in English law, normally requires the injury complained of by the claimant to have been
caused by the defendant, issues of likelihood coming up one step removed through the
operation of the ‘balance of probabilities’ test (that is, it must be established that it is more
likely than not that the defendant’s wrongful conduct was the cause of the claimant’s loss).

20 It is telling that the phrasing of the Act surreptitiously tries to make financial loss a
component part of the injury to reputation itself. This would reconcile the rule with first principles
but, as explained, it is simply impossible to accept.
21 Admittedly this continues a very old theme within the law of defamation. The same remark
can, for instance, be applied to instances of slander not actionable ‘per se’; see Descheemaeker (2009) 29
OJLS 603, 616 f.
22 Media 24 v SA Taxi Securitisation (437/2010) [2011] South Africa Supreme Court of Appeal
(ZASCA) 117.
23 Descheemaeker (2013) 130 SALJ 435, 441 f.

5
 University of Edinburgh School of Law Research Paper 2014/45

But defamation combines the ideas of causation and likelihood in a markedly different way,
which is liable to cause confusion. Historically, the injury to the claimant’s reputation was
always part of the cause of action: indeed it would have been extraordinarily puzzling if a
person could have sued in defamation without alleging that their reputation had been
violated. But, perhaps because such injuries are invisible (in that they exist in the mind of
third parties), courts rapidly came to regard the allegation as irrebuttably presumed from the
publication of the statement complained of: in the old language of the law it became ‘non-
traversable’, that is, it was not open to the defendant to rebut the presumption by proving that
no-one had actually thought the lesser of the claimant because of his statement.24 This was
bound to have a knock-on effect on the very definition of what counts as defamatory: instead
of saying that a defamatory statement was one that caused harm to reputation, such harm
being irrebuttably presumed, the law switched to the functionally equivalent proposition that
a defamatory statement is one that has a tendency to cause (that is, is capable of causing)
such injury. 25 The issue of actionable loss was turned into one of (prospective) causation,
rendering defamation an attack on reputation rather than an injury thereto.26
The drafters of the new Defamation Act picked up on this old theme but, perhaps because
they misunderstood its origins, took it in a different and, again, perplexing direction. In the
wording of sec 1(1), it seems that ‘has caused or is likely to cause’ no longer means ‘has a
tendency to cause’ but rather ‘has already caused or will probably cause in the future’. 27
There are at least two difficulties here. The first is how the new rule relates to the old
presumption of injury. If the latter has been retained, the injury to reputation would still be
irrebuttably presumed from the publication of the defamatory statement (hence at the point of
publication). But then – unless we accept that some injury has already been caused but not yet
a serious injury – the wording of subsec 1 would make no sense: an injury cannot be likely to
happen in the future which is deemed to have already occurred.
The second difficulty is that if, on the other hand, the old presumption has disappeared, we
are faced with a different – but if anything greater – problem. If the (serious) injury to
reputation (and/or the economic interests) is not conclusively established from the publication
of the defamatory statement, it follows logically that it may or may not have occurred at the
point when the action is brought. This makes sense of the alternative ‘has caused … or will
probably cause’, 28 but we now find ourselves in the highly surprising situation where a
claimant might be able to sue successfully, hence recover compensatory damages, for a loss
that has not yet occurred – and of course might never occur at all. The law is familiar with
injunctions looking ahead to a wrong that has not been committed, and it also routinely
compensates yet-to-be-suffered losses alongside already-suffered ones as part of the
valuation of the quantum of damages; but compensating ex ante an injury which has not
happened yet (and may never materialise), as opposed to ‘guesstimating’ its as-yet-unknown
extent, is a novel idea that is very difficult to reconcile with the most basic tenets of tort law.
On what basis can the claimant be regarded as having a complete cause of action when he
does not allege, and the law does not imply, the injury which is consubstantial to the tort?

24 D Ibbetson, A Historical Introduction to the Law of Obligations (1999) 116.

25 Ibid.
26 This idea is pursued in A Fagan, The Gist of Defamation in South African Law, in:
Descheemaeker/ Scott, Iniuria and the Common Law (2013) 169.
27 See Price/McMahon (fn 2) §§ 2.34-2.35 and references cited therein.
28 Though ‘probably’ seems to be redundant, at least if the injury is now to be established on the
balance of probabilities.

6
 University of Edinburgh School of Law Research Paper 2014/45

III Publication on matter of public interest

The second issue to consider is the revamped version of the defence formerly known as the
‘Reynolds privilege’ (or ‘responsible journalism’) and now referred to as ‘publication on
matter of public interest’. Section 4(1) of the Act reads:
4 Publication on matter of public interest
(1) It is a defence to an action for defamation for the defendant to show that –
(a) the statement complained of was, or formed part of, a statement on a matter of public interest;
and
(b) the defendant reasonably believed that publishing the statement complained of was in the
public interest.
This new formulation is, to the present writer, highly objectionable. At best it restates, in an
ambiguous manner, what the common law had been painstakingly trying to clarify over the
previous 15 years; at worst it is capable of being interpreted in such a way as to sever the
defence from the rationale that brought it into existence in the first place.

A Historical background
In order to understand the issue, some background is essential. Historically, the English law
of slander (as indeed the Roman law of iniuria) was from the earliest days grounded in
malice – an ill-defined concept which essentially took the form of malevolence, in the sense
of spite (as in French malveillance). Malice was presumed but could be rebutted: it is from
such attempts to rebut the presumption of malice that most of the organising categories that
underpin the modern law of defamation came into being.
One – principal – way of rebutting the presumption was to show that the situation was
covered by what the law called a ‘privilege’, which itself was defined essentially through a
nexus of duties and interests.29 What happened was that, in a number of circumstances (to
simplify, when the defendant had a ‘good reason’ to speak to the third party), the
presumption of malice was rebutted with the effect that it fell back on the claimant positively
to prove it. Accordingly the law of defamation was divided into two: situations covered by
privilege, where the standard of liability would be proven malice, and situations not covered
by such privilege, where malice was – conclusively – presumed and, thus, the defendant
would be strictly liable (unless of course he could avail himself of a different defence). As a
result, defamation constantly toed and froed between a standard of liability highly favourable
to the claimant and one very favourable to the defendant – arguably, in both cases, overly so.
The story was one of meandering between two extremes.
The rise of the modern press (and later of other media of mass communication) exposed the
untenable character of this situation. Early on, the courts decided that defendants who had
published to the world at large could not avail themselves of the defence of privilege because,
in all but the most exceptional circumstances, not all of their readers would have had an
interest in the communication of the defamatory statement to them. The result was consonant
with the logic of privilege but unduly harsh: when it came to statements making a claim to
truth, the alternative was that they could either prove the truth of the defamatory statement
(thus escaping liability) or not, in which case they would be strictly liable for having injured
the reputation of the claimant – no matter how much care they had taken in gathering the
information or how ‘newsworthy’ it was. This outcome was acceptable to very few and,
indeed, was a principal cause of the ill repute that English defamation acquired around the
world for being unconscionably tilted towards the claimant. It called for reform.
There were several possibilities. One of them would have been to hold that every member of
the public had an interest in receiving information – even if it was defamatory and possibly

29 P Mitchell, The Making of the Modern Law of Defamation (2005) 151 ff.

7
 University of Edinburgh School of Law Research Paper 2014/45

untrue – on a matter of general interest: far from being far-fetched, it would seem that this is
exactly what the (never defined) notion of ‘public interest’ must mean. Had such an approach
been followed, extending the duty/interest analysis to the world at large, English law would
have ended up with a Sullivan-type defence, whereby publishers-at-large would have been
shielded from liability on such matters unless proven malicious. This is the option for which
Lord Lester, then counsel for The Times, pushed in the seminal case of Reynolds 30 that
concerned the claim brought by the then Irish prime minister against the London newspaper
for having alleged that he had misled his national parliament: a claim which The Times was
not prepared to defend as true.
In the Court of Appeal – where the new defence was designed – Lord Bingham, while
recognising that the position of English law, as sketched above, was wrong and required
reform, held that more should be expected of people who publish indiscriminately before
they are allowed to avail themselves of the protection of the law. As a result, turning his back
on Lord Lester’s suggestion, he borrowed from the High Court of Australia – which had
recently handed down its Lange decision31 –what is essentially a standard of responsibleness
of conduct. In order to escape liability the defendant (typically a newspaper, hence the name
the defence subsequently acquired of ‘responsible journalism’) must show that the subject
matter of the incriminating statement was objectively in the public interest and, also, that he
took proper care when gathering and publishing the information. 32 Within its scope of
application, and subject to some fine-tuning, the defence is therefore one of ‘reasonable
truthfulness’, the essence of which is to argue that, even though he got the facts wrong, the
defendant should nonetheless ‘get off the hook’ because he made a reasonable endeavour at
getting them right.
It is important to see how this represents the importation of what is in essence a negligence
standard into an area of the law, defamation, which had always been controlled by malice
(and its rebuttal). In essence it was nothing short of a change of paradigm, giving effect to the
vision, set out as early as 1909 by Charles Wade, the then premier of New South Wales, that
‘a man, while exercising freedom of speech, must take care not to defame his neighbour’.33

B Responsible publication in the Defamation Act

It is against this background that we need to consider the new statutory version of the
defence, which underwent considerable change during the parliamentary procedure. It seems
to the present writer that, although very little (if anything) has in fact changed as a result of
the Act, the new formulation is at best ambiguous and, at worst, in danger of taking the law in
the wrong direction. In part, as will be argued, this is the outcome of contemporaneous
judicial developments which interfered with the debates. These points are perhaps best seen
by comparing the final version in the Act of 2013 with its earlier forms in the Bill considered
by Parliament.

30 Reynolds v Times Newspapers Ltd [1999] United Kingdom House of Lords (UKHL) 45, [2001] 2
Law Reports, Appeal Cases (AC) 127, 160 ff. The reference in the text is to New York Times Co v
Sullivan (1964) 376 United States Supreme Court Reports (US) 254, shielding untrue statements
concerning public officials from liability in defamation unless the claimant can establish that they
were published in the knowledge that they were untrue or with ‘reckless disregard’ as to their truth.
31 Lange v Australian Broadcasting Corporation (1997) 189 Commonwealth Law Reports (CLR) 520.
32 Reynolds [2001] 2 AC 127, 177 f.
33 New South Wales Parliamentary Debates, 2nd Series, vol 35 (1909) 3103; see E Descheemaeker,
‘A man must take care not to defame his neighbour’: The Origins and Significance of the Defence of
Responsible Publication (forthcoming).

8
 University of Edinburgh School of Law Research Paper 2014/45

In its earliest version, the Bill simply stated, in statutory form, what had become the
dominant judicial interpretation of what it means for a defamatory statement to be a
‘responsible publication on matter of public interest’. In trying to flesh out the two aspects of
the defence (its subject-matter and the requirement of reasonable belief in truth), Lord
Nicholls had famously articulated a ten-pronged test in the original House of Lords case, in
which he identified such relevant factors as the ‘seriousness of the allegation’, the ‘nature’
and ‘source’ of the information, the ‘urgency of the matter’, and any ‘steps taken to verify the
information’ or to include ‘the plaintiff’s side of the story’.34 This was essentially an attempt
to clarify what it means to have acted like a reasonable man (without fault) in the particular
circumstances of the publication of potentially untrue defamatory matter. The first draft of
the Bill contained a provision that looked very much like a codified version of the ‘ten-
pronged test’. But this test had long caused dissatisfaction, in particular because lower courts
had a tendency to treat the criteria as cumulative hurdles to be cleared, with the effect that
they imposed on the defendant a much higher standard than was intended when the test was
devised (and also encouraged an unhelpful box-ticking mindset on the part of the media).
This dissatisfaction resurfaced during the Bill’s consideration in Parliament and is clearly one
reason why it decided to do away with the test in the end.35

C Flood and reportage

But that is not the only, or even probably the main reason why Parliament ultimately did
away with the Reynolds test. The final wording of the second limb of subsec (1), concerning
what we could term the ‘depth’ of the defence (that is, the condition on which it applies once
the statement is found to be on a matter of public interest), not only departs from Lord
Nicholls’ test but adopts a strikingly different perspective, whereby the defendant must have
‘reasonably believed that publishing the statement complained of was in the public interest’.
To anyone familiar with the law in this area, the new formulation will appear mystifying.
Regardless of what exactly ‘public interest’ means, it was always concerned (in the context of
defamation) with the subject matter of the statement: this is still reflected in the first limb of
sec 4(1). Saying that the defendant must have believed the publication itself to be in the
public interest uses the word in a very different sense, which is still undefined but has to
mean something akin to ‘reasonable’ or ‘responsible’. This seems to be no more than a
general token of approval: ‘a good thing’ as opposed to ‘a bad thing’. It is so vague as to be
capable of any number of interpretations, but the one thing that is clear is that the drafting
takes it as far as it possibly could be from the requirement of reasonable belief in truth, which
was best captured by the Supreme Court of Canada when it recognised its own defence of
responsible publication in Grant v Torstar: ‘the defendant must show that publication was
responsible, in that he or she was diligent in trying to verify the allegation(s), having regard
to all the relevant circumstances’.36
On the face of it (although arguably not on closer analysis) this condition of ‘reasonable
truthfulness’ has now disappeared, being replaced with a new touchstone: ‘publishing in the
public interest’.37 There is only one possible explanation – which is in fact visible on the face
of the record – namely, the interference created by the judgment of the House of Lords in the

34 Reynolds [2001] 2 AC 127 , 205.

35 Price/McMahon (fn 2) § 5.38.
36 Peter Grant v Torstar Corporation, 2009 Supreme Court of Canada (SCC) 61, [98].
37 The terminology seems to find its origin in a comment made by Lord Hobhouse in the
original Reynolds case, which was not picked on at the time: ‘The publisher must show that the
publication was in the public interest and he does not do this merely by showing that the subject
matter was in the public interest’ (Reynolds [2001] 2 AC 127 , 239).

9
 University of Edinburgh School of Law Research Paper 2014/45

Flood litigation, which was handed down in early 2012,38 half way through the parliamentary
procedure.
Flood concerned an as-yet not fully-fledged defence which is typically referred to as
‘reportage’, and whose relationship with responsible publication is still in the process of
being worked out. In order to understand the ‘contamination’ of the new statutory defence by
this parallel doctrine, we need to go back a couple of centuries in time. In the early 19th
century, for reasons which need not be examined here, English law settled on the proposition
that passing on a defamatory statement amounted to making it oneself, even when the
defendant distanced himself from it: as far as the law was concerned, the propositions ‘Peter
is a thief’, ‘I heard [from John] that Peter is a thief’ and ‘I doubt that Peter is the thief John
makes him out to be’ were to be treated as equivalent: in particular, in order to avail himself
of the defence of justification, the defendant would have to prove the truth of the ‘sting’
allegedly made by the statement, namely, that Peter really was a thief (not that he had heard it
or that he had good reasons to believe it). This became known as the ‘rule of repetition’, still
in force today.39
When one thinks about it, the rule of repetition is untenable, at least when phrased so broadly.
It is one thing to say that the defendant should not have complete licence to spread
defamatory statements simply because they did not originate with him and their truth value
was not endorsed; but common morality suggests very strongly that it cannot be invariably
wrongful – not even prima facie – to repeat incriminating statements that one has not made
one’s own. Naturally such statements could always be protected indirectly, for instance
through the doctrine of qualified privilege, but – surprisingly enough – it is only recently that
English law has come to the explicit realisation that the rule of repetition itself is too broad. It
has done so obliquely, by carving out a defence known as ‘reportage’. The details are not
important here 40 but, essentially, the defence protects those who have ‘reported’ (that is,
repeated) incriminations made by others, even to the world at large, provided the repetition
was itself in the public interest, regardless of the truth value or even plausibility of the
allegations.41
As can be seen, the rationale of reportage is entirely different from – indeed, incompatible
with – the ‘reasonable truthfulness’ which underpins the judge-made (Lange/Reynolds/Grant)
defence of responsible publication, where ‘responsible’ means ‘based on a reasonable belief
in the truth of the matter’.42 But, because English courts are averse to plucking innovations
out of thin air – part of the common law mentality is that judicial development must be seen
to be only one step forward from the previous state of affairs – reportage could not be
recognised straight away for what it is. Rather, it had to ‘piggyback’ on a pre-existing

38 Flood v Times Newspapers Ltd [2012] United Kingdom Supreme Court (UKSC) 11, [2012] AC
273.
39 See Mitchell (fn 29) 123 ff for the historical background.
40 For a full account, see Gatley (fn 1) §§ 15.15–15.18. There have long existed specific reporting
privileges, whether statutory or at common law, but these were tied to particular fact scenarios.
41 Al-Fagih v HH Saudi Research & Marketing (UK) Ltd [2001] EWCA Civ 1634, eg [49], [52];
Roberts v Gable [2007] EWCA Civ 721, [2008] Queen’s Bench (QB) 502, especially [61]; Flood [2012] AC
273, [122] ff.
42 It is possible to say that reportage is a form of responsible journalism in a broader sense of
being ‘non-culpable’. The present writer does not mean to suggest that reasonable belief in truth is the
only basis on which a publisher-at-large can defend its conduct. Being warranted in passing on a
defamatory statement is another such basis (although arguably what this really calls for is a
reformulation of the rule of repetition). The one point that matters is to keep the two strands distinct;
labels are comparatively unimportant as long as they remain stable.

10
 University of Edinburgh School of Law Research Paper 2014/45

doctrine which, in that instance, could only be the Reynolds privilege: just as Reynolds, a
decade earlier, had grown out of qualified privilege, despite clearly being a different
jurisprudential creature (which was recognised some years later), so reportage was – and
continues to be43 – regarded as a species of the Reynolds privilege, even though it is entirely
unconcerned with issues of reasonable belief in truth on a matter of public interest. Indeed,
the Defamation Act 2013 perpetuates this confusion, codifying it as a subspecies of the new
statutory defence abolishing and replacing Reynolds:
4 Publication on matter of public interest
…
(3) If the statement complained of was, or formed part of, an accurate and impartial account of a
dispute to which the claimant was a party, the court must in determining whether it was
reasonable for the defendant to believe that publishing the statement was in the public interest
disregard any omission of the defendant to take steps to verify the truth of the imputation
conveyed by it…
(6) The common law defence known as the Reynolds defence is abolished.
Through a most unfortunate turn of events, just as the Bill was undergoing examination, the
Supreme Court of the United Kingdom handed down its judgment in the case of Flood v
Times Newspapers. Flood concerned an action brought by a police officer against a
newspaper for publishing the statement that Mr Flood was under investigation following
allegations of corruption. The Times took no stance as to the guilt or otherwise of the
claimant:44 this was a pure case of ‘reporting’ (repeating) an incrimination rather than making
one but, as explained, it was litigated under the doctrine of Reynolds. Mechanically this had
the effect of steering the court away from issues of verification (which are, by construction,
irrelevant); yet, because the decision failed to recognise the distinct nature of reportage, its
language had to be sufficiently open-ended to apply across the breadth of responsible
publication. The result was that the court resorted to the – incredibly vague – language of
‘public interest’ as the one overarching principle under which the reasonable truthfulness of
Reynolds could be subsumed as a mere species of the genus of bonus scriptor ephemeridis.45
Understandably eager to take into account the latest judicial developments, but lacking the
time to digest their significance, Parliament simply took this open-textured language and
engraved it in the marble of sec 4 of the new Act, which applies to the entire defence of
‘publication on matter of public interest’. Time and again the preparatory works emphasise
that the statutory defence does not aim to alter the law but, simply, to restate it in its ‘most
recent’ – that is, post-Flood – version, which itself displays no intention to alter the substance
of the defence outside of reportage cases.46

D Whither ‘responsible journalism’ after the Defamation Act 2013?

Where does this leave us? Courts will not be able to change the wording of the Act but of
course they will have to interpret it. As mentioned the statutory requirement that the

43 See eg Gatley (fn 1) § 15.15; contrast J Bosland, Republication of Defamation under the
Doctrine of Reportage (2011) 31 OJLS 89, 101 ff.
44 In the end Mr Flood was cleared of the charge. That allegations had been made was of course
true but, for the reasons explained above, this was immaterial.
45 Flood [2012] AC 273, especially [54], [60], [68], [113], [123], [158]. The analysis of the language
used by the court would deserve a paper of its own.
46 Price/McMahon (fn 2) §§ 5.28, 5.37 ff (especially 5.46 f). It is true that other remarks suggest a
desire to ‘strengthen freedom of speech’ (at § 5.45) but this can be reconciled with the status quo by
interpreting it as moving towards the removal of the 10-pronged test, which itself had been
misapplied by lower courts: the intention was therefore to recapture the essence of the Reynolds
privilege – that is, not to change the law but to remove a misinterpretation.

11
 University of Edinburgh School of Law Research Paper 2014/45

publication must have been (reasonably believed to be) in the public interest is under-
determined. It must be hoped that, faced with a range of possible interpretations, the courts
will return to the understanding of Reynolds as it existed prior to the Act – that is, reasonable
belief in the truth of a matter in the public interest – supplemented by a separate defence of
reportage (or, more broadly, of warranted republication). The latter urgently needs to be
severed from mainstream Reynolds if it is to stop causing interference. Despite the baffling
removal of the word ‘responsible’ in the heading of the defence (in its final form,
‘publication on matter of public interest’ tout court), this is not only a plausible interpretation
of the black letter of the law, but the only one to give effect to parliamentary intent. Courts
must not allow the process of identification and interpretation of the negligence-culpa
standard that underpins the defence, which started with Reynolds, to be derailed.47

IV Honest opinion
Turning to the third point under consideration, sec 3 of the Act begins as follows:
3 Honest opinion
(1) It is a defence to an action for defamation for the defendant to show that the following
conditions are met.
(2) The first condition is that the statement complained of was a statement of opinion.
(3) The second condition is that the statement complained of indicated, whether in general or
specific terms, the basis of the opinion.
(4) The third condition is that an honest person could have held the opinion on the basis of—
(a) any fact which existed at the time the statement complained of was published;
(b) anything asserted to be a fact in a privileged statement published before the statement
complained of.
(5) The defence is defeated if the claimant shows that the defendant did not hold the opinion.
Readers familiar with the English law of defamation will recognise a new version of the
common law defence known, for most of its history, as ‘fair comment’.48 The doctrine of fair
comment is rarely given the attention that it deserves, but it is possibly the least
straightforward of all the defences recognised by English law. In fact, the difficulties that
English law has observably experienced, both in pinning down the specific requirements of
the defence and in stabilising them, evidence a lack of clarity about the purposes the defence
is intended to serve. In this writer’s mind, the Defamation Act 2013 constitutes a further
illustration of this equivocation and represents a missed opportunity to take what would be
the most coherent – if radical – course of action, namely, to get rid altogether of a defence
that, as will be explained, is really rooted in the idea that comment cannot be defamatory in
the first place: fair comment, correctly understood, is not so much about justifying the injury
to the claimant’s reputation as it is about denying that it has occurred.

A Opinion and comment

The very hesitancy concerning the name of the defence is a strong indication of the fact that
English law has two different rationales in mind. Leaving aside for now the qualifier
‘honest’, the new name is ‘opinion’. While the concept of opinion is not a straightforward
one, it seems fair to say that we ordinarily expound it by contrasting it with fact, on the basis
that facts are falsifiable whereas opinions are not. With the emphasis being squarely on the
possibility of proving truth, we might move the line slightly and treat as opinion assertions
which are not practically verifiable even though they could in theory be proved or disproved,

47 While some early commentators were led astray (KY Low, Reynolds Privilege Transformed
(2014) 130 Law Quarterly Review 24), it is reassuring to see the leading practitioners’ works making
this interpretation their own (Price/McMahon (fn 2) § 5.56; Gatley (fn 1) § 15.4).
48 For the historical background see Mitchell (fn 29) ch 8.

12
 University of Edinburgh School of Law Research Paper 2014/45

a prime example being the ascription of a mental state (such as an intention) to the claimant.
The dividing line thus appears to be between what we agree to disagree on and what we do
not: for the latter, truth (justification) is the relevant defence; for the former, fair comment.49
The drawing of such a parallel between truth (veritas) and fair comment (quasi ex veritatis)
has an old pedigree. 50 But if we accept it, several features of the latter defence become
puzzling. One such feature is that, whilst English law has accepted, since the 16th century,
that the truth of an incriminating assertion of fact justifies it, no matter how gratuitous it
might be,51 prior to the 2013 Act fair comment only availed for statements on a matter of
public interest. In the same way, fair comment was – and still is – defeated by malice
(understood as the fact that the defendant did not really hold his alleged view), whereas truth
has always justified the statement, no matter how spiteful. 52 This is reminiscent not of
justification but of qualified privilege, which protects the assertion of non-truths provided
they are made in good faith and in furtherance of a good purpose. The doctrine of fair
comment can thus be seen to straddle two conflicting logics; arguably its refusal to choose is
the principal cause of its being, in the words of Lord Phillips, ‘one of the most difficult areas
of the law of defamation’53 – a doctrine that is extremely difficult to make sense of from the
outside because it is not coherent inside.
What makes things even more complicated is that it is not even clear that the defence of fair
comment in English law really is about opinion. Or rather, while this has become a popular
view in the more recent past – a process culminating in the statutory change of name – this
was emphatically not the case in earlier days. Comment has historically never been about
opinion: it was, as the name states, about commenting (etymologically, ‘devising by careful
thought’). Essentially a comment is a reasoning: in the words of the leading authority, the
Australian case of Clarke v Norton, ‘something which is or can reasonably be inferred to be a
deduction, inference, conclusion, criticism, remark, observation, etc.’54 It is something that
can be seen to originate in the defendant’s mind, as opposed to being the assertion of a state
of affairs about the world to be regarded as independent from the person who makes it. An
opinion, in the above sense, will by construction be comment; but comment is wider. In
particular, it has always extended to what English law describes as inferences of fact, that is,
(secondary) facts derived by the defendant from earlier (primary) facts.55 To give an example,
for the defendant to assert, without further ado, that ‘Oliphant is a thief’ is a statement of fact
that must go to justification. But if instead he asserts some primary facts (for example, ‘my
laptop disappeared’; ‘only Oliphant was around when this happened’; ‘he blushed when I
next saw him’) and then goes on to assert that ‘therefore Oliphant is a thief’, the latter will
now count as comment (despite being evidently falsifiable): it is an inference of fact made by
the defendant, based on facts disclosed by him as the basis of the comment. Despite the
change of name, this doctrine has been retained in the new Act,56 which demonstrates that the
law has not really switched over to a different paradigm: it might have adopted a new label,

49 This is the dominant idea de lege lata, but English law has not committed itself to one
particular dividing line; for a survey of authorities see Gatley (fn 1) §§ 12.7–12.14.
50 Mitchell (fn 29) 169, 185 f; see also Gatley (fn 1) §§ 12.5.
51 E Descheemaeker, ‘Veritas non est defamatio’? Truth as a Defence in the Law of Defamation
(2011) 31 Legal Studies 1, 5 f.
52 There is one small statutory exception concerning the disclosure of spent convictions:
Rehabilitation of Offenders Act 1974, sec 8(5).
53 Joseph v Spiller [2010] UKSC 53, [2011] 1 AC 852, [1].
54 [1910] Victorian Law Reports (VLR) 494, 499.
55 Gatley (fn 1) § 12.8.
56 Price/McMahon (fn 2) § 4.42.

13
 University of Edinburgh School of Law Research Paper 2014/45

which was thought to be more attractive, but when it refers to ‘opinion’ English law still
means ‘comment’.

B Authoritativeness
What the above suggests is that the division between fact and opinion, or falsifiable and
unfalsifiable statements, is a red herring. While superficially attractive it does not in fact
reflect the reality, and therefore the underlying logic, of the law. To the present writer’s mind,
this logic – which surfaces in many judicial authorities but has never gained the upper hand57
– is to distinguish between statements which are meant to be authoritative and statements
which are not. When the defendant makes an assertion of a primary fact about the world –
that is, one that is not ‘devised by thought’ on the basis of others – then (leaving aside any
possible qualifications about its truth value, which the law discounts) he intends his word to
be taken for it: for this reason the law holds him prima facie liable when the result is to cause
others to think less well of the claimant. But when he utters something that can be recognised
as a comment, he does not intend the third party to believe the statement by virtue solely of
its having been made: rather, he invites the recipient to agree with him (‘yes, the only
possible explanation is that Oliphant did it’); but the recipient can just as well disagree (‘no,
this is too rash a conclusion’).
Read in the light of this distinction, a requirement starts to make sense which otherwise
would not, namely, the necessity for the comment to have a basis in fact which should also be
disclosed to (or else sufficiently known or knowable by) the third party – a requirement
which was never pinned down with precision but has been retained by the 2013 Act.58 If the
basis of the defence is that opinions cannot be verified and thus must be allowed as long as
they are honest (that is, quasi-true), there is absolutely no reason why they should disclose
their factual basis or, indeed, be grounded in one: if anything, this would seem to undermine
the logic of the defence. But, in reality, the rationale is that it is this disclosure which
identifies the statement as comment (unless it is self-evidently so because the facts are
already ‘out there’).
This is entirely intelligible: on the basis of facts which are true (or protected in their potential
untruth) I am at liberty to draw whatever inference – of opinion or further facts – that I wish;
on the other hand, if the facts are incorrect and not otherwise protected, the law holds me
liable – not because of the inference but because of the untruth of the underlying facts which
I represented as true. There is in fact a third possibility, which represents the best test case for
the above theory, namely the doctrine of ‘bare comment’: 59 if no facts are disclosed, or
otherwise known, the defence of fair comment will not avail even if the statement is clearly
recognisable as comment; this is because my inference can be taken as implicitly asserting
primary facts that are sufficient for me to draw such an inference, and these will not be
covered by a defence of ‘free reasoning’. Difficult though fair comment is, the elements of
the defence start to fall into place when we recognise that it is not primarily (if at all) about
‘unfalsifiability’; it is about authoritativeness.

C Lingering malice
There remain two features of the defence (one now abolished) which are difficult to make
sense of within any framework. The first is the continuing existence of a replicatio – that is,
an affirmative answer to the defence – of malice: the comment has to be ‘fair’ or ‘honest’ if it
is to escape liability. Unless one interprets honesty as being the truth equivalent of statements

57 Mitchell (fn 29) 186–190 and cases cited; Fagan (fn 26) 188–191 and cases cited.
58 Gatley (fn 1) § 12.24 f.
59 Gatley (fn 1) § 12.9.

14
 University of Edinburgh School of Law Research Paper 2014/45

incapable of proof (quasi veritas), this would seem to be no more than a hangover from the
time when malice was regarded as the gist of the wrong of defamation. But, while it is
demonstrably true that the defence of comment historically emerged – like qualified privilege
– from the ossification of the rebuttal of malice, 60 hardly anyone today would argue that
defamation is or should be grounded in malice: in that sense, the requirement is a legal
hysteresis, the perpetuation of an effect after its cause has disappeared. This does not mean
that the doctrine of comment cannot be justified on another basis, but the relevant question is
whether that other justification allows room for a condition of non-malice.61
The second feature is the requirement of public interest (which originally would have been
part of the rebuttal of malice: defamatory comment would be regarded as spiteful if there was
no ‘good reason’ to comment in the first place). In a legal system such as South Africa (or
New South Wales until the reforms of 2005), where only truth for the public benefit justifies,
the requirement would have sat well with that other defence; but not so in England. Moving
from the logic of qualified privilege to that of truth, the Defamation Act 2013 finally
abolished the requirement of public interest; but little was provided by way of an explanation
besides the assertion that it represented an ‘unnecessary complication’.62

D Abolition?
As appears from the above developments, the 2013 Act continues the process of the gradual
expansion of the defence of comment or opinion, together with its automatic corollary, the
diminution of the scope of liability-triggering speech. While this is all good news from the
perspective of freedom of expression, it seems to the present writer that an opportunity was
missed to re-think the doctrine again from first principles. On that basis, it would seem that
the best thing that could happen would be for it to disappear as an autonomous defence. The
argument for abolition would deserve fuller consideration than space permits, but the gist can
easily be stated: if comment is, by definition, a statement which can be recognised as non-
authoritative, then it follows that no ‘right-thinking member of society’ should change their
mind in reliance on it.
The emphasis here is on ‘should’: it would be disingenuous to argue that no one in fact ever
changes their mind on the basis of comment. (This is especially true when it emanates from
people in a position of authority or holding recognised expertise.) Rather, the point is that
they ought not to, the basis of the prohibition being to protect individuals’ absolute freedom
to reason. In other words, unless perhaps I can be regarded as having assumed responsibility
for the consequences, my ability to apply my mind to facts so as to derive new facts or
opinions ought not to be constrained by what others will do with these new facts or opinions:
what they do with them is their responsibility not mine.63 When, expressing myself, I say or
imply ‘take my word for it’ (expression of primary facts), the law rightly holds me
accountable for the effects of my statement on others; but if the statement is of such a nature
that it invites others to agree with me (‘that’s my take on it’ = comment), then I cannot be
held responsible for others’ agreement. The logical implication is that comment, deemed
incapable of causing others to think less of the claimant, cannot by construction cause an
injury to his reputation, that is, be defamatory.
Two final points conclude this section. The first is that, although the above proposition
cannot be said to reflect English law as it stands (for, even if it is true that it already

60 Mitchell (fn 29) 170.

61 See fn 63 below.
62 Cited in Gatley (fn 1) § 12.36.
63 Note that, on this understanding, the requirement of malice could be reintegrated on the basis
that reasoning is only worth protecting when it is genuine (ie believed by the person making it).

15
 University of Edinburgh School of Law Research Paper 2014/45

recognises, behind the surface, comment as being a non-authoritative statement, there is one
further – normative – step to take before reaching the conclusion that it is incapable of being
defamatory), it is nonetheless derived from English law’s own logic. The reason is this: what
other ground could the law have to distinguish between the assertion of primary facts, on the
one hand, and the expression of one’s reasoning, on the other? The fact that it has done so for
centuries cries out for an explanation, even if one imperfectly articulated by the law: to the
present writer’s mind, this can only be the fact that the third party is justified in relying on the
latter in a way that it is not with regards to the former: otherwise the law would treat both
types of statements in a similar way and would not have developed separate defences in the
first place. But then it is difficult to see how the law could treat the justifiability of reliance as
a matter of degree rather than an all-or-nothing decision. Thus it would appear that English
law really has been groping for the proposition that comment should not be relied on; it
should now carry forward that idea by deeming that, in law, a comment is incapable of
injuring another’s reputation.
The second remark is that the term ‘abolition’ is used in a loose sense. Logically abolishing
fair comment would mean that comment is always actionable rather than never. As explained,
what is really meant is that comment ought not to be regarded as defamatory in the first
place, which, if accepted, would translate into the definition of the cause of action itself. On
that reading, only an assertion of fact – in the sense of a primary fact, that is, not one derived
from others through one’s own reasoning abilities – would be capable of being defamatory.
This would move English law very much closer to a system like French law, which
consistently requires the imputation of a ‘precise fact’ (fait précis) for a statement to qualify
as defamatory. The rationale behind the French position is that only a precise fact is capable
of being proved, which in turn indicates that an unfalsifiable proposition cannot be
defamatory: something very much like the modern understanding of fact (as opposed to
opinion) in English law.
The counterpoint of French law helps us see that there is something rather extraordinary –
and arguably wrong – in English law inasmuch as it follows a two-stage enquiry whereby
defamatoriness is first defined in an extremely broad way (anything liable to cause third
parties to think less well of the claimant), only to almost completely eliminate all that is not
an imputation of facts at the defences stage. What this calls for is a redefinition of what
counts as defamatory; on the above analysis a necessary requirement would be that the
statement should be recognised as authoritative on the defendant’s part. The two stages
would thus be merged into one.

V Conclusion
That the English law of defamation is in need of in-depth re-examination and reform is not
something with which those who are familiar with it will disagree. But regardless of exactly
what reform might have been called for, the Defamation Act 2013 does not deliver it. The
new Act has failed to sort out the existing difficulties and, indeed, has introduced new ones.
Three principal examples were singled out in the present article – the requirement of serious
harm and the revamped versions of both responsible publication and fair comment/honest
opinion – but several others could have been mentioned: for instance, the new power to order
publication of the court’s judgment (sec 12); the failure to abolish trial by jury altogether
(sec 11); or the retention of an absurdly broad definition of ‘publisher’ (sec 10). What
Parliament appears to have done is to take the law as it found it and to attempt to insert a little
more protection of free speech as and where it could find an ‘opening’.
Regardless of whether or not Parliament succeeded in doing this, it is clear that it has brought
yet more complexity and uncertainty into an area of the law that it is already a full-time
occupation to try to master. Given this, one can safely predict that another Defamation Act

16
 University of Edinburgh School of Law Research Paper 2014/45

will follow in the not too distant future, followed by another and another. What defamation
needs is not yet more patches which, at best, relocate problems, but for the law – which
emerged through hundreds of years of decentralised decision-making, only to be tweaked
from time to time by statutory reforms – to be thought through again from first principles.
While judges cannot be blamed for working in an incremental fashion, because this is how
the constitutional consensus understands their role, more could and should be expected from
a Parliament that takes great pride in its ability to do anything it wishes short of turning a
man into a woman (though even that is no longer clear). The 2013 Act is a missed
opportunity.

17