Considering Causation in Criminal Law
| Author | Erik Witjens |
| Published date | 01 April 2014 |
| Date | 01 April 2014 |
| DOI | http://doi.org/10.1350/jcla.2014.78.2.909 |
| Subject Matter | Article |
Criminal Law
Erik Witjens*
‘I think it was the wind’, said Piglet. ‘I think
your house has blown down’.
‘Oh is that it? I thought it was Pooh’.
‘No’, said Pooh.
‘If it was the wind’, said Owl, considering
the matter, ‘then it wasn’t Pooh’s fault.
No blame can be attached to him’.
A. A. Milne, The House at Pooh Corner, p. 137
Abstract According to the dominant view, causation in criminal law is a
bifurcated test consisting of cause in fact and proximate causation. In the
first section of this article the theoretical underpinnings of causation are
explored, for instance concerning the underlying interrogations of causal
questions in the law. In the second section, the traditional view on bifurcation
is challenged, weaknesses are uncovered, and the shortcomings of counter
factual tests as a heuristic test for factual or ‘empirical’ causation are assessed.
By reviewing R v Williams and R v Hughes, the third section of the article
seeks to elucidate the nature of the causal requirement in law. It is suggested,
that legal influences permeate the causal requirement in law to the extent
that it is dominated by them. The article consequently concludes that the
bifurcation needs to be rejected to better reflect the role of (empirical)
causation in criminal law better.
Keywords Cause in fact; Foreseeability; Proximate causation; Sine
qua non test; But for test
Causation in criminal law is something of an enigma. The precise scope of
this notion has been subject to scholarly attention for many years, yet
continues to elude precise definition. As I will put forward in this article,
the difficulty stems from the hybrid background of the notion of causation
in criminal law.
My argument will be that in causation in criminal law an empirical
component, which caters for empirical notions of causation, is married to
a legal component, that dominates the empirical component in the sense
that the legal perspective of apportioning blame, of assigning responsibility
to the act of a specific perpetrator, ultimately dictates the role of the
empirical component, for instance as regards the level of certainty that is
required about the empirical causal chain for the purpose of criminal
liability.
* Associate Professor of Criminal Law at the University of Aruba; email: erik.witjens@
ua.aw.
164
The Journal of Criminal Law (2014) 78 JCL 164–183
doi:10.1350/jcla.2014.78.2.909
Considering Causation in Criminal Law
This article consists of four sections. The first deals with the ideas of
influential scholars that have sought in recent years to elucidate causal
reasoning in (criminal) law. In the second section those ideas are contrasted
with my own ideas on aspects of causation that are at the root of the
difficulties that causation poses for theorists. In doing so, common ground
between the different approaches will become apparent. The third section
consists of a review of recent case law that illustrates the findings from the
previous two sections and ultimately shows that a different approach
might be in order. Finally, in the fourth section, I draw conclusions and
put forward some additional thoughts that might further discussion on
this subject.
1. Three views on causation
This section deals with three different approaches to causation in the law.
My intention is to give a short outline on each approach, focused on
specific aspects that I will consider later in the second section. As this is the
case, a more comprehensive description in the scope of this article seems
redundant.
(a) Moore’s causation and responsibility
In this tome on causation and responsibility, Moore exhaustively explores
many issues that are, in his view, prerequisite to an understanding of
causation in the law.1 Moore believes that cause is univocal, which leads
him to assert a metaphysical notion of cause should be applied in the law.2
The sheer volume of his observations effectively prohibits me from going
into all of the many interesting points he raises. However, a few of his
observations are especially relevant for the subject of this article and
consequently I will turn my attention to those.3
It is widely accepted that the causation requirement in law actually
consists of two components; the first being ‘cause in fact’—purportedly a
‘scientific’ notion of causation—and the second being ‘proximate cause’,
to be resolved by arguments of policy. The dominant test for cause in fact
is the ‘sine qua non’ or ‘but for’ test, a counterfactual approach to causa
tion.4 Moore convincingly illustrates the many problems that are inherent
to counterfactual tests of causation such as sine qua non.5 The unfettered
scope, rendering many irrelevant factors causal, is a prime example. When
a vase falls off a table and shatters, the fact that it was not made of unbreak
able material can be considered a sine qua non for the vase shattering.
Another type of problem concerns the trouble the sine qua non test faces
when it tries to make sense of overdetermination cases. Concurrent cause
cases are wellknown examples of this problem: if two shots are fired
1 M. S. Moore, Causation and Responsibility (Oxford University Press: Oxford, 2009).
2 Ibid. at 5.
3 For instance, I will not address some of the weaker points in his argument. For a more
indepth analysis of his book, see amongst others L. Alexander and K. K. Ferzan,
‘“Moore or less” Causation and Responsibility’ (2012) 6 Criminal Law & Philosophy 81.
4 The alleged result would not have occurred, or would not have occurred at the time or
in the way it did, ‘but for’ the defendant’s act or culpable omission.
5 Moore, above n. 1 at 84–6, 371–470.
165
The Journal of Criminal Law
simultaneously, each of which is capable of killing the victim, counter
factually speaking neither shot was necessary for the result, seeing as it
can be said for each of them that the other shot was sufficient.
Moore subsequently explores the rationale behind different tests of
proximate causation and finds some are predominated by the flaws in
counterfactual reasoning and do not really add anything to the equation.6
On the foreseeability test, he is of the opinion that as it incorporates moral
blameworthiness, it can be considered superfluous on account of culpability
performing the same task. This brings him to ‘prune’ away this theory
from his later analysis. In Section 2 below, I demonstrate that this might
have been an oversight. Moore is slightly more positive on the ‘harm
withintherisk’ test, but is again of the opinion that this grading of
culpable mental states should not be taking place under the banner of
causation.7 His conviction that causation should be a metaphysical concept
is at odds with legal causation as being guided by interests and policies.
Moreover, taking into account the problems that are associated with
counterfactual causation, he is drawn to look into other possibilities than
the common bifurcated approach.8
In doing away with counterfactual theories on causation and opposing
proximate cause as a legal construct that should not be allowed to play a
role in apportioning blame, Moore then has to find a theory that might be
feasible as a metaphysical concept of causation in the law.9 It should be
noted that Moore is of the opinion that causation is scalar in nature.10 In
other words, something can be more or less of a cause for a given result.
This, however, is contradictory with his metaphysical starting point, for
Moore seemingly derives this attribute from the law.11 In the final two
chapters of his book he looks into two possible ways to advance: generalist
and singularist theories of causation. A generalist theory seeks to make
statements of causal generalisation—‘sparks cause fires’—and thereby
identifies a causal relation by means of asserting the relation is an
instantiation of said law.12 However, Moore argues that the generalist
theories are overly broad and therefore not feasible.13
This leaves the singularist theories. Moderate singularism holds that
while some law lies behind every singular causal relation (for instance,
behind every spark that does cause a fire), the existence of such a law is
not sufficient for the truth of a singular occurrence. In other words, not
every spark will lead to a fire. Although Moore remains vague about his
precise preferences, he does profess to favouring a singularist approach as
such. However, perhaps we can deduce a preference from the appendix to
the book, about contract law and causation.14 Moore writes:
6 Moore identifies ‘ad hoc’ and ‘rulebased’ policy tests, above n. 1 at 96–7.
7 Moore, above n. 1 at 100–2 and 178–228.
8 Moore, above n. 1 at 95 and 104–6.
9 Which would then, by all accounts, be a unified theory. The bar for such a theory in a
sense would be higher, for it cannot use proximate cause as a catch all for unwanted
results; see Moore, above n. 1 at 104.
10 Moore, above n. 1 at 118–21, 153 and 508.
11 Ibid. at 118: ‘In various places, the law assumes that the causal relation is in the latter
category [i.e. continuous variation —EMW]’.
12 In more detail, see Moore, above n. 1 at 471–2.
13 Moore, above n. 1 at 475–95.
14 Ibid. at 513.
166
Considering Causation in Criminal Law
If causation is as this book describes it, it is a scalar relation that admits of
degrees and it is so understood in popular understanding … There is an objection
that should be dealt with here. This is the objection that has been voiced in legal
circles...
To continue reading
Request your trial