NOTES / AANTEKENINGE THE DEFENCE OF PROVOCATION – WHERE ARE WE NOW?

The legal development towards treating provocation and emotional stress as factors that may create reasonable doubt as to the presence of criminal capacity is not only novel but revolutionary (Burchell Principles of Criminal Law 5ed (2014) 328). Essentially, the law has recognised that a person who causes the death of another can be acquitted of murder, despite suffering no mental illness or defect at the time of killing, if evidence points to provocation and/or severe emotional stress at the time of the commission of the killing leading to a loss of criminal capacity (Carstens and Le Roux “The Defence of Non-Pathological Incapacity with Reference to the Battered Wife Who Kills Her Husband” 2000 SACJ 180).


OBITER 2019
Before its introduction, South African courts looked at provocation through the lens of mens reathat is, they saw the possibility of provocation negating the element of intention (Hoctor "A Peregrination Through the Law of Provocation" in Joubert (ed) Essays in Honour of CR Snyman (2008) 110 118). The notion of criminal capacity came to the fore in South African criminal law when it was the subject of investigation by the Rumpff Commission in the wake of the assassination of Hendrik Verwoerd by Demetrio Tsafendas, who was held not to be capable of standing trial by reason of mental illness (Rumpff Report par 1.9). The Rumpff Report informed the drafting of the test for criminal capacity set out in section 78 of the Criminal Procedure Act (51 of 1977). While this test was formulated in the context of establishing possible pathological incapacity, the test came to be employed in respect of the test for non-pathological incapacity. In this regard, the landmark case of S v Chretien (1981 (1)  This development occurred naturally and inevitably. Once it was held in Chretien that voluntary intoxication could exclude criminal liability on the basis of rendering the accused's conduct involuntary, or by negating the accused's criminal capacity, or by excluding the accused's intention (although intoxication does not negate negligence, as the reasonable person may drink, but never gets drunk), the expansion of the defence of incapacity was entirely foreseeable. After all, in terms of the prevailing psychological theory of criminal liability, wherever one (or more) of the elements of liability (identified above) was excluded as a result of some external factor, the accused perforce must be entitled to a defence (the term "external factor" is used to distinguish this type of defence from an analogous defence based on mental illness). While this approach is anathema to those who would prefer that policy considerations should be decisive in determining which factors should be entitled to provide the basis for a defence excluding criminal liability, there are, arguably, strong and principled arguments for supporting the approach adopted in Chretien. It is surely an overriding concern that no one who genuinely lacks one or more of the elements of criminal liability, on whatever basis, should be regarded as blameworthy, and subjected to punishment. How could this not be an egregious infringement of the accused's rights?
On this fertile theoretical soil, the defence of non-pathological incapacity soon expanded beyond the basis of voluntary intoxication. Thus, the Appellate Division in S v Bailey (1982 (3) SA 772 (A) 796C-D) noted that fear could exclude capacity. Shortly thereafter, the same court in S v Van Vuuren (1983 (2) SA 12 (A) 17G-H) held that a combination of intoxication and provocation (or severe emotional stress) could be the basis for successful reliance on the defence of non-pathological incapacity (authors' own emphasis). Indeed, a theoretical framework for the defence was carefully formulated by the Appellate Division in S v Laubscher (1988 (2) SA 163 (A) 166G-167A), along with the classic two-stage test for the defence of non-pathological incapacity, which is: (1) the accused's ability to distinguish between the wrongfulness or otherwise of his or her conduct (cognitive capacity), and (2) the capacity to act in accordance with such an appreciation. This development of the law by the then-highest court in the land culminated in a successful reliance on this defence in S v Wiid (1990 (2) SACR 561 (A)), in which, following a humiliating and traumatic assault by the deceased (the accused's husband), she had shot him dead.

The case of Eadie
While the formulation of the defence in Laubscher (supra) remains the classic statement on the matter to this day, despite the well-established nature of the defence, the Supreme Court of Appeal sought to redefine the defence in Eadie. The case was originally heard in the Provincial Division of the Cape High Court, where the accused pleaded not guilty to murder and defeating/obstructing the course of justice (reported at S v Eadie (1) 2001 (2) SACR 172 (C)) and unsuccessfully raised the defence of non-pathological incapacity primarily due to provocation and road rage.
Snyman has stated that the judgment is "one of the most enigmatic judgments of the Supreme Court of Appeal in the field of the general principles of criminal law during the past half century" (Snyman Criminal Law 6ed (2014) 161). It is unfortunate that the judgment may be considered as a "good example of a correct decision arrived at for the wrong reasons" (Snyman "The Tension between Legal Theory and Policy Considerations in the General Principles of Criminal Law" 2003 Acta Juridica 1 14).
While the confirmation of the appellant's conviction was not in doubt, the court (per Navsa JA) saw fit to engage in an analysis of the defence of nonpathological incapacity in the context of provocation. Much has been written about this judgment, and given that the judgment is indeed extraordinary in many respects, all the attention is warranted (see, inter alia, Snyman Unfortunately, much of the commentary on the judgment has been rather less than complimentary, again, for good reason. Two notable features of the judgment are the scant attention paid to the development of the defence of non-pathological incapacity, and the rather selective (and sometimes inaccurate) citation of previous authority (Hoctor in Joubert (ed) Essays 135-138). Nevertheless, the judgment from the outset seemed to be less about a careful examination of the law, with the court describing its quest (par 3) as being to investigate whether "[t]he boundaries of the defence … have been inappropriately extended, particularly in decisions of Provincial or Local

OBITER 2019
Divisions of the High Court, so as to negatively affect public confidence in the administration of justice". If the court had in fact confined itself to the terms of this analysis, much confusion and academic gnashing of teeth could have been avoided. Instead, it drew (par 56) on the writing of Louw ("S v Eadie: Road Rage, Incapacity and Legal Confusion" 2001 SACJ 206 210-211) where it is stated: ogic dictates that we cannot draw a distinction between automatism and lack of self-control … if the two were distinct it would be possible to exercise conscious control over one's actions (the automatism test) while simultaneously lacking self-control (the incapacity test)".
As Snyman has cogently pointed out (Criminal Law 162), the law allows for precisely this possibility in the context of children under 14 years of age, who may be able to perform voluntary acts, but may nevertheless be held by the courts not to incur criminal liability as a result of lack of capacity. Nonetheless, the court found this line of reasoning apposite to its critical assessment of the non-pathological incapacity defence, and duly concluded that "there is no distinction between sane automatism and non-pathological incapacity due to emotional stress and provocation" (par 57).
It follows that it would have to be established that an accused was acting involuntarily in order for her defence of lack of conative capacity to prevail. The court did not shy away from this conclusion (par 57): "It appears logical that when it has been shown that an accused has the ability to appreciate the difference between right and wrong, in order to escape liability, he would have to successfully raise involuntariness as a defence … the two are flip sides of the same coin." While the court's patchy treatment of the existing case law makes it difficult to accept that the defence of non-pathological incapacity based on provocation or emotional stress has been summarily jettisonedcan such a drastic step really occur solely by implication?this seems to be the only conclusion that can be drawn from the unequivocal dicta cited above. Having said this, the court itself provided some contra-indications that this was what it was seeking to achieve. Hence, Navsa JA stated (par 57) that he was "not persuaded that the second leg of the test expounded in Laubscher's case [the test for conative capacity] should fall away", and then (par 59) commented that "[w]hilst it may be difficult to visualize a situation where one retains the ability to distinguish between right and wrong yet lose the ability to control one's actions it appears notionally possible".

Critique of Eadie
There are two major difficulties arising from the Eadie judgment handed down by the Supreme Court of Appeal. The first is undoubtedly the court's conflation of the defence of non-pathological incapacity with the defence of sane automatism (par 57, 70). Ultimately, according to the court, the second leg of the test for criminal capacitythat is, the test for conative capacityand the test to determine if conduct was voluntary are the same. As Snyman cogently points out (Criminal Law 162), these are two completely different concepts: "If the conative leg of the test of criminal incapacity is not complied with, it means that X is indeed able to control his bodily movements by subjecting his muscular contractions to the control of his will or intellect, but that he is unable to resist the temptation to act in a way that differs from what his insights have taught him." Essentially, the crux of the matter is that even though a person performs a voluntary act he may still lack the ability to act in accordance with his appreciation of right and wrong. In terms of established legal principles, the two defences are fundamentally distinct, sane automatism relates to the ability to exercise control over muscular movement of an individual and assesses if the muscular movements were subject to the will of an individual, while the defence of non-pathological incapacity relates to the ability to act in accordance with insight into right and wrong.
Ultimately, a person may still lack capacity even where the conduct is voluntary. Hence, voluntariness of an act cannot and should not be equated with self-control. If it is accepted that there is no difference between the two defences, then the result is that an accused person would have to show involuntary conduct in order for the defence of lack of conative capacity to succeed.
Equating the enquiry into voluntariness with the enquiry into conative capacity is irreconcilable in the light of the same court's statement in Chretien, where it was held that if a person, owing to the effects of intoxication, is not able to differentiate between right and wrong and also does not have the capacity to act in accordance with this appreciation, then the individual does not possess criminal capacity.
While it is true that the defence of non-pathological incapacity owing to provocation and emotional stress has had its controversiesnamely, the acquittals in the cases of S v Arnold (1985 (3) SA 256 (C)), S v Nursingh (supra), and S v Moses (supra)it is the conflation of the defence of nonpathological incapacity with the defence of sane automatism by the court in Eadie that has added the most uncertainty regarding the nature and operation of the defence.
The effects of this approach are not limited just to this defence but have the potential to cause a ripple effect on other areas of law as well. (Snyman Criminal Law 163 argues that the entire defence of non-pathological incapacity has been abolished by implication.) The fusing of two distinct elements of criminal liability unfortunately changes the legal principle; this new development is not limited to cases involving provocation but may extend to cases involving emotional stress and other cases where the presence of criminal capacity is disputed (Burchell Principles of Criminal Law 331). Snyman states that in effect the defence has been abolished, not only in circumstances where provocation has caused the incapacity, but may also extend to cases where incapacity was caused by stress, shock, concussion, panic or fear, since these states are closely tied to emotional stress brought on by provocation and which may be difficult to separate (Criminal Law 164).
The purpose of the test to determine if conduct was voluntary is to assess if the individual was capable of directing muscular movement according to will and intellect (Snyman 2003 Acta Juridica 15). Principally, there is a distinct difference between making a decision and having the ability to execute the decision. As Snyman points out, to treat a defence based on absence of criminal capacity owing to extreme provocation as being identical to a defence based on absence of a voluntary act "contradicts the elementary principles of the construction of criminal liability" (Snyman Criminal Law 162).
The revision of the test for conative capacity is unwelcome, as it will have direct implications for those individuals who have a genuine need for the defence of non-pathological incapacity owing to provocation and emotional stress. The changes brought about by Eadie are especially prejudicial towards the battered woman or any victim of abuse who suffers a loss of self-control and kills her abuser, since a conflation of the concepts of conative capacity and sane automatism requires the accused to prove that her actions were involuntary. It seems that, as a result of this judgment, loss of self-control resulting from provocation and emotional stress is therefore no longer acknowledged. This development is prejudicial to the battered woman especially since the defence of sane automatism is narrower and more difficult to prove. (Burchell Principles of Criminal Law 332 argues that the Eadie judgment should not apply to domestic abuse cases.) The second problematic aspect of Eadie relates to the possible introduction of an objective test into the capacity inquiry. Upon a holistic reading of the Eadie case, it is clear that the court wished to curtail the operation of the defence in the light of the perceived facility of the acquittals in Arnold, Nursingh and Moses.
The introduction of an objective test alters the nature of the defence; the test has refashioned the defence to the extent that it becomes a different defence. This distortion is a direct result of the court in Eadie basing its judgment on policy considerations rather than the established and welldeveloped foundational principles of criminal liability.
If one accepts the dicta in Eadie at face value, the impact of the judgment is extraordinarythat is, the abandonment, by implication, of an innovative yet well-established development in South African criminal law founded on extensive precedent, much of which derives from the SCA (or in its previous guise, the Appellate Division). South African criminal law had eventually thrown off the shackles of an objective approach to intention in R v Nsele (1955 (2)  ). The subjectivisation of criminal capacity, and the development of the defence of non-pathological incapacity (toerekeningsvatbaarheid) followed, founded on the landmark decision of S v Chretien (supra). Thus, at a time when human rights were being violated on a massive scale in apartheid South Africa and criminal law was being misappropriated to ensure political order, the substantive criminal law being developed through the courts upheld the foundational justification of criminal liability leading to punishment: that it is as autonomous moral agents, with an entitlement to freedom of action and the ability to exercise self-determination in their choice of actions, that persons are assessed (see Hoctor "Dignity, Criminal Law and the Bill of Rights" (2004) SALJ 304 309).
It is submitted that the import of an entirely objective enquiry (assessing the voluntariness of an act) into an entirely subjective enquiry thus does violence to the principled and constitutionally sensitive development of the notion of non-pathological incapacity in South African criminal law. Moreover, if we accept that the legitimacy of punishment is founded on responsible moral agency and freedom to choose to do wrong, then any attempt to assess criminal liability without the subjectively assessed capacity to choose how to act is a clear infringement of the right to dignity, which is the constitutional basis of the principle of culpability (see Hoctor in Joubert (ed) Essays 166-167, and sources cited therein).
It may be argued that the introduction of an objective test in the context of criminal capacity is inherently unfair and unjust as it requires a uniform standard of behaviour from individuals in a society that is extremely diverse in terms of education, cultural and racial background. One of the most persuasive arguments against the implementation of an objective test is therefore that it essentially imposes dominant cultural values on others. It seems that the court in Eadie wittingly redefined the defence by adding objective factors to the test for capacity, with the aim of shrinking the boundaries of the defence.
The psychological or principle-based approach to provocation and emotional stress developed by the South African criminal law was a positive development, as it preserves the integrity of the law by applying an approach that is in line with logic and the rights and values enshrined in the Constitution of South Africa. Unfortunately, the well-established principled nature of the defence has been drastically curtailed by Eadie. In light of the problems generated by Eadie, it may be argued that the law must be restored to the two-stage assessment of criminal capacity set out in Laubscher (supra). The principle-based or psychological approach to criminal liability is preferred as it is compatible with the fundamental values of the criminal justice system and is in line with the ethos of the Constitution.

4
What is the current position in relation to the defence of provocation?
In light of the somewhat equivocal statements about whether the defence continues in its present form (as in par 57, cited above), found cheek by jowl with seemingly categorical statements denying such possibility (as in par 56-57, cited above), the question then arises: what is the state of the law on the defence of non-pathological incapacity based on provocation or emotional stress? Has the defence indeed been entirely negated by the Eadie judgment? Is it now merely a historical footnote in the annals of criminal liability, a once-subjective spectral apparition now found making occasional ghostly appearances in the solidly objective automatism defence? Or (despite the apparently plain-speaking dicta to the contrary) has the defence indeed been retained in its original form, and does all that the judgment amounts to is a less-than-helpfully-worded warning about the perils of poor application of the law?
It is plain that the expressed wisdom since Eadie overwhelmingly favours the former interpretation. As indicated above, Snyman (Criminal Law 163) states that the defence of non-pathological incapacity has been abolished. Burchell (Principles of Criminal Law 329) agrees that the effect of the judgment is that provocation could only amount to a defence if it led to involuntary conduct, as do Kemp et al (Criminal Law in South Africa 199,201) and Jordaan ("The Principle of Fair Labelling and the Definition of the Crime of Murder' 2017 TSAR 569). One of the present writers has also made this point in some detail (Hoctor in Joubert (ed) Essays 110ff). It is not only writers that have adopted this interpretation. In High Court judgments in S v Beukes (2003 JDR 0788 (T)), S v Ngobe (2004 JDR 0216 (T)), S v Hughes (2004 JDR 0263 (T)), S v Scholtz (2006 (1) SACR 442 (E)) and S v Marx ([2009] 1 All SA 499 (E)), the courts have dutifully echoed the refrain from the Eadie casethat sane automatism and non-pathological incapacity are one and the same thing, and that the latter defence could only succeed if there were evidence of involuntary conduct.
Is the matter not settled then? Well, not entirely. It appears that the concept of non-pathological incapacity lives on, in its original pre-Eadie form, where it is mentioned in the context of cases discussing diminished capacity (see DPP, Tvl v Venter 2009 (1) SACR 165 (SCA) par 21; S v Mathe 2014 (2) SACR 298 (KZD) par 16), but the concept is also mentioned without comment, or any indication that it no longer exists (or even that it exists in a different form) in cases such as S v Engelbrecht (2005 (2) SACR 41 (W)), S v Volkman (2005 (2) SACR 402 (C)), and S v Longano (2017 (1) SACR 380 (KZP)). None of these cases turned on the content of the notion of nonpathological incapacity, unlike S v Ramdass (2017 (1) SACR 30 (KZD) par 6), in which, despite citing the statement in Eadie (par 57, cited above) conflating sane automatism and conative capacity, the court specifically refused to apply this to cases of intoxication, holding that S v Chretien (supra) still remains the leading authority. Thus it is notable that the concept is apparently still extant. This perception is further strengthened by recent Supreme Court of Appeal cases such as S v Van der Westhuizen (2011 (2) SACR 26 (SCA)) and DPP, Grahamstown v Peli (2018 (2) SACR 1 (SCA)). In the Van der Westhuizen case, in the context of discussing diminished capacity, Cloete JA referred to the defence of "temporary non-pathological criminal incapacity", citing (par 39) the classic Laubscher test in the case of S v Ingram (1995 (1) SACR 1 (A) 4e-g). There is thus no indication in Van der Westhuizen that the Eadie judgment has changed the legal position regarding either the content or availability of this defence. In the Peli case, the court, again discussing diminished capacity, once again (par 9) reiterated the standard test for non-pathological incapacity, this time citing, inter alia, a passage from the Eadie case (notably), where Navsa JA refers to the doctrine on the basis of the standard test in Laubscher (Eadie par 26).
In light of the differing indications in the case law (the academic conclusion seems reasonably monolithic), what exactly is the status in our law of the defence of non-pathological incapacity based on provocation or emotional stress? Perhaps it is fitting to let Navsa JA have the last word in this regard. In the recent bail decision of S v Oosthuizen (2018 JDR 0725 (SCA)), Navsa JA (par 30) briefly referred to the debate around the defence without any unequivocal statement that the Eadie decision has changed the law in any way. Instead, he concluded the brief reference to the defence in a supremely enigmatic way: "Commentators have stated that since this court's decision in S v Eadie … provocation leading up to a lack of criminal capacity as a defence has been limited, if not dealt the death knell." So much for the views of commentators, but the questions around the defence in the light of the Eadie decision remain.

Sandhya Maharaj
University of KwaZulu-Natal, Pietermaritzburg