Automatism

Published by Geoff Harrison | 3 September 2023

The defence of automatism is viewed with some scepticism however, once raised it is for the prosecution to negate automatism. If automatism is not negated, the accused is entitled to an outright acquittal as the prosecution would not have proven that the accused's acts were voluntary. In the decision of R v DB [2002] NSWCCA 87 (below) Brereton J at [11] - [45] sets out the history and rationale for the defence of automatism.

The defence of sexomnia is becoming a more prevalent defence.

Cases:

• R v Mendis [2024] NSWDC 420

_____________________________________________________________________

R v DB [2022] NSWCCA 87 (29 April 2022)

Court of Criminal Appeal

Supreme Court

New South Wales

Case Name:

R v DB

Medium Neutral Citation:

[2022] NSWCCA 87

Hearing Date(s):

26 November 2021

Date of Orders:

29 April 2022

Decision Date:

29 April 2022

Before:

Brereton JA at [1]

Wilson J at [72]

Ierace J at [204]

Decision:

1 Appeal dismissed

Catchwords:

CRIME – Appeal by Crown against acquittals – charges of sexual touching – respondent suffering from “sexsomnia” – question of volition – construction of phrase “mental health impairment” – s 4 Mental Health (Forensic Provisions) Act 2020 – whether Act codifies or alters the common law concerning the mental illness defence - relevance of common law defence of non-insane automatism –whether sexsomnia is a mental health impairment – question of correctness of construction and application of Act at trial

Legislation Cited:

Crimes (Appeal and Review) Act 2001 (NSW)

Crimes Act 1900 (NSW)

Interpretation Act 1987 (NSW)

Mental Health (Forensic Provisions) Act 1990 (NSW)

Mental Health and Cognitive Impairment Forensic Provisions Act 2020 (NSW)

Mental Health Cognitive Impairment Forensic Provisions Bill 2020 (NSW)

New South Wales Legislative Assembly, Parliamentary Debates (Hansard), 3 June 2020

Cases Cited:

Alcan (NT) Alumina Pty Ltd v Commissioner of Territory Revenue (2009) 239 CLR 27; [2009] HCA 41

Bratty v Attorney-General for Northern Ireland [1963] AC 386

Certain Lloyd’s Underwriters Subscribing to Contract No 1H00AAQS v Cross [2012] HCA 56; 293 ALR 412

Commissioner of Taxation v Consolidated Media Holdings Ltd (2012) 250 CLR 503; [2012] HCA 55

Fain v Commonwealth (1879) 78 Ky 183

Fang v The Queen [2018] NSWCCA 210

H.M. Advocate v Fraser (1878) 4 Coup 70

Project Blue Sky v Australian Broadcasting Authority (1998) 194 CLR 355; [1998] HCA 28

R v Boshears (Essex Assizes, 17 February 1961, unrep)

R v Cogdon (Supreme Court of Victoria, Smith J, December 1950, unrep)

R v Cottle [1958] NZPoliceLawRp 16; [1958] NZLR 999

R v DB [2021] NSWDC 213

R v Falconer [1990] HCA 49; (1990) 96 ALR 545

R v Holmes [1960] WAR 122

R v Joyce [1970] SASR 184

R v JS [2007] NSWCCA 272; (2007) 175 A Crim R 108

R v King [2004] ACTSC 82; (2004) 155 ACTR 55

R v M’Naghten [1843] EngR 875; (1843) 10 Cl & Fin 200; 8 ER 718

R v Paltridge (Devon Assizes, 19 February 1952, unrep)

R v Pantelic (1973) 1 ACTR 1; 21 FLR 253

R v PL [2009] NSWCCA 256

R v Porter (1933) 55 CLR 182; [1933] HCA 1

R v Price (Kent Assizes, 19 November 1949, unrep)

R v Quick [1973] EWCA Crim 1; [1973] QB 910

R v Radford (1985) 42 SASR 266

R v Stone (Bristol Assizes, Charles J, 24 November 1936, unrep)

R v Tolson [1889] UKLawRpKQB 85; (1889) 23 QBD 168

R v Tsigos [1964-5] NSWR 1607

R v XHR [2012] NSWCCA 247

R v Youssef (1990) 50 A Crim R 1

Rasic v The Queen [2009] NSWCCA 202

Ryan v The Queen [1967] HCA 2; (1967) 121 CLR 205

Secretary, Department of Family and Community Services v Hayward (a pseudonym) [2018] NSWCA 209

Stapleton v The Queen (1952) 86 CLR 35; [1952] HCA 56

Watmore v Jenkins [1962] 2 QB 572

Williams v The Queen (1986) 161 CLR 278; [1986] HCA 88

Woodbridge v The Queen [2010] NSWCCA 185; (2010) 208 A Crim R 503

Texts Cited:

Canadian Journal of Psychiatry 2003; 48:311-307

N Morris, “Somnambulistic Homicide: Ghosts, Spiders, and North Koreans (1951) Res Judicatae 29

NSW Law Reform Commission, ‘People with Cognitive and Mental Health Impairments in the Criminal Justice System: Criminal Responsibility and Consequences’ (Report, 2013)

Macquarie Dictionary, 8th ed (2022)

Sir Owen Dixon, “A Legacy of Hadfield, M’Naghten and Maclean” (1957) 31 ALJ 255

Vol. 1, The New Shorter Oxford English Dictionary, Clarendon Press, Oxford 1993

‘Alleged Bayonet Attack: Recruit for Trial’, The Manchester Guardian, 26 August 1949

‘Sleepwalking Defence’, The Times, 26 November 1936; ‘Sleepwalker’s Plea Succeeds’, Daily Telegraph, 25 November 1936

The Manchester Guardian, 21 November 1949

‘U.S. Sergeant is Cleared of Murder’, The Times, 18 February 1961.

‘Sleepwalker Attacked Wife with Axe’, Daily Telegraph, 20 February 1952

Category:

Principal judgment

Parties:

Regina (Applicant)

DB (Respondent)

Representation:

Counsel:

H Roberts SC & K Heath (Crown) (Applicant)

P Strickland SC & G Lewer (Respondent)

Solicitors:

Solicitor for Public Prosecutions (NSW) (Crown) (Applicant)

Hugo Law Group (Respondent)

File Number(s):

2019/345129

Publication Restriction:

None

Decision under appeal:

Court or Tribunal:

District Court of New South Wales

Jurisdiction:

Criminal

Date of Decision:

25 May 2021

Before:

Weinstein SC DCJ

File Number(s):

2019/345129

JUDGMENT

1 BRERETON JA: It is a fundamental principle of the criminal law that a person is not guilty of a crime if the act which would constitute it was not done in exercise of the accused’s will to perform it.[1] For, as it was put by Lord Denning in Bratty v Attorney-General for Northern Ireland:[2]

No act is punishable if it is done involuntarily: and an involuntary act in this context — some people nowadays prefer to speak of it as “automatism” — means an act which is done by the muscles without any control by the mind, such as a spasm, a reflex action or a convulsion; or an act done by a person who is not conscious of what he is doing, such as an act done whilst suffering from concussion or whilst sleep-walking. The point was well put by Stephen J. in 1889: “Can anyone doubt that a man who, though he might be perfectly sane, committed what would otherwise be a crime in a state of somnambulism, would be entitled to be acquitted? and why is this? Simply because he would not know what he was doing,” see Reg. v. Tolson [(1889) [1889] UKLawRpKQB 85; 23 QBD 168, at p 187].

2 In this Crown appeal,[3] the immediate issue is whether the respondent, who was acquitted at a judge-alone trial in the District Court of two counts of sexual touching of his daughter[4] because he was asleep and therefore acting involuntarily, ought instead have been the subject of a special verdict of “act proven but not criminally responsible” under s 30 of the Mental Health and Cognitive Impairment Forensic Provisions Act 2020 (NSW) (“2020 Act”).[5] However, the larger question which arises is whether the effect of the 2020 Act is effectively to abolish the “defence” of “sane automatism”, by providing that a person who lacks volition by reason of being asleep at the time of the charged act has a mental health impairment.

BACKGROUND

3 The respondent was tried on an indictment which charged him with three offences, to the following effect:

(1) that on 30 June 2019, at Camden in the State of New South Wales, he had sexual intercourse with the complainant, a child then under the age of 10 years, namely 9 years, contrary to s 66A(1) of the Crimes Act;

(2) that on or about 2 November 2019, at Wollongong in the State of New South Wales, he had sexual intercourse with the complainant, a child then under the age of 10 years, namely 9 years, contrary to s 66A(1) of the Crimes Act; and

(3) that on or about 2 November 2019, at Wollongong in the State of New South Wales, he intentionally sexually touched the complainant, a child then under the age of 10 years, namely 9 years, contrary to s 66DA(a) of the Crimes Act.

4 At the trial it was not in issue that the respondent had, on two occasions, sexually touched the complainant, his young daughter. There was an issue, which was resolved in favour of the respondent, as to whether the touching involved penetration so as to amount to sexual intercourse; accordingly, the statutory alternative to the offence the subject of each of counts 1 and 2, being sexual touching contrary to s 66DA(a) of the Crimes Act, fell to be considered.[6] The respondent’s defence was that his conduct was involuntary: he was asleep, and this was a manifestation of sexsomnia, a parasomnia or form of somnambulism (sleepwalking) in which a person engages in sexual activity while asleep. The Crown accepted that when he performed the charged acts, the respondent was asleep and that his acts were not voluntary.[7] It is unsurprising that this was uncontentious, given that it was the complainant’s evidence that the respondent was snoring at the time of at least one if not both of the episodes.[8] The chief issue at the trial was whether the accused was entitled to an outright acquittal by reason that his acts were involuntary, or whether he had a mental health impairment within s 4 of the 2020 Act with the consequence that there should be a special verdict of “act proven but not criminally responsible" pursuant to s 30 of that Act.

5 Conformably with the position conceded by the Crown, the trial judge found that the Crown had failed to prove to the requisite standard that any of the charged acts was voluntary.[9] Turning to the issue of whether the respondent had a mental health impairment, his Honour concluded that he did not.[10] The first reason for this conclusion was that, accepting expert psychiatric evidence that parasomnias are involuntary and not the product of a disordered mind, as those suffering from them typically have no psychopathology when awake, the respondent had an absence of volition rather than a disturbance of volition, and so did not fall within s 4(1)(a) of the 2020 Act.[11] A second reason was that his Honour also concluded that any “disturbance” was not significant for clinical diagnostic purposes, as there appeared to be no consensus that parasomnias result in a true mental health diagnosis.[12]

6 Further, his Honour concluded that even if the respondent had a mental health impairment within s 4, s 28 of the 2020 Act was not engaged in any event, as the respondent performed the acts unconsciously and involuntarily whilst he was asleep, without choice or decision, while s 28 was reserved for those who are conscious of their actions but either do not know the nature and quality of their acts or do not know that their acts were wrong.[13]

7 Thus, having found that sexual intercourse had not been proved, verdicts of not guilty were returned on counts 1 and 2. On the basis of the finding that the acts of sexual touching were not voluntary, his Honour concluded:[14]

“In my opinion, the Crown has failed to prove, on the balance of probabilities, that [the] accused, at the time of his actions in June and November 2019, suffered from a mental health impairment pursuant to section 4 of the MHCIFP Act. Further, in my opinion, the Crown has failed to prove, on the balance of probabilities, that any mental health impairment the accused may have suffered at those dates had the effect that he did not know the nature and quality of those acts or that he did not know that those acts were wrong within the meaning of section 28(1) of the MHCIFP Act. Thus that Act has no application in the circumstances, and I enter verdicts of not guilty to the statutory alternatives to counts 1 and 2 and count 3.”

8 The Crown appeals on grounds to the following effect:

(1) that the trial judge in erred in failing to find that the respondent’s sexsomnia was a ‘mental health impairment’ pursuant to s 4 of the 2020 Act;

(2) that the trial judge erred in finding that “disturbance of ... volition” in s 4(1) does not include an absence of volition; and

(3) that the trial judge in erred in finding, if at the time of carrying out the act constituting the offence the person had a mental health impairment pursuant to s 4(1), that unconscious and/or involuntary acts cannot fall within s 28 of the Act.

9 These grounds are said to raise the following questions of law relating to the proper construction of the 2020 Act and its relationship with the common law defence of insanity:

(1) whether it is necessary that psychiatric treatment be available for a particular condition before that condition is capable of being “significant for clinical diagnostic purposes” within the meaning of s 4(1)(b) of the 2020 Act;

(2) whether a “disturbance of ... volition” in s 4(1)(a) can include an absence of volition; and

(3) whether s 28 of the 2020 Act can have application to unconscious or involuntary acts, which are unconscious or involuntary as a result of a mental health impairment.

10 The answers to these questions are informed, in part, by the relationship between the defences of automatism and insanity at common law, and in particular in the field of somnambulism.

AUTOMATISM

11 Although sometimes viewed in some quarters with a degree of scepticism,[15] it is now well-established at common law that the “defence” of automatism, if not negated by the prosecution, entitles an accused person to an outright acquittal (as distinct from an acquittal on the grounds of mental illness), because the prosecution has failed to establish that the relevant act was a voluntary one of the accused. In Bratty, Viscount Kilmuir said:[16]

“The Court of Criminal Appeal rejected that “first portion of the argument” on the ground that the learned judge was right in not leaving to the jury the defence of automatism, in so far as it purported to be founded on a defect of reason from disease of the mind within the M'Naughten Rules. In this I think that they were right. To establish the defence of insanity within the M'Naughten Rules the accused must prove on the preponderance of probabilities, first a defect of reason from a disease of the mind, and, secondly, as a consequence of such a defect, ignorance of the nature and quality (or the wrongfulness) of the acts. We have to consider a case in which it is sought to do so by medical evidence to the effect that the conduct of the accused might be compatible with psychomotor epilepsy, which is a disease of the mind affecting the reason, and that psychomotor epilepsy could cause ignorance of the nature and quality of the acts done, but in which the medical witness can assign no other cause for that ignorance. Where the possibility of an unconscious act depends on, and only on, the existence of a defect of reason from disease of the mind within the M'Naughten Rules, a rejection by the jury of this defence of insanity necessarily implies that they reject the possibility.

The Court of Criminal Appeal also took the view that where the alleged automatism is based solely on a disease of the mind within the M'Naughten Rules, the same burden of proof rests on the defence whether the “plea” is given the name of insanity or automatism. I do not think that statement goes further than saying that when you rely on insanity as defined by the Rules you cannot by a difference of nomenclature avoid the road so often and authoritatively laid down by the courts.

What I have said does not mean that, if a defence of insanity is raised unsuccessfully, there can never, in any conceivable circumstances, be room for an alternative defence based on automatism. For example, it may be alleged that the accused had a blow on the head, after which he acted without being conscious of what he was doing or was a sleep-walker. There might be a divergence of view as to whether there was a defect of reason from disease of the mind (compare the curious position which arose in Reg. v. Kemp [[1957] 1 QB 399; [1956] 3 WLR 724; [1956] 3 All ER 249; 40 Cr App R 121]. The jury might not accept the evidence of a defect of reason from disease of the mind, but at the same time accept the evidence that the prisoner did not know what he was doing. If the jury should take that view of the facts they would find him not guilty.”

12 His Lordship later held that if, after considering evidence properly left to them by the judge, the jury are left in real doubt as to whether or not the accused acted in a state of automatism, then they should acquit because the necessary mens rea — if indeed the actus reus — has not been proved beyond reasonable doubt.[17]

13 Reference has already been made to the speech of Lord Denning. Lord Morris said:[18]

"This raises the question whether a person who by legal tests and standards is sane and who is charged with a criminal offence could be held to be non-accountable for his actions so as to be not guilty of the offence charged against him on the basis that his actions had been unconscious ones and in that sense involuntary. My Lords, I can conceive that this could be so, though the cases where such a situation could arise must be very rare. Each set of facts must require a careful investigation of its own circumstances but if, by way of taking an illustration, it were considered possible for a person to walk in his sleep and to commit a violent act while genuinely unconscious, then such a person would not be criminally liable for that act. Apart altogether from any question whether some particular criminal charge requires proof of some particular intent, in the possible case that I have postulated there would be immunity from any conviction for the reason that the act in question could not really be considered to be the act of the person concerned at all.”

14 Bratty was considered by this Court in R v Tsigos,[19] in which Moffitt J, as he then was, explained that where the accused’s involuntary act was attributable to a defect of reason arising from disease of the mind, the M’Naghten[20] Rules applied to cast on the accused an onus which if discharged entitled the accused to an acquittal on grounds of mental illness; whereas automatism not arising from mental illness – on which the accused bore only an evidential burden - resulted in a complete acquittal. His Honour said (citations omitted):[21]

“It has been claimed from time to time that “automatism” can exist in cases where it does not arise from any disease of the mind and examples have been given, such as sleep-walking, activity in a dream or when suffering concussion. That involuntary conduct of a sane person can exist cannot be doubted. It could, no doubt, cause harm to others as where a person after being concussed or in an epileptic fit runs his car off the road or otherwise causes injury to another in an uncontrolled physical way. However, it is much more difficult to conceive the possibility of a sane person performing acts with the external semblance of intentional acts which would constitute a crime, while acting in a state of automatism. Judges have been reluctant to exclude the possible existence of such occurrences, but have doubted the credibility of examples at times claimed or envisaged, e.g., as of a man strangling another in an epileptic fit or a man killing his wife in his sleep, dreaming he was fighting a wolf. Before such a condition can be considered by a jury as a reasonable hypothesis of innocence, a proper foundation for such an inference must first appear. ... It may be added that, by hypothesis, the person has no consciousness or, at least, no memory of the relevant time and so has no capacity to give evidence of his then condition. ... Thus the proper foundation could not be laid except by means of proper evidence, which in most cases would need the support of some qualified scientific opinion, properly based, explaining the cause of the mental incapacity. At this point, ordinarily, it can be expected it will appear whether the condition arises from disease of the mind or not and then, as indicated, it is only in the latter case that the consideration mentioned could arise for the jury in relation to a complete acquittal.”

15 In Ryan,[22] Barwick CJ, explaining the distinction between voluntariness and intention, emphasised that a deed which is not the result of the accused's will to act cannot be made the source of criminal responsibility:

“That a crime cannot be committed except by an act or omission of or by the accused is axiomatic. It is basic, in my opinion, that the "act" of the accused, of which one or more of the various elements of the crime of murder as defined must be predicated must be a "willed", a voluntary act which has caused the death charged. It is the act which must be willed, though its consequences may not be intended. In the ordinary run of cases the voluntary quality of the deed physically related to the accused is not in question. The presumption to which their Lordships refer in Bratty v Attorney-General for Northern Ireland (1963) AC 386 operates sub silentio. Mostly the contest will concentrate upon the question of the intent with which the accused acted. Consequently, there has not been any frequent need to express with technically expressed precision the difference between that element of mens rea which relates the will to act to the deed in question and that element which relates to it the general intent with which that will was exercised. Often, when absence of the requisite intent as distinct from the will to act is sought to be stressed, there is said to be an "accident", or that the physical act which the evidence relates to the accused was "accidental". But, as I have pointed out, such language is ambiguous for it may equally cover an unwilled act and is often so used. Cf Woolmington v Director of Public Prosecutions (1935) AC 462, at p 472. In that case, (1935) AC, at p 482 the description "unintentional" appears to be used to cover an act which was involuntary or unwilled, descriptions of it which for my part I would prefer. However, by whatever adjective or adjectival phrase it is described, the deed which was not the result of the accused's will to act cannot, in my opinion, be made the source of criminal responsibility in him.”

16 His Honour also contrasted “an unwilled act and a willed act the product of a diseased mind”:[23]

“I would then observe that a distinction must be maintained between an unwilled act and a willed act the product of a diseased mind which knows not the nature or quality of the willed act. To express it with what may well be technical inexactitude, it may be said that in the latter case the act is willed by a diseased will in contradistinction to the act which is not willed at all. That to my mind is the core of Bratty v Attorney-General for Northern Ireland (1963) AC 386, and the essence of the distinction between the case of a sane and an insane accused. Whether or not the distinction was fully maintained in that case in applying its basic principles to its particular facts need not be presently considered, for in the instant case no question of any condition of the applicant's body or mind as the cause of the lack of the will to act is raised. Nor is there any need to discuss the validity of Lord Denning's reasons for separating lack of will due to a disease from other acts involuntary for other causes.”

17 After considering the speeches in Bratty of Viscount Kilmuir, Lord Denning and Lord Morris, his Honour concluded:[24]

“In my opinion, the authorities establish, and it is consonant with principle, that an accused is not guilty of a crime if the deed which would constitute it was not done in exercise of his will to act. The lack of that exercise which precludes culpability is not, in my opinion, limited to occasions when the will is overborne by that of another, or by physical force, or the capacity to exercise it is withdrawn by some condition of the body or of the mind of the accused. ... If voluntariness is not conceded and the material to be submitted to the jury wheresoever derived provides a substantial basis for doubting whether the deed in question was a voluntary or willed act of the accused, the jury's attention must be specifically drawn to the necessity of deciding beyond all reasonable doubt that the deed charged as a crime was the voluntary or willed act of the accused. If it was not then for that reason, there being no defence of insanity, the accused must be acquitted.”

18 Windeyer J distinguished “the probable but unpredictable reaction of a man when startled” – which would not be involuntary in the relevant sense – from conditions such as epilepsy and sleepwalking, which would be:[25]

“But assume that the applicant's act was involuntary, in the sense in which the lexicographers use the word, would that, as a matter of law, absolve him from criminal responsibility for its consequences? I do not think so. I do not think that, for present purposes, such an act bears any true analogy to one done under duress, which, although done by an exercise of the will, is said to be involuntary because it was compelled. Neither does it, I think, bear any true analogy to an act done in convulsions or an epileptic seizure, which is said to be involuntary because by no exercise of the will could the actor refrain from doing it. Neither does it, I think, bear any true analogy to an act done by a sleep-walker or a person for some other reason rendered unconscious whose action is said to be involuntary because he knew not what he was doing.”

19 The judgment of the High Court in Falconer confirmed the entitlement of an accused to an unqualified acquittal if the charged act occurred independently of the accused’s will by reason of involuntary conduct not arising from a disease of the mind or natural mental infirmity. Toohey J said that a person was not criminally responsible for an act or omission which occurred independently of the will; that the jury should first consider whether the Crown has disproved, beyond reasonable doubt, non-insane automatism, and if the Crown has failed to do so, then the accused is entitled to an unqualified acquittal. Gaudron J said that the jury should have been directed to consider whether the prosecution had proved beyond reasonable doubt that the accused’s will accompanied the act; and that it should have been explained to the jury that the prosecution would not have proved the issue beyond reasonable doubt if it was a reasonable hypothesis that she acted while experiencing a particular mental state described in the evidence as one that may be experienced by a normal and healthy mind and in which the personality is segmented so that acts are performed independently of the will. Deane and Dawson JJ agreed generally with the reasoning of Toohey J and Gaudron J, and said that in a case where an issue of sane automatism was raised, an accused would be entitled to an acquittal if the prosecution failed to disprove sane automatism beyond reasonable doubt. Mason CJ, Brennan and McHugh JJ differed from the majority in that their Honours were of the view that the accused bore an onus to prove that any claimed malfunction of the mind was transient, caused by physical or psychological trauma which the mind of an ordinary person would be likely not to have withstood and was not prone to recur[26] (that is, sane automatism), whereas the majority required the prosecution to disprove sane automatism beyond reasonable doubt.[27]

20 Consequently, at common law, if there is evidence to raise the question of voluntariness, then unless the prosecution proves that the accused’s act was involuntary, other than by reason of mental illness, there should be an acquittal simpliciter; whereas if on the balance of probabilities the involuntary act was attributable to a defect of reason due to mental illness such that the accused was incapable of appreciating the nature and quality of the act, or of knowing that it was wrong, there should be an acquittal on ground of mental illness.

21 As to the distinction recognised in the authorities between cases of “insane automatism” (where the unconscious act was attributable to a “defect of reason from disease of the mind”), entitling the accused to an acquittal by reason of mental illness, and those of “sane automatism” (entitling the accused to an outright acquittal), in R v Cottle, Gresson P said:[28]

“It would appear that automatism raised as a defence to a criminal charge may be something quite different and distinct from insanity. In a particular case, it may be that the automatism relied on is due to some "disease of the mind" but it is not necessarily so. Automatism, which strictly means action without conscious volition, has been adopted in criminal law as a term to denote conduct of which the doer is not conscious — in short doing something without knowledge of it, and without memory afterwards of having done it — a temporary eclipse of consciousness that nevertheless leaves the person so affected able to exercise bodily movements. In such a case, the action is one which the mind in its normal functioning does not control. This may be due to some "disease of the mind" or it may not; it may happen with a perfectly healthy mind (e.g. in somnambulism which may be unaccompanied by any abnormality of mind), or it may occur where the mind is temporarily affected as the result of a blow, or by the influence of a drug or other intoxication. It may on the other hand be caused by an abnormal condition of the mind capable of being designated a mental disease. What are known as the M'Naghten Rules can have no application unless there is some form of "disease of the mind", which is not necessarily present in all cases of automatism.

...

The test under the M'Naghten formula is "defect of reason" due to disease of the mind, whereas under the New Zealand statute the test is simply "disease of the mind" producing incapacity to understand the nature of the act. But "disease of the mind" is essential both in England and in New Zealand to permit of a finding of insanity. Automatism, that is action without conscious volition, may or may not be due to or associated with "disease of the mind" — a term which defies precise definition and which can comprehend mental derangement in the widest sense whether due to some condition of the brain itself and so to have its origin within the brain, or whether due to the effect upon the brain of something outside the brain, e.g. arterio sclerosis. The adverse effect upon the mind of some happening, e.g. a blow, hypnotism, absorption of a narcotic, or extreme intoxication all producing an effect more or less transitory cannot fairly be regarded as amounting to or as producing "disease of the mind".”

22 North J also accepted that a sleepwalker who committed a criminal act could not be said to have done so while insane,[29] adopting what had been said by Stephen J in R v Tolson.[30]

23 In R v Joyce,[31] the Full Court of the Supreme Court of South Australia appears to have accepted that the acts of a sleepwalker were to be regarded as involuntary:

“We do not mean that apparently willed acts committed by a sleepwalker or a dreamer or a person who has received a blow on the head are necessarily acts for which he is criminally responsible. In such cases it may be that whatever directs the acts is so subordinated to the conscious will that the act can fairly be said to be involuntary. In other words in such cases all the deliberative functions of the mind may be said to be absent; see Burr's Case [[1969] NZLR 736], per North P. at p. 745. It may be relevant also that in such cases the condition is produced by well known causes or in well known states for which the subject is not responsible. Everyone must go to sleep and a man does not choose to be knocked on the head; cf. R. v. Carter [[1959] ALR 335].”

24 In R v Pantelic,[32] Fox J in the Supreme Court of the Australian Capital Territory said:

“It was next submitted on behalf of the Crown that I should not leave to the jury the question of “sane automatism”. This is a term which has been used from time to time and which was used in Bratty v. Attorney-General for Northern Ireland. The submission was put on two different grounds. One was, in effect, that the categories of “sane automatism” are limited and Reg. v. Tsigos; Reg. v. Tsigos and Reg. v. Joyce were referred to. It may be possible, depending on the way one defines “automatism”, to limit its possible causes in some way, but the issue is not simply whether there was “automatism” but whether the acts of the accused were voluntary, in the sense that they were the result of his conscious volition. An absence of voluntariness can be due to a number of causes. The fact that some, such as somnambulism, epileptic convulsions, and concussion have been dealt with in decided cases does not mean that there are not others. Some of the causes can be classified medically, or in accordance with the M'Naghten rules, as diseases of the mind, and some cannot, but the ultimate question is whether the act was voluntary. In my view there was evidence to go to the jury that the acts of the accused were not voluntary, and the view was open that this was not due to insanity, in the legal sense.”

25 In R v Quick,[33] a distinction was identified between “a malfunctioning of the mind of transitory effect caused by the application to the body of some external factor such as violence, drugs, including anaesthetics, alcohol and hypnotic influences”, and a “disease of the mind”:

“In this quagmire of law seldom entered nowadays save by those in desperate need of some kind of a defence, Bratty v. Attorney-General for Northern Ireland [1963] A.C. 386, 403, 412, 414 provides the only firm ground. Is there any discernible path? We think there is. Judges should follow in a common sense way their sense of fairness. This seems to have been the approach of the New Zealand Court of Appeal in Reg. v. Cottle [1958] NZPoliceLawRp 16; [1958] N.Z.L.R. 999, 1011 and of Sholl J. in Reg. v. Carter [1959] VicRp 19; [1959] V.R. 105, 110. In our judgment no help can be obtained by speculating (because that is what we would have to do) as to what the judges who answered the House of Lords' questions in 1843 meant by disease of the mind, still less what Sir Matthew Hale meant in the second half of the 17th century. A quick backward look at the state of medicine in 1843 will suffice to show how unreal it would be to apply the concepts of that age to the present time. Dr. Simpson had not yet started his experiments with chloroform, the future Lord Lister was only 16 and laudanum was used and prescribed like aspirins are today. Our task has been to decide what the law means now by the words “disease of the mind.” In our judgment the fundamental concept is of a malfunctioning of the mind caused by disease. A malfunctioning of the mind of transitory effect caused by the application to the body of some external factor such as violence, drugs, including anesthetics, alcohol and hypnotic influences cannot fairly be said to be due to disease. Such malfunctioning, unlike that caused by a defect of reason from disease of the mind, will not always relieve an accused from criminal responsibility. A self-induced incapacity will not excuse (see Reg. v. Lipman [1970] 1 Q.B. 152), nor will one which could have been reasonably foreseen as a result of either doing, or omitting to do something, as, for example, taking alcohol against medical advice after using certain prescribed drugs, or failing to have regular meals while taking insulin. From time to time difficult border line cases are likely to arise. When they do, the test suggested by the New Zealand Court of Appeal in Reg. v. Cottle [1958] NZPoliceLawRp 16; [1958] N.Z.L.R. 999, 1011 is likely to give the correct result, viz., can this mental condition be fairly regarded as amounting to or producing a defect of reason from disease of the mind?”

26 In R v Youssef, Hunt J (as he then was), with whom Wood and Finlay JJ agreed, said:[34]

“Automatism, however, may be relevant to two different issues in criminal trial. It may be relevant to whether the Crown has established that the actions of the accused were voluntary and, if the evidence of the existence of that condition raises a reasonable doubt as to whether the actions were of that character, the accused is entitled to be acquitted outright. Automatism may also be relevant to the defence of mental illness, in that it may establish that, as a result of a defect of a reason proceeding from a disease of the mind, the accused did not appreciate the nature and quality of his physical act (the first branch of the M’Naghten Rules), leading to the special verdict of not guilty on the ground of mental illness but not to an outright acquittal. If the effect of any disease is to impair a person’s mental faculties of reason, memory and understanding so that he does not know what he is doing, it matters not whether the impairment is permanent or transient or intermittent or whether the cause of that impairment is organic (as in epilepsy) or functional Regina v Sullivan [1984] AC 156 at 172-173.”

27 Elaborating on the distinction, Hunt J continued:[35]

“This duality of relevance has led to a distinction being made in the cases between insane and sane (or non-­insane) automatism, an intellectually unhappy and artificial distinction which at times produces both unsatisfactory and inconvenient results and a distinction which is often difficult of application. The clearest statement of the distinction is that of King CJ in Radford (1985) 20 A Crim R 388 at 397-398.

‘The significant distinction is between the reaction of an unsound mind to its own delusions or to external stimuli on the one hand [the insane automatism] and the reaction of a sound mind to external stimuli, including stress producing factors, on the other hand [the sane automatism].’

Obvious examples of the latter category may be loss of control caused by physical factors (a blow to the head or the effects of chemical substances or intoxication) or by psychological or emotional stress. Another example often given in the cases is that of a sleepwalker; it appears to have been given for the first time in Tolson [1889] UKLawRpKQB 85; (1889) 23 QBD 168 at 187.

28 The distinction was confirmed by the High Court in Falconer. Deane and Dawson JJ said (emphasis added):[36]

“However, there is no such presumption to be overcome in the application of s 23. The onus of proving guilt remains with the prosecution and that onus is not discharged if an accused is able to raise a reasonable doubt. Of course, common experience teaches us that a person's will ordinarily accompanies his actions and evidence will be required to establish the extraordinary circumstance that an accused's acts occurred independently of the exercise of his will or to raise a doubt whether that was so. It is sometimes said in that situation that the accused is required to rebut an evidentiary presumption or to discharge an evidentiary burden of proof, but it is merely a requirement that there be evidence to displace ordinary human experience. And it will not be enough for an accused merely to assert that his acts were involuntary or that he suffered a loss of memory. Evidence of his condition at the time of the alleged offence supported by some expert medical opinion will be required before an issue of sane automatism can realistically be said to be raised. Moreover, those conditions which will admit of involuntariness that is not the product of disease or natural mental infirmity will be quite confined. The few suggested instances would seem to include: sleepwalking in some circumstances, some cases of epilepsy, concussion, hypoglycaemia and dissociative states. Nevertheless, given these constraints, an accused in setting up a defence under s 23 does not have to prove his condition on the balance of probabilities in order to succeed; he merely has to raise a reasonable doubt that his actions were the result of an involuntary reaction of a sane mind.”

29 Toohey J observed:[37]

“Undoubtedly, automatism implies conduct which is involuntary. Some examples, such as acts performed while sleepwalking or while concussed, readily enough answer the description of automatism.”

30 After referring to the above-quoted passage from R v Quick, his Honour said:[38]

“But there are real difficulties with the “external factor” test. Sleepwalking and the diabetic condition of hypoglycaemia stem from internal malfunctioning and yet have traditionally been treated as instances of automatism.”

31 Gaudron J said (emphasis added):[39]

“Despite the differences directed by the language of the Code, that distinction is based on the same considerations as those that distinguish between the common law concept of a sane mind and the concept of a mind that has a defect of reason due to a disease of the mind. See Hitchens v R [1959] TASStRp 25; [1959] Tas SR 209 at 249–50; Hitchens v R (No 2), at 49; Williams v R, at 106. The distinction has sometimes been expressed in terms of mental states having an external cause and those that proceed from internal causes: R v Quick [1973] EWCA Crim 1; [1973] QB 910 at 922. Sometimes the distinction has been expressed by reference to the transient or recurring nature of the particular mental state: Bratty, at 412; R v Carter [1959] VicRp 19; [1959] VR 105 at 110; R v Meddings [1966] VicRp 42; [1966] VR 306 at 309–10. In general terms, a recurring state which involves some abnormality will indicate a mind that is diseased or infirm, but the fundamental distinction is necessarily between those mental states which, although resulting in abnormal behaviour, are or may be experienced by normal persons (as, for example and relevant to the issue of involuntariness, a state of mind resulting from a blow to the head) and those which are never experienced by or encountered in normal persons. That point was made, although in a quite different context, by Dixon J in Porter, at 188, where his Honour observed that the diseased mind is to be distinguished from the “mere excitability of a normal man, passion, even stupidity, obtuseness, lack of self-control, and impulsiveness”. See also the reference by Sir Owen Dixon in his paper “A Legacy of Hadfield, M'Naghten and Maclean”, (1957) 31 Australian Law Journal 255 at 260, to “transient states attributable either to the fault or to the nature of man”. And in Radford, King CJ (at 274) distinguished between “an underlying pathological infirmity of the mind, ... which can be properly termed mental illness” and “the reaction of a healthy mind to extraordinary external stimuli”.”

32 The minority (Mason CJ, Brennan and McHugh JJ) considered that the discriminator was that involuntariness attributable to some malfunction of the mind was a defence of unsoundness of mind (or insanity), unless the malfunction was transient, caused by physical or psychological trauma which the mind of an ordinary person would be likely not to have withstood and was not prone to recur.[40]

33 From these cases it appears that unconscious acts attributable not to mental illness but to the transitory effects on a person of ordinary sound mind of conditions such as concussion, post traumatic epilepsy, hypoglycaemia, and somnambulism, characterise, although they do not necessarily define, “sane automatism”. In any event, for present purposes it is a notable feature of the cases to which reference has been made that “sleepwalking” is treated, effectively, as a paradigm case of sane automatism. There are numerous illustrations of its application.

34 In 1879, in Fain v Commonwealth,[41] the Kentucky Court of Appeals held that evidence tending to show that the accused may have committed a homicide unconsciously while asleep or between sleeping and waking, had been wrongly excluded. Speaking for the Court, Cofer J said (emphasis added):[42]

“If the prisoner, when he shot the deceased, was unconscious, or so nearly so that he did not comprehend his own situation and the circumstances surrounding him, or that he supposed he was being assailed and that he was merely resisting an attempt to take his life or do him great bodily injury, he should be acquitted – in one case, because he was not legally responsible for any act done while in that condition, and in the other, because he is excusable on the ground of self-defence; for although it is clear that he was not in danger and had no reasonable grounds to believe he was, yet if through the derangement of his perceptive faculties, it appeared to him that he was in danger, he is as free from punishable guilt as if the facts had been as he supposed them to be.”

35 In 1889, in the case of R v Tolson referred to by Lord Denning in the passage quoted above, Stephen J in the Court of Crown Cases Reserved suggested that an accused who performed the charged act while sleepwalking was entitled to an acquittal. Observing that while the mental element of most crimes was typically described by such words as “maliciously,” “fraudulently,” “negligently,” or “knowingly,” the general if not invariable practice of the legislature was to leave some of the mental elements unexpressed, his Lordship continued:[43]

“With regard to knowledge of fact, the law, perhaps, is not quite so clear, but it may, I think, be maintained that in every case knowledge of fact is to some extent an element of criminality as much as competent age and sanity. To take an extreme illustration, can any one doubt that a man who, though he might be perfectly sane, committed what would otherwise be a crime in a state of somnambulism, would be entitled to be acquitted? And why is this? Simply because he would not know what he was doing.”

36 In H.M. Advocate v Fraser,[44] the accused killed his child while dreaming that he was struggling with a wild beast; the jury were directed to return a verdict that he killed the child, but that he was in a state in which he was unconscious of the act which he was committing by reason of the condition of somnambulism, and that he was not responsible. He was acquitted “simpliciter”.[45]

37 In R v Stone,[46] the prosecution case was that during the night the accused, who slept on the ground floor of a house, entered a bedroom on the first floor occupied by two sisters and committed what was coyly described as “an offence against a girl”; reading between the lines of the newspaper report, it sounds very much like what would now be sexual touching. His defence was that he was sleepwalking. He said he only woke up when the girl shouted. There was evidence that he had walked in his sleep on a previous occasion. Charles J directed the jury to acquit him.

38 In R v Price,[47] a Royal Marine dreamt that his company sergeant major was meting out punishment to him, while his corporal was standing by laughing; he grabbed his rifle with bayonet fixed and attacked the corporal. There was expert evidence that his subconscious mind may have controlled him while he attacked the corporal, and that he was awoken by the intervention of two soldiers who came to the aid of the corporal. He was acquitted of wounding with intent to murder.

39 In R v Cogdon,[48] the accused said that she was dreaming that the Korean War was “all around the house”, that soldiers were in her daughter’s room, and that one was on her bed attacking the daughter; during the dream she arose from her bed, fetched an axe from the woodheap, entered her daughter’s room and struck her two accurate forceful blows on the head with the blade of the axe, killing her. She was tried for murder. Her story of the dream was supported by evidence of her physician, a psychiatrist, and a psychologist, to the effect that she was suffering from a form of hysteria with an overlay of depression, and was of a personality in which somnambulistic acts were to be expected; they agreed that she was not psychotic and that had she been awake, no defence under the M’Naghten Rules could have been established. The judge directed the jury that if they accepted her story, she was not responsible for her acts during her sleep, and she was acquitted – not on ground of insanity, but because the act of killing was not, in law, regarded as her act at all.

40 In R v Paltridge,[49] a naval officer was alleged to have attacked his wife with an axe. He said that he had no recollection of the event until he woke up and found his hands around her throat. There was expert evidence that somnambulism was common and that there had been two cases in the last 18 months of a husband having attempted to strangle his wife while asleep; that a person in such a state would not have conscious purpose, is living out a dream, and more often than not has no memory of it; and that somnambulism was not a mental illness and was not associated with mental deterioration. He was acquitted of attempted murder.

41 In R v Holmes,[50] Jackson SPJ, in directing the jury that a verdict of not guilty would be the proper verdict if the accused was acting without any conscious volition or without willpower, said:

“Another case in which the will does not go with the deed is where a man is unconscious and acts in that state. There are numerous examples of that; for instance, unconsciousness in sleep — sleepwalking. An act done during that time carries no criminal responsibility. Of course it could not. It would be barbarous if it did. There was a case many years ago of a man acting under the influence of a nightmare during his sleep and attacking his child in the belief in his nightmare that it was a wild animal. No criminal responsibility. His will did not go with the deed.”

42 In R v Boshears,[51] the accused, a staff sergeant in the United States Air Force, was charged with the murder of a 20-year-old woman who had accompanied him back to his flat after New Year’s Eve drinks in public houses. He claimed that he was asleep when he had strangled her. In a summing up which appears to have expressed a considerable measure of scepticism as to the accused’s version, Glyn-Jones J nonetheless directed the jury that if the accused strangled the girl while he was asleep it was not a voluntary act and he was entitled to be acquitted, as was also the case if the jury were in doubt whether he was asleep or not. He was acquitted.

43 It is, therefore, plain that in law the acts of a person who is asleep and engaging in somnambulistic activity are not willed acts; and the accused is not legally responsible for them. It makes no difference whether they would amount to murder, rape, or assault if the accused were awake and conscious. Given the incidence of sexual dreaming, one might think that sexsomnia might be less implausible than some of the cases described above. It is in my view therefore plain that, at common law, the respondent would have been entitled to the outright acquittal that he received from the trial judge. The question is whether the 2020 Act, which commenced on 27 March 2021, requires a different result.

THE 2020 ACT

44 Under the 2020 Act, s 28 provides a defence of mental health impairment or cognitive impairment, as follows:

28 Defence of mental health impairment or cognitive impairment

(1) A person is not criminally responsible for an offence if, at the time of carrying out the act constituting the offence, the person had a mental health impairment or a cognitive impairment, or both, that had the effect that the person—

(a) did not know the nature and quality of the act, or

(b) did not know that the act was wrong (that is, the person could not reason with a moderate degree of sense and composure about whether the act, as perceived by reasonable people, was wrong).

(2) The question of whether a defendant had a mental health impairment or a cognitive impairment, or both, that had that effect is a question of fact and is to be determined by the jury on the balance of probabilities.

(3) Until the contrary is proved, it is presumed that a defendant did not have a mental health impairment or cognitive impairment, or both, that had that effect.

(4) In this Part, act includes—

(a) an omission, and

(b) a series of acts or omissions.

45 Section 30 provides that a jury must return a special verdict of “act proven but not criminally responsible” if the jury is satisfied that the defence of mental health impairment or cognitive impairment has been established, and s 33 provides that such a special verdict has the following effect:

33 Effect of special verdict

(1) On the return of a special verdict of act proven but not criminally responsible, the court may make one or more of the following orders—

(a) an order that the defendant be remanded in custody until a further order is made under this section,

(b) an order that the defendant be detained in the place and manner that the court thinks fit until released by due process of law,

(c) an order for the unconditional or conditional release of the defendant from custody,

(d) other orders that the court thinks appropriate.

(2) Before making an order for the release of a defendant, the court may request a report by a forensic psychiatrist or other person of a class prescribed by the regulations, who is not currently involved in treating the defendant, as to the condition of the defendant and whether the release of the defendant is likely to seriously endanger the safety of the defendant or any member of the public.

(3) The court must not make an order for the release of a defendant unless it is satisfied, on the balance of probabilities, that the safety of the defendant or any member of the public will not be seriously endangered by the defendant’s release.

46 Further, s 34 provides that the person found to have committed an offence for which he or she was not criminally responsible is, unless released unconditionally, to be referred by the court to the Mental Health Review Tribunal.

47 The term “mental health impairment” is defined in s 4, as follows:

4 Mental health impairment

(1) For the purposes of this Act, a person has a mental health impairment if—

(a) the person has a temporary or ongoing disturbance of thought, mood, volition, perception or memory, and

(b) the disturbance would be regarded as significant for clinical diagnostic purposes, and

(c) the disturbance impairs the emotional wellbeing, judgment or behaviour of the person.

(2) A mental health impairment may arise from any of the following disorders but may also arise for other reasons—

(a) an anxiety disorder,

(b) an affective disorder, including clinical depression and bipolar disorder,

(c) a psychotic disorder,

(d) a substance induced mental disorder that is not temporary.

(3) A person does not have a mental health impairment for the purposes of this Act if the person’s impairment is caused solely by—

(a) the temporary effect of ingesting a substance, or

(b) a substance use disorder.

Legislative history

48 The 2020 Act gave effect to the recommendations of the NSW Law Reform Commission in its 2013 report People with Cognitive and Mental Health Impairments in the Criminal Justice System: Criminal Responsibility and Consequences.[52] The terms of reference required the Commission:[53]

“to undertake a general review of the criminal law and procedure applying to people with cognitive and mental health impairments, with particular regard to:

1. s 32 and s 33 of the Mental Health (Criminal Procedure) Act 1990;

2. fitness to be tried;

3. the defence of "mental illness";

4. the consequences of being dealt with via the above mechanisms on the operation of Part 10 of the Crimes (Forensic Procedures) Act 2000; and

5. sentencing.”

49 The terms of reference did not refer to the defence of automatism. The Report does not deal, at all, with the defence of automatism.[54] The focus of the report, and particularly of Chapter 3, is the defence of mental illness (which does not include sane automatism). Its relevant recommendations involved the revision and codification of the M’Naghten Rules, including clarification that cognitive impairment was included. Although noting that the historical terms of “defect of reason” and “disease of the mind” were regarded as “outdated”, “offensive” and “problematic”, far from abandoning the common law concepts relevant to mental illness, it largely retained them, “using more contemporary language”.[55] This is reflected in the following passage from the Commission’s reasons for recommending the provision which became s 28:

“3.36 We recommend that the revised M’Naghten test be incorporated into NSW law, for the following reasons.

3.37 First, we conducted extensive consultation with stakeholders on this issue and the response of those stakeholders was clearly in favour of retention and updating of the M’Naghten rules.

3.38 Second, retention would be consistent with other Australian jurisdictions and cognate jurisdictions:

• Most Australian states and territories, and the Commonwealth, have a version of the M’Naghten rules. Harmonisation of criminal laws in Australia is important and has been promoted for some time. For practitioners in NSW, consistency between NSW and Commonwealth legislation is important because they commonly work across both jurisdictions.

• Psychiatrists who are called upon to give evidence in NGMI cases noted in consultation the problems caused when the same issues are dealt with in different ways in different jurisdictions.

• Cognate jurisdictions such as NZ, the UK, Canada, the US (federal code) and the Statute of the International Criminal Court also have a version of the M’Naghten rules.

3.39 Third, the M’Naghten rules have the advantage of longevity. M’Naghten’s case was decided in 1843. There have been many opportunities for NSW to move to a different test since that time, and none of these opportunities have been taken. Although there are some strong arguments for substantial modification of the M’Naghten rules, most cognate jurisdictions have not taken a fundamentally different route in dealing with this legal issue. For example, the New Zealand Law Commission examined this issue in December 2010 and recommended no change from a M’Naghten based definition.

3.40 There is a substantial amount of existing case law that elaborates and interprets the M’Naghten rules. This case law was cited at length and with approval in submissions. Although the proposed changes would render some of the case law irrelevant, much of it would be retained.”

50 In introducing the Mental Health and Cognitive Impairment Forensic Provisions Bill 2020 on 3 June 2020, the Attorney General the Hon Mark Speakman MP said that the Bill was intended to implement the principal reforms suggested by the Law Reform Commission in its 2013 report, and in an earlier report of 2012, and that:[56]

“The reforms create a statutory definition for “mental health impairment” and “cognitive impairment” which provides ... a statutory defence of mental health impairment or cognitive impairment, previously the common law defence of mental illness; and a revised statutory special verdict of “act proven but not criminally responsible”.

51 Mr Speakman explained (emphasis added):

“Part 3 of the bill updates and legislates what was the common law test for the defence of mental illness and rewrites the special verdict from "not guilty by reason of mental illness"—or NGMI for short—to "act proven but not criminally responsible" due to mental health impairment or cognitive impairment. Currently, under the common law M'Naghten's test, a defendant can raise the defence of mental illness at common law for any criminal offence if, at the time of carrying out the relevant act, the defendant was "labouring under a defect of reason caused by a disease of the mind" and, due to that "disease of the mind", the defendant did not know the nature and quality of the act or did not know the act was wrong. The defence can be raised at a criminal trial or at a special hearing when a person has been found to be unfit to be tried. A successful defence will result in the "special verdict" of "not guilty by reason of mental illness". Very few defendants are found NGMI in New South Wales. For example, 30 people were referred to the tribunal for review following a special verdict of NGMI in 2018‑19.

Clause 28 of the bill applies to matters in the District or Supreme courts. It provides for the defence of mental health impairment or cognitive impairment, which closely mirrors M'Naghten's test but with updated terms. The statutory test of the bill enables a person to show that they were not criminally responsible for an offence if, at the time of carrying out the act constituting the offence, the person had a mental health impairment or cognitive impairment, or both, that had the effect that the person "did not know the nature and quality of the act", or "did not know that the act was wrong (that is, the person could not reason with a moderate degree of sense and composure about whether the act, as perceived by reasonable people, was wrong)". It is important to victims of those forensic patients who have successfully raised the defence of mental illness that the defence recognises that the person did do an act.”

52 There is nothing – in the Law Reform Commission’s Report, or in the Second Reading Speech – to indicate that there was the slightest intention to affect the ability of an accused person to engage a defence of sane automatism. All the indications are that that defence was not considered to be within the scope of the review, the report or the ensuing legislation.

53 Moreover, as will appear, I do not think that textually, the definition of “mental health impairment” in s 4 has that effect.

Section 4: mental health impairment

54 As has been seen, the trial judge held that the respondent did not have a mental health impairment within s 4 for two reasons: first, that he did not have a “disturbance of ... volition” within s 4(1)(a), and secondly, that any disturbance was not of clinical significance so as to fall within s 4(1)(b).

Section 4(1)(a): no disturbance of volition

55 Relevantly, s 4(1)(a) refers to “a temporary or ongoing disturbance of ... volition”. “Disturbance” means an interruption or discomposure of a settled condition or of proper functioning. Section 4(1)(a) appears in the context of a definition of “impairment”, which is a state of diminished, deteriorated, or damaged functioning.

56 In my view, an absence of volition due to being asleep involves no disturbance of a settled condition or proper functioning. Volition is the power of willing.[57] No-one has volition when they are asleep. As was said by the Full Court of the Supreme Court of South Australia in Joyce, “Everyone must go to sleep”.[58] The circumstance that a person does not have volition when asleep cannot be said to be a disturbance of an ordinary condition; it is an ordinary condition. It involves no diminution, deterioration or damage of ordinary functioning; it is ordinary functioning. There is no disturbance of volition. This conclusion is supported by, though not dependent on, the views of Dr Ellis and Dr Fernando, which were accepted by the trial judge, that parasomnias are involuntary, and not considered the product of a disordered mind because those suffering from them typically have no psychopathology when awake and conscious.[59]

57 Thus in this context, it is not correct to say that the unimpaired person has unimpaired volition; when asleep, the person without a parasomnia has just as much an absence of volition as a person with a parasomnia. That, I think, is what the trial judge intended when his Honour said that “when a person does something whilst asleep there is an absence of volition rather than a disturbance of volition. In my opinion, there is a significant difference between the two concepts, on the evidence in this particular case.” Contrary to what is suggested by Ground 2 and Question 2, his Honour did not hold that “disturbance of ... volition” in s 4(1) did not or could never include an absence of volition. His Honour said:

“[279] First, in my view a person suffering from a parasomnia such as sexsomnia, unaccompanied by other psychopathology, does not have a disturbance of volition, temporary or otherwise for the purposes of the definition in section 4. I accept Dr Ellis’s formulation, that in the ordinary sense volition is the act of making a choice, and that when a person does something whilst asleep there is an absence of volition rather than a disturbance of volition. In my opinion, there is a significant difference between the two concepts, on the evidence in this particular case. On this basis alone, I am not satisfied that the Crown has proved on the balance of probabilities that the accused had a mental health impairment.

[280] I also rely on other evidence adduced during this trial from Dr Ellis, who was a member of a committee involved in the drafting of the legislation. His unequivocal view was that the section 4 definition is used for the purpose of imposing involuntary medical treatment on individuals and that no medical institution offers medical treatment for parasomnias. In Dr Ellis’s (and Dr Fernando’s) opinion, parasomnias are involuntary, and not considered the product of a disordered mind because those suffering from them typically have no psychopathology when awake and conscious. I accept Dr Ellis’s evidence in its entirety.”

58 His Honour was there stating, not that an absence of volition could never be a disturbance of volition, but that a person who is asleep is naturally and ordinarily without volition, rather than having his or her ordinary volition disturbed. That is not a holding that an absence of a particular faculty is never a mental health impairment and cannot be caught by s 4 (and s 28) of the 2020 Act. The holding is confined to the faculty of volition in a person who is asleep. Properly understood, his Honour did not hold that a condition that results in the deterioration, to the point of complete destruction, of the relevant faculty – be it thought, mood, volition, perception, or memory – is not a disturbance for the purposes of s 4(1)(a), as in my view it would be. Thus if a condition has caused a complete loss of memory, then the memory is disturbed and diminished. However, a sleepwalker has suffered no such loss: he or she never had volition while asleep, any more or less than a non-sleepwalker. Parasomnia does not lead to a loss of volition; no-one has volition while asleep.

59 It is also a mistake to ask whether a parasomnia such as sexsomnia, unaccompanied by other psychopathology, constitutes a disturbance of volition. The labels parasomnia and sexsomnia, like somnambulism, describe the phenomena of a person performing, while asleep, various acts usually associated with willed behaviour of an awake person; they do not describe a person lacking while asleep the volition which every person lacks while asleep. The true issue is not whether sexsomnia is a mental health impairment, but whether the respondent had a disturbance of volition within s 4(1)(a). That issue is not assisted by argument over whether sexsomnia is a physiological or psychological disorder. Whatever accounts for the acts of the sleepwalker, it is not a lack of volition, but presumably the action of the subconscious mind.

60 Accordingly, in my opinion, the trial judge was right to hold that the applicant did not have a “disturbance of ... volition” within s 4(1)(a); the only intrusion on his volition was that he was asleep, which is not a “disturbance” of volition at all. Although that suffices to dispose of the appeal, I shall address the remaining grounds.

Section 4(1)(b): not significant for clinical diagnostic purposes

61 The trial judge also held that s 4(1)(b) was also not engaged. In this respect, his Honour said:[60]

“As to the requirement that the “disturbance” would be regarded as significant for clinical diagnostic purposes, I note the Attorney-General’s comments set out elsewhere in this judgment that the temporary or ongoing disturbance must be serious enough to result in a mental health diagnosis. Dr Ellis’s view was that to have significance, the disturbance needed to be accepted in the medical community as a diagnosis and that there was a need for standards around that diagnosis. In this case, the Crown relied on DSM-5 as establishing sexsomnia as a mental health diagnosis. DSM-5 is a manual that has been created to assist medical practitioners in treating patients. It includes disorders that could not possibly be classified as mental health impairments pursuant to section 4 of the MHCIFP Act. The examples of insomnia and of erectile dysfunction were noted on several occasions. In any event, I note Dr Fernando’s evidence that parasomnias were classified not only by DSM-5 but by the ICCB, and that persons other than psychiatrists treated persons with sleep disorders. Thus in my opinion, there appears to be no consensus that parasomnias result in a true mental health diagnosis or with respect to their treatment. I note the full content of the Cautionary Statement (exhibit 20). DSM-5 is not determinative of mental health impairment in the legal context.”

62 Section 4(1)(b) requires that for a “disturbance” to constitute a “mental health impairment”, it is necessary that “the disturbance would be regarded as significant for clinical diagnostic purposes”. That directs attention back to the “disturbance” referred to in s 4(1)(a). In this case, that is the disturbance of “volition”. Again, it is a mistake to focus on the condition of sexsomnia; the relevant “disturbance” is not the performance of acts while asleep, but the lack of volition while asleep. It does not require expert evidence to observe that there is absolutely no clinical significance in a person lacking volition while they are asleep, as it is a universal feature of the human condition. There may be clinical significance in the subconscious urges that propel parasomnias, but they are not within the scope of a “mental health impairment”, because they do not involve a disturbance of thought, mood, volition, perception or memory.

63 Thus on this basis also, his Honour rightly concluded that the respondent did not have a mental health impairment, because his lack of volition while asleep was of no clinical significance.

Conclusion – no mental health impairment

64 No “disturbance” of volition is involved in the absence of volition that is a universal incident of being asleep. In the context of this case, the judge did not err in finding that a somnambulist’s absence of volition is not a “disturbance of ... volition” within s 4(1). It follows that his Honour was right to hold that the respondent did not have a mental health impairment, because he did not have a disturbance of volition within s 4(1)(a), and his lack of volition while asleep was of no clinical significance for the purposes of s 4(1)(b). Grounds 1 and 2 fail.

65 I do not regard this outcome as inconsistent with the purposes of the 2020 Act, which were to contemporise and codify the law relating to the defence of mental illness, but did not include resolving any questions about the “defence” of sane automatism. Nor do I consider that the interpretation of s 4, contained in Part 1 (Preliminary) is informed by the objects of Part 5 (Forensic patients and correctional patients) stated in s 69.

Section 28: application to involuntary acts

66 In those circumstances, Ground 3, which complains that the trial judge erred in finding, if at the time of carrying out the act constituting the offence the person had a mental health impairment within s 4(1), that unconscious and/or involuntary acts cannot fall within section 28 of the Act, does not strictly arise.

67 In this respect, his Honour said:[61]

“In my opinion, it being the case that the accused performed the acts unconsciously and involuntarily whilst he was asleep – without choice or decision – means that he cannot fall within the ambit of section 28, which is reserved for those who are conscious of their actions but either do not know the nature and quality of their acts or do not know that their acts were wrong. So much has informed our law for almost 200 years. As was expressed by Mr Strickland, if Parliament meant to interfere with the long-recognised common law entitlement to acquittal in cases of involuntary acts (non-insane automatism), the principle of legality - a rule of statutory interpretation - would have required there to be clear and unambiguous words in the MHCIFP Act. No such words are found anywhere in the MHCIFP Act. I observe that the Act’s most significant change appears to be the inclusion of cognitive impairment.”

68 As it seems to me, involuntary acts resulting from a mental health impairment would be within s 28, just as insane automatism has historically attracted the application of the M’Naghten rules. I do not understand his Honour to have held to the contrary, given his Honour’s reference in the passage just quoted to “non-insane automatism”.

69 However if, contrary to my view, the respondent’s lack of volition were a mental health impairment, then it seems to me that s 28 would be engaged. Section 28 must be read with the definition of “mental health impairment” in s 4, and applies if at the time of carrying out the act alleged to constitute the offence, the person has a mental health impairment within s 4. Its application in such a case would not be affected by the circumstance that the accused performed the acts unconsciously and involuntarily.

70 To that extent, Ground 3 may have merit. However, given that the respondent has no mental health impairment, that could make no difference to the outcome in this case, and its ultimate resolution should be left to a case in which it matters.

CONCLUSION

71 Grounds 1 and 2 fail: the trial judge was right to hold that the respondent did not have a mental health impairment, because his lack of volition while asleep was not a disturbance of volition within s 4(1)(a), and was of no clinical significance for the purposes of s 4(1)(b). In those circumstances, the respondent was entitled to the outright acquittal he received. Ground 3 could not result in a different outcome. The appeal should be dismissed.

72 WILSON J: From 26 April to 30 April 2021 his Honour Judge Weinstein SC, sitting without a jury, heard evidence and submissions in the trial of the respondent before the District Court of New South Wales for child sexual assault offences. On 25 May 2021 the trial judge returned verdicts of not guilty to the offences charged on indictment and to all relevant statutory alternatives. The respondent was acquitted.

73 By Notice of Appeal filed on 22 June 2021, amended on 8 October 2021, the Crown appeals pursuant to s107(1)(b) of the Crimes (Appeal and Review) Act 2001 (NSW) (“CAR Act”) against three of the acquittals entered by his Honour, each with respect to an offence contrary to s 66DA(a) of the Crimes Act 1900 (NSW).

74 Section 107 is, relevantly for present purposes, in the following terms:

107 Directed jury acquittals or acquittals in trials without juries

(1) This section applies to the acquittal of a person—

(a) by a jury at the direction of the trial Judge, or

(b) by a Judge of the Supreme Court or District Court in criminal proceedings for an indictable offence tried by the Judge without a jury, or

(c) [...].

(2) The Attorney General or the Director of Public Prosecutions may appeal to the Court of Criminal Appeal against any such acquittal on any ground that involves a question of law alone.

[...].

75 The Crown raises the following three grounds of appeal:

(1) That the trial judge erred in failing to find that the respondent’s sexsomnia was a ‘mental health impairment’ pursuant to section 4 Mental Health and Cognitive Impairment Forensic Provisions Act 2020 (NSW).

(2) That the trial judge erred in failing to find that “disturbance of... volition” in section 4(1) Mental Health and Cognitive Impairment Forensic Provisions Act 2020 (NSW) does not include an absence of volition.

(3) That the trial judge erred in finding, if at the time of carrying out the act constituting the offence the person had a mental health impairment pursuant to section 4(1) Mental Health and Cognitive Impairment Forensic Provisions Act 2020 (NSW), that unconscious and/or involuntary acts cannot fall within section 28 Mental Health and Cognitive Impairment Act 2020 (NSW).

The Proceedings Before the District Court

76 To prevent the identification of the complainant, the respondent was referred to in the District Court proceedings by the pseudonym “DB”. For the same reason, the complainant was referred to as “CD”. Those pseudonyms have been adopted by this Court for the purposes of these proceedings.

77 The respondent was arraigned before the District Court on 26 April 2021 upon an indictment that charged him with three offences, as follows:

Count 1: On 30 June 2019, at Camden in the State of New South Wales, had sexual intercourse with [the complainant, CD], a child then under the age of 10 years, namely nine years. 66A(1) Crimes Act

Count 2: On or about 2 November 2019, at Wollongong in the State of New South Wales, had sexual intercourse with [the complainant, CD], a child then under the age of 10 years, namely, nine years. S 66A(1) Crimes Act

Count 3: On or about 2 November 2019, at Wollongong in the State of New South Wales, intentionally sexually touched [the complainant, CD], a child then under the age of 10 years, namely nine years. s 66DA(a) Crimes Act

78 He entered pleas of not guilty to each count and his trial commenced.

The Evidence

79 The trial was conducted expeditiously as the scope of the matters in dispute was limited. Other than the issue of penetration, relevant to proof beyond reasonable doubt of the element of sexual intercourse with respect to counts 1 and 2 of the indictment, the only real dispute between the parties was the issue of voluntariness as it applied in the context of a sleeping disorder from which the respondent suffered. There was no dispute that the respondent had sexually touched the complainant, his young daughter.

80 Since the issue on this appeal relates only to the acquittals entered against the statutory alternatives to counts 1 and 2, and count 3, all offences contrary to s 66DA(a) of the Crimes Act, it is not necessary to set out the evidence of the factual allegations in great detail.

81 The Crown tendered the content of a recording of an interview that had been conducted between CD and a police officer from the Joint Investigation Response Team on Sunday 3 November 2019 (“the JIRT interview”) and some very short evidence was called from her.

82 In her JIRT interview CD told the interviewing officer that she was 9 years of age and lived at two separate addresses in Wollongong, at one address with her mother and at the other with her father, DB. She described two separate occasions when her father had sexually touched her.

83 The first occasion was on a night when she was staying with her father at the house of family friends in Camden. The complainant slept with her sister and father in a large bed. The sisters went to bed at an earlier time than the respondent. CD was later woken by her father putting his hand inside her underwear and touching her vagina (count 1). The complainant thought that her father continued to touch her for as much as 45 minutes. She was crying throughout the incident but did not speak. The respondent did not say anything to CD; he was snoring, and CD thought that “he could have been asleep”.

84 Although the complainant was very upset, she did not tell anyone about what had happened.

85 CD said that on the Friday night before her interview she had been at her father’s address, an apartment located near a noisy bar. The complainant was woken by patrons of the bar at around 1am on 2 November 2019 and, frightened by the noisy people, she went into her father’s bedroom and asked him if she could sleep with him. He told her to hop into his bed, and the complainant then climbed into DB’s bed with him. The complainant tried to sleep, but her father started touching her. She said that her father “sort of rolled over” and “started touching [her] in rude places”, being her “vagina and [...] boobs”. CD said that DB put his hand inside her underwear and touched her vagina (count 2). He mumbled something, including the word “pussy”. CD told him to stop, remaining still when the respondent told her to move her leg. The respondent then moved his hand from the complainant’s underwear to her breasts and began to fondle them (count 3). CD got out of the bed, telling her father that she had to go to the bathroom. He said “okay”, and CD returned to her own room.

86 Unable to discuss what had happened with her father the complainant wrote a note that she gave to her mother, which said in part (and as written):

“[...] something HORRIBLE happend [sic]. Dad touched me on the girls part & on perpos [sic]. That was the second time this year he had done something like that to me!”.

87 CD was 11 years old when she was called to give evidence before the trial judge on 26 and 27 April 2021. She agreed in cross-examination that she was not sure if her father had put his hand or fingers inside her vagina on the two occasions she had described to police. The Crown thereafter relied upon the statutory alternative to counts 1 and 2 of sexual touching pursuant to s 66DA(a) of the Crimes Act.

88 Other evidence went to set out the family background of DB and CD and CD’s age; to establish the date of count 1, and the sleeping arrangements in place on that occasion; the complaint made by CD by note to her mother almost immediately after the occasion reflected by counts 2 and 3 and the subsequent notification to police; the level of intoxication of the respondent on the two occasions complained of; and that others regarded the respondent as an attentive father, and an honest man who was of good character.

89 The respondent’s case was that he had acted involuntarily due to a parasomnia, sexsomnia, from which he suffered. His evidence went to the truth of the history given to the specialists who had assessed his condition, and to his good character.

90 The most relevant oral evidence for present purposes was that directed to the defence raised by the respondent, at the heart of which was the question of the voluntariness of the acts of sexually touching CD, acts that the respondent did not dispute had occurred.

91 It was accepted by the Crown that the respondent was experiencing an episode of “sexsomnia” at the time of the acts and that, because of that condition, his acts were not voluntary. The dispute was as to the legal significance to be attributed to sexsomnia and whether it was caused by a mental health impairment as defined by s 4 of the Mental Health Cognitive Impairment Forensic Provisions Act 2020 (NSW) (“MHCIFP Act”), thus dictating the entry of a special verdict pursuant to s 30 of that Act; or whether it did not meet the definition, and thus an outright acquittal was indicated. The Crown argued that the condition was caught by the MHCIFP Act, whilst the respondent contended that it was not.

92 Since neither party disputed that the respondent suffered from sexsomnia, or that he had experienced episodes of it at the relevant times, it is not necessary to give a detailed account of the history given by the respondent to those medical specialists who assessed him, or of his particular circumstances. It suffices to note that the respondent had experienced “parasomnia activity” such as sleep walking when asleep over many years, and sexual partners had reported to him that, when apparently asleep, he had performed sexual acts, of which the respondent had no memory.

93 Some articles or extracts detailing research into or information concerning sexsomia were in evidence before the trial judge, including a study published in the Canadian Journal of Psychiatry, Vol 48, No 5, June 2003, in which the term “sexsomnia” was coined. That study, of 11 patients who exhibited sexual behaviours during sleep, led the authors, Dr Colin Shapiro, Dr Nikola Trajanovic, and Dr J Paul Federoff, to conclude that sexsomnia might be less rare than have previously been thought. The study noted that:

“[...] a good measure of primary and seconder (higher-order) functioning is preserved during sleep, which may give an impression that, during a parasomnic episode, something exists that could be viewed as purposeful act. However, a person experiencing a parasomnic event does not have a fully “awakened” brain – some of the cortical structures, such as those responsible for [memorising] and learning, those that help us to distinguish events from objective reality and intrinsic experiences, remain inactive, making some of the higher-order functions, including the consciousness, impaired. As an example, a person with parasomnia can walk, operate a motor vehicle, eat, perform a sexual act, or even kill, without the ability to, if we simplify, (fully) control his action. This implies that wakefulness and sleep may occur in a fragmented way and may be concurrent. The result is parasomnic behaviour with either complex motor or, as in sexomnia, motor and autonomic activity. At the same time, there is an impairment of consciousness and awareness [...]” (Footnotes and Table references omitted).

94 Whilst the aetiology of the condition was unknown the authors of the Canadian study observed that neurological disorders, brain insults or lesions, neurodegenerative disease, and genetic inheritance could all be relevant. It was suggested that, like other parasomnias, sexsomnia “lies between certain types of sleep-specific seizure disorders on one side of the continuum and dissociative psychiatric disorders on the other”.

95 Dr Anup Desai, a consultant physician in respiratory and sleep medicine, assessed the respondent in May 2020, after the respondent had been charged by police with the assaults upon his daughter, and prepared a report on 22 May 2020. He was thereafter provided with further information as to the respondent’s history of parasomnia, ultimately concluding in a report of 2 February 2021, that the history supported a conclusion that DB’s actions on the two occasions complained of “were due to sexsomnia”.

96 An assessment of the respondent’s sleep was conducted at the Sleep Centre of St Luke’s Private Hospital on 13 May 2020. He was observed to exhibit some insignificant sleep apnoea, teeth grinding, and mild limb movements. No parasomnia activity was noted, although this was not regarded as necessarily significant.

97 A report from Dr Antonio Fernando was tendered to the Court, and the doctor gave oral evidence. Dr Fernando is a consultant psychiatrist and sleep specialist in practice in New Zealand. His expertise was accepted.

98 Dr Fernando reported that parasomnias form a wide group of sleep disorders by which sufferers experience “undesirable events and sleep-related behaviours” when asleep, or immediately before or after sleep. Sexsomnia is a recognised parasomnia categorised in the Diagnostic and Statistical Manual of Mental Disorders 5th Edition (“DSM-5”) as a “non-rapid eye movement sleep arousal disorder, sleep-walking type, with sleep-related sexual behaviour”. There are six diagnostic criteria for the condition. Dr Fernando considered the respondent to have sexsomnia and other parasomnias, such as sleepwalking.

99 By reference to the DSM-5, sexsomnia, like other recognised parasomnias, including insomnia, is regarded as a “mental disorder”. However, the doctor observed that there is a debate within the medical profession as to whether parasomnias are properly categorised as mental disorders, as some in the psychiatry profession would have it, or a medical or physical disorder, as sleep specialists would suggest. Dr Fernando’s conclusion is that sexsomnia is:

“[...] an abnormality that occurs in the brain in a similar way that sleep talking, insomnia, depression, anxiety [...] are all abnormalities in a normal brain.”

100 Sexsomnia, like sleep walking is an act done at a time when the brain is unconscious or partly conscious and the sexual act performed is not voluntary. It can be triggered or worsened by alcohol consumption. Treatment with clonazepam, a benzodiazepine with sedative affects, can be effective. Lifestyle change, such as avoiding the consumption of alcohol, or sleeping alone, can be useful in reducing the occurrence of sexsomnia.

101 Whilst a significant percentage of the population will suffer from “sleep symptoms”, such as an inability to sleep, from time to time, that is to be distinguished from a disorder. A disorder, Dr Fernando deposed, “has to cause a significant impact in terms of a person’s functioning”.

102 Dr Fernando observed that the respondent’s condition was unlikely to require treatment in a hospital. He noted that persons with parasomnia are rarely treated in hospital, although they may be admitted to a sleep clinic overnight for assessment purposes.

103 Dr Andrew Ellis is a forensic psychiatrist whose report of 2 April 2021 formed part of the evidence before the District Court. In his report Dr Ellis considered the historical definitions of “defect of reason” and “disease of the mind”, concepts central to the defence of insanity. Delusional thinking in turn was central, if not essential, to a “defect of reason”. Delusional thinking was, in Dr Ellis’ opinion, “one of the key differentiating factors between an ordinary mind and a mind diseased” (AB159) and served to distinguish somnambulism, or sleepwalking, from a disease of the mind or insanity. He observed that:

“There is considerable modern epidemiological evidence that sleepwalking and related disorders are not associated with psychiatric conditions where delusions feature. Delusions are not a feature of sleepwalking or parasomnic disorders. This is a key reason why they are not currently considered within the modern concept of the disease of the mind” (footnotes omitted).

104 Further distinguishing parasomnia sleep disorders from typical mental illnesses is the nature of the treatment available for each. Dr Ellis noted that:

The treatment of the disorders is significantly different. Psychiatric hospitalisation is not a required treatment for sleep disorders. The treatment of sleep disorder is generally limited to lifestyle changes, with some medications, not those prescribed for schizophrenia, having effectiveness. In this sense it has been considered that sleepwalking is a temporary abnormality of sleep function with the ordinary mind being asleep, rather than a manifestation of an underlying disease of the mind.”

105 Forensic hospitals and the forensic treatment network do not, in Dr Ellis’ experience, treat sleep disordered patients.

106 Dr Ellis gave evidence on the issue of volition and disturbance of volition, opining that a person who does an act when asleep had not made a conscious decision to perform the act and, in that sense, there was an absence of volition. The question of whether an absence of volition was or was not a disturbance of volition was more fraught, as this evidence highlighted:

“Q: The person who engages in sleepwalking has not made a decision or a choice to sleepwalk, do you agree with that?

A: I agree, yes.

Q: So in that example, there is an absence of volition, is that right?

A: I would put it that way, yes. An absence of volition.

Q: Not a disturbance of volition, do you agree with that?

A: I think that – that becomes more definitional. An absence might be a type of disturbance, but it’s I think this is where it’s perhaps from medical to legal. But there’s certainly a difference between a disturbance or distortion of volition as opposed to it being completely absent.

Q: If there’s a distinction between a distortion and an absence, doesn’t it follow there’s also a distinction between a disturbance and an absence?

A: Yes. It could be that the absence is a sub-type of disturbance, but I think that that – this is where it becomes very difficult to define.”

107 Dr Ellis noted that, from a medical point of view, a parasomnia was not regarded as the product of a disordered mind and was not treated as a mental health impairment. The following evidence was given:

“Q: So can you explain the difference between someone suffering from a disordered mind, such as schizophrenia or suffering from delusions, and someone who is suffering from a parasomnia such as sleepwalking.

A: Well, they present very differently, delusions and the other associated symptoms of schizophrenia develop over months and years and persist in a chronic form. Even when a person is treated and well, there are usually some observable signs of behavioural change and change in a mode of thinking that – they are present and that delusional beliefs in particular are commonly persistent and will – they might not be actively thought about all the time but they – if the person’s mind is brought to those delusional thoughts, they’ll recur.

And I think that the level of functional impairment is often significant in those mental conditions. And I think in a significant contrast, the parasomnia – a person is otherwise ordinary in their day-to-day function and it’s really only during these parasomnic episodes that there’s any behavioural change. I think – there – there really, in medicine, in a [sic] very very distinct conditions that are really very different and have – require different treatment.”

108 The doctor was asked what was required for a condition to be significant for clinical diagnostic purposes and deposed:

“I’m aware it’s not further defined, but I think that it would have to have general acceptance in the medical community as a diagnosis and that there are standards around that diagnosis and that it’s of clinical utility in that there are many kinds of mental traits that might not be of any clinical interest, that there’s a clinical need for treatment in that it improves someone’s prognosis of function. So I think that that would be my understanding of what a clinical diagnostic relevance is.”

109 He agreed that a person who was asleep did not, at the time the relevant act was done, know the nature and quality of the act carried out because of a lack of conscious awareness.

The Submissions before the Trial Court

110 It is not necessary to consider the submissions with respect to proof of the element of sexual intercourse relating to counts 1 and 2. The real issue was whether the respondent should be found to be not criminally responsible for the acts he committed of sexually touching his daughter because of a mental health impairment, being the Crown’s position; or whether, as the respondent argued, he should be found not guilty on the basis that, although his sexsomnia deprived him of the capacity to choose to act as he did, and thus his acts were not voluntary, the condition did not amount to a mental health impairment, and the correct verdicts were verdicts of not guilty because the Crown could not prove that the acts were deliberate or intentional.

111 The Crown relied upon the evidence of Dr Fernando to argue that sexsomnia was an abnormality occurring in the brain that was classified by the DSM-5 as a “mental disorder” and submitted that the condition satisfied the definition of mental health impairment at s 4 of the MHCIFP Act. It also pointed to Dr Ellis’s acceptance of the proposition that a person who acted in sleep did not know the nature and quality of his or her act.

112 The respondent argued that the Crown had not proved beyond reasonable doubt that the acts of sexual touching were voluntary, and thus the Crown case failed, entitling him to acquittals. The line of authority exemplified by R v Falconer (1990) 171 CLR 30 was relied upon. Insofar as the Crown contended that sexsomnia was a mental health impairment, the respondent argued that the expert evidence did not support such a conclusion. Rather, the condition was caused by an absence of volition, rather than a disturbance or distortion of volition, and did not meet the s 4 definition in the MHCIFP Act.

The Decision of the Trial Judge

113 On 25 May 2021 the trial judge returned verdicts of not guilty to all three counts on indictment, and not guilty to the statutory alternatives to counts 1 and 2. He delivered his reasons for those verdicts: R v DB [2021] NSWDC 213 (“the verdict judgment”).

114 His Honour’s reasons are careful and detailed, but it is necessary to consider only those parts of the judgment that deal with the issues of voluntariness and the construction and application of the MHCIFP Act.

115 Having summarised the medical evidence given in the trial, his Honour considered the admissibility of evidence concerning the availability of treatment for sexsomnia, evidence which had been admitted during the proceedings only on a provisional basis. The trial judge held that the evidence was not relevant and not admissible; it was rejected.

116 On the basis of the admitted evidence his Honour concluded that the Crown had failed to prove that any of the acts complained of had been committed voluntarily. He observed at [220] that:

“Having found that the acts were involuntary, the parties agree that I must next consider the issue of mental health impairment before the mens rea: see R v Minari (2005) NSWCCA 226 and in particular the judgment of Hunt AJA at 32 and para 264-267 [...]”.

117 Having detailed the competing submissions of the Crown and the respondent his Honour concluded that the respondent was not suffering from a mental health impairment for the purposes of the MHCIFP Act. Referring to the definition of “mental health impairment” in s 4 of the Act the trial judge accepted the respondent’s submission that an absence of volition could not be regarded as a disturbance of volition, and it was not caught by s 4. He continued, at [280]:

“I also rely on other evidence adduced during this trial from Dr Ellis, who was a member of a committee involved in the drafting of the legislation. His unequivocal view was that the section 4 definition is used for the purpose of imposing involuntary medical treatment on individuals and that no medical institution offers medical treatment for parasomnias. In Dr Ellis’s (and Dr Fernando’s) opinion, parasomnias are involuntary, and not considered the product of a disordered mind because those suffering from them typically have no psychopathology when awake and conscious. I accept Dr Ellis’s evidence in its entirety.”

118 His Honour was also not satisfied that any “disturbance” would be significant for clinical diagnostic purposes.

119 The trial judge next turned to the defence of mental health impairment provided by s 28 of the MHCIFP Act, but concluded that it had no application, at [286]:

“In my opinion, it being the case that the accused performed the acts unconsciously and involuntarily whilst he was asleep – without choice or decision – means that he cannot fall within the ambit of section 28, which is reserved for those who are conscious of their actions but either do not know the nature and quality of their acts or do not know that their acts were wrong. So much has informed our law for almost 200 years. As was expressed by Mr Strickland, if Parliament meant to interfere with the long-recognised common law entitlement to acquittal in cases of involuntary acts (non-insane automatism), the principle of legality - a rule of statutory interpretation - would have required there to be clear and unambiguous words in the MHCIFP Act. No such words are found anywhere in the MHCIFP Act. I observe that the Act’s most significant change appears to be the inclusion of cognitive impairment.”

120 Having found that the Crown had failed to prove the element of sexual intercourse to the requisite standard, verdicts of not guilty were returned against counts 1 and 2. Based on his conclusion that the acts of sexual touching were not voluntary, his Honour said, at [291]:

“In my opinion, the Crown has failed to prove, on the balance of probabilities, that [the] accused, at the time of his actions in June and November 2019, suffered from a mental health impairment pursuant to section 4 of the MHCIFP Act. Further, in my opinion, the Crown has failed to prove, on the balance of probabilities, that any mental health impairment the accused may have suffered at those dates had the effect that he did not know the nature and quality of those acts or that he did not know that those acts were wrong within the meaning of section 28(1) of the MHCIFP Act. Thus that Act has no application in the circumstances, and I enter verdicts of not guilty to the statutory alternatives to counts 1 and 2 and count 3.”

The Proceedings on the Appeal

Jurisdiction

121 The jurisdiction to hear an appeal against acquittal in certain circumstances is found in Division 3 of Part 8 of the CAR Act. Section 107 imposes some limits to jurisdiction, including a temporal one, although that is satisfied in the present case, the Crown’s appeal having been filed in compliance with s 107(3), within 28 days of the entry of the acquittals. Subject to s 107(2), the appeal is brought as of right.

122 Section 107(2) of the CAR Act provides a right of appeal to the Attorney-General or the Director of Public Prosecutions “on any ground that involves a question of law alone”. The Director raises three grounds, contending that each complies with s 107(2). The respondent disputes that contention with respect to ground 1, a ground complaining that the trial judge was in error in failing to find that the respondent’s sexsomnia condition was a mental health impairment within the meaning of s 4 of the MHCIFP Act. If the respondent is correct, this Court has no jurisdiction to determine ground 1.

123 The respondent argued that the determination made in the District Court as to whether sexsomnia was caught by the definition of mental health impairment in s 4 of the MHCIFP Act was reliant upon the evidence of the presentation of the condition in his individual case, rather than of the condition more generally. In that the decision rested on evidence relevant to the respondent, his Honour’s conclusion was a factual one, and any review of it necessarily raised matters of fact. Thus, it could not be said that ground 1 involves a question of law alone, and there is no jurisdiction for this Court to determine that ground.

124 The Crown submitted that the conclusion of the trial judge that the respondent did not have a mental health impairment complained of by ground 1 rested on his Honour’s incorrect construction of a statute. The question of the correctness of the construction of a statute involves a question of law alone, and the jurisdiction granted by s 107(2) of the CAR Act is enlivened.

Determination of the Preliminary Question of Jurisdiction

125 There is no doubt that s 107(2) confines any ground of appeal to one which does not require consideration of matters of fact. Although the word “involve” in the provision appears to give some broader scope for matters other than a legal question to form part of the issue under consideration, the word that follows it in the section, “alone”, unquestionably confines the exercise to consideration of a purely legal issue. That interpretation of the legislation has been consistently applied to the provision in this Court: R v JS (2007) 175 A Crim R 108; [2007] NSWCCA 272 at;[24] – [25]; Rasic v The Queen [2009] NSWCCA 202 at [12]; R v PL [2009] NSWCCA 256 at [12] – [14]; R v XHR [2012] NSWCCA 247 at [20] – [23].

126 Although the distinction between a question involving a ground of law alone and one which mixes both law and fact is frequently opaque, I am satisfied that ground 1 falls into the former category and meets the requirement of s 107(2) of the CAR Act.

127 The finding of fact made by the trial judge was that the respondent experienced sexsomnia at the material times and, as a consequence, did not act voluntarily. That finding was not challenged at trial or before this Court. What is under challenge by ground 1 is his Honour’s construction of s 4 of the MHCIFP Act. Although it was the factual conclusion of the trial judge that the respondent did not have a mental health impairment, underlying and essential to that conclusion was the construction of s 4, a legal question.

128 It is akin to the circumstances discussed in R v JS, which Spigelman CJ distinguished from those discussed in Williams v The Queen (1986) 161 CLR; 278; [1986] HCA 88, a case on appeal from Tasmania, where the equivalent provision is differently formulated. The then Chief Justice observed, at [82]:

“In any event, Williams is distinguishable. The issue there under consideration concerned the discretionary exclusion of evidence. That is quite different from what happened in the present case, which involved separate steps. The first was her Honour's determination of the proper construction of the statutory offence. The second was her Honour's determination, on that construction, that there was no evidence capable of supporting the charge. This was not a mixed question of fact and law. There was a question of law, followed by a question of mixed law and fact.”

129 The issue raised by ground 1 is the construction of s 4 of the MHCIFP Act, the “first step” undertaken prior to consideration of whether or not the respondent’s condition met the definition as his Honour construed that definition to be, that being the “second step”. The question of the correctness of the proper construction of s 4 is a purely legal one.

130 It follows that this Court has jurisdiction to consider and determine each of the grounds of appeal pleaded by the Crown.

The Submissions of the Parties

131 The Crown argues that, by the three grounds of appeal advanced, three questions of law relating to the proper construction of the MHCIFP Act and its relationship to the common law defence of insanity are raised for determination. Those questions are:

i. Whether it is necessary that psychiatric treatment be available for a particular condition before that condition is capable of being “significant for clinical diagnostic purposes” within the meaning of section 4(1)(b) of the MHCIFP Act.

ii. Whether a “disturbance of...volition” in section 4(1)(a) of the MHCIFP Act can include an absence of volition.

iii. Whether s 28 of the MHCIFP Act can have application to unconscious or involuntary acts, which are unconscious or involuntary as a result of a mental health impairment.

132 The Crown contends that the trial judge erred in construing the MHCIFP Act, and thus in his ultimate conclusion that the correct verdicts were outright acquittals. The Crown argues that the correct verdicts ought to have been special verdicts of acts proven but the respondent not criminally responsible pursuant to s 30 of the Act.

133 The Crown accepted at trial that it could not prove beyond reasonable doubt that the respondent’s acts were voluntary. It argued, however, that the involuntary nature of them was a result of a mental health impairment and thus that the defence available under s 28 of the MHCIFP Act should be applied, and the special verdict returned.

134 The trial judge rejected that contention and held that s 28 had no application. The Crown submits that his Honour reached that conclusion, itself erroneous, because of his erroneous construction of s 4 of the Act. The Crown argues that the trial judge was guided by the common law terminology which the Act replaced and proceeded on the incorrect basis that the Act codified the common law and applied in broadly the same way. The Crown urges a different approach to properly construe s 4, with the starting point the language of the text, unencumbered by concepts imported from the common law. It argued in written submissions:

“Correctly construed, the word “disturbance”, read with regard to the context and purpose of s 4(1), must be able to accommodate a circumstance where the disturbance is sufficiently severe that the relevant capacity is said to be absent. “Volition” is one such capacity, within a class which includes thought, mood, perception and memory. It would be incongruous if a poor or impaired memory could qualify a person for a mental health impairment, but a more severe case of memory loss (in other words, an absence of memory) would not. The trial judge’s interpretation of the word “disturbance” would have the consequence of excluding a finding of mental health impairment when a mental illness completely, rather than partially, impairs a relevant capacity.”

135 On that approach it is the Crown’s submission that a complete absence of volition due to sexsomnia, as was the respondent’s condition on the two occasions relevant to the charges, would properly fall within s 4(1)(a) of the Act as a “temporary or ongoing disturbance of” volition.

136 The Crown contends that the trial judge further erred in his construction of s 4 of the MHCIFP Act by reading into s 4(1)(b) a requirement that any “disturbance” conform to the common law terminology of “disease of the mind” for which treatment was available in a hospital.

137 Referring to the lack of decided cases in which sleepwalking has been held to be insane automatism, as opposed to sane automatism, the respondent submits that a disturbance of volition refers to an impairment or a disorder of a person's mental capacity to make a conscious choice or decision, not to the complete absence of that capacity. In sleep, it is argued, there is an abeyance of consciousness.

138 The respondent argues that the common law remains relevant in understanding and applying the MHCIFP Act, submitting that:

“The common law approach to sleepwalking or automatism we say is relevant in informing whether one construes disturbance of volition in the manner that [the Crown] says. We say in Australia the common law in Australia is clear, an act performed while sleepwalking is an example of true automatism because the act is not regarded as the act of a conscious mind, that is a person who is not conscious of what he's doing. A person who is not conscious of what he's doing is not making a choice, he's not making a decision.” (T23-24, 26.11.21)

139 On that basis a complete absence of volition cannot be regarded as a disturbance of volition and thus sexsomnia could not be regarded as a mental health impairment under the Act. The respondent points to the common law principle, stated in R v Falconer (1990) 96 ALR 545; [1990] HCA 49 (per Mason CJ, Brennan and McHugh JJ), citing what was said by King CJ in R v Radford (1985) 42 SASR 266 at 272, that “It is a basic principle of the criminal law that a person is not guilty of a crime if the deed which would constitute it was not done in exercise of his will to act”.

140 The respondent argues that distinguishing between the common law concepts of sane (or non-insane) automatism and insane automatism, as the common law did, remains a function of the MHCIFP Act.

141 If the MHCIFP Act is to be construed as changing long-standing common law principle, the respondent argues that absolute clarity as to the intention to depart from accepted legal principle, affecting the rights of individuals, is required. It is submitted that clarity is not to be found in the MHCIFP Act on this issue, militating in favour of the narrow construction of s 4 given the provision by the trial judge, consistent with the common law.

142 The respondent pointed to the evidence of Dr Ellis to submit that neither ground 1 or 2 can be made good, because parasomnias do not involve any abnormality of the mind:

“Q: And do you agree that a person who engages in parasomnia, or parasomnic activity such as sleepwalking, and if that is the only relevant condition, that is not generally considered to be a product of a disordered mind or an abnormal mind?

A: No, because – and particularly because the person when they [are] awake, as they should – there’s often no psychopathology that’s able to be discerned at all.”

143 Dr Ellis deposed that, from a medical perspective, parasomnias are not regarded as a mental illness:

“From a medical point of view, the legal definitions of mental illness or mental health impairments are for the purposes of involuntary treatments and we don’t do involuntary treatment of parasomniacs. And so, from a medical point – and we don’t have, as part of our psychiatric hospitals and community psychiatric services for people with parasomnia. So, from a medical point of view, they’re not usually considered – well, they’re not considered in our current organisation as mental health impairments.”

144 With respect to ground 3, the respondent argues that the conclusion of the trial judge, that the acts were performed unconsciously and involuntarily whilst the respondent was asleep, was correct, as was his Honour’s finding that s 28 of the MHCIFP Act has no application.

145 The Court was referred to R v Youssef (1990) 50 A Crim R 1, where Hunt J, with whom Wood and Finlay JJ agreed, said, at 4 - 6:

“Automatism, however, may be relevant to two different issues in criminal trial. It may be relevant to whether the Crown has established that the actions of the accused were voluntary and, if the evidence of the existence of that condition raises a reasonable doubt as to whether the actions were of that character, the accused is entitled to be acquitted outright. Automatism may also be relevant to the defence of mental illness, in that it may establish that, as a result of a defect of a reason proceeding from a disease of the mind, the accused did not appreciate the nature and quality of his physical act (the first branch of the M’Naghten Rules), leading to the special verdict of not guilty on the ground of mental illness but not to an outright acquittal. If the effect of any disease is to impair a person’s mental faculties of reason, memory and understanding so that he does not know what he is doing, it matters not whether the impairment is permanent or transient or intermittent or whether the cause of that impairment is organic (as in epilepsy) or functional Regina v Sullivan [1984] AC 156 at 172173.”

Determination

146 This appeal centres on the proper construction and application of s 4 and s 28 of the MHCIFP Act, and the question of the relationship of those provisions to the common law. Some background as to the common law concerning mental illness, and non-insane and insane automatism is helpful.

The Common Law and the MHCIFP Act

147 The now repealed legislation, the Mental Health (Forensic Provisions) Act 1990 (NSW) (“MHFP Act”), which the MHCIFP Act comprehensively replaces, provided for the return of a special verdict of “not guilty by reason of mental illness” pursuant to s 38. The term “mental illness” used in the MHFP Act was understood by reference to the common law test or rules as given by the House of Lords in M’Naghten [1843] EngR 875; (1843) 10 Cl & Fin 200; 8 ER 718. That is:

“[...] to establish a defence on the ground of insanity, it must be clearly proved that at the time of committing the act, the party accused was labouring under such a defect of reason, from disease of the mind, as not to know the nature and quality of the act he was doing, or as not to know that what he was doing was wrong.”

148 The phrase “disease of the mind” used in the M’Naghten Rules has been the subject of much consideration over the years. In Radford, King CJ said, at 274:

“The expression ‘disease of the mind’ is synonymous, in my opinion, with ‘mental illness’ [ ...] I do not think that a temporary disorder or disturbance of an otherwise healthy mind caused by external factors can properly be regarded as disease of the mind as that expression is used in the M’Naghten rules. As Lord Denning pointed out in Bratty v Attorney-General (Northern Ireland) [[1963] AC 386 at 412] ... any ‘mental disorder which has manifested itself in violence and is prone to recur is a disease of the mind’.”

149 This statement of the meaning of the phrase has been repeatedly applied in Australia; see for example R v Porter [1933] HCA 1; (1933) 55 CLR 182; Stapleton v The Queen [1952] HCA 56; (1952) 86 CLR 358; and Fang v The Queen [2018] NSWCCA 210.

150 There was a presumption at common law that an accused person was sane and had acted voluntarily in the commission of the alleged offence. The presumptions of sanity and of voluntariness could each be rebutted by evidence.

151 In Falconer, the interplay between the presumptions concerning sanity and voluntariness, and insanity and involuntariness, was discussed, at [552]:

“The foundation for the inference that an act done by an apparently conscious actor is willed or voluntary can be removed by evidence that the actor was not of sound mind or was insane when the act was done, but there are some cases where an act can be shown to be unwilled when it is done by an actor of sound mind. To take some obvious examples: if the act be a reflex action following a painful stimulus or if it be a spastic movement, an inference that the act was willed or voluntary would not be drawn though the actor be of sound mind when the act is done. In Bratty, Lord Denning made the point, at 409:

‘No act is punishable if it is done involuntarily: and an involuntary act in this context - some people nowadays prefer to speak of it as 'automatism' - means an act which is done by the muscles without any control by the mind, such as a spasm, a reflex action or a convulsion; or an act done by a person who is not conscious of what he is doing, such as an act done whilst suffering from concussion or whilst sleep-walking. The point was well put by Stephen J. in 1889: “Can anyone doubt that a man who, though he might be perfectly sane, committed what would otherwise be a crime in a state of somnambulism, would be entitled to be acquitted? And why is this? Simply because he would not know what he was doing,” see R v Tolson ((1889) [1889] UKLawRpKQB 85; 23 QBD 168, at 187).’"

152 The opposition between a voluntary act done by a person of unsound mind and an involuntary act done by a person of sound mind was considered in Youssef, at [5]-[6]:

“This duality of relevance has led to a distinction being made in the cases between insane and sane (or non-­insane) automatism, an intellectually unhappy and artificial distinction which at times produces both unsatisfactory and inconvenient results and a distinction which is often difficult of application. The clearest statement of the distinction is that of King CJ in Regina v Radford (1985) 20 A Crim R 388 at 397 – 398:

‘The significant distinction is between the reaction of an unsound mind to its own delusions or to external stimuli on the one hand [the insane automatism] and the reaction of a sound mind to external stimuli, including stress producing factors, on the other hand [the sane automatism].

Obvious examples of the latter category may be loss of control caused by physical factors (a blow to the head or the effects of chemical substances or intoxication) or by psychological or emotional stress. Another example often given in the cases is that of a sleepwalker; it appears to have been given for the first time in Regina v Tolson [1889] UKLawRpKQB 85; (1889) 23 QBD 168 at 187’.

[...]

Those propositions were authoritatively stated by the (Irish) House of Lords in Bratty v AG for Northern Ireland at 403­404, 410, 415, 417-­419. Many of the other authorities are collected by King CJ in Regina v Radford at 395. The distinction is justified in both Bratty (per Lord Denning at 410) and Sullivan (per Lord Diplock at 172) upon the basis that, if the person who commits the crime suffers from a disease of the mind which caused him to do so, he must be kept secure so as not to be a danger to others. The distinction between sane and insane automatism is well enshrined in the law of this State: Regina v Tsigos [1964­ 51 NSWR 1607 at 1608­-1609, 1620­-1621, 1629­-1630; Regina v S [1979] 2 NSWLR 1 at 56­-60. It is now beyond this Court's proper function to depart from that distinction and, until the High Court or a legislature intervenes to remove it, the distinction must continue to be regarded as authoritative.”

153 In his consideration of the respondent’s case at trial, the trial judge regarded these and like statements of the common law as of continuing relevance in construing the MHCIFP Act.

154 The MHCIFP Act followed the release in May 2013 of a report by the NSW Law Reform Commission entitled “People with Cognitive and Mental Health Impairments in the Criminal Justice System: Criminal Responsibility and Consequences”. Chapter 3 of the report dealt with the defence of mental illness. The report, which recommended the revision and codification of the M’Naghten Rules, was critical of the common law terms of “defect of reason” and “disease of the mind” as outdated, problematic, and offensive to some sectors of society. New legislation, drawing on similar interstate legislation and abandoning the common law concepts relevant to mental illness, was proposed.

155 In introducing the Mental Health Cognitive Impairment Forensic Provisions Bill 2020 to Parliament on 3 June 2020, the Attorney General and Minister for the Prevention of Domestic Violence noted that the new legislation was intended to recognise the different legal response required to the commission of crime by persons with a mental health or cognitive impairment, than to those without impairment; take into account the safety and experiences of victims of crime; and prioritise the safety of the community. The Attorney said that the Bill was intended to implement the principal reforms suggested by the Law Reform Commission in its 2013 report, and in an earlier report of 2012. Mr Speakman continued:

“The reforms create a statutory definition for “mental health impairment” and “cognitive impairment” which provides [...] a statutory defence of mental health impairment or cognitive impairment, previously the common law defence of mental illness; and a revised statutory special verdict of “act proven but not criminally responsible” [...]”.

156 Later in his address to the Parliament introducing the Bill the Attorney said:

“Part 3 of the bill updates and legislates what was the common law test for the defence of mental illness and rewrites the special verdict from "not guilty by reason of mental illness"—or NGMI for short—to "act proven but not criminally responsible" due to mental health impairment or cognitive impairment. Currently, under the common law M'Naghten's test, a defendant can raise the defence of mental illness at common law for any criminal offence if, at the time of carrying out the relevant act, the defendant was "labouring under a defect of reason caused by a disease of the mind" and, due to that "disease of the mind", the defendant did not know the nature and quality of the act or did not know the act was wrong. The defence can be raised at a criminal trial or at a special hearing when a person has been found to be unfit to be tried. A successful defence will result in the "special verdict" of "not guilty by reason of mental illness". Very few defendants are found NGMI in New South Wales. For example, 30 people were referred to the tribunal for review following a special verdict of NGMI in 2018‑19.

Clause 28 of the bill applies to matters in the District or Supreme courts. It provides for the defence of mental health impairment or cognitive impairment, which closely mirrors M'Naghten's test but with updated terms. The statutory test of the bill enables a person to show that they were not criminally responsible for an offence if, at the time of carrying out the act constituting the offence, the person had a mental health impairment or cognitive impairment, or both, that had the effect that the person "did not know the nature and quality of the act", or "did not know that the act was wrong (that is, the person could not reason with a moderate degree of sense and composure about whether the act, as perceived by reasonable people, was wrong)". It is important to victims of those forensic patients who have successfully raised the defence of mental illness that the defence recognises that the person did do an act.”

157 The MHCIFP Act was given assent on 23 June 2020 and commenced on 27 March 2021. By s 167(a) the MHFP Act was repealed.

Sections 4(1) and 28 of the MHCIFP Act

158 Section 4(1) of the MHCIFP Act provides the definition of a “mental health impairment”, relevantly, as follows:

4 Mental health impairment

(1) For the purposes of this Act, a person has a mental health impairment if—

(a) the person has a temporary or ongoing disturbance of thought, mood, volition, perception or memory, and

(b) the disturbance would be regarded as significant for clinical diagnostic purposes, and

(c) the disturbance impairs the emotional wellbeing, judgment or behaviour of the person.

159 Section 28 provides a defence to a criminal charge where the person has a mental health impairment:

28 Defence of mental health impairment or cognitive impairment

(1) A person is not criminally responsible for an offence if, at the time of carrying out the act constituting the offence, the person had a mental health impairment or a cognitive impairment, or both, that had the effect that the person—

(a) did not know the nature and quality of the act, or

(b) did not know that the act was wrong (that is, the person could not reason with a moderate degree of sense and composure about whether the act, as perceived by reasonable people, was wrong).

(2) The question of whether a defendant had a mental health impairment or a cognitive impairment, or both, that had that effect is a question of fact and is to be determined by the jury on the balance of probabilities.

(3) Until the contrary is proved, it is presumed that a defendant did not have a mental health impairment or cognitive impairment, or both, that had that effect.

(4) In this Part, act includes—

(a) an omission, and

(b) a series of acts or omissions.

160 In interpreting each of s 4 and s 28 the trial judge proceeded on the basis urged upon him by the respondent that the common law that applied to a mental illness defence and, importantly, a defence of non-insane automatism, remained of relevance in interpreting and applying the MHCIFP Act.

161 Grounds 1 and 2 both focus on the interpretation of s 4 that was applied to the evidence by the trial judge, with the Crown contending for error in his Honour’s construction of the provision. As the parties did before this Court, it is convenient to address these grounds together. Ground 3 will be addressed separately.

Ground 1: That the trial judge erred in failing to find that the respondent’s sexsomnia was a ‘mental health impairment’ pursuant to section 4 Mental Health and Cognitive Impairment Forensic Provisions Act 2020.

Ground 2: That the trial judge erred in failing to find that “disturbance of... volition” in section 4(1) Mental Health and Cognitive Impairment Forensic Provisions Act 2020 (NSW) does not include an absence of volition.

162 Central to the determination of the trial judge that the respondent did not have a mental health impairment was his conclusion, guided by common law principles, that “a temporary or ongoing disturbance of [...] volition” referred to in s 4(1)(a) did not include an absence of volition. His Honour was of the view that a parasomnia such as sexsomnia, being unaccompanied by other psychopathology, could not constitute a disturbance of volition.

163 That conclusion rests on the meaning given by his Honour to the phrase “disturbance of volition” in s 4(1)(a).

164 In construing the provision, it is the text of it which is important. In Alcan (NT) Alumina Pty Ltd v Commissioner of Territory Revenue (2009) 239 CLR 27; [2009] HCA 41, French CJ said, at [4]:

“The starting point [...] is the ordinary and grammatical sense of the statutory words to be interpreted having regard to their context and the legislative purpose.”

165 At [47] of the same case Hayne, Heydon, Crennan and Kiefel JJ observed in similar vein, at [47]:

“This Court has stated on many occasions that the task of statutory construction must begin with a consideration of the text itself. Historical considerations and extrinsic materials cannot be relied on to displace the clear meaning of the text. The language which has actually been employed in the text of legislation is the surest guide to legislative intention. The meaning of the text may require consideration of the context, which includes the general purpose and policy of a provision, in particular the mischief it is seeking to remedy” (footnotes omitted).

166 Section 33 of the Interpretation Act 1987 (NSW) provides:

33 Regard to be had to purposes or objects of Acts and statutory rules

In the interpretation of a provision of an Act or statutory rule, a construction that would promote the purpose or object underlying the Act or statutory rule (whether or not that purpose or object is expressly stated in the Act or statutory rule or, in the case of a statutory rule, in the Act under which the rule was made) shall be preferred to a construction that would not promote that purpose or object.

167 A purposive construction is ordinarily one which is determined from the text and structure of the statute: Certain Lloyd’s Underwriters Subscribing to Contract No 1H00AAQS v Cross (2012) 293 ALR 412; [2012] HCA 56, at [25] – [26] per French CJ and Hayne J; Project Blue Sky v Australian Broadcasting Authority (1998) 194 CLR 355; [1998] HCA 28; at [78]; Commissioner of Taxation v Consolidated Media Holdings Ltd (2012) 250 CLR 503; [2012] HCA 55, at [39].

168 Section 4(1)(a) deals with impairment, that is, a condition that is different to the undamaged operation of “thought, mood, volition, perception or memory”. The unimpaired person, who would not be subject to the operation of the MHCIFP Act, has unimpaired volition (or thought, mood, perception, or memory). That is, the person without a relevant impairment has the capacity to make and exercise choice or free will, to act with deliberation. If the unimpaired condition is to act deliberately as a consequence of choice, to be without that capacity due to some irregularity in ordinary function must, in my conclusion, be a state of impairment or disturbance of volition. If the text of s 4(1)(a) is viewed without regard to historical considerations, as the authorities dictate, a “disturbance of volition” must include volition that is so disturbed as to be in abeyance altogether at the material time.

169 Comparative analogy can be used as a means of assessing the correctness of that conclusion. It would be an odd and surely unintended outcome if a person with a condition leading to the complete absence of memory was regarded as unimpaired, whilst a person with some but incomplete memory was considered to be impaired.

170 “Disturbance” simply means an interruption or discomposure of a settled condition or of proper functioning.[62] The word should not be treated as the statutory equivalent of “disease of the mind”; there is no basis in the text or in the objects of the legislation to confine the meaning of a disturbance of a relevant faculty in such a way.

171 A condition that leads to an absence of the relevant faculty – thought, mood, volition, perception, or memory – must be a disturbance for the purposes of s 4(1)(a). Were it otherwise, those persons who would under the common law have had available a mental illness defence on the basis of insane automatism, would not have that defence available under the MHCIFP Act, and there could be no regulation of conduct and mandated treatment of such persons.

172 If that construction is viewed in the context of the purpose of the MHCIFP Act, and specifically Part 3 of the Act, it holds good. Part 3 of the Act governs the treatment of impaired persons who commit a physical act that, when accompanied by the relevant mens rea, would constitute a crime. Consideration of the provisions within Part 3 points to its purpose: to provide for a means of recognising the commission of the physical act by the accused person; to acknowledge the absence of criminal responsibility due to impairment; and to empower a court and specialist tribunal to regulate the person’s future conduct as appropriate to ensure the safety of members of the public.

173 A conclusion that an absence of the particular faculty is a disturbance for the purposes of s 4(1)(a) is consistent with the purpose of Part 3 of the MHCIFP Act; the opposite conclusion is not since it defeats the purposes of the Act.

174 It is clear from the evidence in this matter, including both that of the complainant and of the literature concerning sexsomnia, that a person acting without volition due to a parasomnia can commit acts which, if carried out deliberately, would constitute serious crime. A conclusion that an absence of a particular faculty, here volition, is not a mental health impairment and not caught by s 4 (and s 28) of the MHCIFP Act, prevents the achievement of the purposes of the Act. On that construction, there can be no regulation of the conduct and treatment of persons who commit a physical act that would ordinarily constitute a crime, and no means by which the safety of members of the public can be protected.

175 Viewing the provision in light of the purpose of the Part and of the Act supports a textual construction of it.

176 Section 4(1)(b) provides the second limb of the definition of “mental health impairment”, that “the disturbance would be regarded as significant for clinical diagnostic purposes”. The trial judge approached the interpretation of this part of the definition by reference to the distinction between psychiatric and non-psychiatric disorders at [280]-[281], his Honour said:

“I also rely on other evidence adduced during this trial from Dr Ellis, who was a member of a committee involved in the drafting of the legislation. His unequivocal view was that the section 4 definition is used for the purpose of imposing involuntary medical treatment on individuals and that no medical institution offers medical treatment for parasomnias. In Dr Ellis’s (and Dr Fernando’s) opinion, parasomnias are involuntary, and not considered the product of a disordered mind because those suffering from them typically have no psychopathology when awake and conscious. I accept Dr Ellis’s evidence in its entirety.

As to the requirement that the “disturbance” would be regarded as significant for clinical diagnostic purposes, I note that Attorney-General’s comment set out elsewhere in this judgment that the temporary or ongoing disturbance must be serious enough to result in a mental health diagnosis. Dr Ellis’s view was that to have significance, the disturbance needed to be accepted in the medical community as a diagnosis and that there was a need for standards around that diagnosis. In this case, the Crown relied on DSM-5 as establishing sexsomnia as a mental health diagnosis. DSM-5 is a manual that has been created to assist medical practitioners in treating patients. It includes disorders that could not possibly be classified as mental health impairments pursuant to section 4 of the MCHIFP Act. The examples of insomnia and or erectile dysfunction were noted on several occasions. In any event, I note Dr Fernando’s evidence that parasomnias were classified not only by DSM-5 but by the ICCB, and that persons other than psychiatrists treated person with sleep disorders. Thus in my opinion, there appears to be no consensus that parasomnias result in a true mental health diagnosis or with respect to their treatment. I note the full content of the Cautionary Statement (exhibit 20). DSM-5 is not determinative of mental health impairment in the legal context.”

177 In taking this approach to the construction of the provision his Honour appears to have been swayed by the evidence of the experts called at trial concerning a divergence of opinion within the medical profession as to whether parasomnias were physiological or psychiatric disorders. His Honour clearly gave great weight to the opinion of Dr Ellis in that regard and accepted the doctor’s medical opinion that parasomnias were not mental illnesses.

178 That, however, is not the legal test established by s 4(1)(b); the provision does not rest upon whether members of the medical profession attach a particular label or classification to a condition. The test is to be found in the text, the meaning of the words used in s 4(1)(b), in the context of the purpose of the statute.

179 As with the construction of s 4(1)(a) there is nothing in the text of s 4(1)(b) that would confine its meaning in the way in which it was confined by the trial judge. There is no basis to interpret the provision through the lens of the medical profession as to whether a condition is “the product of a disordered mind” or is accompanied by “psychopathology” or “delusions”. Those are considerations relevant to whether an individual has a disease of the mind, but that phrase was swept away by the MHCIFP Act for the purposes of the operation of the Act.

180 The aetiology of a condition will continue to be of relevance to clinicians, but it is immaterial when applying what is a legal definition. Dr Ellis gave evidence to the effect that parasomnias are not treated in psychiatric hospitals, and this evidence was regarded as significant by the trial judge, who referred to it - although by reference to treatment offered in a “medical institution” rather than a psychiatric institution - at [280] of his judgment.

181 Although his Honour ultimately excluded the evidence concerning treatment that was provisionally admitted on this point, he clearly had regard to the absence of treatment in a psychiatric facility for parasomnias in determining the application of s 4. This is not, however, a relevant consideration when applying s 4(1)(b). Properly construed, s 4(1)(b) refers in my conclusion to a “disturbance” or recognised condition with standard diagnostic criteria by which to diagnose it, where the symptoms are sufficiently significant to warrant a diagnosis being made. Considerations of the aetiology of the disorder, and whether it is a disease of the mind amenable to psychiatric treatment, have no role to play in s 4(1)(b).

182 The trial judge appears to have adopted what was a medical view expressed by Dr Ellis concerning the unavailability of psychiatric treatment of parasomnia in psychiatric facilities to conclude that “the section 4 definition is used for the purpose of imposing involuntary medical treatment on individuals and that no medical institution offers medical treatment for parasomnias” (at [280] of his Honour’s judgment). There is nothing in the Act to support that conclusion as to the purpose of s 4.

183 Part 3 of the Act provides for differing outcomes where a special verdict is returned. Section 33 sets out the effect of the return of a special verdict:

33 Effect of special verdict

(1) On the return of a special verdict of act proven but not criminally responsible, the court may make one or more of the following orders—

(a) an order that the defendant be remanded in custody until a further order is made under this section,

(b) an order that the defendant be detained in the place and manner that the court thinks fit until released by due process of law,

(c) an order for the unconditional or conditional release of the defendant from custody,

(d) other orders that the court thinks appropriate.

(2) Before making an order for the release of a defendant, the court may request a report by a forensic psychiatrist or other person of a class prescribed by the regulations, who is not currently involved in treating the defendant, as to the condition of the defendant and whether the release of the defendant is likely to seriously endanger the safety of the defendant or any member of the public.

(3) The court must not make an order for the release of a defendant unless it is satisfied, on the balance of probabilities, that the safety of the defendant or any member of the public will not be seriously endangered by the defendant’s release.

184 It is noted that, for the purposes of s 33(2) the regulations provide for a report to be prepared by a registered psychologist who has, in the opinion of the court, appropriate experience or training in forensic psychology or neuropsychology. Significantly, report providers are not confined to psychiatrists.

185 Section 34 provides for the person found to have committed an offence for which he or she was not criminally responsible to be referred by the court to the Mental Health Review Tribunal, unless released unconditionally. The Tribunal has wide powers on reviewing a person with respect to whom a special verdict was returned, powers not confined to ordering the detention of the person in a psychiatric facility, whether custody or community based.

186 Section 69 of the Act gives the objects of Part 5 of the Act relevant to the treatment of those persons found not criminally responsible for the commission of a crime due to a mental health impairment. Those objects are much wider than to achieve the involuntary psychiatric treatment of such individuals, although such an outcome is clearly available under the legislation. The objects of Part 5 include the protection of the safety of members of the public (s 69(a)) the provision of an opportunity for those persons found not criminally responsible pursuant to s 28 to have access to appropriate care (s 69(e)), and to protect the safety of victims of forensic patients and acknowledge the harm done to victims (s 69(f)).

187 The definition of mental health impairment is also significant to the operation of other parts of the MHCIFP Act that do not result in involuntary psychiatric admission, such as the diversionary orders a magistrate is empowered to make pursuant to s 14 of the Act.

188 It is thus incorrect to view the purpose of s 4(1)(b) as preliminary only to the imposition of involuntary psychiatric institutionalisation and treatment.

189 The text of s 4(1), viewed within the context of the purpose of the Act, has wider application than does the common law concept of mental illness. The respondent argued before both the trial court and this Court that, in the absence of a clear statement of an intention to alter long-standing common law principle, it cannot be concluded that those principles have been abrogated. However, the Parliament did not, in introducing the legislation, simply codify the M’Naghten Rules, as was open to it. The statutory definition of mental health impairment is not the same as the common law concept of insanity or mental illness, and there is no basis to read the latter into the former.

190 The principle of legality does not operate to read down or artificially confine the clear meaning of a statutory text. The principle was discussed by the Court of Appeal (per Bathurst CJ, Beazley P, Basten JA, Gleeson JA, and Payne JA) in Secretary, Department of Family and Community Services v Hayward (a pseudonym) [2018] NSWCA 209, at [28] – [35], with the Court observing “The principle of legality, also known as the clear statement rule, does not follow its own injunction” (at [28]).

191 In Lee v New South Wales Crime Commission (2013) 251 CLR 196; [2013] HCA 39, Gageler and Keane JJ said:

“Application of the principle of construction is not confined to the protection of rights, freedoms or immunities that are hard-edged, of long standing or recognised and enforceable or otherwise protected at common law. The principle extends to the protection of fundamental principles and systemic values. The principle ought not, however, to be extended beyond its rationale: it exists to protect from inadvertent and collateral alteration rights, freedoms, immunities, principles and values that are important within our system of representative and responsible government under the rule of law; it does not exist to shield those rights, freedoms, immunities, principles and values from being specifically affected in the pursuit of clearly identified legislative objects by means within the constitutional competence of the enacting legislature.”

192 On my reading of the MHCIFP Act, its object was not to simply translate the common law into statute; it was to codify and modernise the law that relates to persons with a mental health or cognitive impairment, taking “into account the safety and experiences of victims, as well as [to] prioritise the safety of the community”. So much is plain from the text of the legislation, even setting aside what was said by the Attorney in the Second Reading Speech on 3 June 2020.

193 The principle of legality cannot, in my opinion, be relied upon to read the MHCIFP Act as no more than the codification of the common law. The legislation is broader than that.

194 I am persuaded that there was error in the conclusions of the trial judge that led to him construing the meaning of s 4(1)(a) and (b) consistent with the common law concepts of a defect of reason caused by a disease of the mind, rather than in accordance with the plain meaning of the words, consistent with the purpose of the statute.

195 His Honour was not asked to, and did not, consider s 4(1)(c).

196 If a textual and purposive construction is applied to s 4(1), the evidence establishes that, on the two occasions when he sexually touched his 9-year-old daughter, the respondent had a temporary disturbance of volition (s 4(1)(a)); his condition was regarded as clinically significant in that it met the diagnostic criteria for the disorder of sexsomnia (s 4(1)(b)); and the disturbance impaired his behaviour (s 4(1)(c)).

197 I would uphold grounds 1 and 2.

Ground 3: That the trial judge in erred in finding, if at the time of carrying out the act constituting the offence the person had a mental health impairment pursuant to section 4(1) Mental Health and Cognitive Impairment Forensic Provisions Act 2020, that unconscious and/or involuntary acts cannot fall within section 28 of the Act

198 Although it was not strictly necessary to do so, given the construction he had placed upon s 4(1) of the Act, his Honour went on to consider whether the s 28 defence would have been established had he concluded that sexsomnia was a mental health impairment. He held that it would not, explaining, at [284] – [286] of the verdict judgment:

“In my opinion section 28 codifies the M’Naghten test. I note that the Attorney-General, in the Second Reading Speech, said: -

‘Clause 28 of the bill applies to matters in the District or Supreme courts. It provides for the defence of mental health impairment or cognitive impairment, which largely mirrors the test but with updated terms. The statutory test of the bill enables a person to show that they were not criminally responsible for an offence if, at the time of carrying out the act constituting the offence, the person had a mental health impairment or cognitive impairment, or both, that had the effect that the person “did not know the nature and quality of the act”, or “did not know that the act was wrong (that is, the person could not reason with a moderate degree of sense and composure about whether the act, as perceived by reasonable people, was wrong).”’

So much is clear from the language of section 28, which in fact picks up material parts of M’Naghten’s test and updates some language: see also Anina Johnson Forensic Provisions Act Amendments, [2021] (Autumn) Bar News at 22 and Justice Mark Ierase [sic] Introducing the New Mental health and cognitive Impairment Forensic Provisions Act 2020 (2021) 33(2) JOB 15 at 17.

In my opinion, it being the case that the accused performed the acts unconsciously and involuntarily whilst he was asleep – without choice or decision – means that he cannot fall within the ambit of section 28, which is reserved for those who are conscious of their actions but either do not know the nature and quality of their acts or do not know that their acts were wrong. So much has informed our law for almost 200 years. As was expressed by Mr Strickland, if Parliament meant to interfere with the long-recognised common law entitlement to acquittal in cases of involuntary acts (non-insane automatism), the principle of legality - a rule of statutory interpretation - would have required there to be clear and unambiguous words in the MHCIFP Act. No such words are found anywhere in the MHCIFP Act. I observe that the Act’s most significant change appears to be the inclusion of cognitive impairment.”

199 In so saying, his Honour was in error in his understanding that the common law defence of insanity or mental illness had only ever applied to those who had conscious awareness of the nature and quality, and moral wrongness of their act. The common law defence of insane automatism stands for the contrary proposition: see Woodbridge v The Queen [2010] NSWCCA 185; (2010) 208 A Crim R 503, and those cases quoted therein at [65] – [69].

200 His Honour was also in error, in my opinion, in giving greater weight to the principle of legality than to the text and purpose of s 4 and s 28 of the MHCIFP Act.

201 The two sections must be read together. Taking that approach, it follows from the conclusion that s 4(1) operates such that sexsomnia is a mental health impairment, that a person who commits a criminal act when impaired by the condition could not know the nature and quality of the act, or that it was wrong. In this instance, his impairment prevented the respondent from considering those questions and it must be concluded that he did not know at the time he did the acts their nature and quality, or that they were wrong.

202 I would uphold Ground 3.

Conclusion

203 I propose that the following orders are made:

(1) Each of grounds 1, 2, and 3 be upheld;

(2) The respondent’s acquittals for three offences pursuant to s 66A(1) are set aside;

(3) A retrial of the respondent is ordered for the statutory alternatives to counts 1 and 2, and count 3.

204 IERACE J: I agree with Brereton JA.