R v Klamo [2008] VSCA 75

The applicant told police he knew he was in the wrong and that it was a dangerous and stupid thing to do.

13 After the shaking, about which the applicant said nothing to Ms Sowter, the baby's behaviour was normal. The applicant said that there was a 'check up' (with maternal and child care nurses) probably four days after he shook the child. The applicant was then asked:

Q163 And how had he been over the last week? Was he the same baby that he was ...?

Q167 Did you do anything else around - touch his head - in any other way?

A Just pat his head, that's about it. Bathed him ...

14 The applicant was asked why he did not tell Ms Sowter. He responded:

... I couldn't, because I was feeling guilty what I'd done, and I realised what I done, and never again. And that's my son I was fuckin' with and I shouldn't have done it.

15 The applicant denied having ever hit his son. He continued:

Q313/314 I would not dare to do that. No way. That was the furthest I ever went, and that - you know. Yeah, just ... shaking him. That was the furthest I have ever done, and I can say that out loud. Honestly, everything. I would - wouldn't take it - you know, a mate of mine's son - I would not - I would not, no.

**16 Professor Cordner concluded that the cause of death was a subdural haemorrhage. The bleeding which had caused the death 'was fresh and consistent with having occurred some time in ... the 24 hours before death.' Professor Cordner also observed that there had been an earlier haemorrhage, which he estimated had occurred 'a couple of weeks previously'. These conclusions were not challenged.

17 Professor Cordner identified three possible causes of the fatal haemorrhage. The first was 'a spontaneous bleeding, a phenomenon which is regarded as occurring with minimal or no trauma after a previous subdural haemorrhage.' The second possible cause was direct trauma - 'the head hitting something'. The third was indirect trauma, 'that is, perhaps possibly by shaking.'

18 As to a spontaneous re-bleed, Professor Cordner noted that there had been fresh bleeding in three separate areas of the brain. Because the two hemispheres of the brain are separate, bleeding around one hemisphere would not normally pass into the other. This led Professor Cordner to express his conclusion in these terms:**

... I cannot stand here and rule out the possibility of a spontaneous nature to these haemorrhages, but I think because they occur in three separate spots that I feel quite comfortable myself that it is not a spontaneous re-bleed, but from a factual point of view, I can't rule it out.

19 As to the possibility of direct trauma, Professor Cordner said that there was no bruising to the scalp to indicate direct trauma, but this did not completely exclude the possibility.

The recent haemorrhage could have been precipitated by trauma. If more severe direct blunt trauma to the head occurred there is no sign of it, there is no bruising in the scalp. Significant head injury, however, can occur without obvious scalp bruising and one hits one's head against a firm but not hard material such as firm furniture, particularly over a broad area, you could imagine having a significant impact to the head without necessarily bruising what is very tough tissue, the scalp.

**20 Under cross-examination, Professor Cordner agreed that if a child was suspected of having been struck, a medical examiner would look for a bruise or a fracture, and that neither was present in this case. Small bruises on the cheek had been noted but they did not, in Professor Cordner's view, suggest that the child had been 'obviously and severely struck'.

21 As to the possibility of shaking having caused the fatal haemorrhage, Professor Cordner said:**

... [O]bviously from my point of view, anyway, the possibility of a shaking event shortly before death exists, but there are no bruises to the arms or chest or retinal haemorrhages to support this as a real possibility.

Also there were no bruises to the arms or chest that are sometimes seen in association with a shaking indicating, as they might, gripping of the baby by the arms or chest. Also there were no haemorrhages or bleeding in the eyes, and these are regard[ed] as commonly accompanying shaking, the sort of shaking that results in the sort of haemorrhage that we have seen here.[1]

22 Under cross-examination, Professor Cordner was questioned about competing views in the field of paediatrics and forensic pathology about the link between shaking and subdural haemorrhage. Professor Cordner said:

I don't think in middle of the road forensic pathology there is actually real disagreement about the fact that shaking is a possible cause of subdural haemorrhage. Where the controversy lies is in the mechanism by which that occurs. And that goes to the question, is it always the case that the shaking has to be so severe that the person doing it would have some understanding that some harm might result. So the controversy has been in the area of how is it that shaking causes subdural haemorrhage.[2]

23 Professor Cordner agreed that there was 'a lot of controversy' about what level of force was required for a shaking to cause serious injury. He also agreed that the shaking which the applicant demonstrated in the course of his video record of interview was such that no expert could say whether the force used would, or would not, be sufficient to cause subdural haemorrhage. Professor Cordner also agreed with the following proposition put to him by defence counsel:

Fatal cases of the shaken baby syndrome are not likely to occur from the shaking that occurs during play, feeding or in a swing or even from the more vigorous shaking given by a caretaker as a means of discipline.

**24 The effect of Professor Cordner's evidence as to the three possible causes of the fatal haemorrhage can be summarised as follows:

(a) spontaneous re-bleed: Professor Cordner was satisfied that this was not the cause, but he could not rule it out;

(b) direct trauma: there was no bruising to indicate that this had occurred, but that did not mean the possibility could be excluded altogether;

(c) a shaking event: there were no bruises to the arms or chest or retinal haemorrhages to support this as 'a real possibility', but it could not be excluded altogether.**

25 The case against the applicant was that he was guilty of manslaughter by unlawful and dangerous act. The Crown case was that the relevant act was a shaking of the baby just before 3 o'clock on the morning of the death. This was an assault and, hence, an unlawful act. As to dangerousness, the prosecutor submitted to the jury that a reasonable person in the applicant's position

must necessarily have realised that the act or actions could expose the child to an appreciable risk of serious injury. Where you have an adult of a fair size shaking an infant between 14 and 12 days old or up to 28 days, vigorously shaking, that is a conclusion which I say to you you could readily draw. A reasonable person much necessarily appreciate that there is an appreciable risk of serious injury to a child if you do that.

26 The Crown rejected the possibility of a spontaneous re-bleed. The prosecutor told the jury:

The Crown clearly says there is only one cause here and that's the shaking ... [T]he burden of proof is on the prosecution here to satisfy you beyond reasonable doubt that [it] is the shakings that caused the bleeding and ultimately death as a result.

27 As to the re-bleed possibility, the prosecutor told the jury:

Well, such things do happen in life, spontaneous re-bleeds can happen, but Professor Cordner says no. Whilst he acknowledges theoretically these things can happen, you would think that is commonsense, these things can happen, but the subdural haematoma that occurs in this child on the night he dies is a major bleed for a start, it is a major bleed, it is not a minor bleed, and it is in three distinct places ...

because it occurred in three different parts of the brain, it wasn't a spontaneous re-bleed at 3 o'clock in the morning. Something caused that. It wasn't spontaneous.

28 The prosecutor nevertheless addressed the possibility that the jury might conclude that a spontaneous re-bleed was 'possible'. After reminding the jury, again, that Professor Cordner had 'discounted' a re-bleed as the cause, he said:

If you accept and are satisfied beyond reasonable doubt that there was a shaking episode of such vigour on the first occasion to cause a subdural haemorrhage, and subsequently, for an unknown reason, that spontaneously re-bled and caused death, the first episode would still be a substantial cause of death in this case.

**29 The Crown's shaking case rested on the fact that Professor Cordner had not ruled out the possibility of a shaking event causing death. The prosecutor also relied on answers which Professor Cordner had given in cross-examination, to the effect that the fatal haemorrhage had probably commenced 'a very short time prior to the breathing difficulties being apparent.' Professor Cordner agreed that, if the baby had been able to feed normally shortly before the breathing difficulty becoming apparent, this was 'a pretty good sign of relative health'.

30 The jury's starting-point, the prosecutor said to them, was that the baby had been**

the subject of, at least, a vigorous, if not violent, shake within a time frame that relates to the first subdural bleed.

You might think that within an hour [of feeding normally] when he suffers this subdural bleed which causes his death, was this just a coincidence that it occurred in this way at this time, bearing in mind the previous incident, or does the subdural clearly relate to another shaking episode? I say you can infer beyond reasonable doubt that it happened again, and that's what has caused this second fatal bleed.

a fair and solid basis for you to conclude that the vigorous shaking on that first occasion caused the first subdural haematoma ... I am asking you to conclude that there was a second shake at or about [3 o'clock on the morning of the death]. There is no admission in relation to it and, indeed, a denial that any such thing ever occurred. It has to be in the early hours of the morning after that last feed. There is the opportunity, he is there, he brings the child to the mother, there is a subdural haematoma and that is a second fatal one. The medical evidence ... is compatible with that shaking event in the days prior to it, entirely consistent with it, and the second one, as Dr Cordner said, that the possibility of shaking did exist. It's not ruled out. There we have a second subdural bleed, the fatal one.

The accused did shake his baby in the timeframe related to the first subdural haematoma and the baby has died as a result of the [second] subdural haematoma.

31 More than once, the prosecutor sought to portray the applicant as someone who could not cope with the pressure of parenting. Speaking of the admitted shaking incident he said:

Being a proud father, one who does his best most of the time doesn't mean that he could cope with the pressure he faced with this crying infant. The pressure got to him. He could not cope with the crying infant. It's as simple as that. He, in his own words, said he lost his patience, he was irritated, and a little "pissed off". That's how he described his own attitude to the infant when the child was crying at the time of the first shaking episode. A man who could not cope under pressure.

Implicitly, in my view, this was an invitation to the jury to regard the applicant as a person who had a propensity to react badly when stressed by the child's crying. I deal further with the propensity issue later in these reasons.

32 The trial judge began his charge to the jury by noting that there was no direct evidence as to what had caused the fatal subdural event. As a result, the jury would have to ask themselves

whether upon the facts as you find them following your examination of the evidence you are able to draw inferences that enable you to in the end come to a verdict of guilty.

33 His Honour then distinguished between the function of the expert witness, on the one hand, and that of the jury on the other:

Professor Cordner is approaching his task from a different standpoint to yours. He is a medical practitioner in a particular field of medical practice. He is seeking to determine in medical terms the cause of death. In a sense he has done that, his evidence is that Izaiah died as a result of a subdural haemorrhage. It is not his job to tell you how that haemorrhage was caused, except to the extent that he can do so consistent with his medical expertise, and that limits his field of enquiry. He, within those limits, has formed the somewhat tentative conclusions about which you heard and to which I will return in my summary of the evidence. But remember, that his approach in coming to his conclusions, which you might think are tentative, is very different to yours. In particular, you take into account all the evidence that you have heard in this trial, in a way in which Professor Cordner would not be expected to do.

... [H]e has not heard the evidence as you have heard it, and it is not his job to make any finding at all about the guilt of Mr Klamo. That is your job, it is your job alone, it is nobody else's. So while you should properly take into consideration the points that [defence counsel] made to you, it is not sufficient simply to say Professor Cordner came to such and such a tentative conclusion and we cannot go beyond that. You can go beyond that. You examine all the evidence and if, but only if, you think that evidence points to guilt beyond reasonable doubt, then you return a verdict of guilty. Otherwise you return a verdict of not guilty. So that is the difference between your task and that of Professor Cordner.

**34 His Honour then described the Crown case as having advanced two alternative hypotheses. The first was that the 'very sudden turnaround in the baby's condition' was to be explained by a traumatic incident on the night in question. The second was that there was a causal link between the admitted shaking and the re-bleed. His Honour pointed out that other hypotheses had been advanced by the defence in cross-examination, relating the fatal haemorrhage to the child's birth or to some other disease. Those hypotheses had been rejected by Professor Cordner.

35 His Honour told the jury that, just as a death is only ascribed to Sudden Infant Death Syndrome when there is no medical explanation for it:**

So in a sense medically is Izaiah's death inexplicable because the medical evidence does not allow us to conclude with confidence that the subdural event as a medical proposition was caused by anything identifiable by the doctors. It is your job to ask whether as a jury you can identify a cause for the subdural haemorrhage and do so beyond reasonable doubt.[3]

[Y]ou are the sole judges of the facts and nothing I say about the facts or the evidence is itself evidence; and you should not be influenced in any way by any view you think I might take of the evidence. But you might think that there are three broad areas through which one can examine the causes of Izaiah's death. First, trauma. Secondly, a re-bleed from an earlier haematoma. Third, a completely spontaneous event. I have taken you to these three sets of circumstances because you are faced with the conundrum that we know something with certainty about Izaiah's death and that is that it was caused by a subdural haemorrhage. That is a conundrum because you are asked to examine the causes of that subdural event. It seems to me that as a matter of logic those causes can be divided into those three sets of circumstances; trauma, a re-bleed or a completely spontaneous event.[4]

36 When exceptions were called for, defence counsel said:

[T]he area that troubled me the most was the three broad areas that they ought to consider in approaching the matter generally. That was an area which was outside of counsel's addresses, although dealt with, of course, in a general way in it, but what I was concerned about was this: that you put trauma, one trauma, that is presumably trauma, meaning trauma immediately prior to the child becoming distressed and after the feeding; secondly, a re-bleed of an earlier haematoma and then thirdly, a completely spontaneous event. But the jury may have taken it that they ought to go for and conclude one of those three was the way in which it occurred. If that's what they did think, if that's the impression they did get, it makes it too much of a detective story and is inclined in my submission to turn it from a search for proof into a search for the actual cause. [5]

His Honour responded by saying that the jury were 'certainly not bound to form a conclusion about the death'. He said that he had intended to say something to the jury to that effect. The prosecutor argued, however, that there was no need for a redirection on the point. In the event, there was no redirection.

The first ground of appeal: verdict unreasonable

37 The first ground of appeal was in these terms:

The jury's verdict was unreasonable, in that it was incapable of being supported having regard to the evidence led at the Applicant's trial.

Upon the whole of the evidence adduced by the Crown, it was not open to the jury to be satisfied beyond reasonable doubt of the Applicant's guilt in respect of the charge.

**38 The approach required of appellate courts in considering the 'unsafe and unsatisfactory' ground involves the following steps:

1. The court of criminal appeal must ask itself whether, upon the whole of the evidence, it was open to the jury to be satisfied beyond reasonable doubt that the accused was guilty.

2. In considering that question, the appeal court must bear in mind that the jury has the primary responsibility of determining guilt or innocence and has had the benefit of seeing and hearing the witnesses.

3. In most cases a doubt experienced by an appellate court will be a doubt which a jury ought also to have experienced.

4. It is only where a jury's advantage in seeing and hearing the evidence is capable of resolving a doubt experienced by a court of criminal appeal that the court may conclude that no miscarriage of justice occurred.[6]

39 A guilty verdict can only be said to have been 'reasonably open' to the jury if there was no aspect of the evidence which obliged - as distinct from entitled - the jury to come to a different conclusion. In Libke v The Queen,[7] Hayne J (with whom Gleeson CJ and Heydon J agreed) said in relation to the 'unsafe and unsatisfactory' ground:**

But the question for an appellate court is whether it was open to the jury to be satisfied of guilt beyond reasonable doubt, which is to say whether the jury must, as distinct from might, have entertained a doubt about the appellant's guilt.[8] It is not sufficient to show that there was material which might have been taken by the jury to be sufficient to preclude satisfaction of guilt to the requisite standard.[9]

40 In other words, the question posed in M v The Queen,[10] namely,

Was it reasonably open to the jury to be satisfied beyond reasonable doubt of the accused's guilt?

whether the state of the evidence was such as to preclude a jury acting reasonably from being satisfied of guilt to the requisite standard.

To adopt some helpful metaphors from recent interstate appellate decisions, the question is whether there was a 'solid obstacle to reaching a conclusion beyond reasonable doubt'[11] or whether, instead, the 'path to a conviction was open'. [12]

**41 The central contention advanced for the applicant was that there was no forensic evidence to support either a shaking or the application of blunt force on the night that the infant died. The medical evidence, so it was contended, could not be the basis for the drawing of inferences of guilt to the requisite standard.

42 As noted earlier, the thrust of Professor Cordner's evidence was, essentially, negative. That is:**

• he was satisfied that the cause of death was not a re-bleed;
• there were none of the usual signs (bruising or retinal haemorrhages) to make shaking 'a real possibility'; and
• there was no bruising to indicate blunt trauma.

Professor Cordner's only affirmative conclusion was that the cause of death was a subdural haemorrhage. His evidence provided no affirmative support for either of the Crown's hypotheses. As to the re-bleed hypothesis, the spread of the bleeding was a significant contra-indication. As to the shaking hypothesis, there was a complete absence of the physical signs normally associated with a shaking of sufficient force to cause a subdural haemorrhage.

43 The question for determination, therefore, is whether it was open to the jury, in the face of that expert evidence, to reach a conclusion beyond reasonable doubt that the applicant caused the baby's death, either by the admitted shaking having led to a fatal re-bleed on the night of the death, or by a second shaking (or other trauma) on that night.

**44 Juries are not bound to accept and act upon expert evidence. At the same time, they are not entitled to disregard it capriciously. As the New South Wales Court of Criminal Appeal said 20 years ago,[13] these two propositions 'found expression and support' in a line of English authority dealing with the defences of insanity and diminished responsibility.

45 The starting point is R v Matheson.[14] In that case, Lord Goddard CJ said:**

[A] verdict must be founded on evidence. If there are facts which would entitle a jury to reject or differ from the opinions of the medical men, this court would not, and indeed could not, disturb their verdict, but if the doctors' evidence is unchallenged and there is no other on this issue, a verdict contrary to their opinion would not be a "true verdict in accordance with the evidence".

46 In R v Bailey, Lord Parker CJ accepted and applied Matheson, saying:[15]

The court has said on many occasions that of course juries are not bound by what the medical witnesses say, but at the same time they must act on evidence, and if there is nothing before them, no facts and no circumstances shown before them which throw doubt on the medical evidence, then that is all that they are left with, and the jury, in those circumstances, must accept it.

47 These principles were subsequently applied by the Privy Council in Walton v R.[16] In the particular case it was held that the jury had been entitled to depart from the conclusion of the experts. The Privy Council said:

[I]nsofar as [their Lordships] can judge of the medical evidence from the trial judge's notes, the jury were entitled to regard it as not entirely convincing ... It is plain that the quality and weight of this medical evidence fell a long way short of that in Matheson and in Bailey.[17]

48 In Taylor v The Queen, the Full Federal Court held that the jury should not have rejected unchallenged medical evidence regarding the diminished responsibility of the accused. Connor and Franki JJ treated as applicable the following remarks of Gibbs J in R v Dick:

However, in spite of differences in detail, all medical witnesses were in full agreement in their conclusions, namely, that the accused was suffering from an abnormality of mind, that this arose from epilepsy, and that it was at least probable that at the time of the killing it substantially impaired the capacity of the accused to control his actions. The jury may reject unanimous medical evidence that an accused person was of diminished responsibility if there is other evidence to support the verdict. Here, however, the evidence showed that the actions of the accused before, at, and after the time of the shooting were quite irrational, and the evidence as to his words and conduct does not conflict with medical evidence.

In those circumstances, it seems to me that the case comes within the words used by the Court of Criminal Appeal in R v Matheson.[18]

This passage is a clear illustration of the inability of the jury to reject unanimous medical evidence unless there is other evidence which can displace or throw doubt on that evidence."[19]

49 In Hall, the NSW Court of Criminal Appeal accepted as correct the line of authority beginning with Matheson and Bailey. Importantly for present purposes, Roden J distinguished between cases (like Matheson) where the medical evidence was led to establish a specific defence (on the balance of probabilities), and cases (like Hall) where the medical evidence bore on the question whether the defendant's guilt had been affirmatively established:

Whereas in those other cases the medical evidence had to establish a fact [abnormality of mind] on balance of probabilities, here it had merely to give rise to a reasonable doubt. In that circumstance, more is required to justify a verdict against the medical evidence.[20]

50 In Hall, the defence had led medical evidence to show that the accused was in a state of automatism at the time of the offending. The Crown led no expert evidence to contradict this view. The defence expert was a person of recognised competence; his opinion was unequivocal; it was based on unchallenged accounts of the accused's behaviour at the time; and he was unshaken in cross-examination. The Court of Criminal Appeal upheld the appeal against conviction on the ground that the verdict was against the medical evidence and could not be sustained. It was not open to the jury to reject the medical evidence which - at the least - raised a reasonable doubt about whether the Crown had established voluntariness and specific intent.

**51 It was common ground on the hearing of the appeal that these principles were applicable to the present case. Senior counsel for the Crown contended that, contrary to the applicant's argument, there was other evidence which entitled the jury to be satisfied beyond reasonable doubt that cause of death was a re-bleed or a fresh shaking, for either of which the applicant was criminally responsible. Counsel identified the following matters: the applicant's admission that he had shaken the baby on an earlier occasion; the medical evidence showing that the first haemorrhage resulted from that shaking; the fact that there had been a second - and fatal - haemorrhage; and the fact that the applicant was present on the night of that fatal haemorrhage.

52 The present case is importantly different from those discussed above. In each of those cases, the relevant expert evidence was led by the defence to establish - affirmatively - the defendant's (abnormal) state of mind at the relevant time. The question in each case was whether it was open to the jury to reject the affirmative expert evidence that the mental abnormality existed.

53 In this case, by contrast, the medical evidence did not establish - did not purport to establish - any affirmative conclusion beyond the fact that the baby had died from a subdural haemorrhage. As I have said, there was no medical evidence to support the prosecution case that death resulted either from a re-bleed of the first haemorrhage or from a second shaking causing a second haemorrhage. The medical evidence was to the opposite effect. In Professor Cordner's opinion, the death had not been caused by an event of either kind. All the prosecution had to rely on was the fact that Professor Cordner had left open the possibility that, contrary to his view, one or other of those events had occurred.

54 Hence, in order to convict the applicant, the jury not only had to reject the unchallenged expert opinion of the Crown's witness - that the death did not result from a re-bleed or a second shaking - but had to reach an affirmative conclusion to precisely the opposite effect, and do so beyond reasonable doubt. This was no ordinary circumstantial case. The jury were not simply being asked to reason to a finding of guilt by inference from the circumstances of the death. Rather, they were being asked to attribute to the applicant - by inference - one or other of two unlawful acts, when the Crown's expert had told the jury that in his opinion the death did not result from an act of either kind.

55 I am not persuaded that there was any other evidence which could have justified the jury in reaching an affirmative conclusion, beyond reasonable doubt, that the cause of death was either a re-bleed or a fresh shaking. I deal first with the question of the re-bleed. This hypothesis did not depend on anything which had occurred on the night of the death. It turned entirely on whether the fatal bleeding was a re-bleed of the first haemorrhage. Professor Cordner's opinion was that it was not. There was simply no other evidence which could have justified the jury in reaching a different conclusion. It was a purely medical question.

56 The position may be contrasted with, for example, Walton, where the evidence relevant to the question of the defendant's mental abnormality included the nature of the killing and the defendant's behaviour before, at the time of and after the killing. The jury were in a position to judge for themselves, by reference to the evidence of all of those matters, whether they accepted the medical opinion as to the existence of mental abnormality.[21]

57 In short, there was no evidence to support the re-bleed hypothesis. That part of the Crown case should not have been left to the jury. The no case submission should have been upheld.

58 His Honour gave the following reasons for rejecting the no case submission in relation to the re-bleed hypothesis:**

My reading of Professor Cordner's evidence is that the first subdural bleed occurred between two weeks and five days before the death of the baby. That is the period during which, as the evidence seems to indicate, the admitted shaking took place.

That being the state of the evidence it would, it seems to me, be open to the jury to draw the inference that the second and fatal bleed was a re-bleed following the first episode of subdural haematoma. In coming to that conclusion (which of course is tentative, but I need do no more for present purposes) I accept that Professor Cordner expressed real doubt as to its likelihood. Nevertheless, he was not asked whether, given the contraindications relating to other causes of the fatal intracranial bleed, the theory about a re-bleed might be strengthened. We, therefore, do not know whether his reservations about the re-bleed would have been quite as they were had he been asked whether, given the absence of any other clear indicator of the cause of the fatal bleed, a re-bleed might not have been more likely.

**59 With great respect, this analysis was erroneous. First, the importance of Professor Cordner's evidence on this issue was not that he 'expressed real doubt', but that he had expressed a firm opinion that there had not been a re-bleed. (The prosecutor said, more than once, that this was why the Crown did not contend that a re-bleed was the cause of death.) Secondly, it was not open to the judge - or to the jury - to speculate about what Professor Cordner might have said in evidence had he been asked different questions. The evidence had to be considered in the terms in which it had been given. Thirdly - and most importantly - the ruling overlooked the fact that Professor Cordner's evidence was the only evidence capable of bearing on the re-bleed question, and the jury therefore had no alternative but to accept it.

60 As to the alternative hypothesis - of a second shaking on the final night - the evidence said to bear on that question was that the applicant was present on the night, had the opportunity to shake the baby and had admitted to having shaken it earlier. In my opinion, there was no way of viewing that evidence which could have justified an affirmative conclusion, contrary to Professor Cordner's, that the applicant had shaken the baby to death.

61 The position would have been different had there been some physical sign that a shaking had occurred shortly before the death. The applicant's defence in those circumstances might well have been that any shaking which had occurred was not done by him. It would then have been appropriate for the jury to reason on the basis of probability - in the light of the earlier admitted shaking - that it was the applicant (rather than the mother) who had done the fatal shaking.[22] But here - as defence counsel correctly pointed out in support of the no case submission - there was nothing to show that a shaking (or other trauma) had occurred on that night. The medical evidence was to the contrary. Neither the presence of the applicant in the room at the time, nor the fact that he had on an earlier occasion shaken the baby, tended to prove that - contrary to Professor Cordner's opinion - there had in fact been a shaking on that night.

62 During argument on the no case submission the learned judge pointed out to the prosecutor that there was no evidence of any trauma subsequent to the admitted shaking. The prosecutor answered:**

[T]he evidence is there that the mother was asleep according to him and that he was the one who went to the child and at that stage the child has between the time the child was asymptomatic after having been fed and the time that the child is in real distress at 3 o'clock that child has clearly suffered the second subdural haematoma. This child cannot have done anything to itself to sustain that and if you exclude a re-bleed which we say on the evidence of Dr Cordner is excluded, if you exclude that, then there has occurred a substantial subdural haematoma and the possibility - and he doesn't have to say it, any more than that - the possibility of the shaking exists. Having put it on that basis the jury are entitled to infer that there is damage which is consistent with a shaking, the possibility of a shaking exists, the opportunity for a shaking exists and he is the only person with access to the child in that brief period of time. And a fairly massive subdural haematoma is sustained by the child in circumstances where it is not a re-bleed. With respect, it's open for them to infer that in relation to that fatal episode.

63 The learned judge questioned the probative value of the admitted earlier shaking. The following exchange took place:

PROSECUTOR: ... If the case [was one] where you [had], for example, an absence of any concessions or admissions as to having inflicted any shaking of the child - in such a case if there was evidence from Professor Cordner that well, there was a possibility it occurred and an absence of other evidence apart from a subdural, my learned friend's [client] might have been on a lot stronger ground. But here we do have an admitted shaking within the relevant timeframe.

HIS HONOUR: That is only relevant if you accept the possibility of a re-bleed. Unless you say that the first episode is evidence of a propensity that the jury could take as strengthening the hypothesis that in the period immediately before the baby exhibited signs of distress the only person in the vicinity was the accused.

PROSECUTOR: Yes, I don't think it's a matter of propensity, certainly a matter of opportunity. He had done it once and there was an opportunity and the way in which he quietened the child on that occasion is one that is set out clearly in the material and this is an opportunity where the child again shortly prior to that had been asymptomatic and all of a sudden is [symptomatic] and if you exclude a re-bleed there must necessarily be an episode of trauma because there is no other means of a bleed occurs.

HIS HONOUR: Without the evidence of the shaking to which the accused admitted in his record of interview would there be sufficient evidence to go to a jury given that in those circumstances the only explanation for any association with the accused and the death was that he happened to be with the baby in the few minutes between the baby being put down and the baby exhibiting signs of distress?

PROSECUTOR: He being the only potential author of that situation it's submitted that it is so, because he is the only person who can do it, and this subdural haemorrhage if it is accepted not to be a re-bleed did not occur spontaneously.[23]

**64 For the reasons I have given, I consider that it was not reasonably open to the jury, on the evidence led by the Crown, to conclude that the applicant had caused the child's death by shaking it in the period immediately preceding the death. It follows that the no case submission should have been upheld on this aspect of the case as well. The evidence was not capable of supporting a conviction.[24]

65 It is not apparent from his Honour's ruling why the no case submission on this point was rejected. Having summarised Professor Cordner's evidence, his Honour proceeded to express the opinion that, on the inferences most favourable to the Crown, there was**

material upon which the jury could infer that [the applicant] was criminally associated with and, indeed, responsible for, the death of the baby.

The ruling did not, however, specify what the other material was which could support the conclusion that there had been a second shaking.

66 Apart from the lack of evidence as to the alleged unlawful acts, I also consider that the evidence raised doubts with respect to the mental element of the crime. As noted earlier, Professor Cordner had stated that there was continuing controversy amongst experts about the mechanism by which shaking a baby could cause a brain haemorrhage and - equally - about the degree of force required for that to occur. Professor Cordner added that this controversy was directly relevant to

... the question, is it always the case that the shaking has to be so severe that the person doing it would have some understanding that some harm might result.[25]

**67 The Crown's 'second shaking' hypothesis required the jury to postulate a shaking event about which nothing was known other than that it had - on this hypothesis - caused the child's death. In the face of Professor Cordner's evidence, I do not see how the jury could have been satisfied beyond reasonable doubt that the force used in the postulated shaking was such that a reasonable person in the applicant's position must have known that it would expose the baby to an appreciable risk of serious harm.

68 Two further matters need to be addressed. First, as I mentioned earlier, the Crown relied on the admitted earlier shaking, and the circumstances in which the applicant said it had occurred, to support its 'second shaking' case against the applicant. As noted above, the Crown told the jury that their starting-point was that the applicant had already shaken the baby once, vigorously. In addition the Crown relied on the fact that he was present on the night of the death and had the opportunity to shake the baby again.

69 The prosecutor disavowed any invitation to the jury to engage in propensity reasoning but, in my view, that is how this part of the Crown's case, as presented to the jury, must be viewed. There was no evidence that there had been a second shaking. Plainly, the presence of the applicant on the night of the death proved nothing about the cause of the fatal haemorrhage. That being so, the Crown was forced to rely, impermissibly, on the fact of the admitted earlier shaking. Implicitly, the Crown was inviting the jury to reason that the applicant was the kind of person who was likely to have shaken the baby a second time, on the night of the death, and that a second shaking by him must therefore have taken place, causing the death. Propensity evidence cannot be used for this purpose.[26]

70 Finally, I turn to the judge's repeated directions to the jury that it was their task to determine the cause of the fatal subdural haemorrhage.[27] With great respect, that was a mischaracterisation of the jury's function, and the directions are likely to have misled the jury into thinking that they had to form a view as to what caused the death and, for that purpose, to settle upon one of the three 'logical' possibilities which the judge had identified for them. The Court was not conducting an inquest into the child's death. The jury's function was not to find the cause of death but to decide whether they were satisfied beyond reasonable doubt that the applicant had caused the death by an unlawful and dangerous act. Had it been necessary to do so, I would have concluded that this misdirection had, in the context of the medical evidence which had been led, produced a miscarriage of justice.**

The learned trial judge erred in his directions to the jury in failing to direct the jury that in relation to the question of unanimity of verdict for a finding of guilt the jury had to be unanimous as to the cause of death.

72 In view of my conclusion on ground 1, it is not strictly necessary to consider this ground, but I do so as it raises a matter of importance and was fully argued. As already described, the jury were presented with two alternative paths to conviction. The first was to convict on the basis that the applicant had caused the death of the baby by shaking him dangerously (in the relevant sense) on the night of his death. The alternative was to convict him of having caused the death of the baby by shaking him dangerously on the earlier occasion - approximately a week before the death - causing the first subdural haemorrhage which spontaneously re-bled on the night of his death.

73 The trial judge instructed the jury that they must be unanimous in their verdict:**

You doubtless will look at the evidence through different eyes. In fact, that's a necessity since you have got 12 pairs of eyes and you are all different. It does not matter that you come to your conclusions by different routes. But in the end your verdict must be unanimous. Unless you have unanimity you do not have a verdict. So do not be concerned that the reasoning process of one of your number is different to your own, or even that the other 11 have a different reasoning process to yours.

The only exception taken to this direction was by the prosecutor, who pointed out that the jury must be unanimous either to convict or acquit. His Honour then re-directed the jury as follows:

I spoke about the necessity for a unanimous verdict. I should have added that that means that you cannot bring down a verdict of either guilty or not guilty unless you, all 12 of you, are of that view. So unanimous means either way.

**74 The contention for the applicant is that the judge erred in failing to direct the jury that they must be unanimous as to which of the two alternative acts - the admitted first shaking or the disputed second shaking - was the basis of conviction. In my opinion, this contention must be upheld. My reasons are as follows.

75 In Walsh, Phillips and Buchanan JJA (with whom Ormiston JA agreed) concluded that the authorities dealing with jury unanimity disclosed two distinct types of case. In a case of the first type, alternative legal bases of guilt are proposed by the Crown but depend substantially upon the same facts. In such a case, their Honours said -**

[T]here is no need for a direction on "unanimity" about one or other or more of these bases, at least if they do not "involve materially different issues or consequences".[28]

is where one offence is charged, such as obtaining property by deception, but a number of discrete acts is relied upon as proof and any one of them would entitle the jury to convict. If those discrete acts go to the proof of an essential ingredient of the crime charged, then the jury cannot convict unless they are agreed upon that act which, in their opinion, does constitute that essential ingredient.[29]

76 In my opinion, the present case is clearly of the second type. Although the presentment charged only that the applicant had killed the child, the case against him was that he was guilty of manslaughter by dangerous and unlawful act. Proof of an act having both those characteristics was essential to conviction. The prosecution relied on - and the judge left to the jury - two 'discrete acts', some days apart, each of which was said to be capable of constituting the crime charged. On the authority of Walsh - and the cases on which the analysis was based - the jury could not convict unless they were 'agreed upon that act which, in their opinion, does

77 No direction to that effect was given. Regrettably, neither counsel at the trial recognised the necessity for such a direction. (Nor, as noted earlier, did those representing the applicant in the preparation of his grounds of appeal.) In the absence of such a direction, it is entirely possible that there was no unanimity amongst the jurors as to which act founded the guilty verdict and, therefore, that the applicant was not lawfully convicted. For that additional reason, the conviction would have to be quashed.

78 In view of my conclusion on ground 1, I would order that the conviction be quashed and that a verdict of acquittal be entered.**

**79 I agree for the reasons advanced by Maxwell P that this application should be granted, the appeal allowed and a verdict of acquittal entered.

80 I have had the advantage of reading the draft judgment of Maxwell P and agree with his Honour that the second ground of appeal is made out. In the circumstances of this case I consider that the learned judge below should have directed the jury that they must be unanimous as to whether they were satisfied beyond reasonable doubt that the applicant killed Izaiah by shaking him dangerously about two weeks before he died, or alternatively, by shaking or inflicting some other form of trauma on him, on the night of his death.

81 That brings me to the first ground of appeal, which was that the jury verdict was unreasonable. That ground of appeal has caused me considerable difficulty. However, having considered Maxwell P's reasons my view is that a verdict of acquittal should be entered. In making some additional comments on this ground of appeal, I gratefully rely on the excerpts of evidence set out in Maxwell P's judgement**

**and also refer to some other parts of Professor Cordner's evidence.

82 The uncontested evidence was that Izaiah died from a subdural haemorrhage. Professor Cordner's evidence-in-chief was that because there was no sign of inflammation or the breakdown of blood to produce iron, the haemorrhage was fresh and 'that means is in the perhaps 24 hours before death. It could be 26, 28, it could be one or two.'

83 In cross-examination, defence counsel put it to Professor Cordner that it was rather more likely that the haemorrhage occurred shortly before the baby's breathing difficulties became apparent. Professor Cordner said that he thought this was probably the case. He also agreed with defence counsel's proposition that it was more probable that the bleeding occurred at a time 'really quite proximate to the death.'

84 Professor Cordner also gave evidence that there 'wasn't any obvious to the naked eye natural disease or abnormality otherwise seen in the brain' and that there were no signs in the baby of 'pre-existing natural disease to cause his death.'

85 In cross-examination Professor Cordner agreed with defence counsel's proposition that some babies suffer subdural haemorrhages as the result of a vaginal delivery. He referred to a paper showing that 7 out of 111 babies delivered through the use of forceps and 2 or 3 children delivered without use of forceps had small subdural haemorrhages. However none of these haemorrhages had any consequences and at 4 weeks all had completely healed. It may be noted that Izaiah was 4 weeks old when he died. Professor Cordner said that he 'wouldn't be suggesting there is really any significant possibility that the subdural haemorrhages that we saw on the photos has been there since birth'.

The fresh haemorrhage could possibly be due to a spontaneous bleeding, a phenomena that [is regarded] as occurring with minimal or no trauma after a previous subdural haemorrhage. A spontaneous haemorrhage can occur when there is a generalised tendency to bleed. Some people just have thin blood and bleed more easily, and in those circumstances a spontaneous re-bleed of the previous subdural haemorrhage would clearly be more likely.

87 Counsel for the Crown then asked Professor Cordner whether the baby had a tendency to bleed easily. He responded that:

Apart from the bleeding over the brain and the small bruises in the cheek there were in other haemorrhages or bruises and, of course with the thin blood or the tendency to bleed then bruises can occur spontaneously anywhere.

It seems that the word 'in' in the transcript should read 'no' and that Professor Cordner considered that there was no indication that the baby's blood was thin.

**88 As the President sets out in his reasons, Professor Cordner also said that, although he could not rule out the possibility of a re-bleed from a factual point of view, he was 'quite comfortable [himself] that it [was] not a spontaneous re-bleed' because the bleeding occurred in three separate parts of the brain.

89 On the basis of that evidence I agree with the President that it was not open to the jury to conclude beyond reasonable doubt that the baby died as the result of a re-bleed of the haemorrhage caused by Mr Klamo's admitted shaking of Izaiah about two weeks before the baby died.

90 The question then is whether the evidence was sufficient to leave it open to the jury to infer that Izaiah's death occurred as the result of an unlawful and dangerous act committed by Mr Klamo, on the evening of the baby's death. The President refers in his judgment to the evidence relied upon by the Crown to support the inference that this was the case. This included the fact that the applicant was the last person to have contact with the child, that he had the opportunity to injure the baby and that he had admitted to shaking the baby about 2 weeks previously.

91 Maxwell P's judgment extracts the portions of Professor Cordner's evidence on the question of whether the baby was shaken or suffered some other trauma on the evening of his death. His Honour describes the conclusion that Izaiah died as the result of a shaking event occurring on the night of his death as 'contrary' to that medical evidence. I would not go quite so far.

92 Professor Cordner said that significant head injury could occur without 'necessarily bruising what is very tough tissue, the scalp.' He also said that 'the possibility of a shaking event before death exists, but there are no bruises to the arms or chest or retinal haemorrhages to support this as a real possibility.' In re-examination he conceded that 'a baby being swaddled in clothes would reduce the likelihood of any particular bruising ... being made.'

93 There are two other matters, which I think the jury could legitimately have taken into account.

94 The first was Professor Cordner's evidence that there was no sign of any disease which could have caused the baby's haemorrhage on the evening of his death and that subdural haemorrhages caused during vaginal delivery healed by the time the baby was 4 weeks old. Dr Andrew Watkins, a neonatologist from the Department of Paediatrics at the Mercy Hospital, gave similar evidence about the effects of birth-related subdural haemorrhages.

95 Secondly, in his police interview, the applicant told the police that he was 'in the wrong' to shake the baby on the occasion that he did so. When asked whether that shaking was 'a dangerous thing to do' he agreed, saying 'Yes it was. It was stupid.' If the evidence had been sufficient to permit the jury to find beyond reasonable doubt that the applicant shook the baby shortly before he died, those admissions were evidence from which the jury could have inferred that the applicant knew that his shaking of Izaiah exposed the baby to an appreciable risk of serious harm.

96 In my opinion, however, it was not open to the jury on the whole of the evidence to conclude that the baby's death occurred as a result of shaking or other trauma, which the applicant inflicted on the evening of the baby's death. Professor Cordner's evidence on the cause of the subdural haemorrhage put it no higher than a possibility that there had been a second shaking on the evening of the baby's death or that some other form of trauma had been inflicted on the baby by the applicant, up to 24 hours before his death.

97 Professor Cordner found no sign of pre-existing natural disease and did not consider that the death could have been caused by a birth-related subdural haemorrhage. However, he was not asked whether the baby could have suffered a spontaneous subdural haemorrhage, without any trauma being inflicted upon him. Thus, the possibility that the child died as the result of unknown natural processes was not excluded. Although Professor Cordner did not completely exclude the possibility that the child was shaken or suffered other trauma shortly before his death there was no positive evidence that this had occurred. Professor Cordner's evidence raised only a possibility that this was the case. The absence of evidence as to the cause of Izaiah's death cannot be converted into an inference that the applicant must have caused it, by shaking him or inflicting some other form of trauma. Further, Professor Cordner's evidence left open the possibility that the event which caused Izaiah's death occurred as long as 24 hours before he died, so that there was a possibility that the fresh bleed was not caused by any act of the applicant.

98 This case differs from situations in which there is sufficient circumstantial evidence to permit the jury to infer that an accused has unlawfully caused the death of a victim. If the body of a deceased person is not found, there may nevertheless be evidence which allows the jury to infer that the victim is dead and that the accused unlawfully caused that death.[30] The same applies where the state of a victim's body does not allow the precise cause of death to be determined, but there is evidence suggesting that the accused must have been implicated in the death of the victim. If the cause of a victim's death is known, but there is no direct evidence about the acts of the accused, there may be evidence about the motives or behaviour of an accused, which allows a jury to infer that the accused unlawfully caused the victim's death.

99 In this case, by contrast, there is neither clear evidence as to the cause of the baby's death nor evidence that it was caused by an act of the applicant. As the President observes in his judgment, the inference that the baby died as the result of the applicant's unlawful and dangerous act was based almost solely on propensity reasoning.

100 I therefore agree with the orders proposed by Maxwell P.

[6] M v The Queen [1994] HCA 63; (1994) 181 CLR 487, 493-4 (Mason CJ, Deane, Dawson and Toohey JJ); Jones v The Queen (1997) 191 CLR 439, 451-2 (Gaudron, McHugh and Gummow JJ); Weiss v The Queen [2005] HCA 81; (2005) 224 CLR 300, 316 [41] (The Court); R v Tiburcy [2007] VSCA 124, [5] (Nettle JA).

[7] [2007] HCA 30; (2007) 235 ALR 517, [113].

[8] Citing M v The Queen [1994] HCA 63; (1994) 181 CLR 487, 492-3 (Mason CJ, Deane, Dawson and Toohey JJ).

[9] See also R v Tiburcy [2007] VSCA 124, [17] (Vincent JA).

[10] [1994] HCA 63; (1994) 181 CLR 487, 493 (Mason CJ, Deane, Dawson and Toohey JJ).

[11] R v Shah [2007] SASC 68, [4] (Doyle CJ).

[12] Morabito v R [2007] NSWCCA 126, [34] (Mason P).

[13] Hall (1988) 36 A Crim R 368, 370 (Roden J).

[14] [1958] 1 WLR 474, 478.

[15] (1977) 66 Cr App R 31, 32.

[16] [1978] AC 788.

[18] Taylor v The Queen (1978) 45 FLR 343, 363, citing R v Dick [1966] Qd R 301.

[19] Ibid 364. See also J D Heydon, Cross on Evidence (7th ed, 2004) [29075].

[20] (1988) 36 A Crim R 368, 371.

[21] See also R v Michaux [1984] 2 Qd R 159, 162-4 (Connolly J); R v Gemmill [2004] VSCA 72; (2004) 8 VR 242, 253 [49] (Eames JA).

[22] Cf Pfennig v The Queen [1995] HCA 7; (1995) 182 CLR 461, 485-7 (Mason CJ, Deane & Dawson JJ); see also R v DCC [2004] VSCA 230; (2004) 11 VR 129, 132 [8] (Callaway JA).

[24] Cf Doney v R [1990] HCA 51; (1990) 171 CLR 207, 214-5 (The Court).

[25] See [22] above.

[26] Pfennig v The Queen [1995] HCA 7; (1995) 182 CLR 461, 480-1 (Mason CJ, Deane & Dawson JJ); R v DCC (supra n 22).

[27] See [35]-[37] above.

[28] [2002] VSCA 98; (2002) 131 A Crim R 299, 316.

[30] See, for example, Weissensteiner v R [1993] HCA 65; (1993) 178 CLR 217.