---

IN THE KWAZULU NATAL HIGH COURT, PIETERMARITZBURG

REPUBLIC OF SOUTH AFRICA

CASE NO.: AR342/10

In the matter between:

HILTON CRAWFORD SHAW …........................................................................APPELLANT

and

THE STATE …..................................................................................................DEFENDANT

JUDGMENT

D PILLAY J

Introduction

1. On 15 June 2009, the appellant was convicted and sentenced to twelve years imprisonment for the murder of his wife Susan Shaw on 3 June 2007. Without direct evidence as to how the deceased met her demise, three possibilities emerged at the trial:

   1. The deceased shot herself.

   2. An intruder shot the deceased.

   3. The appellant shot the deceased.

2. The learned Judge in the trial court rejected the first two possibilities and found on the circumstantial evidence that the only reasonable inference to be drawn was that the appellant possessed the direct intention to kill the deceased.

Appellant’s version

3. The appellant presented a picture of his life with the deceased as that of a normal middle-aged couple with their share of troubles and triumphs. They had been married to each other by antenuptial contract for about three years. They lived on his homestead at Lake Lyndhurst. Each had children from previous marriages. The deceased’s sons, seventeen year old Nicholas and fourteen year old Cameron, were in the care of her ex-mother-in-law, Mrs Felicity Smith.

4. On Thursday before she died, the deceased discovered that Mrs Smith had taken half the deceased’s Workman’s Compensation Pension. She telephoned Mrs Smith. They argued. The deceased remained upset. Later, whilst at the home of friends in Nottingham Road, the deceased had a telephonic conversation with Nicholas, which upset her further to the extent that she cried most of the drive home.

5. On Friday morning, the deceased did not feel well. She was deflated and concerned about issues with Mrs Smith. She spent most of the day in bed reading and arose in the early evening. The couple had much to discuss as the following week they were busy with burning firebreaks and receiving paying guests.

6. On Saturday morning, as usual, the appellant gave the deceased coffee followed by breakfast in bed before setting out to Nottingham Road to receive a fax from his son. His son had information about property the appellant was trying to buy and the appellant wanted to know whether Investec had approved the finances for the purchase.

7. He received the fax. He also made an appointment with the deceased’s attorney for 4 June 2007 to discuss her claim against Vodacom for damages arising from its failed attempt to prosecute her for fraud. When he returned home, the deceased was reading in the lounge. They discussed their plans for the following week and the deceased retired early after supper.

8. On Sunday morning, he gave her coffee and breakfast in bed again. He set her up with her book and her glasses. She complained of a stomach ache. She was menstruating. He drove to point K which was a spot about three kilometers from his house where he could get a signal and exchanged several cellular phone calls with his son and his advocate friend about the property transaction.

9. Back at the homestead he fixed a broken window pane. The deceased emerged from the bedroom with her book and complained of being cold. He made a fire whilst she lay on the couch in front of the fireplace reading her book.

10. In the afternoon, the deceased got up, declared that she was bored and tired of reading and was going to make supper. Shortly thereafter she was at the bar with a five litre box of wine from which she poured a glass for herself. He commented to her that she was ‘starting early’. She replied that she was going to use the wine in the food. He cautioned her not to drink as it would exacerbate her stomach complaint.

11. The appellant continued to repair the window. He put the tools away and returned to the house to find the deceased still sitting at the bar. She invited him to join her with a glass of wine and to talk to her.

12. Although the deceased had been upset after her discussion with Mrs Smith and her son, she appeared to have recovered, as those issues did not arise over the weekend until Sunday afternoon when the deceased said: ‘Do you think Felicity knows what she is doing to my children?’ They had a brief discussion during the course of which he told her that if she wanted to go back to town to start life again and to look after her children as her mother had suggested days before, she would have his blessing.

13. Just as the appellant was about to return to point K to check on more short message services (sms) from his son so that if there was a problem with the property transaction he would have the evening to think it over, the deceased got up from the bar stool, hugged and kissed him, told him that she loved him very much and that she was not going anywhere without him.

14. He drove to point K and waited there for about 10 to 15 minutes. He received no sms and returned to the house. As he put down his keys and cellular phone, he saw the deceased lying face down on the veranda. He thought that she might have had an epileptic fit or had slipped and fallen. He turned her around and saw a bullet wound in her right shoulder. Instantly he thought that she had tried to ‘do something to herself’.

15. At the time he did not think that an intruder shot her because in the thirty-odd years that he owned the farm, there had never been a problem. There had been many break-ins when they were not in residence.

16. As he held and kissed her he asked why she had done this to herself. Feeling her breath on his cheek he assumed that she was still alive. Realising she needed medical attention he started to drive back to point K. In his haste to call Mr Kobus Kruger, his friend and neighbour, he omitted to take his cellular phone. Remembering this halfway up the driveway, he returned to the house, searched under the car seat and finally found it still inside the house.

17. Distressed and hysterical he called Mr Kruger at 16:59. As the chairman of the Property Owners’ Association he knew the GPS co-ordinates of the property; he would therefore have been able to direct a helicopter or air ambulance to the property.

18. On returning to the house, the appellant knelt on the floor beside the deceased. Her Croc shoes got in the way and he threw them aside. He could not feel her breath anymore but he was not sure that she was dead. He tried to take her pulse, but he was shaking too much himself.

19. At about 17:50 he returned to point K to telephone Mrs Smith, Michael Lambert the deceased’s brother-in-law, and Marcel Du Preez an inspector at Nottingham Road Police Station. Neither Mrs Smith nor Inspector Du Preez answered his calls. Mr Kruger informed him that as the air ambulance would not take off after 17:00 another ambulance had been arranged.

20. When he returned to the house he knew that the deceased had passed away. He looked for a weapon around the veranda and in the room next to their bedroom where the safe was. The safe door was unlocked. Opening it he noticed that the firearm was missing. In the main bedroom, he found the firearm on the floor next to the bathroom. For the first time he noticed a trail of blood on the cream carpet leading through to the tiled thoroughfare from the house onto the veranda.

21. He returned to point K where he received a sms from Mr Kruger that the ambulance and the police would be there shortly. Remembering that he had to open the gate, he drove to it and found the deceased’s brother-in-law Mr Craig Bricknell, the ambulance and the police already waiting there. As the driveway was a one-car lane which made it difficult to turn around his vehicle, he got into Mr Bricknell’s car and travelled back with him to the house.

22. His sons arrived later that evening. It occurred to him then that when he first found the deceased he did not recall seeing the dogs around. (Again the appellant was asked to speak loudly.) The dogs returned sometime later.

23. Inspector Du Preez arrived at some stage; he ignored the appellant’s greeting. The appellant assisted the mortuary attendants to put the deceased into a bag and onto a stretcher. Constable van de Merwe tested the appellant’s hands for primer residue after 17:00 that evening. The following day the police took the clothes that he had worn for testing.

24. The next morning when he had to go to the mortuary to identify the deceased, he searched for her handbag to find her identity book. Her handbag was missing. He also could not find any of her jewellery and her watch. His sons drove him to Nottingham Road Police Station where Officer Marius van de Skyff informed him that Mr Bricknell had taken the handbag and its contents which were given to the deceased’s sons. They were returned to him after the funeral.

25. As a habit, whenever the appellant left the deceased alone, he left a loaded firearm for her protection. He had trained her to use it. On Saturday morning when he went to Nottingham Road, he had left the firearm beneath her clothing in her cupboard. That was the last he saw of the firearm until the shooting. He assumed that she had put it back into the safe after he returned.

26. He denied having an acrimonious or abusive relationship with the deceased. When they lived in Durban, she used to have two to four epileptic episodes a week. Over indulgence in alcohol also induced epileptic attacks. That afternoon, he did not watch her all the time and did not see her drink the quantity of wine that was found in her system. Furthermore, it was not normal for her to drink on her own. Because of her epilepsy, he was always aware of her state of sobriety and signs of inebriation. Hence, if she was inebriated when he left for point K the first time that afternoon he would have known. It did not occur to him whilst he was talking to her at the bar that she had over-consumed.

27. The appellant denied that Mr Bricknell searched him. He would have objected to being searched. When he returned to the house with Mr Bricknell, he stood next to the bar but he did not consume alcohol. He denied going down to the lake that afternoon either before or after discovering the deceased.

The state’s case

Craig Denton Bricknell

28. Craig Denton Bricknell, who was married to the deceased’s sister, was the first visitor to the scene after the incident. He testified that on Thursday before she was shot, the deceased and the appellant visited him and his wife at their home in Dargle about 30 kilometres away from Lake Lyndhurst. On this occasion she visited because she was angry about Mrs Smith taking her money.

29. On Sunday he received a call from his brother-in-law, Michael Lambert, informing him that the deceased had been involved in an accident with a firearm and that he should go to her. He arrived at the gate to the appellant’s property, found it chained and padlocked. He hooted for what seemed like an eternity. He had just started walking to the homestead when he received a call from Mr Lambert informing him that the deceased had passed away.

30. Just then he saw the appellant driving up to the gate. As the appellant walked to open the padlock, Mr Bricknell told him to stay where he was as he wanted to search him, first, because it was by then pitch dark; second, he had just been told his sister-in-law died; third, he was a bit nervous; and last, he wanted to know where the firearm was. In his evidence in chief, he did not say whether he in fact searched the appellant. He also asked Donovan Wilson, his daughter’s boyfriend, to search the appellant’s bakkie.

31. The appellant opened the gate and tried to reverse his bakkie, which got stuck. He told the appellant to leave the bakkie and accompany him to the homestead in his car. The police had also arrived.

32. On the drive to the house, the appellant narrated the circumstances of the shooting. Mr Bricknell alleged that the appellant’s version ‘just jump[ed] around’. He then summarised the appellant’s version for the trial court. Although Mr Bricknell conceded that he was struggling to remember when the appellant said that she died, he nevertheless concluded that the appellant had various versions. Once the appellant said that she had died in his arms; on another occasion he said she was dead when he got back from getting his sms’. Furthermore, in order to show that the deceased was in a good mood, the appellant reported that the deceased was being typically herself on Sunday by lying on one spot and directing him to rearrange the house for the following weekend’s guests. This contradicted his evidence that she was depressed.

33. Mr Bricknell described the Shaw residence as private and secluded, with no neighbours to speak of. When they reached the homestead they walked around the grass in front of the porch. The appellant cried. Mr Wilson sat with the appellant whilst Mr Bricknell remained outside with the police and Captain van de Skyff. Mr Bricknell asked Captain van de Skyff to ask the appellant to stop drinking and to leave the bar because he, Mr Bricknell, noted blood in the bar and considered it inappropriate that the appellant should be sitting there. The appellant then moved to the back of the house.

34. Using the appellant’s torch, Mr Bricknell walked around the premises, found the firearm in the deceased’s bedroom and noted the blood trail. Later that evening, Mr Wilson informed him that the accused was acting strangely by going to the toilet all the time. Out of curiosity, Mr Bricknell investigated and found tissue with blood on it in the waste basket in one of the bathrooms at the back of the house.

35. Mr Bricknell remained at the house from about 6 pm to 11 pm. At first, the appellant was shaky and distressed but communicative. He observed blood on the seat and knees of the appellant’s pants and the tips of his sleeves. There was a bullet hole in the door of the shower and blood in the bar. The window pane in the bar had been freshly replaced with the glass lying outside. A pot of very stodgy pasta was on the stove. Two dirty plates were in the kitchen. A fire had been made in the lounge. The house was neat with no signs of any struggle having taken place.

36. Mr Bricknell did not get the impression that anything had gone missing. The deceased’s handbag was in her bedroom along with her diamond earrings. Her clothing was neatly folded on a chair in her bedroom. Her bed was made. Wet washing hung in the lounge. No drawers were open. The deceased’s body was ice cold.

37. Noticing that there was blood splattered beneath the bar stool and at the bottom of the bar Mr Bricknell asked the appellant why he had tidied the bar; the appellant emphatically denied cleaning up anything before Mr Bricknell arrived.

38. About the end of his examination in chief, Mr Bricknell was asked whether he had questioned the appellant about his firearm at any stage. He replied that he did so the following day. The learned Judge reminded him that he had mentioned earlier in his evidence that he had asked him for a firearm when they met at the gate. Furthermore, only when the Judge enquired whether he actually searched the appellant did he clarify that he did so.

39. Under cross-examination he testified that his relationship with the appellant was ‘fine’. He was the closest family member to assist the appellant and had done so in the past. He confirmed that he removed the deceased’s handbag, with her identity document and wallet, which he gave to his wife, the deceased’s sister. She had asked him to get hold of her identity document because she believed the appellant would not cooperate with the family. Later that week, when the appellant telephoned him about the handbag, he informed the appellant that he had given it to the deceased’s son. He also admitted telling the appellant that the appellant could charge him with theft; however, he had mentioned to a police officer that he had taken the handbag. He did not trust the appellant because of ‘a gut instinct’ and he remained ‘sceptical’. He conceded that there were problems in his relationship with the appellant.

40. Cross-examined as to whether the appellant resisted, Mr Bricknell evaded the question. Mr Matthews put to him that the appellant recalled Mr Wilson sitting in the back seat of Mr Bricknell’s car. Mr Bricknell persisted that Mr Wilson travelled in the police car. When it was pointed out to him that the appellant’s evidence accorded with Mr Bricknell’s statement to the police, Mr Bricknell said that he must have made an error in his statement. Pressed further, Mr Bricknell conceded that because he saw the appellant holding a glass at the bar, he assumed that he was drinking alcohol. As for the blood stained tissues in the waste basket he conceded that he did not point them out to the photographer but mentioned it to Inspector Marcel Du Preez.

Evidence of Neighbours

41. Forty-eight year old Mr Simon Madlala worked as a caretaker for the neighbouring cottage for twenty years. On 3 June 2007, he was on the upper level of the cottage painting a membrane onto the ridging on the roof to waterproof it. As he worked, he continuously looked around and could see the neighbours. He could see shadows of people in front of the appellant’s house, next to the veranda on the side facing the lake. Although he could not see how many people were there, it looked to him ‘like there was quite some movement.’

42. After some time and at about 17:00 he heard the sound of an explosion. A short while later he saw the appellant’s white bakkie leave the premises; he did not see where it went. About thirty minutes later he got off the roof. It was a windy day with the wind blowing over the dam wall toward him.

43. Sixty-four year old Mrs Dorris Ndlovu was employed at Mr Kruger’s cottage at Lake Lyndhurst. That afternoon, she was in the garden cutting off dried flowers when she heard dogs barking and noise emanating from the Shaw residence. The noise sounded like children chasing each other outside around the house. She thought it was children because at times there were children who visited the house and played there. It was the sound of ‘running, footsteps, etcetera and the (Shaw’s) dogs were also barking at the same time.’ Because it was windy it was not easy to ascertain whether it was children or adults making the noise.

44. In her written statement she made to Inspector Mchunu dated 7 June 2007 she had said that she heard the sound of ‘very high noise of cry (screaming) at the Hilton’s house I (presume) there were crying simultaneously I heard dogs b(ar)king all of a sudden the gunshot heard.’(sic) After her employer told her that the deceased died she ‘realized the screaming noise was produced by the same Susan (deceased) when she met up with her death’.

45. Mr Kruger’s house is across the lake from the Shaw residence. She testified that the incident happened ‘before five but anywhere towards there’. At around ‘five exactly or within that time’ she heard the sound of gunfire coming from the Shaw residence. The noise she had heard earlier ended. The dogs also stopped barking and started howling. The dogs did not chase anyone but the barking seemed to emanate from one place at all times, even after the shot.

46. She saw the appellant go down the stairs towards the lake. A short while later he walked up the stairs, went to the dogs and quietened them down. Not long after she heard the gunshot, the appellant proceeded into the veranda of his house.

47. In her written statement she mentioned nothing about the appellant walking down to the lake. She explained that this omission occurred because she was in shock when she wrote the statement and might have forgotten to mention it. It was put to her that in a joint consultation between the prosecution and the defence with Mr Kruger who was listed as a state witness, Mr Kruger did not mention anything about her reporting that she saw the appellant walking to the lake. All she mentioned to him was the sound of children playing outside. She persisted that she did see the appellant walk to the lake.

48. She was not wearing glasses that afternoon but started wearing them sometime after the incident. She has diabetes. Even after it was put to her that it was common cause that from where she stood she could not make out the gender or race of a person at the Shaw residence, she persisted that she saw the appellant.

Evidence of the Experts

49. Dr Perumal, a pathologist engaged by the deceased’s mother, discovered that the deceased was ‘heavily intoxicated’ with the equivalent of 0.30 grams of alcohol in her blood, implying that she must have consumed about 1.5 litres of wine. Her liver had undergone mild fatty change, which was most likely a result of alcohol use or abuse as the deceased had developed some tolerance for it.

50. In explaining the track of the wound from the front of the deceased to her back and slightly downwards, Doctor Perumal presented two variables: (1) the gun was placed at ninety degrees to the body but the body could have moved by the deceased flexing her knees and (2) if the deceased shot herself, she would have used her left hand to hold the firearm with the thumb or, with some difficulty, her index finger on the trigger. Using the right index finger would have been ‘almost impossible’ but using the right thumb would have been possible. The deceased sustained a contact gunshot wound meaning that there was no gap between the muzzle of the firearm and the deceased’s garments to allow gases to escape.

51. As regards the fresh injury to the left side of her forehead and her left middle finger, in Dr Perumal’s opinion, these injuries would not have been sustained from her having fallen face down on the veranda because, first, she would not have collapsed suddenly and, second, the injuries were not in the prominences, that is, on her eyebrow, cheek or chin. These injuries would have been inflicted at or about the time of her death.

52. The deceased had a healing bruise on her left leg anteriorly above her heel. This bruise could have been sustained a few days before her death. She could have injured herself with the cross-members of the barstool, which were about the height of her injuries from the ground.

53. According to Jakobus Steyl, a private ballistic specialist engaged by the state, the location of the wound on the deceased’s right shoulder indicated that if she shot herself she must have used her left hand to hold the firearm pressed against that shoulder. Holding the firearm in her right hand and pulling the trigger would have been awkward and difficult because she would have had to apply pressure to the trigger. It was also possible that she fired the shot using both thumbs to pull the trigger. Eight pounds or 3.5 kilograms of pressure had to be applied to the trigger to fire a shot.

54. Mr Steyl testified that primer residue cannot be traced once a shottist washes his hands. Primer residue could also fall simply off by movement of the hand. In any event within 10 minutes of hand activity primer residue could be disregarded. The same does not apply to primer residue on clothing. Mr Steyl discounted any inaccuracy in the state laboratory’s finding of no primer residue in the microscopic examination of the clothing.

55. Mr Steyl acknowledged that the splatter of blood on the veranda could have been caused by the appellant trampling on the blood when he initially arrived on the scene or by the paramedics when they moved the deceased. According to Dr Perumal the splatter was caused either by the deceased being lifted and dropped onto the blood or because someone ‘just stomped onto the blood, for reasons unknown’. When Mr Steyl’s version was put to Dr Perumal, the latter downplayed the significance of the blood splatter on the veranda by acknowledging that it was unrelated to the shooting which did not occur in that spot.

56. In the corroborating opinion of Dr Perumal, the deceased would have found it ‘a little difficult’ to pull the trigger herself; suicide was therefore not the reason for the shooting. He cited the results of research recorded in Gunshot Wounds,¹ which show that in cases of suicide by firearms, 72% of females shot themselves in the head, 22% in the chest and 6% in the abdomen. In contrast, Mr Steyl questioned the validity of that survey of 700 men and only 200 women.

57. Whilst conceding that the evidence did not exclude suicide, Mr Steyl pointed out that it was not a typical suicide. It was not typical because suicides usually involved multiple gunshot wounds and, considering all the facts of this case, it did not lean towards suicide.

Police Witness

58. Jabulani Mfanafuti Mchunu, a detective inspector in the South African Police Service stationed at Nottingham Road, was assigned the docket for investigation on 4 June 2006. He visited the scene with Inspector van de Skyff and Inspector Du Preez at about 21h00 on 3 June 2006. He examined the rooms and was surprised to find that there was blood on the wall of the guest room. He discovered this by following a trail of blood. He also found blood on the carpet, on the sink and droplets on the carpet in the back room. He acknowledged that Mr Bricknell informed him about the blood-stained tissues but he himself did not observe them.

59. Under cross-examination he conceded that he did not direct the photographer to take photographs of the substance he saw in the backroom. Pressed that the two experts who testified before him did not give evidence about this, he replied that other police officers who were present at the scene with him would corroborate him. He did not take swabs of the substance he saw on the floor because it was no longer in liquid form. This indicated to him that it had been cleaned off by 4 June 2007.

60. When Mr Paver put to him that the appellant’s statements suggest that the deceased might have been attacked by an intruder, Detective Inspector Mchunu replied that he knew nothing about this and that it was new to him.

Trial court’s findings

61. The trial court rejected the suicide hypothesis for the following reasons: the appellant disavowed reliance on this defence. Furthermore, although the deceased was a bit deflated after her altercation with Mrs Smith, there were positive developments in her life. She was angry with her ex-mother-in-law rather than depressed. She was not on any medication. The appellant had left her alone previously without any incident. A meeting was set up with her attorney for Monday. All in all, the appellant and the deceased had a normal weekend.

62. In the opinion of the two experts, it was unlikely that the deceased fired the shot. If the deceased was determined to end her own life, she had enough time to fire again into a more definitive location. Other fresh injuries were inconsistent with suicide. For these reasons, the trial court excluded suicide as a possibility.

63. The trial court could not understand how the appellant could have jumped to the conclusion that she committed suicide, having regard to her relaxed disposition immediately before the shooting. It found that his explanation that he did not see anybody else, and that confrontation with intruders was not normal flew in the face of his evidence that he was aware that the deceased might be attacked in his absence; after all, he had trained her to use a firearm for that very purpose.

64. His evidence that it was still light when he discovered the deceased conflicted with his evidence that because it was dark he could not see the abrasions on the deceased’s forehead and on her eye, which, in the opinion of the trial court, were obvious. It found it inconceivable that he did not notice the abrasions on her face and her finger which would have indicated to him that she did not shoot herself and that those injuries could not have been self inflicted.

65. On seeing the bullet hole in the bathroom door, the appellant ought to have recalled immediately their contingency plan and realised that the deceased had been surprised by an intruder and had tried to lock herself in the bathroom. However, the intruder theory occurred to him only a few days later. The trial court rejected this theory as false, as well as the appellant’s theory that the deceased could not have been surprised in the bathroom.

66. It also rejected the appellant’s explanation as to how the deceased’s blood got onto the safe door as ‘wholly improbable, particularly given the position of the blood smudges which were said to go from right to left across the front’. It also found it significant that there was no blood on the handle of the safe door or on its opening edge. It found that if the appellant wanted to find the firearm, he should obviously have followed the blood trail from the veranda to the main bedroom where the firearm was lying on the floor.

67. The trial court disbelieved the appellant’s reason for making his trip at about 4:30 pm to point K on Sunday. As nothing further could arise with Investec over the weekend, it found it strange that the accused thought there could be a problem on Sunday afternoon that had not already been discussed between himself and his advocate either on the Saturday or Sunday morning. It could not understand why he sat at point K ‘achieving nothing’ for ten minutes for a sms to download without dialling his son to clarify whether there was a problem with the property transaction.

68. The trial court found that the appellant clearly tried to convey that there had in fact been an intruder. It based this finding on the appellant’s evidence as to how the dogs had behaved if an intruder entered the house and shot the deceased. It found that the appellant could not have known how the dogs would have behaved because on his own evidence there had never been any confrontation with an intruder whilst they were in residence, all burglaries having occurred in their absence.

69. Dr Perumal’s evidence that there was no back splatter explained the absence of back splatter on the appellant’s clothing. However, the trial court considered it a ‘remarkable feat’ to get the deceased’s blood only on the tips of his sleeves after holding her whilst her clothing was blood drenched.

70. It considered the evidence of the contingency plan to be an afterthought because the appellant did not give this evidence on the first day of his testimony. It was an attempt at explaining how the pistol got into the bathroom

71. It was not surprising that Mr Bricknell searched the appellant. As the Shaws were alone he would have thought that the appellant had been involved in the deceased’s death. It discounted the likelihood of Mr Bricknell fabricating this evidence by noting that this evidence was of no significance. It found him to be a good witness who, unshaken in cross-examination, honestly endeavoured to remember as accurately as possible the events of that evening. It could not understand the appellant’s denial that Mr Bricknell searched him. It preferred Mr Bricknell’s evidence over the appellant’s.

72. It agreed with Mr Bricknell that the appellant’s version that the deceased was depressed on Friday and Saturday but in a good mood on Sunday, was inconsistent. It also found that Mr Bricknell did not lie about Mr Wilson finding blood-stained tissues in a waste basket in the bathroom; if he wanted to lie it was ‘unlikely that he would have said that initiative came from his daughter’s boyfriend’ but would have presented it as his own observation.

73. Correctly, with respect, it doubted the usefulness of Mr Madladla’s observation following the in loco inspection, the lapse of 18 months after the incident and the environmental changes since. Cautioning itself that Mrs Ndlovu was a single witness, it nevertheless could not understand why she would add the apparently worthless snippet of evidence that a person at the appellant’s residence walked down to the lake and up again if she did not see this happen.

74. The significance for the trial court of the evidence of Mrs Ndlovu and Mr Madlala was that they both observed and heard some kind of noisy activity at the Shaw residence immediately before hearing the gunshot. Mr Madladla’s evidence that the appellant left the premises for the first time only after the shot was fired was inconsistent with the appellant’s evidence.

75. Inspector Mchunu confirmed that from the outset the appellant presented that the deceased committed suicide and he had never been informed that the police should start investigating an attack by an intruder. He was aware of the blood stained tissue in the waste paper basket.

76. Accordingly, having excluded as reasonable possibilities the suicide and intruder theories, the trial court accepted as the only reasonable inference that the appellant shot the deceased and that he inflicted the fresh injuries found on her face and on her finger. It concluded that the appellant pressed his firearm up against the deceased and shot her whilst forcing her against the open bathroom door, with the intention of not merely injuring her but to kill her because, if she lived to tell the tale, she would testify against him. From her position and her fresh injuries, the trial court concluded that there was some sort of struggle between the appellant and the deceased which resulted in him shooting her in the position that he did. It concluded that the only reasonable inference was that he had the direct intention to kill her.

Demeanour as a guide to credibility

77. The trial court drew inferences favourable to the credibility of state witnesses and fatal to the appellant on the basis of their demeanour. How do the authorities approach demeanour evidence?

78. Demeanour means much more than the appearance of a witness in the box. It includes witnesses’ manner of testifying, character, personality and the impression they create. Whether they are candid or evasive, ready or reluctant in giving their version, whether they hesitate unnecessarily, fidget nervously, twitch facially in response to straight forward questions, and ‘a thousand other considerations’ cumulatively contribute to shaping demeanour. ²

79. The triers of fact are best placed to make findings on demeanour and to factor such findings in assessing credibility. However, in making such findings, they should be mindful that demeanour in itself is a fallible guide to credibility. It can mislead.³

80. Demeanour is seldom ever decisive in determining the outcome of a case. On their own, findings on demeanour have limited value.⁴ Demeanour should be considered with all other factors,⁵ including the probability of the witness’ story, the reasonableness of his conduct, his memory, the consistency of his version, and his interest in the matter.⁶

81. As the Constitutional Court has pointed out,⁷ it is dangerous to assume that all triers of fact have the ability to interpret correctly the behaviour of witnesses; a trier of fact may miss entirely the nuances in the testimony of someone whose life experience differs fundamentally from that of the trier of fact. Conduct that elicits empathy from one adjudicator may seem irrational, inexplicable, odd and even suspicious to another adjudicator. Harder still is to anticipate let alone judge how people behave in the trauma of life and death situations. Some comfort can be drawn from the fact that judicial discipline, the rule of law and the right to appeal itself filter out as far as possible the subjective personal predilections and sensitivities of individual adjudicators. However, these tools are not foolproof.

82. The risks of accepting demeanour evidence is diminished if the evidence accords with the inherent probabilities, is corroborated, is not contradicted, or if it is contradicted then only by evidence of poor quality given. ⁸ Demeanour should be measured against adequate facts and tested against probabilities and improbabilities of the case as a whole.⁹ An appellate court must attach weight but not excessive weight to a trial court’s finding on demeanour because that court is in a better position to make such findings.¹⁰ However, it may interfere with a trial court’s evaluation of testimony in exceptional circumstances.¹¹ In that event, it is not obliged to accept a trial court’s finding on demeanour as conclusive. In this case, the trial court helpfully recorded its impression of the demeanour of some of the witnesses.¹² Were the inferences the trial court drew from its observations of witness’ demeanour the only reasonable inferences?

Credibility and reliability

83. Mr Bricknell’s relationship with the appellant was not ‘fine’. It was palpably tense and riddled with suspicion. He arrived on the scene with a request from his wife to take the deceased’s identity book because she expected the appellant to be difficult. Mr Bricknell proffered no rational basis for his suspicion other than his ‘gut instinct’. For no reason other than that he had just learnt that his wife’s sister had died as a result of a firearm accident he wanted to search the appellant immediately after meeting him at the gate. He presented as if he was close family to the Shaws but he led no evidence to support the insinuation that the appellant abused the deceased.

84. He assumed the role of investigator notwithstanding the presence of the police. He cast the appellant in a bad light, without fully exploring all the possibilities. For instance, he assumed that the appellant was consuming alcohol without making any attempt to ascertain whether the glass he held in his hand at the bar contained alcohol.

85. Mr Bricknell discovered the blood-stained tissues after Mr Wilson told him that the appellant was acting strangely by going to the toilet frequently. Although Mr Wilson initiated the suspicion about the frequent toilet visits, it was Mr Bricknell who linked them to the blood-stained tissues. That Detective Inspector Mchunu did not trouble himself to even view the tissues let alone subject them to forensic analysis confirms that the police did not consider this material to their investigation. Furthermore, the fact that the appellant was not cross-examined to explain this reinforces my view that even the prosecution placed little store on the bloodied tissues. Notwithstanding his suspicions, Mr Bricknell made no effort to ascertain when and who threw them there. Many people were in the house from 6pm to 11pm; anyone could have thrown them there. The trial court heard the appellant denying that he even had a box of tissues in the house; others who were on the scene did not mention the tissues.

86. Mr Bricknell tried to imply that there was a sinister reason for the appellant not contacting him first as he was a relative living closest to them. The appellant’s explanation for contacting Mr Kruger first was logical: the latter knew the co-ordinates of the area to direct an air ambulance.

87. Mr Bricknell contradicted himself about whether he questioned the appellant about the firearm and about where Mr Wilson sat when they drove from the gate to the house. He was also evasive and contradicted himself about searching the appellant for a firearm. Initially he testified that he told the appellant that he wanted to search him. Only in response to questions from the court did he say that he actually searched him. Under cross-examination he evaded the question whether the appellant consented to being searched. He contradicted himself about when he asked the appellant about the firearm, until the learned Judge corrected him.

88. Mr Bricknell was unlikely to have searched or asked to search the appellant because he had no authority or basis to do so other than being irrationally suspicious from the moment he met the appellant at the gate, if not before. Furthermore, the police were already there. Being reluctant to ask the appellant to even move away from the bar, he was unlikely to indulge in the more invasive and offensive act of searching him, especially as the appellant would have resisted being search. Contrary to the trial court’s finding that this evidence was insignificant, whether Mr Bricknell searched the appellant was material not only to fortifying the reasonableness of his suspicions but also to defining his relationship with the appellant, and consequently, his impartiality and credibility as a witness.

89. Significantly, Mr Bricknell tried to influence the court by opining that the appellant’s narrative of the events of that afternoon was inherently inconsistent. He proffered little detail to demonstrate the basis of his opinion. He persisted in this view, notwithstanding his concession that he was ‘battling to remember’. Such evidence as he did proffer proved the opposite. The appellant’s narrative immediately after the incident and in court remained consistent. His uncontradicted evidence was that the deceased’s mood was initially deflated and depressed but she improved to being relaxed and loving. Mr Bricknell’s recollection of the appellant’s account that she died in his arms after he retrieved his sms’ is also consistent with the latter’s evidence in court. Inadvertently, Mr Bricknell corroborated the appellant’s evidence that he cared for the deceased when he testified that he went out to buy the deceased medication for her menstrual pains.

90. Mr Madlala was on the roof of a house about 1.1 kilometres away. He saw the appellant’s bakkie being driven out once but not when it returned. Following the in loco inspection the parties agreed that from point H where Mr Madlala was, no vehicles could be seen or heard leaving point G, the appellant’s house. From point H only the movement of a vehicle could be observed at point G if it was against the white background of the house. Therefore, he could have seen the roof of the appellant’s bakkie as it went up the driveway for about ten to twelve metres only. He could not have seen how many people were in the house and what gender or race they were. He would not have been able to see the main entrance, the kitchen door or a person on the veranda but he would have been able to see a person exiting the house from the veranda.

91. Mr Madlala did not observe the Shaw residence the entire time that he was on the roof; waterproofing the roof on a windy winter afternoon with the sun setting would have required concentration. If he had been attentive to the goings-on at the Shaw residence he would have noticed the white bakkie go up the driveway from about 4:30 pm on at least three occasions.

92. Mr Madlala made his statement to the police three months after the incident. There were material differences between his statement to the police and his evidence in court. Even if his evidence is accepted, it has to be weighed against the appellant’s evidence. His evidence does not assist the court to draw reliable inferences that point to the appellant killing the deceased intentionally. However, he corroborates the appellant in two respects, namely, that the appellant drove out in his bakkie just before 5 pm and that there were shadows over the Shaw residence.

93. If Mrs Ndlovu did see the appellant walk to and from the lake, it would shatter his credibility. For that, Mrs Ndlovu’s evidence has to be accepted as true beyond a reasonable doubt. She had not mentioned seeing the appellant walk down to the lake either in her statement to the police or in discussion with Mr Kruger. Arising from the in loco inspection it was common cause that from point J where she stood neither the main entrance nor the appellant’s kitchen door could be seen; a person exiting the house from the veranda could be seen; however, it was not possible to distinguish between gender, race or identity of a person at the appellant’s house.Therefore she could not positively identify that person as the appellant from that distance, especially as she, as a diabetic with poor eyesight, did not wear glasses at that stage. If she saw a person she could only assume that it was the appellant. She could be wrong in making such an assumption, just as she wrongly assumed that the noise she heard came from children playing. Her evidence about hearing children chasing each other around the house at such a distance is so improbable that it taints the reliability of other observations she allegedly made.

94. To the extent that her evidence is corroborated by the inspection in loco it is admissible. Her evidence that a person exiting the house from the veranda could be seen, is therefore admissible. However, on the whole her evidence is so unreliable that no inferences can safely be drawn from it. However, like the trial court, I doubt that Mrs Ndlovu had a motive to be deliberately untruthful. Whether she testified about matters she could not have witnessed because she wanted to be helpful or for other reasons are best known to her.

95. Mr Madlala and Mrs Ndlovu corroborated each other to the extent that they confirmed that it was windy that day; so it was possible that the wind carried the sound of the gunfire in their direction, as the in loco inspection confirmed. They did not corroborate each other about whether a man exited the Shaw’s house and walked to the lake and whether there were people or children running around the house. Little reliance can be placed on the evidence of Mr Madlala and Mrs Ndlovu to contradict the appellant’s version.

96. Turning to the experts, the general rule is that opinion evidence is inadmissible because it lacks probative value and is therefore irrelevant.¹³ However, as an exception to the general rule, opinion evidence is readily received on ballistics, medicine, psychiatry and other fields in which the skills and knowledge of the court could be supplemented.¹⁴ In a murder case which must be decided substantially on circumstantial evidence, credible and reliable expert opinion evidence on ballistics and forensic pathology is indispensable. ¹⁵ However, the relevance and admissibility of their opinions are confined exclusively to matters specifically within their expertise. Having established facts within the realm of their expertise, the inferences they draw from those facts to found their opinions are relevant only to the extent that they fall within the scope of their expertise. Opinions expressed beyond the scope of their expertise are unreliable, irrelevant, and inadmissible.¹⁶

97. The qualifications and expertise of both experts in this case were accepted. On the whole both experts were independent and impartial. However, the weight to attach to their findings, opinions and inferences fall exclusively upon the court to evaluate in the context of all the evidence and the law.

98. Detective Inspector Mchunu should have been another key witness as the investigating officer in a case built exclusively on circumstantial evidence. His contribution has been shoddy and haphazard. He did not investigate the possibility of the deceased being shot by an intruder. He heard about the intruder theory for the first time when it was put to him in examination in chief. Despite being present from about 21:00 on the night of the shooting, and finding blood on the carpet and the safe curious, he took no samples of it for forensic testing. Neither did he trouble himself to investigate the blood-stained tissues in the wastebasket. He did not test the firearm for fingerprints. Without a full investigation of the intruder hypothesis, the evidence for and against such a possibility is incomplete. His evidence about the blood stains in the back room was not corroborated by the experts. If those stains could have been corroborated by other policemen they were not called. Disappointingly, Detective Inspector Mchunu added little value in a case where the investigating officer had a critical role.

99. Except for his ‘excessive and irritating’ use of the word ‘certainly’ the trial court could not fault the appellant’s demeanour. As a ‘highly intelligent individual’ who had ‘considered every aspect’ and ‘all the objective evidence’, he had a ‘ready answer for all questions’, the trial court opined. Inferring that his ‘intelligence and sophistication’ enabled his faultless demeanour, the trial court questioned his truthfulness. It found his evidence to be ‘patently a fabrication’.

100. With respect, the appellant’s intelligence, sophistication and ready answers cannot without more count against his credibility. Usually, having ready answers count in favour of the credibility of a witness. Guilty or not he would have considered every aspect of the incident. His evidence should be rejected if it is materially contradictory, improbable, incredible, unreliable or otherwise unsatisfactory.

101. I find that his recollection of the events of that afternoon was lucid notwithstanding his distress and trauma. Nowhere in the entire record did the appellant contradict himself. The same cannot be said of the evidence of Mr Bricknell and the neighbours. The appellant’s evidence in cross-examination is almost identical to his evidence in chief and his report to Mr Bricknell. He repeatedly resisted invitations to speculate when responding to questions for which he himself might have been searching for answers.

102. As elucidated below, corroboration for the appellant’s version emerges from Mr Bricknell, the cellular phone records, the photographic exhibits, the in loco inspection report, Dr Perumal, the blood stains, the absence of primer residue on his hands and clothing, and, to a lesser extent, the neighbours. Overall, the appellant’s evidence is consistent both within the content and structure of his own evidence and with the objective facts.¹⁷

103. Even if he did shoot the deceased, some aspects of his version ring true. He narrated such details of the incident that they must have happened. For instance, her Croc shoes getting in the way when he knelt on the floor to kiss her, was an irrelevant detail that stood out in his recollection of the events of that afternoon. Notwithstanding the brevity of his statement in terms of s 115 of the Criminal Procedure Act 51 of 1977, he mentioned it there and repeated it in his examination in chief. He was shaking too much to take her pulse. Panicked, he did not pick up his cellular phone with his keys. He wiped her blood from his hands on the seat of his pants in the urgency of the moment. She was not dead when he found her. He volunteered evidence that Inspector Du Preez who knew him and the deceased cold-shouldered him, an irrelevant detail which did not support him.

104. In so far as the appellant’s evidence conflicts with evidence of Mr Madlala and Mrs Ndlovu, his evidence should be preferred as I have found the latter to be unreliable. As Mr Bricknell was suspicious of the appellant from the outset, he was not independent; his evidence has to be approached with caution. In any case, like the investigating officer, he shed little light on how the shooting occurred. In summary, mainly the evidence of the experts establishes a case for the appellant to meet.

Circumstantial evidence

105. On the facts, many permutations of possible inferences arise to fire the imagination of any murder mystery writer. However, speculating and fantasising about facts not proved or selecting some reasonable inferences and ignoring others is firmly disallowed by the authorities discussed below.

106. Circumstantial evidence is any fact from which a fact in dispute may be inferred.¹⁸ Such facts have to be proved by direct evidence.¹⁹ Conclusions drawn from evidence not proven or admitted are speculation not inference.²⁰ The challenge is to draw the most reasonable inferences from the proven facts to establish the guilt of the appellant beyond reasonable doubt, without overlooking the possibility of other equally probable or reasonably possible inferences.

107. This approach to circumstantial evidence constrains adjudicators of fact from over-zealously exercising their imagination by filling the information gaps to construct theories to explain their conclusions. Such creative enterprises risk overlooking inconsistent circumstances or assuming facts which have not been proved or cannot legitimately be inferred.²¹

108. Referring to R v Blom 1939 AD 188 at pp 202 and 203,²² The South African Law of Evidence encapsulates an enduring rule of logic applied to circumstantial evidence in criminal trials in the following extract: ²³

‘a. The inference sought to be drawn must be consistent with all the proved facts. If it is not, the inference cannot be drawn.

b. The proved facts should be such that they exclude every reasonable inference from them save the one sought to be drawn. If they do not exclude other reasonable inferences, then there must be a doubt whether the inference sought to be drawn is correct.’

109. It follows that it is not for the trier of fact to speculate ‘as to the possible existence of facts which together with the proved facts, would justify a conclusion that the (appellant) is innocent.’²⁴ In a murder case in which the state has not established the cause of death and the guilt of the appellant rested on circumstantial evidence, the majority in the erstwhile Appellate Division held that other indications of an intent to kill had to be very strong if they are to make up for serious deficiency and leave no reasonable doubt. Inferences cannot be drawn from conjecture or speculation.²⁵

110. In R v De Villiers 1944 AD 493 at 508 – 9, it was held that a court should not consider each circumstance in isolation and draw inferences from each single circumstance. The onus on the state is not to prove that each separate item of evidence is inconsistent with the innocence of the accused, but that taken as a whole, the evidence is beyond reasonable doubt inconsistent with such innocence.²⁶

111. Other commonwealth jurisdictions adopt a similar approach to circumstantial evidence in criminal cases. In Shepherd v R [1990] HCA 56; (1990) 170 CLR 573 (19 December 1990), the High Court of Australia confirmed that in cases based on circumstantial evidence, juries cannot use a fact as a basis for inferring guilt unless that fact is proved beyond reasonable doubt. Likewise, in the United Kingdom, facts which establish a link in the chain of reasoning towards an inference of guilt must be proved beyond reasonable doubt.²⁷

112. Chamberlain v R (2) [1984] HCA 7; [1984] 153 CLR 521, colloquially renowned internationally as ‘the dingo case’, cautioned that facts considered in isolation may not lead to any inference of guilt or innocence; however, when evidence is considered together, its probative force is greatly increased.²⁸ As this was not a unanimous decision, it poses another question: Can guilt be found to be proved beyond a reasonable doubt if there is a dissent? If the dissent is based on inferences drawn from the circumstances by adopting the approach outlined above, the answer might well be ‘no’. Put differently, the dissent has to be wholly irrational in order for the majority guilty finding to prevail.

113. Adducing evidence in chief and cross-examination effectively in cases where the evidence is entirely circumstantial assumes special importance. Testing all reasonable possibilities fully is indispensable not only for discrediting evidence and for enabling the witnesses to refute attacks on their credibility,²⁹ but also for eliciting for the trier of fact, not speculation, but explanations which, if reasonable, would enable appropriate inferences to be drawn. Failure to examine and cross-examine effectively may not only bar a party from later seeking to draw inferences from facts not attested to or disputing the truth of a witness’ evidence,³⁰ but also impair the ability of the trier of fact to draw the most reasonable inferences.

114. Does the circumstantial evidence elevate the possibility that the appellant shot the deceased intentionally to a certainty? In other words, do the material circumstantial facts, viewed cumulatively, establish this as the only reasonable inference? To answer this question the facts and inferences for and against each of the three possibilities are assessed.

The intruder theory

115. On the available evidence this theory is weak. Nothing was missing from the house, not even the firearm used to shoot the deceased. Mr Bricknell who accounted for the deceased’s handbag and its contents had no knowledge of the missing jewellery, which could have been mislaid or stolen by someone other than an intruder.

116. There were no signs whatsoever of a third party having entered the premises. The best evidence of this fact came from the appellant himself when, on discovering the deceased, he entertained no thought of a third party being involved. He agreed that it would have been fortuitous that an intruder would break in during the 15 minutes he was away checking his sms. An intruder would have had to observe the house to time his entry at just the moment when he was away. Having observed the house an intruder would also know that the appellant would be away for no more than 20 minutes. Furthermore, the gruelling terrain counted against the intruder theory, which I agree with the trial court, should be rejected as improbable.

The suicide theory

117. The deceased had attempted suicide twice before by overdosing herself with tablets. In the past, she had consulted a psychotherapist and a psychiatrist. She suffered from depression. Although she was no longer on medication, her condition had been serious enough for her to have received medication and professional treatment before.

118. She was a mother of two teenage sons from whom she was unhappily separated. Anxiety, perhaps even guilt, could have contributed to her distress. Her conflict with Mrs Smith ran deep, especially as it related to her children. She invited herself over to discuss it with her sister, Mrs Bricknell. On Friday she was deflated. Even though she appeared to have recovered from her altercation with Mrs Smith and seemed relaxed, she also mused over the issue again on Sunday afternoon with the appellant. She was so deeply conflicted about Mrs Smith caring for her sons that the appellant was moved to suggest that she return to Durban to be with them.

119. She was menstruating, in pain and could have been more emotional than usual. She either had an unusually high tolerance for or she abused alcohol, which also suggests emotional and psychological instability.

120. Although her work and financial interests were on a positive path, there is no evidence that she showed any interest or enthusiasm in these developments. It was the appellant who moved matters along for her by setting up appointments with her lawyers. When they discussed their plans for the following week shortly before she died, her dilemma about her children continued to trouble her to the point that she reintroduced the topic.

121. Although shooting herself on her right shoulder would have been awkward, it was not impossible, especially if she used both thumbs to pull the trigger. The angle and direction of the shot also did not exclude suicide. The opinion evidence that firing only one shot is atypical of women who commit suicide by shooting themselves is a generalisation which cannot, without more, be admitted as a fact from which any reasonable inference can be drawn in this case. Besides, to be admissible, such evidence must emanate from experts in the field if say psychology or psychoanalysis. Furthermore, if the deceased was as intoxicated as Dr Perumal testified, to predict how she would have held the gun would amount to pure speculation.

122. The experts on ballistics and pathology did not know the deceased and were not aware of her mental state. They were especially not aware that the deceased had twice attempted suicide. Or that the appellant had saved her on both occasions. Surprisingly, the judgment of the trial court also does not refer to the deceased’s failed suicide attempts. For reasons not apparent from the evidence, the previous suicide attempts were downplayed before the trial court.

123. The trial court found that shooting herself once in the shoulder did not signal a suicidal intention. Having regard to her two previous suicide attempts, it is an open question as to whether she really intended to end her life. As with her two previous suicide attempts, when the appellant rescued her, she knew that he would return soon to save her. In the circumstances, the suicide theory is a real possibility.

Murder

State’s Case

124. The high water mark of the state’s case is the circumstantial evidence and the opinions of the experts based on it. The inference they drew that the shooting was inconsistent with suicide is admitted insofar as it is drawn exclusively from the narrow perspective of their respective expertise. However, the court must weigh their opinion against all other evidence before it concludes that the only reasonable inference is that the deceased did not commit suicide but was shot.

125. Neither expert was qualified to express a reliable and admissible opinion on the mental and emotional state of the deceased and, from that perspective, her propensity to commit suicide. If they had been aware that the deceased had attempted suicide twice before, they might have drawn other inferences or been less convinced that she did not commit suicide. Furthermore, unlike the court, they did not have the benefit of all the available facts from which to draw the most reasonable inferences. Ultimately, the experts could neither exclude the possibility of suicide nor point to the appellant as the murderer.

126. Upon my finding that the suicide theory is a possibility it follows automatically that there is reasonable doubt about whether the appellant shot the deceased. Weighing in the defence below fortifies this conclusion.

The Defence

127. Critical to the credibility of the appellant is proof of the nature of his relationship with the deceased. The appellant painted a picture of an ordinary middle-aged couple living together, sharing the pleasures and problems that many couples experience. He loved and cared for her. The slightest evidence to gainsay this image of normality and harmony would destroy his defence.

128. For reasons unknown to the court, independent witnesses gave evidence that shrouds appellant in suspicion; in the case of Mrs Ndlovu, it is even incriminating. The deceased’s mother was prepared to incur the cost of a forensic pathologist instead of accepting the appellant’s explanation. This milieu of suspicion suggests that there could be more to the appellant’s relationship with the deceased than he disclosed to the court. However, although the state put to him that he abused the deceased, it led no evidence in support of this proposition. Even Mr Bricknell, who was disposed to painting the appellant in a bad light, failed to advance this evidence. On his version too, the appellant was not the cause of her emotional distress, Mrs Smith was. Without reliable evidence to the contrary, I must accept that the relationship was loving and harmonious.

129. Other than a vague insinuation that the appellant abused the deceased, no motive for killing her was canvassed at all. Proof of motive for committing a crime is highly desirable to establish the intention.³¹ Without proof of any motive the state fails to prove an intention to kill which is an element of the crime of murder. Why the appellant would shoot the deceased when he had saved her life twice before remains unexplained. Instead, corroboration for the appellant emerged from several sources.

130. Evidence that he pampered the deceased emerged from an unexpected source. Mr Bricknell was neither surprised nor did he dispute that the deceased directed the appellant to ready the house for paying guests whilst she lay on the couch. That the appellant served her coffee and breakfast in bed went unchallenged. Mr Bricknell confirmed that the fire had been lit in the lounge. The appellant testified that he lit the fire because the deceased felt cold. Mr Bricknell seemed to accept the appellant’s description of the deceased being her usual self. Evidence that the appellant reported that he went to Nottingham to buy medication to ease the deceased’s menstrual pains came from Mr Bricknell himself and not the appellant who did not repeat that evidence in court even though it favoured him.

131. Blood stains on the safe door also corroborated the appellant’s testimony that he did not immediately see the trail of blood from the veranda into the house where the deceased was shot and where he found the pistol. In his evidence in chief the appellant was not invited to explain the blood stains on the safe door. When cross-examined about it, he seemed to search for an explanation. If he deliberately smeared blood on the safe door to corroborate his version that he went in search of the weapon, he would have put that version upfront at the first opportunity in examination in chief instead of hoping for an opportunity to give this explanation under cross-examination.

132. Photographic Exhibits B 11 and 12 show the smudges on the opening edge of the safe door, extending from left to right for some 150 millimetres, and not ‘from right to left’ as the trial court erroneously found. Furthermore, the fact that there was no blood on the handle of the door corroborates the appellant’s evidence that the safe door was unlocked and could be opened by holding the edge; hence he had no need to touch the handle of the door. As for the apparent absence of blood on the edge of the door, holding it on its narrow edge to pry it open might not have left visible signs of blood. Other than the photographic evidence of the door, no other evidence was adduced to prove that there was no blood on the edge of the door.

133. The appellant proffered the only explanation for the blood stains on the safe. No other version was canvassed during the trial. His explanation that he opened the safe in search of the gun is plausible and has not been gainsaid. To reject this explanation the court must find that it is inherently improbable and that the appellant is lying. The court has no basis to do that. Conjuring any other explanation would be speculation. Photographic exhibits 14 and 15 show that the deceased had no blood on her clothing on her left side. This explains why he had no blood stains on the front of his jacket as a result of holding and kissing the deceased, he must have knelt on her left as he bent over her.

134. Dr Perumal also corroborated the appellant in at least two respects: after looking at the photographic exhibits, he confirmed that it was not easy to see the high velocity splatter of blood on the dark paving where the deceased lay. Furthermore, his evidence that she would have survived for about 10 to 15 minutes corresponds with the appellant’s evidence that he returned to the house in about 15 minutes to find her still breathing.

135. The appellant’s cellular phone records also corroborate his version not only about the calls he made but also about the times he was at point K. Although they do not confirm that at the most crucial moment of the shooting he was at point K because he neither sent nor received messages, it does confirm that from 16:59:10 he made the calls he attested to.

136. Corroboration also emerges from the in loco inspection which confirmed that from point K one could not see the Shaws’ house (point G). More importantly, gunshots fired at the house from the bathroom with the windows closed and open could not be heard at point K. This explains why the appellant did not hear the shooting. As only a dull sound was heard at point J (Mrs Ndlovu’s viewpoint) and at point H (Mr Madlala’s viewpoint), it is doubtful that Mr Madlala and Mrs Ndlovu recognised it at the time as a gun shot or been particularly disturbed on hearing the ‘explosion’. Neither reported nor investigated the cause of the explosion probably because neither considered the sounds and movement unusual.

137. If Mr Madlala is to be believed he also corroborated the appellant about seeing shadows at the Shaw homestead. However, how he could distinguish those to be of people from a distance of 1.1km is hard to fathom.

138. The absence of primer on the appellant’s hands and clothing corroborates his evidence that he did not shoot the deceased. However, this is not decisive proof because the deceased sustained a contact shot and his hands were tested long after primer residue became untraceable.

139. As for the quality of the police investigation, the state failed to investigate and adduce better evidence. The police did not even entertain the possibility of an intruder shooting the deceased let alone investigate it. Amongst other things, the police failed to take finger prints off the firearm, to test the appellant’s clothing sooner, to assess the crime scene immediately, to take samples of the blood that Inspector Mchunu said was smudged by the following day, and to photograph the bar which Mr Bricknell alleged had been cleared by the appellant. No evidence was led to rebut the appellant’s evidence of the two previous suicides. The insinuation that the appellant abused the deceased was not supported by any evidence.

140. As for the prosecution’s failure to cross-examine on material issues, it indicates not only short comings in the state’s case, but also acceptance of the appellant’s version. The prosecution failed to cross-examine the appellant about waiting for a sms instead of dialling his son for information; and about his contingency plan in the event of his homestead being attacked; about how his dogs would have behaved at the time of the shooting, all facts from which the trial court drew adverse inferences without his evidence being challenged and without him having an opportunity to explain himself fully. They were not the only or most reasonable inferences. Crucially, he was not cross-examined about the two prior suicide attempt, why he could not explain the deceased’s high level of inebriation, despite consenting to the ethanol test being admitted without proof and why he persisted that he was always aware of her level of inebriation when he did not explain how she could have consumed such a huge quantity of alcohol without him knowing.

141. As for the fresh injuries to her forehead and finger, the deceased could have sustained them if she stumbled from the bathroom where the shooting occurred onto the veranda. He did not observe these injuries initially possibly because there was not enough light; her hair might have obscured the injuries to her face or, because he was so traumatised, he did not notice such detail. In his own search for explanations he assumed that she had sustained these injuries when she fell because he knew that he did not cause them.

Are there inconsistencies in the appellant’s evidence?

142. The appellant described his state on discovering the deceased as shocked and extremely traumatised, unable to apply logic, distressed, hysterical and distraught. Mr Bricknell corroborated that he was distressed and crying. In that mental state his ‘instant conclusion’ was that ‘she had done something to herself’. He could not understand why but he came to that conclusion because he had not seen any one else around; nor had there been a confrontation while they were in residence in the thirty years he had the property. Hence he factored out a third person. And he knew where he had been and that he had not shot her.

143. By the end of the week he revised his initial deduction that she had ‘done something to herself’ and entertained the possibility of ‘something else’. As he had not been cross-examined on this issue, the trial court erred in finding that he abandoned the theory that she shot herself after seeing that the experts did not support it. He did abandon it but there was no evidence that he did so because he had seen the experts’ reports. His counsel nevertheless persisted with it as a possibility. Although his search for answers led him to consider the intruder theory, he showed no conviction in it. Unequivocally, he conceded that an intruder shooting her would have been purely fortuitous. His instant assumption that she had ‘done something to herself’ is corroborated by his search for the firearm which, in turn, is corroborated by the blood on the safe door; if he shot her he would have known where the firearm was without having to search for it. Viewing his evidence as a whole, he had not totally relinquished this assumption.

144. The appellant noticed the blood stains for the first time on the cream carpet near the bathroom when, almost an hour after he found the deceased, he went in search of the weapon. That was when he saw the trail lead on to the quarry tiles. His explanation for not seeing the trail earlier was that the blood was on dark quarry tiles, a fact corroborated by Dr Perumal. Furthermore, when he entered the house he was not looking at the floor but ahead for the deceased. As it was getting dark the shadows were also long on the floor inside, a fact corroborated by Mr Madlala. Therefore this does not contradict the appellant’s earlier evidence that with the sun setting at 17:05 it was light outside.

145. The appellant attested in chief to the contingency plan if they were attacked on the first day of his testimony. The trial court first erred in finding that he attested to this only on the second day of his evidence. Second, based on this erroneous finding, it also inferred that he must have contrived the contingency plan. His explanation that as a habit he left a loaded firearm for the deceased’s protection whenever he left her alone and went to Nottingham Road was neither unreasonable nor contradicted.

146. The trial court rejected the appellant’s theory that the deceased could not have been surprised in the bathroom because if she had been surprised she would not have had an opportunity to take out the firearm from her cupboard. Invited to speculate under cross-examination, the appellant opined that the deceased might have taken the firearm from her cupboard as she ran into the bathroom but the intruder caught up with her. That either she or the intruder already had the pistol when she got to the bathroom is an objective fact. It is unlikely that an intruder would have found it concealed in her cupboard in such a sort time. Therefore, his theory that she could not have been surprised in the bathroom because she must have taken the pistol from her cupboard is plausible.

147. Undoubtedly, the shot was fired in the doorway of the bathroom. The experts agreed that she was not immediately incapacitated. The droplets of blood suggest that the deceased moved from the en suite bathroom through the main bedroom, past the adjoining bedroom, past another bathroom, into the bar adjoining the lounge, where she u-turned and exited onto the veranda. With no signs of a struggle, the droplets of blood, the absence of more blood stains on the appellant’s clothing and the fact that the deceased was not immediately incapacitated tend towards proving that the deceased made her way unassisted from the bathroom to the veranda. This supports an inference that the appellant was not there when the shot was fired. Importantly, it also corroborates the appellant’s denial that he caused the fresh wounds. If the appellant was not there when the shot was fired and there was no evidence of any sign of a struggle, better evidence was required to exclude the possibility that the deceased did not injure herself by, for example, bumping herself against the walls and furniture as she stumbled on to the veranda. The trial court erred in finding that the appellant inflicted the fresh injuries.

148. According to Dr Perumal, the deceased bled to death from the gun shot wound. There is no evidence that the appellant allowed her to bleed to death or that he did anything to expedite her demise. On the contrary, on discovering that she was shot but alive, the appellant concentrated on keeping her alive. Discovering who shot her and where the weapon was were not his primary concern. After telephoning Mr Kruger the first time the appellant contacted him twice more within the hour to check whether an air ambulance was on its way. He was anxious to get medical attention to her. If he had shot her he would not want her alive to tell the tale. Insofar as the shot in the shoulder suggests that he might have shot her unintentionally, perhaps in a drunken brawl, this was not canvassed at the trial. An accidental shooting was not the state’s case. The only mention of an accident came from Mr Bricknell who testified that Mr Lambert had initially informed him that the deceased had been shot accidentally. Mr Lambert was not called to expatiate on this hearsay evidence.

149. The trial court found that if the deceased was determined to end her own life, she had enough time to fire again into a more definitive location. On analogous reasoning, if the appellant had killed her intentionally, he would have chosen a more fatal spot to shoot her so that her death would have been instant and certain.

150. The reason he went to point K on Sunday morning was to discuss the property transaction with his advocate and his son who had taken documents to the advocate. Whilst it is improbable, though not impossible, that he would have access to an Investec employee on a Sunday, access to his advocate, who was also his friend was not improbable. His cellular phone records show that he contacted his advocate and his son on that Sunday morning. Expecting a communication form them on Sunday afternoon is possibly true. Furthermore, as he was not cross-examined as to why on a non-business day he went up the hill to point K, to wait for a sms that did not come, no adverse inference should have been drawn by the trial court, especially as his evidence that his trips to point K were routine was also not challenged.

151. The appellant volunteered information that he became aware that the dogs were not around when he found the deceased. At the invitation of his counsel to describe the dogs he explained that they were not vicious. In expatiation, he speculated that they would have barked and followed someone off the property. He could only speculate because something like this had not happened before. His speculation should therefore not have attracted an adverse credibility finding. The more obvious explanation that dogs usually flee at the sound of an explosion such as gun shots and fire crackers was not canvassed in the trial court.

152. A careful trawl through the appellant’s evidence suggests that he downplayed the possibility that the deceased abused alcohol and committed suicide. Notwithstanding his puzzlement about the high level of alcohol, the appellant accepted the results of the ethanol tests without evidence being led to prove them. Consequently, the court is bound to also accept Dr Perumal’s inference that she had consumed about 1.5 litres of wine.

153. The deceased either abused alcohol or had a high tolerance for it. Her drinking habits were a concern to the appellant as it induced her epileptic fits. His comment that she was starting rather early also suggests that he might have had more reasons than her epilepsy to caution her about drinking. She could consume as much as twenty glasses without collapsing. Dr Perumal also did not discount alcohol abuse. Significantly, one of the first questions Mr Bricknell asked the appellant was whether he and the deceased had been drinking. The consumption of alcohol by the deceased and the appellant was a concern for Mr Bricknell for reasons he did not explain.

154. Notwithstanding his persistence that he was always aware of her level of inebriation, the appellant did not proffer any explanation for the deceased being so inebriated without his knowledge. He avoided acknowledging that she abused alcohol. He maintained that it was not normal for her to drink on her own. Although he found her standing at the bar when he returned from putting away his tools, he immediately clarified that the bar was a ‘convenient sitting talking place…not necessarily a place where we only…drank.’ Acknowledging her alcohol abuse would have presented the deceased as emotionally unstable and fortified the suicide theory. This is not how he articulated his defence when he testified. He invoked it in his affidavit in support of his petition for special leave to appeal. Significantly, his affidavit was made in August 2009 after a lapse of more than 2 years after the deceased died and when he faced the prospect of imprisonment.

155. He disclosed her previous suicides only at the end of his evidence in chief and on being asked specifically about them. As a result, state witnesses were not cross-examined about it. The state neither cross-examined him nor rebutted this evidence. From the outset he said that the deceased had done something to herself rather than that she had committed suicide. His description kept open possibilities other than suicide to explain the shooting. He was not probed either in examination in chief or cross-examination to explain his choice of terminology.

156. Notwithstanding the appellant downplaying and rejecting the suicide theory even when his counsel relied on it, it was his instantaneous assumption she had done something to herself. Only later when the ‘blur’ abated and he was psychologically more stable did he entertain the intruder theory in his own pursuit for an explanation. This was his uncontroverted evidence from which the trial court drew an adverse inference.

157. The deceased’s apparent good mood was as much for him as for Mr Bricknell and the trial court, irreconcilable with the suicide. However, according to ‘Gunshot Wounds’, to which Mr Pitman referred, it is not uncommon for suicide victims to be relaxed and happy immediately before committing suicide because they have found a solution to their troubles in the suicide. Therein might be an explanation but not one from which the court can draw any inference in the absence of expert evidence.

158. The appellant’s apparent reluctance to rely on the deceased’s previous suicides and his inability to explain her level of inebriation, both of which ironically reinforce his defence, calls for explanations which are not forthcoming from the record. Shock, denial, disbelief that a loved one is dead, and in the case of suicide, even shame, guilt and a sense of having wronged the deceased could overwhelm survivors to the point of them falsifying the reasons for the death. An explanation for the link between appellant’s conduct, his emotions and his thoughts immediately after the incident to his testimony in court might have been forthcoming if he had been examined and cross-examined to that effect, or better still if he had been assessed by a psychologist, psychiatrist or ideally a thanatologist. It might also explain whether and why he was reluctant to accept the deceased’s level of inebriation despite the forensic evidence, and whether his conduct and observations could be a consequence of trauma, distress and nervous shock.

159. The appellant almost did not proceed with this appeal because he had no funds to pay for his legal representation. Consequently, the possibility of being psychologically evaluated and leading that evidence eluded him. As the evidence stands, the court has no explanation as to why he would downplay evidence that reinforced his defence.

160. Acknowledging that she abused alcohol and committed suicide implied that she was unhappy and emotionally unstable, which in turn reflected badly on him as her husband. It implied that he might have contributed to or been responsible for her distress, that he did not take care of her properly, that he failed to save her as he did previously. The possibilities are endless. As he was not examined or cross-examined about why he downplayed her possible emotional instability, the court cannot draw any inferences.

Conclusion

161. Having accepted the suicide theory, it follows automatically that there is reasonable doubt as to the guilt of the appellant. The appellant’s own version fortifies this conclusion. As a credible witness, his evidence that he did not shoot the deceased must therefore be accepted. The state has not proved the appellant’s guilt beyond a reasonable doubt. Cumulatively, the evidence supports the conclusion that the appellant is not guilty. The appeal succeeds.

162. The order I propose is the following:

1. The appeal is upheld.

2. The conviction and sentence are set aside.

DHAYA PILLAY J ______________________

C PATEL DJP ______________________

I agree

GRAHAM LOPES J _______________________

I agree

---