Khua Kian Keong and Another v Public Prosecutor
[2003] SGHC 238

1          The two appellants were convicted under ss 67(1)(a) and 67(1)(b) of the Road Traffic Act (Cap 276) (“RTA”) respectively. The district judge sentenced the first appellant, Khua Kian Keong (“Khua”) to two weeks’ imprisonment and a fine of $4,000. Additionally, Khua was disqualified from holding or obtaining all classes of driving licences for 48 months after his release from imprisonment. The second appellant, Pang Ee-Zian (“Pang”) was sentenced to pay a fine of $3,000, and was disqualified from driving for 24 months. The police had conducted the prosecution in the trial below. Both appellants appealed against their conviction and sentence. I allowed their appeals against conviction and now give my reasons.

Background

2          Khua claimed trial to the following charge:

You, Khua Kian Keong, Male, 33 years old, NRIC No: S6830009F, are charged that you, on or about the 8^th day of August 2001 at about 3.00 am, along Mountbatten Road towards ECP, Singapore, when driving motor car SCN 8780 U on the road was unfit to drive in that you were under the influence of drink to such an extent as to be incapable of having proper control of your vehicle, to wit, by driving in an unsteady manner, and weaving in and out and nearly grazed the left road kerb, and you have thereby committed an offence punishable under Section 67(1)(a) of the Road Traffic Act, Chapter 276.

And further that you, before the commission of the offence, that is to say on the 9^th day of May 1996, at Court 21 of the Subordinate Courts, Havelock Square, Singapore, had been convicted of an offence under Section 67(1)(b) of the Road Traffic Act,  Chapter 276 in TAC 120/96, which conviction had not been set aside.

Under s 67(1) of the RTA, mandatory imprisonment is prescribed upon a second conviction.

3          Pang faced the following charge:

You, Pang Ee-Zian, Male, 33 years old, NRIC No: S6800819J, are charged that you, on or about the 8^th day of August 2001, at about 3.02 am, along Mountbatten Road towards ECP, Singapore, when driving motor car SCN 8780 U did have so much alcohol in your body that the proportion of it in your breath exceeded the prescribed limit and you have thereby committed an offence punishable under Section 67(1)(b) of the Road Traffic Act, Chapter 276.

The prosecution’s case

4          According to the testimony of Senior Staff Sergeant Sairi Bin Aman (“Sairi”), he was manning a road block along Mountbatten Road, which was situated about 200 metres after a bend. At about 3.00 am, he saw the appellants’ car make a sudden turn into a bus bay after navigating the bend. It then stopped and its headlights were switched off. Sairi alerted Sergeant Arman Bin Mohd Ali (“Arman”), who was stationed at the bus bay, to the presence of the car. Before Arman could approach it, the car moved off towards the road block. Sairi observed that it was being driven in a “zigzag” manner between two lanes, and that it nearly grazed the left kerb at one stage. After signalling to the car to stop, he asked the driver, Khua, for his particulars before instructing him to pull the car over to the side of the road.

5          The car moved for about three car lengths to the side, while Sairi walked behind it. When the car stopped, Khua and Pang stepped out to change seats. The car was then driven for another two to three metres. Sairi ran to the car, shouting at them to stop. Khua was by then seated at the front passenger seat and he denied being the driver of the car. Sairi noticed that Khua smelt of alcohol and was unsteady in his gait. He subjected Khua to a breath analyser test and the result was 32 micrograms per 100 millilitre of breath. The legal limit is 35 micrograms according to s 71C of the RTA. Subsequently, Pang was interviewed by Sairi and he failed the breath test (result of 52 micrograms). Both were then arrested. Sairi tendered his log book in which he recorded this event after the arrest.

6          The prosecution’s case hinged solely on Sairi’s account, as the other prosecution witness, Arman, could not assist the court on any pertinent facts. He testified that he did not observe the car after it moved off and he then left the scene.

The defence

7          The appellants claimed that Khua had arranged to meet Pang and a friend, Chang Yoke Wooi (“Chang”), at a pub. Khua intended to hand over his car to Pang for the purpose of Chang’s wedding that day. Although Pang was supposed to drive, he and Chang drank brandy while waiting for Khua. They thus decided to call Chang’s brother, Shang Chung Yoke Choy (“Shang”), to come down and drive the car. While waiting for Shang, Khua consumed two glasses of brandy diluted with water. They decided that Khua would drive to his Khua’s home first before Shang took over. All four of them testified that Khua was not in a drunken state when leaving the pub. He had been convicted once before of drink driving and knew the dire consequences of another conviction. Pang sat in the front passenger seat while both Chang and Shang were in the back passenger seats.

8          Khua drove normally till he negotiated a bend and suddenly saw a road block ahead. He started panicking as he recalled that he had consumed alcohol, so he swerved into the bus bay located just after the bend. After being assured by Chang that he had not drunk much, Khua then proceeded to drive normally towards the road block. There were many police officers at the road block and there was at least one car being inspected in front of them. When it was the appellants’ turn, Sairi asked Khua for his driving licence and identity card. Khua also admitted to drinking alcohol upon being questioned. After the car was pulled over at the side of the road, Khua was unreasonably told to repeat the breath test at least three times. A heated argument between Khua and Sairi then ensued. Pang was also told to do a breath test, after which both Khua and Pang were handcuffed to be transported to the police station.

The decision below

9          The trial judge found that Sairi was a truthful witness and his evidence had no material discrepancies despite extensive cross-examination. Additionally, Sairi’s observations in his log book were given due weight as contemporaneous records. The judge dismissed the argument that it was ridiculous for the appellants to switch seats, since “one cannot account what desperation and panic can make a man do”.

10        The judge then found that the appellants’ version was not believable. Khua had probably panicked because he knew he was not in proper control of his car. His reason for driving was not convincing as he could have easily given Shang directions. It was also remarkable that Pang claimed that he could not remember many things and yet he could distinctly remember that he did not switch seats. The passengers, Chang and Shang, had reason to assist the appellants since they were close friends and drinking buddies. Their evidence was therefore treated with caution by the judge.

11        Finally, the judge refused to draw an adverse inference against the prosecution for failure to call other witnesses, under s 116, illustration (g) of the Evidence Act (Cap 97). He did not find that these witnesses were material and indispensable, since the prosecution’s case had been sufficiently proved by other evidence. Neither could he find any ulterior motive behind the failure to call other witnesses. He took cognisance of the seriousness of drink driving as an offence, and sentenced the appellants accordingly.

The appeal

12        Both appellants raised several arguments in common: -

(a) The trial judge failed to ascribe the correct evidential value to each witness. In particular, he relied solely on and placed undue weight on Sairi’s evidence;

(b) The evidence proffered by the appellants was rejected without any apparent reason. The trial judge did not address his mind to the testimony of the passengers, Shang and Chang; and

(c) An adverse inference under s 116, illustration (g) of the Evidence Act (Cap. 97), should be drawn against the prosecution, for its failure to offer to the defence the 11 other police officers present at the scene as witnesses.

Principles of appellate intervention   

13        It is beyond dispute that an appellate court will be slow to overturn findings of fact by the trial judge especially when an assessment of the credibility and veracity of the witnesses has been made: Ang Jwee Herng v PP [2001] 2 SLR 474 and PP v Hendricks Glen Conelth [2003] 1 SLR 426. I reiterated the requirements for appellate intervention in Ang Jwee Herng. The trial judge’s decision must be “plainly wrong or against the weight of the objective evidence before the court”.

14        If however the trial judge’s assessment of a witness’ credibility was based not so much on his demeanour as a witness, but on inferences drawn from the content of his evidence, the appellate court is in as good a position as the trial court to assess the same material: Awtar Singh s/o Margar Singh v PP [2000] 3 SLR 439. The appellate court might not be in a position to assess the witness’ demeanour, but it should not refrain from evaluating the conclusions of the trial judge based on all the facts known to him. The Court of Appeal in Bala Murugan a/l Krishnan & Anor v PP [2002] 4 SLR 289 provided an another expression, namely that interference is justified when “inferences drawn by the trial court were found to be not supported by the primary facts on the record”. I had summed up this principle in Sahadevan s/o Gundan v PP [2003] 1 SLR 145:

Nevertheless, while an appellate court should be reluctant to interfere with a finding of fact, it is always free to form an independent opinion about the proper inference to be drawn from a finding of fact.

15        My view is that the inferences drawn by the trial judge, independent of his observation of demeanour, were incorrect. These errors had led to convictions that were against the weight of the evidence. It was highly questionable to me whether the prosecution’s case was proven beyond a reasonable doubt.

Whether undue weight was given to Sairi’s testimony

16        Both appellants highlighted the danger of relying solely on Sairi’s evidence. There is no prohibition against relying on one witness. A conviction may be warranted on the testimony of one witness alone, so long as the court is aware of the dangers and subjects the evidence to careful scrutiny: Low Lin Lin v PP [2002] 4 SLR 14. Counsel for Pang claimed that the trial judge, despite acknowledging that this case hinged totally on the credibility of the witnesses and that a minute examination of their testimonies must be done, failed to make such detailed scrutiny. I found that this minute examination was indeed lacking. A careful scrutiny of Sairi’s testimony would have revealed many weaknesses which rendered the convictions extremely unsafe.

17        My greatest difficulty with Sairi’s testimony was his allegation that the appellants swapped seats after being stopped at the road block. According to his examination-in-chief, Sairi obtained Khua’s identity card and driving licence, ascertained their authenticity by looking at him, and shone a torch into the car which enabled him to have a good look at both appellants. Yet, as Sairi walked alongside the car which was moving to the side of the road, both appellants emerged from the car to switch seats in full view of Sairi. It defied logic to switch seats in the hope of deceiving Sairi, when Sairi had already obtained Khua’s particulars and recognised his face. If Khua had planned to switch seats, he would logically have done so surreptitiously at the bus bay before the road block, where he had switched off his headlights. This was all the more likely as both Khua and Pang did not see Arman at the bus bay and would have perceived this moment as an opportune time to switch seats. Furthermore, it was ludicrous for Khua to ask Pang instead of Shang (who was perfectly sober) to replace him, when Pang was more drunk than him. Above all, it was futile for Pang to proceed to drive merely two to three metres after Sairi had already stopped the car and seen Khua at the driver’s seat.

18        Sairi’s testimony was also not consistent with the fact that Khua was alert enough to react almost instantaneously after noticing the road block, and to drive into the bus bay. Sairi himself conceded that Khua must have been quite alert and also admitted that Khua was driving slowly when turning into the bus bay. It was then incredible that, after being alert and driving at a safe speed, Khua proceeded to drive dangerously while fully conscious that he was being watched by the police. The objective evidence that Khua was capable of taking such swift action was thus in stark contrast to Sairi’s description of his driving. I had earlier held in Ang Kah Kee v PP [2002] 2 SLR 104 that reasonable doubt could be cast over the prosecution’s case by pointing to inherent flaws or improbabilities in its logic and evidence. The facts before me were inherently illogical and undermined the prosecution’s case. It was dangerous to convict Pang based only on Sairi’s observation of the “zigzag” movement of the car for 200 metres.

19        I might have accepted the above remarkable facts if the trial judge had made the correct inference that there was no material discrepancy in Sairi’s evidence. Khua’s counsel sought to highlight discrepancies in Sairi’s evidence, but some of these were immaterial. One such argument was that Sairi estimated the distance between the bus stop and road block wrongly, but this shortfall was not pertinent to the question of whether Sairi saw the car being driven in a zigzag manner. It was also submitted that Sairi could not have observed Khua swaying when walking from one point to another, when he had not asked Khua to walk in a straight line. It was not inherently improbable for a police officer to observe a person’s unsteady gait without conducting an actual test of walking in a straight line. Nonetheless, from my review of the notes of evidence, I found that Sairi was a hopelessly confused witness. The trial judge would not have convicted the appellants if he had detected the flaws in Sairi’s evidence.

20        First, Sairi seemed unsure as to when he decided to use the two cars as a barrier, or at which stage he saw the appellants driving dangerously. The trial judge found that he had stopped two other cars after Khua’s car moved off from the bus bay. However, Sairi had categorically stated in examination-in-chief that two cars were stopped at the road block earlier as a barrier, before Khua’s car was spotted. Evidently, there was no reason for him to take this precaution before he spotted the appellants’ car swerving into the bus bay. Yet, he maintained that the two cars arrived at the road block when Khua’s car was seen coming round the bend (before the bus bay). When questioned as to why he had to take this precaution, he gave the answer that the car had suddenly turned into the bus bay. When asked if he was changing his evidence as to the sequence, he denied it but did not give a satisfactory explanation.

21        This confusion was accentuated later when he testified that he asked the two cars to stop because he saw that Khua’s car was being driven in a zigzag manner. This contradicted earlier evidence that he only saw this unsteady driving when the car drove from the bus bay to the road block, and that the two cars were there even before the car turned into the bus bay. If this last statement was true, it implied that he saw Khua driving in a zigzag manner even before Khua stopped at the bus bay, which could not be so as he only noticed the car after it emerged from the bend. The appellants’ counsel had argued below that the two cars were present earlier, as part of Sairi’s routine check. While Sairi denied checking the vehicles, it was a likely situation since the alternative was quite remarkable - that the two cars arrived at exactly the same time as the moment Sairi saw Khua’s car turning round the bend, and Sairi decided to stop them as a barrier though Khua had yet to exhibit any suspicious behaviour.

22        Next, I detected similar vacillations in Sairi’s testimony regarding his alerting of the other officers. Initially, he testified that when he saw the car (which was at the bus bay, before it reached the road block), he alerted other officers to observe it. Yet, subsequently, he stated that he only informed Arman to be alert, when the car turned to the bus bay. He did not alert anyone after this. However he altered his evidence by stating that, when he saw Khua’s car being driven in a dangerous manner from the bus bay to the road block, he alerted Arman and the others would have heard it. Upon being confronted with this inconsistency, he reverted to his previous version that he only contacted Arman when the car drove to the bus bay and did not alert anyone after that. The glaring discrepancy was not explained.

23        If Sairi had indeed informed one officer, and the others heard it on the personal radio system, it is most puzzling why no other officer seemed to have taken note of the car. After all, according to Sairi, the car was “posing a danger to the driver, passenger and any other road users”. Arman, who was instructed to watch the movement of the car at the bus bay, strangely decided to ignore what happened after the car left the bus bay. He walked back to his station and did not keep observation of the car after that, despite the explicit instructions. Ten other police officers were on duty and Sairi conceded that there were at least two police vehicles present.

24        The whereabouts of the other ten police officers was also an enigma to me. Sairi, despite being section leader of the patrol team, was slow to supply details as to the number of officers present. Initially, he testified that there were more than five other officers, though he could not remember the exact number. When asked to look at his log book, he said that 11 others were on duty and could have been attending to other traffic violators or updating diaries. During cross-examination, Sairi said that they had been dispatched to attend to other messages. The following exchange then ensued:

Q: What about the other 11 officers. Where were they doing at that time?

A: The other officers may have been dispatched by our Ops Room to attend to other messages. I did not write down or observed [sic] the movements of the other officers.

Q: You mentioned earlier that there were some officers who were checking other cars?

A: I did not say that the other officers were checking other cars.

Q: But there were some officers who were engaged with other violators?

A: Yes.

Q: So there were other cars that had been stopped?

A: At the particular time when Accused’s car was stopped, no. These cars had been stopped before the Accused’s car was stopped.

Q: At the material time, some officers were engaged with other violators. Therefore they were still around the road block?

A: Yes. These violators could be motor-cyclists and cyclists. Not pedestrians.

25        Sairi, after giving a different explanation for the absence of other officers, then denied that there were other cars being stopped by them. Upon realising that this answer was not consistent with his previous statement that they were dealing with other violators or were updating diaries, he had to offer a reason - some officers were dealing with motorcyclists and cyclists. He could no longer state that they were attending to cars, as he had said earlier that only two cars were stopped before the appellants arrived. It was difficult to believe that there were so many motorcyclists and cyclists at 3.00 am, and none of the officers showed any sign of noticing Khua despite the alert. There were at least two officers, Staff Sergeant Safaruan Kasman and Sub-Inspector Ong Chang Leong, present at the scene who administered the breath test for the two appellants and made conditioned statements. Also, out of all these 11 officers, none tended the road block after Sairi left his place to question the appellants. Sairi’s account seemed to create the impression that he was the only police officer at the road block and who witnessed the events, which was plainly untrue as there were at least two others assisting him in tending to the appellants. Why Sairi would want to diminish the involvement of any other police officer was a mystery.

26        In short, Sairi’s entire account lacked persuasiveness due to multiple vacillations. Since his account of how the appellants switched seats was inherently improbable, his discrepancies further undermined the credibility of his testimony which was, unfortunately, the sole evidence relied on for conviction. Therefore, in the light of the tenuous evidence of  Sairi, I could not agree with the trial judge’s inferences that Khua had driven his car in a dangerous manner, Khua had switched seats with Pang and Pang had driven the car for a few metres.

27        I noted that there were other aspects of Sairi’s testimony which cumulatively cast reasonable doubt on the prosecution’s case. First, some weight should have been ascribed to Khua’s breath test results. His result of 32 micrograms per 100 millilitres of breath was below the prescribed level of 35 micrograms. Admittedly, s 67(1)(a) of RTA, unlike s 67(1)(b), does not stipulate the proportion of alcohol which must be found to exceed the prescribed limit. The difference in the drafting of the two sections was probably meant to cater to a situation when, according to the police officer’s discretion, the driver was unfit to control his vehicle even though he did not fail the breath test, or if a breath test was not conducted. Since the finding of an offence under this section rested heavily on the police officer’s discretion, that discretion must be thoroughly examined to ensure reliability. My opinion was that this discretion was not accurately exercised by Sairi. On the contrary, the fact that Khua did not fail the breath test lent credence to his claim that he only consumed two cups of diluted brandy. The extent of alcohol consumed might have come dangerously close to the limit, but the lack of infringement, coupled with Sairi’s weak evidence and the improbability of his account, cast reasonable doubt on the prosecution’s case.

28        Finally, Khua’s counsel submitted that the trial judge was wrong to treat the log book records as contemporaneous records that buttressed Sairi’s testimony. I decided that the log book records could not be given much weight in corroborating Sairi’s evidence or in neutralising the above deficiencies in his testimony.  The “important notes” in the log sheet specifically stated that if a police officer were to use the log book when appearing as witness, it could only be used if the notes were made at or shortly after the event. Sairi testified that he only recorded the facts on the same day after he lodged the arrest report. In the latter, the time stated was 3.15 am, while the time Sairi spotted the appellants was 3.00 am according to his testimony. It was not known how soon after the report Sairi wrote in his log book. In any event, it held little weight as being contemporaneous evidence, a decision to charge a person should be based on contemporaneous written evidence adduced before charging. The basis of the charge was undermined once the evidence which prompted the police to charge was only manufactured after the arrest report.

Whether the trial judge gave due weight to the evidence of the appellants

29        I did not discount the possibility that there might have been reason for the appellants’ friends, Shang and Chang to assist the appellants, but that possibility alone could not form the basis for rejecting their evidence. As I held in Soh Yang Tick v PP [1998] 2 SLR 42, the mere fact that the appellant`s witnesses were in some way related or connected to appellant did not render their testimonies suspect. “There must be additional grounds for rejecting the evidence of such witnesses, or alternatively the testimonies of these witnesses were so littered with inconsistencies that they could not be believed.”

30        If all the appellants and their witnesses were lying, it would have been easy to detect inconsistencies. Yet, having perused the notes of evidence, I found their versions to be consistent and cogent. Unlike the inadequacies found in cross-examination of Sairi, there were no such discrepancies in their testimonies. Both Pang and Shang had earlier made statements to the police, which were similar to their examination-in-chief. The prosecution did not list any discrepancies in their previous statements. Also, all their testimonies corresponded in stating when they observed police officers appearing, and that there was more than one police officer present. All of them noted that there were at least two officers at the road block. Khua and Chang both mentioned that there was one officer each at the right and left of the road block stopping point. Pang also testified that there were at least two officers, while Shang said that there were other police officers at the road block while one talked to Khua.

31        Moreover, if they were colluding in their testimony, instances of embellishment of evidence would abound, but they were absent. Shang, who was perfectly sober that night, could have easily said that he had kept his eyes open throughout the entire journey and thus could have noticed if Khua was driving dangerously. Yet, he admitted that he had closed his eyes to rest, though he did not sleep. Also, instead of asserting that there was a dispute between Khua and Sairi, he said that he did not know if there was a dispute, but only noticed they were talking very loudly and Khua had his breath tested a number of times.  Likewise, Chang conceded that the headlights were switched off at the bus bay, which Khua said was done to avoid detection. This statement might have incriminated Khua as it could reflect a guilty conscience, but they did not refrain from testifying about this detail.

32        The unusual circumstance here that appeared to weaken the appellants’ account was that Khua chose to drive the car despite having drunk two cups of diluted brandy and despite his previous conviction. The trial judge emphasised this, arguing that he had probably panicked because he knew that he was not in proper control of the vehicle. Nevertheless, Khua’s account of being seized with sudden panic due to having a previous conviction was more believable in the light of the consistencies in the appellants’ accounts. On the contrary, the prosecution, which was tasked with the burden of proving their case beyond reasonable doubt, submitted a case that lacked logic and was riddled with inconsistencies. In the light of the appellants’ consistent account, and the weakness in the sole prosecution witness’ account, the convictions of both Pang and Khua were against the weight of the overall evidence.

Whether an adverse inference should have been drawn against the prosecution for its failure to offer other witnesses to the defence

33        This adverse inference can be drawn when the prosecution fails to call certain witnesses, or fails to offer them to the defence: Satli bin Masot v PP [1999] 2 SLR 637. It is founded on s 116 illustration (g) of the Evidence Act:

The court may presume… the existence of any fact which it thinks likely to have happened, regard being had to the common course of natural events, human conduct, and public and private business, in their relation to the facts of the particular case.

Illustration: The court may presume:

(g) that evidence which could be and is not produced would if produced be unfavourable to the person who withholds it; (emphasis added)

34        Clearly this is a discretionary and not a mandatory inference: Chua Keem Long v PP [1996] 1 SLR 510, Chia Sze Chang v PP [2002] 4 SLR 523 and Satli bin Masot. The appellate court has to evaluate if the trial judge had exercised his discretion correctly. Strict criteria have been stipulated before this adverse inference can be drawn against the prosecution, namely that: 

(a) the witness not offered was a material one: Chua Keem Long, Lau Song Seng & Ors v PP [1998] 1 SLR 663 followed by Satli bin Masot;

(b) the prosecution was withholding evidence which it possessed and which was available: Ang Jwee Herng v PP and Amir Hamzah bin Berang Kuty v PP [2003] 1 SLR 617; and

(c) this was done with an ulterior motive to hinder or hamper the defence:  Ang Jwee Herng, Wong Leong Chin v PP [2001] 1 SLR 146 and Chia Sze Chang.

35        The court is generally reluctant to draw this inference against the prosecution. I explained the rationale in Chua Keem Long:

The discretion conferred upon the prosecution cannot be fettered by any obligation to call a particular witness. What the prosecution has to do is to prove its case. It is not obliged to go out of its way to allow the defence any opportunity to test its evidence. It is not obliged to act for the defence.

36        Various guidelines have been formulated to decide what constitutes a  “material witness”, such as whether the absence of the witness(es) would lead to the demolition of the prosecution’s case: Ang Jwee Herng, and whether the witnesses called sufficiently established the prosecution’s case so that any other witnesses were not essential: Chia Sze Chang, R Yoganathan v PP [1999] 4 SLR 264 and Chua Keem Long. Although I found that Sairi’s evidence alone did not establish the prosecution’s case, I decided that the other police officers were not material or essential witnesses. In previous cases when I drew adverse inferences, I found the witnesses material only when the witnesses called did not sufficiently establish the prosecution’s case. In addition, the evidence of the other witnesses who were not called was ascertainable. For instance, in Sahadevan s/o Gundan v PP [2003] 1 SLR 145, there was clearly another person who witnessed the incident concerned. The prosecution had relied solely on the complainant’s evidence, which had serious inconsistencies affecting his credibility. The complainant alleged that a passer-by had witnessed the incident, but the prosecution did not call him to corroborate the complainant’s account. Likewise, in Khoo Kwoon Hain v PP [1995] 2 SLR 767, the victim of a sexual offence had testified about complaining to her aunt, but the prosecution failed to identify this aunt or call her as witness to corroborate the victim’s account. In another case, Lau Song Seng v PP, I held that there was scant evidence of a conspiracy, and the prosecution did not call a witness who undoubtedly had personal knowledge of the alleged conspiracy. In all these cases, there was no doubt that other witnesses had cognisance of the incidents concerned, but the prosecution had failed to call on them to testify.

37        On the facts before me, there was no clear indication that the other officers, though present at the scene, had witnessed the incident. I was inclined to this conclusion since Arman, who was at the bus bay, did not observe how the car was driven. The content of their testimony was a matter of pure speculation, as the court did not know if they were eyewitnesses. Additionally, the other police officers might have been of no assistance to the court as they could have been as confused as their patrol leader, Sairi. There would have been stronger ground to draw an adverse inference if Sairi had alleged that the other police officers witnessed the incident and could corroborate his account. However, since the prosecution chose to rely on only one witness and there was no certainty of there being other eyewitnesses, I found that the other officers were not material witnesses. Consequently, once the testimony of the others was unascertainable, I could not conclude that the prosecution was withholding material evidence which it possessed.

38        The appellants also could not show that the failure to call other witnesses was motivated by an ulterior motive to hamper the defence. This condition is normally difficult to establish with concrete evidence, especially since the prosecution’s discretion in calling witnesses should not be unduly fettered. The court would only infer the presence of ill intent when the witness was essential, the court could ascertain the witness’ evidence and the witness was available to the prosecution at the time of the trial. Only then can interference be warranted due to a miscarriage in justice, as I held in Chua Keem Long.   In Ang Jwee Herng, the prosecution did not have the means of contacting the witness, but clearly set out his role in the case. No intentional concealment could be found. In Chia Sze Chang, the witness called was found credible, and the other witnesses who were not called were no longer available, having departed from Singapore. Before me, the appellants did not convince me that any miscarriage of justice occurred, as the witnesses were not material to begin with.

39        The above principles are well crystallised in this statement, which I made in Lim Young Sien v PP [1994] 2 SLR 257, and reiterated in Chua Keem Long:

In our judgment, the law is well settled that, in a criminal case, the prosecution has a discretion whether or not to call a particular witness, provided that there is no ulterior motive, and the witness, who is available to, but not called by, the prosecution, is offered to the defence.

40        The other police officers were presumably available at the time of trial and the prosecution could have offered them to the defence. However, as in Roy S Selvarajah v PP [1998] 3 SLR 517, the defence knew about these witnesses and could have reminded the prosecution to identify and offer them. The failure by the defence to do so could not, in the absence of incontrovertible evidence of ill intent on the prosecution’s part, be reason for the appellate court to now draw an adverse inference. Though I upheld the trial judge’s decision in this respect, my conclusion above, that the prosecution’s case was not proven beyond a reasonable doubt, was still valid. Drawing an adverse inference would not have made a discernible difference on the facts before me.

Conclusion

41        In my opinion, Khua gave the trial judge room to suspect that he had committed the offence, as he had consumed alcohol (albeit less than the limit) and then decided to drive despite having a previous conviction. I had emphasised the seriousness of drink driving in Sivakumar s/o Rajoo v PP [2002] 2 SLR 73, in which I commented that a motor car in the hands of an inebriated person is a potentially devastating weapon. Notwithstanding that Khua’s imprudence warranted rebuke, the particular inadequacies of the sole prosecution witness rendered the conviction of both appellants highly unsafe. In the event, I allowed both appeals, and ordered the convictions and sentences to be set aside. The fines paid were refunded to the appellants.

Appeals against conviction allowed.