Jagatheesan s/o Krishnasamy v Public Prosecutor
[2006] SGHC 129

Case Number	:	MA 28/2006
Decision Date	:	24 July 2006
Tribunal/Court	:	High Court
Coram	:	V K Rajah J
Counsel Name(s)	:	Subhas Anandan and Sunil Sudheesan (Harry Elias Partnership) for the appellant; April Phang (Deputy Public Prosecutor) for the respondent
Parties	:	Jagatheesan s/o Krishnasamy — Public Prosecutor

24 July 2006

V K Rajah J:

1          The appellant was tried and convicted in a district court on two charges of trafficking in controlled drugs under s 5(1)(a) of the Misuse of Drugs Act (Cap 185, 2001 Rev Ed). He was sentenced to five years’ imprisonment and five strokes of the cane on each charge. The sentences were ordered to run concurrently. This is an appeal against the convictions.

Factual matrix

2          Sometime in late July 2005, Central Narcotics Bureau (“CNB”) officers arrested Rohaizman bin Rahmat (“Rohaizman”) for possession of controlled drugs. Rohaizman upon being queried by CNB officers implicated Gunaprakash s/o Thuraisamy (“Guna”) as his supplier.

3          Later that month, an undercover CNB officer contacted Guna introducing himself as “Sky”. He told Guna that he was interested in purchasing ecstasy tablets (“tablets”). Guna agreed to procure the tablets for Sky.

4          On 4 August 2005, according to Sky, Guna called him just past noon on his mobile phone and confirmed that he had procured a new stock of tablets. Guna insisted on meeting Sky in person to ascertain the genuineness of the proposed transaction. At about 1.40pm, they met up at the void deck of Block 313 Shunfu Road. Guna, satisfied with Sky’s purported credentials, thereupon confirmed that he would deliver the tablets that evening at Newton Hawker Centre. Guna subsequently called Sky and informed him that he could only supply 90 tablets as he had already agreed to sell ten tablets to other customers. Guna also proposed reducing the transaction price from $1,800 to $1,700. Sky agreed to this.

5          At 5.50pm, Sky conducted a briefing for the CNB officers involved in this operation. He would be accompanied only by Sergeant Bukhari (“Ayie”). Upon receiving confirmation from Ayie that the exchange had taken place, the covering party of CNB officers would promptly arrest Guna.

6          The CNB party reached Newton Hawker Centre at about 6.45pm, and promptly commenced surveillance of the proposed meeting area. Some of the CNB officers noticed two male Indians sitting at a table. One of the Indians matched the description of Guna they had earlier been given. The other person was the appellant. These two persons were later observed leaving their seats and moving in the direction of the main road.

7          At about 6.55pm, Sky and Ayie arrived at Newton Hawker Centre and sat at a table. Guna then joined them. Sky introduced Ayie to Guna. Guna then handed a black plastic bag (“P4”) to Ayie. Ayie brought P4 to a nearby toilet to confirm that it contained the tablets. Once inside the toilet, Ayie opened P4. He found a yellow plastic bag and a red “ang pow” packet. Within these articles were yellow pills (which have been confirmed to be ecstasy tablets). Upon locating the yellow pills, Ayie promptly instructed the covering party of CNB officers to arrest Guna. The appellant, who was waiting near the main road, was also arrested as he had earlier been observed in Guna’s company. At that juncture, unlike Guna, he was not suspected of being implicated in the drug transaction.

8          The appellant was promptly searched. No drugs were found in his possession. His residence was also raided and searched. Again, no drugs were found.

9          During his interrogation by the CNB officers, the appellant consistently and steadfastly maintained his innocence and asserted complete lack of knowledge of Guna’s drug trafficking activities. He was adamant that he had met Guna purely to collect an outstanding loan of $100. Guna, on the other hand, unequivocally implicated the appellant as his supplier.

Guna’s testimony

10        Guna claimed that he had known the appellant for just over a year. They met initially while working for Lux Singapore. Thereafter they became friends. In 2004, he borrowed $100 from the appellant which he later repaid. This, he claimed, was the only loan from the appellant. He had, however, borrowed other sums of money from a loan shark. Guna admitted that he was in dire financial straits.

11        Guna had been unable to secure any permanent employment since 2004. Needing money, he had turned to drug trafficking in 2005. He also claimed to have recently paid substantial telephone bills amounting to between $5,000 and $6,000 on behalf of friends. He did not explain how he obtained funds to pay such large bills. He did not know where these friends currently were nor did he explain how such large telephone bills had been incurred.

12        On 4 August 2005, he allegedly received a call from Sky, literally out of the blue, in connection with a purported transaction for the purchase of 100 tablets. He informed Sky that he needed to call his supplier to ascertain the availability of the tablets. He testified, “I know that [the appellant] has ecstasy tablets because he has told me before that he has them. I needed the money. I did not get any work. I had debts to repay”. After purportedly having ascertained that the appellant would supply him with the tablets he arranged to meet Sky at Newton Hawker Centre at 6.45pm. He stated that he had instructed the appellant to arrive at 6.30pm but later clarified that he had actually fixed the appointment for 6.45pm.

13        When the appellant arrived at about 6.45pm, they both promptly proceeded to a nearby phone booth where the appellant handed P4 over to him. On receiving P4, Guna went alone to the toilet, removed ten tablets and returned to a table where he sat down. He also claimed that the appellant during this interval remained alone at the phone booth.

14        Between 6.55pm to 7.00pm, Sky arrived with a friend. Guna handed P4 containing the 90 tablets to Sky’s companion. Soon after that, Guna was apprehended by the CNB officers. Ten tablets were found in his waist pouch.

15        In the course of cross-examination, Guna initially maintained he was employed by Lux Singapore when he first met the appellant. Later he claimed he was merely a trainee who worked for “one to two months”. He conceded that he was not paid for his “work”. Finally, he acknowledged that he merely went for a training course at Lux Singapore, which lasted for about two to three weeks.

16        Guna admitted having supplied tablets to Rohaizman but declared only in the course of cross-examination that the appellant was also the supplier for this transaction.

17        He maintained during cross-examination that after Sky had called him on 4 August 2005, he had contacted the appellant to make arrangements for the purchase of the tablets. He further clarified that when the appellant joined him at Newton Hawker Centre at about 6.45pm, they had sat at a table for about ten minutes before proceeding to a nearby phone booth. After entering the phone booth the appellant handed P4 to him. Guna then returned to the table alone after visiting the toilet. Departing from his evidence in chief, Guna also asserted that he had visited the toilet to check on the contents of P4. The appellant was still seated at the table when he returned from the toilet. He returned P4 to the appellant. Later he asked the appellant to walk in front of him to the phone booth upon learning that Sky had arrived. At the phone booth he took possession of P4 again.

18        Guna acknowledged having been informed by CNB officers that if he immediately identified his supplier his co-operation would be viewed favourably. Guna also conceded that he had declared to a judge in Court 26 during a mention hearing on 5 August 2005 that he had made a false statement implicating the appellant as his supplier. He explained that he had acceded to this because the appellant was “pestering” him. When pressed by the Defence counsel whether “[b]ecause of the pestering you were prepared to lie to the Court?” he answered affirmatively. When asked why he had given in to the appellant his response was, “I was afraid the friendship will split.”

Appellant’s testimony

19        The appellant was at the material time gainfully employed, drawing a basic salary of $1,500. He claimed to be “financially sufficient”. The appellant emphatically asserted that he had proceeded to Newton Hawker Centre for the sole purpose of collecting the sum of $100 from Guna as repayment of a loan he had extended in July 2005. This was the third such loan he had extended to Guna. The two earlier loans had been repaid. They had first met at Lux Singapore and the friendship with Guna was cemented when the latter was briefly employed as a security officer at Bar None, Marriott Hotel. Guna, on account of their friendship, would allow him to enter the bar without any payment of the requisite cover charge. The appellant vehemently denied having supplied any drugs to Guna.

20        The appellant testified that on the morning of 4 August 2005 he had sent Guna a short message service (“SMS”) message requesting him to call back. When Guna responded, he promptly requested repayment of the outstanding loan which had been extended a month earlier. He needed the money, inter alia, for some urgent alteration to a pair of pants. Guna told him that he was cash strapped as he had just returned from a holiday in Thailand. The appellant was upset that while Guna had delayed repaying the loan, he could nevertheless fritter money away by holidaying in Thailand. Guna subsequently agreed to meet him at Newton Hawker Centre at 6.45pm to repay the loan. The appellant testified that he had also arranged to meet his fiancée at 7.00pm that evening at Scotts Road. (This fact was corroborated by his fiancée who testified that she had also been informed by the appellant that he had a prior appointment with Guna to collect a sum of $100.)

21        The appellant arrived at Newton Hawker Centre at about 6.45pm. He saw Guna approaching him from the direction of the toilet area. They both sat at a table and ordered some refreshment. Guna informed him that the friend who would put him in funds was on the way. Shortly after that, Guna received a call and went to the toilet again. When Guna returned, he informed the appellant that his friend would be late. A little later, Guna received a SMS message. He immediately left the table and requested that the appellant wait for him at the roadside near the entrance to the car park. The appellant acceded to this request. The appellant maintained that he never entered the phone booth. He denied walking together with Guna to the main road, and therefore neither paid attention to Guna’s subsequent whereabouts nor noticed Guna carrying P4.

22        At about 7.00pm, the appellant received a phone call from his fiancée who was upset that he had been delayed. The appellant explained to her that he had not yet been repaid. He thereafter called Guna to inform him that his fiancée was getting impatient with him. When Guna replied that he would be with him very shortly, the appellant decided to wait a little longer.

23        He was completely taken aback when several CNB officers suddenly detained him. In response to his query as to why he had been arrested, he was notified that he was suspected of consuming a controlled drug. No drugs were found on him. Nor was there any trace of drugs at his residence or in a urine sample taken from him.

24        The very next day, on 5 August 2005, at the Subordinate Courts lock-up, while awaiting the mention of their matters, the appellant pleaded with Guna to immediately acknowledge having falsely implicated him. Guna agreed to do so and informed the court interpreter that he had decided to retract his earlier statement to the CNB officers implicating the appellant. This was duly communicated to the presiding judge in Court 26.

25        When closely queried by the learned district judge why he had pressed Guna for the return of the loan the appellant explained that his cousin was to be married on 5 August 2005. He had nothing suitable to wear other than an old pair of black pants that required alteration. The alteration fee was $25. (It bears mention that a pair of black pants was found in his haversack.) He also informed the court that he had only $9.90 in coins on his person at the material time. Responding to further queries on his whereabouts after he left the table, the appellant confirmed that they both departed together but that he alone went to the main road. He was not aware whether Guna had been trailing him.

26        It is noteworthy that the Prosecution did not challenge the appellant’s evidence in relation to how he met Guna, his evidence in relation to the earlier loans he had purportedly extended to Guna, the reason why he urgently needed the money, his alleged appointment to meet his fiancée at 7.00pm on the evening of 4 August 2005 and his communication to her relating to his meeting with Guna for the purpose of collecting $100.

The trial court’s decision

27        The learned district judge correctly observed that the entire case for the Prosecution against the appellant rested solely on Guna’s testimony. Guna had by then commenced serving a seven-year sentence for drug trafficking and possession of controlled drugs. The learned district judge also recognised that Guna was an accomplice and that his testimony needed to be treated with caution.

28        Nevertheless, at [32] of his grounds of decision (PP v Jagatheesan s/o Krishnasamy [2006] SGDC 48) (“GD”), the learned district judge assessed that Guna “was forthright in admitting that he had trafficked in the drugs in question”. He considered all inconsistencies in Guna’s evidence such as his employment history as minor. What was crucial, in his view, was that Guna and the appellant had indeed met while working at Lux Singapore and that “thereafter a bond of friendship grew between them” (see [33] of the GD). As for the discrepancy between the appellant’s and Guna’s versions on the other alleged loans, the learned district judge reasoned that this was a “non issue” since the appellant had not disputed that the loans had been repaid.

29        What is significant is that the learned district judge was invited by Defence counsel to consider whether or not the appellant had indeed lent $100 to Guna. This is a crucial fact as this was the sole reason proffered by the appellant for his presence at Newton Hawker Centre that evening. The learned district judge determined that there was no outstanding debt of $100 due to the appellant. His decision was principally anchored to his wholesale and unqualified endorsement of Guna’s testimony and tied to an unequivocal rejection of the appellant’s version of events. In essence, the learned district judge’s main reasons for convicting the appellant were:

(a)        Guna was a forthright and honest witness. He had openly admitted to trafficking in drugs.

(b)        There was no logical reason for Guna to lie about the number of loans he had taken especially when the two earlier alleged loans had been repaid. He had not lied about the loan or the amount he had borrowed from the appellant. The learned district judge observed (at [36] of the GD): “If indeed he intended to frame up the accused about this inconsequential fact, he could have denied taking the loan which he did not do. This only reflects his honesty and truthfulness.”

(c)        The appellant stated he had only $9.90 in coins when he was arrested. This undermined his claim to be financially sufficient and his assertion that his fiancée received a good allowance from her father.

30        The learned district judge further determined that any inconsistencies in Guna’s testimony in relation to the price of the drugs were minor and wholly immaterial to the charge preferred against the appellant.

31        In rationalising why none of the CNB officers saw Guna leaving the table for the toilet after being purportedly handed P4 by the appellant (as alleged by Guna) the learned district judge reasoned at [41] of the GD that:

From the narcotics [officers’] evidence it is apparent that they all must have commenced observing [Guna] and the [appellant] seated at the table only after the [appellant] had handed the drugs to [Guna] and after [Guna] had counted the drugs in the toilet and returned to the table where he handed the drugs in a black plastic bag back to the [appellant].

32        As to why none of the CNB officers saw P4 although Guna maintained that it was clearly visible throughout, the learned district judge concluded at [42] that this was because they were all focused on the exchange of the drugs between Guna and Sky.

33        As for Guna’s explanation that he had retracted his statement incriminating the appellant during a hearing at Court 26 because the appellant had been pestering him, the learned district judge chose to accept such an explanation on the basis that Guna stood to gain absolutely nothing by falsely implicating the appellant for the offence.

The basis for appellate intervention

34        It is necessary at the outset to restate the limited nature of the review afforded to an appellate court. In Yap Giau Beng Terence v PP [1998] 3 SLR 656 (“Terence Yap”) Yong Pung How CJ noted at [24]:

It is trite law that an appellate court should be slow to overturn the trial judge’s findings of fact, especially where they hinge on the trial judge’s assessment of the credibility and veracity of witnesses, unless they can be shown to be plainly wrong or against the weight of the evidence. [emphasis added]

35        Such judicial restraint in relation to overturning or modifying findings of fact is established and entrenched in numerous decisions: see PP v Poh Oh Sim [1990] SLR 1047; Ng Soo Hin v PP [1994] 1 SLR 105; PP v Azman bin Abdullah [1998] 2 SLR 704; Ang Jwee Herng v PP [2001] 2 SLR 474.

36        The observations in these decisions ought to be read in conjunction with the Court of Appeal decision in Bala Murugan a/l Krishnan v PP [2002] 4 SLR 289 (“Bala Murugan”) which states at [21]:

The intervention of the appellate court would be justified only where the findings below were clearly wrong or the balance of evidence was against the conclusion reached by the trial court … [emphasis added]

37        That said, it must be noted that the position apropos the proper inferences to be drawn from findings of fact is quite different. Yong Pung How CJ in Terence Yap observed in this context (at [24]):

[W]hen it comes to inferences of facts to be drawn from the actual findings which have been ascertained, a different approach will be taken. In such cases, it is again trite law that an appellate judge is as competent as any trial judge to draw any necessary inferences of fact from the circumstances of the case.

38        In short, intervention by an appellate court is justified when the inferences drawn by a trial district judge are not supported by the primary or objective evidence on record: see Bala Murugan at [21]; Sahadevan s/o Gundan v PP [2003] 1 SLR 145 at [17]; see also s 261 of the Criminal Procedure Code (Cap 68, 1985 Rev Ed) (“CPC”) that stipulates that the appellate court should not intervene unless “it is shown … that the judgment … was either wrong in law or against the weight of evidence”.

Witness credibility

39        The decision in Farida Begam d/o Mohd Artham v PP [2001] 4 SLR 610 at [9] is instructive. Yong Pung How CJ observed:

A judge can make a finding on the credibility of a witness based on some or all of the following:

(1)        His demeanour.

(2)        The internal consistency (or lack thereof) in the content of his evidence.

(3)        The external consistency (or lack thereof) between the content of his evidence and extrinsic evidence (for example, the evidence of other witnesses, documentary evidence or exhibits).

40        The same restraint governing appellate review in respect of findings of fact applies in relation to a trial judge’s assessment of a witness’s credibility. Indeed, an appellate court should be even more restrained in such circumstances. The trial judge has had the benefit of viewing and observing the witnesses in court: Lim Ah Poh v PP [1992] 1 SLR 713. There is, however, a difference between an assessment of a witness’s credibility where it is based on his demeanour and where it is based on inferences drawn from the internal consistency in the content of the witness’s testimony or the external consistency between the content of the witness’s evidence and the extrinsic evidence. In the latter two situations, the supposed advantage of the trial judge in having studied the witness is not critical because the appellate court has access to the same material as the trial judge. Accordingly, an appellate court is in as good a position as the trial court in such an instance to assess the veracity of the witness’s evidence: see PP v Choo Thiam Hock [1994] 3 SLR 248 at 253, [12]. An apparent lack of appreciation of inconsistencies, contradictions and improbabilities can undermine the basis for any proper finding of credibility: see Kuek Ah Lek v PP [1995] 3 SLR 252 at 266, [60]. The real tests are how consistent the story is within itself, how it stands the test of cross-examination and how it fits in with the rest of the evidence and the circumstances of the case; per Lord Roche in Bhojraj v Sita Ram AIR (1936) PC 60 at 62.

41        I must caution, however, that even when the trial judge’s evaluation of a witness’s credibility is based on his demeanour, this will not invariably immunise the decision from appellate scrutiny. In PP v Victor Rajoo [1995] 3 SLR 417, the Court of Appeal disagreed with the trial judge’s findings of fact which were, in that case, primarily based on his impression of both the accused and another witness as well as the manner in which they gave their evidence. Writing for the court, L P Thean JA held as follows at 431, [49]–[50]:

The learned trial judge’s acceptance of the accused’s evidence was based mainly on his impression of AB and the accused and the manner in which AB and the accused gave evidence. These factors are of course important and play a vital role in the determination of the veracity and credibility of their evidence. However, it is equally important to test their evidence against some objective facts and independent evidence. In PP v Yeo Choon Poh at p 878 Yong Pung How CJ delivering the judgment of this court said:

As was held by Spenser-Wilkinson J in Tara Singh & Ors v PP [1949] MLJ 88 at p 89, the principle is that an impression as to the demeanour of the witness ought not to be adopted by a trial judge without testing it against the whole of his evidence.

It is also helpful to remind ourselves of what Ong Hock Thye FJ said in Ah Mee v PP, at p 223:

To avoid undue emphasis on demeanour, it may be well to remember what was said by Lord Wright, and often quoted, from his judgment in Powell & Anor v Streatham Manor Nursing Home [1935] AC 243, at p 267 of the possibility of judges being deceived by adroit or plausible knaves or by apparent innocence.

[emphasis added]

42        I should also add that, in my view, reliance on the demeanour of witnesses alone will often be insufficient to establish an accused’s guilt beyond reasonable doubt. In this respect, the astute observation of Lord Bridge of Harwich in the Privy Council decision of Attorney-General of Hong Kong v Wong Muk Ping [1987] AC 501 at 510 is apposite:

It is a commonplace judicial experience that a witness who makes a poor impression in the witness box may be found at the end of the day, when his evidence is considered in the light of all the other evidence bearing upon the issue, to have been both truthful and accurate. Conversely, the evidence of a witness who at first seemed impressive and reliable may at the end of the day have to be rejected. Such experience suggests that it is dangerous to assess the credibility of the evidence given by any witness in isolation from other evidence in the case which is capable of throwing light on its reliability[.]

43        The appropriate balance to be struck between the advantages admittedly available to the trial court and the concomitant need for an appellate court to discharge its constitutional duty in ensuring that a conviction is warranted is perhaps best captured by the Canadian Supreme Court in Her Majesty The Queen v RW [1992] 2 SCR 122 at 131–132:

The trial judge has the advantage, denied to the appellate court, of seeing and hearing the evidence of witnesses. However, as a matter of law it remains open to an appellate court to overturn a verdict based on findings of credibility where, after considering all the evidence and having due regard to the advantages afforded to the trial judge, it concludes that the verdict is unreasonable.

A verdict is unreasonable if, in the words of s 261 of the CPC, the trial judge’s decision is against the weight of the evidence or wrong in law. If the Prosecution has not proved its case beyond any reasonable doubt a conviction would be wrong in law.

Convictions based on the evidence of a single witness

44        There is no absolute prohibition or legal impediment in convicting an accused on the evidence of a single witness: see Yeo Eng Siang v PP [2005] 2 SLR 409 at [25] (although in Tan Wei Yi v PP [2005] 3 SLR 471 (“Tan Wei Yi”) at [23] Yong Pung How CJ expressed his reservations in doing so). Indeed, one wholly honest and reliable witness on one side may often prove to be far more significant or compelling and outweigh several witnesses on the other side who may be neither reliable nor independent.

45        The court must nevertheless be mindful of the inherent dangers of such a conviction and subject the evidence at hand to close scrutiny: see Low Lin Lin v PP [2002] 4 SLR 14 at [49]. This is true whether the witness is an accomplice (see Chua Poh Kiat Anthony v PP [1998] 2 SLR 713, or an interested witness (see Kwang Boon Keong Peter v PP [1998] 2 SLR 592). In such situations, a conviction can only be upheld if the testimony is so compelling to the extent that a conviction can be founded entirely and exclusively on it. This means no more than that the witness’s testimony evidence is so compelling that the Prosecution’s case is proved beyond reasonable doubt, solely on the basis of that witness’s testimony: Teo Keng Pong v PP [1996] 3 SLR 329 (“Teo Keng Pong”) at 340, [73].

Reasonable doubt

46        The requirement that the Prosecution has to prove its case against an accused beyond reasonable doubt is firmly embedded and entrenched in the Evidence Act (Cap 97, 1997 Rev Ed) (“EA”) as well as in the conscience of the common law. In fact, this hallowed principle is so honoured as a principle of fundamental justice that it has been accorded constitutional status in the United States (In re Winship, 397 US 358 (1970) (“Winship”) and in Canada (R v Vaillancourt [1987] 2 SCR 636). It is a doctrine that the courts in Singapore have consistently emphasised and upheld as a necessary and desirable prerequisite for any legitimate and sustainable conviction: see, for example, Teo Keng Pong ([45] supra) at 339, [68]; most recently applied by the Court of Appeal in Took Leng How v PP [2006] 2 SLR 70 (“Took Leng How”).

47        While the raison d’être for this burden of proof is never questioned, much controversy continues to cloud attempts to devise a working definition. A myriad of definitions have been postulated by learned commentators and jurists, each with its own inherent advantages and disadvantages. The magnitude of the debate as to what constitutes reasonable doubt is perhaps best highlighted by the 12 volumes of authorities, replete with over a hundred decisions together with numerous and copious articles and writings, which were submitted to the British Columbia Court of Appeal on the issue: see R v Brydon, (1995) 2 BCLR (3d) 243 (“Brydon”). Many courts have abandoned any further attempt to define what constitutes reasonable doubt, calling it an “impossible” task: R v Yap Chuan Ching (1976) 63 Cr App R 7 (“Ching”) at 11. Many authors have similarly despaired over the futility of such an exercise: see Wigmore on Evidence vol IX (Little, Brown & Company (Canada) Ltd, 1981) at paras 414–415; Note, “Reasonable Doubt: An Argument Against Definition” 108 Harv L Rev 1955 (1995). Perhaps, as is often said, while one cannot precisely define an elephant to adequately convey its peculiarities, it is easily recognised when seen; it is “self-evident” or “self-defining”.

48        Nevertheless, it is in the fundamental interests of the criminal justice system that this issue be sufficiently clarified, (with as much precision as is permitted) and that the nature of the obligation incumbent on the Prosecution in having to prove its case beyond reasonable doubt be adequately explained and understood. Every conviction must hew to an identical touchstone. Such a standard is not so stringent as to mean that every item of evidence adduced should be isolated, considered separately and rejected unless the Prosecution satisfies the trial judge that it is credible beyond reasonable doubt: See Nadasan Chandra Secharan v PP[1997] 1  SLR 723 at [85]. All the principle requires is that upon a consideration of all the evidence presented by the Prosecution and/or the Defence, the evidence must be sufficient to establish each and every element of the offence for which the accused is charged beyond reasonable doubt.

49        Denning J’s dicta in Miller v Minister of Pensions [1947] 2 All ER 372 (“Miller”), at 373, has since its pronouncement been accorded a venerated position and has indeed been very recently cited as correctly encapsulating the law on reasonable doubt; see Took Leng How at [28]. Denning J had declared:

That degree is well settled. It need not reach certainty, but it must carry a high degree of probability. Proof beyond a reasonable doubt does not mean proof beyond the shadow of a doubt. The law would fail to protect the community if it admitted fanciful possibilities to deflect the course of justice. If the evidence is so strong against a man as to leave only a remote possibility in his favour which can be dismissed with the sentence “of course it is possible but not in the least probable,” the case is proved beyond reasonable doubt, but nothing short of that will suffice.

50        The Prosecution’s burden of proof has also been stated in similar terms by Yong Pung How CJ in Teo Keng Pong at 339, [68]:

It bears repeating that the burden on the prosecution is to prove its case beyond reasonable doubt. It is not to prove the case beyond all doubts. That standard is impossible to achieve in the vast majority of cases. In almost all cases, there will remain that minutiae of doubt. Witnesses, apparently independent, could have conspired to ‘frame’ an accused. Alternatively, an accused could be the victim of some strange, but unfortunate, set of coincidences. The question in all cases is whether such doubts are real or reasonable, or whether they are merely fanciful. It is only when the doubts belong to the former category that the prosecution had not discharged its burden, and the accused is entitled to an acquittal.

51        Miller and Teo Keng Pong articulate in measured terms two important concepts intrinsic to the principle of reasonable doubt. The first concept is that it would be wrong to set up a standard of absolute certainty that the Prosecution must meet before an accused can be found guilty. All that is necessary to sustain a conviction is that the evidence establishes guilt beyond a reasonable doubt. Or, to put it another way, if the evidence presents reasonable doubts, the Prosecution has not discharged its burden of proof and the accused must be acquitted. The second concept is that not all doubts about the Prosecution’s case are reasonable doubts. One must distinguish between a “real or reasonable” doubt and a “merely fanciful” doubt. The key question in every case however remains: how should the distinction between a real and a merely fanciful doubt be drawn?

52        Among the many instructive authorities on this issue, the admirably lucid and incisive judgment of Wood JA in Brydon ([47] supra) deserves special mention. Despite Wood JA’s dissent in relation to the result of the case, his astute and thorough analysis of the various formulations of reasonable doubt won him the express support of at least four of the five judges hearing that case. Wood JA concluded in Brydon, that one should not fall back on quantitative descriptions that tend to be both circular and meaningless: at [82] and [83]. In other words, to characterise a reasonable doubt as a “substantial” or “strong” or “serious” doubt is merely to substitute one obscure word for another. None of these adjectives reveal anything about the operation of the reasonable doubt principle at the conclusion of a trial. Indeed, defining reasonable doubt in such terms has, as Wood JA rightly pointed out, only succeeded in meeting with universal criticism: at [84].

53        Instead, Wood JA, at [44], advocated a qualitative definition of reasonable doubt that he found both apt and meaningful:

[I]t is difficult to think of a more accurate statement than that which defines reasonable doubt as a doubt for which one can give a reason, so long as the reason given is logically connected to the evidence.

54        This dictum provides in my view a useful anchor for a working definition of reasonable doubt for two reasons. First, it is a definition the essence of which has already been endorsed by the Court of Appeal in Took Leng How at [29]:

We would also allude to Prof Tan Yock Lin’s work, Criminal Procedure vol 2 (LexisNexis, 2005) at ch XVII para 2952, where he makes the following comments with reference to reasonable doubt in general and the decision of the High Court in Chua Siew Lin v PP [2004] 4 SLR 497 in particular:

It needs no elaboration to state that reasonable doubt is a doubt which is material, which counts. Not any mere possibility of the prosecution case being false will amount to a reasonable doubt in the prosecution case.

He further adds in the accompanying footnote:

A mere doubt, as opposed to a reasonable doubt, must frequently be conceded in the nature of things but because it cannot yet concretely be articulated in relation to the evidence in the case, it remains an untested hypothesis and may be rejected.

[emphasis in original]

55        As such, the Court of Appeal has accepted that a reasonable doubt is one which is capable of distinct expression and articulation and has support and foundation in the evidence submitted which in the circumstances is essential to a conviction. As Prof Larry Laudan puts it, “What distinguishes a rational doubt from an irrational one is that the former reacts to a weakness in the case offered by the prosecution, while the latter does not.”: see, Larry Laudan, “Is Reasonable Doubt Reasonable?” (2003) 9 Legal Theory 295 (“Larry Laudan”) at 320. Reasonable doubt is, in other words, a reasoned doubt.

56        The second reason why I am partial to this particular formulation of reasonable doubt is that it correctly shifts the focus from what could potentially be a purely subjective call on the part of the trial judge to a more objective one of requiring the trial judge to “[reason] through the evidence”: Larry Laudan at 319. Therefore, it is not sufficient for the trial judge merely to state whether he has been satisfied beyond reasonable doubt. He must be able to say precisely why and how the evidence supports the Prosecution’s theory of the accused’s guilt. This process of reasoning is important not only because it constrains the subjectivity of the trial judge’s fact-finding mission; it is crucial because the trial process should also seek to “persuade the person whose conduct is under scrutiny of the truth and justice of its conclusions”: R A Duff, Trials and Punishment (Cambridge University Press, 1986) at p 116; T R S Allan, Constitutional Justice: A Liberal Theory of the Rule of Law (Oxford University Press, 2001) at p 81.

57        Two main concerns have been expressed relating to the desirability of this definition of reasonable doubt. The first is that it would pose a problem for the inarticulate. While expressing and explicating a reason for conviction or acquittal might pose a problem for some jurors (even so, see Wood JA’s observation in Brydon ([47] supra) at [45]), such a scenario is entirely irrelevant in the context of Singapore where the finders of fact are legally trained judges. In any event, judges have an obligation upon an appeal being filed, to justify their decisions. The second concern is that the formulation might be perceived as having the effect of reversing the burden of proof because the burden of furnishing reasons for not finding guilt is shifted to the accused. This, however, is not a legitimate concern. Requiring a trial judge to furnish the reasons for his decision does not limit or compel him to seek or extract those reasons purely from arguments or testimony from the Defence. This is a point to which I will return shortly. Finally, though I am fully aware that the English Court of Appeal has on two occasions rejected this particular formulation: Ching ([47] supra) and R v Dennis Stafford; R v Michael Luvaglio (1969) 53 Cr App R 1, I remain convinced that, neither case provided any considered analysis as to why such a definition was inherently objectionable in a non-jury setting. It appears to me that their unarticulated concerns relate primarily to the possibility of sowing doubt in the minds of jurors; see Phipson on Evidence (The Common Law Library) (Hodge Malek, gen ed) (Sweet & Maxwell, 16th Ed, 2005) at para 6-51. Indeed the editors of Archbold: Criminal Pleading, Evidence and Practice 2005 (James Richardson & David A Thomas, eds) (Sweet & Maxwell, 2005) at para 4-385, submit (presumably to avoid confusing jurors) that a judge should not volunteer an explanation of the expression “reasonable doubt”.

58        In deciding whether the evidence supports a conviction beyond reasonable doubt, it is not only necessary to clarify the conceptual dividing line between reasonable doubts and mere or fanciful doubts. It is also vital to appreciate that the principle that the Prosecution bears the burden of proving its case beyond reasonable doubt embodies two important societal values.

59        First, it “provides concrete substance for the presumption of innocence”: Winship at 363. It is axiomatic that the presumption of innocence is a central and fundamental moral assumption in criminal law. It cannot be assumed that an individual is guilty by mere dint of the fact that he has been accused of an offence, unless and until the Prosecution adduces sufficient evidence to displace this presumption of innocence. That threshold below which society will not condone a conviction or allow for the presumption of innocence to be displaced is the line between reasonable doubt and mere doubt. Adherence to this presumption also means that the trial judge should not supplement gaps in the Prosecution’s case. If indeed gaps in the evidence should prevail so that the trial judge feels it is necessary to fill them to satisfy himself that the Prosecution’s burden of proof has been met, then the accused simply cannot be found legally guilty. In short, the presumption of innocence has not been displaced.

60        Second, the principle of reasonable doubt connotes and conveys the gravity and weightiness that society equates with punishment. It would be wrong to visit the indignity and pain of punishment upon a person (and his family) unless and until the Prosecution is able to dispel all reasonable doubts that the evidence (or lack thereof) may throw up. Therefore, it is critical that trial judges appreciate that inasmuch as fanciful conspiracy theories, often pleaded by the Defence, will not suffice to establish reasonable doubt, the Prosecution’s theory of guilt must be supportable by reference to the evidence alone and not mere conjecture that seeks to explain away gaps in the evidence. Suspicion and conjecture can never replace proof.

61        To summarise, the Prosecution bears the burden of proving its case beyond reasonable doubt. While this does not mean that the Prosecution has to dispel all conceivable doubts, the doctrine mandates that, at the very least, those doubts for which there is a reason that is, in turn, relatable to and supported by the evidence presented, must be excluded. Reasonable doubt might also arise by virtue of the lack of evidence submitted, when such evidence is necessary to support the Prosecution’s theory of guilt. Such a definition of reasonable doubt requires the trial judge to apply his mind to the evidence; to carefully sift and reason through the evidence to ensure and affirm that his finding of guilt or innocence is grounded entirely in logic and fact. A trial judge must also bear in mind that the starting point of the analysis is not neutral. An accused is presumed innocent and this presumption is not displaced until the Prosecution has discharged its burden of proof. Therefore, if the evidence throws up a reasonable doubt, it is not so much that the accused should be given the benefit of the doubt as much as the Prosecution’s case simply not being proved. In the final analysis, the doctrine of reasonable doubt is neither abstract nor theoretical. It has real, practical and profound implications in sifting the innocent from the guilty; in deciding who should suffer punishment and who should not. The doctrine is a bedrock principle of the criminal justice system in Singapore because while it protects and preserves the interests and rights of the accused, it also serves public interest by engendering confidence that our criminal justice system punishes only those who are guilty.

Evaluation of the evidence

Guna’s employment

62        Guna was cross-examined at length on his employment record. When initially asked how long he had worked at Lux Singapore, Guna replied that he worked there for nearly a year. Later he changed the period to “about one to two months”. When asked to explain the discrepancy, he said that his English was poor. The Defence then pointed out that all the questions had been translated into Tamil. Guna countered that he had misunderstood the questions.

63        When the Defence produced an e-mail from Lux indicating that there was no record of Guna’s employment at the company, Guna again altered his testimony, feebly attempting to explain that he had only attended a training course at Lux Singapore. Even then, he inexplicably modified, on no less than three separate occasions, the purported training period from three weeks to two, and finally, to a single week.

The price of the drugs

64        Guna initially testified that he had asked for $1,500 for 100 tablets. Later he changed his testimony to $1,700 for 90 tablets. According to Guna, Sky had said he would pay $1,500 for the tablets and $200 as commission for the deal.

65        However, according to Sky, it was Guna who, after the meeting at Shunfu Road, told him that the price would be $1,800. Later that afternoon, Guna had called again to tell him that he could only sell 90 tablets for $1,700. When confronted with the discrepancy between his testimony on the one hand and Sky’s on the other, Guna’s response was that he could not remember the price.

The loans between the appellant and Guna

66        The learned district judge concluded that the number of loans was a “non issue” because the appellant had acknowledged that the loans had been repaid. I respectfully beg to differ. The number of loans is relevant in assessing the credibility of the two protagonists. The learned district judge chose to disbelieve the appellant’s evidence that Guna borrowed and repaid a total of $200 on two separate earlier occasions but accepted with alacrity Guna’s testimony that he had borrowed from the appellant in 2004 but not in July 2005. The learned district judge reasoned in his GD ([28] supra) at [36] that, “If indeed [Guna] intended to frame up the appellant about this inconsequential fact, he could have denied taking the loan which he did not do. This only reflects his honesty and truthfulness.”

67        With respect, it is difficult to comprehend why the same reasoning cannot also apply with equal force to the appellant – if the appellant had intended to undermine Guna’s credibility, he could have denied repayment of any of the loans. By parity of reasoning, should not the fact that he did not be similarly viewed as reflecting the appellant’s “honesty and truthfulness”? Sauce for the goose is sauce for the gander!

68        On this particular issue, it is pertinent to note that no extrinsic evidence or objective facts have emerged to establish whether, when and how the loans (as the appellant contends) or loan (as Guna contends) were actually made to Guna. This is a case of directly conflicting uncorroborated evidence.

69        The appellant’s testimony was rejected essentially because the learned district judge preferred to rely on Guna’s testimony. (This will be revisited when Guna’s credibility is assessed; see [78]–[81] below). In addition, the learned district judge was not persuaded when the appellant asserted that he needed the money to pay for the tailoring of trousers, given that he had earlier claimed to be “financially sufficient”. The learned district judge reasoned that financial sufficiency would preclude the need to recover $100. With respect, this is an entirely untenable adverse inference to make against the appellant. Self-sufficiency can neither be equated with nor imply substantial financial resources. First, the appellant only claimed that he was financially sufficient in order to refute any suggestion that his financial circumstances had propelled him to traffic in drugs. His testimony on this issue was not challenged by the Prosecution. Second, it is undisputed that when the appellant was apprehended, he had a mere $9.90 in his possession together with the black pants that required alteration. The presence of the pants and the meagre amount on his person both lend some support to the appellant’s testimony as to why he sought urgent repayment of the loan. Third, the appellant’s fiancée, Kelly-Jo Coney, testified that she received a phone call from the appellant at around 5.25pm that evening informing her that he was going to meet Guna for the recovery of a $100 loan. The Prosecution did not dispute this aspect of her testimony. Evidence therefore prevails to support the appellant’s testimony as to why he had pressed Guna for the return of the loan. I am persuaded that the appellant’s version of events should not have been so cursorily dismissed without a proper evaluation of all these considerations.

Events at the Newton Hawker Centre

70        At the Newton Hawker Centre, most of the ten CNB officers only saw the appellant and Guna sitting at a table before they rose from the table.

71        The learned district judge determined on that basis that the CNB officers only commenced their observations after Guna and the appellant had sat at the table and exchanged P4. There is, however, evidence to the contrary which the learned district judge failed to consider (possibly because counsel failed to advert to it). According to Staff Sergeant Puay Bak Yong, one of the CNB officers conducting surveillance:

At about 6.45pm, the covering officers were deployed accordingly at Newton Hawkers [sic] Centre. I positioned myself at the vicinity of the public toilet beside Newton Hawker Centre where I spotted a male Indian subject fitting the description of “Guna”. I then informed Insp William Tan that “Guna” was wearing a dark blue shirt. “Guna” then sat down on [sic] a table together with another male Indian. [emphasis added]

This testimony severely undermines a central premise of the learned district judge’s hypothesis. It indicates that Guna was seen alone in the vicinity of the toilet. Could Guna have met someone else at the public toilet for the receipt of P4? The learned district judge failed to consider that if Guna was to be believed then the timing of his material sequence of events commenced almost precisely when the CNB officers actually began their surveillance. Why did the learned district judge determine that the doubt arising from what the CNB officers had not observed could and should be explained only in a manner that was consistent with Guna’s evidence?

72        The learned district judge remarked that the CNB officers had not observed the exchange of P4 between Guna and the appellant as they were more concerned with the impending transaction between Guna and Sky. P4 is a black plastic bag the length of which is comparable to that of an average person’s arm. I had asked the Prosecution to make available P4 for my viewing during the appeal. I noted that even when folded in half it remained clearly visible. Guna himself testified in response to cross-examination:

Q:         If the accused was carrying the plastic bag when he joined you at the table, anybody observing you and the accused would have seen the accused carrying this black plastic bag?

A:         Yes.

Q:         When you both were walking, anybody observing you both would have seen the accused carrying this plastic bag?

A:         Yes.

Q:         When you both entered the phone booth and when you came out carrying the black plastic bag, anyone observing you would know that the accused gave the black plastic bag to you?

A:         Yes.

Q:         Would it surprise you if people who were observing you, none of them saw the accused carrying the black plastic bag?

A:         Yes.

Q:         It could be that the accused was not carrying the black plastic bag?

A:         But he had a plastic bag in his hand.

Q:         [It’s] not a small plastic bag?

A:         It’s a carrier plastic bag.

Q:         [It’s] easy to see someone carrying the black plastic bag?

A:         Yes.

Q:         All those who did not see the accused carrying the bag must either be blind or the accused was not carrying the black plastic bag?

A:         No, he was having the black plastic bag in his hand. How could they have not seen.

[emphasis added]

A re-examination of Guna’s credibility

73        As shown, Guna’s testimony was riddled with inconsistencies and discrepancies on several issues. His purported period of employment with Lux shrank inexplicably when he was closely queried and pressed to reconcile the discrepancies in his evidence from a year to a month to a week. Guna’s evidence morphed conspicuously when he was confronted with an e-mail confirming that he had not worked at Lux Singapore after all. He also claimed that Sky had contacted him on 4 August 2005 in relation to the subject transaction. He had purportedly informed Sky that he had to check with his supplier on the availability of the tablets. Sky, on the other, hand unequivocally testified that it was Guna who had contacted him on 4 August 2005 claiming that he had received his stock. This is a material point. There is no reason to doubt Sky’s unchallenged evidence on this point. Why then did Guna attempt to create this evidential facade of having to contact his supplier (see [12] above) when he already had the stock by then? Guna’s evidence on the agreed price of the drugs was also far from satisfactory. He struck me, on the basis of the record, as a dodgy witness prepared to mould and modify his evidence so as to address and surmount pressing and difficult queries.

74        The learned district judge dismissed these inconsistencies as inconsequential and trivial. If indeed trivial, why should Guna deign to be confused or worse to lie about so many relevant issues? More importantly, if he could be mistaken or lie about inconsequential facts, what about crucial ones?

75        Guna maintained adamantly that P4 switched hands once at the table and twice at the phone booth. There is no extrinsic evidence corroborating such an assertion. That no one saw P4 at any time raises serious doubts about the veracity of Guna’s testimony on this point. Could all these events have taken place well before the CNB officers began their surveillance? Or are Guna’s assertions inculpating the appellant purely fabricated because he feared revealing his real supplier? In any event, why didn’t the CNB officers observe at least one of the two exchanges at the telephone booth? According to Guna both he and the appellant had made a second exchange at the phone booth just before the appellant proceeded to the main road. It cannot be disputed at that juncture that the CNB officers had already commenced their surveillance.

76        It must be emphasised that not one of the CNB officers witnessed Guna or the appellant enter a phone booth to make the exchange. This creates a reasonable doubt as to whether the second exchange at the phone booth took place at all. A further train of inquiry is raised by the alleged exchange at the phone booth – why return P4 to the appellant at the table only to walk again all the way back to the phone booth once again to repeat the exchange? Guna’s evidence on how the exchanges took place appears rather curious and defies any logical explanation, to say the least. The appellant’s evidence on the other hand was straightforward and remained unshaken albeit by a rather cursory cross-examination which did not challenge his testimony on all material aspects. Having made the observations about Guna’s curious testimony I cannot ignore the possibility that fact may on occasion prove to be stranger than fiction. But is that the case here?

77        It bears emphasis that the conviction of the appellant was founded not simply on the uncorroborated evidence of an accomplice. It was grounded entirely on the testimony of an accomplice who chose to implicate the appellant as his supplier immediately after his apprehension only to retract the accusation the very next day in open court before choosing again to reinstate the accusation during the trial proceedings.

78        Guna’s explanation as to why he withdrew his accusation against the appellant the following day is far from satisfactory. The learned district judge should have refrained from accepting such a bare explanation at face value. The learned district judge placed far too much emphasis on the fact that Guna had acknowledged his guilt at the outset. This, as far as the learned district judge was concerned, signified that he was an honest and forthright witness. This is far too simplistic an interpretation and such conclusion, without more, was in my view wholly unwarranted. Guna was apprehended in an undercover sting operation by CNB officers. He had no conceivable defence or plausible explanation as to why a large amount of tablets was found in his possession. As Guna was caught red-handed, he ought not in the instant case to have scored any points on the credibility front merely for his acknowledgment of guilt. His guilt was incontrovertible and his plea of guilt was to that extent inevitable. It should not have been used as a legal yardstick to assess his credibility. Can the fact that he had already commenced his term of imprisonment and therefore had nothing further to gain or lose genuinely serve to support his credibility? The learned district judge appeared to think so. I do not agree. It seems that the learned district judge further failed to appreciate that Guna also ran the risk of a prospective indictment for providing false information to the CNB if he subsequently chose to resile from his initial statement implicating the appellant. This statement, it also ought to be appreciated, had been made soon after he was informed by the CNB officers that his co-operation in naming his supplier would be viewed favourably; see [18].

79        While an appellate court is bereft of the trial court’s advantage of assessing the demeanour of a witness, it can certainly consider the internal and external consistency as well as the overall intrinsic credibility of a witness’s testimony in deciding whether or not the testimony of a witness ought to have been accepted. The Court of Appeal incisively stated in Bala Murugan ([36] supra) at [25] that “there is no requirement in law that a witness’s credit has to be impeached for his evidence to be disbelieved”. In addition, one must bear in mind that it was solely Guna’s testimony that engendered the Prosecution and ultimately the conviction of the appellant. Such testimony should by no means have been employed to found a conviction unless it was unusually compelling. Guna’s evidence should have been subjected to close and relentless scrutiny given his earlier public retraction of his initial accusation against the appellant. The learned district judge ought to have rigorously probed, assessed and analysed the reason for the retraction in open court. Was the explanation furnished by Guna for this extraordinary retraction in open court plausible? Why had not Guna identified and implicated the appellant during his direct testimony as his supplier for the earlier transaction with Rohaizman? Why did he merely assert in his direct evidence that he knew that the appellant had tablets for sale because he had told him so before (see [12] above) and not because the appellant had supplied tablets to him earlier? Guna was a self-confessed drug trafficker with very substantial telephone bills. He never asserted that the appellant was his only “source” for drugs. It seems rather implausible that these were Guna’s only transactions. All these queries and issues were neither scrutinised nor explored by the learned district judge.

80        In Tan Wei Yi ([44] supra), the appellant was charged with voluntarily causing grievous hurt to the victim. The District Court convicted him on the sole testimony of the victim. On appeal, Yong Pung How CJ set aside the conviction. He observed:

24         In the present case, the district judge never made a finding as to how compelling the victim’s testimony in relation to the appellant was. …

[S]ince it was obvious that the victim’s testimony regarding the appellant was uncorroborated, the district judge should have then applied his mind to consider if the victim’s testimony was so compelling that the appellant’s conviction could be based solely on it.

25         The district judge’s failure to do so rings alarm bells as to whether he had actually exercised the appropriate level of caution when relying solely on the victim’s testimony to convict the appellant. Indeed, there was in this case a very real possibility that the district judge convicted the appellant on the basis of the victim’s testimony without even realising that he had to find that the victim’s testimony was of such a compelling nature as to warrant the conviction. Whatever the possibilities, the fact remains that the law required the district judge to make this finding, and his not doing so was an error of law that could not be rectified.

…

33         In the circumstances, I found that the victim’s testimony was riddled with assumptions and inconsistencies, and was hardly of such a compelling nature that the appellant’s conviction could be based solely on it. In fact, even if the district judge had made the specific finding that the victim’s testimony in relation to the appellant was of a very compelling nature, from my close scrutiny of the notes of evidence, I would have found otherwise and overturned the district judge’s finding as clearly incapable of being supported on the objective evidence.

[emphasis added]

81        These observations of Yong Pung How CJ apply with considerable cogency to the instant facts.

Conclusion

82        It is trite law that minor discrepancies in a witness’s testimony should not be held against the witness in assessing his credibility. This is because human fallibility in observation, retention and recollection is both common and understandable: Chean Siong Guat v PP [1969] 2 MLJ 63 (“Chean Siong Guat”) at 63–64; Ng Kwee Leong v PP [1998] 3 SLR 942 at [17]. Inconsistencies in a witness’s statement may also be the result of different interpretations of the same event: Chean Siong Guat. In fact, a witness may even lie but need not be completely distrusted if he lies not out of guilt but because of a misguided desire to bolster his case, or in other cases, to prevent shameful information from being revealed: PP v Yeo Choon Poh [1994] 2 SLR 867 (“Yeo Choon Poh”). In such circumstances, the court is not obliged, as a matter of course, to dismiss the credibility of the witness and reject his entire testimony out of hand. Confronted with such a witness, the court should, naturally, be more circumspect than ever when scrutinising the rest of his testimony with care. But a court is perfectly entitled, notwithstanding minor inconsistencies, to hold that a particular witness is in fact a witness of truth and to accept the other aspects of his testimony which are untainted by discrepancies.

83        This by no means signifies that a judge can or should ignore any and all discrepancies. Where the inconsistency affects a material part of the witness’s testimony, it may well be safer not to rely on that witness’s evidence. Such an approach is aptly summed up in the words of Hallam, “to pull a stone out of an arch: the whole fabric must fall to the ground”: see Nandia v Emperor AIR 1940 Lahore 457 at 459. It was held in PP v Yeow Beng Chye [2003] SGHC 74 at [27] that while a “one-off inconsistency due to memory lapses” is generally not fatal to the acceptance of the witness’s evidence, a “systematic and widespread pattern of many inconsistencies coming together” can (and, in my view, often will) destroy the credibility of that witness. Where a witness is caught lying deliberately in relation to a material issue, and where the motive for the lie is a realisation of guilt, and if objective evidence further reflects that his testimony is a lie, that lie may very well be used by the Prosecution as corroboration of the witness’s guilt: Yeo Choon Poh at 876, [33], citing with approval Regina v Lucas (Ruth) [1981] QB 720 at 724; PP v Chee Cheong Hin Constance [2006] 2 SLR 24 at [92].

84        The retraction of his own statement by a witness may or may not be treated with circumspection by the court depending on the circumstances. For instance, it is settled law that an accused can be convicted solely upon his own confession even though that statement is subsequently retracted: Lim Thian Lai v PP [2006] 1 SLR 319 (“Lim Thian Lai”) at [43]. It has also been held by the Court of Appeal that a retracted confession of a co-accused implicating the accused in the offence may also be relied upon to establish the accused’s guilt: Panya Martmontree v PP [1995] 3  SLR 341 (“Panya Martmontree”) at 354, [50]. By parity of reasoning, the fact that a witness (in this case, an accomplice) may have retracted his statement inculpating the accused does not, ipso facto, render the statement of little evidential weight.

85        However, both Lim Thian Lai and Panya Martmontree have cautioned that the evidential weight to be assigned to the retracted statement should be assiduously and scrupulously assessed by the courts. In particular, I would add, if the retracted statement forms the only evidence upon which the Prosecution’s case rests, such statements should attract painstaking if not relentless scrutiny. Therefore, in Lim Thian Lai at [43], it was held that it was necessary for the court to be satisfied that the retracted confession is voluntary, true and reliable. In fact, the court in Lim Thian Lai cited Taw Cheng Kong v PP [1998] 1 SLR 943 as an example of where it was correct for the court to have accorded precious little weight to the accused’s statements because of how he had changed his story repeatedly.

86        I pause here only to emphasise that the requirements of the proviso to the general principle that a retracted statement may still be relied upon as being true, viz, that the statement should be voluntary and objectively reliable should be required conjunctively. Therefore, it is not sufficient for the Prosecution merely to prove beyond reasonable doubt that the statement was made voluntarily. A statement by a witness (or even an accused) even if it was given voluntarily may or may not be reliable depending on the circumstances of the case and the cogency of the statement itself and may to that extent, be dubious.

87        In my view, it is neither productive nor meaningful to treat retracted statements as a separate class of evidence attracting its own peculiar rules of analysis. Rather, I prefer to regard retracted statements as an instance of inconsistency in the witness’s testimony. In other words, the fact that a witness admits to a statement and later withdraws it constitutes, both, in principle and in effect, a discrepancy or inconsistency in his evidence. Accordingly, the weight to be assigned to such statements and the assessment of the witness’s credibility falls to be determined by the general corpus of case law relating to inconsistencies, discrepancies and falsehoods in a witness’s statement. In other words, whether the fact that a witness has retracted his statement should be allowed to cast about the credibility of that witness and the veracity of his statement depends on whether a reasonable and reliable explanation can be furnished for the retraction; see, in this regard, the Court of Appeal decision in Syed Abdul Mutalip bin Syed Sidek v PP [2002] 2  SLR 405 at [22] where it was held, in the context of an accused retracting his confession, that “While the court should consider any explanation that the accused person gives for his change of position, the explanation can be rejected if it is found to be untrue.” I would respectfully add that if the explanation for the retraction is unsatisfactory then this may cast doubt on the entire evidence of that witness.

88        Taking into account the glaring incongruities in Guna’s testimony, his vacillating responses and the complete lack of extrinsic evidence supporting his testimony on the crucial issues I can only conclude that Guna’s evidence is not only unconvincing but also devoid of the unusually compelling quality necessary to found a conviction.

89        I am also puzzled as to why the Prosecution omitted to adduce objective evidence that could and would unassailably have secured either a conviction against or acquittal for the appellant. First, the actual content of the text messages that were exchanged between Guna, Sky and the appellant were never adduced. These text messages could effectively and conclusively support or refute the assertions by the appellant as to why he was at Newton Hawker Centre. Nor were the appellant’s telephone records produced even though Guna’s own abbreviated records were tendered for one handphone line. The complete records for all the relevant telephone lines could have composed a comprehensive picture of their contact and relationship. Telephone contacts made in July 2005 (when Guna’s transaction with Rohaizman was effected) would have also assisted in illuminating the dealings (if any) between them. Secondly, neither P4 nor its contents were at any point apparently dusted for fingerprints (nor, if they were dusted, the outcome report adduced). The appellant would have found it impossible to protest his innocence if his fingerprints had been detected on either P4 or any of the other articles in which the tablets had been concealed. On the other hand, the absence of the appellant’s fingerprints on P4 may have supported and sealed his innocence.

90        It is most unfortunate that the learned district judge decided to rely solely on the testimony of a convicted accomplice whose unconvincing testimony could not be corroborated by any objective and/or extrinsic facts. A catalogue of doubts and a pattern of inexplicable incongruity cloud and undermine Guna’s evidence. Unresolved evidential ambiguities have been papered over. Regrettably and rather inexplicably the learned district judge chose to resolve every existing doubt in the Prosecution’s case in its favour. This is decidedly not the correct approach. As explained above, it is a cornerstone of our criminal jurisprudence that in a situation where the evidence may lead to two equally reasonable results, that suffices to establish reasonable doubt. The Prosecution had not in fact discharged the burden of proving the appellant’s guilt beyond any reasonable doubt. A trial judge is not at liberty to bridge gaps in the Prosecution’s case by resorting to and relying on unverifiable inferences and suppositions.

91        In the result, I had no alternative but to conclude that the Prosecution had not discharged its burden of proving the charges beyond reasonable doubt. I set aside the convictions and acquitted the appellant of the two charges he had earlier been convicted of.