1 This Registrar’s Appeal (State Courts) No 9 of 2016 (“HC/RAS 9/2016”) raises a novel point of general importance in the law of limitation that has yet to be conclusively dealt with by our courts. The appellant plaintiff instituted Magistrates’ Courts Suit No 357 of 2015 (“MC/MC 357/2015”) on 7 January 2015 against the respondent defendant for damages in respect of personal injuries suffered by him on 7 January 2012 when the respondent’s vehicle collided into the rear of the appellant’s stationary vehicle. The question for determination in HC/RAS 9/2016 is whether the date on which the appellant’s cause of action accrued is to be excluded or included for the purpose of computing the three-year limitation period under s 24A(2)(a) of the Limitation Act (Cap 163, 1996 Rev Ed).

2 The computation of time relating to the limitation timeline for causes of action is essential. If even a delay of one day is not a good reason to allow a time-barred claim to proceed (see Ang Sin Hock v Khoo Eng Lim [2010] 3 SLR 179 (“Ang Sin Hock”) at [78]), precise rules for the calculation of limitation periods are required. One day, ie, the period of 24 hours beginning with one midnight and ending with the next or the time it takes the earth to revolve once on its axis (see Black’s Law Dictionary (Bryan A Garner ed) (West Group, 9th Ed, 2009) at p 453; Jowitt’s Dictionary of English Law vol 1 (Daniel Greenberg ed) (Sweet & Maxwell, 3rd Ed, 2010) at p 638), may make all the difference in determining whether an action is time-barred or not.

3 This judgment examines the ambit and interpretation of provisions on the computation of time in the Rules of Court (Cap 322, R 5, 2014 Rev Ed) and the Interpretation Act (Cap 1, 2002 Rev Ed) – as well as relevant common law principles in the common law – and their application to the provisions stipulating limitation periods in the Limitation Act.

4 I agree with the position taken by the appellant that in computing limitation periods as provided for in the Limitation Act, the date the cause of action accrued should be excluded. Accordingly, the appellant’s action that was filed on 7 January 2015 is not time-barred.

5 The appellant and respondent were involved in a road traffic accident on 7 January 2012. It is common ground that, at about 4:00 a.m. to 4:50 a.m., a collision occurred at the junction of Victoria Street and Ophir Road involving motor car no. SGB 8876T driven by the appellant and motor lorry no. GBA 7313Y driven by the respondent. The appellant claims that he had brought his car to a stop at the junction when the traffic light turned amber, and that the respondent’s lorry had collided into the rear of his stationary car shortly after.

6 On 7 January 2015, the appellant instituted MC/MC 357/2015 claiming damages in respect of personal injuries suffered by him, alleging negligence of the respondent in the driving, management and control of his vehicle. The respondent entered an appearance to the suit on 20 January 2015 and filed a defence on 17 March 2015, primarily relying on s 24A(2) of the Limitation Act to plead that the suit was time-barred as it was not commenced within three years from the date on which the cause of action had accrued.

7 The respondent then applied in Summons No 4188 of 2015 (“MC/SUM 4188/2015”) to strike out the appellant’s Statement of Claim on the ground that the respondent’s action was time-barred at the time of the issue of the writ on 7 January 2015, the last day to bring the action purportedly being 6 January 2015. The Deputy Registrar hearing MC/SUM 4188/2015 on 18 August 2015 held that the action was time-barred and struck out the appellant’s action.

8 The appellant appealed against the Deputy Registrar’s decision and at the hearing of Registrar’s Appeal No 50 of 2015 (MC/RA 50/2015) on 16 September 2015, the District Judge dismissed the appeal. The District Judge also heard and dismissed the appellant’s application for leave to appeal against the decision in MC/RA 50/2015 to the High Court on 29 October 2015.

9 Both parties then appeared before me on 1 March 2016 upon the application of the appellant in Originating Summons No 1048 of 2015 (HC/OS 1048/2015) seeking leave to appeal to the High Court against the District Judge’s decision in MC/RA 50/2015. I granted the appellant leave to appeal, bringing us to the present appeal in HC/RAS 9/2016.

10 The Deputy Registrar recognised that there were contrary positions in both the case law and material presented to her which were “not reconciled”. ¹ She then decided to follow the apparent approach of the High Court in two relatively recent cases (Yan Jun v Attorney-General [2014] 1 SLR 793 (“Yan Jun”) and Management Corporation Strata Title Plan No 2827 v GBI Realty Pte Ltd and another [2014] 3 SLR 229 (“GBI Realty”)) and computed the limitation period to include the day the cause of action accrued. The limitation period was thus held to have expired on 6 January 2015, and the Deputy Registrar struck out the appellant’s action.

11 In her grounds of decision reported at Suresh s/o Suppiah v Jiang Guoliang [2015] SGMC 31 (the “GD”), the District Judge similarly took the position that the High Court decision of Yan Jun had dealt with the computation of time for the limitation period under s 24A(2)(a) of the Limitation Act (which is the same provision that applies in the present case), where the Judge had included the date when the cause of action accrued in computing the limitation period. Hence, the District Judge held that as the accident in the present case had occurred on 7 January 2012, the last date for instituting the action was 6 January 2015. She thus dismissed the appeal against the Deputy Registrar’s decision.

12 In the proceedings below, the respondent made the argument that only cases dealing specifically with s 24A of the Limitation Act would be relevant for consideration, and that cases interpreting the computation of time for s 6 of the Limitation Act (or s 2 of the Limitation Act 1939 (c 21) (UK) (the “UK Limitation Act 1939”) which was also the provision that related to actions founded on contract or tort) cited by the appellant were not, relying on the Court of Appeal’s holding in Lian Kok Hong v Ow Wah Foong and Another [2008] SGCA 30 at [14] that the two sections cannot apply concurrently (see [5] of the GD). Although the District Judge did not expressly state so, she seemed to have implicitly agreed with this argument when she held that s 24A(2)(a) of the Limitation Act applied and that the decision of Yan Jun applied without addressing the other cases cited by the appellant at all (see [9]–[10] of the GD). This treatment of the contrary authorities seems to be jumping the gun. Although the exact provisions in the Limitation Act that were engaged in those cases may have been different due to the different causes of action, the phrasing of the time limits is largely similar in the format of “action(s)…shall not be brought after the expiration of [time period] from the date on which the cause of action accrued”. With the issue at hand being one of computation of time and not about which limitation period applied, those contrary authorities are prima facie also relevant. Holding otherwise would be an implicit and counterintuitive recognition that the principles relating to the computation of time periods under different provisions in the Limitation Act could differ.

13 The focus of the appeal in HC/RAS 9/2016 is on the issue of whether the action was time-barred under s 24A(2)(a) of the Limitation Act, and the question that determines this issue is whether the date on which the appellant’s cause of action accrued is to be excluded or included for the purpose of computing the three-year limitation period under s 24A(2)(a) of the Limitation Act.

14 Both parties agree that the cause of action arose on the date of the accident on 7 January 2012, and it is also common ground that the appellant had the knowledge required to bring the action on the date of the accident. Thus, the relevant limitation period under s 24A(2)(a) of the Limitation Act is three years. If the date on which the appellant’s cause of action accrued were to be excluded, the relevant limitation period would have started running on 8 January 2012 and would hence have expired on 7 January 2015 (this date being the last day for bringing the action). His action instituted on 7 January 2015 would then not be time-barred. If it were to be included though, the relevant limitation period would have expired on 6 January 2015 and his action would therefore be time-barred. The timeline below summarises these two interpretations:

15 Section 24A(2)(a) of the Limitation Act reads:

16 As the relevant limitation period provided for is a statutory provision providing for a time period to bring an action before the courts, there are other statutory interpretative provisions that may apply to assist in the computation of time. Section 50 (in particular s 50(a)) of the Interpretation Act, which was cited by the appellant, may be one such provision:

17 I am also cognisant of O 3 r 2 of the Rules of Court, which states:

18 The appellant also cited a few English authorities (after contending that provisions in the UK Limitation Act 1939 were similarly worded as the ones in ours, where the preposition “from” is used with reference to the date of the accrual of cause of action) that seemed to point to a general principle as to computation of time that would exclude the day from which time begins to run for limitation periods. These cases support the definition of “from” in Stroud’s Judicial Dictionary of Words and Phrases (8th Ed) (Sweet & Maxwell, 2015) at p 1158:

19 Thus, these are the few questions that I will have to consider before I can address the overarching question of whether the date of the accrual of the cause of action should be excluded or included in the computation of time:

(a) Does s 50(a) of the Interpretation Act and/or O 3 r 2(2) of the Rules of Court apply in the interpretation and computation of time periods in the Limitation Act?

(b) Are there relevant principles in the common law relating to the computation of time and interpretation of statutory provisions providing for time periods that would apply?

(c) Is s 50(a) of the Interpretation Act inapplicable to computation of periods of time not expressed in days (ie expressed in weeks, months, years, etc., as in the present case being that of three years under s 24A(2)(a) of the Limitation Act) due to the express phrase used in the provision being “period of days” (emphasis added)?

20 Before going into the substantive analysis, I have to state that the conflicting local authorities cited by both parties do reflect that this issue has not yet been conclusively or authoritatively determined. Although the more recent cases that guided the Deputy Registrar’s and District Judge’s decisions (see above at [10]–[11]) seemed to indicate that the date of the accrual of the cause of action should be included, I note that these decisions, though relating to the issue of limitation, ultimately did not address or deal directly with the present (and narrower) issue where a day’s delay had mattered and submissions relating to this were not made. The Judge in both Yan Jun and GBI Realty was not actually considering or addressing this narrow legal question. The precise computation of time was not actually an issue that was argued upon as a day’s difference was not critical in both cases; the dates on which the actions were brought had well exceeded the relevant limitation periods. Instead, Yan Jun’s limitation issue was mainly whether the suit was subject to a three-year limitation period (see Yan Jun at [31]) while that in GBI Realty related to when the cause of action had actually accrued.

21 For reference, the Judge in Yan Jun stated that under s 24A(2)(a) of the Limitation Act, the plaintiff’s cause of action would have accrued by 20 July 2009 and the limitation period would have expired on 19 July 2012 (at [48]). In GBI Realty at [28], the Judge postulated a six-year limitation period under s 24A(3)(a) of the Limitation Act:

22 Similarly, the other local authorities that seemed to exclude the date of accrual of causes of action from the computation of limitation periods did not expressly consider the narrow issue here. The Court of Appeal in Ang Sin Hock (at [78]) seemed to have excluded the date of accrual of cause of action:

23 In Ang Sin Hock, the Court of Appeal seemed to have interpreted the limitation period to have expired on 12 April 2006 (the last day to bring the action), with the action commenced on 17 April 2006 being a total of five days late, ie, the date of accrual of the cause of action on 12 April 2000 was excluded from the computation, and time only started running on 13 April 2000 until it expired on 12 April 2006. This is by no means clear, as the Court of Appeal also commented in the same paragraph that time was “running from 12 April 2000”, which seemingly contradicts this computation. However, this apparent contradiction is of course not an issue if one interprets time running “from” a certain date to mean to exclude that date.

24 In Sun Fook Kong Construction Ltd (formerly known as Sung Foo Kee, Ltd) v Housing and Development Board [2004] SGHC 69 at [37], the Judge also appeared to have excluded the date of accrual of cause of action:

25 Lastly, in Vijayakumar a/l Subramaniam v Peng Jun Qing and another [2009] SGDC 158 (“Vijayakumar”) at [4], the District Judge expressly referred to the Interpretation Act in her computation of time in relation to limitation to exclude the day the cause of action arose:

26 Though neither party had submitted otherwise, a preliminary point must be made that the day of the date of accrual of the cause of action referred to in s 24A(2)(a) of the Limitation Act must either be wholly excluded or included. The precise time of the accrual of the cause of action is not taken into account. In the present case, for example, it cannot be argued that a writ hypothetically issued at 4.30 a.m. on 6 January 2015 is not time-barred since the accident occurred at 4.50 a.m. on 7 January 2012 and time ran from exactly then and expired after three years at exactly 4.50 a.m. on 7 January 2015. Generally, the word “date” is understood as meaning a 24-hour division of time shown on the calendar from midnight to midnight and the law does not take into account fractions of the day unless some special reason requires it or unless it is necessary in order to settle which of two acts done on the same day were to prevail: see Trow v Ind Coope (West Midlands) Ltd and another [1967] 2 QB 899 (“Trow v Ind Coope”) at 927; and Migotti v Colvill (1879) 4 CPD 233 (“Migotti v Colvill”) at 234, per Denman J. Lord Mansfield CJ expressed this point in Pugh v Duke of Leeds (1777) 2 Cowp 714 (“Pugh v Duke”) at 720 succinctly:

27 Applying this principle in Trow v Ind Coope where the plaintiffs had argued that the word “date” should be construed as “time” so that 12 months would run from 3.05 p.m. on 10 September 1965 to 3.05 p.m. on 10 September 1966, it was held that no account of the time was to be taken in interpreting a statutory provision that referred to the date of the issue of a writ. More importantly, Lord Denning MR commented that this principle should similarly extend to the interpretation of “date” in provisions stipulating limitation periods that refer to the “date” on which causes of action accrue (at 915):

28 Thus, the whole day of the date of accrual of the cause of action must be either included or excluded in the computation of the three-year limitation period in s 24A(2)(a) of the Limitation Act.

29 I am of the view that s 50(a) of the Interpretation Act is applicable in construing the computation of time in s 24A(2)(a) of the Limitation Act. On the other hand, O 3 r 2(2) of the Rules of Court is not.

30 Subject to a contrary intention in the relevant provision in the Limitation Act, s 50(a) of the Interpretation Act should apply to determine the computation of limitation periods as provided for in the Limitation Act. The primary provision governing the relevant limitation period here is s 24A(2)(a) of the Limitation Act, stipulating a three-year limitation period within which an action where the damages claimed consist of or include damages in respect of personal injuries to the plaintiff must be brought. The computation of this three-year period is governed, unless a contrary intention appears, by s 50(a) of the Interpretation Act.

31 O 3 r 2(2) of the Rules of Court, appears to be in line with—and arguably broader than (due to the absence of the problems of interpretation with the literal narrow phrase of “period of days”)—s 50(a) of the Interpretation Act. However, given that the Rules of Court are secondary legislation enacted pursuant to s 80 of the Supreme Court of Judicature Act (Cap 322, 2007 Rev Ed) (“SCJA”), O 3 r 2(2) does not govern the computation of time for the purposes of computing limitation periods under the Limitation Act.

32 In Thomas & Betts (SE Asia) Pte Ltd v Ou Tin Joon and another [1998] 1 SLR(R) 380, the Court of Appeal decided that s 50 of the Interpretation Act (Cap 1, 1997 Rev Ed), rather than O 3 r 2(5) of the Rules of Court 1996, applied to s 34(1)(c) of the Supreme Court of Judicature Act (Cap 322, 1985 Rev Ed) in force then, as the Rules of Court must be read subject to its primary legislation. The statutory time period here is located within the Limitation Act, and not under the SCJA. A fortiori, O 3 r 2(2) would not govern the computation of time for the purposes of the Limitation Act. Instead, s 50(a) of the Interpretation Act applies to determine how limitation periods ought to be reckoned, subject to any contrary intention in the provisions in the Limitation Act.

33 To this end, I have reached a position different from that of VT Singham J in Muhamad Solleh bin Saarani & another v Norruhadi bin Omar & others [2010] 9 MLJ 603 (“Muhamad Solleh bin Saarani”), a Malaysian Kuantan High Court case which was referred to by the respondent. The learned Judge concluded at [8] that the Limitation Act 1953 (No 254 of 1953) (M’sia) was a “complete code by itself” that excluded the operation of the Interpretation Acts 1948 and 1967 (No 388 of 1967) (M’sia) to construe the period of limitation prescribed in s 6(1)(a) of the Limitation Act:

34 Section 54 of Malaysia’s Interpretation Acts 1948 and 1967 is in pari materia with s 50 of our Interpretation Act. Resorting to the Interpretation Act provisions on the computation of time for the purposes of the Limitation Act would not be “illegitimate” unless the Limitation Act indicates a contrary intention or provides otherwise. Unlike what the Judge in Muhamad Solleh bin Saarani had stated, the provisions on computation of time would in no way “determine when [a] cause of action [arises]” or “define the period of limitation”, which the Judge rightfully considered to be solely within the ambit of the Limitation Act. Interpretation Act provisions would instead assist in construing the running and calculation of the defined periods of limitation as stipulated in the Limitation Act. The Limitation Act is “written law”, like all other statutory legislation having the force of law in Singapore (including even the Constitution) (see s 2 of the Interpretation Act), and not some “special law” that automatically excludes the operation of provisions in the Interpretation Act.

35 In Mansource Interior Pte Ltd v Citiwall Safety Glass Pte Ltd [2014] 3 SLR 264 at [13], it was held that s 50(a) of the Interpretation Act provided guidance on the computation of time relating to the seven-day timeline for lodging an adjudication response under s 15(1) of the Building and Construction Industry Security of Payment Act (Cap 30B, 2006 Rev Ed) (the “SOP Act”) in the absence of provisions under the SOP Act to explain the computation of time. Although the appeal against the Judicial Commissioner’s decision was allowed in Citiwall Safety Glass Pte Ltd v Mansource Interior Pte Ltd [2015] 5 SLR 482, this point was not an issue on appeal.

36 Likewise here then, the Limitation Act does not contain any provisions to explain the computation of time periods referred to in its provisions. Thus, s 50(a) of the Interpretation Act would apply to compute the limitation period stipulated under s 24A(2)(a) of the Limitation Act, in the absence of contrary intention found in the Limitation Act.

37 There are commentators that appear to treat the provisions in both the Interpretation Act and Rules of Court as enacting the same principle of computation of time. The commentary on O 3 r 2(2) of the Rules of Court in Singapore Civil Procedure 2016 vol 1 (Foo Chee Hock JC gen ed) (Sweet & Maxwell, 2016) (“Singapore Civil Procedure”) at para 3/2/2 reads:

38 Jack Lee Tsen-Ta in “Getting Called: Recent Developments” (1998) 19 Sing LR 298 also takes a similar view and further traces the principle to its common law roots:

39 From my survey of the historical development of s 50(a) of the Interpretation Act (see below at [81]–[87]), relevant older English cases (see below at [41]–[50]) predating the genesis of the first version of the provision in the Straits’ Settlements’ Interpretation Act 1867 (Act 14 of 1867) as well as other similar statutory provisions in other Commonwealth jurisdictions (see below at [88], I have come to agree with this view regarding a uniform principle relating to the computation of time enacted consistently in the Interpretation Act to interpret all written law and in the Rules of Court to govern civil procedure. Essentially, the same principle is enacted in both s 50(a) of the Interpretation Act and O 3 r 2(2) of the Rules of Court.

40 In the United Kingdom (“UK”), rules on computation of time are not statutorily provided for. The Interpretation Act 1889 (c 63) (UK) and its successor Interpretation Act 1978 (c 30) (UK) do not have provisions equivalent to s 50 of Singapore’s Interpretation Act. As such, guidance on the computation of time periods provided for in other UK statutes would be located in case law.

41 It appears that in construing the words “from”, “within” or “after” where questions of time are concerned, the English authorities have largely interpreted it to indicate that the first date should be excluded from the computation of the time period.

42 In Marren v Dawson Bentley & Co Ltd [1961] 2 QB 135 (“Marren v Dawson”), Havers J considered the interpretation of s 2(1) of the UK Limitation Act 1939, as amended by the Law Reform (Limitation of Actions, etc.) Act 1954, which provided for a relevant limitation period stipulating that actions “shall not be brought after the expiration of [three] years from the date on which the cause of action accrued” (which is similarly phrased as the provisions in Singapore’s Limitation Act). He held that the day of the accident was to be excluded from the computation of the period within which the action should be brought. Marren v Dawson was affirmed by the UK Court of Appeal in Trow v Ind Coope and Pritam Kaur v S Russell & Sons Ltd [1973] 1 QB 336 (“Pritam Kaur”). The decision drew support from various older cases:

(a) In Radcliffe v Bartholomew [1892] 1 QB 161, the statutory provision in issue was s 14 of the Cruelty to Animals Act 1849 (c 92) (UK), where complaints were to be made “within one calendar month after the cause of such complaint shall arise” (emphasis added). It was held that the day on which the cause of the complaint arose was to be excluded from the computation of the calendar month within which the complaint was to be made. The complaint brought on June 30 in respect of an act of cruelty allegedly committed on May 30 was therefore made in time.

(b) In Williams v Burgess (1840) 12 Ad & E 635, the statutory provision in issue was s 1 of the Warrants of Attorney Act 1822 (c 39) (UK) which stipulated that warrants of attorney to confess judgment shall be filed “within-twenty one days after the execution” (emphasis added). It was held that the day of execution was to be reckoned exclusively. The warrant executed on 9 December and filed on 30 December was thus in time.

(c) In Hardy v Ryle [1828] 9 B & C, the statutory provision in issue was s 8 of the Continuance of Acts 1740 (c 34) (UK) which prescribed that “no action shall be brought against any justice of the peace for anything done in the execution of his office unless commenced within six calendar months after the act committed” (emphasis added). It was held that the day of the seizure (the act committed) was to be excluded in computing the six months. The action commenced on 14 June after the relevant act on 14 December was thus in time.

43 In Pritam Kaur, the UK Court of Appeal considered the computation of time for the purposes of s 2(1) of the UK Limitation Act 1939 as well: “The following actions shall not be brought after the expiration of three years from the date on which the cause of action accrued…” (emphasis added). The UK Court of Appeal dismissed the cross-appeal and affirmed Marren v Dawson’s construction of the computation of time to exclude the first day on which the cause of action accrued, with Lord Denning MR also pointing out (at 348) that nothing turned on the difference in wording, regardless of whether the relevant provision was phrased as “after the expiration of [a time period] from [a date]” or “within [a time period] after [an act or event]”:

44 In contrast, the UK Court of Appeal had earlier held in Trow v Ind Coope that the date of issue was to be included when the statutory provision in question was phrased in the manner of “within [a time period] beginning with [a date/an act or event]”. This case concerned the construction of O 6 r 8(1) of the Rules of the Supreme Court 1962 (SI 1964 No 1213) (UK), which provided that “[f]or the purpose of service, a writ … is valid in the first instance for 12 months beginning with the date of its issue.” The majority held that the ordinary meaning of the statutory words where the period of 12 months was expressed as “beginning with the date of its issue” meant that the computation of time included the date of the issue. The court affirmed the finding in Hare v Gocher [1962] 2 QB 641 where the words of the statutory provision were phrased as “at the expiration… of one month beginning with the date on which it is passed” (emphasis added), it meant that the day upon which the Act was passed was to be included in the computation of time. I should note that Lord Denning MR dissented (see Trow v Ind Coope at 915) as he felt that distinguishing between “beginning with” and “beginning from” was too subtle a distinction that would be “out of touch with the common man”.

45 Fortunately, this purported distinction is not an issue in the present case. In general, I am of the view that there ought not to be an absolute rule as to the consequences of the use of the different prepositions without first considering the context in which it is found. The rational mode of computation is to have regard in each case to the purpose for which the computation is made: see Lord Esher MR’s comments in In re North; Ex parte Hasluck [1895] 2 QB 264 (“In re North”) at 269. However, the plain meaning or settled interpretation of the prepositions used in the statutory language would of course be highly relevant.

46 For our purposes, Lord Denning MR’s survey of the authorities dealing with the word “from” (in Trow v Ind Coope at 916–917) in this area of law is instructive:

47 Although in Lester v Garland (1808) 15 Ves 248, Sir William Grant MR had not definitively laid down an absolute rule of computation, his comments favouring an exclusion of the first day in the computation of time periods have come to be adopted as the general rule subsequently. Notably, Lord Mansfield CJ had observed just a century before in Pugh v Duke (at 725) that the word “from” could mean either inclusive or exclusive (albeit in the construction of a lease, and not a statutory provision), according to the context and subject matter:

48 In Webb v Fairmaner (1838) 3 M & W 473 at 476, Parke B (like Lord Denning MR in Trow v Ind Coope) treated the decision in Lester v Garland as laying down a “sound rule” as to computation of time: where computation is to be made from a certain day, that day is to be excluded:

49 By 1903, Mathew LJ sitting in the UK Court of Appeal treated this rule as “well established” in The Goldsmiths’ Company v The West Metropolitan Railway Company [1904] 1 KB 1, at 5:

50 And in Halsbury’s Laws of England vol 97 (Butterworths, 2015) at para 336, this common law principle is expressed as such:

51 I note that this common law approach is also recognised in cases from Australia and Canada. Their courts construe a period of time which runs “from” a certain date as excluding that date: see Hughes and another v NM Superannuation Pty Ltd and another (1993) 29 NSWLR 653 at 667; Dennis C Pearce & Robert S Geddes, Statutory Interpretation in Australia (LexisNexis Butterworths, 8th Ed, 2014) at pp 295–297; Walton v Cote (1989) 16 ACQS (3d) 288 at [2]; and McCann Milling Co v Martin (1907) 15 OLR 193 at [4].

52 In Marren v Dawson, Havers J declined to follow Gelmini v Moriggia and another [1912] 2 KB 549 (“Gelmini v Moriggia”) where it was apparently held that the date of the accrual of the cause of action was included and not excluded from the computation of a limitation period. In Gelmini v Moriggia, the statutory provision in issue was s 3 of the Limitation Act 1623 (c 16) (UK) which reads that “[a]ll actions… shall be commenced and sued within the time and limitation hereafter expressed, and not after (that is to say)… within six years next after the cause of such actions or suit.” In that case, the last day for payment of a promissory note was on 22 September 1906, and the writ in an action upon the note against the makers was issued on 23 September 1912. Channell J held that since the cause of action was complete at the commencement of 23 September 1906, that whole day (23 September 1906) was to be included in the computation of the six-year limitation period which expired on 22 September 1912. Hence, he held that the writ issued on 23 September 1912 was too late.

53 However, it would seem that the decision in Gelmini v Moriggia could be reconciled with the general rule as followed in Marren v Dawson and Radcliffe v Bartholomew and Channell J’s holding could be interpreted as being in line with it (see Andrew McGee, Limitation Periods (Sweet & Maxwell, 7th Ed, 2014) at pp 27–28). As the general rule to exclude was based on disregarding parts or fractions of a day and allowing potential claimants the whole of the stipulated limitation period to bring their actions, the decision to include the whole day of 23 September 1906 in the computation of time in Gelmini v Moriggia does not flout that principle as the cause of action had technically accrued between 22 and 23 September 1906 or at 0000 hrs on 23 September 1906. At the last moment of 22 September 1906, the time for payment on the promissory note had not yet expired. However, at the first moment of 23 September 1906, it had expired and the cause of action had accrued by then. Hence, there would be no effective backdating in holding that time, for the purpose of the limitation timeline, ran from the commencement of 23 September 1906. No part or fraction of a day was involved at all. However, situations where causes of action can be said to accrue right at the beginning of a day would arise only in rare technical situations.

54 Thus, it is not totally clear that Gelmini v Moriggia is no longer good law, notwithstanding Haver J having expressly declined to follow it in Marren v Dawson. As discussed above, the two cases are arguably reconcilable. Nonetheless, an exceptional Gelmini-type situation is clearly not present here.

55 After examining the development of the common law principles in this area, I now turn to the justification and rationalisation of the “general rule” as applied to the context of limitation periods.

56 The principle that the law does not take into account fractions of a day (fractionem diei non recipit lex) has been used to justify and explain this rule adopted to exclude the first day. In Lester v Garland at 257, Sir William Grant MR observed that because a day is treated as indivisible, an act done in the compass of a day cannot properly be said to be passed until the day is passed:

57 A day is but punctum temporis—an indivisible period of time / a point of time (see Nichols v Ramsel (1677) 2 Mod 280 at 281). Windeyer J’s comments in the High Court of Australia case of Prowse v McIntyre and others (1961) 111 CLR 264 (“Prowse v McIntyre”) at 277–278 are instructive:

Similarly, Lord Mansfield CJ in Pugh v Duke said that reference to a date does not mean the hour or the minute, but the day of delivery, and in law there is no fraction of a day (see [26], above).

58 However, this technical rule and legal fiction is ultimately a convenient rule for the sake of temporal certainty, meant to overcome the practical and evidentiary problems with identifying the precise time of an event with certainty.

59 Another justification for the general rule is derived from considering the hypothetical situation of a one-day time period. If one considers a case of a one-day limitation period (the “one-day test”), the reasonableness of the conclusion to exclude the date of accrual of the cause of action becomes apparent.

60 In this vein, Parke B in Young v Higgon (1840) 6 M & W 49 at 54 remarked: “[a]pply the criterion which has been before suggested—reduce the time to one day, and then see what hardship and inconvenience must ensue if the principle I have stated is not to be adopted.” This “test” was subsequently also adopted in Radcliffe v Bartholomew by Wills J (and cited by Havers J in Marren v Dawson) who commented that “the result of reducing the time to one day would be that an offence might be committed a few minutes before midnight, and there would only be those few minutes in which to lay the complaint, which would be to reduce the matter to an absurdity” (at 163).

61 Underlying the conclusion from this notional test to exclude the first day is the principle that the statutory stipulation of a period of time has to be fully adhered to, ie, the whole of three years to bring an action must pass before the limitation period is held to have expired, giving plaintiffs their full limitation period to bring their actions. Thus in Young v Higgon, Alderson B remarked that “where there is given to a party a certain space of time to do some act, which space of time is included between two other acts to be done by another person, both the days of doing those acts ought to be excluded, in order to ensure to him the whole of that space of time” (at 54). Similarly, Gurney B in the same case relied on this principle “so as not to be in restriction of the general right of bringing actions” (at 55).

62 This brings us to what seems to have been either rarely discussed or disregarded in the authorities after Lester v Garland: the observation that Sir William Grant MR seemingly accepted from counsel for the plaintiff (see Lester v Garland at 256) that where the act done from which the computation is made is one to which the party against whom the time runs is privy, the day of the act done is included; but where it is one to which he is a stranger or foreign to, it may be excluded. Although Bayley J subsequently adopted and applied this distinction in Hardy v Ryle (see 607–608), this distinction and reasoning was doubted by Parke B in Webb v Fairmaner at 477 as being “not…quite satisfactory”. In Hardy v Ryle, it was rationalised that in the case of a seizure of the plaintiff’s goods under the authority of a warrant granted by the defendant, the plaintiff was not privy to the seizure and thus the day of the seizure would be excluded in the computation of time. In the present case, since the appellant, against whom the limitation timeline runs, was not privy to the act done on the day of the accident, ie, the alleged collision against the rear of his stationary vehicle by the respondent (in the sense that a wrong was alleged to have been committed against him), the day of the accrual of the cause of action would be apparently excluded under this rule. However, considering that Sir William Grant MR was reluctant to lay down a general rule upon this subject, his comments on this distinction should not be interpreted as doing so.

63 Another rule of construction has been expressed by the courts to be “fair”: where the “computation is to be for the benefit of the person affected as much time should be given as the language admits of, and where it is to his detriment the language should be construed as strictly as possible” (see In re North at 270). In In re North, it was thus rationalised that this rule would lead to excluding the first day (or part of a day) since the provision in question affected debtors in a very detrimental manner and the provision thus ought to be construed as much for the debtor’s benefit as possible. Without commenting on the soundness of such a rule, it cannot be gainsaid that, at first blush, this rule may not assist the computation of limitation periods, since the affected positions of both the claimant and the defendant are at diametrically opposite ends and have to be balanced. Of course, if one takes the simpler view that only the claimant’s right to bring an action is being directly affected, the conclusion to exclude the date of accrual of the cause of action would be reached to construe the statute as much as possible for the claimant’s benefit.

64 However, considering the context and subject-matter of limitation law, I am of the view that excluding the date of the accrual of the cause of action would be the more reasonable construction that affords potential claimants the full periods of limitation as stipulated by the Limitation Act.

65 Singapore’s Limitation Act was enacted in 1959 (No 57 of 1969), modelled after the UK Limitation Act 1939 and then subsequently amended to include the reforms of the UK Limitation Act 1980 (c 58) and UK Latent Damage Act 1986 (c 37) (see Singapore Parliamentary Debates, Official Report (2 September 1959) vol 11 at col 587 (K M Byrne, Minister for Labour and Law); and Singapore Parliamentary Debates, Official Report (29 May 1992) vol 60 at col 32 (Professor S Jayakumar, Minister for Law)). The Limitation Act currently stipulates limitation periods categorised based on particular causes of actions.

66 The three-fold functions underlying statutes of limitation was laid out in the Report of the Law Reform Committee of England and Wales entitled Limitation of Actions in Cases of Personal Injury (Cmnd 1829, 1962), as follows (at para 17, which was also quoted by the Singapore Court of Appeal in Ang Sin Hock at [77]):

67 As I mentioned earlier, limitations law has to balance what is fair to claimants who would wish to have as long a period as possible to bring their claims, and what is fair to defendants who should be protected from stale claims. With these considerations in mind, I am of the view that it is reasonable, after looking at the one-day test and the fact that it would be unfair to potential claimants, to effectively backdate the commencement of the limitation period and the running of time (ie to 0000 hrs of the day of the accrual of the action), to exclude the date of the accrual of the cause of action in the computation of time for the purposes of limitation.

68 The one-day test which was purportedly ascribed to Lord Tenterden CJ in Pellew v The Inhabitants of the Hundred of Wonford in the County of Devon (1829) 9 B & C 134 (“Pellew”), and subsequently applied in Webb v Fairmaner at 477, Young v Higgon at 54, In re Railway Sleepers Supply Company (1885) 29 Cd D 204 at 207, and Radcliffe v Bartholomew at 163–164, was raised as an example (on close reading) by Lord Tenterden CJ in Pellew not actually as an arbitrary rule but as a test of reasonableness to aid the construction of the statutory provision in question in that case. Like Sir William Grant MR in Lester v Garland, Lord Mansfield CJ in Pugh v Duke and Lord Esher MR in In re North (see above at [45] and [47]), Lord Tenterden CJ had eschewed an absolute and inflexible rule and had in fact arrived at his conclusion by “[l]ooking at the Act of Parliament, and the object with which it was made” (at 144).

69 If, hypothetically, the accident at Victoria Street had occurred at 11.59 p.m. and the relevant limitation period was only one day, holding that the date of the accrual of cause of action should be included in the computation of time would be logically absurd. Including the first day would mean that the claimant would have only a minute to bring the action. Even if the relevant period was a week or seven days, including the first day would lead to the effective backdating of the commencement of the running of time by a whole day less a minute. Depriving claimants of the fraction of the day that has passed from midnight to the precise time of accrual of the cause of action would unduly prejudice them by not affording them the whole of the limitation period clearly provided for. Considering the severe effect of a system of limitation that bars claims offending the requisite limitation timeline and which has no sympathy for claimants where even “a delay of one day would not be a good reason to allow a claim offending the requisite limitation timeline to proceed” (see Ang Sin Hock at [78]), the balance to be struck in this situation should fall in favour of potential claimants.

70 With the passage of time and the ebbs and flows of the tide of the common law, Sir William Grant MR’s comments in Lester v Garland have come to be seen as laying down a general principle as to the computation of time despite his express reservations regarding an absolute rule. However, blind applications of technical rules without considering the context of statutory language would be unwise and may result in injustice. In the present case, the conclusion to exclude the date of the accrual of the cause of action is sound and fair, and is one that accords with the context and purpose of the Limitation Act.

71 Thus, even without the benefit of resorting to s 50(a) of the Interpretation Act, I would still hold that the date on which the appellant’s cause of action accrued is to be excluded for the purposes of computing the three-year limitation period under s 24A(2)(a) of the Limitation Act.

72 The next issue relates to a narrow point of construction regarding s 50(a) of the Interpretation Act. Would the fact that the express phrase used in the provision to refer to time periods is a period of “days” bar the application of s 50(a) to computation of periods of time not expressed in days (ie, weeks, months, years, etc., as in the present case, being three years under s 24A(2)(a) of the Limitation Act)?

73 Dr Choong Yeow Choy considers the general rule to exclude the date of the accrual of cause of action in the English cases of Marren v Dawson and Pritam Kaur to be “consistent” with s 54(1) of Malaysia’s Interpretation Acts 1948 and 1967 (which are in pari materia with s 50 of Singapore’s Interpretation Act) (see Choong Yeow Choy, Law of Limitation (Butterworths Asia, 1995) at pp 45–46). Similarly, Jack Lee and the author of Singapore Civil Procedure have expressed the same view that this principle of computation of time in the common law has been given statutory recognition in the Interpretation Act (see above at [37]–[38]).

74 The appellant submits that a “month” or “year” constitutes a period of days, and therefore for the computation of time, s 50(a) of the Interpretation Act applies. ² I agree that the expression “period of days” does not bar the application of s 50(a) of the Interpretation Act to statutory provisions that refer to time periods not expressed in units of days.

75 It would also be useful to consider the interpretation of provisions found in Interpretation statutes in foreign jurisdictions that are in pari materia with s 50(a) of the Interpretation Act. Apart from Malaysia and Hong Kong, many other Commonwealth jurisdictions have adopted the same provisions on computation of time. For our purposes though, the foreign authorities that are on point are scant.

76 In Malaysia, the Johor Bahru Federal Court in Setali Development Sdn Bhd & another v Lim You Keng [1984] 1 MLJ 26 applied s 54(1)(a) of the Malaysian Interpretation Act 1967 (exactly worded like s 50(a) in Singapore’s Interpretation Act) in the computation of the time period of “one month” under r 56 of the Rules of the Federal Court 1980 (M’sia) which disallowed appeals without leave after the expiration of one month. The court did not seem to find any difficulty with applying s 54(1)(a) in spite of the unitary reference to “days”.

77 However, the courts in Hong Kong have reached a different conclusion. In Li Tat Kong v Official Receiver & another [2001] HKEC 35 (“Li Tat Kong”), the Hong Kong Court of Appeal interpreted the reference to “period of days” in s 71(1)(a) of the Interpretation and General Clauses Ordinance 1966 (Cap 1) (HK) (“Hong Kong’s Interpretation Ordinance”) literally to mean period of “days” only and “not a period of weeks, months or years” in its interpretation of the reference to “period” in a subsequent sub-section.

78 Similarly, in Secretary for Justice v Maxim's Caterers Ltd [2009] 4 HKLRD 723 which dealt with the interpretation of s 26 of the Magistrates Ordinance (Cap 227) (HK) that required any information in respect of an offence to be laid “within 6 months from the time when the matter of such information …arose”, the Deputy Judge adopted the reasoning in Li Tat Kong that s 71(1) of Hong Kong’s Interpretation Ordinance was “relevant clearly to the computation of a period of days rather than weeks, months or years”. However, she eventually allowed the appeal and held that the six-month time limit excluded the day of the offence by resorting to the common law position (at [9]–[10]):

79 To a certain extent, it would not totally fly against logic to consider that any time period could be a “period of days”, with a day being a smaller unit that makes up time periods expressed in weeks, months, years, etc.. In Prowse v McIntyre, Windeyer J referred to the coming of age of a person as the “end of a period of days” in, apparently treating all time periods (that may extend to years and decades) as essentially “period[s] of days” (at 280).

80 However, it would also not be inaccurate to consider the fact that larger units of time vary in length and are not the same. Months and years vary in length and ought to be taken as months and years respectively in computing the passage of time. When computing time by calendar months for example, time is reckoned by looking at the calendar and not by counting days. Thus, one calendar month from any given day of a given month is held to expire upon arriving at the first moment of the corresponding day in the next month. A period of one month starting on 2 January will expire upon the first moment of 2 February (ie, the end date of the period is 1 Feburary). This principle is also termed the “corresponding date rule” (see Jeow Fong Mei v Chong Mee Yoke [1996] 1 MLJ 387, following Migotti v Colvill, and also the House of Lords’ decision in Dodds v Walker [1981] 2 All ER 609). For further illustration of how the unit of expression used is important, the time periods expressed as 30 days, four weeks and a month from 1 January 2016 (assuming 1 January 2016 is excluded) would all actually result in different end dates: namely 31 January 2016, 29 January 2016 and 1 February 2016.

81 Looking at the historical development of the provision though, it would seem that the phrase “period of days” in s 50(a) was intended to refer to time periods in general, regardless of the base unit of time they are expressed in.

82 Section 50(a) can be traced back to the first legislative provision dealing with the computation of time in s 16 of the Interpretation Act 1867 (SS Act No 14 of 1867) passed by the Straits Settlements Legislative Council, which seemed to codify the common principle then to exclude the first day (by 1838 in Webb v Fairmaner, Parke B had treated the 1808 decision in Lester v Garland as laying down this “sound rule”), with reference to time as “computed by days”:

83 The enactment of the General Clauses Ordinance 1888 (SS Ord No 1 of 1888) subsequently repealed the Interpretation Act 1867 and the provision was redrafted as such in s 10(1):

84 In this iteration of the principles relevant to the computation of time in statutes (which continued to be in this form in the subsequent Interpretation Ordinance 1912 (SS Ord No 6 of 1912) which replaced the General Clauses Ordinance 1888), it is clear that, at least by 1888, the exclusion of the first day in the computation of a time period is a principle that applies to “any period of time”, regardless of the unit of time used to express it in.

85 The current statutory language of s 50(a) in the Interpretation Act only came about when s 36(a) of the Interpretation and General Clauses Ordinance 1948 (Ord No 7 of 1948) was enacted in the Malayan Union:

86 This same phrasing was then adopted in s 40(a) of the Interpretation and General Clauses Ordinance 1951 (No 4 of 1951), which was passed by the Legislative Council of the Colony of Singapore, repealing the Straits Settlements’ Interpretation Ordinance 1912. The provision as currently worded was then preserved from the Interpretation Act 1965 (No 10 of 1965) onwards.

87 From this brief historical survey, it would seem that the phrase “period of days” should not be narrowly interpreted as, without any indication otherwise, I would assume that the principles on computation of time in the interpretation statutes are intended to remain unchanged. Section 10(1) of the General Clauses Ordinance 1888 clearly refers to time periods in general, and even statutorily references the common law general rule in construing the word “from” in the computation of time. Thus, I find that s 50(a) refers to time periods in general and applies to the computation of the three-year period in s 24A(2)(a) of the Limitation Act.

88 More importantly, this construction accords with the context and purpose of the Limitation Act. I also note that this is consistent with other statutory provisions in Commonwealth jurisdictions which express the same principle but with clearer and express reference to “period[s] of time” or “time” in general, instead of “periods of days”: see s 35(2) of the Interpretation Act 1999 (No 85) (New Zealand); s 26(4) of the Interpretation Act, C 1985, (c I-21) (Canada); s 39(2) of the Interpretation Act 1954 (Cap 33) (Northern Ireland); and s 36(1) of the Acts Interpretation Act 1901 (No 2) (Australia).

89 In conclusion, s 50(a) of the Interpretation Act is applicable in determining how to compute time for the purposes of s 24A(2)(a) of the Limitation Act. However, even if I am wrong on the point that s 50(a) should not be narrowly construed to only apply to time periods expressed strictly in units of days, I would reach the same conclusion by following the general rule at common law, which excludes the day of the cause of action in computing the limitation period (see [41]–[51] and [64]–[71] above). Hence, in computing the period of three years from the relevant event, that is, the date of the accrual of the cause of action, the day that the event happened shall be excluded.

90 In the present case, the appellant’s cause of action arose on the day of the accident on 7 January 2012. With 7 January 2012 excluded from the computation, time only started running on 8 January 2012 with 7 January 2015 being the last day for the appellant to bring the action against the respondent. Thus, the action in MC/MC 357/2015 which was instituted on 7 January 2015 is not time-barred.

91 For the reasons stated, the appellant’s appeal in HC/RAS 9/2016 against the decision of the District Judge is allowed.

92 I will hear parties on costs if no agreement is reached.