Salmizan bin Abdullah v Crapper , Ian Anthony [2023] SGHC 75

This is the defendant’s application pursuant to O 33 r 2 of the Rules of Court (2014 Rev Ed) (“ROC 2014”) for the preliminary determination of the following three questions of law (“the Questions”) prior to the assessment of damages (“AD”) for the underlying matter. The defendant has framed the Questions as follows:

(a) whether causation can be reserved in toto to the AD Stage (“Question 1”);

(b) if the answer to (a) is no, whether causation can be reserved to the AD Stage, if parties accept that the claimant suffered one or more types of special damages causally connected to the defendant’s breach of duty (“Question 2”); and

(c) if the answer to (b) is no, whether causation can be reserved to the AD Stage, if parties accept that the claimant suffered one or more types of general damages causally connected to the defendant’s breach of duty (“Question 3”).

Having considered the parties’ submissions, including those of the learned young independent counsel, Ms Leo Zhi Wei (“Ms Leo”), I answer Question 1 in the negative. Further, I consider that Question 2 and Question 3 should be reframed, and that the gist of those questions is whether causation can be reserved to some extent to the AD Stage. I also answer this reframed question in the negative but with some qualifications. I provide the detailed reasons for my answers below.

Background context to the Questions

The general context in which the Questions have arisen: PIMA cases

In Singapore, the Protocol can be found in Appendix B of the State Courts Practice Directions 2021 (“PD 2021”). Pursuant to para 39(2) of the PD 2021, claimants in all PIMA cases must comply with the Protocol before commencing court proceedings. The Protocol was previously found in Appendix E of the State Courts Practice Directions 2014 (“PD 2014”). While the present case concerns an accident which occurred prior to the application of the PD 2021, I will deal primarily with the PD 2021 in as much as the material parts of the Protocol remain unchanged across the two Practice Directions.

The specific context in which the Questions have arisen: MC/MC 8815 of 2020

Against this general context, the Questions have arisen in the scenario where the parties have agreed on liability and entered into a consent interlocutory judgment on such. The present case arose from MC/MC 8815 of 2020 (“MC 8815”). This is the plaintiff’s claim against the defendant for personal injury, loss, and expenses arising out of an accident on 29 March 2019. The defendant was driving a motorcycle which was insured with Direct Asia Insurance (Singapore) Pte Ltd when it collided with the plaintiff’s motor car along Loyang Lane. As a result of the accident, the plaintiff allegedly suffered neck pain and back pain. Presumably after going through the Protocol (albeit under the PD 2014), the plaintiff filed his Statement of Claim on 21 August 2020. Pursuant to O 18 r 12(1A) of the ROC 2014, the plaintiff also annexed to his claim a medical report and statement of the special damages claimed. The plaintiff claimed for general damages and special damages. The special damages claimed included loss of income amounting to $434.00, medical expenses amounting to $66.65, and transport expenses amounting to $30.00. In his Defence filed on 24 September 2020, the defendant resisted the plaintiff’s claim on the basis of, among other grounds, the lack of causation of the plaintiff’s losses.

It is noteworthy that despite the terms of the consent interlocutory judgment making no mention of issues of causation, in the completed Form 9I dated 8 January 2021 (which is the equivalent of Form 7 under the PD 2014), the parties had ticked the box which indicated that by consent, interlocutory judgment is entered for the plaintiff against the defendant at 90%, “leaving the issues of damages and causation to be assessed and costs reserved to the Registrar assessing the damages” [emphasis added]. The current version of Form 9I no longer contain the words “and causation” in the equivalent sentence. Neither does Form 7 in the PD 2021 contain those words. I understand that the prevailing practice at one time in the State Courts was to allow the defendant to challenge the causation of injuries in toto at the AD Stage despite the entering of interlocutory judgment by consent in relation to liability.

Subsequently, pursuant to an order of court dated 24 March 2021, the plaintiff’s proposed medical doctor, Dr Stephanie Ong, and the defendant’s proposed technical specialist, Mr David James Hunter (“Mr Hunter”), were appointed as Single Joint Experts (“SJEs”) under O 108 r 5(3)(a) of the ROC 2014. Their respective SJE reports were duly filed. Mr Hunter’s report stated that the damage profiles of the vehicles involved were not consistent with the level of force transference required to have caused the plaintiff to suffer neck and lower back pain from the accident. Therefore, relying on the SJE reports, the defendant challenged the causation of the plaintiff’s injuries. More specifically, from Mr Hunter’s report, the defendant had disputed the causal connection between the accident and the plaintiff’s injuries. By extension, the defendant also challenged the plaintiff’s heads of claims for general damages. As a result, in the Joint Opening Statement dated 13 June 2022 filed for the AD hearing in MC 8815, the defendant submitted a “nil” position in respect of the plaintiff’s claim for pain and suffering for neck pain and back pain. In particular, the defendant had stated “causation disputed” in relation to both of these injuries. However, the defendant did agree to pay the sum of $66.65 claimed for the plaintiff’s medical expenses and the sum of $10.00 for the plaintiff’s transport expenses.

At the AD hearing for MC 8815 before the learned Deputy Registrar Lewis Tan (“DR Tan”), the parties confirmed that they were willing to proceed with the AD despite the defendant disputing the causation of the plaintiff’s general damages. However, DR Tan was reluctant to proceed with the AD hearing on this basis in light of the Court of Appeal decision of Tan Woo Thian v PricewaterhouseCoopers Advisory Services Pte Ltd [2021] 1 SLR 1166 (“Tan Woo Thian”). In that case, the Court of Appeal held ex tempore that (at [8]), in a bifurcated trial, the plaintiff at the liability stage would need “to show that he did, in fact, suffer one or more types of loss that was causally connected to the alleged breach” [emphasis in original]. DR Tan expressed the view that “there does not seem to be any decision whereby assessment went ahead when parties only consented to [special damages]”.

As a result, the defendant filed HC/OA 301 of 2022 (“OA 301”) for MC 8815 to be transferred to the General Division of the High Court (“the High Court”) under s 54B(1) of the State Courts Act 1970 (2020 Rev Ed) (“SCA”) for the determination of the Questions. According to the defendant, the rationale for OA 301 was to seek a High Court pronouncement to settle the extent to which causation might be contested at the AD Stage after the entering of interlocutory judgment against a defendant. The defendant says that this would affect more than the immediate interest of the parties in MC 8815 and would benefit future litigants, insurance companies, and their legal advisors in planning their litigation roadmap and strategies in relation to causation. On 5 August 2022, an Assistant Registrar (“AR”) allowed the transfer of MC 8815 to the High Court for the Questions to be answered.

The problem raised by the Questions summarised

Reduced to its core, the problem raised by the Questions concerns the extent to which causation may be challenged at the AD Stage after the entering of a judgment on liability by consent. Specifically, the “causation” that the Questions refer to is the causal connection between a defendant’s breach of duty and a claimant’s damage or injuries. Also, the nature of the judgment is not usually material: it may be by consent or not, and it may be interlocutory or final. This problem arises because Tan Woo Thian has reminded us of the fundamental principle that causation is an element of liability in the tort of negligence which PIMA cases are based on. Parenthetically, I should say that while Tan Woo Thian was not a PIMA case, its statements on the relevant principles of the tort of negligence are clearly universal and apply to PIMA cases as well. However, the effect of Tan Woo Thian in practice has not been determinatively settled, especially in relation to the process by which PIMA cases are dealt with. Yet, it would appear that the profession has always had sight of this problem. Thus, as the learned authors of “What to do in a motor accident” anticipated several years ago (at footnote 43):

The problem can be illustrated more clearly by a hypothetical example concerning a matter that has been bifurcated along liability and AD. Suppose that a claimant claims to have suffered injuries A, B, and C in an accident in which a defendant breached his duty of care owed to the claimant. However, the claimant can only establish that injury A was caused by the defendant’s breach of duty. Despite this, the parties enter into a consent interlocutory judgment for the claimant in the tort of negligence. At the AD Stage, the defendant wishes to challenge the issue of causation. Specifically, the defendant wishes to challenge whether injuries B and C were caused by his breach of duty. Can the defendant do this?

According to the defendant, in a scenario where a claimant suffers injuries A, B, and C in an accident that was indisputably caused by a defendant, the parties can enter into a consent interlocutory judgment in favour of the claimant, without the claimant even proving that any injury was caused by the defendant’s breach of duty. By this approach, since the claimant has not established causation at the liability stage, the defendant would be able to challenge causation in toto at the AD Stage. For convenience, I will term this as the “Total Causation at AD Stage Approach”.

In contrast to the defendant’s approach, the approach taken in some cases is slightly different. Suppose again that a claimant suffers injuries A, B, and C in an accident that was indisputably caused by a defendant. The claimant can only establish that injury A was caused by the defendant’s breach of duty. By the approach advanced in some cases, if consent interlocutory judgment is entered in favour of the claimant, the defendant retains the right to challenge causation with respect to injuries B and C at the AD Stage. This is because the claimant can justify the consent interlocutory judgment for liability, having made out a cause of action in the tort of negligence on the basis of injury A alone. However, because the claimant has not established causation for injuries B and C, the defendant retains the right to challenge causation with respect to those injuries at the AD Stage. For convenience, I will term this as the “Partial Causation at AD Stage Approach”.

In my respectful view, neither of these approaches is correct. In the scenario that I have set out above, the correct solution is that the claimant can make out a cause of action in the tort of negligence so long as he can establish, among others, causation in relation to one of the injuries he has suffered. The claimant needs to do this to obtain an interlocutory judgment on liability so as to proceed to the AD Stage. This is because causation and damage are necessary elements to make out liability in the tort of negligence. At this point, that there is liability is not in doubt. The question is: liability for what? In this aspect, if the claimant manages to establish causation only in respect of, say, injury A, his cause of action and any resulting judgment for liability is limited to that injury. As such, the claimant would only be able to claim damages for injury A. It follows that the defendant would not be able to challenge causation at the AD Stage in respect of injuries B and C because it is not necessary for him to do so: the claimant has not even established liability with respect to those other injuries. For convenience, I will term this as the “No Causation at AD Stage Approach”. However, the defendant cannot challenge causation in respect of injury A at the AD Stage because the claimant has already established liability with respect to that injury and interlocutory judgment had been entered on that basis. The claimant also cannot bring a subsequent claim in respect of injuries B and C because all three injuries emanate from the same cause of action and the claimant must bring his action once and for all.

I will explain the reasons for my broad conclusions above. I will organise the discussion below along the following broad segments for ease of explanation: (a) the conceptual points in the tort of negligence, (b) why the Total Causation at AD Stage Approach and the Partial Causation at AD Stage Approach are wrong, whereas the No Causation at AD Approach Stage is correct, (c) a summary of my answers to the Questions posed, and (d) some consequential outcomes as a result of the answers. In discussing these broad points, I will also canvass the effect of a consent interlocutory judgment on liability at the appropriate points. This is because the effect of such a judgment is to restrict the issues that the defendant can challenge at the AD Stage, and this has a direct bearing on the correctness of the Approaches described above. I also take the opportunity to comment on some procedural aspects of an application made under O 33 r 2 of the ROC 2014 which, for brevity, I will hereafter refer to as “O 33 r 2”.

Because of the potential consequences of my answers to the Questions, as well as the fact that the plaintiff and defendant have advanced a unified view on what the answers should be, I invited Ms Leo to provide an independent (and potentially competing) opinion as a young independent counsel. I wish to state at the outset that Ms Leo’s oral and written submissions have been exceptional and greatly assisted me in the present case. I am most grateful for her able assistance. I shall have occasion to refer to her submissions in due course.

Procedural aspects of an application made under O 33 r 2

The interaction between s 54B(1) of the State Courts Act and O 33 r 2 of the ROC 2014

Before I turn to the Questions, I first address some procedural issues that have arisen from the manner in which the Questions have come to me. I begin with the interaction between s 54B(1) of the SCA (“s 54B(1)”) and O 33 r 2. The basis for the defendant’s application to transfer MC 8815 to the High Court under OA 301 was s 54B(1), which provides as follows:

From a plain reading of the provision, s 54B(1) contemplates that proceedings can be transferred from the State Courts to be tried in the High Court if the case, among others, involves some important question of law or is a test case. It does not, however, contemplate a transfer of such proceedings to the High Court only for the court to answer those questions (as questions or issues arising in a cause or matter) without trying the proceedings. The defendant recognised this lacuna and therefore did two things procedurally. First, so as to place the Questions squarely before the High Court, the defendant took out the present application under O 33 r 2 for the High Court to determine the Questions. This is because the mere fact of transfer to the High Court does not place the Questions before the High Court for preliminary determination. Second, so that the High Court does not need to carry out the AD after determining the Questions, the defendant had also included a second prayer in OA 301 for MC 8815 to be transferred from the High Court back to the State Courts for the AD to continue after the determination of the Questions. The learned AR hearing OA 301 made no order as to this second prayer after the defendant indicated that he would make an application later under s 54C of the SCA to transfer MC 8815 back to the State Courts.

While what the defendant has done is not impermissible, it does bring into question how s 54B(1) interacts with O 33 r 2 when the High Court is being asked to preliminarily determine questions that form the basis for the transfer in the first place. This question arises because the very reason for the transfer under s 54B(1) is that there are important questions of law for the High Court to decide. Yet, when the defendant invokes O 33 r 2 of the ROC 2014 to ask the High Court to preliminarily determine those questions, it would appear that the High Court is left with little choice but to determine those questions since the transfer was premised on the High Court having to answer those questions in the first place.

In my view, as the learned AR pointed out in OA 301, this issue has arisen because there is no express procedure for the High Court to determine such questions of law whilst still retaining the rest of the proceedings in the State Courts. So long as such a statutory lacuna persists, it may be preferable for the same judge to hear the transfer application under s 54B(1), as well as to decide whether to allow for the preliminary determination of questions under O 33 r 2. This would avoid a situation where the judge who eventually hears the transferred matter disagrees that there is an important question of law to be determined. In that instance, the judge may transfer the case back to the State Courts, returning to square one, so to speak. However, this is likely to be a rare case as in most cases, such as the High Court decision of Lee Chye Chong and others v SBS Transit Ltd [2021] 5 SLR 821, the important questions of law would be answered in the course of the trial of the matter being transferred.

The manner in which an application under O 33 r 2 should be made

In the present case, no such problem arises because I respectfully agree with the learned AR that there are indeed important questions of law to be determined. Put in the terms of the application before me, I am of the view that the Questions should be preliminarily determined pursuant to O 33 r 2. While the parties do not address me on this, I take this opportunity to make some observations on the manner in which an application under O 33 r 2 for the preliminary determination of questions should be made.

In my view, an application for a question to be preliminarily determined under O 33 r 2 should proceed in two stages. First, before any question which is the subject of the application can be preliminarily determined, the logically prior question is whether the court should grant permission for the question to be determined in the first place. Second, if the court grants permission, it will proceed to the second stage to consider the merits of the question submitted for preliminary determination. Whether the two stages should proceed in one hearing, or in separate hearings, would depend on the directions given by the court. Therefore, to enable the court to properly manage such applications, it would be helpful for the applicant to make it clear in its application whether it is seeking only the court’s permission for a question to be preliminarily determined, or if it is seeking the determination of that question as well in the same hearing. Indeed, all too often, applications for the preliminary determination of a question under O 33 r 2 are framed ambiguously and it is not clear if the applicant is seeking permission only or the further determination of the question at the same hearing.

I turn then to the first stage of the application of O 33 r 2, which is whether I should grant permission for the Questions to be determined. In this regard, O 33 r 2 provides as follows:

To begin with, O 33 r 2 confers upon the court the power to order any question or issue to be determined preliminarily (see Singapore Civil Procedure 2021 vol 1 (Cavinder Bull gen ed) (Sweet & Maxwell, 2021) (“White Book”) at para 33/2/1). In deciding whether to exercise this power, it is well-established that the court would consider whether a preliminary determination would save substantial time and expenditure (see the Court of Appeal decision ACB v Thomson Medical Pte Ltd and others [2017] 1 SLR 918 (“ACB”) at [22]). However, this consideration should be regarded as one, albeit important, factor that forms part of the broader inquiry of whether a preliminary determination would allow for fuller treatment of the issues and allow for a more just outcome.

It is clear from the above extract from ACB that the court was ultimately concerned about whether the preliminary determination of the questions there would achieve a just outcome in the case. In my view, this suggests that the overarching inquiry in deciding an application under O 33 r 2 is whether a preliminary determination would be in the interests of justice. To this, I would add that considerations of justice need not be limited to those concerning the parties in the dispute. Indeed, there is nothing in the wording of O 33 r 2 which limits the considerations that the court may take into account. In some cases, it may be in the interests of justice to make a preliminary determination where it would benefit third parties. This is likely to be so if questions of law of public importance are involved or if the dispute in question is a “test case” that would provide guidance to other similarly situated litigants.

Applying these principles to the present case, I am of the view that a preliminary determination of the Questions would be in the interests of justice. I agree with the defendant that the determination of the Questions would affect more than the immediate interest of the parties here and would benefit future litigants, insurance companies, and their legal advisors in planning their litigation roadmap and strategies in relation to causation. And significantly, the plaintiff in the present case does not object to the application. Thus, in my judgment, it is appropriate for the Questions to be preliminary determined pursuant to O 33 r 2.

Conceptual points in the tort of negligence

I turn now to the Questions. But before I answer them, it would be helpful for me to go through some conceptual points on the tort of negligence.

A “cause of action”

I begin with the definition of a “cause of action”. This is a useful starting point as the material holding in Tan Woo Thian (at [6]) is that “[a] cause of action in negligence is inchoate absent evidence of actual loss” [emphasis added]. In my view, the fulfilment of the constituent elements of the tort of negligence being framed as establishing a “cause of action” is simply another way of saying that “liability” in the tort has been established. This gives effect to the trite principle that before a defendant can be held liable for the claimant’s damage, the claimant would have to establish a cause of action in the tort of negligence against him. In this regard, Black’s Law Dictionary (Bryan A Garner ed) (Thomson Reuters, 11th Ed, 2019) defines a cause of action in the following terms (at p 275):

Indeed, the High Court in Zhang Run Zi v Koh Kim Seng and another [2015] SGHC 175, citing Multi-Pak Singapore Pte Ltd (in receivership) v Intraco Ltd and others [1992] 2 SLR(R) 382, held (at [42]) that a cause of action could also refer to “the facts which the plaintiff must prove in order to get a decision in his favour” (see also White Book at para 15/1/3).

Similarly, in Andrew McGee, Limitation Periods (Sweet & Maxwell, 8th Ed, 2018) (“Limitation Periods”), the learned author notes (at para 5.008) that, in relation to single torts that require proof of damage (of which the tort of negligence is an example), the basic rule is that the cause of action accrues when damage is suffered. This basic rule flows from the principle that the cause of action is complete only when there is a claimant who can sue and a defendant who can be sued, and when the ingredients of duty, breach, and damage are all satisfied (see the decision of the English Court of Appeal in Coburn v Colledge [1897] 1 QB 702).

Most of the difficulties that have arisen in relation to this definition of a “cause of action” concern whether the necessary elements have been fulfilled. These difficulties are exacerbated by the rule that generally only one action may be brought in respect of any cause of action (see, eg, Henderson v Henderson (1843) 3 Hare 100). Thus, in the English Court of Appeal cases on economic loss, such as Forster v Outred & Co [1982] 1 WLR 86 (“Forster”) and D W Moore & Co Ltd v Ferrier [1988] 1 WLR 267 (“D W Moore”), where the same breach of duty gave rise to an immediate diminution of the plaintiff’s estate, and the risk of further diminution (which later materialised), the courts have consistently held that there is only one cause of action. However, these cases may be contrasted with the first instance decision of Post Office v Official Solicitor [1951] 1 All ER 522 (“Post Office”). In Post Office, a postman was injured by the defendant’s negligent driving. The Post Office sued the driver successfully for the loss of the postman’s services. The Post Office also gave the postman a disability pension, which it later sought to sue the defendant’s estate for (at least in respect of the amounts that were paid out). Barry J held that this latter claim was a separate cause of action, distinct from the claim relating to the loss of the postman’s services. As such, the Post Office was not precluded from bringing the latter action to recover the disability pension.

In Limitation Periods (at para 1.017), McGee has opined that Post Office was incorrectly decided. In his view, there was only one breach of duty on the facts of Post Office even if the damage had occurred in stages rather than all at once. He argues that “cause of action” must be regarded as covering all the relief that can be claimed by any one claimant in respect of any one breach of duty by the defendant. However, where a case involved a continuing breach of duty, it may be possible to accept that a fresh cause of action accrued every day. I respectfully agree with McGee’s view. Indeed, such a view is more consistent with the appellate decisions of Forster and D W Moore. Practically, when transposed to the tort of negligence, this means that all the elements of the tort, including damage, must be established so as to constitute the cause of action. I will return to this point subsequently.

The elements of a cause of action in the tort of negligence

With the definition of a “cause of action” in mind, I turn to the elements of a cause of action in the tort of negligence. In this regard, Menon CJ held in Tan Woo Thian (at [6]) that “[a] cause of action in negligence is inchoate absent evidence of actual loss”. Indeed, it is trite law that in order for a defendant to be liable in the tort of negligence, four elements have to be fulfilled: (a) the defendant owed a duty of care to the claimant, (b) the defendant acted in breach of that duty, (c) there was a causal connection between the defendant’s breach and the claimant’s damage, and (d) that particular kind of damage suffered by the particular claimant is not so unforeseeable as to be too remote (see the Court of Appeal decision of Spandeck Engineering (S) Pte Ltd v Defence Science & Technology Agency [2007] 4 SLR(R) 100 at [21], citing an earlier edition of Clerk & Lindsell on Torts (Sweet & Maxwell, 23rd Ed, 2022) (“Clerk & Lindsell”) at para 7-04, as well as Gary Chan, The Law of Torts in Singapore (Academy Publishing, 2nd Ed, 2016) at para 07.001). The elements are no different when the negligence has occurred in a PIMA case (see the recent High Court decision of CXN (a minor suing by her father and litigation representative) v CXO and another [2022] SGHC 311 (“CXN”) at [12]). Indeed, Clerk & Lindsell further elaborates in the same paragraph that when these four elements are fulfilled, the defendant is liable in negligence. It is only then that it becomes relevant to consider the assessment of damages, ie, the compensation for the loss for which the defendant is liable.

In contrast to these points above, counsel for the defendant, Mr Tan Seng Chew Richard (“Mr Tan”), argued before me that questions of damage, causation, and remoteness only go towards the assessment of damages and not liability. In support of his position, he relied on the Court of Appeal decision in Ngiam Kong Seng and another v Lim Chiew Hock [2008] 3 SLR(R) 674 (“Ngiam Kong Seng”) (at [146]), where the court commented that “[i]n so far as the issue of liability is concerned, the plaintiff is only required to establish that a duty of care was owed and that a breach of such duty occurred. The other requirements listed above are, strictly speaking, concerned with the assessment of damages and not with liability…” [emphasis in original]. In other words, Mr Tan’s position is that, based on the authority of Ngiam Kong Seng, it is sufficient to establish the first two elements of duty and breach and leave questions of causation between the defendant’s breach and the claimant’s damage, as well as questions of remoteness, to the AD Stage.

I disagree with Mr Tan’s submission. In my view, it is clear that proof of damage, subject to the rules of causation and remoteness, is a necessary element to establish liability for the tort of negligence. This is clear in light of the authorities discussed above (see [35] above). More crucially, I do not think that the Court of Appeal in Ngiam Kong Seng intended to make a definitive pronouncement on whether such elements went towards liability, given that the court there was primarily concerned with how the presence of liability for negligence in cases of psychiatric harm should be analysed. In any event, any uncertainty in Ngiam Kong Seng is now resolved by the express pronouncement in Tan Woo Thian reiterating that causation and damage are necessary elements to establish liability in the tort of negligence. For these reasons, it is clear that the elements for establishing liability in the tort of negligence follow those which were set out in Spandeck and reiterated in Tan Woo Thian.

The element of causation

Having established that a cause of action in the tort of negligence comprises the four elements set out above, I come then to the element of causation, which is central in the Questions posed to me. At the outset, the defendant attempts to draw a distinction between “responsibility for an accident and responsibility for personal injuries”. As I understand it, the defendant’s argument is that a defendant may admit “responsibility” (ie, liability) for having caused the accident, enter interlocutory judgment on that basis, but challenge his liability for the claimant’s injuries resulting from the accident at the AD Stage.

I disagree with the defendant’s argument because he has misconstrued the causation element in the tort of negligence to be whether the defendant caused the accident, rather than whether the defendant’s breach caused the claimant’s injuries. Indeed, as Viscount Simonds put it in Overseas Tankship (UK) Ltd v Morts Dock and Engineering Co Ltd, The Wagon Mound [1961] 2 WLR 126 (at 141), it is not the negligent act itself but the consequences of the negligent act on which tortious liability is founded. As such, in CXN, in the context of discussing the elements of the tort of negligence in a PIMA case, Teh Hwee Hwee JC referred to the “causal connection between the collision and the plaintiff’s injuries” (at [13]). The learned judge was clearly concerned with whether the accident caused the claimant’s injuries, rather than merely whether the defendant caused the accident. Thus, in so far as the court in Arunachalam Balasubramanian v Lion City Rentals Pte Ltd and another [2020] SGMC 33 (“Arunachalam”) suggested (at [23]) that the relevant causal connection is that “between the negligent act and the accident”, I would respectfully disagree.

In my view, it is more accurate to say that the relevant causal connection to establish liability is that between the negligent act and the alleged damage. Fundamentally, this follows from causation being an essential element of liability in negligence. As the learned authors of Clerk & Lindsell explain (at para 1-15), in a system of corrective justice, a sufficient causal link between the defendant’s conduct and the claimant’s damage is required to justify compensation. Being an action on the case, negligence is consummated only upon the occurrence of damage. Indeed, without damage, there can be no claim against a person whose conduct may well be deficient in some significant manner (see Christian Witting, “Physical Damage in Negligence” (2002) 61(1) CLJ 189). To prove causation, the claimant would have to adduce factual evidence to show that there was a physical connection between the defendant’s breach and his damage (factual causation) and that the defendant’s breach was an operative or effective cause of his damage (legal causation) (see Clerk & Lindsell at paras 2-01 to 2-05).

Additionally, by virtue of the symbiotic relationship between causation and damage, causation is also related to damages, which generally quantifies, in monetary terms, the extent of the damage caused to the claimant, or the loss suffered by the claimant. It would follow that causation, as an element of liability, “must be established before the question of quantification of damages arises is obvious” (see the decision of the Court of Appeal in Salcon Ltd v United Cement Pte Ltd [2004] 4 SLR(R) 353 (“Salcon”) at [24]). This is why Menon CJ in Tan Woo Thian rejected the appellant’s argument that since the trial was bifurcated, he would only need to adduce evidence relating to the types of losses caused by the respondent’s alleged breaches only at the AD Stage. Menon CJ clarified (at [7]) that this argument “wrongly conflates the separate questions of whether the appellant is able to establish that the respondent’s breach has caused loss, with the quantum of that loss” [emphasis in original]. This is because, in order to make out the tort of negligence, the appellant must show that the respondent’s alleged breach has in fact caused damage. As such, while the appellant was not obliged (at [8]) “to adduce evidence at the liability stage of the trial as to the quantification of the losses and injuries he claims he suffered”, he would “nonetheless have been obliged to show that he did, in fact, suffer one or more types of loss that was causally connected to the alleged breach” [emphasis in original].

The distinction between damage, loss, and damages

As such, it is clear from the foregoing that damage and causation are essential elements of the tort of negligence. However, I pause to observe that, when speaking of the elements of negligence, the terms “damage”, “loss”, and “damages” are often used interchangeably. Therefore, at this juncture and as a preliminary point, it is important to distinguish between the different terms carefully so that a consistent set of terminology can be used in the future.

The concept of “damage”

In explaining the concept of “damage”, I begin with the concept of a “wrong”. A wrong is a breach of legal duty (see Peter Birks, “The Concept of a Civil Wrong” in Philosophical Foundations of Tort Law (David G Owen ed) (Oxford University Press, 1997) at p 37). This legal duty, in the context of the tort of negligence, is not a duty to act carefully but a duty not to inflict damage carelessly on someone else. As a corollary to the idea that a wrong is committed against someone, “intrinsic to the notion of damage in negligence is the idea that the damage is suffered by someone” [emphasis in original] (see Donal Nolan, “Rights, Damage and Loss” (2017) 37 OJLS 255 at 260). Therefore, it has been said that damage involves the intangible notion that there has been an interference with a legally protected interest of the plaintiff (see ACB at [44], citing the decision of the High Court of Australia in Cattanach and another v Melchior and another (2003) 199 ALR 131 at [23]). By way of example, if a claimant’s car is dented as a result of the defendant’s negligent conduct, it is not the fact of the car being worse off that establishes “damage”; rather, it is the fact that the claimant had, for instance, legal ownership of or a possessory title to the car at the time the defendant dented it (see Winfield & Jolowicz on Tort (James Goudkamp and Donal Nolan gen eds) (Sweet & Maxwell, 20th Ed, 2020) (“Winfield & Jolowicz”) at para 7-006).

However, this understanding of damage has not made the task of formulating a test for damage any easier. Indeed, to attempt to exhaustively define damage would be a difficult task (see ACB at [45]). For instance, Lord Hoffmann sought to define damage as “an abstract concept of being worse off, physically or economically” in the House of Lords decision of Rothwell v Chemical & Insulating Co Ltd [2008] 1 AC 281 at [7]. One problem with this definition, as observed by the Court of Appeal in ACB at [45], is that it restricts harms to the physical or economic, leaving out categories of damage such as psychiatric harm. Yet, even if one removes these restrictions, it is also not accurate to simply define damage as an abstract concept of being “worse off”. On the one hand, it would be over-inclusive because not all forms of being worse off would amount to damage, such as where a claimant suffers from mental distress that falls short of a recognised psychiatric illness. On the other hand, such a definition would be under-inclusive as there are instances where a claimant is able to establish damage and complete his cause of action even where he was not worse off. One common example given by academic commentators is as follows (see Donal Nolan, “Rights, Damage and Loss” at 271):

Rather than formulate an exhaustive definition of damage, I agree with the authors of Winfield & Jolowicz (at para 7-008) that the prevailing approach is to consider whether the particular harm suffered by a claimant is actionable or not actionable in negligence. This refers to the requirement of actionable damage, which the Court of Appeal accepted in ACB at [47]. As Professor Jane Stapleton explained in “The Gist of Negligence: Part 1 Minimal Actionable Damage” (1988) 104 LQR 213, actionable damage forms the “gist of the action” which gives rise to a cause of action in the tort of negligence. This means that the harm suffered by the claimant must be one that the law recognises as being capable of compensation. Accordingly, by this analysis, the characterisation of the harm suffered by the claimant as “damage” expresses a conclusion that liability should attach in respect of the harm (see Donal Nolan, “Damage in the English Law of Negligence” (2013) 4 JETL 259 at 267).

The concept of “loss”

Having considered the concept of “damage”, I now come to the related concept of “loss”. In this regard, to suffer a “loss” is to be factually worse off (see Donal Nolan, “Rights, Damage and Loss” at 256; Robert Stevens, Torts and Rights (Oxford University Press, 2007) at p 43). In cases of personal injury, it is not difficult to see how damage and loss will usually co-exist. For instance, if the claimant’s leg is broken as a result of the defendant’s negligent driving, he has suffered damage because his interest in bodily integrity has been interfered with, and he has also suffered loss because he is now worse off with a broken leg. It bears emphasis, however, that damage and loss are distinct concepts. As I explained earlier (see [44] above), it is possible for a claimant to have suffered damage and complete his cause of action even though he has suffered no loss in the sense of being worse off. Therefore, it is conceptually inaccurate to speak of “loss”, instead of “damage”, as being an element of the tort of negligence.

Finally, the law also recognises that an award of damages does not necessarily have to be tied to “loss”. This is because torts are civil wrongs which “themselves dictate no fixed measure of response” (see Peter Birks, “The Concept of a Civil Wrong” at p 51; Clerk & Lindsell at para 1-02). In this regard, I would respectfully set out the view in James Edelman, McGregor on Damages (Sweet & Maxwell, 21st Ed, 2021) (“McGregor on Damages”) at para 1-023:

Indeed, similar sentiments were expressed by the Court of Appeal in ACB where it recognised that “punitive damages may be awarded in tort where the totality of the defendant’s conduct is so outrageous that it warrants punishment, deterrence, and condemnation” (at [176]). This shows that, in Singapore, compensation for loss would not always be the sole aim of an award of damages in tort. Nevertheless, I would caveat that it is only in exceptional situations that damages in tort are awarded for purposes apart from compensation and that, generally, compensation for loss remains the usual aim in an award of damages.

The relationship between “damages” and “damage/loss”

Related to but distinct from the concepts of “damage” and “loss” is the concept of “damages”, which refers to a monetary award for a civil wrong (see McGregor on Damages at para 1-001). The concept of “damages” is related to the concept of “damage” because no damages can be awarded in the absence of a wrong (see the House of Lords decision of Bourhill v Young [1943] AC 92 at 106). And since the wrongfulness of negligence is tied to the existence of damage, it follows that no damages can be awarded without proof of damage.

However, while these concepts are related, they must not be conflated. Indeed, Menon CJ alluded to the difference between the concepts of “damage” and “damages” in Tan Woo Thian (at [7]) when he said that the appellant had wrongly conflated “the separate questions of whether the appellant is able to establish the respondent’s breach has caused loss, with the quantum of that loss” [emphasis in original]. The importance of this distinction has also been clearly expressed by the Court of Appeal in ACB (at [44]):

The concept of damages is also related to the concept of loss because the usual aim of damages is to eradicate, through a monetary award, the consequences that fall within the scope of duty breached (see McGregor on Damages at para 1-001). Indeed, the court would usually award “that sum of money which will put the party who has been injured, or who has suffered loss, in the same position as he would have been in if he had not sustained the wrong for which he is now getting his compensation or reparation” (see the decision of the Court of Appeal decision in Minichit Bunhom v Jazali bin Kastari and another [2018] 1 SLR 1037 at [30] and the House of Lords decision of Livingstone v Rawyards Coal Co (1880) 5 App Cas 25 at 39). In other words, damages are generally aimed at compensating the loss suffered by the claimant in so far as it can be quantified in monetary terms.

Summary of terminology

At this stage, it is helpful to summarise the terminology that should be used. Without intending to be exhaustive, the following describe the relationship between the concepts of damage, loss, and damages in the context of the tort of negligence:

(a) damage is an essential element of the tort of negligence;

(b) however, damage that is capable of fulfilling the elements of the tort of negligence needs to be of a type that is actionable or recognised in law. For most cases, this is not an issue because the claimant would have suffered clearly observable physical damage to his person (ie, injuries, which I will use interchangeably with “damage” in this judgment) or property;

(c) a claimant who has suffered damage is able to claim damages, which is the monetary sum payable that is usually aimed at compensating the claimant for being made worse off by the defendant’s negligence;

(d) generally, in order for the claimant to claim damages, he must prove that he has suffered loss from the damage, in the sense that he has been made worse off as a result of the damage caused by the defendant’s negligence; and

(e) accordingly, it is possible for a claimant to have suffered damage caused by the defendant’s negligence, but for the defendant not to pay the claimant any damages because the claimant has suffered no loss due to the damage.

With these conceptual points on the tort of negligence in mind, I turn to explain the correct approach to take in relation to causation at the AD Stage.

Why the Total Causation at AD Stage Approach is wrong

I turn first to the Total Causation at AD Stage Approach, which is the approach argued for by the defendant. By this approach, even though the claimant may not have proved causation in relation to the injuries (or damage) he has suffered, the parties can enter interlocutory judgment by consent and reserve causation in toto to the AD Stage.

The Total Causation at AD Stage Approach in practice

Similarly, in the Magistrate’s Court decision of Eliora Yow (an infant suing by her father and litigation representative, Yow Tuck Meng Jerry) v Kwa Kian Peng [2020] SGMC 44 (“Eliora Yow”), the learned Deputy Registrar (“the DR”) agreed with the defendant’s submission that the causation of the claimant’s injuries in that case remained a live issue at the AD Stage (at [10]). The DR cited Muhammad Shaun Eric bin Abdullah alias De Silva Shaun Eric v Ng Ah Tee (Chua Seng Thye, Third Party) [2004] SGHC 268 (“Muhammad Shaun Eric”) in support of his conclusion.

In Muhammad Shaun Eric, the claimant was involved in a road traffic accident with the defendant and the third party. By consent, interlocutory judgment was entered against the defendant with the claimant agreeing to bear 10% of the liability and the third party agreeing to indemnify the defendant for 15% of the damages that the defendant had to pay the claimant. The parties did not reserve the issue of causation of injuries to the assessing registrar in the interlocutory judgment. When the matter went for assessment, the learned Assistant Registrar found that save for one injury, the claimant had failed to discharge his burden of proving that his injuries were caused by the accident. The High Court affirmed the decision of the Assistant Registrar in Muhammad Shaun Eric bin Abdullah alias De Silva Shaun Eric v Ng Ah Tee (Chua Seng Thye, Third Party) [2005] SGHC 180 (“Muhammad Shaun Eric (on appeal)”). Crucially for present purposes, the High Court regarded that causation remained a live issue at the AD Stage when it said (at [5]) that “[t]he main issue in the appeal, as it was in the assessment hearing, was that of causation” [emphasis added]. As such, it may be that there has been a High Court decision on how to deal with causation, which is binding on practitioners and the lower courts.

The Total Causation at AD Stage Approach is wrong

In my view, the Total Causation at AD Stage Approach is wrong as this approach is conceptually unsound and goes against the key principles of law reiterated in Tan Woo Thian.

An interlocutory judgment prevents the reservation of causation in toto to the AD Stage

To begin with, in a bifurcated trial, parties would be required to enter interlocutory judgment for liability before proceeding to the AD Stage. While such interlocutory judgments in the context of PIMA cases would usually be by consent, it is not material whether this is the case. Since causation is an essential element in establishing liability (or a cause of action) in the tort of negligence, then, as Ms Leo points out, once interlocutory judgment is entered, the defendant would not be entitled to challenge liability (and hence causation) in toto during the AD Stage.

In this regard, the court’s power to enter interlocutory judgment is governed by, among others, O 13 r 2 of the ROC 2014, which provides:

In this context, an interlocutory judgment is defined as a judgment that is “interlocutory only as to the amount” but is “final as to the right of the plaintiff to recover damages and costs” [emphasis added] (see White Book at para 13/2/1). Ordinarily, as the court in Arunachalam held (at [25]), the ambit of an interlocutory judgment would have to be construed together with the underlying Statement of Claim. However, in the context of an interlocutory judgment entered so that the parties can proceed to the AD Stage, it must be the case that the interlocutory judgment has fully determined the defendant’s liability. This is because the entire premise of damages being assessed is that liability for those damages has already been established by the interlocutory judgment.

Thus, in the Privy Council decision of Strachan v The Gleaner Co Ltd and another [2005] 1 WLR 3204, Lord Millett had said this (at 3209):

I would only repeat the final sentence, which is that the claimant obtains his right to claim damages from the judgment on liability. It must therefore follow that that judgment has determined liability fully between the parties.

As such, I agree with Ms Leo that the final nature of the claimant’s right to recover damages pursuant to the interlocutory judgment on liability means that liability cannot be challenged in its entirety at the AD Stage. Indeed, the Court of Appeal held in U Myo Nyunt (alias Michael Nyunt) v First Property Holdings Pte Ltd [2021] 2 SLR 816 (“U Myo Nyunt”) that after a judgment in default under O 13 r 2 of the ROC 2014 is entered with damages to be assessed, the defendant is not entitled to dispute liability at the AD hearing (at [47]). It would follow that after interlocutory judgment is entered pursuant to the same O 13 r 2, the defendant should similarly not be allowed to challenge liability at the AD Stage. Consequently, causation cannot be challenged or re-opened in toto at the AD Stage after interlocutory judgment is entered, as doing so would potentially lead to the inconsistent consequence that the defendant was never liable in the first place.

I also agree with Ms Leo’s submission that the approach taken in U Myo Nyunt serves an important policy objective of guarding against the risk of inconsistent judgments. If both an interlocutory judgment and final judgment make different pronouncements on the extent of a defendant’s liability, this would give rise to significant difficulties for a party wishing to appeal the outcome of either judgment. As such, it is important that the distinct issues of liability and damages be kept separate to avoid the possibility of liability being dealt with twice over in the interlocutory judgment and the final judgment.

Quite apart from the effect of an interlocutory judgment in preventing a defendant from reopening issues of liability (including causation) at the AD Stage, the purpose of the AD Stage is not to determine issues of liability. This is simply looking at the situation from the perspective of the AD Stage. In this regard, the AD Stage, as its name suggests, is meant for damages to be assessed. Thus, as the Court of Appeal in ACB pointed out (at [44]), when the issue of quantum of damages is engaged, it must mean that prior questions of liability, including causation, would already have been settled. Thus, the very premise of the AD Stage makes it clear that issues of liability, including causation, cannot be “reserved” in toto to the AD Stage.

The court has no power to “reserve” causation in toto to the AD Stage under O 33 of the ROC 2014

For completeness, I deal with the defendant’s argument that the court retains the power to “reserve” all issues of causation for determination at the AD Stage under O 33 of the ROC 2014. The defendant contends that O 33 r 2 permits “different questions and issues to be determined by different modes of trial, including… the mode of assessment of damages by the registrar under Order 37”. As such, the defendant says that there is no conceptual hurdle to the mode being utilised to determine all issues relating to the element of damages, if these are reserved or otherwise not canvassed before the trial judge during the liability stage. The outcome of the defendant’s submission is that, as long as the issue of causation of a claimant’s injuries was not addressed at the mode of trial, it remains open for causation to be determined at the AD Stage.

I disagree with the defendant’s argument. In the first place, it is conceptually unsound for the reasons I have given above for issues of liability, including causation, to remain open to challenge at the AD Stage. I do not think that the provisions of the ROC 2014 should be construed in such a way as to result in conceptual incoherence.

Second, I agree with Ms Leo that the defendant’s argument rests on an incorrect interpretation of O 33 of the ROC 2014, which I set out as follows:

In my view, the defendant is not correct in describing “assessment of damages” as a “mode of trial” under O 33 of the ROC 2014. Instead, I agree with Ms Leo that the different “modes of trial”, based on a plain reading of O 33 r 1, refers to the manner in which the trial is conducted, whether by a presiding judge or registrar and with or without the assistance of assessors. By this interpretation, the different “modes of trial” do not refer to the difference between the questions of “liability” and questions of “assessment of damages”. Therefore, the correct interpretation of O 33 is that questions of “liability” or “assessment of damages” are merely “issues” that can be tried by different “modes of trial”. Thus, there is nothing in O 33 that would permit the court to reserve issues of causation in toto to the AD Stage.

Reconsideration of prior cases adopting the Total Causation at AD Stage Approach

I would therefore respectfully suggest that authorities which suggested that it was possible to reserve causation in toto to the AD Stage (ie, the Total Causation at AD Stage Approach) be reconsidered. For example, I respectfully disagree with the learned DR’s view in Eliora Yow (at [9]–[10]) that interlocutory judgment on liability deals only “with the defendant’s liability in respect of causing the accident and does not deal conclusively with the causation of the injuries suffered by the plaintiff” [emphasis in original]. This, with respect, mistakes the causation element in the tort of negligence to be concerned with the defendant’s act as opposed to whether that act caused any actionable damage. Instead, I agree with Ms Leo that an interlocutory judgment on liability cannot stand if the defendant did not cause the claimant any damage. Simply causing an accident, without causing any injury or damage, would not be sufficient to render a defendant liable in the tort of negligence.

Further, while a number of High Court decisions such as Muhammad Shaun Eric (on appeal), Lee Mui Yeng v Ng Tong Yoo [2016] SGHC 46 (“Lee Mui Yeng”) and Noor Azlin bte Abdul Rahman and another v Changi General Hospital Pte Ltd and others [2021] SGHC 10 (“Noor Azlin”) seem to suggest that causation can still be challenged in toto at the AD Stage, I would respectfully point out that these cases were decided before Tan Woo Thian. As such, it may well be that, just like in the State Courts, the High Court was not presented with arguments as to the effect of Tan Woo Thian and therefore continued to regard causation as a live issue at the AD Stage. In my view, for the reasons I have explained above, a plain application of Tan Woo Thian would preclude the consideration of causation in toto at the AD Stage.

Why the Partial Causation at AD Stage Approach is wrong

The Partial Causation at AD Stage Approach in practice

After Tan Woo Thian was decided, one approach taken by some judges seems to recognise that Tan Woo Thian has made it impossible to challenge causation in toto at the AD Stage, but that prior authorities still make it possible to dispute causation as to certain parts of the alleged damage at the AD Stage. This approach, which I have termed the “Partial Causation at AD Stage Approach”, can be found in DR Tan’s comprehensive and clearly reasoned decision of Lim Mei Choo (Lin Meizhu) v Muhammad Azham bin Razak (Direct Asia Insurance (Singapore) Pte Ltd, intervener) [2021] SGMC 74 (“Lim Mei Choo”). Having surveyed the authorities extensively, DR Tan concluded as follows (at [20]):

Thus, DR Tan accepted that the fundamental proposition laid down in Tan Woo Thian (ie, it is only after a claimant has successfully established liability by, among others, establishing causation, that the assessment of damages come into consideration) means that it is no longer possible for a defendant to dispute causation entirely at the AD Stage. However, in view of a number of binding High Court decisions which show that “causation can reasonably be disputed at the assessment of damages phase as long as at least some damage or loss was caused by the defendant’s negligence” [emphasis in original], DR Tan held that a defendant may continue to dispute causation “as to certain (but not all) heads of claims at the AD Stage” (see Lim Mei Choo at [20]–[21]). This therefore leaves open the possibility of challenging causation to some extent at the AD Stage.

A similar approach was taken by the learned Deputy Registrar Hairul Hakkim (“DR Hakkim”) in Krishnamoorthy s/o Chellappan v Ramasamy Arivazhagan [2021] SGDC 283. After agreeing with the approach taken by DR Tan in Lim Mei Choo, DR Hakkim proceeded to observe as follows (at [12]–[13]):

Like DR Tan in Lim Mei Choo, DR Hakkim also relied on prior authority that seemingly allowed causation to be disputed to some extent at the AD Stage. In particular, DR Hakkim closely interpreted one paragraph in Tan Woo Thian as standing for the proposition that a plaintiff need not prove causation for all the heads of losses claimed; it suffices that he is able to prove causation for some of those heads. The relevant paragraph from Tan Woo Thian is [11], which I reproduce for ease of explication:

As such, DR Hakkim relied on Menon CJ’s use of the word “any” in this paragraph to infer that if the appellant in Tan Woo Thian had been able to prove causation in respect to some of his losses, then he would have crossed the causation hurdle and been able to proceed to the AD Stage.

In summary, the approach taken by DR Tan and DR Hakkim preserves the ability to challenge causation to some extent at the AD Stage provided that the claimant is able to prove causation in respect of some of his damage at the liability stage. According to the learned judges, this approach is permissible (and indeed, binding on them) because of: (a) the Court of Appeal decision of Tan Woo Thian at [11], and (b) a number of High Court decisions such as Muhammad Shaun Eric (on appeal), Lee Mui Yeng, as well as Noor Azlin. This approach is also in line with that taken by the English authorities. Thus, Martin Spencer J in the English High Court decision of Justyna Zeromska-Smith v United Lincolnshire Hospitals NHS Trust [2019] EWHC 630 (QB) said this (at [33]):

What appears on its face to be a differing approach, which focuses on the quantum of damages as opposed to causation, can be seen in at least two decisions from the learned District Judge Sheik Umar Bin Mohamed Bagushair (“DJ Sheik Umar”). In Fobrogo Loreen Vera Mrs Sandosham Fobrogo Loreen Vera v MCST Plan No 1614 [2021] SGMC 75 (“Fobrogo”), the plaintiff sued the defendant for the injury and losses she suffered when she fell while on an escalator at the Adelphi Shopping Centre. The defendant was the manager of the premises. Among others, the defendant argued that the accident did not cause the plaintiff’s injuries (at [57]). In particular, the defendant argued that the subsequent injuries recorded in a medical report (such as a fracture) were only diagnosed when the plaintiff went for further appointments, and it could not be said that these subsequent injuries were caused by the accident (at [63]).

DJ Sheik Umar dismissed the defendant’s arguments against causation having been established. Noting that the trial had been bifurcated, the learned DJ held that, for the purposes of the trial on liability and in accordance with Tan Woo Thian, he did “not need to find that all of the [p]laintiff’s damages were caused by the Accident” [emphasis in original] (at [60]). Rather, the learned DJ said that “in order to establish a cause of action in negligence, it must first be shown that some loss had been caused, though the quantum of the loss can be assessed at a later stage” [emphasis in original] (at [60]). Later, in two explanatory paragraphs, the learned DJ said this (at [64]–[65]):

Accordingly, while framed in terms of damages, DJ Sheik Umar’s approach is tied to causation as it was concerned with the existence of actionable damage, namely, the fracture. His approach is therefore the same as the approach taken by DR Tan and DR Hakkim. This is because while the learned DJ said that a defendant cannot argue that the plaintiff’s damages should be zero, I understand that to refer to the global quantum of damages potentially due to the claimant. I do not think that the learned DJ meant to say that a defendant cannot argue, to take the example of the fracture in Fobrogo, that the damages with respect to a particular injury should be zero. Indeed, the entire premise of DJ Sheik Umar’s approach is that the claimant had to prove that he had suffered some damage. Once the claimant has done this, it should follow that the global amount of damages would not ordinarily be zero. But the learned DJ seemed to suggest that a defendant can still challenge the causal relationship between the defendant’s breach of duty and the particular injury at the AD Stage. If the defendant succeeds in doing that, it must follow that the damages claimable in respect of that injury must be zero. This must therefore mean that, even at the AD Stage, a defendant can still challenge causation to some extent so long as the claimant has established liability by proving causation of some of his losses.

Indeed, in the District Court decision of Lim Ai Bee v Da-Cin Construction Co Ltd (Singapore Branch) and another [2021] SGDC 227 (“Lim Ai Bee”), which was decided just ten days before Fobrogo, DJ Sheik Umar articulated his approach in a similar way. In that case, the claimant sued the defendants for the injury and losses she suffered when she witnessed and heard aluminium bars falling around her as she was walking with her daughter out of her condominium’s lobby. The aluminium bars had dropped from a gondola being operated by the second defendant’s workers. The second defendant argued, among others, that the accident did not cause the full extent of the claimant’s injuries. This was because there was some evidence showing that she was suffering from a pre-existing condition. The second defendant also argued that some of the claimant’s health conditions were only diagnosed several years after the accident.

DJ Sheik Umar dismissed the second defendant’s argument. In particular, he held (at [188]) that he “do[es] not need to find that all of the [p]laintiff’s damages were caused by the Accident” [emphasis in original]. The learned DJ explained (at [188]) that, in line with Tan Woo Thian, in order to establish a cause of action in negligence, “it must first be shown that some loss had been caused, though the quantum of the loss can be assessed at a later stage” [emphasis in original]. On the facts, the learned DJ found that the claimant did suffer some loss as a result of the accident. Accordingly, he held that liability was established but that the defendants remained free to argue at the AD Stage what the claimant’s damages should be, except zero.

As with my analysis of Fobrogo, I do not understand DJ Sheik Umar to be applying a different approach from that of DR Tan and DR Hakkim. Indeed, I see the learned DJ’s approach in Lim Ai Bee to be ultimately premised on the claimant being able to prove causation as to some of his damage at the liability stage, which would allow the defendants to challenge causation as to other damage (and hence the quantum of damages) at the AD Stage. Thus, when the learned DJ said that “the quantum of loss can be assessed at a later stage”, I infer that he meant that the quantum of damages payable in respect of some of the damage can be disputed by challenging the underlying causal connection.

The Partial Causation at AD Stage Approach is wrong

In my respectful view, the Partial Causation at AD Stage Approach is wrong. As such, I respectfully disagree with the approach taken in the lower courts, as exemplified by the learned decisions of DJ Sheik Umar, DR Tan, and DR Hakkim. In this regard, I also disagree with the similar view taken by Ms Leo, who submitted that a defendant can raise issues of causation at the AD Stage in so far as it would not pertain to the injuries he had conceded liability for at the liability stage. I justify my conclusion based on principle, precedent, and policy.

Principle: An interlocutory judgment on liability sufficient for parties to proceed to AD Stage would preclude a defendant from challenging causation to any extent

I turn first to principle. Bearing in mind the final nature of an interlocutory judgment on liability as discussed above, I am of the view that such an interlocutory judgment, whether by consent or not, which is sufficient for the parties to proceed to the AD Stage, would also prevent the defendant from challenging causation to any extent at that stage. This follows from the effect and purpose of an interlocutory judgment in a bifurcated matter being to establish liability so as to give the claimant the right to claim damages, which are to be assessed at the AD Stage.

In explaining this view, it may be helpful to divide the cases into “single injury” and “multiple injury” situations. In the former situation, there should be no dispute that the claimant has to establish causation in respect of the single injury to obtain interlocutory judgment on liability. Thus, as Ms Leo points out, in a straightforward case, the causative inquiry would end at the liability stage. For instance, if a car driver had knocked into a motorcyclist in an accident and caused him to suffer a head injury, causation is satisfied on the application of a simple but-for causation test: but for the car driver’s negligent driving, he would not have knocked down the motorcyclist and caused him to suffer the head injury. During the AD Stage, causation for the establishment of liability is no longer relevant as the assessment focuses primarily on the severity of the motorcyclist’s injuries, which in turn determines the quantum of damages to be paid by the negligent car driver.

However, problems arise when the Partial Causation at AD Stage Approach is applied to the “multiple injury” situation, which I had outlined at the start of this judgment. First, the Approach as applied is conceptually unsound. It is conceptually difficult to accept that a defendant, in a multiple injuries situation, can effectively be made globally “liable” for all the injuries on the basis that the claimant has established causation with respect to just one of those injuries, so as to make out a cause of action in the tort of negligence for the matter to proceed to the AD Stage. It is even more difficult to accept that the defendant, despite already being technically liable for all of the claimant’s injuries, retains the liberty to challenge the causation of those injuries for which causation has not been established at the AD Stage. This, in effect, allows the defendant to escape the effect and purpose of the interlocutory judgment on liability, which is to make him liable for all of the claimant’s injuries so that the matter can proceed to the AD Stage.

Second, this approach is also inherently inconsistent. How would a court decide consistently which injury to require the claimant to prove causation for, out of the multiple injuries he has suffered, so as to enter interlocutory judgment on liability? Moreover, if this approach is based on the premise that the claimant is only able to adduce evidence to prove causation with respect to one or some of the injuries he suffered, then why should he be allowed to obtain interlocutory judgment with respect to all of the injuries and for the matter to proceed to the AD Stage? As a matter of consistency, if the claimant is not able to prove causation with respect to some of his injuries, then he should not be allowed to enter interlocutory judgment for those injuries. Importantly and practically, moving the matter to the AD Stage for the defendant to “challenge” causation effectively reverses the burden of proof, since the claimant bears the burden of proving causation in the first place.

Third, this approach is also difficult to apply. If the law is that it is possible for parties to challenge causation “to some extent” at the AD Stage despite entering into an interlocutory judgment on liability, this raises questions as to what that extent might be. And because that extent is not based on any sound conceptual premise, there would be no principle to guide the consistent application of this approach. While it is sometimes said that conceptual purity is a luxury in practice, this is one of those instances where clear conceptual clarity is needed. I do not think that the courts should accord flexibility to parties by allowing them to dispute causation “to some extent” at the AD Stage based on unclear principles. I also think that drawing a clear line and saying that parties cannot challenge causation at the AD Stage at all would provide for certainty.

In essence, bearing in mind the existence of the interlocutory judgment on liability in this context, it is idle to speak of a defendant being liable in the tort of negligence at large upon proof of just one of many actionable damage suffered by the claimant. Indeed, Lord Hoffmann explained the following in the House of Lords decision of Harding v Wealands [2007] 2 AC 1 (at [24]):

As the learned Law Lord says, the defendant must be liable for something. It is therefore unprincipled to say that a defendant is “liable” in the tort of negligence, without specifying the damage which he is liable for. Yet, when one considers what a defendant should be liable for, in the context of an interlocutory judgment entered to establish liability so that the matter can proceed to the AD Stage, the answer must be that the defendant has been made liable for all the damage that the claimant is seeking damages for. It must therefore follow that the defendant cannot challenge causation at the AD Stage because the interlocutory judgment has determined that the defendant’s breach had caused all the said damage.

Precedent: The Partial Causation at AD Stage Approach is not supported by English and local authorities

(1) Local authorities do not address whether the defendant is precluded by the interlocutory judgment from challenging causation at the AD Stage

I turn then to precedent. To begin with, I do not think that the Partial Causation at AD Stage Approach is supported by the local authorities. More specifically, the local authorities do not address the specific issue at hand, which is whether the defendant is precluded by the interlocutory judgment on liability from challenging causation at the AD Stage. I refer first to Tan Woo Thian, where Menon CJ had said this (at [11]):

Much reliance has been placed on Menon CJ’s use of the word “any” in this paragraph to infer that if the appellant in that case had been able to prove causation in respect to some of his alleged damage, then he would have crossed the causation hurdle in respect of all of his alleged damage and would have been able to proceed to the AD Stage.

In my respectful view, this is an incorrect reading of the passage. I do not think that Menon CJ intended to leave open the possibility that the appellant in Tan Woo Thian could have moved on to the AD Stage in respect of all of his alleged damage if he could prove “some” of his alleged damage. Indeed, this would go against the overall tenor of Tan Woo Thian in emphasising the importance of causation as a crucial element in making out a cause of action in the tort of negligence. Instead, in my view, what Menon CJ meant by the use of the word “any” is that a claimant has to prove that at least one part of the alleged damage had been caused by the defendant’s alleged breach of duty in order to establish any liability in the tort of negligence at all. If the claimant is able to do so, then the defendant would be liable only in respect of the alleged damage that has been proven.

As for the High Court authorities which seemingly allowed causation to be challenged in whole or in part at the AD Stage, I am of the view that they should not be regarded as laying down any definitive position on this issue. Indeed, it appears that these High Court decisions did not expressly consider the issue of whether causation of parts of the alleged damage can be challenged at the AD Stage despite an interlocutory judgment on liability. It might well have been the case that this issue was not brought to the court’s attention in those cases. Finally, as for the District Court decision of Lim Ai Bee, in so far as it advocates the same approach as that of DR Tan and DR Hakkim, I have already explained why (with respect) it should not be followed.

(2) English authorities that establish that causation can be challenged in part despite an interlocutory judgment on liability should not be followed

I also do not think that the Partial Causation at AD Stage Approach is supported by the English authorities. In particular, I do not think that Singapore law should follow the English authorities that supposedly establish the principle that causation can be challenged in part despite an interlocutory judgment on liability.

(A) The development of the relevant English law

The development of the relevant English law can be viewed through the English High Court decision of Symes v St George’s Healthcare NHS Trust (2014) 140 BMLR 171 (“Symes”) (see generally John McQuater, “Case Comment” (2015) Journal of Personal Injury Law C54 (“McQuater”), which discusses this decision). In that case, the claimant, Symes, was referred to the hospital by his doctor because of a lump on his face. In January 2009, Symes was seen by Williamson, an ear, nose, and throat consultant working for the defendant, NHS Trust. Williamson reported that the lump was a pleomorphic adenoma, but it was in fact a malignant tumour. Symes alleged that the defendant was negligent because of the failure to advise him that his lump was “suspicious of malignancy”, which would have warranted an urgent detailed examination to be carried out within two weeks. Symes further alleged that these failures resulted in the metastasis of the tumour to the lungs and invasion of the facial nerve, both of which were diagnosed in May 2009. Symes then developed inoperable lung cancer and had only a short time to live.

Before Symes commenced legal proceedings, the defendant admitted that Williamson’s report had been wrong and that the detailed examination should have been carried out within two weeks of the initial consultation rather than four months later. However, the defendant denied that the delay had affected the nature or extent of Symes’ surgery or post-operative treatment, or his later development of lung cancer that affected his life expectancy. Relying on the defendant’s admissions, Symes commenced proceedings. The defendant did not serve a defence and judgment in default was entered against it. Symes then served a schedule of losses, and the defendant served a counter-schedule, which accepted that the delay in having the surgery had caused Symes pain and discomfort but denied the other consequences of the delay as claimed.

In March 2014, Master Roberts struck out the parts of the defendant’s counter-schedule that were inconsistent with the particulars of claim. In his view, it would be contrary to the overriding objective for the defendant to allow judgment to be entered against it, only then to allow the defendant to file a counter-schedule that challenged aspects of causation that should have been addressed within a defence filed weeks earlier. The defendant appealed. It submitted that, pursuant to the English Court of Appeal decision of Lunnun v Singh [1999] CPLR 587 (“Lunnun”), the default judgment should be regarded as only establishing that it had acted negligently, such that Symes had suffered some unspecified damage and loss. The defendant further contended that it was not correct to think that the default judgment had already determined that the defendant was liable to Symes for all the damage claimed, since that presupposed a determination of causation that the default judgment did not entail.

Upon the defendant’s appeal, Simon Picken QC, sitting as a Deputy Judge of the High Court, allowed the appeal. He did so on the primary basis that any other conclusion would be contrary to authorities that were binding on him. Pursuant to these authorities, the learned judge explained (at [58]) that in a tort context where either summary or default judgment had been entered in respect of liability, the defendant would be precluded from arguing that no damage at all was sustained because “such an argument would be inconsistent with a judgment on liability in circumstances where, in a tort context, there has to be some damage caused by the tort for the cause of action to be complete” [emphasis added]. As such, “beyond this the defendants were permitted to take issue with causation”. In any event, the learned judge also found that the defendant had accepted that Symes had suffered at least some of the damage that had been alleged in the particulars of claim. The default judgment should therefore be regarded as having determined only that there was some damage and nothing else. Although Symes concerned a default judgment on liability, I do not think that this is material: the broader question remains whether a defendant can challenge causation despite a judgment (in whatever form) on liability.

Symes is simply one of the latest cases in a line of English decisions that have allowed the defendant to contest causation at the AD Stage despite summary or default judgment having been entered into in relation to liability. It is therefore unsurprising that Ms Leo relied on these decisions in support of the Partial Causation at AD Stage Approach. In this regard, the key decision that Ms Leo relied on is the English Court of Appeal decision of Turner v Toleman (Unreported, 15 January 1999) (“Turner”). In that case, the particulars of claim alleged negligence causing injury, loss, and damage. The accompanying medical report described the claimant as having suffered a frozen shoulder. The defence admitted liability to compensate any injuries, loss, or damage suffered but denied that the claimant had suffered the alleged, or indeed any, injury. Despite summary judgment having been entered in favour of the claimant, the defendant made a renewed oral application for permission to appeal. Simon Brown LJ held that the question was “[w]hat loss and damage was caused by this Defendant’s negligence must be part of the exercise of assessing damages”. This seems to suggest that causation can be partially considered at the AD Stage. Simon Brown LJ justified his position as such:

The approach taken in Turner was followed in the English Court of Appeal decision of Lunnun. There, the claimant obtained default judgment against the defendants for damage caused by water draining from the defendants’ premises into the claimant’s cellar. The defendants sought to argue at the assessment of damages hearing that some part of the damage suffered by the claimant as a consequence of the influx of water and sewage into the premises was attributable to some other source other than the defendants’ sewer. In other words, the defendants sought to argue that they were not liable for some part of the damage suffered by the claimant. The learned judge at the assessment of damages hearing held that the defendants were not entitled to make that argument. On appeal, the three judgments delivered by the Court of Appeal took varying approaches. Jonathan Parker J opined that at the AD Stage, all issues are open to the defendant save to the extent that they are inconsistent with the earlier determination of the issue of liability, a rule with which Clarke LJ and Peter Gibson LJ agreed. In particular, Parker J opined:

Clarke LJ further explained that at the AD Stage, the defendant may not take any point inconsistent with the liability alleged in the statement of claim. Such points would include contributory negligence, failure to mitigate, causation, and quantum. Specifically, on causation, Clarke LJ followed Maes Finance Ltd v Al Phillips & Co (Unreported, 12 March 1997) and held that while the defendant cannot contend that the breach of duty was not causative of any loss, the defendant might still be able to argue that such breaches were not causative of any particular items of alleged loss. Finally, Peter Gibson LJ held as follows:

In the English context, Turner and Lunnun pre-dated the introduction of the Civil Procedure Rules (UK) (“CPR”). Crucially, the CPR requires the defendant to plead in the defence any positive case. In justifying the rule in Lunnun, Ms Leo also cited cases that were decided after the introduction of the CPR. For example, she cited the English High Court decision of Carbopego-Abastecimento de Combustiveis SA v Amci Export Corporation [2006] EWHC 72 (Comm) (“Amci Export”). In that case, Aikens J held that where a judgment is given in default, the issues that are thereby determined, and which cannot be challenged subsequently, are those that constitute the bare essence of what the default judgment must necessarily have decided. Such issues are points that were pleaded in the Statement of Claim (at [15]). Therefore, outside of the issues determined by the default judgment, Aikens J opined that the defendant was entitled to argue points on, among others, the causation of damage (at [17]). In this regard, Ms Leo submitted that the rule in Amci Export is equally applicable to interlocutory judgments, and that the effect of the aforementioned decisions is that it is open for a defendant to challenge issues that go towards liability, including causation, at the AD Stage, so long as the defendant’s contentions are not inconsistent with the Statement of Claim.

Indeed, this line of English cases, dating after the introduction of the CPR, can perhaps be summed up by Carr J’s succinct observations in the English High Court decision of New Century Media Ltd v Makhlay [2013] EWHC 3556 (QB) (“New Century Media”) in the following terms (at [30]):

(B) Reasons for not following the relevant English law

With respect, I do not think that these cases should be followed in Singapore to support the Partial Causation at AD Stage Approach. As a preliminary matter, it is not clear to me that the English cases speak with one voice. In Turner, Simon Brown LJ had simply insisted that whether damage was caused by the defendant’s negligence must be part of the exercise of assessing damages. However, by the time New Century Media was decided, the proposition had morphed into one that the defendant can only challenge causation at the AD Stage if this was not inconsistent with the judgment on liability. Indeed, in the even more recent English Court of Appeal decision of Seabrook v Adam [2021] 4 WLR 54, Asplin LJ interpreted Lunnun to say that it is not always open to a party to contest causation after having admitted to liability and that it all depended on the pleadings (at [20]). This appears to be a further erosion of the approach taken in Turner. While this may have been caused by the introduction of the CPR in the English context, such inconsistency can also be observed across the three judgments delivered in Lunnun. This unclarity in the precise proposition from this line of English cases is a good reason in itself not to follow what is otherwise unclear authority. However, it is useful to consider the two broad propositions that have emerged from the English cases pertaining to the issues at hand.

To begin then, I would not follow Simon Brown LJ’s approach in Turner that causation can be partially considered at the AD Stage. In this regard, the learned judge had justified his approach on the basis of his own “experience” and the “practice” of the English courts. Despite the weight that one ascribes to the experience of a Lord Justice, and even if one accepts that the practice was as such at the time of Turner, it is still the case that the right principles should take precedence. In sum, I do not find Simon Brown LJ’s approach to be persuasive since no principled justification could be found in Turner. In my respectful view, a bare assertion that causation is a part of the assessment process requires stronger justification. Unlike the judge in Symes, who appeared to have some concerns with Turner but regarded himself as bound by a higher authority, I am not bound to follow Turner and the subsequent English cases. I accordingly reject Simon Brown LJ’s statement of the law in Turner to the effect that whether damage was caused by the defendant’s negligence is part of the AD process.

Moving on now to consider the subsequent proposition of law in Amci Export, New Century Media, and Symes that it is open to the defendant to challenge issues of causation that have not been decided in (and hence not inconsistent with) a judgment on liability, I am not convinced that this should be applied in Singapore for the following reasons.

First, I am not convinced that these cases relied on good authority to advance the proposition that it is open to the defendant to challenge issues of causation that have not been decided by the judgment on liability. For example, turning to Amci Export, I do not think that the authorities that Aikens J relied on provide the basis to support his conclusion that the defendant is entitled, in the same action, to challenge issues that go toward liability at the AD Stage so long as the challenge did not pertain to points pleaded in the Statement of Claim. The Privy Council decision of Kok Hoong v Leong Cheong Kweng Mines Ltd [1964] 2 WLR 150 and the House of Lords decision of New Brunswick Railway Co v British & French Trust Corporation [1939] AC 1 (“New Brunswick”), which Aikens J relied on (at [15]), stood instead for the proposition that a default judgment creates a limited estoppel, which prevents a defendant from setting up in a subsequent action a defence which was necessarily, and with complete precision, decided by the previous judgment (see New Brunswick at 21). To say an estoppel operates in respect of a subsequent action is, however, quite different from the question of whether there is any inconsistency in allowing causation, an issue that is an essential element for liability, to be challenged in the same action even when liability has been established by an interlocutory judgment.

Second, I return to my points of principle explained above. Even if I were to accept that it is open to the defendant to challenge issues of causation that have not been decided, the question then becomes what is decided when interlocutory judgment on liability is entered in favour of the claimant? More specifically, in asking this question, I am not concerned with just any judgment for liability, but with an interlocutory one that is entered into in a bifurcated matter so that the parties can proceed to the AD Stage. This added qualification is important because the ability of the parties to proceed to the AD Stage presupposes that the claimant has proved the elements of a cause of action in the tort of negligence. In sum, and as I have explained above, the extent of an interlocutory judgment on liability in this context must be considered with regard to the elements needed to make out a cause of action (or liability) in the tort of negligence. Because an essential element is that the defendant’s breach of duty must have caused the claimant’s damage, the extent of the interlocutory judgment on liability must be understood with regard to the damage that the claimant can prove was caused by the defendant’s breach. As such, I do not think it accords with principle to say that an interlocutory judgment on liability can be taken to have decided liability for all the damage for the only purpose of proceeding to the AD Stage, but not to prevent the defendant from challenging causation in respect of the damage he has not admitted to. Either the claimant has established liability (and with that, causation) in respect of all the damage that he is claiming damages for, or he has not. There is no half-way house that justifies the Partial Causation at AD Stage Approach.

Third, even if I were to (again) accept that it is open to the defendant to challenge issues of causation that have not been decided despite the interlocutory judgment on liability, it seems to me that, at least in so far as PIMA cases are concerned, the Protocol ensures that all the issues of causation would have been sufficiently ventilated between the parties even before legal proceedings are commenced. In this regard, para 3.2 of Appendix B to the PD 2021 provides that the letter of claim, which the claimant must send to the potential defendant and his insurer, must set out the “full particulars of his claim”, including “(b) a brief description of the nature of the property damage and/or injuries suffered” and “(c) an estimate of general and special damages with a breakdown of the heads of claim”. This deliberately adopts a specific definition of “damage” that would only be satisfied upon the provision of the specific damage suffered. Indeed, it would not be sufficient, nor would this be consistent with the courts’ practice, for a bare assertion to be made that a claimant has suffered “physical damage” but leave it to the AD Stage to flesh out exactly what that entails.

Assuming that the Protocol does not result in an early settlement between the parties, it is conceivable that the claimant will base his eventual Statement of Claim on the letter of claim. If the interlocutory judgment on liability is entered on the basis of the Statement of Claim, it would be likely that the defendant would be precluded by the breadth of the stated claim from challenging any residual issue of causation at the AD Stage, if this was even conceptually permissible. In any event, as I observed above, O 18 r 12(1A) of the ROC 2014 requires the plaintiff to annex a medical report and a statement of the special damages being claimed to his Statement of Claim. This is therefore similar to the argument raised in Symes, which the court had some sympathy for (see Symes at [61]), that, viewing the particulars of claim in that case as a proxy for the default judgment, the damage which is necessary in order for there to be a cause of action must be the damage alleged in the particulars of claim and not some vague, unpleaded notion that the claimant merely suffered “some damage”.

Finally, and more broadly, while not always determinative, there is a distinct absence of secondary authority for the underlying proposition that a cause of action for all damage in a tort of negligence is made out once the claimant proves causation for some of the damage. Indeed, Turner and Lunnun are not even cited in the leading tort textbooks. Instead, these textbooks have tended to focus on the need for “actionable damage” to make out a cause of action in the tort of negligence, without specifying the extent to which this must be shown (see, eg, Clerk & Lindsell at paras 7-04 and 7-05, and Charlesworth & Percy on Negligence (Mark Armitage gen ed) (Thomson Reuters, 15th Ed, 2022) at para 6-01). The same might be said of the articles that I have referred to above, where the focus has tended to be on the need for actionable damage (see, eg, Donal Nolan, “Rights, Damage and Loss” at 274; Donal Nolan, “Damage in the English Law of Negligence” at 267; and Christian Witting, “Physical Damage in Negligence” at 206). Indeed, in an important article, Professor Donal Nolan simply notes that “[t]hat damage is an element of the tort of negligence is not in doubt: actionable injury completes the cause of action, so that time begins to run for limitation purposes only from the moment it occurs” (see Donal Nolan, “New Forms of Damage in Negligence” (2007) 70 MLR 59 at 59). There is otherwise no elaboration of the extent of the actionable damage that needs to be proved to constitute the cause of action.

For all the reasons given above, I conclude that neither the local nor English authorities support the Partial Causation at AD Stage Approach. In so far as the English cases establish that a defendant is able to challenge causation at the AD Stage despite the entering of a judgment on liability, I would respectfully not follow those cases in the present context.

Policy: The Partial Causation at AD Stage Approach is inconsistent with prevailing practice and policy

(1) Inconsistency with prevailing practice

Finally, as a matter of policy, the Partial Causation at AD Stage Approach is inconsistent with how the local courts have been apportioning liability in PIMA cases. In CXN, Teh JC explained that in determining the apportionment of liability, regard must be held to the extent of the defendants’ individual responsibilities for the claimant’s injuries so as to arrive at a just and equitable apportionment (at [54]), referring to s 16(1) of the Civil Law Act 1909 (2020 Rev Ed). Indeed, as the learned judge also alluded to, the Court of Appeal in Cheng William v Allister Lim & Thrumurgan and another and another appeal [2015] 3 SLR 201 held that the apportionment exercise involved the comparison of the relative significance of the acts or omissions of the parties in causing the claimant’s injuries, and of the relative culpability of the parties. Consistent across CXN and the authorities cited within is the reference to the claimant’s injuries. This presupposes that the apportionment of liability takes place against the backdrop of the defendant’s liability for those injuries having already been established.

Transposed to the entering of interlocutory judgments for liability in PIMA cases, I note that Form 7, as with Form 9I, both record “Consent Judgment on liability” in terms of percentage liability in favour of the claimant:

In entering a consent judgment on liability, it would be odd to think that the defendant has not admitted to liability with respect to all the damage or injuries that are the subject of the consent judgment. Absent such a vital admission, it would be difficult to apportion liability if the specific damage or injuries that the defendant is liable for are still left to be determined. In the same vein, it follows that if liability for part of the admitted injuries or damage can be challenged at the AD Stage, it would make any earlier determination of apportionment at the liability stage meaningless. Therefore, it cannot be that the defendant is allowed to challenge causation with respect to some of the claimant’s damage or injuries at the AD Stage.

In saying this, I am aware that the courts rely on the liability apportionment guidelines provided in the Motor Accident Guide (2nd Edition) (“MAG”) issued by the State Courts and the Motor Accident Claims Online (“MACO”) in both effecting the Protocol and also deciding on the actual liability proportions. I understand that the guidelines are based primarily on the manner in which the accident happened, rather than the injuries suffered by the claimant. However, I do not think this affects my observations above. First, the guidelines only provide an estimate of the apportionment of liability so as to aid in early settlement in simple cases. Thus, they should not apply so strictly in cases where the core elements of liability, such as causation, are contested. Second, as Teh JC held in CXN (at [74]), citing the Court of Appeal decision of Ng Li Ning v Ting Jun Heng and another [2021] 2 SLR 1267, the guidelines in the MAG and the MACO are meant only as estimates and should not override the otherwise highly fact-sensitive apportionment exercise.

(2) Inconsistency with prevailing policy

Further, as aptly put in McQuater, raising issues about causation in relation to liability, after judgment, is not a satisfactory way of approaching matters in the modern era where case management has placed greater emphasis on proportionality and predictability. In my view, the Partial Causation at AD Stage Approach would not be in line with the broad policy that undergirds the Protocol and the simplified dispute resolution process that is aimed at encouraging parties to settle their PIMA cases as early as possible. Indeed, it does not appear consistent with this broad policy to allow the defendant to, in essence, kick the can down the road by reserving some issues of causation to be contested at the AD Stage, despite having consented to an interlocutory judgment for liability. As the experience in the English cases shows, this would give rise to further disputes as to the content of the judgment for liability and reignite issues of causation at the AD Stage. There is also no reason why the defendant cannot deal with the causation issues at the liability stage if it remains willing to do so at the AD Stage (indeed, as I explained above at [85], on a more difficult basis since the burden of proof would effectively be reversed).

Indeed, the problem of defendants and their insurers changing their position despite having, in effect, admitted to liability can also be seen in the UK. In a response to a consultation on pre-action admissions, the UK Association of Personal Injury Lawyers (“APIL”) pointed out that defendants and their insurers should not be allowed to withdraw their admissions of liability in so-called multi-track cases as a result of the English Court of Appeal decision of Sowerby v Charlton [2006] 1 WLR 568 (see Ministry of Justice Consultation on Pre-Action Admissions: A Response by the Association of Personal Injury Lawyers (APIL) (2006)). While not the same as in the present case, the response does point out that claimants make decisions on how to run their lives based upon known aspects of their claim and on how they expect to be compensated. As such, knowing that liability is admitted and that they will be compensated aids in their physical and mental recovery process. However, if the extent of the admitted liability is not clear, the claimant may suffer additional worry and anxiety. Further, the APIL’s response also referred to the trend where insurers admitted to primary liability without raising any issues of contributory negligence, but later did so despite having made a full admission of liability (at p 5). It is important that the Singapore approach avoids such uncertainty for the claimant. In my view, the Partial Causation at AD Stage Approach has the potential to introduce such uncertainties because it is inconsistent for interlocutory judgment to be entered in respect of liability but then to allow the defendant to contest causation, which is an element of liability, at the AD Stage.

Reconsideration of prior cases adopting the Partial Causation at AD Stage Approach

Accordingly, for all of these reasons, I am of the view that it should not be possible to challenge causation to any extent at the AD Stage. I would respectfully suggest that the Partial Causation at AD Stage Approach, in so far as it has been adopted in prior cases, be reconsidered.

Why the No Causation at AD Stage Approach is correct

I turn finally to explain why the No Causation at AD Stage Approach is correct and should be adopted. In doing so, I shall have occasion to return to some of the conceptual points in the tort of negligence that I had discussed above.

The No Causation at AD Stage Approach accords with the conceptual points in the tort of negligence

To begin with, I have explained that a cause of action refers to “the facts which the claimant must prove in order to get a decision in his favour”. In the context of a cause of action in the tort of negligence, a claimant would need to prove the facts associated with each injury in a multiple injuries situation in order to complete the cause of action in the tort of negligence that encompasses those injuries.

More specifically, as was held in Salcon, the claimant would need to establish causation, being a crucial element of the cause of action, before the question of quantification of damages can arise. And since the quantification of damages only arises in respect of damage that is proven, then it must follow that the claimant must prove causation in this manner with respect to each part of the alleged damage he has allegedly suffered. If he cannot establish causation as to one particular part of the alleged damage, then he would not have crossed the liability hurdle with respect to that part of the damage. No assessment of damages would arise for that damage. This is a legal conclusion that cannot be overcome by the parties agreeing to a consent interlocutory judgment.

Therefore, if a claimant suffers injuries A, B, and C in an accident that was indisputably caused by a defendant, and the claimant can only prove causation in relation to injury A, he can make out a cause of action in the tort of negligence but only in respect of injury A. It follows that the claimant can only claim damages for injury A. However, it may be asked, if the claimant finds out later that he can prove causation for injury B, having already obtained interlocutory judgement for injury A, can he bring a separate claim framed as a different cause of action?

The claimant must bring his action for injuries arising for the same breach once and for all

In my view, while this does not arise for direct consideration in the present case, the claimant cannot do this. Where there are multiple injuries caused by a single breach of duty, all the injuries would form a single cause of action. Crucially, it is trite that the claimant must bring his claim once and for all against the defendant in respect of all of the alleged damage, even if there is a possibility that further damage might be discovered by the claimant in the future (see Clerk & Lindsell at para 30-15). It therefore follows that, in respect of the same breach of duty, the claimant cannot split his cause of action by suing separately for different parts of the alleged damage. He also cannot bring a subsequent claim to increase or decrease the award made on his initial claim if the damage is subsequently discovered to be greater or less than the terms of the initial award or settlement (see Fleming’s The Law of Torts (Carolyn Sappideen and Prue Vines gen eds) (Thomson Reuters, 10th Ed, 2011) at para 10.20).

The English Court of Appeal decision of Bristow v Grout (The Times, 9 November 1987) illustrates this point. In that case, the appellant was struck by a motor car that was driven by the respondent and there was no doubt that the respondent was at least negligent. The parties entered into a binding compromise and settlement. The appellant subsequently brought another claim for damages in respect of further injuries and symptoms that were a consequence of the road accident, but which were unknown to the appellant at the time he entered into the compromise and settlement. The court rejected the appellant’s claim because there was previously a full and final settlement of the appellant’s claim in negligence. Thus, he could not establish a subsequent claim by alleging that he sustained further damage of which he was earlier unaware. It follows that the court regarded that there was only a single cause of action in negligence in respect of the same accident, even though there were multiple injuries or multiple parts of the alleged damage. Given that the single cause of action had been compromised, the appellant could not establish a separate claim in negligence in respect of the further injuries.

In contrast, where there are two separate breaches of duty by the same defendant towards the same claimant at different times, there may be two different causes of action (see Limitation Periods at para 5.008). In Birmingham Midshires Building Society v JD Wretham [1999] Lloyd’s Rep PN 133, the third defendants (“D”) were a firm of solicitors who acted for the claimant in a mortgage transaction in 1990. In 1995, the claimant became aware that there were substantial differences between their understanding of the nature of the mortgage transaction and what it actually was. At that time, they also discovered that D knew the claimant’s understanding was inaccurate. As such, by a writ issued on 28 February 1997, the claimant sued D in negligence and breach of trust on the ground that D were under a duty to report those differences to the claimant but failed to do so. However, D sought to argue that the claim in negligence was time-barred. Their case was that, at the time of the mortgage transaction in 1990, they had negligently failed to inform the claimant that a demolition order had been made in respect of the house in the mortgage transaction. As the claimant had discovered the existence of the demolition order in April 1991, D argued that it was at that point where the three-year limitation period began to run and that, accordingly, the claim in negligence was time-barred when the writ was issued in 1997.

However, Hicks J held that the negligent failure to disclose the existence of the demolition order in question was of a different kind from the negligent failure to disclose the nature of the transaction. The latter negligence therefore gave rise to a different cause of action which was not time-barred since the claimant had not been aware of this particular breach until 1995. Therefore, in my view, the test in determining whether there are one or multiple causes of action is whether there is in substance one or multiple breaches of duty by the same defendant. The sole fact that there are multiple parts forming the total alleged damage claimed for is irrelevant to answering this question.

For completeness, it follows from the foregoing that multiple causes of action in the tort of negligence do not arise solely because different types of rights have been interfered with. Indeed, if the existence of different parts of damage in respect of the same type of right do not by themselves give rise to multiple causes of action, it is difficult in principle to accept that there should be any difference even where different types of rights are involved. Moreover, there appears to be no explanation for why different causes of action should accrue just because different types of rights are interfered with (cf the English Court of Appeal decision of Brunsden v Humphrey (1884) 14 QBD 141 and see the High Court decision of Ng Kong Choon v Tang Wee Goh [2016] 3 SLR 935 at [72]). On the contrary, to accept that a cause of action might be split because different types of rights were interfered with is, in my view, an unwarranted distinction that goes against the longstanding policy objective of promoting finality to litigation and discouraging a claimant from bringing a multiplicity of proceedings. It might encourage claimants to bring subsequent proceedings in respect of a different type of damage so that they can have an opportunity to reargue their case. It would also disincentivise claimants from raising all relevant facts in their claim in the hopes that they might hedge their bets by adapting their arguments in subsequent proceedings depending on how well received their arguments were.

My answers to the Questions

Question 1: Whether causation can be reserved in toto to the AD Stage

With the above background in mind, I turn to consider Question 1. For all the reasons discussed above, I answer Question 1 in the negative, that is, causation cannot be reserved in toto to the AD Stage.

Question 2 and Question 3: The extent to which causation may be challenged at the AD Stage

I turn now to Question 2 and Question 3. At the outset, I am of the view that Question 2 and Question 3 need to be reframed. After hearing parties, I come to the respectful view, in agreement with Ms Leo, that Question 2 and Question 3 have not been framed accurately.

The extent to which causation may be challenged at the AD Stage

In particular, by framing the issues as whether a claimant had suffered “general damages” or “special damages” causally connected to the defendant’s breach, the defendant reveals a fundamental conceptual misapprehension by conflating the concepts of “damage” and “damages”. As I have explained above, while the claimant’s injury in a personal injury claim would constitute the “damage”, general damages and special damages are actually the “damages” which are awarded to a successful claimant as monetary compensation for his loss. Put differently, general damages and special damages do not constitute the elements necessary to establish a cause of action in the tort of negligence. Therefore, it only makes sense to speak of damages if liability is first established. Based on this analysis, as Ms Leo points out, the practice in PIMA cases of conceding liability to pay special damages in order to complete the cause of action in tort and thereafter contest causation for the claimant's injuries at the AD Stage reveals a conceptual conflation between “damage” and “damages”. I respectfully suggest that this practice should not be continued.

Accordingly, I do not think it is appropriate to answer Question 2 and Question 3 as they have been framed by the defendant as they are based on a fundamental conceptual misapprehension. However, from the defendant’s submissions, I discern that the true import of Question 2 and Question 3 is really to ask, if causation cannot be reserved in toto at the AD Stage, to what extent can it be challenged at the AD Stage? To the extent that I have explained why the Partial Causation at AD Stage Approach is wrong, and that the No Causation at AD Stage Approach is correct, I have answered the substance of the question inherent in Questions 2 and 3.

What can be challenged at the AD Stage

Furthermore, in so far as Question 2 and Question 3 relate to what may still be challenged at the AD Stage following an interlocutory judgment on liability, I provide my answer as follows. In essence, in the context of bifurcated proceedings for PIMA cases, where any question relates to whether the alleged damage exists or should be legally attributed to the defendant’s breach of duty through the doctrines of causation and remoteness, this must be dealt with at the liability stage. On the other hand, where the question relates to the consequences flowing from proven damage and their extent, this may be reserved to the AD Stage.

Generally speaking, in a bifurcated trial, the typical enquiry at the AD Stage is on measuring the extent of the loss suffered by the claimant as a consequence of the damage that was proven at the liability stage. For instance, in cases of personal injury, the enquiry at the AD Stage may involve a calculation of the claimant’s loss of future earnings from not being able to work, which represents the consequence that his physical injury will have on him. Similarly, the claimant may be compensated for the non-pecuniary consequences that flow from his injuries, such as pain and suffering. This contrasts with the inquiry at the liability stage where, in respect of the alleged damage (which must be actionable), the enquiry would be on whether the damage was caused by the defendant’s negligence, and whether such damage is too remote.

To be clear, it is important to highlight that these heads of loss, which go towards assessment, are conceptually distinct from the element of “damage” that makes out a cause of action in negligence. These heads of loss may not in themselves be actionable, but recovery for them is permitted once a cause of action is established and it is proven that these losses are consequential to proven actionable damage. For instance, distress or being upset (absent recognised psychiatric injury) is a head of loss that does not in itself ground a negligence claim (see the House of Lords decision of Hicks and another v Chief Constable of South Yorkshire Police [1992] 2 All ER 65 at 69), but if it is consequential on actionable damage, such as a broken leg, then compensation will generally be payable in respect of the claimant being distressed or upset (see Winfield & Jolowicz at para 7-004).

Following from the above, the consideration of other possible subsequent events (that is, subsequent to the initial suffering of the damage) is also permissible at the AD Stage. This is because such consideration is not a question of causation in the sense of proving that the defendant’s breach of duty was the cause of the damage. Indeed, once the cause of action is completed, causation is taken to have already been established in respect of the claimed damage and this would not be affected by subsequent events. Instead, the consideration of possible subsequent events belongs to the AD Stage because it seeks to ascertain the extent of the claimant’s loss that the defendant should be made liable for, and thereby quantify such loss in monetary terms.

Thus, in the English Court of Appeal decision of Smithurst v Sealant Construction Services Ltd [2011] EWCA Civ 1277, the appellant suffered a disc prolapse and the respondent accepted that this was a result of its negligence in providing him with a defective van. The trial judge accepted the respondent’s expert evidence that the appellant was likely to have suffered a major prolapse with similar consequences within two years. The trial judge treated this expert evidence as being relevant to the question of causation, and he regarded the issue as whether or not the appellant’s injury would have occurred but for the admitted breach of duty by the respondent. However, this approach was rejected by the English Court of Appeal, which held that the trial judge wrongly treated the question he had to decide as one of causation, in the sense of proof that the respondent’s breach of duty had caused the appellant’s injury (at [13]). Instead, the expert evidence was relevant to the question of assessment of damages (at [10]). This was because it was not in dispute that the appellant had suffered injury as a result of a breach of duty on the part of the respondent; the only remaining question was how great a loss he had suffered as a result (at [11]).

However, any further actionable damage that flows from the defendant’s breach of duty must be proved at the liability stage. This has to be distinguished from the consequences that flow from proven damage, which are represented in heads of loss and may be proven at the AD Stage. The English Court of Appeal decision of Smith v Leech Brain & Co Ltd and another [1962] 2 WLR 148 is an example where further actionable damage flowing from the defendant’s breach of duty was in issue. There, the claimant’s husband was a labourer and galvaniser of the defendant who suffered a burn on his lip that was caused by the defendant’s breach of duty in failing to provide adequate protection. The burn was the promoting agent of cancer, which developed at the site of the burn, and from which he died three years later. In finding that the defendant was liable to compensate the claimant for her husband’s death from the cancer, the English Court of Appeal held that since the defendant could have reasonably foreseen the type of injury suffered, which was the burn, the defendant should have reasonably anticipated the cancer which was merely an extension of the burn (at 156). This reasoning of this case has been regarded as falling under the question of remoteness (see Clerk & Lindsell at para 2-172), which in turn goes towards the question of liability and not the question of assessment of damages.

In sum, in deciding whether any question ought to be properly dealt with at the liability or the AD Stage, the test is whether the question goes towards the elements of negligence. In particular, where the question relates to the existence of any part of the alleged damage, this must be dealt with at the liability stage. On the other hand, where the question relates to the consequences flowing from proven damage and its extent, this may be reserved to the AD Stage. These principles are succinctly summarised by Stephanie Jackson-Haisley J in the Jamaica Supreme Court decision of Vera Dallas v L P Martin Co Ltd [2018] JMSC Civ 78 (at [45]):

In sum, as Eyre J put it in the English High Court decision of Nedjla Surer v Stuart Driver [2021] EWHC 3595 (TCC) (“Stuart Driver”), the issue is the level of generality at which to consider the question of causation. Specifically, as the court in Stuart Driver asked (at [13]), is the court considering “a question of whether damage has been caused where proof of damage is an essential ingredient of a particular disputed tort or considering questions of the causation of particular loss flowing from already established torts”? In my view, for clarity, it is preferable to reserve the future usage of the term “causation” to questions of liability, and not questions of assessment of damages. This is to avoid further confusion about whether issues of “causation” may be challenged at the AD Stage.

Practical considerations

All that said, I recognise that there are practical merits to the State Courts’ previous practice of allowing the parties to challenge causation in toto (ie, the Total Causation at AD Stage Approach) (and even the current approach of allowing them to challenge causation to some extent (ie, the Partial Causation at AD Stage Approach)). In this regard, the DR in Eliora Yow helpfully explained the practical reasons why causation of the alleged injuries should remain a live issue at the AD Stage. He had said as follows (at [14]–[15]):

It is important to bear in mind these practical reasons for the practitioners’ treatment of causation prior to Tan Woo Thian. I am cognisant that the Protocol provides for the resolution of PIMA cases on the (eventual) basis of a trial bifurcated between the liability and AD Stage. This would reflect the existing practice of parties and the court in assessing the progress of a personal injury claim in terms of how far the claim has advanced through the liability and AD Stage. Clearly, there are case management concerns that need to be addressed.

In a related vein, it may be worthwhile for the Protocol and later processes in the resolution of PIMA cases to define precisely what “liability” means. In this regard, the Singapore approach may be contrasted with the UK Pre-Action Protocol for Low Value Personal Injury Claims in Road Traffic Accidents (“RTA Protocol”). In the RTA Protocol, para 1.1(1) expressly defines an “admission of liability” to mean that the defendant admits that (a) the accident occurred, (b) the accident was caused by the defendant’s breach of duty, and (c) the defendant caused some loss to the claimant, the nature and extent of which is not admitted. This has been understood under the English system to amount to an admission of “primary liability”, which is in turn usually understood to mean that the defendant reserves the right to argue contributory negligence because an unqualified admission of liability would preclude that right. If the defendant only admits to breach of duty but disputes causation, this will not amount to an admission of liability at all (see John McQuater, “Case Comment” (2017) Journal of Personal Injury Law C226, commenting on the English High Court decision of Blake v Croasdale and another [2017] EWHC 1336 (QB) (“Blake”)).

Finally, since I have been asked the Questions in the context of PIMA cases, I expressly restrict my conclusions to PIMA cases that have been bifurcated. Assuming that they are not overturned by an appellate court, it would be better for the applicability of these conclusions outside of the PIMA cases to be considered when the occasion arises specifically. Also, in so far as cases have already commenced before this judgment and conducted in a manner inconsistent with the conclusions within, I would reserve the question of whether the conclusions should only apply prospectively to when the occasion arises specifically.

Conclusion

For all of the reasons above, I answer Question 1 in the negative, in that causation, in the sense of the causal connection between the defendant’s breach and the damage claimed, cannot be reserved in toto to the AD Stage.

I decline to answer Question 2 and Question 3 as they were originally framed by the defendant as being based on a fundamental conceptual misapprehension. However, bearing in mind the import of these Questions, I further hold that causation cannot be challenged to any extent at the AD Stage.

In summary, and for clarity, I have reached the following conclusions in relation to the Questions, which are as follows.

(a) While Tan Woo Thian was not a PIMA case, its statements on the relevant principles of the tort of negligence are clearly universal and apply to PIMA cases as well. Amongst others, Tan Woo Thian established that a claimant needs to prove the following four elements in order to make out a cause of action in the tort of negligence: (a) the defendant owed a duty of care to the claimant, (b) the defendant acted in breach of that duty, (c) there was a causal connection between the defendant’s breach and the claimant’s damage, and (d) that particular kind of damage suffered by the particular claimant is not so unforeseeable as to be too remote.

(b) As such, a claimant can make out a cause of action in the tort of negligence so long as he can establish, among others, causation in relation to one of the injuries he has suffered. Specifically, causation in this context refers to the causal connection between the defendant’s breach of duty and the damage alleged. If the claimant has made out a cause of action, he is said to have established liability in the tort of negligence.

(c) The claimant needs to establish causation in this sense because causation and damage are necessary elements to make out liability in the tort of negligence, so that an interlocutory judgment on liability can be entered. However, if the claimant manages to establish causation only in respect of, say, injury A, when he claims to have additionally suffered injuries B and C, his cause of action is limited to injury A. As such, the claimant would only be able to claim damages for injury A. It follows that the defendant would not be able to challenge causation at the AD Stage in respect of injuries B and C because it is not necessary for him to do so: the claimant has not even established liability with respect to those other injuries. As a matter of principle, the type of the alleged damage (ie, whether they are personal injury or property damage) should not affect the analysis so long as the damage all emanate from the same breach of duty so as to constitute one cause of action.

(d) In this context, the effect and purpose of an interlocutory judgment in PIMA cases that are bifurcated is to establish liability so as to give the claimant the right to claim damages, which are to be assessed at the AD Stage. In other words, the interlocutory judgment, whether by consent or not, determines finally the defendant’s liability in respect to the claimant’s cause of action in the tort of negligence. This necessarily means that the interlocutory judgment has determined that the defendant is liable for all of the damage that the claimant is claiming for, leaving the extent of such damage and the resulting quantum of damages to be determined at the AD Stage. The defendant is not able to challenge the liability that has been established pursuant to the interlocutory judgment. Following from this, the Total Causation at AD Stage Approach and The Partial Causation at AD Stage Approach are both wrong.

(e) At the AD Stage, the defendant may raise issues that relate to the consequences flowing from proven damage and their extent. The consideration of these other possible subsequent events is not a question of causation in the sense of proving that the defendant’s breach of duty was the cause of the damage. Instead, such questions belong to the AD Stage because they seek to ascertain the extent of the claimant’s loss, and thereby quantify such loss in monetary terms. However, these events must flow from proven damage. As such, it would not be conceptually correct for a claimant to justify liability on the basis of consequential loss (eg, loss of future earnings) without proving the underlying damage (in this case, the specific personal injury).

Needless to say, these points in the preceding paragraph are not exhaustive of the reasoning I have employed in this judgment. But they relate to the main points in relation to the Questions.

In closing, I wish to reiterate my immense gratitude to Ms Leo for her able oral and written submissions, which have been extremely helpful in the course of coming to my decision.