1 This was an appeal against the decision of the High Court judge (“the Judge”) with regard to the Appellant’s application that the Respondent be compelled to produce legal advice (“the Advice”) which he had received from his in-house legal department on 3 April 2008. This appeal arises out of an application filed in Suit No 350 of 2014 (“the Suit”). As will soon be clear, the Suit was itself a sequel to prior legal proceedings that were the subject of this court’s decision in Comptroller of Income Tax v AQQ and another appeal [2014] 2 SLR 847 (“AQQ”). We will come to the details shortly, but it suffices to state for now that the Advice had been referred to in an affidavit affirmed by the Respondent’s employee, one Ms Christina Ng Sor Hua (“Ms Ng”) on 28 August 2014, and that the Appellant had sought production of the Advice to demonstrate that the Suit is time-barred.

2 Before the Judge, the Appellant argued that the Advice was not privileged as it was created before amendments to the Evidence Act (Cap 97, 1997 Rev Ed) (“the EA”) in 2012 extended legal professional privilege to advice given by in-house counsel. Alternatively, it argued that privilege had impliedly been waived by reason of the reference made to it in Ms Ng’s affidavit. After hearing the parties, the Judge did not make an order for production. He ordered that the Respondent was to elect whether he:

Subsequently, Ms Ng affirmed an affidavit in which she deposed that the Respondent “sought to refer only to the fact that advice was sought … and received on 3 April 2008 … and does not seek to rely on the substance or content of the [Advice]”.

3 Dissatisfied, the Appellant appealed, reprising the same arguments it had presented in the court below. The Appellant also argued that the Judge had erred in law by failing to render a decision and had, instead, left it to the Respondent to “decide for himself” what the outcome of the application should be. After careful consideration of the arguments presented by both parties, we dismissed the appeal. It was clear to us that the Judge had held that the Advice was privileged under the common law and that there was no basis for abrogating privilege in this case. We affirmed his holdings on these points. It was also clear to us that what the Judge had in mind when he made the order giving the Respondent the option to elect was that the Respondent was to clarify whether (and if so, how) he would use the Advice in the future. This was, in our judgment, a prudential exercise in case-management and did not amount to having the Respondent decide the outcome of the present application for himself. We therefore dismissed the appeal, delivering judgment as follows:

4 The parties then proceeded with the hearing of one of the two applications which had been held in abeyance pending the outcome of this appeal. However, disputes soon arose as to the precise ambit and basis of our decision and the parties sought clarification from us. In order to address their concerns, as well as to set out our decisions on the many important points of principle in relation to the scope of legal professional privilege which were raised in this case – which included whether the doctrine of legal professional privilege extended to communications made by in-house counsel before 2012 and the circumstances under which privilege may be impliedly waived by reason of actions taken during litigation – we now give the detailed reasons for our decision, beginning with a more detailed recitation of the relevant facts.

5 We begin with the events in AQQ. In 2002, significant changes were introduced to Singapore’s tax regime. Following these changes, the Appellant, a company listed on the Kuala Lumpur Stock Exchange which had significant business interests in Singapore, took steps to restructure its Singapore operations. Among the steps that were taken was the incorporation of a new subsidiary in Singapore, “AQQ” (“the Subsidiary”), which acquired the Appellant’s interests in most of its Singapore subsidiaries through a complex financing scheme. We shall, as in AQQ, refer to this as the “Corporate Restructuring and Financing Arrangement”.

6 Between 2005 and 2007, the Respondent paid approximately $9.6m to the Subsidiary in tax refunds. Subsequently, the Respondent concluded that the Corporate Restructuring and Financing Arrangement was a tax avoidance arrangement which was designed, among other things, to induce the payment of the aforementioned tax refunds. The Respondent then issued additional notices of assessment (“Additional Assessments”) to recover the sums paid out in refunds. On appeal, this court accepted that there was a tax avoidance arrangement within the meaning of the Income Tax Act (Cap 134, 2008 Rev Ed) (“the ITA”) but it held that the Respondent had acted ultra vires in attempting to claw back the payments through the assessment of additional sums in tax so it set aside the Additional Assessments. As a consequence, the Respondent came away with only “a purely symbolic victory” because while he succeeded in demonstrating that there was a tax avoidance arrangement, he had not succeeded in exercising his statutory powers in an appropriate manner to counteract the tax advantage gained (see AQQ at [30]). In the concluding paragraphs of its judgment in AQQ, this court observed that the Respondent might nevertheless be able to avail himself of alternative means of recovering the tax refunds that it had paid to the Subsidiary (see AQQ at [162]).

7 This was precisely what the Respondent sought to do in the Suit, which it commenced on 1 April 2014 to seek recovery of the tax refunds. The gist of the Respondent’s case was that the Subsidiary had falsely represented that certain interest payments incurred under the Corporate Restructuring and Financing Arrangement were legitimate interest expenses, when they had in fact been contrived expressly to reduce the extent of its tax liability and allow it to claim tax rebates. As a result, the Respondent pleaded, he had mistakenly paid a total of $9.6m to the Subsidiary in tax refunds. The following heads of claim were pleaded: unjust enrichment, fraudulent misrepresentation, and conspiracy by unlawful means. The Respondent also pleaded that the Appellant was liable as it had both been unjustly enriched as a result of the tax refunds paid to the Subsidiary in addition to having conspired with the Subsidiary to induce the payments of the tax refunds.

8 On 6 June 2014, the Respondent filed an application to serve his writ out of jurisdiction on the Appellant at its registered address in Malaysia. This application was accompanied by an affidavit sworn by Ms Ng. In a section of the affidavit entitled “Full and Frank Disclosure”, she alerted the court to the possibility that a limitation defence might be raised as the tax refunds “took place from 2005 to 2006”. However, she proceeded to express her belief that the claims were not time-barred because the Respondent only became satisfied that the Corporate Restructuring and Financing Arrangement was a tax avoidance arrangement “on or about 7 April 2008”. Although it was not explicitly stated, it is clear that what Ms Ng was alluding to was s 29(1)(c) of the Limitation Act (Cap 163, 1996 Rev Ed) (“the LA”), which provides that where an action is for relief from the consequences of a mistake, the period of limitation would not begin to run until the plaintiff had discovered the mistake or until such as he could, with reasonable diligence, have discovered it. This application was granted on 9 June 2014 and the writ was duly served.

9 The Appellant entered an appearance in the Suit and subsequently filed Summons No 3709 of 2014 (“the Setting Aside Application”), seeking to set aside the order granting the Respondent leave to serve the writ out of jurisdiction. This application was supported by an affidavit affirmed by Ms Ong Swee Took (“Ms Ong”) on 31 July 2014, in which she expressed the view that the Respondent knew that his claim was time-barred but deliberately sought to conceal this fact by omitting crucial details. She pointed out, among other things, that the Respondent had not disclosed that he had carried out an audit of the Subsidiary nor had he provided details of how the Respondent came to form the view that the Corporate Restructuring and Financing Arrangement was a tax avoidance arrangement or the reasons why the Respondent decided to exercise its statutory powers the way he did.

10 In response, Ms Ng affirmed an affidavit on 28 August 2014 (“Ms Ng’s 2nd Affidavit”) in which she reiterated that the Respondent had only come to the conclusion that the Corporate Restructuring and Financing Arrangement was a tax avoidance arrangement “on or about” 7 April 2008. She further explained that the audit was commenced in July 2007 but that documents had been sought from the Company all the way through to the end of March 2008. In a crucial passage, Ms Ng then deposed as follows:

11 Thereafter, the parties continued to file further affidavits in which they continued to maintain their respective positions: namely, the Respondent asserted that he had only formed the view that there was a tax avoidance arrangement after the Advice had been received in April 2008; while the Appellant maintained that this view must have been formed earlier. On 9 September 2014, the Appellant served a notice demanding production of the Advice, but the Respondent resisted production on the ground of privilege. On 23 September 2014, the Appellant filed Summons No 4769 of 2014 (“the Production Application”) to seek production of the Advice for inspection under O 24 r 11 of the Rules of Court (Cap 322, R 5, 2014 Rev Ed) (“the Rules”).

12 The Judge noted that, even before the passage of the Evidence (Amendment) Act 2012 (Act 4 of 2012) (“the 2012 Amendment Act”), it had always been clear that under the common law, privilege attached to communications exchanged between an entity and its in-house legal advisers in so far as these communications related to legal, rather than administrative, matters. The only qualification was that the two requirements of independence (of the in-house counsel) and confidentiality (of the communications) had to be satisfied. He held that there was “no reason to think” that these two requirements had not been satisfied in this case.

13 Turning to the issue of whether privilege had been waived, the Judge held that neither the so-called “state of mind exception” nor the doctrine of “disclosure waiver” applied on the facts. In respect of the former, the Judge held that it was “not apparent” that the [Respondent had] put his state of mind in issue. In respect of the latter, he held that that the Respondent had not disclosed any part of the Advice which might be said to be a misrepresentation of the document as a whole. We pause to note that while the Judge later said (with respect, perhaps somewhat confusingly), that he “[did] not have to decide on the issues concerning the state of mind exception and disclosure waiver”, it is clear to us that what he had in mind was the applicability of these two legal doctrines in the local context and he did not seek to resile from his earlier decision, which was that neither of them applied on the facts.

14 Finally, the Judge dealt with the argument that the Respondent had waived privilege by deploying the Advice through the references made to it in Ms Ng’s 2nd Affidavit. He first began by observing that it was “undisputed that mere reference to the fact that legal advice has been taken does not in itself waive privilege because the fact that legal advice has been taken is not of itself privileged” [emphasis in original]. However, he noted that it was unclear if (a) the Respondent was advancing the contention that “the Advice was to the effect that the [Appellant] was in a tax avoidance arrangement” and that he had relied on the contents of the Advice in reaching that conclusion or (b) whether the “timing of the advice received merely form[s] part of a sequence of relevant events which provide[s] the backdrop” and therefore there had only been a reference to the fact that the Advice had been received.

15 Ultimately, the Judge noted (in the penultimate paragraph of his judgment) that the burden of proof lay on the Appellant to demonstrate that it was necessary for the court to order production of the Advice. It was evident that he did not think that this burden had been discharged, for he made no order for production and instead put the Respondent to an election (see above at [2]). Dissatisfied, the Appellant sought and was granted leave to appeal on 9 October 2014. It filed its notice of appeal shortly thereafter.

16 Before we proceed to examine the arguments advanced on appeal, we will first explain why we did not agree with the Appellant’s contention that in making the order to elect, the Judge had “effectively allowed the [Comptroller] to determine for himself if he had waived privilege” [emphasis in original]. The short answer to this submission is that in deciding not to make any order for production of the Advice, the Judge had essentially dismissed the Production Application. In Sinwa SS (HK) Co Ltd v Nordic International Ltd and another [2015] 2 SLR 54 at [38], which concerned an application for summary judgment, this court held that the making of “no order” amounted to a dismissal of the application. By parity of reasoning, in declining to grant the order for production, the Judge had clearly made “no order” and had thereby dismissed the Production Application. Rightly or wrongly (and in our judgment he did not err), he had made his decision, and it was binding on the parties until reversed on appeal. This then raises the question: what was the Judge attempting to achieve with the order to elect? We shall explain this in greater detail at a later part of our grounds.

17 For now, we will first summarise the parties’ arguments on the substantive issues. The Appellant contended that the Advice was not covered by privilege as it was created before legal professional privilege extended to advice made by in-house counsel. Alternatively, it was argued that privilege had been waived either because the Respondent had “deployed” the Advice to advance a position: namely, that the claims in S 350/2014 were not time-barred because the Respondent “discovered that the arrangement was a tax-avoidance arrangement only on or after 3 April 2008” or because the so-called “state of mind exception” to legal professional privilege should apply in Singapore. Finally, and on the premise that privilege was not a barrier to production, it was argued that the production of the Advice was necessary under the Rules because it was probative of whether the claims in the Suit were time-barred and whether the Respondent had provided full and frank disclosure in its application for leave to serve the writ out of jurisdiction.

18 The Respondent – not surprisingly – took a contrary position on all these issues. He submitted that privilege attached to the Advice and that such privilege had not been waived by the references made in Ms Ng’s affidavits. He contended that the Respondent had only ever referred to the fact that the Advice had been made and not to its contents and that this was insufficient to constitute a waiver. He also submitted that the “state of mind exception” should not be accepted as a part of our law because it constitutes a “fundamental inroad into the sanctity of privilege”. Finally, he submitted that the Advice was not necessary for the fair disposal of the action and, hence, production should not be ordered, even if the Advice was not found to be privileged.

19 Against that background, the following four broad issues arose for our decision:

(a) whether legal professional privilege extended to communications exchanged with in-house counsel;

(b) whether privilege had been waived by reason of the references made to the Advice in Ms Ng’s 2nd Affidavit;

(c) whether the state of mind exception applied here; and

(d) whether disclosure of the Advice was necessary for the fair disposal of the Production Application.

20 After careful consideration of the competing arguments, we held that the Advice was protected by privilege and that no applicable exception to privilege existed in this case (we use the expression “exception” loosely to cover both the concept of waiver and the “state of mind exception”). Thus, the final issue – whether production was necessary for the fair disposal of the Production Application – did not arise for discussion. We now proceed to set out our reasons with regard to the first three issues in turn.

21 Legal professional privilege is a statutory right which is enshrined in ss 128 and 131 of the EA. Taken together, they cover both legal advice privilege as well as one element of litigation privilege (see the decision of this court in Skandinaviska Enskilda Banken AB (Publ), Singapore Branch v Asia Pacific Breweries (Singapore) Pte Ltd and other appeals [2007] 2 SLR(R) 367 (“Skandinaviska”) at [27]). They provide as follows:

22 Section 128 of the EA (“s 128”) enjoins anyone who might be an “advocate or solicitor” from revealing confidential communications exchanged with his client without his client’s consent whereas s 131 of the EA (“s 131”) protects a client from having to disclose any confidential communication that might have been exchanged between him and his “legal professional advisor” outside of specified circumstances. It is clear that privilege cannot be asserted on the basis of s 128, because in-house counsel are not advocates or solicitors in the sense originally intended pursuant to that particular section (see also below at [23]–[24]). The question, therefore, is whether they may be considered “legal professional adviser[s]” within the meaning of s 131. Linguistically, it would appear to be possible. However, when one examines the history of the provision, this argument, in our view, loses much of its force (see also below at [23]–[24]).

23 The EA traces its history to the Indian Evidence Act of 1872 (Act I of 1872) (India) (“Indian EA 1872”) (see Skandinaviska at [28]–[30]). Section 126 of the Indian EA 1872 (the legislative precursor of s 128 of our EA) enjoined any “barrister, attorney, pleader, or vakil” from disclosing any communication made to him by his client in the course of his employment. These four occupations were those which were available to those who practised law in India at the time (pleaders and vakils being legal occupations which were specific to the Indian legal practice). Section 129 of the Indian EA 1872 (the predecessor to s 131 of our EA), however, uses the composite expression “legal professional adviser”. In a previous edition of his treatise on the law of evidence, Prof Jeffrey Pinsler explained that Sir James Fitzjames Stephen, the drafter of the Indian EA 1872, intended to use the expression “legal professional adviser” as shorthand for the four occupations listed individually in s 126 of the Indian EA 1872 (see Evidence and the Litigation Process (LexisNexis, 3rd Ed, 2010) (“Pinsler 2010”) at para 14.17).

24 What this would appear to suggest is that s 131 only applied to advocates or solicitors (namely, those who were in independent legal practice), as these were occupations specifically mentioned in s 128. On this reading, in-house counsel fall outside the ambit of the EA, for they are not in practice per se (see Pinsler 2010 at para 14.18). We do not find this surprising. The EA was enacted in 1893 and it reflected the legal reality which existed then, when almost all legal advisers were in independent practice. Hence, it would be an understatement of the highest order to note that that statute could not possibly have provided for legal communications proffered by in-house counsel simply because in-house counsel (at least in the form we understand it today – that is to say, legal professionals who are salaried employees) did not exist then. Furthermore, as pointed out in Chen Siyuan and Lionel Leo, The Law of Evidence in Singapore (Sweet & Maxwell, 2016) at para 8.034, if in-house counsel fell within the ambit of s 131 there would have been an anomaly. Clients would have a right to refuse to disclose confidential communications exchanged with in-house counsel by virtue of s 131 but there would be no corresponding provision prohibiting in-house counsel from disclosing communications exchanged with their clients as s 128 did not apply to them.

25 While the position under statute was unclear, it was common ground that the common law has long protected confidential communications with in-house legal counsel. The only dispute was whether this common law rule applied in Singapore before the enactment of the 2012 Amendment Act, given the existence of the EA. In our judgment, there is no reason, in principle, why the common law rule ought not to have applied in Singapore prior to the enactment of the 2012 Amendment Act and, for the reasons set out below (at [30]–[42]), the existence of the EA does not detract from this proposition.

26 We will first begin by sketching the contours of the common law rule. The leading decision is that of the English Court of Appeal in Alfred Crompton Amusement Machines Ltd v Customs and Excise Commissioners (No 2) [1972] 2 QB 102 (“Alfred Crompton”) where Lord Denning MR said as follows (at 129A–D):

This is also the position across the Commonwealth: see, for example, the decision of the High Court of Australia in Waterford v Commonwealth of Australia (1987) 71 ALR 673 (“Waterford”), which discussed this in the specific context of legal officers in government employment (at 677), and the decision of the Supreme Court of Canada in R v Campbell [1999] 1 SCR 565.

27 However, the common law rule was not without limits. First, it was not so liberal as to cover any communication by in-house counsel of any stripe. In our view, there must be reasonable limits placed on the scope of the common law rule which are consistent with the underlying rationale and spirit of the doctrine of legal professional privilege itself. Under the common law, it has always been well established that privilege only attaches to communications between a client and a legal professional (see the UK Supreme Court decision of Regina (Prudential plc and another) v Special Commissioner of Income Tax and another (Institute of Chartered Accountants of England and Wales and others intervening) [2013] 2 WLR 325 at [30] per Lord Neuberger of Abbotsbury PSC). We stress that what is important is not the in-house counsel’s designation, but whether he possesses professional legal expertise. In the Singapore context, this would usually mean that he is a “qualified person” who is eligible to be admitted as an advocate and solicitor pursuant to the Legal Profession Act (Cap 161, 2009 Rev Ed) (“the LPA”). This does not necessarily preclude foreign qualified lawyers (although it seems to us that such counsel must not provide advice in contravention of the relevant provisions of the LPA). However, as this particular issue did not arise in the present case, we will not express any definitive view, and will leave this to be decided when the matter is directly in issue.

28 Second, the limits of the scope of the common law rule would also include the proposition, echoed in the judgment of Karminski LJ in Alfred Crompton at 136D that the in-house counsel must have been consulted qua legal adviser. As Lord Denning recognised in Alfred Cromptom at 129E–F, legal advisers can play either an executive or legal function and it is only communications made by them in their capacity as legal advisers that should be protected. What this means is that the communication must be in relation to a professional matter and the in-house counsel concerned must have been acting independently (or, as some cases put it, “professionally”) in rendering the legal advice in question (see Waterford at 677).

29 Third, it is only confidential communications which may be the subject of privilege. This requirement is not unique to communications with in-house counsel, but is a fundamental tenet of the law of privilege: “[i]t is axiomatic that there can be no privilege without confidentiality” (see HT SRL v Wee Shuo Woon [2016] 2 SLR 442 (“Wee Shuo Woon”) at [21] and Skandinaviska at [35]). For this reason, if the communications between the in-house counsel and its client were made publicly available (for example, if it was freely disseminated to the clients of the corporation) then the relevant confidentiality would not subsist and privilege cannot be claimed in respect of the advice.

30 We now turn to explain why we were of the view that the common law rule applied in Singapore. We begin with s 3 of the Application of English Law Act (Cap 7A, 1994 Rev Ed), which empowers the Singapore courts to apply the English common law, subject to such modifications as might be necessary for local circumstances. This provision reads as follows:

31 However, the caveat is that the common law must be “applicable to the circumstances of Singapore”. As the Advice in this case was proffered prior to the enactment of the 2012 Amendment Act, the issue arises as to whether the common law rule as to privilege with regard to in-house counsel was applicable in Singapore in general (and to the Advice in particular) in light of the EA. The relevant provision in this regard is s 2 of the EA, which reads:

32 It is clear from the language of s 2(2) of the EA that the common law rules of evidence continue to apply so long as they are not inconsistent with any of the provisions of the EA. Put simply, these common law rules continue to supplement the law of evidence in Singapore (which is contained primarily within the EA itself). As the Court of Three Judges observed in Law Society of Singapore v Tan Guat Neo Phyllis [2008] 2 SLR(R) 239 (at [117]), “[t]he EA only codified the law of evidence existing at the time of its enactment; therefore, new rules of evidence can be given effect to only if they are not inconsistent with the provisions of the EA or their underlying rationale” [emphasis added]. Applying that approach here, it is clear, contrary to what the Appellant submitted, that the common law rule relating to the privilege that is conferred upon advice proffered by in-house counsel is not inconsistent with any of the provisions of the EA. The common law rule is, in fact, wholly consistent with the rationale as well as spirit undergirding the existence of the doctrine of legal professional privilege.

33 The raison d’etre of legal professional privilege is that full, free, and frank communication between persons and their legal advisors, without which the effective administration of justice would not be possible, can only take place if such communications can be carried out in confidence (see Skandinaviska at [23]). This rationale applies with equal force to communications with in-house legal advisors, whose ability to provide legal advice would be greatly stymied if communications with their clients were not privileged (see also, the observations of Lord Denning in Alfred Crompton cited above at [25]). It would be artificial, unjust and unfair to draw a distinction between the advice proffered by in-house counsel on the one hand and advice proffered in the more traditional context of practice on the other.

34 What, then, is the effect of 2012 Amendment Act? In our judgment, it does not detract from the then existing position at common law as set out above. Let us elaborate.

35 As we explained above, before the 2012 Amendment Act was passed, it was unclear if in-house counsel fell within the ambit of s 131. Indeed, we note that the Law Reform Committee of the Singapore Academy of Law concluded in 2011 that “[t]here are serious doubts, however, whether legal advice privilege extends to communications with in-house counsel for the purposes of obtaining legal advice” (see Law Reform Committee, Singapore Academy of Law, Report of the Law Reform Committee on Reforming Legal Professional Privilege (October 2011) (Chairman: Harpreet Singh Nehal, SC) (“LRC Report”) at para 107). However, the authors of this report did not elaborate in any detail as to why they thought that the position was unclear, save to state the following (LRC Report at para 121):

36 The reference to “19th century conditions” in the observation quoted in the preceding paragraph is probably an allusion to the drafting history of the EA and, in particular, to the fact (as noted above at [24]) that it understandably dealt only with circumstances and conditions that existed at a particular (and much earlier) given point in time. The authors of the report proposed, after careful consideration of the position at the common law and a review of the positions taken in different jurisdictions, that legal professional privilege should extend to communications with in-house legal advisors provided the advice in question was rendered by a qualified lawyer and the requirements of confidentiality (of communications) and independence (of the legal advisor) are met (see LRC Report at paras 125–126). This was the genesis of the 2012 Amendment Act, which swiftly followed.

37 The 2012 Amendment Act wrought significant changes to the EA but for present purposes, we can focus on the two which relate to the matter in issue. The first was the introduction of s 128A of the EA which set out for in-house counsel (which the EA now refers to as “legal counsel in an entity”) what s 128 had always provided in respect of advocates or solicitors: namely, that they were not to disclose any communication made to them “in the course of and for the purpose of [their] employment”. The second was the introduction of s 131(2) of the EA. This new definitional sub-section provided that the expression “legal professional adviser” was to mean either (a) an advocate or solicitor (covered by s 128) or (b) a legal counsel in an entity (provided that he was acting in his capacity as counsel at the time and dispensing legal advice). Taken together, these provisions have the effect of extending legal professional privilege as it was codified in the EA to communications exchanged with in-house counsel.

38 However, the question that needs to be asked is this: Was the 2012 Amendment Act enacted because Parliament was of the view that privilege did not extend to communications with in-house counsel before the 2012 Amendment Act but should (whether because the common law fell afoul of s 2(2) of the EA or otherwise – a point we have already expressed our views on above) or because Parliament simply wanted to put the matter beyond doubt and clarify that it did. Understandably, this was a point which divided the parties. The Appellant argued that it was the former, and pressed the point that the 2012 Amendment Act could not be “applied retrospectively” to the Advice, which had been rendered in 2008. The Respondent submitted that it was the latter and argued that the 2012 Amendment Act merely codified what had always been the position under the common law.

39 In order to answer this question, we need to consider what precisely was said during the relevant parliamentary proceedings. One of the stated purposes of the 2012 Amendment Act, as disclosed in the Explanatory Statement to the Evidence (Amendment) Bill (No 2/2012) (“the Bill”), was to “extend legal professional privilege to legal counsel” (at p 20). In the speech he delivered during the Second Reading of the Bill, the Minister for Law, Mr K Shanmugam (“Mr Shanmugam”) stated as follows (see Singapore Parliamentary Debates, Official Report (14 February 2012) vol 88 at col 1128 (Mr K Shanmugan, Minister for Law) (“Amendment Act Debates”)):

40 Immediately after Mr Shanmugam delivered his speech, Mr Hri Kumar Nair (Member of Parliament for Bishan-Toa Payoh) rose to address the House and he observed as follows (see Amendment Act Debates at col 1129):

41 Predictably, each party to the present proceedings focused on that part of the debates which was most advantageous to its case. The Appellant focused on the references to the 2012 Amendment Act “extend[ing]” legal professional privilege to in-house legal counsel to support its argument that legal professional privilege did not extend to communications made with in-house legal advisors prior to the passage of the 2012 Amendment Act. The Respondent, on the other hand, submitted all that was meant by this was that the ambit of the statutory right to legal professional privilege under the EA was being extended to in-house counsel where previously it did not. However, the Respondent stressed that it was clear from Mr Hri Kumar Nair’s speech that the common law had long “filled that gap”.

42 With respect, the relevant statements on the subject in the Parliamentary Debates (as set out above) are, at best, neutral. There are arguments going both ways and this reflects the general conceptual confusion and lack of clarity that beset the operation of s 2 of the EA, which has given rise to an innumerable number of difficulties. However, we are of the view that, even if it were the case that Parliament thought that the common law did not apply, this would be irrelevant. At the end of the day, it is quintessentially the function of the judiciary to state what the law is. As we pointed out above, the EA only codified the law of evidence as it stood at the time of its enactment. The crux of the issue is whether the common law rule that privilege extends to in-house counsel applied in Singapore before the 2012 Amendment Act. In our judgment, the answer to that question is in the affirmative. The common law that recognised that privilege attaches to communications by in-house counsel and that existed prior to the enactment of the 2012 Amendment Act was applicable in the Singapore context. Of course, that common law position has now been superseded by the 2012 Amendment Act, and the EA henceforth governs the assertion of privilege over communications exchanged with in-house counsel.

43 The question which then arose was whether the Advice possessed the prerequisites necessary for privilege to be claimed. As noted above at [27]–[29], there are three requirements: (a) the Advice must have been rendered by a legal professional; (b) the legal professional must have been acting qua legal adviser when he/she provided the advice; and (c) the communications must have been made in confidence. It was common ground between the parties that the party who is asserting privilege (in this case, the Respondent) has the burden of proving that that the aforementioned preconditions have been satisfied. The Appellant submitted that this burden had not been discharged because the Respondent had not adduced any positive evidence either of the confidential nature of the communications or of the independence of the lawyers who rendered the advice. We did not agree.

44 In Australian Hospital Care (Pindari) Pty Ltd v Duggan (No 2) [1999] VSC 131, the Supreme Court of Victoria was faced with a summons for the production of certain documents which had been produced by a qualified lawyer who was the general counsel of a company. Gillard J stressed that in order for privilege to attach to documents produced by in-house counsel, the party asserting privilege had the burden of proving that the requirements of independence and confidentiality had been satisfied. However, he went on to explain that this burden was discharged, at least in the first instance, by the swearing of an affidavit in which privilege is asserted because the assertion of privilege implies also the assertion that the requirements for privilege to subsist had been satisfied, namely, the communications were made confidentially and the legal adviser had exercised independent judgment in rendering the advice (at [62]).

45 Gillard J stressed that the law did not require a ritual of formalities. In particular, there was no need for a party to include a specific averment that the legal advisor was acting independently (at [65]). At [67]–[68], he stressed:

46 Of course, if the court is not satisfied with this, it is always open to the court to look behind the affidavit to the documents themselves to ascertain if privilege was rightly asserted and the court would reach a decision after examining the evidence (at [71]). This approach was also applied by the Supreme Court of Queensland in Aquila Coal Pty Ltd v Bowen Central Coal Pty Ltd [2013] QSC 82 where it was explained that there ought not to be any presumption that in-house counsel lacked independence (at [9]). Ultimately, the question is whether the counsel in question brought a disinterested mind to the subject matter of legal advice and whether the relationship between the parties was such as to secure the advice the independent character necessary for privilege to attach. This is a question of fact (at [11]).

47 We endorse this approach. In our judgment, this is in keeping with the general approach of the common law, which has been to treat in-house counsel in the same way as lawyers in independent practice (see The Law of Privilege (Bankim Thanki QC gen ed) (Oxford University Press, 2nd Ed, 2011) (“The Law of Privilege”) at para 1.49). With that, we turn to the facts of this case, and to Ms Ng’s 2nd Affidavit in particular. As a starting point, it is important to remember that Ms Ng had filed this affidavit in response to Ms Ong’s allegation that the Respondent must have formed the view that the Corporate Restructuring and Financing Arrangement was a tax avoidance arrangement within the meaning of s 33 of the ITA before 7 April 2008 (see above at [9]–[10]). Ms Ng’s response was that the Respondent conducted an audit to inquire into the Subsidiary’s affairs in July 2007 as part of its “review of cases which involved claims (and payments of) significant amounts of tax refunds”. In the subsequent paragraphs of her affidavit, she set out in detail the steps which were taken in the audit into the “full circumstances of the [Corporate Restructuring and Financing Arrangement]”. She explained that after collecting all the relevant documents, the Respondent sought advice from its Law Division (see above at [10]).

48 Viewed against this background, it is clear that the Respondent had approached the Law Division for legal advice – this much was clear from the context, as Ms Ng was trying to explain (specifically) how the Respondent had come to form the view that there was a tax avoidance arrangement within the meaning of the ITA. From the manner in which Ms Ng’s 2nd Affidavit was structured, there was no suggestion that the Law Division (and the Appellant never suggested otherwise) was involved in any of the investigative aspects of the task. Instead, its role appeared to us to be limited only to the rendering of legal advice. In our judgment, these points, when coupled with the proviso at the end of para 17 of Ms Ng’s 2nd Affidavit where she asserted that privilege subsisted in the Advice and had not been waived, sufficed to present a prima facie case that the Advice was privileged.

49 To be fair, more could perhaps have been done, and it was clear that the Respondent realised this as well, for in a later affidavit sworn by Ms Ng on 27 April 2015 (“Ms Ng’s 9th Affidavit”) in reply to a separate application taken out by the Subsidiary for discovery of, among other things, the Advice, she included the following paragraph in her affidavit:

50 The Appellant had said much about the circumstances under which this affidavit came to be filed and whether it was proper for this to be taken into account for the purposes of the present appeal, but we will put that aside for the present. The point, it seemed to us, was simply this. The position in the authorities suggest that while the legal onus lies on the party asserting privilege to demonstrate that the preconditions for privilege to subsist are present, the bar (at least to make out a prima facie case) is not high and we were satisfied that it had been crossed in this case. The Appellant had not – in response – adduced any material to rebut this prima facie case. It had not said, for example, that the officers from the Law Division of IRAS might not have been acting in their capacity as professional legal advisers in rendering the Advice, nor had it been asserted that the Advice had not been communicated confidentially. If the Appellant’s complaint was simply that there were no specific averments in Ms Ng’s 2nd Affidavit that the communications were “confidential” and that the Law Division was “independent”, then it seemed to us that the objection was not so much one of substance as it was about form and we would reject it. For good order, we accept that a paragraph such as that which was found in Ms Ng’s 9th Affidavit should have been included in her 2nd Affidavit, but we preferred to take a broader view of things in this case and accepted that the requirements for the assertion of privilege were satisfied.

51 We now turn to discuss the issue of waiver. The starting point is that privilege is understood as “a right to resist the compulsory disclosure of information” (see the New Zealand Privy Council decision of B and others v Auckland District Law Society and another [2003] 3 WLR 859 (“B v Auckland”) at [67] per Lord Millett). Like all rights, it may be waived (see Phipson on Evidence (Hodge M Malek gen ed) (Sweet & Maxwell, 18th Ed, 2013) (“Phipson”) at para 26-01). This can be done either expressly or, as is more common, it may be done impliedly, such as where a client sues his solicitor (see, for example, the English Court of Appeal decision of Lillicrap and another v Nalder & Son (a firm) [1993] 1 WLR 94 at 98B). We are concerned, for present purposes, with implied waiver. Given that there is no local authority which has discussed the doctrine of the implied waiver of privilege at any length, we will first begin by summarising the applicable legal principles.

52 The law on the implied waiver of privilege is of tremendous complexity. The situations in which the courts have found there to be an implied waiver are multifarious, and it has been said that it is “difficult, if not impossible, to identify a coherent principle that underpins all the various circumstances in which a waiver of privilege can occur” (see Colin Passmore, Privilege (Sweet & Maxwell, 3rd Ed, 2013) (“Passmore on Privilege”) at para 7-002). For that reason, many commentaries approach this issue by dividing the cases according to the categories of situations in which previous courts have found that a waiver has occurred (see, for example, The Law of Privilege at para 5.01). We will also adopt this approach, and refer only to that subset of cases which relates to the loss of privilege arising out of a reference to privileged material because of a step taken in litigation. The cases in this area usually concern references made in court documents filed before a trial (generally, in affidavits or in pleadings) or references to privileged material in the course of the trial itself.

53 It is well-established that not all references to legal advice in court documents would amount to a waiver of privilege (see, for example, the English Court of Appeal decision of Roberts v Oppenheim (1884) 26 Ch D 724 (“Roberts”) at 735 per Fry LJ). However, there are circumstances where such a reference will give rise to a waiver of privilege on the ground that the making of such a reference is inconsistent with the subsequent maintenance of privilege and that it would consequently be unfair, in those circumstances, to allow the party to withhold production of the legal advice. The difficulty lies in determining precisely when this takes place. We will summarise the two approaches presented by the parties before turning to discuss what we consider to be the appropriate approach.

54 The traditional test, which the Respondent submitted ought to apply here, is that in order for there to be a waiver, it is insufficient to show merely that the party has referred to the fact that legal advice has been given; instead, it must be shown that the contents of the legal advice have been relied on. This test has been adopted throughout the Commonwealth (see the cases cited in Passmore on Privilege at para 7-151) and we shall refer to this as the “fact/contents distinction”. A succinct statement of this doctrine may be found in Phipson, where the learned editors put it as follows (at para 26-05):

55 The basis of this distinction, as it emerges from the extract above, is that privilege only subsists in respect of the confidential content of the communications between a client and his solicitor. The mere fact that such communications had taken place is not in itself privileged and a reference to the receipt of legal advice does not therefore give rise to any basis for waiver. However, in many of the earlier cases, a somewhat different reason was articulated. The concern was that deponents are often called to give details of the source of their information and beliefs as a pro forma requirement in an affidavit or deposition, and it was thought that compliance with such procedural requirements could never give rise to a waiver (see the decision of the High Court of Australia in Attorney General for the Northern Territory v Maurice and others (1986) 69 ALR 31 at 42–43 (“AG (NT) v Maurice”) per Deane J). In the English Court of Appeal decision of Government Trading Corporation v Tate & Lyle International Ltd (Unreported), 19 October 1984, Goff LJ (as he then was) put it in the following terms:

56 The trouble, however, is that the line between a mere reference to the “fact of legal advice” and “reliance on its contents” is so slippery that it is almost impossible to apply. Take, for example, the decision of the English Court of Appeal of Marubeni Corporation v Alafouzos (Unreported), 6 November 1986 (“Marubeni”). The plaintiffs sought leave to serve a writ out of jurisdiction on defendants in Japan in respect of a claim for an unpaid contractual debt. Appreciating that the defendants might raise a defence of illegality, the plaintiffs’ solicitor affirmed an affidavit in which he said, “the Plaintiffs have obtained outside Japanese legal advice which categorically states that this agreement does not render performance of the sale contract illegal in any way whatsoever”. The application was granted. The defendants then sought to set aside the writ and also applied for the production of the Japanese legal advice. The High Court granted the application for production but this decision was reversed on appeal. Lawton LJ explained his decision in the following way:

Lloyd LJ, the only other member of the court, put the point as follows:

57 With the greatest respect to the English Court of Appeal, it is difficult to understand how it could be said that the plaintiffs were not relying on the contents of the Japanese legal advice. What the plaintiffs were saying was that any putative defence of illegality that might be raised by the defendants would be unmeritorious based on the authority of what was said in the advice they had received from their Japanese counsel. This would seem to be a reference to the contents of the legal advice. Separately, we note that what had begun as a dichotomy between “fact” and “contents” has now spawned at least two further pairs of distinctions: the first is a distinction between “reference” to the advice (which is acceptable) and “reliance” on it (which is not); the next pair of distinctions is between reference to the “effect” of a document (which is permissible) and a reference to its “contents” (which is not). Of these two pairs of distinctions, we have particular discomfort with the second, viz, the effects/contents distinction. It seems to us that a reference to the effect of privileged material usually (and we acknowledge that each case must turn on its particular facts) entails an implicit reference to its contents and, on the facts of Marubeni, it most certainly did. We will return to this point later.

58 The other approach, which was favoured by the Appellant, is to draw a distinction between cases of “deployment” (which results in an implied waiver of privilege) and cases where there had been a mere “reference” to the privileged document (where there is no waiver). This distinction was drawn by the English Court of Appeal in Dunlop Slazenger International Ltd v Joe Bloggs Sports Ltd [2003] EWCA Civ 901 (“Dunlop”) at [11], where the issue was whether there had been a waiver of privilege over an expert report by virtue of a reference made to the report in an affidavit. However, the test of “deployment” was first articulated in the oft-cited dictum of Mustill J (as he then was) in the English High Court decision of Nea Karteria Maritime Co Ltd v Atlantic & Great Lakes Steamship Corpn and others [1981] Com LR 138 where he stated as follows (at 139):

59 This passage was expressly approved by the English Court of Appeal in Great Atlantic Insurance Co v Home Insurance Co and others [1981] 1 WLR 529 (“Great Atlantic”), which is often held up as a paradigmatic case of implied waiver. In that particular case, the plaintiff’s counsel had read out the first two paragraphs of a privileged document at the opening of the trial. It later became clear that what had been read out was only a part of the document and the defendants sought production of the rest on the ground that privilege had been waived. The court agreed with the defendants and ordered production. Templeman LJ (as he then was) explained that the selective disclosure of privileged material was not permitted as it was “unfair or misleading” and risked leaving the court and the other party with a partial and misleading picture of the purport of the document (at 538E–539A). This is commonly referred to as the “cherrypicking” principle (see Dunlop at [11]).

60 It is clear from the foregoing that the selective deployment of privileged material is not permitted on the ground of fairness. However, this begs the question: When will material be taken to have been relevantly “deployed”? Originally conceived, the concept of “deployment” was used in relation to the introduction of material into evidence at trial: in the Great Atlantic at 537H, Templeman LJ wrote about the “deliberate introduction by the plaintiffs of part of the memorandum into the trial record”. The concept of “deployment” is particularly suited to the trial context because deployment is effected by the entry of the material into the trial record. However, outside of the trial process, it is difficult to determine when a mention of privileged material in an affidavit or a pleading crosses the line from mere “reference” to “deployment”. Furthermore, it seems to us that the difficulty is not just practical, but conceptual as well. The concept of “deployment” is really just a statement of a legal conclusion (that the reference in question amounts to an implied waiver of privilege) rather than a useful aid to analysis. It does not enable the court to decide exactly when a party would be taken to have impliedly waived privilege over a document.

61 The Appellant contended that the test ought to be “whether the party asserting privilege has deployed the contents of the privileged document in order to advance a position in the relevant cause or matter” – if this was the case, privilege would be taken to have been waived. However, with respect, this test is at once too wide and too narrow. It is too wide because it fails to capture the essence of the decision in Great Atlantic, which is that it is not just the use of the privileged material per se which is objectionable, but the use of it in circumstances when it might give rise to an incomplete, partial, and therefore unfair picture of events. It is also too narrow because the question of whether there has been waiver cannot simply be governed by the purpose of the reference. For example, in the Great Atlantic, the plaintiffs’ counsel had wrongly (though excusably) read out the opening two paragraphs of the memorandum thinking that it had to be disclosed and not knowing that it was an incomplete part of the memorandum (at 537D–E). In these circumstances, it is difficult to say that the plaintiffs had sought to achieve any particular purpose by reading those passages – it was simply a mistake. In our view, the deployment/reference distinction is little better than the fact/contents distinction as a guide to determining whether there has been a waiver.

62 In our judgment, neither of the two approaches suggested by the parties is particularly satisfactory. We think it is useful to recall Lord Goff’s warning against what he called the “temptation of elegance” (see Robert Goff, “The Search for Principle” (1983) 69 Proceedings of the British Academy 169 (reprinted in William Swadling and Gareth Jones (Eds), The Search for Principle – Essays in Honour of Lord Goff of Chieveley (Oxford University Press, 1999), pp 313-329 at p 318). We should not, in the rush to develop an elegant and pithy test, reduce the principle of implied waiver to a few reductive dichotomies – for example, fact/contents; reference/deployment; effect/contents; reference/reliance – when the inquiry as to whether there has been an implied waiver is one which is complex, nuanced, and textured.

63 As the United Kingdom Employment Appeal Tribunal cautioned, after an extensive review of the authorities in England and Australia, in Brennan and others v Sunderland City Council and others [2009] ICR 479 (“Brennan”) at [65], words used in judgments do not have the lapidary quality of a statutory enactment. The various expressions used by the courts shade into each other and the application of the doctrine of implied waiver cannot depend merely on the outcome of a “labelling exercise”. For this reason, there has been a shift away from an application of these rigid tests towards a single inquiry, which is whether fairness demands disclosure. As the court put it at [67] of Brennan:

64 Having reviewed the authorities for ourselves, we agree that fairness is the theme that runs through the authorities on implied waiver in all its various forms, whether it be waiver in the context of references made in court documents, “deployment” in the course of litigation, or the waiver of privilege that is imputed by operation of law when a client sues his solicitors. However, as the Employment Appeal Tribunal clarified at [63] of Brennan, it is not “fairness at large” that drives the operation of the doctrine of implied waiver. Mere unfairness, in the sense of a disadvantage accruing to one side due to the withholding of information on the ground of privilege, can never be the touchstone by which the court determines whether there has been an implied waiver of privilege (see the English Court of Appeal decision of Paragon Finance plc (formerly National Home Loans Corporation plc) and others v Freshfields (a firm) [1999] 1 WLR 1183 at 1194 (“Paragon Finance”) per Lord Bingham CJ (as he then was)). There is always some unfairness in allowing information to be withheld on the ground of privilege but in so far as there is a tension between disclosure and privilege, the balance, as pointed out by the Privy Council in B v Auckland at [51], has long been struck in favour of the preservation of privilege.

65 The doctrine of implied waiver has always been concerned with fairness of a very particular sort. The principle of the matter, simply put, is that a party cannot have his cake and eat it. If a party voluntarily puts privileged material before the court, he cannot rely on the advantageous aspects of it to advance his case but claim privilege in respect of the other less advantageous aspects of the documents for fear that it might damage his case. In AG (NT) v Maurice, Deane J explained it as follows (at 42–43):

66 It has been suggested that the approach in Australia has evolved, and that the touchstone is now inconsistency rather than fairness (see Phipson at para 26-04). In Mann v Carnell (1999) 168 ALR 86 (“Mann”), the majority of High Court of Australia explained that it was important to remember legal professional privilege existed for the benefit of clients – to protect the confidentiality of communications they exchanged with their lawyers – and it was only the client who could give up that entitlement. This may be done expressly, or it may be done impliedly, where the client acted in a manner which was “inconsistent with the maintenance of the confidentiality which the privilege is intended to protect” (at [29]). One example of this was where a client gave evidence during legal proceedings concerning instructions she gave her barrister. This was conduct which was inherently inconsistent with the maintenance of confidentiality, for matters in legal proceedings are a matter of public record and a reference to privileged material in such a context constituted an implied waiver of privilege. It did not matter that the client did not subjectively intend to waive privilege; all that mattered was the inconsistent of her actions. At [29], they put the matter as follows:

67 In our judgment, the concepts of inconsistency (in the Mann sense) and fairness (in the restricted sense in which it is used in Brennan) are not in tension, but instead work in tandem. In every case of implied waiver there must be inconsistency, for the essence of implied waiver is the implicit relinquishment of a right through inconsistent conduct. However, it is not every case of inconsistent conduct that will warrant a finding of implied waiver. Taken too far, this could lead to the conclusion that all references to the fact of the receipt of legal advice (see above at [55]) might lead to an implied waiver. This was certainly not what was intended and it is notable that the majority in Mann did not give any indication that they intended to depart from the previous authorities on this subject. That was why the High Court of Australia was careful to stress that, where necessary, the inquiry would be “informed by considerations of fairness” (at [29], and cited above at [66]). What was meant by this, in our view, is that not all instances of inconsistent conduct lead to an implied waiver of privilege – it is only where fairness demands disclosure that the law bars reliance on privilege. Likewise, it is not mere unfairness in a broad sense that justifies waiver. As we said at above, the unfairness that justifies a finding of implied waiver is of a very particular sort: it is the unfairness that arises from the inconsistency of the posited act with the subsequent maintenance of privilege that impels a remedy (see above at [64]). Much depends, in the final analysis, on the precise facts and context concerned (see also below at [69]).

68 In our judgment, McLachlin J (as she then was) summarised the point elegantly when she stated, in the Supreme Court of British Columbia decision of S & K Processors Ltd and another v Campbell Ave Herring Producers Ltd and others [1983] 4 WWR 762 at 764, that:

69 Given the importance of legal professional privilege, waiver is not to be easily implied – this is a point which is constantly reiterated in the authorities (see, for example, Brennan at [66]). A court tasked to determine whether there has been an implied waiver of privilege by reason of a reference made to privileged material should approach the matter by examining all the circumstances of the case including what has been disclosed (the materiality of the information in the context of the pending proceedings); the circumstances under which the disclosure took place (in particular, the position in the authorities appears to be that disclosures of privileged material during trial almost invariably results in a waiver); whether it may be said (albeit only as a relevant factor as opposed to a single test) that the party had “relied” or “deployed” the advice to advance his case; and whether it can be said that there is a risk that an incomplete and misleading impression had been given. This list is not exhaustive, and no one factor is determinative of the issue. Ultimately, the court should ask itself whether, in all the circumstances of the case, it may be said that – given what has already been revealed – fairness and consistency require disclosure. This is a fact-sensitive exercise of judgment and the inquiry is objective and not subjective: it is the objective role played by the legal advice which is relevant, not the subjective intention of the party who is asserting privilege so a profession that one is not relying on the contents of legal advice is no bar to the court finding that there has been an implied waiver (and see the English High Court decision of Digicel (St Lucia) Ltd (a company registered under the laws of St Lucia) and others v Cable and Wireless plc and others [2009] EWHC 1437 (Ch) at [31]).

70 One point which divided the parties was whether it was possible to consider the “purpose” for which the privileged material had been referred to. The Appellant argued that it was; the Respondent disagreed. It follows from what we have said that, in determining whether there has been an implied waiver, the court will have to look at the role that the legal advice plays in the context of the document in which reference to it was made. If this is the sense in which the Appellant talked about having regard to the “purpose” of the reference, then an inquiry as to “purpose” is clearly relevant. However, if what was meant was that the court should have regard to the subjective intentions of the party which made reference to the advice then we would disagree. As noted above, the inquiry as to waiver is objective and not subjective.

71 If the court concludes that there has been an implied waiver of privilege, it will then have to consider the extent of the disclosure required. In certain circumstances, disclosure of only a part of the document might suffice if that is all that is needed to correct the unfairness; in other cases, remedial action can be taken to obviate the need for disclosure (see Great Atlantic at 539H per Templeman LJ). For instance, if reference was made only to a part of a document which is clearly severable from the whole, then disclosure of just the implicated section might suffice to remedy the prejudice. In other instances, for example if the reference was inadvertent and the material had yet to enter the trial record, it might be possible for the references to be deleted and privilege to be preserved (see The Law of Privilege at para 5.34).

72 Looking at the matter against the background of these principles, we were satisfied that there had not been any implied waiver of privilege. The reference to the Advice was scant. There was certainly no reference to what was said in the Advice nor was there a deposition to the effect that it was because of the Advice that the Respondent formed the view that the Corporate Restructuring and Financing Arrangement was a tax avoidance arrangement. In our judgment, fairness and consistency did not require disclosure of the Advice in this situation. This case can be contrasted with the English High Court decision of Mid-East Sales Limited v United Engineering & Trading Company Limited, The Islamic Republic of Pakistan [2014] EWHC 892 (Comm) (“Mid-East Sales”), which the Appellant cited. There, default judgment had been obtained against the defendants and they subsequently sought to set it aside. In attempting to explain why the defendants had entered an appearance late, one Mr Piracha affirmed an affidavit to explain that, “acting on the advice of [its] solicitors”, the second defendant had returned the claim form and its enclosures to the British High Commission”. In a later affidavit, he reiterated that the decision had been “premised on incorrect legal advice” and that he would “demonstrate that it was a result of [its solicitor’s] advice” that the claim form had been returned. In response, the plaintiff sought an order for the production of the legal advice and it was granted. Males J explained his decision as follows (at [18]):

73 It is pertinent to note that the second defendant was the applicant in that case and the very premise upon which he sought to persuade the court to exercise its discretion to set aside the default judgment was that its delay in responding to the claim was attributable to the contents of legal advice which it had received. It seemed to us, therefore, that Mid-East Sales was a clear case in which, applying the legal framework we have set out above at [69], disclosure would have been necessary in the interest of fairness and consistency. Having relied on the advice in this way, it would have been inconsistent for the defendants to have refused production on the ground of privilege; furthermore, there was no way the plaintiff could have resisted the second defendant’s application, at least in so far as it was premised on the receipt of incorrect legal advice, if it did not have sight of the advice in order to verify what it said and to cross examine the second defendant on his reliance on its contents.

74 However, these were not the facts here. The present matter can usefully be looked at in terms of how it might proceed at the hearing of the Setting Aside Application. The Appellant would doubtless contend on that occasion that it cannot be inferred, from the fact that the Advice was rendered on 3 April 2008, that the Respondent only acquired the relevant knowledge (ie, that there was a tax avoidance arrangement) in the month of April 2008. That is surely correct. If the Respondent does not disclose the Advice, then he cannot invite the court to infer that the contents of the Advice were at all relevant to the acquisition of the relevant knowledge. As the Judge rightly pointed out (see above at [2]), the court cannot at present make any assumptions as to the contents of the Advice and therefore cannot rely on it to establish when the Respondent had formed the view that there was a tax avoidance arrangement. In other words, the references to the Advice in Ms Ng’s 2nd Affidavit are, for all intents and purposes, neither here nor there. As matters stood at the time of the hearing before us, we were satisfied that the Advice had not been specifically referred to in order to advance any particular point and the passing references to it in Ms Ng’s 2nd Affidavit were not sufficient – in and of themselves – to constitute a waiver of privilege.

75 This brings us to the subject of the Judge’s order to elect. Thus far, our analysis has been entirely consistent with that of the Judge, who likewise found that the Advice was privileged and that there was no basis to order production. However, this, then, begs the question: what, then, was the Judge trying to achieve with the order to elect? The Judge did not, with respect, explain this in the short oral judgment he delivered, but it was clear to us, from an examination of the record, that what he had in mind was that the Respondent should be asked to indicate his future position on the use of the Advice, in order that the issue of waiver might be placed beyond any doubt.

76 It is helpful to begin with the decision of the English Court of Appeal in Buttes Gas and Oil Co and another v Hammer and another (No 3) [1980] 3 WLR 668 (“Buttes”), which was cited to the Judge and which he clearly drew on in making his order. This is a seminal decision which traversed many areas of the law of privilege. (It was appealed to the House of Lords, which disposed of the appeal on the basis that the claims in question were non-justiciable and so the questions of privilege raised did not arise for decision: see Buttes Gas and Oil Co and another v Hammer and another and another appeal [1981] 3 WLR 787.) For present purposes, we can confine ourselves only to what was said in respect of legal professional privilege. The facts of that case are involved but it suffices to state that after the pleadings had been completed, the defendants in that case sought production of certain documents which had been referred to in the plaintiff’s pleadings. The plaintiff resisted the application on the ground of legal professional privilege and public interest privilege and, in response, the defendant contended that privilege had been waived by reason of the references made to the documents in the pleadings.

77 The English Court of Appeal was divided on the question as to whether there had been a waiver of privilege and, consequently, they approached the matter differently. Lord Denning MR opined that in referring to the documents in its pleadings, the plaintiff had evinced an intention to rely on them and should either make them available for production or strike out all references to them (at 683B). The other two members of the court disagreed. Both Donaldson LJ (at 688F–G) and Brightman LJ (at 703B) agreed that the mere reference to a document in a pleading did not amount to a waiver of privilege. At 703C, Brightman LJ remarked, obiter, that the plaintiffs would eventually have to decide if they wished to forego privilege in the documents by using them at trial (in which event they would have to produce it) or if they wished to abandon reliance on it. He observed that if this was not done in advance, an adjournment of the trial might have to be sought at the plaintiffs’ expense.

78 We agree with the majority of the court in Buttes that the mere reference in a pleading to a document does not constitute an automatic waiver of privilege over that document. As we noted above at [53], this has been the case at least since the 19th century decision of Roberts and it is consistent with the modern approach towards implied waiver, which applies the test of fairness and consistency. However, what is important to note for present purposes is that there were two different “election orders” proposed by the different members of the court:

(a) The first, which was espoused by Lord Denning MR, begins from the premise that there had been a waiver of privilege. In order to remedy the waiver, Lord Denning MR said that the party would be forced to elect: it would either have to produce the document or it would have to strike out the offending portions of its pleadings. This is consistent with the position in the authorities, which does not demand the production of a document the moment there has been a waiver of privilege. Instead, it is open to the party which has waived privilege to take appropriate remedial measures (see above [71]).

(b) The second, which was proposed by Brightman LJ, proceeded from the premise that there had not yet been any waiver of privilege. What he was saying is that even though there was no basis for an order of production to be ordered at this point, things would eventually come to a head and that, sooner or later, the plaintiff would have to decide whether it wanted to rely on the document or not, and if it were the former, it would of course have to offer the document for inspection. The election he envisaged was purely voluntary, for he explicitly stated that the question whether the defendants could force the issue by making an application to strike out the offending references to the document did not arise for decision (though he expressed the view that it was likely that the defendants would be able to force the plaintiff to decide through the bringing of such an application).

79 It is clear that the Judge did not put the Appellant to election in the sense that Lord Denning MR had proposed in Buttes, for the order to elect did not the take the form of “produce or strike out”. Rather, what the Judge said was that the Respondent should clarify, once and for all, whether he wished to rely on the Advice. If he did, he should disclose it; if he did not, he ought to state so on affidavit in order to put the matter beyond any doubt. In other words, what the Judge had in mind was an order along the lines suggested by Brightman LJ. The order to elect was, when viewed in this light, a prudential exercise in case-management – the Judge had clearly thought (and we agree) that it was sensible that the Respondent indicate his position in advance in order to save time and costs. In this connection, it is worth mentioning that the Judge had been docketed to hear this Suit and he had heard a multitude of interlocutory applications arising from it. It was not, as the Appellant submitted, an attempt to have the Respondent “determine for himself” the outcome of the Production Application.

80 When he appeared before us, Mr Davinder Singh SC (“Mr Singh”), counsel for the Appellant, took particular issue with the manner in which para 18 of Ms Ng’s 2nd Affidavit had been phrased. Paragraph 18 states that it was “[f]ollowing receipt and consideration of Law Division’s advice” that the Respondent had formed the view that there was a tax avoidance arrangement and issued the Additional Assessments to recover the tax refunds. Mr Singh contended powerfully that para 18 had been carefully crafted to allow the Respondent to argue that the effect of the Advice was that the Respondent was only able to conclude that there was a tax avoidance arrangement after 3 April 2008 (and that, therefore, the Suit was not time-barred). Furthermore, he argued that even if this particular argument was not run, the references to the Advice in para 18 of Ms Ng’s 2nd Affidavit were “insidious” and had the potential to colour the mind of the Judge who heard the Setting Aside Application. These were concerns, he submitted, which the Judge’s order to elect did not address.

81 As we understood his argument, his concern lay with the possible prospective use of the document. That was not a matter which we could adjudicate on at the time. The fact of the matter was that, as matters stood, there was no basis for imputing that there had been a waiver of privilege. That said, Mr Singh’s contentions were not without force. As we noted above at [57], we are uncomfortable with the effect/contents distinction drawn in the authorities. The decision of the Federal Court of Australia in Bennett v Chief Executive Officer, Australian Customs Service (2004) 210 ALR 220 (“Bennett”) is instructive in this regard. In that case, the respondent, the Australian Customs Service (“ACS”), was taking disciplinary action against the appellant, the President of the Customs Officers’ Association, for certain public remarks he had made. The ACS had sent him a letter proposing a settlement and in it, it was stated that it had been advised that the appellant was “not correct in asserting that he is not subject to the Act and Regulations [ie¸ the legislation governing the conduct of public service officers] if he makes public statements about Customs-related matters” and that the terms of the settlement were informed by this understanding. The appellant sought production of the advice under the Freedom of Information Act 1982 (Cth) and this was resisted on the ground of privilege. By a majority, the Full Court of the Federal Court of Australia ordered production.

82 As Tamberlin J explained at [5], this was a case in which the “substance and effect of the advice was being communicated in order to emphasise and promote the strength and substance of the case to be made against [the appellant]”. Thus, even though the reference might on one reading be seen as being limited only to the effect of the advice, it clearly constituted an implied waiver of privilege and disclosure was necessary. At [6], Tamberlin J put the matter as follows:

83 Likewise, in the present case, if the Respondent seeks to rely on the effect of the Advice to avoid falling within the proscription of a statutory time bar pursuant to the LA (for example, by saying that the effect of the receipt of the Advice was that the Respondent was able to form the view that there was a tax avoidance arrangement) then disclosure should be ordered. In those circumstances, it would only be fair that the Appellant and its Subsidiary, both of whom are defendants in the Suit should have sight of the Advice in order that they might have an opportunity to present a contrary case. It seems to us that it would be inconsistent and unfair, in that situation, for the Respondent to rely on the effect of the Advice to advance his case while simultaneously withholding it from disclosure on the ground of privilege.

84 However, it also seemed to us that this concern was allayed somewhat by what was said by the Respondent. In its written submissions, the Respondent stated that the events set out in Ms Ng’s 2nd Affidavit (including the fact that the Advice was received on 3 April 2008) served as “a summary chronology of events, which are nothing more than time-markers” [emphasis in original removed]”. What we understood by this was that the Respondent was not saying that he had relied on the Advice in reaching the conclusion that there was a tax avoidance arrangement. When he appeared before us, Mr Alvin Yeo SC, counsel for the Respondent, explained that there were two primary issues which would be ventilated at the Setting Aside Application. The first was when the audit had been completed. In relation to this issue, the Respondent would seek to rely on the fact that the Advice had only been received on 3 April 2008 to support his contention that the audit had only been completed on 24 March 2008. However, even if the Respondent were successful in this, he would still have to show that he had only formed the view that there was a tax avoidance arrangement sometime after the completion of the audit (and, crucially, sometime in the month of April 2008, rather than before that). On this second issue he candidly admitted that mere reference to the fact of the Advice having been received (assuming that he could not refer either to the effect or the substance of the Advice) might not get him very far.

85 In the light of the foregoing, we did two things (see above at [4]). First, we modified the Judge’s order to the extent that we specified that if the Respondent wished to maintain its position that the Advice was privileged, then it would have to clarify that it did not intend to rely on “the contents, substance, and/or effect” [emphasis added] of the Advice. This modification was intended to set out our position that, on the facts of this case, reliance on the effect of the Advice having been obtained (more specifically, that the effect of the receipt of the Advice was that the Respondent formed the view that there was a tax avoidance arrangement) would necessitate disclosure. Second, we recorded our observation that para 18 of Ms Ng’s 2nd Affidavit should be disregarded. What we intended by this was to put on our record our understanding that para 18 of Ms Ng’s 2nd Affidavit could not be used to support the contention that the Suit was not time-barred without such use amounting to an implied waiver of privilege. This is consistent with what we said about the references being “neither here nor there” (see above at [74]).

86 In our judgment, these two modifications made it clear that disclosure was not necessary. Of course, things might well change. It might be open for the Respondent to contend, despite what he has hitherto maintained, that it was because of the Advice that he concluded that there was a tax avoidance arrangement. (The question of whether the Respondent can still do this even though he has confirmed by way of Ms Ng’s 9th Affidavit that he will only be relying on the fact of the Advice did not arise for decision and we therefore express no concluded view on it.) If so, the Advice would undoubtedly have to be disclosed. But if the Respondent elects not to disclose the Advice, then he cannot use it to support his case that the Suit is not time-barred. Ultimately, this is entirely a matter that is for the Respondent to decide on. However, what the Respondent cannot do is act inconsistently: he cannot rely on the Advice to his advantage while seeking to withhold production. This would be unfair and it is precisely this unfairness which the law of implied waiver is concerned with. This was also the position taken in Brennan where the court refused to order production on the ground that the defendant had not yet relied on their contents in the course of the litigation but held that production might be ordered if the material were subsequently relied on (see Brennan at [69]).

87 We now turn to the third issue, which concerns the so-called “state of mind exception” to legal professional privilege. The Appellant submits that this doctrine provides that where a party puts into contention an argument in respect of which his state of mind is a material issue, then he should be taken to have waived privilege over all legal advice which is probative of the question whether that state of mind exists. In his skeletal submissions, the Appellant described this as a species of implied waiver which arises where a party “deploy[s] his statement of mind as to legal knowledge… to advance his case” and is therefore a species of implied waiver. This is also commonly referred to as the doctrine of “issue waiver” in the authorities and we shall use that particular expression interchangeably with the expression “state of mind exception” in the remainder of these grounds. The parties focused on whether the doctrine ought to be a part of Singapore law. However, when considering the facts of this case, we were more concerned with the point which was raised by the Judge, which was that the doctrine did not even apply on the facts.

88 We will begin by sketching the broad contours of the doctrine as it has emerged in the Australian authorities which were cited to us. For this purpose, we will refer to three decisions of the Federal Court of Australia, one of which preceded and two of which came after the decision of the High Court of Australia decision in Mann. The significance of this will be discussed later.

89 Telstra Corporation Ltd and another v BT Australasia Pty Ltd and another (1998) 156 ALR 634 (“Telstra”) concerned a tender for the management and development of an integrated telecommunications network for the State of New South Wales’s public sector. This was won by the first plaintiff, BTA, a subsidiary of the second plaintiff, BT. However, a dispute soon arose and the plaintiffs alleged that the State had represented during the negotiations that, amongst other things, the network would be used by all the State’s agencies and that BTA would have the exclusive right to supply all agencies with telecommunications services when this was not true. On this basis, the plaintiffs brought a claim for deceptive and misleading conduct under the Fair Trading Act 1987 (NSW) and the Trade Practices Act 1974 (Cth). Following general discovery, the defendants sought production of legal advice which BT had allegedly received relating to its entitlement to rely on these representations and in respect of which the plaintiffs had claimed legal professional privilege. By a majority, the Full Court of the Federal Court of Australia ordered production of the advice.

90 Branson and Lehane JJ, who comprised the majority, drew an analogy with cases involving undue influence and legal professional negligence (at 647). In relation to the former, they explained that a party “who initiates an undue influence case puts in issue in the proceeding the quality of his or her consent or assent”, which would “ordinarily be affected by relevant legal advice received”. To allow a party to put in issue the quality of his consent “while, at the same time, withholding evidence relevant to that issue would be to allow him or her unfairly to handicap the other party to the proceedings”. In relation to the latter, viz, claims for legal professional negligence, they likewise held that clients who commenced suits against their solicitors are taken to have “consented to the use of privileged material” and that it would be unjust for the court not to have access to all the legal advice rendered in order that it might determine whether the solicitor had been negligent. Drawing on these two categories of cases, they put the matter thus (at 647):

On the facts, they held that BT had put their state of mind in issue because the essence of their claim was that they had relied on misleading and deceptive representations made by the State in bidding for the contract (at 649).

91 Beaumont J, dissenting, held that there was no “absolute rule that, whenever it appears, pre-trial, that a party’s state of mind may be relevant to an issue, privilege is lost” (at 639). Instead, cases involving allegations of implied waiver must depend on their particular circumstances, with the touchstone to be applied in such cases being that of fairness. On the facts, Beaumont J noted that that BT had neither pleaded the legal advice as an element of its claim nor had it asserted that it had relied on the advice. He concluded that there was nothing to demonstrate that there was any unfairness in BT’s insistence in claiming privilege although he accepted that this might change at trial if, for example, BT were to lead evidence as to the content of privileged communications which concerned the State’s alleged representations (at 640).

92 As noted above, Telstra was decided before the High Court of Australia delivered its decision in Mann (Telstra itself was appealed to the High Court and fully argued but the appeal was withdrawn before judgment was rendered). The “state of mind exception” was not discussed in Mann itself but it was revisited in the post-Mann decision of the Federal Court of Australia in DSE (Holdings) Pty Ltd v Intertan Inc [2003] FCA 384 (“DSE”). The plaintiff in that case brought a case for rectification and pleaded, among other things, that the amended terms it proposed (a) reflected the common intention of the parties at the time they concluded the agreement and (b) that the parties both believed (perhaps mistakenly) that the text of their agreement reflected this understanding. The defendants joined issue on this with a simple denial and the question before the court was whether the defendants had, in so doing, put their state of mind in issue and thereby waived privilege over all relevant legal advice which they received during the material period.

93 After an extensive review of the authorities, Allsop J (as he then was) opined that the broad statement of the court in Telstra, based as it was on a general notion of fairness, was probably inconsistent with the subsequent decision of the High Court in Mann (at [5]). For that reason, he stated that he would approach the matter “with the necessary recognition from Mann v Carnell that inconsistency is the key to understanding the application of the principle” (at [112]–[113]). Reformulating the Telstra test in terms of the Mann language of inconsistency, he put the matter as follows (at [58]):

94 He held that the doctrine of issue waiver, as it emerged from Telstra and the decisions which preceded it, had always been confined only to circumstances in which “a positive case [had been] raised by the holder of the privilege” [emphasis added] (at [116]). This was the gist of the undue influence cases relied on by the majority. He explained that where a plaintiff brings a claim based on undue influence, he necessarily lays out for examination his capacity for independent action and, in so doing, waives privilege over legal advice which he might have received during that time (at [46]). Applying this to the facts of Telstra, BT had positively pleaded that they had been misled by the representations made by the State and had in so doing waived privilege over such legal advice as might be probative of the existence of that state of mind. However, Allsop J held that the doctrine did not cover instances in which a party simply joined issue with a point raised by the other side, as was the case here. He further held that since a waiver was dependent on the conduct of the holder of the privilege, it was not open to the other party in the litigation to “force” a waiver by making assertions about the other’s state of mind, in an attempt to put the matter in issue (at [121]).

95 Allsop J’s analysis was endorsed by the Full Court of the Federal Court of Australia in Commissioner of Taxation v Rio Tinto Ltd (2006) 229 ALR 304 (“Rio Tinto”). That case concerned the decision of the Commissioner of Taxation’s decision that a dividend payment in the sum of approximately A$100m that was paid to the respondent, Rio Tinto, was taxable and that he would not exercise his discretion to remit tax. Rio Tinto challenged this decision and under the statutory regime, the Commissioner was required to furnish reasons why he was satisfied that the dividends were taxable and why he elected not to exercise his discretion to remit tax. The Commissioner did so was by way of a letter in which it was stated that he had taken into account matters which were “evidenced by” certain listed documents, which included legal advice. Rio Tinto filed an application seeking production of the legal advice claiming that the Commissioner had, in referring to the advice in the way he did, necessarily laid them open to scrutiny and had thereby waived his right to claim privilege. This submission was accepted by the Full Court.

96 As a starting point, the court held, agreeing with DSE, that the broad statement of principle articulated by the majority in that case was probably inconsistent with Mann (at [56]). Instead, the “governing principle” was whether the “privilege holder had directly or indirectly put the contents of an otherwise privileged communication in issue in litigation, either in making a claim or by way of defence” [emphasis added] (at [61]). The court was careful to note that just because the validity of the Commissioner’s state of satisfaction and the exercise of his discretion were in issue, this did not suffice to waive privilege (at [71]). However, on the facts, the Commissioner had gone beyond simply saying that the privileged documents were relevant to these key issues. Instead, he had “identified as his bases for satisfaction and exercises of discretion as the matters evidenced in the scheduled documents” with the consequence being that he had thereby put the contents of these documents in issue and necessarily opened them up for scrutiny (at [72]). In so doing, he had acted in a manner inconsistent with the maintenance of privilege.

97 It can be seen that the doctrine of issue waiver has largely been subsumed within the general principles set out in Mann. The DSE-Rio Tinto approach towards issue waiver has since been consistently applied across the various Australian states: see, for example, the decision of the New South Wales Supreme Court in In the matter of Idoport Pty Ltd (in liq) (recs apptd); National Australia Bank Limited (& Ors) v John Sheahan (& Ors) [2012] NSWSC 58 (“Re Idoport”); the decision of the Supreme Court of Victoria in BHP Billiton Petroleum (Bass Strait) Pty Ltd v Esso Australia Resources Pty Ltd and others [2007] VSC 281; the decision of the Supreme Court of Western Australia in Riverstone Asset Pty Ltd v Ammendolea (No 2) [2013] WASC 351; and the decision of the Supreme Court of South Australia in Elders Forestry Ltd v Bosi Security Services Ltd & Ors (No 2) [2010] SASC 226.

98 To summarise, we can distil the following propositions from the foregoing cases (see Re Idoport at [67]):

(a) The basis of issue waiver is the act or omission of the holder of the privilege. It is not open to another party to the litigation to seek to force a waiver by making assertions about or by seeking to put in issue the state of mind of the holder of the privilege. Instead, it only arises where the holder puts forward a “positive case”: DSE at [116] and [121]).

(b) Where a particular state of mind is asserted against the holder of privilege, merely joining issue with that and denying the existence of the alleged state of mind will not amount to a waiver: DSE at [115].

(c) The mere fact that a holder of privilege raises an issue as to their state of mind will not, without more, amount to a waiver over such legal advice as might have contributed to the creation of that state of mind or be probative of its existence: Rio Tinto at [67].

(d) Where the state of mind of the holder has been put in issue, an acknowledgement that privileged documents were relevant to the formation of that state of mind does not, without more, amount to a waiver of privilege: Rio Tinto at [71].

(e) Instead, waiver will only be found if the holder of privilege, in explaining or justifying his/her state of mind, puts in issue the contents of the privileged material and/or opens it up to scrutiny and, in so doing, acts in a manner which is inconsistent with the subsequent assertion of privilege: DSE at [58]; Rio Tinto at [61].

99 Against that background, we turn to explain why we considered that the Judge was correct in saying that the state of mind exception could have no application on these facts. But before we proceed into the detail of the matter, we emphasise that we undertake this exercise assuming, but not deciding, that the doctrine is part of Singapore law.

100 The crux of the inquiry is whether the Respondent has brought a positive case which relates directly to the contents of the Advice or which necessarily opens it up for scrutiny. In our judgment, he had not. The Respondent had neither alluded to the contents of the advice nor relied upon it in establishing the existence of his state of mind. As we explained above, because of the way in which Ms Ng’s 2nd Affidavit had been crafted, the references to the Advice are irrelevant to the question of when the Respondent formed the view that there was a tax avoidance arrangement (see above at [74]). If this was not already clear as a matter of construction, it has been put beyond any doubt by the clarification we issued (see above at [85]). The most that can be said is that the Respondent had raised an issue as to his state of mind and had, in this connection, alluded to the existence of privileged material which could be probative as to the existence of this state of mind. However, this is not sufficient – in and of itself – to constitute a waiver. This case is very different from Rio Tinto, where the Commissioner had explicitly stated that he took into account the matters evidenced by the privileged documents in order to satisfy himself that the dividend payments were taxable and that he ought not to exercise his discretion to remit tax.

101 In its case, the Appellant, relied on the decision of the Supreme Court of British Columbia in Camosun College v Levelton Engineering Ltd [2014] BCSC 1190 (“Camosun”) at [13], in advancing the proposition that “where a party relies on the postponement provisions for limitation, it automatically puts its state of mind as to its legal knowledge into issue” [emphasis added]. On that basis, it argued that the Respondent had, in alluding to the postponement provisions in the LA, automatically put its state of mind into issue and had thereby waived privilege over the Advice. In our judgment, this proposition is much too broad and we did not agree that it applied here.

102 The decision in Camosun has to be read in light of the statutory context in which it was decided. Section 6(3)(f) of the British Columbia Limitation Act (c 266) (Can) (“BC Limitation Act”), which was the relevant statute then in force, provided that the limitation period for a cause of action might be postponed in cases of mistake. In such a situation, time would only begin to run when the plaintiff either subjectively knows or when a reasonable person in the plaintiff’s position “knowing those facts and having taken the appropriate advice a reasonable person would seek on those facts” would be able to bring a cause of action (see s 6(4) of the BC Limitation Act). Crucially, s 6(5)(a) of the BC Limitation Act defined “appropriate advice” to include legal advice. In this context, a plaintiff who pleads reliance on the postponement provisions in the BC Limitation Act necessarily, in the words of Rio Tinto, “opens up for scrutiny” the contents of any legal advice which he/she might have received. While this might be the position in the State of British Columbia, it is not applicable in Singapore, where our LA does not contain definitional provisions of the sort seen in the BC Limitation Act.

103 Furthermore, the facts of Camosun are also quite different. There, the plaintiff had not only expressly pleaded reliance on the limitation periods but it had also, in its supporting affidavits, made specific reference to the contents of its communications with its solicitor, whom it alleged had been negligent in not advising it properly. This was not the case here, where there was no explicit reference what the Advice said, either in the affidavits or in the pleadings. In these circumstances, we agreed with the Judge that the Respondent had not, thus far, acted in such a manner as would attract a waiver of privilege under the state of mind exception and we therefore found that the doctrine, even if it was a part of Singapore law, could not assist the Appellant.

104 For these reasons, the question whether the state of mind exception should be a part of Singapore law did not arise for decision. We therefore do not propose to decide that issue here save to make several brief observations which might guide the approach taken by future courts towards this question. First, as we have already explained above, the broad formulation of the doctrine has been greatly attenuated following the decision in Mann. At present, it seems to us that it no longer exists (if it ever did) as a free-standing exception to legal professional privilege but is instead best thought of as a specific form of implied waiver that is subject to the Mann test. If this is right, one might question the extent to which its “acceptance” would add anything to the existing doctrine of implied waiver as it currently exists.

105 Second, the state of mind exception – in the form in which it was articulated in Telstra – appears not to be based on a party’s use of the privileged material, but on the raising of an issue to which the privileged material might be relevant. Juridically, this sits uneasily with the doctrine of implied waiver, which has never been about the relevance of the privileged material, but about the unfairness which arises from inconsistent conduct when a party seeks, on the one hand, to rely upon the contents of privileged material to advance his case while seeking, on the other hand, to prevent the opposing party from inspecting the document. The notion of “fairness” referred to in Telstra does not appear to be the restricted notion of fairness described above, but about the general unfairness associated with any assertion of privilege (see above at [64] and D L Matheison and Julian Page, “Implied Waiver of Privilege” [2000] NZLJ 355 at 359). As we noted above, there is always some unfairness associated with the assertion of privilege, but this is no bar to the assertion of privilege.

106 In the English High Court decision of Farm Assist Ltd (in Liquidation) v Secretary of State for Environment, Food and Rural Affairs [2008] EWHC 3079 (TCC) (“Farm Assist”), the claimant applied to set aside a settlement agreement on the ground that it had been entered into under economic duress. The defendants contended that the claimant ought to disclose all legal advice it received during the relevant period because it had, in putting forward an application based on economic duress, put its state of mind in issue. After an extensive review of the relevant authorities, Ramsey J rejected this argument, holding that English law had never recognised such a broad exception to privilege. At [53], he put the matter in the following terms:

107 Third, the approach in Telstra, if accepted, would apply in a multitude of situations. Indeed, it would apply in any case in which the state of mind of the tortfeasor is a constitutive part of the action (for example, misrepresentation, where proof of reliance is necessary) or in any case where it is an essential part of any defence (for example, trademark suits, where “good faith” is a defence): see Andrew Brown, “Deemed Waiver of Privilege – is nothing sacred?” [1999] Law Inst Jnl 64. Arguably, the only area in which (English) law recognises such a general category-based exception to privilege is where a client attempts to sue his solicitor. However, even in such cases, the underlying explanation is one based on fairness and consistency. As Lord Bingham CJ explained in Paragon Finance at 1188:

108 It is perhaps for these reasons that many jurisdictions like New Zealand (see, for example, the decision of the New Zealand Court of Appeal in Shannon v Shannon [2005] 3 NZLR 757 at [39]–[45]), Hong Kong (see, for example, the decision of the Hong Kong Court of First Instance in Chinachem Financial Services v Century Venture Holdings Ltd [2014] HKCFI 453 at [135]), and Ireland (see, for example, the decision of the Supreme Court of Ireland in Redfern Limited v Larry O’Mahony and others [2009] IESC 18 at [29]–[32]), have, like the English High Court, rejected the state of mind exception in the broad form as it was articulated in Telstra.

109 At the end of the day, any court that considers whether the state of mind exception should be a part of our law will have to confront the difficulties that we have outlined above. It is likely that in the course of such an inquiry the court will also have to consider the deeper normative question which we alluded to earlier: namely, the balance that should be struck between (a) ensuring full and frank communications between persons and their legal advisers (which ostensibly supports a robust doctrine of privilege) and (b) allowing all relevant information be placed before the court in order that accurate adjudication can be carried out (which favours greater disclosure). This involves a delicate balancing exercise and we did not think it appropriate for us to consider it at this juncture. We would prefer to defer the consideration of a definitive answer to such a question to another occasion when it squarely arises on the facts and it is essential for us to decide it.

110 In the premises, we dismissed the appeal save for the modifications made to the order to elect set out above at [4].

111 We also ordered that the costs of the appeal be in the cause and that the usual consequential orders would apply.