Zhu Yong Zhen v AIA Singapore Pte Ltd and another

Chan Seng Onn J:

[1] The 1 Defendant (“AIA”) is an insurance company and the Plaintiff (“Mdm Zhu”) a policyholder with AIA. Mdm Zhu brought this suit against AIA for breach of contract. AIA counterclaimed for defamation relating to statements made on an internet blog set up by Mdm Zhu. Mdm Zhu’s claim against the 2 Defendant was previously struck out and is no longer relevant to this dispute.

[2] Three discrete issues arise for my decision. First, the correct interpretation of Mdm Zhu’s insurance policy, viz, whether it was a term of the policy that annual premiums would only have to be paid for a specified number of years, up to a cut-off date known as the “Critical Year”. Second, whether AIA had in breach of contract caused Mdm Zhu’s insurance coverage to lapse. Third, whether AIA’s claim for defamation is established and, if so, whether Mdm Zhu’s defence of justification is borne out. Unfortunately, perhaps owing to Mdm Zhu’s lack of representation, there was a proliferation of irrelevant issues in submissions and at trial, resulting in rather prolix proceedings.

[3] On or about 29 April 1993, Mdm Zhu met an AIA insurance agent named Oscar Huang. Mr Huang utilised a document to aid in advising Mdm Zhu about a product known as the Singapore Financial Guardian (“SFG”) policy. AIA refers to this document as the Policy Benefit Illustration (“PBI”), while Mdm Zhu names it the Original Policy Quotation (“OPQ”). AIA’s terminology will be adopted and the document referred to as the “PBI” in this judgment.

[4] The PBI comprised a table setting out projected annual policy values for a woman of Mdm Zhu’s age at the material time then with an insured sum of $100,000.00. Of general interest to many potential clients are the columns headed “Death Benefit” and “Total CSV Available”. Of greater relevance to this dispute are the columns headed “Premium Paid by Assured”, “Premium Paid from CD Accumulation”, “Current Year Dividend” and “Balance CD”. The acronyms “CSV” and “CD” are not defined, but refer to “cash surrender value” and “current year dividends” respectively. The cash surrender value is the amount of money that a policyholder would receive on terminating the policy in a given year. The current year dividends are the monies paid to the policyholder annually out of the profits of the fund of paid-up premiums.

[5] At the heart of this dispute are the words “Critical Year: 16”. The PBI does not explain what “Critical Year” means. However, the table described above indicates that the annual premium of $2,091.50 would be paid by the policyholder for the first 16 years. Thereafter, this sum would instead be paid out of the accumulated dividends. This is evidenced by the values under the column “Balance CD”. For the first 16 years, the figure reflected under that column is the accumulated dividends as of the previous year, with the addition of 7% p.a. interest and the current year dividend. For the 17 and subsequent years, the annual premium is deducted from the “Balance CD” value for the year. It can be surmised that the Critical Year refers to the last year that a policyholder would have to pay the premium out of her own pocket. It is Mdm Zhu’s contention that the Critical Year is a fixed feature of her policy and that she would not under any circumstances have to pay the 17 and subsequent annual premiums.

[6] Also relevant are the following notes at the bottom of the PBI:

[7] On 3 May 1993, Mdm Zhu completed and signed a policy application form for an SFG policy with a sum insured of $200,000.00 and handed Mr Huang the annual policy premium of $3,883.00. The date of the application form was later changed from 3 May 1993 to 14 May 1993 as Mr Huang only became an AIA insurance agent officially on the latter date. The policy was also backdated to 23 March 1993 to give Mdm Zhu the benefit of a lower premium as her birthday was the day after (viz, 24 March).

[8] The policy application form comprised a questionnaire on the applicant’s personal and health details as well as the identity of the policy applied for. Of note are the following declarations included in the form:

[9] AIA approved the application on 17 May 1993 and a policy booklet setting out the terms of the insurance contract was delivered to her by Mr Huang on 19 May 1993. The preceding facts set the stage for the contractual dispute between the parties. Subsequent events are relevant to the defamation claim made by AIA.

[10] In 2003, AIA established what it calls the Critical Year Support Program (“CYSP”). Like Mdm Zhu, other policyholders were of the impression that they would not have to pay premiums beyond the Critical Year stipulated in the policy benefit illustrations shown to them. This problem was exacerbated by the fact that it was not an industry practice or regulatory requirement for policy benefit illustrations to be provided to potential policyholders. As a result, non-standard explanatory documents and advice might have been given to policyholders. The CYSP was started to address this issue. AIA communicated with the general public through announcements on its website, where the concerns of policyholders were noted and AIA’s position on the Critical Year – that the Critical Year was not a guaranteed feature of its policies – was set out. AIA also informed the public of the steps it intended to take to resolve the issue through its website.

[11] In addition, AIA contacted policyholders individually by letter, beginning with those whose Critical Year had already passed. On 12 May 2004, Mdm Zhu received a letter from AIA regarding the CYSP. Mdm Zhu was told that her projected Critical Year was the 15 year according to AIA’s records. She would be sent a support program package around March 2008, about the time Mdm Zhu’s 16 annual premium became payable. This would have been the first annual premium to be paid from accumulated dividends, had the Critical Year taken effect on the 15 year. Whether the Critical Year was the 15 or 16 year is a point of contention between the parties.

[12] As promised, Mdm Zhu received a second letter from AIA regarding the CYSP on 11 January 2008. In it, AIA presented Mdm Zhu with alternative courses of action which she could elect to pursue. In particular, Mdm Zhu was asked to provide AIA with the policy documents in her possession to enable AIA to make her a “support offer”, ostensibly a favourable variation of her original contract. It appears that the precise offer made to each policyholder would depend on what documents were provided to AIA and their contents. As will be seen, AIA’s request for Mdm Zhu’s policy documents aroused her suspicions. The other options presented in the letter included continuing to pay policy premiums as before, paying the premiums out of accumulated dividends until these were depleted and resuming payment thereafter, or referring the matter to dispute resolution before an adjudicator appointed by AIA or to the Financial Industry Disputes Resolution Centre.

[13] Mdm Zhu replied on 9 April 2008 but declined to select any of the options presented to her. Instead, she asserted that the PBI was in unqualified terms and that everything stated therein was fixed and not variable. Puzzlingly, she also asserted that the PBI was unclear and difficult to understand. In any event, Mdm Zhu submitted copies of a number of documents to AIA, including the PBI, but expressed the view that AIA ought to have these documents already in its records. She asked AIA to send her the details of the support offer to be extended to her as soon as possible.

[14] AIA replied in a letter dated 15 May 2008, stating its position that the Critical Year was merely a projection dependent on the cash dividends declared each year and the interest rate applicable to the accumulated dividends. AIA further stated that Mdm Zhu’s policy was for $200,000.00 and not $100,000.00 and that the projected Critical Year was 15. Accordingly, AIA declined to take Mdm Zhu’s PBI into account when formulating the support offer to be extended to her. This support offer was eventually despatched to Mdm Zhu on 16 June 2008 after repeated inquiries by her.

[15] There followed an exchange of correspondence between the parties, with increasing dissatisfaction on Mdm Zhu’s part regarding AIA’s management of the matter, and no progress towards a resolution of the dispute. On 20 August 2008, Mdm Zhu wrote a letter to AIA’s then Executive Vice President and General Manager, Mark O’Dell, setting out “research findings” that she had made and proposing what she called a “win-win solution”. The research findings were essentially a litany of Mdm Zhu’s suspicions and complaints. Among other things, Mdm Zhu asserted that:

(a) the words “Critical Year: 16” on her PBI were unqualified;

(b) the qualifications on the PBI (referred to at [6] above) were incomprehensible and the table of figures inadequately explained;

(c) AIA was lying when it said that her Critical Year was the 15 year;

(d) AIA was unwilling to fulfil the terms of the PBI and was therefore attempting to disclaim the document in order to substitute terms favourable to itself;

(e) AIA had the PBIs of policyholders in its possession, but pretended not to have them in order to further its objective of displacing these documents;

(f) AIA did not keep policyholders adequately informed, and what information was provided to them was incorrect and/or misleading; and

(g) AIA’s conduct amounted to cheating.

[16] These assertions are the precursors of the statements subsequently made on Mdm Zhu’s blog that AIA submits are defamatory. The “win-win solution” was for AIA to invest $1.5million in a company to be started by Mdm Zhu producing a health supplement which she had developed that promised to be a panacea for a wide variety of diseases and ailments. In exchange, Mdm Zhu would settle the Critical Year issue on AIA’s terms. It was also intimated that, if an amicable resolution were not reached, Mdm Zhu could commence legal proceedings and that information relating to the dispute, presumably the “research findings”, would be revealed to the public.

[17] Mdm Zhu did not receive a substantive reply from AIA. On 3 September 2008, she wrote again to Mr O’Dell, this time threatening to publish her “research findings” if AIA did not make any objections known to her by 10 September 2008. AIA replied on 10 September 2008 declining to invest in Mdm Zhu’s business and inviting her to contact its employees, but did not make reference to this threat. On 13 October 2008, Mdm Zhu sent an email to AIA’s customer care address providing access to her as-yet unpublished blog and expressed the intention to publish it if an agreement on compensation could not be arrived at. Having received no response from AIA regarding the blog, Mdm Zhu opened public access to her blog on 15 October 2008.

[18] On 4 November 2008, AIA’s lawyers, Rajah & Tann LLP, wrote to Mdm Zhu objecting to the publication of the blog and asserting that it contained baseless and defamatory statements. Mdm Zhu was asked to deactivate the blog within five days. Mdm Zhu did so six days later on 10 November 2008.

[19] There followed an attempt at adjudication, which did not commence due to Mdm Zhu’s unwillingness to sign a form agreeing to the terms of the adjudication. Mdm Zhu finally sought legal representation, culminating in the initiation of this action.

[20] Mdm Zhu contends that it is a term of the insurance contract that she would not have to pay premiums beyond the 16 year. Accordingly, she argues that AIA was in breach of the contract in providing her with a CYSP that entailed continuing payments beyond the Critical Year.

[21] The basis of Mdm Zhu’s claim is that the PBI forms part of the contract between her and AIA. However, Mdm Zhu’s understanding is incorrect: the terms of the contract are governed by the policy booklet that was delivered to her on 19 May 1993. It has been observed that a contract of insurance normally arises from an offer made by an insured and an acceptance of that offer by the insurer (Poh Chu Chai, General Insurance Law (LexisNexis, 2009) at 154; see also International Testing Co Pte Ltd v Public Prosecutor [1998] 2 SLR(R) 1026 at [22]).

[22] The facts of the present case fit within the usual situation. The document that Mdm Zhu completed on 3 May 1993 was titled “Application for Life Insurance”, indicating that an offer was being made to AIA which could be rejected. Mdm Zhu was required to provide details of the state of her health to enable AIA to determine whether or not to accept this offer. The identity of the insurance policy and riders being applied for had to be stipulated, indicating that the terms of the contract were not yet a given. Declaration C (referred to at [8] above) makes plain that the policy would not take effect until issued and delivered to the applicant. Mdm Zhu herself agreed that AIA would only know which policy a customer wanted to buy from reading the application form itself. Accordingly, the terms of the contract must be determined by reference to Mdm Zhu’s offer. No mention is made of the PBI. Instead, Declaration D states that the declarations made in the application form as well as the answers provided by Mdm Zhu to any questionnaire or medical examination would together with the “relevant policy” constitute the entire contract between the parties.

[23] This reference to the “relevant policy” cannot be mistaken for the PBI. Indeed, Mdm Zhu’s insistence that the PBI was a description rather than an illustration (the assumption apparently being that a description is of a more definitive character than an illustration) entailed that the substance of the contract would have to be found elsewhere, a proposition I do not accept. Further, although features of the PBI could possibly reflect the terms of the contract, the PBI in Mdm Zhu’s possession itself cannot constitute the contract. This is because it relates to different riders and a different insured sum from that eventually purchased by Mdm Zhu. Moreover, as AIA rightly points out, any number of PBIs could have been printed for Mdm Zhu to advise her of the various policies, riders and sums insured from which she could choose.

[24] It should also be noted that Mdm Zhu cannot rely on the fact that she did not read the policy booklet to assert that it did not constitute the contract. In Rust v Abbey Life Assurance Co Ltd [1979] 2 Lloyd’s Rep 334, Brandon LJ (as he then was) said (at 339):

[25] This is eminently sensible. An insured person may choose not to verify the terms of the policy being entered into, but that does not negate that individual’s agreement to those terms. Mdm Zhu’s complaint at trial that AIA had never told her that the terms of the contract were contained in the policy booklet is not only irrelevant but doubtful as well. It was Mr Huang’s evidence that he had personally handed her the policy booklet and that he had spent an hour or two explaining it to her. He averred that it had always been his practice to go through the policy booklet with his clients, and he was especially enthusiastic with Mdm Zhu because she was his first client and he had just received training. Indeed, Mdm Zhu had the opportunity to cancel the policy even after AIA’s acceptance as it was a term of the policy that she could do so within 14 days after the policy document had been received. The obvious intent of that was to allow Mdm Zhu sufficient time to read and review the policy terms and decide whether or not to cancel or to continue with the policy. If she could not understand any term, she could easily have contacted Mr Huang for clarification. In any event, based on my observation of Mdm Zhu, she does not seem to have difficulty in reading and understanding English despite learning the language only at university. She conducted the trial herself in person, gave evidence in English and cross-examined Mr Huang at length also in English.

[26] Crucially, the policy booklet foreclosed the incorporation of extrinsic terms by way of an entire agreement clause which mirrored Declaration D of the application form. That clause stated as follows:

[27] It has been authoritatively established that an entire agreement clause will confine the terms of a contract within the four corners of the document in which they are found. In Lee Chee Wei v Tan Hor Peow Victor and others and another appeal [2007] 3 SLR(R) 537, the Court of Appeal cited the following passage with approval (at [26]):

[28] The Court of Appeal further held that an appropriately worded provision would be acknowledged and upheld if it clearly purported to deprive any pre-contractual or collateral agreement of legal effect. To hold otherwise would render an entire agreement clause meaningless and fail to give effect to the parties’ expressed intent and their legitimate expectations (at [33] and [35]).

[29] In the present case, the purport of both Declaration D and the entire agreement clause is clear: the policy (of which the clause is a part) and the application for the policy constitute the entire agreement between Mdm Zhu and AIA. There is no scope to contend that they are also bound by the PBI. As the Critical Year is not mentioned in either the policy booklet or the application form, it is not a term of the contract. Mdm Zhu has also not alleged that the entire agreement clause and Declaration D have been oppressively employed against her as a consumer by AIA such that they should not be given effect to. Nor has she alleged that the entire agreement clause and Declaration D are unfair or unreasonable such that they are rendered unenforceable under the Unfair Contract Terms Act (Cap 396, 1994 Rev Ed). Accordingly, Mdm Zhu’s claim for breach of contract must fail.

[30] Additionally, AIA attempted to rely on Declaration A, arguing that since the PBI was not presented to and approved by an officer specified in the policy, it was not binding on AIA. Given my finding above, it is not necessary to address this argument.

[31] For completeness, it is noted that even if the PBI did form a part of the contract, Mdm Zhu would receive no aid from it. This is because the content of the PBI clearly indicates that there was no guarantee that she would only have to pay premiums up to the Critical Year. The import of the qualifications noted at [6] above and their relationship with the Critical Year is clear. The column headed “Premium Paid from CD Accumulation” indicates that premiums paid after the projected Critical Year will be paid out of accumulated dividends. The PBI did not indicate that Mdm Zhu’s obligation to pay premiums would lapse altogether after the projected Critical Year. It follows that there must be sufficient accumulated dividends for Mdm Zhu to be relieved of paying the premiums for any particular year.

[32] Further, the PBI expressly states that there was no guarantee that the dividends actually issued by AIA would be the same as that projected in the table. Nor would the dividends that were issued necessarily increase as favourably from year to year since the interest rate paid on them was not fixed. At trial, Mdm Zhu admitted that she understood that dividends were not guaranteed to be the same as that reflected in the table and that the amount of accumulated dividends was also dependent on the interest rate. She also agreed that Mr Huang had taught her how to calculate the accumulated dividends. It must follow that she also understood that there could well be insufficient accumulated dividends to pay for the policy premiums. Instead, Mdm Zhu ran the fanciful argument that the policy values in the PBI table and, consequently, the Critical Year, were guaranteed regardless of whether the dividends and interest rate were guaranteed or not.

[33] Mdm Zhu made comparisons with other policy benefit illustrations that were issued to her together with the CYSP in 2008. In these documents, each column in the table of policy values was expressly stated to be non-guaranteed. Mdm Zhu reasoned that by contradistinction this must mean that the values in the PBI issued to her in 1993 were guaranteed since similar stipulations were not made then. This argument has no merit. It is trite law that contracts are to be interpreted in the context in which they were entered into. These later documents do not form part of that context and in any event do not detract from the clarity of the PBI given to Mdm Zhu and the qualifications contained therein.

[34] Mdm Zhu also made the argument that the economy was buoyant through the 1980s and at the time she purchased the policy. Accordingly, both she and AIA must have been of the view that the policy values in the PBI were sustainable and that the Critical Year would therefore be a guaranteed feature of the policy. However, it is also trite law that the intentions of contracting parties are to be assessed objectively, and nowhere is this alleged subjective understanding objectively manifested. It need hardly be said that it is highly unlikely that AIA was under the same misapprehension as Mdm Zhu as regards the prospects of the economy. A company with an institutional memory and professional employees would not have been under the mistaken notion that the hitherto cyclical performance of the economy had been discarded with the 1980s. Nor could it have been so foolhardy as to believe that the rosy economic situation in the 1980s allowing AIA to pay out high dividends and a high interest of 7% p.a. on the accumulated dividends would continue unchanged for the 65 years for which policy values were reflected in the PBI.

[35] A casual analysis of the PBI reveals that the concept of a Critical Year was nothing more than a puff made in the context of an illustration based on certain assumptions. It reflects the truism that if certain favourable conditions prevail which allow AIA to continue to pay out high dividends and a high interest of 7% p.a. on the accumulated dividends, a policyholder would then accumulate dividends at a sufficiently high rate and receive sufficient returns on the premiums already paid, such that further payments out of the policyholder’s own pocket would no longer be necessary as the policy would become self-sustaining after a certain number of years, that being the Critical Year. But the Critical Year (be it the 15 or 16 year) is premised on certain assumptions prevailing and if those assumptions do not prevail, the Critical Year (whether the 15 or 16 year) will simply not hold true any longer. An assumption of favourable conditions prevailing throughout the life of the policy thereby allowing the projected rates in the PBI to be given for the next 65 years clearly does not amount to a contractual term that the assumption is to be a guaranteed fact for the next 65 years. In fact, the PBI itself negates any possibility that the assumption is to be elevated to a guaranteed fact because the footnotes to the PBI have clearly stipulated in capital letters for emphasis that future dividends are not guaranteed and that the interest rate of 7% p.a. on accumulated dividends is not guaranteed and is used for illustration purposes only (see [6] above).

[36] It should be noted that Mdm Zhu did not advance the case that Mr Huang had misrepresented that the PBI was part of the contract or that the Critical Year feature contained therein was guaranteed or that under no circumstances would the policyholder pay anymore policy premiums after the Critical Year or that the interest rate illustrated of 7% p.a. was a fixed and guaranteed interest rate throughout the life of the policy. It appears that Mdm Zhu’s reluctance to make such a claim was two-fold. First, pleading misrepresentation would have been inconsistent with her case that the PBI was part of the contract and that the statement “Critical Year: 16” contained therein was unqualified. Second, she was of the view that Mr Huang would be unwilling to give evidence detrimental to AIA. It may be for this reason that her evidence on her interactions with Mr Huang was extremely brief. However, when cross-examining Mr Huang, Mdm Zhu asserted:

[37] First, it is to be noted that a question from Mdm Zhu in cross-examination per se does not become her evidence in court on oath. I will therefore not treat those assertions within her question as part of her evidence in court where they are self-serving. However, where any assertion of hers is against her interest, it may be relevant as an admission under s 21 of the Evidence Act (Cap 97, 1997 Rev Ed) and admissible as such. It would appear from the verbatim court transcript of the question framed by Mdm Zhu during her own cross-examination of Mr Huang, that she had let it slip out that Mr Huang might well have correctly advised her that interest rate and dividends were variable and not guaranteed. While he might have told Mdm Zhu not to worry because the values in the table were likely to be met or fulfilled in view of the favourable past performance, this contradicts any suggestion or alleged misrepresentation that they were fixed and guaranteed. If the dividend interest rate was supposed to be fixed and guaranteed at 7% p.a. every year, then how could Mr Huang possibly be mentioning that the dividend interest rate was increasing every year since 1986. At no stage did I understand Mdm Zhu to be saying that Mr Huang had misrepresented to her that the dividend interest rate was variable but nevertheless guaranteed at a minimum of 7% p.a. for the entire life of the policy. This is quite different from saying that it was guaranteed at a fixed rate of 7% p.a. for the entire life of the policy, which apparently is what Mdm Zhu is now alleging as having been represented to her by Mr Huang. In any event, since Mdm Zhu has consciously declined to plead that there was misrepresentation, it is not necessary for me to consider those admissions against her own interest made in the course of her cross-examination of Mr Huang as set out above. They would have been relevant only if misrepresentation had been specifically pleaded.

[38] Mdm Zhu separately claims that AIA refused to execute her instructions to pay her policy premiums out of the accumulated dividends since 2009, thereby causing her SFG policy to lapse in breach of contract and contrary to her instructions. This claim has no merit. In a letter dated 6 October 2010, AIA’s lawyers told Mdm Zhu:

[39] Similar assurances were given in a letter dated 5 January 2012:

[40] At trial, AIA’s Acting Head of Service Quality, Ms Eileen Ong, confirmed that Mdm Zhu’s policy remained in force providing coverage of the insured sum of $200,000.00. There is therefore no basis to the claim that Mdm Zhu’s insurance policy has lapsed.

[41] In its counterclaim, AIA pleaded that 10 separate statements made on Mdm Zhu’s blog were defamatory. However, AIA has not established on the evidence adduced before me that the blog and the allegedly defamatory statements therein have been published, and I am accordingly of the view that this counterclaim is not made out. The learned authors of Evans on Defamation in Singapore and Malaysia (Doris Chia and Rueben Mathiavaranam) (LexisNexis, 3rd Ed, 2008) (“Evans on Defamation”) note the following (at p 59):

[42] These observations are of general application and are equally relevant to defamation on the internet. In The Law of Defamation and the Internet (Matthew Collins) (Oxford University Press, 3rd Ed, 2010), it is observed at 5.04:

[43] It is therefore insufficient for material on the internet to be technically accessible – it must actually have been visited by a third party for the requirement of publication to be met. AIA may not have adequately grasped this distinction. In its Reply Submissions, it stated (at [35]):

[44] As regards proof of publication, the case of Al Amoudi v Brisard [2007] 1 WLR 113 is instructive. The court observed that, following from the general rule that the claimant bears the burden of proving that the words complained of were read or seen by a third party, it was for the claimant in the case of an internet libel to prove that the material in question was accessed and downloaded (at [32]). Publication to a particular individual can be proved by calling that individual to say that he or she accessed the items and downloaded them within the jurisdiction; a wider publication may be proved by establishing a platform of facts from which the tribunal of fact can properly infer that substantial publication within the jurisdiction had taken place (at [33]). There is no presumption of law of substantial publication in regard to internet libels (at [37]).

[45] In the present case, no direct evidence has been adduced to show that Mdm Zhu’s blog had been accessed by third parties. No witnesses have been called to give evidence to that effect. The blog itself appears to have been reproduced by printing it off from the web browser and making copies of the printout. As the blog did not have a web counter that logs the number of visitors it had received, it is not possible to ascertain whether the blog had been accessed by the time the printout was made. The blog also included hyperlinks to “comments” pages where third parties could air their views, but each link shows that no comments were made. AIA could have gone beyond the blog itself for direct evidence of publication, perhaps by seeking discovery of web analytics data that might have been available to the blog owner as part of the suite of tools provided by the host website, or by demonstrating that third party websites were hosting links to the blog. This was not done.

[46] Nor has there been adduced any evidence from which it can be inferred that the blog had been accessed by third parties. AIA could have provided such evidence by, for example, establishing the prominence given to the blog by internet search engines when relevant search terms were entered. In Steinberg v Pritchard Englefield (a firm) [2005] EWCA Civ 288, the Court of Appeal of England and Wales held that an inference of substantial publication of a defamatory letter on the defendant’s website was “irresistible” because it was accessible to anyone who fed the claimant’s name into a standard search engine or who accessed the defendant’s website (at [21]).

[47] Indeed, it appears highly likely that the blog was not in fact accessed by third parties. The blog is notable for its utter lack of success. In view of the forceful assertions made by Mdm Zhu and Singapore’s vocal internet audience, it would be reasonable to expect comments expressing views one way or the other to be left on the blog had it gained any attention whatsoever. It is also doubtful that the blog could ever have garnered an audience without active promotion, which Mdm Zhu did not engage in. It is not likely that anyone would have been searching the internet for information on the Critical Year issue in 2008, some four years after AIA launched the CYSP. Mdm Zhu opened public access to this blog on 15 October 2008 and it was deactivated soon thereafter on 10 November 2008, just one day past the deadline set by AIA’s lawyers for her to take down the blog. The window for access to this blog of some 25 days was thus relatively short, which minimised the likelihood of a third party accessing the blog. In my view, Mdm Zhu’s blog was probably one of the multitudes of dormant pages on the internet that are technically accessible but never visited. I should not be understood as saying that blogs are generally poorly patronised websites or that publication will be difficult to establish in such cases. Certainly, some blogs are among the most popular websites on the internet. In other cases, there may be direct evidence of publication in respect of websites which have negligible readership.

[48] In the final analysis, the publication of defamatory material should not be assumed; each case must be determined on its facts. Analogies may helpfully be drawn to the traditional modes of communication. For example, an email that has not been accessed is not very different from a letter in an unopened envelope. Similarly, defamatory material that has been placed on the front page of a mass media outlet such as an online newspaper are analogous to articles in a traditional newspaper, and the inference that it has been published to the readers of that outlet can reasonably be drawn.

[49] However, the internet may well disclose new scenarios in which the element of publication is not clearly established. The present case with a short window period for access is one such scenario. Another possible situation is where defamatory material is made accessible on a website that is generally well patronised, but rather than being on the front page, lies at the end of a series of hyperlinks that receive little or no attention. A defendant may well be able to produce data to show that the (allegedly) defamatory page did not receive any visitors.

[50] As publication has not been established in the present case, the issues of the natural and ordinary meaning of the allegedly defamatory statements and Mdm Zhu’s defence of justification do not arise.