WANG DIN SHIN v. NINA KUNG

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF APPEAL

CIVIL APPEAL NO. 460 OF 2002

(ON APPEAL FROM HCAP 8 OF 1999)

__________________

____________________

_________________

J U D G M E N T

_________________

Hon Yeung JA:

Introduction

1. This case concerns the genuineness or falsity of "Wang Teh Huei" (王德輝) -the questioned signature of the man so named- written with a broad-nib calligraphic fountain pen in liquid ink in each of the four homemade documents, "A", "B", "C", and "D" (collectively referred to as "the 1990 documents").

2. Wang Teh Huei (Wang)'s Wife - Nina Kung Ru Xin (the Wife)- said Wang had duly executed "the 1990 documents" as his last will ("A", "B", and "C" were testamentary in nature). Wang's 93-year-old father -Wang Din Shin (the Father)- said they were forgeries.

3. If the Wife were right, probate of Wang's estate should be granted to her upon his death. Otherwise, the Father would be entitled to such grant by virtue of an earlier will made in 1968 ("the 1968 Will").

4. As the propounder of "the 1990 documents," the Wife had the burden to prove - on a balance of probability - that Wang had duly executed them so that the court could pronounce them, as his last will in solemn form. If she failed, the Father succeeded, as the due execution of "the 1968 Will" was not disputed.

5. In discharging her burden, the Wife was bound to satisfy the conscience of the court that Wang did execute "the 1990 documents" (i.e. Wang signed them). The Father did not have the burden to prove Wang did not.

6. The approach of the court was inquisitorial in nature, as the court must give effect to the true intention of Wang and not be bound by the wishes of the parties.

7. If the Father wished to establish that "the 1990 documents" were forged, he had to adduce strong and cogent evidence, as the standard of proof should commensurate with the gravity of the allegation and the seriousness of the consequences. (See the judgments of Lord Nicholls of Birkenhead in In re H (Minor) [1996] AC 563 at p. 587, of Bokhary JA (as he then was) in Aktieselskabet Dansk Skibsfinansiering v Wheelock Marden & Co Ltd & Ors [1994] 2 HKC 264 at p. 270 and of the Court of Final Appeal in HKSAR v Lee Ming Tee FACC 1 of 2003 at p. 14)

8. If "the 1990 documents" were forged, obviously they could not have been duly executed, although the reverse was not necessary true as technically it involved a different burden and/or standard of proof.

9. The simplicity of the issue did not lead to a short and guileless trial. The resourcefulness of the parties or the ingenuity of counsel or both resulted in a record-breaking 172-day trial, which took place before Yam J (the judge) over the 14-month period between 6 August 2001 and 15 October 2002. On the relevant issues, Mr. Edward Chan SC leading Mr. Albert Tsang and Mr. Victor Luk for the Father; Mr. Martin Lee SC leading Mr. R. Sujanani and Mr. Erik Shum for the Wife.

10. Unparalleled in the legal history of Hong Kong, the trial was a chimera, with a fire-breathing mouth that had devoured a significant part of our judicial capacity and a serpent's tail in the form of a 600-page judgment.

11. In the judgment handed down on 21 November 2002, the judge found the questioned signatures of Wang to be forgeries.

12. The judge gave judgment for the Father with costs and granted him a decree of probate for "the 1968 Will" in solemn form. The Wife appealed against the judgment (CACV 460 of 2002-the main appeal) and the costs order (CACV 67 of 2003-the costs appeal).

13. Due to time constraints, the parties agreed to adjourn the costs appeal.

Background Facts

14. The background facts are set out only in such details as to explain how the issues arose.

15. Wang, a multi-billionaire and the head of "Chinachem" (a family business started by the Father and substantially expanded into a property development conglomerate by Wang with the assistance of the Wife), was kidnapped in 1983.

16. He was released after the payment of $75 million ransom.

17. On 10 April 1990 Wang was kidnapped again and had not been seen or heard of since (subject to the Wife's assertion to the contrary).

18. One month preceding the second kidnap, on 10 March 1990, Wang was injured in a riding accident at the Jockey Club Clubhouse at Beas River, Fanling. He stayed in a hospital for two days before discharging himself against medical advice.

19. On 22 September 1999, the judge (in separated but related proceedings) granted leave to the Father to swear to Wang's death to have occurred on or since 10 April 1990, despite objection from the Wife who had insisted that Wang was still alive.

20. In an obvious attempt to defeat the Father's application to swear to Wang's death, the Wife claimed to have spoken to Wang twice since his kidnap (on 23 April 1990 and 14 December 1996). She also claimed that on 2 September 1996, she had seen Wang's photograph supposedly taken after the second kidnap. The judge rejected her assertions and she accepted the decision.

21. Wang first made a will in 1960 ("The 1960 Will") by Messrs. P C Woo, solicitors, bequeathing his entire estate to the Father and the Wife in equal shares. The Wife was to hold her half share on trust for any children of the family also in equal shares.

22. When "the 1960 Will" was made, Wang was 26, and the Wife 24 years old. They did not have children and Wang was not aware of his infertility.

23. The Father kept "the 1960 Will".

24. In March 1968, Wang made "the 1968 Will," revoking "the 1960 Will" and bequeathing his entire estate to the Father. By that time, Wang was aware of his physiological defects and the resulting infertility.

25. Mr. Donald Cheung of Messrs. F Zimmern, solicitors prepared "the 1968 Will." His firm kept a copy in a wax-sealed envelope.

26. Both "the 1960 Will" and "the 1968 Will" were written in English and executed by Wang in English.

27. Wang made "the 1968 Will" to sever the Wife completely from his estate because he believed the Wife had an adulterous relationship with another man. The Father was aware of the existence of "the 1968 Will," but the Wife was not until about February 1985 at the earliest.

28. There was positive evidence that Wang was still concerned about the Wife's "unfaithfulness" in 1974.

29. "The 1968 Will" was kept in a safety deposit box rented by Wang from a branch of the then Bank of Canton ("the Bank") on 12 February 1968 with a P.O. Box address of Ping Hing Hong -a company acquired by Wang from the Father also on 12 February 1968. That branch of the Bank was close to Wang's office.

30. The Father paid the rental for the safety deposit box initially. Wang then paid it. There was a period from 1982 to 1991 when the rental was not paid. The Father only paid the entire outstanding rental in cash on 16 July 1990.

31. Wang only authorized Yih Lee Kwong (Yih) - his trusted cousin and colleague - to have access to the safety deposit box. The Bank issued two keys, but only one was found in Wang's belongings after his 2nd kidnap. It appeared that the Father was given the other key, but he was unable to locate it.

32. Wang told the Father in 1968 about "the 1968 Will" inside the safety deposit box. Wang also expressly instructed Yih not to open the safety deposit box except at the Father's request.

33. When the safety deposit box was finally opened on 22 April 1997 in the presence of representatives of the Inland Revenue Department, "the 1968 Will" was found inside with other items, including the revocation of the Power of Attorney granted to the Wife, a Power of Attorney granted to Yih, photographs of Wang's girlfriends with negatives, receipt from the private detective who obtained evidence of the Wife's "infidelity," and some banking documents (all relating to or arising out of the Wife's "infidelity").

34. The Bank's records indicated that, between February 1968 and April 1997, the safety deposit box was opened only once, on 22 October 1975. If the Bank's records were accurate, Wang could only have deposited the documents in the safety deposit box on 22 October 1975 as they came into existence after the safety deposit box was rented.

35. However, the evidence (on the number of times that the safety deposit box had been opened) was not entirely satisfactory due to changes in the ownership of the Bank.

36. In February 1985, Wang and the Wife had a dispute with Mr. Donald Cheung, who then suggested the return of "the 1968 Will" to Wang. A meeting was proposed for such purpose but the meeting did not take place.

37. Messrs. Deacons, acting for Mr. Donald Cheung of Messrs. F Zimmern, eventually returned the copy of "the 1968 Will" in 1991.

38. If the Wife were aware of the existence of "the 1968 Will", the earliest would be the occasion in February 1985. There was, however, no clear indication that she was aware of the contents of "the 1968 Will" until after the opening of the safety deposit box in April 1997.

39. In September 1999, the Father sought probate of "the 1968 Will". The Wife objected and counterclaimed for probate of Wang's estate to be granted to her. The Wife claimed that "the 1990 documents" were Wang's last valid will and she was thereby made the sole beneficiary of his estate. The suggestion was that "the 1968 Will" had been revoked by "the 1990 documents".

40. The Wife first lodged the sealed envelope containing "the 1990 documents" on 16 January 1998, nine months after the safety deposit box was opened. "The 1990 documents" were only lodged pursuant to an Unless Order and after the Wife had said she needed time to search for testamentary scripts. In her affirmation of testamentary scripts, the Wife said the following:

"In about February or March 1990 my husband made a will and a few days later he gave me a sealed envelope which he told me contained the will and various confidential instructions relating to his estate. He gave me this envelope for safekeeping and I have kept it ever since in my safe. When he gave me the envelope he told me that it should not be opened until after his death. Out of respect for his wishes I have never opened the envelope."

41. When the Wife lodged the sealed envelope, she also applied for and was granted an order prohibiting its opening. The Wife then suggested that should the kidnappers be aware of the contents of the envelope, Wang's safety would be put at risk. The reasons she advanced for such suggestion were imprecise and not easy to comprehend.

42. The Wife withdrew her objection at a hearing on 6 September 1999 and the judge opened the sealed envelope in court. Inside the sealed envelope, there were 4 smaller sealed envelopes separately containing "A", "B", "C", and "D". None of the envelopes had been marked to indicate their contents. The home address of Wang and the Wife was printed on the outer envelope.

43. "A", "B", and "C" were very thin pieces of paper whereas "D" was slightly thicker.

44. "A" (dated 12 March 1990 and captioned "Will") contained the following writing in Chinese:

"I, Wang Teh Huei, of 15 Bluff Path, The Peak, Hong Kong, solemnly make this will that after my death, all my property shall be bequeathed to my Wife Nina Kung."

45. "B", and "C" (both undated) each contained a declaration in Chinese as follows:

"B:

'After the death of me Wang Teh Huei, all (my) property shall be left to my Wife Nina Kung for her management and no one shall disagree with that. I love my Wife. She is my dearest in the world. After my death, all my property, real estate and my body shall all belong to my beloved Wife. Although my father and mother disappointed me, I still insist that my Wife must take care of them and also that ailing younger sister who can never possibly support herself.'

C:

'After my death, all my property shall be left to my Wife and no one shall object to that. My Wife shall manage all the property, but shall never deliver or give any monetary benefit or property to any others of my Wang's family. They are all disappointing. But neither shall (my) Wife share out the property to anyone of your Kung's family, because everyone of your Kung's family is also disgusting.'"

46. "D" (also undated) consisted of 4 English words- "One life one love".

47. "One life one love" (of the same style as the one on "D") appeared faintly on "C". This suggested that "C" was underneath "D" when "D" was written.

48. "One life one love" (of a different style from the one on "D") together with the additional word "Nina" also appeared faintly on "A", and "B".

49. These suggested that there should be another document "E" with "One life one love Nina" written on it. "E" was not part of "the 1990 documents" and it was never located. It appeared that when the missing "E" was written, somehow (due to the thinness of the paper or the presence of carbon papers) the writings ended up on "B" and then "A" as well.

50. "A", "B", and "C" were very sloppily prepared with the use of at least 3 different writing instruments. The text of "A" and "B" was written with the same ink. The text of "C" and "D" were written with different inks (also different from those of "A" and "B"). Some of the characters appeared to have been written by defective ball-pens, as there was uneven ink-flow. There were also over-writing and re-writing of characters.

51. In addition to the questioned signature of Wang, each of "the 1990 documents" bore three Chinese characters -Tse Ping Yim (謝炳炎) purportedly the signature of the man so named. The questioned signature of Tse (written with similar ink) appeared between two crosses marked with pencil.

52. On "A", there was a further name (Lee K Sg) written in English under the characters- Tse Ping Yim (謝炳炎).

53. Tse Ping Yim (謝炳炎) (Tse) was the name of a manservant. Tse first worked for Chinachem between 1972 and 1974. He rejoined Chinachem in 1982 as a building cleaner. After Wang's kidnap in 1983, the Wife asked Tse to work as a butler/cook in the family house. Since 1989, Tse also worked as a messenger in the Chinachem office on the instruction of Wang.

54. Lee K Sg could be the name of Lee Kin Sang (a senior Chinachem staff and a subcontractor who had regular contacts with Wang). Lee Kin Sang died on 21 June 1996.

55. It was however made clear on the first day of trial by Mr. Martin Lee SC (leading counsel for the Wife at trial) that it was not the Wife's positive case that Lee K Sg was Lee Kin Sang. Her case was that she did not know who signed Lee K Sg in "A" and she could not recognize the writing (i.e. Lee K Sg in "A").

56. There was no evidence of any attempt to find out if the writing (Lee K Sg in "A") belonged to Lee Kin Sang although the Wife obtained documents with his signatures long before the trial commenced. There was also no indication of any attempt to find out who Lee K Sg really was.

57. The respective characters -Wang Teh Huei (王德輝) and Tse Ping Yim (謝炳炎) (the questioned signatures)- on each of "the 1990 documents" were written by the same two persons. The questioned signatures appeared in more or less the same positions irrespective of where the text of the documents ended.

58. Wang did not write the text of "the 1990 documents". The author of the text of "A" was not the same as the author of the text of "B", or "C", which were probably written by the same person. There was no clear indication as to who authored "D". However, "D" and the missing "E" appeared to have been written by the same person.

59. Tse had made two statements supporting the Wife's case. Shortly after the opening of the sealed envelope by the judge on 6 September 1999, Tse wrote a statement in the presence of Mr. Alex Lam (Mr. Lam) in the same evening.

60. On 9 September 1999, Mr. Allan Leung (Mr. Leung) took a similar statement from him. Both Mr. Lam and Mr. Leung were solicitors closely connected with the Wife.

61. Mr. Leung (of Messrs. Lovell, solicitors) was the solicitor acting for the Wife at the hearing on 6 September 1999. He informed the Wife on that day the importance of taking a statement from Tse. However, he was not asked to do it and he did not find it necessary to do it himself.

62. At the hearing on 6 September 1999, Mr. Joseph Poon (Mr. Poon), the principal of Mr. Lam and described by the judge as "the solicitor acting as an adviser and a confidant of the Wife" was also present in court.

63. Mr. Lam claimed to have been told by Mr. Poon that the Wife was unable to get hold of Mr. Leung on 6 September 1999 and that Mr. Poon himself was unavailable; hence Mr. Poon asked Mr. Lam to take a statement from Tse, although Mr. Lam knew nothing about the case and had not been given any proper briefing on it.

64. In the statement dated 6 September 1999, Tse wrote neatly six paragraphs of Chinese without any significant clerical or grammatical error.

65. Tse stated that he had been asked by Wang to sign some documents without knowing their contents. Tse also stressed that Wang and the Wife were an "affectionate couple".

66. References were made to the 10 March 1990 horse-riding accident of Wang, his subsequent hospitalisation, and the Wife's advice that Wang should avoid exercising, as his wound had not healed.

67. Tse identified the three characters, "Tse Ping Yim," (謝炳炎) in each of "the 1990 documents" as his signature. Tse emphasized that the text of the documents had been written before he appended his signatures to them. He said that he was alone with Wang when he witnessed Wang's signing the documents. He also said that he himself signed at positions designated by Wang.

68. Tse then affirmed the veracity of his statement by way of a statutory declaration. Mr. Lam himself attested to Tse's statutory declaration instead of asking another solicitor firm to do it (contrary to proper practice and conduct, as Mr. Lam admitted).

69. In the 9 September 1999 statement, Tse again confirmed "Tse Ping Yim" (謝炳炎) on each of "the 1990 documents" as his signature. He repeated the circumstances in which he appended his questioned signatures without knowing the contents as he had full confidence in Wang. He said it was the first time Wang had asked him to sign any document.

70. Shortly after making the two statements, Tse took a paid leave on 14 September 1999. He left Hong Kong for the Mainland the same day and returned on 18 September 1999.

71. On 23 September 1999, the judge granted leave to the Father to report to the police that "the 1990 documents" were forged although the report was only made on 6 October 1999.

72. The very next day after the judge granted leave, on 24 September 1999, Tse left Hong Kong. He was diagnosed with liver cancer in November 1999. Between 28 October 1999 and 2 November 1999, Tse was in Thailand on holiday, flying directly between Thailand and Guangzhou.

73. Tse returned to Hong Kong on 4 December 1999 in critical conditions. Soon after his return, Tse became unconscious and the police could not interview him. He died two days later.

74. Two relevant entries in Tse's employment records kept by Chinachem suggested that Tse took leave on 14 September 1999 and that he was not paid his wages in November 1999 because "he was sick".

75. Apart from her affirmation of testamentary scripts, the Wife also filed other affirmations in her attempts to propound "the 1990 documents" as the last will of Wang. The Wife said in those affirmations that "the 1968 Will" had been superseded by "the 1990 documents" and she was made the sole beneficiary of Wang's estate under their terms.

The Father's Case

75. The Father put forward "the 1968 Will" as the last will of Wang. He put the Wife to proof of due execution of "the 1990 documents" by Wang.

76. It was the Father's case that that Wang could not have executed "the 1990 documents." The Father pointed out the continuous existence the "1968 Will" and suggested that there was no reason for Wang to make a fresh will in 1990 to cut his father off completely from his estate.

77. The Father suggested that the alleged execution of "the 1990 documents" and the documents themselves were inconsistent with the character of Wang. His allegation was that the questioned signatures of Wang and Tse in "the 1990 documents" were forgeries.

78. Handwriting experts' evidence was adduced to support the Father's case.

The Wife's Case

79. The Wife accepted (only after the trial had started) the due-execution of "the 1968 Will" by Wang but maintained that it had been revoked by "the 1990 documents".

80. The Wife suggested that the general probabilities of the case supported the due execution of "the 1990 documents" in her favour by Wang. She also adduced handwriting experts' evidence to prove the genuineness of the questioned signatures of Wang and Tse in "the 1990 documents".

81. The Wife emphasized that she was capable and she worked closely with Wang in his real estate business throughout their marriage and therefore, to ensure the smooth transfer of Chinachem, Wang had good reasons to make her sole beneficiary of his estate.

82. Reference was made to occasions in the 1980s when Wang made wills at Chinese New Year before going on holidays ("the holiday wills"). Such holiday wills, so the Wife claimed, indicated that Wang was not unused to making homemade wills and that "the 1968 Will" no longer reflected Wang's true testamentary intention in 1990.

83. The Wife also referred to attempts by Wang to make trust arrangements in 1987 and in 1989-1990 without informing the Father. The effect of the trust arrangements was to make her beneficiary of Wang's assets with the added advantage of avoiding or reducing estate duty. The Wife claimed that such trust arrangements were consistent with Wang's wish to make her sole beneficiary of his estate.

84. The Wife emphasized that she and Wang were extremely close to substantiate her case that she had long been reconciled with Wang. The implication was that her infidelity in the mid-1960s had already been forgotten and forgiven.

85. It was pointed out that the development of Chinachem was the result of the couple's joint efforts and it was their "baby". Wang therefore had no reason to allow anyone but the Wife to inherit Chinachem.

86. It was also suggested that the riding accident on 10 March 1990 must have triggered Wang's "post-mortem" thoughts and could have caused him to execute "the 1990 documents" on the spur of the moment.

87. The urgency of the matter (to ensure the smooth transfer of Chinachem should Wang die unexpectedly) also explained why Wang had to execute "the 1990 documents" in such a casual, sloppy manner and without the assistance of a solicitor.

88. The Wife, through her counsel, alleged that Wang had good reasons for excluding the Father from his estate. The reasons were, inter alias, the Father's "womanising" (including his relationship with a concubine) and his failure to financially support Wang's studies. It was suggested that Wang had to abort his studies to help repaying the Father's debts.

89. It was also suggested that Wang must have forgotten about "the 1968 Will" when he executed "the 1990 documents".

Examinations by Handwriting Experts

90. Three experts (two engaged by the Father plus an independent one) were provided with Wang's known signatures for comparisons. Those known signatures originated from various documents which came into existence at different times, including: 668 (March 1958), 669 (August 1958), 735 (October 1967), 675 (July 1975), 695 (December 1980), 672 (May 1983), and 671 (July 1983).

91. The experts also referred to known signatures of Tse in various documents, including 674 (a statement taken from Tse by the Police in May 1990 containing 9 signatures), 696, and 736 (a statement taken from Tse by the Police in July 1990 containing 7 signatures), 700 (an application for a identity card in June 1990 with a signature) and 1829 (a signature card for a bank saving account date November 1993).

92. The experts were subsequently provided with additional known signatures of Wang and Tse for comparison with their questioned signatures.

93. The additional known signatures of Wang include 6006 (March 1984), 6005 (March 1984), 5997(1) and 5997(2) (October 1984), 5998, 5999 and 6000 (March 1985), and 6001, 6002, 6003 and 6004 (June 1985).

94. The additional known signatures of Tse were those contained in or connected to his two statements taken on 6th and 9th of September 1999, namely 5995 and 5996.

95. The handwriting experts engaged by the Father said the questioned signatures of Wang and Tse were forged.

96. Apart from the aforesaid known signatures of Wang and Tse, the independent expert examined further signatures of Tse, including 5992 (1992), 5993 (1994), and 5994 (1979).

97. The independent expert initially said the questioned signatures of Tse were forged and those of Wang "might not have been written by him". After studying further known signatures of Wang, he said the questioned signatures of Wang "were probably not written by Wang".

98. The Father's handwriting experts also compared "A" with the writings on two documents said to contain writings of the Wife [a Power of Attorney by Wang to the Wife's sister- Gong Yan Shum in connection with his application to immigrate to USA (the Power of Attorney) and a "Fotomax" envelope] containing the following writings:

"Teh-han,

Sorry that the formalities for the requisite documents are rather troublesome and it is hard to handle matters in China. Regarding the birth certificate, you have to write a letter of authorization first before a form is obtained and then your signature shall be affixed thereto. Please write a letter authorizing my sister Gong Yan Shum to get an application form, which form will then be sent to you for your signature. Your date of birth may be incorrect."

 

"Teh Shyan,

Happy Easter (!)

T-shirt(s) will be sent again after (they) have been purchased. Would go out again to purchase those for Mui Mui (transliteration)."

99. Both Teh-han and Teh Shyan were transliterations of the Chinese name of Wang's sister, Teresa. Both documents ended with what appeared to be the initials of the Wife, "Ru Xin" (如心).

100. The experts engaged by the Father suggested that the Wife could have written the text in "A". The independent expert was unable to come to a definite conclusion.

101. An ink-dating expert engaged by the Father suggested that "A" and "C" only came into existence after 1996.

102. The Wife had six handwriting experts from the Mainland to study the questioned signatures of Wang and Tse. They were provided with more known signatures of Wang and Tse than those available to the Father's experts, including two disputed payment vouchers (100 and 101), dated 10 March 1990 and 31 March 1990 respectively, and a Home Visit Permit (issued in 1992), all said to contain additional signatures of Tse.

103. Three of the six experts engaged by the Wife, including Professor Jia Yu-wen ("Professor Jia"), produced a joint report on their findings and, after studying additional known signatures of Tse, a brief supplemental report. They opined that the questioned signatures of Wang and Tse in "the 1990 documents" were all genuine.

104. The joint report suggested that the questioned signatures of Wang in "the 1990 documents" -written by the same person- contained no features of imitation such as stroke pauses or rewriting, "rigid and awkward" stroke movements, or "apparently fast but in fact slow" stroke movements.

105. An overall comparison between Wang's questioned signatures and his known signatures indicated that the "calligraphies were similar" and that the "layouts of the characters, the locations of the signatures, and their styles were the same". The "bending" of the last vertical stroke in the character "Huei" 輝 towards the right was explained as a "writing habit".

106. The joint report compared Wang's questioned signatures with 668, 669, 5999, 6000, 6002, 6003, and 6004 to demonstrate the points of conformity.

107. Similarities in the writing details between Wang's questioned signatures and his known signatures of different decades up to the late 1980s were identified to support the suggestion that Wang's questioned signatures were genuine.

108. The joint report identified a number of differences between Wang's questioned signatures and his known signatures. It described these differences as "not fundamental but natural variations in the writing movements". It said that such differences "do not illustrate that the handwriting is not written by one and the same person."

109. The only significant difference identified by the Wife's experts was the obvious tremors in some of the strokes in Wang's questioned signatures. Such tremors were attributable to the injuries Wang suffered in March 1990. Hence such tremors (which could be indications of forgery) actually supported the genuineness of Wang's questioned signatures, as they were consistent with the injuries that Wang suffered at around that time.

110. Tse's questioned signatures in "the 1990 documents" (written by the same person) were described as normal handwriting, carefully and neatly written, with very few connected strokes and of the same level with his known signatures in terms of calligraphy.

111. Special efforts were made to show the similarities between Tse's questioned signatures with his known signatures made in 1999.

112. The joint report identified many points of conformity between Tse's questioned signatures and his known signatures. The differences in individual strokes between the two were said to have been caused by his known signatures "being written more quickly".

113. It was said that the differences were "natural variations of one person writing the same strokes at different times and in different conditions", not "differences of a fundamental nature".

114. The joint report identified tremors in Tse's questioned signatures. The tremors also appeared in his other known signatures (only in those of 1999). The tremors were described as a reflection of Tse's writing habit and not the result of imitation.

115. The joint report suggested that the questioned signatures of Tse were neatly written and the pen movements were smooth with no features of imitation such as "pauses in and rewriting of strokes, rigid and awkward stroke movements or apparent fast but in fact slow stroke movements".

116. The joint report concluded that the questioned signatures of Wang and Tse in "the 1990 documents" were all genuine.

117. The Wife's three experts prepared a supplemental report after examining three further documents purportedly bearing Tse's known signatures ["Home Visit Permit issued in 1992" (5992), "Hong Kong Certificate of Identity issued in 1994" (5993), and "Hong Kong Return Permit issued in 1979" (5994)].

118. The supplemental report identified a very high level of conformity between Tse's questioned signatures and his known signatures in the further documents. The differences were described as the result of different writing speeds.

119. The supplemental report concluded that the questioned signatures of Tse in "the 1990 documents" had almost identical characteristics with his known signatures on the further documents with no differences worthy of notice.

120. The further documents did not change, but fortified, the conclusion of the joint report.

121. The Wife's experts made no attempt to determine if the Wife authored any of "the 1990 documents".

The Trial

122. Apart from the experts' evidence on handwriting and ink dating, the Father and the Wife adduced evidence on the "probability" of Wang executing "the 1990 documents".

123. In fact, a lot of the evidence was common evidence and both parties tried to put a different gloss on such evidence to strengthen their respective cases.

124. The Father contended that the overall circumstantial evidence was inconsistent with the probability of Wang executing "the 1990 documents" and that the Wife had not proved their due execution. The court therefore should not pronounce them in solemn form.

125. The Wife's case was that the overall circumstantial evidence supported the probability of Wang's execution of "the 1990 documents" and that she had discharged the burden of proving their due execution. The court therefore should pronounce them as Wang's last will in solemn form.

126. Neither party had set out the "opposing" or the "supporting" circumstantial evidence in the pleadings. Both parties appeared content to proceed with the case on such basis. There had, of course, been extensive exchanges of witness statements relating to all the relevant circumstantial evidence.

127. Clearly, the genuineness or falsity of "the 1990 documents" stood or fell together. The genuineness or falsity of the questioned signatures of Wang and Tse did not necessarily stand or fall together. However, if Tse's questioned signatures were forged, the genuineness of Wang's questioned signatures would be very much in doubt.

Evidence For The Father

128. In addition to witnesses whose evidence was not substantially disputed, the Father, Teresa, and Yih gave evidence with a view to demonstrate the characters of Wang and the Wife, their relationship, and the general family background.

129. It was suggested that Wang was not a romantic person. He was, however, very careful about proper execution of documents and would not tolerate any documents that were not meticulously prepared. He was also a very private person and would never tell others of his own affairs. Evidence was led to show that Wang enjoyed very good relationships with his parents (particularly with his mother), and his siblings.

130. Wang provided his parents with a Chinachem flat and gave them a monthly maintenance of $19,000 (not an insignificant sum by Wang's standards, as he was known to be an extremely frugal person).

131. The Father said Wang gave him US$10,000 in 1988 when he went to America to visit his other children. Wang also gave, via the Father, US$100,000 and US$50,000 respectively to each of his two brothers on the same occasion.

132. The alleged execution of "the 1990 documents" was therefore inconsistent with Wang's character. There was no reason for Wang to make a fresh will in 1990 to leave his entire estate to the Wife and cut off the Father all together.

133. It was improbable for Wang, a very private person, to have Tse witnessing his signing "the 1990 documents". Further, being a very careful person, Wang would not have executed "the 1990 documents" in the careless manner suggested by the Wife and demonstrated by the documents themselves.

134. Teresa said that she had the opportunities to observe the Wife's writing (between 1954 and 1959 when they were practicing Chinese calligraphy and in the 1970s/1980s when they were holidaying in USA). There were also written correspondences between them. Teresa produced the Power of Attorney and the "Fotomax" envelope and said they contained the Wife's writing.

135. Teresa pointed out the signatures of the Wife- "Ru Xin" (如心) in the two documents. She said she was familiar with the handwritings of both Wang and the Wife. She was adamant that the two documents came from the Wife and that she recognized the writing as hers. Teresa also said the Wife had imitated Wang's signatures in the past and had shown such imitated signatures to her and Wang.

136. Teresa said the Wife wrote the text of "A".

137. Yih gave evidence on the development of Chinachem up to 1970, when he left the company. He said that the Wife did not play any significant role in running the company.

138. Far from being hostile to the Wife as suggested by Mr. Lee, Yih very often tried to protect her interests, e.g. by persuading Wang not to be suspicious of her faithfulness.

139. The Father, Yih, and Teresa were subjected to exceptionally lengthy cross-examinations, which lasted 10 or more days, not just to discredit them but also to show that the Wife had played an important role in Wang's business success. It was expressly suggested (with reasons) that Wang hated Yih and the Father.

140. The development of Chinachem was recounted. The Wang family history dating back to 1911, including jealously kept family secrets, was revealed. Wang was shown to be infertile. He was dismissed from secondary school for falsifying school reports in a university application. This incident, and not the lack of financial support, caused the cessation in his education. Further, there was evidence of infidelities of both Wang and the Wife.

141. The cross-examinations of the Father, Yih, and Teresa were extremely lengthy although they were restricted to the pre-1970 period (the Father was already semi-retired by mid 1960s and knew very little about the business of Chinachem; Yih left Chinachem in January 1970; Teresa left Hong Kong in 1959 to study in USA and had not been residing in Hong Kong since).

142. There was, however, no cross-examination of Teresa on the origin of the Power of Attorney and the "Fotomax" envelope or on her suggestion that the Wife was the author of "A".

143. Mr. Lee indicated reservation on the admissibility of the Power of Attorney and the "Fotomax" envelope when they were first introduced. The evidence was then introduced de bene esse. The objection on admissibility was formally raised only in his final written submissions. The judge overruled the objection.

144. Mr. Lee did not even suggest to Teresa that she was wrong to say that the Wife authored "A".

145. The aggressiveness, oppressiveness, unfairness, and aimlessness of the cross-examinations were exceptional. In the end, the Wife failed to establish what she had wished to establish. The judge dismissed much of the cross-examinations undertaken on her instruction as "irresponsible, vicious, opportunist" and "just a waste of time".

Evidence For The Wife

146. The Wife did not give evidence, although her lengthy witness statement was filed with court and was to stand as her evidence in chief pursuant to an earlier court order.

147. The Wife called a number of witnesses to show that she had indeed taken an active role in the development of Chinachem and that she and Wang were a "loving couples" (at least since the early 1970s). There was evidence on the riding accident on 10 March 1990 and documents were produced to support the trust arrangements in 1987 and in 1989-1990.

148. The Wife's brother, Dr. Kung, gave evidence on "the holiday wills".

149. Mr. Lam and Mr. Leung gave evidence on how Tse made the statements on 6th and 9th of September 1999.

150. The Wife made applications for and was granted leave to serve witnesses' statements out of time (in connection with the disputed vouchers 100 and 101 and the preparation of the "holiday wills"). In the end, most of those witnesses were not called.

Experts' Evidence

151. A number of experts gave evidence on handwriting. Two experts (Mr. Gus Lesnevich and Mr. David Tsui) gave evidence for the Father.

152. An independent expert (Mr. Patrick Cheng Yau Sang of the Government Laboratory) gave evidence on subpoena issued by the Father. His evidence also supported the Father's case.

153. Mr. Lesnevich examined and analysed the questioned signatures of Wang in "the 1990 documents" with reference to his known 1958 signatures (668, 669) as well as his post-1967 ones.

154. Mr. Lesnevich opined that Wang's 1958 signatures (668 and 669) were more controlled, more artistically written than the post-1967 ones, in which significant deteriorations and degenerations were visible.

155. Wang's questioned signatures in "the 1990 documents" bore similarities to his 1958 signatures but were completely different from his post-1967 ones.

156. Mr. Lesnevich opined that the author of the questioned signatures on "the 1990 documents" could not have been Wang.

157. Mr. Lesnevich also did a careful analysis of Tse's questioned signatures in "the 1990 documents" and compared them with his known signatures. His conclusion was that the questioned signatures of Tse contained such divergences in both character formation and manner of execution that the questioned signatures of Tse must have been forged.

158. Mr. Lesnevich concluded that the differences simply could not be reasonably explained.

159. Mr. Lesnevich had the further chance of examining Tse's twelve known signatures in two additional documents written in September 1999, shortly before Tse died. While the normal free-flowing and spontaneous writing habits and abilities shown in his earlier signatures were no longer present due to the obvious effects of aging and/or illness, the dissimilarities between the additional known signatures and the questioned ones were still obvious.

160. Mr. Lesnevich was asked if a single significant dissimilarity (one that was inexplicable and could not be accounted for) could lead to a finding of forgery, he said:

"It could lead to a finding of forgery. What I am saying is that you cannot identify. If you have that one major dissimilarity, I cannot tell you that signature is genuine. I am also not able to tell you that the signature is a forgery at that time."

161. Based on writing on the Power of Attorney and the "Fotomax" envelope, Mr. Lesnevich concluded that he was unable to rule out the possibility of the Wife being the author of "A". His conclusion on this aspect was quite innocuous.

162. Mr. David Tsui also compared the questioned signatures of Wang and Tse with their known signatures.

163. Mr. Tsui said that the questioned signatures of Wang in "the 1990 documents" bore some resemblance to his 1958 ones. However, he identified 18 points of discrepancies between the questioned signatures and his other known signatures.

164. In his initial report, Mr. Tsui concluded that the questioned signatures of Wang in "the 1990 documents" were likely to be forged.

165. In a supplemental report prepared after the examination of further known signatures of Wang, Mr. Tsui confirmed his earlier conclusion.

166. As for the questioned signatures of Tse in "the 1990 documents", Mr. Tsui was more definite. He positively concluded that those signatures were forged.

167. He said the overall writing of Tse's questioned signatures was slow, cautious (showing some tremors), and unnatural. He also said Tse's questioned signatures on "A", "B", "C", and "D" were found to be quite "superimposable" to each other (an indication that a tracing method had been used in forging the signatures).

168. The examination of Tse's further known signatures did not change Mr. Tsui's opinion.

169. Mr. Tsui also took the view that the text of "A" "could well have been written by the Wife". On a scale of -10 to +10, he put such possibility at +2.

170. Mr. Patrick Cheng Yau Sang also found significant discrepancies between Wang's questioned signatures and his known signatures. He identified slight tremors in Wang's questioned signatures and said that there were indications of slow writing and hesitation.

171. Mr. Cheng's conclusion in his initial report was a measured one. He said that Wang's questioned signatures "might not have been written by Mr. Wang Teh-huei".

172. After examining more known signatures of Wang (those of 1984 and 1985), Mr. Cheng's initial view was slightly reinforced. He said, "the questioned signatures were probably not written by Mr. Wang".

173. Mr. Cheng's opinion on Tse's questioned signatures in "the 1990 documents" was more definite. He said that those signatures were carefully and slowly written with "tremors" on a number of places. They were inferior to Tse's known signatures in fluency of writing and in line quality of writing. He also found significant discrepancies between Tse's questioned signatures and his known signatures.

174. Mr. Cheng concluded that Tse did not write his questioned signatures in "the 1990 documents".

175. The examination of further known signatures of Tse made in 1999 revealed significant changes from his earlier signatures, but Mr. Cheng's opinion on the questioned signatures of Tse in "the 1990 documents" remained unchanged.

176. Mr. Cheng offered the further opinion that the similarities between Tse's questioned signatures and his known signatures in 1999 suggested that the forger was modelling the forgery on Tse's 1999 or contemporaneous signatures. He said there were enormous changes in Tse's signatures between 1990 and 1999.

177. Mr. Cheng was unable to either eliminate or confirm if the Wife wrote the text of "A".

178. The Father's ink-dating expert (Mr. Speckin) took the view that "A" and "C" were not written in 1990 but "at least in 1996 or later".

179. At trial, only Professor Jia gave evidence for the Wife (the other two experts were present in court, but no clear indication was made that they would not be called until the very last moment).

180. In the course of his evidence (lasting almost 40 days), Professor Jia used a microscope to carefully examine the questioned signatures of Wang and Tse. He also used a projector to demonstrate the points he was making. Computer images were captured and produced as evidence.

181. Although Professor Jia confirmed the contents of the joint report, he did not rely on the similarities between the questioned signatures and the known signatures identified in that report. He said that wrong signatures were chosen in the joint report for comparison. He ostensibly tried to alienate himself from the joint report; hence his evidence-in-chief alone lasted 17 days.

182. In his evidence, Professor Jia referred to other known signatures of Wang to show similarities in "alignment" with Wang's questioned signatures. He concluded that Wang's signatures on "the 1990 documents" were genuine.

183. Professor Jia took the view that Wang had two styles of Chinese signature, a more cursive style and a more classical and formal style. He said that Wang's post-1967 signatures were signed in the former style, whereas his questioned signatures were signed in the latter style.

184. Professor Jia sought to explain the special features in Wang's questioned signatures by suggesting that Wang was actually signing them in the "Lishu" (隸書) style with "Xingshu" (行書) characteristics. He emphasized that Wang used a calligraphic pen for the questioned signatures to explain their unusual features.

185. Professor Jia denied that there were any signs of retouching and overwriting in Wang's questioned signatures. He sought to explain the indicia of such phenomena as a reflection of Wang's personal habit or the different conditions in which he wrote.

186. In his evidence, Professor Jia also put forward a "matching variation" theory. He pointed out minor natural variations within the questioned signatures of Wang and Tse. He also identified similar natural variations scattering among their known signatures.

187. Professor Jia suggested that the matching variations could not have been resulted by forgery, as the supposed forger could not have noticed such variations, let alone reproduced them. Therefore these matching variations supported the genuineness of the questioned signatures.

188. On the other hand, Professor Jia agreed with the suggestion put to him that one significant but unexplained difference could lead to non-identification of a signature.

189. Professor Jia carefully examined Tse's questioned signatures and tried to explain features of unnatural writing.

190. Professor Jia accepted the indications of slow writing movement, the changes in stroke direction, and the lack of smooth turning in Tse's questioned signatures. However, he disagreed with the suggestion that these features indicated forgery by tracing.

191. Professor Jia accepted what appeared to be differences between Tse's questioned signatures and his other signatures in 1990. He said the differences were caused by Tse's two ways of signing ("a fast written one" and "a slow written one"). Tse's questioned signatures belonged to the "slow written group," which included a few of his other known signatures as well.

192. The Wife attempted to introduce two vouchers with purported signatures of Tse (100 and 101, said to be made in 1990) to support her contention that there had not been significant changes in Tse's signatures between 1990 and 1999. As the questioned signatures were slowly written, they resembled the slowly written signatures of Tse in the two vouchers and in other documents.

193. Professor Jia sought reliance on Tse's Home Visit Permit (5992) to support part of his conclusion. However, there was no evidence adduced by the Wife to show the custodian of the Home Visit Permit. There was also dispute as to when Tse's purported signature was put on it.

194. It appeared that two different ball pens (both defective) were used to write Tse's purported signature and the first entry in his Home Visit Permit.

195. The Father challenged the two vouchers and Tse's Home Visit Permit (on both admissibility and authenticity).

Medical Evidence

196. The presence of tremors in Wang's questioned signatures prompted the Wife to call Dr. Brian Chao to show that the tremors were caused by the injuries that Wang suffered in the riding accident on 10 March 1990.

197. The suggestion was that the presence of the tremors (which could indicate forgery) actually supported the Wife's case.

198. However, Dr. Chao was unable to confirm how the accident affected (or could have affected) Wang's writing.

199. Further reports from Dr. H T Wu on the nature of Wang's injuries and their likely effects were read out. According to Dr. Wu, Wang was examined and investigated for concussion with abrasions and bruises. The concussion was not severe and did not require surgical intervention. Wang stayed in the hospital for two days and on 12 March 1990, he discharged himself against his doctor's advice.

200. Wang was reminded that he must return to the hospital immediately if he felt any fresh symptoms. He was also told to return after five days in any event because haemorrhage might occur subsequently. Wang, however, never returned to Dr. Wu after 12 March 1990.

201. Dr. Edmund Woo, called by the Father, disagreed with the suggestion that Wang's injuries could cause any tremors in his handwriting. He said, "if Wang had tremor in his hand on 12 March 1990 as a result of his head injury two days earlier, his tremor would have been rhythmically so severe that his signatures would have been much more illegible with greater distortion of the characters".

202. Despite repeated questioning under cross-examination, Dr. Woo was adamant that there was neither pathological basis nor clinical cases supporting the Wife's attempt to establish a correlation between the severities of a head injury with the severity of hand tremors.

203. Subsequent to Dr. Woo's evidence, Dr. Chao produced a second report to support the contention that a severe head injury would cause severe tremors and therefore a mild head injury would cause milder forms of tremors. Such contention was based on the following analogy:

"...There is no example in medicine where, when a severe insult results in a serious complication, a milder complication cannot occur either from a lesser insult of the same type, or in a less susceptible patient. Two examples will suffice to illustrate this point(s):

a) If a severe head injury can cause paralysis and dementia, milder forms of the same type of injury can lead to partial paralysis and mild cognitive dysfunction and an even milder one might just cause intermittent headache and impaired concentration.

b) If cigarette smoking can cause lung cancer in some patients, it is known also to cause lesser damage to the lungs, such as emphysema, chronic bronchitis, or even just a morning cough, or indeed nothing.

For the above reasons, far from disproving that mild head injuries might result in mild tremors, the Biary and Goetz paper would support such a connection."

The Judge's Findings

204. The judge found the Father, Yih, and Teresa to be truthful and reliable witnesses. The suggestion that Wang hated the Father and Yih was rejected as a "completely groundless and shifting" accusation.

205. The judge found that Wang was particularly close to his mother and "always had a deep gratitude and respect for the Father". His relationships with his siblings were also good.

206. The judge found that the Wife wrote the text of "A" and "D". His finding on the authorship of "A" was based on the evidence of Teresa and the handwriting experts. The judge, however, placed little weight on this aspect in reaching his final conclusion.

207. The judge's finding on the authorship of "D" was somewhat unusual. He took the view that "Nina," in relation to "One live one love," in "E" was positioned as if it was a signature. Since Nina was the Wife's name, the judge inferred that the Wife wrote in "E", "One life one love" and then signed her name, "Nina" underneath. As "D" and "E" appeared to have been written by the same person, the judge concluded that the Wife must have written "D".

208. The judge found Wang to be a meticulous businessman who was never slow in employing the services of lawyers. He also found Wang an unromantic person and would not have written "D" with "One life one love". The judge also found Wang a private person who would not have allowed Tse to know about his testamentary arrangements.

209. The couple's "loving relationship" was considered insignificant because it was based on outsiders' observations made after the 1970s. The judge further found that the Wife was adequately provided for and that Wang had only wished to "benefit the Wife by inter vivo provisions and not by testamentary provisions".

210. The judge expressly rejected the suggestion that Wang could have forgotten "the 1968 Will". He also rejected the suggestion that Wang had ever asked the Wife to empty the safety deposit box containing "the 1968 Will". He pointed out that the Wife was not authorized to access the safety deposit box.

211. The judge dismissed the evidence relating to "the holiday wills" as complete fabrication by the Wife, Dr. Kung, and their collaborators.

212. The judge found the circumstances of Tse's giving the statements and his subsequent movements suspicious. He had reservations about Mr. Lam's credibility, writing in his judgment, "Mr. Lam appeared to be very nervous and timid ... [He] was hesitant and evasive and was not a credible witness".

213. The judge pointed out Mr. Lam's inability to produce the attendance note or the bill of costs in connection with the statement taking on 6 September 1999. He also referred to Mr. Lam's attestation of Tse's statutory declaration, which was contrary to proper practice and conduct.

214. The judge took the view that Tse's statement on 6 September 1999 could have been a pre-prepared statement.

215. The judge also said Tse (having made the statement on 6 September 1999) would not find it difficult to repeat the same exercise on 9 September 1999.

216. The judge referred to Tse's suspicious departure from Hong Kong on 24 September 1999. In the end, he rejected the content of Tse's statements.

217. The judge made adverse remarks on the Wife's failure to give evidence after making, through Mr. Lee, serious attacks on the Father and his witnesses. The judge also commented negatively on the ways in which the various stages of the proceedings were conducted on her behalf.

218. The judge pointed out the absence of evidence to challenge the allegations made against her, her failure to explain the suspicions, and contradictions in her affirmations. He concluded that these did not support the Wife's case.

219. The judge extensively dealt with the handwriting evidence. He adopted both a macro- and a microanalysis of the questioned signatures. He conducted character-by-character/stroke-by-stroke comparisons of the questioned signatures with the known signatures of Wang and Tse.

220. He examined both the similarities of and the differences between the questioned signatures and the known signatures of Wang and Tse.

221. In particular, the judge identified all the similarities of Wang's questioned signatures and his genuine signatures, on which the Wife's experts relied to show that Wang had executed the questioned signatures. The judge rejected the conclusion of the Wife's experts.

222. The judge carefully analysed Wang's questioned signatures and his genuine signatures. He identified eight significant differences in the character "Wang" (王), six significant differences in the character "Teh," (德) and five significant differences in the character "Huei" (輝). He said that there was no acceptable explanation for those differences.

223. The judge further identified other significant differences in each of the three characters in Wang's questioned signatures.

224. The judge pointed out that there was no acceptable explanation as to why Wang's questioned signatures resembled his 1958 signatures instead of his post-1967 signatures. Wang had no reason to revert to the former on the occasion in question.

225. The judge rejected Professor Jia's suggestion that Wang, when signing "the 1990 documents," was exercising calligraphy and trying to sign in Lishu (隸書) style with Xingshu (行書) characteristics.

226. The judge considered the arguments of Professor Jia and his colleagues. He concluded that their arguments were unconvincing and evasive. Labelling them "hired guns," he expressed doubt about their credibility and reliability.

227. The judge rejected the suggestion that the riding accident supported the genuineness of Wang's questioned signatures, saying that the tremors in Wang's questioned signatures could have been "caused at least partly by the psychological condition of the forger".

228. Apart from the unnaturalness of Tse's questioned signatures, the judge identified more than 10 significant differences between those questioned signatures and Tse's genuine signatures.

229. The judge also identified six less conspicuous features in some of Tse's questioned signatures that suggested forgery, including incidences of retouching/rewriting, unnatural pen movements, lack of smooth turning, tremors, and slow writing.

230. The judge rejected Professor Jia's suggestion that the similarities between Tse's questioned signatures and his genuine signatures indicated that Tse wrote the questioned signatures. He also rejected Professor Jia's explanation about the discrepancies between Tse's questioned signatures and his genuine signatures.

231. The judge pointed out that the questioned signatures of Tse resembled those he signed in 1999 and not those he signed in 1990, despite the enormous changes in his signatures in the intervening years due to aging and/or illness. He thus opined that the forger had overlooked such changes.

232. The judge found Tse's questioned signatures in "the 1990 documents" to be highly "super-imposable".

233. Chinachem employees used the type of vouchers (those similar to 100 and 101) to apply for reimbursements; the applicants' supervisors normally countersigned them. The judge pointed out that the countersigning officer of 100 and 101 was not called to give evidence, although he had given a witness statement. Instead, another witness with no personal knowledge of the 2 vouchers and closely connected to the Wife (being her nominee to enable her to conceal her beneficial ownership of a major company) was called to give evidence. Also, his evidence turned out to be highly unsatisfactory.

234. Further, although there should have been corresponding numbered payment vouchers (issued when the requested reimbursements were made), 100 and 101 were unnumbered.

235. The judge therefore did not accept the two vouchers (100 and 101) to be vouchers bearing Tse's genuine signatures.

236. He found Tse's Home Visit Permit suspicious. In any event, as the purported signature was signed with a defective ball pen, the document should not be used for comparison.

237. In the end, the judge ruled that the experts could not use the two vouchers and Tse's Home Visit Permit for the purpose of comparison.

238. Based on the respective evidences of Mr. Lesnevich, Mr. Tsui, and Mr. Cheng, the judge found the questioned signatures of Wang and Tse in "the 1990 documents" to be forged (Wang's by simulation, Tse's by superimposition tracing).

239. Expressing his sentiments on the issue, the judge wrote in his judgment:

"Apart from being cogent and strong, they [the evidences] are to the extent that the only conclusion I can draw is that I have no doubt at all these eight signatures are nothing but forgeries."

240. He rejected the ink-dating evidence adduced by the Father. He also rejected the alternative argument of "lack of knowledge or approval" put forward by Wang's leading counsel, Mr. Chan in his final written submission.

241. In reaching his conclusions on the factual issues, in resolving the handwriting issues against the Wife and ink-dating issues against the Father, the judge relied extensively on counsel's respective submissions, which he reproduced in his judgment (sometimes word for word, pages after pages) without acknowledgement.

242. In addition, the judge referred to the following matters that he termed "suspicious circumstances":

1. Wang had no reason to change his mind, having made the Father a major beneficiary in the "1960 Will" and the sole beneficiary under the "1968 Will".

2. Wang was a sharp and prudent person. He was never slow in engaging the services of solicitors. There was no reason for him to prepare a homemade will when the legal services to him were so readily available. Wang was not the sort of person who would allow strangers to know about his affairs, particularly when the 1990 documents contained such strong sentiments against both his family and that of the Wife.

3. "A" was supposed to have been written by one person and "B" and "C" by another. Wang could have asked them to witness his signatures on those documents. The judge took the view that if Tse did go to Wang's office on 12 March 1990 to attest Wang's signatures, his presence would have been noticed. The judge concluded that Wang could not have caused any of the documents to be written on 10, 11 or 12 March 1990. He also concluded that the use of three pieces of thin paper and one piece of thicker paper was to facilitate the production of forged signatures of Tse by superimposition tracing.

4. The 1990 documents, particularly "A" were sloppily prepared in Chinese with characters over-written and/or written with faint colour ink. "One life one love" not only appeared on "D", but also faintly on the other three pieces of paper. The judge said such sloppiness was inconsistent with Wang's character, not when he was supposed to be making a will. The judge also said the "One life one love" declaration was inconsistent with the character of Wang as he was not known to be a romantic person whereas the Wife appeared to be so.

5. The judge found it difficult to understand why Wang wanted "D" to be attested when it was not testamentary in nature but just an expression of love and devotion and why he did not sign in English. The judge also expressed surprises that "B" and "C" were not combined into one and that the contents were conflicting with one another. It was pointed out in one document Wang showed concern for his parents and sister while in another prohibited the Wife from giving them any money or property. The judge took the view that the contents of "B" and "C" were unreasonable and untrue when there was no reason for Wang to be disappointed with his family or that of the Wife. The judge pointed out that Wang's relationship with his parents, particularly his mother had always been good and that the Wife's mother and brother had also been nice to him. The judge further took the view that the Wife had more reasons to find the Wang family disappointing and the Kung family disgusting. The judge pointed out that Wang was never a romantic person and did not in the past find it necessary to justify his decisions.

6. The absence of the revocation clause in "A" was more consistent with it being forged by someone who did not know the existence of the earlier wills or their contents.

7. On the evidence adduced by the Father, which the judge accepted, Tse was a talkative person. The judge took the view that Wang would not have liked nor trusted Tse to witness the execution of his will. The judge referred to the special relationship between the Wife and Tse. He had also carefully analyzed the contents of the two statements given by Tse and the circumstances in which such statements were given. The judge took the view that the content of the 1st statement was so neat and so befitting the Wife's case that it must have been prepared/written as a result of coaching. The judge also referred to the suspicious movement pattern of Tse after he commenced his leave on 14 September 1999.

8. The Wife appeared to have known that the signature of Lee K Sg also appeared on "A" on 6 September 1999. In the light of what transpired that day, the Wife must have seen "A" before the envelope was opened.

9. The Wife's behaviour in propounding "the 1990 documents" as the last will of Wang. The judge made special reference to various affirmations by the Wife. Those affirmations indicated that she was aware of the contents of "A" before it was opened on 6 September 1999. She was able to say that "the 1968 Will" had been superseded and revoked and that she was appointed the sole beneficiary of Wang's later will. Such matter was inconsistent with the suggestion that she had been asked by Wang not to open the envelope until after his death.

243. The judge found these "suspicious circumstances" supportive of the Father's case. In reaching his conclusion on the "suspicious circumstances," the judge again reproduced substantial parts of the submissions by counsel for the Father.

244. The judge emphasized that the Wife had failed to dispel the "suspicious circumstances," upon which basis the judge also concluded that "the 1990 documents" were forged.

245. A fortiori, from the experts' handwriting evidence together with the "suspicious circumstances", the judge concluded that he had "no doubt that ...the 1990 documents are forged documents".

Grounds of Appeal

246. We had been provided with hundreds of pages of written submissions from counsel for the Wife, together with significant amounts of annexure.

247. Mr. Michael Thomas SC and Mr. Denis Chang SC, leading counsel for the Wife, amplified the written submissions with extensive oral arguments on factual issues and on experts' evidence respectively.

248. The complaints by the Wife were directed mainly at the judge. It was suggested that the judge had wrongly identified the issues involved in the case.

249. Mr. Thomas said the judge made three fundamental mistakes. Firstly, instead of confining himself to the only two issues, namely, whether Wang had duly executed "the 1990 documents" and whether the Father had established the allegation of forgery, the judge wrongly cast the burden on the Wife to dispel "suspicious circumstances." Secondly, such "suspicious circumstances" had not been provided as particulars of fraud alleged against the Wife and therefore the judge should not have used them to establish forgery. Thirdly, in finding forgery on the basis of such inadequate evidence as the "suspicious circumstances," the judge applied the wrong burden and standard of proof.

250. Mr. Thomas suggested that the length of the trial, the emotional intensity of the case, and the extensive media coverage might have created difficulties for the judge to the extent that he was unable to maintain impartiality or to deal with the case properly.

251. Mr. Thomas said that the Wife had not been given a fair trial because the judge accepted the Father's entire case and rejected the Wife's without giving adequate reasons.

252. On the issues of circumstantial evidence, Mr. Thomas contrasted the extensive adoption by the judge of the submissions made on behalf of the Father and the little reference to those made on behalf of the Wife.

253. The judge was criticized for his discouraging oral submissions and demanding detailed written closing submissions.

254. The judge was further criticized for failing to properly weigh the parties' cases and their related evidence. The suggestion was that the judge, "without any sense of balance," had pre-judged the issues and ignored "the overwhelming probabilities in favour of genuineness" of the questioned signatures.

255. Insofar as the judge made unfair findings of facts and drew unreasonable inferences against the Wife, it was the result of bias as demonstrated in his lengthy judgment that adopted substantially the written submissions made on behalf of the Father and rejected those made on behalf of the Wife without adequate explanations.

256. In the written submissions for the Wife, the judge was accused of "wholesale adoption, copying (substantially verbatim)". An annexure highlighting identical parts of the closing submissions of counsel at trial and of the judgment was included to demonstrate the extent of the judge's reproduction of counsel's submissions.

257. Such reproduction, Mr. Thomas argued, demonstrated the lack of impartial and independent evaluation of the parties' contentions; the trial was therefore not a fair one.

258. The written submissions devoted almost 60 pages to matters said to be supportive of the probability that Wang made a will in favour of the Wife in 1990. The suggestion was that the judge had completely ignored these matters.

259. Mr. Thomas suggested that most of the "suspicious circumstances" (those identified by the judge and perhaps others) were irrelevant because they were not at all probative to the issue of due execution. He further argued that the judge had wrongly made use of such "suspicious circumstances" to reach his conclusion on forgery.

260. Mr. Thomas emphasized that the "suspicious circumstances" could not by themselves amount to proof of forgery; hence when the judge so found, he was in error.

261. Insofar as the "suspicious circumstances" were relied on to support an allegation of "want of knowledge and approval" (Order 76 Rule 9), neither the allegation nor the "suspicious circumstances" had been pleaded.

262. Mr. Thomas said, "Had the 'suspicious circumstances' been pleaded and particularized, the evidence would have been more extensive and the case would have been fought differently".

263. Mr. Thomas complained that the Father did not declare his intention to rely on the "suspicious circumstances" to prove that Wang did not duly execute "the 1990 documents" until the conclusion of the trial. Mr. Thomas said that the particulars in support of such argument had not been pleaded.

264. In any event, Mr. Thomas argued, the judge was wrong to interpret the "unpleaded" "suspicious circumstances" in such a way to cast doubt on the due execution of "the 1990 documents".

265. Another 70 pages of the written submissions were devoted to the analysis of the "suspicious circumstances" to demonstrate the fallacies of the judge's approach. Mr. Thomas then expanded upon such written submissions orally.

266. Mr. Thomas emphasized the following:

(i) The Wife was the obvious heir, having reconciled with Wang;

(ii) Wang must have forgotten about the "1968 Will";

(iii) The riding accident on 10 March 1990 unsettled Wang, who then made a provisional will on the spur of the moment out of worry of an unexpected death;

(iv) The homemade will was to ensure the continuity of Chinachem;

(v) Wang had the opportunity on 12 March 1990 to prepare "the 1990 documents";

(vi) The sloppy drafting, convoluted expressions, and unreasonable and untrue content of "the 1990 documents" were curiosities more consistent with the documents' genuineness than with forgery;

(vii) The absence of a revocation clause was innocuous;

(viii) The statements of Tse and the circumstances in which they were taken should not be lightly rejected. There could well be innocent personal reasons for Tse to leave Hong Kong on 24 September 1999 rather than the nefarious one speculated upon by the judge;

(ix) The Wife had the opportunities to learn of the presence of "Lee K Sg's" signature in "A"; and

(x) Her behaviour in propounding "the 1990 documents" as the last will of Wang was consistent with her honest belief that Wang was still alive.

267. Mr. Thomas suggested that if "the 1990 documents" were the result of fraud, the fraudster could have done a much better job. The fraudster would not want to make things more difficult for him or herself by forging more than one signature with a calligraphic pen.

268. Mr. Thomas also pointed out that if the Wife could manipulate Tse as suggested, there would be no need to forge Tse's signatures on "the 1990 documents". It would have been easier to have Tse sign "the 1990 documents".

269. The judge's adverse findings against Mr. Lam were highlighted to show that the judge had not been fair to the Wife's case.

270. The judge was further criticized for placing undue reliance on Wang's character when the supporting evidence only came from observations made about him over 20 years ago.

271. Mr. Thomas submitted that the evidence of the handwriting experts must be considered together with the overall probabilities of the case and that the conclusion of forgery should not be made on the handwriting evidence alone. Mr. Thomas, however, sought to argue that the "suspicious circumstances" identified by the judge were not probative to the issue of forgery.

272. The judge was said to have adopted a wholly artificial approach of "splitting and slicing up the issues into self-contained packages".

273. A similar criticism was made against the judge's handling of the handwriting evidence in that he concentrated on individual points of similarity or difference and failed to take a consolidated evaluation of all the evidence.

274. Insofar as the judge relied more on differences than on similarities between the questioned signatures and the known signatures, his approach constituted an "inherent and structural bias" in favour of finding forgery and thus had the effect of shifting the burden of proof on "an operational level" against the Wife.

275. The approach adopted by the judge was said to be mechanical in that any "slight variation beyond the acceptable range" was viewed as a significant difference indicating forgery.

276. Mr. Chang suggested that when the judge ignored the "matching variations" on the basis that the forger might have copied from more than one model, he was converting evidence of genuineness into that of forgery.

277. Mr. Chang argued that both Wang's questioned signatures and his known 1958 signatures (668 and 669) were modelled on the same styles of Chinese calligraphy- Lishu (隸書) with Xingshu (行書) characteristics- and therefore the similarities between them were natural, particularly when fountain/calligraphic pens were used in both cases.

278. Mr. Chang emphasized that the use of a broad-nib calligraphic pen could have produced the differences and special features in the questioned signatures. He further complained that the judge not only ignored the effects of the calligraphic pen but also relied on the differences and special features caused by the use of the calligraphic pen to find forgery.

279. Mr. Chang suggested that a person could have more than one style of signature. In determining a signature's genuineness, it was important to restrict comparisons to signatures of the same style. It was also important to consider such factors as writing instrument and material, writing speed, and nature of document.

280. The suggestion was that in signing "the 1990 documents," Wang must have abandoned his usual cursive style in favour of the formal and artistic style. The use of a calligraphic pen not only demonstrated Wang's intention to sign formally and artistically but also explained the upward tilt of the first horizontal stroke in the character Wang "王" (indeed a most distinct feature) and other unusual features in his questioned signatures.

281. The judge was criticized for underestimating the importance of knowledge of Chinese in the examination of Chinese signatures (specifically legible ones).

282. Insofar as the judge identified further and other significant differences not mentioned by the Father's experts and contrary to Professor Jia's evidence, the judge indulged in speculation and "played expert".

283. Each of the judge's comments on Professor Jia was put to microscopic examination. The judge's criticisms of the professor were said to be unfair; the rejection of his evidence was described as symptomatic of an indiscriminate adoption of the submissions by the Father's counsel.

284. Detailed and lengthy analyses of the questioned signatures-character-by-character, stroke-by-stroke, dot-by-dot-were made to validate the "matching variation" theory, to demonstrate that the "significant differences" identified by the judge as indicative of forgery were present even among Wang's genuine signatures, and hence to support the contention that Wang's questioned signatures must be genuine.

285. Mr. Chang said that the judge was wrong to find that the Wife authored "A" and "D" because of the following reasons:

(i) The allegation that the Wife authored "A" and "D" was never pleaded.

(ii) The authenticity and authorship of the documents (the Power of Attorney and the "Fotomax" envelope) for comparison with "A" were doubtful.

(iii) The judge wrongly placed the burden of proof on the Wife.

(iv) The above points notwithstanding, the experts' evidence did not support the judge's conclusion.

286. Mr. Chang also suggested that the judge was wrong to exclude the two vouchers (100, 102) and Tse's Home Visit Permit from the samples for comparison.

287. Mr. Thomas invited this court to review the evidence comprehensively, in accordance with the principle set out in Onassis and Calogeropoulos v Vergottis [1968] 2 Lloyd's LR 403, Benmax v Austin Motor Co Ltd [1955] All ER 326 and Assicurazioni Generali Spa v Arab Insurance Group [2003] 577.

288. Mr. Thomas suggested that in considering the question of due execution, the court should adopt a "benevolent approach". Relying on the presumption omnia rite esse acta, he argued that the mere production of "the 1990 documents" was good enough evidence to prove their due execution. (Emphasis added)

289. The above points notwithstanding, Mr. Thomas said, there was substantial evidence supporting the due execution of "the 1990 documents" by Wang, specifically Tse's affirmations. He suggested that this court should either allow the appeal by giving judgment for the Wife or at least order a re-trial.

The Approach of This Court

290. In identifying the issues in this appeal, I am slightly troubled by the variance among the notice of appeal, the written submissions filed on behalf of the Wife, and the oral submissions of Mr. Thomas.

291. Mr. Thomas did not deal with a number of points set out in the notice of appeal and/or the written submissions. He did not clearly indicate that these points had been abandoned either.

292. For the sake of succinctness, I have tried to integrate the notice of appeal, the written submissions, and counsel's oral submissions in my summary of the Wife's grounds of appeal. I hope that this approach has not done any injustice to counsel or the parties; it can be assured that every point made by counsel has been taken on board.

293. Regarding the approach of an appellant court, I here set out the principles laid down by the authorities to which Mr. Thomas referred. Mr. Robert Tang SC, leading counsel for the Wife, raised no objection to such principles.

294. In the judgment of Onassis (supra), Lord Pearce writes,

"The function of a Court of Appeal is to set aside a judgment that should not be allowed to stand because it occasions a substantial wrong or miscarriage of justice. That wrong or miscarriage of justice may consist of a judgment in favour of the wrong party. It may also consist of a failure in the judicial process to which both parties are entitled as of right, namely, the weighing of their respective cases and contentions. Such failure may constitute a wrong or miscarriage of justice even though it may appear that the appellant may in the end fail to secure a judgment in his favour. But the fact that the right party seems to have succeeded in the Court below will naturally make a Court of Appeal extremely reluctant to interfere, and it would only do so in the rarest cases. Such matters are questions of degree." (p. 430)

Viscount Dilhorne writes in the same judgment,

"Usually the power to order a new trial is only exercised where a trial has been with a jury and the usual consequence of a successful appeal from a judge sitting alone is that the judgment is reversed or altered. It must, however, be recognized that in some cases, and it is said that this is one, although there may have been error in the course of the trial or in the judgment sufficient to give rise to a substantial wrong or miscarriage, it is not possible to do justice by reversing or amending the judgment. In such a case the only possible order is an order for a new trial (see Jones and Others v Hough and Others, (1879) 5 Ex. D. 115, per Lord Justice Cotton, at p. 125)." (p. 405)

In Benmax (supra), Lord Reid adopts the following passages from other leading cases:

"The appellate court, either because of the reasons given by the trial judge are not satisfactory, or because it unmistakably so appears from the evidence, may be satisfied that he has not taken proper advantage of his having seen and heard the witness, and the matter will then become at large for the appellate court. It is obvious that the value and importance of having seen and heard the witnesses will vary according to the class of case, and, it may be, the individual case in question." (p.329 D-F)

"But in cases where there is no question of the credibility or reliability of any witness, and in cases where the point in dispute is the proper inference to be drawn from proved facts, an appeal court is generally in as good a position to evaluate the evidence as the trial judge, and ought not to shrink from that task, though it ought, of course, to give weight to his opinion," (p.329 G-H)

"Upon appeal from a judge where both fact and law are open to appeal, it seems to me that the appellate tribunal is bound to pronounce such judgment as in their view ought to have been pronounced in the court from which the appeal proceeds, and that it is not within their competence to say that they would have given a different judgment if they had been the judge of first instance, but that because he has pronounced a different judgment they will adhere to his decision." (p.329 H-I)

In Assicurazioni (supra), Clarke LJ says,

"The approach of the court to any particular case will depend upon the nature of the issues (and the) kind of case determined by the judge...In some cases the trial judge will have reached the conclusions of primary fact based almost entirely upon the view which he formed of the oral evidence of the witnesses. In most cases, however, the position is more complex. In many such cases the judge will have made findings of primary fact based entirely or almost entirely on the documents. Some findings of primary fact will be the result of direct evidence, whereas others will depend upon inference from direct evidence of such facts.

In appeals against conclusions of primary fact the approach of an appellate court will depend upon the weight to be attached to the findings of the judge and that weight will depend upon the extent to which, as the trial judge, the judge has an advantage over the appellate court; the greater the advantage the more reluctant the appellate court should be to interfere.

Some conclusions of facts are, however, not conclusions of primary fact of the kind to which I have just referred. They involved an assessment of a number of different factors which have to be weighed against each other. This is sometimes called an evaluation of the facts and is often a matter of degree upon which different judges can legitimately differ. Such cases may be closely analogous to the exercise of a discretion and, in my opinion, appellate courts should approach them in a similar way." (p. 580)

295. If an appellant alleges a failure of the trial judge to give adequate reasons, the approach of the Appellate Court is set out in English v Emery Reimbold & Strick (supra) on page 2419:

"Where permission is granted to appeal on the grounds that the judgment does not contain adequate reasons, the appellate court should first review the judgment, in the context of the material evidence and submissions at the trial, in order to determine whether, when all of these are considered, it is apparent why the judge reach the decision as he did. If satisfied that the reason is apparent and that it is a valid basis for the judgment, the appeal will be dismissed."

Procedural Irregularities

296. Earlier, I have expressed my sentiments on the length of the trial. I do not think the judge alone should be blamed. Clearly, Mr. Lee's cross-examination of the Father's witnesses and Professor Jia's evidence greatly prolonged the trial.

297. The Father, Yih, and Teresa were called for the limited purpose of providing background information on Wang's character, his family, and the pre-1970 development of Chinachem. Yet, they were subjected to many days of cross-examination on irrelevant or, at best, peripheral matters based on untrue or unsubstantiated allegations.

298. The pre-trial exchanges of experts' reports should have shortened the hearing time. However, Professor Jia, who turned out to be the only live expert witness for the Wife, almost completely abandoned the joint report that he co-authored and spent 17 days on his evidence-in-chief alone.

299. Two to three days were completely wasted because Mr. Lee had trouble finding a photocopier that would produce copies of a certain desired quality.

300. Because the due execution of "the 1968 Will" was not admitted, Mr. Donald Cheung had to give evidence in Canada via video-link. His evidence on the execution of "the 1968 Will", however, was never challenged.

301. Without a doubt, the trial was difficult, but there was no indication that the judge was unable to deal with it properly. The judge showed a keen interest in the case, particularly the experts' evidence (as Mr. Thomas agreed). He reminded counsel frequently of the issues to focus on and dealt with every conceivable point arising out of the dispute.

302. It is not unusual for counsel to submit detailed closing submissions in writing. In a case of exceptional length, such submissions are particularly helpful.

303. At trial, counsel on both sides agreed to the judge's request for such written submissions. Moreover, counsel had been given adequate opportunities to further assist the court with viva voce submissions and there was no request for more time.

304. The complaint that the judge discouraged oral submissions and asked for detailed closing submissions should not have been raised. (Mr. Thomas did not pursue this complaint in his oral submissions.)

Relevance of the "Suspicious Circumstances"/Circumstantial Evidence

305. The case started and ended with the same issues. The pleadings were short and simple; they did not give rise to any complication.

306. The Father wished to propound "the 1968 Will". The Wife resisted on the basis that "the 1968 Will" had been revoked by "the 1990 documents" and filed a counterclaim. The Father said "the 1990 documents" were forged.

307. As the propounder of "the 1990 documents," the Wife had to establish its due execution. Apart from experts' evidence, she adduced circumstantial evidence to demonstrate that she was a reasonable and deserving heir to Wang's estate.

308. The Father challenged the Wife's case by adducing circumstantial evidence to demonstrate the improbability of Wang's execution of "the 1990 documents". In addition, he adduced experts' evidence to prove that such documents were forged.

309. However, at trial, Mr. Lee downplayed the importance of circumstantial evidence on the issue of due execution of "the 1990 documents," writing in his final submissions that "[the issue was] forgery and nothing else, experts' evidence and nothing else".

310. Mr. Thomas was right to draw a distinction between the issue of due execution and the issue of forgery. But if there were forgeries, there could not have been due execution. In the context of the present case, if the court was not satisfied with due execution, it necessarily follows that the court was not satisfied with the genuineness of Wang's questioned signatures in "the 1990 documents".

311. Hence, the two issues proffered by Mr. Thomas are in fact interrelated, although each involves a different burden of proof and a different standard of proof.

312. Mr. Thomas made lengthy submissions on Order 76(9), repeatedly stressing that the "suspicious circumstances" encapsulated in the issue of "want of knowledge and approval" must be pleaded with sufficient particulars and could not be introduced by way of written submissions.

313. Mr. Thomas cited leading textbooks on probate matters and cases after cases to support the contention that the issue of "want of knowledge and approval" must be expressly pleaded with full particulars before it could be relied on.

314. In the cases referred to by Mr. Thomas, whether the deceased had executed the purported wills was not in issue; there was in fact no dispute over the deceased's signatures.

315. The issues were whether there had been due execution in accordance with the statutory requirements and whether the deceased knew or approved the contents of the purported wills.

316. In cases involving the issue of "knowledge or approval," allegations that the deceased did not know or did not approve the contents of the purported will must be pleaded, together with the full particulars in support thereof (duress, undue influence, unsound mind, signature without reading or explanation, etc.).

317. This principle, however, has no bearing whatsoever in a case in which the issue is whether or not the deceased had executed the purported will.

318. I am sure that Mr. Thomas did not intend it, but the reference to Order 76(9) was a complete red herring. It is not a live issue in this appeal.

319. The Father's primary case was not that Wang executed "the 1990 documents" or signed his names without knowing or approving of their contents. The Father's primary case was that Wang did not execute "the 1990 documents" as Wang's questioned signatures were forged.

320. Apart from experts' evidence on the authenticity of the questioned signatures, the Father adduced "opposing circumstantial evidence" to demonstrate that it was improbable that Wang executed "the 1990 documents," and the Wife adduced "supporting circumstantial evidence" to demonstrate that it was probable.

321. Mr. Edward Chan SC, leading counsel for the Father at trial, did not raise until his final written submissions, the alternative argument that if Wang indeed signed "the 1990 documents," he did so without knowing or approving their contents.

322. The alternative argument was not meant to be a substitute of the Father's primary case, nor did it preclude him from relying on the issues raised on the pleadings.

323. Mr. Lee objected to the alternative argument of Mr. Chan on the basis that it had not been pleaded in accordance with Order 76(9). The objection was valid and was accepted by the judge. This issue did not need to be resurrected.

324. However, the "suspicious circumstances" must not be regarded as only relevant to the issue of knowledge or approval. They are equally relevant to the issue of due execution.

325. In a case of due execution or forgery of a will in which only the dead and/or the forger has direct knowledge of the authenticity or falsity of the will, circumstantial evidence supporting or opposing due execution is highly relevant.

326. Mr. Tang quoted extensively from the judgment of Scarman J (as he then was) in In the estate of Fuld, Dec'd (No.3) [1968] P 675.

327. Fuld's case concerns the issue of knowledge and approval. Nevertheless I agree with Mr. Tang's observation that the reasoning applies equally on the issue of execution.

328. The court must ensure that the decision to grant probate accords with the true intention of the deceased. Even when there is no dispute over the genuineness of the signature of the decreased, the court must be vigilant and cautious so as to ascertain the deceased's knowledge and approval of the content of the purported will. In such cases, circumstantial evidence is relevant and important.

329. Concerning this point, I refer to the wisdom of Scarman J in the judgment of Fuld (supra);

"It is sufficient now to indicate that it is a rule which in certain cases requires of the court vigilant care and circumspection in investigating the facts of a case. It is a rule which calls upon the court not to grant probate without full and entire satisfaction that the instrument did express the real intentions of the deceased." (p. 698 A-B)

"The English court is being asked to grant probate in solemn form. Upon it falls the responsibility of deciding whether the instruments propounded express the real intentions of the testator. In my judgment the discharge of this responsibility is a matter for the judicial conscience of the court, guided in the business of investigation and proof by its own lex fori." (p. 699 A-B)

"His argument, as I understand it, is that the 'vigilant and jealous scrutiny' called for in Wintle v Nye is required when the instrument under challenge has been drawn by one who takes a substantial benefit under it.....Even if authority were lacking, I would not accept the submission. I would have expected the law to require the court to exercise scrutiny whenever a case revealed reasonable grounds for suspicion. But fortunately, there is authority. In Tyrrell v Pinton, the Court of Appeal held that the rule throwing upon the party propounding a will the burden of showing that it expresses the true will of a deceased is not confined to cases where the will is prepared by a person taking a benefit under it. The weight of the burden will, I readily accept, vary with the weight of the suspicion to be dispelled: but in a matter as vital as this the law wears no cramping straitjacket. The court's vigilance is called for whenever circumstances reasonably excite suspicion." (p. 712 D-F)

"The court must not allow itself to be led to conclusions upon the strength of presumptions-in blinkers, as it were. When circumstances put it upon inquiry, as they assuredly do here, the judicial conscience must be satisfied before the instrument can be admitted to probate." (p. 716 A-B)

"Those who know are not prepared to reveal the whole truth. Darkness and suspicion are common features in will cases: the truth too often is the secret of the dead or the dishonest. Because it is often difficult, and sometimes impossible, to discover the truth, the law insists on two types of safeguard in will cases. The first type of safeguard is part of the substantive law- the requirements of proper form and due execution. Such requirements which in this case have to be decided in accordance with German law, are no mere technicalities. They are the first line of defence against fraud upon the dead. The second type of safeguard is the second line of defence. It is invoked where there are circumstances which give rises to suspicion; it is the safeguard of strict proof. In cases where no suspicions reasonably arise the court will allow inference- presumptions as they are sometimes called- to be drawn from the regularity of a testamentary instrument upon its face, or the fact of due execution. But if there are circumstances, whatever be their nature, which reasonably give rise to suspicion, the court must be on its guard. It must ensure that the burden of proof rests upon the party propounding the will: and 'he must satisfy the conscience of the court that the instrument so propounded is the last will of a free and capable testator." (p. 719 B-E)

330. If the genuineness of the deceased's signature is in issue, a fortiori the court must be equally, if not more vigilant, and cautious so as to ensure that the purported will truly represents his testamentary intention, i.e. that the purported signature of the deceased is indeed genuine. Therefore, the "suspicious circumstances" present in this case should not be regarded as irrelevant.

331. Section 5(2) (now amended) of the Wills Ordinance stipulates that a Chinese will, unlike other wills, does not have to be witnessed by two or more witnesses both present at the same time to be valid. In the case of a Chinese will, the safeguards identified in Fuld are even more important. It is not difficult for a Chinese will with forged signature to be put forward as the last will of a deceased.

332. The circumstantial evidence supporting or opposing the due execution/authenticity of "the 1990 documents" had not been pleaded. Neither party took issue with it at the trial.

333. When the Wife adduced supporting circumstantial evidence, the Father was of course entitled to adduce opposing circumstantial evidence to challenge the Wife's case. It would be illogical to forbid the Father from relying on circumstantial evidence when the Wife herself did so.

334. Theoretically speaking, the Father might not even have full knowledge of what evidence to adduce to challenge the Wife's case on due execution until her case was presented in court. He could not possibly set out the "full particulars" of the supporting circumstances in the pleadings.

335. The authorities suggest that collateral circumstances relevant to the execution of a will may be pleaded. Sir John Nicholl wrote in Saph v Atkinson 164 ER 57:

"Now, where the inquiry is, whether an asserted will was, or was not, executed by all alleged testator, all such collateral circumstances may be pleaded and proved on either side, as have a tendency to shew, on the one hand, the probability, and on the other, the improbability, that it should have been so executed."

336. There is no suggestion that such circumstances must be pleaded. In a probate action-an inquisitorial hearing when the court tries to establish the true intention of a deceased-a rigid approach to the rules of evidence may not always be necessary. [See the judgment of Karminski J in Trotman v Trotman (infra)]

337. The Father and the Wife relied on a lot of common facts in furtherance of their respective cases. Each, however, tried to interpret such facts in ways most favourable to their respective cases.

338. In light of the above points, Mr. Thomas's complaint against the Father's failure to plead the collateral circumstances is unjustified.

339. Whether collateral circumstances were pleaded or not, there is no doubt that both parties were fully aware of the other party's position at the outset and that they had ample opportunities to meet each other's case. There had been extensive exchanges of witnesses' statements prior to trial. Also, the trial lasted a very long time.

340. The Father, Yih, and Teresa gave evidence on the circumstantial evidence not just to support "the 1968 Will," but also to show the improbability of Wang executing "the 1990 documents". There was no objection to such evidence.

341. Mr. Lee subjected the Father, Yih, and Teresa to days and days of cross-examination in order to undermine their credibility and hence the veracity of the circumstantial evidence supporting the Father's case. His cross-examination did not touch upon "the 1968 Will", the due execution of which turned out to be undisputed.

342. Heavy reliance, on the other hand, was placed on the circumstantial evidence adduced by the Wife to establish her case.

343. The Father's case was opened with reference to the circumstantial evidence, although not in details. The judge had reminded Mr. Lee that the case depended on such matters and not only "signature and handwriting".

344. Yet, curiously, Mr. Lee did not deal with the Father's submissions on the circumstantial evidence, seeking instead to dismiss such evidence on the basis of Order 76 (9). Mr. Lee was even bold enough to suggest that the case hinged on experts' evidence alone.

345. Initially, Mr. Thomas suggested that experts' evidence and the circumstantial evidence must be considered together to determine the overall probabilities of the case. However, in his reply, he seemed to have modified his position, suggesting that the only relevant evidence on the issue of forgery was that of the experts.

346. Nevertheless, Mr. Thomas repeated the attack on the Father's case on the basis that the circumstantial evidence adduced by him was not pleaded. Mr. Thomas further suggested that if it had been pleaded and particularized, the Wife would have adduced more evidence and the case would have been fought differently.

347. In my view, Mr. Thomas was wrong to suggest that the Father only made clear his case (i.e. that the suspicious circumstance was inconsistent with Wang's due execution of "the 1990 documents") at the end of the trial. Quite the contrary, it was evident in the statements of the witnesses called by the Father. Further, throughout the trial, Mr. Chan had clearly indicated his reliance on circumstantial evidence, although he also put forward a further alternative suggestion at a very late stage.

348. The judge had reminded counsel at the beginning that the circumstantial evidence was relevant.

349. Mr. Lee must have been aware of the relevance of circumstantial evidence to the issue of due execution. Otherwise there could be no reason for the extensive cross-examination he directed at the Father, Yih, and Teresa.

350. It was indeed quite surprising that Mr. Lee decided not to address the court on the circumstantial evidence relied on by the Father. Mr. Tang, leading counsel for the Father, suggested that the reasons for Mr. Lee's decision were "forensic". I do not wish to speculate on the matter. I assume that there must be "good reasons" for Mr. Lee's approach.

351. However, it is wrong to suggest that the Father could not rely on such circumstantial evidence. It is also wrong to suggest that the use of such evidence caught the Wife by surprise.

352. The two grounds of the appeal relating to circumstantial evidence-inadequate pleadings and "surprise"-are unjustified and hence rejected.

Fair Trial

353. Any party to litigation is entitled to a fair trial by an independent and impartial tribunal. A judge must of course identify the issue(s), consider the evidence and submissions from both sides, fairly and independently adjudicate upon the issue(s), and give reasons for his or her decision. The above proposition hardly needs any supporting precedents.

354. The requirement of judges to give reasons for their decisions under common law is explained in English v Emery Reimbold & Strick Ltd [2002] 1 WLR 2409 from p. 2417 onwards:

"There is a general recognition in the common law jurisdictions that it is desirable for judges to give reasons for their decisions...

Reasons are required if decisions are to be acceptable to the parties and to members of the public...

Henry LJ in Flannery's case [2000] 1 WLR 377 observed that the requirement to give reasons concentrates the mind of the judge and it has even been contended that the requirement to give reasons serve a vital function in constraining the judiciary's exercise of power....

We would put the matter at its simplest by saying that justice will not be done if it is not apparent to the parties why one has won and the other has lost...

A judge should give his reasons in sufficient detail to show the Court of Appeal the principles on which he has acted and the reasons that have led him to his decision. The need not be elaborate. I cannot stress too strongly that there is no duty on a judge, in giving his reasons, to deal with every argument presented by counsel in support of his case. It is sufficient if what he says shows the parties and, if need be, the Court of Appeal the basis on which he has acted...(see Sachs LJ in Knight v Clifton [1971] Ch 700, 721.)..

It follows that, if the appellate process is to work satisfactorily, the judgment must enable the appellate court to understand why the judge reached his decision...the issue the resolution of which were vital to the judge's conclusion should be identified and the manner in which he resolved them explained...

Whatever the explanation may be, it should be apparent from the judgment...

The essential requirement is that the terms of the judgment should enable the parties and any appellate tribunal readily to analyse the reasoning that was essentials to the judge's decision."

355. At trial, there were four main contentions or issues between the Father and the Wife, namely (1) the "circumstantial evidence issue," (2) the "knowledge and approval issue," (3) the "handwriting issue," and (4) the "ink-dating issue". The "circumstantial evidence issue" involved a significant number of factual disputes.

356. Clearly, the judge had correctly identified the issues and had dealt with each one of them. He found for the Father in respect of two of the issues and against him in respect of the remaining two.

357. The judge was well aware of the parties' cases. The record of proceedings indicates that the judge was, throughout the trial, on top of the issues in dispute. The judge resolved those issues and also gave detailed reasons for his decisions.

358. What and how much a judge should do must depend on the issues involved in the disputes, the extent of the disputes, the parties' respective cases, and how such cases are presented.

359. The point was made clear in English's case (supra) by Lord Phillips of Worth Matravers MR on p. 2415-2417:

"The question is always, what is required of the judge to do so; and that will differ from case to case. Transparency should be the watchword" (page 2415 C-D)

"It requires that a judgment contains reasons that are sufficient to demonstrate that the essential issues that have been raised by the parties have been addressed by the domestic court and how those issues have been resolved." (Page 2416 F)

"We would put the matter at its simplest by saying that justice will not be done if it is not apparent to the parties why one has won and the other has lost.

As to the adequacy of reasons, as has been said many times, this depends on the nature of the case: se for example Flannery's case [2000] 1 WLR 377 382. In Eagil Trust Co Ltd v Pigott-Brown [1985] 3 All E R 119, 122 Griffiths LJ stated that there was no duty on a judge, in giving his reasons, to deal with every argument presented by counsel in support of his case." (Page 2417 G-H)

In Hiro Balani v Spain (1994) 19 EHRR 566, it is stated on p. 574 of the judgment:

"The Court reiterates that Article 6(1) [similar to Article 10 of the Hong Kong Bill of Right Ordinance] obliges the courts to give reasons for their judgments, but cannot be understood as requiring a detailed answer to every argument. The extent to which this duty to give reasons applies may vary according to the nature of the decision. It is moreover necessary to take into account, inter alias, the diversity of the submissions that a litigant may bring before the courts and the differences existing in the Contracting States with regard to the statutory provisions, customary rules, legal opinion and the presentation and drafting of judgments. That is why the question whether a court has failed to fulfill the obligation to state reasons, deriving from Article 6 of the Convention, can only be determined in the light of the circumstances of the case."

360. It must be remembered that the evidence of the Father, Yih, and Teresa took about 50 days, mainly because of the lengthy cross-examinations, which were very often directed at peripheral matters based on untrue and unsubstantiated allegations. In the end, Mr. Lee chose not to make any submissions on those matters.

361. It must also be remembered that the circumstantial evidence adduced by the Wife was relatively short and with little dispute (except the "holiday wills"). Mr. Lee suggested that the case was about "forgery and nothing else; experts' evidence and nothing else".

362. Mr. Lee clearly relied more heavily on the experts' evidence. It is not at all surprising that, on the circumstantial evidence, the judge dealt more extensively with the Father's case than with the Wife's.

363. In any event, the judge sufficiently dealt with the evidence adduced by the Wife to support her case on the due execution of "the 1990 documents," such as "the holiday wills", the riding accident, the abortive trust arrangements, the Wife's contribution to Chinachem, and her "loving relationship" with Wang.

364. The judge made as many rulings against the Father as he did against the Wife. He set out matters favourable to and supportive of the Wife's case. There is no basis for the suggestion that the judge had "ignored" the Wife's case without any or any adequate explanation.

365. Mr. Thomas complained about "plagiarism" of the part of the judge.

366. The conspicuous and extensive reproduction of counsel's submissions in the judgment is unconventional, surprising, and unnecessary. If the judge accepted a party's case entirely, he could have paraphrased its submissions and stated his agreement, and vice versa. He would not then be the subject of any criticism.

367. On principle, however, a judge is entitled to accept or reject a party's case in its entirety. A judge's total adoption or rejection of counsel's submissions per se does not imply the lack of independent adjudication, nor does it constitute a valid ground for upsetting the judgment on the basis of an unfair trial.

368. The record of the proceedings clearly indicates that the judge was fully aware of and keenly interested in the issues and the parties' arguments (as Mr. Thomas agreed).

369. The judge dealt with the respective cases of the parties thoroughly. He accepted most of the arguments advanced on the Father's behalf, but he also rejected some of them.

370. The judge accepted the Wife's case on the ink dating. He also dealt sufficiently with the circumstantial evidence put forward by the Wife.

371. That the judge did not deal explicitly with every piece of circumstantial evidence supportive of the Wife's case does not mean that the judge was neglectful of such evidence. It was neither realistic nor reasonable to expect the judge to deal with all the points made Mr. Lee when the judge considered many of them "a complete waste of time".

372. Moreover, the judge expressly said in his judgment (Para. 32.2), "However if any point which has not been mentioned herein but at the same time not in line with any of the findings or decisions I made herein, it should be taken that the point has impliedly been rejected by the Court. In fact vice versa is also true."

373. The judge extensively dealt with the handwriting evidence. He clearly considered and analysed the experts' evidence fully and carefully before he reached his conclusions on the matter.

374. The suggestion that the judge had abdicated his responsibility by not properly evaluating the parties' cases and evidences (indeed a most serious allegation) is not justified.

375. The grounds of appeal and written submissions filed on the Wife's behalf objected to the reference by the judge to her previous affirmations. Mr. Thomas abandoned this point at the commencement of the appeal. In fact, he resolutely sought to rely on such affirmations as evidence of due execution. (There was no other evidence on due execution except the documents themselves)

376. That the judge dealt with a related proceeding in which he made adverse findings against the Wife is not a valid ground of complaint.

377. Allegations of bias (structural or otherwise) are not established. There is no risk of a fair-minded and informed observer to conclude from the available materials that there was a real possibility or a real danger that the judge was biased.

378. The allegation that the Wife had not received a fair trial is not established; this ground of appeal also fails.

The Compartmentalization Argument

379. The judge was criticized for considering the handwriting evidence separately and independently from the circumstantial evidence. Mr. Tang responded by suggesting that the approach was adopted only in response to Mr. Lee, who wrote in his submission that the issue was "forgery, and nothing else; experts' evidence, and nothing else".

380. Mr. Tang suggested that, if the judge were to adopt a different approach, his approach to the experts' evidence would no doubt be criticized as being coloured by his view of the circumstantial evidence.

381. A judge should not adopt an incorrect approach based on counsel submission or for fear of criticism. No criticism, however forcefully made, is valid if a judge adopts the correct approach.

382. It was clearly out of excessive cautiousness that the judge took the unusual step of making the finding of forgery based separately on the handwriting evidence and on the circumstantial evidence.

383. The judge did reach an overall conclusion in the end, both on the circumstantial evidence and on the handwriting evidence, that "the 1990 documents" were forged.

384. This approach was not absolutely necessary but is not objectionable. It is certainly not unfair or prejudicial to the Wife's case.

385. On the handwriting evidence alone, the judge had "no doubt at all" that the questioned signatures were forged. On the circumstantial evidence alone, the judge-rightly or wrongly-also came to the "irresistible conclusion" that "the 1990 documents" were forged.

386. On the judge's findings, the overall conclusion of forgery based on both circumstantial and handwriting evidence together was inescapable.

The Presumption "Omnia Praesumuntur Rite Esse Acta"

387. Mr. Thomas relied heavily on the leading textbook Executors, Administrators and Probate, Williams, Mortimer, and Sunnucks, 18th Ed. (2000) paras. 12-25 and 39-11:

"The presumption that everything was properly done (Ominia rite et solemniter esse acta), arises whenever a will, regular on the face of it and apparently duly executed, is before the court, and amounts to an inference, in the absence of evidence to the contrary, that the requirements of the statute have been duly complied with."

"The burden of proof in probate cases lies, in general, upon those who propound a will. But the burden of proving fraud, undue influence and forgery lies on those who allege it."

388. Mr. Thomas suggested that once a will valid ex facie was produced, the onus on the propounder had been discharged and strong proof was required to establish that there had not in fact been due execution. He referred to Harris v Knight (1890) LR 15 PD 170, Wright v Rogers (1869) LR 1 P & D 678 and Wright v Sanderson (1884) 9 PD 149.

389. Mr. Thomas also suggested that the court should adopt a benevolent approach in favour of the propounder on the question of due execution. He said that a low level of proof was adequate lest difficulties frequently resulting from death, delay, disappearance of witness would defeat an apparently valid testamentary intention for lack of proof, particularly so when s.5 (2) of the Will Ordinance Cap. 30 relax the strict requirements on the form of execution of a Chinese will.

390. Mr. Thomas sought to rely on leading English textbooks on probate and leading authorities, including Harris v Knight, (1890) 15 P.D. 170, and Burgoyne v Showler, 1 Bobertson Ecclesiastical 5, 163 ER 945.

391. It must be borne in mind, however, that leading textbooks and authorities on the subject refer to the equivalent of s.5 (1) of the Wills Ordinance, which stipulates that:

"no will shall be valid unless-

(a) it is in writing, and signed by the testator, or by some other person in his presence and by his direction;

(b) it appears that the testator intended by his signature to give effect to the will;

(c) the signature is made or acknowledged by the testator in the presence of 2 or more witnesses present at the same time; and

(d) each witness either-

(i) attests and signs the will; or

(ii) acknowledges his signature, in the presence of the testator (but not necessarily in the presence of any other witness),..."

392. Obviously, if the aforesaid provisions were complied with, it would almost be impossible to make any successful allegation that the signature of the testator was forged. Indeed, such allegation is rare.

393. Section 5 (1) of the Wills Ordinance provides clear statutory safeguards against forgery of the deceased's signatures.

394. Hence contested probate actions rarely concerned the authenticity of the testator's signature. In most cases, evidence of genuine signatures of the deceased and his witnesses was available and undisputed. The cases mostly concerned the issue of knowledge and approval or the issue of whether a will was made by a free and capable testator.

395. The presumption omnia praesumuntur rite esse acta apply in cases of "English wills" because of the inherent statutory safeguards, including the need for attestation clauses.

396. Some of the cases relied on by Mr. Thomas clearly emphasized the importance of attestation clauses. Lord Penzance in Wright v Rogers (1869) LR 1 P&D 678 said at p. 682:

"The court ought to have in all cases the strongest evidence before it believes that a will, with a perfect attestation clause, and signed by the testator, was not duly executed, otherwise the greatest uncertainty would prevail in the proving of wills. The presumption of law is largely in favour of the due execution of a will, and in that light a perfect attestation clause is a most important element of proof." (Emphases added)

Halsbury's Laws of England (4th Edition) also provides at para. 304:

"The principle omnia praesumuntur rite esse acta applies where the will is regular on the fact of it, with an attestation clause and the signatures of the testator and witnesses in their proper places. This presumption of due execution applies where there is a proper attestation clause, even though the witnesses have no recollection of having witnessed the will, and even though the attestation clause appears only on the completed draft of a last will. .... It may be rebutted by evidence of the attesting witnesses or otherwise, but the evidence as to some defect in execution must be clear, positive and reliable, since the court ought to have the strongest evidence before it believes that a will, with a perfect attestation clause and signed by the testator, was not duly executed. Where there is only an incomplete attesting clause it seems that the presumption applies, but with less force than where the attestation clause is in proper form. Where both the attesting witnesses are dead and the will is in regular form the principle is applicable on proof of the handwriting." (Emphases added)

397. In my view, the presumption of regularity must not be applied blindly, but with circumspection. Lindley L.J. in Harris and Knight (1890) 15 PD 172 said at page 179;

"The maxim, 'Omnia praesumuntur rite esse acta,' is an expression, in a short form, of a reasonable probability, and of the propriety in point of law of acting on such probability. The maxim expresses an inference which may reasonably be drawn when an intention to do some formal act is established; when the evidence is consistent with that intention having been carried into effect in a proper way; but when the actual observance of all due formalities can only be inferred as a matter of probability."

Lopes, L.J. made a similar observation at page 183-184:

"The presumption 'Omnia praesumuntur rite esse acta' is an inference of fact- an inference of reasonable probability. The testator clearly knew he must sign; he knew that there must be two witnesses; is it not more probable than not, that he also knew that those witnesses must be present together and is it not more probable than not, that the signature which purports to be Heningham's signature, is his signature?

The inference to be drawn in cases of this kind depends upon a number of circumstances peculiar to the cases in which they arise, and the presumption 'Omnia rite esse acta' applies with more or less force according to the circumstances of each case. In every case of this kind the Court should be influenced by a desire that the intention of the testator should not be frustrated, when the execution of the testator is sufficiently proved, and the will, on the face, complies with the requirements of the statute."

398. Without the statutory safeguards in the case of a Chinese will, the extent to which the presumption omnia praesumuntur rite esse acta (heavily relied on by Mr. Thomas) applies, is questionable. There has not been any authority, which suggests that the presumption applies in the absence of any attestation clause and/or when the signature of the testator is in dispute.

399. There is, in my view, no logical support for the operation of this presumption on the due execution of a Chinese will when the signature of the deceased is alleged to be forged. An inference of reasonable probability of due execution cannot be drawn at all, not at least until the signature of the deceased is proved to be genuine.

400. "The reasons stated by Dr. Lushington for the decision of the Judiciary Committee in Lloyd v Roberts 12 Moo P.C. 165 proceed upon the same principles, and upon the presumption in favour of the due execution of a holograph will on the face of which everything is regular and where there is no question of fraud." [Per Earl of Selborne L.C. in Wright v Sanderson (1884) 9 PD 149]

401. In his reply, Mr. Thomas brought to our attention the case of Chetty v Chetty (1916) PC 113 in support of his contention that the court should not consider the question of why Wang would want to leave his entire estate to the Wife and other matters of "suspicious" nature. In particular, he relied on the following passage at page 1916:

"These are the reasons that convince their Lordships that, whatever may be the true cause of the inconvenient way in which some of this writing appears, the explanation is not and cannot be the one that the learned judge has assigned; it would indeed be most unsafe and most undesirable in circumstances such as these to try to spell out from the peculiar form in which a document written in the vernacular appears, a hypothetical answer to the clear, distinct, and trustworthy evidence of the doctor who witnessed the will.

Their Lordships only desire to add, in conclusion, this: When a will has once been made and is apparently in perfect form, and the evidence of the attesting witness is to be trusted, few things can be more dangerous than to attempt to recreate the kind of will that the man ought, in the opinion of the Court, to have made. Once the man's mind is free and clear and is capable of disposing of his property, the way in which it is to be disposed of rests with him, and it is not for any Court to try and discover whether a will could not have been made more consonant either with reason or with justice."

402. I fail to see how the reasoning in the quoted passage supports the points advanced by Mr. Thomas. Such reasoning premises again on the undisputed or undisputable facts that the deceased had indeed executed the will and that a trustworthy witness witnessed such execution.

403. I need only to repeat the first few lines of the head-note of the judgment from which Mr. Thomas cited the passage: "When a will is apparently in perfect form and the evidence of the attesting witness is trustworthy, the Court ought not to attempt to recreate..."

404. Insofar as Mr. Thomas sought reliance on Chetty (supra), the very premise of his reasoning-that Wang did execute "the 1990 documents" and that a trustworthy witness witnessed such execution-are the subject matters of the present dispute.

405. Mr. Thomas would have been correct in suggesting that the court should adopt a benevolent approach and a low standard of proof to give effect to a will if its due execution were not disputed or not disputable (e.g. supported by formal attestation clause). I do not subscribe to this view on the issue of execution of a Chinese will under section 5(2) of the Wills Ordinance, particularly when there was an allegation of fraud.

406. Even in non-contentious probate matter, due care must be exercised before granting probate (See rule 10 Non-Contentious Probate Rules, Cap. 10).

407. Contrary to the suggestion of Mr. Thomas in his reply, I am of the view that the court must be vigilant and cautious on the issue of execution of a Chinese will (particularly when there is suspicious circumstance coupled with the allegation of fraud). I reject the suggestion that the presumption Omnia praesumuntur rite esse acta applied to the instant case at all.

408. The court should take into consideration all the relevant circumstantial evidence in deciding if Wang had duly executed "the 1990 documents" and not to find due execution merely on their production as suggested by Mr. Thomas.

"Suspicious Circumstances" as Evidence of Proof of Forgery

409. I think that Mr. Thomas was right in his attack on the judge's conclusion that the "suspicious circumstances" alone established forgery of "the 1990 documents".

410. Firstly, the Father did not expressly plead the case that the "suspicious circumstances" per se supported the allegation of forgery. In fact, the pleadings did not mention any of the "suspicious circumstances".

411. The "suspicious circumstances" were presented at trial to support the contention that it was improbable for Wang to have executed "the 1990 documents" in order to defeat the Wife's attempt to prove their due execution in an inquisitorial hearing.

412. If the Father intended to rely on the "suspicious circumstances" to support a case of fraud, those "suspicious circumstances" must be pleaded.

413. Bokhary JA (as he then was) said in no uncertain terms in Aktieselskabet (supra) on p. 270:

"To the generality of the foregoing must be added the special rule that allegations of fraud must be pleaded distinctly and with the utmost particularity."

414. As the propounder of "the 1990 documents," the Wife had to prove, on a balance of probability, that Wang had duly executed those documents as his last will.

415. In order to successfully discharge such onus, the Wife might have to dispel the "suspicious circumstances" if such "suspicious circumstances" were inconsistent with the due execution of "the 1990 documents". The Wife might have to prove that Wang's questioned signatures in "the 1990 documents" were genuine as part of her duty to establish due execution satisfactorily.

416. However, the Wife had no burden to disprove the allegation of forgery based on the "suspicious circumstances," even if they were expressly pleaded (which they were not). In fact, it was the Father who had the burden to establish forgery with strong and cogent evidence.

417. The distinction between the aforementioned two points (i.e. the burden to dispel the "suspicious circumstances" to prove due execution and the burden to disprove the allegation of forgery based on such "suspicious circumstances") may be a fine one but should not be overlooked.

418. Mr. Tang was correct in suggesting that fraud could be established by drawing inferences from circumstantial evidence.

419. By defining the issue as "forgery and nothing else; experts' evidence and nothing else," Mr. Lee might have misled the judge.

420. However, a case of fraud based on circumstantial evidence must be expressly raised and properly dealt with. In my view, neither had been done in the present case. The judge was not entitled to find forgery on the "suspicious circumstances" alone.

421. In any event, given the higher standard of proof required, albeit a civil standard, it was uncertain whether the "suspicious circumstances" were "strong and cogent" enough to support a positive finding of forgery.

422. In fact, there was no need for the judge to make a finding of forgery or a finding of the Wife's participation in the forgery based on the circumstantial evidence. The Father did not make such allegation in his pleadings. Despite the allegation of forgery based on experts' evidence, which the Father did make, the primary issue was still whether the Wife successfully established due execution.

The Relevance of the "Suspicious Circumstances"

423. Although I rule that the judge was not entitled to find forgery based solely on the "suspicious circumstances," this does not of course mean that the "suspicious circumstances" identified by the judge were irrelevant.

424. The judge found "the 1990 documents" to be forged based on the "suspicious circumstances". A fortiori it must be his conclusion that the "suspicious circumstances" were inconsistent with the Wife's case of due execution of "the 1990 documents" by Wang.

425. Such "suspicious circumstances," and indeed other circumstantial evidence (both supporting and opposing), were relevant to the issue of whether Wang had executed "the 1990 documents".

426. In Tyrrell v Painton [1894] P 151, Davey L.J. made this point clear on p. 159:

There rests upon that will a suspicion which must be removed before you come to the plea of fraud. It must not be supposed that the principle in Barry v. Butlin 2 Moo P.C. 480 is confined to cases where the person who prepares the will is the person who takes the benefit under it- that is one state of things which raises a suspicion; but the principle is, that wherever a will is prepared under circumstances which raise a well-grounded suspicion that it does not express the mind of the testator, the Court ought not to pronounce in favour of it unless that suspicion is removed. Here the circumstances were most suspicious, and the question a judge has to ask himself is whether the defendants have discharged themselves of the onus of shewing the righteousness of the transaction, and without going again over the circumstances which have been referred to, I am compelled to say that they have not."

427. In the later part of his submission, Mr. Thomas appeared to have conceded that his initial objection to the Father's reliance on the circumstantial evidence to support his case was not justified.

428. He then sought to argue that such circumstantial evidence, before it could be relied on, must be established and proved to carry probative value.

429. He suggested that the circumstantial evidence advanced by the Father was not relevant to the issue and did not, in any event, support the Father's case.

430. Mr. Thomas also submitted that events subsequent to the questioned execution of "the 1990 documents" could not be relevant to the issue of due execution because such events bore no causal relationship to the questioned execution. He sought reliance on Davis v Mayhew [1927] P. 264.

431. Properly understood, Davis (supra) does not support Mr. Thomas's submission. One needs only to refer to the judgment of Lawrence L.J. on p. 287:

"In the present case the matters which have excited the suspicion of the learned judge are matters which occurred only after the execution of the will. Subsequent events may in some cases no doubt give rise to suspicion that the will was not properly executed or that the testator did not know or approve the contents, but in the present case the facts relied upon by the learned judge are not, in my opinion, of such a nature as to form a good ground for any such suspicion."

In Harris v Knight (1890) 15 PD 170, Lindley LJ said on p. 181:

"I can find no case that goes so far as this; but authorities are only of use as guide when a case turns on inferences of fact. The extent to which the supposed will was acted upon after the death of the deceased tends to support the inference in favour of its validity; whilst its destruction by the widow (if indeed she destroyed it) tends the other way."

432. As Mr. Tang rightly pointed out, circumstances such as how a purported will was discovered could certainly be relevant to the issue of the due execution. (In all fairness, Mr. Thomas did agree on the relevance of the circumstances in which the Wife came by "the 1990 documents".)

433. Having found the circumstantial evidence to be relevant, I now proceed to deal with it.

Circumstantial Evidence

434. The circumstantial evidence (both supporting and opposing the due execution of "the 1990 documents") is clearly relevant. The judge made his findings of fact on part of the circumstantial evidence albeit only in connection with the issue of forgery.

435. Counsel subjected the judge's reasons to microscopic examination and expressed their views on the circumstantial evidence to attack or support the due execution of "the 1990 documents". Their analyses went on for pages after pages in their respective written submissions.

436. Mr. Thomas pointed out the supporting circumstances. He suggested that if Tse were willing to lie in his statements given to Mr. Lam and Mr. Leung to assist the Wife, there would be no need to forge his signatures on "the 1990 documents".

437. Mr. Thomas further submitted that a forger would have done a much better job and behaved more rationally than demonstrated in "the 1990 documents". He cited the poor quality of Tse's questioned signatures, the number of signatures involved, the sloppiness and varieties of handwriting, the use of different writing instruments and types of paper, and the unusual sentimental expressions, particularly that expressed in "D".

438. Citing the Wife's reluctance to produce the sealed envelope containing "the 1990 documents" and her efforts to prevent the disclosure of their contents, Mr. Tang contended that the Wife did not intend to involve Tse unless it was absolutely necessary. That explained why, hypothetically speaking, even if Tse were willing to lie for the Wife's benefit, she might still have been motivated to forge his signatures.

439. Mr. Tang emphasized that "The 1990 documents" were only lodged nine months after the opening of the safety deposit box.

440. Mr. Tang further retorted by pointing out that one must not assume a forger to be rational. He commented, with obvious reference to the Wife, "A person who continues to contend or pretend that Teddy (Wang) was alive might not be rational either."

441. More fundamentally, it is difficult for a bystander with limited information to determine via post facto examination of a person's behaviour if a certain decision of that person at an earlier time had been rational or not. I am not convinced by Mr. Thomas's suggestion that the hypothetical forger would necessarily have behaved "more rationally".

442. Also, I would like to add that the Wife did not allow herself to be examined in court to determine or estimate within reason whether she was capable of the (irrational) conducts alleged against her.

443. I accept that if Tse were prepared to be one of the conspirators to the fraud, it would have been simpler to have his genuine signatures in "the 1990 documents".

444. On the other hand, I must make one point very clearly. That a collaborator, at the point of collaboration, is a willing one does not eliminate the risk of involving him or her in a fraud. Among other things, the collaborator's subsequent betrayal, acts of blackmail, or even mere negligence can expose the fraud and/or endanger the fraudster(s). Hence obviously there is risk for the fraudster(s) in involving even a willing collaborator.

445. In the present case, at the point of deciding whether to forge or not forge Tse's signature, the hypothetical forger was faced with two opposing considerations, namely the (aforesaid) risk of involving one additional collaborator and the risk entailed in forging additional signatures. It would be ideal to avoid both risks, but when this was impossible (as it obviously was), then the hypothetical forger's apparent preference for the latter risk over the former only means that he or she considered the latter risk less significant. It does not mean that the hypothetical forger acted irrationally. The forgery of Tse's signatures could well be in the forger's interests even if Tse was willing to provide genuine signatures to the forged documents.

446. Mr. Thomas's reasoning was premised on his incorrect assumption that the former risk did not exist. Only with this incorrect assumption would it appear that the hypothetical forger had taken the latter risk superfluously and hence irrationally.

447. Mr. Thomas argued that the hypothetical forger would have done "a better job" and therefore the apparently sloppily prepared "1990 documents" could not have been forged. This argument is unacceptable because it was both tautological and two-sided, mainly because of the inherent vagueness of what "a better job" means. Assuming indeed that "a better job" was done and a more genuine-seeming (but nonetheless imperfect) will were forged, Mr. Thomas could still make the same argument. The only scenario in which his argument would not work would be when the documents were "perfectly" forged (whatever that might mean), in which case the issue of forgery would not have arisen all together. If the forgery was well executed, Mr. Thomas could then argue that a forger could not have done such a good job to defeat any suggestion of forgery.

448. One could argue that Wang had no more reason than the forger to be sloppy. Hence the sloppiness of "the 1990 documents" could equally be an indication of forgery. Also, there could be no logical explanation for Wang to create so many "curiosities" (to adopt Mr. Thomas's term) in expressing his testamentary intent.

449. The Wife carried the burden of establishing the due execution of "the 1990 documents". The suspicious and/or sloppy nature of the documents simply cannot be a factor supporting her case.

450. Mr. Thomas heavily relied on the suggestion that the Wife (not the Father) was Wang's natural and logical choice to inherit his estate. He seized upon Mr. Tang's comment made in the course of his submission that the Wife was a worthwhile beneficiary of Wang's estate.

451. Of course, a spouse is normally a worthwhile beneficiary of a deceased's estate. However, the court must not be unduly influenced by this consideration in determining the true intention of the deceased. It is, but one of the factors to take into consideration.

452. It must be borne in mind that Wang had no children and that the Wife was not much younger (in 1990, Wang was 56, the Wife 54 years old). Wang had also made adequate financial provisions for the Wife, who was independently very well off.

453. There was a marital dispute in the mid-1960s resulting in the execution of "the 1968 Will". Despite the improvement in their relationship as observed by some witnesses, Wang was still concerned about her "faithfulness" in 1974.

454. "The 1968 Will" remained "untouched" throughout the 1970s and 1980s. It was still kept in the safety deposit box when Wang was kidnapped again in April 1990.

455. If the Father were to inherit Wang's estate, the business that the Father started and Wang developed would eventually go into the hands of Wang's siblings or their descendents. If the Wife were to inherit his estate, the business would go into the hands of the Wife's maternal family or even total strangers.

456. Viewed from a traditional Chinese perspective (a conservative one according to Mr. Thomas), the choice of the Wife to inherent the entire estate over the Father might not be as obvious as Mr. Thomas suggested. I would also add that, on matters of inheritance, minimizing estate duty might not be a prime consideration.

457. The evidence suggested that Wang had been very generous to the Father and was always very concerned about his wellbeing. Under "the 1960 Will," the Father was to inherit half of Wang's estate, the Wife and their children (if any) the other half. Under "the 1968 Will," the Father was to inherit his entire estate.

458. In view of Wang's generous provisions to his parents and his siblings, it was improbable that he would have completely excluded them from benefiting, directly or indirectly, from his estate.

459. Despite the riding accident, Wang was still a relatively young man in 1990 and death was but a remote possibility. The suggestion that Wang had to execute "the 1990 documents" as an urgent and temporary measure is in my view unconvincing.

460. The background of the case and the continued existence of "the 1968 Will" are more consistent with the suggestion that in 1990 Wang still wanted the Father to inherit his estate despite the improved relationship with the Wife. They are also more consistent with the suggestion that Wang did not then find it necessary, for whatever reason, to revoke "the 1968 Will".

461. The purported trust arrangements in 1987 and in 1989-1990 were heavily relied upon to support the Wife's case that Wang had decided to leave his estate to her.

462. It is common knowledge that the political instability in the mid-1980s had prompted many Hong Kong residents (particularly those with substantial assets) to emigrate and/or transfer their assets overseas. That Wang had considered making trust arrangements does not indicate an intention to leave the Wife his entire estate.

463. Moreover, the documents produced in support of the trust arrangements contained only proposals and did not constitute Wang's preferences, let alone commitments.

464. The proposals were never materialized. If Wang had intended the Wife to inherit his estate, he would unlikely be deterred by the costs involved.

465. The Wife's counsel suggested that "the 1990 documents" were a provisional measure taken before the trust was properly set up.

466. Such suggestion was inconsistent with the Wife's assertion that she was told not to open the envelope containing "the 1990 documents" until after Wang's death. If the documents were temporary, they would have to be replaced and/or made ineffective within a short period of time.

467. The suggestion also contradicted the contents of one of the trust documents dated 26th March 1990, which indicated that there was no immediate intention to proceed with the proposed trust arrangements.

468. The Wife chose not to reveal in court what the couple intended when they consulted advisers to discuss the proposed trust arrangements.

469. I agree with Mr. Tang's observation that if the 1987 and the 1989-1990 trust documents constituted the "most important part" of the Wife's case, they reflected badly on the strength of her case.

470. Mr. Thomas argued that the findings that Wang was unromantic and too meticulous to accept any sloppily prepared document were based on obsolete "ancient history".

471. Wang's character was assessed by his family members and close associates, who knew him well.

472. It is improbable that Wang's character had changed so drastically-from unromantic to very romantic, from meticulous to very sloppy-despite the lapse of twenty or even more years.

473. There was clear evidence on which the judge could reject "the holiday wills" as complete fabrications.

474. Dr. Kung in his evidence suggested that "the holiday wills" were made in 1984 and 1985 and that he had mentioned them to the Wife in 1997. On the other hand, the Wife never mentioned such wills in any of her affirmations of testamentary scripts.

475. In both such affirmations, dated 16 January 1998 and 18 November 1999, the Wife did not mention, "the holiday wills". She said instead:

"No testamentary scripts of Mr. Wang, whose estate is the subject of this action, being a will or codicil or draft thereof, written instructions for a will or codicil made by or at the request of or under the instructions of Mr. Wang, or any document purporting to be evidence of the contents or to be a copy, of a will or codicil of Mr. Wang which is alleged to have been lost or destroyed, has at any time come into my possession or knowledge..."

476. The Wife was clearly aware of her duty to give information relating to all testamentary scripts that she knew of. She said so in her affirmation dated 24 April 1998 in which she explained why she took so long to produce "the 1990 documents":

"The last reason that the Plaintiff gives for speculating that the Second Envelope is a forgery is that if my husband had really made a Will in 1990, I would not have instructed my solicitors that I needed time to search for testamentary scripts. The Plaintiff has clearly overlooked the fact that parties to contentious probate action are obliged to give information relating to all testamentary scripts that they know of. The fact I required time to carry out the search is not evidence that I was unaware of those documents that I have lodged at court."

477. The Wife made a late and unsuccessful application to amend the defence and counterclaim to include reference to "the holiday wills". She suggested "the holiday wills" had revoked "the 1968 Will".

478. The affirmations in support of the application suggested that "the holiday wills" were executed several times from 1986 to 1990.

479. If "the holiday wills" truly existed, there was no conceivable reason for the Wife to fail to mention them in her affirmations.

480. In any event, I am not convinced by the suggestion that the couple made a will every year before their annual vacation. If the wills were made out of worry about mishaps during the vacations, one such will would have been sufficient.

481. The finding by the judge that "the holiday wills" was a fabrication casts a very unfavourable light on the Wife's case.

482. Although the judge might have criticized Mr. Lam harshly, the circumstances in which Mr. Lam took the statement from Tse on 6 September 1999 were indeed highly suspicious. Even more so was Tse's sudden and unexplained departure from Hong Kong.

483. The important issue is not so much the credibility of Mr. Lam and Mr. Leung, but that of Tse and the veracity of the statements that he made on 6 and 9 September 1999.

484. There were valid reasons for rejecting Tse's statements.

485. Circumstantial evidence is by nature indirect and can only lead to "inferred conclusions". Either party can put a different gloss on the same piece of circumstantial evidence to render it favourable to his or her case.

486. In this very long trial, the judge had the advantage of seeing and hearing the witnesses directly. His conclusion thus reached is entitled to the utmost deference by the appeal court.

487. In my view, the judge had valid bases to conclude that the circumstantial evidence identified by him did not support the due execution of "the 1990 documents".

Evidence of "Due Execution" relied on by the Wife

488. To argue that the proof of the due execution of "the 1990 documents" was strong, Mr. Thomas referred to the following supporting evidence:

(i) "The 1990 documents" with the expression of testamentary intention were apparently signed by Wang and handed to his Wife. They were witnessed (though unnecessarily) by two identified persons who would be known to Wang and available so to act;

(ii) The Wife, being childless, and a longstanding business partner was the obvious beneficiary;

(iii) The only witness available to give evidence (Tse) had given two supporting statements before his own death, proving execution;

(iv) Wang had given the will to the Wife for safe custody, saying it was his will;

(v) Wang had previously made "home made" will(s);

(vi) Wang had lately suffered an accident which would have been likely to trigger "post-mortem" thoughts;

(vii) "Post-mortem" arrangements, only for the benefit of the Wife, were currently in contemplation; and

(viii) There was the expert evidence showing that the questioned signatures were that of Wang.

489. (i) and (iv) were only in the form of affirmations (by the Wife and by Tse) and were therefore untested by cross-examination. Indeed, as the judge observed, the contents of such affirmations left a lot of unanswered questions. Little weight could be attached to such affirmations.

490. I have already discussed why (ii), (iii), (v), (vi), and (vii) are not sufficient support for the Wife's case. In particular, the judge correctly rejected the "holiday wills" evidence.

491. The expert evidence (viii) is a separate issue all together. I will discuss this issue below.

492. Since the Wife did not give evidence, Mr. Thomas obviously had to rely on her affirmations pertaining to the circumstances in which "the 1990 documents" came into her possession.

493. However, Mr. Thomas made a rather astonishing submission that the Wife's affirmations should be given full weight because the Father did not apply to cross-examine her on her affirmations. (Emphasis added)

494. The Wife did not indicate that she would not give evidence until late in the trial. As early as 27 September 1999, in a letter to the police (the parties agreed that the court could refer to it although it did not form part of the evidence before the judge), Messrs. Lovells, solicitors then acting for the Wife, set out her position, inter alias, as follows:

"We would ask you to conclude that the Police should keep a close watch upon the conduct of the probate action, during the course of which all will be revealed about the authenticity and validity of the contending wills, and that the Police should not involve themselves at this stage in conducting an investigation which will have the effect of intervening...

In the meantime our client, Mrs. Wang, has been advised by us and by leading counsel in plain and unequivocal terms not to say anything at all to anyone outside the context of this Probate Action until it has been determined by the High Court..." (Emphases added)

495. The Wife had filed her witness's statement, an indication that she would be giving evidence at the trial. More importantly, the Wife never applied to use her affirmations as evidence in the trial. The affirmations were available to court only as affirmations of testamentary scripts, not as evidence for the purpose of the trial.

496. Under Order 38 (2) of the Rules of the High Court, the court may allow the affidavit of a witness to be read as evidence at trial if it is reasonable to do so. The court may also order the production of the deponent for cross-examination.

497. In the present case, it was inconceivable that the court would admit the Wife's affirmations as evidence without requiring her to appear for cross-examination. The father would certainly have objected to such course.

498. Mr. Tang fairly conceded that the court could still refer to the Wife's affirmations despite the non-compliance with the rules, as the court's approach was inquisitorial in nature. He cited a short passage from the judgment of Karminski J in Trotman v Trotman (1964) 108 Dol. Jo. 159:

"The probate court had always done its best to elicit the facts of the cases it had to try without too rigid an adherence to the rules of evidence practised in other courts." (p. 160)

499. A fundamental feature of our system of the administration of justice is the right, as well as the need, to cross-examine witnesses on facts relevant to the issues in dispute. It is "beyond any doubt the greatest legal engine ever invented for the discovery of truth". (See Wigmore on Evidence 1974 Ed., Vol. V, Para 1367)

500. By choosing to avoid this "legal engine," the Wife had deprived the court of one of the major means to test the veracity of her assertions. While the court was entitled to refer to her affirmations, I am of the view that very little weight should or could be attached to them.

501. The judge had good reasons to reject the purported attestation evidence of Tse, i.e. (i) and (iii).

502. It is quite inconceivable that Tse could have, by his own volition and without prompting or coaching of any kind, come up with the statements in question, the contents of which were a perfect match with the Wife's case.

503. The four questions allegedly put to him by Mr. Lam could hardly have resulted in Tse's producing the contents of the statement on 6 September 1999 on his own initiative. There are strong internal marks of untruth, unfairness, and partiality in Tse's statements.

504. Mr. Thomas's comment on another witness, "[the witness] had overcooked it by providing too much details when there was no obvious need for it," applies equally aptly to Tse.

505. Further, Tse did not volunteer the information to the police when they interviewed him in May and July 1990, very shortly after Wang's kidnap and very shortly after Tse allegedly appended his signatures to "the 1990 documents" on Wang's instruction without knowing their contents.

506. That a matter of such peculiarity and immediacy should have escaped Tse's mind and attention when the police interviewed him after Wang's second kidnap was highly improbable. If they did not escape Tse's mind and attention, he would have no doubt mentioned them to the police.

507. Mr. Thomas was incorrect to say that Tse was the only witness available to give evidence. Mr. Thomas was also not correct to say that all witnesses had been identified.

508. Mr. Thomas argued that "Lee K Sg" was in all probability Lee Kin Sang, who died in 1996, to justify why another purported witness to the execution of "the 1990 documents" could not be called. This argument contradicted Mr. Lee's suggestion at trial that it was not the Wife's case that "Lee K Sg" was Lee Kin Sang.

509. The Wife could have identified "Lee K Sg" at least to the extent of determining if "Lee K Sg" was or was not Lee Kin Sang.

510. However, the Wife made no attempt to find out who "Lee K Sg" in fact was. If her case was that "Lee K Sg" was Lee Kin Sang, with the number of handwriting experts at her service she could have established this with handwriting evidence. There were documents with Lee Kin Sang's signatures readily available.

511. If her case was that "Lee K Sg" was someone else, she could have made an effort to find out his or her true identity and, if successful, call him or her as a witness. The Wife did not do this either.

512. There was absolutely no attempt to find out who the authors of "the 1990 documents" were. A simple advertisement could have done the trick-after all, not many people had the opportunity of writing a will and similar documents for a multi-billionaire or witnessing his execution of them.

513. No one who wrote any of "the 1990 documents" or witnessed their execution could have forgotten about the events or the contents of the documents.

514. If someone had written any of "the 1990 documents" or witnessed their execution one month before Wang's kidnap, he/she probably would have come forward on his/her own initiative when Wang's kidnap was published.

515. The written submissions for the Wife suggested that when Wang was hospitalised after his riding accident, nurse(s) in the hospital could have written "the 1990 documents". If such suggestion were correct, the task of locating such author(s) would have been a very simple task indeed.

516. In Coles v Coles and Brown (1866) LR 1 P&D 70, it was held that "a party propounding a will is bound to call one at least of the attesting witnesses, if he can be produced, to prove the due execution, if such witness fails to prove due execution he is bound to call the other, although he may know him to be an adverse witness".

517. The Wife's failure to produce or attempt to produce any witness who could have written any of "the 1990 documents" or witnessed their execution is most surprising, casting a grievous doubt on their due execution.

518. Tse's highly suspicious departure from Hong Kong on 24 September 1999 was never explained. Mr. Thomas said he could not offer any explanation, only speculating that Tse might have been concerned about his safety or might have feared interference for his involvement with the case.

519. Mr. Thomas tried to brush the issue aside by calling it unimportant. I disagree with this view.

520. After the court granted leaves to the Father to report to the police the forgery of "the 1990 documents", the police was clearly interested in interviewing Tse about those documents. Tse could not have been unaware of this. The Wife and/or her legal advisers would certainly expect it. After all, it was not long ago that Tse gave statements supporting the Wife's case to two different solicitors acting for her.

521. Tse left Hong Kong on 14th September 1999 and returned on 18th September 1999, only to leave again on 24th September 1999, the day after the leave was granted to the Father to make a report to the police. Tse's behaviour gives rise to a compelling inference that he wanted to avoid interview by the police and that those controlling him also wanted to prevent him from such interview.

522. If Tse did not want to be interviewed by the police and those controlling him did not want him to be interviewed, it is definitely legitimate to ask why.

523. On the other hand, there was no identified cause for Tse's alleged fear of interference or for his safety. If Tse indeed had such fears, that he chose to remain in Hong Kong until 14th September 1999 and to return to the city on 18th September 1999 is strange.

524. Although, as a crucial witness, Tse's whereabouts must have been an important concern for the Wife and Chinachem, neither the Wife nor anyone from Chinachem offered an explanation on such matter.

525. Both the Wife and Chinachem must have been able to contact Tse when he was still on the company's payroll and was officially on leave. No attempt was made to bring back Tse. He only returned when he was on the verge of death.

526. The judge's questions on the above suspicions concerning Tse's movement after 24th September 1999 were certainly legitimate and pertinent to the case, particularly in determining the weight to be attached to Tse's statements. Hence such questions are highly relevant to the Wife's case of the due execution of "the 1990 documents".

527. Proof of a will in solemn form can never be taken lightly. The court owes a duty to the deceased to ensure that his testamentary intention will be carried out. In the discharge of such duty, the court must be serious, cautious, and vigilant. [See the suggested approach of Sir John Nicholl in Robson v Rocke (1824) 2 Add 53 and Saph v Atkinson (supra)]

528. It is na?ve to suggest that simply because the best available evidence in support of the Wife's case had been given (itself highly questionable), the court must be satisfied with the due execution of "the 1990 documents" and hence propound them as Wang's last will. Further, the so-called best available evidence consisted of merely untested assertions in the form of affirmations or statements, and the Wife's case was shrouded by the many unanswered questions hereinbefore identified.

529. The reliance on s. 15 of the Oaths and Declarations Ordinance Chapter 11 did not advance the Wife's case at all.

530. A propounder of a will must adduce sufficient evidence by witness(es) to satisfy the court of its due execution. What is required to discharge such burden depends on the specific circumstances of each case. There are no "hard and fast rules".

531. There may well be a case involving a Chinese will written by the testator that a propounder, in proving its due execution, can do very little other than producing the will itself and perhaps presenting evidence on the circumstances in which the will was found.

532. However, "the 1990 documents" were written, not by Wang, but by two or three others. They appeared to have been witnessed by two witnesses, of whom one disappeared suspiciously at a crucial moment. No attempt was made to identify or trace the other. No attempt was made to identify any of the authors of "the 1990 documents" although a simple attempt could have been highly productive.

533. The Wife failed to give evidence in court on the circumstances in which she obtained "the 1990 documents" so as to allow her evidence to be tested by cross-examination. The judgment of Sir John Nicholl in Saph v Atkinson (supra), excerpted below, demonstrates the inappropriateness of such failure:

"I allude to such circumstances as the place in which the alleged will was found- the parties through whose agency it was prepared- its conformity, or the contrary, with the deceased's avowed or presumed testamentary intentions; and so on.

And first, what is the evidence as to the circumstances connected with the finding of this instrument? Now, the circumstances under which a will is found may, as we all know, tend very materially to its authentication. If, for instance, it is found locked up in repositories, to which the deceased only had access, and is produced from such by indifferent and disinterested parties, these are circumstances which, as tending to connect the will with the testator, are strongly confirmatory of its genuineness and authenticity. The disputed will, in this cause, has no advantage of that description: its first production, after the deceased's death, is by Mrs. Saph, the sole executrix, and party principally benefited under- from where remains to be answered."

534. On a proper analysis, the "strong evidence" proffered by Mr. Thomas not only failed to achieve what he suggested (i.e. sufficient to establish due execution) but gave rise to more suspicions, and hence was detrimental to the Wife's case.

535. Nevertheless, the court have to consider all the matters put forward by Mr. Thomas and other relevant matters in deciding the main issue, i.e. whether it was probable that Wang executed "the 1990 documents". If it was improbable, the Wife had failed to prove due execution.

536. I accept Mr. Thomas's suggestion that, in making such decision, we are in as good a position as the trial judge.

The Probability of Wang's Execution of "The 1990 Documents"

537. I have considered carefully all the relevant factors, both supporting and opposing Wang's due execution of "the 1990 documents". I must say that I do not have much difficulty in reaching a conclusion. In my view, the circumstantial evidence overwhelmingly supports the Father's case. The Wife's case is unreasonable and improbable.

538. Wang clearly was a very careful person, as demonstrated by the evidence adduced by the Father and, in particular, by his handling of the "sensitive documents" in the safety deposit box.

539. It is inconceivable for Wang to have executed a will (one concerning assets worth billions of dollars) by way of "the 1990 documents" and without the help of lawyers as he did for his two previous wills.

540. Wang had teams of lawyers and professional advisers to assist him with the attempted trust arrangements and indeed with any other matter. In the afternoon of 12th March 1990 (even at 6 p.m.), Wang should have been able to summon with a phone call many an available and willing lawyer.

541. Further, there was no plausible reason for Wang to enlist two or three unknown persons to prepare "the 1990 documents" and then ask Tse to be a witness. If Wang were indeed "in hurry," it would have been simpler to have a lawyer preparing and witnessing his will.

542. It was unreasonable and improbable for Wang to have someone else write "the 1990 documents" when he could have done it himself. No medical evidence suggested that he was unable to write. If the Wife's case was correct, Wang was even able to manoeuvre a calligraphic pen and sign his name in a classical style, namely "Lishu (隸書) style and Xingshu (行書) characteristics".

543. Wang, in any event, would not have two or three persons to prepare "the 1990 documents". Wang had no reason to let so many people know about his testamentary arrangements.

544. It was improbable that Wang would have tolerated the sloppy preparation of "the 1990 documents," which was inconsistent with Wang's character and with his alleged use of a calligraphic pen.

545. In any event, Wang would very unlikely have signed in a formal and artistic style with a calligraphic pen if he were indeed in a haste to prepare "the 1990 documents" under "post-mortem" thoughts.

546. It was improbable for Wang to simply put "the 1990 documents" respectively into four envelopes and then collectively into a larger envelope without any external markings.

547. It was improbable that Wang would have a manservant (Tse) witnessing his signatures in "the 1990 documents".

548. In any event, it was improbable that Wang found it necessary to make a will in such haste. If he was worried about not surviving the riding accident to the extent of needing a provisional will urgently, he would unlikely have discharged himself from hospital early and returned to work the same day.

549. The medical evidence certainly did not support the contention that his injuries from the accident were life threatening.

550. Wang had the opportunities to prepare a proper will subsequent to 12 March 1990. The suggestion that Wang gave the hastily and sloppily prepared "1990 documents" to the Wife a few days later was extremely unconvincing.

551. Wang was kidnapped in 1983 and according to the Wife, he was concerned that he would be kidnapped again. Also, Wang had an earlier riding accident in 1987.

552. However, Wang did not find it necessary to make a will after either of the two earlier traumatic experiences.

553. The presence of "D" ("One life one love") among serious documents was strange enough in itself. Further, it was particularly improbable and strange for Wang to sign such a saying and to have Tse witnessing his signing of it.

554. There was absolutely no reason for Wang to express the sentiments about his family and/or the Kung family in "B" and "C".

555. Moreover, the sentiments expressed in the two documents were incompatible with the facts. Wang clearly enjoyed good relationships with his parents, particularly with his mother. That Wang was very close to his mother was undisputed.

556. Objectively speaking, Wang had no reason to be disappointed with his siblings, who all had excellent academic achievements and were gainfully employed in USA.

557. The suggested ill feelings for the Kung family were also artificial.

558. Wang had no reason to express such sentiments against either the Wang and Kung families in testamentary documents. What was he trying to prove or justify?

559. There was no reason for Wang to completely sever the Father from his estate, nor would such an action be consistent with their relationship. The provisions Wang made for his parents, considering his "frugal" character, were generous.

560. In my view, the unusualness of the contents of "B" and "C" is very telling and betrays the intentions of whoever prepared them. To borrow Mr. Chang's metaphor, he/she "was adding legs to a drawing of snake" (劃蛇添足).

561. There is also inherent inconsistency in the contents of "B" and "C". For example, "B" requested that the Wife take care of Wang's parents and his younger sister despite the disappointment about them expressed in "C".

562. Mr. Thomas agreed that "A", "B", "C", and "D" stood or fell together. The contents of "B", and "C" certainly cast grave doubts on the validity of "A".

563. Mr. Thomas suggested that the marital dispute in the mid-1960s had long been forgotten-presumably, the Wife's "infidelity" forgiven as well. Mr. Tang disagreed. He said, "forgiven perhaps, but certainly not forgotten".

564. Only Wang knew if he had forgiven the Wife, but there was positive evidence that he was still concerned about her "faithfulness" as late as 1974.

565. It is in my view improbable for Wang to have forgotten the Wife's infidelity despite the obvious improvement in their relationship. It is even more improbable for Wang to have forgotten "the 1968 Will," which he kept in the safety deposit box away from the Wife together with the other "sensitive documents".

566. It must be remembered that Wang's safety deposit box in question was not used for the storage of general valuables. Wang used it specifically to hide from the Wife sensitive items against her.

567. Mr. Thomas's suggestion in the written submissions that Wang could have overlooked "the 1968 Will" because the safety deposit box "was full of other documents as well" is highly unconvincing. Almost bizarre is his suggestion that Wang must have forgotten what he kept in the safety deposit box.

568. The reason in support of the suggestion that Wang had wanted to cancel "the 1968 Will" in 1985 is far-fetched. If instructions to cancel it were given to any solicitors, such instructions surely would have been carried out.

569. No evidence from any solicitor was given to support the cancellation instructions or explain why the instructions, if given, were not carried out.

570. If Wang had indeed wanted to cancel "the 1968 Will", he probably would have destroyed the original copy in the safety deposit box.

571. The incident in 1985 is certainly inconsistent with the suggestion that Wang had by then forgotten about "the 1968 Will".

572. The Wife herself said in her affirmation that Wang had mentioned "the 1968 Will" to her in late-1979.

573. The suggestion that Wang had forgotten about "the 1968 Will" and/or the safety deposit box is inconsistent with the evidence and, in any event, unconvincing.

574. If Wang had not forgotten about "the 1968 Will", it is improbable, and indeed comical, that he should make a will in 1990 in favour of the Wife (with the dying declaration, "One life one love") while keeping "the 1968 Will" in favour of the Father together with photographs of his ex-girlfriends.

575. I find it even more improbable and comical that Wang would instruct the Wife not to open the envelope containing "the 1990 documents" until after his death, when the jealously kept evidence of his own infidelities would be exposed in juxtaposition with his "One life one love" declaration.

576. The suggestion that Wang prepared "the 1990 documents" because he was concerned about his well being after the riding accident is again unconvincing.

577. The medical evidence did not show any life-threatening conditions. Wang could and did return to work after he discharged himself against doctors' advice from the hospital, to which he did not return for further examination despite Dr. Wu's express instruction to do so.

578. If Wang did execute "the 1990 documents", he would have at the first available opportunity removed the sensitive documents that he kept in the safety deposit box of a nearby bank to avoid the embarrassment to the woman whom he declared to be his "One life one love".

579. Wang might have wanted to keep the photographs of his girlfriends for their sentimental value, but if he had indeed forgiven and forgotten and/or had intended the Wife to inherit his estate, he would have had "the 1968 Will" destroyed long before 1990.

580. If Wang were indeed concerned that the Wife should inherit his estate in case he died unexpectedly, he would have made a proper will in her favour after his 1983 kidnap or the 1987 riding accident or in the many years before his second kidnap. He certainly had the opportunity to do so even after the riding accident on 10 March 1990, as he was not kidnapped again until one month later.

581. On her case, the Wife could have reasonably inferred from the circumstances that "the 1990 documents" (then unopened) contained the latest will of Wang that superseded "the 1968 Will". However, without knowing the contents of "the 1990 documents," she could not have known (and claimed, as she did in her affirmations) that she would be the sole beneficiary of Wang's estate under the new will. The judge was entitled to conclude that the Wife must have prior knowledge of the contents of "the 1990 documents".

582. It was neither reasonable nor probable that Wang would make no provision for his parents and (if indirectly) his siblings.

583. The evidence of the couple's "loving relationship," the Wife's contributions to the Chinachem business, and the abortive trust arrangements do not in my view tilt the balance in favour of the Wife at all.

584. In conclusion, the circumstantial evidence does not support the Wife's case. Considering the sum of all circumstances, the execution of "the 1990 documents" by Wang is improbable and was rightly rejected.

The Wife's Silence

585. Serious complaints were made of the judge's drawing adverse inference against the Wife from her failure to give evidence.

586. As Mr. Thomas rightly pointed out, the Wife might not have been able to explain everything, not on her case anyway. However, it was the Wife who sought to propound "the 1990 documents" as the last will of Wang. Therefore she should have at the very least told the court the circumstances in which she obtained "the 1990 documents" to satisfy the court's conscience that Wang did execute it.

587. After all, the circumstances in which the Wife came to possess "the 1990 documents" was an important factor on the issue of due execution, particularly given the absence of statutory safeguards identified earlier in this judgment.

588. Moreover, the Wife should have allowed her assertions to be properly tested by cross-examination instead of simply relying on her affirmations.

589. As Mr. Tang rightly pointed out, if the court were to accept the suggestion that Wang handed the "the 1990 documents" to the Wife, much suspicion about the genuineness of Wang's questioned signatures would have been dispelled.

590. When such evidence was available but withheld, and when there was so much other circumstantial evidence opposing the Wife's case, the court could not in good conscience be satisfied that "the 1990 documents" were executed by Wang.

591. One must not overlook that the Wife only produced "the 1990 documents" reluctantly and many years after Wang's disappearance and after the safety deposit box was opened.

592. If Wang and the Wife was indeed a "loving couple", she must have been concerned about his safety.

593. The police was informed of the kidnap on the very day it happened. The documents-supposedly given to the Wife only a month before the kidnap-could have provided clues for the crime. In any case, there was no conceivable excuse for the Wife to keep them in secret.

594. It is therefore very strange and suspicious that "the 1990 documents" were never brought to the attention of the authorities investigating the kidnap.

595. It was suggested that the Wife refused to reveal "the 1990 documents" because she believed Wang to be still alive and only adhered to his instruction not to open the sealed envelope until after his death. Such suggestion is highly unconvincing, if not outright ridiculous.

596. At one point, the Wife claimed that she believed that if the kidnappers had knowledge of "the 1990 documents," Wang's safety would be put at risk. Her belief was incomprehensible when she supposedly knew nothing about the contents of the documents. Moreover, her belief would in fact be a strong motivation to reveal the existence of the documents and their contents to the police.

597. The evidence clearly suggested that Tse was very much under the Wife's influence. He was a Chinachem employee and had been the couple's "house servant" for a long time.

598. Tse's sudden departure on 24 September 1999 begged explanation. If it did not occur at the Wife's instigation, it was unreasonable that she made no enquiries about it. On the other hand, if the Wife had made enquiries, the court should have been provided with the results and/or related information.

599. The Wife failed to explain what appeared to be discrepancies between her case and her affirmations.

600. She could have told the court what happened in the afternoon of 6 September 1999 and provided answers to many other troubling questions.

601. She could have given evidence on "the holiday wills," any one of which, if accepted, would revoke "the 1968 Will". She must have knowledge about the contents of at least some of "the holiday wills".

602. If "the holiday wills" truly existed, she would not have failed to mention them in her affirmations when they were, as pointed out by her counsel in the application to amend her defence and counterclaim, an important part of her case.

603. If the proposed trust arrangements in 1987 and in 1989-1990 constituted the importance in the Wife's case as purported, it was reasonable to expect her to tell the court the exact nature of the proposed arrangements.

604. Beyond those just mentioned, there were many other legitimate questions regarding the issue of due execution that the Wife could have provided answers or at least information on.

605. Such questions were not mere unfounded suspicions or "curiosities" as Mr. Thomas lightly suggested, but could lead to reasonable inferences to be drawn against the Wife. They did have probative value and they were not irrelevant. The court was entitled to have regard to such matters on the issue of due execution.

606. Mr. Thomas relied on Wisniewski v Central Manchester Health Authority [1998] Lloyd's LR 223 to support his contention that the Wife's silence should lead to neither adverse inference nor adverse comment by the judge.

607. Brooke LJ in Wisniewski (supra) made the following observation at p. 240 of the judgment:

"From this line of authority I derive the following principles in the context of the present case:

(1) In certain circumstances a court may be entitled to draw adverse inferences from the absence or silence of a witness who might be expected to have material evidence to give on an issue in an action.

(2) If a court is willing to draw such inferences they may go to strengthen the evidence adduced on the issue by the other party or to weaken the evidence, if any, adduced by the party who might reasonably have been expected to call the witness.

(3) There must, however, have been some evidence, however weak, adduced by the former on the matter in question before the court is entitled to draw the desired inference: in other words, there must be a case to answer on that issue.

(4) If the reason for the witness's absence or silence satisfies the court then no such inference may be drawn. If, on the other hand, there is some credible explanation given, even if it is not wholly satisfactory, the potentially detrimental effect of his/her absence or silence may be reduced or nullified."

608. The aforesaid comments were made following the reference to the judgment of Lord Lowry in T C Coombs v IRC [1991] 2 AC 283 on p. 300;

"In our legal system generally, the silence of one party in face of the other party's evidence may convert that evidence into proof in relation to matters which are, or are likely to be, within the knowledge of the silent party about which that party could be expected to give evidence. Thus, depending on the circumstances, a prima facie case may become a strong or even an overwhelming case. But, if the silent party's failure to give evidence (or to give the necessary evidence) can be credibly explained, even if not entirely justified, the effect of his silence in favour of the other party may be either reduced or nullified."

609. Mr. Thomas argued that there was no issue that reasonably required any explanation from the Wife. He also suggested that the Wife's silence could have been caused by the police investigation against her and was hence excusable.

610. Mr. Thomas particularly pointed out that the Father had, in the course of his evidence, made serious allegations of criminal conduct against her, on which the police showed keen interest.

611. I am surprised by the suggestion that there was no issue upon which the Wife could or might be expected to give material evidence on. I have identified some of those issues and do not wish to repeat them. Those issues were within the Wife's knowledge and she could be expected to give evidence on them.

612. I disagree with Mr. Thomas's submission that as the only issue in question was due execution and, because the Wife was not present when Wang's allegedly executed "the 1990 documents," she could not and might not be expected to give any material evidence on it.

613. I also disagree with Mr. Thomas's suggestion that, insofar as the circumstances in which the Wife obtained "the 1990 documents" were relevant, her affirmations alone had adequately dealt with them.

614. I would like to emphasize that the issues identified related not to the question of forgery, but to the question of due execution, on which the Wife bore the burden of proof as the propounder of "the 1990 documents".

615. She had the duty to satisfy the court's conscience by witnesses that Wang duly executed "the 1990 documents" as his last will, so that the court could safely pronounce them in solemn form.

616. Mr. Baron Parke in Barry v Butlin 2 Moo PC 480 said on p. 482,

"The rules of law according to which cases of this nature are to be decided, do not admit of any dispute, so far as they are necessary to the determination of the present Appeal: and they have been acquiesced in on both sides. These rules are two; the first that the onus probandi lies in every case upon the party propounding a Will; and he must satisfy the conscience of the Court that the instrument so propounded is the last Will of a free and capable Testator.

[...]

The strict meaning of the term onus probandi is this, that if no evidence is given by the party on whom the burden is cast, the issue must be found against him. In all cases the onus is imposed on the party propounding a will, it is in general discharged by proof of capacity, and the fact of execution, from which the knowledge of assent to the contents of the instrument are assumed..."

617. I will now discuss the Wife's claim to the right of silence together with the suggestion that her silence should be excused because of the police investigation.

618. In the notice of appeal, the Wife raised the issue that "the trial was unfair and prejudicial because of the inherent conflict between the Wife's right as a party to give oral evidence in support of her case at the trial and her constitutional right of silence" (Emphasis added). The conflict was said to have arisen from the police investigation against her for alleged fraud.

619. The notice of appeal clearly confirms the relevance of her evidence and contradicts Mr. Thomas's claim that the Wife could not be expected to give any material evidence. To be fair, Mr. Thomas did not pursue this point raised in the notice of appeal, although he did not abandon it either.

620. In my view, that the Wife was under police investigation could admit neither excuse nor justification whatsoever for not doing what was obviously expected of her.

621. The right to silence (i.e. the rule against self-incrimination) allows a party to refrain from making admissions in the witness box that may expose him or her to any criminal charge.

"...the rule is that no one is bound to answer any question if the answer thereto would, in the opinion of the judge, have a tendency to expose the deponent to any criminal charge, penalty or forfeiture which the judge regards as reasonably likely to be preferred or sued for. Goddard LJ in Blunt v Park Lane Limited [1942] 2 KB 253"

622. Here it is perhaps helpful to refer to the classic statement of Cockburn CJ in R v Boyes (1861) 1 B. & S. 311 on p. 330:

"To entitle a party called as a witness the privilege to the privilege of silence, the Court must see, from the circumstances of the case and the nature of the evidence which the witness is called to give, that there is reasonable ground to apprehend danger to the witness from his being called to answer...The danger to be apprehended must be real and appreciable with reference to the ordinary operation of law in the ordinary course of things- not the danger of an imaginary and unsubstantial character, having reference to some extraordinary and barely possible contingency so improbable that no reasonable man would suffer it to influence his conduct...in merely remote and naked possibility, out of the ordinary course of the law and such that no reasonable man would be affected by, should not be suffered to obstruct the administration of justice."

623. If Wang had duly executed "the 1990 documents," there would be no reasonable ground for the Wife to apprehend danger of any kind in testifying in court to such effect and to the circumstances in which the documents came to her possession.

624. There could be no risk of the Father's bringing in irrelevant information to incriminate the Wife. Any attempt to make allegations of unrelated criminal conduct against her would no doubt be stopped.

625. On the other hand, if the Wife chose not to give evidence for fear that she might be caught lying or for fear that her evidence on the relevant issues might lead to true incriminating matters, she could not legitimately complain about the judge's adverse comment against her.

626. Public policy requires that a party must not be allowed to make use of his/her own wrongdoing to advance his/her case as Templeman LJ observed in Rank Film Distributors Limited v Video Information Centre [1982] AC 380 on p. 423:

"Where a defendant in a civil action relies on the doctrine against self-incrimination and insisted on remaining silent and on concealing documents and other evidence relevant to the action, he is relying on his own wrongdoing or on his own apparent or possible wrongdoing to hamper the plaintiff in the proof of his just claims in the suit."

627. If the court were to rule that the Wife's silence was justified or excusable because of the police investigation, it would be a judicial sanction of attempts to rely on a party's possible wrongdoings to hamper the other party's case against him/her.

628. On this aspect, it is perhaps relevant also to refer to the judgment of Lloyd J in Hart v Dabbs & Ors, unreported, Ch. D. 6 July 2000 at para. 72:

"That was a very different case. It seems to me that, where the silent person is he who propounds the will in the probate action itself, it is legitimate for the court to have regard to his silence, not to the extent of finding the opposite case proved by that silence, but to the extent indicated by the passage quoted above from Scarman J. in Re Fuld deceased (No. 3), as contributing to the cloud of suspicion which may surround the will and which it is for the Claimant, by whatever evidence he does adduce, to dispel if the will is to be admitted to probate."

629. In summary, if the Wife were an innocent person telling the truth, as she claimed herself to be, giving evidence on oath would not expose her to any legitimate risk. Therefore she had no claim to invoke the right to silence, and her failure to give evidence could not be justified or excused in the circumstances of the present case. The judge was entitled to draw an adverse inference against her.

Teresa's Evidence and the Judge's Finding that the Wife was the author of "A"

630. Mr. Thomas complained about the judge's finding that the Wife was the author of "A". His complaint had two aspects: (i) that the judge was wrong to allow Teresa to produce the Power of Attorney and the "Fotomax" envelope, when it had not been established that the writings on these two items belonged to the Wife and (ii) that the judge was wrong to act on Teresa's handwriting identification.

631. Unless the writings on the Power of Attorney and on the "Fotomax" envelope were established to be by the Wife, the experts' comparison of those writings with the text of "A" to determine the latter's author would be meaningless.

632. In the course of her evidence, Teresa sought to produce the "Fotomax" envelope, which she claimed to have been sent to her together with photographs taken by either Wang or the Wife during the couple's trip to the Niagara Falls in 1984. Teresa identified the writing on the "Fotomax" envelope to be that of the Wife for the following reasons:

(i) Only Wang or the Wife could have sent the photographs, which were taken with their camera;

(ii) The writing was not that of Wang, which she recognized;

(iii) The contents of the writing matched an earlier conversation that Teresa had with the Wife but not any conversation that she had with Wang; and

(iv) On the document were the initials "Ru Xin," which Teresa recognized to be the Wife's.

633. When Teresa sought to identify the Wife's initials, "Ru Xin" (如心), the judge asked Mr. Lee if he had any objections. Mr. Lee replied that he wished to reserve his objection until he heard more. Teresa's evidence relating to the "Fotomax" envelope was then received de bene esse.

634. Teresa also produced the Power of Attorney, which, she claimed, was sent to her after a few telephone conversations with the Wife about Wang's application to emigrate to USA. The judge accepted this claim.

635. Mr. Lee did not indicate any objection to the above claim. Teresa again confirmed that the writing on the Power of Attorney was that of the Wife.

636. When Teresa sought to identify the text of "A" to be the writing of the Wife, Mr. Lee again indicated no objection to receiving such evidence de bene esse.

637. Mr. Lee's initial objection was not pursued orally. The judge was not asked by either party to adjudicate on the admissibility of Teresa's evidence on the Power of Attorney and the "Fotomax" envelope. Teresa was not cross-examined at all on such evidence.

638. The objection to admissibility was raised again only in Mr. Lee's final submission. In the end, the judge found the writings on the Power of Attorney and the "Fotomax" envelope to be the Wife's.

639. If Teresa's evidence was received de bene esse, the only relevant issue was admissibility. If Teresa's evidence was disputed, she should have been cross-examined with her evidence challenged.

640. Mr. Thomas argued that, even if no proper objection had been raised at trial, we could still exclude Teresa's evidence if it was proper to do so. He also suggested that, even if the Wife did send those documents, their contents could have been written by a secretary on her dictation.

641. I can see no valid reason to object to the admissibility of the Power of Attorney and the "Fotomax" envelope as containing the Wife's writings.

642. Here we are not interested in what was possible or what "could have been," but in drawing reasonable inferences based on Teresa's evidence. Mr. Thomas's suggestion that the contents of the Power of Attorney and the "Fotomax" envelope could have been written by a secretary on the Wife's dictation is not a suggestion that can be taken seriously in the circumstances of this case.

643. Based on Teresa's unchallenged evidence and the contents of the documents themselves, in the absence of any opposing evidence from the Wife, the judge was perfectly entitled to receive Teresa's evidence that the writings on two documents were the Wife's.

644. It is perhaps worthwhile to repeat the judge's sentiment expressed on p. 391 of his judgment:

"I just find the way the defence was conducting her case by putting the plaintiff to strict proof without saying whether it was the defence case that those writings were the defendant's or not, is a complete waste of the Court's time."

645. Mr. Thomas also objected to Teresa's evidence on the authorship of "A" on the basis that Teresa was not competent to give such evidence. He further said that Teresa's recognition of the handwriting on "A" had not been pleaded and that it was against the public policy that disallows identification by a layperson, which he said was haphazard and could thereby cause injustice.

646. Mr. Thomas cited Pitre v The King [1933] 1 D.L.R 417, Rex v Arthur [1947] 4 D.L.R. and R v O' Brien [1911] 7 Cr App Rep 29 to support his contention that the witness's regular correspondence with the person in question and frequent sighting of that person's handwriting were prerequisite to the witness's competence to identify that person's handwriting.

647. In Duke v Duke [1975] 12 SASR 106, Sangster J drew a distinction between the admissibility of and the weight to be attached to identification evidence. He wrote on p. 108:

"The passage in Cross on Evidence to which Miss Nyland referred me concludes as follows: 'There must, however, have been a sufficient opportunity for the witness to acquire such knowledge of the handwriting in question as to make it worthwhile receiving his evidence.' This passage is supported by a reference to two cases, one English, the other Canadian, neither of which I find to be very helpful. In Taylor on Evidence the paragraphs which follow that to which Miss Nyland referred me suggest (partly on the authority of Lewis v Sapio) that a very slight acquaintance is sufficient to justify the evidence, but that the weight to be attached to the evidence once admitted is another matter. Of these two views I prefer that expressed by Professor Cross."

648. From the aforesaid authorities, the correct position appeared to be that if a layperson's identification evidence was based on tenuous and slight acquaintance, such evidence was inadmissible.

649. According to Teresa, she could recognize the Wife's handwriting because of their calligraphic lessons in the 1950s, the holiday shopping lists written by the Wife, the Power of Attorney, and the "Fotomax" envelope. In my view, such acquaintance was not "tenuous and slight".

650. Section 17 of the Evidence Ordinance Cap. 8 provides:

"Comparison of a disputed writing with and writing proved to the satisfaction of the court to be genuine shall be permitted to be made by witnesses in any proceedings, and such writings, and the evidence of witnesses respecting the same, may be submitted to the court and to the jury, if any, as evidence of the genuineness or otherwise of the writing in dispute."

Phipson on Evidence (15th Edition) on p.984 comments with reference to a similar English provision:

"In a criminal case a judge may permit disputed handwriting to go before the jury for the purposes of comparison, pursuant to section 8 o the Criminal Procedure Act 1865 only if he is satisfied beyond reasonable doubt of the genuineness of the handwriting. In civil cases, of course, the civil standard of proof on the balance of probabilities applies.

A statement that the witness is acquainted with the party's writing is generally sufficient in chief, it being for the opponent to cross-examine as to means and extent. Such knowledge may be acquired: (a) By having at any time seen the party write, though the value of the opinion will, of course, vary with the frequency and recentness of the occasions and the attention paid to the matter by the witness; or (b) by the receipt of written communications purporting to be in his handwriting, in reply to documents addressed to him by or on behalf of the witness...."

651. Footnote 67 on the same page further stipulated, "But the threshold requirements for admissibility are not high." (Emphasis added)

652. The pleading point is not valid. The authorship of "A" was one piece of the circumstantial evidence, and I have already established in the above text that circumstantial evidence did not need to be pleaded to be admissible.

653. In my view, the judge had valid bases in admitting Teresa's recognition of "A" as the Wife's handwriting. However, it is still necessary to consider the extent to which the court could rely on such evidence, i.e. the weight to be attached to it.

654. Clearly, Teresa was not an impartial witness. Her description of the Wife demonstrated that she harboured substantial dislike, if not outright hatred, for her.

655. Although Teresa said she was able to recognize the Wife's handwriting and explained why, Teresa did not have too many opportunities to observe the Wife's handwriting, particularly in recent years. Further, Teresa's recognition of the Wife's handwriting was partly based on extrinsic evidence.

656. The judge correctly did not rely heavily on Teresa's recognition of the Wife's handwriting. He said, "The evidence of Teresa is not as strong as the experts' opinion [...] I do find on the experts' evidence alone, the writing in Document A is probably the handwriting of the defendant herself".

657. The evidence of the Father's experts on the authorship of "A" could hardly be described as strong. Mr. Lesnevich simply said the Wife "might" be the author of "A". Mr. Cheng said he could not eliminate the possibility of the Wife being the author of "A". Their evidences were not certain or determinative.

658. Of all the experts, Mr. Tsui was, at first glance, the most definite, suggesting that the Wife "could well have been" the author of "A". Even so, Mr. Tsui only put the possibility at +2 on a scale of -10 to +10. According to this scale, +1 to +3 simply indicated possibility with the use of the word "might".

659. That the Wife was the author of "A" was indeed a serious finding with possibly far-reaching consequences and could only be justified if supported by strong and cogent evidence.

660. Mr. Tang fairly conceded that even if Teresa's evidence was not challenged, the court did not necessarily have to accept such evidence.

661. Although Teresa's identification evidence was not challenged and despite Mr. Lee's remark, "Even if the Wife was the author of "A", so what?" I am not persuaded that the evidence was strong and cogent enough to justify the judge's finding that the Wife authored "A".

662. If the Wife were indeed the author of "A", there would be nothing left of her case. However, such finding was not sufficiently supported by evidence or justified in all the circumstances, it must be set aside.

The Judge's finding that the Wife was the Author of "D"

663. The judge found the Wife to be the author of "D" based on the writings on the perceived document "E".

664. The judge argued that since "Nina" was written beneath "One Life One Love" and positioned as if it were a signature, the Wife, whose name was Nina, must have signed "E".

665. I have expressed my surprise at such finding, which I found to be unreasonable and unsubstantiated. His inference-leading to a serious and important finding of fact possibly with far-reaching consequences-was haphazard and poorly supported. Hence I set aside such finding as well.

666. That said, I do not rely on either of the judge's findings that the Wife authored "A" or "D" in resolving this appeal.

667. Further, my rejection of the aforementioned findings does not affect my earlier conclusion on the overall circumstantial evidence (that it was inconsistent with the Wife's case).

668. Nevertheless, as mentioned before, I accept Mr. Thomas's suggestion that, in determining whether Wang duly executed "the 1990 documents," the court must also consider the handwriting evidence of the experts together with the overall probabilities of the case. With this in mind, I now turn to the handwriting evidence.

Handwriting Evidence

669. The judge had valid reasons to disallow the experts' reference to Tse's Home Visit Permit and the two vouchers for comparison purposes. His reasoning could not be faulted.

670. The handwriting evidence took no less than 70 days before the judge. Over 200 pages of the judgment were devoted to such evidence alone.

671. The judge did not only review comprehensively the experts' evidence but also referred to leading textbooks on document examination. He set out in detail the relevant considerations for handwriting examinations, as well as the respective roles of the judge and the experts.

672. Assisted by the experts' evidence, the judge analysed the questioned signatures first macroscopically and then microscopically.

673. In this appeal, both parties again presented the court with very lengthy submissions that contained many charts and diagrams and excerpts from leading textbooks on document examination. In no less than 15 days, Mr. Chang and Mr. Chan both took us through their respective arguments painstakingly to demonstrate their validity.

674. The case involved a significant amount of experts' evidence. It must be borne in mind that the task of an expert is to present and not to proselytise, to give evidence and not to decide. Expert evidence is to be confined to theoretical aspects. The judge decides if an expert opinion should be accepted.

675. Jacob J in Routestone Ltd v Minories Finance Ltd [1997] BCC 180 commented:

"I think this provision makes the opinion admissible. It by no means follows that the court must follow it. On its own (unless uncontested) it would be 'a mere bit of empty rhetoric' (Wigmore, Evidence (Chadbourn rev.), para. 1920). What really matters in most cases is the reasons given for the opinion. As a practical matter a well-constructed expert's report containing opinion evidence sets out the opinion and the reasons for it. If the reasons stand up the opinion does, if not, not." (p.188)

676. The judge was entitled and indeed bound to contribute his own knowledge, experience, and common sense to reach a conclusion, particularly in an inquisitorial hearing. In so doing, the judge must not consider expert evidence in a vacuum but within the context of the entirety of the evidence.

677. The attack that the judge "played expert" was premised on his findings of further and other significant differences (e.g. retouching, retracing, and unnatural writing) in some of Wang's questioned signatures although neither the Father's experts nor Mr. Cheng had identified those differences.

678. Professor Jia gave evidence before the judge with the use of a microscope and a projection screen for almost 40 days. A lot of his evidence was not mentioned in the joint reports that he co-authored, or to the Father's experts. Some of the slides used by Professor Jia were converted into computer images available to the judge. He therefore could not have been unaware that the evidence by the Father's experts did not mention, let alone support, the further and other significant differences.

679. However, the judge had himself raised the further and other significant differences to Professor Jia. The judge certainly was entitled to and indeed must do so in an inquisitorial approach to the issues before him. Professor Jia then had the opportunity to offer his explanations, which the judge took into consideration.

680. The judge was entitled and justified to form his own view based what he saw and heard in court-he had himself looked at the relevant characters through the microscope and followed closely the demonstrations in Professor Jia's testimony. He was entitled and justified to reject Professor Jia's explanations and to arrive at a contrary conclusion.

681. That there was no evidence from the Father's experts did not preclude the judge from disagreeing Professor Jia's opinions.

682. The further and other significant differences identified by the judge were based on common sense and/or reasonable inferences.

683. The judge's conclusion on such differences was well reasoned.

684. Mr. Chang suggested that the indicia of retouching and retracing identified by the judge could be the result of "optical illusion", but the expert Professor Jia did not offer such explanation. The judge could not be criticized for not considering the suggested "optical illusion".

685. I fail to see any merit in the suggestion that the judge "played expert" as a ground to vitiate any of his findings.

686. When the judge said on the experts' evidence alone, the only conclusion was that the questioned signatures were forged, he must be referring to what could be made out of such evidence, including his own evaluation of such evidence.

687. The judge clearly set out a proper and undisputed approach to expert evidence. The judge's decision to accept Western handwriting expertise and methodology in the analysis of Chinese signatures is not, in my view, objectionable.

688. The judge was entitled and correct to take the view that knowledge of Chinese was not an essential requirement in the identification of Chinese signatures, which concerned the quality of dots, strokes, and lines-features of writing not unique to Chinese.

689. Nevertheless, the judge was aware of the possible shortcomings of such evidence. He was entitled to determine the weight to attach to such evidence.

690. Mr. Tsui and Mr. Cheng are both experienced leading experts in Chinese handwriting, particularly Chinese signatures.

691. The approach to take in case of conflicting experts' evidences is no different from that of conflicting evidences from other witnesses. Black-Kerr J in Cathay Textile Corpn v Eastern Zone Industrial Co. Ltd. [1962] HKLR 139 said the following:

"The present case depended upon what view was to be taken by the trial judge of the conflicting testimony of the expert witnesses on each side. We see no difference between conclusions arrived at by a trial judge on expert evidence and conclusions arrived at upon any other form of oral testimony. Applying the principles laid down in Watt v Thomas, as it appears to us, there is no foundation laying it open to this Court to say that the learned trial judge erred in his conclusions, there being nothing on the record to indicate that he failed to take due advantage of his opportunities of assessing the expert testimony given on each side." (p.148)

Viscount Simon in Watt v Thomas [1947] 1 All E R 582 laid down, in no uncertain term, the approach that an appellate court should take:

"I desire to make some observations as to the circumstances in which an appellant court may be justified in taking a different view on facts from that of a trial judge....Apart from the classes of case in which the powers of the Court of Appeal are limited to deciding a question of law (e.g., on a Case Stated or on jurisdiction to review the County Courts Acts) an appellate court has, of course, jurisdiction to review the record of the evidence in order to determine whether the conclusion originally reached on that evidence should stand, but this jurisdiction has to be exercised with caution. If there is no evidence to support a particular conclusion (and this is really a question of law), the appellate court will not hesitate so to decide, but if the evidence as a whole can reasonably be regarded as justifying the conclusion arrived at the trial, and especially if that conclusion has been arrived at on conflicting testimony by a tribunal which saw and heard the witnesses, the appellate court will bear in mind that it has not enjoyed this opportunity and that the view of the trial judge as to where credibility lies is entitled to great weight. This is not to say that the judge of first instance can be treated as infallible in determining which side is telling the truth or is refraining from exaggeration. Like other tribunals, he may go wrong on a question of fact, but it is a cogent circumstances that a judge of first instance, when estimating the value of verbal testimony, has the advantage (which is denied to courts of appeal) of having the witnesses before him and observing the manner in which their evidence is given. What I have just said reproduces in effect the view previously expressed in this House.....that a court of appeal should 'attach the greatest weight to the opinion of the judge who saw the witnesses and heard their evidence' and, consequently, should not disturb a judgment of fact unless they are satisfied that it is unsound". (p.583 G-H)

In English v Emery Reimbold & Strict (supra), Lord Philips MR suggested a similar approach to conflicting experts' evidence:

"The first two appeals with which we are concerned involved conflicts of expert evidence. In Flannery's case [2000] 1 WLR 377 Henry LJ quoted from the judgment of Bingham LJ in Eckersley v Binnie (198) 10 Con LR 1, 77-78 in which he said that 'a coherent reasoned opinion expressed by a suitably qualified expert should be the subject of a coherent reasoned rebuttal'. This does not mean that the judgment should contain a passage which suggests that the judge has applied the same, or even a superior, degree of expertise to that displayed by the witness. He should simply provide an explanation as to why he has accepted the evidence of one expert and rejected that of another. It may be that the evidence of one or the other accorded more satisfactorily with facts found by the judge. It may be that the explanation of one was more inherently credible than that of the other. It may simply be that one was better qualified, or manifestly more objective, than the other. Whatever explanation may be, it should be apparent from the judgment." (p.2418 E-G)

692. By the aforesaid principle, I fail to see the justification for most of the attacks on the judge's approach to the experts' evidence.

693. If a group of signatures are genuine, there will of course be similarities. Even in a forged signature, there are bound to be significant similarities with the genuine one (if the forgery is not a complete failure).

694. One must bear in mind that people properly trained in Chinese writing write in similar, if not identical, stroke movements and sequences. A worthwhile forger would definitely try to forge with as many similarities and as few variations with the genuine signatures as possible.

695. In contrast to signature examination, handwriting examination does not generally concern deliberate attempts to produce identical or identical-seeming copies of a certain piece of writing. In handwritings by two different persons, there are bound to be more variations than similarities.

696. It is logical to start with identifying differences first in the comparison of signatures and with identifying similarities first in the comparison of handwriting. In any case, this is just a matter of convenience, as Mr. Tsui tried hard to explain in his evidence.

697. The "controlling factor" theory advanced by Mr. Tsui was not such that a "controlling factor" should be given preference or predominance in determining the genuineness of handwriting or signature. He clearly emphasized that there must still be overall assessments of both differences and similarities.

698. Mr. Tsui's theory is not illogical and unfair as Mr. Chang sought to persuade us. It constitutes a reasonable working platform and is not inherent and structural bias as Mr. Chang suggested.

699. The term "controlling factor" in its literal translation into Chinese might have misled Professor Jia (who appeared not to be proficient in English) into attacking Mr. Tsui's theory before us in the same way Mr. Chang did.

700. Counsel had not been and should not have been thus misled. The judge certainly was not misled.

701. In fact, in the joint report, the Wife's experts appeared to have adopted the same approach when they relied on the similarities in six common characters to reach their conclusion that "B" and "C" were written by the same person and that another person wrote "A".

702. Experts called by both sides agreed that when determining if a signature was forged, significant and unexplained differences might well be more important than similarities.

703. Mr. Lesnevich was confident enough to say that a single inexplicable dissimilarity in a signature alone could lead to a non-finding of authenticity, a suggestion Professor Jia also subscribed to.

704. Mr. Xu, a co-author of the joint report and one of the Wife's experts, made similar point in his book, in which he wrote, "There must not remain significant differences that cannot reasonably be explained".

705. In his testimony, Professor Jia had virtually abandoned the contents of the joint report. After a careful analysis of the contents of the joint report, it is not difficult to see why.

706. To show that Wang's questioned signatures were genuine signatures, the joint report selected a few known signatures (without indicating the criteria of such selection) of Wang and then carried out both a "macro" (overall) and a "micro" (detailed) comparison.

707. In the overall comparison, the joint report rested its findings on the following:

"The level of calligraphy are similar; the layouts of the characters, the locations of the signatures and their style are the same."

"On the (questioned signatures), the character "huei" was written slightly towards the right, and this is in conformity with Wang Teh Huei's habit in writing the character "huei" slightly towards the right in his usual vertical written signatures. In addition, from the (documents containing the questioned signatures), it has been found that Wang The Huei was accustomed to putting down is signature on documents after other people had written the text. It appears that the text of the (documents) were not written by Wang Teh Huei and Wang Teh Huei only signed after the text had been written and this is in conformity with the above-mentioned habit."

708. In the detailed comparison, the joint report suggested that the character "Wang" (王) in Wang's questioned signatures resembles its counterpart in his genuine signatures because:

1. The first horizontal stroke is the longest, moving from left to right and then slightly upward;

2. The vertical stroke is connected with the second and third horizontal strokes.

709. The character "Teh" (德) in Wang's questioned signatures resembles its counterpart in his genuine signatures because:

1. The left radical "double men" (彳) of "Teh" (德) is abbreviated, as is the radical "Shui" (水), which is represented by three dots;

2. The conformity in the features in connected pen wielding in the first and second strokes on the right radical, "four" (四), in "Teh" (德) and the connected pen-wielding, direction, stroke arrangement and proportions in the lower part of the character "heart" (心).

710. The character "Huei" (輝) in Wang's questioned signatures resembles its counterpart in his genuine signatures because of:

1. The connected stroke movement of the radical 'Kwong' (光);

2. The features of connected strokes and directions of the character 'Jun' (軍) on the right;

3. The non-connected last two horizontal and vertical strokes and their ending directions".

711. The so-called reasons advanced in the joint report to support the genuineness of Wang's questioned signatures are most unconvincing and bordered on absurdity. There is no detailed discussion of the differences between the two sets of signatures, although such differences are obvious even to the untrained eye.

712. On the overall comparison, I fail to see the importance of the layouts, the locations, and the general style of the signatures in determining their genuineness. The emphasis on the "habit" of signing after the text had been completed is absurd.

713. How did Professor Jia and his colleagues know that Wang's questioned signatures on "the 1990 documents" were only put after the text had been completed? On what basis did they emphasize this?

714. In the detailed comparison, the reasons advanced for the genuineness of the character "Wang" (王) are again superficial and misleading.

715. The execution of this character with a long first horizontal stroke and then a cursive stroke in place of the vertical stroke and the second and third horizontal strokes is most common and unworthy of mention.

716. The left-to-right movement of the pen in the first horizontal stroke is a standard manoeuvre and hence hardly unique to Wang.

717. The suggestion that the first horizontal stroke of the character "Wang" (王) in the Wang's known signatures moves slightly upward to the right is factually incorrect. From all but one of the samples, the first horizontal stroke of the character "Wang" (王) is a firm and horizontal stroke, often slightly tilting down towards the right. Such feature is obvious even to untrained eyes.

718. The only exception appeared to be 668. A more careful examination, however, shows that the first horizontal stroke in 668 still starts with a downward movement and only tilts upwards slightly after the first half of the stroke has been completed. 735 and 5998, in my view, are no different. The first horizontal stroke of the character "Wang" (王) is written with a horizontal movement with a general downward tendency towards the right.

719. The blinkered approach in the joint report is surprising.

720. The treatment of the characters "Teh" (德) and "Huei" (輝) is convoluted and vague. There is no proper analysis of the similarities or differences and no proper reference to the correct approach to determine the genuineness of a signature as suggested by leading textbooks on document examination, namely:

"An imitation of a writing will naturally resemble the writing imitated and most errors in identifying writing are due to the improper assumption that this 'general character' is proof of genuineness. The sympathetic, senile neighbour who consents to help the claimant, and the retired banker with failing vision who is complimented by being asked to testify; both say, "Yes, that looks like Dave's signature.' To them it has the same 'general character.' In opinions of the highest courts, or some highest courts, this kind of testimony has been highly extolled.

Every questioned writing should be carefully examined with a view of determining whether by itself and without comparison with any writing it shows evidence of forgery as indicated by line quality, suspicious retouching, unnatural hesitation, pen lifts, interrupted and unnatural movements, identity of forms of several disputed signatures, or any evidence of hesitating, drawn writing." (Questioned Documents 2nd Edition by Albert S. Osborn- Sweet & Maxwell)

721. The joint report appears to have altogether overlooked the important matters identified in the quotation.

722. It must be borne in mind that Wang's questioned signatures contain obvious discrepancies from his genuine signatures, especially from his post-1967 genuine signatures. Again, the discrepancies are apparent even to the naked eyes of non-experts.

723. Indeed, Professor Jia himself must have acknowledged such obvious discrepancies. Otherwise, there would be no reason for him to advance in his testimony the theory of "two styles of signature" and the suggested effects of the use of a calligraphic pen.

724. Such obvious discrepancies were simply dismissed in the joint report as insignificant and "natural variations in the writing movements" without elaboration.

725. The blinkeredness of the joint report's approach is further demonstrated by the attempt to justify the tremors in Wang's questioned signatures with the riding accident:

"In accordance with the understanding from the investigation and enquiries, Wang Teh Huei fell from horseback on 10th March 1990 and sustained injuries to his head and arms etc. Such fact confirms, in the opinion of the three appraisal experts, the relationship of cause and effect between trembling in the strokes of the signatures "Wang Teh Huei" on the four sets of Appraisal Materials and the physical injuries sustained by Wang Teh Huei. From such relationship of cause and effect, it can also be confirmed that the time when the signatures "Wang Teh Huei" were written by Wang Teh Huei on the four sets of Appraisal Materials was close to the time when Wang Teh Huei sustained his injuries. The three appraisal experts are of the view that it is highly probable that Wang Teh Huei's signatures on the four sets of Appraisal Materials were written by him after he had sustained his injuries in March 1990."

726. The contents of the joint report fail miserably once subjected to logical analysis and careful scrutiny.

727. In my view, the joint report by the Wife's experts clearly demonstrates that they had assumed the role of a proselytiser instead of an impartial witness to assist the court.

728. The shortcomings of the joint report were clearly recognized in the course of the trial. The discrepancies between Wang's questioned signatures and his genuine signatures, particularly the post-1967 ones, were too blatant to be dismissed as "natural variations" as the joint report attempted to do.

729. In his testimony, Professor Jia sought to alienate himself from the joint report and explain such discrepancies by suggesting that Wang had "two styles of signature" and that it was "Wang's intention to write more artistic signatures with the use of a calligraphic pen". He also sought to show the genuineness of the questioned signatures with his "matching natural variations" theory.

730. Note that there was no reference to any of the above theories in the joint report prepared for the Wife. There was no mention of "two styles of signature" or the effect of the use of a calligraphic pen. The variations were described as differences not indicative of forgery (defensively) but not supportive of genuineness (assertively), and the "matching variations" theory was altogether absent.

731. Such "theories" were not put to Mr. Lesnevich. They were partially revealed in the cross-examination of Mr. Cheng and more significantly revealed in the cross-examination of Mr. Tsui after a break of over 40 days.

732. The "theories" were clearly conceived only after the trial had commenced. Their development gained momentum during the trial and reached its peak when Professor Jia gave his evidence in court.

733. The "two styles of signature" theory is unconvincing. As the judge rightly pointed out, a genuine signature is a reflection of the spontaneous writing habits of an individual.

734. Obviously, one person may have different signatures in different languages (e.g. Chinese and English). Notwithstanding the strangeness of having multiple styles of signature in the same language, even if a person does indeed have such multiple styles, these styles should be significantly different.

735. Professor Jia's suggestion that Wang could have two different styles of and yet similar signatures in the same language was difficult to comprehend and impossible to accept as convincing.

736. Such details aside, it is unclear whether Professor Jia's so-called "theory" makes sense more fundamentally. To explain the difference in two sets of signatures by saying they were of "two different styles" is merely playing with terminology.

737. All in all, Professor Jia's reasons for his opinion do not stand up to logical analysis.

738. A person's signature does not remain the same and normally gets more cursive with passage of time (at least for adults). Such changes may be described as deterioration, change of habit, or loss of ability, although a more neutral term is perhaps "evolution".

739. Old age (with its associating physical conditions) and other forms of physical disability may of course affect the manner and quality of signature.

"It has already been mentioned that the design of most people's signatures is slightly but steadily modified with the passage of the years, a fact which appears to be unknown to many otherwise quite skilful forgers. As this is so, there is always the chance that the document which may be said to 'frame the signature in time' will bear a date or furnish evidence from which it can be dated, which is t variance with the style of the signature it bears. Considerable doubt can be cast upon the authenticity of a suspect signature when it can be shown to be of a pattern which had been discarded by the writer before the date attributed to the document on which it appears, and even less credence will be given to a signature which considerably antedates the style of genuine signature to which it conforms.

The importance of this preliminary investigation cannot be over-estimated, because even the most skilfully executed simulated forgery is not proof against attack on the grounds of style related to date. This line of attack is of the greatest utility when applied to documents which have conveniently come to light some years after they are supposed to have been drawn up. It is of less general application when contemporary signatures on contemporary documents are the subject of investigation." (Suspect Documents by Wilson R. Harrison-Sweet & Maxwell)

"Many writers are not aware of the fact that even a few years, especially with those of advanced age, may make a great change in a signature and one who sets about making a tracing may select a model writing of the wrong date. It is often assumed that any genuine writing is good enough and a model may be taken that was written many years before or after the date of the questioned writing.

In every case of this kind a chronological study should be made of the standard writing; changes should be carefully noted and their application to the investigation in hand carefully considered. The selection of a model of the wrong date may show conclusively that a disputed signatures is not genuine." (Questioned Documents 2nd Edition by Albert S. Osborn-Sweet & Maxwell)

740. On the other hand, despite the virtually infinite ways of writing Chinese characters, there are basic rules of stroke movements and sequences that anyone properly trained in Chinese writing follows. The "evolution" has little impact on such basic stroke movements and sequences.

741. Comparing Wang's 1958 signatures (668 and 669) with his post-1967 signatures, there are clear indications of "evolution". The post-1967 ones are more cursively, and possibly more speedily, written. The stroke movements are more spontaneous; individual strokes are less clearly defined.

742. The last two dots of the radical "Xin" (心) in the character "Teh" (德) are linked up to the "Xie Gou" (斜勾). The two radicals of the character "Huei" (輝) are not linked up with the last stroke of the radical "Kwong" (光) and the last horizontal stroke of "Jun" (軍), to cite just two obvious examples.

743. On the other hand, the structure and alignment of the Wang's genuine signatures remain the same throughout the years (see appendix 3 of the judgment).

744. With the exception of 671 where the character "Teh" (德) was written in simplified Chinese, all of Wang's genuine signatures contain the same components, as well as identical stroke movements and sequences with almost identical shape, alignment, and appearance.

745. In particular, the first horizontal stroke of the character "Wang" (王) in all the signatures is a firm, definite stroke slightly tilting down towards the right. There is almost invariably a significant loop formed by the first vertical stroke and the second horizontal stroke.

746. In 668, the first horizontal stroke of "Wang" (王) appears to tilt upward, but in fact it starts with a downward movement and only tilts upwards after the first half of the stroke.

747. The last vertical stroke of the radical "Jun" (軍) in the character "Huei" (輝) always ends slightly towards left.

748. I do not wish to set out all the examples of commonalities among Wang's genuine signatures. Suffice it to say that there are plenty of such examples provided by the Father's experts.

749. In my view, the suggestion that Wang signed both his questioned signatures and his 1958 signatures in "Lishu (隸書) style and Xingshu (行書) characteristics" is far-fetched. It is also unconvincing to say that his 1958 signatures are of a different style from his post-1967 signatures.

750. The differences between Wang's 1958 signatures and his post-1967 ones are just the result of natural evolution and not, as Professor Jia suggested and Mr. Chang emphasized, his deliberate attempts to sign in two different styles.

751. To justify the significant differences between Wang's questioned signatures and his post-1967 signatures identified by the judge, Mr. Chang relied heavily on the suggestion that Wang signed the questioned signatures in "Lishu (隸書) style with Xingshu (行書) characteristics" and the effect of a calligraphic pen. He complained that the judge had not dealt with the effect of a calligraphic pen at all.

752. When asked which parts of Wang's questioned signatures were of Lishu (隸書) style and which were of Xingshu (行書) styles, Mr. Chang responded that the character "Wang" (王) was Xingshu (行書) with the first horizontal stroke in Lishu style (隸書) with a "silkworm head and wild goose tail" (蠶頭雁尾).

753. The only other parts in Lishu style with Xingshu characteristics were the left radical [the "double men" (彳)] of the character "Teh" (德) and the last horizontal stokes of the radical "Huei" (輝). The latter part again had "silkworm head and wild goose tail" (蠶頭雁尾), according to Mr. Chang.

754. The way Wang wrote the character "Wang" (王) is indeed very common. I find the suggestions about "silkworm head and wild goose tail" (蠶頭雁尾) to be a desperate and ill-reasoned attempt to explain why Wang's questioned 1990 signatures resembled his 1958 signatures and not his post-1967 despite the lapse of over 30 years.

755. In fact, the feature "silkworm head and wild goose tail" only specifies what the similarity was but hardly explains why.

756. The general impression gained from Professor Jia's evidence and Mr. Chang's submissions is that they tried to dress up vacuous and unsound arguments with expert terminology like "Lishu style with Xingshu characteristics" and "matching variations".

757. In fact, there is absolutely no indication that Wang was a calligrapher or had the habit of signing in any classical style, let alone the apparently invented style that counsel termed as "Lishu style with Xingshu characteristics".

758. That the judge did not specifically dealt with the effect of a calligraphic pen does not mean that he had overlooked it.

759. The effect of a calligraphic pen was not a point that was initially relied on. There was no mention of it at all in the joint report and the Father's experts were hardly questioned with particular reference to it. It was just briefly mentioned in conjunction with the effect of the use of a fountain pen.

760. In any event, the judge clearly was familiar with calligraphic pens as he said he used one on a daily basis.

761. In a case of great length and meticulous attention to details such as the present one, it was neither necessarily nor possible for the judge to deal with all the points raised. By necessary implication, the judge must have rejected the argument that the effect of a calligraphic pen explained the significant differences between Wang's questioned signatures and his post-1967 ones. There was a valid basis for such rejection.

762. Since the judge had made adverse findings on the reliability of Professor Jia, it was hardly necessary to address every point made by him.

763. In my view, the use of a calligraphic pen could not have resulted in the significant differences between Wang's questioned signatures and his known post-1967 signatures.

764. It could not cause the first horizontal stroke of the character "Wang" (王) to tilt up to the right to the extent as consistently demonstrated in Wang's questioned signatures. It also could not result in the complete/almost complete closure of the "eyelet" at the connection point of the vertical and second horizontal stoke of the same character.

765. The use of the calligraphic pen could not account for the awkward and unnatural pen-wielding movement of the right radical of the character "Teh" (德) in Wang's questioned signatures.

766. It did not explain the "V" shape formed by the last stroke of the radical "Kwong" (光) of the character "Huei" (輝) or the right-tilt of the last vertical stroke of the radical "Jun" (軍) in the character "Huei" (輝).

767. The reliance on the effect of a calligraphic pen was an afterthought and yet another desperate and ill-reasoned attempt to "explain" the significant differences between Wang's questioned signatures and his post-1967 ones. So was the suggestion of two styles of signatures and Wang's signing "the 1990 documents" in "Lishu style with Xingshu characteristics".

768. In sum, there is no convincing explanation for the ostensible similarities between Wang's questioned 1990 signatures and his 1958 signatures despite the "evolution" and for the obvious differences between Wang's questioned signatures and his post 1967-signatures.

769. The judge was right to reject the Wife's case on the genuineness of Wang's questioned signatures. I, too, would reject it.

770. The joint report was highly selective in choosing the appropriate known signatures of Tse for comparison.

771. In the overall comparison, the joint report used four samples [5996(3) (1999), 736(4) (1990), 736(7)(1990) and G100 (one of the vouchers rejected by the judge)].

772. The comment was simply that they conformed to Tse's questioned signatures in terms of the "level of calligraphy, the few connected strokes, and the overall structure as well as style".

773. There was no reference to the obvious differences in the radical "Chuen" (寸) in the character "Tse" (謝) with the vertical stroke connecting to the dot, the radical "For" (火) in the character "Ping" (炳) when the two dots were written as a horizontal short line or the connecting dots in the two radicals of "For" (火) comprising the character "Yim" (炎).

774. When comparing Tse's questioned signatures with his other known signatures, particularly those of 1990, the differences are even more apparent.

775. In the comparison of the detailed writing features, two known signatures of Tse from 1990 and 1999 were used in connection with each of the characters of "Tse" (謝), "Ping" (炳), and "Yim" (炎).

776. For the character "Tse" (謝), 5995(1), 5996(4) both of 1999 and 736(2), 736(3) both of 1990 were used. The joint report commented abstrusely as follows:

"The features of the character "Tse" in terms of stroke arrangement, proportion, points of connection and appearance among the three parts and in each part of the character conform with the samples."

777. For the character "Ping" (炳), 5996(2), 5996(10) both of 1999 and 736(6), 736(7) both of 1990 were used. Similar abstruse comment was made, namely, "The features of the writing movement of the lateral radical "For" (火) in the character "Ping" (炳), and the features of the stroke arrangement and proportion of the character "Ping" (炳) also conform with the samples.

778. For the character "Yim" (炎), DPH 1829 of 1993, 736(2), 736(6) and 736(7) of 1990 were used. The comment was, "The features of arrangement and proportion of the upper and lower parts of the character 'Yim' (炎) and those of the writing movement and pen-wielding of the character 'For' (火) also conform with the sample."

779. There was no explanation or justification of the criteria for selecting the samples for comparison. There was no mention of the obvious differences even in the selected samples (e.g. the two dots of the radicals "For" (火), present in both "Ping" (炳) and "Yim" (炎), were written as a connecting strokes in most of the samples).

780. Instead, the report identified other differences and tried to explain such differences in the following vague and ultimately meaningless terms:

"These differences are not reflections of different people's handwriting habit, but show the natural variation of one person writing the same stroke at different times and under different condition. When comparing the signatures of "Tse Ping Yim" on the four sets of Appraisal Materials with Tse Ping Yim's carefully and neatly written signatures on the samples, it has been found that the above-mentioned differences are in fact differences not of a fundamental nature but are manifestations of a natural variation in the handwriting features of Tse Ping Yim." (Emphases added)

781. The joint report further identified the presence of tremors in Tse's questioned signatures and sought to show the genuineness of those signatures by pointing out the presence of similar tremors in 5996(4), 5995(1), and 5995(2).

782. What the report failed to mention was that Tse signed 599(4), 5995(1), and 5995(2) in 1999. There was no indication of any tremors in Tse's known signatures in 1990 at all.

783. The joint report sought to justify such suspicion (i.e. that tremors were present in the 1999 genuine signatures and the questioned 1990 signatures but not in the 1990 genuine signatures) by suggesting that the tremors were only found in those signatures "carefully and neatly" written by Tse. As Tse's 1990 signatures were all written quickly, the absence of tremors was explained.

784. What the joint report had again failed to mention was another relatively slowly written signature of Tse in 1979, i.e. 5994, in which there was no indication of any tremors. The tremors in Tse's questioned signatures (supposedly written in 1990) could not be justified by similar tremors in his known signatures of 1999 when he was 69 years old and clearly suffering from serious illness.

785. The joint report concluded that there were many points of conformity between Tse's questioned signatures and his known signatures without identifying them in detail. The joint report again failed to refer to obvious differences.

786. There are indeed obvious differences in Tse's questioned signatures and his known signatures in 1990, although both sets of signatures were supposedly written only about one month apart.

787. Mr. Lesnevich, Mr. Tsui, and Mr. Cheng had identified those differences. The differences are again obvious to the naked eyes of any layperson.

788. Professor Jia, in his testimony sought to explain the differences by suggesting that Tse's wrote the questioned signatures "carefully and neatly" whereas most of his known signatures were written quickly.

789. On the evidence in its entirety, the judge was right to reject Professor Jia's evidence. There were valid bases for his conclusion that Professor Jia and his colleagues were just "hired guns" (although I would not necessarily subscribe to such an unkind labelling), and their arguments illogical and evasive.

790. The judge also had valid grounds to conclude that Tse's questioned signatures were forged by way of superimposing-tracing.

791. Mr. Chang relied on what he termed as "significant matching variations" (or "significant matching natural variations") in the questioned signatures of Wang and Tse with their respective known signatures to demonstrate that the questioned signatures were genuine.

792. The suggestion was that if the questioned signatures were forged, there would not be such significant "matching variations".

793. A "variation" was defined as one component of Chinese characters (e.g. a vertical line, a sabre-stroke) executed in two or more ways.

794. For example, if the "vertical line" in character X were written differently from the "vertical line" in character Y, the difference was a "variation". (X and Y could be instances of the same ideograph or of two different ideographs.)

795. A "matching variation" is a variation present in two groups of signatures. A "significant matching variation" is a matching variation that is not easily noticeable, and hence unlikely to be imitated.

796. According to Mr. Chang, if the two groups of signatures contained a sufficient number of "significant matching variations," then the same person must have written them.

797. In the course of his submission, Mr. Chang produced many charts and diagrams purportedly containing the following particulars:

i. Variations among Wang's questioned signatures;

ii. Variations among his known signatures;

iii. Variations between Wang's questioned signatures and his known 1958 signatures;

iv. Features in Wang's questioned signatures not found his known 1958 signatures but found in his more recent known signatures.

798. Such particulars were identified to demonstrate the following three points:

799. First, that there were "significant matching variations" common to both Wang's questioned signatures and his known signatures. This point was the crust of the "matching variation theory".

800. Second, that Wang's questioned signatures could not have been modelled on his 1958 signatures because of the variations between them. One such variation could be the "vertical line" written differently in character X in one of Wang's questioned signatures and in character Y in one of Wang's 1958 signatures. (Again, X and Y could be instances of the same ideograph or of two different ideographs.)

801. Third, that the questioned signatures were in fact genuine because they contained features also found in Wang's recent known signatures.

802. I am impressed by the ingenuity, diligence, and thoroughness of Mr. Chang, and his assistants as demonstrated by his submission on handwriting. I am equally impressed by Mr. Chan's effort in responding to Mr. Chang's submission.

803. However, it is neither possible nor necessary to repeat their exercises in detail in this judgement.

804. In the course of counsel's submissions, I felt that I was being led to individual trees and branches when the issue was whether two forests were the same.

805. The significance of the particulars identified by Mr. Chang set out in paragraph 797 were neither mentioned in the joint report nor put to all of the experts called by the Father. Clearly they were only conceived in the course of the trial.

806. As Mr. Chang pointed out, human hands are not machines. Therefore there are bound to be variations (natural or accidental or both) even among genuine signatures written by the same person. Similarly, there must also be variations among signatures forged by the same person.

807. The variations that Mr. Chang identified are indeed very common variations in the writing of Chinese characters. Those variations can be accidental or natural, and in my view they are entirely insignificant. They are nothing but different (yet common) ways of executing the same characters. Such common variations do not prove the genuineness of Wang's questioned signatures at all.

808. The suggestion that the variations in Wang's questioned signatures were matched by corresponding variations in his known signatures, including those post-1967 ones, in my view is farfetched and unconvincing.

809. Professor Jia called the downward pen-wielding movement in the first vertical strokes of the character "Wang" (王) and the "Kwong" (光) radical of the character "Huei" (輝) a "significant similarity" between the questioned and known signatures of Wang.

810. I cannot think of any other way of writing the first vertical strokes of the two characters or by someone proficient in writing Chinese. The aforementioned is doubtlessly a similarity, but it is hardly "significant".

811. Another "significant similarity" is the clockwise pen-wielding movement that abbreviates the top horizontal and right vertical strokes in the radical "Si" (四) in the character "Teh"(德). The "connecting stroke" / abbreviation itself is a common feature in fast or cursive Chinese writing. Since horizontal strokes in Chinese are executed from left to right, vertical strokes top-down, there is no reason for the abbreviation not to be executed by a clockwise pen-wielding movement.

812. A third "significant similarity" is the small space between the first horizontal stroke and the first vertical stroke of "Wang" (王). This is simply factually wrong, as this space is present consistently in the questioned signatures, but absent in all save a few of the genuine signatures.

813. Professor Jia further suggested "significant similarities" between the two groups of signature in the structure of the character "Wang" (王) and in the pen-wielding movement and connection strokes in the character "Teh" (德). These suggested similarities are again betrayed by a simple visual comparison of the two groups of signatures.

814. Based on the "significant similarities" like those mentioned above, Professor Jia concluded that the two groups of signatures were by the same author, and hence that the questioned signatures were genuine. This attempt was courageous but in my view unsound and unconvincing.

815. Professor Jia attempted to demonstrate similarities in character structure and alignment between Wang's questioned signatures and his known signatures with a two-part approach, different from the approach adopted by the judge. First, a vertical straight line was drawn along the first vertical stroke of "Wang" (王) through the entire signature. Second, a slanted line was drawn linking the rightmost point of the radical "Si" (四) of the character "Teh" (德) and the rightmost point of the "shelter" of the radical "Jun" (軍).

816. In Professor Jia's approach, the vertical line was meant to show essentially that the different parts of the characters were in same/similar overall layout as their counterparts in the genuine signatures. The angle formed by the vertical line and the slanted line was meant as indication of the consistency in the relative positions of "Teh" (德) and "Huei" (輝) across both sets of signatures.

817. In clear contrast, the approach adopted by the judge required first drawing a rectangle around the four outermost points in the entire signature and then drawing a vertical straight line through the middle of the rectangle.

818. Professor Jia's approach was suspicious in several respects.

819. Most glaringly, it completely ignores the obvious deviations caused by a signature's four outermost points (which the other approach's rectangle took into account). Instead of the spatial axes of the characters, his approach considers their calligraphic axes, whose alignment is a basic point of good handwriting. Any regular vertical Chinese signature is expected to be reasonably aligned in this latter sense.

820. The second slanted line seems to be an arbitrary choice. If another slanted line is drawn between two different points (such as the top-right corner of the "Si" radical (四) and the rightmost point of the last horizontal stroke of the "Jun" radical (軍), one can see conspicuous deviations in the questioned signatures.

821. One can draw a similar slanted line between any of the hundreds of pairs of points in the signatures; whether one particular pair shows consistency means little and certainly des not prove the genuineness of the questioned signatures.

822. In effect, Professor Jia's approach selected only features-trivial and arbitrary features, I might add-that supported the genuineness of the questioned signatures and ignored those that did not. Mr. Chang's term, "structural bias," immediately comes to mind.

823. In fact, even under Professor Jia's scheme, one can see deviations in the questioned signatures. In particular, the vertical line tends to cut through top part of the radical "Kwong"(光) in "Huei" (輝) in the questioned signatures, but this never happens in the genuine signatures.

824. In any event, the approach adopted by the judge, reflected in Appendix 3 of his judgment, could be applied universally to all Wang's genuine signatures and show consistency among them. There was no compelling reason that the same approach, when applied to the questioned signatures, could show such obvious deviations if they were genuine.

825. Even if Professor Jia's approach were sound, neither Professor Jia nor Mr. Chang attempted to explain why the court should favour it in place of the approach adopted by the judge. There was no suggestion that the latter approach was in any way unfair or unreliable.

826. As a non-expert in handwriting or calligraphy but someone familiar with Chinese writing, I have studied each of the characters in the questioned signatures and known signatures of Wang, guided by the principles set out in the leading test books recommended by counsel, opinions of the experts, and counsel's submissions.

827. I am satisfied to the required standard the Wang's questioned signatures in "the 1990 documents" were indeed forged signatures.

828. The evidence supporting such conclusion is strong and cogent. I do not wish to repeat the evidence of the Father's expert witnesses, but only mention a few blatant features that were obvious to naked eyes.

829. The characters "Wang" (王) in all of Wang's questioned signatures have the following special and consistent features:

1. The starting and the ending horizontal strokes all tend to tilt upward to the right in pen-wielding movement;

2. There is a significant gap between the first horizontal stroke and the vertical stroke;

3. The first vertical stroke tends to slant downward to the left before the commencement of the second horizontal stroke forming the Arabic numeral "2" which tilts upward towards the right top corner;

4. The connecting point of the vertical stroke and the Arabic numeral "2" leaves no eyelet; and

5. Generally the character inclines from the right top to the left bottom.

830. Such special and consistent features are absent in Wang's genuine signatures, where:

1. The starting and ending horizontal strokes all tend to be horizontal and often tilt downward to the right in pen-wielding movement; (the only exception appears to be 668 of 1958 which only tilts upward after half of the stroke, which tilts downward is completed);

2. There is no (or no significant) gap between the first horizontal stroke and the vertical stroke in Wang's post 1967 signatures at all. Such significant gap appears in 668 and 669 only.

3. The first vertical stroke is always a straight downward pen-wielding movement. The Arabic numeral "2" tends to be horizontal or tilting down to the right;

4. There is invariably (with 1 or 2 exceptions) an obvious eyelet in the connecting point of the vertical stroke and the Arabic numeral "2".

831. The characters "Teh" (德) in Wang's questioned signatures have the following special features:

1. The starting first horizontal and second vertical strokes of the right radical (+) are oddly executed with unnatural pen-wielding movements. There are variance amongst the questioned signatures with ostensible similarities with 668 and 669;

2. The center part of the right radical, namely the "Si" (四) radical is executed in a more convoluted manner again similar to 668 and 669, but different from Wang's post 1967 signatures;

3. The "Xie Gou" (斜勾) in the bottom radical "Xin" (心) ends with an inward movement curling back to the center of the radical, similar to 668 and 669, but different from the post 1967 signatures; and

4. The last dot of the radical "Xin" (心) takes the form of a short and distinct slanting stroke.

832. The same characters in Wang's post 1967 signatures are different:

1. The starting horizontal stroke moves from left to right in a natural and smooth manner, then turns upward to execute the 2nd vertical stroke forming a distinct loop;

2. The center part of the right radical, the "Si" (四) radical consists of much more abbreviated and cursive pen-wielding movements with all the strokes connecting to one another;

3. The "Xie Gou" (斜勾) in the radical "Xin" (心) moves back out away from the center to link up with the last dot of the radical and the last dot takes the form of a relatively longer slanting stroke; and

4. Generally, the character is written with much quicker pen-wielding movements in one continuous stroke.

833. The characters "Huei" (輝) in Wang's questioned signatures have the following special common features:

1. The left radical "Kwong" （光） is written with one continuous stroke forming a very narrow V shape at the left bottom corner of the radical;

2. The structure of the character is more complicated, similar to 668 and 669 but completely different from Wang's other known signatures; and

3. The last vertical stroke of the radical "Jun" (軍) is a downward stroke pointing to the right.

834. The aforesaid features are different in Wang's genuine signatures:

1. The left radical "Kwong"(光) sometimes consists of more than one stroke and the left bottom corner of the radical form the shape of a U instead of a V;

2. The structure of the character is more simple with a smoother and more natural pen-wielding movement; and

3. The last vertical stroke of the radical "Jun" (軍) is a straight downward stroke with the tip of the stroke pointing to the left.

835. The Father's experts no doubt could identify more differences than the aforesaid. However, in identifying those differences, I have not referred to any of the reports by the experts (at the risk of being accused of "playing expert"). I do not see the need, as those differences are obvious even to untrained eyes.

836. In my view, such differences are not explainable-not by the theory of two styles of signatures, not by the suggested effect of a calligraphic pen, nor by the suggestion that Wang deliberately tried to sign in a particular classical style.

837. The only reasonable and logical explanation is that Wang's questioned signatures were indeed forged signatures and modelled on his 1958 signatures.

838. I do not wish to repeat the exercise of examining Tse's questioned signatures against his genuine ones. There had been more than enough of such exercises by the experts, by the judge, and by counsel both at trial and in the course of this appeal.

839. The "matching variations" identified between Tse's questioned signatures and his known signatures again are not significant. Again, the variations are common variations resulting from common ways of executing the relevant strokes. In my view, they simply do not prove genuineness of Tse's questioned signatures.

840. There are clearly significant differences between Tse's questioned signatures and his known signatures-particularly those executed in 1990, within two months of the alleged execution of the questioned signatures.

841. The Father's experts were justified to conclude that Tse's questioned signatures were forged.

842. There was certainly evidence to support the judge's findings that Tse's questioned signatures were forged by way of superimposing-tracing.

843. The judge carefully evaluated the handwriting evidence on the questioned signatures both macroscopically and microscopically. He considered both different and similar features. He was well aware of the importance of natural variations and significant differences.

844. The judge was also aware of the Wife's case that Wang's questioned signatures were written not in his usual style, but allegedly in a more formal and classical "Lishu (隸書) style with Xingshu (行書) characteristics"; that were deliberately written artistically and with a calligraphic pen. He was also aware of the suggestion of two styles of signatures.

845. The judge never said the variations beyond the acceptable range were conclusive signs of forgery. He, however, was entitled to consider them as indications of forgery.

846. Mr. Chang complained about the absence of comparing "like with like". The complaint was unjustified, because there was simply no genuine post-1967 signature of Wang resembling 668 and 669 or his questioned signatures.

847. There could be no dispute that there are significant differences between Wang's 1958 and 1985 signatures. Professor Jia tried to explain the differences as the result of deliberate attempts to sign in different styles, while the Father's experts attributed them to loss of motor abilities or change of writing habits.

848. However, why did Wang decide to sign "the 1990 documents" in a way that he had not done for 30 years? Why did he use a calligraphic pen although there was no indication that he used one any other time? Why were the questioned signatures so carefully and slowly written on documents so hastily and sloppily prepared?

849. These are pertinent questions and the judge rightly took them into consideration.

850. I see no valid basis for most of the criticism directed at the judge's handling of the handwriting evidence. If any criticism were to be made, it would perhaps be his excessive attention to details. He did not have to entertain Mr. Lee's misleading suggestion that the case depended on handwriting evidence alone.

851. The judge's findings that the questioned signatures of Wang and Tse were forged were fully justified on the evidence before him. I support such findings.

852. I accept, on the other hand, that some of the judge's approaches and reasoning were questionable, as was his extensive reproduction of counsel's submissions.

853. The logic for his conclusion that the Wife authored "D" somewhat defied comprehension. I have also expressed my reservation on his finding, based on the experts' evidence alone, that the Wife authored "A".

854. Some of the differences in the characters that the judge identified might not be as significant as he suggested. He was perhaps too rigid to find that the questioned signatures must have been forged by way simulation of Wang's two known 1958 signatures (668 and 669). After all, there might well have been other signatures of Wang similar to 668 and 669.

855. The judgment may not be perfect; no judgment is. However, the judge's mistakes and the shortcomings of his judgement do not enhance the probability of the Wife's case. They do not prove the genuineness of the questioned signatures or the due execution of "the 1990 documents".

856. That the judge misunderstood and/or misinterpreted some of the evidence does not imply that there was no evidence to establish the falsity of the questioned signatures. That the judge had overstated the Father's case does not mean all his findings had to be set aside.

857. The extent of the attacks on the judge and the time spent on such attacks in this appeal are unprecedented, at least in my experience.

858. But the judge was not on trial. He was not the issue.

859. The issues were:

i. Whether the judge was right in refusing the Wife's application to propound "the 1990 documents" as the last will of Wang;

ii. Whether the Wife was able to satisfy the court's conscience that Wang did execute "the 1990 documents";

iii. Whether there was strong and cogent evidence to show that the questioned signatures of Wang and Tse were forged signatures.

860. In my view, on the evidence, no reasonable tribunal could be satisfied that Wang had duly executed "the 1990 documents". No reasonable tribunal could feel save to pronounce them, in solemn form, as the last will of Wang.

861. In my view, the experts' evidence was strong and cogent enough to support the findings that the questioned signatures of Wang and Tse were forged.

862. "On an appeal against a judgment of a judge sitting alone, the Court of Appeal will not set aside the judgment unless the appellant satisfies the Court that the judge was wrong and that his decision ought to have been the other way" (per Lord Sankey LC in Powell v Streatham Manor Nursing Home [1935] AC 243 at p.249). The Wife failed to persuade me that the judge was wrong in refusing to grant probate of Wang's estate to her.

863. In my view, the judge was right in refusing to propound "the 1990 documents". The judge was right in granting the decree of probate for "the 1968 Will" in solemn form in favour of the Father. The judge was also right in finding that the questioned signatures of Wang and Tse were forged signatures.

864. I would dismiss the main appeal.

Hon Yuen JA:

Main issue

865. Before starting this judgment, I have to say, with respect to counsel appearing below, that the trial had been bedevilled by far too many peripheral issues. There was and is only one main issue in this case: viz. were the documents being propounded signed by Wang Teh-huei ("the Deceased") as his will?

866. In respect of this main issue, none of the witnesses of primary fact gave oral evidence at the trial.

- The persons who wrote the text of the testamentary documents being propounded were never identified.

- Tse Ping-yim ("Tse"), who asserted that he had witnessed the Deceased signing the documents, had died by the time of trial.

- The identity of the other purported witness Lee K__ S__g whose purported signature appears on the document entitled "Will", was never positively asserted, and no attempt was made to trace the purported witness.

- The Deceased's wife ("the Wife"), from whose custody the will being propounded came, did not give oral evidence at the trial.

867. Although there were persons - including the Deceased's father ("the Father"), the Wife's brother Dr Kung, and the Wife's solicitors Alex F.V. Lam and Allan C.Y. Leung - who gave evidence of facts which were relevant (in varying degrees) to the main issue, the evidence they gave was not on the primary factual issue.

868. As for the expert witnesses, the strength or weakness of their evidence does not depend on an assessment of their performance in the witness box.

869. Accordingly, notwithstanding the length of the trial and the number of witnesses called, this court is in as good a position as the trial judge to evaluate the evidence given at trial and to draw inferences from it.

870. Before I deal with the evidence however, I shall first consider the legal issues.

Burden of proof

871. It is clear that the legal onus of proving a will is on the party propounding it, in this case, the Wife. It is also indisputable that as the Father was alleging that the signatures of the Deceased and Tse were forged, he had the evidential burden of proving those two allegations (Williams Mortimer and Sunnucks, Executors, Administrators and Probate 18th ed par. 39-11).

Standard of proof

872. The standard required of the Wife, to prove the will, is on a balance of probabilities. The standard required of the Father, to prove that the signatures of the Deceased and Tse were forged, is also on a balance of probabilities, but by reason of the seriousness of the allegations, "the more improbable the event, the stronger must be the evidence that it did occur before, on the balance of probabilities, its occurrence will be established" (Re H [1996] AC 563, 586-7).

Effect of evidence on parties' cases

873. The court must of course consider all the evidence before determining whether it can pronounce in solemn form for a will being propounded. The same approach applies when the court is considering the evidence adduced by the Father in seeking to prove forgery. Some of the evidence pressed upon the court, whether it be by the Wife in trying to prove the will being propounded, or by the Father in trying to prove forgery, may be common, but the parties' cases are not opposite sides of the same coin.

874. If the Father succeeds in proving that the signature of the Deceased has been forged, then obviously the will has not been proved. However, even if the evidence is not strong enough for the Father to prove forgery, the evidence he has adduced towards that end, together with other evidence, may still instill such doubts in the court's mind that the court finds that it cannot grant probate of the will being propounded.

Formal validity

875. For probate to be granted, a will must satisfy the requirements of both formal and substantial validity. Formal validity depends on compliance with statutory requirements. Substantial validity depends on matters such as testamentary capacity and knowledge and approval (Williams Mortimer and Sunnucks par. 12-01).

876. In a typical Wills Act or Wills Ordinance case, the requirements of formal validity are basically:

(a) that the will is in writing and signed by the testator, and

(b) it appears that the testator intended by his signature to give effect to the will;

(c) the will is signed by the testator in the presence of at least two witnesses present at the same time;

(d) each witness attests and signs the will in the testator's presence but not necessarily in the presence of each other

(section 5(1) Wills Ordinance cap 30).

877. In Hong Kong, however, part of these statutory requirements for formal validity are relaxed for Chinese testators. Under s. 5(2) of the Wills Ordinance applicable at the material time, any will of a Chinese testator written wholly or substantially in Chinese and signed by the testator "shall be valid and duly executed" although not executed in accordance with the rules set out in s.5(1).

Due execution

878. It has been suggested that there is no requirement for "due execution", but only for "execution", for Chinese wills. This contention is relevant to, amongst other things, the issue whether the presumption of due execution can apply to Chinese wills.

879. As the language of s.5(2) Wills Ordinance shows, "due execution" is still pertinent. The requirements for formal validity are still that (a) the will is in writing and signed by the testator and (b) it appears that the testator intended by his signature to give effect to the will. It is simply that where the testator is Chinese and where the will is written wholly or substantially in Chinese, due execution is established without the need to comply with the requirements for witnesses set out as (c) and (d) in paragraph 876 above.

Presumption of omnia rite esse acta

880. How the onus (of proving that a will has been duly executed) can be discharged must depend on the individual circumstances of each case.

881. In seeking to discharge this onus, the presumption omnia rite esse acta is often called in aid by the party seeking to prove the will. This is a rebuttable presumption of law that all things are presumed to have been done in the right way. Where an act appears on the face of a document to have been done, one assumes that that act has been done in the right way in the absence of conflicting evidence.

882. In the context of proving the due execution of wills, this presumption serves a useful function. In many cases, it may be impossible to have actual evidence of the primary facts of due execution by the time a will falls to be proved. This is because the maker of the will being propounded would obviously be dead and the witnesses may be unavailable. Then, if there is proof that the signatures appear to be those of the testator and the witnesses, the court will presume due execution, in the absence of conflicting evidence (Williams Mortimer and Sunnucks par. 12-26). Even where the witnesses are available, confirm their signatures on the will and can recall having seen the testator sign it, they may, due to passage of time, be unable to recall certain acts which are necessary for validity under the statutory requirements (eg whether both were present at the same time when the testator signed the will).

883. In some cases, the application of the presumption may be pivotal in discharging the onus of proving that a will has been duly executed. In other cases, it may be of little use at all. In all cases, what the presumption can do - what assistance it can provide the party seeking to propound the will - towards discharging the onus of proving the will obviously depends on what other evidence there is in the case and the nature of that evidence.

Applicability of presumption to Chinese wills

884. It has been contended that the presumption cannot apply at all to a case involving a Chinese will because of the absence of the statutory requirement of attesting witnesses. I do not agree with that contention. Of course in a typical Wills Act case, the presence of an attestation clause signed by two witnesses would demonstrate a wish on the part of the maker of the document to comply with the statutory requirements for due execution. Clearly the presumption would then apply with some force. But it does not follow from that that the presumption can only apply in, or is parasitic upon, a situation where there were witnesses to the testamentary document.

885. The presumption is a rebuttable presumption of law. As such, it is a legal rule which is applicable by the court to acts and documents of all kinds, unless its applicability has been expressly or impliedly excluded. There is no express exclusion of its applicability to Chinese wills in the Wills Ordinance. If anything, its applicability is affirmed by the proviso to rule 10 of the Non-Contentious Probate Rules cap. 10:

"(1) Where a will contains no attestation clause or the attestation clause is insufficient or where it appears to the Registrar that there is some doubt about the due execution of the will, he shall, before admitting it to proof, require an affidavit as to due execution from one or more of the attesting witnesses or, if no attesting witness is conveniently available, from any other person who was present at the time the will was executed:

Provided always that if a will is in Chinese characters and appears to be satisfactorily executed by the testator the Registrar may assume without further inquiry that the will has been properly executed".

886. Nor is there a sufficient case for an implied exclusion. If the legislature had seen fit to relax some formal requirements, it is not for the court to hinder this objective by excluding the applicability of the presumption in all cases, however simple and non-contentious.

Force of presumption in present case

887. The real issue on the presumption is this - given the other evidence in this case, to what extent can the presumption go towards discharging the Wife's onus of proving that the Deceased signed the will being propounded?

888. In my judgment, the conflicting evidence in this case (which will be discussed in detail later in this judgment) is such that the presumption applies with much less force that it might do in other cases, and certainly does not discharge the Wife's burden of proving the will being propounded.

Evidence of "suspicious circumstances"

889. This leads me to the issue of evidence of "suspicious circumstances". Generally speaking, where it has been proved that the deceased had signed a will and had testamentary capacity at the time of execution, it may be assumed that the deceased had knowledge and approval of the contents of the will, which is one of the requirements of substantial validity (Barry v Butlin (1838) 2 Moo. PC 480, 484).

890. However, where there is evidence of circumstances which excite the vigilance and suspicion of the court, the party seeking to propound the will must prove positively that the testator knew and approved the contents of the will. In other words, he cannot rest on the general assumption referred to in the previous paragraph, and he must dispel the suspicions arising from circumstances such as where the will has been written or prepared by the party who takes a substantial benefit from it (Barry v Butlin, Wintle v Nye [1959] 1 All ER 552), or where he suggests the terms to the testator, chooses a solicitor and supervises the execution of the will by the testator (Williams Mortimer and Sunnucks par. 13-26).

891. There are authorities to the effect that there is no exhaustive list of such instances (Tyrrell v Painton (1894) P 151, 157, In the estate of Fuld, dec'd (No.3) [1968] P 675, 712), but whether or not that is so does not fall to be determined in the present case.

892. That is because one thing is indisputable, and that is that Order 76 rule 9(3) Rules of the High Court requires any party who wishes to assert that at the time a will was alleged to have been executed, that the testator did not know and approve of its contents must "specify in his pleadings the nature of the case on which he intends to rely".

893. That was not done in this case. It was not until final submissions that leading counsel for the Father sought to amend the pleadings to include a plea that the Deceased did not know and approve of the contents of the will being propounded. The trial judge rejected the application for amendment. There is no appeal from that decision.

894. That being the case, the body of case law relating to "suspicious circumstances" in the context of want of knowledge and approval is not, in my view, relevant.

895. However, the body of evidence which might be termed "suspicious circumstances" in a "want of knowledge and approval" case may still be relevant. I do not agree with the trial judge that it was for the Wife to dispel the "suspicious circumstances" such that failure to do so would prove forgery. But the body of evidence is still relevant to the issue of due execution as "all collateral circumstances" have to be considered (Saph v Atkinson (1822) 1 Add 162, 163). The question is: tested against that evidence, has the Wife discharged the burden of proving that the Deceased had signed the propounded documents as his will?

Approach to evidence in general

896. As in all cases, the court should look at all the evidence adduced by the parties as a whole. In the context of this case, that would include both the non-expert and the expert evidence. Insofar as the trial judge dealt with the expert evidence on forgery and ink-dating separately, that was an attempt to deal with it in a purely technical way. Given the rest of his findings against the Wife, that methodology cannot be said to be unfair to her. It would be recalled that applying his methodology, the trial judge rejected expert evidence on ink-dating adduced by the Father.

897. This case is no different from all cases which come before the court. As always, the court should look at all the admissible evidence, considering their relevance and reliability. There have been many attempts in the course of this appeal by both sides to ask the court to speculate why "the Deceased" or "the forger" (depending on whether one is advancing one case or the other) should (or would not) have done this or that act. In my view, the court should resist invitations to indulge in speculation where there is no evidence in support of a contention. There may be all kinds of reasons (whether personal or otherwise), not known to the court, why a person did (or did not do) this or that act, which acts may have been done (or not done) under all kinds of contingencies or circumstances.

898. The court's only function is to look at the evidence established before it and make only reasonable inferences from that evidence. Beyond that, it should not go.

Criticism of trial judge

899. Before I deal with the evidence however, I have to deal with one of the grounds of appeal raised by the Wife. It is to the effect that the judge had failed to bring an independent mind to bear on the case and it is said that this is shown by the fact that he had copied substantial parts of the submissions of counsel for the Father.

900. It is clear to anyone comparing the judgment and the submissions that the trial judge did borrow substantially from them when writing the judgment. However it does not necessarily follow from that that the trial judge had failed to bring an independent, fair and unbiased mind to bear on the case.

901. It would appear from the way the action was run by both sides that the trial judge had a difficult trial on his hands. The pleadings were sketchy, both sides played their cards close to their chest, there was an abundance of evidence which came in bits and pieces and under different conditions, and concessions which were retracted. The evidence on virtually every point, as well as every non-point, was pursued to extremes by both sides.

902. After a trial which overran substantially, it was probably in the hope of producing a judgment speedily that the trial judge extracted parts of the submissions for his judgment. I have to say that this is unusual and undesirable, but I am not persuaded that it meant that no independent thought had been exercised. It is clear from the transcript that the trial judge followed the case closely, he probed the Father's witnesses as hard as the Wife's, he rejected the Father's case on ink-dating, he refused the Father's application to amend to plead want of knowledge and approval and he made points of his own in the judgment that were not in the submissions. I do not therefore consider that this ground of appeal has been made out.

Both Father and Wife natural legatees

903. With that, I turn to the substance of the case which is discussed below in a general chronological order. To start, the evidence shows that neither the Father nor the Wife would be a surprising legatee. In his 1960 will, the Deceased had made both his wife and his father legatees in equal shares. The Wife was a natural legatee by reason of the bond of marriage. The Father was also a natural legatee because the evidence showed that not only was he personally close to his father, his business had developed from his father's business. It is an indication of the Father's position in the Deceased's eyes that the Deceased provided in his will that any children he had would only take from the Wife's share of the estate.

904. The change made by the 1968 will, in cutting off the Wife, was understandable because the Deceased was angry with her for what he perceived to be her unfaithfulness. He knew by then that he would be childless.

905. Although even as late as 1974, the Deceased was still concerned about the Wife's faithfulness to him, that was 16 years before the date of the will being propounded. By 1990, the couple was in their mid-50's.

906. The evidence was that by then the Deceased was close to his wife and he remained close to his parents. As far as the Wife was concerned, the evidence from the couple's friends and godson was that they displayed their affections openly. There was some evidence that the couple had been looking into trust arrangements on and off, with a view to minimizing estate duty and that each might have been intending to benefit the other. By this time, the Wife had taken a more active role in developing the business. On the other hand, this also meant that she had already become very wealthy in her own right.

907. As far as the Father was concerned, the evidence was that the Deceased remained close to his father and indisputably fond of his mother. The evidence also showed that the Deceased enjoyed cordial relationships with his siblings who had settled in the United States after graduation, going on holidays together with them. He was also proud of the achievements of their children. He had shown generosity to them, giving them gifts e.g. for winning a place at medical school. Benefiting the Father, the patriarch of the Wang family, would preserve part of the Deceased's business empire for the family from which it had started.

908. The evidence thus shows that it would have been as natural for the Deceased to benefit the Wife as to benefit the Father, as both were obviously close to him. The court should not have been asked to determine what was "the only rational decision" that the Deceased could make to favour one or the other.

909. In any event, the issue is not what the Deceased might have decided to do if he had been asked (prior to his disappearance in unexpected circumstances in 1990) what testamentary arrangements he wished to make, but whether in fact he made them in the will being propounded.

Retention of 1968 will in safe deposit box

910. There were lengthy submissions from both parties on the retention of the 1968 will in the Deceased's safe deposit box and whether the Deceased might have forgotten about it. It is unlikely that the Deceased would have forgotten about the 1968 will, having been reminded of it when Deacons offered to return the copy in 1985. However, the retention of that will in the safe deposit box is in itself of little value in determining the question whether the Deceased had made the will being propounded, because it is generally understood that when a new will is executed, a former will is revoked.

911. As for the photographs of the girlfriends in the same safe deposit box, they might cause some discomfort to a spouse but other than that, it is difficult to see how this evidence could be of much weight in considering the question whether the Deceased had made the will being propounded.

Unlikelihood of intention to dispose of estate in manner inviting doubts and delay

912. The will being propounded purports to be a home-made Chinese will, document "A" being written in an apparently hasty hand with a defective pen. The question is whether, on the evidence, it is likely that this testator would have chosen to dispose of his very large estate in this fashion. The answer to this question is, in my judgment, of significance to the issue of the proving of the will being propounded.

913. First, it is obvious that a person who is making a will to dispose of his estate would not wish to do it in such a way as to invite doubt and dispute which may arise from the testamentary document having been prepared in a way otherwise than would be expected of him. This would particularly be so for a businessman with a large business empire, who would be expected to take all reasonable steps to avoid delays in the administration of his estate. Of course, if a person was about to go down in a plane crash, he would have had no choice but to scribble a will hurriedly in the exigencies of the situation, but that was not the situation here.

914. Secondly, the Deceased had in fact demonstrated his choice to make wills in the conventional way twice before, at times when his estate was not as large as it had become.

915. It was suggested on behalf of the Wife that perhaps by 1990, the Deceased had become so successful that he had more confidence in himself than when he made the first two wills and might have thought that he did not need lawyers. However, the more successful the businessman, the greater would be his need to ensure that there should be no unnecessary delays and the greater the understanding of the need for the safeguards of formal documents. There was no evidence that indicated that the Deceased had ever acted otherwise.

916. Monday 12 March 1990, the date of the will being propounded, was a working day. According to Tse's statutory declarations, it was only late afternoon when the Deceased allegedly signed the documents being propounded. It is unlikely that solicitors were unavailable to the Deceased.

917. Insofar as it was suggested that the Deceased might have no longer trusted solicitors after his litigation with Donald Cheung, there was no evidence to support this suggestion. In fact, there is affirmation evidence from the Wife the other way, when she alleged that Ronald Arculli, a partner of a well-known law firm, was one of the Deceased's oldest and closest friends and often advised them on their personal matters (Nina TH Wang, 2nd affirmation, 15 July 1999 par. 29-30, HCMP 3454/99).

918. In any event, whichever solicitor the Deceased might have chosen to draft and attest a will would have been bound by the profession's code of conduct to maintain confidentiality. That safeguard would have been known to the experienced businessman that the Deceased was. It was therefore unlikely that he would have chosen as the authors of his will not just one, but two persons, whose identities (according to the Wife's case) are unknown even to her, and therefore not someone close to the Deceased, whether in his personal or business life. If it is suggested that the Deceased did not even trust any one of the solicitors that he had instructed in the course of his business dealings, then there would have been less reason for him to have trusted a stranger or casual acquaintance, such as a nurse at the hospital (as suggested to the court), who owed him no obligations, legal or moral.

919. Thirdly, it has been suggested that the will being propounded may have been the result of sudden "post-mortem" thoughts on the part of the Deceased who had had a riding accident a few days before. However, the evidence does not support that theory. The Deceased had felt confident enough to discharge himself from hospital instead of staying for further observation, and according to Tse's statutory declarations, was eager to return to his physical exercise routines. The evidence does not suggest that the Deceased was having prescient thoughts about his mortality.

920. Further, in the nearly one month between the date of the will being propounded and the Deceased's kidnap, he took no steps to follow-up on the various schemes for estate duty avoidance that various agents had tried to interest him in.

921. Finally, there was no reason why, even if the Deceased did have "post-mortem" thoughts, he did not instruct lawyers to draft a will in the conventional way in the period of nearly one month after the date of the will being propounded.

Surprising features of propounded documents

922. The testamentary documents being propounded contain certain features unconventional in form and surprising in substance. These features are surprising for a prudent and experienced businessman as the Deceased was. It may be said that they would also be surprising if prepared by a forger setting out to fabricate a will. However, as discussed in paragraph 897 above, it is not for the court to try to speculate why a forger would or would not have done something, as the court has no evidence of the relevant circumstances at the time of commission or omission. The court's function is to see whether the Deceased would or would not have done something, on the evidence concerning him as adduced before the court.

923. It was accepted by counsel for the Wife that the authenticity of these four documents stood or fell together. It could hardly be otherwise, as Tse had stated in his statutory declarations that all four were signed by the Deceased and by himself at the same time. Further, the Wife's case was that they were given to her by the Deceased at the same time.

924. The unconventional or surprising features of the four documents were as follows:

"A" - in this document entitled "Will", there is no appointment of a personal representative, contrary to the previous two wills made by the Deceased; further, this document was purportedly signed by Tse and Lee K__ S__g under the characters for "Witness(es)", but these signatures appeared at the start of the document, before the text, and before the purported signature of the Deceased, an unconventional position for witnesses to wills as the Deceased would have seen from the previous two wills that he had made;

"B" - in this document which contained an appointment of the Wife to administer the estate, Tse's purported signature appears but not that of Lee K__ S__g; the Deceased's parents are described as being "disappointing" but no reasons are given for that criticism of them;

"C" - in this document, also signed purportedly by Tse but not by Lee K__ S__g, the Wife is prohibited from giving any money, interest or property to the Wang family because they were "all disappointing", nor to the Kung family as they were "disgusting", but again no reasons are given for those criticisms;

"D" - although this document would not be a testamentary document, is of no possible formal significance, and the contents are of a purely personal and indeed intimate nature, it was also purportedly signed by Tse for no discernible reason.

925. There was no evidence why a prudent and experienced businessman like the Deceased would have wished to create by way of a "home-made" will in this form doubts and delays for a will by which a large estate was to devolve.

Unsatisfactory evidence of "holiday wills"

926. The Wife sought to establish, through the evidence of her brother Dr Kung, that the Deceased had made "holiday wills" before the date of the will being propounded.

927. There was no evidence about the form of these "holiday wills". It was simply assumed that they had not been prepared by solicitors and so would have been informal documents prepared either by the Deceased himself or with the help of others. There was no evidence whether they were Chinese wills. If they were not, there would still have to be witnesses to the Deceased's execution, but there was no evidence of any attempts to trace any such witnesses.

928. The evidence that was adduced was most unsatisfactory. First, although Dr Kung claimed to have remembered the "holiday wills" and reminded the Wife about them at the end of 1997, they were not mentioned at all in his witness statement made in January 2001. Nor were they mentioned in the Wife's affirmation of testamentary scripts. In the present case, where no stone was left unturned, there was no satisfactory explanation how this evidence about these "holiday wills" was omitted.

929. There was also the expectation of evidence from Ng Shung-mo and Andrew Wong on the alleged "holiday wills". In the event, neither of them was called.

930. Further, Dr Kung's evidence was that the envelopes he was given contained the wills of the Deceased and the Wife. In that case, the Wife would have had to liaise with the Deceased for them both to have their respective wills ready for delivery to Dr Kung. However there was no evidence at all from the Wife in support of Dr Kung's evidence. That obviously detracts from her case on this issue.

931. Mr Thomas SC, leading counsel for the Wife on appeal, accepted that not all the evidence that could have been called on this issue had been called.

932. In my view, that leaves the evidence on this issue with hardly any value. Put into the balance against the evidence that the Deceased had previously made two wills which were drafted and attested by solicitors, the balance is clearly heavily against the likelihood of the Deceased having chosen to dispose of his very large estate in a way that would arouse doubt and dispute, and engender delay in its administration.

933. In my judgment, given the evidence discussed above, the form of the will being propounded is a weighty factor against the likelihood that it had been the Deceased's instrument.

Authorship of "Will" (document "A") - not proven to be the Wife

934. As for the authorship of the text of the will being propounded, this is of course relevant to the proof of execution (Saph v Atkinson, 58). The further significance of the point would have been that if the Wife is proved to have written the will being propounded, then the issue of want of knowledge and approval would also have been engaged if it had been pleaded.

935. At the trial, the Deceased's sister Teresa Tak-shyan Sun ("the Sister") gave evidence which was taken de bene esse that the text of the "Will" (document "A") was in the Wife's handwriting. She was not cross-examined on this by the Wife's leading counsel, who objected to admissibility.

936. The trial judge ruled at the end of the case that the evidence was admissible and found that it proved that the Wife had written the text of the "Will". I do not consider that it was helpful to defer the question of admissibility to the end of the case when there were some preliminary disputed facts on which the question of admissibility depended (Phipson on Evidence 15th ed par. 1-24). In any event, the evidence was in my view insufficient to prove that the Wife had written the text of document "A". The Sister's evidence was that she recognized the Wife's handwriting from calligraphy lessons which took place over 30 years ago and from shopping lists scribbled during occasional holidays thereafter. The less recent and less frequent the occasions, and the less likely that attention would have been paid to the handwriting, the less strong the evidence of recognition would be (Phipson on Evidence par.37-73).

937. There were produced by the Sister only two short pieces of Chinese handwriting viz. the "Fotomax" envelope and the document pertaining to the power of attorney, written on the face of it by the Wife.

938. However, these two documents did not produce much evidence to go by. The Government handwriting expert Patrick Cheng could neither eliminate nor confirm that the handwriting on the text of the "Will" (document "A") was that of the Wife. The Father's expert David Tsui's evidence was that on a scale of -10 to +10, he would only put the possibility of the Wife having written the document at a level of +2. In the face of this expert evidence, and as the Sister's sympathies obviously lay with the Father and may have clouded her opinions, it was unsafe to find the allegation that the Wife had written the text of the "Will" (document "A") proved.

939. As a matter of completeness, I would add here that the trial judge's finding that the Wife wrote document "D" was unjustified on the evidence before him. The word "Nina" written at the end of the postulated document "E" may or may not have been a signature. Even if it were, there was no evidence of who signed it. In any event, since document "D" does not have any testamentary relevance, no further time need be spent on it.

No attempt to trace authors of documents

940. However what in my view is important about the authorship of the testamentary documents is that they were not written by the Deceased but by at least two authors (one for document "A" and the other for documents "B" and "C"), possibly at the same place because the paper was similar.

941. Since the Deceased would (whether directly or indirectly) have had to instruct these persons to write the text of the documents being propounded, they would have been important witnesses for the party seeking to prove the will. However it would appear that no attempt was made to trace these persons.

942. It must be remembered that the burden is on the party seeking to prove a will in solemn form. Where one purported witness (whose evidence has been untested) is dead and the other witness's identity is unknown, the court would expect the party seeking to propound the will, especially one made in such an unconventional form, to seek out the persons "through whose agency the will was prepared" (Saph v Atkinson, 57-8). There was no evidence of attempts to do so.

Criticism of parents and siblings contradictory to evidence

943. I now come to the contents of the documents being propounded. I do not consider that the lack of a revocation clause is relevant, as it is generally understood that a will made later in time would revoke a former will.

944. In my judgment, the significance of the contents of the documents being propounded lies not so much in that the Father was disinherited, but that in light of the evidence that the Deceased was close to his parents and fond of his siblings, the documents contained hurtful statements as to how they have so "disappointed" him as to cause him to expressly prohibit the Wife from giving them anything from his estate.

945. The sentiments purportedly expressed were contradictory to the evidence. The evidence was that the Deceased and the Father had been particularly close. The attempts during trial to show that the Deceased harboured some secret resentment against the Father because the latter had been unable to finance the former's further education at university in America backfired when the evidence from the Deceased's school showed that the reason for his not going on to university was because he had been expelled from school. The Father kept the fact of the expulsion to himself and took the Deceased under his wing, giving him a job in his business from which the Deceased flourished. Further, the attempt to show that the Deceased may have been "disappointed" with his father because the latter had a long-standing relationship with a woman was also contradicted by evidence that the Deceased had long accepted the relationship and saw the woman socially. The Deceased's concern for his father was shown in evidence of the minutiae of everyday life given at trial; there was no material evidence the other way.

946. As far as the mother was concerned, the undisputed evidence at the trial was that he was very fond of her. There was unchallenged evidence that the Deceased would personally drive to her home in Kowloon to pick her up for dinner, and would ask her to stay at his home from time to time, the last occasion being shortly before his kidnap in April 1990. In the light of this unchallenged evidence, Mr Thomas SC accepted that it was "surprising" that the Deceased would have said that he was "disappointed" with his mother.

947. The references (in the documents being propounded) to the Deceased's parents having "disappointed" him was thus clearly contradictory to the evidence, and it is on the evidence (not speculation) that the court should make its determination on these documents.

948. Whilst document "B" contained a request to the Wife to take care of the parents and a sister who was mentally ill, document "C" expressly prohibited her from giving anything to the Wang family, because they were also said to be "disappointing". Not only was this sentiment contradictory to the evidence as far as the parents were concerned, it was also contradictory to the evidence in relation to other members of the Wang family.

949. The evidence was that the Deceased's relationship with his siblings was cordial. The evidence disclosed they went on holidays together, his making of cash gifts to them for no particular occasion and his recognition of their children's achievements. There was no evidence of any events which might have caused him to feel disappointed with them, let alone any evidence of actual expressions of disappointment.

950. The question is not so much whether the Deceased might or might not have bequeathed any part of his estate to his siblings or their children. That is one thing. It is another question whether the Deceased would have been hurtful by criticizing them as being "disappointing" when that was contradictory to the evidence of a lifetime's cordial relationship. It is so extraordinary that it serves as a serious warning to the court to be cautious when examining the issue whether the documents were really signed by the Deceased.

Witnesses

951. As the will being propounded purports to be a Chinese will, there is no requirement in law for witnesses to sign it, but the fact is that these documents purport to bear Tse's (and Lee K__ S__g's) signatures. Therefore, if these persons were available, they would have been the most relevant witnesses in respect of the issue whether the document being propounded was signed by the Deceased as his will.

952. If the court accepts that Tse actually saw the Deceased sign the will being propounded, that would (on the case before the court) prove that it was his will, there being no dispute as to testamentary capacity, or want of knowledge and approval. If the court rejects the veracity of Tse's account, that does not prove that the Deceased did not sign the documents, but his false statements would have to be taken into account when the court considers whether it would be safe to grant probate in these circumstances.

Consequence of Tse's death to case

953. In the present case, the most important piece of evidence for the Wife was in my view Tse's statements. The greatest difficulty with the case was caused by his death. Tse's death did not make his statements any more or less credible, but its significant consequence to the case lies in their being untested by cross-examination.

Alex FV Lam's evidence

954. The court has been left with only Tse's two statements verified by way of statutory declarations. Before I consider the first statement, I should first deal with the trial judge's rejection of the testimony of Alex F.V. Lam ("Lam"), the solicitor before whom Tse made his first statement, which rejection was in my view unjustified by the evidence.

955. It was not disputed that Lam was at that stage not involved with any of the Wife's court cases. There was no evidence to contradict his evidence that he had simply been told at short notice towards the end of a working day by Joseph Poon ("Poon"), his senior partner and a consultant of the Wife (though not the solicitors representing her in the probate litigation), to take a statement from Tse in Chinachem's office.

956. Whether Lam did or did not bring with him the photocopy documents he had been shown by Poon is not important. Lam's evidence was that the Wife gave him photocopies of the same documents when he arrived at Chinachem's office (Transcript, Day 150 p.45). If she had not done so, then he might have had to get other copies, but in the event, that was not necessary. There was no basis, on this evidence, for deciding that Lam had something to hide in failing to produce the photocopies he had been shown by Poon, documents that were kept, in any event, not by him but by Poon (Transcript, Day 150 p.78).

957. Given that Lam was only filling in for Poon in a matter that he did not know much about, it was understandable that he did not make an attendance note or draw up a bill after the event. Nor was it suggested that he was in a firm where great importance was placed on filling in time sheets.

958. Further, the fact that Lam signed to attest the statutory declaration could as much be due to the absence of other solicitors or unfamiliarity with the rules of practice as to something sinister.

959. Nor can Lam be criticized for only having asked Tse four questions: (i) when he started to work at Chinachem; (ii) what responsibilities he had; (iii) how the documents came about; and (iv) how his signatures came to be on the documents.

960. Those were exactly the sort of questions someone in Lam's position would reasonably be expected to ask in the circumstances. He was not there to interrogate Tse as one would a hostile witness. He would not have been in a position to do so in any event, given his lack of knowledge and preparation. Nor was he to know that Tse would be unavailable in the future.

961. Accordingly, although I am aware that the trial judge was not impressed with Lam's demeanour while giving evidence, where there was no real challenge to the evidence he gave, it should not have been rejected on the basis of unreliability.

962. The trial judge's finding that the statement was "pre-prepared", meaning that it had already been written prior to Lam's meeting Tse, was based on tenuous evidence regarding the writing surface (judgment par. 5.88) and the question why the paper showed deep indentations was not directly put to Lam. In my view, evidence of that quality did not justify what was tantamount to a finding that Lam (a professional person with no relationship to any of the parties) had committed perjury.

Unexpected contents of Tse's first statement

963. Coming then to the contents of Tse's first statement, the first point that immediately strikes a reader is the fact that it is very well-structured and composed. Even though the basic structure could be explained by Tse following Lam's four questions (par. 959 above), and even allowing for the time taken to write it and care noticeably taken by a person who has obviously practised calligraphy, it remains a remarkable work for an unscripted first draft. This is the more so given Lam's evidence that Tse refused his offer to write the statement for him and did not need to ask for help at any stage when he was writing the statement (Transcript, Day 150 p.48-9).

964. What however was more remarkable was the fact that in his statement, Tse conspicuously volunteered information in a passage about how he was in a position to often see the Deceased and the Wife behaving in a loving way towards each other. 我在王德輝家中工作時間不短。感到王德輝先生夫婦對我像一家人一樣。我經常見到他們夫婦非常恩愛，出雙入對，時有拖手走路。 Translation: "I have been working in the household of Mr Wang Teh Huei for quite a long period of time and feel that Mr and Mrs Wang treated me like their family members. I always saw them going out and coming back together and walking hand in hand. They were an affectionate couple" ("the passage"). In the context of the statement he was writing, the passage, especially in the vocabulary of the original, was curiously incongruous.

965. There was no evidence which explained why Tse would have taken it upon himself to volunteer the passage when he was writing the first statement. It was not pertinent to the subject-matter of the statement, which was about the appearance of his signatures on the documents.

966. Nor was it Lam's evidence that he had asked Tse about the relationship between the Deceased and the Wife. Indeed, Lam's evidence was that Tse wrote "a lot more" in his statement than had been in his answers to Lam's four questions.

967. Lam's evidence in cross-examination on this point included the following: (Transcript, Day 150 p.99-100, Day 151 p.91-2):

"Q.	Did he [Tse] write anything there which was not covered by his oral answers to your oral questions?
A.	You mean, he wrote something more or wrote something less? He wrote a lot more.
Q.	Than what you originally asked him?
A.	That is right.
Q.	Were you able to identify what was the further information that he volunteered which had not been given to you in answer to your oral questions?
A.	Now I do not remember, but I want to say this: at that time, I only asked him very few questions, so he wrote out a lot more".
"Q.	In the last sentence of this first paragraph, he gave a reason as to why he had to stay in the family, or in the home, of Mr Wang, right?
A.	Right
Q.	Did you ask him any question as to why he should stay in the home of Mr Wang?
A.	No, did not ask him.
Q.	Could you think of any particular reason in the context of your asking him to make a statement to you, why should he volunteer such information as an explanation as to why he stayed in the house of Mr Wang?
A.	I do not know".
"Q.	Again, he mentioned that he very often had seen Mr and Mrs Wang as a very loving couple; they went in and out as a couple; and very often, they held hands when walking along the road. Again, did you ask him any questions about how the Wangs walked in and out, whether they were a loving couple?
A.	At that time, he wrote out the statement himself. He was sitting down writing it out. I walked away, as I did not want to affect him.
HIS LORDSHIP:
	Mr Lam, we understand that part of your evidence. Mr Chan wanted to know, before that, did you ask him any questions about how the Wangs walked in and out, whether they were a loving couple? Did you ask him that sort of question?
A.	I do not remember.
MR CHAN:
	Had that sort of question any relevance to the task that you had to perform, to take a statement from him on his being a witness?
A.	No relevance.
HIS LORDSHIP:
	Would you have asked that sort of question, that is, the relationship between Mr and Mrs Wang?
A.	I do not remember. Perhaps I did not. I do not remember" (emphasis added).

968. It is clear from the above testimony that since the Deceased's relationship with the Wife was not relevant to the subject matter of the statement, Lam would not have asked Tse about it. It was also clear from Lam's evidence that Tse did not even ask or consult Lam (the solicitor having the conduct of the interview) as to whether it would be appropriate to add the passage.

969. As to the question whether Tse might have been asked by someone other than Lam to volunteer the information in the passage when writing the statement, it would be noted that Tse alleged that he had not read the documents when he was asked by the Deceased to sign them, so he would not have had discussed the documents with anyone as a result. Further, in the statement that Tse gave to Allan Leung ("Leung"), Tse said only that he had seen the documents when Lam showed them to him, from which it may reasonably be inferred that he had not seen them earlier than that (par. 3).

970. Although it may be assumed that Tse was aware of litigation between the Wife and the Father from newspaper reports, it remains an extraordinary act for a housekeeper, supposedly just shown documents "A" to "D" for the purpose of confirming his signatures on them, and without any prompting or coaching from Lam, and without asking Lam first, to consider it appropriate to take the initiative to add the passage about the relationship between the Deceased and the Wife.

971. No explanation for the addition of the passage was sought by Leung, the Wife's solicitor, who took a second statement from Tse three days later. That was not surprising because Leung was not given Tse's first statement.

972. Therefore there remained no evidence before the court that affords a satisfactory explanation for the addition of the passage in Tse's first statement.

Tse's absence from Hong Kong

973. On 23 September 1999, the Father obtained leave from the court to issue a probate action and to make a report to the police on "matters concerning ... documents inside the envelopes containing testamentary scripts filed by [the Wife] with [the] court on 16th January 1998". The Wife's legal representatives were present at the time the order was made.

974. The next day Tse left Hong Kong for the Mainland and stayed there for 2 1/2 months until virtually his dying day. He was admitted to hospital in a serious condition and the police were unable to interview him.

975. There was no evidence called (e.g. from Tse's wife or other members of his family) to explain why he had stayed on the Mainland during that period. He remained in Chinachem's employ until November 1999, so the Wife must have, at least, acquiesced in his staying away from September to November.

976. There was some suggestion before the court that, perhaps on legal advice, she wished to protect Tse from interference. That may or may not be so, as no evidence was called. But if so, that was unsatisfactory because there is no property in witnesses, and particularly so in a probate action, where witnesses to a will are witnesses of the court (Re Fuld [1965] P 405, 409; Tristram and Coote's Probate Practice par. 27.08). It would be improper for a party seeking to propound a will to keep a witness out of reach, for whatever motives.

977. More importantly, Tse's departure also meant that the Hong Kong police had no opportunity to interview him.

978. There may have been many possible reasons why Tse decided to stay on the Mainland, but in the absence of evidence of the actual reason, the fact that he made himself unavailable to the Hong Kong police, immediately the day after leave was given to make a report to the authorities, meant that he never allowed himself to be questioned by an independent investigator. In the hands of a skilled examiner, Tse would have had to explain for instance, how he could claim (and why he claimed) not to have seen the characters for "Will" when he purportedly signed document "A", given that the characters were virtually immediately above his signature (since Tse had not suggested that the Deceased had covered up the contents of the document).

979. Tse's departure preventing questioning by the Hong Kong police, together with the addition of the volunteered passage referred to above, does not inspire confidence that Tse was a person with no interest in the outcome of the case whose statements could be accepted without question, and this must be put into the balance against what would otherwise be important evidence for the case of the party seeking to prove the will.

Failure to trace other purported witness

980. If Tse's statements successfully proved due execution, then it would not have been necessary to consider the identity of the other purported witness to the will being propounded. However, for the reasons set out above, it would be unsafe for the court to act on his statements alone.

981. That leads me to the other purported witness. It is well-established in probate practice that

"at the hearing of any cause, the court must be satisfied, upon the examination of one or more of the attesting witnesses (if available), of the due execution of the will; although if a satisfactory explanation is given for the failure to trace the witnesses, so that no adverse inference can be drawn from their absence, the will must be pronounced for without their evidence. To prove the due execution of a will, it is necessary to examine one only of the attesting witnesses, provided that his evidence establishes due execution. If the witness called fails to prove its due execution, then the party propounding the will is bound to call the other attesting witness, notwithstanding that he may be an adverse or hostile witness.

...

If the court is dissatisfied with the evidence of the attesting witness examined and the other attesting witness is not called, it is competent to decline to grant probate of the instrument propounded" (emphasis added)

(Tristram and Coote's Probate Practice par. 27.08)

982. The evidence adduced on behalf of the Wife in relation to the identity of the other purported witness was most unsatisfactory. It was not clear from Leung's evidence when he was first instructed that this was Lee Kin-sang who had died. But what is important is the fact that Lovells' written advice to the Wife on 9 September did not mention taking a statement from Lee Kin-sang, or if it was not Lee Kin-sang, trying to ascertain the identity of this person. That shows that by that date, Lovells' instructions were that this was Lee Kin-sang, or at least a person who it was known could not be traced.

983. The Wife's case that the other purported witness was Lee Kin-sang was consistent with her attempt to include evidence of Lee Kin-sang's signature. However, that case was not persisted in at trial. On the contrary, the case advanced at trial was that the identity of the other purported witness was not known.

984. However if that were the case, attempts should have been made to trace persons the Deceased had seen in the short period between 12 March 1990 (the date of the will being propounded) and a few days later when the envelopes were allegedly given to the Wife. There was no evidence that any such attempts were made (e.g. by checking the Deceased's appointments).

985. The court expects a party seeking to prove a will in solemn form to make a better effort to trace a purported witness to due execution. If the Wife's case was that the person was Lee Kin-sang, and if he had died by the time of trial, then the court would expect to see evidence of the likelihood that he had seen the Deceased at the material time, of his signature and of his death. If the Wife's case was that the person may not have been Lee Kin-sang, then the court would expect to see no stone unturned, in the way the case had been fought, directed towards ascertaining the identity of and tracing this person. There was a complete absence of such evidence and no satisfactory explanation for this startling choice of inactivity.

No evidence of transfer of documents from Deceased to Wife

986. There is authority that where due execution is in issue, the location where the will being propounded was found is relevant (Saph v Atkinson, 57-8). In the present case, the Wife's case was that possession of the documents being propounded was transferred by the Deceased to herself a few days after the date appearing on the purported will.

987. On the Wife's own case therefore, she was the last available living link to the res gestae (Tse having died, and the other purported witness and the authors being untraced). If she had given evidence at trial of this transfer of documents and if her evidence had been accepted, this would have been a very strong piece of evidence in support of her counterclaim.

988. Although she had filed affirmations of testamentary scripts, there had not been any orders from the court that they could be used in place of oral evidence at trial.

989. As for the reasons given before this court as to why she might not have wished to give oral evidence, a witness may of course have all sorts of reasons why he or she might not wish to give evidence in court and be cross-examined. But apart from a situation where he could claim the privilege against self-incrimination, a person who chooses not to give evidence and be cross-examined has to bear the consequences of that decision.

990. Of course, given the lines along which the Father's case was run, questions incriminating the Wife might have been put to her in cross-examination. No doubt, her legal representatives would be attentive to such a risk and if that occurred, she would of course be advised to refuse to answer such questions. In case of dispute, the court would have to decide on the point. But by not offering herself as a witness at trial at all, she did not even get to that stage.

991. It cannot be said that since the Father had not pleaded that the Wife was the author of document "A", there would have been no real material questions for her. There was the conflicting evidence, discussed above, working against the presumption of due execution (cf. Davis v Mayhew [1927] P 264, 280). The Father was of course entitled to put the Wife to strict proof of her allegation that the Deceased had given her the documents. There was also at least one very material question about the circumstances in which the documents were said to have been kept. Counsel on behalf of the Father would have been entitled to question her on how she could have made an affirmation on 15 July 1999 (Nina TH Wang, 2nd affirmation, par. 51(n), HCMP 3454/99) that she believed she was the "sole beneficiary" of the Deceased's estate, before the documents were opened by the court on 6 September 1999. One could understand an assertion that she believed she would be one of the beneficiaries, as the documents were allegedly given to her, but there would have to be other evidence to explain the belief that she was the "sole" beneficiary because her allegation had been that the envelopes had been sealed when the Deceased gave them to her.

992. Therefore, the circumstances are that (1) the court has before it an unconventional "home-made" will (2) purportedly made by a person who would be reasonably expected to make one in the conventional way, (3) containing sentiments which contradict undisputed evidence, (4) where the purported witness's statement contained surprising elements and has been untested, (5) where no effort has been made to trace the other purported witness and the authors, and (6) where a vital living link has chosen not to testify as to how she came to possess it.

Hand-writing expert evidence - general points

993. With that I turn to the handwriting expert evidence. I would first make some general points on this aspect of the case. First, the expert evidence was not as helpful to the trial court as it could have been because the oral evidence of Professor Jia called on behalf of the Wife was quite different from the contents of his report. During his oral evidence, points came up in bits and pieces in a way which the rules of court governing the exchange of expert evidence were designed to avoid.

994. Secondly, minute points continued to threaten to swamp the court even on appeal. It was easy to get bogged down by details of individual dots and strokes, and then to get carried away by them, instead of considering all the evidence of similarities and differences as a whole.

995. Thirdly, it is important to remember that the human hand is not a machine, and none of the experts has ventured to suggest that handwriting authentication is a complete, let alone exact, science. As Mr Jules Sher QC (sitting as Deputy Judge of the High Court) observed of handwriting analysis in Fuller v Strum (2000), Transcript p.11,

“In my judgment, however, there is a world of difference between the type of expert evidence led in Re B [radiologists' evidence regarding injuries] and the evidence contained in Dr Giles' [the handwriting expert] report in this case. The training of experts enables them to identify facts which a lay witness or a judge could not identify, without expert help. Such evidence may truly be described as scientific and the radiologists' evidence as to when an injury occurred falls plainly within this category. But some expert evidence may amount to no more than the drawing of inferences from facts observable as much by the expert as by a lay witness; and the inferences to be drawn from those facts may be capable of being drawn as much by the expert as by a lay witness. Of course in such a case the views of the expert are entitled to be given great weight. After all, the expert's training and experience will have equipped him or her to draw these inferences. But in relation to this type of expert evidence the judge, I think, is entitled to form his own view, having regard to, and balancing, the other evidence available to him in the case" (emphasis added).

That there is a distinction between the scientific part and the opinion part of the evidence of a handwriting expert is now recognized judicially. The question of which features of the handwriting are significant and the inferences to be drawn from them are questions of judgment, assessment and opinion (Phipson on Evidence 15th ed par. 37-10).

996. Fourthly, at the end of the day, the handwriting evidence is but one part of the evidence in this case and should be viewed together with the rest in the round.

Questioned Wang signatures

997. I turn first to the purported signatures of the Deceased on documents "A" to "D" ("the Questioned Wang signatures"). Compared with the known signatures of the Deceased, it is obvious that they resemble more or less the two 1958 signatures but not his signatures since the 1980's.

Resemblance to older, not recent, signatures

998. This is important to the question of authenticity. As stated in Questioned Documents by Osborn, "considerable doubt can be cast upon the authenticity of a suspect signature when it can be shown to be of a pattern which had been discarded by the writer before the date attributed to the document on which it appears, and even less credence will be given to a signature which considerably antedates the style of genuine signatures ... ".

Not two different styles

999. I do not accept the submission made on behalf of the Wife that the Deceased had two different styles of signatures, one shown in the 1958 signatures and the other shown in the later signatures. There is no evidence of a marked choice of two different styles, and a comparison of the known signatures is more indicative of a gradual evolution with the passage of time.

No evidence explaining choice of older style

1000. Even if the Deceased had two different styles of signatures, one which he preferred to use in the 1950's and the other which he preferred to use in the 1980's, there is no evidence why he should have chosen to revert to the older "more artistic" one when signing the will being propounded. If the Deceased had drawn a picture instead of a will, that might have provided an explanation for the choice of reverting to a "more artistic" signature, but it was not a picture that he was signing but a will disposing of a large business empire.

1001. The purpose of any testator making a will is to enable his estate to be dealt with without dispute, and the single, vital purpose of the signature on it is for identification. It is difficult to see why a testator would have chosen to create doubts over the authenticity of the will by reverting to a signature that he had not used for more than 30 years. There is no evidence that provides a reasonable explanation.

Calligraphic pen

1002. This takes me to the issue of the calligraphic pen. It was suggested on behalf of the Wife that this was (at least partly) responsible for the appearance of the Questioned Wang signatures. The same question arises as to why the Deceased would have chosen to use a pen that would not have allowed him to sign in a way that would have caused no doubts in recognition and acceptance of his signature. There is no evidence that provides a reasonable explanation. What is in evidence is the fact that Tse's purported signatures were in ballpoint. So, even on the Wife's case, another writing instrument was available.

1003. I shall revert to the Wife's case that the calligraphic pen provided an explanation for some of the differences between the Deceased's known signatures and the Questioned Wang signatures later in this judgment.

Approach

1004. The court was taken at great length to a number of features of imitation, such a tremors and super-imposability (in the case of Tse), and submissions were made on the presence or absence of some or all of these features and possible explanations for them. There were also submissions made on similarities, a distinctive aspect of which were "matching variations", by which it was sought to show that variations found in some of the known signatures were also found in some of the Questioned Wang signatures. There were of course many submissions made on differences between the known signatures and the Questioned Wang signatures, and explanations sought to be provided for their appearance.

1005. It would only further lengthen this lengthy judgment if I were to rehearse these submissions. The evidence has to be considered in the round, and a weighing exercise undertaken. In undertaking this weighing exercise, it is not simply a question of totting up the number of similarities against the number of differences, because similarities which are visually obvious and easy to copy would provide less strong evidence of authenticity as compared with unexplained differences which would provide stronger evidence of non-authenticity. In my view, the more important points are those set out below.

Layout of the characters in the Questioned Wang signatures

1006. The first main point that struck me was the layout of the three characters forming the vertically-written signatures. This point is best shown in the chart entitled 王德輝簽名比對表 ("Wang Teh-huei signature comparison chart").

1007. It is clear from the known signatures that when a line (shown in red in the chart) is drawn vertically through the middle of a rectangle formed by the signature as a whole,

(i) the vertical stroke of the character 王 ("wang") is nearly always straight down the red line;

(ii) the right part of the character 德 ("teh") also is nearly always bisected by the red line;

(iii) the left and right parts of the character 輝 ("huei") are always on different sides of the red line.

1008. This shows that the Deceased had a good eye for vertical alignment. That however is not the case with the Questioned Wang signatures, as appears clearly from the chart referred to above, where

(i) the vertical stroke of the character 王 ("wang") falls entirely to the left of the red line;

(ii) nearly the entirety of the character 德 ("teh") falls to the left of the red line;

(iii) there is not such a strong demarcation of the left and right parts of the character 輝 ("huei").

1009. Insofar as Professor Jia sought to negate this point by drawing different lines, there was no satisfactory objective justification of his choice of points between which his lines were drawn.

Unexplained significant differences

1010. It was not disputed by the experts that unexplained significant differences were indicative of forgery. Having examined the evidence and the submissions made by both parties, I consider there were the following unexplained differences between the known signatures since the 1980's and the Questioned Wang signatures. In referring to them as "unexplained", I do not suggest that no attempt has been made to offer explanations. On the contrary, attempts have been offered at length, e.g. by reference to the nature of the pen used, or by reference to the postulated writing speed, etc. However, in respect of the points set out below, the attempts did not provide reasonable explanations.

1011. In all the Questioned Wang signatures, the first horizontal stroke of the character 王 ("wang") was written slanting upwards in a somewhat stylised manner. Since the 1980's, this stroke in all the known signatures has been bolder, firmer and either flat or slanting downwards.

1012. The Wife has sought to explain this difference by reference to the use of the calligraphic pen. It is clear from handwriting text books that the appearance of a signature may be affected by the nature of the pen used. Further I am aware that Mr Tsui agreed with the trial judge when he suggested that persons using calligraphic pens may have a "tendency" to write "a bit upwards". However, this was no more than a generalised suggestion which Mr Tsui did not accept explained the angle of slant seen in the Questioned Wang signatures. The evidence went as follows (Transcript Day 70 p.63-5):

"HIS LORDSHIP:
	Usually, people hold calligraphic pens with the broad nib at 45 degrees to the vertical axis. That is the usual way of holding a calligraphic pen?
A:	Yes.
HIS LORDSHIP:
	It is easier to write with the calligraphic pen a little bit upwards than horizontally, because the friction will be less?
A:	Yes, it would be slightly slanting upwards usually, yes.
HIS LORDSHIP:
	That may explain why the first stroke is a little bit tilted upwards?
A:	Bu the slant is too much when you compare .. Suppose that stroke is horizontal, but this stroke is quite slanted, quite deviated from the horizontal; it is too much.
HIS LORDSHIP:
	Yes it is upwards rather than horizontal.
A:	Yes.
HIS LORDSHIP:
	But that ease of writing upwards than horizontally applies both to the forger and to Mr Wang?
A:	And it also depends on the paper position. ...
HIS LORDSHIP:
	Yes but do you agree that in using a calligraphic pen, the tendency to write a bit upwards would apply to both the forger and to Mr Wang?
A:	Yes.
...
MR LEE:
	In one of your answers to my Lord's questions, you said that the initial stroke in the four questioned signatures were slanting too much when compared with the 668 and 669, I suppose. Is that what you are saying?
A:	Yes.
Q:	As to the slant, the forger got it wrong already, right?
A:	Yes, the general slant".
	(Emphasis added)

1013. The fact remains that the Deceased's habit was to write a bolder, firmer, flat or downward-slanting horizontal stroke as the very first stroke of his signature. There is no evidence to explain why he should have chosen to use a pen that would require him to change that habit. Further the generalised "tendency" to write "a bit upwards" is not seen in all the horizontal strokes in the signatures, e.g. in the signature on document "B", the lowest horizontal stroke of the character 王 ("wang") does not exhibit such a tendency, and Mr Tsui did not accept that that tendency would have accounted for the angle of slant seen in all the Questioned Wang signatures.

1014. Further, in virtually all the known signatures since the 1980's, the Deceased has written the vertical stroke of the character 王 ("wang") at a position nearly at the centre of the horizontal stroke above it. This was clearly not so in the Questioned Wang signatures, where the vertical strokes start to the left of centre. Also in all the known signatures since the 1980's, the last horizontal line in this character is written without an initial loop upwards, unlike in three of the four Questioned Wang signatures. There are no reasonable explanations for these differences.

1015. Further, in the character 德 ("teh") in all the Questioned Wang signatures, the lifting of the hand before writing the last dot is very pronounced (by which I do not mean the emphasis of the last dot, but the blank space left before that dot is written). This was not the Deceased's habit as shown in the known signatures since the 1980's. It was suggested that this could be due to the "more artistic" style. I have set out above my reasons for the rejection of the "two different styles" and the "more artistic style" arguments and I will not repeat them here.

1016. As for the "matching variations" exercise undertaken on behalf of the Wife, this comprised combing through the Questioned Wang signatures and the known signatures, and finding occasional common variations in writing a dot or a stroke. It was suggested that the scattering of these "matching variations" proved genuineness. However none of these variations in the known signatures was shown to be unique to the Deceased and discoverable only by scientific methods or experts. Nor is there any reason to assume that the person who wrote the Questioned Wang signatures had access to only some of the Deceased's known signatures. As such, I do not consider that there was such weight in this exercise as to counteract the effect of the unexplained significant differences.

1017. Of course the point has been made on behalf of the Wife that a "master forger"would not have chosen to copy signatures knowing them to be outdated, or would not have missed significant differences. Mr Cheng and Mr Tsui were cross-examined at length along these lines. As noted above, it is not for the court to speculate why a forger would have chosen to present these four signatures and not others (which might or might not have been less successful imitations). The court's function is simply to examine the evidence before it and to decide the question whether the Deceased had made the signatures in the documents being propounded.

1018. There are in my view significant and unexplained divergencies in fundamental parts of the structure of the signatures in the documents being propounded. They have not been satisfactorily explained as natural variations in the Deceased's signature.

1019. In my judgment, the evidence shows that the signatures purporting to be those of the Deceased in the documents being propounded have been forged.

Questioned Tse signatures

1020. I will deal with Tse's signatures briefly. As far as his signatures are concerned, I do not consider that there is sufficient cogent evidence to prove that he did not make the signatures appearing on the documents being propounded. I do not accept that the super-imposability of the signatures in the documents being propounded (relied on heavily by Mr Tsui) are indicative of forgery. As was demonstrated to this court, even Tse's known signatures were highly super-imposable.

1021. Further, the burden of the Father's case was that Tse's signatures in the documents being propounded were written in a more reserved (or less free-flowing) way than in his known signatures in 1990. That is indeed the case for the signatures in his police statements i.e. No.674 (1-9) dated 24 May 1990. However I note that the application for identity card renewal (No.700) dated 6 July 1990 contained a more reserved way of signing.

1022. Further, the Home Visit Permit No.5992 also contained a more reserved way of signing. There was some dispute between the parties as to whether this signature had been made in 1992, when the permit was used. It was suggested that a signature was not required before the document could be used. There was no evidence that an official document such as that could be used without the bearer's signature. I am not prepared to accept that it could. Accordingly, the signature would have been put on by 1992. Two other signatures No.1829 dated 1993 and No.5993 dated 1994, which were made before Tse's hospitalization in 1995, were also signed in a more reserved way.

1023. I am therefore not satisfied that there has been sufficient cogent evidence of forgery of Tse's signatures whether by super-imposition or tracing. I should however clarify that this should not be taken as an acceptance of the submission that Tse's signatures on the documents being propounded were made in 1990. The position simply is that in 1990 and thereafter until 1995, he was capable of both the free-flowing and more reserved ways of signing, but from 1995, he almost exclusively adopted the more reserved way of signing.

Conclusion

1024. I have set out above my finding that the Questioned Wang signatures are forgeries. I would add that if I am wrong in that there was insufficient cogent evidence to prove forgery, I would nevertheless have no hesitation in holding that the party propounding the will has not proved it to the standard required for probate in solemn form, taking into account, not only the handwriting evidence, but more importantly that

(1) the court has before it an unconventional "home-made" will

(2) purportedly made by a person who would be expected to make one in the conventional way to avoid doubt and delay

(3) containing sentiments directed against his parents and siblings which contradict undisputed evidence

(4) where the purported witness's untested statement contained surprising elements

(5) which witness had taken himself out of the jurisdiction immediately after the court granted leave for a report to the Hong Kong police to be made

(6) where no effort has been made to trace the other purported witness and the authors, and

(7) where the last living link to the primary facts (the making of the documents as the will of the Deceased) has not testified as to how she came to possess them.

1025. For the reasons set out above, I agree that the appeal should be dismissed.

Hon Waung J:

1026. It is regrettable that the heat and emotion generated throughout this legal process have clouded the essential questions which should be asked and answered for a proper resolution of the dispute between the Respondent/Plaintiff ("Father") and the Appellant/Defendant ("Wife"). The fundamental complaint of the Wife against the Judgment is that the Judge failed to recognise the true issues of the case, namely Execution and Forgery and that the Judge committed serious errors under each and every important issue in this case.

1027. In my view the issues which this Court should resolve are the following, which I consider in the following paragraphs (in brackets):-

(A) The approach of the Court of Appeal to this appeal from a first instance judgment after a long Trial (Para. 1029).

(B) In the context of modern probate litigation, whether Execution of the Will has been established or not, by reference to:-

(I) history of the formal probate requirements of Wills (Para. 1031-1038);

(II) whether the finding of Forgery by the 9 Suspicious Circumstances ("SC") should be set aside (Para. 1039-1042);

(III) whether the SC to be dispelled principle ("Dispel SC Principle") is applicable not only to a case of Knowledge and Approval but also to the question of Execution (Para. 1043-1060);

(IV) if the Dispel SC Principle is not applicable to Execution, whether the presumption of Omnia praesumuntur rite esse acta ("Omnia Presumption") applies in the circumstances (Para. 1061-1104);

(V) whether, if Omnia Presumption execution does not apply, Execution has been proved (Para. 1105-1112);

(VI) whether any adverse inference is to be drawn from Nina not giving evidence and if so, what is the adverse inference and with what effect on (IV) and (V) above (Para. 1113-1130);

(C) Whether Forgery has been established by reference to all the relevant facts including the handwriting evidence (Para. 1131-1279) by reference to:-

(I) Preliminary Observations on Comparison Handwriting Evidence (Para. 1132-1145);

(II) Preliminary Observations on Comparison Chinese Handwriting Evidence (Para. 1146-1151);

(III) No Methodology, No Proper Methodology and Fatal Methodology (Para. 1157-1178);

(IV) Wrong approaches, Wrong Application of Methodology and Misconceptions (Para. 1179-1213);

(V) Serious Errors of Micro Analysis (Para. 1214-1279)

(D) Whether, even if there is no Judgment in favour of the Wife, there had been a fair trial, and if not whether there should be a Re-Trial by reference to:-

(I) No Fair Judicial Process by reason of Plagiarism (Para. 1282-1297);

(II) Wife's Points not considered and no reason given for rejection (Para. 1298-1301);

(III) Unpleaded case made the subject of Judgment (Para. 1302)

(E) Conclusions (Para. 1303).

1028. Yeung JA has set out in considerable detail the history of this matter and most of the relevant facts. I will therefore only amplify particular aspects of the facts or his general description of the case, when it is necessary for the purpose of this my judgment.

(A) APPROACH OF THE COURT OF APPEAL

1029. The approach of the Court of Appeal on appeal from a Judgment after a full Trial was set out in the relevant principles extracted from the authorities cited by Yeung JA. I would only like to add that:

(a) the Court of Appeal does not allow an appeal lightly and should only enter judgment in favour of the appellant, when it is satisfied that this is the correct outcome to the Trial and in this connection I accept what was said by Kirby J in Whisprun Pty Ltd. v Dixon (2003) 77 ALJR 1598 at 1615 that:-

"the appellate court must reconcile the rule of restraint (enhanced where a decision rests on a credibility assessment what was reasonably open to the primary decision-maker) and the need to protect parties against clearly flawed primary decision, illogicality in reaching them and injustice that demands that the appellate court exercise the functions of review conferred on it by the legislature";

(b) the Court of Appeal should not order a Re-Trial unless it is persuaded that this is the only fair and just order to make in the circumstances.

(B) EXECUTION OF THE 1990 DOCUMENTS

1030. A large part of the time during the hearing of the appeal was taken up by the subject of Execution of the 1990 Documents. Execution in this context of probate litigation is a technical term applicable to "the performance of the formalities required to validate a legal document" (Shorter Oxford English Dictionary page 885). A brief history of the formal legal requirement to validate wills, may assist us in understanding the subsequent discussion of the Execution of the 1990 Documents.

(I) HISTORY OF THE FORMAL REQUIREMENT OF WILLS

1031. In England prior to 1540 there was no right of testamentary alienation of land. By the Statute of Wills 1540 of Henry VIII, testamentary disposition of land was permitted. By the Statute of Frauds 1677 testamentary disposition of land was required to be in writing signed by three witnesses.

1032. By the important Wills Act 1837, it was provided that no will is valid unless:-

(i) it is in writing

(ii) it is signed by the testator

(iii) at the foot of the will

(iv) in the presence of 2 witnesses

(v) who must be both present at the same time and

(vi) such witnesses attest and subscribe by their signatures in the presence of the testator.

1033. The strict requirement of the position of the signature of the testator was relaxed somewhat by the Wills Act Amendment Act of 1852.

1034. Upon the founding of Hong Kong, the strict formal requirements of wills prevailing in England also became the law of Hong Kong but exception was made for Chinese wills as provided by Ordinance No. 3 of 1854 (see The Wlls Ordinance 1970 of Prof. Evans (1971) 1 HKLJ 107, 110-112). The special position of Chinese wills was reflected in section 3 of the Wills Ordinance Cap. 30.

1035. By section 5(2) of the Wills Ordinance of 1970 which is the provision governing the disputed 1990 Documents, Chinese wills are given the long-held privilege of a low statutory requirement for its formal validity. Section 5 of the Ordinance is in the following terms:-

"(1) Subject to subsection (2), no will shall be valid unless it is in writing and executed in accordance with the following rules:-

Rule 1 It shall be signed at the foot or end thereof by the testator, or by some other person in his presence and by his direction.

Rule 2 Such signature shall be made or acknowledged by the testator in the presence of two or more witnesses present at the same time, and each witness shall attest by his signature the signature of the testator, or of the person signing for him, in the presence of the testator, but no form of attestation shall be necessary.

Rule 3 It is sufficient if the signature of the testator, or of the person signing for him, is so placed at or after, or following, or under, or beside, or opposite to the end of the will, that it is apparent on the face of the will that the testator intended to give effect by the signature to the writing signed as his will.

Rule 4 Without prejudice to the generality of Rule 1, a will shall not be affected if-

(a) the signature does not follow or is not immediately after the foot or end of the will; or

(b) a blank space intervenes between the concluding word of the will and the signature; or

(c) the signature-

(i) is placed among the words of the testimonium clause or of the clause of attestation; or

(ii) f ollows or is after or under the clause of attestation, either with or without a blank space intervening; or

(iii) follows or is after, or under, or beside the names or one of the names of the attesting witnesses; or

(c) the signature is on a side or page or other portion of the papers containing the will on which no clause or paragraph or disposing part of the will is written above the signature; or

(d) there appears to be sufficient space on or at the bottom of the preceding side or page or other portion of the same paper on which the will is written to contain the signature.

Rule 5 A signature shall not be operative to give effect to any disposition or direction which is underneath or which follows it, nor shall it give effect to any disposition or direction inserted after the signature is made.

(2) Any will of a Chinese testator written wholly or substantially in Chinese and signed by the testator shall be valid and duly executed although not executed in accordance with the rules set out in subsection (1)."

1036. The strong contrast between on the one hand, the elaborate statutory requirements of the formality of testator's signature in the presence of and with the attestation of witnesses of all wills (except Chinese wills) under Section 5(1) of the Wills Ordinance and on the other hand, the minimal requirement of formality of only testator's signature of Chinese wills under section 5(2), clearly points to the statutory intention (long accepted in Hong Kong) that Chinese wills are to be easily recognized with legal informality not known to English law. Rule 10(2) of the Non- Contentious of the Probate Rules is good indication of the low statutory requirement of Chinese wills.

1037. The history of probate litigation in the English and Hong Kong courts disclose that the propounding of wills in the Probate Court have been the subject matter of the following categories of defences or disputes:-

a. Execution;

b. Testamentary capacity;

c. Lack of Knowledge and Approval;

d. Undue Influence;

e. Fraud;

f. Forgery.

1038. Execution in probate litigation is concerned with ensuring that the statutory requirements such as set out earlier and in section 5(1) of the Wills Ordinance are satisfied. This is to be contrasted with the other 5 disputes of (1) testamentary capacity, (2) Lack of Knowledge and Approval, (3) undue influence, (4) fraud and (5) forgery, all of which are concerned with judge-made legal obstacles to validity of wills not derived from statutory requirements.

(II) SET ASIDE FINDING OF FORGERY BY 9 SUSPICIOUS CIRCUMSTANCES

1039. Suspicious Circumstances ("SC") gave this probate litigation its unique flavour. Much of the time taken up at the hearing of this appeal was over the disputed question of the relationship between what I will call Dispel SC Principle (derived from the second rule in Barry v Butlin) with firstly Execution and secondly Lack of Knowledge and Approval. This arose because of the finding of the Judge that there were 9 Suspicious Circumstances which caused the Judge to hold that these 9 Suspicious Circumstances had not been dispelled, leading to his conclusion that Forgery has been established. (see paragraph 5.11 of Judgment)

1040. Forgery is a separate and independent defence from the defence of no Execution or the defence of Lack of Knowledge of Approval. The Judge did not find that there was no Execution or that there was Lack of Knowledge and Approval. Instead the Judge found that there was Forgery because of the suspicion from the 9 Suspicious Circumstances which had not been dispelled and for convenience I will hereinafter call this finding of the Yam J. the "Forgery By 9 SC". The Judge also found that there was Forgery based on Handwriting Expert evidence which for convenience I will hereinafter call the "Handwriting Forgery".

1041. Suspicious Circumstance is relevant to the defence of Lack of Knowledge and Approval but is not known to be relevant to the defence of forgery. In the annals of probate litigation, there is no known case let alone any authority showing that a finding of forgery has been or can be established by the failure to dispel Suspicious Circumstances.

1042. Yeung JA has pointed out (Para. 409-422) the error of the Judge's conclusion on Forgery By 9 SC. I agree with his view and I will only like to add that this is more than merely a matter of finding of Forgery without any proper pleading or by applying the wrong burden of proof. It goes to the heart of the proper judicial process in probate litigation. In my view, the finding and conclusion of Forgery By 9 SC, set out at paragraphs 5.127 and 5.128 of Chapter 5 of the Judgment cannot be allowed to stand and must be set aside.

(III) WHETHER DISPEL SC PRINCIPLE IS APPLICABLE TO EXECUTION

1043. The question which was not asked in the Judgment but which was at the heart of the hotly contested difference between Mr. Thomas for the Wife and Mr. Tang for the Father is whether the Dispel SC Principle is relevant to Execution. Mr. Tang contends that both on principle and on authority the Dispel SC Principle is relevant not only to Lack of Knowledge and Approval but also to Execution. Mr. Thomas disagrees with that argument both on authority and on principle. I will first consider the question based on authority.

1044. In Barry v Butlin (1838) 11 Moore 480, Baron Parke touched on "suspicion of the Court" when he said at page 482-5 (underlining supplied):-

"The rules of law ..... These rules are two; the first that the onus probandi lies in every case upon the party propounding a Will; and he must satisfy the conscience of the Court that the instrument so propounded is the last Will of a free and capable Testator.

The second is, that if a party writes or prepares a Will, under which he takes a benefit, that is a circumstance that ought generally to excite the suspicion of the Court, and calls upon it to be vigilant and jealous in examining the evidence in support of the instrument, in favour of which it ought not to pronounce unless the suspicion is removed, and it is judicially satisfied that the paper propounder does express the true Will of the deceased." (pp. 482-3)

"The strict meaning of the term onus probandi is this, that if no evidence is given by the party on whom the burden is cast, the issue must be found against him. In all cases the onus is imposed on the party propounding a will, it is in general discharged by proof of capacity, and the fact of execution, from which the knowledge of assent to the contents of the instrument are assumed; and it cannot be that the simple fact of the party who prepared the Will being himself a Legatee, is in every case, and under all circumstances to create a contrary presumption, and to call upon the Court to pronounce against the Will, unless additional evidence is produced to prove the knowledge of its contents by the deceased...." (pp. 484)

".... But in no case amounting to more than a circumstance of suspicion of the Court in investigating the case, and calling upon it not to grant probate with full and entire satisfaction that the instrument did express the real intention of the deceased." (pp. 485)

1045. At page 482-3 when Parke, B. was discussing the second rule, of the circumstance to excite the suspicion of the Court and requiring the suspicion to be removed, he was addressing the question of Knowledge and Approval of the contents of the will by the testator (namely the internal mental intention of the testator) and not the matter of execution (the external act of signing by testator in the presence of witnesses). The second rule of Parke, B. is on the assumption that there was execution by the testator (namely Execution not in issue) but there are facts which caused the Court to have suspicion (namely SC) as to whether the testator Knew and Approved the contents of the will (namely raising the issue of Lack of Knowledge and Approval) which suspicion therefore had to be removed by the party propounding the will. The situation of suspicion therefore puts the propounder of the will in a different situation from the normal case where "the fact of execution, from which the knowledge of and assent to the contents of the instrument are assumed" (pp. 484). The Dispel SC Principle is therefore narrowly related to the issue of Lack of Knowledge and Approval and is not in any way connected to the issue of Execution.

1046. In Tyrrell v Painton [1894] P 151, it was not Painton (the beneficiary under the disputed will) but the son of Painton who prepared the will and who witnessed the execution in the middle of the night by the sick old lady testatrix. Execution was therefore not in issue. What was in issue was whether the testatrix really knew and approved the contents of the will. The question there was whether the 2nd rule in Barry v Butlin of suspicion was confined only to a situation where the legatee was the preparer of the will or could include the situation where the son of the beneficiary was the person who prepared and witnessed the will. The question was confined to SC in relation to the defence of Lack of Knowledge and Approval. That knowledge and approval of the testatrix was the issue and not execution could be seen from what Lindley, LJ said:

"suspicious that I am in the greatest doubt whether she knew the effect of the document she was signing" (pp. 156)

"wherever such circumstances exist ..... it is for those who propound the will to remove such suspicion, and to prove affirmatively that the testator knew and approved of the contents of the document .... " (pp. 157).

Davey, LJ was also of the judgment that the suspicion cast doubt on the "mind of the testator" when he said at page 159

"raise a well-grounded suspicion that it does not express the mind of the testator".

The Terrell judgment did not extend the Dispel SC Principle to Execution. All it did was to make clear that the Lack of Knowledge and Approval special rule was not confined to suspicion where the beneficiary prepared the will but could extend to all circumstances such as in Terrell where Painton's son prepared the will in favour of his father and came to the testatrix in the middle of the night to ask her to sign in front of a stranger and then took the signed will away.

1047. In Davis v Mayhew [1927] P 264 the Court of Appeal took the view that the Dispel SC Principle is applicable only to the question of Lack of Knowledge and Approval. Sargent, LJ said at page 283

"the principle is, that wherever a will is prepared under circumstances which raises a well-grounded suspicion that it does not express the mind of the testator ...".

At page 286 Lawrence LJ said

"... whether, apart from the evidence as to the formal execution of the will, the circumstances were of such a nature as to arouse a well grounded suspicion that the alleged will did not in fact represent the testamentary wishes of the testator ... ".

Further at bottom of page 286, Lawrence LJ said

"... primarily circumstances existing ... having a direct bearing on the question whether the testator then knew and approved of its contents."

1048. Re Fuld Deceased (No. 3) [1968] P 675 is what might be called the anchor sheet of the Father's legal authorities. A superficial reading of dicta in that judgment might give hope to the Father's case. But this long judgment must be properly understood in the context of what was being decided.

1049. The second and fourth Codicil gave substantial legacy to Dr. Tarnesby who attended to the very sick testator Fuld at the time and who was propounding these two Codicils which the Judge held to be invalid for a variety of reasons including Lack of knowledge and Approval. At page 712 in respect of the second Codicil, after holding that there was no valid execution and no testamentary capacity, the Judge went on to consider the case of Lack of Knowledge and Approval on the basis that he might be wrong on the other two issues. The argument for Dr. Tarnesby was that the Doctor did not draw the second Codicil and therefore there could be no operation of the Lack of Knowledge and Approval vigilance because (it was argued) that was confined to a situation where the person who drew the instrument took a benefit under the instrument. At page 712E the Judge said;-

"I would have expected the law to require the court to exercise scrutiny whenever a case revealed reasonable grounds for suspicion. But fortunately there is authority. In Tyrell v Painton .... the rule is not confined to cases where the will is prepared by a person taking a benefit under it ..... The court's vigilance is called for whenever circumstances reasonably excite suspicion."

The whole of what were quoted above and in fact all of what was said by Scarman J. from page 712 C to F was in the context of whether the testator Knew and Approved the transaction. The apparently wide language used by Scarman J. therefore had to be understood in the context of what he was considering and not be transposed as to widen the scope of Dispel SC Principle to be applicable to Execution. Turning to the 4th Codicil, page 716 F to G is again a discussion of the question of Knowledge and Approval and concluded at 716G that Dr. Tarnesby has not proved that the testator knew and approved of the 4th Codicil. Then at page 719 the general expression of Scarman J. in respect of the rejection of the three codicils are relied upon by Mr. Tang as widening the scope of operation of Dispel SC Principle to include Execution. The Judge referred to the first kind of safeguard being the requirement of Execution, namely what is required by statute. Then the Judge went on to discuss the second kind of safeguard, namely there are Suspicious Circumstances when the court must ensure that the burden of proof rests upon the party propounding the will. The Judge was there again expressing the view in the context of Lack of Knowledge and Approval. The Respondent in this case reads too much into the passage of Scarman J. at page 719C to F. Scarman J. (as he then was) did not suddenly out of the blue decide to create new law. There was no reason for him to do so. Re Fuld (No. 3) does not assist the father.

1050. Fuller v Strum [2002] 1 W.L.R. 1097 is the latest authority from the English Court of Appeal on the proper modern approach of the court to probate litigation. The Judge at trial (Sher, QC) held against part of the will on the basis that the suspicion had not been dispelled and that therefore the claimant failed to show that there was knowledge and approval. Although a large part of the Judgment of the Court of Appeal was to consider why the Judge had gone wrong with his holding that the dispel of the suspicion required a beyond reasonable doubt standard of proof, the Judgment is instructive not only for the modern approach to probate litigation but also for a proper understanding of the Dispel SC Principle applicable to Lack of Knowledge and Approval.

1051. Peter Gibson LJ at page 1107 said:

"Probate proceedings peculiarly pose problems for the court because the protagonist, the testator, is dead and those who wish to challenge the will are often not able to give evidence of the circumstances of the will. The doctrine of "the righteousness of the transaction" whereby the law places a burden on the propounder of the will, in circumstances where the suspicion of the court is aroused, to prove affirmatively that the deceased knew and approved of the will which he was executing, is a salutary one which enables the court in an appropriate caser properly to hold that the burden has not been discharged." (para. 32)

Chadwick, LJ said at page 1118 and 1120:-

"... The question is not whether the court approves of the circumstances in which the document was executed or of its contents. The question is whether the court is satisfied that the contents do truly represent the testator's testamentary intention. This is not to suggest that the circumstances of execution or the contents may not, in the particular case, be of the greatest materiality in reaching a conclusion whether or not the testator did know and approve of the contents of the document-and did intend that they should have testamentary effect. But their importance is evidential. There is no overriding requirement of morality. If Lord Hatherley's reference to "the righteousness of the transaction" in a speech delivered in the late 19th century leads to misunderstanding at the beginning of the 21st century, then the time has come to consider whether the phrase is still helpful. For my part, I think it is better to avoid it. (para 65)

"There is, to my mind, nothing in the statement of the law by the Privy Council in Barry v Butlin which suggest that ... the standard of proof required in relation to knowledge and approval in a probate case is other than the civil standard-that is to say, that the court must be satisfied, on the balance of probability, that the contents of the will do truly represent the testator's intentions." (para. 70)

Longmore, LJ said at page 1123:-

"The vigilance and jealousy of the court is directed to being satisfied that the testator did know and approve of the contents of his will; no less but also no more." (para. 78)

1052. All the passages quoted above from Fuller v Strum again demonstrate convincingly that the Dispel SC Principle is applicable in probate law to only the defence of Lack of Knowledge and Approval and is applicable to no other issue and certainly not to the issue of Execution.

1053. The above review of the relevant authorities can leave no doubt that Mr. Thomas is correct in his submission on the plea of Dispel SC Principle being not applicable to Execution. I will now consider this question from the point of view of principle.

1054. The history of probate litigation provides lessons as to why the Probate Court has developed the judge made law in the way it did. Undue influence, fraud and forgery are all serious charges where the burden of proof is on the challenger to the will, not on the propounder. The burden of proof on Execution, testamentary capacity and Want of Knowledge and Approval is on the propounder of the will. Proof of Execution is required by statute to be on the propounder by reason of the provisions in the various Wills Acts. Testamentary capacity does not normally create problem for the propounder as it is presumed that all men have testamentary capacity unless there exist facts which might cast doubt on such capacity. Knowledge and Approval is presumed from the fact of execution by the testator and it is only when there are facts giving rise to suspicious circumstances which create doubt for the Court as to whether the testator knew and approved the contents of the will, that the principle of Lack of Knowledge and Approval comes into play requiring the propounder to dispel suspicion and to show that in all probabilities, the testator did know and approve the contents of the will. The principle established by Barry v Butlin and followed by Tyrell v Painton therefore enables the burden of proof on those alleging no knowledge or approval to be reversed so that the burden falls onto the propounder, upon the challenger to the will establishing Suspicious Circumstances which create suspicion of the Court that there was Lack of Knowledge and Approval. However, the propounder will only assume that burden when there were proven facts amounting to sufficient Suspicious Circumstances as to create real suspicion.

1055. What then can be the reason for extending the Dispel SC Principle to Execution. The burden of proof on Execution is already on the propounder. The statutory requirements have to be met by the Propounder. If the Propounder does not prove the statutory requirements, then the Will must fail. But if the Propounder meets the statutory requirements, is there any legal principle which imposes upon the propounder an extra second burden to dispel suspicion created by suspicious circumstances. There is no necessity to impose this second burden for Execution, which will place greater difficulty on legitimate claimants to the will and give licence to vigorous challenger to frustrate the intention of the deceased testator.

1056. What is even more objectionable in principle is that there is no logical probative connection between suspicious circumstances and Execution. Execution is an external act of the testator and of the witnesses of attestation as opposed to Knowledge and Approval of the will being the internal mental intention of the testator. The suspicious circumstances affecting Knowledge and Approval all have a directly probative relation with the fact of whether the testator really intended what was in the will, even though he signed. There can be very few suspicious circumstances which has a directly probative relation to execution. Take for example the contents of a will whereby half of the estate was given to a casual friend. Independent persons witnessed the will and execution is proved convincingly. The fact that the will gave half of the estate to a casual friend (creating alleged suspicious circumstance) has no probative relation to the external act of Execution although it might have a probative bearing on whether the testator really knew what was in the will and really intended to give half of estate to such casual friend.

1057. Finally in principle, it would be wrong to add to the certain and well known requirements of Execution laid down by statute, an uncertain and novel requirement of Dispel SC Principle applicable to Execution. There is no reason or logic or necessity to introduce such a rule and the legal authorities clearly show that the probate law has not found it necessary to rely on such a rule.

1058. It is therefore clear on authority and on principle that there is no basis in law for the Dispel SC Principle to be applicable to Execution. As far as I am concerned, the 9 Suspicious Circumstances were the most unfortunate creation of Yam J. and Mr. Edward Chan, SC (Leading Counsel for the Father). The 9 Suspicious Circumstances requiring to be dispelled were wholly irrelevant to Execution and to Forgery and had bedevilled both the Trial below and in the Court of Appeal.

1059. Mr. Tang bravely sought to defend the 9 Suspicious Circumstances as being relevant to Dispel SC Principle applicable to Execution. These 9 Suspicious Circumstances were neither pleaded nor relied upon in the Court below as being relevant to the Dispel SC Principle applicable to Execution. The Judge did not find against Execution by reason of the application of the Dispel SC Principle. On any basis, it will be wrong in the Court of Appeal to affirm the Judgment below on a newly advanced Dispel SC Principle applicable to Execution.

1060. In the circumstances, I regret I am unable to agree with the conclusion of Yeung JA on the relevance of the Suspicious Circumstances to Execution (Para. 324).

(IV) APPLICATION OF PRESUMPTION OF OMNIA PRAESUMUNTUR RITE ESSE ACTA

1061. The presumption expressed by the old Latin maxim of Omnia Praesumuntur Rite Esse Acta [Co. Litt. 6] (all things are presumed to be correctly and solemnly done) as applied in probate law (it will be presumed in favour of an apparent testamentary document that all formalities have been satisfied) was the subject of considerable submission in the Court of Appeal but nowhere featured in the Judgment of Yam J. Presumption of Omnia Praesumuntur Rite Esse Acta (hereinafter referred to shortly as "Omnia") did not feature in the Judgment because Execution as an issue was not considered by the Judge and furthermore because the Judge did not address himself to the unique statutory provision of section 5(2) of the Wills Ordinance and its impact on the presumption.

1062. Halsbury's Laws of England, 4th edition, Volume 17(2) said this of the Presumption of Due Execution at para. 304:-

"The principle omnia praesumuntur rite esse acta applies where the will is regular on the face of it, with an attestation clause and the signatures of the testator and witnesses in their proper places. .... It may be rebutted by evidence of the attesting witnesses or otherwise, but the evidence as to some defect in execution must be clear, positive and reliable, since the court ought to have the strongest evidence before it believes a will with a perfect attestation clause and signed by the testator, was not duly executed...."

1063. Williams, Mortimer on Executors 18th edition said this at para. 12-25 on "Presumption of the due execution of a will":-

"There is no absolute necessity for positive evidence of due execution in order to enable the court to pronounce for a will. The court will take into account the circumstances and judge from them collectively whether or not there was due execution ......

The presumption that everything was properly done (Omnia rite et solemniter esse acta) arises whenever a will, regular on the face of it and apparently duly executed, is before the court, and amounts to an inference, in the absence of evidence to the contrary, that the requirements of the statute have been duly complied with. The presumption applies with more or less force according to the circumstances of each case. When there is a regular attestation clause, with the name of the two witnesses appended thereto, leading to the conclusion that the will was duly executed by a person who knew the requirements of the Wills Act, the principle applies directly; for the court will assume that no one would have signed his name to the statement contained in such clause unless it were true.......

Probate judges have been long accustomed to give great weight to the presumption of due execution arising from the regularity, ex facie, of the testamentary paper produced where no suspicion of fraud has occurred. And Fry, L.J. in Wright v Sanderson stated that in doing so they have, in his opinion, acted rightly and wisely. It seems, therefore, that it is not strictly necessary to prove a testator's intention to make a will before the presumption applies. Indeed, in many cases nothing whatever is known about the testator's testamentary intention at the date when he placed his signature on the paper in question."

1064. In probate law, the application of the presumption of Omnia by the Probate Court is a matter of long history and of high acceptance and has been applied by the most eminent of Judges in a large variety of probate cases.

1065. It was applied by Dr. Lushington in Burgoyne v Showler (1844) 1 Rob. Ecc. 5 (strong case of application at 13), by Sir John Wilde in Vinnicombe v Butler (1864) 164 ER 1400 (presumption not defeated by defective memory), by Lord Penzance in Wright v Rogers (1869) LR 1 PD 678 ("presumption of law largely in favour of due execution" and requires the strongest evidence before it believes that a will ... was not duly executed"), by the Court of Appeal in Wright v Sanderson (1884) 9 PD 149 ("judges who have presided over the Court of Probate have long been accustomed to give great weight to the presumption of due execution"), by the Court of Appeal in Harris v Knight (1890) 15 PD 170 (source of classic dictum by Lindley), by President Jeune in Re Peverett [1902] P 205 (applied even though document was informal and irregular), by Faulks, J. in Re Webb [1964] 2 AER 91 (applied based on draft of will, even though will destroyed) and by Sachs, J. in Re Denning [1958] 2 AER 1 (applied because only practical reason name on will was for purpose of attesting the will).

1066. Lindley L.J. explained the maxim in Harris v Knight (1890) 15 P 170 at page 179 thus:-

"The maxim "Omnia praesumuntur rite esse acta" is an expression, in a short form, of a reasonable probability, and of the propriety in point of law of acting on such probability. The maxim expresses an inference which may reasonably be drawn when an intention to do some formal act is established: when the evidence is consistent with that intention having been carried into effect in a proper way; but when the actual observance of all due formalities can only be inferred as a matter of probability. The maxim is not wanted when such observance is proved, nor has it any place where such observance is disproved. The maxim only comes into operation where there is no proof one way or the other; but where it is more probable that what was intended to be done was done as it ought to have been done to render it valid; rather than that it was done in some other manner which would defeat the intention proved to exist, and would render what is proved to have been done of no effect."

Lopes, L.J. said this at page 183-4:

"..... What is the fair and probable inference to be drawn from all the circumstances of the case. The presumption "Omnia rite esse acta" is an inference of fact-an inference of reasonable probability ...... (pp. 183)

The inference to be drawn in cases of this kind depends upon a number of circumstances peculiar to the cases in which they arise, and the presumption "Omnia rite esse acta" applies with more or less force according to the circumstances of each case. In every case of this kind the Court should be influenced by a desire that the intention of the testator should not be frustrated, when the execution of the testator is sufficiently proved, and the will, on its face, complies with the requirements of the statute. (pp. 184)"

1067. Sir Anthony Mason in the property case of Leung Kwai Lin v Wu Wing Kuen [2001] 1 HKLRD 212 said this of the maxim Omnia at page 223:-

"In the application of the common law maxim Omnia praesumnuntur rite esse acta to the due execution of wills lost or destroyed, significance has been attached to the fact that the will "on its face" complies with the requirements for due execution (Harris v Knight (1890) LR 15 PD 170 at 184, per Lopes LJ; see also Re Phibbs [1917] P 93; Re Webb [1964] 2AER 91 at 9, per Faulks J.) In other words, the common law maxim has been applied to a will which has been lost or destroyed where one of the circumstances has been that the will appears on its face to have been duly executed...

On the other hand, the maxim Omnia praesumuntur rite esse acta is applied, as demonstrated by the authorities already cited, in the light of all the circumstances of the case. So much emerges from the judgments of Lindley and Lopes LJJ in Harris v Knight (1890) LR 15PD 170 at pp. 179, 183-4."

1068. Having regard to the long history of the application of the Presumption of Omnia, was there a different application of the Presumption in Hong Kong under section 5 of the Ordinance. Doubt was expressed as to whether the presumption applies to Chinese wills because of the weaker statutory safeguard under section 5(2). I regret I am unable to share this doubt. The presumption is that there is regularity and that the statutory requirements are complied with and this presumption was applied repeatedly in probate because (a) of the inherent probability that events occurred regularly rather than irregularly and (b) that teststors generally take steps so that the law is complied with rather than the law is not complied with and (c) that the law has a strong inclination not to frustrate the intention of testator. If the statutory requirement under section 5(2) is low, that is no reason either in principle or logic for the presumption not to apply. If the application of presumption is to depend on the degree of formal requirements for execution demanded in law, then there will be uncertainty whenever there is any change in the law on execution (for example the position now in Hong Kong under the changes brought about by the Wills (Amendment) Ordinance 1995) and there will be the most bizarre situation of the Presumption applying in respect of section 5(1) wills but not in respect of section 5(2) wills.

1069. There is in addition to the above consideration, a good reason why it is wrong for the court to refuse to apply this well known probate principle of Presumption. The legislature in Hong Kong has seen fit to enact section 5(2) of the Wills Ordinance giving a much lower legal formal requirement for Chinese wills than wills under section 5(1). By not applying what is accepted probate law of Presumption, the court will in fact be frustrating the legislative intention set out in section 5(2) of the Ordinance. In the case of Chinese wills, it should be easier for the Propounder of the Chinese wills, with the appropriate application of the Presumption, to establish the validity of the Chinese wills which are intended after all, to be made informally and to be given recognition and effect easily, specially in the light of the Chinese traditional distaste for litigation.

1070. I therefore proceed on the basis that the Presumption of Omnia is applicable to Chinese wills. The question then is whether in the light of all the circumstances of the case, the presumption of Omnia does apply. Mr. Thomas says it applies and Mr. Tang disputes this. They each put forward somewhat different facts and what is more important their different inferences drawn even from the same facts as being worthy of consideration for the application or non-application of the Presumption of Omnia.

1071. It seems to me that a proper consideration of the various facts said to be relevant in the circumstances would be essential for the determination of whether the Presumption of Omnia applies in this case. What is important is to bear in mind is that the facts in support or in opposition to the application of the Presumption should be relevant to and probative on the issue of Execution, namely on the reasonable probability that what was required to be done for Execution under section 5(2) of the Ordinance was in fact done.

(1) 1990 Documents is formally testamentary

1072. The contents of Document A, B and C are clearly formally testamentary. Document A, by comparison with Document B and C, shows a more formal testamentary intention. It is headed Will. The word witness appears. It has a date of 12th March 1990. In the word of Lindely, L.J. in Harris v Knight it shows an intention to do a formal act.

(2) Home made 1990 Documents in Chinese of Chinese testator

1073. The 1990 Documents is clearly the Will of a Chinese Testator (Teddy Wang) which was written wholly in Chinese. It was home made but that would be normal for wills in Chinese in Hong Kong. The 1990 Documents therefore clearly shows an intention to do the formal act as provided in Section 5(2) of the Wills Ordinance relating to Chinese Wills.

(3) Childless Wife Beneficiary of the 1990 Documents

1074. The fact that the Wife was the beneficiary and the only beneficiary under the 1990 Documents was the subject of intense debate during the hearing of this Appeal. There is no dispute that the Wife was childless at all times, including in March 1990 when the 1990 Documents were made. The Father's lawyers sought to draw an adverse inference from the Wife being the sole beneficiary and the Judge relied on the Wife being the sole beneficiary under the heading "No reason to change intention" as being the 1st of his 9 Suspicious Circumstances (paragraph 5.12 to 5.19). Mr. Thomas submits that the Wife being the sole beneficiary is not only positively in support of Execution but is also positively in support of the application of the Presumption of Omnia.

1075. When a man dies is it more natural to leave his estate to his aged father or is it more natural to leave his estate to his wife and does it make any difference if the wife is childless. But for the extraordinary way this case had been argued by the Father, the answer would be obvious.

1076. When a man marries, he makes promises (if in church) as part of his marriage vows to his wife to share his worldly goods and to look after each other (through sickness and health) until death. The wife shares his bed, shares his home, shares his life and bears his name. They live together day in and day out and there is no relationship more intimate or more binding than that between a man and a woman. The wife is normally the first person who observes if he has a cold or an illness and is the first person who knows if something is wrong. No father can ever be in that position.

1077. The wife is therefore not merely a worthy beneficiary of the husband's estate as conceded by Mr. Tang but is the most natural beneficiary of the husband's estate. In the very old days in England, prior to the William & Mary Act, a man's testamentary disposability is limited to only one third of his estate, with the wife being entitled in any event to one third and the children entitled in any event to one third. European continental laws such as that prevailing now in Italy and France still provide that a wife is entitled to at least a minimum percentage of the estate.

1078. The intestate law of Hong Kong provides that the wife is entitled to half of the estate (subject to the initial sum), if the husband is survived by the wife and children of the marriage and half of the estate (subject to the increased initial sum) if he is survived by the wife (with no children) and his parents. The fact that the marriage is childless does not point to the wife getting less but points to the wife getting more. The above is what is normal for marriages in general and that normal situation is a strong fact which points to the probability or the presumption that the 1990 Documents were properly executed in favour of Nina.

(4) Wife of long marriage in loving relationship with Teddy

1079. By 1990, Teddy had been married for 35 years to Nina. It is a long marriage. The marriage hit its lowest point in 1968 when he suspected Nina of infidelity and made his 1968 Will in favour of the Father. Nina however was forgiven although it is impossible to know the full intimate private details between a man and his wife. To all outward appearances, there was a strong loving relationship between Teddy and Nina. They were always seen together and they were seen often to hold hands. The marriage not only survived the suspected infidelity but had apparently grown stronger after the trouble of 1968. The marriage survived the first kidnapping of Teddy and the riding accident of 1990 when Nina was at his hospital bedside afterwards. The fact that the marriage did not produce any children might have caused the bonds of matrimony to grow stronger because each spouse relied more on each other as there was no distraction made by the demands of children. It was certainly an intimate marriage as described by his godson Mr. Anthony Cheung. That intimate and loving relationship is another strong factor in favour of the regularity and probability of Execution of the 1990 Documents.

(5) Wife was business partner & Chinachem was their baby

1080. What is however the most unique aspect of the relationship between Teddy and Nina is their long-term close business relationship. There is no doubt that Nina was a close working partner of Teddy in Chinachem, the huge property development business which they had built up over some three decades. The Father was not involved in that property business empire in anyway. Nina was the junior partner and the only partner of Teddy in Chinachem and they spent their working and waking lives on Chinachem, this "baby" of theirs. It is therefore not only natural but the only logical act of Teddy for the long-term continuation of the life of their joint "baby" to give all he has to Nina by the 1990 Documents. This is another powerful fact in favour of the regularity and probability of the Execution of the 1990 Documents.

(6) Previous 1960 & 1968 Wills and No Need to change by 1990 Will

1081. Mr. Tang placed weight on the previous two wills and the fact that there was no change of will by Teddy after 1968 and no reason for him to change his 1968 Will. The 1960 Will gave 50% of the estate to Nina. The 1968 Will gave all to the Father. The 1968 Will was made in circumstances of anger and betrayal when Teddy thought that Nina was unfaithful. It was not a will of a husband in a normal situation. By 1990, some 20 years later, the matrimonial relationship was totally different and the question which can be asked is if the 1968 Will was not made, would Teddy have provided a will in 1990 giving everything to the Father or would he, as one would normally expect a loving husband to do, give all to his wife. The question has only to be asked for the absurdity of the argument to be exposed. The Judge got it upside down and did not even regard it as the most natural act for a loving husband to give all to his wife. Once Mr. Tang concedes that Nina as the wife is a worthy subject of his estate, then it must follow that there is nothing suspicious about his giving all to his wife. It is the most natural act which a loving husband can make and it powerfully supports the presumption of regularity and probability of the 1990 Documents. The fact that subsequent to 1968 Will, Teddy did not seem to have made another will until 1990, is not something which points probatively to anything. One simply does not know why he did not make any will between 1968 and 1990. The fact that he did not do so, has no probative connection to whether he did or did not make the 1990 Documents.

(7) Tse Signatures on 1990 Documens (ex facie)

1082. In each of Document A, B, C and D of the 1990 Documents, there appeared the signature of Tse, the family housekeeper. In Document A, with Tse's signature appearing at a position just next to the word "Witness" it was clear that Tse was signing as witness of attestation of Document A. The position where Tse's signature appeared between pencil crosses in Document B, C and D and the context of the documents also suggest that the Tse's signatures were made as witness to the signing by Teddy of each document. The appearance therefore of Tse's signature in the 1990 Documents give strong support to the regularity and probability of the Execution of these Documents.

(8) Tse as Witness of attestation

1083. Tse was a trusted housekeeper of the Wang family in 1990. By 1990 he had served a combined total of some 16 to 17 years in the domestic household of Teddy and Nina and as the personal office messenger/servant of Teddy. A man working in such close proximity to both Teddy and Nina would be privy to all the intimate details of their lives, both domestic and business. To be able to work, for so long, in such close proximity of two difficult and successful persons such as Teddy and Nina, Tse must have been a well-trusted servant of ability, intelligence and discretion (notwithstanding what Teresa said in evidence). It was said that it would be out of character for Teddy to ask Tse to witness his 1900 Documents Will. This seems to me to be perverse with the reality when it was clear that Teddy chose to keep Tse as the most intimate staff for some 16-17 years. If Tse was indiscreet, it would be inconceivable that Teddy would have kept him on not only in the house but later in the office. It is precisely because Teddy trusted very few persons (I believe after the row with Donald Cheung he probably did not even trust lawyers) that it would be wholly in character for him to trust Tse. The best secrets of very important persons tend to be known to their long serving office or domestic staff, whether they be personal maids, drivers, personal assistants or secretaries. Many of the most private documents of persons in high position are witnessed by their staff close to them. So instead of the Judge finding it suspicious that Tse was a witness (see SC VII), on true analysis Tse's signature (as long serving staff of Teddy) on the 1990 Documents as witness should powerfully confirm and add to the presumption of regularity and probability of Execution.

(9) Tse's Statements and Affirmations of Witnessing Teddy's Execution

1084. Tse made Statements and Affirmations to two different solicitors on 6th and 9th September 1999 that he witnessed the signing of the 1990 Documents by Teddy. But instead of these being accepted as the most convincing evidence of Execution of the 1990 Documents by Teddy, the Judge turned this most important evidence on its head and regarded them as Suspicious Circumstance VII. Was injustice done to both Tse and Teddy. In my view the lack of critical analysis of this crucial evidence has caused a totally erroneous conclusion to be drawn.

1085. Tse's Statement written out by him in the presence of Alex Lam, the solicitor dated 6th September 1999 at 8.30 p.m. was in a neat Chinese handwriting and expressed in Chinese language which suggested that the writer was a man of some literary skill and also some intelligence. (It was agreed at the appeal hearing that Tse's Chinese handwriting was far superior to that of the two solicitors). There was no evidence that Tse did not have the literary skill or the intelligence to write out the 6th September 1999 Statement.

1086. What was held against this Tse Statement of 6th September 1999 and by extension also against the Tse Statement of 9th September 1999 is the allegation that it was too good to be true. But that is speculation of the worst kind because it was without foundation and made when it was not capable of being answered by a dead man. What is held against Tse and what he said in his Statements and his Affirmations is that what he wrote out (on 6th) and said (on 9th) were too good to be true (too close to the Wife's case) and that therefore it must have been untrue. The reasoning unexpressed presumably was that the Wife's case was false and it could not be a coincidence that what Tse said was very similar to what the Wife said and very supportive of the Wife's case and that therefore what Tse said must also be false. There was no analysis of why the contents of what Tse wrote out and sworn to was inherently or probably untrue or why it was either textually improbable or unreliable.

1087. It is of course beyond doubt that these Statements and Affirmations of Tse were made on the 6th and 9th September 1999 and were made to and in the presence of Alex Lam and Alan Leung respectively. There is absolutely no reason for these two professional men to tell any untruth about what happened on the 6th and 9th September. It is possible that Alex Lam as a nervous witness might have got some of the details of the giving of the Statement and the making of the Affirmation not totally accurate when he was giving evidence. But that is not an adequate reason to dismiss the Tse's Statements and Affirmations. They were there and could not be ignored and that evidence must be properly weighed.

1088. The heart of the case on Tse Affirmations and Statement is whether the contents of what Tse said are true and is there any good valid reason to believe that what Tse said was not true. The Judge for example never addressed himself the question as to the probative value of the what was written by Tse for Alex Lam (as part of the Attachment to his Statement) on pages C273, 274, 275 and 276 saying in each case that the signature of Tse Ping Yim identified with an arrow was his signature and then signed his name Tse Ping Yin. The Judge never asked himself the inherent improbability of forgery of Tse signatures, identified by Tse as being his signatures in four documents which Tse specifically recognised.

1089. Apart from the assertion that the Tse Statements and Affirmations were too good to be true, there was advanced another reason for dismissing totally Tse Statements and Affirmations, namely the second departure of Tse from Hong Kong (24th September 1999) the day after the Father obtained leave from the Judge to report forgery of the 1990 Documents (23rd September 1999). There was no evidence that Tse knew about the Judge's Order of the 23rd September 1999. There was no evidence of the circumstances of Tse's departure from Hong Kong or whether prior arrangements had been made by Tse before the 23rd September 1999 to leave Hong Kong on 24th September. On such a slim foundation, a case of unalleged and unspoken conspiracy was made against Tse, a dead man. The thrust of that allegation is that Tse knew he made untrue Statements and untrue Affirmations on 6th and 9th September 1999 and that he was deliberately avoiding the police as soon as he heard about the Judge giving permission to the Father to report forgery to the police.

1090. The seriousness of the unspoken allegation deserves proof of an appropriately high standard (In Re H [1996] AC 563, 586-7). The adverse inference sought to be drawn (by the Judge and by the Father's lawyers) against Tse and against his Statements and Affirmations is not warranted in law and is wholly unjust in the circumstances. It is regrettable that the serious allegation was made by way of unspoken insinuation, driven by a desire for a result in favour of the Father. The evidence in this Trial does not support such a result.

1091. The third and final inference sought to be drawn against the Tse Statements and Affirmations is that Tse did not volunteer to the Police in May 1990 when he was interviewed, that Tse was witness to the 1990 Documents. The two events were wholly unrelated. It cannot even be legitimately suggested that Tse was aware of the connection between his witnessing the signing of the 1990 Documents and the kidnapping of Teddy or that there was an obligation for him to volunteer the information of witnessing the 1990 Documents to the Police even if he was aware of any connection. Assuming that Tse knew the connection but kept his silence, was that not to be expected of a faithful long service family housekeeper. How could his silence create the compelling inference that there was no truth in the Tse Statements and Affirmations.

1092. Stripped of the aforementioned three unsustainable objections to Tse's Statements and Affirmations, there is nothing left of the Father's resistance to Tse's evidence, set out so clearly in Tse's Statements and Affirmations. What the Judge ought to focus on is the inherent probability of the contents of Tse's Statements and Affirmations.

1093. The inherent probability is that Tse was a trusted housekeeper of Teddy and was asked to witness the signing by Teddy of the 1990 Documents which he did (as evidenced by the signature of Teddy and the signature of Tse on the 1990 Documents) and that what Tse said in his September 1999 Statements and Affirmations were all true. (including the fact of holding-hands domestic bliss, which was naturally stated, in the context of Tse being treated as part of a happy family). The inherent improbability is that Tse was part of a conspiracy involving the creating of the false 1990 Documents and making subsequent false Statements and Affirmations in support of the false 1990 Documents.

1094. Tse had no reason to lie and to commit not only the crime of perjury but also the crime of participation in the making of a forged will. It was never openly suggested (or backed up by any cogent evidence) that Tse was part of a conspiracy. It was not open to the Judge so to conclude either expressly by open express finding to that effect or covertly by insinuation and by implication.

1095. The Tse Statements and Affirmations of September 1999 being unimpeachable can therefore only add to the circumstances leading to the Court to draw the conclusion that the Presumption of Omnia should be applied in this case.

(10) Tse's Death and Presumption

1096. Tse died in December 1999 and was therefore incapable of giving evidence. His signatures on the 1990 Documents create the Presumption (even in the absence of the Statements and Affirmations). His Statements and Affirmations (confirming that that he saw Teddy sign and that the signatures on the 1990 Documents are his) are further powerful evidence which cannot be ignored and which add to the Presumption. The death of the attesting witness is in probate law a classic case of the application of the Presumption. The strength of the Tse Signatures on the 1990 Documents and of Tse's Statements and Affirmations did not diminish on his death but they strengthen on his death by reason of the application of the Presumption.

(11) Teddy Signatures on 1990 Documents (ex facie)

1097. Each of the Document A, B, C and D of 1990 Documents bears on the face of the Document the apparent signature of Teddy Wang and in the context of the documents, the signature of Wang Teh Huei was clearly made as testator, signing in his testamentary capacity.

1098. As Sir Anthony Mason said in Leung Kwai Lin v Wu Wing Kuen in relation to the application of the Presumption that "significance has been attached to the fact that the will "on its face" complies with the requirement for due execution." On its face, the 1990 Documents being a Chinese will bearing the apparent signature of Teddy Wang, a Chinese testator comply with the Execution requirement of section 5(2) of the Ordinance and the Omnia maxim should be applied.

(12) Other Miscellaneous Facts

1099. A number of other facts were also mentioned during the hearing to support the issue of Execution. The fact that in the months prior to March 1990 Teddy and Nina were actively together engaged in tax and trust planning suggest that each would give to the other the respective assets, does not directly bear on the issue of Execution although it suggests that Teddy would have the inclination to make the 1990 Documents. Similarly the riding accident in March triggering off possible post-mortem thought and therefore inclination to make the 1990 Documents would suggest the probability of Teddy causing the making of the 1990 Documents although it does not bear directly on the issue of Execution.

Conclusion on Application of Presumption

1100. A detailed consideration of the facts and circumstances described above strongly support the application of the Presumption. This is hardly surprising having regard to the extremely low threshold for validity of Chinese wills and the strong case of this particular Chinese Will contained in the 1990 Documents, giving his estate to his Wife, which was signed apparently by the Chinese testator Teddy and which was also witnessed apparently by Tse as evidenced by Tse's signature.

1101. If the Presumption applies, where stand the contrary evidence against Execution, such as the handwriting evidence pointing to forgery and a number of adverse inference evidence pointing against the 1990 Documents being validly executed The Presumption of Omnia is strong and tenacious once applied, but may be rebutted by evidence of the attesting witness or otherwise, but the evidence of some defect as to execution must be "clear, positive and reliable" (see Halsbury's Laws of England, Vol. 17(2), para. 304 and Note 6).

1102. It was argued that because of the allegation of forgery, the Presumption will not apply. This is not supported by authority. All the old cases suggest that when there is contrary evidence against execution they are weighed against the Presumption and it is a matter for the Court as to whether the Presumption is rebutted and there is no question that the Presumption does not apply. In my judgment, neither the fact of allegation of forgery nor the evidence of forgery can in law render the Presumption not applicable. The difference made by the evidence of forgery, is that if they are sufficiently strong they might be able to rebut the Presumption, in which case Execution will fail to be proved.

1103. It was suggested that the fact of a home-made will and the particular contents of the 1990 Documents point to the non-validity of the 1990 Will. These include the fact that a careful businessman like Teddy would not expose his estate to the possibility of probate litigation by a home-made will, the fact that the 1990 Documents being sloppily written which it was said that Teddy would not tolerate, the fact that the 1990 Documents contained romantic expression which was out of character for Teddy, the fact that there were strange and false expressions relating to Teddy's parent and the Kung family which was out of character for Teddy. In my view, none of these considerations are relevant to the issue of Execution although they might be relevant to the disallowed issue of Lack of Knowledge and Approval. If I am wrong on this and these matters are relevant to the issue of Execution and can be relied upon to rebut the Omnia Presumption, then it seems to me that the probative value of these matters are so slight being in the nature of speculative inferences (having regard to the unpleaded general and imprecise nature of allegations of "out of character" which was not much thought of in Fuller v Strum) compared to the weighty facts giving rise to the strong Presumption, that I will hold that the Presumption has not been rebutted.

1104. So far as the handwriting evidence is concerned, as appears from what I say hereinbelow, I am of the view and I hold that the Handwriting Forgery findings of the Judge were seriously flawed and that he ought to have rejected the allegation of Handwriting Forgery having regard not only to the conflicting handwriting evidence but having in regard to the overall probabilities of the case. In the circumstances, I am of the judgment that the Omnia Presumption should be applied and that Execution has been established by the Wife and that Judgment ought to be entered in favour of the Wife.

(V) WHETHER EXECUTION PROVED, IF OMNIA PRESUMPTION DOES NOT APPLY

1105. Should I be wrong on the application of Presumption, it will then be necessary to consider, in the light of all the relevant evidence on Execution whether Execution has been established on a balance of probability.

1106. Of the 12 factors which I have discussed earlier pointing to the application of the Presumption, most are also relevant to the issue of proof of Execution. In particular I place emphasis on the fact that;-

(1) this was a Chinese Will therefore subject to the low statutory requirement of section 5(2) of the Ordinance;

(2) the signatures of Teddy appeared to be regular on the face of the 1990 Documents;

(3) the signatures of Tse, his long service housekeeper appeared to be regular on the face of the 1990 Documents;

(4) the Wife is the most natural beneficiary, as she was both a wife in a loving marriage as well as the long established business partner of Teddy;

(5) Tse's Statements and Affirmations confirming that Tse witnessed Teddy signing the 1990 Documents.

1107. The Father has put forward various matters which it was said point to the 1990 Documents not being executed by Teddy including in particular the various inferences which ought to be drawn against the Wife such as:-

(1) the Wife failing to give oral evidence confirming that Teddy gave the 1990 Document to her;

(2) the Wife knew contents of the Will;

(3) the Wife's behaviour in late propounding the Will;

(4) the Wife's fabrication of holiday wills;

(5) the Wife's failure to give evidence.

(6) the wife's failure to put evidence before the court of efforts or lack of efforts to trace the unknown K.S.Lee and the unknown writers of the 1990 Documents.

1108. It seems to me that all these matters are wholly irrelevant to the question of execution. In the detailed submission made to the Court on Suspicious Circumstances, Mr. Thomas and Mr. Tang were in considerable disagreement on the relevance of subsequent events to Execution. In so far as subsequent events relevant to Lack of Knowledge and Approval is concerned, both Re R [1950] 2 AER 117 at 121 and Davis v Mayhew [1927] P 264 at 280 point to suspicious circumstances attaching only to the preparation and execution of the will and not to subsequent events. In Chan Sai Lun and Chan Wai Wah [1999] 1 HKLRD 22 Barnett J was of the view at page 291I before quoting those earlier authorities that suspicious circumstances was so restricted in its relevance. In the next section, I will consider the law and its effect by the Wife's silence or choice of not giving evidence where I conclude that no adverse inference can be drawn against the Wife by her failure to give oral evidence.

1109. At the hearing of the appeal, a great deal of time and energy was spent on the question of Execution. Even more time was spent on Forgery. The Trial took so long because of the case of forgery put forward by the Father.

1110. What will strike any reasonable person is how a Court could be misled into accepting what was apparently a very weak case of the Father to pronounce judgment of Forgery both by Suspicious Circumstances and by Handwriting. The case of the Father was apparently weak because there was a 1990 Will giving all to the Wife who was a natural beneficiary specially as she was a loving wife and a long term business partner of Teddy. The case of the Father was weak because the 1990 Documents bore on the face of the Documents not only the signatures of Teddy but also the signatures of the family housekeeper Tse. The case of the Father was weak because being a Chinese Will, only the minimal formal requirement under Hong Kong law was required. The case of the Father was weak because there were not only the signatures of Tse on the Documents but also the Statements and Affirmations made by Tse to two different solicitors affirming that he saw Teddy sign and confirming his signatures on the 1990 Documents. The Father's case was weak because handwriting evidence when contested will rarely if ever prevail over evidence from attestation witness. The Father's case was extraordinary weak when what was purported to be challenged was not only the handwriting signatures of the testator Teddy on the 1990 Documents but also the handwriting signatures of the witness Tse on the 1990 Documents in the face of Tse saying that these were his signatures.

1111. In the annals of probate history it is unheard of for an ancient father to contest a will giving the estate to the wife. It is certainly legal history for a judgment to be entered against the wife on the basis that the wife was not a natural beneficiary and that the will the wife was propounding was a forgery. The extraordinarily large fortune left by Teddy Wang was what caused this case to take its extraordinary course.

1112. In my judgment the evidence as well as the justice of this case, compel this Court to come to the only possible conclusion, namely that the Tse Statements and Affirmations are true and it follows therefore that in the circumstances, there is no room for forgery and the only judgment open to this Court is to pronounce in favour of the Appellant Wife.

(VI) WHAT ADVERSE INFERENCE FROM SILENCE OF WIFE

1113. It was argued for the Father that the failure of the Wife who was propounding this 1990 Will gave rise to an adverse inference which can substantially undermine her case on Execution.

1114. It is first necessary to examine the authorities cited by the parties. In my view, the authorities of Rank Film Distributors Limited v Video Information Centre [1982] AC 380 (per Templeman, L.J. at 423C/F) and Den Norske Bank v Antonatos [1999] Q.B. 271 (per Waller, L.J. at 296A/B) are concerned with discussion of privilege against self-incrimination, namely where a party is obliged and compelled by order of the court to give discovery or evidence but refuses to answer or give the required evidence and claims the privilege of being excused from so doing by reason of fear of self-incrimination. Similarly, the authority not cited to us (but referred to by my Lord Yeung) of R v Boyes (1861) 1 B. & S. 310 (per Cockburn C.J. at 330) relates to the privilege of a witness who was given a pardon in not answering on the ground of incriminating himself. The question facing this court is not whether the Wife was entitled not to answer certain question on the ground of self-incrimination but whether exercising her right not to give evidence, what inference could be drawn from her not giving any evidence.

1115. If the Wife was called at the Trial (whether by the Plaintiff or the Defendant) and during her evidence, she had refused to answer a particular question, then the Rank Film and Den Norske authorities might become relevant for consideration. The Wife simply did not step forward to give evidence, notwithstanding the fact that she was the party propounding the 1990 Will and that she had filed her witness statement in the Probate Action. The question in this situation is what adverse inference if any is to be drawn by her not taking the witness stand, apart from her witness statement not becoming part of the evidence.

1116. Wisniewski v Central Manchester Health Authority [1998] Lloyd's LRM 223 is the authority relevant to the point. After a review of the authorities, Brooke, L.J. said at page 240 in the passage already quoted by my Lord Yeung:-

"From this line of authority I derive the following principles in the context of the present case:

(5) In certain circumstances a court may be entitled to draw adverse inferences from the absence or silence of a witness who might be expected to have material evidence to give on an issue in an action.

(6) If a court is willing to draw such inferences they may go to strengthen the evidence adduced on the issue by the other party or to weaken the evidence, if any, adduced by the party who might reasonably have been expected to call the witness.

(7) There must, however, have been some evidence, however weak, adduced by the former on the matter in question before the court is entitled to draw the desired inference: in other words, there must be a case to answer on that issue.

(8) If the reason for the witness's absence or silence satisfies the court then no such inference may be drawn. If, on the other hand, there is some credible explanation given, even if it is not wholly satisfactory, the potentially detrimental effect of his/her absence or silence may be reduced or nullified."

1117. It is to be noted that the Wisniewski case was concerned with a substantial body of evidence against the hospital that the doctor of the hospital was negligent in the various respects alleged and that it was for the doctor to give evidence to dispute the negligence evidence. The question on appeal was whether the trial judge was wrong, in addition to finding negligence for the plaintiff based on the evidence called by the plaintiff, to draw adverse inference against the defendant for the failure of the doctor to give evidence to answer the criticism made against him. In the Wisniewski case there was already much evidence of negligence against the hospital and the failure of the hospital defendant to call the doctor merely added an adverse inference to the already strong case against the hospital. Applying that Wisniewski case to this Action, the complication we have is that the Wife was, both in the position of the defendant in terms of resisting Forgery (the burden on which lies on the Father) and in the position of Counterclaimant to propound the 1990 Will ( the burden on which lies on the Wife). What adverse inference is to be drawn against the Wife by her failure to give evidence, firstly in relation to the Execution issue and secondly in relation to the Forgery issue.

1118. Proceeding on the assumption that the principles stated by Brooke, L.J. cited above are correct, it seems to me the first question which should be asked is what relevant evidence on execution she was expected to give and secondly what relevant evidence she could give on Forgery.

1119. So far as Execution is concerned, as the Wife did not actually witness the signing of the 1990 Documents, her serious relevant evidence is, as Mr. Thomas submits, really limited to how Teddy Wang gave her these Documents and what he was alleged to have said to her about them. That evidence if accepted would certainly reinforce her case that Teddy acknowledged the 1990 Documents. But it is to be remembered that her case on Execution was based on (1) what are on the face of the 1990 Documents including the testator's signatures of Teddy Wang and the signatures of Tse as witness, (2) what Tse had provided by way of Statements and Affirmations on 6th and 9th September 1999, (3) inherent probability of a loving wife inheriting the estate of the husband and (4) the low probate requirement of Chinese wills under section 5(2). It was therefore in the circumstances reasonable for her legal team and for her to believe that she had a strong case for Execution and there was no necessity for her to take the stand and give evidence to booster a strong case.

1120. So far as the Forgery issue is concerned, as no case of Forgery had been pleaded against her (in fact Forgery was disavowed against her), there was certainty no reason whatsoever for her to give evidence on Forgery. This is what distinguishes this case from such authorities considered by Brooke, L.J. at pages 237 to 240, in which there was good prima facie evidence against the defendant who was expected to answer the criticism made by the plaintiff.

1121. The dispute between the parties here is whether in the circumstances, there was a reason for her not to give evidence which satisfies the court and secondly if the explanation for her silence is not totally satisfactory, to what extent the adverse inference to be drawn is to affect the case on Execution.

1122. In my judgement the Court can legitimately take into account the following as being valid reasons for her to not take the stand:-

(1) the Court can take note that from the beginning she was neither silent nor intended to be silent. In fact the Wife filed Affidavits and Witness Statements;

(2) the Trial proceeded with strong resistance by the Father to the validity of the 1990 Will and with serious allegation of Forgery in respect of the signatures of Teddy and of Tse on the 1990 Documents;

(3) the bitterness of the dispute between the parties intensified with the progress of the 172 Days Trial spanning over one year and with vicious personal attacks (much of it wholly unjustified) being made against her;

(4) if driven to give evidence, the whole private life of the Wife would not only be probed but be explored in the greatest graphic detail by the Father and his family with all the attendant publicity;

(5) the Wife was a high-power business lady running a very large business empire and a protracted time spent in the witness box would be detrimental to her and her companies;

(6) the manner the Trial developed, resulted in the factual witnesses for the Defendant giving evidence last and thereby giving rise to the assessment by the Wife's legal team as to the necessity and wisdom of her giving evidence;

(7) the legal team of the Wife could reasonably take the view that the Wife had a strong case on Execution and that there was little to be gained by the Wife taking the witness stand;

(8) prudence would dictate that in the face of the serious charge of Forgery which had been reported to the Police by the Father (with the Police investigating the matter actively), the Wife should take the safest course which would expose her to the least overall danger.

1123. In the circumstances, there are certainly good and legitimate reasons for the Wife not to take the witness stand. It follows therefore that in my judgment no adverse inference is capable of being drawn against the Wife by her not giving evidence.

1124. Assuming I may be wrong and that she did not have a fully legitimate excuse not to give evidence, what difference to the case of Execution is made by her not giving evidence. Obviously the Court could not place the full weight on what she said in her Affidavit about the 1990 Documents being given to her by Teddy Wang. But can the Court go beyond that and hold that her not giving evidence will result in an adverse inference and if so, what is the compelling adverse inference of fact which will be drawn in the circumstances. It certainly could not be that there was no Execution. Nor could it even be that she knew there was no Execution. The problem with adverse inference is the difficulty of being able to formulate or state the precise fact, which compellingly can be adversely inferred by the silence of the Wife.

1125. I used the word compellingly because it seems to me that where adverse inference is to be drawn against the Wife in these circumstances, the precise fact to be adversely inferred must be compelling so as to be able to carry the necessary weight on the Execution issue, specially if when that adverse inference might make a material difference to the outcome on the Execution issue.

1126. The large categories of what I might call peripheral circumstantial evidence which Mr. Tang said would be suitable subjects of the wife's oral evidence (see Para. 591 to 605 herein) illustrates the difficulty and dangers inherent in drawing adverse inference from the non giving of evidence by a witness. Is the Court for example to infer that because the Wife did not give evidence that, arising from the allegation that the Wife did not inform the Police about the 1990 Documents (Para. 593) :-

(1) that she knew there was a connection between the 1990 Documents and the kidnapping, or

(2) that she knew that informing the Police about the 1990 Documents could have provided police with clues for the kidnapping crime, or

(3) that she knew that there was no proper Execution of the 1990 Documents or

(4) that she knew that at the time of the kidnapping the 1990 Documents did not exist and had not been handed to her.

1127. The Wife is entitled to rely on a strong case of Execution where the Tse Statements and Affirmations can link the 1990 Documents to Teddy Wang and to his signing them in 1990. Of course she carried the burden of proof on Execution but in modern probate litigation, I certainly take the view that she can satisfy the Court on the balance of probabilities on the materials available to the Court without having to walk into Lion's Den for a marginal gain by giving her oral evidence (on Teddy giving her the 1990 Documents).

1128. Throughout the hearing heavy emphasis was placed on what I might call the various religious and moralistic dicta from Re Fuld. In my view, to follow on the sentiment expressed by Chadwick, L.J. in Fuller v Strum, the modern approach to probate litigation would be better served by a consideration on a balance of probabilities (fully taking into account the burden of proof), firmly grounded on legal principle, applied to the concrete evidence directly relevant to the issue, rather than motivated by ecclesiatical conscience (an expression much repeated) which like the proverbial Chancellor's foot is less susceptible to reviewable measurement or standard.

1129. In my judgment, even if any adverse inference can be drawn in this case, the adverse inference will be so imprecise and insufficiently compelling as to make no difference to the outcome of the Execution issue. I am therefore of the judgment that the failure of the Wife to take to the Witness stand does not make any difference to the conclusion I have earlier reached or to the outcome on Presumption or Execution.

1130. Finally for the sake of completeness, I would just mention that in my view, Hart v Dabbs & Dabbs [2001] WTLR 527 does not add anything to the proper application of the law on adverse inference. Hart is a case of Lack of Knowledge and Approval and because of the necessity to dispel suspicion, the failure to give evidence was argued to have an impact on the failure to dispel suspicion. Our case is not about Lack of Knowledge and Approval and there was no case of failure to give evidence to dispel suspicion.

(C) PROOF OF FORGERY (HANDWRITING & OVERALL PROBABILITIES)

1131. The central dispute and what took a great deal of the court time at the Trial was the allegation of handwriting Forgery of not only Teddy's signatures in the 1990 Documents but also the Tse signatures in the 1990 Documents. What had given further complication was the wholly unnecessary and unpleaded case which the Judge found that the Wife was the author of Document A because of the handwriting evidence. I agree with the Judgment of Yeung JA and of Yuen JA that the finding of the Wife being the author of Document A could not stand. I will not burden this Judgment with my separate and additional reasons for such view. I will confine myself to where I differ from Yeung JA and Yuen JA on the substantial dispute on the Forgery issue. The Judge found in favour of the Father on the Handwriting Forgery issue based on the evidence of the three handwriting experts of the Father (GL [Gus Lesnevich], Tsui and Cheng) and rejected the evidence of the Wife's handwriting expert (Jia). Before I turn to the fundamental errors including errors of methodology, a few preliminary observations touching on comparison handwriting expert evidence will serve as a useful introduction to the discussion on this part of the appeal.

(I) PRELIMINARY OBSERVATION ON COMPARISON HANDWRITING EVIDENCE

1132. In the Judgment, there was no analysis of the comparative value of evidence from an attesting witness of execution of his signing of a document as opposed to that from handwriting expert opinion evidence to the contrary. Nor was there any assessment in the Judgment of the weight, which a Court should place on heavily contested expert opinion.

1133. On the value of expert opinion generally, this was said in Phipson on Evidence, 15th edition, at para. 37-51:-

"Formerly, the testimony of experts was often considered to be of slight value. Even today, the principles governing the admission of their evidence and judicial attitudes to witnesses exhibit traces, and sometimes more than traces, of this old suspicion. It is proverbial that expert witnesses are, perhaps unwittingly, biased in favour of the side which calls them, as well as over-ready to regard neutral facts as confirmation of pre-conceived theories; moreover, support or opposition to give hypotheses can generally be multiplied at will. Indeed, where the jury accept the mere untested opinion of experts in preference to direct and positive testimony as to facts, a new trial may be granted."

1134. In Re Dyce Sombre (1849) 1 Mac. & G. 116, Lord Cottenham said this at 128:-

".... I have seen enough of professional opinions to be aware that in matters of doubt upon which the best constructed and best informed minds may differ, there is no difficulty in procuring professional opinions upon either side."

In The Tracy Peerage (1843) 10 Clark & Finnelly 154 Lord Cambell said this at 191:-

"... really this confirms the opinion I have entertained, that hardly any weight is to be given to the evidence of what are called scientific witnesses; they come with a bias on their minds to support the cause in which they are embarked ...."

Jessel M.R. an outstanding judge even by the high standards of Masters of Rolls, had this to say in Abinger v Ashton (1873) L.R. 17 Eq. 358 at page 373:-

"... I may say what I think I have often said before, that in matters of opinion I very much distrust expert evidence, for several reasons. In the first place, although the evidence is given upon oath, in point of fact the person knows he cannot be indicted for perjury, because it is only evidence as to a matter of opinion. So that you have not the authority of legal sanction. A dishonest man, knowing he could not be punished, might be inclined to indulge in extravagant assertions on an occasion that required it. But this is not all. Expert evidence of this kind is evidence of persons who sometimes live by their business, but in all cases are remunerated for their evidence. An expert is not like an ordinary witness, who hopes to get his expense, but he is employed and paid in the sense of gain, being employed by the person who calls him.

Now it is natural that his mind, however honest he may be, should be biased, in favour of the person employing him, and accordingly we do find such bias. .... Undoubtedly there is a natural bias to do something serviceable for those who employ you and adequately remunerate you. It is very natural, and it is so effectual, that we constantly see persons, instead of considering themselves witnesses, rather consider themselves as the paid agents of the person who employs them.

..... Well, we have the witnesses giving evidence for the Plaintiffs' view of the matter, or the Defendants' view, according as they are sought out and paid by each. It is very natural, and just what one would expect, but it leads one to distrust their evidence. There is also this to be said against them, namely, that their evidence is not the evidence of fair professional opinion. The men are selected according as their opinion is known to incline.

Suppose a person wants to sell a house, and as he wants a very high value put upon it, he send to ten valuers, and out of these he selects the three who have put the highest value on the house. The purchaser wants a very low value, and selects out of a number of valuers three of the lowest. Each set of valuers values high or low, according to the requirements of the person who employs them. I have known the same sort of thing done even as regards medical evidence. The consequence is, you do not get fair professional opinion, but an exceptional opinion by evidence selected in this way.

That being so, when I have expert evidence I am, as I said before, very distrustful a priori; and I am anxious to ascertain the character of the experts, and to see the position they occupy."

1135. When one comes to handwriting expert opinion evidence, there are two initial considerations which are of considerable relevance. The first is that the nature of handwriting expert opinion is far less reliable than other scientific expert opinion. The science of DNA, the science of chemical analysis, or even finger-print science are far more advanced and precise and therefore renders the expert opinion on those matters of greater reliability.

1136. On the other hand, opinion evidence by handwriting comparison is in a different category. Historically at common law, the admission of evidence of witness as to handwriting was limited to cases where the witness had seen the person write (as Tse did) and those where the witness had acquired some previous familiarity with the person's handwriting in another way (as Teresa did not). Holdsworth's History of English Law, vol. 9 at page 213-4 described the position up to the 1854. In the important case of Doe d. Mudd v Suckermore (1836) 5 Ad. & E. 703, 4 eminent judges of the King's Bench were evenly divided (Coleridge J. & Patteson J. against and Denman C.J. & Williams J. in favour) as to whether the opinion evidence of handwriting based upon a handwriting comparison formed for the purposes of trial was admissible. As result of this legal impasse, section 27 of the Common Law Procedure Act 1854 (subsequently replaced by section 8 of the Criminal Law Procedure Act 1865, the Hong Kong equivalent of which being section 17 of the Evidence Ordinance) was passed allowing the giving of handwriting comparison evidence. This statutory provision applies to both civil and criminal proceedings and it provides as follows:-

"Comparison of a disputed writing with any writing proved to the satisfaction of the judge to be genuine shall be permitted to be made by witnesses; and such writings, and the evidence of witnesses respecting the same, may be submitted to the court and jury as evidence of the genuineness or otherwise of the writing in dispute." (underlining supplied)

Yam, J. cited this statutory provision at paragraph 6.17 of the Judgment, except that he was mistaken in believing (having copied paragraph 37 of Mr. Edward Chan's Written Submissions) that the right of the court's admission of comparison handwriting evidence was derived from common law. It was not, as explained by my earlier historical account. It required a statute in 1854 to permit comparison handwriting evidence to be admitted.

1137. A careful study of the reasons advanced against such handwriting comparison evidence in Doe v Suckermore (see for example), per Coleridge J. at 714:-

"danger is in the principle itself, that selected specimens may be made the standard from the which the witness is to judge"

and per Patteson J. at 735:-

"other minute circumstances of similitude or discrepancy, which every man in his own experience must know may arise from the different pen or ink, or haste, or deliberation, with which the same person writes at different times"

will inform this Court of the dangers inherent in the nature of such evidence. It can be seen from this historic account that comparison handwriting evidence had generally been regarded as not very satisfactory and that it required a statute to make such evidence admissible, leaving of course the weight of such evidence to be decided in each individual case.

1138. In Fuller v Strum, at first instance when Deputy Judge Sher Q.C. held against forgery despite the expert opinion in favour of forgery, he said at page 11 of the Transcript this:

"Mr. Mitchell submits that I cannot reject the expert evidence of Dr. Giles. In my judgment, however, there is a world of difference between the type of expert evidence led in Re B [medical expert evidence] and the evidence contained in Dr. Giles's report in this case. The training of experts enable them to identify facts which a lay witness or a judge could not identify, without expert help. Such evidence may truly be described as scientific and the radiologists' evidence as to when an injury occurred falls plainly within this category. But some expert evidence may amount to no more than the drawing of inferences from facts observable as much by the expert as by a lay witness; and the inferences to be drawn from these facts may be capable of being drawn as much by the experts as a lay witness. .... But in relation to this type of expert evidence the judge, I think, is entitled to form his own view, having regard to, and balancing, the other evidence available to him in the case. The way Miss Rich, counsel for the claimant put it in her skeleton argument is this:

"Handwriting analysis is not a complete science and an analyst's conclusions, however carefully reasoned and supported by experience and reputation, are ultimately impressionistic and susceptible to error."

...... But the central reasoning why she found that the will was forged was the number and nature of the differences between that signature and the genuine signature of Mr. Strum. In that respect, her evidence falls outside the purely scientific category in respect of which the judge would be helpless without expert assistance...."

1139. This healthy scepticism which a Court should show to comparison handwriting expert opinion was displayed in Public Prosecutor v Mhamed Yatim [1977] 1 M.L.J. 64 where Hashim Yeop Sani J. said at page 66:-

"It is settled law that evidence by a handwriting expert can never be conclusive because it is only opinion evidence - see Ishwari Prasad v Mohd. Isa. The assessment of evidence of handwriting experts was also dealt with in Indar Datt v Emperor. In that case In re B. Venkata Row was cited and also a quotation from Dr. Lawson's work the Law of Expert and Opinion Evidence which runs as follows:-

"The evidence of the genuineness of the signature based upon the comparison of handwriting and of the opinion of experts is entitled to proper consideration and weight. It must be confessed however that it is of the lowest order of evidence or of the most unsatisfactory character. We believe that in this opinion experienced laymen unite with the members of the legal profession. Of all kinds of evidence admitted in a court this is the most unsatisfactory. It is so weak and decrepit as scarcely to deserve a place in our system of jurisprudence."

In Srikant v King Emperor, two learned judges of the Allahabad High Court observed that:

"To base a conviction on the evidence of an expert in handwriting is, as a general rule, very unsafe."

And this observation was approved of by a bench of two other learned judges of the same court in Kali Charan Mueerji v Emperor."

1140. The nature of the handwriting expert evidence is such that in a case where the competition is between Forgery based on contested evidence of handwriting expert evidence and proper Execution based on the evidence of the attesting witness, a Judge must cast not only a critical eye, but a very critical eye, because handwriting analysis is less than a complete science and contains in its analysis many inherent errors and unreliabilites.

1141. My second initial observation on comparison handwriting evidence is that the Court generally preferred the evidence of an attesting witness (when the handwriting of a document is in dispute) to the expert opinion evidence on handwriting. This is borne out both by authorities and for good reason.

1142. First on authority. In Robson v Roche (1824) 2 Add 53, Sir John Nicholl said at page 65:-

"Evidence to handwriting is, at best, inconclusive; but of the species I will say that witnesses here produced have furnished as strong and as satisfactory a sample as well could be furnished".

It then went on to say the opposite about the value of handwriting evidence in support of the will, at page 98:-

"the rule of this Court being that evidence to hand-writing only is incapable of substantiating any disputed instrument as a will "

In Newton v Ricketts (1861) 10 H.L.C. 263 Lord Campbell said at 266:-

"There was the clear and distinct evidence of the surviving witness, who was not cross-examined, as he might have been, and who swears that he and the other attesting witness "were respectively present, and saw Sir Tristram Robert Ricketts and Dame Rebecca Ricketts his wife, both of whom they well knew, severally and respectively sign and seal, as their respective acts and deeds deliver, the said parchment writing or deed poll of appointment." That evidence being unimpeached, was it a ground for directing an issue that certain person, looking at the handwriting of the signature, intimated their opinion, which they might sincerely entertain that it was not the character of Sir Robert Rickett's handwriting? That could be no counterbalance to the direct evidence of the attesting witness ....."

1143. In Fuller v Strum at first instance Deputy Judge Sher Q.C. held (there was no appeal on this aspect) that there was no forgery and preferred to rely on the evidence of the attesting witness.

1144. In Hong Kong there was the probate case of Iskandar, where the contest was between Execution and Forgery, with evidence from competing handwriting experts. At page 11 of the unreported judgment of Iskandar v Leo (1987) P. 9 of 1987, Liu J. said:-

"The possibility of destroying or undermining the combined effect of these cogent accounts given by Mr. Lui and Miss Ivy Neo*, by any handwriting expert evidence, even if unanimous, must be remote. Handwriting experts' examination "is purely a matter of practical training and experience". It is an inexact science practised as a recondite art. No counsel in this case was bold enough to enquire into its margin of error." (*being the attesting witnesses)

In Iskandar the Court was firstly of the view that the discipline of comparison handwriting is less than an exact science and therefore of somewhat doubtful reliability (with unknown margin of error) and secondly that comparison handwriting evidence (even if unanimous) could hardly ever overcome acceptable attesting witness evidence.

1145. I now turn to the question of principle. The reason why the Court should generally prefer the attesting witness evidence to that of the comparison handwriting expert evidence can be seen in the wisdom of history, namely the heavy statutory reliance on the necessity for testator's signature to be witnessed and for the witnesses themselves to sign. When everything concerning a will is capable of being disputed, the best evidence is therefore always the contemporaneous signature of the witness on the will and the second best evidence is the evidence of the attesting witness that he did witness the signing by the testator. A particular event such as witnessing the signing of a document is not something which a person might easily forget although he might get the precise details wrong and this is far more reliable and preferable to the uncertainties of forensic comparison handwriting evidence. It is precisely because of the unreliability of the comparison handwriting evidence (when presumption of Omnia does not apply) that the Court both historically and on principle, generally give greater weight to the eye-witness account of the attesting witness and his contemporaneous signature on the will.

(II) PRELIMINARY OBSERVATION ON COMPARISON CHINESE HANDWRITING EVIDENCE

1146. The Chinese language is an ideogram and pictorial based language with Chinese characters generally built up from radicals (a total of some 214) and phonetics (see Mathews' Chinese-English Dictionary, page xxii). Traditionally it was written with a brush dipped in ink. Chinese language is closely connected to Chinese calligraphy and Chinese painting.

1147. Western European languages are largely alphabet based and are not ideographical or pictorial in their composition. The writing of European languages is culturally and analytically very different from writing in the Chinese language.

1148. There was no reference in the Judgment of Yam J. to any difference between analysis of European language handwriting and the analysis of Chinese language handwriting and notwithstanding the reference to the Kung case, there was an underlying assumption that occidental handwriting expertise and methodology is accepted by the Court to be equally applicable to Chinese handwriting analysis. I have considerable reservation whether that assumption is really justified or that it serves the cause of justice where this is of critical importance as in this case.

1149. Assuming however that there is a proper basis for that assumption, the Judge nevertheless apparently made a large jump from accepting Occidental handwriting analysis methodology to accepting that the person applying the methodology does not need to read, write or have long exposure to the Chinese writing (both genuine and forged). Such a person was of course GL, an American forensic expert who had acknowledged himself that he could not read, write, understand Chinese and that he did not have long and full professional exposure to Chinese handwriting or its comparison analysis.

1150. If the Judge had been more critical or analytical, he would have appreciated that the analysis of Chinese handwriting is far more complex than the analysis of European writing, consisting of 26 alphabets with a limited combinations (low in numerical terms compared to Chinese) of how these alphabets are joined together to form words.

1151. Chinese characters with their makeup from radicals and phonetics present many different factors which create far greater complexity and therefore room for uncertainty. The first uniqueness is, the spacing between characters and within the parts of the character. Secondly, the infinite ways the individual parts of a character can be joined together gives each piece of writing a unique character. Third uniqueness of Chinese handwriting, is the freedom within a defined space, a character can be built up. Fourthly it is the built-up of each character which gives it a unique ascetic or pleasing appearance. What is special about the writing of Chinese character is that it is closely related to Chinese painting dictated by the using of brush dipped in ink applied to space. Because of the complexity of Chinese writing, it might be said that the unmodified blind application of Occidental handwriting methodology to seriously disputed Chinese handwriting written in the manner of the ink brush (with a calligraphic pen) is of doubtful value if not wholly inappropriate.

1152. In the context of the 8 Disputed Signatures alleged by comparison handwriting evidence to be forged, bearing in mind the aforesaid considerations of the peculiar nature of comparison handwriting expert opinion, it is obvious that the Trial on the Father's case of Forgery is unusually complex and that it requires extremely careful consideration, on methodology, on approaches, on application, on evaluation and on analysis at the micro level.

1153. The Judge held on Forgery that each of the 4 Wang Signatures were forged by simulation (paragraph 31.36) and each of the 4 Tse Signatures were forged by superimposing tracing (paragraph 31.39). The Judge arrived at these conclusions on forgery by a process of :-

(1) Macro or overall view;

(2) Micro detailed analysis divided into;-

(a) Similarities;

(b) Significant Differences (in practically all cases on a difference based on one feature in relation to only one of the 4 Signatures in question);

(c) Judge-Observed Further Differences (in practically all cases based on a difference based on one feature in relation to only one of the 4 Signatures in question).

1154. On the so called Micro Analysis (which is copied virtually verbatim from Volume 1 of Mr. Edward Chan's Written Submission dated 20th September 2002 and not from Mr. Chan's Volume 2, the significance of which will be analysed later), the Judge held that:-

(1) The Similarities relied on by the Wife were all irrelevant;

(2) The Significant Differences relied on by the Father were all relevant and pointing to forgery;

(3) The Judge-Observed Differences were all relevant and pointing to forgery.

1155. The submission on behalf of the Wife is that the Judge erred on many levels and the elaborate arguments presented to the Court of Appeal over many days by both sides very often clouded some of the critical issues on this difficult case of Forgery. I found that by going back to the original Written Submissions given to the Judge (details and chronology of which I will set out later), I have begun to understand why this case had gone seriously wrong and that it is vitally important to formulate the critical issues on Forgery in this appeal.

1156. In my view, the proper resolution of this aspect of the appeal on Forgery requires the following issues (additional to issues (I) and (II) considered above )to be properly considered:-

(III) NO METHODOLOGY, NO PROPER METHODOLOGY & FATAL METHODOLOGY

(1) No Methodology or Principle for determination and evaluation of Forgery.

(2) Failure to formulate a proper rational and coherent Methodology based on consolidated evaluation of Signatures evaluated according to rational propositions on handwriting and having regard to the overall probabilities on Forgery and applying the appropriate high standard of proof required for Forgery.

(3) Adoption of a Fatal Methodology by accepting a novel theory of simulation from multiple models and thereby dismissing all similarities and rendering meaningless any consolidated evaluation of handwriting evidence.

(IV) WRONG APPROACH, WRONG APPLICATION OF METHODOLOGY & MISCONCEPTIONS

(1) Atomisation (No consolidated evaluation) and Wrong Controlling Factor

(2) Mechanical approach to dissimilarities and elevating dissimilarities to significant dissimilarities pointing to Forgery

(3) GL not Chinese Handwriting Expert and Judge played Chinese Handwriting expert.

(4) Misconceptions about Implications of Calligraphic Pen

(5) Misconceptions about Tremors

(V) SERIOUS ERRORS OF MICRO ANALAYSIS

Re Wang Teh Huei (1) Wang - Similarities and Dissimilarities (2) Teh - Similarities and Dissimilarities (3) Huei - Similarities and Dissimilarities Re Tse Ping Yim (4) Tse - Similarities and Dissimilarities (5) Ping - Similarities and Dissimilarities (6) Yim - Similarities and Dissimilarities.

(III) NO METHODOLOGY, NO PROPER METHODOLOGY AND FATAL METHODOLOGY

No Methodology

1157. The first serious methodological shortcoming in the Judgment is the failure to state what was the principle or methodology for the resolution of the Forgery dispute. At the Trial, the question of Forgery involved the respective experts and lawyers contending the opposite sides of two questions, (1) whether the Wang and Tse Signatures were genuine (Wife said yes, Father said no) and (2) whether the Wang and Tse Signatures were forgeries (Father said yes, Wife said no). Although these two questions of Genuine Signatures and Forged Signatures are intimately related, the methodology, principle and test governing each question are not quite the same. It is unfortunate that in the long Judgment, the Judge got the two questions confused and sometime transposed the test for one question to the other question. This took place because in the Judgment it was not stated or formulated what was the proper Principle or Methodology applicable to proof of these 8 Signatures being forgery. Without such Methodology or Principle being stated, there could be no proper application of Principle or Methodology. A proper review of the Judgment will show that the Judgment did not state what was the Principle or Methodology that the Judge was adopting or following for the analysis and proof of Forgery of these 8 Signatures.

1158. Section I of the Judgment does not deal with Methodology or Principle applicable to the contested case of 8 Signatures Forgery. The Introduction in paragraphs 6.1 to 6.4 of Section IIA described the sequence of coverage of the Judgment but did not state what was the Principle or Methodology which was applicable. The question of standard and burden of proof was discussed from paragraph 6.8 to 6.15 but otherwise Chapter 6 of Section IIA does not throw light on the proper Methodology and Principle applicable.

1159. Chapter 7 of Section IIA is the Chapter dealing with theories and methods and principles relating to handwriting comparison and is logically the Chapter which should set out the Methodology and Principle governing Forgery Signatures and Genuine Signatures. Apart from paragraph 7.4, 7.5, 7.15 and 7.29, the many relevant passages in Osborn, Huber, Harrison, Hilton and the evidence of Jia (which I will refer to later) discussing proper evaluation and the principle of Consolidated Evaluation (which is the foundation of examination of disputed documents) were not referred to by the Judge in Chapter 7 of the Judgment.

1160. After the long discussion of individual handwriting evidence in Section II (IIA and IIB), the Judge in Section IV of the Judgment under Final Conclusion said at paragraph 31.37 :-

"All the similarities pointed out by the defendant's experts were only superficial similarities produced by simulation. There are many unexplained but significant differences. There are also other features including tremors, retouching and retracing, and different writing movements with the sample signatures of Wang. All these are features and indications of forgery. The evidence in support of such a finding of forgery is very strong and cogent."

This concluding paragraph indicates that the Judge did not apply any Principle or Methodology governing the proof of forgery of the 8 Signatures. His conclusion of forgery was based on his view of a collection of differences which he then added up to forgery. There was no critical method of consolidated evaluation (which I will set out below) applied. The Judgment was therefore fatally flawed for not stating what Principle or Methodology was to govern the determination of Forgery.

No Proper Methodology

1161. The complaint by Mr. Chang on behalf of the Wife on this appeal is that the Judge applied the wrong methodology. That of course begs the question of what is the proper Methodology. Many passages from the leading textbooks as well as from the evidence of Jia are relevant for this purpose. I will first set them out (with text highlighted with bold where appropriate) before proceeding to discuss what is the proper Methodology which should be formulated arising from these passages and other passages referred to in the Judgment.

A. Albert S. Osborn ("Osborn") on "Questioned Documents" 2nd Edition

(1) At page 205 under Chapter 12 - Systems of Writing/"Osborn Quotation 1"

"The conclusion that a writing is not genuine is only properly reached when it contains divergences in amount and quality beyond the range of variation in the standard writing that cannot reasonably be accounted for by changed condition in the writer or surrounding the writer."

(2) At page 241 under Chapter 15 - Process of Comparison/" Osborn Quotation 2"

"Mere similarities do not necessarily prove genuineness anymore than mere superficial differences necessarily prove lack of genuineness. The incompetent witness, notwithstanding this fact, by dependence upon similarities alone reaches the conclusion of genuineness, or by dependence upon differences alone reaches the conclusion of lack of genuineness.....

As all kinds of thing are shown to be the same by significant similarities or different by significant differences, there arises the necessity of distinguishing what is significant. This is of course the essential element in a process of comparison of any kind and the key to the problem, and a certain amount of knowledge and a certain degree of intelligence are therefore necessary before intelligent comparison can be made that leads up to correct identification."

(3) At page 243-244 under Chapter 15 - Process of Comparison/" Osborn Quotation 3"

"The "dents" and "scratches" in a handwriting are its individual characteristics. One without knowledge and without experience does not know what they are nor how to look for them. .....

This matter of analysis and reasoning assumes special importance in connection with certain often proposed restrictions of expert testimony when it is argued that the ruling should be made that, "The witness be allowed to point out the similarities or dissimilarities but make no comment thereon." When this procedure is followed it puts all testimony, good and bad, practically on one dead level and suppresses the vital element in comparison which is the rational interpretation of likeness and differences.

The sensible discussion of this subject by Professor Wigmore deserves special emphasis. He says:

"on the direct examination, the witness may and, if required must, point out his grounds for belief in the identity of the handwriting, on the principle already considered. Without such a re-enforcement of testimony the opinion of experts would usually involve little more than a counting of the numbers on either side. The progress of modern chirographic science makes it all the more possible, as well as desirable, to discriminate between witnesses according to the convincingness of the reasons that may be given by them for their conclusions.....

The method of weighing the testimony here suggested, that is a consideration of "the convincingness of the reasons", at once removes the stock objection to expert testimony, that it is always in conflict. The usual argument criticizing expert testimony is based on the assumption that it is a mere opinion devoid of all "convincingness". There is expert testimony of this class but it should not be confused with that expert testimony, the "convincingness" of which can be compared by an intelligent hearer who himself has the ability to reason regarding similarities and differences. There are those, however, who will not do the necessary reasoning; they, in effect, count the witnesses.

As is pointed out in detail in the following chapter, in order to reach the conclusion that two writings are by the same hand there must not only be present class characteristics but also individual characteristics or "dents and scratches", in sufficient quantity to exclude the theory of accidental coincidence; to reach the conclusion that writings are by different hands we may find numerous likenesses in class characteristics but divergences in individual characteristics, or we may find diverges in both, but diverges must be something more than mere superficial differences." "

(4) At page 245 under Chapter 15 - Process of Comparison/"Osborn Quotation 4"

"In identifying a person, for example, scars, deformities, finger-prints, or a series of accurate measurements, must be depended upon the finally, if the conclusion of identity is reached, either in a person or a handwriting, there must not remain significant differences that cannot reasonably be explained. This ignoring of the differences, or the failure properly to account for them, is the cause of most of the errors in handwriting identification. The dishonest advocate and the dishonest witness sometimes make a labored and elaborate effort to "explain away" differences that are as plain as the nose on a man's face."

(5) At page 250-1 under Chapter 16 - Individual Characteristics in Handwriting/"Osborn Quotation 5"

"No general set of infallible rules can be formulated but some general principles can be stated that apply in most cases. One of the first of these principles is that those identifying or differentiating characteristics are of the most force which are most divergent from the regular system or national features of a particular handwriting under examination.

The second principle, perhaps more important than the first, is that those repeated characteristics which are inconspicuous should first be sought for and should be given the most weight, for those are likely to be so unconscious that they would not intentionally be omitted when the attempt is made to disguise and would not be successfully copied from the writing of another when simulation is attempted.

A third principle is that ordinary system or national features and elements are not alone sufficient on which to base a judgment of identity of two writings, although these characteristics necessarily have value as evidence of identity, as stated above, if present in sufficient number and in combination with individual qualities and characteristics.

Any character in writing or any writing habit may be modified and individualized by different writers in many different ways and in many varying degrees, and the writing individuality of any particular writer is made up of all these common and uncommon characteristics and habits. As in identifying a person, as we have already seen, it always is the combination of particulars that identifies, and necessarily the more numerous and unusual the various elements and features the more certain the identity"

(6) At page 262 under Chapter 16 - Individual Characteristics in Handwriting/"Osborn Quotation 6"

"It also needs to be emphasized that two writings are identified as being by the same writer by the absence of fundamental divergences as well as by a combination of a sufficient number of similarities. The process is always a double operation, positive and negative, and if error is to be avoided neither part of the process should be overlooked. In order to reach the conclusion of identity of two sets of writings there must not be present significant and unexplained divergences. These divergences must, however, be something more than mere trivial variations that can be found in almost any handwriting."

(7) At page 364 under Chapter 20 - Proof of Genuineness in a Disputed Document/"Osborn Quotation 7"

"In investigating alleged forgery it is important to keep in mind the fact that evidence of genuineness of all kinds is in all ways just the opposite of evidence of forgery. The most common symptom of forgery is not, as is incorrectly thought by many, divergences in form but a drawn and hesitating quality of stroke or line, particularly at certain parts of letters that should be freely written, indicating that during the writing act, attention was given to the details of writing and naturally, just the opposite conditions are those that indicate genuineness.

Flying starts and flying finishes where the motion of the pen precedes the beginning of the stroke and continues beyond the end to a vanishing point, are found in free natural writing and, as a rule, are important indications of genuineness. Intermediate strokes also where the pen comes off the paper but is not stopped and shows continuity of motion are, as a rule, indications of unconsciousness of the details of the operation and point toward genuineness. This analysis requires close observation and accurate reasoning."

(8) At page 368-9 under Chapter 20 - Proof of Genuineness in a Disputed Document/"Osborn Quotation 8"

"Another indication of genuineness in a holographic document or a considerable amount of writing, or in two or more disputed signatures, are certain natural variations in the details of the writing. It is difficult for the inexperienced or unthinking examiner to understand that a certain extent of variation in a group of several signatures and variation in repeated words and letters in a continuous holographic document can be evidence of genuineness. The forger does not understand this necessity for natural variation and, as nearly as he can, makes words and letters just alike.

As is well known, one of the evidence of forgery in a tracing is its identity with model or the damaging identity of several traced forgeries when compared with each other, or of several simulated signatures copied from one model or written from one mental design. As is easily demonstrated by experiment, when even two short signatures or any two words are written, there is possibility of slight divergence at any point in the line making up the words or signatures, and this is the genuine divergence which should appear in genuine signatures or in any continued genuine writing. The endings of repeated words in an extended writing are almost certain to show a natural variation and similar words in a simulation will usually show an unnatural similarity.

It necessarily follows, therefore, that if the several lines of a disputed document, or several signatures under investigation, show these natural variations of writing of the same word or letter, all of course within the scope of variation of the genuine writing, this variation itself, surprising and paradoxical as it may appear, is as strong evidence of genuineness as the opposite condition is evidence of forgery.

Imitation is almost certain to conform closely to a fixed type and is not apt to show that unconscious abandon that in genuine writing produces natural variation. Here again arises the necessity for logical reasoning applied to the facts as an essential part of a thorough examination. If this reasoning is not permitted in testimony it is easy to understand how testimony is weakened or practically destroyed.

There is about genuine writing viewed as a whole that which indicates that it is genuine and a close analysis of this evidence shows that one of the elements is this slight variation in proportions, size, and the care given to the formation of certain letters, especially letters in certain positions in words, or words in a line, and in general that indifference that shows that the writing was not the result of following a copy but of following a habit."

(9) At page 370 under Chapter 20 - Proof of Genuineness in a Disputed Document/"Osborn Quotation 9"

"The successful simulation of a group of signatures is also a task of great difficulty, and in this task also, strange as it may seem, one of the main difficulties is the necessity of incorporating this free and natural variation. When therefore this natural variation does appear it is strong evidence of genuineness." (page 370 under Chapter 20 - Proof of Genuineness in a Disputed Document/"Osborn Quotation 10")

(10) At page 388 under Chapter 20 - Proof of Genuineness in a Disputed Document/"Osborn Quotation 10"

"....identity is proved when the two handwritings both contain a sufficient number of significant characteristics, qualities and elements so that it is unreasonable to say that they would all accidentally coincide in two different handwritings."

B. Huber on "Handwriting Identification: Facts and Fundamentals" at pages 34/5/"Huber Quotation" :-

"ANALYSIS: The unknown and the known items, must by analysis, examination, or study, be reduced to a matter of their discriminating elements.

COMPARISON: The discriminating elements of the unknown, observed or determined through analysis, examination or study, must be compared with those known, observed, or recorded of the standard items.

EVALUATION: Similarities and dissimilarities in discriminating elements will each have a certain value for discrimination purposes, determined by their cause, independence, or likelihood of occurrence. The weight or significance of the similarities or difference or each element must then be considered and the explanation(s) for them proposed".

C. William R. Harrison ("Harrison") on "Suspect Documents" at pages 343-5/"Harrison Quotation":-

"The rule is simple - whatever features two specimens of handwriting may have in common, they cannot be considered to be of common authorship if they display but a single consistent dissimilarity in any feature which is fundamental to the structure of the handwriting, and whose presence is not capable of reasonable explanation.

This rule must not be applied blindly. For example, the dissimilarity must be present in some feature which is known to the examiner to be fundamental to the structure of the handwriting.

....Again, the dissimilarity must be truly consistent; the presence of a few accidentals should never be interpreted as denoting the existence of a consistent dissimilarity not capable of reasonable explanation. The dissimilarity must clearly be shown to be a natural feature of the writing.

....The type of dissimilarity which is regarded as most significant is that which has to be sought for in the structure of the handwriting and whose presence may well be overlooked by those inexperienced in the critical examination of handwriting. There is little chance of such inconspicuous dissimilarities having been deliberately introduced as disguise."

D. Ordway Hilton on "Scientific Examination of Questioned Documents" at pages 174, 184-5/"Hilton Quotation":-

"This natural variation between several signatures by the same writer plays an important role in an identification. No two specimens are absolutely identical. .....The process of identification requires that the known signatures establish accurately the extent of variation, and then it is necessary to ascertain whether the identifying qualities of the unknown signature fall within these limits. Thus, the identification of a signature consists not of matching it exactly with a particular known signature, but of determining that it contains the characteristics of and is written in the same way as the standards and also fits within the extremes of variation established by the collection of known signatures."(page 174,184-5/"Hilton Quotation")

E. Jia evidence at Transcript 73:27:6-28:9/"Jia Evaluation Quotation":-

"It is only on this basis that a comparison could be made between the questioned handwriting and known handwritings. Of course, when I said "handwriting", that includes signatures as well.

The purpose of doing the comparison is to find out collectively the similarities between the two. At the same time, one also needs to find out the differences between the two. In this process, the principle behind it is that the approach must be objective. One cannot just rely more on similarities or just rely more on differences.

After laying out all the similarities and differences, the next step, which is the most important step, is to give an overall evaluation. By saying "overall evaluation", that means to conduct an in-depth analysis on the similarities and differences, to evaluate the value of these similarities; at the same time, one has to find out the reasons for having these differences, and then to sum up all these similarities and differences to see whether similarities are more important or whether differences are more important.

After weighing all these factors, one can then reach a reasonable and scientific conclusion. During this process, one has to insist on adopting this approach for evaluation. However, during the examination process, one needs to go all over it again so as to have a more in-depth knowledge of the documents. These procedures and methods have been adopted all along in the past based on our experience."

And at Transcript 73:46:12 - 73:46:25/"Jia Quotation 2":-

"Because Chinese handwriting is different from English handwriting, because the subject matter of examination of Chinese handwriting is Chinese characters, because Chinese characters are idiographic characters, and as for English, they are formed by letters in the way of phonetics. So in the examination of Chinese handwriting, one has to study different styles, writing styles, which include the regular style, the running style, the cursive style, the clerical script, or sometimes even seal script. Structure is very important in Chinese handwriting. How can one put all together the lateral radical as well as the strokes together in a nice manner?....."

1162. What does this large collection of passages quoted above amount to. In my view (and here one is trying not to play expert), an attempt can be made to reduce the wisdom in these passages into essential elements and to distil from them a rational and hopefully coherent Principle or Methodology applicable to the proper determination and evaluation of Forgery of two sets of 4 Signatures each (namely the 4 Wang Signatures and the 4 Tse Signatures). This is put forward as a contrast to regrettably (in my view) firstly no methodology adopted by the Judge in arriving at the conclusion of Forgery or secondly a partial methodology (if at all) which was in any event flawed.

1163. I suggest that the following Propositions can be extracted and extrapolated from the passages quoted:-

(1) Significant differences and significant similarities (not ordinary regular differences or ordinary regular similarities) are the key to evaluating handwriting;

(2) Individual and Personal Characteristics rather than national regular system characteristics are the important features which might be called significant features;

(3) Individual and Personal Characteristics which are major departures from the national regular system characteristics are given the most force and value in terms of significant features;

(4) Individual and Personal Characteristics which are inconspicuous should be given much weight and value in terms of significant features;

(5) Individual and Personal Characteristics which are fundamental to the structure of that person's writing is given much weight and persistent departure from those individual and personal fundamental structural characteristics is of importance, specially when these individual and personal fundamental structural characteristics are inconspicuous;

(6) The range of variations of a person's writing is very relevant and is of particular significance when it includes in its variations individual and personal characteristics;

(7) Matching variations which includes individual and personal characteristics must be given great weight and significance and can be called significant matching variations;

(8) Free natural writing such as flying starts and finishes or intermediate strokes with pen coming off paper are significant indications of genuineness;

(9) Natural slight variations in writing indicating writing from habit rather than from copying is significant indication of genuineness;

(10) Natural writing described in (8) and/or (9) above when combined with individual and personal fundamental structural characteristics strongly suggest genuineness;

(11) Simulation of a group of signatures with variations amongst the group is a task of great difficulty specially when the group incorporates natural variations and containing individual and personal characteristics;

(12) Evaluation of handwriting is never a matter of adding up the numbers of similarities or differences;

(13) Evaluation of handwriting for Forgery is a matter of:-

(a) qualitative assessment of the convincingness of

(b) persistent

(c) divergences from and/or conformity with

(d) individual and personal (inconspicuous) characteristics

(e) which are fundamental to the structure

(f) beyond the range of natural variations

(g) without reasonable explanation

(h) in the presence/absence of free natural writing

(i) in the presence/absence of matching variations

(j) in the presence/absence of one set of multiple signatures containing matching variations and individual and personal characteristics

(14) Overall evaluation of handwriting evidence is essential and calls for the in-depth analysis of the value and reasons of similarities and the value and reasons of differences, weighing all factors including the Propositions above, in order to arrive at one of three options, firstly Genuineness, secondly Not Sure and thirdly Forgery.

1164. I t is to be noted that the convincingness of evaluation is directly related to the value of the assessment of the significance of the similarity or difference. In the absence of a powerful convincingness in the evaluation, the reasonable result of consolidated evaluation should be the second option, namely Not Sure.

1165. What is always essential for a full and proper determination and evaluation for the purpose of handwriting and in particular for the purpose of Forgery is a consolidated evaluation which includes not only the Propositions stated in the paragraphs immediately above but also includes two further important considerations, namely the overall probabilities of Forgery and secondly the appropriate convincingness of proof required.

1166. The overall probabilities of Forgery will in this case, apart from the evidence of handwriting, also involve the probability of:-

(1) Forgery when the natural person to inherit was the Wife;

(2) Forgery when the Wang Signatures on the 1990 Documents were witnessed and there was Affirmation evidence in support of valid execution;

(3) Forgery when the Tse Signatures on the 1990 Documents were sworn by Tse in the Tse Affirmations as being those signed by him;

(4) Forgery alleged to be perpetuated was not one forged signature of Teddy but four forged signatures of Teddy contained in 4 Documents as a set;

(5) Forgery alleged to be perpetuated was not just one forged signature of Tse but four forged signatures of Tse contained in 4 Documents as a set;

(6) Forgery of the 4 signatures of Teddy signed not with an ordinary ball-point pen (which will be much more difficult for the detection of handwriting genuineness) but with a calligraphic pen.

1167. The appropriate high standard of proof is important in the overall evaluation and determination in the consolidated evaluation for Forgery not only because of the well-known principle stated in Re H but because in this case there is a substantial disagreement between the experts as to what are significant or not significant for similarities and for dissimilarities (which the Judge resolved without stating by what methodology and to what margin of errors he allowed). What is even more controversial is the Judge's personal involvement in a matter of Chinese handwriting, of which the Judge claimed no expertise and cannot be cross-examined on his assumptions and beliefs. But the Judge's involvement was very much causative in the final determination by him of Forgery. A proper and full Methodology therefore should include in the consolidation evaluation:-

(1) the handwriting evaluation;

(2) the overall probabilities; and

(3) the appropriate levels of proof required specially having regard to the controversial nature of the disagreement between the experts and the involvement of the Judge (whether termed as Judge playing expert or not).

1168. In my view, the lack of a properly formulated Methodology doomed the Judgment. In a way, this was inevitable because the Judge was copying from Volume 1 of Mr. Edward Chan's Written Submission and that Volume 1 contained no Methodology to apply. This fundamental flaw in the Judgment of lack of a properly stated Methodology to apply in the determination and evaluation of the 8 Signatures Forgery ought to result in the Judgment being wholly unsustainable and should compel this Court to hold that Final Judgment should be given in favour of the Wife and against the Father on Forgery.

Fatal Methodology - Multi Models Simulation

1169. The vital key to the ultimate finding of Forgery of the Wang Signatures by simulation is the dismissal of all similarities, even significant similarities, on the basis that the forger could have copied from several models.

1170. After referring to the Osborn Quotation 8 which described copying from one model or one mental design, the Judge at paragraphs 7.6 and 7.7 said:-

"7.6 It should be noted that the proposition made by Osborn was predicated on a forger copying from one single model. Plainly when this assumption is applicable, the forger could not have introduced into the questioned signatures the natural variations found within the genuine signatures of the true author. Hence, an inference of genuineness could be drawn from the presence of the same natural variations being present in both the questioned and the genuine signatures. However the strength of the inference of genuineness would also depend on the nature of the variations found. If those variation features are personal to the author, then the inference would be stronger. If the features of the variations are those normally found in writings of a large number of persons, then the inference would be a weak or very weak one.

7.7 On the other hand if the evidence would give rise to the possibility that the forger may be copying from say, two or three selected samples, then the presence of the same features of natural variations, and more significantly, accidental features, found only amongst the two or three samples and the questioned signatures might well be the result of the copying. This inference would be more strongly drawn if amongst the samples available, the variations found within the questioned signatures are only found within the two or three particular known samples and not found from the rest"

1171. By postulating copying from several models, the Judge at one stroke totally neutralised and cancelled out all evidence of natural variations and matching variations which are strong insignia of genuineness. The theory of copying from several models was totally novel and is not in the handwriting textbooks and manuals.

1172. Simulated forgery is one of the two methods of forging signatures but simulated forgery is always based on one model simulation. Agreeing with one model simulation of Osborn, Harrison described the simulated forgeries in the following way at page 380:-

"The first involves the careful study of a genuine specimen of the signature it is intended to copy, so that the forger may acquire a mental picture of the letter design, letter spacing, shading and any other prominent feature which endows the genuine signature with its general pictorial effect. This study is combined with many practice attempts at the reproduction from memory, first of portions of the signature and then, when these elements have been mastered, of the signature written as a whole."

1173. The theory propounded by the Judge was a composite model simulation based on copying various parts of different models so that the natural variations appearing in the different models would then appear also in different places in the group of Wang Signatures. It was a novel theory, without any foundation, advanced by Mr. Edward Chan in his Written Submission Volume 1. This is what Mr. Chan wrote in paragraph 60 and 61 of his Volume 1:-

"60. The logic of Osborn is easy to follow. However, it is important to note that this work was written about nearly a century ago and during this time the wisdom of the forger may or may not have improved particularly after this feature has been revealed in the work and possibly in other works on handwriting. Furthermore, it is important to note that the proposition made by Osborn was predicated on forger copying from one single model. Plainly when this assumption is applicable, the forger could not have introduced into the question signatures the natural variations found within the genuine signatures of the true author. Hence, an inference of genuineness could be drawn from the presence of the same natural variations being present in both the questioned and the genuine signatures. Of course the strength of the inference of genuineness would also have to be dependent on the nature of the variations found. If those variation features are personal to the author, then the inference would be much stronger. If the features of the variations are those normally found in writings of a large number of persons, then it is submitted that the inference would be a very weak one.

61. If on the other hand the evidence should give rise to the possibility that the forger may be copying from say 2 or 3 selected samples, then it is submitted that the presence of the same features of natural variations, and more significantly, accidental features, found only amongst the 2 or 3 samples and the questioned signatures may well be the result of the copying. This inference would be more strongly drawn if amongst the samples available, the variations found within the questioned signatures are only found within the 2 or 3 particular known samples and not found from the rest."

Apart from the first sentence of paragraph 60 being omitted, paragraphs 7.6 and 7.7 of the Judgement were embarrassingly and uncritically copied, word for word, from paragraphs 60 and 61 of Volume 1 of the Written Submissions of Mr. Edward Chan.

1174. But what really called for the greatest objection by the Wife is the disregard or contempt shown to the Opening Written Submission of Mr. Martin Lee submitted on 6th September 2002. Paragraph 28 and 29 of Wife's Written Submission reads:-

"28. According to the Plaintiff's three experts, all the four questioned Wang signatures were forged by simulation; and according to two of them, Cheng and Tsui, all the four questioned Tse signatures were forged by tracing or superimposition. (Lesnevich did not say how the Tse signatures were forged).

If so, the forger would have done a perfect job in relation to the four questioned Wang signatures, because, as pointed out by Professor Jia, there were at least 9 significant variations found among the 18 specimen Wang signatures which were reflected in different ways in one or other of the four questioned Wang signatures.

And when it comes to the four questioned Tse signatures, it would be wholly inconceivable for a forger to be able to incorporate not one but 38 significant variations from among the 35 specimen signatures in different ways into one or other of the four questioned Tse signatures, and he did all that while tracing! The forger must have been a devil incarnate!

It is submitted that this could well be the single most important point in this case. For unless the Plaintiff can come up with a good answer to it, the Court would find it difficult, if not impossible, to conclude that the four questioned Wang signatures and the four questioned Tse signatures were forgeries."

The highlighting in bold font of that last paragraph by Mr. Lee of the words "single most important point" is very crucial.

1175. Volume 2 of the Written Submissions of Mr. Edward Chan was in answer to the Written Submission of Mr. Lee dated 6th Spetember 2002. There was no proper answer in Volume 2 (see page 68 where there ought to be an answer) to Mr. Lee's single most important point about impossibility for the forger to produce the large number of natural variations. Volume 1 was written by Mr. Edward Chan not in answer to Mr. Lee's 6th September 2002 Submission. Although Volume 1 and Volume 2 of Mr. Chan's Written Submissions were both dated 20th September 2002, Volume 2 was really the important Volume because that was in response to Mr. Lee's Written Submission dated 6th September, containing the words "single most important point". The Judge copied Volume 1 as has been seen. He ignored for all practical purposes Volume 2 and furthermore ignored Mr. Martin's Written Submissions dated 6th September 2002 and 11th October 2002 (in Reply).

1176. Mr. Lee's Reply Submission dated 11th October set out in paragraph 12 its strong objection to paragraph 60 of Mr. Edward Chan's Written Submission. It reads:-

"The Plaintiff attempted to belittle this difficulty as he would have to. Faced with the authoritative quote of Osborn (p. 368) cited in paragraph 59 of the Plaintiff's Main Submission that the incorporation of matching variations is "strong evidence of genuineness", the Plaintiff boldly asserted that "it is important to note that this work [of Osborn] was written about nearly a century ago and during this time the wisdom of the forger may or may not have improved particularly after this feature has been revealed in the work and possibly in other works on handwriting. Furthermore, it is important to note that the proposition made by Osborn was predicated on forger copying from one single model". The Plaintiff has adduced no evidence, whether through his three handwriting experts or otherwise, to show that the said authoritative quote from Osborn was wrong. Nor was there any evidence where the Plaintiff was able to point to one case where a forger, who was more competent and skilful than those forgers living a century ago, was able to incorporate matching natural variations of a writer in a series of forged signatures. The Plaintiff's attempt to dispute this part of Osborn which fully accords with common sense shows how desperate the Plaintiff is in trying to salvage his case in the light of the convincing evidence of Professor Jia on these points of striking matching internal variations."

The Judge did not deal with the point or even referred to it, probably because paragraph 7.6 and 7.7 of the Judgment had already been written (see later discussion of the sequence and chronology of Written Submissions and the admitted sequence of writing of different Sections of the Judgment). Therefore what the Wife considered "the single most important point" was simply ignored, unanswered and not judged upon.

1177. Plagiarism or not, the question is whether there is any substance in the multiple copying or the composite model new theory advanced by Mr. Edward Chan in his Written Submission and adopted by the Judge in paragraph 7.6 and 7.7 of the Judgment. I regret to say that I cannot find any basis in either factual or expert evidence or logic or authority or common sense to accept this novel theory. I believe the whole case went astray as result of the Judge accepting this novel theory and from that point onwards there could be no consolidated evaluation. It was simply a road down to disaster with all emphasis by the Judge and Mr. Chan on dissimilarities, significant or otherwise and whether found by the experts or by the Judge.

1178. Fundamentally what is objectionable to the new theory of multi model copying is that carried to its logical conclusion, it destroys natural variations as a means of positive identification of genuineness. It destroys meaningful value being given to significant similarities and renders illusory any final consolidated evaluation as there will be nothing on the positive side to evaluate. By this novel theory, the Judge has created a built-in loading in favour of forgery and against genuineness and this vice poisoned the whole of the handwriting analysis in the Judgment. This Fatal Methodology of Multi Model Simulation undermined the whole of the Judgment and in my view, the Forgery Judgment must be set aside with Judgment entered against the Father on Forgery.

(IV) WRONG APPROACH, WRONG APPLICATION OF METHODOLOGY & MISCONCEPTIONS

Atomisation (No Consolidated Evaluation)
& Wrong Controlling Factor

1179. Mr. Chang complained that the Judge adopted consistently a wrong approach or methodology in singling out individual points of difference as indications of forgery without "consolidated evaluation" of both similarities and differences and Mr. Chang called this flawed approach Atomisation. Mr. Chang complained further that the Judge as part of this consistent Atomisation gave in effect primacy to differences over similarities as controlling factors in the comparison of signatures. He called this Controlling Factor. This Controlling Factor as applied by the Judge was even more unfair when the opposite technique was applied in the comparison of alleged Nina handwriting. Mr. Edward Chan disputed that there was any Atomisation or that there was no proper consolidated evaluation and said there was a misunderstanding of the term Controlling Factor.

1180. The Index to Chapter 10 as well as the structure of Chapter 10 of the Judgment shows that the micro analysis (other than Tremors and Conclusions) was divided into 3 parts, namely:-

(1) Similarities (paragraph 10.4 to 10.53)

(2) Significant Differences (paragraph 10.54 to 10.57) and

(3) Judge Found Significant Differences (paragraph 10.58 to 10.85).

The Similarities Part is again divided by following a sequence of analysing under Similarities (1) the character Wang in Document A, then B, then C then D. After that under heading Similarities (2) with the character Teh, there was analysis of that character in Document A, then B, then C then D. Finally under the Similarities (3) there was the sequential analysis of the character Huei in A, B, C and D.

1181. A similar pattern is followed for the analysis of significant differences namely by each character in each Document sequentially. The same pattern is repeated for the analysis of what I will call Judge observed significant differences.

1182. It will be observed that the approach adopted in the Judgment was not:-

(1) to take Document A and then consolidate the similarities and dissimilarities of Wang and draw a mini conclusion of Wang, then do the same for Teh, then do the same for Huei and finally leading to an intermediate consolidated evaluation of the whole Signature in Document A and then proceed to do the same for Document B and then C and then D leading to a Consolidated Evaluation by the method of what I will call "Document by Document Analysis";

(2) to take all four characters Wang in turn and then consolidate the similarities and dissimilarities of the four Wangs and draw a mini conclusion on the four Wang characters and then do the same for the four Tehs and then do the same for the four Hueis leading to a consolidated Evaluation by the method of what I will call "Character by Character Analysis".

1183. The method adopted by the Judge to evaluate all the similarities and then all the differences was apparently, with each part analysed to give already value judgment pointing to forgery. The final paragraph at paragraph 10.87 was simply a conclusion and not a consolidated evaluation. There was no proper consolidated evaluation where all the factors were discussed and analysed so that the whole picture is looked at. In the light of the presence of many similarities, some of striking natural variations (with significant combination and distribution of the similarities), what might originally been thought of as a significant dissimilarity, on consolidated evaluation, could turn out to be in fact no dissimilarity but more like a variation, although slightly out of range, or capable of explanation such as accident or having been written under very different circumstances or under different conditions. The consolidated picture might then give rise to a rational and convincing well-argued reasons for a final decision which might well be Genuineness or even Don't Know and not necessarily Forgery. That is the virtue and the absolute necessity for consolidated evaluation. In my view, Mr. Denis Chang is right in his complaint that "the Judge was preoccupied with finding differences ..... and took up each individual point of apparent disparity in isolation rather than as part of a consolidated evaluation of similarities and differences at the completion of the examination." (paragraph 2.6 of Part A of Wife's Submission on Handwriting). The real complaint of the Wife is that the Judge did what Osborn said should never be done, namely adding up numbers, in this case of dissimilarities all pointing to forgery and paying only lip service (if any regard at all) to consolidated evaluation.

1184. The Atomisation (or evaluation of forgery dissimilarities before consolidated evaluation) is of course made worse by the combination of two other sins namely:-

(1) the misconception of Controlling Theory which emphasised even more the value of dissimilarities and

(2) the Fatal Methodology of Multi Models simulation which devalued all natural variations similarities.

The built-in structural bias with this Atomisation approach renders the whole exercise of consolidated evaluation of handwriting impossible and it was never done.

1185. As Mr. Chang quite rightly pointed out, the error of the method adopted by the Judge could be demonstrated by the inevitable finding that would be made by the Judge against an unknown but genuine signature of Wang. Even with a genuine signature, the Judge would on the one hand dismiss all similarities even though they may be matching variations (on the basis of multi models simulations) and on the other hand any difference of that genuine signature from the samples would be condemned if that difference falls outside the range of variations of the samples. It was by no means certain that the Samples given to the Court was comprehensive or even representative across the whole period specially those closer to the date in question. When a person has a very large range of variations with his writing as was in this case of Mr. Wang, even taking away one of the 18 Samples in Court could make a material difference to the range of variations. If one takes out say for example Sample 668 which is an important Sample and ask the experts to compare Sample 668 which is suspected to be a forgery and compare that with the other 17 Samples, then Sample 668 as a suspect, by the Judge's method, would be condemned as a Forgery on the basis that the features in Sample 668 are significantly different from the range of variations in the 17 Samples and any significant similarities were merely the product of multi models simulations. This demonstrates not only the error in the approach of the Judge but also is a lesson in warning us about the uncertainty of expert handwriting comparison. So much depends on the total and complete adequacy of the Samples (which would of course affect the range of variations), the quality of the consolidated evaluation and the honesty and the courage of the experts to say "Don't know or I am not sure".

Mechanical Approach To Dissimilarities

1186. As the finding of significant dissimilarities was the foundation of the conclusion of Forgery, understandably the fire power of Mr. Chang was directed to the way the Judge evaluated and assessed dissimilarities. The complaint is that he did it mechanically so that he treated uncritically every departure from the control Samples as significant difference pointing to Forgery.

1187. Mr. Edward Chan disputes this but I find what he said in paragraph 34 of his Written Submission very revealing. He said:-

"... When a range of variations of a feature is determined based on sufficient representative known samples, it can be used to determine whether or not the same feature in a questioned signature is out of range. If it is, then this feature is suspicious of forgery because it is dissimilar to the known samples. If sufficient number of these dissimilarities in combination is found, the questioned signature would be found to be a forgery."

1188. Mr. Chang attacked this submission and said that it revealed the very vice which affected the Judgment. The fact that there was a difference is only the beginning of a long exercise and not the end because what is necessary to consider is the quality and nature of the dissimilarity, whether it is of a nature fundamental to the structure of the writing, and whether such fundamental structural dissimilarity is consistent and is of such quality which is not capable of any reasonable explanation. In the consolidated evaluation exercise it has to be considered what is on the other side of the coin, namely do they exhibit such significant similarities of consistent and fundamental nature (such as matching variations) and that when it is taken all together, this gives rise to a conclusion not of forgery and if not of genuineness, at least to a possible answer of Don't Know. The vice exhibited in Mr. Edward Chan's paragraph 34 is the vice of adding up numbers and taking the easy route to a conclusion of Forgery. This is the approach which gives handwriting expert evidence a bad name because it lacks the sophistication of a complex reasoning which at the end must be convincing, even on a balance of probability let alone to reach the suitable high standard of proof required for Forgery.

GL Not Chinese Handwriting Expert
And Judge Playing Handwriting Expert

1189. GL, Gus Lesnevich would not have given comparison Chinese handwriting expert evidence if the Judge had set a limit of say 2 handwriting experts on each side. GL who knew no Chinese language and who had no proper full professional exposure (see his CV and notwithstanding what he said in Transcript 17:3-4) to disputed Chinese documents and its analysis should not have been allowed to give expert Chinese writing evidence. He was clearly embarrassed on many occasions and it is right that he felt he did not belong there. The Judge should have totally rejected the evidence of GL but he did not and even relied on it. He said nothing about the weight he could attach to the evidence of GL, when it was clear that in this case which demanded a high standard of proof with cogent and compelling evidence, very little weight could be given to the evidence of GL.

1190. The essence of expert handwriting evidence is the analysis of the quality of similarities and dissimilarities. The significance and value attached to analysis of similarity and dissimilarity is intimately connected not only with the knowledge of Chinese writing or the writing of its basic 8 strokes but how most persons write Chinese, not only the regular order of strokes, but the appearance of strokes, the spaces left between, the turn, the uplift, the pen movements, the irregular way of writing, the unusual and the individual, the subtle and the inconspicuousness of Chinese handwriting. All this requires years of experience and exposure to every kind of Chinese handwriting, with pen of every kind including the traditional writing with brush and ink. With all respect to GL, he simply was not equipped to tell the Court when a dissimilarity is significant or not because he could not even say whether or how much it is a departure from the regular way of writing. He could not say whether a feature is a departure from a unique/individual way of writing which is so individual and inconspicuous in its combination of "dent" and "scratch", so that great value must be placed on such significant dissimilarity of such feature in that particular Chinese signature. For me, GL was there to add to the numbers (of experts), a constant vice of the Father's team. He not only did not add to the acceptability of the Judgment but his presence and his evidence and the acceptance of it, made the Judgment unacceptable.

1191. The Judge did not confine his error to relying on GL as a Chinese handwriting expert. The Judge participated actively in the handwriting contest. Mr. Denis Chang accused the Judge of playing expert and then getting it wrong. In this part of the Judgment, I limit the discussion as to whether the Judge was playing as expert and whether that is objectionable.

1192. The unsuitability of relying on GL as expert on Chinese handwriting is his ignorance of the Chinese writing and his lack of experience in examining over long years large number of Chinese documents as part of the professional life of a Chinese handwriting expert. The Judge did not suffer from the first handicap but he does suffer from the second handicap. The critical aspect of the case is the value to be attached to the existence or absence of certain features in the questioned documents compared to the Samples. The essence of the consolidated evaluation is the value and quality evaluation of the handwriting comparison of similarities and dissimilarities. The reason why Tsui, Cheng and Jia were allowed to give expert evidence is because of their long exposure to Chinese handwriting so that they could say, not like any ordinary Chinese, what are the features relating to the writing of the radical say double man. It is for the Chinese handwriting experts to say not just what is the way to write double man, but how usual it is to write double man with the unusual one slant (diagonal stroke falling from right to left) and not the usual two slants, and how common or uncommon (what percentage) it is for Chinese writers to combine two ways of writing with one slant and also with two slants.

1193. A Judge might have some limited exposure to Chinese handwriting (varying from Judge to Judge) but he does not have the training as a handwriting expert and certainly not the long exposure and experience to Chinese documents (suspect or non-suspect). Furthermore his view cannot be cross-examined and it is invidious for Counsel or for the Wife's expert to be saying to the Judge all the time (not on a matter of law) but on a matter as personal as the Judge's own experience of handwriting that the Judge is wrong. A Judge should never allow himself to wander into that kind of swamp. A Judge should not play expert and make findings of dissimilarities or retouching based on his observed and personally initiated features of the handwriting which he termed significant. This is specially so when none of the experts expressed any view on such Judge observed features thereby suggesting either that all the experts are incompetent (if the Judge is right) or if they are professional that the Judge was wrong. The implication of the former case is that there can be little weight given to the evidence of the experts bearing in mind particularly the high standard of cogency required. The implication of the latter case is that the Judge should not be playing handwriting expert and playing against the experts and then declaring that he has won against the Wife's expert. One is either a player or a Judge but not both.

1194. This is not a situation where it was an isolated one-off observation of the Judge. The Judgment was based on a large collection of Judge played expert significant dissimilarities and re-touchings pronounced as such by the Judge on his own initiative. The Written Submission of Mr. Chang contained chapter and verse of the instances when the Judge played expert (9th Submission and Section 5 of Wife's Submission). The Judge believing himself to be superior to not just the Wife's experts but even the Father's experts, never warned himself as to the reasons for the silence of the Father's experts on features which only the Judge emphasied nor did he ask the Father's experts to give him expert assistance on those aspects. This is serious. But what is even more serious is that as result of his playing expert and reaching a wrong view, he took very adverse views on Jia and made findings against Jia which are clearly wrong in the circumstances as well as being unfair (10th Submission, Section 10 of Wife's Submission).

1195. Once the case has reached the stage where the Judge was himself contributing to the evidence of what are contested significant dissimilarities and what are contested retouchings, the case was obviously no longer one with the Judge acting as the impartial adjudicator but with the Judge acting as a witness who however could not be cross-examined and who was no expert and what is worst who was going to decide (judge) that he was right. No Judgment given with such participation by the Judge could be allowed to stand. It must be set aside.

Misconceptions of Implications of Calligraphic Pen

1196. Considerable criticism was directed towards the Judge for his failure to properly appreciate the implications of the use of the calligraphic pen and for his signal failure even to consider in his Judgment the evidence on the effect of the calligraphic pen. There was of course no dispute that all four Wang Signatures were written with a calligraphic pen (to be contrasted with no Sample having been written with similar calligraphic pen, the closest writing instrument being Sample 668, 669 written with fountain pen and Sample 695 probably written with a fountain pen).

1197. Harrison as usual sheds light on the importance of the writing instrument. He said at page 331-2 under the Heading "Influence of the Writing Instrument":-

"The question as to what extent handwriting is influenced by the nature of the writing instrument is often asked, because it is sometimes impossible to obtain handwriting written with a similar writing instrument for comparison with that of anonymous letters or other suspect documents. It is obvious that the writing instrument must have an effect which extends beyond that of mere difference in appearance; a fine flexible pen can be used to produce handwriting characterised by a clear outline, marked shading and the reproduction of the finest details of letter design, no matter how small the scale on which it has been written. On the other hand, a blunt pencil will defeat the efforts of the most skilled to include either shading or the fine detail in the construction of the letters. The limit is reached when a ball-point pen is used; this implement has all the faults of the blunt pencil together with others peculiar to itself."

Then at page 333 he said:-

"...... As is pointed out in some detail in the chapter on forged signatures, the use of the ball-point pen makes the detection of forgery much less certain, for here the finest details of handwriting structure are all-important."

Further he said at page 345 under the Heading "The Comparison of Handwriting" about the relevance of the writing instrument:-

"Apart from disguise and variation, there are a number of "reasonable explanations" for the appearance of consistent dissimilarities in specimens of handwriting written by the same person. These are generally concerned with the conditions under which the writing act was carried out. The physical conditions, such as the nature of the pen and other writing materials as well as the position taken up by the writer, play their part to say nothing of the physical and mental state of the writer himself."

1198. The full graphic details of writing including shading and the contour of turning are clearly shown when calligraphic pen with a broad nib is used. Harrison described the significance and difference of the usual ball-pen as opposed to ordinary foundation pen (calligraphic pen would of course take shading one stage further) in the following passage at page 404-407:-

"Shading

Some signatures of today, like many of those of a previous generation, are written with due regard being paid to the capabilities of the pen as a medium of artistic expression. These signatures not only exhibit a graceful succession of letter designs, but also utilise to the full the ability of the pen to produce variations in the width of the ink line. This variation in width, produced by control of the pen position with respect to the direction of the pen motion and of the pressure exerted, is known as shading. Not only do these shaded signatures grace any document on which they are found, but they provide a most difficult problem for the would-be forger, however skilful he may be.

........

Unfortunately, shading in most contemporaneous handwriting is most conspicuous by its absence. The advent of the ball-point pen, with its symmetrical writing point which makes a deeper rather than a wider line on the application of excess pressure, has done much to banish shading from contemporary handwriting and so to facilitate the task of a forger.

It matters little how a pencil or a ball-point pen is held; the line remains practically unaltered whatever the angle, and the amount of turning of the pen in the fingers as the writer proceeds. This is yet another reason why important signatures should never be written with other than the conventional type of pen. Many forgeries made with a ball-point pen remain undetected through lack of evidence to show how they were written."

1199. What can be drawn from Harrison's observations are the following;-

(a) Forgers often use ball-point pens to write the forged suspect documents because documents written with calligraphic pens show the greatest true details. It is almost axiomatic that no forger will use a calligraphic pen to set himself an impossible task and to give himself away. To use a calligraphic pen and to execute four signatures with internal variations suggests not a forger but the real person, who had no fear to write naturally and without concern, displaying all the shading and calligraphic flavour in four separate signatures and with no regard for being found out as his writing is the real thing;

(b) The writing with different instruments will produce different results not only in terms of appearance but also in terms of details including shading;

(c) Consistent dissimilarities by the same writer are explainable by the difference in the nature of the writing instruments;

(d) Often it is impossible to compare documents written with similar writing instruments but comparing like with like is important.

(e) When comparing like not with like, then caution must be exercised in the analysis and evaluation of dissimilarities; the difference in the writing instrument often furnishes the reasonable explanation for the dissimilarities.

1200. By definition, because of the ability of the calligraphic pen to show not only shading but the finest details, a document written with a calligraphic pen will display more embellished or artistic writing than a document written by the same person with say a ball-point pen. It is for that reason we always see important public documents such as treaties executed by heads of state in public view being signed with special foundation pens or pens with suitable nibs and never, never with ball-point pens.

1201. The evidence also show that ball-point pens write faster than calligraphic pens and the tendency with calligraphic pen is to write a bit upwards (per Tsui at T70:63:14-64:2).

1202. In relation to the important first horizontal stroke in the character Wang which was characterised by the Judge as a significant difference at paragraph 10.55(1), the Judge ignored the evidence and implication of the calligraphic pen. The Judge made no allowance for reasonable explanation because of the different writing instrument and because no other Sample was written with a calligraphic pen. Another example of the Judge failing to take into account the calligraphic pen is at paragraph 10.55 of the Judgment where the Judge characterised the middle loop being virtually closed in the questioned Wang signatures as significant dissimilarity. The Judge ignored the evidence of Tsui that it was more difficult to produce an unlinked space with a broad nib calligraphic pen.

1203. It was said by Mr. Edward Chan that the Judge was unfairly criticised in relation to the subject of calligraphic pen because he was very familiar with the usage and special features of calligraphic pen. If that is the case, then it is all the more important for him to state in his Judgment what the Judge considered to be the limit and scope of the use of the calligraphic pen and squarely deal with the point raised by the Wife that furnished a reasonable explanation for some at least of the significant dissimilarities found by the Judge. When a Judge resorted to his own personal experience and knowledge (not generally shared by everyone and accepted by everyone) then justice demands that in his Judgment he should state his findings based on his own personal experience and knowledge so that the point can be properly examined for its validity. No one can assume that a Judge is always right in either his experience or observation or in the logical connection between his experience/observation and his conclusion or inference drawn from such experience.

1204. What is disturbing and undoubtedly open to justified criticism is that the Judge in copying extensively and indiscriminately from Volume 1 of the Father's Written Submission, had totally ignored the implications of the calligraphic pen. It is extremely regrettable that in the most important Handwriting Chapters 6, 7, 8, 9, 10 and 11 in Section II of his Judgment, there was no single reference to calligraphic pen, let alone the effect of the calligraphic pen on the questioned Wang signatures. The Wife cannot be blamed in the circumstances for complaining that justice had not been done to her case.

Misconceptions of Implications of Tremors

1205. Had the Judge not found it necessary to add Tremors as an additional reason for his finding of Forgery of the Wang Signatures (paragraph 11.18), there would be no necessity to ponder over the implications of tremors found in the Wang questioned Signatures. The experts placed no importance on them. Even the much copied Volume 1 of Mr. Edward Chan's Written Submission did not go so far as the Judge did towards Forgery. It stopped at linking tremors to "psychological cause such as being anxious or exciting, which would be just the sort of psychological condition of a forger". The Judge took it one stage further and drew the inference from the existence of the tremors to psychological cause of anxiety and excitement and then drew the further inference that anxiety and excitement was the psychological condition of a forger.

1206. I regret to say that for me it was a non-sequitur. The psychological condition of anxiety and excitement could be caused by any number of reasons, angry or annoyed for something which had happened, anxious because of waiting for some important news or any number of reasons which made a person anxious or excited. Anxiety and excitement is not an exclusive monopoly of the forger. Unfortunately at various times we all shared those emotion and tremors might be reflected in our writing as a result.

1207. The Judge was of course entitled on the medical evidence to find that the fall did not cause the tremors. The Judge was also entitled to find that the tremors were caused by psychological causes such as anxiety and excitement. But what he was not entitled to do was, having regard to the high standard of proof required and that the experts did not speak in favour of tremors pointing to forgery, to infer that the tremors were caused by the anxiety and excitement of a forger when he was writing slowly.

1208. The flaw in the reasoning of the Judge was that he did not appreciate that the evidence has to point to the presence of tremors being indicative of Forgery. If there was no such evidence then the mere presence of tremors was no value and was certainly not probative of Forgery.

1209. Further in respect of the fall, the Judge was too ready to draw the wrong inference. Although the medical evidence indicates that the fall physically would not cause the minor tremors in the questioned signatures, that does not mean the trauma of the fall could not have caused anxiety and excitement in Wang after he left hospital with minor injuries. In my view, the evidence was really too insubstantial (not cogent or compelling) for the Judge to draw the inference of Forgery from the presence of tremors.

1210. In relation to tremors in Tse Signatures, the errors of the Judge were of a different order. He came to the conclusion that the disputed Documents contained tremors and that the contemporaneous 1990 Samples did not exhibit tremors and therefore that the disputed signatures exhibiting tremors are Forgeries. Again the error was caused by a non- appreciation of the complexity of the analysis required. The tremors present in the Disputed Signatures were in the long radicals "san" (身) and "chuen" (寸) in "Tse" (謝). The presence of tremors in these common strokes, mostly in the long vertical strokes, was indicative of genuine signatures rather than anxiety tremors of a forger and certainly inconsistent with forgery by tracing.

1211. What the Judge ignored was that there were tremors also found (as admitted in the evidence of GL, Cheng and Tsui) in the same common downward slow strokes in the Sample signatures. So in the words of Mr. Denis Chang there was symmetry in the location of the tremors which is strongest indication of Genuineness and not Forgery. This is again one of the many instances in this case, where the Judge turned what was strong evidence of Genuineness into conclusion of Forgery.

1212. What the Judge failed to realise is that these tremors occurred in the down strokes in both the Samples and the Disputed Signatures were written in relatively slow speed. What the Judge considered to be the significant 1990 Samples with no tremors were written in faster speed (undisputed evidence). The Judge paid attention only to identity of time (1990) but paid no regard to identity of speed.

1213. It was the same slow speed in the same down strokes which exhibited the tremors. That identity of tremors in the similar down strokes could not possibly be produced by tracing action. As Osborne said at page 329:-

"It is almost impossible by the tracing method successfully to imitate natural tremor."

Not only was the Judge wrong in his conclusion of Forgery by the presence of the tremors in the Tse Signatures, he was wrong in failing to find that the presence of these similar down strokes tremors was highly indicative of Genuineness and thereby disproving his own theory of Forgery by Tracing.

(V) SERIOUS ERRORS OF MICRO ANALAYSIS

1214. Having regard to the serious errors of Methodology, of Wrong Approaches, of Wrong Applications of Methodology and of Misconceptions, the Judgment on Handwriting with a Conclusion of Forgery of both the Wang and Tse Signatures cannot possibly stand and must be set aside. In this part of the Judgment, it only remains for me as briefly as I can to indicate as illustration and not as comprehensive review, the errors of the Judgment on a micro level (namely the detail analysis of the individual parts of each of the 8 questioned Signatures). Having regard to the amount of material put to this Court and to Yam, J. a comprehensive review on the micro level of the errors of handwriting analysis would take up an inordinate large amount of time and space. I am therefore confining my analysis on the micro level to a small selection as illustration that the Judge was in error on his analysis of every one of the 6 characters (similarities and dissimilarities) where he held against the Wife - (1) Wang (2) Teh (3) Huei (4) Tse (5) Ping (6) Yim.

Wang Signatures - Preliminary Observations

1215. The preliminary observation to be made about the Wang Signatures is firstly, that the 18 Samples of Wang Signatures all display a large range of variations. Teddy Wang was obviously not a man of tediously uniform or fixed writing movement.

1216. The second point which is of particular significance is that the four questioned Wang Signatures agreed to be all from one hand, exhibit very substantial range (intra the four Signatures) of variations. This suggests not only that all four of these Signatures are from the same hand but that the range of variations from the same hand is more likely to be consistent with the range of variations from the 18 Samples and furthermore is more likely to be the result of natural variations of the same person rather than studied variations deliberately introduced by way of Forgery simulations. The reason for this is that simulation is difficult and to simulate deliberately four signatures with different variations would call for skills and daring of extraordinary level. This is of course the "single most important point" of Mr. Martin Lee. The minute micro analysis of each of the 4 Wang Signatures must be done in the context of the above observations.

1217. Thirdly it is to be noted that because of the large variations in the genuine signatures of Wang, it is vitally important to pick out features of Wang signatures which are individual and unique to him rather than non- unique features which are commonly seen in the writing of these characters by other people. This is particularly important in relation to dissimilarities. It is not enough merely to say that there is a dissimilarity which is outside the range but it is necessary to give significance to the particular feature in the Samples which cannot be found in the Disputed Signatures. The more unique that feature which can be said to be an "identifiable dent" or an "identifiable scratch", the more significant is the dissimilarity. And if that feature is structurally fundamental to the writing in the Samples of that character and is consistently to be seen in all the Samples throughout the whole period then the absence of such unique identifiable structural fundamental feature will carry weight in the assessment. Equally the feature that is said to be found in the Disputed Signatures but not in the Samples must also be of uncommon characteristic.

1218. Fourthly in my view, the failure of the Judgment is that nowhere does it appear an analysis of any feature of Wang Signature which is said to be so unique and individual to him and so structurally fundamental to his writing that such feature can be said to be an "identifiable dent" or "identifiable scratch". I agree with my Lord Yeung that many of the features discussed in the Judgment are common features and value can never be given to dissimilarity of a common feature and it follows that such dissimilarity of common feature can never be elevated to significant dissimilarity. The impression given by the analysis of dissimilarities in the Judgement (apart from retouching or so called unnaturalness) is that it was really a method of adding up numbers in terms of one dissimilarity of a minute common feature followed by another dissimilarity of another minute feature. Adding these dissimilarities can hardly be convincing as to persuade that there was Forgery.

1219. Fifthly, in respect of Matching Variations, the uniqueness of the individual feature is not the key factor but the uniqueness of the combination of features (even common) is what is important. If an analogy is to be drawn I might refer to the features of a face. Thin lips is not uncommon. Nor is long lips. Thick eyebrows is quite common. So is short eyebrows. But a combination of short thick eyebrows and long thin lips would be rare and can be of great value in identifying a person. The same can be said about Matching Variations. It is not the commonness of each feature in Matching Variations but the pairing of common features in combination which gives significance to Matching Variations. Of course if in the combination, there is also a unique identifiable feature coupled with say a common feature, then such Matching Variation could be a clincher and one such Matching Variation might well win the match not just the game or the set.

1220. Sixthly, it seems to me that reference to a different or second style of writing of Wang is distracting the attention from the true issues. Stylistic labelling of particular strokes does not help in the evaluation of the individual and unique characteristics of Wang signatures and it is a proper understanding of the individual and unique characteristics of Wang signatures which lies at the heart of the evaluation exercise.

1221. It is said that none of the 4 suspect Signatures exhibit any significant similarities to the 18 Samples but that they exhibit many significant dissimilarities. Instead of the methodology adopted in the Judgment of analysing sequentially all the similarities of each character and then in a separate part, sequentially all the significant differences and then in yet another part all the Judge observed Further Differences, I propose to consider each character in turn: to consider the similarity and dissimilarity and evaluate whether in relation to that character what value judgment can be placed on the suspected Signatures in relation to that character. The exercise will then continue with the 2nd character and then the 3rd character.

(1) Wang - Similarities and Dissmilarities

1222. The Judge was very dismissive of the similarities of the character Wang and said there was no significant similarity all. I believe the Judge was wrong in that assessment. A striking combination of features is:-

(1) the first horizontal and the second vertical stroke of Wang in separate strokes and did not join together ("1st Feature"), and

(2) the first horizontal stroke is always longer than the bottom horizontal stroke ("2nd Feature").

The combination of the 1st & 2nd Features can be seen in Slide D11 below:-

1223. As my Lord Yeung has observed, to write the first horizontal stroke always longer than the bottom horizontal stroke is not the normal way of writing. By far the majority of Chinese writers would write the first horizontal stroke shorter than the last stroke which tends to be the longest of the three horizontal strokes. This is just in the nature of Chinese writing so as to give the character Wang a better appearance. A writing of Wang with the first horizontal stroke longer than the last horizontal stroke is a perfect application of Osborn's first Principle set out in Osborn Quotation 5, namely the characteristics which is most divergent from the regular system has the most force.

1224. What is even more important is the fact that as observed by Jia and agreed by Tsui, that this very unusual way of the writing the 1st and 3rd horizontal stroke of Wang was in the structure of the writing and this is an application of the fundamental principle of Harrison set out in the Harrison Quotation.

1225. From the above observations, what emerges is that not only is there a combination of the 1st Feature and the 2nd Feature which is observable in all the Questioned Signatures and all the Samples but that the 2nd Feature is of a unique and individual characteristics in the application of the first Principle of Osborne and of fundamental structural nature in the application of Harrison principle so that the combination containing a feature of the writing which is individual and unique of a fundamental structure is very strong proof of Genuineness.

1226. But is this powerful similarity combination to be ignored because of the theory that a forger will notice this unique combination. But here we are not talking about forgery tracing but forgery by simulation. If Sample 668 was the model then why would a forger know that first horizontal stroke is longer than the last horizontal stroke and if he did not know then it would not be possible for such forger to produce Signature B and C where the last horizontal stroke was clearly very much shorter than the first horizontal stroke. If the model is Sample 669 then how does one account for the almost equally long last horizontal stroke in Signature A and what is even more remarkable is how does one account for the last horizontal stroke undulation in Signature A, C and D which is so absent from Sample 669.

1227. This is where Mr. Edward Chan was driven to his multi models simulation theory because of the variations in the four Signatures which is impossible to obtain from a forger, if a forger is stupid enough in the first place to write four separate Signatures. The Father simply had a totally non-viable case on Forgery and that is why Mr. Chan was driven to his bizarre multi models simulation theory. In my view the unique combination of the 1st and 2nd Feature in the Wang character proves persuasively the uniqueness of the four questioned Wang Signatures.

1228. I now turn to the dissimilarities. The Judge found at paragraph 10.55(2) the large gap between the first horizontal stroke and the beginning of the vertical stroke to be a dissimilarity. In all the control Samples there were gaps only in Sample 668, 669, 5997 (small gap), 5999 (small gap). If one writes with two separate strokes the first horizontal stroke followed by the vertical stroke then it is common to have a gap between the horizontal stroke and the vertical stroke. It is to be noticed that the gap in the 4 disputed Signatures vary, largest in D smallest in A. In the same way there is a range of gap in the control Samples, largest in 669, smallest in 5999. Jia said in his evidence that there was nothing remarkable in the fact of the gap which is perfectly consistently with both Cheng and Tsui who said nothing about the gap in their reports. The gap and no gap and the size of the gap is simply a feature of the wide variations in the Sample signatures of Wang. There is therefore nothing remarkable about the existence of the gap let alone this being an unique feature which is very noticeably different from the usual way people write. There is for me no significant difference. The Judge in copying from Volume 1 was simply creating a number to add.

1229. At paragraph 10.59 to 10.61 the Judge made stinging findings of 2 Judge-Observed significant dissimilarities of the same document (Signature A) and further accused Jia of giving dishonest answer. The features in question are firstly (see Slide E1-2) the downward movement vertical stroke appearing to be on top of the upward stroke ("Intersection Feature") and secondly (see Slide E1-6) the eyelet of the lowest stroke being composed of two writing movements ("Eyelet Feature"). In relation to the Intersection Feature, the observation of the Judge was that there was retouching of the vertical stroke over the intersection point. In relation to the Eyelet Feature, the Judge observation (not based on microscope) was that the eyelet was in fact written in two strokes, as indicated by the darker colour of the upward strokes.

1230. Jia who examined the original under the microscope, did not observe any difference in the ink colour for both of these strokes and could not see any sign of retouching. He gave cogent evidence on both features to the effect that there was nothing sinister in these two features. He said in relation to the Intersection Feature that the ink was heavy and thick for the vertical stroke whereas the horizontal stroke written faster and lighter and hence could not cover the vertical stroke. In relation to the Eyelet Feature, Jia said that the eyelet could not have been caused by two writing movement because it was not possible that the strokes could be linked up in such perfect manner. The Judge was looking at the enlarged photograph (used by the Plaintiff's counsel) which might not have been accurate because much depends on how the colours came out on each development. The Judge was also relying on the image which was projected onto the screen in Court. The difference in the colour seen under the microscope or on the screen could be changed by adjusting the lighting source under a well-known optical illusion (Harrison at page 226, Osborn at page 510).

1231. In these circumstances, the Judge should have been extremely careful of the possible distortion caused by the lighting. But he failed to take that factor into account in his findings. It was therefore entirely wrong for the Judge to reject Jia's evidence, particularly when the Judge was relying on the projected images on the screen, while Jia was testifying on the basis of what he saw from looking at the microscope. Typically instead of jumping to conclusion that the witness Jia might be honest and there was a difference in perception because they were watching the feature through different media, the Judge accused Jia of being dishonest or less than a satisfactory witness which assessment coloured his view of Jia. This was both unfair to Jia and to the cause of the Wife. There should not be any finding of dissimilarity.

1232. In coming hastily to his conclusion of significant dissimilarity, the Judge should take heed of the fact that none of the experts regarded these aspects as being significant either in their reports or in their oral evidence. It was a cause espoused only by the Judge. Secondly, in playing expert the Judge had overlooked the concession of Counsel for the Father that on a point where there was no expert evidence from the Father, if Jia disagreed with the point suggested to Jia, "that is the end of the matter" (T96:67:24-68:12). There simply was no evidence to contradict what Jia said except the observation of the Judge who could not be cross-examined. Thirdly the Judge should not draw adverse conclusion of Forgery retouching when handwriting expertise could have informed him what Osborn and Harrison said on retouching:-

"When there is no pen-lift and no second application of ink the greater quantity of ink at a shading or the slightly retraced line at an angle will show a sudden change in density that may be incorrectly interpreted as a retouching or a second application of the pen." (Osborn, page 372)

"It must be emphasized that the presence of retouching or of modified letter design in a signature is by no means an infallible sign that the signature is fraudulent." (Harrison, page 403)

Mr. Edward Chan says that the Optical Illustration passages and possibly retouching passages were not put before the Judge but these are all well known handwriting literature which the Judge if he was playing as expert should know. The whole point of this is, that the Judge had no business playing expert and could not on the one hand hide behind his ignorance of the literature and yet still played expert. Chinese Handwriting expert territory is where "angels fear to tread". Fourthly, the Judge ought to remind himself that this is an isolated alleged feature in only one Disputed Signature and was not a phenomenon observed in any of the other 3 Disputed Signatures. The Judge should ask himself why there was the need for retouching of only Signature A and not the same need to retouch the other 3 Disputed Signatures and further if there was the need of the forger to retouch Signature A, why would Signature A be put forward rather than simply rely on the other 3 Suspect Signatures which had not been retouched.

1233. It is clear that even on a micro level, a proper analysis of the 4 Wang Signatures requires a careful evaluation involving critical study and analysis of all the Wang signatures (suspect and non suspect) and an overall consideration having regard to the significance of the various features observed. It is never a matter of adding up numbers of dissimilarities. If that exercise of consolidated evaluation of only Wang character had been done, then no doubt a conclusion of genuineness would have been reached. Even taking into account the differences from observation between the Judge without the microscope and Jia with the microscope, the Judge should not in the circumstances point to Forgery against the character of Wang. Further Consolidated Evaluation of all four characters would undoubtedly indicate more forcefully towards Genuineness.

(2) Teh - Similarities and Dissmilarities

1234. The "double men" radical (彳) in Signatures A, B, C, D show two ways of writing the initial slant. From the left, as in Signature C and Signature D and as seen in Sample 669, 6003. From the right, as in Signature A and Signature B and as seen in 668, 672, 5997, 6002 and others. This is a good illustration of Matching Variations as illustrated in slide D1 of Mr. Denis Chang which is reproduced below:-

1235. The Judge made no reference to this Matching Variation of the double men radical but instead at paragraph 10.34 accused Jia of dishonesty because of what Jia said about the double men radicals in Signature D compared to Sample 668 and 669. What caused the Judge to accuse Jia of dishonesty was what the Judge perceived to be the alleged double standard applied in relation to the "direction of the commencement of the writing movement of a stroke". He referred to the direction of the vertical stroke of the character Wang (王) and the direction of the initial stroke of the double men radical. It was a false point of the Judge because he really did not properly understand the subtlety of the points being made by Jia arising out of the subtleties of the Matching Variations in relation to at least 2 features of the double men radicals. The following extract passage from the Transcript at Day 102 page 68 to 69 is significant.

"Q. So if the way that a stroke is being initiated is of importance in deciding on the authenticity of handwriting, then surely this would be an important point of difference?

A. You just look at 668 and 669, and they are written in different ways, and would you say that these two were in fact not written by Mr Wang?

Q. So you disagree with that proposition of mine; right?

A. That is correct"

1236. What is not properly appreciated by the Judge is that Jia was talking in terms of the admitted variations of Wang in the Samples, of ways of writing the double men radicals as can be seen from Sample 668 and 669, if not also in many of the other Samples. Given that, it is therefore not right to confine comparison of Signature D to 668, although of course a strict comparison of the initial stroke of the double men of these two (Signature D and Sample 668) does show a difference. But it is a difference of no significance and in fact as I said earlier, the whole exercise is a demonstration of genuineness by reason of Matching Variations. The Judge focused on a point from a different context and then accused Jia of being dishonest.

1237. What the Judge ought to have done on the very subject of the double men radical, the evidence on which caused him to make the unfair accusation, was to consider the many Matching Variations arising from that radical. Slide D1 referred to earlier was the first Matching Variation. Slide D2 as reproduced below demonstrates the second variation. That variation consists of the way the down stroke ends, either by pen pressure at the end with no hook or down stroke ends with a hook. Signatures B, C and D are with hooks but Signature A is without hook and with pen pressure at the end. Sample 668, 695 and 6005 are with hooks whereas Sample 6002, 6603 are with pen pressure and no hook.

1238. This earlier discussion serves to illustrate the danger of the narrow focus of the Judge on minute detail and the error which can be made by his failure to see the larger picture. What is required is of course consolidated evaluation of differences and similarities and to consider at every point (where he got repeatedly confused) of whether apparent difference is to be treated as variation (indicating Genuineness) rather than difference (indicating Forgery). Variation and specially Matching Variation is indicative of Geuineness, the opposite of Forgery. What the Judge was confronted with in this unique case is a set of 4 Signatures A, B, C, D from the admitted same hand but containing variations. The road signs were repeatedly showing the Judge Matching Variations and therefore Genuineness, not significant differences indicative of Forgery.

1239. The Judge placed emphasis on the significant difference of the last dot of "心" being separately written and that the dot is large and emphatic specially for Signature A and Signature B. The point of course has to be considered firstly in the context of the variations of the last dot already contained in the 4 Suspect Signatures. Samples 668, 669 and 5999 (for me not a link dot) all exhibit the separate dot and this is to be contrasted with no separate dot in the other Samples such as 695, 6006, 6003. There is nothing remarkable having regard to the variations for the separate dot to appear. The 4 Suspected Signatures were all written with a calligraphic pen therefore producing slower speed in the same way as Sample 668 and 669 produced with slower speed contained the separate dot. I do not attach that much emphasis to the fact that the pen wielding movement of the radicals were all the same. But I do place emphasis on the fact that even Cheng agreed that this separate dot is within the range of variation in the Sample controls and that it is consistent with the broad nib of the calligraphic pen. The separate dot is simply not a difference let alone a significant difference or a consistent significant different of a fundamental structural nature.

1240. The dot on the other (left) side of the radical "sum" (心) also gave rise to the Judge-Observed Further Significant Difference at paragraph 10.62-66 that the left dot in (心) of Signature A (nothing affecting the other 3 Signatures) was retouched creating a bulge. It is of course to be noted that none of the Father's experts said anything about retouching at that location in that particular document. This is yet again another example of the Judgment going against the concession of the Father that the Father's side will accept Jia's evidence in the absence of any from the Father's experts.

1241. The evidence in this case was somewhat involved, at times confusing and straddled over two days (Day 96 and 97). The alleged extra bulge in the left dot was the point in question and was suspected to be the subject of unnatural retouching. In fact the explanation was, in the end, quite reasonable and with the aid of two reverse photocopies of the Signature A at page 2C 419 and 420 (see also Slides E1-7 and E1-8) and a slow reading of the relevant transcript, there is really every good reason to accept that evidence and the explanation of Jia.

1242. The evidence of Jia was that there was retracing at the horizontal stroke at the base of "four" (四), which then formed the slant and ended at the so-called "bulge". The left dot was written just slightly above the slant giving rise to the small triangular shape above the "bulge" which was the start of the dot. This was then followed by the 3rd stroke, namely the slant hook "che ou" (斜勾) in the middle of the radical "sum" (心) which started slightly above the left edge of the slant (see T97:6:13-18 & T97:30:7-11). The description of the way the strokes were written was based on the form and shape of the dot, because there was no difference in the colour tone, which is a good indication of no re-touching. According to Jia, the distribution of the ink was fairly even, and there was no part which was displaced or unmatched, caused by writing twice (T96:82:7-10).

1243. To support the contention that there was nothing unnatural about that part of Signature A, reliance was placed on the 2 Samples of Wang's signature which showed this same feature of an "extra" left dot, namely Sample 6002 and the "written name" of Wang on the 1968 Will. These Samples (at least 6002) show that this very subtle feature found in Signature A was in fact a significant similarity - and one that was not found in either Sample 668 or Sample 669. The Judge however turned what was significant similarity into significant difference on the basis that in both those cases the left dot was written separately from the slant (paragraph 10.66). But regard should be had to the wide range of variations of Teddy and it was quite consistent with that wide range that he wrote without the extra dot most times and with the extra dot on few occasions. The fact that even with the extra dot, it would practically merge with the slant as in Signature A or not quite merge with the slant (more visible on the slant) is missing the subtle point of the variation of the extra dot. It is to be noted that both in Sample 6002 and in Signature A, the dot straddles over the slant and it is this subtle feature which therefore calls for consolidated evaluation of quality assessment as opposed to the mechanical approach to difference.

1244. To elevate this feature to significant dissimilarity in the circumstances is quite erroneous, specially in the absence of positive evidence from the Father's experts and against the steadfast and reasoned evidence of Jia and in the face of valid Samples and this being an one off situation applying only to Signature A. The Judge was therefore wholly unjustified in his conclusion of significant dissimilarity and even more erroneous in his grossly unfair criticism of Jia. Yet again, the Judge turned what ought to be a finding of Genuineness into a finding of Forgery.

1245. Finally it seems appropriate just to say a brief word about the compressed loop of the Huei in the Signature C which the Judge found at paragraph 10.82 to 10.85 to have been retouched at the right top of that compressed loop. (see Slide E1-10A for illustration). Again there was no evidence from the Father's experts on this Judge-Observed re-touching. Jia explained that it was due to ink fusion with the use of the calligraphic pen which caused the patch and that there was no retouching. (T98:37:17-47:8; T109:28:8-31:15). There was no evidence to contradict what Jia said and there was no expert evidence of retouching. But the Judge preferred to do his detective work and then by playing the expert on questions such as ink infusion and behaviour of broad-nibbed pen, quite unjustifiably rejected the evidence of Jia. Basing himself on his own expertise which cannot be cross-examined, the Judge went on to find that there was significant dissimilarity because of his conclusion of alleged retouching.

1246. The whole approach of the Judge was wrong. He tried to do experiment which was not part of the Father's case. He did not put his own final theory to the expert. Then he forgot large part of the evidence for the Wife. He came to a conclusion which was against the evidence. Then finally he went too far when he did not copy Volume 1 (paragraph 10.85 was not copied, 10.82 to 10.84 were copied). This is yet one more illustration of how again and again, the Judge went out alone, in the absence of Plaintiff's expert evidence, in a section of his Judgment called Further Significant Dissimilarities, allowed a case of Forgery to be built based on the Judge's expertise and not based on the evidence of the experts. No compelling case of Forgery can in these circumstances be properly established. In my view, the case on the character of Teh of the Father is weak and on evaluation it should have been held that no case of significant dissimilarities had been made out. Rather the reverse should have found, namely a strong case of Genuineness.

(3) Huei - Similarities and Dissmilarities

1247. It is surprising that even with this character, there was substantial debate as to whether there were significant similarities, matching variations and/or significant dissimilarities.

1248. In terms of similarities, Slide D10 purports to show similarities in the down stroke of 光 in 輝. The range of the down stroke, from right to left in Signature A, B and D which echoes the same from right to left in Samples 668 and 6006 is to be contrasted with those from left to right in Signature C which is echoed in Sample 669 and 6002. The point might not be overwhelming but nevertheless valid.

1249. However, when one turns to the matching variations, the case for significant similarities is strong. Chart 47 at 1C 215 (see Slide D8 reproduced below) is a Matching Variation involving 3 ways to write connected movement in the middle of 軍 as can be seen in the 4 questioned Signatures.

The point is made clear when read with the Transcript at T77: 40:-

"Q. Chart 47, please.

A. This chart is to show the connected stroke movement feature in this part in the middle of the radical "gwan" in this character in the four questioned signatures. If we look carefully at the connected stroke movements found in these four questioned signatures, generally speaking, such feature is exhibited in three ways. Such feature is consistent with those found in Mr Wang's signature samples. For example, in A, the connected stroke movement is this: a compressed loop in the middle. This connected stroke movement feature is the same as samples 5-1 and 5-3. The second way, as shown in B and D, is this: going up in a connected stroke movement and then to execute the stroke like an Arabic numeral 3. Such feature is also consistent with that shown in 669 and sample 2-1. The third way is this, as shown in C: the connected stroke movement is executed in a slanting manner, in a slanting manner, and then goes downwards. It is consistent with DPH-668 and 695. They are the same.

Q. Is this variation easy or difficult to pick up if there was a forger?

A. It is not easy for anyone to be able to take note of this variation. Relatively speaking, it would be easier for him or her to be able to pick up this movement shown in A, because, normally speaking, people tend to execute the stroke in a continuous movement like this. It is quite rare to find such connected stroke movement as shown in B, C and D. Without comparing them in minute detail, it is not easy for anyone to spot the difference of C with B and D. So it is difficult to forge; especially to imitate all these three variations -- it is almost impossible."

1250. Even Tsui agreed with the variations. What is however even more telling and against the constant refrain about copying from Sample 668 and Sample 669, is that Variation 1 is not in Sample 668 or Sample 669 but is in Sample 5998 and 5999 ( both 1985 Samples) so there can be no question of simulation based on the model from the wrong period. A Matching Variation such as D8 (Chart 47) is a powerful evidence of Genuineness, not just lack of sufficient proof of Forgery.

1251. Turning to the dissimilarities, in the radical "Kwong" (光) in Signature B, the first short vertical stroke was coming down from right to left thus the stroke as a whole was tilting towards the upper right. This the Judge held to be a significant dissimilarity at paragraph 10.57(3). Although Cheng made the point (others did not), he in cross-examination (T30:169:3-172:21) accepted that there were variations in the controls, sometimes the stroke slanted to the right and sometimes to the left. In a copied passage, however, which omitted reference to Cheng's change of stance, the Judge held that it was a point of significant difference. What the Judge failed to consider was firstly that there was natural variations amongst the 4 questioned Signatures on this vertical stroke and secondly that the same natural variations also appear in the control Samples again indicating Genuineness not only of Signature B (which was condemned as Forgery) but also of the other 3 questioned Signatures. This is yet one further example of the Judge turning Genuineness into Forgery.

1252. On the character of Huei, the same observation as formerly made in relation to the other two characters of Wang and Teh can be repeated, namely that there was no basis for the Judge to come to a view of significant dissimilarities suggesting Forgery. In fact the evidence when properly analysed and evaluated would have revealed a case of Genuineness not a case of Forgery.

Conclusion on Micro Analysis of the 4 Questioned Wang Signatures

1253. A fair conclusion after the above exercise should point to the case of significant dissimilarities not made out even by a relaxed standard of proof, let alone that which is required in this case. There are numerous examples of significant similarities if not clinching Matching Variations. Analysis shows the repeated errors in findings of significant dissimilarities. Even on a micro analysis therefore, the weak case of the Father on handwriting Forgery has failed. This being so, the Plaintiff's Action deserves to be dismissed with Judgment both on the Claim and on the Counterclaim to the Wife.

Tse Ping Yim Signatures - General Remarks

1254. It seems appropriate to first make a few general remarks on the Judgment relating to the Tse Ping Yim Signatures. The first and foremost consideration which should not be overlooked at all times during the course of the discussion about the dispute of the Tse Signatures is that Tse himself confirmed under statutory oath that the disputed Signatures were signed by him on these documents. It is therefore, a very very unusual case (even by the standards of forgery cases) of the experts trying to prove that a person either could not recognise his own signatures or that he deliberately lied about his signatures swearing that these disputed Signatures to be his, when they were not. It is with this fundamental consideration in mind that this handwriting evaluation has to be undertaking.

1255. Secondly, it seems to me that as the case found by the Judge is that the Disputed Tse Signatures were all written by the method of Forgery Tracing from one model, the analysis of whether the Judge was correct in his conclusion is somewhat different from the analysis in relation to the multi-models simulation Forgery of the Wang Signatures.

1256. The question of speed was discussed in the Judgment in relation to the changes in the Tse Signatures. At paragraph 12.19 it was said that in relation to writing of signatures that the variation in speed between a person's signatures at a given point of time is unlikely to be great and it was explained that signature was a spontaneous and natural act of the person in the production of his name. With respect, the view in the Judgment is difficult to accept. A person throughout his life signs his name thousands of times. But he does not always sign them with the same speed. Even in any given week, he may sign many documents at various speeds, ranging from slow to fast to very fast. If a stack of routine documents are put in front of a person and he is in a hurry, he will likely sign them very fast. In a hurry, he would probably sign faster and faster as he went through the pile and the signature would probably also deteriorate with very fast speed of signing. An important contract or a very large cheque will obviously call for a more careful signing. This is simply general common sense applicable to most people. Therefore I cannot accept that variation in speed of a person's signatures is not great. The condition under which a person signs a document very often dictates the speed and manner of his signing and with that the shape of the signatures also changes. Ten signatures of mine even signed on a single day can vary a great deal but of course another person's signature signed on a single day might not vary so much. Not only therefore does speed and shape of signature vary from time to time but also vary from person to person.

1257. At paragraph 12.20 it was said that Tse would not sign more slowly before his boss because he was a special employee and treated as a member of the family. That Tse was regarded very favourably by the Wang family is undoubtedly true but that does not mean he was not an employee doing largely domestic duties or that Wang was not a big boss. Furthermore Tse was not in the habit of witnessing his big boss signing documents and this was certainly a sufficiently important occasion when his involvement was required. Even senior employees of big bosses would generally want to please their boss. Tse on this occasion, would definitely want to ensure that he was doing the right thing. He would not want to get it wrong. It is perfectly logical for Tse to sign the witnessing documents carefully and therefore slowly.

1258. It was generally recognised that whereas Wang wrote much more cursively and with wider range of variations, Tse wrote neatly and more formally and with smaller range of variations. It is therefore against that background one has to assess the question of superimposability. The first question which should be asked is to what extent are the Sample signatures of Tse superimposable. Only if the answer to that question is in the negative namely that there are not superimposable, then do we need to ask the next question of the significance of superimposability of these 4 disputed Tse Signatures.

1259. In the Judgment, this was not fully explored. At paragraph 15.5 it was said "the degree of superimposability is certainly much more than other sets of signatures, such as the four small signatures in DPJ 5993(3) to (6).." At paragraph 15.8 there was the reference to Tse's supposed ability to write consistently but again there was nothing specifically said about his consistency being such that many of the Samples signatures are superimposable. A study of the evidence does suggest that Tse was a person who could write very superimposable signatures as can be seen from his Sample signatures. This question of superimposability of the other Samples was never seriously considered nor properly analysed. The basis of the probative value of superimposability was never properly laid. The assumption was unconsciously made in the Judgement that the other Samples signatures are not superimposable or do not exhibit such characteristics and that therefore when it can be shown that these 4 disputed Signatures are superimposable, it has a high probative value to show that all these 4 disputed Signatures were copied by tracing from one model. It loses its probative value when that assumption is removed because if one persons writes always in a manner which is very consistent and superimposable, then the fact that four Suspect Signatures signed as a set are highly superimposable does not prove that they are forged or copied by tracing. In fact that they are more likely to be Genuine as being from the same hand whose signatures are always superimposable. The reason for this is the forgers may not generally have that high skill of consistency and superimposability.

1260. The next matter which is of significance is that with the exception of one, all the Sample signatures of Tse were all written horizontally from the traditional left to right. The 4 disputed Signatures (between pencil markings) were all written vertically. This of course makes the case of forgery by tracing that much more unlikely as there would have to be considerable adjustment of spacing between characters and from horizontal spacing to vertical spacing. The exercise of alleged tracing in this case could be by only one of two methods, either by tracing from one model and make spacing adjustment with each of the disputed Signature, so that there would be 2 vertical space adjustments (between Tse and Ping and then between Ping and Yim) to write the Tse Signature A, and then another two vertical space adjustments to write Signature B, and then another 2 vertical space adjustments to write Signature C and then another 2 space adjustments to write Signature D. This process would therefore involve eight space adjustments. The second possible way for this Forgery by Tracing to be carried out is to photocopy each of the characters from a model, then cut out each copied character and lay all three of them out in vertical space and make a photocopy of the vertical newly layout words and write by tracing 4 times (for A, B, C and D) based on this photocopied vertical lay-out words.

1261. The Judgment did not make it clear exactly which method the experts or the Judge believed was adopted for the Forgery by Tracing. But a careful reading would suggest that the Judge was focusing on the first method namely the 8 vertical space adjustment method. I say this because the Judge said for example at paragraph 12.2 that "spacing between the characters would unlikely be consistent and superimposable." At paragraph 15.12 he said "the spacing between the characters of the various questioned signatures would not be superimposable or otherwise the same for the four signatures." What was said at paragraph 15.14 also suggests that the mind of the Judge was focused on tracing with 8 vertical adjustments.

1262. What the Judge did not address his mind to is that a tracing from a horizontal model and writing out in a vertical signature would require not just 8 vertical adjustments of spacing between characters but also 8 adjustments of vertical alignments, which operation would be subtle, difficult but important. Slides F6 to F10 illustrate the problem. When tracing from a horizontal model and turning it into a vertical writing, what is involved is not merely a character to character vertical spacing adjustment but also a vertical alignment adjustment because the original model is a horizontal model and the character written on a horizontal model of course has its horizontal alignment but not a vertical alignment. What is required for every disputed Signature is therefore 4 space adjustments, two space adjustments for vertical spacing and two space adjustment for vertical alignment adjustment. If the Judge had thought it through, he would appreciate that the speculated Forgery by tracing from horizontal model is a scenario of a forger gone bad. It is an impossible task involving 16 space adjustments (8 vertical spacing adjustments and 8 vertical alignment adjustments).

1263. There is yet one more aspect which the Judgment did not seem to address, namely to identify which model was used for the tracing. There was no suggestion in the Judgment that out of the 38 Samples, a particular one was chosen as the model for tracing. Nor was it demonstrated at any time that the 4 disputed Signatures were highly traceable or imposable on one identified Sample. This being the case, the speculation of the Judge must be based on tracing of a model which was deliberately withheld from the Court, creating yet one further uncertainty for the case on Forgery.

1264. The superimposability of the 4 suspected Signatures was attacked by Mr. Denis Chang. He made a good point about the unworkability of this superimposability theory by referring to the character Ping in Signature B being much larger than the Ping in Signature A and Signature C. He also referred to character Yim in Signature D being larger compared to that in Signature A.

1265. But the really unanswerable point Mr. Denis Chang made on this superimposability is that it was an illegitimate superimposability exercise for the Judge to chop off the ends (which he admitted are not superimposable, see paragraphs 15.6 and 15.7 in version as corrected or not corrected). I must say I agree with Mr. Chang. The whole point about tracing is that the forger did not have the skill to write out the signature by simulation and that therefore it had to be done by tracing to enable all parts of the character fully traced out. The ends of Chinese handwriting such as ticks, slants and sabres are particularly important and sometimes are of identifiable unique nature and can give rise to proof of matching variations. It can hardly be called tracing if one has the ends not properly traced. Imagine having the president's double who has different and larger head or feet. This would be ridiculous. Instead of accepting that the non-imposability of ends demonstrates the non viability of the Forgery by tracing and to make no finding of Forgery, the Judgment chose to find wholly unconvincing excuses for the non-imposability of ends (such as poor tracing) and persist with the case of Forgery by Tracing. I regret to say that for me the theory of Forgery by Tracing simply does not work and as I see it, the whole case of the Father on Tse Forgery was unworkable and untenable from the very beginning.

(4) Tse - Similarities and Dissmilarities

1266. The Judgment is noticeably silent on the similarities in the character of Tse in the disputed Signatures with the Samples. Presumably all similarities are dismissed and discounted and not even considered on the basis that there was forgery by tracing and it was therefore pointless to consider the similarities. If I may say so, the methodology is again all wrong because it is going back to front, namely to reach conclusion of tracing and superimposability and therefore to discount similarities rather than carry out properly the exercise of observing all similarities and all differences and then do the full and careful consolidated evaluation and if necessary and if so warranted by compelling evidence, draw the conclusion of Forgery by Tracing. The way the Judgment was structured and this clearly comes out again and again by a detailed analysis of the Judgment it seems that the Judge had already decided on Forgery by Tracing and was really looking for reasons to back up the decision.

1267. What is contended by the Wife on appeal is that leaving aside superficial similarities which can be dismissed on the ground of tracing, what about the Matching Variations. Matching Variations arise because we have not one suspected documents but four suspected documents in a set and because intra these four suspected documents there are a lot of natural variations, may be more with the Wang signatures but still considerable even for the Tse Signatures. If the four Tse Signatures which are agreed to be from the same hand, contain amongst themselves variations which match the variations amongst the Tse Samples then it is strong evidence of Genuineness and even more proof against Forgery by Tracing.

1268. Slide F14 below demonstrates the first of the Matching Variations of Tse. The variation of pen pressure in the execution of the horizontal strokes of the radical "yin" (言) of the word "Tse" (謝) reflected in the Samples (with variable delicate retracing of the 3 horizontal strokes in the words of Mr. Denis Chang). The retracings at the end of the horizontal stroke, as can be seen in F14, vary from retracings in all 3 horizontal strokes in Signature A, to retracing in one horizontal stroke in Signature B, to retracing in 2 horizontal strokes in Signature C, to no retracing of horizontal stroke in Signature D. Tsui agreed with the presence of the retracing feature at T70:134:25-136:21. These 4 variations of horizontal stroke retracings in the 4 Suspect Signatures are reflected in the same variations in the Samples. The retracing at the end of strokes are delicate movements, difficult to observe and very difficult to imitate. This is a perfect application of the second principle of Osborn set out in Osborn Quotation 3. This Matching Variation is therefore powerful evidence of Genuineness but the Judge simply ignored all Matching Variations or evidence which will disprove Forgery by Tracing.

1269. Slide F16 provides another good example of Matching Variations of the character Tse. Here the variations lies in the slight bringing up of the pen at the end of the vertical stroke of the "hau" (口) (as to be found in Signatures A, B and C) and the curved downward of that vertical stroke without brining the pen up (as in Signature D). The same variations appear in the Samples. Jia said this at T82:1:22-24; T80:33:25-34:4 There is no doubt that these are Matching Variations. Again the same point as before can be made that these are delicate movements, subtle not easy to notice and difficult to imitate. Another perfect application of the just stated Osborn principle pointing to Genuineness which again the Judge chose to ignore.

1270. In the impressive List 2 of Mr. Jat, there are no less than 14 items of Matching Variations relating only to the character Tse. It is unnecessary to analyse further the Matching Variations relating to Tse except to say that they are so cogent and compelling and that they cannot be ignored. The cogency and compellability here is of course the other way, namely towards Genuineness, not towards Forgery.

1271. In relation to significant difference on Tse, the Judge held in a Judge-Observed Significant Difference at paragraph 15.65-15.69 that there was retouching on the long vertical line stroke before the tick in the radical "san" (身). The reason for his conclusion was, notwithstanding no expert of the Father remarked anything adverse about this feature, that unlike the striation marks in the radical "chuen" (寸) there was no striation marks coming up from the end of that vertical stroke in (身) along the direction of the tick, as in the case of the corresponding stroke in the radical "chuen" (寸). The Judge observed that there were clear striation marks all the way down to the end of the vertical stroke. It is difficult to understand how with the presence of the whilte unlinked striation lines that there could be any over-writing on that stroke. It was simply impossible to have both the original stroke and the over written stroke having exactly the same striation lines, one superimposing on the other. The Judge seemed to have misunderstood and distorted the evidence of Jia and again in playing expert came to a totally unjustified conclusion. The case of the Wife on this point is set out fully in paragraphs 230 to 239 of Part B Submissions and in my view the criticism of the Judgment on this part is fair.

1272. A study of either the Master Chart of all the Tse signatures or the Chart relating only to the Tse character simply do not show that the Tse character as written in the disputed Signatures bear the sign of Forgery as found by the Judge. In fact the clearest and very powerful evidence of Matching Variations suggest compellingly that the 4 disputed Tse Signatures are Genuine.

(5) Ping - Similarities and Dissmilarities

1273. Matching Variations is again the dominant feature in showing significant similarities in the writing of the character Ping which can be seen in all 4 suspected Signatures. Slide F20 shows the variation in the retracing with a return stroke of the horizontal stroke of the right radical "丙"in Signature A and Signature B which feature is absent in Signature C and Signature D. In the Samples, the same variations can be found as shown in Slide F20 below. Again this is a subtle feature which is difficult to imitate and is a classic application of Osborn's second principle.

1274. Another equally persuasive Matching Variations for Ping can be seen in Slide F22 below, where in respect of the 2 strokes constituting the radical "人" in "丙", the contrast was between the smooth left slanting stroke of "人" as can be seen in Signature A, B and C and the angular turn to the left in Signature D which occurred at a place coming to the middle part of character "人". The Samples in 674(5), 674(8) and 700 show quite dramatically the angular turn to the left which are absent in the other Samples illustrated in Slide F22. We have another application of Osborn's second principle pointing to Genuineness.

1275. In relation to significant differences of the character Ping in one or other of the disputed Signatures, the comprehensive Table C (item (2), (3), (5), (10)) and Table D (item (2)) of the Wife's Submission fully deal with the Judge's errors and I will refrain from repeating what are said there.

1276. In my judgment, whatever minor defect which the Judge might find against one or other of the 4 Signatures in respect of any particular feature, upon proper evaluation of the character Ping it will be seen that the minor differences even if so observed are far eclipsed by the overwhelming case of the various and convincing Matching Variations which point strongly in favour of Genuineness. In relation to the character Ping, a case of Forgery by tracing from a post 1994 model cannot possibly be sustained.

(6) Yim - Similarities and Dissmilarities

1277. With the character Yim, Matching Variations again give us a picture of Genuineness. Slide F23 below shows how the final stroke in the lower radical "火" varied retracing from below the stroke for Signature B and D, to retracing above the stroke for Signature C and with no retracing for Signature A. The same variation can be found in Sample 736(4), 5995(1) comparable to Signature B and D. Sample 736(3) and 5996(1) is comparable to Signature C. Sample 700 and Sample 8 are comparable to Signature A. The subtle features which constitute these variations again serve to remind us the value of Osborn's second principle in the identification of Genuineness.

1278. At paragraph 15.62 the Judge allowed speculation to take flight and said that in relation to Signature A, the slants in both radicals of "火" being not smooth, gave rise to the inference that in tracing to the point of the both dots the writer slowed down and unconsciously allowed the pen to have some minor change of direction after some pausing and hesitation. This is not a feature that any of the Father's expert said anything adverse. The assumption at paragraph 15.62, of such feature of change of direction, is incorrect and reliance was placed by Mr. Chang on other Samples to show that the Judge was incorrect. There was simply no evidence of the ability of Tse to write in that way whether in 1990 or otherwise. There was in my view, nothing in that feature which could compel the Judge to draw the inference. On a matter where even the expertise of the handwriting specialists are less than wholly reliable at the best of times and certainly when they were at great variance in this case, a Judge who chose to play expert and draw inferences from features without cogent and compelling evidence or reason, could not pass the high hurdle of Re H.

Conclusion on Micro Analysis of the 4 Questioned Tse Signatures

1279. My conclusion on a Micro level analysis is that there was neither sufficient evidence nor sufficient convincing and compelling reasons shown in the Judgment to justify a conclusion of Forgery. In fact in my view, the Matching Variations examples again and again demonstrate powerfully why the 4 Tse Signatures, even though they look very much alike, contain sufficient Matching Variations amongst them as to give rise to the most persuasive and probative proof that these Tse Signatures are Genuine and not Forgeries.

(D) RETRIAL EVEN IF NO JUDGMENT FOR WIFE

1280. For reasons given earlier, I am of the clear and firm view that the Judgment below ought to be set aside and that Judgment should be entered for the Wife against the Father both on the Claim and on the Counterclaim. In other words this Court should pronounce in favour of the 1990 Will and not in favour of the 1968 Will.

1281. But in the event that this Court of Appeal is of the opinion that Judgment in favour of the Wife on the Claim and the Counterclaim should not be given for any reason, then in my view, the least that this Court of Appeal should do and it is a duty which it is bound to do, is to order a Retrial for reasons which I will state under the following headings:-

(I) Plagiarism

(II) Did not properly consider the points advanced by the Wife

(III) Unpleaded case allowed to be the Basis of Judgment

PLAGIARISM

1282. The complaint of plagiarism is serious but it requires a close examination because the Wife says that no independent mind had ever been judicially applied to the proper resolution of the dispute and that the Judgment was not the finished product of an independent judicial mind.

1283. It is accepted that all parties to litigation in Hong Kong are entitled to a fair trial before an independent court which in turn entails an independent judge bringing to the case at all times an independent mind so that the cause before the judge will be fairly and without bias be judicially adjudicated. It is also part of the tradition in Hong Kong that lawyers for each party will strive to advance the best case of their client irrespective of whether that cause of the client has merit or not. It is for the judge sitting impartially and independently to apply his mind to determine whose cause has the merit in law. A fair trial therefore requires a judge bringing his independent mind as a judicial decision-maker to the adjudication of the case before him and not to adopt too easily and/or uncritically the advocacy of the lawyer for one party.

1284. The case of Miranda v Bennett (2003) 322 F. 3d 171 contains a good passage of the danger of a court adopting too much of what is advanced by lawyer for one party. At page 177 this is said:-

"... we nonetheless heed the cautionary note repeatedly sounded by the Supreme Court as to the imprudence of wholesale adoption of a party's position even in making findings of fact after the court's decision has been announced:

"District judges should avoid as far as they possibly can simply signing what some lawyer puts under their noses. These lawyers, and properly so, in their zeal and advocacy and their enthusiasm are going to state the case for their side in these findings as strongly as they possibly can. When these findings get to the courts of appeals they won't be worth the paper they are written on as far as assisting the court of appeals in determining why the judge decided the case." United States v El Paso Natural Gas Co. 376 U.S. 651, 657 n. 4 (1964)."

1285. The reason for the objection to the wholesale adoption by a court of what one side's lawyer said was so obvious that the reason for its unacceptability is seldom properly articulated. A good articulation came recently from as one would expect an Australian court. In Huluba v Ministry for Immigration [1995] 59 FCR 518, a case involving the refusal of refugee status which was attacked on the basis that no independent mind to the decision making process was given by the second delegate of the Minster who was supposed to review the reasons of the first delegate of the Minister for refusing the application. The attack was successful.

1286. At page 529D the following was said by Beazley J:-

"Procedural fairness requires a decision-maker to apply an independent mind to the application subject of administrative action. A decision-maker is entitled to have regard to research and investigations carried out by others as well as to assessments and reports and recommendations prepared by others in the course of the administrative process. A decision-maker may have regard to and adopt, if thought appropriate, the reasoning of some other person involved in the administrative process. Thus a decision-maker could accept the reasoning of an officer whose function it had been to provide a recommendation and could adopt verbatim, such report or recommendation, provided at all times that the decision was the independent decision of the decision -maker. This case is different. The second decision-maker's task was to make a new determination. In doing so there would have been no breach of the rules of procedural fairness for the second decision-maker to read and consider the findings of the first decision-maker. However, procedural fairness required that she reach an independent decision in the matter.

It is obvious from the passages set out above that the second decision-maker used substantial portions of the report of the first decision-maker. The coincidence of the language makes any other conclusion improbable. Those passages contain critical findings. The question arises, therefore, whether this coincidence of language demonstrates a failure by the second decision-maker to bring an independent mind to the determination of the application."

Then at page 530G he said:-

"In the present case, I consider that the use of the same language, sometimes in florid terms, on critical aspects of the decision-making process, makes it more probable than not that the second decision-maker did not apply an independent mind to the decision-making process. It follows that the applicant was denied procedural fairness by the second decision-maker."

1287. The question is therefore whether the substantial adoption of the Written Submission of the Father by the Judge rendered the Judgment unacceptable as being a denial of a fair trial to the Wife. It was said by both Counsel for the Father as well as by my Lord Yeung that the attack on the Judge was most unfair and as also said by Yuen JA that the Judge was following the proceedings and was on top of the material.

1288. The attack is not on the Judge personally but on the judicial process as finally concluded by his Judgment. The complaint is that there was no fair judicial process because the Judge failed to bring an independent mind to that important judicial decision-making process and the way to prove that lack of independent judicial decision-making process is to examine what is called during the Appeal, the pink passages in the Judgment.

1289. There is no doubt from a study of the 2 Volumes produced by the Wife on Appeal (marking the copied passages with pink colour) that this Judgment, handed down less than 6 weeks after the final oral submission (lasting one day), consisted of pages and pages of pink colour. An estimate of not less than 50% pink passages copied from the Written Submissions would not be unfair.

1290. A chronology of the Written Submissions of the parties is revealing in what happened. The Written Submissions were divided into 3 parts, namely firstly the Ink Dating Submissions, secondly the Handwriting Submissions and thirdly Submissions on Factual aspects including forgery by Suspicious Circumstances and the law. In August 2002 the main Submissions on ink dating had been given to the Judge and he must have, around that time in late August/early September or shortly afterwards, written his Section III Judgment on Ink Dating. In September 2002 the main Submissions on handwriting were given to the Judge and he must have written his Section II Judgment (Section IIA and Section IIB) on Handwriting in late September/early October 2002 copying heavily Vol. 1 of the Father's Written Submissions on Handwriting dated 20th September 2002. In late September and early October 2002, the Factual Submissions were given to the Judge and he must have written in late October and November 2002 his Section I Judgment on the Factual and Suspicious Circumstances, again copying heavily on the Volume 1 of Father Written Submission on Factual Issues dated 4th October 2002. The sequence of the writing of the Judgment was Section III, followed by Section II (A and B) and followed by Section 1 as the Judge was at pains to point out in paragraph 6.4 of the Judgment.

1291. Because of the serious complaint against the Judge of his extensive and indiscriminating copying of the written Submissions of the Father on Handwriting and on Suspicious Circumstances, a chronology of the way the Written Submissions were given to the Judge is hereby set out in Table form. It is very likely that this chronology of the handing in of the Written Submissions to the Judge, had contributed to the sequence and the timing of the writing of the Judgement, first on Section III, followed by Section II, followed by Section I.

Chronology of Written Submissions on Handwriting

Chronology of Written Submissions on Forgery by Suspicious
Circumstances and Law

1292. What is of course critical to the decision-making in this long judicial process is the analysis of the main issue between the parties, namely the allegation of Forgery of the signatures with vast evidence given by the handwriting experts. Volume II of the Judgment contains that critical judicial decision-making by the Judge. It is therefore vitally important, as well as being absolutely fair to the Judge, to examine carefully the vital Chapters in Volume II as to the ways that hard fought handwriting dispute was resolved by the independent decision-maker, the Judge. The two critical Chapters are Chapter 9 and Chapter 10.

1293. A comparison paragraph by paragraph of the Chapter 9 with Volume 1 of Mr. Edward Chan's Written Submission shows the following:-

1294. A comparison paragraph by paragraph of the Chapter 10 with Volume 1 of Mr. Edward Chan's Written Submission shows the following:-

1295. It can therefore be seen that the copying was not casual, not minimal or even reasonable (in terms of adoption of argument of counsel) but was so substantial that in respect of the key analysis and evaluation section of the Judgment on handwriting, the copying was something like 95% or more. Every one of those relevant pages was a sea of pink colour which suggests, I hope not unfairly, that the Judge has abdicated his independent decision-making or adjudication functions which involved the giving of his own reasons why he accepted or rejected the main arguments of the parties.

1296. What I find painfully puzzling is why did the Judge choose to copy only Volume 1 if he was going to copy the Father's Written Submissions. Mr. Martin Lee's Opening Submission on Handwriting consisted of 256 pages with 220 paragraphs. Mr. Edward Chan's Volume 1 on Handwriting consisted of 290 pages with 377 paragraphs. Mr. Edward Chan's Volume 2 on Handwriting consisted of 169 pages with 138 paragraphs. Mr. Lee's Reply Submission on Handwriting consisted of 78 pages with 91 paragraphs. Mr. Edward Chan's Reply Volume 3 Submission (handed in the day after the Oral Submission) consisted of 36 pages with 51 paragraphs. And yet out of these 5 volumes (2 from Wife and 3 from Father), the Judge chose to copy so extensively from Volume 1 of Mr. Chan. I do not understand but the point is squarely in front of us that even in copying, what the Judge did was puzzling.

1297. I hope I am not being unfair in saying that the repeated use of the word Plagiarism during the hearing of the Appeal was not too harsh or unfair in the circumstances. If two examination papers were handed in by two pupils with this kind of identity of not just contents but language and even layout, the two pupils will likely find themselves, as a minimum, being asked to take their examination again. I do not believe that our Court in a case such as this can do anything less. A Retrial is unfair in my judgment to the Wife but that is the least she deserves if no Final Judgment is given in her favour.

POINTS OF WIFE NOT PROPERLY ADDRSSED BY THE JUDGE

1298. The complaint of the Wife however is not confined to mere Plagiarism because arising from Plagiarism is the failure of some of the crucial judicial obligations in a fair trial by an independent court. In English v Emery [2002] 1 W.L.R. 2409, the English Court of Appeal explained the extent of the judicial duties under the European or Strasbourg jurisprudence based on the European Human Rights Convention at page 2416:-

"The Strasbourg court, when considering article 6, is not concerned with the merits of the decision of the domestic court that is under attack. It is concerned to see that the procedure has been fair. It requires that a judgment contains reasons that are sufficient to demonstrate that the essential issues that have been raised by the parties have been addressed by the domestic court and how those issues have been resolved."

1299. Reference was made in the English case to the decision of Hiro v Spain where the European Court of Human Rights held that there was a violation of the right to fair trial under Article 6(1) of the Convention. In Hiro v Spain [1994] 10 E.H.R.R. 567 the Court said at 574 and 575 the following:-

"The Court reiterates that Article 6(1) obliges the Court to give reasons for their judgments, but cannot be understood as requiring a detailed answer to every argument. The extent to which this duty to give reasons applies may vary according to the nature of the decision. .....

It therefore required a specific and express reply. In the absence of such a reply, it is impossible to ascertain whether the Supreme Court simply neglected to deal with the submission based on the prior right to the mark "Creacions Orient" or whether it intended to dismiss it and, if that were its intention, what its reasons were for so doing."

1300. Mr. Thomas submits that there are four requirements under Hong Kong Law for a fair trial by an independent court, namely (1) a competent and impartial court (2) the court reaches its independent decision (3) that it gives reasons why it reaches its independent decision and (4) the reasons given for reaching its independent decision deals with the main point addressed to the court. These four requirements are not challenged by Mr. Tang. What is possibly at issue is to what extent they have been infringed.

1301. Plagiarism which I have addressed earlier is the infringement of the second requirement of independent decision making. The Wife complains that there was also the infringement of the fourth requirement of the reasons given must deal with the main points addressed to the court. What is submitted on behalf of the Wife is that by copying in the way the Judge did, the Judge failed to deal with the main points advanced by the Wife. Take just for example the Written Submissions of the Wife on the Handwriting. As said earlier the Wife put in the Opening Submission dated 6th September which was responded to by the Father's Volume 2 dated 20th September and replied to by the Wife's Written Submission dated 11th October. There were therefore at least 3 documents dealing directly with the points advanced by the Wife, namely the Opening Submission of the Wife, the Volume 2 of the Father and the Reply Submission of the Wife. Unfortunately, the Judge copied from Volume 1 (which because it was in anticipation could not possibly and did not in fact deal with all the Wife's points) whereas he should properly in his Judgment deals with all the major points advanced by the Wife covered by the aforesaid three documents. The path followed by the Judge inevitably resulted in his Judgment not dealing with many of the important points advanced by the Wife. Important subjects such as the Calligraphic pen, Matching Variations, Impossibility of Multi Model simulation (Mr. Lee's "single most important point") did not even get a proper mention let alone serious consideration. In these circumstances, no Court in Hong Kong can allow this Judgment to stand and it seems to me that the minimum that must be done is to order a Retrial if the Court is of the view that Judgment cannot be given to the Wife on the Claim and Counterclaim.

UNPLEADED CASE - BASIS OF JUDGMENT

1302. The final basis in my view that this Court will be compelled to order a Retrial (if not giving Final Judgment to the Wife) is on the basis that this case had gone wholly wrong by reason of inter alia the Judge allowing an unpleaded case of Forgery to form the basis of his Judgment. The importance of pleadings is sometimes devalued in this modern age. But its importance must not be overlooked, specially when it is critical in a heavy case such as this when so much sought to be made by the Father's legal team with so little. The salutary wisdom of Bokhary JA (as then was) in ADS v Wheelock [1994] 2 HKC 264 at 270 is worth re-reading:-

"..... and especially where fraud is alleged

To the generality of the foregoing must be added the special rule that allegations of fraud must be pleaded distinctly and with the utmost particularity. The word 'distinctly' is the one used by Thesiger LJ in his well-known statement in Davey v Garrett (1878) 7 Ch D 473 at 480 as to how fraud is to be pleaded. And the expression 'utmost particularity' as one sees from note 18/8/8 at p. 297 of Volume 1 of the 1993 Supreme Court Practice, is the one chosen by its learned editors and has the authority of their combined experience. That special rule arises in this case.

Never before as much a now

None of the basic rules of pleading have anything to do with technicality. All of them have everything to do with practical justice. They have always been of importance. Never before as much as now. For these days, there are more and more cases so vast and so complex that they push practically to its limit our system's capacity effectively to cope with them. Any laxity in their proper management, whether in regard to pleadings or anything else, can all too easily result in such cases spinning into confusion if not chaos-even before trial but especially at trial. The present case, for the trial of which half a year has been reserved, is such a case.''

It seems that these prophetic words were spoken specially for this disaster of a case, where after more than one year of Trial, the Judgment given did not reflect the fair process which was the entitlement of every party who comes before our Hong Kong Court. If this Court is unable to give Judgment against the Father and in favour of the Wife on the Claim and Counterclaim (which will be most regrettable) then there must be at the least an order of Retrial.

(E) CONCLUSIONS

1303. The conclusions which I have reached can be summarised below:-

ON EXECUTION

Execution of the 1990 Documents is proved with Judgment on the Counterclaim to the Wife because:-

(1) The finding of Forgery by Suspicious Circumstances is to be set aside;

(2) The Dispel SC Principle is not applicable to Execution;

(3) The Presumption of Omnia applies in these circumstances to Execution which in any event has not been rebutted;

(4) Execution is proved even if Presumption of Omnia does not apply;

(5) No adverse inference is to drawn against the Wife by her failure to give evidence at the Trial.

ON FORGERY

Forgery claim of the Father is dismissed because the Judgment on Handwriting was seriously erroneous in that:-

(1) there was no Methodology, no Proper Methodology and a Fatal Methodology;

(2) there were wrong approaches, wrong application of Methodology and Misconceptions;

(3) there were serious errors of Micro Analysis.

ON RETRIAL

If no Final Judgment is to be entered for the Wife against the Father on Execution and on Forgery, then there should be at the least an Order for Retrial because:-

(1) There was no independent judicial decision-making as evidenced by Plagiarism;

(2) The Wife's points were not addressed by the Judge in his Judgment

(3) The Judgment was wrongly based on an unpleaded case of fraud.

Hon Yeung JA:

Conclusion

1304. By a majority decision, the main appeal is dismissed. The parties have liberty to apply on the costs appeal and on the costs of this appeal.