List of Abbreviations

Chapter 18 - Ink-dating - The Analysis and Method

The Analysis

The Method

The Results

The Defence

Chapter 19 - The True Proficiency or Blind Test

Chapter 20 - The Law

Chapter 21 - The Underlying Scientific Theory of the Dye-ratio Method

The First Hypothesis (Extractability)

The Second Hypothesis (Decomposition)

Conclusion

Chapter 22 - Scientific Validity

Validation Studies

Peer-review

Chapter 23 - Acceptance By the Scientific Community and the Court

Acceptance of the Techniques by Courts

Conclusion

Chapter 24 - Arbitrary Separation of Fused/Overlapped Chromatographic Peaks

Higher Resolution should be used

Conclusion

Chapter 25 - Illogical Results of the Dye-ratio Method

Criticism 1

Criticism 2

Criticism 3

EJS's Defences

Conclusion

Chapter 26 - Mass Independence

What EJS could have done

EJS's Claim

Conclusion

Chapter 27 - One Standard Deviation

2 STDs (95%) or 3 STDs (99.7%) should be used

The Student's T-test

Conclusion

Chapter 28 - Procedure

VAM Procedures

Protocol

EJS's Laboratory Technique as shown on the Video

(1) Hygiene and Contamination

(2) Time-keeping

Ink Thickness

Conclusion

Chapter 29 - EJS's Qualification and Integrity

Qualifications

The Defence Experts

EJS's Integrity

Chapter 30 - Conclusion on Ink-dating

Section IV - Final Conclusion

Chapter 31 - Final Conclusion

Background

Suspicious Circumstances

Handwriting

Ink-dating

The 1990 Documents

Orders

Chapter 32 - Epilogue

Chapter 33 - Costs

The Application

Costs on Indemnity Basis

Ink-dating

The Direction on the Cost on LiveNote and the Written Submissions

The Plaintiff's Costs to be Paid Out from the Estate of the Deceased

Appendices

Appendix 1 - Table on the Progress of Trial

Appendix 2 - The Four Questioned Documents and their Translations

Appendix 3 - Master Chart of Wang Teh Huei's signatures (E6-654(37B))

Appendix 4 - Master Charts of Tse Ping Yim's signatures (E6-654(18-20))

Appendix 5 - Judgment on the Application by the Plaintiff for Leave to Swear Death dated 22 September 1999 in HCEA 79/1999

Appendix 6 - Judgment on the Application by the Defendant for Stay of Proceedings dated 20 August 2001 in HCAP 8/1999

 
 

"It cannot be that axiom established on argumentation can suffice for the discovery of more new works, since the subtlety of nature is greater much more than the subtlety of the argument."

- Francis Bacon (1561-1626)
Lord Chancellor of England (1618-21)

List of abbreviations
 

ADR	Alternative Disputes Resolution
ASTM	American Society for Testing and Materials
Mr Brunelle	Mr Richard L. Brunelle
Daubert	Daubert v Merrel Dow 509 US 579 (1993)
EJS	Mr Erich J. Speckin
EJS 1	First Report of Mr Erich J. Speckin dated 15 November 2000
EJS 2	Second Report of Mr Erich J. Speckin dated 19 February 2001
EJS 3	Third Report of Mr Erich J. Speckin dated 3 August, 2001
Frye	US v. Frye 293 F 1013 (1923)
JSM	Messrs Johnson Stokes & Master
MAFS	Mid-Western Association of Forensic Scientists
Mr Lee	Mr Huy Lee
PVT	Mr Peter Tytell
R-ratio	The rate of extraction
SOFIA	Society of Forensic Ink Analysts
STD	Standard deviation
TLC	Thin Layer Chromatography
VAM	Valid Analytical Measurement Programme
VNA	Dr Valery N. Aginsky

 

18.1 The plaintiff has instructed one Mr Erich J. Speckin ("EJS") from Michigan of the United States, who professed that he could identify whether a writing in ink was still in the drying process or was completely dry.

THE ANALYSIS

18.2 On 1 August 2001 a quantity of small microplugs was removed by EJS from designated characters in Documents A, B, C and D where there was ink-writing and they were separated into vials containing six plugs each.

18.3 Upon analysis, EJS found (which is not disputed by the defence) that Documents A and B were written by the same type of ink viz. Formulab ink, and Document C was written by ink of a different formulation viz. Papermate. The ink used for Documents A and B were manufactured in 1982. Both the Formulab ink and the Papermate ink used were commercially available in 1990.¹ The Papermate ink is still available commercially today.²

18.4 I accept the plaintiff's submission that, although there is no direct evidence available on this point, if it is established that both Documents A and C were written sometime in or about 1996, then it is extremely unlikely that Document B would have been written in 1990 or before. In fact for the same reason it is probable that Documents A and B were written at or about the same time. This is because :
 

(a) On the defendant's case especially the suggestion that it was Tse Ping Yim who witnessed the deceased's signatures on all four documents at the same time in March 1990, the four documents should have come into existence at more or less the same time; and

(b) It is most likely that Documents A and B were written by the same pen at about the same time since :
 

(i) the ink on both documents was of the same made and the same year; and

(ii) the writing on both Documents A and B showed the same phenomenon that the flow of the ink in the pen was not smooth. This could be seen from the much lighter colour in the characters after the pen had been used for writing for a short while. Thus in Document A the characters "產, 全, 部, 遺, 贈, 我, 妻, 子, 龔" showed evidence of bad ink flow resulting in re-writing of certain strokes in the characters "產, 部, 贈, 妻, 龔". In Document B, the same phenomenon is found in the characters "愛, 妻, 我, 的, 父, 母, 雖, 然, 令, 我, 失, 望".

18.5 EJS dissolved the ink from the plugs taken from each of the documents by using a strong solvent pyridine and the dissolved ink in the solvent from each of the documents were spotted on a TLC plate (TLC plate No. I). It was found that the ink on the plugs from Document D could not be extracted by the solvent and the conclusion of EJS was that the ink from Document D was probably either a gel ink or India ink.³ The TLC plate No. I was then developed in a TLC chamber and the bands of each of the spots of ink from Documents A, B and C were separated by chromatography. From the ink library kept by him, EJS was able to come up with possible matches.⁴

18.6 EJS further tried to identify the ink to find out if there was any date tag in the ink. He dissolved the ink on the plugs taken from the first three Documents in another solvent. The dissolved ink was then spotted onto a TLC plate (No. III) and chromatography was done to separate their components. The control for the four date tags was also run on the plate. The result showed that the ink from Documents A and B was separated in the same way.⁵ The result also showed that the ink in Documents A and B contained a date tag which identified the ink as being produced by Formulab in 1982.⁶

18.7 In order to identify the ink on Document C, another chromatography was done on TLC plate No. IV in relation to the ink and four known samples of ink from the ink library. The result showed that the ink was manufactured by Papermate which was one of the more common ink formulations which was available both in 1990 and also today.⁷ In fact this ink formulation has been in existence since about the 1950s.⁸

THE METHOD

18.8 In the procedure of finding out whether or not the ink was dry, EJS used the accelerated aging method. Acceleration was achieved by heating the vials in a laboratory oven at 100^oC for approximately 30 minutes, uncapped, and was then allowed to cool down. Two of the vials were aged by accelerated aging method of heating whilst the other two were not.

18.9 The cooled vials would then be placed in a rack next to the corresponding unheated samples for the relative ink-age comparison tests. It was contended that the artificially aged sample would represent an ink that was approximately three years old. If a significant statistical difference was found between the aged and the unaged samples, it would be concluded that the ink was still in a drying process and would most likely be less than 3 years old.

18.10 Three different methods were used to measure the relative dry levels of the heated and unheated samples : i.e. R-ratios, Percent Extraction, and Dye-ratio.
 

(a) The rate of extraction (i.e. R-ratios) would measure the rate at which the dyes in the ink were extracted in a weak solvent by taking samples at different time intervals.

(b) The percent of extraction would measure the ratio of the amount of ink that could be extracted in a weak solvent and the same would be compared to the total amount of ink on the paper which was extracted by using a strong solvent.

(c) The Dye-ratio would measure the relative concentration of each dye to one another.

18.11 EJS contended that if at least one of these tests showed a significant statistical difference between the heated and the unheated samples, it would be concluded that the ink was still in the drying process, i.e. less than 3 to 3 1/2 years old.

THE RESULTS

18.12 The data from the R-ratios, Percent Extraction and Dye-ratio were analysed statistically to determine if the differences were statistically significant or not. The standard threshold used was one standard deviation ("one STD"). EJS contended in his First Report ("EJS 1") dated 15 November 2000 that⁹ :
 

"the data obtained for the rate of extraction, percent extraction and dye-ratio meets this threshold of analysis 'significant statistical difference', meaning in this case that the samples tested from documents A and C were not dry. Based on these differences it is my opinion, the question documents were not written on or near the purported dates but at a later time at least in 1996 or later."

18.13 EJS submitted a Second Report ("EJS 2") dated 19 February 2001 and a further Third Report ("EJS 3") dated 3 August 2001. I shall consider them hereinbelow when the validity of EJS's method and results are considered.

18.14 EJS contended that his conclusion was drawn upon results he obtained from the Dye-ratio method, which he claimed to be a well-established method having been in use worldwide (including law enforcement agencies in the United States) for many years, and accepted universally by the scientific community at large and numerous courts at various levels in the United States and in other countries.¹⁰ He said that the research has already been done to show the reliability of the relative ink-dating procedures using the Dye-ratio method of analysis,¹¹ and that the data collection in this case was performed on accepted standards in the field of forensic ink analysis and in a manner that has been tested and proven reliable on many occasions in the past.¹²

18.15 He claimed that there are many publications that show the method utilized by him in this case is valid, reproducible, and reliable.¹³

18.16 He contended that he has personal knowledge of at least two cases in which the United States Secret Service has performed testing in the same manner as he performed in this case and it has used this technique between 1995 and today.

18.17 In the end, EJS also contended that his testing is based upon sound objective scientific testing and his opinion is based upon sound scientific principles.

THE DEFENCE

18.18 The defence contended otherwise. They said that the contrary is true, EJS was cross-examined in depth and the defence called two experts, Dr Valery N. Aginsky ("VNA") and one Mr Peter Tytell ("PVT"). They did not perform any tests on the documents but they have put in reports in contradiction to EJS's reports with their critical analysis.

18.19 Before I consider the validity of the methods and findings of EJS and their scientific theory and validity, I would like to approach the problem from a layman's point of view first which I think ordinary people would understand. This is what the defence called "the true proficiency or blind test".
 
 

¹ p.409, E4-E5
² p.409, E4-E5 and T35:62:9
³ p.409, E4-E5
⁴ T35:54-55
⁵ p.923, E13
⁶ T35:58-61
⁷ T35:61:17-62:10
⁸ T35:62:18
⁹ p.413, EJS 1, E4-E5
¹⁰ para.5, p.909, EJS 3, E13
¹¹ para.15, p.885, EJS 3, E13
¹² para.42, p.897, EJS 3, E13
¹³ para.47, p.898, EJS 3, E13
 
 

19.1 The defence solicitors JSM instructed a firm called fJ Cleveland, who is a patent agent. fJ Cleveland wrote to EJS and under the disguise that they had a certain problem in 12 documents, or 12 signatures in 12 documents, which they would like to date, i.e. whether they were made recently or some years ago. In short, this is, to put it crudely, a sort of "set-up" made by the defence in order to find out whether or not the method used by EJS for ink-dating is valid, reliable and accurate.

19.2 The test was conducted in January 2001. EJS used two out of the three relative aging methods, namely, R-ratio and Percent Extraction methods. There was no mention of the Dye-ratio method until much later after the examinations.

19.3 The results of EJS's "proficiency" tests are as follows¹ :
 
 

There was no challenge by the plaintiff as to the dates of each and every document used by fJ Cleveland and they were sampled from letters of these various dates they have kept in their respective files. I accept the defence submission that Mr Cleff's affirmation (for fJ Cleveland) has established the dates of the aforesaid documents beyond any doubt. He was even offered to be cross-examined by Counsel for the plaintiff when his affirmation was already admissible because he was beyond the sea, but this offer was not taken up.² Those appalling allegations against Mr Cleff and the solicitors handling the case from Messrs Johnson, Stokes & Master amounted to allegations of dishonesty, fraud, forgery, and/or attempting to pervert the course of justice, should not have been made by the plaintiff's side. They are made without a shred of proof at all.

19.4 EJS concluded that³ :
 

"A relative ink aging examination was conducted on each ink found on each piece of paper within the 12 envelopes. Since most inks take approximately 3 to 3 1/2 years to dry completely on paper, if the writings were written within the last three years the ink should still be in the drying process and show differences dry or near dry. In this case, the [sic] all inks were found to be dry and showed no evidence of still being in the drying process except sample #11. Sample #11 showed some indication that it was still in the drying process, [sic] however due to the amount of overall evidence it cannot be concluded to any degree of certainty that this ink is still in the drying process. The difference in values obtained in the relative ink aging analysis between the heated and unheated samples (except for sample #11) showed no indication that the ink was still in the drying process.

Based on the above testing, it is my opinion that the writings were written at least 3 to 3 1/2 years ago." [emphasis added]

19.5 The defence contended that EJS was marked at 67% as the probability that he would wrongly date an ink is one out of three. However, in my view the matter is more serious than this. The result should be an accurate one and there should not be any mistake at all in order to be certain that the method is reliable. Otherwise when the dates of documents are not known, it would not be certain which result is reliable and which result is not. The result of this so-called "proficiency" test shows that the conclusion of EJS cannot even satisfy the civil standard of a balance of probability, i.e. it cannot be said that "all 12 signatures in these 12 documents were probably dry except Sample #11". The aforesaid contention containing two elements are both wrong! Thus, in my view, the defence is very lenient in their submission that EJS was marked at 67%. In fact when both elements in the aforesaid contention are wrong, EJS is 100% incorrect in his concluding statement.

19.6 Sample #11 is of particular importance and significance to this case. EJS said :
 

"The sample of ink in sample #11 is more likely than not, still in the drying process at the time the testing was performed, [sic] this means that the ink is less than three and one half years old."

and
 

"Based on this and the ink dating results that were obtained, the ink in sample #11 was not dry. Therefore, the conclusion is that the ink was not written eleven years ago, but was written sometime in the last three and one-half years."

19.7 The document from which Sample #11 was obtained was dated 5 June 1990, which was about three months after the stated date on the 1990 Will. EJS concluded that the ink on Sample #11 was still in the drying process and must therefore be less than three to three and half years of age. But in fact, the document and the signature thereon had been in existence for 11 years. I fully accept the defence submission that on the basis of this error alone, this court should be slow in accepting EJS's evidence in this case, in particular, his conclusion that the 1990 Will was not written on or near 10 April 1990 "but at a later time at least in 1996 or later".

19.8 Further, the ink-dating examinations were conducted under similar circumstances in both cases, in particular :
 

(a) no reference sample nor sample of known date was available; and

(b) the accelerated aging technique was allegedly used to mimic the natural aging of ink.

19.9 Further, the same statistical analysis was performed and the conclusions were both drawn at "one standard deviation" to determine if there was a "significance statistical difference".

19.10 EJS did not use the Dye-ratio method in his "proficiency" test. He never advised fJ Cleveland of its availability⁴ even when he was asked what additional examinations he could undertake to date ink on the documents. He expressed the following view of the Dye-ratio method only when fJ Cleveland explicitly asked of "the dye method", namely⁵ :
 

"The dye ratio is another method of relative ink age analysis that can be used in conjunction with the rate and percent extraction that were used in this case. This would have added to the expense and is not typically the most discriminating test. It could have been run and could still be run. It is possible that this method would show differences that the other methods did not show in the other 11 samples. The primary reason for not performing this test is cost to benefit ratio."

19.11 However, in his cross-examination, he gave the following contrasting view6 :
 

"Q. Before you started this test, or any test for that matter, do you know which of the three tests is likely to be more discriminating?

A. No.

Q. You do not?

A. No.

Q. You do not say, because of the age of the ink, this one is likely to be more discriminating than the other?

A. I would say if you have an ink that is very close to the end of the drying process that typically the dye-ratios would be, but not always. For instance, the Bic black example that I just gave - it would not be the dye-ratios. So you have an idea, but you cannot be certain that one would be more discriminating than others."

19.12 But in fJ Cleveland's first letter to EJS dated 3 October 2000, it was stated that⁷ :
 

"The purpose of this preliminary enquiry is to ask whether your laboratory undertakes this type of work, and secondly, what the approximate costs are likely to be in examining 10 to 12 signature inks on various documents which are alleged to have been signed over a period of 12 years or so."

19.13 In short, EJS was or should be aware that he might be dealing with "old" inks. In other words, EJS would or should have examined the 12 samples with the Dye-ratio method or at least, advised fJ Cleveland of its availability and use, should he be concerned with the "cost to benefit ratio".

19.14 In EJS3, he said⁸ :
 

"... each test (rate of extraction, percent extraction, and dye ratios) is independent of the others. The tests do measure different parameters of the ink."

In that case, that would be an additional reason for EJS to conduct the examinations with the Dye-ratio method, which he did not.

19.15 In EJS2, he said⁹ :
 

"The data from all three methods was evaluated and the Dye-Ratio method is the method that leads to the conclusion that the documents were not written on the dates that they purport to have been written. It is common for only one method to show a significant statistical difference while the other two methods overlap."

19.16 However, in his evidence, he gave the following contrary views¹⁰ :
 

"If the rate and per cent [sic] extraction do not show a difference, typically it is not. The only time it ever is found to be - or I should not say "the only time", but in general - is when you are dealing with an old ink, the rate of extraction and per cent [sic] extraction may be at the end of their curves and then it can be more discriminating. But typically, the rate or per cent [sic] would show a difference."

and¹¹ :
 

"Because I just said that, based on my experience, [the Dye-Ratio method] is not typically the most discriminating test. In some instances, it is. Typically, if you see a difference with the dye-ratios, you will see it with the rate or the per cent [sic] as well. You do not always, but typically that is the case."

19.17 There is no justification for these different and opposite approaches as submitted by the defence and, as EJS submitted, there is nothing atypical about this case.

19.18 Further, EJS admitted that it was possible that "the heat was unable to induce age in the ink by using the heating process".¹² There is no evidence to substantiate EJS's claim that accelerated aging did mimic natural aging in this case. It is therefore uncertain whether or not heat did induce aging in this case, and as a consequence, EJS's opinion in this case cannot be accepted. In view of his failure in the "proficiency" test, EJS made four contentions thereto. First, he claimed that the conclusion was not strong to a reasonable degree of certainty.¹³ However, in EJS 3 when he adopted one standard deviation he said that the opinion stated was "to a reasonable degree of scientific certainty".¹⁴

19.19 Secondly, EJS blamed the fact that the instructions given to him were unclear. However, from the documents before me, it is quite clear that the instructions were clear or as clear as EJS stated : "That all these documents are questioned, and let us see if we can find anything about them".¹⁵

19.20 Thirdly, EJS blamed fJ Cleveland for not giving him "a relevant timeframe and information about the samples, as to what the sides are, how to interpret it."¹⁶ However, if the accelerated aging method is accurate and reliable and it can determine the age of the ink as alleged, in the absence of any evidence of such prerequisite, it should not be necessary for the examiner to be given the "relevant timeframe ... as to what the sides are".

19.21 Lastly, EJS claimed that he was under the "misapprehension" that all 12 documents were prepared around the same time, which he claimed was the basis on which he drew his conclusion for his "proficiency" test.¹⁷ There was in fact no "misapprehension" in EJS's conclusion for his "proficiency" test, as it was stated by him that¹⁸ :
 

"all inks were found to be dry and showed no evidence of still being in the drying process except sample #11."

19.22 The sequences of exchanges between the parties all show that EJS changed his opinion to his client's, or what he perceived to be his client's need as follows :
 

Re : Sample #11 :

(a) EJS's initial conclusion was that¹⁹ : despite the indication that it was still in the drying process, it could not be concluded that the ink on Sample #11 was still in the drying process due to the amount of the overall evidence.

(b) Letter from fJ Cleveland, dated 25 May 2001²⁰ :
 

"... However, as you have probably gathered, the question is raised as to whether your tests can show that any of these documents are less than four years old. A finding that any of these documents were written within the time period you have indicated of 3 1/2 years would be of considerable significance.

Your finding that one signature is still drying is of particular interest, especially bearing in mind the nature of the document itself. Because of the particular nature of this document, we would like to pursue the matter as far as possible and as indicated in our earlier letter, it is agreed that the remainder of the signatures could, if necessary, be submitted for further study if that is likely to assist your analysis."

(c) Letter from EJS, dated 25 May 2001²¹ :
 

"3. The testing showed that one of these documents is still in the drying process.

...

... The sample of ink in sample #11 is more likely than not, still in the drying process at the time the testing was performed, this means that the ink is less than three and one half years old."

(d) Letter from fJ Cleveland, dated 8 June 2001²² :
 

"Point 3 - Hopefully the knowledge that this document is said to be eleven to twelve years old may assist further. ..."

(e) Letter from EJS, dated 12 June 2001²³ :
 

"3. Based on this and the ink dating results that were obtained, the ink in sample #11 was not dry. Therefore, the conclusion is that the ink was not written eleven years ago, but was written sometime in the last three and one-half years."

19.23 Re : The 11 remaining results :
 

(a) EJS's initial conclusion on the remaining 11 samples was²⁴ :
 

"In this case, the [sic] all inks were found to be dry and showed no evidence of still being in the drying process except sample #11. ... The difference in values obtained in the relative ink aging analysis between the heated and unheated samples (except for sample #11) showed no indication that the ink was still in the drying process.

Based on the above testing, it is my opinion that the writings were written at least 3 to 3 1/2 years ago."

(b) Letter from fJ Cleveland, dated 27 April 2001²⁵ :
 

"We are primarily interested in your conclusions [sic] and note that you indicate that all inks except sample 11 were found to be dry and therefore more than 3 - 3 1/2 years old.

We would however ask for further clarification in relation to sample 11 as this is of particular importance in this matter. We would therefore appreciate any further elaboration that you can make on this sample."

(c) Letter from fJ Cleveland, dated 25 May 2001²⁶ :
 

"The documents are all purported to be from various letters and related documents in support of patent files, many of which are said to be more than four years old, and indeed in some cases considerably more than four years old. However, as you have probably gathered, the question is raised as to whether your tests can show that any of these documents are less than four years old. A finding that any of these documents were written within the time period you have indicated of 3 1/2 years would be of considerable significance."

(d) Letter from EJS, dated 25 May 2001²⁷ :
 

"4. However if you want an opinion based solely on the ink dating results and the results of each documents to stand alone and independent from the next sample, I would state the following:

...
 

... None of the other samples show any indication of being in the drying process; this does not prove that they were done more than three and one half years ago."

5. The inability, based on circumstance, for a complete examination of the documents does not allow for a conclusive opinion as to the date of the documents."

(e) Letter from fJ Cleveland, dated 8 June 2001²⁸ :
 

"We also note that you state that 'none of the other samples show any indication of being in the drying process; this does not prove that they were done more than three and one half years ago.' We would request clarification on this point so that all parties fully understand the implications of what you are saying. In the fourth paragraph of page 2 of your report of 19 January 2001, you state :
 

'However, if the inks were not fully aged, then at least one of the relative ink age comparison methods that was used would show a significant statistical difference between the aged and unaged samples' [emphasis added].

Would you please clarify this further: if all the samples that we sent you are said to be more than 3 1/2 years old, and (excepting sample 11) you found that they were all "dry", does this not mean that they could not have been produced in the last 3 1/2 years, and that therefore your analysis has authenticated their alleged 'old age'?"

(f) Letter from EJS, dated 12 June 2001²⁹ :
 

"4. With respect to the other eleven samples of ink analyzed, the inks showed no indications of being in the drying process. This does not confirm that they are completely dry. Three possibilities exist with this type of testing:
 

(a) The ink is completely dry.

(b) The testing did not discriminate a difference in age using the testing parameters chosen.

(c) The heat was unable to induce age in the ink by using the heating process."

19.24 In conclusion, in light of EJS's failure in the "proficiency" test, I accept the defendant's submissions that the court should reject EJS's opinion offered in this case for the following reasons, inter alia :
 

(a) EJS's method has proven to be unreliable in light of his failure in the "proficiency" test.

(b) EJS failed to date an ink on a letter which was written about three months after the date of the 1990 Will.

(c) The Dye-ratio method was not "typically discriminative", and there was no sound scientific justification as to why in this case, a conclusion drawn on the basis of this method should be received in favour of the overwhelming negative results obtained from the Dye-ratio method as well as the other four methods.

(d) Given that it was possible that the heat was unable to induce age in the ink by using the heating process,³⁰ and there is no evidence that it did take place in this case, EJS's conclusion is unreliable.

19.25 Before I shall consider other aspects of EJS's method and its reliability and validity, I shall first of all consider the legal position of expert's evidence first in the next chapter.
 
 

¹ pp.2327-2328, G9 and p.3075, G10, JMC-14
² T47:60:7-13
³ p.3075, G10, JMC-14
⁴ pp.3023-3024, 3028-3029 and 3033, G10, JMC-10
⁵ pp.3105-3107, G10, JMC-15
⁶ T38:23:6-20
⁷ p.3020, G10, JMC-10
⁸ para.46, p.898, EJS 3, E13
⁹ p.433, E4-E5, EJS 2
¹⁰ T47:44:13-20
¹¹ T47:110:7-12
¹² pp.3088-3108, G10 at p.3104 and confirmed on T47:87:17-22
¹³ T47:74:12-20
¹⁴ para.20, p.888, EJS 3, E13
¹⁵ T47:60:5-6
¹⁶ T47:79:15-17
¹⁷ T47:91:10-19
¹⁸ p.3075, G10, JMC-14
¹⁹ p.3075, G10, JMC-14
²⁰ pp.3093-3094, G10, JMC-15 at p.3093
²¹ pp.3097-3098, G10, JMC-15
²² pp.3099-3101, JMC-15, G10
²³ pp.3103-3104, JMC-15, G10
²⁴ p.3075, JMC-14, G10
²⁵ pp.3089-3090, JMC-15, G10
²⁶ pp.3093-3094, JMC-15, G10
²⁷ pp.3097-3098, JMC-15, G10
²⁸ pp.3099-3101, JMC-15, G10
²⁹ pp.3103-3104, JMC-15, G10
³⁰ pp.3088-3108, G10 at p.3104 and confirmed on T47:85:5-87:22
 
 

20.1 In The Ikarian Reefer [1993] 2 Lloyd's Reports 68, at 81-82, Cresswell J formulated the duties of an expert and what is expected of him as follows :
 

"The duties and responsibilities of expert witnesses in civil cases include the following :

1. Expert evidence presented to the Court should be, and should be seen to be, the independent product of the expert uninfluenced as to form or content by the exigencies of litigation (Whitehouse v. Jordan, [1981] 1 W.L.R. 246 at p.256, per Lord Wilberforce).

2. An expert witness should provide independent assistance to the Court by way of objective unbiased opinion in relation to matters within his expertise (see Polivitte Ltd. v Commercial Union Assurance Co. Plc., [1987] 1 Lloyd's Rep. 379 at p.386 per Mr. Justice Garland and Re J, [1990] F.C.R. 193 per Mr. Justice Cazalet). An expert witness in the High Court should never assume the role of an advocate.

3. An expert witness should state the facts or assumptions upon which his opinion is based. He should not omit to consider material facts which could detract from his concluded opinion (Re J sup.).

4. An expert witness should make it clear when a particular question or issue falls outside his expertise.

5. If an expert's opinion is not properly researched because he considers that insufficient data is available, then this must be stated with an indication that the opinion is no more than a provisional one (Re J sup.). In cases where an expert witness who has prepared a report could not assert that the report contained the truth, the whole truth and nothing but the truth without some qualification, that qualification should be stated in the report (Derby & Co. Ltd. and Others v Weldon and Others, The Times, Nov. 9, 1990 per Lord Justice Staughton).

6. If, after exchange of reports, an expert witness changes his view on a material matter having read the other side's expert's report or for any other reason, such change of view should be communicated (through legal representatives) to the other side without delay and when appropriate to the Court.

7. Where expert evidence refers to photographs, plans, calculations, analyses, measurements, survey reports or other similar documents, these must be provided to the opposite party at the same time as the exchange of reports (see 15.5 of the Guide to Commercial Court Practice)."

20.2 The aforesaid observations had been adopted by the Court of Appeal in England in Stanton v. Callaghan [1998] 4 All ER 961, at 991a and, specifically, post-Woolf reforms, in Stevens v. Gullis [2000] 1 All ER 527. In the Chief Justice's Working Party on Civil Justice Reform, paragraphs 496 and 497, it was said :
 

"496. After the CPR came into operation, in the English Court of Appeal's decision in Stevens v Gullis [2000] 1 All ER 527, Lord Woolf pointed out that the duties in The Ikarian Refeer continue to be reflected in, and given emphasis by, the new rules.

497. Reforms adopted in various jurisdictions all begin with the same premise, namely, that the function of the expert is to help the court by providing independent and impartial advice - not to act as an advocate for his client. This is reflected in rules which emphasise that the expert's duty to the court overrides his duty to his client."

20.3 The Court's approach to the reception and evaluation of expert testimony was stated in Davie v. Magistrates of Edinburgh [1953] SC 34, at 40 :
 

"Expert witnesses, however skilled or eminent, can give no more than evidence. They cannot usurp the functions of the jury or Judge sitting as a jury, any more than a technical assessor can substitute his advice for the judgment of the Court. Their duty is to furnish the Judge or jury with the necessary scientific criteria for testing the accuracy of their conclusions, so as to enable the Judge or jury to form their own independent judgment by the application of these criteria to the facts proved in evidence. The scientific opinion evidence, if intelligible, convincing and tested, becomes a factor (and often an important factor) for consideration along with the whole other evidence in the case, but the decision is for the Judge or jury." (my emphasis)

This authority was followed in Hong Kong in a case of R. v. Yeung Kwok Fai [1996] 1 HKC 754, at 757H-1 and also in Australia in :
 

(a) R. v. O'Callaghan [1976] VR 676, and

(b) R. v. Lucas [1992] 2 VR 109.

20.4 In the United States, two cases have been referred to in evidence as setting the standards for admissibility of scientific evidence in the United States :
 

(a) US v. Frye 293 F 1013 (1923) ("Frye"), and

(b) Daubert v. Merrell Dow 509 US 579 (1993) ("Daubert").

20.5 Frye was a case concerning the admissibility of polygraph or lie-detector tests. What is referred to as "the Frye test" is satisfied if the scientific theory advanced is established to have gained general acceptance in the particular field in which it belongs. Thus, the "general acceptance" theory is still the underlying basis for the admission of scientific theories in the United States. The Frye test has been stated to be in accord with the approach in England in the following cases :
 

(a) R. v. Gilfoyle, The Times, 13 Feb 2001 per Rose LJ, and

(b) R. v. Clarke [1995] 2 Cr. App. R 425.

20.6 Daubert, to some extent superseded Frye in the United States, because thereafter the Federal Rules of Evidence were promulgated which had some limiting effect. The US Supreme Court in Daubert held that the trial judge must ensure that any and all scientific testimony or evidence admitted is not only relevant but reliable. Proposed scientific testimony must be supported by appropriate validation, that is good grounds, based on what is known to science. The law requires that expert scientific testimony pertains to "scientific knowledge" and establishes a standard of evidentiary reliability, i.e. trustworthiness. Evidentiary reliability is based on scientific validity. Thus, in effect, Daubert added reliability to the requirement of general acceptance in Frye.

20.7 In the more recent case of Nelson v. American Sterilizer Co., 223 Mich. App. 485 (1997), the Court emphasised on the proposed testimony of an expert must be derived from "recognized scientific knowledge" and the inferences or assertions must be supported by appropriate objective and independant validation based on what is known in scientific and medical literature.

20.8 From the aforesaid authorities, I accept the submissions of the Counsel for the defendant, Mr Clive Grossman, SC, who appeared together with Ms Alice Lee (in this area of ink-dating only), that certain principles may be derived, and each and all of which must be fulfilled if the Court is to accept the evidence of a specific scientific theory, novel or not :
 

(a) The person propounding the scientific theory must have the necessary qualifications, expertise, experience and integrity to ensure that the Court can have confidence that his testimony is worthy of consideration.

(b) The theory must have a sound scientific basis, comprehensible to the Court.

(c) The theory should have gained widespread support amongst that sector of the scientific community which would be likely to utilise it or its results.

(d) The methods used to carry out the scientific test should be safe and reliable, and follow an established protocol, i.e. one that has been published, disseminated and acknowledged to be reproducible.

20.9 The defence submitted that EJS's evidence fails at every single hurdle. I shall consider in the next chapter whether that is the case and I shall start with the underlying scientific theory of the Dye-ratio method. But before I do, I should point out that Counsel for the plaintiff did not submit otherwise contrary to the defence submissions hereinbefore.

20.10 They only drew the Court's attention to the decision of the Court of Final Appeal in Aktieselskabet Dansk Skibsfinansiering v. Brothers & Others (2000) 3 HKCFAR 70, where the Court of Final Appeal reiterated the importance of putting to witnesses things which the Court is invited to make findings adversed to the opposite party. In that case, because the allegation was not sufficiently and clearly put to the party Mr Brothers, the Court of Final Appeal had taken the unusual course of reviewing concurrent findings of fact in the courts below and reversed it. Lord Hoffmann NPJ said, at p.91J :
 

" In addition to particularity in the pleadings, fairness requires that the adverse findings which the Judge will be invited to make should have been put squarely to the witness in cross-examination, so that he can have the opportunity to offer an explanation."

20.11 Lord Hoffmann NPJ in fact considered all other relevant pieces of evidence concerning the issue of whether Mr Brothers had dishonestly concealed the cash flow position of the borrower company. His Lordship said at p.92A that :
 

"The Court has examined in detail the relevant parts of the cross-examination of Mr Brothers and I cannot find any place in which the two allegations about his state of mind were adequately put to him."

20.12 His Lordship then examined in detail all the pieces of evidence (see pp.92-96) and came to the conclusion that the evidence to contradict Mr Brothers is insubstantial, and thus the lender company Aktieselskabet Dansk Skibsfinansiering had failed to discharge the burden of proof that Mr Brothers "deliberately withheld the cash flow and, notwithstanding his belief in support, he was dishonest in doing so".

20.13 The plaintiff relied on the case of ADS v. Brothers and submitted that some of the allegations against Mr Speckin have not been put to him. I shall examine the evidence in the following chapters the effects of such failure.
 
 

21.1 The Dye-ratio method was developed by a Mr Richard Brunelle ("Mr Brunelle") and a Mr Huy Lee ("Mr Lee"), who was then only a student. The scientific basis upon which the Dye-ratio method is founded can be found in "Determining the Relative Age of Ballpoint Ink Using a Single-Solvent Extraction, Mass Independent Approach".¹ This paper appears to be the only paper which discussed the underlying theory of the Dye-ratio method.

21.2 At pages 746 and 748, Brunelle and Lee said :
 

"Methanol is a fairly strong solvent for extracting ballpoint inks, yet it still produced a good aging curve in Fig.10. It was this finding that led to the speculation that perhaps subtle dye deterioration or fading or both may also be a factor of aging.

.......

Therefore, the authors believe that extractability and subtle fading/decomposition of dyes are both aging factors ..."

21.3 Thus, the two hypotheses presented for the Dye-ratio method are :
 

(a) The ratio of dye-ratios changes as a consequence of changes in their extractability as the ink ages on paper (the "First Hypothesis"); and/or

(b) The ratio of dye components changes as a consequence of some of them decompose with age (the "Second Hypothesis").

In short, according to Brunelle and Lee, the Dye-ratio method purports to measure the extractability of dye components, or the decomposition of dye components, or both, with a view to determining the age of ink.

21.4 However, whilst no results were presented by Brunelle and Lee to verify the First Hypothesis, VNA pointed out² that their experimental data presented in this paper did verify the Second Hypothesis. Each of the two Hypotheses are considered in turn hereinbelow.

THE FIRST HYPOTHESIS (EXTRACTABILITY)

21.5 VNA considered that the First Hypothesis should be rejected.³ He said that generally ballpoint inks consist of three main ingredients, namely :
 

(a) Dyes;

(b) Solvents (which are used to dissolve or disperse the dyes); and

(c) Resins (which are used to thicken the inks).

21.6 As soon as the ink is placed on paper, various processes of aging will start. These processes would include :
 

(a) gradual decomposition of some or all ink dyes;

(b) evaporation of solvents; and

(c) hardening of ink resins.

21.7 The ink will become harder to extract in weak solvents as it ages, as a consequence of the age-transforming process of the hardening of ink resins. VNA considered that only the solubility of ink resins changes (decreases) with time, whilst the solubility of dye components remains constant as the ink ages. The extractability of dye components also decreases with time, but this is only because the dye components become more difficult to be extracted from the hardening or hardened (depending on the age of the ink) "matrix"-ink resin.

21.8 Extraction of ink in weak solvents is a complex process. Dye components are sort of "trapped" in ink resins. Therefore, before the dye components may be extracted, the ink resin has to be dissolved. However, the solubility of ink resin differs by a number of factors including the type of ink resin, the age of the ink and the strength of the solvent used for extracting the ink. As the ink ages, its ink resin also hardens. It follows that the older the ink, the harder thus the longer it will take to dissolve the ink resin. But when the ink resin is gradually dissolved, the dye components start to come into contact with the extracting weak solvent. VNA described it as "layer by layer" by using a simplified example illustrated at G11, p.3229. The trouble is if the solubility of dye components in weak solvent :
 

(a) is not too weak to make it an irrelevant factor for consideration, and

(b) does not differ dramatically for different ink dye components (bearing in mind that solubility is a physical property which remains constant irrespective of the age of the ink),

the dye components will be dissolved very quickly.

21.9 In other words, as VNA explained in evidence, when an aging ballpoint ink is being extracted in a (weak) solvent, the main limiting factor for it to be extracted quickly is the dissolution of the hardening or hardened, if the ink has already ceased aging, ink resin. As far as the ink dye components are concerned, their rate of extraction or their extractability is a priori significantly larger than the rate of extraction of the hardening or hardened ink resin. For this reason, the rate of extraction of the ink dye components would depend only on the solubility of the ink resin in the weak solvent (provided, of course, the dyes are also soluble in the weak solvent), which in turn, depends on the age of the ink. In short, the extractability i.e. the rate of extraction of the ink dye components will practically be equal to the extractability or rate of extraction of the ink resin.

21.10 As VNA pointed out, the experimental results published in Brunelle and Lee's paper show that the Dye-ratio method was not measuring changes in the extractability of the dye components, but the decomposition of dye components, as the ink ages⁴ (i.e. the Second Hypothesis as explained hereinbelow). This is because the extractability of dye components is not an aging factor.⁵ These will be considered further below.

21.11 Thomas Andermann and Roderick Neri published a paper in which they said they also had difficulty in understanding why the extractability of a dye component, methyl violet homologues (which are used by Mr Brunelle in all his papers to demonstrate the validity of the Dye-ratio method) should change as the ink ages on paper, given that the methyl violet homologues are chemically closely-related substances, and therefore have almost identical physical and chemical properties, including their solubility in butanol and other solvents.⁶

21.12 By reason of the aforesaid matters, the First Hypothesis that the Dye-ratio method measures the difference in the extractability of dye components in ink on the premise that the extractability of dye components will change or decrease differently as the ink ages, must not be accepted. There is no scientifically sound explanation and it has not been supported by any published experimental data.

THE SECOND HYPOTHESIS (DECOMPOSITION)

21.13 The Second Hypothesis is that some ink dye components decompose as the ink ages, hence the change in ratios of dye components in an ink. Using this property of the dye components, EJS claimed that one can, therefore, determine the age of an ink by measuring the change in dye ratio due to the decomposition of dye components.

21.14 Thus, the Dye-ratio method measures the decomposition of ink dye components as follows⁷ :
 

(a) VNA showed that all published experimental data related to the Dye-ratio method clearly indicate that what this method actually measures is changes in the relative content of ink dye components as a consequence of the gradual decomposition of some ink dyes as the ink ages.⁸

It is suggested therefore, that the Second Hypothesis can be accepted. However, whilst this is the case, the decomposition of dye components could not be used for the purpose of determining the age of an ink as this process continues indefinitely and does not depend on the age of the ink.

(b) The experimental data obtained by Brunelle and Lee verified the validity of the Second Hypothesis. They obtained the two dye-ratio aging curves for Formulab 587 black ballpoint ink as shown in Figure 11 at p.748 in E7-E10.

(c) The upper curve shows the dye-ratios using pyridine, a "strong" solvent, and the lower curve shows the dye-ratios using n-butanol, a "weak" solvent. Both curves show similar shape. It means that the strength of the extracting solvent has no effect on the relationship between the dye-ratio and the age of the ink. Therefore, the Dye-ratio method does not measure the extractability of dye components in ink, but their decomposition.

(d) In VNA's research paper : "A Microspectrophotometric Method for Dating Ballpoint Inks - A Feasibility Study".⁹ VNA's findings are in agreement with the results published in Brunelle and Lee's paper. VNA conducted the experiments which measured the ratios of ink dye components as a function of aging of ink as Brunelle and Lee did, using a different analytical method, i.e. microspectrophotometry. He found that whilst the ratio of comparatively unstable methyl violet and stable copper phthalocyanine dyes gradually decreases with age during a six-year period i.e. (up to six years old writings) made by a Parker blue ballpoint ink were available in that study, no satisfactory correlation was established between the age of the ink analysed and the content of its dye components.

21.15 Even in the aforesaid paper of Brunelle and Lee, it also confirmed that ink takes more than five years to "dry".

21.16 VNA pointed out in his evidence that neither of the aging curves in the said Figure 11 show any indication of levelling off after a five-year period of time.¹⁰ This is a clear indication that the process of decomposition of ink dyes components will continue and does not stop (or in the case of Formulab 587, does not even slow down) after five years contrary to EJS's claim that this process will stop after three to three and a half years. This is also in agreement with VNA's own finding that the decomposition of dye components does not stop after six years.¹¹ As VNA said¹² :
 

"But the problem is that this process does not stop after 3 years, 3 1/2 years, or 4 years that are used by Mr Speckin to draw a conclusion that the ink [on the document] is still in the aging process. This decomposition goes further and it does not stop after 5, 6, 10, 20 or 30 years; it is a process that we cannot even find out when it will stop."

21.17 In light of the continuing process of decomposition of dye components, which process has no means of determining its termination, it is not possible to obtain a reference (artificially aged) sample of an unknown ink (which has reached the end of its natural aging curve with the age of the ink known to us), to which samples under examination could be compared. Thus, although the Dye-ratio method does measure the decomposition of dye components, it could not measure the age of the ink or it cannot be certain whether certain ink is beyond five or other years of age and definitely could not determine whether the ink is or is not more than three or three and a half years old.

21.18 Further, the Dye-ratio method has other sources of error. They are as follows :
 

(a) Dye components in ink, as discussed above, decompose as the ink ages. However, such decomposition may be attributed to different factors, independently or collectively. Varying these factors would therefore vary the rate of decomposition of dye components. One of such factor is the thermal instability of some dye components. Some dyes decompose quicker when subject to increased temperature, whilst some dyes are less affected by the difference in temperature. Methyl violet is known to be a relatively unstable dye when it is subjected to increased temperature.

(b) It is common ground that ink on Documents A and C contain methyl violet dye, and that reference samples were prepared using the accelerated aging technique. When these samples were heated, the methyl violet components were further decomposed. We should, therefore, expect to receive a statistically significant difference between the heated and the unheated samples. Yet, it is not possible to conclude that the ink is, therefore, less than three, or five, or six years old, as we do not know how old the heated sample represents and cannot therefore draw any conclusion on the basis of such reference sample. Indeed, any statistically significant difference between the heated and unheated ink samples would likely be the result of, at least, the thermal instability of the methyl violet components in the ink.

(c) VNA gave an example in his evidence.¹³ Supposing two samples of ink-on-paper are taken from a 50-year old ballpoint ink entry on a document. The ink contains, inter alia, methyl violet. The samples are then examined using the accelerated aging techniques (i.e. one of the samples is heated to prepare an artificially aged reference sample) and the Dye-ratio method as EJS did in this case. At the end of the examination, we should expect to receive a statistically significant difference in the dye ratios between the two samples. This is because when the reference sample is being heated, some part of the methyl violet components will inevitably decompose further. One cannot, however, conclude that the ink on the document is not "dry" as, as we know, it is 50 years of age.

(d) Another source of error is the photo-instability of some dye components. According to the experiments conducted by Andermann and Neri¹⁴, there were significant changes in the mass ratio of three methyl violet homologues after the samples were exposed to light.

21.19 Thus in conclusion in respect of the Second Hypothesis, on one hand, the Dye-ratio method measures decomposition of dye components and this hypothesis can be accepted because its validity has been established and verified by the published experimental results obtained independently by Brunelle and Lee and by VNA. However, on the other hand, the decomposition of dye components cannot be used as a reliable parameter for the purpose of determining the age of an ink because :
 

(a) The process of decomposition will not stop and does not even slow down as shown by Brunelle and Lee, and VNA after five or six years; and/or

(b) In light of the existence of the sources of errors as discussed above.

21.20 In reply, counsel for the plaintiff submitted that the allegation of the behaviour of the resin and the ink dye in solvent was not in VNA's Report nor was it fully put to EJS. Whilst I would accept that it is something less than what would be desired, the circumstances of this case is somewhat different from the case of ADS v. Brothers.

21.21 EJS was in Court whilst VNA was cross-examined. EJS's answer to VNA's allegation of the behaviour of resin can be put to him. (This was done in respect of other points not specifically or fully put to EJS by the defence.) In the end, it would depend on the soundness of the reason or the lack of it and not on the demeanor of the witness. Thus, in spite of the failure of the defence in putting fully this allegation to the plaintiff's expert, I find the contention of VNA has casted grave doubt on the scientific validity of the Dye-ratio method.

CONCLUSION

21.22 By reason of the aforesaid matters, I accept the submissions of the defence as follows :
 

(a) The Dye-ratio method does not measure the extractability of dye components. It measures changes in the relative content or ratio of ink dye components, which change is a result of the gradual decomposition of some ink dye components as the ink ages.

(b) The decomposition of ink dyes is a long-lasting process which does not stop even after scores of years of natural aging of ink. As the duration of this process cannot be determined, any attempt to prepare a reference sample by artificially aging a sample (which allegedly corresponds to an ink of certain age) would a priori be meaningless.

(c) Indeed, if an ink contains a dye which decomposes gradually during the natural aging of ink at normal temperatures, this dye would decompose at a significantly higher rate when a sample of it is being artificially aged using elevated temperatures. The extent to which this dye would decompose when the accelerated aging technique is applied does not however depend at all on the ink age but depends solely on the elevated temperature and the duration of heat applied to prepare the artificially aged sample.

(d) Thus, if an ink contains dyes capable of gradual decomposing with time like methyl violet as we have in this case, the Dye-ratio method coupled with the accelerated aging technique would almost certainly produce false evidence that the ink is still aging, irrespective of the actual age of the ink.

Therefore, a conclusion based on results obtained by the Dye-ratio method coupled with the accelerated aging technique, as we have in this case, cannot be accepted.

21.23 Further, the method is yet to be validated and accepted by the scientific community and the courts which I shall discuss in the next chapter.
 
 

¹ pp.736-753, E7-E10
² T57:20:11-22:24
³ T51:11:17-12:11
⁴ T54:28:7-31:15, T57:20:11-22:24
⁵ See Figure 11 at p.748, E7-E10 (VNA 2)
⁶ pp.781-797, E7-E10 (VNA 2)
⁷ T51:12:12-14:3
⁸ T54:28:7-31:15, T57:20:11-22:24
⁹ pp.693-697, E7-E10 (VNA 2)
¹⁰ T54:30:20-31:11
¹¹ p.694, E7-E10, Figure 2 in VNA-4 (VNA 2), and T57:23:15-24
¹² T51:12:25-13:6
¹³ T51:13:8-18
¹⁴ p.794, E7-E10 (VNA 2)
 
 

22.1 EJS concluded that the ink on Documents A and C was still in the drying process. This conclusion reached must be :
 

(a) adequately supported by reliable data;

(b) that the data was produced by a scientific method which has a sound underlying scientific theory;

(c) the method has been proven to be valid, reliable; and

(d) the same has generally been accepted by the scientific community.

22.2 I shall consider the aforesaid criteria one-by-one hereinbelow. However, it is worthy to look at the outset that the conclusion reached by EJS was based on "significant statistical difference" allegedly found in only five out of 18 results obtained between the heated and the unheated ink samples using the Dye-ratio method. There was no difference detected from the other two relative aging tests, namely the Percent Extraction method and the R-ratio method, which were performed on the same ink samples, before the Dye-ratio method.

VALIDATION STUDIES

22.3 Validation is a crucial step to demonstrate that a scientific method intended for forensic use will work as expected and will provide reliable results. Validation is a term pertaining to studies in ascertaining the validity, reliability and reproducibility of test results. It has been said that¹ :
 

"the purpose of validation is to establish limits within which we are confident of the results obtained. Validation also helps us to establish the variance inherent in the system as a whole, and to gain a further understanding of one kind of explainable difference. ... We learn how much variation may exist between tests, so that we may determine criteria within which we can either conclude that two different samples may share a common origin or exclude them as having a common origin. How sensitive is the system to the quantity of sample? What are the lowest and highest quantities that can reliably be analyzed? What artifacts begin to appear at the extreme ends of the spectrum and can we use them as diagnostics? What do mixtures of similar substances look like in a particular test system, and are there indicators that might alert us to the presence of a mixture? How might various environmental factors to which forensic samples are commonly subjected influence the ability to analyze the samples and to interpret the results obtained? These are the kinds of questions that validation studies attempt to answer."

22.4 The defence witness, PVT said that validating a new technique involves efforts of multiple laboratories, and broad and comprehensive testing using the same samples. The relevant scientific community can then rely on the resulting data which ideally will be peer-reviewed and published. When substantive changes are introduced, the protocol must be revalidated in order to understand the full impact of those changes. This has never been challenged by the plaintiff.

22.5 Falling short of validation, it should be demonstrated, at the very least, that the test methods are reproducible by the scientific community. In order to do so, sufficient experimental data obtained should have been published in peer-review journals. The underlying scientific basis of the methods, the procedure, and the data obtained from the test of different types of ink under different environmental conditions, on different types of paper and stored under different conditions, by using different heating parameters, should be set out to enable other practitioners to follow their experiments and verify their reproducibility.

22.6 The methods, accompanied by appropriate standard protocols, should be endorsed by a consensus of practitioners, with widely disparate backgrounds and interests, such as those from private and public laboratories all agreeing on what should be done and not simply (as in this case), members from a small and young association namely, SOFIA.²

22.7 Mr Larry Stewart and Ms Susan Fortunato, (both of the United States Secret Service), in their paper entitled : "Distinguishing between Relative Ink Age Determinations and the Accelerated Aging Technique" published in 1996, stressed the importance of the validation process before applying a new technique to casework. They expressed that³ :
 

"Every forensic scientist should agree that before applying a new technique to casework, it is important to validate the work, procedures, limitations and to illustrate the soundness of the technique through the close scrutiny of your peers".

22.8 There is no doubt that the disputed methods used here have not undergone any validation studies.⁴

22.9 On the contrary, it appears that SOFIA, consisting of a very limited number of persons including Dr Lyter, Mr Robert Kuranz and Mr Brunelle, and of which EJS is the Vice President, has only just started to validate the other ink-dating procedures namely, the Percent Extraction method and the R-ratio method on comparison between known and questioned ink samples, not involving the use of artificial heating to create reference samples for comparison purposes.⁵ Their attempt in June 2001 inviting other laboratories, including government laboratories, to participate in the project had nothing to do with either the accelerated aging method or the Dye-ratio method. In short, there has never been any validation studies for the Dye-ratio method employed by EJS in this case.

PEER-REVIEW

22.10 EJS contended that the methods have been published in peer-review journals.⁶ Alternatively, he said that the process of validation does not require specific endorsement by any organization, so long as the methods have been accepted by the scientific community.⁷ However, simply by publishing a new technique in a peer-review journal (even with full experimental/research data) does not guarantee that the technique is reliable, reproducible, and is certainly not equivalent to it having been validated.

22.11 A method can only be said to have been peer-reviewed when it has been debated, and later agreed and accepted by the general scientific community that the method has a sound scientific basis, and that it is reliable and reproducible. Here on the contrary reservations have been expressed by other experts including Mr Joel Harris⁸ and Dr Giles⁹.

22.12 There has not been any or any sufficient research data on Dye-ratio method used in conjunction with the accelerated aging method published so far. The first paper on the Dye-ratio method, as stated hereinbefore, was published by Brunelle and Lee in 1989.¹⁰ Data obtained from initial experiments using the Dye-ratio method was published but the authors did not appear to know with any degree of certainty the underlying scientific basis of the Dye-ratio method. It was said at p.628 that :
 

"It was this finding that led to the speculation that perhaps subtle dye deterioration or fading or both may also be a factor of aging."

This was followed by the statement that :
 

"...the authors believe that extractability and subtle fading/ decomposition of dyes are both aging factors." (Emphasis added).

22.13 In 1995, Mr Brunelle published the article entitled : "A Sequential Multiple Approach to Determining the Relative Age of Writing Inks". Only procedures, but not experimental data, in respect of various ink-dating techniques were described. The procedure in respect of the accelerated aging technique was also described. However thereafter, doubts on the reproducibility and reliability of the methods were raised by leading members of the relevant scientific community. They included VNA, Ms Joanne Becker, Mr Larry Stewart and Ms Susan Fortunato, Dr Albert Lyter, Dr Thomas Andermann and Mr Rodric Neri.¹¹

22.14 In particular, Mr Larry Stewart and Ms Susan Fortunato in a paper entitled : "Distinguishing between Relative Ink Age Determinations and the Accelerated Aging Technique", conducted a review of all the papers on the accelerated aging method, including those relied upon by EJS in support of his contention that the methods have been published and peer-reviewed. The authors found that there had been no consistent result that had led multiple examiners to utilize a given procedure and that no reported studies existed which examined the effects of various time and temperature parameters on numerous ink formulations. They contended that research of this type might determine not only the maximum time and temperatures that might be applied to inks without fear of degradation, but also the conditions necessary to ensure that the heated sample was then representative of itself at the end of its natural aging curve.

22.15 Further, they also found that the origins of time/temperature combination were unclear. They felt that the practice of using unverified heating parameters could lead to erroneous conclusions unless further studies were conducted. They concluded that there were many unanswered questions with respect to the impact and the effect of applying heat to an ink entry. The influence from paper type, the time and temperature parameter selections, heat degradation and significant differences were all in need of attention.¹² These questions merit serious consideration and are as yet, unanswered.

22.16 Since then, the only paper that has been published on the subject is : "Technical Report with Case Studies on the Accelerated Aging of Ball-point Inks" by Brunelle and EJS in 1998¹³, but it shed no further light on the aforesaid unanswered questions. At about the same time, a paper by VNA entitled : "Measuring Ink Extractability as a Function of Age" published in the same Journal, raised profound doubts as to the reproducibility of the Dye-ratio method. Before coming to that conclusion, VNA discussed the usability of the relative aging tests which included the other two methods namely, the R-ratio method and the Percent Extraction method.

22.17 The aforesaid paper published by Brunelle and EJS in 1998 did not, however, alleviate the concerns of the scientific community. In its abstract, it was acknowledged that "the data presented does not satisfy the requirements for [it] to be a research paper". Upon review of the same, it did nothing more than setting out results allegedly obtained from real caseworks over the years before EJS's time, some of which were from the use of the Dye-ratio method. In a letter dated 17 September 1997¹⁴, which was addressed to Mr Joel Harris, the then editor of the Journal in response to a reviewer's request for additional research on the effects of comparing inks on old versus new paper, Mr Brunelle confirmed that "the paper was not intended to report the results of research". He said "the recommended study may be the subject of another paper in the future". It goes without saying that it was recognized that further research on the various ink-dating techniques was required. However, no further research appears to have been conducted up to the present day. Furthermore, full data from the Dye-ratio method was absent from the report. The graphs in respect of the Dye-ratio method to be found in the said article contained data of one Dye-ratio only in order to "preserve space".¹⁵

22.18 Other literature cited in support of the claim that the methodology has been published widely and peer-reviewed are :
 

(a) "Chromatographic Analysis of Inks for Forensic Science Application" published in 1991 by Dr I.R. Tebbett; and

(b) "A Sketch of Analytical Methods for Document Dating - Part II, the Dynamic Approach" published in 1995 by Dr A.A. Cantu.

These are, however, no more than reviews of the technical papers published on the subject and are not to be taken as, nor do they purport to be, independant studies for the purpose of validating the methods. In the course of his review of the progress of ink-dating techniques, Dr Tebbett said : "the solution continues to elude researchers and document examiners alike".¹⁶ It is noted that when EJS testified in the case of Utica Square Shopping Centre v. Renberg's Inc in July 1998 he was dismissive of Dr Tebbett and did not consider him to be an authority on ink-dating.¹⁷

22.19 In the end, I must conclude, as submitted by the defence, that not only have the accelerated aging method and the Dye-ratio method not been validated, the procedures described in Mr Brunelle's paper published in 1995 in respect of the Dye-ratio method and the accelerated aging method were not followed by EJS¹⁸ and he admitted in his evidence that he had not followed the procedures "to the letter".¹⁹ This will be discussed further in Chapter 28 on Procedures. Before that, I shall deal with the question of "Acceptance by the Scientific Community" first in the next chapter.
 
 

¹ Chapter 9, Good Laboratory Practice - Establishing Validity and Reliability, extracted from Principles and Practice of Criminalistics, pp.3130-3132, G11
² See PVT, T49:44:6-59:19
³ p.773, E7-E10
⁴ T39:106:16-108:13
⁵ See SOFIA's Ink-dating Research Project, pp.2097-2112, G8
⁶ T35:37-40; T49:56:2-8
⁷ T49:55:7-12
⁸ p.2094, G8
⁹ p.155, C
¹⁰ pp.619-635, F2-3; also pp.737-753, E7-E10
¹¹ They are exhibited to VNA 2, respectively at pp.694-697, pp.761-768, pp.770-775, pp.777-780 and pp.782-797, E7-E10.
¹² p.773, E7-E10
¹³ p.1386, G5
¹⁴ p.1537, F6
¹⁵ pp.1391-1394, 1396, and 1398, G5
¹⁶ p.614, F2-3
¹⁷ p.1012, F5; T44:19:3-25:25
¹⁸ Brunelle : "A Sequential Multiple Approach to Determining the Relative Age of Writing Inks", pp.603-607, F2-3
¹⁹ T37:79:23-80:2, T38:51:18-22
 
 

23.1 A further step in deciding whether the methods are reliable and whether they have been validated is to examine the extent of their acceptance by the scientific community. EJS no doubt is aware of the strictures of Frye/Daubert, by which the boundaries of the acceptance have been laid down as : "those who would be expected to be familiar with its use". Thus EJS himself also stressed the long-time and universal acceptance of his method. However, he confined those private ink chemists to those like-minded members of SOFIA only.

23.2 But then the methods/techniques used by EJS in this case have apparently not been used by a majority of chemists/scientists in private practice and government laboratories. Those chemists/scientists would include the following, namely :
 

(a) Bundeskriminalamt of Germany¹ (The German police laboratory);

(b) Forensic Science Service of the United Kingdom²;

(c) Forensic Science Division of the Zurich Canton Police³;

(d) Dr Audrey Giles, United Kingdom⁴;

(e) Mr Robert Radley, United Kingdom⁵;

(f) The United States Secret Service.

23.3 As far as the accelerated aging method is concerned, none of the chemists/scientists in government laboratories use it and in fact, it is not even one of the methods endorsed by SOFIA itself, as at August 2001.⁶

23.4 EJS claimed that the United States Secret Service used the same techniques. However, this has been unequivocally refuted by Mr Larry Stewart and Ms Susan Fortunato.

23.5 In his letters dated 8 June 2001 and 17 September 2001,⁷ Mr Larry Stewart, the Laboratory Director of the United States Secret Service, clearly stated that the Secret Service Forensic Laboratory does not utilize the accelerated aging technique in the examination of questioned documents in actual casework because the technique has not undergone the rigours of validation/reproducibility studies and peer-review. Further, in an electronic mail dated 2 October 2001,⁸ Mr Stewart said he was of the view that there were still many questions left unanswered in relation to the reliability of the technique and remained adamant that there had as yet been insufficient research conducted to allow its use in caseworks. In a further letter dated 12 October 2001,⁹ Mr Stewart clarified that the in-house testing conducted by his laboratory and other leading laboratories showed that the accelerated aging technique was not reliable.

23.6 Indeed, Mr Stewart and Ms Fortunato expressed their concern about the reliability of the technique as early as 1996 in the aforesaid paper entitled : "Distinguishing Between Relative Ink Age Determinations and the Accelerated Aging Technique".¹⁰

23.7 Before routine utilization of the accelerated aging technique, they requested "more data regarding the number of inks tested, the types of ink formulas tested, the reasoning for the temperature, humidity and time choices as well as the limitations of the conclusions".

23.8 Similarly, the National Forensic Laboratory of the Inland Revenue Service in the States does not use the accelerated aging technique as part of their ink-dating techniques.¹¹

23.9 In a deposition in the very recent case of Murphy v. Murphy,¹² even Dr Lyter (the President of SOFIA) said he did not think that the accelerated aging test was reliable and he did not perform it.¹³ He confirmed this view again in his letter dated 23 October 2001.¹⁴ Dr Lyter did not run the Dye-ratio method in that case at all.

23.10 When EJS was cross-examined on the true "blind" (or "proficiency") test given to him by fJ Cleveland, it transpired that the Dye-ratio method was conducted only occasionally (after "a cost to benefit" analysis) and then really only as a confirmatory test.¹⁵ He explained that this was because "typically", he would not expect to obtain "positive" results with the Dye-ratio method, if no positive result is obtained from the R-ratio method and the Percent Extraction method.¹⁶
 

"A : In the universe of cases that I see, if you see a difference with dye-ratios, typically you see it with rate and percent extraction as well. And if you do not see it with rate and percent extraction, typically you do not with dye-ratios."

23.11 Notwithstanding the claim in a letter dated 10 October 2001 by Mr Brunelle¹⁷ that the accelerated aging method in conjunction with the relative aging tests using the Percent Extraction method, the R-ratio method and the Dye-ratio method have been routinely accepted by courts throughout USA and other countries since 1988, and that he routinely used these methods in caseworks, the Dye-ratio method was not defined as one of the methods of the relative aging tests in determining ink age in this case in his advice to the plaintiff contained in a letter dated 21 February 2000.¹⁸ There was certainly no recommendation for the use of the Dye-ratio method in that letter. In the penultimate paragraph at p.87, Mr Brunelle said :
 

"Relative age comparison tests used to date inks tell you more than when the ink was first commercially available. These tests can estimate when an ink was actually written on a document. This is done by comparing the solubility characteristics of questioned ink with the same properties of known dated ink. The theory behind these techniques is based on the scientifically proven premise that as inks age on paper, changes in the solubility of the ink occur. By measuring the rate of extraction or percent of extraction of the questioned ink into a weak solvent and comparing this result with those of a known dated ink, it becomes possible to determine whether both inks were written at the same or different times. When enough known dated inks are available, it is possible to estimate the time of writing within about 6 months to 1 year."

23.12 In effect, EJS and Mr Brunelle appear to be the only proponents of the use of accelerated aging in conjunction with the Dye-ratio method and no one else. On the contrary, VNA was most critical of the Dye-ratio method. In his paper entitled : "Measuring Ink Extractability as a Function of Age" published in 1998,¹⁹ he found that the Dye-ratio method would give unreproducible results.²⁰

23.13 In respect of Mr Brunelle, there was no evidence adduced to show that the conclusions, if accepted, made by him were based on results obtained from the Dye-ratio method alone.²¹ This Court cannot help but list out certain misgivings about Mr Brunelle, even he did not give evidence in this case, namely :
 

(a) In February 2000, it was stated in the affirmation of the plaintiff's solicitors that Mr Brunelle had been retained in this case and would be responsible for the analysis of data and compilation of the expert report, while EJS would be responsible for collecting the samples.²² This did not happen. The reason appears to be that Mr Brunelle has not done any cases since 1997 as contended by EJS.²³ No explanation has been offered for the change of expert in evidence. Counsel for the plaintiff in their final submission submitted that Mr Speckin would not be in a position to know why. They suggested that cost factor or cost saving factor might well be an explanation. But this cannot explain why EJS said Mr Brunelle has not done any cases since 1997.

(b) However it appears from the list of cases submitted to the court by Mr Brunelle that he had done so.²⁴ It is not clear whether Mr Brunelle had misled EJS or EJS had misled the Court when trying to explain why Mr Brunelle had not changed his methodology by requiring fewer microplugs.

(c) As aforesaid Mr Brunelle stated that in relation to the time it takes for an ink to dry, it would vary from six years to five years and then four years and back to six years in 2000.²⁵ In cross-examination, EJS rejected such statement and confirmed that there were no findings which said six years.²⁶ In one of the letters presented to the Court towards the end of EJS's evidence,²⁷ Mr Brunelle had reduced his claim to "approximately 3 years". This appears to be just a transparent attempt to bring his "evidence" or version in line with EJS who had (at least prior to his evidence in this case) been consistent in his claim that ink takes three to three and a half years to dry.

(d) Mr Brunelle chose not to offer himself as an expert and his late contentions in his letters, many of them are controversial but cannot be tested by cross-examination. Thus, the Court would not place any weight on Mr Brunelle's self-serving letters placed before the Court at the latter part of EJS's evidence.

(e) Mr Brunelle in his paper : "Ink Analysis", published in 2000,²⁸ attempted to rationalise two conflicted statements, namely :
 

"It is important to know that depending on the ink formulation involved and the paper it is on, each of the methods described above may not all have equal ability to discriminate the age of the ink being analyzed. For example, it is not uncommon for one method to detect differences in age, when one or both of the other procedures fail to detect this difference. This fact does not negate the positive results of the one method. Only if the results of one method conflict with the results of another method are the overall results negated." (Emphasis added)

(f) Mr Brunelle's claim that the courts in the United States have routinely accepted the methods²⁹ is untrue, to his knowledge. This is revealed in the newsletter "Scientific Sleuthing" where it was pointed out that as recently as 1993, Longhi J in New Jersey had rejected Mr Brunelle's evidence in somewhat scathing terms.³⁰

(g) In addition, Mr Brunelle's assertion that all advanced members of SOFIA had performed the ink-dating examinations (including accelerated aging and the Dye-ratio methods) for over 10 years³¹ is obviously untrue, at least with regard to EJS and Dr Lyter, which further reduced the number of practitioners using the methods.

23.14 In the end, this Court cannot accept that there has been acceptance by the scientific community in respect of this ink-dating method by using Dye-ratio method after accelerated aging.

ACCEPTANCE OF THE TECHNIQUES BY COURTS

23.15 It is at least an overstatement to claim that the methods that were used in this case have been in use for 15 years³² and routinely accepted by Courts in the United States and in other countries, since the Dye-ratio method was first published only in 1989 by Brunelle and Lee. Even in that paper, the authors cautioned that the method was not free of limitation and could only be used subject to certain conditions. (This statement is only true if it refers to the ink library technique.)

23.16 Of all the cases in which EJS testified and the transcripts of which are before this Court, either the Dye-ratio method was not used or the results obtained were not relied upon and accepted by the courts, let alone in conjunction with the accelerated aging method. None of the transcripts supplied by EJS, presumably to support his track record, and none of those obtained by the defendant, support his contention that the Dye-ratio method has been accepted by the Court.

23.17 In particular, in the case of Aptix Corp. and Meta Systems Inc. v. Quickturn Design Systems Inc.,³³ the Court actually rejected the results obtained by EJS using the accelerated aging method. The Dye-ratio method was not used. According to EJS, as the reliability of the accelerated aging method was in dispute, Mr Brunelle was called to testify on it. Notwithstanding that the Court found that the problem was that "the test tries to draw large conclusions from tiny differences... In all cases, most of the differences at various durations were inconclusive and at most, only a few were conclusive."³⁴ It was held as follows³⁵ :
 

"(e) Relative-Ink-Date Testing.

88. The Court will not place any reliance on the results of Expert Speckin's 'relative-ink-date testing,' also known as 'accelerated aging.' The Court recognizes that the methodology is vouched for by both Experts Speckin and Brunelle, the latter especially preeminent in the field. The problem is that the test tries to draw large conclusions from tiny differences in leach rates and to do so after artificial 'accelerated aging' (i.e., heating in an oven) of part of the test sample (so as to provide a 'known' old sample for comparison). Each was tested at different durations of leaching to detect differences in the leach rates. In all cases, most of the differences at various durations were inconclusive and, at most, only a few were conclusive. The Court agrees that the tests tend to show that the sentence on Page 4 'A programmable system ...' was added recently and is 'still drying.' It is unnecessary, however, for the Court to rely on the test and the Court does not."

23.18 EJS conceded that, with the exception of Nartron v. GM Motors³⁶ in which Mr Brunelle used the Dye-ratio method on some of the ink entries, none of the cases mentioned by him had anything to do with the Dye-ratio method. Even when the Dye-ratio method was performed in that case by Mr Brunelle³⁷ in addition to the percent extraction method and the R-ratio method (by EJS), the Court relied upon the "reverse extraction" method as the basis of its conclusion. It follows that none of the cases cited by EJS had anything to do with the acceptance of the Dye-ratio method used in conjunction with the accelerated aging method by any court.

23.19 In People of the State of California v. Genaway, Mr Kuranz formed his opinion using a combination of the relative aging tests and the accelerated aging method.³⁸ It is not clear, however whether or not the Dye-ratio method was used. Counsel for the plaintiff submitted that this case did not involve Dye-ratio method. In any event, the Court rejected the evidence of Mr Kuranz on the grounds that :
 

(a) the reliability of his technique had not been established;

(b) the technique had very limited publication and inadequate peer-review;

(c) its known or potential rate of error was not established; and

(d) it has not been generally accepted in the scientific community.

23.20 In Learning Curve Toys v. Playwood Toys Inc., the accelerated aging method was not used and the conclusion reached was based on the relative aging tests using the Percent Extraction method and the R-ratio method by Dr Lyter. The Court rejected Dr Lyter's evidence and concluded that he had failed to establish that the relative aging tests were based on methodology that had been subjected to peer-review and publication and generally accepted in the scientific community.³⁹

23.21 Here, Dr Lyter appears to be inconsistent as to whether he has used the Dye-ratio method. In a letter adduced by the plaintiff from Dr Lyter,⁴⁰ he said he "routinely" uses the Dye-ratio method to detect differences between ink entries. However, on 19 October 2001, when he testified in the case of Murphy v. Murphy,⁴¹ he confirmed⁴² that he did not use the Dye-ratio method in that case.

23.22 According to PVT,⁴³ in any event, the four apparent "users" (Mr Brunelle, EJS, Dr Lyter and Mr Kuranz) of SOFIA do not use exactly the same technique. At the meeting held on 14 August 2000, Mr Brunelle said that each of them had modified the technique, each of them were using their own version of it, and each of them had their own personal habits or personal modifications to it. This evidence was not challenged by the plaintiff.

23.23 By reason of the aforesaid matter it cannot be said at all that the methods/techniques have been accepted by the Court.

CONCLUSION

23.24 There is no evidence before this Court that the methods/techniques upon which the conclusion was drawn by EJS in this case, have been in use for 15 years and routinely accepted by courts. This Court is only presented with a technique which may have been followed in general outline by two or three other people, which may have been accepted by some U.S. Courts and which is certainly not accepted by any law enforcement agencies in the United States and all over the world. I am not satisfied that the Dye-ratio method coupled with the accelerated aging method used by EJS in this case are valid, or reliable and from which a responsible conclusion could be drawn.
 
 

¹ p.828, G3
² pp.829-831, G3
³ p.832, G3
⁴ pp.153-155, C
⁵ pp.156-160, C
⁶ pp.1447-1448, G5
⁷ pp.2090, 2096, G8
⁸ pp.2205-6, G8
⁹ p.2308, G8
¹⁰ pp.770-775, E7-E10
¹¹ Letter from Mr James Davidson of 14 June 2001 p.2091, G8
¹² pp.3154-3195, G11
¹³ p.3193, G11
¹⁴ p.1772, F7
¹⁵ T47:105:5-22
¹⁶ T47:109:7-14
¹⁷ pp.1577-1578, F6
¹⁸ pp.84-89, C
¹⁹ pp.705-721, E7-E10
²⁰ p.710, E7-E10
²¹ pp.1496-1525, F6
²² p.11, para.4, B
²³ T43:22:16-17
²⁴ pp.1521-1525, F6
²⁵ p.529, F2-3
²⁶ T44:46:9
²⁷ p.1597, F7
²⁸ pp.528-9, F2-3
²⁹ pp.1577-8, F6
³⁰ p.2303, G8
³¹ p.1578, F6
³² para.5, p.909, EJS 3
³³ pp.790-827, G3
³⁴ p.813, G3
³⁵ para.88, p.813, G3
³⁶ Decision dated 2 January 2001, pp.967-977, F5 and Submissions at pp.1464-1491, F6
³⁷ p.1483, F6
³⁸ 14 May, 1996, pp.340-343, G2
³⁹ See Memorandum Opinion and Order 31 March 2000, pp.2085-2087, G8
⁴⁰ p.1776, F7
⁴¹ pp.3154-3195, G11
⁴² p.3191, G11
⁴³ T48:129:15-130:2