‘Difference and the non-numeric system’

[Iain McKie]

We have heard a great deal at the Fingerprint Inquiry Scotland about the importance of ‘differences’ present in a mark when comparing fingerprints and for a lay person like myself this can become confusing. It certainly appears that some experts are more cavalier than others in explaining away these ‘differences’.

In respect of ‘differences’ it appears as if they can be/are explained away by dipping into a ‘lucky bag’ of distortion, slippage, double touch, photo quality and the whole range of other ‘excuses/reasons for the ‘difference’ that we have heard from the SCRO experts and their supporters over the course of the past 13 years. That many are unable to explain these ‘differences’ to a jury does not seem to be important.

Ms McBride in her Inquiry testimony does not even see this as her role.

“Juries are not qualified to decide on fingerprint identifications. It is a professional task."

Mr Macpherson states in his oral evidence.

“……but if you have 16 ridge characteristics in sequence and agreement it has to be an ident.”

What Mr Macpherson is claiming is that where you have so many features in agreement the print must be an ident. By implication all 'differences' must be capable of explanation and even if the explanation is not immediately apparent, as with Messrs Swann, Leadbetter, McBride, MacPherson and others in respect of print Y7, the identification can still stand because, ‘ fear not’, the explanation will come one day.

Other experts have skirted round this issue but it has become clear to me that if some experts can find enough characteristics in sequence and agreement they are happy to bring out their lucky bag of excuses/reasons to explain away ‘differences’.

Am I right in believing that no matter how many features you have in sequence and agreement, if you accept that you can have even one unexplained 'difference' and claim an ident then you are undermining the whole basis of uniqueness ,which as I understand it is based on the premise that, ‘no two people have ever been found to share the same fingerprint’?

The premise does not appear to state that, ‘no two people have ever been found to share a fingerprint containing 1 or more unexplained differences’.

Surely the acceptance of any ‘unexplained differences’ undermines the basis of uniqueness that underpins the science.

Given we are in the main dealing with partial latents whose surface area is often very small – Y7 is 18 mm – then this sort of logic worries me and I would appreciate some clarity in this 'difference' rule.

This Inquiry has also left me with the impression that in the hands of the incompetent/dishonest expert the non-numeric system is a haven where identification standards do not require to be fully articulated and a great deal depends on the expert’s training, integrity, the culture he/she works in and the court’s view on expert ‘infallibility’ and their willingness to challenge. In the hands of some the non-numeric system appears to have become the same old numeric system by another name.

Hugh MacPherson in his Inquiry evidence stated.

‘I think on the move to non-numeric they seem to hang their coattails on ridgeology as if because you were taking away this 16-point standard you have to have something else along with it. But, as I say, any ridgeology course I was on you were always told you're doing nothing different and to me that's the same with ACE-V. It's basically putting into words what you did.’

The Inquiry evidence suggests that in parts of the UK all that has changed is that the number of ‘points’ now needed for an identification is less than 16. While some experts seem able to ‘talk the talk’ they are unable to ‘walk the walk’ when it comes to explaining their ACE-V analysis in a coherent manner so that they can be understood by the laypersons in our courts and prosecution services.

Is the mantra, “this is the way we have always carried out fingerprint comparisons", just an excuse to allow some experts to pay little regard to changed procedures or standards and continue to do things in the same old way.

From the evidence heard at the Inquiry we are led to believe that in some areas while current trainees are receiving adequate instruction in the non numeric system, the ‘old hands’, the existing experts, appear to receive little in the way of follow up ‘non-numeric’ training.

The position of the so called independent experts like Messrs Swann, Leadbetter and Graham, who have been retired from the police for many years and whose training regime is unknown, gives me even more cause for concern.

[Charles Parker]

The one discrepancy rule has been discussed a number of times on this forum and if you go to the first page and do a search on the 'single discrepancy rule' it will bring up the postings and a number of latent print examiners views on the one discrepancy rule. I think also Professor Thorntons article on the subject is under the Holy Grail on this site.

An interesting topic and one that has a wide diversity of views in the fingerprint community.

[Pat A. Wertheim]

. . . in the hands of the incompetent/dishonest expert . . . .

The entire essense of reliability is summed up in that clause. If an expert is competent and honest, then his/her conclusions are reliable. But both competence without honesty and honesty without competence are dangerous.

Just as it is an expert's task to be both competent and honest in conducting examinations, it is the Inquiry's task to be both competent and honest in reviewing the evidence to determine what happened in the case at hand.

[Colin]

I also share IMcK’s concern re “the position of the so called independent experts like Messrs Swann, Leadbetter and Graham, who have been retired from the police for many years and whose training regime is unknown”. In private e-mail exchanges with the self-confessed “distinguished expert” Martin Leadbetter (his own description of himself to the Inquiry) I asked him about the qualifications and credentials he possessed that permitted him to intervene in important matters of Fingerprint Forensics. His response was extremely evasive, defensive and threatening. My curiosity aroused, I delved a little more into this character’s background and activities and soon concluded that I wouldn’t trust him to tell me the time of day, let alone allow him near Court of Law (unless of course he was in the dock himself).

[Atonement]

A very interesting topic.

As a complete layperson, my understanding before this Inquiry was that if you (conscientiously) found 16 characteristics in agreement between two prints then you could say with absolute confidence that the two prints had the same source, and so any "differences" which you spotted "must" be explainable in some way and should not bother you at all. If you didn't get to 16 characteristics in agreement, you wouldn't be declaring an "identification", so any differences would not be relevant anyway. However if you were identifying only for elimination purposes (or identification of a dead body) on less than 16 points, then you would have to be able to explain away any differences before you could pronounce an identification.


Under the non-numeric system, where there is no threshold of characteristics in agreement to be reached, the apparently subjective "explaining" of apparent differences seems much more important.

[Big Wullie]

Atonement

As a complete layperson, my understanding before this Inquiry was that if you (conscientiously) found 16 characteristics in agreement between two prints then you could say with absolute confidence that the two prints had the same source, and so any "differences" which you spotted "must" be explainable in some way and should not bother you at all.

This however could never explain Two Bifurcations going the wrong way in one to the other, could it ?

Nor could it explain A characteristic moving on its own, from one place to another like the Rosetta, Could it ?

Nor could it explain there being 9 ridges missing from one to the other, could it ?

Could it also explain how some experts claimed to have avoided parts of Y7 yet others found characteristics similar, they claimed ?

Could it explain how some said it was a double tap, triple tap, when others said a single tap ?

Could it explain how some said it took a 66 degree twist, while others said a 180 degree twist ?

Surely with all these differences of opinions it would be "Safe" to say there was nothing that could be relied upon within all the various opinions to identify Y7 as Shirley McKie's.

Surely it would be safe to say this: If they cannot get it right with a 16 point standard, what chance have they of getting it right when there is no required standard at-all, and on as little as 3 points we have heard maybe even none.

I would suggest the standard in Scotland should rise to 20 if they cannot get it right with 16

[David L. Grieve]

Iain,

I think you have raised issues and concerns that are troubling to me as well, and I have been just as disappointed in the responses of the "new guard" as I have in the SCRO four, plus MacKenzie and Dunbar. Many years ago, in a high profile case, the city of Chicago announced they had identified a latent print from the crime scene of a multiple homicide. We at the state lab reviewed this alleged identification and did not agree. We met to reconcile the differing conclusions and the justification given for ignoring differences was attributed to movement. Unfortunately, to explain all the absences or additions required the finger to move four different directions at once.

Yes, I believe some examiners play the distortion card to dismiss what is actual discrepancy, and do so without giving any thought to how that distortion could occur. As mentioned by others, Alice Maceo has an excellent workshop in which movement of a finger on a glass surface is recorded on video from below. The distortion to the skin from various degrees of pushing, pulling and twisting can be observed as it happens, then the resulting latent lift is examined The results of distortion are predictable, thus can be analyzed, not a matter of whim or magic that was offered as sworn testimony in the inquiry. While Alice does not have an example of 66 degree twist, a 45 degree twist does weaken the argument of Leadbetter, et al, as for the "basis" of their immaculate identification that loses 9 or 10 ridges. No accepted text on fingerprints mentions Rosetta points and one hopes thaat is not added to SPSA training.

Non-numeric systems are not without risk. This risk is minimized, never eliminated, by proper training, faithful execution of procedures and the best quality control possible. I see none of that in SPSA programs. The official position of classifying Y7 and Q12 as "unsafe identifications" is frightening. The mutation of ridgeology into the way things have always been done with new terminology is a revealing insight that the spirit of matching points while ignoring ridges is alive and well. If there is a plan to retrain those formerly with SCRO, especially those who thought Y7 and Q12 were identifications that failed to meet the 16-point requirement, I missed it as evidently you did. Nor did I see any mention of eliminating the mind set that Glasgow ruled the fingerprint world. Reading testimony that delegated valid concern about the accuracy of these identifications by other Scots as a bad case of Glasgow envy sent chills down my spine. I fully understand why courthouse staff used one word, arrogant, to describe SCRO wtinesses. If that attitude remains with SPSA in a non-numeric environment, disaster will follow.

So, Iain, your concerns are genuine and your observations are accurate Based upon testimony as to the new forensic service for Scotland, I see cosmetic change with little substantive departure from old ways. The description is full of postive buzz words but there is little meat clinging to the skeleton. Quality assurance appears to be the same tired program of adding signatures with the same bias and peer pressure. If the trainer does not understand the concepts, the future is bleak. I was once informed that a favorite saying of mine is of Scottish origin, probably because Scots have had to endure it too often. You can put lipstick and a wig on a pig, but it is still a pig.

[Gerald Clough]

As a complete layperson, my understanding before this Inquiry was that if you (conscientiously) found 16 characteristics in agreement between two prints then you could say with absolute confidence that the two prints had the same source, and so any "differences" which you spotted "must" be explainable in some way and should not bother you at all. If you didn't get to 16 characteristics in agreement, you wouldn't be declaring an "identification", so any differences would not be relevant anyway. However if you were identifying only for elimination purposes (or identification of a dead body) on less than 16 points, then you would have to be able to explain away any differences before you could pronounce an identification.

Under the non-numeric system, where there is no threshold of characteristics in agreement to be reached, the apparently subjective "explaining" of apparent differences seems much more important.

What you are seeing are some of the various ways to approach the two current realities. One being that fingerprints are very powerful identifiers, and the other being that no valid research adequately provides information on how powerfully any number of consistencies establishes identify. The difference between declaring any number, such as 16, an adequate finding to declare identification and holding that an examiner judging several variable (clarity, quantity, rarity, etc.) makes a determination of when the data is sufficient for identification in each individual case. It's not really two different ideas, just a difference in how sufficiency is established. In one way, the number is set by consensus to be high enough that any reasonable threshold, in the judgment of whatever body sets the number, has been exceeded. In the other, no such number is presumed, but the same confidence is presumed. In the view of the non-numerical assessment, a threshold high enough to be always more than sufficient will prevent many correct identifications from being accepted. In the view of the numerical standard, case-by-case assessments risk misidentifications. This difference cannot today be reconciled, because there is no valid way to test the meaning of any particular amount of agreement against reality. There are a number of reasons for this, not the least of which is the very difficult task of establishing how features and the relationships among features should be characterized so that one can ever judge is they are sufficiently alike. Studies to this end are still in the design stages, being approached as both thought experiments and as tentative numerical studies.

In practical use, there is really no difference in the handling of inconsistencies. The difference is in the direction of logic presumed by the practitioner. If one presumes that some numerical consistency absolutely implies sufficiency, then observed inconsistencies must indeed reflect something other than actual differences in skin features. The presumption of sufficiency drives the logical presumption of false inconsistencies. One then has to ask why the logic shouldn't work the other direction, why the inconsistency shouldn't deny the numerical standard. When you look at the non-numerical standard in the same way, you have to recognize that there's a similar situation. If the dynamic threshold is sufficient, should an inconsistency not have the same status as in the numerical standard? After all, but for an accident of impression, the inconsistency might not even have been revealed, and the sufficiency would never have been questioned. But the fact that sufficiency is dynamic lends more dynamic potential to perceived inconsistencies.

These are not issues that can, at least right now, be settled. Nevertheless, disposing of an unquestionably powerful identifier that, by its nature, appears among the evidence in many crimes, is not palatable. And remember that inconsistencies are in no way different from consistencies. They both require interpretation of impressed features. In a great many actual cases of crime scene latent impressions being compared to record impressions, both the consistent features and inconsistent features are subject to argued comparison. The possible mechanics of how impressions may be made, not made, or distorted are many, from very common mechanisms to very rare mechanisms. Even when the ultimate truth of the impression's source becomes known, the particulars of how that source produces some impression can only be estimated with more or less confidence. It remains, as do so many things requiring expert forensic interpretation, an opinion of certainty, the rationale for which the legal finder of fact must weigh. But neither the nature of a numerical standard nor the nature of a non-numerical standard can be blamed for a revealed error.

A lot of this is driven by the extreme difficulty in conveying the reasons and rationale for expert confidence in identifications. I believe that the reality that a highly critical lay presenter, for instance, a trial attorney, will present any qualification of an opinion as suggesting it is far less than certain tends to inflate statements of confidence. Not so much in any specific case, but in the overall tone of the profession. The confident expert believes the degree of uncertainty is being grossly over-represented. The critical lay person, on the other hand, considers that such certainty is not justified and that the expert status of the analyst lends it undue confidence before a lay jury. The expert believes, with considerable justification, that he has it right. As the science becomes more sophisticated, we will learn more. But science is not a good place to seek neat, final word, happy endings, since its purpose is ultimately to learn to ask more questions. Meanwhile, as in everything else, we will carry on do the best we can with what we have and argue the results, happy if we can simply weed out most of the bad acts and plain mistakes.

[David Fairhurst]

I think it's worth reminding the lay-people that friction ridge detail is unique down to the level of individual ridge units.
When looking at the skin itself therefore, either it all matches or it's all different!
When we look at prints, both latent and inked, the issue we have to grapple with is how well can the unique detail on the skin be determined from the detail in the print. This is, of course, where clarity comes in.
The basic ground truth still applies, however modified, as has been stated by others, either both prints were made by the same piece of skin or they were not.

I am of the opinion that the "one difference rule" applies only where the distinctiveness of the detail in agreement is insufficient to determine a common source. In these cases it is possible to exclude commonality of source using only one difference in the detail. If that difference were absent the examiner would be inconclusive as to the source of the latent.

The uniqueness of the source dictates that you simply cannot have enough detail in agreement to individualise and differences that are not caused by the imperfect reproduction of the detail in the deposition of the print (distortion). To say that you can is the epitome of illogic.

We are therefore left with the situation that if I have observed sufficient detail in agreement to individualise, then any difference I do observe must be caused by distortion.
I know that this sounds very Swannish, so before you all start shouting about it let me go further.

Firstly I must state that Y7 does not have sufficient detail in agreement with Shirley's left thumb print to determine a common source. Y7 has absolutely no detail in agreement, it's all different.

Secondly, what do I do if I have observed sufficient detail in agreement and then come across a difference?

Well, I have to account for that difference. How this is done is vital. You cannot simply say that, "I know there must be distortion because this one characteristic is different. And I can explain the presence of distortion by showing that this one characteristic is in the wrong place." That is the epitome of circular reasoning.

What you have to do is show that you noticed evidence of the distortion in your analysis, before you started comparison. Even if you were unable to perfectly explain the nature of the distortion during your analysis, the simple act of noticing a possible area of distortion and red-flagging it is enough. If, when conducting your comparison, you can determine more accurately the nature of that distortion, then that's a bonus, but there will always be those stubborn ones that, though evident, cannot be fully explained.

So, what do I do when I have sufficient detail in agreement and then I find a difference?
I look at my analysis and see if I red-flagged it. If I did them I'm fine. If not then I need to reanalyse, as objectively and honestly as I can, to see if I simply missed the red-flag or if there really was no way of seeing it.

Finally, what if I think I've found enough detail in agreement and then find a difference that was not red-flagged and is not evident upon reanalysis? Then I need to recalibrate my determination of how much detail is needed to be in agreement to determine commonality of source and, in this particular case, report an exclusion.

[Charles Parker]

David, that is very nice or to say differently beautiful and great.

Step by step practicality.

Do you mind if I steal that?

[printlady]

Big Wullie

I would suggest the standard in Scotland should rise to 20 if they cannot get it right with 16

IMHO a major problem with the 16 point standard - or any other numerical standard - is ‘what do you consider a point’. Scotland was using a point based system when the McKie case occurred, but in some of the charts I have seen from the current fingerprint hearing in Scotland some points marked don’t exist. Henry classification defines a point as an ending ridge, bifurcation or dot (Galton detail) but it looks like the charts must be marking edge shapes or other characteristics in the print (or maybe nothing at all?). Pat Wertheim talked in an earlier thread about a closeted practice of ‘teasing out the points’ which sounds like a way of reaching the 16 point threshold without actually having 16 Galton points. You could make the standard 20 (or 25 or 30) but the same problem could occur. What is the point of a numerical standard if the criteria defining the points changes at the examiner’s whim? An examiner must be able to articulate and demonstrate the clues in a print that indicate distortion is occurring not just jump to a conclusion to explain away something they don’t like in a print. There are no signs in Y7 that would indicate a 66 degree twist as has been asserted by some.

Scotland appears to have adopted the non-numeric standard without any of the theory that goes along with it. Ridgeology uses all the information in a print for comparison using level 1, 2 and 3 details. The impression is viewed in a holistic manner rather than relying on whether the examiner can count to a specific number. This system still depends on the examiner’s training, experience and integrity. You can provide additional education and instruction but if the examiner does not have integrity, no amount of training will ever correct that deficiency. Research is continuing, studies are being conducted, statistical formulas are being evaluated and computer algorithms are being developed but in the end we will always come back to the examiner’s honor and reliability.

[David Fairhurst]

Do you mind if I steal that?

Please, feel free.
Thanks for the endorsement.

[Atonement]

Speaking as a lay person, it is becoming increasingly apparent to me that fingerprint identification is not nearly as straightforward and "infallible" as I previously understood. If you had asked me a year ago (even in the knowledge of the accepted misidentifications in the McKie and Asbury cases but before hearing evidence at this Inquiry) which form of identification I "trusted" more, fingerprints or DNA, I would have said fingerprints beyond question. Yet DNA identification has a proper statistical basis for its conclusions (which fingerprint "science" doesn't have) and DNA identification is devoid of any element of subjectivity on the part of the examiner (which fingerprint identification clearly has). So much relies, in fingerprints, on the competency and integrity of the individual expert. It would be difficult to imagine a DNA practitioner setting out, on his or her own, to mislead without involving investigators, scenes of crime officers etc. How is it that fingerprinting acquired the myth of infallibility, which has never attached to DNA identification? Could it be to do with history, that DNA identification was born in a much more sceptical age?

[ER]

fingerprint identification is not nearly as straightforward and "infallible" as I previously understood

I would say that fingerprint identification is sometime very straightforward and sometimes extremely challenging. While the people working in this discipline are not always perfect, fingerprints are still the best method of personal identification that we have. They are easy to collect and are usually easy and inexpensive to compare.

Yet DNA identification has a proper statistical basis for its conclusions (which fingerprint "science" doesn't have)

The statistical basis for DNA identification lies in the structure of DNA itself. It is a linear strand of molecules, with only 4 choices at each point along the strand. This structure is perfectly made for statistical analysis. Fingerprints follow basic overall patterns, but the major ridge events (points) occur randomly. This leads to an almost infinite number of possible fingerprints and makes statistical analysis... difficult at best. There have been many statistical models developed for fingerprints, but they have all been incomplete in one way or another.

DNA identification is devoid of any element of subjectivity on the part of the examiner (which fingerprint identification clearly has)

You may be surprised that there may be some elements of subjectivity on the part of the DNA analyst as well, especially if the DNA sample is a mixture of donors. In either case there have been many discussions on this forum that the identification is not based on how the examiner may feel about the prints (subjective) but on what is actually in the prints (objective). Other posters on this forum have explained this better and in much more detail.

Fingerprint identification and DNA identification are both powerful forensic tools. They each have their strengths depending on the case and the situation. The DNA people can do amazing things with a couple dozen cells, but can't tell the difference between twins. Fingerprint identification is just as reliable even though no one has developed a complete statistical model yet. Think about this, many DNA collection forms have a place for the subject's fingerprints.

[Taggart]

On a slightly different theme but still connected I’d like to get some views on the following scenario which occurred in the Marion Ross case.

Four experts checked the print Y7. And they all ’independently’ verified the ‘identification’.

We now know they all claim they were fully aware there was some sort of movement, distortion, double touches, slippage etc etc in the upper two thirds of the print. In fact not one of the four experts could ‘identify’ the upper two thirds of the print and openly admit to finding discrepancies which they could not account for.

I want to take you through the four relevant statements and then pose a couple of questions.

Inquiry statement of Hugh Macpherson

150. I remember that I took a long time to make the identification between Y7 and DIC Shirley McKie. This was a distorted mark. It looked to me that it had been twisted and pulled down at the tip. That made the job of comparing it harder than it might otherwise have been. I would say that a mark of Y7s quality requires careful consideration in order to extract as much relevant information from it as possible.

Inquiry statement of Charles Stewart

140. Once I had reached sixteen characteristics in sequence and agreement I would have stopped my comparison. Sixteen points was the requisite standard and there was no need to go beyond that.

141. I did not keep notes. Staff were not encouraged to take notes. When I started as an expert, we did keep notes but time and resource constraints did away with the note keeping. We did not have the staff or resources for such activities.

142. I do not know how long I spent on the mark. I would have spent some time on it as it is a complex mark.

143. Once I had completed my comparison I gave the mark back to Hugh Macpherson and told him my conclusion. I did not have any discussions about the mark.

Inquiry statement of Fiona McBride

78. I do not know how long I spent on Y7 but it was a difficult mark. I spent a long time on it. During my examination Hugh Macpherson came round and asked are you finished? I said no and told him to go away.

79. Y7 was a broken mark. There was a breakage across the mark. At first I thought it was a right thumb. However then I looked at it and determined that it was Shirley McKie's left thumb. In so doing I had regard to both the plain and rolled impressions of the left thumb on Shirley McKie's tenprlnt form. In my opinion one print on the tenprint form did not show the full extent of the scene of crime mark however the rolled and plain impression did increase the available area for comparison when looked at together with the mark. I used linen glasses and not the comparator.

80. At that time I did not think that there was anything special about the exercise. Examining Y7 was just another task for me.

81. I had in the past noticed that one fingerprint expert; Kenny Graham was in the habit of putting his initials on the back of a photograph when he had identified it. I do not believe that this was a formal process. Once I had completed the identification of Y7 against the prints I asked Hugh Macpherson if I should do the same in respect of this photograph. After some thought Hugh Macpherson agreed. Hugh Macpherson tended to think deeply about most things. I placed the initials of Hugh Macpherson, Charles Stewart and myself on the back of the photograph. It was standard practice to mark up the fact that a mark had been reviewed on a comparator screen in this way and I thought that it would be a good thing to do this on photographs which hyad been examined by glass. We did not discuss the detail of my comparison.

82. I did not normally take notes during my examination. Neither do I make marks on the front of the actual photograph or prints if I am not the original identifier. One would not do that. It was not normal practice to take working notes or note up at this stage as to exactly why a conclusion was reached. I did not take any such steps in respect of mark Y7. I cannot remember exactly how many points of similarity I noted. However I suspect I did not find many over sixteen. Occasionally I would include a note when there was something special about a mark or print; e.g. if the impression was a colour reversal. However I cannot remember doing this with respect to Y7.

83. Once I had finished I handed the photograph of the mark and ten print form to Hugh Macpherson. I told him my conclusions. We did not discuss the reasons for my conclusions. He then took it to someone else for examination.

Inquiry statement of Anthony McKenna

25. I remember looking at Mark Y7 in early February 1997. Hugh Macpherson asked me to look at it. He was the team leader for another team. The office was stretched for resources and staff and it was not uncommon to be asked to check marks for other teams.

26. Hugh Macpherson did not give me any information before I looked at the mark. I did not know anything about the mark or prints before starting my comparison. Accordingly I was unaware of any involvement of Alister Geddes in Mark Y7.

27. I was not to my knowledge, involved in eliminating Shirley McKie's mark in the "baby in the bag" case. I was not aware of the link to that case when I worked on Y7 in February 1997. Police marks are eliminated day in and day out. Police often leave marks at crime scenes.

28. I became aware of the link between the two cases in 1998 or 1999. I was involved in the court reports in the "baby in the bag" case. I may have looked at the elimination, but this was not relevant to the court preparation. The focus would have been on the father's or mother's prints. I cannot remember if the case went to trial.

29. If there were marks on the comparator, (by that I mean that when an expert was carrying out a comparison on the comparator he or she would "mark up" the characteristics on the screens) made by another examiner on Mark Y7 before I compared it, I would have erased them before carrying out my comparison ..However that would have had no influence on me. I always formed my own professional view of matters.

30. I removed mark Y7 from the comparator and carried out my comparison at my desk, and when satisfied that it was an identification, I put it back on the comparator to confirm that I had sixteen characteristics in sequence and agreement.

31. I would have stopped when I found sixteen characteristics in sequence and agreement. There was no need to go beyond sixteen. All that was required for court purposes was sixteen characteristics in sequence and agreement.

32. I do not remember how long I took to do my examination of Y7.

33. I noticed the initials of the other three fingerprint officers on the back of the photograph of Y7 and I signed it as well. I signed the back of the photograph to indicate that I had compared and identified the mark.

34. I am shown a copy of a photograph of Mark Y7 with handwritten annotations, (PS_0002). The initials on the rear of the photograph refer to Hugh Macpherson, Fiona McBride and Charles Stewart. The last initials are mine. The signature of the back of photographs was a standard practice in cases I would sign the back of a photograph, especially if I had looked at it on a comparator. That way one could track who had seen it before. It was something I did for every identified mark.

35. In general eliminations would be marked in black ink, and identifications of a suspect's mark in red ink. Four signatures on the back of the mark would indicate identification.

36. After I had completed my work I handed the ten print form and mark back to Hugh Macpherson and told him my conclusion.

If these statements are accurate clearly there was little or no dialogue between the various checkers and Hugh Macpherson.

My problem is this.

If you are second checking a mark, and you realise, as in this case there are disagreements in the upper two thirds of the print and you effectively cannot identify two thirds of that print because of unexplained discrepancies would you not discuss this with the first checker?

Would you not go back and say ‘Hugh, I am able to find 16-points in the lower area but the top two thirds I cannot see?’

Do you believe that four ‘independent’ checks could be carried out, with each expert finding unexplained discrepancies in two thirds of the print and for there to be absolutely no discussion?

Two thirds, not one small part, but two thirds of the whole print.

I also find it strange that in their Inquiry statements not one single expert alludes to unexplained discrepancies in two thirds of the print. Conveniently ignored?

Just curious as I cannot believe that any four so-called competent exeprts would not discuss at any point the upper two thirds that none of them could actually identify.