Introduction
28.1. When SCRO management instructed a verification of 123 serious cases in which each, all or any of the four suspended officers had provided evidence in the year before and after the McKie case, the exercise showed the SCRO identifications to be 100% accurate. The independent examination of all cases over a 13-month period involving SCRO fingerprint evidence, announced by the Lord Advocate in the Scottish Parliament in June 2000, arrived at a similar result.1 How then did SCRO officers make erroneous identifications in these two connected instances?
28.2. Any suggestion that the officers conspired together in order to identify Y7 as being the mark of Ms McKie can be dismissed. With the exception of Mr Stewart - who had been involved in an earlier case and had no contact since then - they did not know Ms McKie and had no reason to do so. There are satisfactory explanations for the factors that led the National Training Centre to report the possibility of collusion.2 Moreover, Mr Swann, an independent expert with many years' experience, was also of the opinion that the mark was made by Ms McKie. Like the SCRO officers he had no reason wrongly to attribute it to his client Ms McKie.
28.3. As the debate during the Inquiry has shown, the lower part of the mark bears a superficial resemblance to the print of Ms McKie making human error the likely explanation.
28.4. The same can be said of QI2 Ross. The identification of this mark as that of Miss Ross on a tin found in the home of Mr Asbury was compelling evidence against Mr Asbury. Mr Graham, an independent expert with many years' experience, was instructed on behalf of Mr Asbury and agreed the identification. Although there is no evidence whatsoever that they did so, if SCRO officers were prepared to put their reputation at risk to secure a conviction one has to ask again why an independent expert for the defence would put his reputation at risk by agreeing with them.
28.5. While I am satisfied that two wrong identifications were made, I am equally satisfied that there was no conspiracy and no impropriety and that there is nothing sinister about the fact that this happened twice within a short period of time. I have no doubt that it was attributable to human error alone. The factors that contributed to the error are not unprecedented. On the contrary, reference to the reports of the INTERPOL European Expert Group on Fingerprint Identification (IEEGFI) entitled 'Method for Fingerprint Identification'3 shows that the same factors have been involved in other cases of misidentification.
Absence of alternative identifications
28.6. Other inquiries into fingerprint misidentifications, including the error made by Nottinghamshire Police in the Lee case4 and the FBI in the Mayfield case, have had the convenience of knowing the true donor of the mark in question. In addition, the OIG's investigation of the Mayfield case had the benefit of admissions by the fingerprint examiners that they had made an error.5 I do not have the consolation of alternative identifications of Y7 and QI2 Ross and the SCRO examiners, and others, remain firmly of the view that these marks were correctly identified.
28.7. In these circumstances I have limited direct insight into the causes of error. I have to infer the causes of error from my own examination of the marks, my understanding of the methodology of fingerprint comparison work and evidence about the practices of SCRO.
Overview: weaknesses in the methodology applied
28.8. Systems of work are designed to guard against error. Systems failures do not inevitably result in error but they do create the risk of error. The method of work described by the four SCRO officers displays a number of recognised risk factors and in the case of Y7 and QI2 Ross it is likely that these risks crystallised into the misidentifications.
28.9. The SCRO officers, Mr MacPherson, Mr Stewart, Mr McKenna and Ms McBride, were all in-house trained and accordingly described similar methods of working. The following descriptions concentrate on Y7 but it is understood that QI2 Ross will have been approached in the same manner.
28.10. Mr MacPherson described a process of analysis of the mark based on the practice developed when searching manually collections of fingerprints some years ago. In the interests of efficient searching, digit determination had been a priority.6 Analysis involved observation of the general pattern (whether whorl, loop or arch) and the direction of these. It also involved consideration of whether there was distortion or a double touch. He spoke of looking for as many characteristics in the mark as possible, although those would not be counted or recorded. In the course of analysis, a group of characteristics ('the target group') would become the focus of the examiner's attention. He gave the example of SCRO point 9 and what he described as an adjacent lake in Y7. As far as he could remember, those were the characteristics that first caught his eye in Y7. His practice was then to work outwards from the target group.
28.11. He would look at the mark and the print simultaneously. He would place one glass on the mark and another on the print. Although some colleagues looked from one to the other, he said that he had developed an ability to look at both at the same time.7 It is plain from his description that he was not looking simply from the mark to the print, but in some way making sense of the two visual images together. When he was satisfied as to identity, he would sometimes take the mark and the print to the comparator and mark on it his points of comparison and he thought this is what he did with Y78 and QI2.9
28.12. Mr Stewart's description of his own method of work was similar. He would begin, after checking the paperwork, by looking overall at a mark, and deciding whether there was enough detail in it to allow him to reach a conclusion. If there was he would go on to consider whether he could see a pattern, and whether he could see the core or a delta to give him a starting point for comparison. He would then look for problems such as superimposition or movement, and consider whether the mark "looked right" for the surface from which it was said to have been recovered. He would consider which digit was involved. He would then pick out an area in the mark with three or four characteristics to give him a starting point. He said he was taught to use both eyes simultaneously, which was quicker and more accurate because he was not relying on memory when comparing one with the other. By using both eyes simultaneously he was doing what was, effectively, a simultaneous comparison. He would use two glasses, with "one eye over the mark and one glass over what digit [he] was comparing in the fingerprint form." He would then look at the fingerprint form to see if he could see the group of characteristics. If he could, he would check they were in correct sequencing relationship to the other. He would go back to the mark and see where the next characteristics were in relation to the starting group, then return to the fingerprint form and see if he could find those in the same relation to the starting group. He would keep going and "build up the process" as he went through it. He did not record the characteristics observed on the mark but would use pointers to assist with counting the ridges intervening between characteristics.10
28.13. The record of the findings on the image of QI2 contains the words "on screen" in relation to both Miss Ross and Mr Asbury. Mr Stewart said that those words meant that the mark had been placed on the comparator for viewing and did not mean necessarily that he carried out any part of his comparison on the machine. He said that he would have carried out a comparison using glasses and doubted that he used the comparator. He conceded that if there were markings on the screen when he first picked up any mark he would have a quick look at them before carrying out his own comparison but he could not recall if there were markings on screen for QI211 .
28.14. Mr McKenna's recollection12 was that he picked up Y7 from the comparator machine and took it back to his desk to check it before returning it to the comparator screen to check that he had 16 points. He could not recall if there were markings on the comparator screen other than signatures when he first collected the mark and print. At his desk he proceeded straight to examination of the mark and print together and gave no account of a separate preliminary analysis of the mark. His normal way of working was to look at the mark and print at the same time, having one eye on each. Use of target groups was something that he described as natural, though his own normal method of carrying out any comparison was to scan the mark and print for some characteristic from which to start and that could be either on the mark or on the print. He would then look for other characteristics in sequence and agreement working between the two images. He kept a mental count as he went along, stopping once he had 16,13 and he would go to the comparator to check that he had that number.14 He accepted that it could happen that he found less than 16 points very easily working backward and forward between mark and print and would then have to work a lot harder to find the points that took him up to 16, but said that he would not put himself under pressure to get 16: "If I could not find 16; I could not find 16."15
28.15. Mr McKenna was not among those who first identified any part of QI2. He would have looked at that mark for the first time either when he signed the case envelope or when the joint report was being prepared. He said that he would have used glass and possibly also the comparator machine.16
28.16. Ms McBride likewise was not among those who first identified QI2. As for Y7, she was understandably uncertain given the passage of time whether she selected a target group for her comparison. It would seem that it was not her invariable practice to begin with a target group (a term that she did not use)17 because the pattern could in some instances be bland. She may have chosen a group of characteristics to start with, but it would have been far easier to start at the core and work out, and she thought it may be more likely that she did so. She could not recall with what features she began when working with Y7.18
28.17. She did, though, give a clear account of her general method of working.19 It was her practice to check the whole mark and thereafter she carried out a comparison in the manner described by Mr Stewart in his evidence, which she termed a "binocular comparison", carried out simultaneously with one linen glass over the print and the other over the mark. Working in this manner she would form pictures in her head of mark and print and would focus her mind, rather than her eyes, when using pointers to check ridge counts. She did not at this stage count the number of matching ridge characteristics but would do so only after she had reached what counsel described as "a personal point of certainty": that is to say, having checked and double-checked and being certain "beyond reasonable doubt or beyond your own doubt" that there was a match.20
28.18. Fingerprint examiners now practise in accordance with the acronym ACE-V and though that acronym was not current at the time when Y7 and QI2 Ross were being considered, the SCRO examiners said that it reflected their approach to their work. Each of the initials ought to21 spell out a distinct stage in the methodology but it is recognised that there is a tendency for the stages to merge, with analysis (A) often overlapping with comparison (C). Without diminishing any of these stages, because practices do vary, greater weight may rest on proper evaluation (E) and verification (V). Fingerprint work is an exercise that calls for the application of subjective judgment and there is a need for careful evaluation by each examiner of the assumptions made in the course of an examination. Finally, the separate requirement for verification is intended to ensure that the reliability of the conclusion is demonstrated by the concurrent findings of a number of individuals (a total of four in SCRO at that time) acting independently.
28.19. No specific criticism falls to be levelled at the fingerprint examiners for not following the ACE-V protocol in 1997. It was not then in common currency in the United Kingdom. That said, the acronym was understood to have been a reflection of common sense and hence a reflection of the practice that was followed at that time. Mr MacPherson's comment is perhaps typical: "ACE-V was just verbalising what you did."22 However, it is clear from the summary above that the SCRO officers did not subject themselves to the disciplines of the separate stages of a literal application of the ACE-V methodology.
28.20. The analysis stage of the process for them involved such matters as identification of pattern type, identification of areas which appeared in some way distorted or unclear, an attempt at digit determination, and the identification of a target group of features which the examiner would then use in the initial stages of determining whether there was a match between the mark and a particular print.
28.21. Other examiners external to SCRO also spoke of the use of a target group of points but there was care as to the stage in the process at which this was done. Mr Chamberlain said that it came after the initial analysis of the whole mark and as a logical start to the comparison stage. As he put it, it is possible to put an 'S' between the 'A' and 'C' of ACE-V to signify that the comparison begins with a 'search' area.23 Mr Grigg also said that examiners will start any comparison with a set of strong features, or 'target group'. As he explained, the analysis stage requires that all features of the mark (and not just the target group) are analysed before comparison24 .
28.22. Although SCRO examiners spoke of looking at the whole mark, there was no systematic effort to assess in its entirety the volume and quality of information available in the mark before moving on to consider the known prints. There was no systematic assessment or noting of the number of characteristics visible, or whether, from the mark alone, the examiner had formed a view that a characteristic was, for example, a bifurcation or a ridge ending as distinct from an 'event' which might be either of the two, or even an artefact on the mark or the image of the mark. There was no systematic assessment or noting of the degree of confidence of the examiner in any view he had formed as to the nature of a particular characteristic.
28.23. The need for comprehensive analysis of the mark before the selection of any target group has relevance to reverse reasoning: that is to say, using clearer detail in the known print to elucidate detail in the mark. The majority view of the experts who gave evidence to the Inquiry was that reverse reasoning is permissible to some extent but needs careful consideration at the evaluation stage. Mr Zeelenberg's description of the evaluation stage of ACE-V is a guard against the premature formulation of an 'inner conviction' and a deliberate strategy to postpone the point at which the examiner reaches a conclusion as to identity; one built into the process in the awareness that it can be difficult for an examiner to alter his view once he has reached a conclusion as to identity. It is a separate stage in which the examiner balances the pros and cons regarding the identification.25 Mr MacPherson did not recognise a separate 'evaluation' stage as is in the ACE-V process26 and, as will be discussed, he did speak of an inner conviction formed part way through, and not at the end of, the process.
Detailed consideration
28.24. In building up a picture of the causes of misidentification it is necessary to give more detailed consideration to the specific risk factors that were present in the working environment and practices prevalent in SCRO in 1997-1999.
28.25. The factors to be addressed are:
(i) practitioners being taught 100% certainty, which could be attained prematurely in the examination process on the basis of relatively few matching characteristics;
(ii) the ethos in the SCRO fingerprint bureau where pride was taken in an ability, particularly on the part of more experienced officers, to identify marks that other bureaux might not consider sufficient for identification;
(iii) an inappropriate hierarchical philosophy;
(iv) the application of inappropriate tolerances in the observation and interpretation of detail in marks and prints;
(v) reverse reasoning;
(vi) the influence of repeated viewing of known prints;
(vii) concentration on target areas;
(viii) circular reasoning in relation to the 16-point standard and the explanation of differences and generalised thinking;
(ix) slack working practices such as the inappropriate use of the comparator machine;
(x) a failure to address the distinction between eliminations and identifications; and
(xi) contextual information from the police.
100% certainty
28.26. Fingerprint examiners are taught that once they have reached a conclusion they can be 100% certain in their own mind that the identification is correct and, moreover, that they can expect any other examiner of similar training and experience to reach the same conclusion.
28.27. It is critical to note that even in the days of the 16-point standard the tipping point of 100% certainty would in all probability arrive before the examiner had found 16 points in sequence and agreement. Evett and Williams explained the 16-point standard as a legal practice only and, having observed that fingerprint examiners reached their conclusion on a lesser number of points, cautioned that "It is important to understand that a fingerprint expert regards any identification as a certainty."27 Mr MacPherson's evidence was that when he got a 'sufficiency' of characteristics in sequence and agreement he knew within himself that the mark and the print had been made by one and the same person.28 Counsel referred to that variously as an 'inner conviction'29 or, in a question to Ms McBride, 'a personal point of certainty'.30 As will be discussed later, that inner conviction led Mr MacPherson to a circular argument discounting differences as matters that must be capable of explanation without knowing what the precise explanation was and the same process of reasoning would inevitably tend to influence an examiner to interpret ambiguous or obscure detail in a manner consistent with positive identification.
28.28. The prevalent belief in 100% certainty may also have diminished the independence of the verification stage of the process because a verifying examiner might tend towards confirming the view of the first examiner, particularly if the first examiner was a senior one, as was the case for Y7 and QI2 Ross where Mr MacPherson was the first examiner. This risk of sub-conscious bias is related to the tendency, noted below, to have undue confidence in the ability of senior SCRO examiners to interpret difficult or complex marks.
28.29. It may also be observed that the belief in 100% certainty was bound to reduce the utility of asking any particular fingerprint examiner to re-consider his or her own findings. Some of the SCRO examiners involved with Y7 personally re-examined the mark on a number of occasions by reference to a variety of source materials but given that each had already reached a conclusion with 100% certainty it is no surprise that each individual's re-examination led to confirmation of the initial conclusion.
The ethos in the SCRO fingerprint bureau in 1997
28.30. The evidence regarding the reputation of SCRO is summarised in chapter 21.
28.31. The HMICS report highlighted what it called "an 'internalised' culture"31 and further insight into this was provided by Mr Luckraft. Though he worked at SCRO between 2000 and 2001 when SCRO was 'under attack' and therefore on the defensive, his evidence was descriptive of an atmosphere that is likely to have existed before then.
28.32. While the SCRO officers took pride in the reputation of the bureau as amongst the best (if not the best) in the world, Mr Luckraft perceived arrogance in this approach. He pointed to a 'culture' at SCRO which was in his view 'insular' and "not willing to listen to other ideas",32 accompanied, in particular, by a belief that senior examiners such as Mr MacPherson and Mr Mackenzie were incapable of making a mistake. He inferred that this ethos could produce peer pressure that could "possibly result in a bias towards fingerprint identifications."33
28.33. An ethos in which officers pride themselves in their work is not necessarily a matter of criticism. The fact that in over 1700 cases identifications made by SCRO examiners were confirmed by other bureaux34 shows that there was some justification for the confidence that the SCRO examiners had in their own ability. However, the belief among the SCRO examiners that they could identify marks that others would regard as unsuitable for comparison carried the risk that with difficult and complex marks, such as Y7 and QI2 Ross, they would be tempted to make identifications where the quality and volume of information did not properly support identification.
28.34. Mr Luckraft highlighted one particular aspect of practice within SCRO that was a matter of particular concern to him and that was "a culture of being able to 'push' a comparison to 'tease' 16 points in agreement."35 He observed:
"The problem with working to 16 points all the time means that your experience becomes limited to pushing and working to 16 points only. Therefore, when you come across a scene of crime mark of poor quality, there will be a tendency to attempt to push and make 16 points, even when there are considerably less points contained within the mark. In my opinion, I believe this type of practice could explain how and why a mistaken identification could occur."36
28.35. Evett and Williams, reporting in 1989, found 'teasing out' of points to be prevalent in the UK as a consequence of adherence to the 16-point standard.37 As they emphasised, it has to be recalled that fingerprint examiners would be personally convinced to the level of 100% certainty probably on substantially fewer than 16 points (perhaps eight) and in that situation it would be understandable if there was a sub-conscious recourse to reverse reasoning, using the known print to guide the interpretation of the mark to complete the 16 points required by the legal standard.38 Significantly, the assessment of Evett and Williams was that this practice posed no particular risk because they were assuming that the examiner was correctly satisfied on the lesser number of points that gave rise to the 'inner conviction' of identity. Mr Zeelenberg's evidence was that experience since Evett and Williams, particularly with the Mayfield case in the United States, had shown that teasing out could be a source of error.39 A tendency to push to 16 points would be one explanation for SCRO being able to report an identification where other bureaux would not; and it would be a particular risk factor in relation to complex marks, such as Y7 and QI2 Ross, where the existence of a lesser number of matching points to form a correct 'inner conviction' was itself open to debate.
28.36. There was confirmatory evidence of a working environment within SCRO to 'push' to find 16 points and the conclusion that that will have contributed to the erroneous identifications here can be supported by the candid evidence of Mr Graham. He agreed with SCRO that Y7 was the mark of Ms McKie but he reached that conclusion perhaps on as few as seven points: "According to Mr Graham, where SCRO went wrong was in then attempting to find 16 points in common."40
28.37. Ms McBride denied 'teasing' but it is significant to observe that she would reach a conclusion before she counted the points and, if she did not find 16 points, she would double check:
"...it is not that I am looking to tease out points as one of the phrases that have been used or whatever, it is just a case of being certain and making sure before you pass it on and say, 'I am not able to get 16'."41
Double checking could include looking at other impressions of the print on different ten-print forms "to see if it was possible [to get 16 points] but if it is not possible, it is not possible." The risk is that, already being certain of identity, any examiner carrying out a double check could interpret ambiguous detail, such as SCRO points 4-6 in Y7,42 in a manner consistent with completing the list of 16.
An inappropriate hierarchical philosophy
28.38. The Inquiry heard evidence of a peer pressure within SCRO attributable to an inappropriate hierarchical philosophy within the bureau and that is summarised in chapter 23. That examiners could be immune to both peer pressure and the sub-conscious influence of a culture of a 'push' to 16 points can be clearly exemplified by Mr Geddes who adhered to his own opinion that he could see only ten points in sequence and agreement in Y7 despite the demonstration by Mr MacPherson of the 16 points that he observed. The consequential issue is why the application of a rigorous verification process did not stop the identification at that stage.
28.39. There are two answers to that.
28.40. The first is that practice within SCRO at that time permitted an examiner like Mr Geddes, who agreed that there was an identity between mark and print but could not find 16 matching characteristics, to be effectively by-passed in favour of verification by other examiners.
28.41. The second answer is more subtle. The by-passing of Mr Geddes was a lost opportunity for each of the four examiners Mr MacPherson, Mr Stewart, Ms McBride and Mr McKenna properly to reflect on their evaluation of the mark by giving due consideration to the view of Mr Geddes. In part at least, the failure to take advantage of that opportunity may itself have been attributable to the ethos within SCRO. Mr Luckraft reported a perception that more years' experience equalled higher skill level43 and that attitude was displayed by Mr Stewart both in his oral evidence to the Inquiry44 and in his witness statement, this quotation coming from the latter:
"Likewise the difference may simply reflect fingerprint experts' own skills. Different fingerprint experts have different tolerances to different marks. As an example a more experienced individual will normally be able to work better with more difficult marks. Some experts could have a particular aptitude for analysing certain types of marks, for example marks made in blood."45
28.42. The belief was not only that SCRO could identify marks that other bureaux could not, but also within SCRO itself there was a belief that the interpretative skills of an examiner grew with experience and hence Mr MacPherson's certainty in his own conclusion is unlikely to have been diminished by the failure of the more junior Mr Geddes to agree his entire list of 16 points. The same may also be said of the other three signatories, each of whom was senior to Mr Geddes.
28.43. The application of different tolerances, whether as a product of years of experience or personal aptitude, leads to the next risk factor.
Tolerances
28.44. Fingerprint examiners are working with impressions (both in the case of known prints and crime scene marks) that may be more or less partial reproductions of the fingerprint or fingerprints and even such detail as is reproduced may be distorted due to the manner of deposition or development. Their work requires them to make assumptions in order to compensate for the incomplete and distorted state of the materials. If the detail is clear the compensating assumptions will be narrowly confined with the result that a more exacting correspondence in detail will be required before an examiner will declare an identification. Where, as in a complex mark like Y7 or QI2, the detail reproduced in the mark lacks clarity there is scope for the examiner to apply wider tolerances in the assumptions that he makes. This carries the risk of an adventitious match because the width of the tolerances may in fact be accommodating genuine points of difference.
28.45. Before reaching a conclusion as to identity of a complex mark an examiner must carefully evaluate the tolerances applied. An organisation which takes particular pride in working with marks which other organisations might not be 'able' to use risks failing properly to evaluate the tolerances that have been applied and there is the added danger that this factor will be compounded by 100% certainty. If an examiner has reached an inner conviction of identity based on wide tolerances applied to a limited number of points, there is the risk that he may apply ever expanding tolerances in the search for the remaining points to complete his list of 16.
28.46. One outward sign of the application of inappropriately wide tolerances by SCRO is the conflicting evidence by Mr MacPherson and Mr Mackenzie regarding SCRO points 11 and 12 in QI2 Ross. Both cited these as points in sequence and agreement but to Mr MacPherson these were a ridge ending (SCRO 11) and a bifurcation (SCRO 12) producing the shape of a bell, whereas Mr Mackenzie saw two bifurcations meeting as a lake.46 These mutually exclusive interpretations are indicative of the application of wide tolerances not only to the mark but also to the corresponding section of the print of Miss Ross. Mr Zeelenberg's slide 16247 shows the upper ridge of the bifurcation SCRO 12 in the mark coming into contact with a vertical swipe and it is difficult to confirm the flow of the ridge thereafter in the mark, leaving scope for an inference of either a ridge ending (Mr MacPherson) or a bifurcation (Mr Mackenzie) at the intersection. Turning to the print, there is the appearance of a connection between the two opposing ridges at SCRO 11 forming the bifurcation as seen by Mr Mackenzie. Mr MacPherson's interpretation of the section of ridge that appears to form that connection was that it was no more than an incipient (a third level detail) and hence, for him, the print properly interpreted showed a ridge ending.48 The result is that, in order to complete a matching bifurcation, one of them (Mr Mackenzie) has had to infer the presence of a section of ridge detail in the mark where such detail was not readily observable; while the other (Mr MacPherson) has had to discount, as an incipient, detail that can be seen in the print in order to conclude that there are matching ridge endings. The same can be said of SCRO points 1, 10 and 16 in QI2 Ross.49
28.47. As for ever expanding tolerances being applied, Y7 serves as an example. In the lower part of the mark (from SCRO 1-9) there is a series of characteristics the majority of which are readily observable and do bear some degree of similarity depending on the degree of tolerance applied but, having concluded that those details do match, the SCRO officers must have applied wider tolerances to detail such as SCRO 10-13 because the ridge characteristics in those locations are not readily observable.
28.48. The fundamental problem with the application of wider tolerances lies in a paradox identified in a passage in the report by the INTERPOL European Expert Group on Fingerprint Identification 'Method for Fingerprint Identification'50
"A difference in appearance between compared fingerprints (or details of them) that is [attributable]51 to normal variations with printing can be tolerated. Tolerances should be applied consistently and honestly. Experts should be aware of the paradox that one may be inclined to accept more differences in bad prints under the umbrella of distortion than one would accept in better quality prints. Distortion not only limits the perception of the similar but also of the dissimilar. The pitfall is that a premature assumption of donorship leads to transplantation of data from the 'original' into the blur of the latent. It is circular reasoning like: this print comes from this donor, prints are unique, thus all data must be the same and subsequently all differences are not real. With identifications proved to be mistaken, it became clear that the involved experts had ignored the differences. Evaluation of those comparisons often contains a long list of excuses why the print does not look like how it should, disguised as demonstration of the skill and experience of the expert. The rule is therefore that: 'Tolerances should not vary dependent on the quality of the impression'." (emphasis added)
28.49. Recognition that the application of wide tolerances can lead to potentially exclusionary differences being treated as possible points of similarity ought to result in greater experience counselling greater caution with marks that lack clarity. However, that is likely to have weighed less with practitioners who had a belief in a personal and collective ability to identify marks that others of less experience could not.
Reverse reasoning
28.50. To some extent this is an alternative perspective on the application of undue tolerances. The application of wide tolerances can be associated with reverse reasoning because both can be responses to lack of clarity in the mark: an examiner being able (because of the application of wide tolerances) to interpret poor detail in the mark in a manner consistent with the clearer detail in the print (reverse reasoning).
28.51. 'Teasing out' can be a form of reverse reasoning and that was the descriptive term applied by Mr Zeelenberg.
28.52. In QI2 Ross he referred to SCRO points 4 and 11.52 SCRO point 11 has already been mentioned in the context of the tolerances that were applied and because the detail is ambiguous in each of mark and print it may be open to question whether the interpretation of the print has guided the interpretation of the mark or vice versa.
28.53. Y7 may provide clearer examples in SCRO points 10-13.53 In the print these appear as a lake above the core (points 10 and 11) and an adjacent incipient ridge or island (points 12 and 13). There is little corresponding detail in the mark and it is difficult to assume that these four characteristics would have been picked out in an unguided initial analysis of the mark.
Repeated viewing
28.54. A more subtle form of reverse reasoning may also have been present in the case of QI2 Ross.
28.55. In the murder investigation it was a logical and necessary step to consider first which of the many marks at the crime scene were those of Miss Ross, so that those marks could be eliminated from the investigation. It followed that Mr MacPherson had to view Miss Ross's prints repeatedly on the many occasions he required to compare them with marks found at the crime scene. In that situation he would come to "memorise" the prints.54
28.56. Mr Stewart also gave evidence to the effect that in investigations where prints or marks were viewed repeatedly, it was possible for examiners to memorise where particular clusters of features were: "It gets to the stage that if you are working on some of these major inquiries for the lengthy periods they run you are effectively picking up the form, it is almost an aide-memoire to you. The same by looking at the marks, you get to see the marks so often you already can remember where the cluster is you are going to start looking at the target group, whatever you want to call it."55
28.57. Mr MacPherson did not see that there was any danger that the visual memory of the prints would come to influence him. There must, however, be a risk that the process of reasoning from mark to print, which is intended to prevent examiners from forming a view about the mark by reference to the known print, will be undermined in a situation where an examiner has formed a strong visual familiarity with and memory of the print.
Target areas
28.58. A proper analysis, comparison and evaluation ought to consider mark and print as a whole. Mr MacPherson, Mr Stewart and Mr McKenna all worked with target groups, and the evidence of Mr Chamberlain and Mr Grigg supports Mr McKenna's statement that this is a natural approach to adopt in a comparison. Still, Mr Chamberlain's and Mr Grigg's evidence emphasised the need for a comprehensive analysis of the mark before the commencement of a comparison based on a selected target group followed by a proper evaluation before a conclusion is reached. The risk associated with a conclusion based on target areas is that the comparison may be confined to that area and will not take into account a proper consideration of the remainder of the detail in mark and print. Alternatively, a preliminary conclusion formed by reference to a target area could give rise to the risk of reverse reasoning when comparing the remaining detail.
28.59. Reverse reasoning has been discussed. As for the risk that the conclusion will be based on an incomplete analysis, there is the fact that all four of the SCRO officers discounted the top part of Y7. To take Mr MacPherson as an example, when discussing the significance of unexplained differences, he drew a distinction between differences that were in his target area and differences lying elsewhere. He said that he could still reach a conclusion of identity if there were unexplained differences outside his target area but not if there was one in that area56 and that attitude in part explains why he attached no significance to the "one or two characteristics which were not in alignment" in the top part of the mark.57 There is no proper justification for discriminating between an arbitrarily chosen target area and the remainder of the mark when it comes to the proper evaluation of a potentially exclusionary difference.
28.60. Concentration on a target area might be thought to have some basis in the basic premise that the pattern of fingerprints is unique not just in the fingerprint as a whole but also to small areas.58 That is a premise which, if correct, applies to the pattern in the skin. Care has to be taken when transposing that premise to an impression, whether a mark or a print.59 That premise cannot justify a failure to analyse the whole mark or to fail to evaluate all the points observed during comparison of mark and print. Apart from anything else, the presence of an unexplained difference in one part of a mark ought to cause an examiner to reconsider whether the characteristics that are thought to be similar in another part of the mark truly are in agreement.60 Taking Y7 as an example, the differences in the upper section of the mark ought to have led the SCRO examiners to consider whether it was appropriate to construe ambiguous detail such as SCRO points 4-6 as matching characteristics.61
Circular reasoning on points of difference where 16 points found and generalised thinking
28.61. Given the propensity for the precise appearance of ridges to alter depending on the manner of deposition even in prints taken under controlled circumstances, it is not realistic to expect that the detail in two prints made by the same individual will coincide exactly. In fingerprint comparison allowance has to be made for inevitable variations in detail and this leads to a distinction between what Professor Champod called 'within source' variations (i.e. the type of variation that can occur when the same individual deposits his print) and 'between source' variations (i.e. any difference indicative of the fact that the prints come from two different individuals). The presence of some differences between mark and print would not exclude a conclusion of identity if the difference can be explained as a within source variation but the presence of an unexplained difference ought to exclude that conclusion.
28.62. The skill of the fingerprint examiner is to ascertain whether any difference is capable of reliable explanation.
28.63. When the 16-point standard prevailed there was a perception that the existence of such a number of points in sequence and agreement itself was sufficient to prove identity beyond any reasonable doubt, and that led to circular reasoning that if such a number of common points was present a difference could be discounted because it could be assumed to be capable of explanation, even if the particular explanation was not immediately apparent.
28.64. Mr MacPherson subscribed to that view: if there were 16 points in sequence and agreement it would be impossible to have an unexplained difference.62 That was a view shared by others outside SCRO including Mr Graham63 and Mr Leadbetter64 and it was reflected in contemporaneous literature. Mr MacPherson referred in his evidence to a 1970 handbook in which proof of identity was taken if 16 points were found in sequence and agreement and the formulation of the 16-point rule in that text made no reference to unexplained differences.65 The same is true of the Crown Office Expert Evidence Manual.66
28.65. The application of such reasoning runs the risk of discounting a true between source variation such as the Rosetta in Y7. At a more fundamental level it also assumes that the examiner has correctly identified 16 points that are truly in sequence and agreement. At the evaluation stage consideration ought to be given to all of the detail in mark and print before a conclusion is reached. If tolerances have to be applied to accommodate differences in detail such as those in the formation of SCRO points 4-6 in Y7, does the application of a similar degree of tolerance account for other observed differences, such as the Rosetta? If not, is it reliable to apply even that degree of tolerance to SCRO 4-6 or is the more correct conclusion that those points cannot be taken to be in sequence and agreement? That was the process of reasoning of Mr Grigg in relation to Y7.67 By assuming an explanation for differences from the mere fact that 16 points are believed to be in sequence and agreement, examiners deprived themselves of the opportunity properly to evaluate the tolerances being applied in deeming the 16 to be truly in agreement.
28.66. Not all of the SCRO examiners subscribed to that specific circular reasoning but there was a related factor also in play. SCRO fingerprint officers were trained to talk in generalities68 with explanations for movement and the like passed on by word of mouth from mentor to student.69 The inability of SCRO officers to advance a satisfactory reason for discounting the top part of Y7 was remarked upon by Lord Johnston in his charge to the jury in HMA v McKie.70 Merely to say that there has been movement of the top part of the mark Y7 relative to the lower part is not sufficient to explain the points of difference in the upper part. It is necessary to consider what particular pattern of movement may have given rise to the distribution of characteristics in the mark and to proceed from there to ask how likely it is that that pattern of movement may have occurred. Take, for example, Mr Mackenzie's assumed distribution of matching characteristics in figure 13 in chapter 25:
(i) What pattern of movement of the finger could have produced such a 'leapfrogging' redistribution of characteristics?
(ii) How likely is it that such a pattern of movement could have occurred?
(iii) What physical evidence (such as smearing and criss-crossing of ridges) might one expect to see in the mark had that pattern of movement occurred?
(iv) Is any of that physical evidence present in the mark?
28.67. Thinking in terms of generalities deprived the SCRO examiners of the opportunity properly to evaluate all of the detail in the mark and to consider properly whether they were overstepping the limits of tolerance in deeming Y7 and QI2 to match the prints of Ms McKie and Miss Ross.
Slack working practices such as the inappropriate use of the comparator machine
28.68. Evidence relating to the use of the comparator machine in SCRO in 1997 is summarised in chapter 23. Two separate matters require consideration. The first is the practice whereby some examiners carried out comparisons on the comparator machine alone. The second relates to markings being left on screen.
28.69. That it was acceptable practice in SCRO to use the comparator to carry out comparisons is demonstrated by the choice of that apparatus by Mr Dunbar for the so-called 'blind testing' exercise. In that exercise he gave no opportunity to the participants to use linen glasses.71
28.70. Examiners sometimes carried out verifications using only the comparator machine. The evidence from the four signatories to the joint report is that they would have examined both Y7 and QI2 Ross under glass. Ms McBride's account of having done so is reinforced by the contemporaneous note on the reverse of the photograph of Y7, recording her use of glasses.72 That evidence is accepted. Nonetheless, it may have been the case that Mr Bruce's verification of QI2 Ross was based solely on work on the comparator machine.73 The problem is that the comparator screen may not display the complete mark and print and though an examiner can move the images to carry out a complete comparison the temptation may be to compare only the parts first displayed and not to look at the complete mark.
28.71. As for the practice of leaving markings on screen, the presence of such markings could be a form of 'peer pressure' (as Mr Luckraft characterised it)74 in the sense of being a source of sub-conscious influence on later examiners, including those who removed the mark from the machine and examined it under glass. That practice potentially compromised the full independence of the verification process as the report on the misidentification by Nottinghamshire Police in the Lee case stated.75
28.72. That practice is of no relevance to Ms McBride and Mr McKenna's conclusions on QI2 because neither was among the initial verifiers of that mark. It may, though, have had some impact on the verifiers, including Mr Stewart, who admitted that, if there were markings on screen, he would take a quick look at them to get an indication of where points had been found before removing the material from the comparator machine and carrying out his examination at his desk.76 As for the possibility that there were markings there to be seen, allowance has to be made for the fact that Mr MacPherson said that he tended to leave two or three characteristics on screen to give others a start, which suggests that at least some markings for QI2 would have been there to be seen.
28.73. The same inferences arise in relation to Y7 where Mr MacPherson was again the first examiner and there is the possibility that Mr McKenna and Mr Stewart (but probably not Ms McBride) would have seen his markings in relation to that mark.
A failure to address the distinction between eliminations and identifications
28.74. The distinction between an 'elimination' and an 'identification' is explained in chapter 32.77
28.75. SCRO, like English bureaux,78 could make an 'elimination' without finding 16 points in sequence and agreement. There were variations among the witnesses as to the lower limit necessary for an 'elimination', ranging from 8-12 points, but the more pertinent fact is that, from the perspective of fingerprint examiners, this did not imply the application of a lower standard because they regarded both an 'elimination' and an 'identification' as a certainty: see Evett and Williams.79 The distinction between the two was of relevance to lawyers, not fingerprint practitioners.
28.76. In February 1997, when Y7 was initially studied at SCRO, Ms McKie fell into the category of those who could be 'eliminated': she was a police officer engaged in the murder investigation. While Mr MacPherson required examiners involved in the ordinary process of examination of marks from the house to work to the 16-point standard, it is clear that, with the exception of Mr Halliday, those involved at the stage of the exceptional additional checks on Y7 (i.e. Mr Mackenzie, Mr Dunbar and the examiners involved in the 'blind test') considered it only from the standpoint of an 'elimination'.80 The Inquiry has established that the examiners were equally divided as between an 'identification' and an 'elimination'. Five found the 16 points necessary for an 'identification' (the four who signed the reports plus Mr Halliday) and the same number (Mr Geddes, Mr Mackenzie, Mr Dunbar, Mr Bruce and Mr Foley) verified an 'elimination' on a lesser number of points as low as eight.81
28.77. The problem is that an individual may be a candidate for 'elimination' one day as a non-suspect and turn out later to require consideration for proof of identity to the full legal standard. That was critical in the case of Ms McKie. Application of different standards can lead to confusion, particularly if (as happened here), there is no record that some have only 'eliminated' on less than 16 points.
28.78. Ms Climie and Mrs Greaves both proceeded under the misapprehension that there was unanimity among the examiners that there were 16 points in sequence and agreement.82
28.79. Within SCRO itself no single individual was aware of the full detail of the various comparisons;83 and that this gave rise to misapprehension on the part of Ms McBride and Mr Stewart can be seen from the evidence that they gave at the trial in HMA v McKie. In the course of cross-examination Ms McBride made the claim that the system of fingerprinting was infallible84 and that was explored in this question and answer:
"Have I got that right; it is infallible? - So far, yes. I would say the system is infallible. One person can make a mistake but four people and the quality assurance officer - it is not possible¿"85
"So far" refers to the fact that there had been to that date no successful challenge to a SCRO fingerprint identification but the point that is relevant for present purposes is that Ms McBride believed that the conclusion reached by the four signatories to the joint reports had been supported by the quality assurance officer, Mr Dunbar. Earlier in the trial Mr Stewart had testified that the mark had been checked not only by the quality assurance officer but also by the deputy head of bureau, Mr Mackenzie.86 There is no reason to think that Ms McBride and Mr Stewart knew then what the Inquiry has since established, which is that at the date of the trial Mr Mackenzie and Mr Dunbar concurred only to the extent of agreeing an 'elimination' on less than 16 points.
28.80. Mr Graham (who eliminated the mark to Ms McKie possibly on as few as seven common points) believed that SCRO made a mistake in proposing that an identification could be established to the 16-point standard.87 The fact that a number of officers, including the most senior fingerprint examiner in the bureau and the quality assurance officer, had not found at that stage 16 points in agreement, ought to have presented an opportunity to SCRO itself to reflect on the robustness of the opinion that proof of identity could be established to the full legal standard. Disclosure of that fact to COPFS would also have given the prosecution the opportunity to reflect on that matter. That neither was afforded that opportunity is attributable to the lack of proper recording within SCRO and a consequent failure to address the implications of the distinction between an 'elimination' and an 'identification'. The fuller information available to the Inquiry supports the conclusion that the system was far from infallible.
Contextual information from the police
28.81. The police provided contextual information to the SCRO fingerprint examiners. This is plain from the note88 recovered from the case envelope,89 which read: "They are hopeful about the tin the money was in. There is an area the same size as the tin, clearly seen with dust round it. Tin recovered at accused's." This was a note taken by Mr Stewart.90
28.82. Mr MacNeil's note on Form 13B when sending the batch of marks, which included QI2, from the Identification Bureau to SCRO said "Ident Required for Deceased".91 None of the fingerprint examiners who gave evidence looked on this as an instruction to provide a positive identification of any or all of the marks as deriving from Miss Ross, but as an indication that the marks required to be compared against those of Miss Ross.92
28.83. The handwritten note about a visit to SCRO by Mr McAllister which referred to the Senior Investigating Officer as 'Stevie Heath' gives the impression of a close and relatively informal working relationship between police and the fingerprint bureau. Mr Heath's evidence was that he did not know the fingerprint examiners well.93 However Mr Stewart's evidence was that the level of personal relationship with an SIO "sometimes decided the level of information and the quantity of information you got."94 Mr Stewart spoke to having worked on many occasions with Mr Heath.95
28.84. Mr McAllister spoke about what he saw as the positive need to keep the fingerprint bureau informed and make them feel as if they were part of a team and he indicated also that this was a method of working promoted under Mr Heath's direction.
28.85. There is some academic research tending to show that contextual information may sub-consciously influence the conclusions of fingerprint examiners, although there is also research which has reached the opposite conclusion.96 There may be risks in the inclusion of fingerprint examiners in the police 'team' in the way described by Mr McAllister. Their proper role is to provide their conclusions in relation to the identity of the donor of marks. Work can be prioritised without the provision of detailed reasons as to why the police consider a particular mark to be of importance. In relation to QI2, there was no need for the fingerprint examiners to be made aware of the significance of the tin from the point of view of the police.
28.86. Mr MacPherson could not recall seeing the handwritten note from Mr Stewart. He could not say with any certainty whether he had seen the Form 13B. His evidence was, however, that he was aware of the background to mark QI2 at the time he examined it.97 His knowledge of what was believed to be the significance of the tin in the context of the police investigation is a factor which may, sub-consciously, have influenced him and contributed to the initial misidentification by him of mark QI2 Ross.
Conclusion
28.87. There is no evidence that the SCRO officers deviated from their normal practices and there is no suggestion that these had led them into error in the past. Nonetheless, there was a number of risk factors present in those practices and in the case of Y7 and QI2 Ross these crystallised into error. An excess of confidence in the skill and ability to work with complicated marks within the bureau is likely to have been the main reason why they fell into error rather than any lack of competence, the principal underlying methodological failure being that the SCRO officers concentrated on only part of the mark and applied an inappropriate degree of tolerance in the comparison of marks that were of poor quality.
Precedent
28.88. In the second of his two reports prepared prior to the trial in HMA v McKie, Mr Wertheim explained that in beginning a comparison of mark and print a 'target group' should be selected and for Y7 he chose a group of five points. In his report he used the numbering in Production 189 but the corresponding numbers as used in the comparative exercise were SCRO points 3-6 and 9. He wrote that though these five points did not match exactly they could be considered within tolerance.98 In their earliest reports on Y7 the National Training Centre at Durham conceded only three points in Y7 as bearing any similarity to characteristics in Ms McKie's left thumb print;99 those being SCRO points 1, 3 and 4. However, in the final report prepared for Mr Gilchrist's investigation they accepted that five points broadly coincided and, again converting to the numbering used in the comparative exercise, they were SCRO points 3-6 and 15.100
28.89. There is, therefore, some support for the view that focusing on a target group could suggest some similarity between Y7 and the left thumbprint of Ms McKie. Of course, neither Mr Wertheim nor the experts at Durham rested there because they tested the superficial similarity between the target group points in the light of the remaining detail and concluded that the two did not match.
28.90. In retrospect it is worth noting the contemporaneous observations made by Mr Wertheim as noted in the precognition that the Crown took from him when he attended Glasgow High Court on 5 May 1999:
"Inevitably as print distortion increases the clarity of the ridge characteristics decrease. It is this fact that means that an element of tolerance [or] latitude has to be brought in by the examiner when making his comparison. It is here I think that things went wrong.
I could find maybe 5 so called features or characteristics from the 16 highlighted by Scottish fingerprint examiners. Even then these were on the outer range of tolerance making due allowance for distortion or pressure variations in the ridge characteristics.
It seems to me that the problem here was in the methodology adopted; having found a comparison of this sort I would then look elsewhere in the characteristic of the print to see if I could match the conclusion or not. In making my examination I did not confine myself to the box area as defined by the Scottish print examiners. I examined the entire latent and then made my own comparison. I noted between Productions 152 and 189 that there were different points charted in each of the two exercises undertaken by the Scottish Police but even then I still could not get sufficient common characteristics to share their conclusion as to identity.
My feeling is that a mind set developed once a target group of common characteristics had been found and from there the examiners were looking for features that would justify or back-up 5 target points rather than looking at the remainder of the print with a fresh eye.
Once the target group had been found in both the locus print and the comparison then greater care has to be taken. It is then essential to work off the latent print to find additional characteristics not to work off the known print seeking confirmation from the latent print.
I do not know how the Scottish fingerprint examiners approach this task; the problem is that when the 5 target characteristics have been identified at the outer range of tolerance, one then looks to other common characteristics but actually extend the limits of one's tolerance. Thus a false conclusion is reached. In a sense once the target points have been established, these were on the outer range of tolerance, the tolerances become even more elastic in order to achieve a result. There is nothing wilful or deliberate in this, it is simply human nature."101
28.91. This correctly points to the error being attributable to the methodology applied, rather than any wilful or deliberate misconduct. It also encapsulates a number of the risk factors discussed, including the focus on a target group, the failure to look at the mark as a whole, reverse reasoning being applied to back up the five target points, working at the outer limits of tolerance even in relation to the five target points and applying "even more elastic" tolerances in order to achieve the result of 16 points in sequence and agreement.
28.92. Reference can also be made to the two reports by the INTERPOL European Expert Group on Fingerprint Identification entitled 'Method for Fingerprint Identification'.102 Mr Zeelenberg was chairman of the group when the first report was written. Two passages in the General Statements section of the first report are pertinent:
"4. Environment
(i) Mistaken identifications have some common causes. The (latent) fingerprints being examined were of bad quality, the expert was biased and there was pressure involved. The expert(s) was (were) sure to be right and could most of the time not be convinced of the opposite. Independent experts investigating the print later judged most of the times the prints to show insufficient detail for identification or even for comparison. Real verification did not take place.
(ii) False identifications are human errors but errors are human. If man were able to judge independently and free of bias, mistakes would be virtually impossible. The fingerprint expert is working in a 'field of force' that generates pressure towards results. Open pressure but mostly hidden, pressure from outside but also from the inside. The need for result can be considerable in high profile cases. The longing for result leads to guided perception and biased evaluation. More subtle is the mechanism of subconsciously deciding while comparing. If one has found 6 points in agreement and gets the 'warm feeling', the perception and validation are guided often leading to upgrading information, ignoring differences and stretching tolerances.
(iii) Everything should be undertaken to keep the pressure off the investigating process. It is the responsibility of the management to create an open and sound culture in the first place. A sound culture starts with proper goals for the organisation. The goals of the forensic specialist is not generating results but scientifically sound conclusions regardless who 'profits' from them.
(iv) The organisation should not be involved in the judicial system as a party and express verdicts in terms of winning or losing.
5. Hierarchy
1. Hierarchy (rank) in scientific decision-making is considered to be inappropriate. The dangers of such a process, which must be recognised and overcome if a hierarchical system is used, are that:
a. The 'junior' tunes his/her opinion to that of the 'senior';
b. The culture of the longer serving expert 'sees more';
c. The pressure on the junior to please the senior."
28.93. The factors present in relation to Y7 and QI2 Ross that find precedent in that report are: (1) the fact that the marks were of poor quality (i.e. complex marks); (2) the experts felt themselves "sure to be right" (i.e. 100% certainty and an over-confidence in the ability to identify marks that others could not); (3) experts sub-consciously deciding while comparing (Mr MacPherson's inner conviction or Ms McBride's personal point of certainty), leading to guided interpretation (or reverse reasoning) that results in ignoring differences and "stretching tolerances"; and (4) a culture in which the views of more junior examiners were not given due weight because a longer serving officer "sees more".
Contextual bias and impropriety
28.94. The INTERPOL report does mention bias. In chapter 5 the background material available to the SCRO officers at the time of the examination of the marks has been explained and some of the key elements have been mentioned in this chapter. Bias in this context is a reference only to sub-conscious influences that may come to bear on an expert. It does not refer to actual bias or any impropriety.
28.95. There are two positive indications in support of that conclusion. The first is that I am satisfied that the SCRO officers had no motive to harm Ms McKie. Nothing in the background information supplied by the police led either the police or SCRO to suspect Ms McKie as the maker of the mark and, far from assisting the police, even at the time of the investigation the misidentification of Y7 complicated the murder investigation and placed considerable strains on individual police officers including Mr Heath and Mr Shields. As for QI2 Ross, the critical fact is that Crown Counsel had authorised the full committal of Mr Asbury on the charge of murder on 30 January 1997, the day before that print was identified.103 The assessment of Crown Counsel was that there was a sufficiency of evidence to place Mr Asbury on trial even before that print was identified and I am satisfied that there is no basis for concluding that QI2 Ross had to be identified in order to make a case against Mr Asbury.
Other allegations of impropriety
28.96. In the report dated 27 September 2000, fingerprint experts at the National Training Centre, including Mr Sheppard and Mr Grigg, reported that they had "grave doubts" that the examinations of Y7 were carried out totally independently and advised that "without adequate explanation there appears to be collective manipulation of evidence and collective collusion" to identify Ms McKie erroneously.104
28.97. The factors that led to that opinion coincided in large measure with the observations that Mr Wertheim made to Mr Gilchrist.105
28.98. These matters were raised with Mr Sheppard in his evidence to the Inquiry106 and setting aside for the moment the question of the degree to which there was a misidentification of the marks, he accepted that a fuller understanding of Scottish practice could provide possible explanations for the specific considerations that led him to suspect collusion.
(i) In England the practice was for each expert to prepare a separate report but the fact that four SCRO officers prepared joint reports and joint exhibits was sanctioned by statute: section 280 of the Criminal Procedure (Scotland) Act 1995.
(ii) The SCRO experts did not rely on the charted enlargements as evidence of their own individual comparisons but merely as illustrations to show the jury generally the method by which officers reached a conclusion on identity.
(iii) The charted enlargements were of poor quality but that was because the computer on which they were prepared itself produced images of poor quality.
(iv) Images were "cropped", that is to say showed only the lower section of Y7 where similarities were said to have been found and not the upper section where the NTC observed differences, because the computer was capable of reproducing only a small area of the mark.
(v) The images in Productions 152 and 180 were identical because they were taken from the computer.
28.99. For completeness, Mr Wertheim also drew attention to the facts that (1) the production of a total of three sets of charted enlargements was unusual and (2) different inked prints were used in Production 189 and his observation was that progressively from production to production more dissimilarities were masked.107 I am satisfied that there is adequate explanation for each of these matters. In particular, the fact that Production 189 used a different print for Ms McKie was because Crown Office instructed that a fresh comparison be carried out relative to the prints that were taken from her when she was arrested, which occurred at a date after Productions 152 and 180 were prepared. That in itself would have resulted in at least two sets of charted enlargements and I am satisfied that the third set were produced for the internal police disciplinary investigation led by Mr Wilson. The number of charted enlargements can be explained but, more significantly, Mr Wertheim, like Mr Sheppard, did not have the benefit of an understanding of the peculiar status of the charted enlargements in Scottish practice. Mr Wertheim understood them to be the 'best evidence' being advanced by the SCRO officers in support of their identification. In fact, as already noted, they were only copies produced for illustrative purposes and the Crown lodged separately the 'best evidence' which was (a) unmarked photographs of Y7, (b) the original fingerprint forms taken from Ms McKie, (c) photographic negatives of Y7 (Production 172) and (d) the door-frame with Y7 in its natural state (Label 102). Access to the original material, coupled with Mr Wertheim's ability to take his own prints from Ms McKie, enabled him successfully to challenge the SCRO identification. He was not impeded by the charted enlargements. On the contrary, it has to be recalled that SCRO was unique by this time in producing charted enlargements, and the fact that the enlargements were available and showed the total of 17 points on which the SCRO examiners relied to illustrate their opinion gave Mr Wertheim a secure basis on which to contest the basis for the identification. Without the lead afforded by the chartings it would have been much more difficult for Mr Findlay Q.C. to have been properly prepared to cross-examine the SCRO witnesses.
28.100. Having dealt with these specific matters, I return to the additional factor cited by the National Training Centre which was, in effect, incredulity that four fingerprint experts acting independently could have identified Y7 when the NTC experts found as few as three (latterly five) points bearing even a superficial similarity. That has already been addressed in the earlier section of this chapter. I am satisfied that the misidentifications are attributable to poor methodology. Among the pertinent considerations is that Mr Graham, instructed independently by the defence in HMA v Asbury, agreed the identifications of Y7 and QI2, and Mr Swann similarly agreed Y7 when instructed as an independent expert for Ms McKie. Mr Swann subsequently also confirmed the identification of QI2 Ross and I would add that Mr Berry and Mr Leadbetter also subsequently endorsed the identification of Y7. My conclusions on Y7 and QI2 Ross mean that I have not accepted the expert opinions of Mr Graham, Mr Swann, Mr Leadbetter and Mr Berry but the very fact that independent experts with their considerable experience could agree with SCRO satisfies me that there was no impropriety on the part of the SCRO officers.
28.101. The final perspective to be addressed is the allegation that what began as a mistake, perhaps due to lack of attention to what was initially in the case of Y7 no more than an incidental elimination print, became over time impropriety as the SCRO officers failed to acknowledge the error in the identifications when presented with the contrary opinions of other experts including Mr Wertheim and Mr Zeelenberg. The premise is readily understandable but I have already highlighted that one of the by-products of experts being taught that they can have 100% certainty in their own conclusions is a significant natural impediment to a subsequent change of view. I am satisfied that the SCRO officers from whom I have heard evidence to this day genuinely, but mistakenly, continue to subscribe to the conclusions that they formed in 1997.
1. SP_0004 Black Report appendix 1(a) paras 3.16-3.18 and PS_0147 - Lord Advocate's Written Answer to Parliamentary Question S1W-16832, Scottish Parliament 6 July 2001
2. See para 96ff below
3. INTERPOL European Expert Group on Fingerprint Identification - IEEGFI (2000) Method for Fingerprint Identification, URL: http://www.interpol.int/public/Forensic/fingerprints/WorkingParties/IEEGFI/ieegfi.asp?HM=1; and IEEGFI II (2004) Method for Fingerprint Identification Part II, URL: http://www.interpol.int/Public/Forensic/fingerprints/WorkingParties/IEEGFI2/default.asp
4. See chapter 38 para 54
5. US Department of Justice, Office of the Inspector General (2006) Review of the FBI's Handling of the Brandon Mayfield Case (Unclassified and Redacted) (US Department of Justice) URL: http://www.justice.gov/oig/special/s0601/final.pdf, pdf pages 221-222
6. Mr MacPherson 27 October pages 47-51. Mrs Tierney gave similar evidence under reference to her training in Northern Ireland - 12 November page 59.
7. Mr MacPherson 27 October page 59
8. Mr MacPherson's description of analysis and comparison is at 27 October pages 45-66.
9. FI_0055 paras 118-119 Inquiry Witness Statement of Mr MacPherson
10. Mr Stewart 5 November pages 34-41
11. FI_0036 paras 173-174 Inquiry Witness Statement of Mr Stewart and Mr Stewart 5 November pages 43-44 and 173-174
12. Mr McKenna 6 November pages 25-31
13. FI_0054 para 31 Inquiry Witness Statement of Mr McKenna
14. FI_0054 para 30 Inquiry Witness Statement of Mr McKenna
15. Mr McKenna 6 November pages 30-31
16. FI_0054 paras 113-114 Inquiry Witness Statement of Mr McKenna
17. Ms McBride 6 November page 128
18. Ms McBride 6 November pages 126-130
19. Ms McBride 6 November pages 126-137
20. Ms McBride 6 November pages 136-137
21. The extent to which this is actually achieved in practice is considered in chapter 36.
22. Mr MacPherson 27 October page 137
23. Mr Chamberlain 18 November pages 23-24
24. Mr Grigg 29 September pages 7-10
25. FI_0201 para 23 Inquiry Witness Statement (Supp.) of Mr Zeelenberg
26. Mr MacPherson 27 October pages 64-66
27. CO_1375 pdf page 30
28. Mr MacPherson 27 October page 134
29. Mr MacPherson 27 October page 135
30. Ms McBride 6 November page 137
31. SG_0375 para 8.14.1
32. Mr Luckraft 20 October pages 13, 44ff
33. FI_0113 paras 1-12 Inquiry Witness Statement of Mr Luckraft
34. PS_0147
35. FI_0113 para 9 Inquiry Witness Statement of Mr Luckraft; see chapter 23 para 49
36. FI_0113 para 10 Inquiry Witness Statement of Mr Luckraft
37. CO_1375. See chapter 23 para 48
38. CO_1375 pages 26-27
39. Mr Zeelenberg 7 October pages 59-63
40. CO_0003 pdf page 18 and Mr Graham 9 July pages 61, 75-81
41. Ms McBride 6 November pages 135-141 at page 139
42. See chapter 25 paras 45-76 and 88-89
43. FI_0113 para 6 Inquiry Witness Statement of Mr Luckraft
44. See quotation in chapter 23 at para 46
45. FI_0036 para 121 Inquiry Witness Statement of Mr Stewart
46. See chapter 26 para 21ff
47. AZ_0061
48. Mr MacPherson 29 October pages 97-102
49. See chapter 26 para 30ff
50. INTERPOL (2000) Method for Fingerprint Identification para 12 and INTERPOL (2004) Method for Fingerprint Identification Part II para 8.10
51. The text says "contributed".
52. Mr Zeelenberg 7 October pages 100-102, 104-105 and AZ_0061 slides 158 and 162
53. Mr Zeelenberg 7 October pages 45-47 and AZ_0061 slides 59 and 60
54. Mr MacPherson 27 October pages 53, 76-79 and FI_0055 para 118 Inquiry Witness Statement of Mr MacPherson
55. Mr Stewart 5 November pages 57-58
56. Mr MacPherson 27 October page 135
57. Mr MacPherson 27 October pages 139-140
58. Ashbaugh, D.R. Quantitative-Qualitative Friction Ridge Analysis, Boca Raton: CRC Press, 1999 pages 85 and 91-92; see chapter 2 para 9
59. See chapter 35 para 80ff
60. Mrs Tierney 12 November pages 69-70; see chapter 35 para 83
61. Mr Grigg 29 September page 96; see chapter 25 para 165
62. Mr MacPherson 27 October pages 132-134
63. Mr Graham 9 July pages 61-63
64. Mr Leadbetter 23 October page 52
65. Mr MacPherson 29 October pages 1-8; see, more generally, chapter 35 para 75ff
66. CO_4342 section 2.4
67. See chapter 25 para 164ff
68. Mr Dunbar 6 October page 67
69. Mr Stewart 5 November page 168
70. CO_1465 pages 18-19
71. Mr Bruce 9 July page 150
72. PS_0002h
73. See chapter 5 para 76ff
74. FI_0113 para 12 Inquiry Witness Statement of Mr Luckraft
75. MP_0007 pdf p 84; and MP_0006 and Mr Pugh 24 November page 51ff correcting FI_0082 para 49 Inquiry Witness Statement of Mr Sheppard and Mr Sheppard on 7 July page 182ff; see chapter 38 para 54
76. Mr Stewart 5 November pages 43-44, 173-174, c.f. FI_0036 para 174 Inquiry Witness Statement of Mr Stewart
77. Paras 30-32
78. Mr Sheppard 7 July pages 176-178
79. CO_1375 pdf page 30
80. See chapter 7
81. See chapter 7 para 192
82. See chapter 10 para 101ff
83. See chapter 7 para 194
84. SG_0528 page 28
85. SG_0528 page 30
86. SG_0526 page 86
87. Mr Graham 9 July pages 80-81 and see also pages 58-62 and 75-81
88. DB_0256
89. DB_0253
90. Mr Stewart 5 November pages 53-54,159-166
91. DB_0251 pdf page 33
92. FI_0056 para 50 Inquiry Witness Statement (Supp.) of Mr MacPherson, FI_0040 paras 13-15 Inquiry Witness Statement of Ms McBride, Mr Stewart 5 November pages 51-52, Mr Geddes 26 June pages 80-81 and FI_0054 paras 54-55 Inquiry Witness Statement of Mr McKenna
93. Mr Heath 9 June page 44
94. Mr Stewart 5 November page 55
95. Mr Stewart 5 November pages 55-56
96. I. Dror and D. Charlton, Why Experts Make Errors (2006) 56 Journal of Forensic Identifications 600-616; I.Dror and others, Contextual information renders experts vulnerable to making erroneous identifications Forensic Science International 156 (2006) 74-78; but see also Hall, Player, Will the introduction of an emotional context affect fingerprint analysis and decision-making? Forensic Science International 181 (2008) 36-39
97. Mr MacPherson 27 October pages 75-76 and FI_0055 para 51 Inquiry Witness Statement of Mr MacPherson
98. DB_0168 para 15
99. CO_1065 para 4.3
100. CO_2003 para 2.18
101. CO_0024 pages 7-8
102. INTERPOL (2000) Method for Fingerprint Identification and INTERPOL (2004) Method for Fingerprint Identification Part II
103. See chapter 5
104. CO_1065 pdf page 8; see chapter 17 para 4ff
105. CO_0003 page 29
106. Mr Sheppard 8 July pages 23-30
107. CO_0003 pdf page 29