Major R - David Wolchover
Visual
Identification
Procedures
Under PACE Code D
David Wolchover
DavidWolchover.co.uk
Visual Identification Procedures
Under PACE Code D
David Wolchover
Of Gray’s Inn, Barrister
Former Head of Chambers
at
[pic]
Seven Bell Yard
DavidWolchover.co.uk
Published on-line at DavidWolchover.co.uk
(Expanded from an earlier version published on-line 26 November 2003)
Previously posted: 23/02/05; 11/03/05; 1/04/05; 5/07/05; 8/09/05; 20/10/05; 18/11/05; 20/6/06; 28/04/07; 18/05/07;
1/08/07; 11/11/07; 27/11/07; 15/01/08; 16/05/09; 10/12/09; 15/12/09; 12/02/11; 27/03/11; 22/09/15; 04/10/15; 16/04/16; 14/06/16; 29/12/2021
This edition posted 02/01/2022
Readers should note that as a result of recent editing there may be some discrepancies in cross referencing between particular passages of text or footnotes and other such text footnotes of which lack of time has precluded amendment. It is planned to complete that exercise in the near future.
© David Wolchover, davidwolchover.co.uk
Set in Microsoft Word 10 point Times New Roman
All rights reserved. UK statutory material in this publication is acknowledged as Crown Copyright. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system without prior permission of the author and webowner
Contents
Table of
Cases i
Chapter 1 Visual Identification in
the Criminal Process 1
Chapter 2 Suspect’s Identity Not Known 29
Chapter 3 Suspect’s Identity Known 59
Chapter 4 Video Identification Parades 100
Chapter 5 The Evidence in Court 158
Table of Cases
PLEASE NOTE THAT EARLIER DISCREPANCIES IN
page references have been corrected
Attorney-General’s Reference (No. 2 of 2002) [2003] 1 Cr.App.R. 21 10, 13, 45
Castorina v Chief Constable of Surrey [1996] LGRR 241 17
Chapman v DPP (1988) 89 Cr.App. R. 190 17
Coulman v D.P.P. [1997] C.O.D. 91, D.C. 18, 46, 55
Cumming and others v Chief Constable of Northumbia Police
[2003] EWCA Civ 1884 99 19
Holland v. H.M. Advocate, The Times, June 1, 2005 158
Kajala v Noble (1982) 75 Cr.App.R. 149, D.C. 28
K v DPP [2003] EWHC 351 (Admin); [2003] 4 Archbold News 1 17, 69, 165
McKenna v. DPP [2005] 5 Archbold News 2 77
O’Hara v Chief Constable of the Royal Ulster Constabulary
[1997] 1 Cr.App.R. 447 17
Perry v United Kingdom (2003) E.C.H.R. No. 63737/00 95, 120
Powell v D.P.P [1992] RTR 270 164
R v Allen [1996] Crim.L.R. 426 30, 161, 164
R v Amin [2015] 631 21, 82
R v Anastasiou [1998] Crim.L.R. 67 72, 79
R v Aurelio Pop (2003) 147 S.J. 692 158
R v Bell [1998] Crim.L.R. 87 71, 74
R v Birch, Bryant and Crowley (1992) unreported CA 90/0947/Z3 152
R v Blenkinsop [1995] 1 Cr.App.R. 7 10, 46
R v Brizey (1994) unreported, 10 March 53
R v Brown [1991] Crim.L.R. 368 46, 73
R v Bush (1997) unreported, 27 January 18
R v Byron (1999) unreported, 12 February 61
R v Callie [209] EWCA Crim 283 75
R v Chen and others [2001] 5 Archbold News 3 69, 70
R v Christie [1914] AC 545 159
R v Clare and Peach [1995 2 Cr.App.R. 333 10, 45
R v Cole (2013) unreported July 5, C.A. 20
R v Conway (1990) 91 Cr.App.R. 142 76
R v Constantinou (1990) 91 Cr.App.R. 74, C.C.C. 36
R v Cook (1987) 84 Cr.App.R. 369 36
R v Crowe, unreported C.A. 6 March 1998 51
R v Dodson and Williams (1984) 79 Cr.App.R. 220 45
R v Downey [1995] 1 Cr.App.R. 547 45
R v Duggan (1998) unreported 16 June 51
R v Edwards (2006) 150 S.J. 570 158
R v El Hannachi and others [1998] 2 Cr.App.R. 226 17, 51, 71
R v Elson (1994) unreported, C.A. 94/0547/Y3 134, 151
R v Emiku [2003] EWCA Crim 2237 65
R v Finley [1993] Crim.L.R. 50 134, 166
R v Folan [2003] EWCA Crim 908 91, 98
R v Forbes [2001] 1 Cr.App.R. 4306 7, 14, 46, 72, 73, 76, 162, 166
R v Fowden and White [1982] Crim.L.R. 588 28, 42
R v Fuller (2000) unreported, 22 May 53, 165
R v Gall [1989] Crim.L.R. 745 116, 165
R v Gasper [2002] EWCA Crim 1764 89
R v Gayle [1999] 2 Cr.App.R. 130 64
R v George [2003] Crim L.R. 282 82, 87, 105, 160
R v Gojra and Dhir [2011] Crim.L.L. 311 66
R v Gornall 149 S.J. 300 18
R v Governor of Pentonville Prison ex p Voets [1986] 1W.L.R. 470 31
R v Graham [1994] Crim. L.R. 414 59
R v Grimer [1982] Crim.L.R. 674, C.A. 9
R v Harley (2000) unreported, 22 February 81
R v Harris [2003] EWCA Crim 174 61, 73, 75, 77, 162, 166
R v Haynes [2004] EWCA Crim 390 76
R v Hickin [1996] Crim.L.R. 584 50,53, 57-58, 165
R v Hope, Limburn and Bleasdale [1994] Crim.L.R. 118 69
R v Jamel [1993] CrimL.R. 52 86
R v Jones and Nelson, The Times 21 April, 1999 81, 165
R v Jones and others (1994) 158 J.P. 293 47
R v Joseph [1994] Crim.L.R. 48 164
R v Kennedy (1992) unreported, C.A. 20 March, 90/02296/Y4 94, 120, 130
R v Kitchen [1994] Crim.L.R. 684 17, 30, 80
R v Lamb (1980) 71 Cr.App.R. 198 30, 32
R v Lambert [2004] EWCA Crim 154; (2004) 68 J.C.L. 285 65, 66
R v Lariba [2015] Crim. L.R. 534, C.A. 14, 41
R v Lennon (1999) 63 J.C.L. 459 58
R v Loveridge et al. [2001] 2 Cr.App.R. 29 96
R v Malashev [1997] Crim.L.R. 587 49
R v McCartney and others [2001] EWCA Crim 2283;
[2003] 6 Archbold News 2 65
R v McKay (1900) 91 Cr.App.R. 84; [1990] Crim.L.R. 338 159
R v McMath [1997] Crim.L.R. 586 51, 52, 69, 73
R v McNamara [1996] Crim.L.R. 750, C.A. 45
R v Maqsud Ali [1966] 1 Q.B. 688, C.A. 28
R v Marcus (Ruel) [2004] EWCA Crim 3387; [2004] All ER (D) 351 86,126
R v Maughan (Bernard), unreported, Wood Green Crown Court, T1998 0680 153
R v Mendili [2001] EWCA Crim 757 131, 166
R v Meredith and Cowan [2001] EWCA Crim 1415 65
R v Miah [2001] EWCA Crim 2281 50
R v Morrisey (2014) 78 J.C.L. 460 45
R v Moss [2011] Crim.L.R. 560 14, 42
R v Muhidinz (2005) 70 J.C.L. 197 69
R v Nagah (1991) 92 Cr.App.R. 344 165
R v Nicholson [2000] Cr.App.R. 182 61
R v Noonan [2003] EWCA Crim 3869 79
R v Nunes [2001] EWCA Crim 2283; 10 Archbold News 1 18, 55, 58, 73, 79
R v O’Brien et al. [1982] Crim.L.R. 746 36, 164
R v O’Leary and Lloyd-Evans 67 J.C.L. 115; CLW/03/16/5 69, 73
R v Okorudu [1982] Crim. L.R. 747 36
R v Osborne and Virtue [1973] 57 Cr.App.R. 297; [1973] Crim.L.R. 178 159
R v Oscar [1991] Crim.L.R. 778 76
R v Perry (2000) unreported, 3 April 120
R v Popat [1998] 2 Cr.App.R. 208 47, 62, 71, 72, 161
R v Popat (No 2) [2000] 1 Cr.App.R. 387 72
R v Quinn [1995] 1 Cr.App.R. 480 116, 131, 166
R v Rogers [1993] Crim.L.R.386 18
R v Rutherford and Palmer (1994) 98 Cr.App.R. 191 61
R v Ryan [1992] Crim.L.R. 187 116
R v Shanmugarajah and Liberna [2015] 2 Cr.App.R. 215(14) 46, 65, 98
R v Slater [1995] 1 Cr.App.R. 584 69
R v Sloan (Brandon), unreported, Inner London Crown Court, T2019 1017 141
R v Smith (Dean Martin) and others [2009] 1.Cr.App.R. 521 (36) 14, 41
R v Tolson (1864) 4 F. & F. 103 28
R v Turnbull (1977) 98 Cr.App.R. 313 2, 46, 69,137, 166
R v Vaughan (1997) The Independent 12 May 49, 50, 58, 165
R v Wait [1998] Crim. L.R. 68 47, 73
R v Walters [2001] EWCA Crim 1261 116
R v Williams [2000] unreported, February 22, C.A. 32
R v Williams [2003] EWCA 3200; (2003) SJ 1305 19
R v Williamson [2002] EWCA Crim 1809 77
R v Willoughby [1999] 2 Cr. App.R. 82 147, 153
R v Wright [1994] Crim.L.R. 131 118
Ryan v D.P.P. (2000) unreported, 10 October 50
Taylor v Chief Constable of Cheshire (1987) 84 Cr,App.R. 191, D.C. 28
Tido v The Queen, unreported, June 15, 2011 158
Chapter 1
Visual Identification in the Criminal Process
A. The Dawning of Enlightenment
Every criminal investigation involves the initial question of the offender’s identity. In some cases the police will have little or no idea who may have perpetrated the crime. In other cases they may quickly discover facts which bring a particular individual under suspicion. Sometimes those grounds for suspicion will trigger a confession. Often forensic science will be vital in proving identity and increasingly science is displacing subjective testimony across a broad range of disciplines. However, a very important instrument of proof of guilt remains, as it has always been, that of establishing the identity of the culprit visually by eyewitnesses. In a less sophisticated age, where the culprit was a stranger to the eyewitness and someone fell under suspicion reliance would be placed on confronting the suspect with the witness as a means of establishing that the suspect and the culprit were one and the same person. Later, in court, the witness might or might not be asked to confirm that he had earlier identified that person as the culprit. However, with or without any refererence to an earlier identification, witnesses would almost invariably be invited to say if they could see the person in the court room. Such a studied avoidance of a leading question was entirely spurious, as if it would neutralise the pointed effect upon the witness of the glaring presence of the defendant sitting in the dock. (The formal identification of the accused in court in this way later became known as a “dock identification.”) With enlightenment it came to be appreciated that the force of suggestion was an innate deficiency of blunt confrontation and in the view of the committee on identification evidence chaired by Lord Devlin it was probably judicial criticism of its crudeness which resulted in the emergence of the identification parade as the best practicable palliative.[1] The commentator C. H. Rolph pointed out that “the mists of antiquity have closed over the date” when identification parades first emerged[2] but rules for conducting them have been traced at least as far back as a Metropolitan Police Order of 24 March, 1860 following some remarks of the Assistant Judge at the Middlesex Sessions.[3] In the wake of the notorious case of Adolf Beck in 1904[4] the Metropolitan Police revised their regulations, which in 1905 were commended by the Home Secretary to all Chief Constables[5] and from time to time thereafter, typically following some cause célèbre or other,[6] the rules(strictly speaking guidance principles(were refined and improved through amendments published by Home Office circular. Then, following two notorious miscarriages of justice involving mistaken visual identification, the Devlin committee was appointed to consider the whole question of visual identification and issued their celebrated report in 1976.[7]
Apart from examining the two cases in detail Devlin reviewed investigative procedures for cases involving visual identification and the management at trial of evidence relating to the issue of identification and focused attention on the growing body of research into the psychology of identification. The report made a number of recommendations, most prominent perhaps of which was that visual identification should not normally be permitted to sustain a conviction unless supported by substantial evidence of another sort. Inspired by Devlin, the Court of Appeal, in R v Turnbull,[8] issued guidance on the withdrawal of weak cases from the jury and on the standard directions which juries were to be given about the potential dangers involved in visual identification. However, the court stopped short of declaring a rule requiring corroboration of such evidence and convictions may be sustained, not merely in theory, but in practice too, on the evidence of a visual identification alone. To attempt to understand what factors may have been behind the court’s attitude it is necessary to move from history to a brief consideration of human nature.
B. Balancing Assumptions on Accuracy and Error
The visual recognition of a particular person from the memory of a previous encounter will typically be based on one or more of a number of general features: gender, age, build, height, hair, complexion, ethnicity, hair type and colour, to specify most of the obvious ones. While any one of these may tend either to remove a person from suspicion or point to the possibility that the person was the offender, even in combination they can rarely if ever by themselves furnish proof of guilt because there will be nothing uniquely distinctive about a person’s appearance when described under these categories. However, it is the human face which, visually (though self-evidently with far less certainty than in the case of a fingerprint or DNA match) marks out an individual from almost every other person on the planet. It may be an innate grasp of this fact which instils in each of us an instinctive feeling that even if we cannot easily call it up into our mind’s eye we never forget a face, once noticed, when we see it again and it may be a reflection of that instinctive understanding that investigators, lawyers and jurors have tended to believe in the general infallibility of face recognition.
In order to comprehend and thereby to try to prevent those breathtaking errors of identification of the kind retailed earlier a growing body of psychologists have been drawn to conduct a wide range of theoretical and empirical research on the mental processes involved in recognition, particularly of the face. With the accumulation of the number of published experimental studies, particularly those based on staged crimes, the realisation has grown that false identifications following dramatic events are by no means uncommon.[9] Large scale studies of actual cases have seemed to bear this out.[10] As a result of this research the old-time blind faith in the absolute reliability of facial identification has given way, in the mind of some populist commentators, to the contrary dogma that practically no one is capable of remembering a face and that identification evidence is “virtually useless.”[11] According to at least one authoritative statement of expert opinion, “[t]his is a view which is both far too pessimistic and too simplistic.”[12] The rather more balanced, if perhaps obvious, judgment has been expressed that psychological research has shown that witnesses’ identification performance can range from total accuracy to total inaccuracy depending upon a multitude of factors.”[13] In a reference to this opinion it has been observed that—
“such a statement is a tacit acknowledgment of the possibility of the unreliability of the testimony of some eyewitnesses, and psychologists have suggested that eyewitness error is a joint product of inherent human cognitive limitations . . . and the methods that are used to obtain information from witnesses.”[14]
While there is little room for complacency and always acknowledging the empirically attested incidence of some human error in the field of visual identification the fact remains that the courts can in the main safely rely on identification by face. At least under laboratory conditions experimental research has demonstrated that people are remarkably accurate at remembering briefly viewed, previously unfamiliar faces. In an important research project undertaken by Bruce, participants typically scored over 90 per cent correct when asked to decide which of a large set of faces in identical views were previously studied (although the figure dropped to 60 per cent when the viewpoint and expression were changed).[15] The study seems to bear out the confidence we tend to feel as a species in our innate ability to recognise faces and the Court of Appeal may therefore have been instinctively right in Turnbull when they refused to introduce a rule precluding convictions based on unsupported visual identification. It will never be possible to remove from the equation the personal limitations on the ability of some witnesses to perceive, recall and recognise faces but it is certainly feasible to try to perfect the general procedures which are employed to test the ability of witnesses to make an identification. This has been the objective in developing and incrementally improving the modern rules of investigation.
C. Variables Determining the Results of Identification Procedures
The result of an identification procedure will be determined or influenced by what are known as estimator and system variables.[16] Estimator variables are those which are specific to the particular witness and the circumstances of the original sighting, for example, the distance between the witness and the person seen, lighting conditions, eyesight quality, duration of the sighting, degree of attention paid, impact of the incident on the witness’s emotional state,[17] and whether the witness and the person seen were from the same or different ethnic backgrounds. In contrast with estimator variables, system variables can be controlled by the criminal justice system and include both the method chosen for testing the ability to make an identification and the rules devised for conducting the particular method adopted. It is essential to balance the sensitivity and fairness of the system variables. If the person originally seen is actually present the procedures must be sensitive enough to allow a reliable witness to be able to make an identification but fair enough to keep the possibility of a witness making a mistaken identification to an acceptably low level.
D. Regulation by PACE
(1) Evolution of PACE Code D
Devlin had made a number of suggestions for improvements in the investigative process where identification was concerned and this resulted in additional improvements being set out in another Home Office circular.[18] However, with the enactment of the Police and Criminal Evidence Act 1984 the circular guidance, further improved and refined, was placed on a statutory footing in the form of Code D, The Code of Practice for the Identification of Persons by Police Officers, the first edition of which was approved by Parliament under s.66 of the Act and came into effect on 1 January 1986. It was substantially revised with the other PACE codes in 1991 and 1995 and then in late March 2002 a further revision was unveiled (although not under a general revision of the codes), coming into effect on 1 April under s.77 of the Criminal Justice and Police Act 2001. It involved a radical restructuring of what was then section 2, together with amendments to Annexes A to E.[19] That edition was originally intended to last for a projected transitional period of two years but a consultation draft revision of the codes was swiftly circulated in June 2002. The earlier amendments to Code D(now further amended in a number of particulars and expressed in the modified linguistic style of the revised codes(were subsequently incorporated into the general revision to the PACE codes which, after being laid in draft before Parliament on 26 February, 2003, came into force on 1 April, 2003. Section 2 of the 1995/2002 version, “Identification by witnesses,” was transposed into section 3. A fifth edition, which came into force on 1 August 2004, made some minor adjustments of a typographical nature to section 3. Further revisions followed in 2006, 2011 and 2017.[20] The latest edition continues the practice of refining and developing the text, which is applied in the current version of this treatise.
(2) The scheme of Code D for visual identification
(i) Distinction between visual identification and recognition
The provisions in Code D for regulating the visual identification by an eyewitness of the perpetrator or person otherwise present at a location suggestive of involvement are primarily concerned with cases in which the individual was a stranger, that is, not somebody known to the witness. Most of the control procedures in Code D are designed to test the ability of eyewitnesses to pick out from a line-up or group an individual they did not know. Self-evidently a procedure designed to identify a stranger will serve little useful purpose if the eyewitness professes to assert that they recognised the person as someone they knew. Placing a suspect whom the witness knows in a line-up with volunteers who are entirely unknown to them will probably tell us little or nothing of value.
Prior to the 2011 revision, Code D, section 3 (the section on visual identification of suspects) applied almost exclusively to regulate identification of a suspected offender not known to the eyewitness prior to the incident in question. With the exception of one paragraph in the 2006 edition declaring that nothing in the code inhibited the showing of films or photographs to the public through the national or local media, or to police officers, for the purposes of recognition and tracing suspects,[21] it contained no explicit measures for regulating recognition of persons known to the witness.That hiatus was rectified in the 2013 edition of Code D, section 3 of which was divided into two parts. Part (A), embracing paragraphs 3.0 to 3.33, regulated “Identification of a suspect by an eye-witness” (excluding recognition cases). Part (B), containing paragraphs 3.34 to 3.37, was captioned “Evidence of recognition by showing films, photographs and other images” and explicitly and purposively dealt with recognition of an individual known to the eyewitness. The 2017 edition of Code D introduced a refinement of the provisions for recognition through imagery by establishing an explicit distinction between recognition “by controlled showing of films, photographs and images,” regulated by an amended Part B, and that by “uncontrolled viewing of films, photographs and images” regulated by discrete provisions in new Part C (paragraphs 3.38 to 3.41).
(ii) Code D identification
D1.2 explains that identification by an eye-witness arises when a witness who has seen the offender committing the crime and is given an opportunity to identify a person suspected of involvement in the offence in a video identification, identification parade or similar procedure. It should be noted that the wording is deficient in failing to include a reference to seeing a person, not actually committing the crime but present at a location in circumstances suggestive of the person’s involvement. The deficiency is remedied by D3.0, which explains that Part (A) of section 3 (identification by an eye-wtiness) applies when an eyewitness has seen a person committing a crime or in any other circumstances which tend to prove or disprove the involvement of the person the witness saw in a crime, for example, close to the scene of the crime, immediately before or immediately after it was committed.
(iii) Testing identification ability and prevention of mistaken identification
Both D1.2 and D3.0 explain that the procedures specified in Part A are designed to be used to test the ability of the eyewitness to identify a person suspected of involvement in the offence (the “suspect”) as the person the witness saw on the previous occasion. However, D1.2 states that the procedures set out in Part A are also designed to provide safeguards against mistaken identification, a purpose which, for some reason, is not replicated in D3.0. That objective would seem to reflect the observation made by the Devlin Committee that(
“An identification parade is not primarily a scientific test of a witness’s memory for faces. It is a device for avoiding a confrontation.”[22]
However, if a parade, whether live or video, is not “primarily a scientific test,” it may be asked what sort of test it is. The short and simple answer is that parades (live or video) do no more than aspire to screen for error. However, lest this suggests the toleration of some flexibility or laxity in the rules requiring their application the importance of compliance with the code provisions could not have been more plainly stressed than it was in R v Forbes when the House of Lords stated that the code was(
“. . . intended to be an intensely practical document, giving police officers clear instructions on the approach that they should follow in specified circumstances. It is not old-fashioned literalism but sound interpretation to read the Code as meaning was it says.”[23]
(iv) Code breach
As with all the PACE codes a breach of a provision will not automatically vitiate the admissibility of evidence obtained following the breach but it may be excluded if to admit the evidence “would have such an adverse effect on the fairness of the proceedings that the court ought not to admit it.”[24] The Code is admissible in evidence and relevant to the question of considering the applicability of the issue of exclusion for unfairness, as well as to any question of the weight to be attached by the tribunal of fact to any evidence of identification obtained in breach of a code provision.[25]
(3) Recognition cases in Code D
(i) Recognition of the actual or suspected perpetrator
as someone known to the witness
It is important to note that D3.0 and Note 3AA stress that Part A does not apply to the procedures specified in Part B, which is used to test the ability of someone who is not an eye-witness, to recognise anyone whose image they see.
In claiming to have recognised the culprit or suspected culprit as a person familiar to them witnesses will usually be giving the person’s name, and possibly address, and an explanation of the context in which they know the person. In many cases the police will encounter little difficulty in using this information to arrest the person so implicated. However, it may be that the witness knows the person quite well by sight but not by name or there will be error or confusion over the name by which the witness believed the person they recognised was generally known. For example, the witness may have described the individual to a third party who confused the description with that of someone else or the witness may originally have confused the name with that of someone else.
Again, the witness may not know the given name of the recognised person but only a nickname or other “tag” which may be in only limited circulation. Name error, confusion or obscurity may offer an obstacle to the police in pursuing their inquiries and they will need to exercise particular care and discretion to avoid causing distress, embarrassment and inconvenience to innocent persons as a result. It will obviously help if the witness happens to possess a photograph of the person or can point to an image posted on social media. Where the police consider that there may be ambiguities or equivocation over the name and they have on record an official photograph of the person to whom they consider the witness was likely to have been referring it will be prudent to show the witness that photograph in order to facilitate the arrest of the right person. Again, they may already have made an arrest on the strength of a name provided by a witness claiming to have recognised the perpetrator as someone well known to him. Here, again, it will be prudent to show the witness an image of the person arrested to eliminate the possibility of error. Code D makes provision for this.
The question whether the witness truly recognised the offender (or suspected offender) is of course dependent on whether that person was actually well known to the witness. In some cases there will be no dispute that they know one another but in other cases the suspect may be denying this. That the suspect claims not to know the witness will hardly mean that the witness cannot know the suspect. After all, such one-way familiarity will be the norm as between celebrities and anonymous members of the public. However celebrities are not usually arrested for street robberies.[26] A more pertinent illustration of the point might feature two members of a community such as the student body in a large secondary school. While a popular and dominating sixth-former “alpha male” may command widespread adulation from the mass of pupils he may well live in blissful unawareness of the existence of a timid and retiring fourth-former, for whom he looms large in the younger boy’s field of awareness. There is a body of case learning in which the question whether a witness truly recognised the suspect as someone well known to him impacted on the issue of whether there should have been a full, controlled identification procedure.[27]
(ii) Two pseudo-categories of recognition: actual
encounter and when viewing imagery
There is a distinction of sorts between on the one hand recognition of the perpetrator by an eyewitness to the crime and, on the other hand, recognition by a member of the public (or a police officer) of a person he knows who is depicted on imagery released into the media, or circulated on a limited basis within the police service, in a quest to find out the person’s identity and whereabouts. However, the distinction may be one without a difference. It is hard to see any contrast in principle between (a) an eyewitness to the crime recognising the perpetrator “in the flesh,” so to speak, as someone he knows, and (b) a witness viewing an image, a series of still images or moving imagery of a person either committing an offence or depicted as present at a particular location from which the individual’s involvement in the offence in question may be inferred, and asserting that he recognises that person as someone known to him.[28] The cogency of professed recognition from imagery will depend on such factors as the clarity and resolution quality of the images and the camera position relative to, and distance from, the person depicted.
(iii) No requirement as to time of onset of knowledge of the person
The admissibility of professed recognition is not dependent on a witness’s knowledge of the person prior to the offence, or even prior to the suspect’s arrest. Two months after interviewing the defendant about a violent disorder for no more than three minutes a police officer viewed stills from a video-recording of the incident and the admissibility of his professed recognition of the defendant based on his presence with him in the interview was not challenged at the subsequent appeal.[29] Nonetheless there will be an obvious need for caution where the opportunity to become familiar with the person is limited and the witness is on the police investigating team.[30]
(iv) No requirement for the recognising witness to have met the offender
In principle it would not even seem to be necessary for the officer actually to have met the person with whom he becomes familiar on the photograph not disputed to be the defendant. We shall return to the topic in the next chapter.[31]
(v) Recognition likely to be facilitated by intense scrutiny of imagery
CCTV recordings are rarely of the highest resolution and quality and the facial and other physical characteristics of persons depicted on it committing an offence will rarely be highlighted with sufficient clarity or proximity to the camera for it to be feasible to establish an immediate match with a person known to the investigator. However, by patiently and repeatedly playing over the footage, both at normal speed and in slow motion, by scrutinising it at length frame by frame and by focusing on, analysing and isolating the individual’s characteristics of build, physiognomy, ethnic type, gait, movements and gestures, it may eventually be possible for an investigator to establish a match with a particular person known to him. It has been held that such diligent scrutiny can give the investigator a degree of special knowledge capable of warranting an avowal of recognition.[32] In Clare and Peach[33] it was accepted that the defendants were present at a football match at which a violent disorder took place, it being agreed that they were depicted on relatively good quality colour images arriving at the ground. From viewing black and white CCTV footage of the disorder forty times, examining it in slow motion and frame by fame, analysing the content and comparing it with the colour images a police officer was held to be capable of recognising the defendants as among those involved in the violence. By virtue of his “lengthy and studious” analysis the officer was possessed of a special knowledge of the recording which the jury lacked and the guidance afforded to them by his expert commentary made up for the impracticability of giving them sufficient time and facilities to conduct their own researches. The defence were not disadvantaged because the recordings were available to them as was the officer for cross-examination.
The facts may be contrasted with the second of two cases in Attorney-General’s reference (No 2 of 2002)[34] in which the officer had spent many hours viewing the video recording of a riot and had then later by chance seen the defendant whom he purported to recognise from the recordings as having been a participant.
(vi) Ambit of Code D recognition procedure: controlled and uncontrolled viewing of imagery
As already mentioned, D3.0 provides that Part A of section 3 does not apply to the procedure described in Part (B) which is used to test the ability of someone who is not an eye-witness, to recognise anyone whose image they see. This is explained further in Note for Guidance 3AA, which advises that the eye witness identification procedures in Part A should not be used to test whether a witness can recognise a person as someone they know and as to whom they would be able to give evidence of recognition from an image seen at a particular date, time and location and in particular circumstances. In such cases, the Note states, the “procedures in Part B shall apply if the viewing is controlled” (which is to say the procedures set out in Part B for establishing whether a witness recognises the person through controlled viewing of imagery). Complementing D3.0 and Note for Guidance 3AA, D1.2A states that separate provisions in Part B apply when any person, including a police officer, is asked if they recognise anyone they see in an image as being someone they know and to test their claim that they recognise that person as someone who is known to them.
The explicit reference to a distinction between controlled and uncontrolled viewing of imagery in the process of visual recognition of offenders was an innovation of the 2017 edition of Code D. Previously, Part B of the 2013 edition had contained paragraphs (D28 and 29) which effectively applied to what was in fact uncontrolled viewing, but without specifically using that adjective. The 2017 edition inserted a new part, Part C, captioned “Recognition by uncontrolled viewing of films, photographs and images” made up of new discrete paragraphs (D3.38 to 3.41) based on the old paragraphs D28 and 29 (which are marked “not used” in the new edition). Heralding Part C, D1.2B states that it applies when a film, photograph or image relating to the offence or any description of the suspect has been broadcast and published in any national or local media or on any social networking site or on any local or national police communication systems. These separate provisions are not subject to the eye-witnesses identification procedures described in D1.2.
(vii) Provisions for controlled viewing of imagery
D3.34 provides that Part B (regulating the controlled viewing of imagery) applies when, for the purposes of obtaining evidence of recognition, arrangements are made for a person, including a police officer who is not an eye-witness:
(a) to view a film, photograph or any other visual medium, and
(b) on the occasion of the viewing to be asked whether they recognise anyone whose image is shown in the material as someone who is known to them.
The paragraph adds that the arrangements for such viewings may be made by the officer in charge of the relevant investigation. It further adds that although there is no requirement for the identification officer to make the arrangements or to be consulted about the arrangements, nothing prevents this.
(viii) Risks involved in controlled viewing
Reference is made in D3.34 to Note 3AA, already mentioned, and to Note 3G, which admonishes that the admissibility and value of evidence of recognition obtained when carrying out the procedures in Part B may be compromised if before the person is recognised, the witness who has claimed to know the person is given, or is made or becomes aware of, information about the person which was not previously known to them personally but on which he has purported to rely to support their claim that the person is in fact known to them.
That oblique expression of caution may embrace a multitude of sins but it is not difficult to conjecture an example of the type of cue which might be supposed to taint a claim of familiarity and would be the sort of mischief against which the Note appears to be aimed. An eyewitness observes a street robbery and believes one of the robbers is someone he knows quite well, although his view was imperfect and he is initially less than certain. However, minutes later he encounters a friend who also knows the individual he suspects to be the culprit and he tells his friend what he has seen. The friend replies that he too has just seen the same person in the vicinity of the robbery and the eyewitness’s belief is elevated to certainty.
A variation on this theme might occur where the witness is mistaken about the name of the person he believes was the robber. Shortly after the incident he gives the incorrect name to a friend who shows him a good photograph of the actual person bearing that name posted on a social media site. The suspected robber and the man in the photograph happen to bear a mild resemblance. Sight of the image coupled with his mistake over the name confirms his erroneous belief that the robber was the man in the photograph and he goes on to identify that person when subsequently shown an official police photograph.[35]
(ix) Controlled viewing safeguards
D3.35 accordingly states that to provide safeguards against mistaken recognition and to avoid any possibility of collusion, on the occasion of the viewing the arrangements for the viewing of imagery (mentioned in D3.34) should ensure:
(a) that the films, photographs and other images are shown on an individual basis
(b) that any person who views the material:
(i) is unable to communicate with any other individual to whom the material has been, or is to be, shown;
(ii) is not reminded of any photograph or description of any individual whose image is shown or given any other indication as to the identity of any such individual;
(iii) is not be told whether a previous witness has recognised any one;
(c) that immediately before a person views the material, they are told that:
(i) an individual who is known to them may, or may not, appear in the material they are shown and that if they do not recognise anyone, they should say so;
(ii) at any point, they may ask to see a particular part of the material frozen for them to study and there is no limit on how many times they can view the whole or any part or parts of the material; and
(d) that the person who views the material is not asked to make any decision as to whether they recognise anyone whose image they have seen as someone known to them until they have seen the whole of the material at least twice, unless the officer in charge of the viewing decides that because of the number of images the person has been invited to view, it would not be reasonable to ask them to view the whole of the material for a second time.
D3.34(d) further stipulates that a record of the decision must be included in the record that is made in accordance with paragraph 3.36, considered below. Reference is made by D3.35 to Note 3G, which explains that the admissibility and value of evidence of recognition obtained when carrying out the procedures in Part B may be compromised if, before the person is recognised, the witness who has claimed to know them is given or is made, or becomes aware of, information about the person which was not previously known to them personally but which they have purported to rely on to support their claim that the person is in fact known to them.
D3.36 requires that a record of the circumstances and conditions under which the person is given an opportunity to recognise an individual must be made and the record must include:
(a) whether the person knew or was given information concerning the name or identity of any suspect;
(b) what the person has been told before the viewing about the offence, the person(s) depicted in the images or the offender and by whom;
(c) how and by whom the witness was asked to view the image or look at the individual;
(d) whether the viewing was alone or with others and if with others, the reason for it;
(e) the arrangements under which the person viewed the film or saw the individual and by whom those arrangements were made;
(f) subject to paragraph D2.8 (requring the recording of an officer’s name and rank where the authority of an officer of specific rank for a particular action is required, unless the exceptions in D2.18 apply) the name and rank of the officer responsible for deciding that the viewing arrangements should be made in accordance with Part B;
(g) the date, time and place images were viewed or further viewed or the individual was seen;
(h) the times between which the images were viewed or the individual was seen;
(i) how the viewing of images or sighting of the individual was controlled and by whom;
(j) whether the person was familiar with the location shown in any images or the place where they saw the individual and if so, why;
(k) whether or not on this occasion, the person claims to recognise any image shown, or any individual seen, as being someone known to them, and if they do:
(i) the reason
(ii) the words of recognition
(iii) any expressions of doubt
(iv) what features of the image or the individual triggered the recognition.
D3.37 provides that the record required under paragraph 3.36 may be made by the person who views the image or sees the individual and makes the recognition; and, if applicable, by the officer or police staff in charge of showing the images to that person or in charge of the conditions under which the person sees the individual.
The person must be asked to read and check the completed record and as applicable, confirm that it is correct and accurately reflects the part they played in the viewing. Note for Guidance 3H advises that the record is made as soon as practicable after the viewing and whilst it is fresh in the mind of the individual who makes the recognition.
In Lariba[36] the appellant and others were alleged to have murdered a rival gang member after pursuing him on their bicycles. The police had used CCTV footage to create a composite recording showing the outward journey of seven cyclists to the vicinity of the attack and their return. To a greater or lesser extent, each group member had disguised his appearance with clothing and face coverings. The composite recording was played on public media and three police officers, having informally viewed the footage on the internet, identified the appellant before attending a formal procedure, when they were again shown the composite recording. Two further officers identified the appellant at a formal procedure only. Immediately after the formal procedure all five officers made witness statements. The officer in charge of the investigation was unaware of paragraph D3.36(k) which had been inserted in the 2011 revision of Code D, requiring a contemporaneous record of a witness’s reaction upon being shown CCTV material and no such records were therefore made.[37]
The appellant had objected to the admissibility of the identification evidence of the five officers, but for the following reasons it was held that the breaches of D3.36(k) did not render the evidence plainly unreliable or unfairly prejudicial. (1) The formal procedure took place not long after the informal viewing of the footage by the three officers who had viewed it and was followed immediately by the making of witness statements. While the statements (made by all five officers) were not a complete record in accordance with the Code D requirements, they did record the factual basis for the recognitions made. In the case of one of the three officers, a virtually contemporaneous record had been made of the circumstances of his informal viewing. (2) The witnesses and the images were available to the jury who enjoyed the advantage of seeing exactly what the witness saw and the image was permanent. (3) The appellant had established that not all police officers who knew him had recognised him from the same images. (4) The judge was able to explain to the jury the respects in which the appellant was disadvantaged by the breaches and to call for extreme caution. (5) There was some supporting evidence for the recognition.
(x) Recognition by uncontrolled viewing of imagery: Part C
D3.38 provides that Part C applies when, for the purpose of identifying and tracing suspects, films and photographs of incidents or other images are:
(a) shown to the public (which may include police officers and police staff as well as members of the public) through the national or local media or any social media networking site; or
(b) circulated through local or national police communication systems for viewing by police officers and police staff;
and the viewing is not formally controlled and supervised as set out in Part B.
D3.39 provides that a copy of the relevant material released to the national or local media for showing as described in D3.38(a), must be be kept. The suspect or their solicitor must be allowed to view such material before any eye-witness identification procedure under paragraphs 3.5 to 3.10, 3.21 or 3.23 of Part A is conducted, provided it is practicable and would not unreasonably delay the investigation. D3.39 explicitly states that the paragraph does not affect any separate requirement under the Criminal Procedure and Investigations Act 1996 to retain material in connection with criminal investigations that might apply to sub-paragraphs 3.38(a) and (b).
D3.40 provides that each eye-witness involved in any eye-witness identification procedure under paragraphs 3.5 to 3.10, 3.21 or 3.23 must be asked, after they have taken part, whether they have seen any film, photograph or image relating to the offence or any description of the suspect which has been broadcast or published as described in paragraph 3.38(a) and their reply recorded. If they have, they should be asked to give details of the circumstances and subject to the eye-witness’s recollection, the record described in paragraph 3.41 should be completed.
D3.41 advises that as soon as practicable after an individual (whether a member of the public, police officer or police staff) indicates in response to a viewing that they may have information relating to the identity and whereabouts of anyone they have seen in that viewing, arrangements should be made to ensure that they are asked to give details of the circumstances and, subject to the individual’s recollection, a record of the circumstances and conditions under which the viewing took place is made. That record must be made in accordance with the provisions of paragraph 3.36 insofar as they can be applied to the viewing in question.
Note for Guidance 3H stresses the importance of making the record referred to in paragraphs D3.36 and D3.41as soon as practicable after the viewing and whilst it it is fresh in the mind of the individual who makes the recognition.
(xi) Implied exclusion of recognition by confrontation
It is to be noted, importantly, that Part B makes no provision for establishing recognition through confrontation in person. Although the Code per se does not explicitly preclude the police from using such a confrontation to determine if the witness recognises the person as someone known to them the controls required by D3.36, whether Annex E or Annex A, as the case may be can only be secured by the use of imagery rather than live confrontation.
E. Identification Method in Part A Determined According to Whether or Not the Identity of a Suspect is “Known”
Generally speaking, the approach which investigators must adopt in seeking to discover evidence, and establish proof, of a culprit’s identity will be determined at a given stage of the inquiry according to whether they are starting from scratch in uncovering clues or whether they have some idea as to who their quarry might be. This rather fluid distinction has been translated in the Code D scheme into two categories of rules: those governing the position where a suspect is “not known” and those where the suspect is known.
A suspect is defined in section 3 of the code as being “known” where “there is sufficient information known to the police to establish in accordance with PACE Code G on arrest that there are reasonable grounds to suspect a particular person of involvement in the offence.”[38] This appears to be a reference to justification in law rather than whether the arrest, technically justifiable in law, might be deemed proportionate or sensible as a matter of tactics. Although it is established that for the purposes of arrest that standard involves a combination of objective and subjective criteria (reasonable grounds and the officer’s actual suspicion)[39] for the purposes of section 3 of the code a purely objective standard has generally been applied[40] and is reflected in the abandonment of a draft amendment originally canvassed for the 2017 edition.[41]
The distinction between suspects “known” and “not known” acknowledges the rather obvious fact that those stringent protective procedures which must be followed where a suspect is known may be irrelevant or otherwise impracticable to follow where there is no known suspect. However, it has proved less than successful as an instrument for applying the generality to the particular because there are many intermediate cases in which a specific individual will come to the attention of the police as being of interest to them but will remain as yet technically outside the category of being a “known” suspect within the meaning of the Code D definition. This is not uncommonly the situation where shortly after the commission of a crime by a culprit who is a stranger to the victim a person who happens to be in the vicinity is stopped or detained by the police as a potential suspect. At that stage the information or evidence may fall short of providing reasonable grounds for suspicion, sufficient to justify an arrest, but that is not to say that the stop is not an essential preliminary measure or that the police can safely rely on letting the detainee go after taking his name and address (and even, as an addition precaution, a photograph) so that formal identification procedures may be followed up later after contacting the person who was stopped and seeking his co-operation. The name and address may be fictitous and they may never see the person again, even if they do have his picture. Characteristically, where the evidence at the moment of the stop is too weak to warrant a lawful arrest, other sources of information are for the moment elusive, and arrest is the only practicable guarantee of the detainee’s continued involvement in the inquiry, the police will have little choice but to hold an impromptu identification, which will in effect amount to a confrontation between the complainant and the detainee. Such persons are peculiarly vulnerable to error on the part of the witness and, as will be argued later, the procedures for conducting such “street identifications” offer little protection.
It is important to note that generally speaking a person will not amount to a “known suspect” merely because he matches the description of an offender circulated to police officers: Coulman v D.P.P.[42] However, a description coupled with other circumstances may furnish sufficient grounds for reasonable suspicion of guilt. In Nunes[43] a police officer observed a burglar in a house and circulated a description but was unable to arrest him. The area was immediately sealed off and shortly afterwards other officers arrested the appellant, who broadly fitted the description. He immediately denied having done anything and the original officer attended the scene of arrest in order to confirm whether or not he was the culprit and positively identified him. It was held that the identification had amounted to a clear breach of the code.[44]
The difficult judgment as to whether there are reasonable grounds for suspicion may lead an officer into making an honest but mistaken decision to arrest. (Of course, the decision may not be so pious.) If an impromptu identification procedure is then held the prosecution may be faced with the prospect of having to concede that the arrest was wrongful in order to justify a submission that the identification was not.[45] They might then have to argue that the unlawful nature of the arrest ought not to vitiate the fairness of the informal identification to which it led on. Such an argument might well succeed since the informal procedure would in any event have been permissible for the very reason that the arrest was not, and no unfairness would have flowed from it.
Two (or more) people may simultaneously become known suspects for a single culprit[46] although it has been rightly pointed out that as the number or class of potential candidates increases so will any claim that they are all known suspects become more attenuated.[47]
F. First and Subsequent Descriptions
(1) The requirement
Where a witness reports having directly observed a crime or otherwise having seen an event which may indirectly furnish evidence material to a crime the first thing the police will automatically do is ask for a description of the perpetrator or the person who is suspected of involvement. This is not only a basic essential in any investigation but any description given may furnish a useful check against retrospective self-corroboration if and when the witness is called upon to identify a suspect as the culprit or the person originally observed. Apart from retro-racionation, it is self-evidently desirable to get a description down in writing while the mental image of the culprit, or other person previously seen, remains fresh in the witness’s mind. If then any deterioration takes place in the witness’s memory through the normal effluxion of time, or if the original memory becomes corrupted by such intervening suggestive influences as “overshadowing” (discussed below) the original record will be there to validate an identification of a person whose appearance matches the early description. Otherwise, if the memory changes with time and only a late description is recorded but on seeing the culprit during the formal procedure the witness’s original memory is revived an entirely misleading impression of inconstancy will have been created. (On the other hand, consistency is not necessarily a reliable indicator of accuracy, as is mentioned later in this section.) It would not be unreasonable to assume that in the light of these considerations and notwithstanding the Code lacuna, it is routine police practice to record a description as soon as practicable and at a early opportunity.
It was therefore often, though by no means invariably, the case that the police would write down, and preserve, the witness’s description of the culprit. However, there was no mandatory requirement to follow what was after all only an axiom of prudent practice, a deficiency which lingered on in spite of proposals by the Criminal Law Revision Committee in their 11th Report[48] and by Devlin[49] and a call for action by the Court of Appeal in Turnbull.[50] It was not in fact until after the Royal Commission on Criminal Justice in 1993 urged the insertion in Code D of a rule requiring first descriptions in effect to be requested and transcribed[51] that this was finally achieved(in the 1995 revision. Since the requirement may arise at any stage of an investigation it is the one provision in section 3 of Code D which is placed outside the general scheme of categorising rules according to whether or not the suspect is known. D3.1 of the current edition provides(
“A record shall be made of the description of the suspect as first given by the eyewitness. This record must
a) be made and kept in a form which enables details of that description to be accurately produced from it, in a visible and legible form, which can be given to the suspect or the suspect’s solicitor in accordance with this Code, and
b) unless otherwise specified, be made before the witness takes part in any [formal] identification procedures . . . .”
Although the description to be recorded is that of the “suspect” and not of the person who was originally witnessed it is clear that the reference is not to the person in respect of whom identification procedures are to be conducted.[52] D.3.1 concludes by providing that a copy of the record must where practicable be given to the suspect or the suspect’s solicitor before any of the formal identification procedures applicable where the suspect is known arte carried out. In Cole[53] an identification parade was held the day after the incident and the appellant’s solicitor was not informed beforehand of a note of a telephone call made to the police by the witness attending the parade in which she gave a description of her assailant which differed from the description in her witness statement two days later. Although this constituted an obvious breach of D3.1 it was held that an application to exclude the evidence of the identification parade under section 78 of the PACE Act for that code breach on its own would have failed. Had the solicitor been informed of the contents of the note when he should have been it was highly doubtful that the appellant would have been advised to withdraw his consent to participate.[54]
The provision lays down no specific requirement for a description to be requested. It may have been thought unnecessary to command that which is second nature to the police. The first draft of the 1995 revision required records of first description to be made “as soon as practicable,”[55] a requirement which was omitted from the final form and not subsequently re-instated. This omission has previously been described as “mystifying”[56] although it may have been intended to provide bureaucratic leeway for fact that many first-time written reports are in the form of computerised entries which are not necessarily made until some time has elapsed after the description is first given to the police. However, some forces in England and Wales have applied the instruction to mean as soon as possible.[57] In the interests of maintaining best practice uniformly a requirement for first descriptions to be recorded as soon as practicable should be incorporated in the next revision of the code. There is no provision for the witness to sign the description, which may inhibit its use as an aide memoire,[58] but which, again, may be explained by the fact that most crime reports are now computer recorded, often simultaneously with the description being given over the telephone to a police telephone operator, rendering a signature impracticable.
Utility of first description contingent on known appearance of suspect at the time It hardly needs stressing that the evidential value of the witness’s first description is necessarily contingent on the known, rather than the presumed, appearance of the defendant at the time the description is given. In Amin[59] the marked disparity between the description of the robber and the appellant’s appearance when arrested was ascribed by the prosecution to the assumption that he must deliberately have altered his look in the intervening three days. The assumption was negated by police photographs the Criminal Cases Review Commission unearthed showing his appearance on the occasion of three previous arrests for matters on which no action had been taken and his conviction was quashed.
Eyewitness who have participated in a formal identification procedure must be asked if they have seen broadcast or published descriptions Every eyewitness involved in a formal identification procedure must be asked, after they have taken part, whether they have seen any description of the suspect which has been broadcast or published in any national or local media or on any social networking site, and if they have they should be asked to give details of the circumstances, such as the date and place as relevant; their replies must be recorded.[60] This will be important in order to indicate whether the description might have influenced the identification.
(2) Research and common suppositions
(i) Descriptive consistency and identification accuracy
Apart from serving as a potential safguard against the risk that a description given later might have been influenced by the apppearance of the suspect when identified, early descriptions are also widely supposed to serve as a test of consistency with later descriptions given, in a statement or in court. However some research has shown that consistency is a poor predictor of accuracy and that witnesses are commonly inconsistent in their statements but still make accurate identifications.[61]
(ii) Descriptive and identification accuracy
It is also important to note research tending to show that there is sometimes little relationship between the detail and accuracy of the description that an eyewitness can supply of a person and the ability of the witness to identify that person.[62] Good describers are not necessary good identifiers.
(3) General research on accuracy of descriptive detail
It is may also be of importance to summarise the findings under the following heads of certain British research into the accuracy of descriptions of perpetrators furnished by witnesses.[63]
Gender Witnesses are normally accurate at identifying and remembering the offender’s gender.
Age Witnesses are characteristically most accurate when estimating the age of a person of like age to themselves. Some witnesses may have developed a special aptitude for estimating the age of those in a certain age group (for example a secondary school teacher at estimating the age of teenagers).
Height Witnesses tend to be poor at estimating height, with inaccuracy increasing with the difference between their height and that of the offender.
Build Witnesses are often poor at estimating build, with estimates often influenced significantly by clothing, such as baggy trousers or padded jackets. It has ben suggested that witnesses may tend to overestimate the build of violent offenders, but this is far from having been established.
Hair Descriptions of hair colour and length tend to be reasonably accurate.
Race Witnesses tend to be accurate at establishing when someone comes from a different racial background to their own but they are less accurate in describing the specific racial group to which an individual belongs. The confusion may arise partly as a result of confusion in the terms used to describe different racial groupings.[64]
Distinguishing features Witnesses tend to be good at remembering distinguishing features which stand out, such as a gold tooth or a tattoo, and unusual faces are more likely to be described accurately.
Clothing Descriptions of clothing style tend to be reasonably accurate but the accuracy of the clothing colour described will diminish as the light conditions are poorer.
(4) Theory and research on the phenomenon of “overshadowing”
Although it is clearly imperative to take a description of a person who may be the culprit research has found that when eyewitnesses provided a verbal description their performance at a subsequent identification test was liable to be impaired.[65] The phenomenon which is believed to cause this, known as “verbal overshadowing,” is thought to occur because people possess both memories of their verbal description and an additional memory trace of the actual person seen. The former will tend to interfere with—or overshadow—the latter and thus impair identification ability.
(5) Methods of eliciting descriptions
The current consensus among psychologists is that the method of the “cognitive interview,” involving an invitation to free recall by the witness, is apt to elicit more information and more accurate information than by value-laden questioning.[66] The proper application of the cognitive interview requires the luxury of time, a commodity which may not be available to a police operations room telephonist, who will often need information quickly so that it can be circulated to beat or patrol officers, or when officers are setting off with a witness on the look-out for the perpetrator in the immediate aftermath of some street crime. Inappropriate questioning on the hoof can distort or even destroy the processes of mental encoding, storage and retrieval, particularly through overshadowing. Suitable techniques in the training of police telephone operators and patrol officers alike need to be developed and applied, yet although this has been urged from time to time,[67] very little has so far been done to implement such a programme. It follows that Code D contains no provisions designed to avoid the dangers implicit in poor technique.
(6) Taking a description of the witness
As it is tolerably well established that descriptive accuracy is liable to be greater as the person described is closer in description to the witness it has been suggested that it may be helpful to record a description of the witness as a guide to the predictable accuracy of the description of the offender, such an exercise presumably being helpful in the investigation and as a means of assessing the weight to be given at trial to the description.[68]
(7) Admissibility in court
The jury can hardly be prevented from comparing the defendant in the dock with a description of the offender or person it is contended was likely to have been the offender, particularly where the features described are distinctive. In doing so they will obviously make allowance for superficial changes of appearance which may have occurred since the alleged offence, hair length or facial hair changes being the most obvious but substantial ageing if the time lapse has been extensive. In the absence of markedly distinctive features a comparison will go only to consistency of appearance.[69]
G. Use of Photographs
(1) Unknown and known suspects
The process (a) of seeking to discover the identity of potential suspects to a crime, presently unknown, from a description of their physical appearance supplied by witnesses to investigators, and (b) that of establishing from his or her physical appearance that a particular suspect, already known, may in fact be the perpetrator, will in many cases necessarily be dependent on the scrutiny of existing official photographs taken of persons who have previously been investigated by the police for an offence.
The facility of being able to show eye witnesses photographs of potential or actual suspects in order to discover if the person depicted is or may be the perpetrator is made possible by a range of statutory powers which the police enjoy for the taking and retention of photographs of certain categories of person who have previously been arrested or otherwise investigated for offences. These powers are mainly furnished by the PACE Act and Code D for the purposes both of keeping records of such persons and for general investigative purposes where identity based on physical appearance may be in issue. For the sake of completeness in this treatise it is proposed to set out the principle powers.
(2) Statutory powers and Code provisions regarding
the taking of photographs of suspects
(i) Detained persons
Section 64A of the PACE Act authorises the police to photograph any person whilst they are detained at a police station and any person who is elsewhere than at a police station and who has been arrested by a constable for an offence or taken into custody by a constable after being arrested for an offence by a person who is not a constable.[70] Photographs may be taken with the person’s consent or without their consent if consent is withheld or it is not practicable to obtain their consent.[71] The officer proposing to take a detainee’s photograph may, for this purpose, require the person to remove any item or substance worn on, or over, all, or any part of, their head or face; if they do not comply with such a requirement, the officer may remove the item or substance.[72] If it is established that the detainee is unwilling to co-operate sufficiently to enable a suitable photograph to be taken and it is not reasonably practicable to take the photograph covertly, an officer may use reasonable force to take the photograph without consent and for the purpose of taking the photograph, remove any item or substance worn on, or over, all or any part of, the person’s head or face which they have failed to remove when asked.[73] If force is used when taking a photograph a record must be made of the circumstances and those present.[74] For the purposes of Code D, a photograph may be obtained without the person’s consent by making a copy of an image taken at any time on a camera system installed anywhere in the police station.[75] When persons are photographed under the foregoing provisions they must be informed of the purposes of the photograph, the grounds on which the relevant authority, if applicable, has been given and the purposes for which the photograph may be used, disclosed or retained.[76] The information must be given before the photograph is to be taken covertly or obtained without consent by making a copy of an image taken on a camera system installed at a police station.[77] A record must be made when a photograph of a person or any identifying marks found on them are taken and must include the identity of the officer taking the photograph, the purpose of the photograph and the outcome, the person’s consent to the photograph or the reason for taking the photograph without consent and giving any authority, the grounds for giving it and the authorising officer.[78] If force is used when taking a photograph, a record must be made of the circumstances and those present.[79]
(ii) Persons at police stations not detained
The aforementioned provisions for taking photographs apply mutadis mutandis when there are reasonable grounds for suspecting the involvement of a person in a criminal offence but that person is at a police station voluntarily and not detained apply.[80] In the case of persons voluntarily attending the police station force may not be used to take a photograph of the person.[81]
(iii) Photography of identifying marks
The PACE Act permits the photographing of any marks on a person detained at a police station for an offence which would tend to identify them or assist in establishing their identify as a person involved in the commission of an offence.[82] The officer must be of the same sex.[83] Such marks may be photographed with the detainee’s consent, or without their consent if it is withheld or it is not practicable to obtain it.[84] Where the person is voluntarily at a police station and not therefore detained there force may not be used to take photographs of any identifying marks that may assist in establishing the person’s identity, or their identification as a person involved in the commission of an offence.[85]
(3) Statutory provision for the use and retention of photographs
Photographs taken under section 64A of the PACE Act may be used or disclosed only for purposes related to the prevention or detection of crime, the investigation of offences or the conduct of prosecutions by, or on behalf of, police or other law enforcement and prosecuting authorities inside and outside the United Kingdom or the enforcement of a sentence or order made by a court when dealing with an offencer; after being so used or disclosed they may be retained but can only be used or disclosed for the same purposes.[86] The same power of use and retention applies to photographs of identifying marks, save that it will not extend to use for sentencing or other court order when dealing with an offence.[87]
(4) Destruction of photographs
Subject to any requirements under the Criminal Procedure and Investigations Act 1996 for the retention of material in connection with criminal investigations, photographs of persons, or of their identifying marks, which are not taken in accordance with the provisions in Code D, section 5, must be destroyed (together with any negatives and copies) unless the person (a) is charged with, or informed that they may be prosecuted for, a recordable offence, (b) is prosecuted for a recordable offence, (c) is cautioned for a recordable offence or given a warning or reprimand in accordance with the Crime and Disorder Act 1998 for a recordable offence, or (d) gives informed consent, in writing, for the photograph or image to be retained for one or more of the purposes specified by the PACE Act, sections s.54A and 64A, as set out in the previous paragraph.[88] D.3.31 applies these conditions for retention to photographs of suspects (negatives and copies) taken for the purposes of, or in connection with, the formal identification procedures authorised by section 3[89] and D3.30 restricts the use of photographs retained under D3.31 to the purposes enumerated by the PACE Act, sections 54A and 64A.
(5) Admissibility
The admissibility of photographs of suspects is almost as old as photography itself,[90] the image being direct evidence of the scene depicted.[91] Moving films and electronic video and audio recordings are admissible on the same basis[92] and need not necessarily be produced by the person responsible for operating or managing the settings on the recording device. Authentic copies are admissible without regard to the best evidence rule.[93] In the the event that an image or a recording is lost, destroyed or erased a witness who viewed it may attest to what it depicted[94] unless doing would be likely to undermine the fairness of the proceedings.[95] In the present context the witness would be constrained by the difficulties of using words to describe physical appearance. In theory there would be no reason in principle to prevent the witness from participating in the making of a likeness construct from his or her memory of the appearance of the individual depicted in the image. This is largely academic nowadays since digital technology is such that stills are invariably available and that ostensibly erased images can be retrieved or reinstated by experts.
Chapter 2
Suspect’s Identity Not Known
A. Routes Towards Discovering the Suspect´s Identity
(1) Searching the “rogues gallery”
(i) Common practice and the Code
A crime has been committed and although the police may have no one in particular in contemplation as the perpetrator, the nature of the offence, its modus operandi, possibly a psychological profile, and the physical description may have led them to draw up a form of “short list” of candidates from their records of known offenders. To what extent this would entitle them to “round up the usual suspects” for questioning is perhaps moot, given modern susceptibilities. Sometimes the factors are so precise that they can focus on one individual, but, as has already been noted, a selection of persons can nevertheless logically be “known” suspects at the same time even though there is only one culprit.[96]
On the other hand, the police may have no particular suspect to a crime in their sights and little or no idea who might have committed it but may have an eyewitness who, having reported it and having given a description of the culprit, purports to be capable or even confident of making an identification. As an investigative tool it has been the time-honoured practice in such cases to invite the witness to peruse a photographic “gallery” of known offenders to see if the culprit might be amongst them. The precise practice obviously varied regionally but every police force maintained albums of photographs of known offenders categorised in general according to race, gender, age, physical description, and type of offence. Albums kept by divisional commands would necessarily concentrate on offenders who had committed offences in their area (and who possibly also lived in the locality) so that the witness would necessarily be examining photographs of offenders who had committed offences in the same area as the offence under investigation. Perusal would usually be limited to albums containing photographs of offenders matching the general physical description of the culprit, and having a record of committing offences similar in character or method of execution (or both) to the offence under investigation. Any wider trawl might be time-wasting. The use of “hard” albums has now widely given way to computerised presentation by means of DVD compilations made up from data bases of the images of known offenders. This has facilitated the narrowing down of possible suspects to those corresponding to the witness´s description as well as to those within other relevant categories, something which was not practicable in the past when permanent hard albums were used.
Where there is no known suspect D3.3 permits eye-witnesses to be shown photographs, computerised or artist’s composite likenesses or similar likenesses or pictures (including “E-fit” images). The showing of such images must be in accordance with the provisions of Annex E to the Code. Conversely, D3.3 forbids the showing to eye-witnesses of photographs, etc., if (a) the suspect’s identity is known and (b) the suspect is available to participate in a video identification, live identification parade or group identification procedure.[97] Stills may therefore be used if the known suspect is unavailable. It has been suggested that the reason for this is that although the format for this procedure is controlled and structured (as described below) it remains a less transparent test than a video identification procedure and may suffer from a lack of real choice in the physical appearance of those depicted in the photographs.[98] The lack of choice, at least as far as the contribution of any potential suspect is concerned, is certainly implicit in the use of photographs as a tool to detect a hitherto unknown suspect but that is inherent in the nature of the exercise and is hardly any argument against their use where a suspect is known. It is difficult to see why the use of stills as against moving images will necessarily suffer from a lack of transparency or element of real choice.
Although the paragraph would seem to allow the showing of still images where a known suspect is unavailable it has been argued that they ought to be shown in video format in conformity with D3.21.[99]
(ii) Deficiencies revealed by research
Research has shown a negative correlation between the accuracy of identification and the number of photographs inspected.[100] As the number of faces inspected by the witness increases the greater the extent to which those images interfere with the witness’s memory of the culprit’s face, the phenomenon of overshadowing. Similarly, viewing faces in photographs can create some confusion in memory that interferes with the ability to make a subsequent identification.[101]
(iii) Amelioration
A partial amelioration of the shortcomings alluded to above may be to restrict the total number of photographs selected for viewing, rather than allowing open ended search through albums until a face is recognized. This would not necessarily mean fixing an absolute maximum but it it is certainly feasible to reduce the number of photographs which are worthy of inspection.[102] Paring down the gallery to this end would necessitate the witness being asked to provide as precise a verbal description of the facial features recalled as possible. Alternatively, a likeness construct such as photofit or E-fit could be employed as a first stage. Then, using the construct, a trained officer could conduct a trawl through the available albums picking out only those faces which correspond to the description or which resemble the construct. However, the problem with this strategy would be that verbal overshadowing or overshadowing as a result of co-operating in the preparation of a composite can, as discussed below in the section on composites, also interfere with the accuracy of an identification. Since the performance of witnesses in attempting to create constructs tends to be poor it may be best to restrict preparation to concentrating on taking as detailed a verbal description as possible, making the selection from that and imposing a strict limit on the number of images viewed.
(iv) Restricting exposure time to each photograph
Any notion of restricting the number of images viewed would inevitably face official resistance. However, an alternative means of avoiding the problem of overshadowing would be at least to limit the exposure time of witnesses to each image shown, a benefit of the use of DVD. Since, as discussed later, research has shown that accurate identifications tend to be swift, this is a limitation which may serve to reduce the effect of overshadowing without diminshing the comprehensive ambit of the “rogues gallery” trawl.
(v) The problem of implicit revelation of previous convictions
The photograph by which a witness identified the defendant when perusing the photographic gallery of known offenders is admissible as an element of the process of proving identification.[103] This will implicitly convey to the tribunal the fact that the defendant has a criminal record, which can be problematical for the defence if identity is in issue and no application has otherwise been made to adduce a defendant’s previous conviction under the bad character provisions of the Criminal Justice Act 2003. If – as more often than not happens – the witness has gone on to identify the defendant via a standard video identification parade the initial identification will have been superceded and there need be no reference to it at the trial. If, on the other hand, no identification is made at a subsequent formal identification procedure it is unclear if the initial, provisional, identification might still stand to be admissible to support other evidence. The suggestion by one commentator[104] that the failure to make a positive formal identification will nullify an initial one is to be doubted although in Lamb it was held that a police photograph may not normally be led.[105]
(2) Witness participating in the preparation of a likeness construct
(i) Methods employed
For the purposes of seeking out and apprehending unknown offenders it has always been elementary to try to convert a witness’s description of the appearance of the culprit into a visual form. A witness may have the skill himself to draw the offender from memory but more usually the police will employ an artist skilled in portraiture. Through a process of question and answer, linguistic and pictorial prompts, two-way cues and a large measure of trial and error the artist will seek gradually to evolve a sketch concordant with the witness’s memory. In recent times the traditional employment of an artist has been supplemented by manual, paper, systems. These have been devised to enable composite likenesses to be constructed by slotting together pictures of specific facial features chosen by the witness from a selection of many different eyes, eyebrows, hairline, jawline, cheekbones, forehead, noses, mouths, and so on.[106] The first such system to be introduced, Identikit, used drawings of facial features. Later, the Photofit system system used photographs of a large selection of different facial features. More recently, computerised systems were introduced, of which E-fit is perhaps the best known,[107] but which included such other systems as Pro-fit, Mac-a-Mug Pro, and FACES.[108] Once the witness expresses satisfaction as to the likeness it can be shown to other witnesses to see if they recognise the person depicted, or it may be released for publication in the media.
(ii) Constructs to be shown in accordance with Code D, Annex E.
Where the suspect’s identity is not known to the police, showing a witness computerised or artists’ composites or similar likeness constructs or pictures (including ‘E-fit’ images) in order to obtain identification evidence must be carried out in accordance with Annex E.[109]
(iii) Witness performance
Studies have consistently shown that these procedures can be poor at creating a likeness of the offender.[110] Witnesses generally find it difficult to select individual features, even when there is no need to rely on memory: composites constructed with a photograph present throughout the exercise are often no better than composites constructed by memory.
In an experiment designed to test performance, participants prepared composites of former high school friends and teachers (that is, individuals whom they had seen repeatedly over a number of years) and when these were then shown mixed up with foil composites to other participants who had attended the same high schools the accuracy of identifications was extremely poor.[111] The study also showed that confidence in the accuracy of the composite was unrelated to the ability of others to identify the composite from the foils. It has been suggested that when a number of different witnesses are asked to make a composite picture of the same person and the same composite picture specialist is used throughout, the specialist will have gained a preconceived idea of what the offender looks like after creating the first representation and this is likely to bias the way he questions other witnesses and interprets their responses when constructing subsequent composites.[112]
(iv) Developing a new generation of systems
The existing computerised systems are little better than the older manual systems and both systems are less effective than a skilled artist. However, a new generation of facial composite systems are under development. According to the following summary of their operation by Professor Tim Valentine of Goldsmith College, University of London, they―
“make use of a statistical model of facial appearance and a genetic algorithm to converge on the desired facial appearance under the guidance of the witness. Initially a set of random photographic-quality synthetic faces are generated by the software and displayed on the screen. The witness is asked to select the face that most resembles the culprit. The software then generates a new ‘generation’ of faces that use the witness’s selection to guide the search through the range of facial appearance (or ‘face space’). The software also introduces random variation to generate the new set of faces. This process continues repeatedly in an interactive manner. At first the faces are dissimilar and none may look similar to the culprit. Gradually the facial appearance of the culprit is selected for. With each successive generation the facial appearances converge on the desired appearance. Systems under development include Eigen-Fit and Evo-Fit. The great advantage of these systems is that the witness only ever has to look at whole faces and does not have to select individual facial features. This is important because experimental evidence shows that face recognition is an holistic process and people find it very difficult to analyse a face into its constituent parts or features. . . . The new technology allows a number of innovations which are currently under development. A 3D facial model would allow the witness to rotate the head to see a three-quarter or profile view. Hand held devices could be used to allow rapid deployment for a witness to construct a facial likeness at the crime scene, within minutes of reporting the crime. The facial appearance generated under guidance by the witness could be used to search a database of known offenders while the police are still at the crime scene.”[113]
In their early evaluations, Valentine reports, the new systems had not improved on the performance of the older systems. However, something of a breakthrough in the use of computer generated constructs was given national prominence at the end of November 2007 with reports that early in 2008 Derbyshire police were to begin piloting a system based on research by psychologists at Stirling University and the University of Central Lancashire.[114] The research had found that successful matching of a facial composite can be improved sevenfold (from an average of 3.7 to 26 per cent) by presenting the composite as a series of computer generated animated caricatures, exaggerating specific facial features.[115]
(v) Overshadowing
Research has also demonstrated that the phenomenon of overshadowing[116] can undermine the ability to make an accurate identification following participation in the preparation of a mechanical or computer generated likeness. Persons who had attempted an Identikit of a suspect did not perform as well as a control group at a later identification parade and in the two hours it took to create a likeness an immense amount of verbal overshadowing could have occurred to contaminate the memory of the witnesses by the time they participated in the identification test.[117]
(vi) Development of an alternative strategy to eliminate contamination factors
In recent years some police forces have tended to eschew the strategy of inviting witnesses to view images from an electronic gallery of known offenders, preferring instead a practice apparently designed to eliminate the contamination factors which undermine the integrity of a later formal identification procedure where a suspect becomes “known.” In particular it overcomes the problem that where a witness makes a tentative identification as a result of a modern version of the rogues’ gallery search a later formal identification procedure may be no more than an identification of the person previously picked out from the gallery. The new procedure involves a police officer acting in effect as a filter by trawling the central data base for images most closely matching the description of the culprit. The witness will then view the shortlist of images selected by the officer.
(vii) Admissibility
Approval by the witness of the evolved likeness renders it his own description of the person depicted and ousts any issue of hearsay. Such constructs differ from narrative, verbal, accounts in that they are not treated as self-serving or previous consistent statements but are admissible in the first instance in their own right as a discrete type of evidence analogous to a photograph of the culprit, albeit imperfect and created after the event.[118] As with CCTV recordings or stills they may be used by the jury to compare with the appearance of the defendant.[119]
(3) Code D, Annex E provisions
(i) Supervising officer
An officer of sergeant’s rank or above shall be responsible for supervising and directing the showing of photographs although the actual showing may be done by another officer or civilian support staff.[120] It is to be noted that there is no requirement for independence from the inquiry, a reflection of the assumption that with no suspect there is no motive to point the witness in the direction of any individual.
(ii) Confirming first description prior to showing of photographs
Annex E.2 states—
“The supervising officer must confirm [that] the first description of the suspect given by the eye-witness has been recorded before they are shown the photographs. If the supervising officer is unable to confirm the description has been recorded they shall postpone showing the photographs.”[121]
This provision is ambiguous. If the officer is able to confirm that a description was not recorded that is perhaps only another way of stating that he will be unable to confirm that it was recorded, so there can be no showing of photographs until the description is recorded. However, what if it is confirmed that the witness never gave a description? The use of the definite article suggests that the embargo only applies where the description given has not been recorded. Conversely it will not apply where “a” description has not been given; confirmation of non-recording will not arise. However, possibly, or even probably, the intended effect of the paragraph’s rather convoluted wording is that photographs should not be shown unless and until a description has been given and recorded. Perhaps the paragraph is worded on the assumption that a first description would normally be requested at an early stage as a matter of elementary police practice—even in cases where the culprit is well known to the witness. In other words the paragraph assumes that there will always have been a description and the only doubt is whether the record was properly preserved. It certainly ought to be second nature for officers to ask. Since normally after an offence or incident involving a sighting no time ought to be wasted before the witness is shown photographs the paragraph arguably requires the prompt recording of any description.
(iii) One witness at a time to be shown photographs
Only one eye-witness may be shown photographs at any one time.[122]
(iv) Requirement for privacy
Each witness must be given as much privacy as practicable.[123]
(v) No communication between witnesses
During the showing of photographs the witnesses in the case must not be allowed to communicate with each other.[124]
(vi) Not less than twelve photographs to be shown
The eye-witness must be shown not less than twelve photographs at a time, which must, as far as possible, all be of a similar type.[125]
(vii) Caution that photograph of culprit may not be among those shown
When an eye-witness is shown photographs they must be told that the photograph of the person they saw on a specified earlier occasion may, or may not, be amongst them.[126]
(viii) Instruction to indicate if positive identification cannot be made
The eye-witness to whom photographs are shown must be told that if they cannot make an identification they should say so.[127]
(ix) Instruction not to decide until at least twelve photographs have been viewed
When the eye-witness is shown photographs they must be told that they should not make a decision until they have viewed at least twelve photographs.[128]
(x) Witness must not be prompted or guided
The eye-witness must not be prompted or guided in any way but must be left to make any selection without help.[129]
(xi) Positive identification precludes showing of photographs to other witnesses
If a witness makes an identification from photographs, unless the person identified is otherwise eliminated from enquiries or is not available, other witnesses must not be shown photographs.[130] However, both they, and the witness who has made the identification, must be asked to attend a video identification unless there is no dispute about the suspect’s identification.[131]
(xii) Selection followed by inability to confirm the identification
If the eye-witness makes a selection but is unable to confirm the identification the person showing the photographs must ask the witness how sure they are that the photograph they have indicated is the person they saw on the specified earlier occasion.[132]
(xiii) Consequence of identification from prepared composite
When the use of a computerised or artist’s composite or similar likeness has led to there being a known suspect who can be asked to participate in a video identification, appear on an identification parade or participate in a group identification, that likeness must not be shown to other potential witnesses.[133]
(xiv) Notifying suspect or solicitor of previous showing of photographs or composite where formal identification procedure is to be conducted
When an eye-witness attending a video identification (or identification parade or group identification) has previously been shown photographs or a computerised or artist’s composite or similar likeness, the suspect and their solicitor must be informed of the fact before the video identification takes place.[134] It is the responsibility of the officer in charge of the investigation to notify the identification officer that this is the case.[135]
(xv) Obligatory destruction of photographs
None of the photographs shown must be destroyed, whether or not an identification is made, since they may be required for production in court.[136] They must be numbered and a separate photograph taken of the frame or part of the album from which the witness made an identification as an aid to reconstituting it.[137]
(xvi) Documentation
Whether or not a positive identification is made, a record must be kept of the showing of photographs on forms provided for the purpose.[138] This must include anything said by the eye-witness about the identification or the conduct of the procedure, any reasons it was not practicable to comply with any of the provisions of Code D governing the showing of photographs and the name and rank of the supervising officer.[139] The supervising officer must inspect and sign the record as soon as practicable.[140] It is noteworthy that these provisions do not preclude the use of audio or video recording of the showing of photographs, although the common practice is to maintain only a written record. A written record will not pick up nuances of vocal intonation and inflexion or the subtleties of facial expression and body language of the kind which may prove significant in the context of assessing the validity of a formal identification made later and the weight to be given to the witness´s evidence in court. There is therefore a very strong argument for making it compulsory to video-record the procedure, more particularly to compensate for the necessary absence of any prospective defence legal representative.
(4) Chance or unofficially occasioned exposure
to images of a possible culprit
Hitherto in this section we have been examining the position where a suspect is not yet known and the showing of photographs to an eyewitness is officially initiated and supervised with a view to his identifying the perpetrator. By contrast, it may happen that a first identification is made when a relative, friend or other associate of an eyewitness informally and casually shows him a photograph of a person which the witness then claims is the perpetrator, or he otherwise by chance sees the photograph of such a person. Provided the jury be adequately informed of the circumstances in which the photograph was shown or observed and the identification was made there ought to be no instrinsic reason why this should necessarily taint a subsequent formal identification procedure. In Alexander and McGill[141] the complainant made an initial identification from images posted on Facebook and the neglect of the police and the Crown Prosecution Service to obtain the relevant Facebook pages had placed the appellants at a considerable disadvantage when it came to assessing the complainant’s identification of the appellants in a subsequent video identification parade. When an initial identification was made through social media websites it was incumbent on the police and CPS to obtain, in as much detail as possible, evidence of the identification, including the original images. However, the jury had been treated to a detailed account of how the initial identification had come about and knew about the Facebook photographs and it was held that properly directed it was possible for them to take account of the appellants’ disadvantage and to consider carefully the reliability of the identification in those circumstances.
B. Circulation to the Media and to the Police Internally
of CCTV, Other Video Footage and Stills of Incidents
(1) Detecting and tracing suspects
(i) Ubiquity of video surveillance installations and mobile telephones
With the ever increasing spread of closed circuit television (CCTV) surveillance systems the police will often have obtained footage relevant to a particular incident at a very early stage in their inquiries into an offence and this may include viable images of a suspect.[142] Indeed, for a number of years it has been standard practice for the police, at the start of any inquiry, to investigate whether the scene of an offence was covered by a CCTV recording. This will customarily involve house-to-house or shop-to-shop inquiries for relevant imagery of the street recorded on domestic or commercial equipment. On occasions this may yield a high resolution still image of a suspected perpetrator sufficient to furnish proof of identity. Again, with the ubiquity of mobile telephones equipped with video and still cameras it has become a commonplace for members of the public present at the scene of a crime to record crucial episodic evidence.
(ii) General publication
Suspect not known Where the identity of the person shown in the images is not known to the police investigating the incident it is the common practice in the case of offences attracting prominent public interest for images of the suspect to be released to the news media in general or specialist programmes such as the BBC’s Crimewatch, in the hope that members of the public might recognise the person shown. These may be actual photographs or artist’s impressions prepared with the help of an eyewitness or manual or computerised constructs (identikit, photofit or e-fit).
Suspect known Similarly, where the identity of the suspect is known but the suspect is elusive it is not uncommon in high profile cases for the police to release a photograph to the media for the purpose of tracing their whereabouts.
(iii) Internal police circulation
As distinct from public dissemination, recorded moving and still images or likeness constructs are commonly circulated internally within the police service to facilitate possible recognition by officers who may happen to have had dealings with the suspect or to assist in the tracing of known suspects. This will often involve comparison between the seized imagery and official photographs of known offenders held on file.
It is now the ubiquitous practice for one or more officers in a borough division to be employed on the dedicated full-time duty of scrutinising CCTV video imagery of recent local offences, or stills derived from such footage, and comparing it with high quality official photographs of known local offenders in order to try to establish a match. If the video imagery is of sufficiently high quality to furnish a clear image of the offender’s face the case may be treated as one of recognition on the basis that the suspect had become known to the officer from a file photograph which was indisputably his. Out of an abundance of caution the officer could enlist a colleague who local research and intelligence indicates has had dealings with the known offender or is otherwise reasonably familiar with him, to obtain a second recognition.
In other cases, however, the moving imagery may afford a less clear view of the offender’s face and any recognition would necessarily need to take account of his facial and bodily appearance and distinctive movements, gait and gestures. On an initial viewing of the CCTV footage the officer compares it with the photographs of a number of local known offenders. Selecting one who most closely resembles the person on the recording the obvious course would then be to call in a local colleague who knows the offender. That officer would then be invited to scrutinise the images, paying particular regard to the totality of personal characteristics which might conduce to his recognition of the person.
In either case D3.35 will be applicable to require compliance “as far as possible” with the principles for identification by recognition through controlled showing of moving or still imagery in Annex E and the stringent provisions for record keeping laid down in D3.36, as considered in Chapter 1.
It was stressed in Smith (Dean Martin) and others[143] that although a police officer who views a CCTV recording in his official capacity in order to see whether he recognises anyone in it is not in the same position as a witness who is asked to identify a person he has seen committing a crime, the safeguards provided by Code D are of equal importance in such a case. Thus, in particular, where an officer does recognise someone in such a situation, there must be some record which assists in gauging the reliability of that assertion of recognition. Accordingly, it is important that the officer’s initial reactions to the recording are set out and available for scrutiny, including whether he fails to recognise anyone on first viewing but succeeds in doing so subsequently, what words he uses by way of recognition, whether he recognises anyone else, whether he expresses any doubt as to the recognition, and what it is about the image that he says has triggered the recognition; it is therefore vital that a protocol is prepared which provides the safeguard of measuring the recognition against an objective standard of assessment, else there can be no assurance that an officer is not merely asserting that which he wishes and hopes, however subconsciously, to achieve, that is to say the recognition of a guilty participant.
However, recognition by police officers of suspects depicted on images will not always take place in purposive, formal, viewing sessions. In Moss,[144] for example, an off-duty police officer was at the police station to check his emails when he passed a computer screen showing an image taken from CCTV footage of a burglary and recognised the appellant in the still. In such circumstances, the same degree of formal record-keeping as in a purposive viewing session could hardly be expected but it was important for the jury to be able to assess the reliability of the recognition and evidence enabling the jury to do that could be given in different ways. In the present case, the officer immediately reported the recognition to his superior, made a note of it when he was back on duty seven days later, and gave a description of his previous contact with the appellant, and thus an explanation of his ability to recognise him, all of which provided a basis upon which the jury could judge the reliability of his evidence.
Other misconduct implicit in recognition by a police officer Evidence of recognition by a police officer either from direct observation of the offence or from scrutinising CCTV imagery of the incident will of course plainly suggest that the defendant is a known offender. In principle this will not preclude its strict admissibility but circumstances may conduce to its exclusion.[145] On the other hand, if the recognition arises from the officer’s involvement as a police witness in relation to the defendant’s commission of an offence of a kind that would normally lead to the prosecution applying to adduce it under the bad character provisions of the Criminal Justice Act 2003, there would on that account be no reason to exclude the recognition evidence. More generally, however, the problem can in most cases be circumvented by not adducing evidence of recognition but leaving it to the jury to decide the issue on the basis of comparison with the defendant in the dock.[146]
(iv) Part (C) provision for recognition by uncontrolled viewing of imagery
As distinct from Part (B) of Code D, which provides for the obtaining of evidence of recognition by controlled showing of films, photographs and (other) images,[147] the practice of exploratory public dissemination and internal police circulation of imagery are given express sanction by Code D, Part (C) which, as already mentioned in detail, regulates the processes of “recognition by uncontrolled viewing of films, photographs and images.” Specifically, Part (C):
“applies when for the purposes of identifying and tracing suspects, films and photographs of incidents or other images are:
(a) shown to the public (including police officers and other police staff) through the national or local media or any social media networking site; or
(b) circulated through local or national police communication systems for viewing by police officers and police staff;
and the viewing is not formally controlled and supervised in accordance with Part B.”[148]
However, when such material is shown to obtain evidence of recognition, the procedures in Part (B) (evidence of recognition by controlled showing of films, photographs and other images) will apply. The issue of recognition was discussed in part in Chapter 1 and will be further considered in Chapter 3 in the context of the topic of whether there is any useful purpose in undertaking a controlled identification procedure.
(v) Tracing suspects
The reference in D3.38 to “tracing suspects” shows that the purpose of disseminating relevant material under the paragraph is not limited to cases where there is no known suspect but embraces the purpose of discovering the whereabouts of suspects whose identity is already known.
(vi) The phenomenon of “unconscious transference”
Academic and research psychologists have developed the concept of a phenomenon described as “unconscious identity transference,” according to which the image of a proclaimed suspect, usually one nationally televised or otherwise disseminated in the media, may blend with, or even oust, the actual culprit’s face in the memory of a witness who has been exposed to the image and the publicity related to it. At a subsequent identification process it may become impossible for the witness to differentiate between the images in his mind and the result may be that the suspect, wholly innocent, may be wrongly identified.[149] It has been suggested that the requirements to preserve images shown in the media and to ascertain whether the witness has previously seen such images will provide a basis for the court to consider whether a positive identification may have been adversely influenced and contaminated by prior sight of the images.[150] Whether or not unconscious identity transference is empirically verifiable, it is difficult to see how exactly a court might go about using the prior images to determine if the phenomenon has played a part in the instant case. Would trial judges in the English and Welsh jursidiction be prepared to admit theoretical expertise into evidence when it may be no more than supposition? It is a hard enough exercise persuading courts in this country to allow evidence of eyewitness phenomena validated by empirical research.
(2) Action following public release of imagery
(i) Retention of copies of material disseminated in the media
Where “relevant material” has been released to the national or local media for showing as envisaged by D.38, a copy of the material must be kept and will therefore be available for later inspection by the suspect or their solicitor who must be allowed to view such material before any video or other formal identification procedure is carried out, provided it is practicable and would not unreasonably delay the investigation.[151]
(ii) Recording an individual’s substantive response to a viewing
As soon as practicable after a member of the public, police officer or police staff, as the case may be, indicates in response to a viewing that they may have information relating to the identity and whereabouts of anyone they have seen in that viewing, arrangements should be made to ensure (a) that they are asked to give details of the circumstances and (b) subject to the individual’s recollection, that a record of the circumstances and conditions under which the viewing took place is made.[152]
(iii) Establishing if a witness participating in a formal Part (A) identification procedure has previously seen any published or broadcast imagery or description of the suspect
Every eyewitness involved in any of the relevant formal identification procedures in Part A must be asked, after they have taken part, whether they have seen any film, photograph or image relating to the offence or any description of the suspect which has been broadcast or published as described in D3.38(a) and their reply recorded.[153] If they have seen any such imagery, they should be asked to give details of the circumstances.[154]
(3) Use of CCTV footage or stills in court
With the ubiquity of static CCTV cameras and mobile telephones very many prosecutions nowadays are supported by images generated by such technology. These may depict a person alleged to be the defendant actually engaged in committing the offence in question or otherwise present at a location, and evincing behaviour, demeanour and deportment, from which his involvement may reasonably be inferred. Although the present work deals with Code D regulation of identification procedures, it is necessary to refer here to an important rule of court practice: provided the images are of sufficient clarity to be capable of furnishing proof that they depict the defendant juries may reach such a finding by comparing the appearance of the defendant sitting in the dock with the images on screen.[155] Indeed it might very well be futile to try to prohibit them from attempting such an exercise.
The images need not be so clear as to permit the jury to come to a conclusion without expert assistance. As in Clare and Peach[156] this can be provided by an officer with special knowledge of the recording acquired through extensively repeated playing of the recording, both at normal speed and in slow motion and on a frame by frame basis, possibly using highlighting and enhancement techniques. Such ad hoc expert analysis might well involve comparison with other imagery proved or not disputed to be that of the defendant, such as a CCTV recording taken of him in the police station custody suite following his arrest or still images taken at that stage. (Such imagery will also assist if the defendant has significantly altered his appearance since his arrest,[157] although he may have altered his appearance between the original incident and his arrest.)
Apart from reference to control imagery, the expert evidence might make specific reference to those facial features, head and body shape and size and to characteristic gestures and movements which are clearly discernible from looking at the defendant as he sits in the dock.[158] Furthermore imagery showing the depicted person’s gait may be compared with that of the defendant as he makes his way to or from the witness box. Comparison between the contentious imagery and the defendant in court has the considerable advantage that it will obviate any risk of conveying to the jury the fact that the defendant had a police record implicit in evidence of recognition by a police witness.
No invariable requirement for caution as to care In Shanmugarajah and Liberna it was held that there is no invariable rule that when the jury are invited to make a comparison between a contentious image and the defendant in the dock they are required to be warned of the risk of mistaken identification; the modern practice was not to require judges to direct juries as to the obvious and it was a statement of the obvious to direct the jury that they must be careful.[159] The careful scrutiny which is incumbent on the jury when comparing real evidence – that is, images and the defendant in the dock – is obviously of a different calibre than than which is necessary for them to deploy when assessing evidence of what a witness observed and remembered. On the other hand, as the editor of Criminal Law Week succinctly observes, the need for caution is every bit as obvious as in a classic Turnbull case:
“Just as it is notorious that honest and convincing witnesses may be mistaken about their identifications . . . so it is notorious among practitioners in the criminal courts that there are few photographs that could not match several people equally convincingly.” [160]
C. Street Identifications
(1) The common scenarios: extrinsic factors of weakness in identifications
The criminal courts frequently have to deal with cases where the police arrive on the scene within a very short time of an alleged offence and the complainant or an eyewitness is then taken on a tour of local streets to search for the culprit or culprits on the supposition that they may still be in the vicinity. A potential suspect is spotted, is positively identified by the witness, and is duly arrested.[161] A variation on the theme occurs where the suspect has been stopped by police nearby on the basis of a description and the complainant/witness is then brought to the location where the suspect is being detained by the police, effectively for a confrontation, and makes a positive identification.[162]
Another not uncommon circumstance of street identification, as it is called, is that of the complainant or witness who by chance sees the alleged offender in the street or elsewhere on some occasion perhaps weeks or days or even hours after the relevant incident and then calls the police.[163] Purposive, as distinct from fortuitous, street identifications after a lapse of time are occasionally arranged by the police where, for example, the culprit is thought to make the same journey regularly or to frequent the same location, in which case officers will take up a speculative observation with the witness.[164] The surveillance need not necessarily be live. Thus, where there were no known suspects following an outbreak of violence at a public house and the police installed a video system which enabled a witness to make an identification from recordings of people subsequently using the premises, the procedure was apparently treated as analagous to a street identification or group identification.[165]
The imperative of prompt search and capture necessitating these exercises are fundamental to law enforcement and are likely to offer the only chance of bringing offenders to book. On the other hand, the obvious weaknesses implicit in confrontational street identification require a balance to be struck between the detection of offenders and the protection of defendants. That balance is expressed in D3.2, which allows street identification where the suspect’s identity is not known.[166] So long as the accumulated information, then, does not furnish reasonable grounds to suspect the person on whom attention has focused street identification is permissible. Note for Guidance 3F draws attention to the fact that the admissibility and value of street identification evidence may be compromised if the suspect’s identity becomes known before the procedure. Not only may it result in exclusion of the street identification evidence but may have the knock-on effect of vitiating the admissibility of a later formal procedure.[167]
The potential pitfalls in street identification are fairly obvious. The environmental conditions in which they often take place may in themselves undermine their reliability. The lighting at the place of identification may be poor, or no better than at the location of the crime, and the witness may be at a distance from the person to be identified when the identification is made.[168] In any event, there will be little prospect of keeping an unassailable record of these factors. Again, the limited recording facilities available to the police may admit of uncertainty and ambiguity as to the manner in which the purported identification was expressed. If it was by pointing at individuals in a group there is obvious scope for error if the exact target was misinterpreted and not clarified in words. It is always to be hoped that the police will take the greatest pains to ensure fairness and accuracy in their recording of detail but, especially where there is a need to watch out for possible escapes, or to prevent further eruption of public disorder, the ideal is often unachievable.
The very presence in the vicinity of a person who may in broad terms resemble the culprit or one of a group of perpetrators may lead the complainant or witness to presuppose that the person is indeed the culprit, particularly where there are few members of the public in the vicinity at the time and the witness is in an excited and suggestible state. In the experience of practitioners identifications “on the hoof” frequently appear to be influenced by a distinctive article of attire (which may in fact be in common currency) rather than the emphasis on precise physical characteristics such as facial appearance, gait or voice and speech pattern. The importance of clothing is clearly not to be discounted, because the shorter the lapse of time the less likely will it have occurred to the culprit to discard a distinctive item or to exchange it with a confederate in the hope of sowing confusion. On the other hand, the more distinctive and unusual the item the more likely will be the formation of reasonable grounds for suspicion and hence the requirement for a formal procedure.
Again, where a street robbery, for example, has been committed by a group the purported identification of individuals within a group observed afterwards often appears suspiciously to lay stress on group resemblance as between the perpetrators and those pointed out afterwards, in terms of number, approximate age, mode of dress and ethnic origin. It may be that in certain circumstances the sighting of no other persons in the vicinity coupled with the short interval of time between the offence and the stop and group resemblance may by a process of elimination justify the inference that the group is one and the same. By the same token, however, these factors may very well form the basis of reasonable grounds for suspicion against each member of the group, necessitating formal procedures.
Apart from the weaknesses discussed above, street identification involves the problem of influence by the force of suggestion which inheres in its very mechanics. This is discussed next in the context of the measures designed to reduce its impact.
(2) Identification of temporary “detainees” (ie potential
suspects who have been stopped by the police)
(i) The fundamental problem of inherent suggestion
One of the benefits of “all cars, all points” rapid response to street crime and the pervading ownership of mobile telephones is that the police often arrive in the vicinity so quickly and in such numbers that they are able to spot and stop possible suspects on the basis of information received over the air even before there has been any opportunity to take the complainant or other eyewitness on a tour of the neighbourhood to hunt down the culprits.
The fundamental problem with identification of persons who have already been stopped by the police when the witness is brought to the location of the stop is that the very fact that the police will manifestly already have targeted a possible culprit and will be standing very near to the individual in itself carries the strongest taint of suggestion that the person would not have been stopped unless the police had very good reason to suspect guilt.[169] The antidote to the problem provided by the 1995 version of Code D was supposedly set out in the well known paragraph D2.17 which perfunctorily instructed:
“Care should be taken not to direct the witness’s attention to any individual.”
The sentiment was pious but the advice offered little protection against the influence of suggestion which inevitably arises where police officers are necessarily having to stand at close quarters to a provisional suspect. The basic problem is always going to exist of balancing the need to keep a close eye on would-be escapers who are not yet in custody because the police are supposedly constrained by the belief they have as yet insufficient information to justify arrest but who may very soon be in custody after being identified,[170] with that of trying to avoid too obviously “nudging and winking” by the mere fact of physically juxtaposing a single individual or a group of persons with police officers in uniform. A new sense of realism in acknowledging the basic conundrum has been injected into the 2003 edition of the code, which now makes the admonishment that care should be taken not to direct the witness’s attention to any individual conditional on this being unavoidable.[171]
(ii) Statutory attempt to mitigate the problem of inherent suggestion
The measure removing from the code the former absolute prohibition on drawing the witness’s attention to a particular individual where it is unavoidable is contained in a somewhat forlorn attempt[172] to introduce features of a more formalised controlled procedure to be followed where “an eye-witness [is] taken to a particular neighbourhood or place to see whether they can identify the person they saw on a previous occasion.” The paragraph (D3.2) continues:
“Although the number, age, race, general description and style of clothing of other people present at the location and the way in which any identification is made cannot be controlled, the principle applicable to the formal procedures under paragraphs D3.5 to D3.10 [ie. where a suspect is known and is available] shall be followed as far as practicable.”
D3.2 then gives a number of examples which are discussed next, although not necessarily in the order in which they appear in the code text.
(a) First descriptions Paragraph D3.2(a) states(
“where it is practicable to do so, a record should be made of the eye-witness’s description of the person they saw on the previous occasion, as in paragraph 3.1(a), before asking the witness to make an identification.”
As already explained in relation to the general topic of the impotance of taking and recording first descriptions, the obvious purpose of taking such a description as early as possible is to serve as a check against the subsequent intrusion of extraneous influences and a yardstick for testing the accuracy of any subsequent identification.[173] As the Court of Appeal said in Vaughan,[174] “the rule is not mere bureaucracy [but] affords the best safeguard that has so far been devised against the possibility of auto-suggestion.” In other words, it will obviate the problem of determining whether a description subsequently included in a statement was based in fact on an observation made at the time of identification rather than at the time of the offence. It has been pointed out, with some justice, that “[t]he risk of this happening, even quite unconsciously, may be heightened if the observation at the time of the offence was for a short period by a frightened victim or a witness at some distance, and if by contrast the later identification provides a better opportunity to view and digest descriptive information about a person who is perhaps static, closer, or in better lighting.”[175]
The failure to take a first description and to make a note of it before an identification will normally constitute a breach of Code D[176] and may result in the conviction being quashed.[177] However, the code is not breached if it is impracticable to make a record owing to the urgency of the situation. Thus, in “a volatile situation” in which “the primary need is to identify suspects before they disperse”[178] and where there may be insufficient officers available it may be regarded as an “affront to common sense”[179] to require a description to have been taken if this would have allowed the suspect to depart the scene. However, the code does not specify a requirement for a written record and while the exigiencies on the ground may preclude the jotting down of a written note, opportunities on the air, so to speak, will routinely furnish a tape-recording of an initial description given during an initial 999 call and hence an “incontrovertible record that could carry the greatest weight.”[180]
(b) Keeping witnesses apart Paragraph D3.2(c) provides that(
“where there is more than one witness, every effort should be made to keep them separate and witnesses should be taken to see whether they can identify a person independently.”
The precise mischief which the provision is designed to prevent warrants attention in some detail because there may be different priorities and only limited resources. If witnesses are allowed to mingle with the detained person in their sight it is easy to see how a positive identification by one witness, in the hearing of the others, may influence one or more of those others to concur in the identification, whether expressed to the police without solicitation or simply noted mentally. In turn this may lead to a tainted video identification.
However, there might be a problem of unconscious influence even where the potential identification witnesses are kept together but do not have the suspect in their view and witnesses are escorted to a place out of sight of the group in order to see if an identification can be made. The main danger here is that the witnesses might compare notes on the appearance of the culprit or culprits while they are waiting. However, contamination might occur in a more complex way. Under D3.2(d), considered next, once a positive street identification is made by a witness any further informal identification procedures involving other witnesses are prohibited. If, until a positive identification is made each witness in turn is taken to the detained person and then returned to the group before the next is taken, that could lead to the influence of suggestion. A report back by one witness to the others that he believes he has successfully identified the person will hardly be capable of influencing the later judgment at a video identification parade of those who are waiting to be called upon and have not yet seen the detained person. However, those who have already been taken and have failed to make an identification might well be influenced by such a report back if they are later asked to participate in a video identification. (Notwithstanding the failure to make a street identification a witness may nevertheless be invited to participate in a formal procedure if he subsequently expresses confidence in his ability to make an identification or if there is a reasonable chance of the witness being able to do so.[181]) Again, failure by one or more of a series of witnesses may, if reported back to those yet to be called upon, may cause a witness to feel that the police ought not to be “let down.”
The main objective, surely, is not to keep the witnesses separate from each other throughout but, first, as far as possible to impose a regime of non-discussion, and, second, to keep them collectively separate from the temporarily detained person while each witness in turn is taken to that person. The latter the paragraph requires in its insistence on the witnesses being taken to see the person independently (suggesting that before they are so taken they will not have had the person in their sight). Further, it will be important to isolate those who have already participated from those who are yet to participate and, as far as possible, from each other. This can be achieved with relative ease by swiftly taking their name and address after participating and asking them to leave the vicinity as quickly as possible. (Two or more witnesses who are companions could hardly be prevented from meeting up afterwards and travelling home together but the paragraph is little able to avoid that anyway.) It is true that comprehensive implementation of the measure must, as has been suggested, depend on the practicability of chaperoning witnesses separately throughout, in terms of the number of witnesses and available police officers and vehicles and the urgency of the need to identify suspects before they disperse.[182] If there are insufficient numbers of officers to keep witnesses separate from each other before they participate the spirit of the measure can probably be catered for adequately enough first of all by a single officer group chaperoning those who are yet to participate. The officer would have to be politely robust in insisting that they refrain from discussing the description of an culprit or culprits and vigilant in noting down any discussion if it did occur, as required by the code.[183] Second, those who who have completed their involvement should be asked to leave the scene immediately.
(c) Street identification by one witness automatically renders known suspect identification procedures applicable for other witnesses Of self-evident purpose and effect is paragraph D3.2(d), which provides(
“once there is sufficient information to justify the arrest of a particular individual for suspected involvement in the offence, e.g. after a witness makes a positive identification, the provisions [for cases in which the suspect is known and available, ie in paragraph D3.4 onwards] shall apply for any other witnesses in relation to that individual.[184]
(d) Recording action taken New to the 2013 edition of the code and responding to a plea by Mitchell J. in Hickin[185] was D3.2(e), which provides(
“[T]he officer or police [‘civilian support’ in the 2006 edition] staff accompanying the witness must record, in their pocket book, the action taken as soon as, and in as much detail, as possible. The record should include: the date, time and place of the relevant occasion the witness claims to have previously seen the suspect; where any identification was made; how it was made and the conditions at the time (e.g., the distance the witness was from the suspect, the weather and light); if the witness’s attention was drawn to the suspect; the reason for this; and anything said by the witness or the suspect about the identification or the conduct of the procedure.”
The requirement to keep a written record of the events will not preclude the propriety of an extra precaution of making an audio-recording of the manner and form of any identification by the use of a hand-held cassette or digital recorder or a mobile telephone. This would certainly be better than a written record, conveying as it would in addition to a record of the exact words used, the inflection and tone of the witness’s voice. Better still would be the use of a camcorder, which will reveal any tell-tale gestures or facial expressions and the general demeanour of the witness. Where several people have been stopped a global video-recording of the possibly confused events might well prevent later dispute as to whether a person arrested had actually been identified. The entry need not be made contemporaneously and the code can be observed by making a written note from an electronic record at a later time.
(e) The giving of cues on where to look While the need for police officers to stand close to a potential suspect may be unavoidable, it will rarely if ever be necessary to point directly at a particular person and to ask whether he or she is the perpetrator. To do so would clearly be regarded as a leading question, as the Court of Appeal observed in Brizey, although in that case the actual effect of the officer’s words used was held merely to have widened the ambit of the complainant’s attention.[186] D3.2(b), which now adds an important qualification to the old rule D2.17, provides(
“[C]are must be taken not to direct the witness’s attention to any individual unless, taking into account all the circumstances, this cannot be avoided. However, this does not prevent a witness being asked to look carefully at the people around at the time or to look towards a group or in a particular direction, if this appears necessary to make sure that the witness does not overlook a possible suspect simply because the witness is looking in the opposite direction and also to enable the witness to make comparisons between any suspect and others who are in the area . . .”[187]
The passage in emphasis nobly aspires to elevate a procedure which inevitably comes close to being a confrontation to a quasi-identification parade. However, in most cases in a deserted street late at night there will be few citizens about, either walking past or stopping to watch, who may be used as foils. In most cases the only people present who are not in police uniform are likely to be members of a group who have been stopped and all or most of whom may be suspects. It is plainly undesirable that they should be used as foils for each other, particularly when the group may well have been stopped on the basis of a group description. This is reminiscent of the practice of some of the states in the USA, in which it is quite permissible for the police to put together a line-up in which all of the suspects of a crime might be placed together and the witness is invited to select the culprit(a procedure which has been referred to as similar to a multiple-choice examination question with no wrong answer.[188] The sanction in the code for directing the witness’s attention to any individual where this is unavoidable amounts to a sensible recognition of the fact that in general it is in practice impossible to avoid drawing a witness’s attention to an individual who has been stopped at the roadside by uniformed police on mobile patrol. The measure has been criticised as being likely to be taken as a warrant for inviting the witness( expressly, it is assumed(to concentrate on a particular person or group.[189] Strictly speaking and for what it is worth, this seems to be precluded by the generalised, almost deliberately vague, terms of the language which are to be used in inviting the witness where to look. In reality, however, the coy hints sanctioned by the passage may convey a stronger message than a blunt question as to whether the person is the culprit. If a casual identification is going to be permitted where there is no option but to place a potential suspect immediately next to a uniformed officer it is farcical to prohibit an express invitation to participate in what would amount to nothing short of a confrontation. Any officer advancing such an argument might well feel justified in expecting to receive a sympathetic hearing from a court.
(iii) A general solution to the problem of inherent suggestion
Where there are multiple witnesses to a street crime the prohibition on further casual identification procedures as soon as one witness has made a positive identification affords some relief from the risk of multiple error. However, this does not eradicate the problem of auto-suggestion in the case of the witness making the first identification. Moreover, in many instances there will only be one witness to an incident(usually the complainant. While it may useful to hold a second, formal, procedure later, since this may cause an honest witness to have second thoughts when the suspect is viewed against an array of near lookalikes, this hardly offers a guaranteed cure for hasty error. One solution which has been suggested is to require officers standing with a potential suspect to be in plain clothes, where possible, and that uniformed police and marked police vehicles should not be in the immediate vicinity.[190] This supposed antidote ignores operational constraints and, although in many instances plain clothes robbery squad officers will be in attendance, the system should not have to depend on the instant availability of plain clothes officers, or on uniformed officers switching tunics and police headgear with anoraks or sports jackets and back-to-front baseball caps conveniently kept in the boot as stage props, or on the hope that witnesses will fail to notice size 12 police issue boots, or the fact that three diminutive youths are flanked by large well-built men in their thirties dressed slightly self-consciously in the latest designer sports casual wear.
What is a fundamentally very serious problem should be addressed by a vastly more decorous solution than the frankly ridiculous measures necessitated by staging a procedure in the street. Fortunately there is a far simpler and more effective solution made available by modern technology, although not one which may be to the taste of everyone associated with liberal opinion.
In the classic situation, if not in the majority of cases, where the police arriving on the scene in answer to a call within a short time of the offence stop a possible culprit or group of individuals on the street it will almost invariably be on the basis of a matching of the physical description of the culprits with the persons stopped, combined with an assessment of the relative times, location, direction of travel of the temporary detainees, and degree to which the vicinity is devoid of pedestrians. Although it has been held that that a person is not a “known suspect” merely because he or she matches the description of an offender circulated to police officers[191] a combination of the above factors may well furnish reasonable grounds for suspicion justifying an arrest, even without a positive identification.[192] Stops are rarely random or even based on description alone but are usually driven by a combination of circumstances which in themselves will almost certainly raise reasonable grounds for suspicion justifying arrest. If there are reasonable grounds for arrest, even without a positive identification, formal procedures for visual identification are obligatory under the code where identification is disputed (D3.12) and arrest (as distinct from reliance on voluntary attendance at the police station) is likely to be the only practical means of securing the suspect’s co-operation in carrying out these procedures. Where the suspect has said nothing of relevance prior to, or on arrest, D3.12 will arguably necessitate a prompt, formal interview at the police station to determine if identification is likely to be in issue. Where the suspect makes no comment in that short interview there will still be scope for conducting a formal identification procedure if this is regarded as useful (D3.12). It can hardly be denied that with its attendant distress and inconvenience, an arrest would deprive an innocent, albeit suspected, person of the opportunity of being cleared forthwith. This is the understandable attraction of street identification and the police might well seek to demonstrate its efficacy by pointing to the presumably significant numbers of people whom it clears.[193]
On the other hand, arrest and formal procedure will afford a strong measure of protection against a hasty identification which may well be mistaken. Since a formal video identification procedure can now be arranged within a matter of hours, and possibly in under one hour in urgent cases (the speed of the system is described later) the suspect’s stay at the police station need not be protracted and may well result in his elimination from the inquiry where he might otherwise be implicated on the basis of a woefully imperfect procedure. Important also is the fact that video identification permits articles of clothing to be removed from the equation, which may otherwise feature disproportionately in a street identification. Although many suspects might understandably prefer in the short run to take their chances on instant elimination, the interests of justice surely require a more objective, if paternalistic, regard for reliability. If the grounds for suspicion are reasonable enough to justify a stop, then they will probably be enough to justify an arrest, and, if so, there should be an arrest and not a street confrontation. Indeed, the police should be discouraged from declining to arrest ostensibly on the basis of insufficient grounds for suspicion and therefore for arrest, since this may furnish an improper pretext for circumventing the protection of a formal procedure in favour of holding an inherently unsatisfactory casual one. Advocates should be prepared to counter any police claim that they lacked grounds for arrest and to seek exclusion of visual identification evidence under section 78 of PACE if it appears that grounds for arrest clearly existed, irrespective of the professed belief or motivation of the officers.
(iv) Redundancy of argument allowing for street identification of known suspects in cases of special urgency
In the days when live parades were the normal mode of formal identification procedure one of the chief attractions of street identification was that despite all its shortcomings it at least eliminated the factor of fading memory which was always a conceivable concomitant of the delay invariably involved in the organising of live parades, particularly where the number of suspects involved was large. With the availability of a swift formal procedure offered by video identification this can no longer enjoy any validity as an argument in support of street identification. However, prior to the new regime the factor of delay confronted the courts with the possibility of admitting an exception to the normal code requirement for a formal procedure where the suspect was “known.” That possibility emerged in Hickin and others[194] in which, shortly after an attack on two people during the course of a serious outbreak of public disorder, the police arrested a large group of men and then brought two eyewitnesses to the group for the purposes of a street identification, as a result of which seven of the group were identified. As the group had been arrested the Court of Appeal proceeded on the assumption that they were known suspects within the meaning of the code. However, declining to hold either that the code requirement for formal parades applied in the particular case or that the police were therefore in breach for conducting street identifications the court stressed the undoubted impracticability of setting up fourteen identification parades that night and the fact that the consequent lapse of time before parades could be held was liable to cloud the clarity of recollection. The court also placed reliance on the fact that recognition of clothing could be valuable as an aid to identification. (In fact, as the men were under arrest it is difficult to see why their clothing could not simply have been seized for later comparison with the recollection of witnesses[195]). The decision seems implicitly to have recognised an exception to the obligation to hold formal procedures in cases where, because of the large number of suspects, it is impracticable to arrange formal procedures before memories are likely to have faded.[196] However, the difficulty of arranging live parades quickly was always a problem for the police, regardless of the number of suspects.[197] In any event it has been observed that the conundrum for the court could have been avoided had it been argued by the prosecution that the arrested men were too large a group to have been classed as known suspects, but this would have meant conceding that the arrests were unlawful.[198]
(v) Stops initiated by an eyewitness on tour
The problem of auto-suggestion arises too, although less acutely, where the witness is taken on a tour of the vicinity and sees a person on foot who matches the description of the culprit and who, at least viewed from a distance, he thinks he can identify. Here again, the description, coupled with such elements as timing, distance, direction of travel, and the fact that the streets of the locality are deserted(factors which may in themselves induce the witness to believe that the person is the culprit(may give rise to reasonable grounds for suspicion without a positive identification. Questioning on prior movements may enhance the suspicion. If there are sufficient grounds for arrest the formalities of VIPER become obligatory and the witness should be kept at a safe distance and not given a chance to make a close facial examination of the suspect before the latter is taken away.
Chapter 3
Suspect’s Identity Known
A. The Statutory Requirement to Hold a Properly
Controlled Identification Procedure
(1) Evolution to the present rule
Where the identity of the suspect is known there can be no question of the police testing the ability of witnesses to identify the culprit by resorting to any of the relatively casual methods for detecting the identity of the perpetrator necessitated by the lack of information available to them. Instead, they will be required, where appropriate under the code, to employ one of the properly controlled methods prescribed for the purpose of testing a witness’s ability to identify the person in question. The methods which may be used are those set out in in paragraphs D3.5 to D3.10, that is, video identification, identification parade, and group identification.[199] If none of these methods is practicable in the circumstances resort may be had under the code to a confrontation, defined as when “the suspect is directly confronted by the witness.”[200] The question of the selection of the appropriate procedure is examined later but suffice it to say that under the scheme of the current code the method will invariably be video identification, a method described in detail later.
As already mentioned the traditional “live” identification parade was originally introduced in order to avert the risk of miscarriage[201] by supplanting the old and inherently flawed method of confrontation. It was only ever considered to be a screening procedure for exposing uncertainty, not a method of validating recognition scientifically.[202] Moreover, it furnished a useful means by which a defendant who was not picked out could seek to lessen the probative impact of a previous street or other impromptu identification by photograph or likeness construct, or incriminatory evidence of another kind. That it was regarded primarily as a protective procedure was reflected in the scheme of the first edition of Code D which, by requiring a parade in cases involving disputed identification “if the suspect asks for one and it is practicable to hold one”[203] laid the onus for initiating a parade on the suspect, albeit the suspect enjoyed entitlement to it as of right. With the third edition of 1995 the question whether a parade ought to be held shifted significantly from being a matter of the suspect’s right to that of a prosecution duty: “Whenever a suspect disputes an identification, an identification parade shall be held if the suspect consents . . .”[204] In an attempt to “resolve some of the lacunae and uncertainties”[205] which had emerged in relation to this brief formulation it received an elaborate expansion under the 2013 edition of the code, amended in the present, 2017 edition. Set out under the caption “Circumstances in which an identification procedure must be held,” D3.12 provides:
“If, before any identification procedure set out in paragraphs 3.5 to 3.10 has been held
(a) an eye-witness has identified a suspect or purported to have identified them; or
(b) there is an eye-witness available who expresses an ability to identify the suspect,
(c) there is a reasonable chance of an eye-witness being able to identify the suspect
and the eye-witness in (a) to (c) has not been given an opportunity to identify the suspect in any o the procedures set out in paragraphs 3.5 to 3/10 [video identification, identification parade and group identification] then an identification procedure shall be held if the suspect disputes being the person the eye-witness claims to have seen on a previous occasion . . . unless:
(i) it is not practicable to hold any such procedure, or
(ii) any such procdure would serve no useful purpose in proving or disproving whether the suspect was involved in committing the offence, for example:
• where the suspect admits being at the scene of the crime and gives an account of what took place and the eye-witness does not see anything which contradicts that; or
• when it is not disputed that the suspect is already well known to the eye-witness who claims to have recognized them when seeing them commit the crime.”
The reference to an identification procedure reflects the fact that the code allows for the selection from one of three methods of identification, that is, video identification, traditional live identification parade, or group identification, according to an assessment of suitability and practicability in the circumstances, albeit with a preference for video identification. As will be demonstrated later, in practice there are likely to be few if any circumstances in which video identification will not be utilised.
(2) Prior and prospective identification[206]
Read literally, the requirement to hold a parade (or to consider an appropriate alternative) in cases involving disputed identification evidence (first, 1986, and second, 1991, editions) or whenever the suspect disputed an identification (third, 1995, edition) might have seemed to be restricted to cases where there had been a prior visual identification, whether in impromptu form in the street or one made, for example, from a photograph album. However, suspects characteristically do not find themselves under arrest with identification in issue only where an impromptu or photographic identification has been made by an eyewitness. For example, a suspect may very well be arrested for a robbery on the basis of a combination of a matching description and evidence of being found in very recent possession of articles stolen in the robbery or as a result of a fingerprint or DNA trawl. Accordingly, the Court of Appeal had delivered dicta asserting that the wording of D2.3 of the 1991 edition of Code D should be construed to include cases where the witness had indicated that he would be able to identify the person he claimed to have seen or where there was a reasonable chance that he would be able to do so.[207] No steps were taken to incorporate this formulation in the 1995 issue of Code D but the earlier judicial opinion favouring the widened construction was cited with approval in at least two Court of Appeal decisions under the regime of the third edition,[208] although a differently constituted court made no reference to it.[209] The expanded formulation was not incorporated in the 2002 Consultation Draft, the only change of wording being to require an identification procedure “[w]henever a suspect disputes an identification made or purported to have been made by a witness.”[210] The proposed continuation of the narrow criteria attracted criticism for exempting the police from the obligation to undertake a procedure where although they had a witness who asserted an ability to identify the culprit the witness had not actually been required to do so, a situation, it was conjectured, which might occur where the police were seeking to rely on scientific evidence in the belief that it was more reliable, or where they lacked confidence in the witness’s ability to make an identification, to the benefit of the defence.[211] Fortunately the Home Office ultimately responded by including the wider obligation.[212]
The term “suspect” in the prior and prospective conditional clauses of D3.12 is self-evidently a reference to the person the witness claims to have seen on the relevant occasion, the phraseology used later in the paragraph, not the detainee under investigation.[213] The use of the same word in a single paragraph to describe two different persons is undoubtedly clumsy. However, strictly speaking its use to refer to the person professedly seen is more appropriate than “culprit”[214] since in many cases where identification is in issue there will be no direct evidence that the person claimed to have been seen by the witness was the perpetrator of the offence in question. Rather, the prosecution case may be no more than that the person the witness claims to have seen was observed at a certain location in circumstances implicating him in suspicion. The position, then, is that it will be incumbent on the police to undertake a controlled procedure (ie in effect a video parade) whether the prior or prospective identification conditions apply, provided (in either case) that “the suspect disputes being the person the witness claims to have seen.” The converse opinion has, however, been expressed – surprisingly in view of its distinguished source – that the proviso applies exclusively to condition (i), that is to say where “an eye witness has identified a suspect or purported to have identified them prior to any identification procedure . . . having been held.” This is clearly an ill-considered interpretation. Self-evidently the “person the witness claims to have seen” (that is, seen originally) does not mean the suspect the witness purports to have identified later.[215]
The qualification to the prospective identification condition that the witness has not been given an opportunity to identify the suspect in a formal identification procedure (in effect video identification) is evidently intended to exempt the police from the duty of arranging a further identification procedure where the witness has already participated in one but has failed to make an identification and has then subsequently expressed to the police an ability to make an identification, after all. Where, as will invariably be the case, a witness has enjoyed the opportunity of unhurriedly inspecting an array of images but has not made an identification there it may be arguable that the police should not be obliged to incur the time and trouble of giving the witness a second bite at the cherry on the basis of what may well be little more than whimsy. Conversely, the reference to “the suspect” plainly indicates that where the witness has failed to make an identification via a controlled procedure but insists on a continuing ability nonetheless to make an identification and a new suspect comes to the fore the police arguably remain obligated to hold a controlled procedure to test if the witness can identify the new suspect as the person originally seen.[216]
The latter saving may give rise to a potential conflict of interest between suspects. A conundrum suggested to the author of the present work at a workshop for Metropolitan Police identification officers in December 2009 was the occasion for robust differences of opinion. Four men are described as having participated in a robbery but the victim states he would only be able to identify one of the culprits, for whom he gives a description. Four men, all of broadly similar appearance and matching the description, are arrested on suspicion of being the perpetrators and it is proposed to hold video parades for each of them in turn. All four allow capture of their moving images. The first suspect to be taken is positively identified. Would it be useful to test that identification by holding further parades for the other suspects? The luckless first suspect might have an interest in further parades being conducted because of the possibility that the witness might resile from the original identification on seeing the other suspects. On the other hand, he might privately prefer there to be no further parades, thus facilitating his argument at trial that the procedure was manifestly unfair because the witness was given no opportunity to resile from the original identification when faced with the other suspects. For their part they would be likely to prefer to avoid the risk of being identified and would probably object, although the investigating officer would not be bound by the objection. Yet carrying on would be unlikely to damage their positions unduly even if the witness changes his mind and opts for the second, third or fourth suspect. That suspect would be able to argue with some force that the effect of the witness’s change of mind renders the two identifications cumulatively weak, if not wholly valueless. From the point of view of the investigating officer it is in the interests of justice to test the witness by holding further parades. It may be argued that to stop after the first suspect has been positively identified is arbitrarily to cut off an avenue of inquiry, because the police are duty bound to pursue both inculpatory and exculpatory investigations.
Although the two components of the prospective condition, the expression of an ability to identify, and a reasonable chance of the witness being able to do so, are set out disjunctively there will be little option but to read them conjunctively. An assessment as to the second component is clearly contingent on the first, since without knowing whether the witness is confident of an ability to identify the police would hardly have any means of assessing the chances of an identification being made. Conversely, the expression of an ability must surely be subject to an assessment of the probability of success. Where a witness could not see any facial features because of the distance, initially stated that he might be able to identify the person but subsequently conceded that he would not be able to do so, and later gave evidence that the chances were 2 per cent, it was held in effect that the slight chance avowed by the witness did not amount to a reasonable one requiring a parade to have been conducted.[217] Where there is a fundamental disparity between a previous description and the physical appearance of the detainee being investigated, the police might well conclude that there was no reasonable prospect of a positive identification.[218]
(3) Suspect disputes being the person seen
by a witness on the material occasion
The duty which the police bear of arranging an identification procedure is conditional on the suspect disputing that he was the person whom the witness claims to have seen on the material occasion. The requirement is no longer limited to cases where, there having already been an identification, the suspect disputes it, a deficiency in the scheme of the 1991 edition (and continued in the 1995 edition) which, as mentioned above,[219] the Court of Appeal had sought to rectify by construing the wording to include cases where, in short, an identification might reasonably be anticipated.
To dispute that he is the person a witness claims to have seen on a material occasion it is unnecessary for the suspect to know of the existence of such a witness at the time he informs the police that he denies presence at a material place and time. Indeed, it is not even necessary for the police to know of the existence of such a witness at that stage. If the suspect is arrested for an offence on the strength of fingerprint evidence and denies he was anywhere near the scene of the crime and a witness is later traced who claims to have seen a man in the vicinity at the time who may well be the perpetrator and from the description could conceivably be the suspect and the witness states that he would recognise the man again, that would be sufficient, other things being equal, to require an identification procedure.[220]
For the police to be obliged to arrange an identification procedure they must be made reasonably aware of the potential dispute about identification. This would normally be conveyed to them by the suspect on arrest or in interview, by formal notification from the suspect’s solicitor, in a defence statement or by way of notice of alibi.[221] However, it has been repeatedly held that no requirement arises if the suspect makes no request for an identification procedure and gives a “no comment” interview,[222] a proposition which the editor of Criminal Law Week argues amounts to the equating of no comment with an admission and a reversal in effect of the onus of proof.[223]
In other cases the question whether the suspect or defendant is disputing identification may not always be easy to discern and the police may need to consider with care all the information available to them in order to determine whether in all the circumstances the suspect’s account amounts to a disputed identification. In Lambert,[224] where these principles were affirmed, the defendants had either made no comment in interview, admitted to being present at the time of the offence, or admitted limited participation and it was held that it had been reasonable on the facts of the case for the police to conclude that participation, not identification, was the issue, albeit it is perfectly possible for there still to be a serious identification issue notwithstanding that presence at the scene is admitted.
The obligation to hold an identification parade under D3.12 is subject to no time-limit on claiming mistaken identity and the relative ease, expedition and convenience with which a video identification can be organised, as distinct from a live parade, may in practice mean that, other things being equal, even a very late assertion of dispute will oblige the police to arrange it.[225] However, it has been suggested that it would be inconceivable for the duty to arise where a defendant announced a defence of mistaken identity for the first time at trial.[226] This may be compared with the opinion of the Court of Appeal that the question whether a suspect disputes being the person the witness claims to have seen on the relevant occasion, such as to require an identification procedure to be conducted under D3.12, falls for consideration at the time that the police are investigating the offence, rather than in the light of the evidence actually given at trial.[227] It is not clear if the court meant that the obligation under D3.12 ceased at the charging stage but it may be observed that an investigation does not necessarily cease at that point but may well continue up until trial, if not actually throughout the trial.)
It is not clear from the authorities how far the logical distinction between disputing an allegation that the defendant was seen at such and such a date, time and place and expressly putting the prosecution to proof on the issue impacts on the application of D3.12. While it might be thought tactically imprudent for a suspect in interview to go on record throwing out a challenge to the police “to prove he was there,” would such a challenge in a solicitor’s letter or in the defence statement, expressed, no doubt, in decorous terms, engage the provision? The issue of identification, as any issue in the proceedings, will be a live one requiring the police to prove it unless and until it is formally admitted. It will be in issue even if the suspect makes no comment. However, the articulation of a “dispute,” that is expressing the converse, or a challenge, involves something more.
Self-evidently, a positive identification of the suspect by one witness to an offence will not relieve the police of the obligation under D3.12 to hold an identification procedure for the benefit of a second eyewitness where the criteria under the paragraph apply. If this were not required the police could with impunity avert the risk of a second witness failing to make another positive identification. That was the rationale for the rule as explained in Gojra and Dhir where in summing-up the judge had merely suggested that an identification procedure would have been desirable.[228] The conviction was quashed on the grounds that this may thereby have left the jury with the impression that the omission was of little consequence, that the appellant was entitled to the imprimatur of the judge in stressing the fact of a code breach in clear and unequivocal terms, that it was not enough to rely on his counsel to make the point to the jury, and that the judge should have directed the jury as to how the breach had arisen and that it was for them to determine the possible effect of the breach.[229]
(4) Adviser’s role where the suspect’s private instructions involve admitting to presence but he wishes to request a video identification parade
The legal adviser in attendance on a client in the police station may face particular difficulties where the suspect’s instructions are that he was indeed present at the scene of the incident in question. For example, the suspect may be saying that he was acting in self-defence or that he was an innocent bystander, or that while guilty of a relatively minor offence, he did not participate in the main offence alleged (for example that while guilty of assisting in robbery he did not aid and abet the stabbing). On the other hand, he may actually be admitting he was guilty of the main offence. Yet there may be no scientific evidence linking him to the scene and he may be a stranger to the complainant or other identification witness so no question of recognition of a familiar face arises. If without visual identification there is little or no admissible evidence against him he would usually be advised to make no comment and wait to see what the police will do. (At this stage he would certainly not want to make any admission as to presence.) In such a situation, without the suspect disputing he is the person seen by the witness, they may decide to invite him to participate in a video identification parade anyway. Although there is no obligation on the police under the main provision of D3.12 to hold a parade where the suspect gives a “no comment” interview another part of the paragraph makes provision for this sort of situation:
“Such a procedure may also be held if the officer in charge of the investigation considers it would be useful.”
A further difficulty will arise where the allegation for example, is that of a multi-handed attack, the suspect has been arrested near the scene shortly afterwards, he is implicated by a co-suspect and the police are contending on the basis of descriptions that the suspect was the principal assailant. For his part the suspect’s case (as yet disclosed only to the adviser) may be that he played a lesser role or was simply a curious bystander. Without evidence of identification there may be no case against the suspect, or only a weak one, and he would not want in that event to admit to presence as part of a “confession and avoidance.” Such an admission might well provide a basis sufficient to proceed against him in the absence of a positive identification.
The suspect can keep his own counsel and wait to see if the evidence “stacks up.” Confident that their witness may well identify the suspect but faced with a no comment interview the police may choose to set up video parades under the D3.12 subsidiary “useful purpose” rule. Conversely, they may be pessimistic about the chances of the witness making a positive identification (despite an avowal of confidence) and would prefer not to risk a failure by the witness to identify the suspect because now or later it may weaken their hand and strengthen the suspect’s. In these circumstances faced with the police declining to request his participation in a video identification parade but confident that he would not be picked out by the witness because he played only a subsidiary role and on that account was only within the witness’s peripheral vision, he may want to adopt a more proactive approach with the aim of securing his unconditional release without charge, or at any rate to ensure that the case remains weak.
In these circumstances the suspect/client may with good reason be anxious to chance his arm. Advising him not to take the risk may be overcautious and may not be serving his best interests. How should his adviser proceed? It is most important to establish through the disclosure of any descriptions or other evidence whether the police are contending on the basis of any particular eyewitness that the suspect played a specific significant role. If this is confirmed the suspect should then tender a prepared statement in which he simply “disputes being the person the witness claims to have seen” on the relevant occasion. This will accurately, if narrowly, represent his case. Of course, the witness will no doubt also be claiming to have seen other persons who were involved. As long as it is established from the police that the person the witness was clearly referring to was someone who performed a specific role and they are contending that that person was the suspect (the person “the witness claims to have seen”) and not someone else present at the time, he could not properly be criticised at trial for untruthfulness if―disappointingly enough for him―he is then identified as that person. If he is identified he can then give a further statement setting out his account.
(5) Possible applicability of Code D to the issue of participation
in the offence by a suspect whose presence is not disputed
Until the 2011 revision of Code D it had been uncertain whether the police were required to undertake a formal identification procedure where the defendant was not disputing presence when an offence took place but denied any involvement or, while admitting limited involvement, disputed that he was the person who performed a particular role ascribed to that person by an eyewitness.
The potential ascription to a suspect can arise even before any identification has been made, for example where following a group attack the suspect is arrested on suspicion of being the knifeman on the basis of a description and the issue of identification arises because he asserts he was an innocent bystander and that the complainant must be confusing him with the assailant. Less innocently, he may be admitting he was one of a number of men on the periphery encouraging, for example, a “one-on-one” attack with fists and not conceiving that the assault would end in a knife being drawn and used. There was certainly a strong case for contending that the requirement to hold a formal procedure ought to apply in such a case.[230] It has been observed that cases decided under the third (1995) edition of the code (D2.3) indicate that if the possibility of mistake exists, an issue of participation is generally to be regarded as synonymous with an issue of identification and that there is no reason to think that the wording used in the pre-2011 code altered the position.[231]
If, however, there was no realistic possibility of confusion between a suspect who admitted being present and another person who could have been the perpetrator the issue of identification would not seem to arise.[232] The question whether the dispute was one of identification or participation might have depended on the nature of the circumstances of the offence and the period over which it was alleged to have taken place. In Chen and others[233] the police maintained surveillance on a flat in which it was believed victims of kidnapping were being held against their will. The appellants were then arrested in the flat, where they were found with the victims, who were subsequently shown a video recording of everyone entering and leaving the flat, in order to assist them in making statements identifying who exactly had participated in the kidnapping. Although with their arrest the appellants had become “known suspects,” it was held, dismissing their appeals against conviction, that there had been no requirement to undertake formal Code D identification procedures since, in the view of the court, a dispute about roles or about the commission of particular acts in the context of criminal activity sustained up to the moment of arrest was not a dispute about identification but rather one as to participation.
Prior to the 2011 revision, the discretion written into D3.12 allowing that an identification procedure need not be undertaken where “it would serve no useful purpose in proving or disproving whether the suspect was involved in committing the offence” could on the face of it have permitted an identification procedure to be employed for the purpose of ascertaining whether a person indisputably present actually participated or did so in the way alleged by a witness if it were considered useful. However, the Code states at D1.2 in the introduction that its procedures(
“are designed to test the witness’s ability to identify the person they saw on a previous occasion.”
Furthermore, a witness participating in a video identification procedure was (and of course continues to be) required to be asked whether “the person they saw on an earlier specified occasion is present.”[234] It is arguable that these provisions demonstrate that prior to 2011 the identification procedures in Code D were directed exclusively towards the issue of presence at the scene of a crime. For the code to have been applied to the question of participation, as distinct from presence, the witness would have had to be asked whether the person who carried out a particular act on an earlier specified occasion was among the images to be shown, an exercise requiring some adaptation of the terms of the code. This was in fact suggested by the appellants in Chen but rejected by the court as “unrealistic,” although without explanation.[235]
The uncertainty of the code position has now been removed. The 2011 revision has made an insertion in D3.12 which requires an identification procedure where there is dispute as regards the degree of participation, if any, in an incident involving more than one culprit at which the suspect admits having been present. Of those instances where a formal identification procedure would not required because it would serve no useful purpose in proving or disproving whether the suspect was involved in committing the offence the paragraph now gives as an example of such a case the position–
“where the suspect admits being at the scene of the crime and gives an account of what took place and the eye-witness does not see anything which contradicts that.”
The corollary of this is there may be a useful purpose where there is a conflict for example between the eye-witness’s account of how many people were involved when he was robbed in the street and the suspect’s account. Suppose the witness says there were three men present, all of whom played an obvious part. On the other hand, the suspect claims that he was one of four men who were present and that he was one of two who played no active role and were innocent bystanders. In such situations the code now deems it to be useful to conduct an identification procedure to determine if the witness can identify the defendant as one of the three alleged robbers.
Paragraph 3 of Code D is confined in its existing purpose to being something of a blunt instrument for testing the issue of whether a suspect was present. As a means of testing the more subtle question of ascribing a particular role, if any, to a suspect who admits presence, the problem still remains that ad hoc adaptation of the code as presently formulated would lead to an undesirable lack of uniformity.
A suspect who wished to take a chance and challenge a provisional police contention that he was major participant A as distinct from minor participant B could arguably use the stratagem set out in section (4) above, the previous section.
B. Inapplicability of the Mandatory Rule Where an Identification
Procedure will Serve No Useful Purpose
(1) Developments up to the changes of 2002
(i) The narrow construction of the utility principle
Paragraph D2.3 of the 1995 edition of Code D declared the mandatory rule that whenever a suspect disputed an identification, an identification parade had to be held if the suspect consented, unless certain relatively narrow exceptions applied and also if the officer in charge of the investigation considered that it would be useful. The code exceptions did not exempt the police from the duty to undertake any formal identification procedure but permitted the recourse to one of the alternative procedures catered for in the code, group identification, video identification, or confrontation, as the case might be.[236]
In spite of the mandatory terms of the requirement to hold a formal identification procedure it was recognised that there would obviously be some cases of disputed identity in which no useful purpose whatsoever would be served by doing so. Obvious and uncontroversial examples were cases in which witnesses asserted that they would be unable to make an identification or where it was common ground that the witness and the suspect were well known to one another, in which case an identification parade would prove nothing. However, the useful purpose issue arose more controversially in cases in which there had been a prior impromptu street identification and the question which fell to be debated was whether a further formal procedure might add anything of value, since the witness would be viewing an array of persons which plainly included one person whom the witness had already purported to identify as the person seen on the relevant occasion. In Popat,[237] the Court of Appeal favoured the view that a formal procedure would add nothing of value to an unequivocal prior identification. Where a previous “street” identification had been “actual and complete,” having been properly carried out under good conditions with no risk of any corruption of its reliability in accordance with D2.17 of the 1995 edition governing identification where the suspect was not known, the court held that there could be no question but that the purpose of the code in ensuring the accuracy of any identification had been fulfilled, that the requirements of the code had been complied with and that there was no obligation thereafter to hold an identification parade for that witness again to identify the same person. The possibility that the witness would make no identification, select a volunteer, or even declare that the person seen was not one of the array, was regarded in effect as so slight that a formal procedure would serve no useful purpose. Although the court did not in terms say so, an identification parade following an informal identification would in effect be a “farce,” a characterisation originally coined by the Devlin Committee.[238]
(ii) The strict approach
Overturning Popat, the House of Lords in Forbes unanimously upheld a strict application of the mandatory approach, effectively holding that the test of usefulness was not determined by an assumption that the witness would inevitably make a second identification of the suspect.[239] Delivering the opinion of the Appeals Committee, Lord Bingham of Cornhill reviewed in detail the development of Code D through its then three editions, examined the authorities, and concluded[240] that it was not old-fashioned literalism but sound interpretation to read the Code as meaning what it said. The Popat approach, he declared, subverted the clear intention of the code, which was a practical document giving police officers clear instructions. It would replace an apparently hard-edged mandatory obligation by an obviously difficult judgmental decision. That decision would be entrusted to a police officer whose primary concern would (perfectly properly) be to promote the investigation and prosecution of crime rather than to protect the interests of the suspect against mistaken identification. Such decisions would be bound to produce challenges in the courts and resulting appeals. Save in circumstances which were specified or exceptional suspects should not be denied their prima facie right to the protection against the risk of a mistaken identification on the decision of a police officer. In the opinion of the Committee, the Popat approach overlooked the important fact that grave miscarriages of justice had in the past resulted from identifications which were “fully satisfactory,” “actual and complete” and “unequivocal” but proved to be wholly wrong.[241] The decision therefore upheld the oft declared principle that an identification procedure is as much for the benefit of protecting a suspect against mistaken identification as for the benefit of the prosecution.[242]
(2) Continuation of the strict approach under the code regime after 2002
The inutility exception is expressly written into the code but the principle laid down in Forbes, that a previous street identification, however, clear-cut, will not disapply the mandatory rule, arguably remains intact. There is nothing in the drafting of D3.12 which permits a wider, more prosecution slanted, application of the no useful purpose exception than was allowed for in Forbes.[243]
An anomaly removed in 2011 Until D3(2)(d) was amended in 2011 it purported to contain an exception to the general mandatory rule in the case of known suspects which excused the first witness to make a street identification from participating in a subsequent controlled procedure under D3.12. In its entirety D3.2(d) decreed:
“once there is sufficient evidence to justify the arrest of a particular individual for suspected involvement in the offence, e.g. after a witness makes a positive identification the provisions [requiring the formal controlled identification procedures] set out from paragraph 3.4 onwards shall apply for any other witnesses in relation to that individual. Subject to paragraphs 3.12 and 3.13, it is not necessary for the witness who makes such a positive identification to take part in a further procedure.”
In short, the second sentence was saying that no useful purpose would be served by inviting the witness who made the first identification in the street from participating in a later formal procedure. As this clearly went against the grain of the essential thrust of Forbes the anomaly was removed.[244]
(3) The nature of useful purpose
(i) A rationale of utility in cases of prior street identification
The decision in Forbes was based on the assumption that even where there has been an unequivocal street identification (or an identification on inspecting a gallery of still photographs) there will always be a useful purpose in following up with a controlled identification procedure (which will now always be video identification). The opposing position(that there will be no such useful purpose(is predicated on the assumption that witnesses will inevitably identify the person whom they have already identified, even if weeks or even months have elapsed, and so the prosecution case will be advanced no further. It can add nothing to the identification already made, since in objective terms it confirms no more than that the person picked out on the parade was the person who was arrested, notwithstanding that the witness avows certainty in professing to identify the culprit. Not only will it be a “farce,” to use the term coined by the Devlin Committee (albeit slightly out of context), so the argument goes, but it might actually be misleading and unfair in giving the impresssion of an identification of the offender rather than merely that of the person arrested. Conversely, if there was some delay before the controlled procedure was conducted and no identification was made, or only a qualified one expressed, would this reflect anything beyond the mere fact that the memory of witnesses often fade with the passage of time?
The opposing argument, that the holding of a controlled procedure will always be useful because exposure to an array of similar looking persons might cause a witness, whose memory of the appearance of the person previously identified was imperfect, to fail to make a positive identification, to make a qualified one, to pick one of the volunteers instead or even to state that the culprit is not among the array. On balance there are distinct potential advantages for the defendant in a formal procedure undertaken days or weeks after an identification made in the heat of the moment, in spite of the risk that a mistaken identification may be reinforced.
Depending on the circumstances an unqualified failure to make an identification, or even a qualified assertion of resemblance of the suspect to the person seen, may warrant an exculpatory inference. If the witness claims to have had a good view of the culprit and the original street identification was made under good conditions and was expressed in terms of robust certainty, one might expect that degree of certainty to be maintained at least over the course of the next few days, or possibly weeks. Consequently the failure to make an identification under the controlled conditions of a video procedure a mere matter of days later could be taken to imply that the original identification may have been precipitate and the result of auto-suggestion. It is significant that in Bell,[245] a decision in keeping with Popat, it was said that the possibility of non-identification was a “useful purpose.” In other words, “the mere possibility that a witness might fail to pick out someone whom he had purported to identify on a previous occasion ought always to preclude the suggestion that holding an identification procedure in such circumstances would serve no useful purpose in determining whether or not the suspect committed the crime.”[246]
In contrast with an unexpressed failure to make an identification it is always possible that the identification procedure may result in the witness conceding doubts in terms, or even actually exonerating the suspect. Once the immediate excitement of the incident has worn off a witness who has had the time, leisure and opportunity for contemplation, may have become less prone to assume that a person not unlike the offender who was in the vicinity must indeed have been the offender. When confronted with the images of a range of similar-looking persons seeds of doubt may take root as to the accuracy of the original identification. Coupled with the statutory caution warning “that the person they saw on a specified earlier occasion may, or may not” appear in the images they are shown or be present on the parade, as the case may be,[247] it might also be sensible for the identification officer to alert the witness to the danger of confusing the person arrested with the offender, and to the danger of assuming too much from a combination of resemblance and presence in the vicinity at the time.[248] Such warnings assume that witnesses will tend to be honest enough with themselves to ignore amour propre or the embarrassment of loss of face. However, although such factors can never be discounted it is arguable that even without the suggested add-on cautions the defendant will have little to lose and much to gain from a formal procedure. The possibility of a change of mind (and not merely a failure to identify) was expressly referred to in Harris as the reason why it would always be useful from the defence perspective to conduct a controlled procedure, although that was a case of purported recognition rather than a street identification.[249]
(ii) General utility of controlled procedure following positive street
identification now empirically validated and implications
The utility of a controlled procedure following a street identification is supported by field research studies and the result of experiments undertaken by the leading team of psychologists working on the subject.[250] Their research demonstrates that a statistically significant minority of witnesses who made a positive street identification and then went on to participate in a formal video identification procedure failed to identify the suspect whom they had originally identified in the street identification. This is a finding which may be taken unequivocally to support the general utility of a video parade follow-up.
However, the questions as to whose interests may be served by a follow-up video identification and whether the suspect should have the right to require or reject a follow-up procedure are separate issues. The researchers[251] make the telling point that although––
“[t]he data in the field studies and the results obtained in experiments demonstrate that it is not inevitable that a witness will identify the suspect in a video identification procedure where he or she has previously identified that person in a confrontation in the street . . . [o]ur findings suggest that in the great majority of cases a witness who has identified an innocent suspect in a street identification will identify that person again if they are presented in a video identification procedure.”
They concede that it is always open to the defendant to argue at trial that the witness was merely picking him out as the person he originally identified in the street. However, they caution that there is a significant risk that the tribunal of fact will treat it as a stand alone identification corroborating and supporting the initial one in the street and attach to it weight which may not be warranted. For the defendant the conundrum is stark: a real but slender possibility of non-identification against the probability of corroboration but it is not open to defendant to elect. The researchers characterise the crrent mandatory rule as paternalistic, contrasting it with the original 1986 version of Code D which provided for an identification parade upon the defendant’s request and the revised version of 1995 which allowed for one where the defendant consented. Now suspects can avoid a parade only by becoming unavailable, an option which places them at risk of being seen by the jury as obstructive. The researchers urge that as a matter of fairness the risk of corroboration makes it prefereable to leave it up to the defendant to request a follow-up procedure.
(iii) Recognition of a person well-known to the witness
In Forbes Lord Bingham gave two examples of disputed identity cases in which a formal procedure would “very probably be futile.”[252] The first was the eyewitness making plain that he would not be able to identify the culprit or person seen on the relevant occasion (although possibly the clothing[253]). It was subsequently incorporated into what is now D3.12 as the converse of one of the conditions for invoking the mandatory principle. The second example of a case in which a procedure would be futile was that of the “pure recognition of someone well known to the eyewitness.” This too was written into D3.12 but as the single illustration, now, of a case in which no useful purpose would be served by a controlled procedure. In a decision on the 2002 transitional provisions the Court of Appeal could not see what the adjective “pure” (in Forbes) added to the word “recognition” and suggested that it was for that reason that it had been omitted from the revised code in the example given there, that is “where it is not in dispute that the suspect is already well known to the witness.”[254] In fact, well before the 2002 transitional provisions it had been understood that what the Law Lords referred to as a “pure” recognition case meant one in which there was no dispute that the suspect was well-known to the witness.[255] If there is no issue between the prosecution and the defence that the defendant was well-known to the witness the witness would simply be identifying the person who was well known to him, not necessarily the person seen on the relevant occasion and it will be for that reason that no useful purpose would be served by the procedure.
Conversely, where the suspect denies being well known to the witness a formal procedure might be thought to be essential as a means of putting the issue to the test. The facts in Harris[256] are typical of a very commonly experienced sort of case demonstrating the importance for the defence which the courts now attach to a controlled procedure (although the approach of the trial judge was unusual). The two victims of a three-handed robbery in a recreation ground claimed to have been able to recognise clearly and under prime conditions one of the robbers, who was wearing a parka jacket with the hood up, as a former pupil at their school, although and they had not seen him for more than two years, when the appellant would only have been 14. One of the victims also recognised his voice. The complainants stated that they knew him by his first name and the headmaster was called to say that the appellant had been the only pupil with that name at the school during the whole of the material period that the witnesses and the appellant would have been pupils there. From the outset the appellant, whose defence was alibi, contended that since he did not know who the victims were he could not say whether he might know them or not. The police opted to conduct no identification parade and the appellant, despite having legal advice, never requested one. There was no corroborative evidence.
The trial judge ruled that as the case was one of what he referred to as “partial” recognition of a person who was well known to the witnesses (by which it is assumed he meant the claim of familiarity was challenged) no useful purpose would be served by an identification parade. He ruled nonetheless that the failure to hold a parade constituted what he termed a technical breach. He then went on to exercise his discretion to allow in the evidence of identification by the combination of the appellant’s name and the headmaster’s evidence. In so ruling he took into account the fact that the appellant, although legally represented at the time, had never requested an identification parade and was entitled to cross-examine the witnesses as to how well they knew him. Having allowed in the evidence he then directed the jury effectively that there had been no code breach and that no useful purpose would have been served by an identification parade because the case was one of recognition.
The Court of Appeal took the view that having found the police to be in breach of Code D (although, it must be said, for reasons which remained obscure) he had not failed to exercise his proper discretion in allowing in the evidence, no relevant consideration having been omitted nor anything obviously wrong with his ruling. However, quashing the appellant’s conviction they held that he had misdirected the jury in telling them that no useful purpose would have been served by an identification parade, since the case was not one where it was not in dispute that the suspect was already well known to the witness.[257] The jury should have been told that the appellant had been deprived of the benefit of the safeguard under the code of putting the eyewitnesses’ identification to the test. Expressing concern that the failure may have caused the mischief which the code was designed to prevent, Potter L.J., giving the judgment of the court, said:
“The judge appears to have founded his view as to the inutility of an identification parade upon the proposition that, if a parade was held and the appellant identified, it would have carried the matter no further from the prosecution point of view, because it would leave open the question whether the identification was of the appellant as the robber or was merely a perpetuation of the ‘recognition’ of the appellant as a former fellow pupil. There is strength in that view. However, from the defence point of view, it ignores the possibility of a change of mind and/or a failure to identify the appellant at the identification parade, of which possible benefit the appellant was, in the event, deprived. It does not seem to us that, in the circumstances of this case where the ‘recognition’ was of a presumably growing 14-year old, not seen for 2 years, such a possibility should have been discounted.”
Unconscious identity transference While there can be little quarrel with the inutility of a video identification test where it is not disputed that the suspect is well known to the witness the fact remains that a purported recognition in such a case must leave the defendant extremely vulnerable to an assertion which it may be almost impossible to test even by the most skillful cross-examination. However, this is not to discount the possible impact of the conjectural phenomenon of unconscious identity transference, described earlier in the context of images televised or disseminated in the media.[258] From the point of view of defendants it would certainly be agreeable to be permitted to adduce evidence from psychologists vouchsafing the “reality” of such a notion. However, in the courts of England and Wales the frontiers of admissibility of expert evidence are unlikely to be pushed back on this topic. For one thing, the phenomenon is conjectural rather than empirically validated. For another, it adds little to the common experience of mankind, the courts being assailed almost ad nauseam with defence reminders to the jury on how we are all only too familiar with the experience of seeing someone in the street whom we are convinced we recognise as a person we know reasonably well but who turns out to be a stranger.
(iv) Recognition from a photograph
The inutility exception was upheld in Noonan,[259] a case which is analagous to undisputed recognition. In order to keep observation on the appellants police officers had been provided with high quality photographs which indisputably depicted them. The only issue was whether the officers had correctly recognised the appellants from the photographs and a controlled identification procedure would have been valueless. Indeed, it might be supposed that the officers were better placed to make a recognition than in the usual sort of recognition case, as they had been able to study the photographs in briefing sessions immediately before taking up observation.
(v) Hot pursuit capture rendering formal procedure pointless
A further example of the type of case in which a formal procedure would be irrelevant even though identity might well be disputed, is “hot pursuit,” that is where an offender is seen in the act of committing the offence and is then pursued by the witness (usually a police officer) and is captured. If the witness claims never to have lost sight of the suspect clearly there can hardly be any question of an issue of identification. However, even a short break in the continuity of sight may have little consequence if on the alleged facts of the case the inference be irresistible that the person seen after the interruption must have been the culprit under pursuit beforehand. For example, the circumstances in which the alleged culprit is seen to be running after the interruption may make it certain that he was the person being chased before the break, by virtue of a combination of the factors of appearance, direction of travel, isolation of the vicinity, timing, and the hour of night. Again, the suspect may have been found hiding in such circumstances as make it certain that he was the person being chased before the police lost briefly sight of him. It is precisely in relation to such a case that it has been characterised as a “farce” to hold an identification parade.[260]
C. The Exclusive Role of Video Identification in Code D
(1) Distinction between known suspects who are available and unavailable
For the purposes of selecting the appropriate identification procedure to be followed where the suspect is known, Code D distinguishes between suspects who are available to participate in an identification procedure and those not available. It will become apparent that although the code theoretically allows for the use of other, alternative, methods (identification parade, group identification or confrontation) video identification is in practice the only procedure which may properly be employed in either case, although, necessarily, covert methods are likely to be necessary in the case of suspects who are unavailable (a category which embraces unco-operative suspects).
(i) Code definition of availability
Suspects are defined as being available if they are “immediately available or will be available within a reasonably short time, in order that they can be invited to take part in at least one of the eye-witness identification procedures under [D]3.5 to 3.10,” that is, video identification, identification parade, or group identification, “and it is practicable to arrange an effective [such] procedure.”[261]
(ii) Availability within a reasonably short time
It has been rightly asserted that availability within a reasonably short time means in a period beginning with the decision to hold the procedure,[262] a decision which necessarily can only be made when both the suspect has become known and a witness purports to be able to identify the person seen on the relevant occasion. What is reasonable will depend on the particular circumstances, including probably the extent to which delay might prejudice the investigation, as well as the reasons for the suspect’s unavailability.[263]
(iii) Willingness to take an effective part
Until the 2017 edition of the code the definition of availability included a reference to the suspect’s willingness to take an effective part in at least one of the three formal eyewitness identification procedures for which provision was made in D3.5 to 3.10.[264] Normal implementation of the three procedures self-evidently requires the full co-operation of the suspect so to that extent if a suspect declines to participate he will be making himself unavailable. However, the willingness component of the definition of availability was removed for the 2017 edition, presumably because it was appreciated as a matter of strict logic that availability and willingness are mutually exclusive: a suspect can be fully available but determined not to participate. Suspects are entitled to withhold their consent to the procedure,[265] and while they may profess to refuse their consent but agree in practice to co-operate “under protest” it will be their declared and actual refusal to co-operate which demonstrates their unwillingness to take an effective part. (The police will presumably not need to test the suspect’s declared intent by proceeding to find out if in the event he means it.) The most extreme manifestation of a refusal to co-operate will be escape from custody, or failing to surrender to police bail. A suspect who professes to be willing to co-operate with a procedure which is impracticable to arrange or is not suitable will not be considered available.[266]
Applicability to juveniles D2.12 provides that if any procedure in the Code requires a person’s consent the consent of a juvenile is only valid if their parent’s or guardian’s consent is also obtained unless the juvenile is under 14, in which case their parent’s or guardian’s consent is sufficient in it own right. The paragraph provides that if the only obstacle to an identification procedure in section 3 is that a juvenile’s parent or guardian refuses consent or reasonable efforts to obtain it have failed, the identification officr may apply the provisions of paragraph 3.21 relating to known but unavailable suspects. Thus, that paragraph provides in effect that the presumption of unavailability in the case of suspects who have evinced an unwillingness to co-operate will apply to juveniles where the consent of their parent or guardian is either refused or reasonable efforts to obtain such consent have failed.[267]
Mentally disordered or vulnerable persons D2.12 similarly provides that the consent of a mentally disordered or otherwise mentally vulnerable person is only valid if given in the presence of an appropriate adult. There will plainly be similar concerns about the quality of any consent so tendered.
(iv) Change of appearance amounting to unwillingness to take an effective part
The suspect will normally have been informed of any description given by the witness of the person seen on the relevant occasion. Any alteration by the suspect in his appearance between being offered an identification procedure and the attempt to hold one may well be indicative of a desire to undermine the ability of the witness to make an identification. It may even be indicative of a guilty conscience, although that is not a necessary inference of any wilful obstructiveness. (On the other hand, evidence that the suspect’s change of appearance dated from well before his arrest and not long after the commission of the offence, may well be incriminating if it was apparent that the suspect appreciated that he was being sought by the police in connection with the allegation.[268])
Before a formal controlled identification procedure is arranged suspects must normally be informed “that if they significantly alter their appearance between being offered an identification procedure and any attempt to hold an identification procedure, this may be given in evidence if the case comes to trial, and the identification officer may then consider other forms of identification.[269] There is no separate provision expressly authorising such admissibility at trial or such an alternative course in these circumstances but they are implied.
A change in appearance may be so substantial that it will be virtually guaranteed to defeat any attempt by a witness to make an identification using the normal application of video identification procedure, that is using moving images (and certainly not in any event by identification parade, group identification or confrontation). Radical plastic surgery to the face is perhaps the most striking and obvious example. This may well be successful in impeding the inquiry but if there is other evidence sufficient to bring the case to trial the fact that the jury will inevitably be informed of the change in appearance and will be able to observe it with their own eyes may prove to be a sufficient disincentive to any suspect to attempt such a ploy. On the other hand, less radical plastic surgery may be too superficial to preclude at least the reasonable chance of the witness seeing through it and recognising the suspect. Indeed, if the fact of the suspect having undergone surgery is apparent to the witness this may of itself be taken by the witness as the suggestion of an attempt to distract and thus strengthen the witness’s confidence that the recognition is accurate.
Other changes may be less extreme or impermanent but nevertheless striking. The suspect may have been able to achieve noticeable (if not miraculous) weight loss over a short period in an attempt to avoid recognition. While both this and, for example, the removal of a bushy beard and moustache may be a potential distraction, ultimately neither ploy may be potentially decisive, since it is the shape and colour of the eyes which is perhaps the one factor which most distinguishes individuals from each other. However, without meaning to be facetious, theatrical costumiers will always have in stock a supply of false beards and these can be used for the suspect and any foils in order to recreate the suspect’s former appearance. For video identification this would necessitate the recording of special footage (as to which see below) or alternatively, if technically feasible, the recreated image of a beard on each face can be superimposed electronically. Conversely, the growing of a bushy beard[270] may be seen as a facile attempt to distract, but the suspect will always have the option of removing it when issued with the above mentioned warning. A refusal to remove the recent beard growth will not necessarily amount to, or be treated as, wilful refusal to participate; it may be met by the selection of foils sporting beards. Again, a temporarily irreversible change in hair style or length can easily be remedied by issuing hats to the suspect and the volunteers whether for the purposes of a video identification or a live parade.
An attempt by a suspect to change his appearance may certainly indicate the desire to avoid being identified. However, this will not necessarily be tantamount to an “unwillingness to take an effective part” in one of the normal procedures for available suspects, such as to permit resort to be had to an alternative method for unavailable suspects. The attempt may be capable of remedy, with his co-operation, so as to permit one of the standard procedures for available suspects to be used after all. In such a situation the change will not have been significant and the suspect’s belated co-operation will mean that he must be treated as available. If the change is reasonably capable of remedy in order to implement a normal procedure, then it will not be “significant” and the identification officer ought to give consideration to the remedy. It will amount to a failure to comply with the code (a) to ignore the possibility of remedy, (b) not to offer the suspect an opportunity of co-operating in the remedial measure, (c) taking the change of appearance as proof without more of an unwillingness to take a willing part in the procedure, (d) treating the suspect therefore as unavailable, and (e) resorting to an inherently less satisfactory method which may be used when the suspect is unavailable.
If after a full assessment of the possibility of remedy it is considered that the the attempt was successful enough to preclude the use of a normal procedure, the alteration will be significant enough[271] to warrant resort to a special method permissible for unavailable suspects. It will also constitute proof of the suspect’s “unwillingness to take an effective part” in a normal procedure, thus rendering him constructively not available within the meaning of the code definition of that term, and for that reason allowing resort to one of the alternative methods for unavailable suspects. As explained below, in the case of a significant (that is irremediable) change of appearance (as in the example of radical plastic surgery) the special method may be adopted of using a photograph of the suspect taken before the change to make up an array of images in video format.[272] (A known photograph of the suspect may have been obtained from official police records or it may have been seized from the suspect’s personal effects, such as his bus pass.)
(2) Scheme of ranking of methods in the case of available suspects
(i) Notional choice with preference for video identification under the the statutory scheme
Video identification, which made its debût in the 1991 edition of Code D, is defined in D3.5 of the 2017 edition, as one “when the eye-witness is shown images of a known suspect, together with similar images of others who resemble the suspect.” Moving images must be used unless the exceptions set out in the paragraph apply.[273] Under the 1995 edition of the code it seems to have been elevated to rank equally with the traditional live parade in the scheme of selection, but under the fourth (2003) and subsequent revisions, it became the first option to be considered in a range of three available methods. However, any notion of choice is purely theoretical. In reality, it has effectively ousted the live parade and group identification (as well as rendering confrontation entirely otiose) and, with the possible exception of group identification in one particular circumstance,[274] is now the exclusive means of testing eyewitness identification where the suspect is known, whether or not “available.” In the following exposition of why this is the case, reference will be made to some of the features and merits of the video identification systems which are described more fully later.
It is because video identification is now the only formal method for identifying known suspects which the police will be using that it is unnecessary to consider those other methods in this treatise. The other procedures do remain in the code as nominal alternatives but to all intents and purposes they have been rendered otiose. To understand why this is so it is necessary to examine the textual framework of the code. D3.14 provides(
“If, because of paragraph [D]3.12 an identification procedure is to be held, the suspect shall initially be invited to take part in a video identification unless:
a) a video identification is not practicable; or
b) an identification parade is both practicable and more suitable than a video identification; or
c) paragraph 3.16 applies [see below].
The identification officer and the officer in charge of the investigation shall consult each other to determine which option is to be offered. An identification parade may not be practicable because of factors relating to the witnesses, such as their number, state of health, availability and travelling requirements. A video identification would normally be more suitable if it could be arranged and completed sooner than an identification parade. Before an option is offered the suspect must also be reminded of their entitlement ot have free legal advice.”
It is appropriate to set out at this point paragraph D3.16, which provides(
“A group identification may initially be offered if the officer in charge of the investigation considers it is more suitable than a video identification or an identification parade and the identification officers considers it practicable to arrange.”
(ii) Practicability
The scheme of the paragraph is to give primary preference to video identification. The first test(in (a)(of whether it may give way to an identification parade is that of practicability. If it is not practicable in the particular case an identification parade may be considered. For example, if a potential identification witness was near death and there happened to be a disruption in contact with, or in the operations of, the only accessible database large enough to provide a sufficient number of suitable foils, that might well warrant resort to an alternative. However, a conjunction of events ruling out the feasibility of video identification seems almost inconceivable.
Even if video identification is a practical proposition a traditional live identification parade may still be considered under (b). The first, self-evident, condition of such resort is the practicability of such a identification parade, which, the wording appears to suggest, means practicability in terms of witness availability (not the suspect’s convenience). It may be noted that the paragraph does not require a comparison of relative practicability between the two methods. It is enough if both are practicable.[275] The point has rightly been made that with the increased use of video and the infrequent use of parade volunteers, the pool of parade volunteers would probably become too limited for the assembly of a viable parade.[276] On the other hand, a large data base of images of volunteers presupposes the willingness of many of those volunteers to stand on a live parade, if requested and available. If practicability is measured in terms of efficiency video would be sure to come out on top with its very large database of instantly available foils drawn from the whole spectrum of racial and ethnic backgrounds and the facility of rapid retrieval. However such a comparison is not required.
(iii) Suitability: video will invariably be more suitable
The next test(in (b)(of whether video identification may give way to an identification parade is the question of comparative suitability. Given that both methods may be practicable (though not necessarily equally practicable) the police are required to utilise a video identification if it is more suitable than a live parade. The primary test of suitability is set out in the paragraph. Video identification is deemed to be more suitable, in the normal case, if it can be arranged and completed sooner than a parade. It is this provision which now effectively makes video identification the exclusive method to be used in virtually all conceivable cases involving an available suspect.[277] The main reason is that with the easy availability of a comprehensive video library of faces and the comparative simplicity of the arrangements for organising a video identification it will always be possible to hold a session well ahead of a live parade. It is noteworthy that in the Government’s White Paper Justice for All[278] it was stated that the use of video identification could frequently take place within one week (and often less) in contrast to an average time for a parade of 10 weeks.[279] The cancellation rate for video was around 5 per cent compared with 52 per cent for parades. The move to video identification reflected the view that the video library would do the job better than the traditional parade.[280]
That the test of suitability in the paragraph, as between video and live parade, is primarily one of comparative expedition is implicit in the use of the word “normally” but that word would seem to permit other subsidiary tests of suitability, exceptionally, to be considered.[281] The factor of delay comes within the principal test of suitability. Whether the reference to other tests is conditional on it being feasible to arrange video and live parade with equal (or nearly equal) dispatch is not clear. It has been suggested[282] that a live parade may be more suitable for eliminating unusual features borne by the suspect. Thus, the suspect’s distinctive hair can be hidden by issuing hats to all on the lin e-up and a facial scar can be disguised by the time-honoured practice of placing a sticking plaster on the suspect’s scar and similar plasters on the same position of the faces of all the volunteers. The first stratagem can be achieved with video by the routine procedure of recording footage for permanent saving to the database showing each volunteer with and without a standard item of headgear. A similar effect can also be achieved through the electronic technique of pixellation, as can that of superimposing the image of a sticking plaster to the image of the suspect and to each of the images selected for the video array. The argument that the live parade may be more suitable in these circumstances is based on the assumption that pixellation is “in its infancy.”[283] This is surely not the case, as can be attested by the most remarkable exercises in computer generated graphics wizardry which have been achieved for very many years now. It is understood that such disguises are well within the capability of the available programmes used for video identification, as they are within the skills of the operators. (Indeed, they are available on many packages sold for use with home computers.) The standard images used in video identification consist of a view of the head and shoulders and the subject being shown full faced then turning to one side, then the other. If, however, pixellation for the purpose of superimposing the image of a hat or a sticking plaster did prove too difficult for such moving images, the decision can be made, where profile views are not essential, to show only full faced views. Equally, special footage might be taken using real sticking plaster. (The possible use of ad hoc moving images for the purpose of showing full body appearance or posture, gait, movement or gesture is discussed in the next paragraph.)
A more telling reason which has been canvassed for resorting to a live parade is where it is thought that the sight of the suspect’s overall appearance is important or that the witness particularly recalls a characteristic posture, gait, movement, or gesture.[284] Normally, the presentation of video identification images is confined to the head and shoulders of the suspect and foils and so height and build (the latter other than in a marginal way) will be out of account. However, this is not a statutory requirement of the method, merely an operational reflection of what the traditional live parade has essentially always been intended to test, that of the ability to make a facial recognition. It is why foils for live parades were selected by their resemblance to the s uspect in terms of height and build(to eliminate irrelevant but potentially misleading differentials.
This potential shortcoming of standard format video identification can present a particular problem for a suspect whose height is significantly at variance with that described by the witness because in focusing on facial appearance to the exclusion of the consideration of height it will necessarily preclude the possibility that the witness might otherwise be less inclined to make a positive identification on the grounds that the suspect’s height is noticeably and significantly different from that given in the previous description. In their old-time format live parades may have been of real assistance to the suspect in this respect because witnesses were invited to “walk the line” close up. By contrast, in their final form live parades were of relatively little assistance to the suspect on this score because the witness would view the array from some distance away, behind a two-way mirror.
If it were thought so essential to test the ability of the witness to recognize posture, gait or a particular movement of some part of the body in accordance with a description given by the witness, that could assuredly be accomplished by departing from the usual method of preparing images and requiring special footage to be supplied. Thus, the suspect would be video-recorded standing or walking or making the particular gesture requested by the witness, and the chosen volunteers, whose images are held on the data base, could be contacted and asked to attend their nearest identification suite, where, on instructions passed on from the investigating officer, perhaps with the footage of the suspect as a convenient model, they would be video-recorded standing, walking or making the same requested gesture. They would not need to assemble on the same occasion (as in the case of a parade) and, subject to their diligent attendance, the compilation could be made up with reasonable speed, certainly more rapidly than it would take to arrange a live parade. So the advantage of swifter arrangement would be matched by an equal capacity to meet special requests.
The opening words of D3.14 in the fourth edition of 2003 had stated “. . . the suspect shall initially be offered either a video identification or identification unless . . .” The words “either” and “or identification” (with “parade” missing) were included as a vestige of the temporary, 2002, modifications to the third edition in which video identification and identification parade were ranked equally in the hierarchy of selection.[285] It has been suggested that since the preference for video identification is already stated in the hierarchy, the preference for video identification if achievable sooner, as stated in the last sentence of the paragraph, is an “incongruous hangover” from those modifications.[286] On the other hand, express inclusion of the “achievable sooner” test is still essential for underscoring the justification of the rejection of any request by the suspect for a live parade. By itself the provision in D3.11 that “[w]hen an identification procedure is required, in the interests of fairness to suspects and witnesses, it must be held as soon as practicable” would have been inadequate to the task of ensuring that expedition outweighed other considerations of suitability.
Group identification option By contrast with video against live parade determination under D3.14 the choice between video and group identification under D3.16 is determined not by the competing speeds with which they can be arranged but by suitability. Although the minimal requirements of group identification allow it to be held swiftly this affords no advantage against video identification under D3.16. In terms of comparison of persons of similar appearance group identification will be far less suitable than video, as the presence of persons similar in appearance to the suspect will be purely fortuitous. In terms of allowing comparisons of posture, gait, movement or gesture to be taken into account there is no prospect of this being achieved with a group identification whereas video identification can be harnessed with relative ease to furnish full body images of foils all of whom can be recorded walking or making a specially requested body movement or gesture. There is no contest.[287]
(iv) Representations on behalf of suspect as to the choice of procedure
D3.15 provides that
“[a] suspect who refuses the identification procedure first offered shall be asked to state their reason for refusing and may get advice from their solicitor and/or if present, their appropriate adult. The suspect, solicitor and/or appropriate adult shall be allowed to make representations about why another procedure should be used.”[288]
After considering any reasons given, and representations made, the identification officer must, if appropriate, arrange for the suspect to be offered an alternative which the officer considers suitable and practicable.[289] The failure to offer an alternative where there is no impediment in terms of practicability and suitability will be a breach of the code.[290] However, with the scheme of the code and the manifest advantages of video identification, as against the other methods, being what they are, it seems unlikely that there is anything which either the suspect, legal representative or appropriate adult, might say which could be capable of obliging the identification officer to agree to an alternative to video identification.
(v) Selecting the method: a nominal impasse
Reference should be made to the small curiosity of the requirement in D3.14 for the Identification Officer and the officer in charge of the investigation to consult with each other in order to determine, in effect, which method to select. It is not clear what is to happen if they cannot agree or whose view is meant to prevail. The potential impasse between identification and investigating officers will of course never arise because the whole scheme of Code D is cleverly designed to oust the alternatives.[291]
(3) Suspects who are not available (including non-coöperating suspects)
(i) The permissible use of covert action against the suspect
(a) General remarks In the previous section it was demonstrated why it is that, in cases where the suspect is available, out of the three relevant procedural options video-identification will in practice by the sole method of choice. Similarly, it can be asserted that video identification will be the only proper method of identification where the suspect is unavailable[292] (as defined in the code). This is because the police may employ appropriate existing video-footage of the suspect or still photographs presented in a video format to prepare a video identification without the suspect’s consent or co-operation and, if necessary, they may also employ covert recording of moving images or the use of a covertly captured still picture presented in a video format. Covert recording or still-photography may be necesssary where they do not have appropriate existing images or a photograph of the suspect and where, if he were aware of an attempt to film or photograph him he might try to conceal or avert his face. If the unco-operative suspect is accessible enough to be video-recorded or photographed, or the police have a still photograph of the suspect from which a video array of stills can be compiled, this will obviate the need for a resort to the invariably less suitable and usually less convenient method of group identification or the inherently unsatisfactory option of a confrontation. A suspect who is available for these will conceivably always be accessible to be filmed or photographed covertly. Where covert means are necessarily employed the suspect has to be given the choice thereafter of co-operating in making more suitable images. If he chooses not to co-operate the penalty is that he will simply lose the opportunity of being able to make representations over the images to be used and the selection of foils. The significance of this is that the suspect who has not become a fugitive will be unable to escape the procedure, one way or another.[293] In the case of fugitive suspects video identification will also be feasible if the police have moving images or still photographs of them.
(b) Video identification without the suspect’s co-operation The code structure reveals how a suspect who chooses not to co-operate will be on a hiding to nothing. D3.21 stipulates that “[w]hen a known suspect is not available or has ceased to be available . . . the identification officer may make arrangements for a video identification . . . [i]f necessary . . . follow[ing] the video identification procedures without the suspect’s consent using suitable moving or still images and these may be obtained covertly if necessary.”[294]
(c) Difficulty of compiling an array to match covert imagery Clearly it will be impracticable to assimilate the neutral background and appearance and movements of volunteers depicted in standard archival comparator images with that of the covertly recorded footage. The best means of furnishing suitably correspondent moving images would be to choose persons of similar appearance from the database and then to contact the shortlisted volunteers with a request to participate in the compilation of ad hoc footage. This might depict them walking for example either in the same location as that in which the images of the suspect were taken covertly and from the same camera position or in a variety of locations in order to avoid drawing attention to the suspect. With relatively low cost high resolution cameras now widely available good facial shots can easily be obtained from afar under conditions which are less than ideal. Paragraph D3.22 provides that “[a]ny covert activity should be strictly limited to that necessary to test the ability of the eye-witness to identify the suspect as the person they saw on the relevant previous occasion.” This is presumably intended to avoid the use of images which convey a negative impression of the suspect, such as for example slyness, and certainly of footage of the suspect engaged in criminal or other reprehensible conduct or behaviour. The option of proceeding in the face of the suspect’s wilful non-coöperation does not give the police carte blanche to use images which will draw attention to suspect. Annex A.3 states(
“The images used to conduct a video identification shall, as far as possible, show the suspect and other people in the same positions or carrying out the same sequence of movements. They shall also show the suspect and other people under identical conditions unless the identification officer reasonably believes(
(a) because of the suspect’s failure or refusal to co-operate or other reasons, it is not practicable for the conditions to be identical; and
(b) any difference in the conditions would not direct a witness’s attention to any individual image.”
Thus, if it is not possible to show images under identical conditions any differences should not be so marked as to highlight the suspect.
(d) Use of still photographs in video format It is the last mentioned proviso which, in non-coöperation cases, may impel the police to resort to the use of still photographs presented in a sequential video format. The provision for using still images is clearly a reflection of the fact that, notwithstanding the theoretical feasibility of shooting ad hoc footage, as suggested above, it may be inconvenient, difficult or impracticable for the police to match footage of the suspect shot covertly, specifically for the purpose of the identification procedure, or which they otherwise happen to have in their possession, with footage of foils shot in anything like the same circumstances.[295] The reasons why identical conditions are not practicable must be recorded on forms provided for the purpose.[296]
(e) Notification regarding consent and non-coöperation D3.17 provides that before an identification procedure is arranged a number of specified matters must be explained to the suspect. Clearly this will not apply if the need for pre-emptive covert action is properly considered necessary.[297] The information which D3.17 requires to be given to the suspect must also be recorded in a written notice to be handed to the suspect which the suspect must be given a reasonable opportunity to read.[298] The suspect should then be asked to sign a second copy indicating willingness to co-operate with the making of a video.[299] Where the police decide to employ a video identification D3.17 requires suspects to be told that a moving image or photograph may be taken of them when they attend for any identification procedure.[300] They must also be told that they do not have to consent or co-operate[301] but that if they do not consent to, and co-operate in, a video identification, their refusal may be given in evidence in any subsequent trial.[302] That the right not to consent is of little intrinsic value will be obvious to suspects because they must also be told that the police may proceed covertly without their consent or make other arrangements to test whether a witness can identify them.[303]
(f) Pre-emptive covert action D3.20 provides that where the identification officer and the officer in charge of the investigation suspect on reasonable grounds that if the suspect was informed, in accordance with D3.17, that in the event of a refusal to co-operate, the police might proceed covertly, and that on being informed the suspect would then take steps to avoid being seen by a witness in any identification procedure, the identification officer may arrange for images of the suspect suitable for use in a video identification to be obtained before giving the information orally required by D3.17 and the requisite written notice.[304] In other words, if it is suspected that a potentially non-coöperating suspect would take steps to avoid being video-filmed or photographed if made aware of the intention to do so the images are permitted to be recorded covertly before the suspect is alerted to the fact that they are going to be recorded or taken.[305]
(g) Futility of non-coöperation While it is true that the whole scheme is designed to engineer a video identification procedure with or without the suspect’s co-operation it is hard to concur with the alarmist concerns of Roberts and Clover that the scheme somehow involves an inherent lack of fairness to suspects.[306] Coercive it may be but the saving is that the suspect will have the final say because the code provides that if images are obtained covertly following a decision to withhold the notice, ultimately “the suspect may . . . co-operate in providing new images which, if suitable, would be used instead.”[307] D3.17––the paragraph which sets out the information which must be given to suspects before a video identification––furnishes the opportunity for last resort co-operation: it provides, under sub-paragraph (vi), that suspects must be informed that for the purposes of the video identification procedure, images of them have previously been obtained and that they may co-operate in providing further, suitable images to be used instead. The reference to images previously obtained is explicitly coupled with a reference to D3.20, indicating that the images in question are those covert images recorded without the initial D3.17 notice procedure. It is clear, therefore, that before a video identification procedure can take place using covertly recorded images made in the first instance by virtue of the D3.20 exception to the notice requirements the suspect must be given the opportunity to co-operate in providing more suitable images. There can be no obstacle in communicating this to reticent suspects in custody. Again, where the suspect is on bail and is legally represented this can be conveyed through the solicitor. Where this has been done, before proceeding with a video identification procedure based on the previously or covertly recorded images the police should wait to give the suspect an opportunity to make himself available at least “within a reasonably short time.”[308] If the suspect cannot be informed of the option because he fails to surrender on the due date and has failed to keep in touch with his solicitor it may be observed that he will little grounds for complaint if the police proceed with a video identification without his co-operation. On the other hand, bearing in mind that the definition of “available” includes “available within a reasonably short time” it is arguable that the police should not act with indecent haste to conduct a video identification in such a case, and might be advised to await a reasonable time either for the suspect to surrender voluntarily or for any bench warrant to be executed. In the end even those suspects who have initially tried to avoid co-operating enjoy the option of securing a presentation most advantageous to their interests.
(ii) Sources of video images recorded previously or covertly
There will be a practical distinction between images of the suspect which the police may already have in their possession and which may or may not have been recorded with a view to a potential identification issue and footage shot covertly, or non-consensually, for the specific purpose of holding a video identification procedure. The police may routinely have recorded CCTV footage of the suspect being brought into the custody suite of a police station and this might be available subject to matching with suitable corresponding images of foils, although its use in an identification procedure may amount to an infringement of the European Convention on Human Rights if Code D is not materially complied with (see below). Again, they may be able to use CCTV footage of the suspect in a public place as long as it does not depict any activity the adducing of which would involve prejudice to him in the instant case. If they do not have satisfactory images already in their possession when the question of non-coöperation arises, it may be envisaged that they will need to film the suspect surreptitiously, for example, walking or sitting in a public place while on bail or walking up to the police station when answering to bail.
The case of Kennedy[309] is an early example(notably dating from just about the time of the 1991 revision, when video identification was included for the first time(of the use by the police of a covert video-recording in what amounted to a video identification procedure. The appellant, who had refused an identification parade and had thwarted police attempts to effect a confrontation, was video-recorded walking down a passageway in a police station handcuffed to a police officer. The police then made similar video recordings of eight men of similar appearance, similarly handcuffed and the recordings were shown to the witnesses. The court said that this was the(
“. . . fairest possible method which in the circumstances could have been devised. It was certainly a great deal fairer than any sort of group identification . . . [or] confrontation.”
The case illustrates how the use of video identification began as a necessary means of thwarting attempts by suspects to avoid a live parade but became in the course of a decade the standard method of first call.
If resort to the use of still images presented in a video format proves to be necessary there are a number of potential sources of photographs of the suspect. The police may already have in their possession a photograph of the suspect, for example an official photograph contained in a criminal record file, one taken of the suspect during surveillance prior to arrest, a photograph of the suspect found among his personal effects and seized on his arrest for evidential use, or one taken of the suspect while detained at the police station for the specific purpose of investigating an offence.[310] Again, stills may be made up from CCTV time lapse recordings. Finally, while it may be as easy to take covert moving images as it is to take still photographs it may be easier to match a covertly taken still with the stills of foils from the database, than it will be to organise the ad hoc taking of moving images of foils recruited from the database to match in context the covertly taken moving images of the suspect. For this reason it is likely the police will usually rely on stills in a video format where the suspect has refused to co-operate.
(iii) Human Rights implications
Roberts and Glover had predicted that any policy of routinely and systematically recording and storing images of arrested persons in police station custody suites might be susceptible to challenge under the right to privacy guaranteed by Article 8 of the European Convention on Human Rights.[311] So it has proved although it seems that in principle such images may be used provided there is proper compliance with Code D. In Perry v United Kingdom[312] the police covertly used a security camera in a custody suite to record the images of the applicant which were used in a compilation video tape, along with the images of 11 volunteers, to show to witnesses, and which was subsequently shown in open court. The European Court on Human Rights held that this amounted to an interference with the applicant’s right under Art. 8(1) of the convention to respect for his private life, and that the interference was not justified under Art. 8(2) because, while Code D provided a legal footing for the actions of the police, the court at first instance had found three significant breaches in the light of which the measures taken had not complied with domestic law. However, the trial judge had found that the breaches did not adversely affect the fairness of the trial and the application under Art. 6 had been found inadmissible. The decision has been criticised on the basis that the domestic legal rules should have been considered in their entirety, including the case law on when a breach should make evidence inadmissible, which was as much a part of the domestic law as the provisions of the code themselves.[313] Covert filming in a public place can amount to an infringement of the right to privacy even when there was no private element in the events filmed.[314]
(iv) Group identification or confrontation now generally otiose
Where the suspect is unavailable or has ceased to be available the identification officer must decide whether it is feasible to arrange a video identification using covertly recorded images of the suspect. The Code provides that if it is not practicable to do so the identification officer may make arrangements for a group identification.[315] If this option is impracticable the officer may arrange for a confrontation.[316] The power to make use of covertly shot images and the relative ease with which such images may be taken have rendered such alternative procedures effectively otiose. Any attempt to justify resorting to confrontation on the grounds of the impracticability of taking suitable images covertly might be difficult to sustain since it is assumed that a suspect who is available for a confrontation will be available to be filmed or photographed. Almost inevitably the ambience of the setting of the covertly recorded images would be qualitatively different from that of the posed library images of the foils. For example the suspect might be surreptitiously “shot” or “snapped” walking towards the police station when answering his bail. However, the suspect would always have the option thereafter of co-operating in the shooting of more suitable images (see above). If the suspect still refused to co-operate, the police would then be perfectly entitled to use what they had in continuing with the procedure[317] and would have little basis for arguing that a confrontation was employed because it was perceived to be fairer. If the police cannot secure co-operation for video identification they would be unlikely to get it for a group identification and non-coöperating group identification would usually be much less convenient and practicable to arrange than a video-identification employing covertly recorded images. On the other hand, there might conceivably be circumstances in which it might be more practicable to undertake a non-coöperating group identification than to record footage of the suspect. Again, if the lines were temporarily down and the matter were extremely urgent a group identification might be necessary as the first instance choice. However, these highly exceptional circumstances do not warrant consideration in this paper of the provisions of Annex C regulating group identification.
(4) The general requirement for moving images
D3.5 requires the use of moving images for a formal video identification procedure, such moving images being those in the prescribed format recorded when they attend at a police station or other official venue in order to participate in the procedure and after they are given the formal notices required by D3.17 and D3.18. However, introduced into the 2017 revision are two categories of exception to the general requirement in D3.5 for the use of such purposive moving images in a video identification procedure. It will be convenient at this stage to set out the first such exception concerning the issue of a significant change in the suspect’s appearance. We have already considered the problem of suspects deliberately altering their appearance in order to defeat the making of a positive visual identification. What is probably a more difficult question, however, is the alteration of appearance through the effluxion of time, a problem which may well arise in historic cases. Provision has been made to address the problem of significant alteration of appears through the first exception. Thus, moving images recorded for the purposes of the identification procedure need not be used where––
(a) the identification officer, in consulation with the officer in charge of the investigation, is satisfied that because of aging, or other physical changes or differences, the appearance of the suspect has significantly changed since the previous occasion when the eye-witness claims to have seen the suspect;
(b) an image (moving or still) is available which the identification officer and officer in charge of the investigation reasonably believe shows the appearance of the suspect as it was at the time the suspect was seen by the eye-witness; and
(c) having regard to the extent of change and the purpose of eye-witness identification procedures, the identific ation officer believes that such an image should be shown to the eye-witness.
In such a case, the identification officer may arrange a video identification procedure using the old moving or still photograph referred to in (b). However, the suspect must first be given an opportunity to provide their own image or images for use in the procedure although it is for the identification officer and the officer in charge of the investigation to decide whether any images provided by the suspect should be used. A video identification using an old image or one provided by the suspect may, in the discretion of the identification officer, be arranged in addition to, or as an alternative to, a video identification using moving images taken for the purposes of the identification procedure.
D. Discretionary Employment of Controlled Procedure Where
the Investigating Officer Considers it will Be Useful
Absent the conditions specified in D3.12 for a mandatory formal identification D3.13 states:
“Such a procedure may also be held if the officer in charge of the investigation considers it would be useful.”
The procedures referred to are those set out in paragraphs D3.5 to D3.10, that is video identification, identification parade and group identification.
Limited nature of the discretion to invoke the paragraph Although the wording of the paragraph is permissive, it is arguable that the officer’s discretion is limited and that if a procedure were plainly to be of use in the circumstances it would constitute a breach to eschew one for example merely because the suspect had given a “no comment interview”[318]
It is to be noted[319] that whereas D3.14 invokes the selection process when a controlled identification procedure is mandatory under D3.12, it contains no reference to any such selection process where under D3.13 an identification procedure would be considered useful. It is has been argued that this should be regarded as a drafting error and that the scheme of D3.14 should be applied in both cases to avoid the anomalous situation in which the selection process could be ignored if the procedure was not mandatory. In support of the contention that the selection process should apply in both situations reliance is placed on the fact that in the 2002 transitional modifications to the Code D both the mandatory and discretionary procedure were combined into one paragraph, D2.14.[320] The suggestion urged is that in removing the discretionary procedure to its own paragraph the draftsman appears to have forgotten to widen the scope of the selection scheme to both paragraphs.
While there is much force in this argument it may be suggested that D3.13 has been deliberately severed from D3.12 and is not subject to mandatory observance of the detailed rules of compliance for a very good reason. It arguably allows a measure of flexibility of practice under the Code in cases where it may be impracticable to follow the mandatory terms of D3.14 but where a controlled procedure following the spirit of the Code is nonetheless desirable if at all possible. The facts and decision in Folan[321] are illustrative. The appellant’s appearance had changed significantly in the 20 years between his wife’s disappearance and the discovery of her remains during demolition work at a hospital. Accordingly, the police showed his old passport photograph to witnesses who had been involved in building works at the site during the period that his wife had gone missing and who recognised him as having worked at the hospital at that time. It was held that in the unusual circumstances application of the Code was inappropriate but that the witnesses should have been shown a series of photographs.[322]
Applying the mandatory requirement of D3.14 for a video identification to the facts in Folan it might be argued that the police ought to have employed still photographs in a video parade format on the basis that D.3.21 permits stills to be used as such in cases when a known suspect is unavailable. Arguably the suspect as he appeared twenty years previously was no longer available. However, if that argument is considered over-contrived and unpersuasive, a discretionary approach would nevertheless seem to call for a video identification using stills on the basis that since it may be useful and will certainly be more protective it ought therefore to be used.
The investigators in Folan no doubt assumed that after twenty years the suspect’s change of appearance was so dramatic that there was no reasonable prospect that an identification would be made but considered that an attempt might prove fruitful under D3.13. However, the results of a research study by Bruck and others suggest that recognition across the passage of many years may not be anything like as forlorn a prospect as may be supposed.[323] Volunteers were asked to attempt to match high-school graduation photographs with pictures of the same people taken 25 years later, when they were in their early forties. Those unfamiliar with any of the people depicted performed with an accuracy of 33 per cent (compared with 10 per cent which would have been expected from guessing). Parti cipants who had been class-mates were significantly more accurate, with 49 per cent giving correct matches. It may be that accuracy would be less where the time lapse was middle age to old age, when the change in appearance would probably be greater.
Chapter 4
Video Identification Parades
A. Video Identification Systems in Operation in England and Wales
(1) VIPER[324]
(i) History and background
In about 1991, when PACE Code D first included video identification, the West Yorkshire Police Video and Photographic Unit (subsequently renamed the Imaging Unit) began to develop a tape library of volunteer individuals for use in video ID parades. In about 1995 approval was given by the Command Team for an online video identification scheme and the resulting VIPER system (standing for Video Identification Parade by Electronic Recording) and its subsequent developments have been in use since 1996. Although originally developed for use within the West Yorkshire Police it soon became known to other forces in the country and was made available for their use in serious crimes where there was no realistic prospect of holding a traditional identity parade because of the suspect’s unusual appearance. Initially, other forces would bring suspects to a VIPER suite in West Yorkshire and more than half of the forces in England and Wales made use of the facility in this way. The obvious development of this was that VIPER suites were established outside West Yorkshire, thereby eliminating the need to bring the suspect there. With the Code D revision of April 2002, giving priority to video identification, VIPER was “fast tracked out to the ten forces participating in the Street Crime Initiative, when the resource was identified as being particularly useful for identifying street robbers.”[325] The Home Secretary, Mr David Blunkett, officially opened the national VIPER Bureau at Wakefield, West Yorkshire, on 14 March 2003, as of which date the system was being used by12 forces in addition to West Yorkshire and there are plans to extend its use to all 34 forces in the United Kingdom. Including the West Yorkshire catchment a total earlier this year of 58 police sites were equipped with imaging units and there were orders for 20 more. In the 12 months to 25 February, 2003, the VIPER suite delivered 14,991 video ID compilations. Wakefield is now a 24 hour facility.
(ii) The system described
At a local VIPER suite the suspect’s head and shoulders are video-recorded in front of a standardised neutral cloth background using a wall-mounted camera and fixed standard lighting. At the start of the sequence the suspect looks straight towards the camera and then slowly turns the head to show first the right full profile and then the left full profile before again facing the camera. A clip of “footage” lasting 30 seconds is recorded and, using simple commands and prompts, it is transmitted via the central network to the VIPER bureau at Wakefield, where its image quality is monitored. There, a trained editor searches the database for the 15-second clip images of volunteers of similar appearance using keyword descriptions based on age, ethnicity, hair type, build, etc. The editor makes up a selection of still frames from the selected moving images and these stills are then transmitted to the remote identification suite where at least eight are chosen from the selection available and a decision is also made as to the order in which the sequence is to be shown to the witness.[326] The unique reference number for each still is noted (names and details are withheld) and these, together with the sequence choice is sent back to Wakefield. The editor then makes up the requested compilation using the moving image versions of the stills, editing down the footage of the suspect to a 15 second clip to match the comparator clips. The compilation is then downloaded at the remote VIPER suite where it can be burnt on to a CD-ROM or DVD disk, to be shown to a witness on a laptop, or recorded on to a DVD or VHS or other standard cassette tape for court use. The images are shown to the witness in sequence. If a video recording is made of the identification procedure a split-screen sound and vision composite can be prepared for court use showing the images as they are seen and the witness inspecting them.
(iii) Database
Depending on which web site is relied upon the database holds image clips of anything from 6,000 to 10,000 volunteers. Images are collated and quality-checked to ensure that they meet stringent requirements of standardization.[327]
(iv) Staffing and training
The Wakefield VIPER Bureau employs 20 trained editors and 10 other staff. As of May 2003 the centre had trained 700 operators for the forces involved in the scheme, including some 51 training officers.
(v) Funding
The Wakefield Bureau was originally funded by £7.6 million from the Home Office Police Support Unit and Street Crime Action Team, with a further £2.8 million committed for the year 2003/2004.
(vi) The system’s technical specification
The system in use at Wakefield is unique and has been described as Europe’s largest video editing facility. It runs on an industry standard platform of X Series Linux processors and intellstations supplied by Sagitta at a cost of about £1.3 million, which use the IBM General Parallel File System, enable the showing of bandwidth necessary for large file sizes of at least 25MB each, and support 1.4TB of storage space. DigiScope 601AX units with remote controls and two specially modified Hamlet 302 WVR waveform monitors maintain signal quality and ensure the essential probity and consistency of image quality across both library files and the recordings of suspects. The Bureau utilises 34 Pinnacle Liquid silver NLE editing machines of the kind long employed by the West Yorkshire Police Imaging Unit for editing CCTV and video of raids and public disorder incidents. Liquid silver is regarded as the perfect non-linear instrument for making such adjustments as colour correction, essential in the task of eliminating differences between the comparator clips and that of the suspect. Supporting MPEG-2, 1394/DV25 and uncompressed video codes, the system offers the same quality as traditional MPEG systems for 40 per cent less disk space.
(vii) Future of the system
The National Video Identification System is the next level up, exploring the use of images of suspects from other sources such as CCTV to match against the database of known offenders. It will be backed by the Police Information Technology Organisation, which has a strategy of putting links into the system to enable forces to share this information.
(2) PROMAT[328]
In competition with the employment of VIPER, a commercially developed system known as PROMAT (standing for “profile matching”) has achieved adoption so far by over thirty, that is more than half, of Britain’s police forces, pre-eminently the Metropolitan Police. In fact, it has now overtaken VIPER as the principal system in use, its success being a measure of its greater manageability and flexibility, larger database, simpler software, and emphasis on speed. In contrast with VIPER, PROMAT is a “stand-alone” system, the heart of which is a customised, built-in, “clip-shared” database tailored to permit rapid search and retrieval of the volunteer “distractor” images. At the time of writing, the shared database of moving images stood at that of 17,000 volunteers and is continuing to grow with volunteer images supplied by the participating police forces. It includes an extensive range of individuals from ethnic groups. An important advantage of the system is that the preparation of the array from capture of the suspect’s image to completion of a video identification parade can be accomplished if necessary within minutes. The system is compatible with a range of media, whether VCD, VHS, SVHS, DVD, CD, or direct from the compiler’s PC monitor. A recent innovation is the use of facial recognition algorithms (biometrics) for searching the volunteer database. The system facilitates the creation of multiple parades with random or selective positioning for multi-witness incidents. Provision is made for on-line software maintenance and technical support, on-site hardware maintenance and comprehensive training with advanced training for “mosaicing” the removal of scars and other distinctive features. The stand-alone nature of the system allows defence solicitors to make one visit to the identification suite, as the capture and compilation is all carried out on one occasion. In the case of suspects who have deliberately made themselves unavailable to participate in the preparation of a moving image compilation the system allows the use of stills compiled, not from the volunteer data base of moving images, but from a separate database of digital still photographs taken of prisoners in custody suites under the Custody Suite Imaging System. It has been observed that whereas conditions for recording video images is standardized the specifications are less stringent than for VIPER.[329]
PROMAT has obviated the need for the old identification suites, although these continue to be used as an administrative convenience in run-of-the-mill cases. For appropriate cases the Metropolitan Police have pioneered the use of mobile identification units.[330] Witnesses can thus view the compilations in a more relaxed and less intimidating atmosphere (for example, at home, at work, or in hospital) and the units have been been used extensively for conducting parades in prison, avoiding the need to escort the suspect to an identification suite.
For superior integrity 9 compilations can be written to a single CD,with the suspect located in each of the 9 positions on the different parades.This is a facility which had not been available with VIPER and means that the selection of the suspect can be made at the time of the witness viewing, a feature of the system which ensures that none of the ID staff is aware of the suspect position until after the viewing. It could usefully be combined with a procedure using two screens placed back to back, one viewed by the witness, the other viewed by the operator, suspect´s legal adviser any other persons present and showing only the number of the participant appearing on the witness´s screen.[331] The use of CDs as a media can have dramatic cost savings over DVDs or video tapes. The database can also carry on managing line-ups if these are still required. A matrix of pictures can be printed out to allow the suspect and solicitor to make a selection of volunteers, if required. Fiscally, the advantage of the system is that no fees are payable for distractors or for a managed bureau service. Until sufficient images became available under PROMAT, the Metropolitan Police for a time resorted to the VIPER database for an expanded selection, the combined system being known as PROMAT-VIPER.
(3) Approval or veto of images
Practice varies between different police forces across the country as to who makes the initial selection of images and decides on the order in which they are to appear in the compilation.[332] In some forces (e.g. the progressive West Yorkshire Police) suspects and their legal advisers make the selection and decide the sequence order, with the advice and assistance of the identification officer. The Metropolitan Police follow a more prescriptive approach. The initial selection is made by the identification officer and, in accordance with the strict letter of Annex A.7 to Code D (see below), the suspect and his or her legal representative are merely given an opportunity, once the selection has been made, to inspect the complete set of images and to voice any objection.[333]
(4) Merits of video identification
Video identification offers a number of advantages over the traditional live parade.[334] The following are the most important of those which have been described in academic studies and on various web sites on the subject.
(i) Economy
A single live parade cost up to £1,250. In contrast a VIPER download of about 10 video images can cost as little as £150. It will eliminate the need to pay volunteers for each attendance at a live parade. It has been reported that police forces expect to save £7 million in 2003 from the use of video identification as against the use of traditional live parades. It is estimated that 1.2 million policing hours will be saved.
(ii) Convenience[335]
It is an easier procedure than the live parade to arrange and conduct, will reduce the burden of marshalling where there are many witnesses, which might require the police to muster and organise the witnesses and perhaps a great many volunteers on a given day, will excuse the need for witnesses to travel long distances and will spare witnesses who are in poor health from the need to attend a live parade.[336]
(iii) Speed
On average live parades took anything between 6 and 10 weeks before they could be held. By comparison the whole process of video identification can take less than 3 hours. Indeed a service level agreement requires preparation and delivery within 2 hours. Urgent cases can be reduced to 1 hour and the fastest recorded took a mere 15 minutes. This degree of speed means that the procedure can be conducted while a witness’s memory is fresh, an important consideration given empirical evidence that a witness’s ability to make an accurate identification diminishes as the interval between the original sighting and the identification procedure lengthens.[337] The failure to make an identification after such an early opportunity to participate in an identification procedure, or a tentative or qualified identification, may well be of considerable value to a defence case of mistaken identification.[338] The facility of speed will also allow a procedure to be conducted before any decision is made about granting or withholding bail, a consideration which may be relevant on the issue, referred to earlier, as to whether the police should eschew street identifications in favour of arrest.
(iv) Elimination of delay through cancellations
Reference has already been made to the fact that in the Government’s White Paper Justice for All it was stated that the cancellation rate for video was around 5 per cent compared with 52 per cent for live parades.[339] Cancellations were largely the result of the failure of a bailed suspect or of the witness to attend or a lack of suitable volunteers on the appointed day.
(v) Extensive range of available comparator images
For the purposes of conducting a live parade finding a sufficient number of volunteers, who in the words of the long established rule, “as far as possible resemble the suspect in age, height, general appearance and position in life,”[340] was always problematic.[341] In contrast, the instant availability of a very extensive library of images facilitates refinement of the selection process to the point at which the foils are as near to being clones of the suspect as it is possible to achieve in humanity.
(vi) Absence of extraneous influences caused by inappropriate volunteer demeanour
With live parades it has long been recognised that the behaviour of the volunteers may consciously or unconsciously serve as a cue in drawing the attention of the witness to the suspect.[342] A graphic instance of such a factor occurred in the notorious case of Major R.O. Sheppard, who in June 1925 was the victim of an allegation of theft based on mistaken identification.[343] Police officers present at the parade all maintained that the witness in question picked out Major Sheppard without hesitation or encouragement but at the later tribunal of inquiry into the case the witness himself said that he had not been sure who the man was and it was only through the “hypnotic influence of the police and the volunteers, all of whom were looking towards” Major Sheppard, that he eventually touched him.[344] With a video array this of course will not be an issue.
(vii) Elimination of unusual features
The suspect may bear an unusual and distinctive facial feature, such as scarring, a birth mark, an area of skin discolouration or other blemish, a deformity or injury, the absence of eyebrows through hair loss, hair of an unusual syle or colour, or a tattoo. While it may not be difficult to find volunteers who in other respects resemble the suspect it will usually be imposible to find those who share the particular feature as well. It will obviously be essential to eliminate the feature from the equation in the interests of achieving a viable degree of resemblance so as to ensure fairness. It has long been the practice with live parades to eliminate the feature, where possible, by placing a sticking plaster over the mark on the suspect’s face and plasters in same position on the faces of all the volunteers. In the case of unsual hair, this may be achieved by having everyone on the parade wear an item of headgear to disguise the distinctiveness of the suspect. Video identification allows such stratagems to be achieved electronically.[345] In one case an uncoöperative suspect wearing unique eyeglasses refused to remove them. The Pinnacle system allowed the editor to put the same glasses on the faces of all the stand-ins.
We have already considered the first of two exceptions introduced into the 2017 revision of the code allowing a formal video identification without the use of bespoke moving images, where the suspect’s appearance had changed significantly since the original sighting. The second exception allows still images to be used where the identification officer––
“does not consider that replication of a physical feature or concealment of the location of the feature can be achieved using a moving image.”
(viii) Concentration on facial appearance
Generally speaking visual identification of an individual is through recognition of the face, the unique visual fingerprint of human individuality. To that end factors such as height and build are extraneous, though they may be relevant in corroborating a facial identification. By concentrating on the head and shoulders of participants video identification removes those irrelevant factors from the equation.
(ix) Portability and elimination of witness stress
It has been argued(notably by Peter Burton, Head of the Wakefield Imaging Unit(that an important advantage of using a laptop, DVD or VHS tape is that it will enable witnesses conveniently to view the line-up in the protective, private and relatively stress-free environment of their own home or from a hospital bed, while the details of the crime are fresh in their minds. In practice, however, this benefit is contingent on the suspect’s solicitor being present or on the procedure being itself video-recorded, so an Identification Officer would have to travel to the witness’s home either accompanied by the suspect’s solicitor or equipped with a camcorder to film the witness making an identification or both, if persuaded to go beyond the strict requirements of Code D.[346]
(x) Increased accuracy of identifications
Since the revised code has effectively made video identification the exclusive procedure to be followed, it has been noted that the one factor which might have been expected to be cited in justifying the change would have been the greater accuracy of video as against the traditional live parade, yet the preamble to the consultation draft curiously contained no reference to any such consideration.[347] In fact the evidence appears to indicate that there is a significantly lower risk of false positive identifications with the method of showing video clips than with traditional parades. This is considered in more detail below.
(xi) Empahsis on sequential as distinct from simultaneous inspection
The VIPER system was devised by the police themselves, originally without any direct input from psychologists.[348] However, as the images are shown to the witness sequentially (albeit with permissible opportunity to freeze images and to switch back and forth between images out of sequence), the method to a greater or lesser extent will tend to accord with the approval given to sequential viewing by psychologists on the basis of theory and research considered later. Certainly video identification naturally lends itself to pure sequential presentation (with no opportunity to concentrate at length on certain of the participants for the purpose of comparison between them but rather a decision to be made on each participant in turn as that person´s image is displayed). Even without a strict sequential manner of proceeding it will reduce the opportunity for the witness to compare the images with each other, an unavoidable concomitant of live parades which is thought to be conducive to false positive identifications.
(5) Possible shortcomings
One conjectured disadvantage of video identification is that it cannot, as can the live parade, furnish the witness with an ad hoc opportunity for observing particular requested aspects of posture and gait and other body movements, or of demeanour and expression when uttering a remembered sequence of words, of the kind which may furnish clues as to identity. On the other hand, parades, whether live or in video format, are primarily intended to test the ability of a witness to make a recognition by the unique factor of facial appearance. It might be thought that, even in conjunction with facial examination, a comparison between participants by asking them to speak certain phrases or adopt various mannerisms or expressions is prone to selection by guesswork.
It is arguable that another disadvantage is that video clips give no opportunity for the comparison of height or build, but this may be a blessing in the light of research considered immediately below, suggesting that comparison between participants on a traditional parade is conducive to false positive identifications.
(6) Research comparing VIPER-type procedure with traditional parades
In an experiment conducted many years ago with the use of colour slides projected to life size instead of a live parade mistaken identifications were higher in the conventional parade than with the colour slides, a finding which, it is thought reflected the fact that, at a time when witnesses were in the same room as the parade (rather than as in the practice developed later of viewing the line-up from behind a one-way glass screen) witnesses tended unconsciously to avoid looking directly into the face of parade participants.[349] However, even with the intervention of a one-way glass screen, debilitating stress levels may well remain high.
A review of research findings led the New Zealand Law Commission to conclude that procedures using live subjects produced more reliable results than those using video recordings.[350] According to some research accuracy rates were comparable when the true “culprit” was present on the parade or was one of the array.[351] However, the results of an experiment conducted by Professor Tim Valentine of Goldsmith’s College, University of London, suggested that when the true culprit was not present video parades were fairer to the suspects than conventional live parades.[352] Briefly, the experiment involved showing mock witnesses video parades or photographs of live parades held as part of the investigation of real criminal cases. For each parade, video or live, the mock witnesses were asked to read the first description of the offender given by the original, real, witness and were then required to guess who the suspect was on the parade. The mock witnesses simulated a notional real witness who (a) had no memory of the culprit at the time of the identification procedure; (b) could remember the description previously given to the police, and (c) purported to make an identification. In a perfectly fair selection of foils, the mock witnesses would have had no basis on which to make their selection and would merely have to guess who the suspect was. From perfectly fairly constructed arrays a large number of mock witnesses would correctly guess the suspect on 11 per cent of occasions because the parades all contained a suspect and eight foils. In fact, however, 1 in 4 mock witnesses (25 pc) identified the suspect in live parades, showing some bias against the suspect, compared with 15 pc of mock witnesses who identified the suspect from video parades. Whereas the vidoe parades were therefore signficantly fairer than the live parades (15 as against 25 pc) the video parades were not significantly less fair than would have been expected by chance (15 as against 11 pc). Other research suggested that video parades of African-Caribbeans and of white Europeans using equal numbers of mock witnesses from both ethnic backgrounds were equally fair to suspects of both ethnic groups,[353] a finding which contrasted favourably with earlier research suggesting that live parades may have been less fair to ethnic minorities than to white Europeans.[354]
B. Rules of Procedure for Video Identification
(1) Annexes A and E
In addition to the relevant provisions of section 3, video identification parades must be carried out in accordance with Annex A.[355] Annex E, dealing with the showing of photographs, is also relevant. The overall purpose of these provisions is to prevent the contamination of the witness’s memory.
(2) Promptness in conducting a procedure
When an identification procedure is required, in the interests of fairness to suspects and witnesses, it must be held as soon as practicable.[356] Coupled with the practical ease of arranging a video identification at very short notice this provision will now make it feasible to implement the suggestion, expounded earlier, for arresting persons who have been stopped in the street and against whom there are reasonable grounds for suspicion, instead of holding an impromptu street identification process ostensibly on the basis of the supposition that there are as yet insufficient grounds to make an arrest.
(3) Identification Officer and delegated duties in the case of video identification
Paragraph D3.11 provides that the arrangements for, and conduct of, formal identification procedures, including video identification, and the circumstances in which such procedures must be held (ie those set out in Annex A) must normally be the responsibility of the “identification officer,” an officer not below the rank of inspector. The officer must not be involved with the investigation, a prohibition which is clearly fundamental to the aim of ensuring the integrity of the process. Before the major restructuring of Code D in 2002, old paragraph D2.2 made no provision for delegation of identification duties by the identification officer to other officers or civilian staff. A change in this respect was signalled in 2002,[357] what is now D3.11 was duly formulated to permit delegation of the arrangements and conduct of identification procedures “[u]nless otherwise specified,” an explicit saving which was deleted for the 2017 edition of the code. The paragraph states that any reference in section 3 of the Code to the identification officer includes the officer or police staff to whom the arrangements or conduct of identification procedures has been delegated. In practice much of the work in arranging and conducting identification procedures is delegated and in the case of identification procedures involving detained suspects staff to whom identification procedures may be delegated must be employed directly by the police force and under the direction of its chief officer or employed by a person with whom a police force has a contract for the provision of services relating to persons arrested or otherwise in custody.[358] In delegating the arrangements and procedures for identification in D3.5 to 3.10, the identification officer must be able to supervise effectively and either intervene or be contacted for advice.[359] Delegated custody officers or staff members given custody remain responsible for ensuring that the procedures and tasks are carried out correctly in accordance with the Codes of Practice.[360]
Annex A.1 states that the arrangements for obtaining and ensuring the availability of a suitable set of images to be used in a video identification must be the responsibility of an identification officer, who has no involvement with the case. This function may be delegated to other officers or police staff in accordance with the power of delegation under D3.11, mentioned earlier. Subject to two conditions the duties of the identification officer with regard to giving the suspect the information required by the code and serving on the suspect the requisite written notice detailing that information may be performed by the custody officer or other officer not involved in the investigation.[361] The conditions are (a) that it is proposed to hold an identification procedure at a later date, e.g., if the suspect is to be bailed to attend an “identification parade,” and (b) that an inspector is not available to act as the identification officer before the suspect leaves the station. The officer concerned must inform the identification officer of the action taken and give them the signed copy of the notice.[362] A Note for Guidance advises that the purpose of these provisions is to avoid or reduce delay in arranging identification procedures by enabling the required information and warnings to be given at the earliest opportunity.[363]
Later, in addressing the question of measures which might be put in place to prevent the communicating of cues to witnesses during video identification sessions, an added safeguard that will be canvassed is the amending of Code D to provide for such sessions to be “double-blind.”[364] In other words, operators should be unaware of the identity and appearance of the suspect and thus in no position to send signals to the witness, whether consciously or unconsciously. Such a measure would require delegation to become obligatory (not merely permissible) under the code in order to guarantee a division of function between (a) the police officer or civilian staff member conducting the actual identification with the witness and (b) the officer or staff member selecting the array of images, who necessarily has to be aware of the suspect’s identity in order to carry out that exercise.
In the consultation process leading to the code revision which came into force on 1 January 2006 Sir Ronnie Flannagan and Her Majesty´s Inspector of Constabulary recommended the removal of the requirement for an inspector to be in charge of identification procedures and recommended the addition of a section providing that the appointment of an identification officer should be a matter for the Chief Constable to decide. In the Schedule of proposals and Home Office responses, in which this proposal is recorded, it is noted that the Home Office were proposing to amend the first sentence in D3.11 to read “. . . an officer or police staff appointed as an identification officer by the chief officer of their police force and who is not involved with the investigation.” It was further noted that the recommendation reflected a move towards using police staff and that the Home Office welcomed the proposal but required clarification from their legal advisors on the need for primary legislation to enable the Designated Identification Officer to take custody of the detainee and/or use reasonable force. This was to be subject to further consultation and discussion with stakeholders.
The Home Office Consultation Paper Modernising Police Powers: Review of the Police and Criminal Evidence Act (PACE) 1984, published in March 2007, posed the question whether there was any need to consider provision of any identification procedures and processes in primary legislation, as distinct from Code D (para 3.41). No hint was given in that terse sentence as to what exactly the authors had in mind but what in fact was being considered was an elaboration upon the Flannagan proposal of 2006. The scheme had in fact been set out in a paper issued by the National Identification Forum.[365] It involved a radical change in the status of Identification Officers, to be effected by the implementation of primary legislation. The changes in contemplation were likely to lead to a serious dilution of the safeguards which have been achieved through evolution of the general PACE scheme.[366]
In place of the requirement for the identification officer to be of the rank of inspector it was intended to create the post of Designated Identification Officer under schedule 4 to the Police Reform Act 2002, which “would ensure that staff appointed to undertake the role would be subject to the existing disciplinary and liability provisions that apply to all police officers and designated police staff.”[367] The Act requires Chief Officers to be satisfied that staff are “suitable, capable and trained.”
The justifications offered by the Home Office for their proposals were tendentious and unconvincing. Thus, it was suggested—ingenuously it might be thought—that “using civilian staff has the potential to increase, rather than reduce, the independence of the post holder.”[368] Of course, many people could, with sufficient training and experience, manage and supervise video identification procedures perfectly competently. However, it may be queried whether civilian auxiliaries would in general be more likely to resist the sort of esprit de corps pressures coming from senior investigating officers to cut corners or misinterpret the code than do inspector rank identification officers at present. Would appointment to “designated” status under the 2002 Act really be sufficient to instill the necessary independent self-confidence to make them more robust than police inspectors? This is very doubtful.
Under the existing rules decision-making in respect of every formal identification procedure lies exclusively with the Identification Officer. Indeed, with the exception of one particular decision which the code requires to be made,[369] there is no obligation on the Identification Officer even to consult the officer in charge of the investigation. In all other respects any discretion allowed for by the code may be exercised by the I.O. without the investigating officer having any recourse to refer the ultimate decision to a more senior officer. The professed belief of the NIF paper authors that Designated Identification Officers (police officers of unspecified rank or civilian staff employees) will possess the self-assurance to withstand pressures from senior investigating police officers is belied by the proposal that “[a]s with the provision for Staff Custody Officers, the Codes will provide the Identification Officer with recourse to a superintendent in the event that his or her authority or decision is questioned.”[370] The reality of course, plainly reflected in the referral proposal, is that, lacking the status of inspector’s rank and facing pressure from senior police investigators to forego the highest standards, D.I.O.s would need to feel able to reach for the support of a high ranking-officer, or, at least, to pass on responsibility for a decision which may displease the investigator.
The problem lies with the personnel who would be likely to serve as ultimate referees. It is true that most superintendents have a good working knowledge of the law on custody and are competent to adjudicate on detention issues referred to them by custody officers. Many may have been custody officers, or if not, they would at least have experienced many years of interaction with custody officers as investigators or supervisors. By contrast, very few would have gained suitable experience of, not to say a necessary level of expertise, in understanding, the relatively complex issues raised by the Code D provisions on visual identification by witnesses. Lacking the necessary expertise divisional officers might feel unqualified to engage meaningfully in an area of law and practice in which they enjoy little experience and might be tempted to take the easy option of simply supporting the senior investigator, whom they would have been likely to know on a personal level. Sensing that reticence, and not wishing to press an argument which they might feel they were likely to lose and to become a nuisance by taking a particular point in vain, D.I.O.s might prefer simply to back down. It is understood, for example, that across the country it is not uncommon for a division to be without its own dedicated inspector rank identification officer, with the result that requests for video identification tend to be met regardless of the legal position.
Uncontroversially, the NIF paper stated:
“Given the expertise many civilian staff have developed and the need for designated staff to be suitable, capable and trained the skills and knowledge they possess will see the post holders develop as specialists within their field. Designation will also provide a career pathway enabling staff to be retained and take on new responsibilities.”[371]
It was proposed to amend “PACE” (ie presumably the Act) in order to place “witness identification procedures” on the same statutory footing as other identification powers (eg fingerprinting, photographing and taking tissue samples) and thus “enable designation to take place” (para 7.8). It is not clear why this could not have been achieved simply by amending Code D to remove the police status and rank qualification from the post of identification officer. The requirement for Chief Officers making an appointment to be satisfied as to suitability, capability and adequacy of training could presumably be written into the code. In fact, the forum paper acknowledges that police staff could be authorised to act as identification officers without primary legislation simply by amending Code D but asserts that this would “not carry the same weight as the Police Reform Act in requiring designated staff to be suitable, capable and trained” (para 7.10). Apparently it requires the imprimatur which only Schedule 4 to the Police Reform Act 2002 can give to ensure that standards are maintained or even raised.
This assessment might well be justified. However, as already stated, it was abolition of the present status and rank qualification which was objectionable. By placing emphasis on the vague notion of “training” and dispensing with that qualification, there was a very real risk that in the interests of cutting costs highly experienced and high salaried Code D police inspectors woudl have been removed from post and transferred to field or more general managerial duties. They would have been replaced by ancillary staff (police or civilian) currently working under the supervision of Code D Identification Officers. The likelihood was that they would have been promoted to the post of D.I.O. with only a modicum of “add-on” training.
The NIF paper envisaged that D.I.O.s could also be empowered to exercise other complementary identification powers and duties under PACE, such as the taking of fingerprints, photographs, samples and footwear impressions currently assigned to designated detention offices (para. 7.9). There were three potential problems with this proposal.
First, the management and supervision of video identification parades and attendant procedures is time-consuming and amounts to a full-time job. To increase the burden with unrelated, undemanding and laborious tasks would only have diminished the time available for work requiring careful thought, skill and experience, to the inevitable detriment of standards.
Second, to conduct those other procedures identification officers would have needed to spend much of their professional time in the custody area. This would have involved considerable inconvenience and time-wasting in coping with demands to be in two places a once, since video identification facilities must be kept well away from the custody area in order to avoid witness contamination and intimidation.
Third, the other tasks required little evaluation and in many forces were performed by staff on a lower pay scale. The result was likely to be a downgrading in the perception of the status and importance of the post, a result directly contradictory of the proclaimed aspiration of investing it with even greater independence than police inspectors had. By abolishing the inspector qualification the status of the post would have been reduced. To diminish it further by associating it with comparatively low-status duties, would have made it that much harder for post holders to persuade high-ranking investigators to accept their view on the conduct of procedure.
(4) Officers involved in the investigation
Of very long established provenance is the principle that no officer or any other person involved with the investigation of the case against the suspect, beyond the extent required by the provisions in paragraphs D3.5 to 3.10 for arranging and conducting identification procedures, may take part in the procedures or act as the identification officer.[372] So important is the rule that it has been very strictly applied.[373] The general prohibition is not incompatible with the process of consultation between the identification officer and the officer in charge of the investigation which the code requires in the determination as to the appropriate procedure to be used.[374] An ancillary safeguard against involvement by investigators in the selection of images and the risk of influencing witnesses is the rule that no one involved in the investigation is permitted to view the images to be used for the procedure before they have been shown to the witness.[375]
(5) Notice to the suspect
Reference has already been made to the provisions of D3.17, requiring the suspect to be given an explanation of various specified matters before arrangements are made for a video identification (or live parade, or group identification). Reference was also made to D3.20 permitting the explanations to be delayed where it might induce the suspect to take steps to avoid being available to participate in the identification procedure. In addition to the explanations already mentioned, an explanation must also be given to the suspect as to:
• the purposes of the identification procedure[376]
• the entitlement to free legal advice[377]
• the procedures for holding the video identification, including the suspect’s right to have a solicitor or friend present[378]
• the special arrangements for juveniles, if appropriate[379]
• the special arrangements for mentally disordered or otherwise mentally vulnerable people, if appropriate[380]
It has already been mentioned that the information which must be given under D3.17 must also be recorded in a written notice to the suspect.[381] Suspects must be given a reasonable opportunity to read the notice, after which they should be asked to sign a second copy to indicate if they are willing to co-operate with the making of a video and the signed copy must be retained by the identification officer.[382]
(6) Details of any “first description” to be disclosed to suspect
before any video identification is carried out
Reference was made in Chapter 1 to the requirement laid down by paragraph D3.1 for any first description given by a witness of a suspect to be recorded.
Before any video identification parade is carried out the suspect or their solicitor must be provided with details and where practicable the record of the first description of the suspect by any witnesses who are to attend the video identification.[383] The code makes no reference to later descriptions (which may be at variance with the first description) but it has been suggested that subsequent descriptions should equally be disclosed.[384] In any event, a later description arguably comes within the meaning of a “first” description by virtue of being an amendment to it.
The provision facilitates an assessment of the suitability of the proposed foils by suspects and their solicitors and it may well have an impact on whether any objection is taken or any requests for adjustments are made. Thus, if the suspect has a facial mark and a similar mark is described by the witness, a request would be certain to be made for the mark to be concealed by the placing of a pixellated blur on the image of the suspect and on the images of each of the foils. However, where no such feature had been described previously, the suspect and solicitor may have a more difficult and delicate decision to make. On the one hand they might decide to forego such a request in order not to forfeit the potential opportunity at trial of being in a position to stress the considerably enhanced exculpatory significance of a failure by the witness to pick out the suspect, or of the witness’s expressed reluctance to make a positive identification. On the other hand, this may be a high risk strategy. As has been cogently pointed out,[385] if the distinctiveness is not suppressed by eliminating it from the suspect’s image and from the foils it may be that the witness will profess to recognise the suspect by the distinctive feature, claiming that the sight of it revived his memory, or that he in fact mentioned it when the original description was given but that it was doubly missed from the record(by the police maker and then by himself when checking the record. In fact he may simply be latching on to a previously unobserved feature and using it in the belief that it will enhance his appearance of confidence. The choice whether to leave the distinction or to suppress it may be a difficult one, but the profession by the witness of a sudden revival of memory or asserting poor record-keeping and careless verification is liable to prove unconvincing in the face of sceptical cross-examination. In the end, if it has not been previously referred to, it may well be worth leaving the feature distinct.
(7) Giving notice of previous showing of photographs or
computerised or composite likeness
Reference has already been made to paragraph D3.3, the effect of which is to permit eye-witnesses to be shown photographs, computerised or artist’s composite likenesses or similar likenesses or pictures (including “E-fit” images) if the identity of the suspect is not known or if known is unavailable to participate in one of the formal identification procedures for which provision is made in D3.5 to D3.10.[386] Where the police have shown a witness an image of a culprit or suspect this may well undermine the value of any subsequent formal identification procedure. Accordingly, when a witness attending a video identification has previously been shown such material the suspect and their solicitor must be informed of the fact before the procedure is conducted.[387] It is the responsibility of the officer in charge of the investigation to make the identification officer aware that this is the case.[388] The code contains no express requirement for disclosure of the material at that stage.[389] However, there would be little point in requiring the police to pass on the fact that the witness had been shown the material if the suspect and representative were not also to enjoy the right (subject to considerations of practicability) to view the material before the identification procedure continued. The evident purpose of giving the information beforehand is clearly to afford the suspect and representative the opportunity to request any adjustments in the presentation of the selected images which they perceive to be necessary in the light of the material shown. Thus, it may be that although the suspect has a distinctive facial scar no such feature was described by the witness and the representative may originally have taken the view that it was therefore unnecessary to request that the scar be masked electronically with identical masking being added to the images of the foils in order to match the suspect.[390] However, if it turns out that one of the photographs previously shown depicted the suspect, then even though the witness never picked out the suspect when viewing the photographs the fact that the witness had been shown a photograph of the suspect would be almost certain to occasion such a request. (Of course, failure previously to pick out the suspect’s photograph might make it unlikely that an identification would be made, so obviating the need for a procedure.)
(8) Measures relevant where moving images, still photographs, constructs and information have previously been disseminated via the media
In Chapter 2 we reviewed Part (C) of Code D, section 3, which deals with the issue of recognition by uncontrolled viewing of images where, for the purpose of identifying and tracing suspects, “films” (a now perhaps archaic usage meaning of course video footage) and photographs are (a) shown to the public (including police officers and police staff) through the national or local media or any social media networking, or (b) circulated through local or national police communication systems for viewing by police officers and police staff, and the viewing is not formally controlled and supervised: D3.38.[391] The self-evident problem with the dissemination of such imagery is that a witness who is to participate in a formal identification but has previously seen material in the media depicting a possible culprit might well be influenced in making an identification. The code affords some protection against this possibility by (i) requiring the preservation of the material (or, presumably, copies thereof) released to the national or local media for showing in furtherance of D3.38(a) and (ii) requiring that the suspect or their solicitor must be allowed to view such material before any formal eye-witness identification procedure under C3.5 to C.3.10, C3.21 or C3.23 is carried out, provided it is practicable and would not unreasonably delay the investigation.[392] The evident intention behind the provision is that suspects and their representatives should have an opportunity to make representations on the selection of foils or the form of presentation in order to request any adjustments which they perceive to be necessary in the light of the appearance of an image depicted in the media material. The example of the scar on the suspect’s face postulated above in relation to photographs previously shown to the witness is equally applicable in relation to viewing material previously broadcaset or published in the media.
It might have been supposed that an additional safeguard would be to ask witnesses before any identification procedure if they have seen any images in the media. The code indeed provides for such an inquiry but only after the witness has participated in the procedure. Thus each witness involved in the procedure must be asked, after they have taken part, whether they have seen any broadcast or published films or photographs relating to the offence or any description of the suspect and their replies must be recorded.[393] The rationale for a delayed request is not immediately obvious but it may be that the draftsman had in contemplation an assumption that the requirement to offer inspection would serve as a sufficient safeguard whereas it was supposed that asking the witness if they had seen such material might trigger a recollection which would otherwise have remained dormant.
(9) Assimilating movements, positions and condition of
suspect and foils to be depicted in the images
Full consideration was given earlier to the requirement that “[t]he images used to conduct a video identification shall, as far as possible, show the suspect and other people in the same positions or carrying out the same sequence of movements.”[394] As already pointed out the suspect and volunteers are conventionally recorded sitting in the same position against the same neutral backdrop and with the same lighting. Each person faces the camera, turns to one side and then to the other and finally faces the camera again. Each aspect is timed to last an equal number of seconds.
The images must be shown under identical conditions unless the identification officer reasonably believes it is not practicable to do so because of the suspect’s failure or refusal to co-operate or other reasons, provided any difference in the conditions would not direct a witness’s attention to any individual image.[395] This ought not to be difficult to achieve. For example, a suspect in custody who has refused to co-operate might conveniently be video-recorded covertly in the precincts of a police station. The police might already have recorded such footage routinely as part of an ongoing CCTV security and surveillance programme covering the custody areas of the police station.[396] Volunteers might then be similarly recorded in the same or similar location to avoid drawing attention to the suspect.[397] A suspect on bail might be recorded surreptitiously walking along a particular road with comparable footage being recorded of volunteers walking along similar roads.
Where a known suspect is not available or has ceased to be available the identification officer may make arrangements for a video identification and if it has not been possible to obtain video footage of the suspect covertly, or at all, the identification officer may follow the sequential video identification format of procedure using still images, which may be obtained covertly if necessary.[398] As suggested earlier the still image of the suspect may conceivably be made up with a photograph taken from criminal records, from one taken during surveillance prior to arrest, from one seized at the suspect’s address or from his among belongings, from one taken while the suspect was initially detained at a police station or from a CCTV time lapse tape.[399] It has been suggested that the use of a photograph from criminal records will allow a high degree of uniformity to be achieved if similar sourced stills of the foils are used.[400] The Custody Suite Imaging System now provides a potentially unlimited database of foils nationally available under the PROMAT system. However, it is not clear whether this would require the permission of those persons shown in such photographs. (The images stored in the various identification data bases such as the VIPER collection are used by the consent of the persons they depict.) In any event, it is understood to be a comparatively straightforward exercise of modern computer technology for still photographs taken from the moving images held in the VIPER database to be adapted to match most supplied portrait photographs in terms of lighting and background.
(10) Requirement for the selection of foils conforming to
the suspect´s “general appearance”
Reference has already been made[401] to the long-established rule for identification parades that the chosen volunteers should “as far as possible resemble the suspect in age, height, general appearance and position in life.” This principle has been applied to video identification, the rule for which formerly was that “[t]he set of images must include the suspect and at least eight other people who, so far as possible, resemble the suspect in age, height, general appearance and position in life”[402] but which under the code revision which came into force on 1 January 2006 was modified to remove the requirement for height conformity on the ground that within reason it is impossible to tell a discrepancy in height from the images displayed.[403]
For reasons which are summarised below psychologists have argued that it is preferable to use foils whose appearance matches a previous description given by the witness rather than the suspect, at least where the description involves no obvious or radical disparity with the suspect.[404] The code emphasis on “general appearance” would certainly permit the use of foils whose facial description may differ on some specific points from that of the suspect and would arguably allow for a choice to be made between match to culprit description and suspect resemblance methodology.
In isolating the rationale for choosing between match to description and suspect resemblance it is necessary to distinguish between cases in which the witness has given a description consistent with the appearance of the suspect in all attributes mentioned in the description and those in which the suspect does not match the witness’s description in respect of some attributes.
(i) Description of person seen matches the suspect
As Valentine has explained, where the description of the person seen (for ease of reference, the “culprit”) matches the suspect it introduces no bias against the suspect if the foils differ on some feature that was not mentioned in the original description:
“For example, imagine a witness who described the culprit as “white, male, mid-forties, with long, dark hair.” The suspect fits this description and has a pale complexion and prominent dark eyebrows. A culprit-description strategy would require that all lineup foils would be white males, say between 43 and 47 years old with long, dark hair. No attempt would be made to match the prominent eyebrows or the pale complexion. Indeed there should be a range of variability around these and other features not mentioned in the description (e.g. build, face-shape etc.). Heterogeneity of features not mentioned in the description will help a witness with a reliable memory to distinguish the culprit from the foils. If the police suspect is the culprit, the witness may be able to identify the suspect, because on seeing the lineup, she recognises the man with the pale complexion and prominent eyebrows. A lineup that consists of a number of people chosen because they closely resemble the suspect in all aspects of their appearance, will make it difficult even for a reliable witness to identify the culprit, if present. However, if the suspect was not the culprit, the pale complexion and prominent eyebrows are no more likely to trigger a mistaken identification than some other features on a foil (maybe thin eyebrows and an olive skin tone) because the witness has not seen the suspect before. A strategy of suspect resemblance, which attempts to match all features including the pale complexion and the prominent eyebrows, will produce a procedure that is less sensitive than a lineup constructed of foils chosen to match the witness’s description of the culprit. As the culprit description strategy does not introduce any systematic bias against the suspect both procedures should be equally fair.”[405]
Although selection by match-to-culprit description may therefore be no less fair to the suspect than selection by suspect resemblance the latter is likely to offer a more tempting option for the suspect since it might well be supposed that, distracted by an array of near-clones, the witness will be confronted with an almost forbidding obstacle to surmount in making an identification. However, selection by suspect resemblance may ultimately prove to be a mixed blessing for the suspect because the closer the resemblance between foils and suspect the more forlorn will be the chances of the defence successfully mounting any challenge if the suspect should happen to be identified.[406] From an objective standpoint although there is some experimental evidence to suggest that foils selected by culprit description may produce more reliable identification evidence than those selected by suspect resemblance, taking all relevant studies into account there is, as Valentine points out, little empirical evidence on which to base a contention that a match-to-description strategy is a superior method to construct an array than a suspect-resemblance strategy.[407]
Since the available empirical evidence places no particular premium on the need to select by culprit description there remains one compelling reason for selecting by suspect-resemblance. The suspect may have a facial feature which though not idiosyncratic, such as a tattoo, is nevertheless uncommon, for example, an abnormally large nose. No such feature is mentioned in the witness’s written description, though that would not necessarily mean that the culprit had a normal size nose. The witness might have forgotten that the culprit´s nose was large and the officer taking a description might have neglected to take the witness through the appearance of particular facial features. If foils are selected with a normal range of nose sizes, an identification of the suspect would raise suspicions that the witness had been told privately about the suspect’s nose size and was therefore on the look-out for a face with such a feature.
(ii) Disparity between description and suspect’s appearance
A stark descrepancy between the description and the suspect on some important attribute (for example, age, complexion or hair length) will create an obvious objection against the selection of foils according to the description, if for no other reason than that it will draw attention to the suspect. On the other hand it might be supposed that any risk that the difference will in itself induce the witness to plump for the suspect rather than foils matching the description will be substantially more than offset by an assumption that witnesses will normally have in mind the description they previously gave to the police and that they will be expecting to identify somebody who matches that description.[408] It is always possible that on the point of discrepancy the description was mistaken and that despite the mistake the witness on seeing the suspect in the array instantly recognised him as the culprit. A witness making such an assertion to the jury might well be believed. Equally, (as in the example of the large nose postulated above in relation to non-disparity cases) such an outcome might not unnaturally raise suspicions that prior to the video identification procedure the witness had been shown a photograph of the suspect or, at least, told of the disparity. In such a case, selection of foils by culprit-description could prove a recipe for miscarriage. It is therefore imperative that a selection should be by suspect-resemblance for those features which do not match the description.
Since a radical discrepancy between some point of the description of the culprit and the suspect´s appearance would raise obvious suspicions about an identification of the suspect it might be expected that no identification would be likely ever to be made, whether employing a culprit description or suspect resemblance methodology. Indeed in the case of a fundamental disparity it is conceivable that no procedure would even be contemplated.[409] Presumably proponents of a culprit description strategy would rarely expect an accurate identification to be made but it might be wondered what purpose they would in that case envisage for a comparator test, be it a live parade or a video method. There is an implied suggestion in the analysis of one commentator with a background in criminal investigation and academic psychology that, although it may be explained as “an additional investigatory advantage,” the real purpose behind the use of foils selected on the basis of a previous description is that where one of the foils is picked out this “will give the investigators a wealth of information about the appearance of the culprit.”[410] Yet this can hardly provide a justification for conducting a video identification procedure. The suggestion by another commentator[411] that the parade might usefully comprise a mix (e.g. some with fair hair to resemble the suspect, others with dark hair to match the description) seems based on a similarly confused approach.
(iii) Annex A provision for electronic manipulation of images
Reference has been made to the technique of pixellation, including the electronic masking out of an unusual facial feature belonging to the suspect with identical masking out of the same area of the faces of otherwise similar foils. Annex A makes full provision for such exercises. Thus Annex A.2A provides that where the suspect has an unusual physical feature, eg, a facial scar, tattoo or distinctive hairstyle or hair colour which does not appear on the images of the other people that are available to be used, steps may be taken to (a) conceal the location of the feature of the suspect and the other people, or (b) replicate that feature on the images of the other people. It is to be emphasised that the provision is discretionary. The paragraph continues that for these purposes, the feature may be concealed or replicated electronically or by any other method which it is practicable to use to ensure that the images of the suspect and other people resemble each other. The identification officer has discretion to choose whether to conceal or replicate the feature and the method to be used. Thus the position of a scar on the image may be pixellated over to cover the scar and pixellation imposed over the same position on the images of the faces of the foils. Alternatively, if it can be done effectively a scar might be digitally added to the images of the foils. It is submitted that a mark, blemish or scar on the suspect should not be eliminated by the application of cosmetic make-up, because that would to mislead the witness into supposing that the suspect does not have such a feature. Annex A.2A formerly provided further that if an unusual physical feature had been described by the witness, the identification officer ought, if practicable, to have that feature replicated but that if it has not been described, concealment might be more appropriate. The advice was deleted under the 2017 edition.
In a description of the culprit (or “the person seen”) the witness may have made no mention of a particular facial feature such as a scar. However, suppose the suspect has such a feature and there are no suitably similar foils available on the data base with a comparable feature. The identification officer will need to decide whether to pixellate over the feature on the image of the suspect and over the equivalent areas of the faces of the foils, or to make no alteration. The suspect would probably press the officer to opt for the latter course as a consequence of assuming that not having noticed or recalled a scar the witness would be likely to assume that a person bearing a scar is not the culprit. The hidden danger for the suspect is that the witness may have been unconsciously aware of a scar, the memory of which is only triggered on seeing the suspect, who in other respects closely resembles the culprit. The revived memory then reinforces what might otherwise be no more than a tentative feeling of familiarity. Being discretionary, the option under A.2A to assimilate the appearance of the suspect and foils, or not, allows the identification officer to accede to the suspect’s request.
If the witness makes a request to view an image where an unusual physical feature has been concealed or replicated without the feature being concealed or replicated, the identification enjoys a discretion to allow the witness to view any one or more such images if they are available and the identification officer should consider whether it would be more appropriate to arrange for the witness to view separate similar images of the usually feature alone: Annex A.2C.
Annex A.2B provides that if the identification officer decides that a feature should be concealed or replicated, the reason for the decision and whether the feature was concealed or replicated in the images shown to any witness must be recorded.
In order to attempt to comply with the relevant provision of Annex A pixellation was employed with the consent of the suspect and his solicitor in an appeal case some years ago to eliminate the suspect’s greying temples and greying goatee beard, where such features could not be found even after a trawl of the 19,000 images available on the particular database in question.[412]
A particular facial feature of the person seen by the witness on the occasion in question may have been sufficiently noticeable to be described by the witness. Where the suspect shares that feature it is especially important to select foils who also share it, otherwise the witness’s attention is likely to be drawn only to those on the array with that feature and the other foils may well be ignored with the result that the relevant array is effectively reduced. It may therefore be necessary to select foils to include a feature which though not normally considered a criterion for selection, such as eye colour, has relevance in the particular case.[413] The suspect’s legal representative should be alert to the problem when inspecting the proposed array and should be prepared to raise the issue with the identification officer.
(iv) Practical impact of match to description and match to suspect debate
The whole issue of whether to select the foils by reference to culprit description or on the basis of resemblance to the suspect is actually less important in practice than the vigour of the debate mentioned earlier might suggest. As between first description and suspect appearance it is not uncommon to encounter fundamental disparities over such characteristics as height, build, age, racial type, hair quality, hair quality and colour, and skin complexion, blemishes or markings. But in very many cases(probably the great majority(there will be little starkly contrasting difference between the description and the suspect. While it is the face which uniquely distinguishes human individuals from each other and is the key feature upon which identifications tend to be based, paradoxically, descriptions rarely contain much detailed facial information. This is because facial characteristics are difficult to describe,[414] particularly for lay persons. Literature abounds in such precisely understood shorthand as “button nose,” acquiline profile,” “well chiselled features,” “beetle-browed,” and “lantern-jawed,” and such phrases as “narrow, deep-set eyes,” “hooded eyelids,” “thin lips,” “high cheek-bones,” “cleft chin” and “prominent ear-lobes” are in common currency outside the hard covers of fictional prose. Yet, even then(
“. . . language is much too imprecise to be of much use in acquiring the subtle information regarding each face we encounter. When asked to describe faces people usually resort to simple general labels (eg crafty looking) or select a few salient features (eg large nose and red hair).”[415]
Whereas in cases of very serious crime the police will have the time and resources to assist witnesses in making up photo-fit or E-fit composites of a suspect, this is a luxury unavailable to police information room operators who need to record descriptions verbally in order to relay instant information to officers on the look-out for offenders. In the heat and excitement of a 999 call few witnesses have the perceptive faculties and language skills at their command to emulate Jane Austen or a Daily Mail crime reporter. In the words of McKenzie and Dunk(
“Th[e] Code of Practice requirement [for the recording of first descriptions] is so recent that it is doubtful whether any serious thought has been given to police telephone operators’ techniques and the impact of them on a witness’s ability to store and then successfully (and accurately) recognise a suspect.” [416]
It might anyway be thought that where a description was wholly at variance with the appearance of the suspect an application of the second, prospective identification, condition of D3.12 might obviate the duty to hold a parade. Thus, despite the profession by the witness of an ability to identify the person previously observed, a fundamental difference between the description given by the witness and the physical appearance of the detainee being investigated would conceivably cause the police to decide that there was no reasonable chance of the witness being able to do identify the detainee.
(11) No more than one suspect in a single array
Annex A.2 provides that “[o]nly one suspect shall appear in any set unless there are two suspects of roughly similar appearance, in which case they may be shown together with at least twelve other people.” This is the counterpart of an identical provision relating to live parades.[417] It has been argued that it is intended to permit doubling up where two people are suspected of being a sole offender, and not where two persons are suspected of joint involvement in an offence.[418] It may be conjectured that the rationale is to reduce the effort and time taken up in searching the data base for suitable foils. Given that the two are mutually exclusive candidates for the role of culprit anyway and are of similar appearance it is considered that they may safely be used as foils for each other, provided that the usual minimum of eight volunteers is increased to twelve (so reducing the number of requisite foils by fifty per cent). In relation to live parades it has been observed that in practice it is very rare for a parade to contain two suspects, whose interests in its arrangements might conflict.[419] While considerations of economy and convenience may have made doubling up an understandable expedient for live parades there would seem to be hardly any justification for it in the conducting of video parades.
The significance of an additional provision relating to live parades (but not replicated in the Annex A provisions for video parades) is unclear:
“When all members of a similar group are possible suspects, separate identification parades shall be held for each unless there are two suspects of similar appearance when they may appear on the same identification parade with at least twelve other members of the group who are not suspects.”[420]
There is nothing implicit in the language to indicate whether the reference to “all members of a similar group [who] are possible suspects” is to the suspected members of a confederacy or whether it merely refers to a group of similar looking persons all of whom individually are candidates for identification as the single culprit. In any event, where there are a number of suspects of similar appeaerance the measure seems to furnish no more than a licence for taking two each at a time in a single parade with 12 volunteers.
(12) Separate foils for each suspect
The rule for live identification parades is that where a witness is to view separate parades for different suspects, the parade for each suspect must be made up of a separate array of volunteers.[421] The rule is clearly designed to preclude the possibility otherwise that the witness may infer the identity of the second and further suspects from noticing the repeated appearance of the volunteers and hence recognising them as such. No such rule is found in Annex A for video identification parades, presumably through a drafting oversight. The interests of fairness and reliability and a recognition of the spirit of Code D plainly require the observance of such a rule as if it were in the Codes.
(13) No discussion with witness as to composition of the images
For obvious reasons there must be no discussion with the witness about the composition of the set of images.[422]
(14) Identification by number
The term “identification parade” originates from the time that the procedure literally involved the suspect and volunteers parading in a static row while the witness walked up and down in front of them. The conventional method of making an identification was for the witness to step forward and touch the person selected on the shoulder. It always seemed as if there was an element of ritual about this, involving a possible test of the witness’s resolve and confidence. On the other hand, touching is unambiguous, whereas being required to describe the position of the person chosen (“third from the left”) might have occasioned dispute about the exact position of the person the witness meant to select. Time and resource constraints have precluded research as to when touching gave way to the use of selection by number but with the introduction of identification suites and the use of a viewing room screened off from the suspect and volunteers by a two-way mirror, identification by physical touching was no longer feasible. With video identification provision must be made for each person shown in an image to be identified by number.[423]
(15) Images of police officers
If police officers are shown, any numerals or other identifying badges must be concealed.[424] This provision is clearly applicable to cases in which officers are suspected of committing offences while on duty.
(16) Prison inmates
In cases in which an accused or suspect was a serving or remand prisoner it had long been the practice to conduct identification parades within the precincts of a prison, if it were considered unsafe to release the prisoner into the custody of the police so that they could be conducted elsewhere for a parade, for example to an identification suite. In such a case other inmates would almost invariably be asked to volunteer themselves as foils. Where a prison inmate is required for identification and there are no security problem about the person leaving the prison establishment, the inmate may asked to participate in a video identification to be conducted elsewhere.[425] As well as permitting video identification procedures involving inmate suspects to be conducted outside the prison the rule appears additionally to make provision for volunteers who happen to be inmates to participate in video identification procedures conducted outside prison. It may be envisaged that the involvement of such volunteers may be necessary for example to assist in the compilation of an array of foils to match footage of a recalcitrant suspect which has been recorded covertly in a police station.[426] The police may have been unable to recruit sufficient suitable volunteers from members of the public and the suspect’s lack of physical stature may have ruled out the participation of police officers as volunteers.
Where the recording of the suspect for a video identification must be made within a Prison Department establishment, other inmates may participate in the provision of comparator images.[427] This may be necessary if it is not possible to record an image of the suspect using a camera angle or against a background or under lighting conditions which together match the standard used to record the images of volunteers drawn from the data base. Portable equipment taken to prison, or equipment permanently retained at the particular prison, may allow such a match but these options may not always be available. Even where it is feasible to create the standard conditions within the prison the suspect’s recalcitrance may preclude their use and covert recording may become appropriate. In cases of the non-availability of suitable equipment or of a non-coöperating suspect it will be necessary to record inmate volunteers in order to secure uniformity with the image of the suspect. In such a case, if a prison inmate is shown, either as a suspect or not, then either all, or none of, the people shown should be in prison clothing.[428] This will cover the situation where the suspect refuses to shed his prison uniform or otherwise has no civilian clothes available and none can be found to fit or for some reason no such suitable clothing can be found for one at least of the inmate volunteers.
(17) The suspect’s right of reasonable veto on the selection of images
Annex A.7 provides that the suspect or their solicitor, friend, or appropriate adult must be given a reasonable opportunity to see the complete set of images before it is shown to any witness.[429] If an objection is raised to the set of images or to any of the participants the suspect must be asked to state the reason.[430] If the objection is reasonable steps must be taken to remove the grounds for objection, if it is practicable to do so. If this is not practicable, the suspect and/or their representative must be told why their objections cannot be met. The objection, the reason given for it, and why it cannot be met must be recorded on forms provided for the purpose.
Constraints of urgency may mean that in order to be able effectively to view the selection of images to be shown to the eye-witness the suspect’s solicitor may have no alternative but to attend the venue where, and on the occasion when, the showing process is to take place. For that reason it is stipulated in Annex A.9 that the suspect’s solicitor, if practicable, must be given reasonable notification of the time and place the video identification is to be conducted.
The 2017 edition inserted in Annex A.7 additional guidance pointing out that the A.2 provision for the images of the other people to resemble the suspect does not require that the images must be identical or more or less identical. The guidance stresses that the purpose of the video identification is to test the eye-witness’s ability to distinguish the suspect from others and it would not be a fair test if the images shown were virtually indistinguishable for each other. It is further provided that the identification officer is responsible for ensuirng that images shown are suitble for the purpose of this test. In reality the innovation is otiose. The very notion that a battery of images of near clones of the suspect might somehow be available from the database of similar imagery is virtually bordering on the absurd. The measure was self-evidently inserted in order to inhibit suspects and their representatives from even thinking about making unreasonable editorial demands.
Reference has already been made to the variation in practice over the initial selection of images and to the fact that in the Metropolitan Police identification officers make the initial selection of images and present the complete set to the suspect for consent. It is evident that this practice is determined by the ambit of Annex A.7, which, notably, bestows suspects with no right to be involved in the initial selection of images from those transmitted from the Viper centre but merely furnishes a right of reasonable veto over the “complete set of images,” subject to the practicability of meeting any objection. Although it was originally envisaged that suspects and their legal advisers would make the initial selection,[431] it is assumed that the terms of the paragraph were deliberately formulated in order to meet concerns on the part of the police that they should retain the right of selection in the first instance, no doubt in the interests of imposing time constraints if not for other reasons also.
There are two conditions for an objection to be met by substitution. The objection must be reasonable and it must be practicable to make any change. The first condition expressly involves an objective test and is clearly therefore justiciable. The test of whether a task is practicable is necessarily whether it is “reasonably” practicable and the second condition is also therefore justiciable. As to the question of practicability it may be envisaged that there could be very little practical difficulty in requesting the Viper centre to supply unobjectionable replacement images. It may be assumed that with trained operators in charge of making selections from the data base and with an adequate number of foils to be supplied as back-up, it will normally be feasible to meet most reasonable objections. Whether the objection to a particular image is reasonable can only be decided by a court inspecting the range of images selected by the police which they claim it is impracticable to replace. This would necessitate a copy of the selected images being made available to the defence and to the court for adjudication. The author’s attention has been brought to one ongoing case involving proceedings which had advanced beyond the pleas and directions hearing in the Crown Court. It was proposed to conduct a video identification but when objection was taken to two out of the nine images selected by the identification officer the police refused to make any change on the purported ground of impracticability. It was not clear why it was regarded as impracticable to request a further small selection of unobjectionable images. The defendant’s solicitor accordingly asked for a copy of the compilation video so that counsel’s advice could be sought but the police refused this on the grounds that the compilation was “data protected,” presumably because it was stored on a data base. It is understood, at the time of writing, that the matter is to be listed for mention so that a judge can be invited to make an appropriate order. If the data protection argument were to prevail this would mean that a right to an objective evaluation available to the defence in relation to live identification parades, in the form of the long established right to have photographs of the volunteers, would not be available in relation to video identification.
If the defence take no steps to request a court adjudication before the witness participates in the identification procedure it is open to question whether the defence would have any basis for seeking the exclusion of evidence of identification under section 78 of the Police and Criminal Evidence Act 1984. If the allegation had resulted in a charge and the case had been sent to the Crown court it is far from clear whether a judge of the Crown Court would have any locus to require the police to make a reselection. Again, it is doubtful whether the Legal Aid Agency would grant a certificate to pursue Judicial Review, even on the strength of counsel’s advice.
Any impromptu assessment which the suspect and legal adviser make of the degree of resemblance of the foils to the suspect in determining whether to approve or object to the array can only be subjective. If, however, the case has already reached preliminary proceedings when it is sought to hold a video identification it may be that the court will entertain interlocutory litigation on the fairness of the selection of images. In that event the provisional selection made by the police, and their resistance to overtures by the defence to make a more suitable selection, may be susceptible to objective assessment through expert evidence. Again, it may be that a trial judge will allow expert evidence to assist the jury on the fairness of the selection, particularly if the defence have unsuccessfully raised early reservations or have failed to have the selection reviewed by an expert during preliminary proceedings. The method of assessing whether the array is slanted against the suspect is to give volunteers a description of the suspect and to ask them to identify the offender from an array of images. Selection of the suspect by a disproportionate number of volunteers will indicate “parade bias” in the composition of the array. A graphic example of such an experiment was conducted in order to demonstrate shortcomings in the identification procedures employed by investigators in the case of John Demjanjuk, convicted in Israel in the late 1970s of being the Treblinka guard “Ivan the Terrible” who had participated in the murder of hundreds of thousands of Jews. A number of witnesses had purported to identify the accused as the mass murderer when they were shown a photograph taken of him for his immigration to the United States in 1951 set out in an array of photographs of a number of men. The accused was the only person depicted who had a round face, short, wide neck and broad shoulders. Wagenaar, a forensic psychologist recruited by the defence to investigate the procedures used, showed the array to a test sample of 25 people with the instruction “We are looking for a man with a full round face, a short wide neck, a bald pate starting.” (Only one other of the array, a thin man, had a noticeably receding hairline). All 25 of the sample picked out the picture of Demjanjuk as the wanted person, showing that a genuine witness with only a vague recollection of the visual appearance of Ivan the Terrible, but strong motivation to see him apprehended, could have picked out the picture on the basis of its close resemblance to this recollection, and nothing more.[432]
(18) Ensuring non-communication between witnesses
An essential and obvious safeguard against mutual suggestion and reinforcement between witnesses is contained in Annex A.11 which stipulates that only one witness may see the set of images at a time.
Annex A.10 provides that the identification officer is responsible for making the appropriate arrangements to ensure that, before they see the set of images, witnesses are not able to communicate with each other about the case.[433] Nor must the witnesses have been able to overhear a witness who has already seen the material. A witness must not be told whether a previous witness has made any identification.
The general use of dedicated identification suites, originally introduced for holding live parades, continues to be of benefit in isolating witnesses from one another although they can be of little utility in preventing communication between witnesses who are not strangers to each other, such as family members who are co-witnesses or work colleagues. On the other hand, video identification does not specifically require witnesses to be assembled for convenience on any one single occasion, as was the case with live parades, and it is perfectly feasible to ask them to attend video procedures at police stations on diverse dates to avoid meeting. The use of portable equipment may certainly permit this, as it will permit sessions to be undertaken at the home of an infirm or frail witness or at a hospital where a witness lies injured.
(19) No unauthorised people may be present
No unauthorised people may be present during the procedure.[434] It is not clear whom this provision is intended to exclude. Clearly officers involved in the investigation may not be present but that is achieved by the general rule in D3.11 and the suspect is excluded by specific provision.[435] While the suspect is implicitly entitled under the code to be legally representated for the purposes of safeguarding the suspect’s interests regarding the identification process in general there is now no unconditional entitlement of the legal representatives to be present on the suspect’s behalf to observe the actual procedure of showing images to the eye-witness.[436] Aside from those categories of person, in the absence of any criteria it is difficult to see what prevents the identification officer from allowing in whomsoever they please. On the other hand, since the session is conducted in private it is difficult to envisage some “unauthorised” person slipping in unnoticed at the back of the room with the identification officer being powerless to eject such a person. The measure is hardly necessary to enforce an eviction.
(20) Suspect not permitted to be present during a video identification session
By contrast with a live identification parade, in which the suspect’s presence is essential, the suspect’s presence at a video identification would defeat the whole purpose of the exercise. However, in losing no opportunity to state the obvious Annex A.9 provides that the suspect “may not be present when the images are shown to the eye-witness(es) and the 2017 edition added the further provision of disentitling the suspect from being informed of the time and place the images were to be shown.[437] There may, of course, be occasions when it will be necessary, for example in cases of urgency, for the suspect to be present at the same time as the witness in the general vicinity of the place where the witness is to view the images, in order to inspect the selection beforehand.
(21) Ensuring that the witness’s attention is not drawn to any individual image or given any indication of the suspect’s identity
Video An obvious stricture lays down that “[c[are must be taken not to direct the witness’s attention to any one individual image or give any indication of the suspect’s identity.”[438]
(22) No reminding the witness about images previously identified or descriptions previously given
The witness may previously have made an identification from a photograph, or computerised or artist’s composite likeness, of a person who appears not to be the suspect. Again, the witness may have given a previous description which is significantly different from the suspect’s appearance. In such a case the suspect would plainly have an interest in seeing that the witness is reminded of the photograph, likeness or description immediately before embarking on the video identification procedure. Such a reminder might well reinforce any disinclination to identify a person resembling the suspect. However, after expressing the warning against directing the witness’s attention to any one individual or giving any indication of the suspect’s identity, Annex A.13 continues:
“Where a witness has previously made an identification by photographs, or a computerised or artist’s composite or similar likeness, the witness must not be reminded of such a photograph or composite likeness once a suspect is available for identification by other means in accordance with this Code. Nor must the witness be reminded of any description of the suspect.”
Although the rule relates to the position where the suspect has become “available for identification by other means” it is assumed that the proscription lasts until the conclusion of the witness’s participation in the identification procedure, and in any event the paragraph comes under the section of the Annex captioned “(b) Conducting the video identification.”[439] It is to be noted that the rule in the final sentence against reminding the witness of a description of the suspect is not confined to a description given by the witness.
It is easy to understand the purpose of the rule where the previously viewed photograph or likeness, or a previous description given by the witness, matches the suspect’s appearance. A prior reminder will be equivalent to self-prompting from a previous statement, a form of leading question. However, the same cannot be said where there is an obvious disparity. In such a case the rule might be seen to be designed almost to suppress the disinhibition to commit an error. The mandatory terms of A.13 are regrettable. If it were expressed in more advisory syntax (“should not”) this would give identification officers some measure of latitude in conceding a defence request for a reminder to be given. There is no leeway at all in Annex A.13. However, there may be some amelioration furnished by Annex A.14, which provides that “[a]fter the procedure, action required in accordance with [C]3.40 applies,” that is to say that “each witness involved in [a formal] eye-witness identification procedure . . . shall be asked, after they have taken part, whether they have seen any film, photograph or image broadcast relating to the offence or any descriptions of the suspect which has been broadcast or published as described in [C]3.38(a) and their reply recorded.”[440] This might arguably be interpreted to embrace a description given by the witness which was then made up into a record seen by the witness for the purpose of validating it. So, if a witness makes a video identification and is then, under C3.40, shown a previous contrasting description given by “them” (to use the politically correct if grammatically incorrect syntax of the new Codes) there is the possibility that the witness will immediately resile from, or at any rate express reservations about, the identification. The argument would not be available where the previous description was given but then computer entered without being thereupon visually checked by the witness. Solicitors should be prepared to ask about the provenance of the record of an inconsistent description and, where appropriate, urge identification officers to use A.14 in this way.
(23) Warning that the image of the person seen on a specified earlier occasion
may not be among those shown
(i) The rule
In order to minimise any perception of pressure on witnesses to make an identification and to reduce the “relative judgment problem,”[441] Annex A imports from identification parade procedure the warning, first recommended by Devlin[442] and incorporated into law by Turnbull[443] that immediately before the images are shown, the witness must be told that the person they saw on a specified earlier occasion may, or may not, appear in the images they are shown.[444]
(ii) The suggested use of “blanks”
To reinforce the caution that the person whom the witness saw on the previous occasion may or may not be among the images to be shown and to reduce the risk of error resulting from a relative judgment approach it has been suggested that the practice could be adopted, and written into the code, of presenting the witness with two sets of images, one containing the suspect and the other, a “blank,” comprised entirely of foils.[445] A scheme (obviously for live parades) was originally proposed by witnesses to the Royal Commission on Police Powers and Procedure as long ago as 1929 but not endorsed by the Commissioners[446] and the Devlin Committee similarly thought that the problem of psychological pressure was insufficiently acute to demand such a measure.[447] If the first array were a blank set but the witness made a choice it is argued that there should be no viewing permitted of the second set containing the suspect’s image because the result of the first round would indicate that the witness was prone to guessing.[448] Conversely, if the suspect did not choose any of the first, blank, set of images, he might feel under even more pressure to identify someone in the second set (containing the suspect) than when viewing only one.[449]
(24) Statements of confidence: the question whether to request witnesses to express their degree of certitude when participating in an identification procedure
Recent research suggests that the expressed confidence of an eyewitness in court may be more diagnostic of identification accuracy than psychologists had previously believed.[450] Further, other research suggests that confidence measured before an attempted identification is not as predictive of accurate identification as confidence measured immediately after the identification attempt,[451] which it is argued is one reason why it is good practice to take a clear “statement of confidence” from the witness immediately after the identification attempt and before the formal procedure has ended.[452] Code D does not preclude the taking of a statement of confidence after the identification procedure has been concluded. However, the pre-2017 Code D procedure for conducting a video identification was criticised for requiring witnesses to be told––
“that if they cannot make a positive identification, they should say so.”[453]
The use of the epithet “positive” was criticised in that it made no allowance for the witness to express lesser degrees of confidence crucially during the procedure itself.[454] It was therefore suggested that although the witness might have harboured such thoughts as “I’m fairly sure,” “it might well be number four,” “it looks like him” or “it could be number three,” the form of the instruction included no cue for uttering them, with the result that a qualitative assessment of the identification might well be precluded. If the witness were not in fact asked any qualitative question until trial, a considerable time later, slight misgivings at the time might not be recalled, or an identification might be fortified by the subsequently acquired knowledge of another identification of the same suspect. Again, the witness might simply feel under pressure to express greater certainty before the jury, especially if the prosecution was dependent on the identification.
Instead of using the formulation “positive identification” the idea was canvassed of using alternative questions such as one in open form as to how sure the witness was, or a request to grade the degree of certainty in terms of whether the person was possibly, probably or certainly the same person as previously seen.[455] Again, it was suggested that the degree of confidence could be expressed in terms of a numerical gradation scale of the kind used by scientific experts.[456] During the consultation process which preceded the code revision due to come into force on 1 January, 2006, the Home Office was urged to amend Annex A by introducing in effect a request for a statement of confidence. Thus, it was proposed that use of the phrase “positive identification” should be reconsidered to allow for degrees of certainty to be expressed and that witnesses should be asked to express their degree of certainty on a scale of 1 to 10.[457] Devlin had considered the question “does anyone on the parade closely resemble the person you saw?” but rejected it as liable to confuse some witnesses.[458] Possibly in ignorance of Devlin´s authoritative view, it was also proposed to the Home Office that on failing to make an identification the witness should be asked if anyone shown on the film was similar to the person previously seen.[459] These proposals were rejected by the Home Office on the ground that “the current wording is sufficient and the Codes deliberately only allow for a positive identification because anything less than that opens the potential to flawed identificiation.”
The admittedly sensitive critique of the supposed inflexibility of the use of the expression “positive identification” in the pre-2017 rule nonetheless overlooked the plain thrust of the instruction “that if they cannot make a positive identification, they should say so”––a clear linguistic form of encouragement to witnesses to speak their minds and not be afraid to express reservations. It is difficult to see why such an exhortation was insufficiently effective in eliciting such commonplace phrases expressive of reservations as those instanced above, with the offering of explanatory reasons (eg “the hair looks different, somehow”). In contrast with the idea of using the three degrees of confidence referred to above or the numerical scale, it is arguable that by contrast the nuanced colloquialisms liable to be elicited by the criticised form of entreaty were more apt to express a range of subtle variation, including at one extreme the negative assertion “none of those shown,” than were the relatively inflexible three options or a 10-point gradation scheme.
Criticism was also levelled at the pre-2017 code scheme for its reference exclusively to positive identification rather than also to positive exoneration of the suspect. There was sakid to be “little reason not to ask whether the witness can exclude the persons viewed,” though it has been conceded that such a direct challenge might provoke a qualified identification.[460] While it might have been argued that the positive identification instruction provided a reasonably effective means of prompting the witness (if so minded) to exonerate the suspect without raising any such counter-productive risk, it has been usefully suggested that the code should require witnesses to be cautioned that it is as important to clear possibly innocent persons from suspicion as it is to identify the guilty and that regardless of whether or not an identification is made, the police will continue to investigate the incident.[461]
The response of the Home Office to the debate on the use of the phrase “positive identification” was simply to drop the word “positive” from Annex A.11 in the 2017 edition, which now therefore reads:
“[T]he witness shall be told that if they cannot make an identification, they should say so.”
Whether the bare deletion counts as a material improvement may be regarded as debateable.
(25) Second thoughts
The 2017 revision introduced in new Annex A paragraph A.13A provisions to cater for the position when, after the viewing procedure has ended, the witness informs any police officer or police staff involved in the post-viewing arrangements, that they have made a mistake in their identification or they have not made an identification when in fact they could have made one. In such circumstances an accurate record of the words used by the witness and of the circumstances immediately after the viewing ended must be made.
The key point about A.13A is that it provides that if the witness has not had an opportunity to communicate with other people about the viewing, the identification officer has the discretion to allow the eye-witness a second opportunity to make an identification by repeating the viewing procedure. There is a noteworthy contrast between the phrase “other people” employed in the actual revision and the phrase “other eye-witnesses” which was used in the draft version.[462] The purpose behind the change is far from clear and introduced an inexplicable obfuscation which led to a first instance challenge to admissibility.[463]
(26) The difficult problem of whether to restrict witness scrutiny of the array
(i) Annex A provisions
Witnesses participating in a video identification procedure are shown the moving images in numerical sequence. Annex A provides that the witnesses should be asked not to make any decision as to whether the person they saw is on the set of images until they have seen the whole set at least twice.[464] Once they have seen the whole set of images at least twice and have indicated that they do not want to view the images, or any part of them, again, they shall be asked to say whether the individual they saw in person on a specified earlier occasion has been shown and, if so, to identify them by number of the image.[465] Witnesses will then be shown that image to confirm the identification. There is no upper limit to the number of viewings permissible, nor is there anything in the code to prevent witnesses, after the initial two viewings, from being permitted on request to be shown particular individuals from the array selectively or to switch repeatedly between from one particular image and another. Moreover, witnesses must be advised that at any point, they may ask to see a particular part of the set of images or to have a particular image frozen for them to study and it should be pointed out to the witness that there is no limit on how many times they can view the whole set of images or any part of them.[466]
(ii) Argument that protracted scrutiny may cast doubt on accuracy
It has been argued that the requirement for the witness to look at the array of images twice before making a decision “may represent an ongoing tendency of those involved in the drafting of the Codes and so on, to use pop-psychology, rather than research, to inform their drafting.”[467] Whether or not there is any conscious element here of “pop-psychology,” the requirement certainly appears to reflect an assumption that the more care the witness takes in scrutinising the array the more likely will it be that the eventual choice is accurate. Yet studies of witness accuracy[468] argue against such conventional wisdom, tending to urge instead that(
• faces are represented as “holistic gestalts,” that is, the face is viewed as a whole structure rather than concentrating on individual features and that identification may involve an immediate and automatic matching between a stored gestalt and the external stimulus face;
• accurate witnesses tend not to be able to give a specific reason for their choice;
• successful facial identification is unlikely to require much reflective thought;
• accurate witnesses take significantly less time to identify someone than those who were inaccurate.[469]
The notion that since human faces are essentially similar our possession of the faculty of instantly distinguishing the multitude of details which go to make up each face must be based somehow on an innate ability to assimilate the totality of the image was aptly described in a nineteenth century discourse which foreshadowed research undertaken a century later:
“The difference in human features must be reckoned great, inasmuch as they enable us to distinguish a single known face among those of thousands of strangers, though they are mostly too minute for measurement. At the same time, they are exceedingly numerous. The general expression of a face is the sum of a multitude of small details, which are viewed in such rapid succession that we seem to perceive them all at a single glance. If any one of them disagrees with the recollected traits of a known face, the eye is quick at observing it, and it dwells upon the difference. One small discordance overweighs a multitude of similarities and suggests a general unlikeness.” [470]
(iii) Simultaneous versus sequential viewing: will repeated viewings encourage a relative judgment process with the consequent heightened risk of error?
When a line of people are paraded together in front of a witness for the purposes of identification it is assumed that the witness will scrutinise the array by focusing on each person in turn. In England the live identification parade originally involved the witness walking along the line of participants at close quarters and then touching the shoulder of any person picked out. The physical act of walking along the parade lent it some of the characteristics of a sequential presentation although the opportunity to switch focus instantaneously from one face to another, particularly before and after what amounted to a close-up military-style inspection, essentially defined the procedure as a simultaenous viewing of the array. In its final form the live parade was viewed by the witness at a distance from behind a two-way mirror (with signification orally by number), so that a more global view of the array was facilitated (at the same time as removing the risk of feedback cues from the suspect and disabling eye contact avoidance). In contrast, video identification involves the showing of the images of the participants in sequence, although switching from image to image on the screen at the witness´s request is permissible and capable, with DVD, of being very rapid. It has been suggested that simultaneous presentation, as in a traditional parade, encourages witnesses to engage in a relative judgment process, comparing each person in the line against the others and selecting the one most closely resembling the trace memory of the culprit; in contrast, sequential presentation of the suspect and foils in the showing of a series of video clips tends to engage the witness in an absolute judgment process, comparing each image as it is shown with the memory of the culprit.[471] There is a notion that the relative judgment process induced by simultaneous presentation occasions a tendency to error[472] particularly when it is compounded by the witness’s assumption that the person whom the police already believe to be the culprit is on the parade, in the light of which “a passing likeness can become a positive identity.”[473] Indeed, many studies have shown that when the culprit is absent from the array sequential presentation is significantly less prone than the simultaneous kind to result in mistaken identifications although when the culprit is included in the array the results indicate at least comparable accuracy or even that simultaneous presentation may lead to greater accuracy.[474] However, reference must be made to the results of other research suggesting that simultaneous presentation may be more reliable overall than sequential.[475]
Although more akin to a sequential presentation than the old live form of parade (and certainly lending itself to a pure sequential procedure) the PACE provision for at least two viewings of the array before a decision is voiced plainly differs from a pure sequential presentation. The latter would involve instructing the witness to make a decision on each participant in the compilation immediately after viewing that person, with no initial run-through. It has been argued that with the absence in Annex A of any restriction on the number of times that witnesses are permitted to view the complete set of images, the more will witnesses scrutinise a set of images, and so the more will they tend to compare the individual images with each other and engage in a relative judgment process with the attendant risk of higher rates of false identification.[476] It is not clear how far this prediction is sustainable in the light of results from recent research which indicate no significant reduction in mistaken identifications using pure sequential viewing instructions (decision to be made after viewing each participant) with culprit-absent video parades but a significant reduction in correct identifications using such instructions with culprit-present video parades.[477] The relatively greater suppression of correct identifications over that of incorrect ones with pure sequential presentation as against standard PACE procedure may tell against the introduction of sequential instructions and viewing restrictions. Another, earlier, study suggests that there may be no basis for predicting a diminution of accuracy with increased viewing; indeed, quite the contrary.[478] The participants were asked to view a staged crime, and four weeks later they viewed either a target absent or target present video identification-style parade. One third of participants saw the parade once, one third saw it twice and the final third saw it twice before being allowed to rewind the tape as they chose. For target present parades, performance improved with additional viewing of the parade, and for the target absent parades, the participants who were allowed to rewind the tape at will were more accurate in their decision than the participants who saw the parade once or twice. These results are not wholly consistent with those of Lindsay and others, who reported no change in decisions made after the first and second viewing of a sequential parade.[479] However, it seems that repeated viewing of sequential parades does not necessarily undermine the advantage of the video identification method, in comparison with simultaneous parades.[480]
(iv) Restriction on number of viewings?
Taking all the theoretical and empirical factors reviewed above into account is there a case for changing the current Annex A provisions on viewing the video parade? Should pure sequential instructions be introduced? If not that, should restrictions be imposed on the number of viewings of the array to be allowed witnesses? It has been suggested that the alleged adverse effects of repeated viewing of the array of images can be prevented by permitting witnesses to view the images no more than twice.[481] Should a limit be imposed on the time to be permitted the witness to scrutinise particular faces, whether by replaying the moving sequence of that person of by way of examining a still frame of the face? Roaming freely among the images may well facilitate comparison between them rather than between each and the trace memory of the person originally seen (the “culprit”). Is there a greater danger of miscarriage from the resultant relative judgment process than there would be of the suppression of what might otherwise be a reliable identification if restrictions were imposed? It would have to be demonstrated on the basis of the theoretical considerations considered here combined with the totality of the available research findings that the risk of miscarriage so outweighed that of suppressing reliable identifications through the imposition of restrictions that the case for such restrictions became strong enough to overcome the undoubted axiom of received wisdom that more scrutiny means more accuracy. The combination of theory and empirical evidence seems inconclusive.
(v) Possible impact of exhortation to examine images without limit[482]
Apart from the reliability issue, the absence of any restriction in Annex A on the number of times the array of images may be viewed may have another adverse effect. It may be interpreted by the police as a licence to encourage witnesses to proliferate the number of repeats and, although Annex A.11 requires witnesses to be warned that the culprit may, or may not, appear in the images they are to be shown, this may in turn induce an assumption that the police believe the culprit to be present. The consequence may be that the witness will feel under pressure and be more inclined to make a guess.
(vi) Impact of research on trial process
Where a witness has taken a noticeably long time to make an identification, it may be feasible to prevail on at least some judges to allow in expert evidence expounding a combination of theoretical considerations and those research results which establish a negative correlation between length of scrutiny and accuracy, since these findings clearly controvert conventional wisdom, or “common sense.”
(27) Preventing reinforcement feedback which may happen when identification witnesses are informed that they have picked out the suspect
A significant shortcoming in the Code which may need to be redressed in any future revision is that it imposes no prohibition on witnesses being told that they have either identified the suspect or have failed to do so. It has been argued that there is a danger in such confirmation that a witness who may have harboured some doubt could feel fortified in an initially tentative identification. The process of such reinforcement of confidence has been described as “bolstering” or “firming up.”[483] Such a contention appears to be supported by research demonstrating not only that receiving feedback that the suspect has been identified, or that another witness has made the same identification, will increase the witness´s confidence in their identification, but also that it may inflate a range of assertions tending to enhance the professed reliability of the original sighting, for example, its duration, the quality of lighting conditions, the proximity of the suspect to the witness, the witness´s attentiveness, and so forth.[484] Other research demonstrates that post-identification feedback tends to make eyewitnesses over-confident, expressing more certainty in their identification than may be merited.[485]
It has also been suggested that informing witnesses of a failure to identify the suspect may induce them to adjust their evidence, although how exactly is not explained.[486] It is assumed that what is meant by “adjustment” in this prognosis is the danger that a witness who thinks he may have recognised the person but says nothing out of a punctilious regard for fairness and propriety may feel less inhibited about implicating the person he thinks he may have recognised once he receives confirmation that the suspect was among the images depicted (whether or not he is told at which number the suspect appeared in the array).
The Court of Appeal have advised that a witness ought not to be told that the person picked out was the suspect until after making a statement about the parade.[487] This will prevent a suspiciously rapid process of firming up between the identification procedure and the making of the routine statement which is usually taken immediately afterwards. However, it is difficult to see what significant protection it will offer against potentially much more insidious long-term bolstering of a tentative recognition as a result of being told subsequently to the making of a statement that the right person was picked out. Thus, it has been persuasively argued[488] that the principle of non-disclosure ought to be maintained right up to trial(
“[F]airness to the suspect surely requires that the witness be not told at any time before trial whether the person identified was the suspect, unless there is a particular reason for doing so. The reality is that any witness is going to be fortified in a belief as to the correctness of the identification by being given such information. It is no answer to say that, if the witness says in a post-parade statement, ‘I think it was the person I identified, but I am not entirely sure’ and at trial says ‘I am in fact 100% sure,’ the evidence at trial may be undermined by cross-examination on the previous statement. It is notorious that honest but mistaken identification witnesses can be convincing. If the witness says that he has given it considerable thought, and has in consequence become sure, this may be extremely plausible. This may be so even if the jury are informed of the fact that the witness has been told that the suspect was the person identified.”
It has been reported that in the light of pronouncements by a number of Crown Court judges the policy has been adopted in some police areas (including, notably Kent) of not confirming or denying the accuracy of any identification made by a witness.[489] This local practice ought to be incorporated in the Code as a provision. As the prevention of secret approaches to a witness would be virtually impossible to enforce it may be worth considering whether to introduce a routine practice of withholding the details of the procedure and of the witness´s performance from officers involved in the investigation, even from the officer in the case. The details would be confidential to the supervising identification officer and the Crown Prosecution Service.
(28) Preventing non-vocal cues during the identification process
The process of reinforcement referred to in the last section may conceivably take place during the identification procedure itself, with the supervising officer indicating or signalling to the witness that a tentative identification had indeed been correct. This may be conveyed in speech or by some other vocal cue (a cough or grunt) or by a non-vocal cue (the raising of an eyebrow or a head tilt) or by a combination. Non-verbal and non-vocal cues may also induce the witness to assert an identification in the first instance, expressed either with certainty or tentatively. Audio recording of the procedure will pick up vocal cues, but only comprehensive video monitoring of the whole procedure, with the positioning of multiple high definition cameras in the room will furnish direct evidence of non-vocal ones. While identification procedure administrators can be deterred from giving conscious signals by the presence of the suspect’s legal representative or by the installation of video/audio monitoring such safeguards will not preclude unconscious ones or those which might be given at a “dead” point out of view of the cameras or when the suspect´s representative, if present, is not looking. Comprehensive video monitoring of the room could be combined with the current best practice use of the random selection by the witness of one from a number of DVDs each with the suspect shown in a different position in the array, and the use of twin back-to-back monitors, one for the witness and the other for the operator (and the suspect’s legal representative, if present) showing only the number identifying the position in the array being presented.[490] It has already been mentioned[491] that as an added safeguard Code D could be amended to require that the officer or member of police civilian staff taking the witness through the array (not the identification officer in overall charge of the operation) should be unaware of the identity and appearance of the suspect. The legal representative should also not know in which position the suspect appears on the array, to avoid any possibility of an unconscious cue coming from that source (such as the drawing in of breath).
C. Documentation and Validation
(1) Written records
A record of the conduct of the video identification must be made on forms provided for the purpose.[492] In particular, a record must be made of all those participating in, or seeing, the set of images whose names are known to the police.[493] It must include anything said by the witness about any identifications or the conduct of the procedure and any reasons it was not practicable to comply with any of the provisions of Code D governing the conduct of video identifications.[494] If identical conditions for each image are not practicable the reasonmust be noted on the form.[495] If the suspect or legal representative raises any objections to the array which cannot be met, they should also be recorded together with the reasons why they were not met.[496]
(2) Image security
Arrangements must be made for all relevant material containing set of images used for specific identification procedures to be kept securely and their movements accounted for.[497] They will therefore be available for later examination by the parties and for court use. (It has already been mentioned that no one involved in the investigation shall be permitted to view the material prior to it being shown to any witness.[498])
(3) Image destruction
Unless the suspect is charged, prosecuted, cautioned, warned reprimanded or if the suspect gives consent to its retention, the images of the suspect must be destroyed save where necessary to retain it pursuant to the Criminal Procedure and Investigations Act 1996.[499]
(4) Flaw in the validation requirements for video identification rectified
(i) Live parades
It is a key component of any reliable identification procedure that an unassailable record be maintained of the behaviour and reactions of the witness to the live participants of a parade or to video images as they are displayed. Written records are subjective, prone to error and can hardly convey the full flavour of the events they describe. Still photography of the volunteers to a live parade can of course provide a record for assessing resemblance to the suspect (although the Devlin Committee thought it might be deceptive as to colour[500]) but can hardly furnish a record of the identifying witness’s behaviour. The obvious solution lies in technology but without explanation the Devlin Committee believed that a “ciné film” would not be as valuable as a still photograph.[501] The original edition of Code D required a video or photograph of the parade where it was held without a solicitor or friend of the suspect being present but, following a recommendation of the Royal Commission on Criminal Justice,[502] the qualification was removed from the 1995 revision and it became mandatory to take a colour photograph or to video-record the parade. In the 2003 version the rule was refined still further and stipulated that a video film of a live parade must be made unless it is impracticable, in which case a colour photograph must be taken.[503]
(ii) The fundamental flaw regarding video parades
Prior to the 2017 Code D revision there was a serious flaw of video identification in that Annex A.9, inter alia, laid down the following requirement(
“In the absence of the suspect´s representative, the viewing itself shall be recorded on video.”
The meaning was clear: if a defending solicitor was available to observe in person the showing of images there was no obligation on the part of the police to maintain a video and sound record. This stood in stark contrast to the mandatory requirement for live identification parades. The qualification for video identification was the same as the pre-1995 rule for live parades. It is not apparent why it should have stuck in a time-warp and it may well be that the distinction was unintentional.[504] It has been noted that some constabularies would make a video recording of the identification as a matter of routine.[505] Ideally, the record for case preparation and court presentation would consist of a split screen in which the images are shown and the witness’s viewing of them are recorded and displayed simultaneously.[506]
Where the viewing was not routinely recorded in all cases irrespective of the presence or absence of a legal representative the only way for the suspect to ensure that the procedure was video-recorded was to instruct the solicitor to stay away from the procedure.[507] It is inconceivable that the rule was designed to force a choice between having a solicitor or a video-recording. After all, a mounted camera is no guarantee against surreptitious or unconscious signalling, which an eagle-eyed defence representative in attendance might very well pick up.
(iii) Illustration of the importance of video identification being itself video-recorded
It has been suggested that unless professionally produced with discreetly positioned cameras, video recording of the video identification procedure would probably not be of sufficient quality to be of real value, would not reduce the number of identification errors, and might inhibit witnesses from making an identification if they knew they were being filmed.[508] Lest it be thought that too much is being made here of the need to have the procedure video-recorded where a solicitor is present, it must be stressed that its importance is not to be underestimated. This can be demonstrated with a dramatic illustration from the author’s own practice. The defendant was accused of being one of two men involved in a road rage attack. Suspicion originally focused on him after the registration number of the assailants’ lorry was traced to his cousin, with whom he was believed to work in scrap metal dealing and he was arrested in the gaoler’s office at a magistrates’ court where he was being released on bail in connection with another unrelated case. On being arrested he was asked a number of questions about the incident and allegedly made incriminating replies. These he subsequently denied making, refusing to verify the officers’ notes of arrest. In any event the questions should not have been asked because, being under arrest, he was not in a lawful place of detention for PACE purposes and in any event the questions should only have been asked on tape.[509] In the presence of his solicitor he was put up on an identification parade, which was video-taped, chose position number 6, and the committal statements disclosed that the witness said “I’m not sure, but I think it’s number six.” The defendant was charged and released on bail, but went on the run. Two years later he was re-arrested. The video of the parade had been mislaid. At the plea and directions hearing the judge was told that it was proposed to seek the exclusion of the alleged admissions and further to argue that the witness’s assertion was equivocal and incapable of sustaining a case to answer. His response was that there was Court of Appeal sanction for the sufficiency of the assertion although he did not specify the authority.[510] The Crown undertook to make every effort to trace the video-cassette of the parade and eventually it surfaced. It revealed that the witness had not in fact said “I think it’s number six.” What he had actually said was “I think it’s number seven.” However, the identification officer apparently misheard the defendant( perhaps he was unconsciously predisposed to hear the witness identify number six, the defendant(and he repeated aloud as he wrote down the remark “I think it’s number six.” The witness made no comment. At a renewed directions hearing the same judge observed, provocatively, that as the witness did not demur from the officer’s “repetition” he could be said to have adopted it as a “correction.” Fortunately, prosecuting counsel furnished a very strong advice which the CPS eventually accepted.[511]
The witness’s actual remark had been missed by the defendant’s legal representative. It was only the video which picked it up. This, then, underscores the singular importance of video monitoring, or at the very least an audio-recording of the proceedings even where a solicitor is present.
(iv) Options open to solicitors on validation prior to the 2017 revision
(a) Requesting the identification officer to video-record the procedure with a solicitor present Although Annex A.9 in the pre-2017 edition of Code D only made a video-recording of the video identification procedure mandatory in the absence of a solicitor’s representative, this would not have precluded the Identification Officer from agreeing that a video recording ought to be made of the procedure. After all, the facility existed for cases where the solicitor did not attend and any solicitor who was proposing to attend ought to have pressed for the facility to be used even though Code D did not require this under the pre-2017 regime. If the request were denied the solicitor could always have attended and could have made an audio-tape of what had been said. Whether this ought advisedly to have been done openly or whether covertly to avoid the risk of an objection from the Identification Officer would have been for the solicitor to decide. A solicitor might have chosen to bring along a camcorder but the identification officer might conceivably have refused to allow it to be used, perhaps on the ground that it would intimidate or inhibit the witness.
(b) Objecting to video identification where video-recording request was declined What if the solicitor (on instructions) objected to a video-identification on the grounds that the police were insisting on sticking to the letter of Annex A.9, and were refusing to make a video-recording of the procedure where the legal representative was going to be present? As it still does, D3.15 stipulated(
“A suspect who refuses the identification procedure first offered shall be asked to state their reason for refusing and may get advice from their solicitor and/or appropriate adult. The suspect, solicitor and/or appropriate adult shall be allowed to make representations about why another procedure should be used. A record should be made of the reasons for refusal and any representations made. After considering any reasons given, and representations made, the identification officer shall, if appropriate, arrange for the suspect to be offered an alternative which the officer considers suitable and practicable. If the officer decides it is not suitable and practicable to offer an alternative procedure, the reasons for that decision shall be recorded.”
Where the suspect’s solicitor requested a live identification parade on the grounds that this was the only way to ensure that a video record was kept of the procedure if the solicitor was to be present it was likely that if the Identification Officer had been so obtuse as to refuse to have the video identification video-recorded the officer would very likely have refused a request for a live parade as an alternative. In obvious contrast with a live parade, a video identification does not require the suspect’s consent or co-operation. If the alternative of a live parade was refused by the police the suspect could always choose not to co-operate in the video identification procedure.
(v) Proposed ending the suspect’s right to have a solicitor present at the video identification
The Home Office Consultation Paper Modernising Police Powers: Review of the Police and Criminal Evidence Act (PACE) 1984, published in March 2007, contained three paragraphs (3.39 to 3.41) inviting submissions in respect of the question whether the suspect’s absolute right to have a solicitor in attendance when witnesses participate in a video dentification should be removed:
“3.39 Currently, PACE Code D stipulates that the suspect must be given a reasonable opportunity to have a solicitor or friend present at the time and place a victim or witness is asked to make a video identification. This is known to have an adverse affect on the ability of some victims or witnesses to make a fair and accurate identification. It also places an additional burden on the police and demands on legal advisers which, if the viewing by the victim or witness is itself videoed, adds little to the safeguards.
3.40 For the identification officer, it creates particular problems when a victim or witness is unable to travel and the officer considers it appropriate to arrange the viewing at the person’s home. Therefore we want to look at ways in which we are able to remove or minimise any sense of fear or intimidation and encourage victims and witnesses to confidently participate in identification procedures.
3.41 We also want to examine how technology can be used to protect the identity of victims and witnesses and how best to use places other than the police station to carry out identification procedures. In all of this, we must ensure that sufficient safeguards are in place which make sure that any new processes are subject to appropriate scrutiny and accountability. . . .”
The author collaborated in a submission to the Home Office upon which the observations which follow are largely based.
In asserting that the presence of a defence solicitor at a video identification “is known to have an adverse effect on the ability of some victims or witnesses to make a fair and accurate identification” the Consultation Paper cited no systematic research. Presumably it was based on informed supposition that some vulnerable witnesses might fear that a solicitor would identify them to the suspect or pass on details of their appearance, and that this might lead to their being harmed. Warranted or not such fears might have posed a very real problem for investigators at what must often have been a delicate and decisive phase of the inquiry. It would certainly have been unfortunate if the presence of the suspect’s legal representative stultified a prosecution before it had even got off the ground. Provided a system of safeguards of the kind described earlier for preventing the communicating of cues were to have been maintained, including video monitoring of the room where the video identification session was to be conducted,[512] there would seem to have been very little practical advantage for the suspect in having a solicitor present, beyond of course the subjective benefit that it might have caused the witness to be inhibited from making an identification!
Portable system For obvious reasons any nervousness on the part of a witness about the presence of the suspect’s legal representative would have been likely to be heightened where the video identification was to take place at the witness’s home. The system described earlier was eminently adaptable for portable use at the witness’s home or in hospital.
CCTV monitoring of the precincts The provision of relatively inexpensive CCTV in the witness waiting area to monitor the arrangements for preventing cross-contamination would have served to obviate any justification for solicitors to attend in order to keep watch at vastly greater expense.
The question of preserving witness anonymity The authors of the Consultation Paper stated, in paragraph 3.41, that they wished to examine how technology might be used to protect the identity of victims and witnesses. If no identification is made there will be little reason to disclose to the defence the video-recording of the process and therefore the appearance of the witness. However, where a positive identification is made and is relied upon by the prosecution it is an essential component of the process of fair trial that defence lawyers, at least, should be able to scrutinise the facial expressions and demeanour of the witness viewing the array of images. To ensure a fair trial through effective legal representation it will not therefore be possible to conceal the witness’s identity, in the sense of appearance, from defence lawyers. The principle will be no different from the provision of screens in court. However, to ensure that the defendant does not become privy to the witness’s identity in a case of vulnerability it may be necessary to make disclosure of the video-recording of the identification process conditional on its non-disclosure to the defendant. This may easily be achieved by non-release of copies, as in some cases of sensitive “ABE” (Achieving Best Evidence) video interviews involving child victims of sexual depredation. In such cases the defence legal team would only be permitted to view the video-recording in private at a police station, in the absence of the defendant. Later, in court, where the witness gives evidence from behind a screen and the defence require the jury to view the video-recording this would be accomplished without the defendant being afforded a view of the video monitor.
(vi) Rectification of the fundamental flaw by the 2017 Code D revision: video and sound record now obligatory for video identification
The passages in this section addressing the erstwhile absence of a requirement for the viewing by eye-witnesses of images in the video identification process itself to be video and sound recorded were set out in previous postings of the present treatise. Whether or not they contributed to the catalyst for change is not known but it is gratifying that the flaw was at last corrected with the issuing of the 2017 edition of Code D. Now, the process must be video-recorded with sound: Annex A.9.
The recording must include a clear view of the face of the eye-witness as they look at the images, with an audio recording of what they say and what is said to them by the identification officer and by any other person present at the showing. If the suspect is identified, a supervised viewing of the recording by the suspect and their solicitor (or both) must be arranged, if requested.[513]
The requirement to maintain a video-recording of the identification process ought to provide an unassailable record obviating any need for additional contemporaneous live monitoring by the suspect’s legal representative. Prior to the 2017 Code D revision the requirement for the process to be video-recorded was conditional on the suspect’s solicitor being absent. The removal of that condition and its replacement by an absolute requirement for a video record has effectively removed any justification for the solicitor to be present observing the process. Thus Annex A.9 states that the solicitor’s rȏle with regard to the images shown to the eye-witnesses is confined to that set out in A.7, namely the raising of any reasonable objection to the suitability of the complete set of images to be used in the identification process. A.9 states that since the solicitor will be able to check the conduct of the procedure by viewing the recording, they may only attend the viewing on request and with the prior agreement of the identification officer if the officer is satisfied the the attendance will not deteror distract the eye-witnesses from viewing the images. A.9 stipulates that in any event defending solicitors are not entitled to communicate in any way with the the eye-wintesses during the viewing procedure and indeed it is difficult to envisage any circumstances in which they could properly seek to request justify such communication.
Chapter 5
The Evidence in Court
A. Admissibility of Evidence of the Signifying
of a Visual Identification
(1) Modes of signification
Street identifications will characteristically and typically consist of a combined gesture of pointing and denotive speech. Identification of a photograph will be spoken. The older procedure for conducting identification parades involved the witness touching the selected person’s shoulder, and words might sometimes be used in addition. Video identification will necessarily involve the use of words to signify the number of the image selected, although witnesses with hearing or speech impairment may well employ gesture in lieu of speech.
(2) Principles of admissiblity
(i) Evidence of positive identification
Nowadays it is considered generally of little evidential value, prejudicial and unacceptable for witnesses to identify the defendant in the dock as a person seen on a relevant occasion, although nominally that might well constitute direct evidence.[514] However, what in contrast will be of real importance is evidence conveying to the jury what a witness said or did by way of identifying the defendant on a particular significant occasion, such as in the street shortly after the perpetration of an offence or during a controlled formal procedure such as a video identification. The courts have long been faced with the problem of having to define a legal basis for the admissibility of evidence narrating such significations.[515]
Conventionally, identification witnesses are permitted to give evidence that they made a street identification or that they identified a particular person during a formal identification procedure. Similarly police officers are conventionally allowed to give evidence that an eyewitness made a positive identification. The admissibility of such evidence would not now seriously be challenged. However, the precise legal basis for the admissibility of the combination of the evidence of the identifying witness and of a supporting witness remains uncertain and undefined. In Christie[516] a child gave evidence that the defendant was his attacker although not that he had identified the defendant at the scene but the fact that he had done so was permitted to be adduced by his mother and a police officer apparently on the basis that it showed consistency on the part of the complainant. In McKay[517] the identifying witness had identified the suspect by number at a live parade and in evidence told of having done so but was no longer able to recall the number and that evidence was adduced by the identification officer. The Court of Appeal found it unnecessary to decide whether dmissibility was on the basis that signification by number was an act of identification and therefore real or original evidence or whether it was a category of exception to the hearsay rule. The court foudn support for admissibility in the structure of Code D itself and there seems little doubt that time-honoured practice has established its own legitimacy, irrespective of the precise legal category of the evidence.
In practice the faltering of the witness’s memory over whether an identification was made or as to the number of the person or image selected can easily be rectified by showing the witness the official contemporaneous record of the procedure. The relevant portions record the witness’s words and those portions must be signed by the witness. The signature is evidence that the entries are the witness’s own and a witness may therefore properly be invited to use it as an aide memoire. Thus, it may be contemplated that the official signed record will normally be utilised in this way to overcome the potential evidential hiatus caused by memory frailty of the kind which occurred in Osbourne and Virtue.[518] Two witnesses had identified one of the appellants at a parade but one stated at trial that she could not remember identifying anyone and the other said that she did not think that the person she had picked was the defendant in the dock. It was held that the evidence of the identification officer that they had each identified one of the appellants was admissible, a decision which it has been observed is difficult to reconcile with any principle of evidence.[519] As the hearsay rule was not referred to in the judgment the court were not professing to recognise an exception to the rule but such an exception would hardly have worked any injustice in a similar case provided the documentation is sufficiently comprehensive to ensure the elimination of error in recording what was said and done. In any event, in allowing proper memory refreshment such documentation will preclude the need to fall back on a hearsay rule exception. Moreover, video-recording of the identification procedure is now in common use and may well be admissible as a specie of real evidence.
(ii) Failure to identify and qualified identification
In the interests of being seen to be fair the prosecution will normally lead evidence of the failure by a witness to make an identification or to the fact that a witness professed to exonerate the suspect (for example by declaring that the culprit was not among the images shown). However, such extreme positions are probably less common in practice than qualified identifications or those expressed with reservations, to which reference has already been made.[520] The whole question of the basis of the admissibility of such intermediate utterances was considered in depth by a cause celebre in the history of modern identification cases, the decision in George[521] In that case a number of witnesses gave a variety of qualified, not to say attenuated, assertions pointing to the possibility of an accurate identification of the suspect as the person seen on the relevant occasion. The Court of Appeal upheld the decision to admit the evidence of the totality of remarks, acknowledging that although a defendant was not to be convicted on the evidence of a qualified identification alone, it might be relevant in at least two circumstances: first, where it supports or at least is consistent with other incriminating evidence, and secondly where the explanation for a non or qualified identification may help to place the non or qualified identification in its proper context and so, for example, show that other evidence given by the witness may still be correct. Otherwise, the court observed, a non or qualified identification could be used to attack the credibility of other evidence given by a witness when the explanation may show that such an attack is unjustified. On the other hand, warned the court, prosecuting counsel must be cautious and avoid conducting the examination of a witness who had failed to make a positive identification in a manner which suggests that “but for this fact or that fact” the witness would have made a positive identification. In the court’s words,
“An identification which is qualified cannot be transformed into one which is unqualified by careful questioning.”
On the facts of the case the evidence was properly admitted as supporting an “underlying unity of description” of the assailant and general consistency of the identification evidence and supported “a considerable volume of other circumstantial evidence.” The prosecution were entitled to place before the jury the qualified identification in the context of the real and substantial problems which faced the witnesses resulting from the delayed procedure and the change in the appellant’s appearance.
B. Exclusion for Code Breach
(1) Discretionary exclusion of identification evidence in the event of code breach
(i) The general principle of exclusion for breach of Code D
The breach of any provision of a PACE code of practice may occasion an application to the court of trial to exercise its discretion under s.78 of the PACE Act by excluding evidence to which the breach relates in the interests of promoting fairness in the proceedings. In assessing the issue of fairness the court is required by s.78 to take account of the circumstances in which the evidence was obtained. Although non-compliance with the Code is not decisive but “only a factor, maybe a cogent factor in the decision to admit or exclude evidence” it has been said that as the overall purpose of Code D is one of adopting fair identification practices and adducing reliable identification evidence, insufficient regard to those purposes is likely to be met by the discretionary exclusion of evidence and convictions are liable to be treated as unsafe.[522] The breach of a code provision requiring a procedure to be conducted in a certain way may impugn the fairness of that particular procedure and vitiate the evidence obtained in consequence. Other breaches may be adjuged to have caused unfairness indirectly, as where the failure to hold a video identification under controlled conditions deprived the suspect of the chance to test a prior informal street identification, albeit one which was in itself code-compliant.
Where the defence raise a complaint of a breach of Code D, the court of trial will be required first of all to rule on whether the police were in breach, and then, if the complaint is upheld, the court will have to determine whether to exclude the evidence to which the breach relates. Reasons must normally be given for allowing in evidence in spite of an admitted or proven breach.[523] On appeal against conviction from a trial on indictment the appeal court will be required to consider whether or not the trial judge was in error in ruling that no breach occurred. If the complaint of breach is upheld it will mean that the trial judge failed to exercise his discretion under s.78 to exclude the evidence of identification.[524] The court will then in effect exercise its own discretion in determining whether the evidence ought to have been excluded. If it holds that the evidence ought to have been excluded this will result in the conviction being quashed if the court concludes that the verdict of the jury might have been different and the conviction was therefore unsafe. If on appeal the complaint is that although the trial judge upheld the breach but declined to exclude the evidence to which it related, as a matter of general practice the appeal tribunal will rarely interfere unless it concludes that the discretion was exercised unreasonably.
(2) Examples of typical breaches requiring the exercise of discretion
(i) Exclusion of evidence not subsequently tested by a mandatory controlled procedure
(a) Street identification Where, following a street identification no formal procedure is carried out in breach of Code D, the court of trial will be required first of all to rule on whether the police were in breach, and then, if such a ruling is given, consideration will have to be given to the question whether to exclude the original identification or the evidence of purported recognition under PACE, s.78.[525] Against the background of these principles the actual decision in the landmark authority of Forbes warrants critical examination, having regard to the fact that the House of Lords (i) laid as much stress as they did on the importance of holding an identification parade in order to put the reliability of an earlier street identification to the test and (ii) emphasised that the mandatory obligation to hold an identification procedure was not displaced merely because there had been a “fully satisfactory” or “actual and complete” or “unequivocal” identification by the witness previously. The complainant had been driven by a friend to a cashpoint machine where he withdrew £10. He was then approached by a man who blocked his path, asked for money on compassionate grounds, became aggressive when he was refused, pursued the complainant, threatened to “cut him up” and, standing very close to him, revealed what looked like the handle of a knife. The complainant managed to evade him and was driven away from the scene in his friend’s car. Subsequently they passed the man in the street and, as they did so, the complainant and the man made eye contact and the man spat towards the car as it went by. The complainant called the police on his mobile telephone and gave a description. A few minutes later he was driven around the streets in a police car and positively identified the appellant as his assailant. The appellant denied the accusation and three times before the trial the defence asked in vein for an identification parade. Holding that the evidence of the street identification was rightly admitted in spite of the code breach the House of Lords concurred with the Court of Appeal that(
“. . . [t]he evidence was compelling and untainted, and . . . did not suffer from such problems or weaknesses as sometimes attend evidence of this kind: as, for example, where the suspect is already visibly in the hands of the police at the moment he is identified to them by the complainant.”[526]
The Appeals Committee expressed the opinion that in the event of an (unwarranted) failure to hold an identification procedure the jury should ordinarily be told that such a procedure enabled a suspect to put the reliability of an eyewitness’s identification to the test, that the suspect had lost the benefit of that safeguard and that they should take account of that fact in their assessment of the whole case, giving it such weight as they thought fit.[527] In the instant case there had been no such direction because the trial judge had ruled wrongly that the decision not to hold a parade did not constitute a code breach. However, surprisingly, the failure to give such a direction likewise did not, in the view of the Lords, infringe the appellant’s right to a fair trial or render the conviction unsafe. So saying, their Lordships adopted the following passage from the judgment in the Court of Appeal(
“As we have already said, the evidence of street identification was compelling and untainted. It would be wholly artificial to suppose that a reasonable jury might have taken a different view if they had been told that the appellant had been deprived of the chance that the complainant might not have picked out the appellant on a parade. Again, we bear in mind that there were in effect two informal identifications.”
What was being said here in effect was that no reasonable jury would have been likely to entertain doubts about the reliability of the street identification even if they accepted that, had there been a parade (as there should have been), the complainant might have failed to identify the appellant on the parade. Cutting to the chase, the use of the phrase “wholly artificial” was a more elegant way of saying that to have held a parade would have been a “farce.” Yet, what was lost was the possibility that, after due time for reflection and confronted by a range of persons resembling the appellant, the complainant might actually have expressed doubts about the original identification, or might even have exonerated him. There may have been two identifications, but this was entirely to beg the question of whether the complainant in purporting to identify the appellant to the police was correctly identifying the man who had spat at him (who it may be presumed was the assailant). In the judgement of the House of Lords while a fully satisfactory, actual, complete and unequivocal identification could not negate the obligation to hold a Code D procedure( because such identifications have proved in the past to be “wholly wrong”(it did not matter that an identification which was compelling and untainted was not tested with the obligatory Code D follow-up procedure. The distinction between the two sets of epithets is obscure.
(b) Identification by photographs Where in the case of a suspect whose identity is not at the time known a witness makes an identification from an album of photographs the failure to follow up with a formal controlled procedure may render the original selection unfair in depriving the defendant of the means of testing it and the breach may be visited by exclusion of the evidence of the original photographic identification.[528]
(c) Dispute that the defendant is well-known to the witnesses Where a witness claims that he recognised the defendant as a person well-known to him but the defendant denies or challenges the assertion the failure to test the claim under formal controlled conditions may result in the exclusion of the evidence of the claim of recognition.[529]
(d) Description Evidence of a description may be excluded if it was accompanied by an expression of ability to make an identification but the police have improperly and unfairly failed to hold an identification procedure.[530]
(ii) Exclusion of purported identification by confrontation
(a) Improper resort to confrontation A court would be almost certain to exclude evidence of an identification following a confrontation which the police held without properly resorting to a controlled procedure.[531]
(b) Improper conduct of confrontation Evidence of identification by a confrontation may be liable to exclusion if a confrontation, though permissible in the circumstances, is carried out without proper compliance with the rules for confrontation.[532]
(iii) Exclusion of evidence of street identification which should not have been held or by reason of defect in its conduct
Evidence of a street identification may be excluded if the suspect was known at the time and it should not have been carried out.[533] On the other hand, while it may not have been improper to hold a street procedure the resulting evidence of identification may be excluded because the failure to comply with one or more of the many applicable rule for carrying it out vitiated its fairness and reliability.[534]
(iv) Formal procedure tainted by flawed street or photographic identification
Just as evidence of a street identification or identification by inspection of a photographic gallery may be excluded because it was not followed up by a formal procedure, so evidence of an identification made as the result of a formal controlled procedure such as a video identification may be excluded by reason of having been tainted by a flawed street identification procedure which preceded it.
(v) Other procedural breaches in conducting a formal controlled procedure
Breaches of the Code in conducting a formal procedure may result in exclusion (or the quashing of a conviction where appropriate) if it is so serious as to vitiate the fairness and reliability of the identification. However, it may also be excluded if the court considers that the transgressed rule is of such importance that as a matter of principle exclusion ought follow in the interests of ensuring the appearance of fairness, in effect the maintenance of standards. A variety of examples is illustrative.
(a) Involvement of investigating officer in the procedure The involvement of the investigating officer in the procedure in contravention of the long established rule may well be regarded as the breach of a fundamental principle serious enough to merit exclusion in spite of a finding that no actual contamination was likely to have occurred.[535]
(b) Communication between witnesses Where witnesses participating in an identification procedure have been in a position to communicate with each other and so influence the selection this may taint the porcess sufficiently to justify exclusion or the quashing of a conviction.[536]
(c) Requirement for resemblance It is of crucial importance that in any array of photographs or images the persons depicted should resemble each other and the suspect. It will obviously be a fundamentally serious code breach of the suspect’s appearance differs markedly from that of the other persons whose pictures or images are shown and this may vitiate the fairness of the procedure.[537] However, where the suspect or the defence legal representative have been able to vet the selection of images to be shown in a video identification it may be difficult to sustain a complaint of unfairness.[538]
C. Directions to the Jury
(1) Regina v Turnbull
The jury must be given a “Turnbull” direction in any case of visual identification.[539]
(2) Directions where identification evidence is allowed despite the upholding of a submission of code breach
It was stressed in Forbes[540] that if the judge upholds a complaint that Code D has been contravened but nevertheless exercises his discretion to allow in the evidence of identification to which the provision relates he must then give the jury an appropriate direction as to the relevance and effect of the breach, explaining that there has been a breach and how it has arisen and that the defendant has lost the benefit of the particular safeguard in question. They must then be told that in assessing the whole of the evidence they should give the breach such weight as they think fit.
-----------------------
[1] Report to the Secretary of State for the Home Department of the Departmental Committee on Evidence of Identification in Criminal Cases, H.C. 338, 26 April 1976, paras 1.10 and 5.31. As will be explained in the course of this treatise video identification parades have now superceded the traditional live parade.
[2] Personal Identity, London: Michael Joseph, 1957, p.32.
[3] Devlin Report, para. 5.29.
[4] See Roughead, W. (ed.), The Trial of Adolf Beck, Notable British Trials, London: William Hodge, 1924; Report of the Committee of Enquiry into the case of Mr Adolf Beck, published 14 November 1904, Cd. 2315; Watson, E.R. (ed.), The Trial of Adolf Beck 1924. For a recent summary of the case see Bogan, P., Identification: Investigation, trial and scientific evidence, London: Legal Action Group 2004, pp. 2-4. Another miscarriage of justice still often cited as an object lesson for mistaken identification was the case of Oscar Slater, convicted of murder in 1909 but eventually cleared after serving 19 years in prison: see Roughead, W. (ed.), The Trial of Oscar Slater, Notable British Trials, London: William Hodge, 1929; Hunt, P, Oscar Slater, The Great Suspect, London: Carroll & Nicholson, 1951. Both the Beck and Slater cases are described and discussed at length in Rolph, C.H., op. cit., caps. 6 and 7. For other works, academic and popular, on the dangers of visual identification see Wall, P.M., Eye-Witness Identification in Criminal Cases, London: 1965; Brandon, R., and Davies, C., Wrongful Imprisonment: Mistaken convictions and their consequences, London: Alen & Unwin, 1973, Chap. 2; Coles P., and Pringle, P., Can You Positively Identify this Man?: George Ince and the Barn Murder, London: Andre Deutsch, 1974; Hain, P., Mistaken Identity, London: Quartet, 1976; Huff, C.R., Ratner, A. And Sagarin, E., “Guilty until proven innocent: Wrongful conviction and public policy,” (1986) 32 Crime and Delinquency 518-544; Cutler, B.L., and Penrod, S.D., Mistaken Identification: The Eyewitness, Psychology and the Law, Cambridge: Cambridge University Press, 1995.
[5] Devlin Report, para. 5.29. Prior to the Devlin Report the identification parade rules received close attention in the report of the Royal Commission on Police Powers and Procedure, 1929, Cmd. 3297, paras. 123-129.
[6] See McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, A., Shepherd, E., and Wolchover, D. (eds.), Analysing Witness Testimony, London: Blackstone Press, 1999, pp 178-193, at p.179.
[7] See above, note 1. The two cases were those of Luke Dougherty and Laszlo Virag.
[8] (1977) 98 Cr.App.R. 313, C.A.
[9] McKenzie and Dunk, supra, p.179.
[10] In a United States study of 1,000 convictions of innocent persons eyewitness identification error was found to be the single most important factor: Wells G.L., ‘What do we know about eyewitness identification?’ (1993) American Psychologist, vol. 48, pp.553-571. The US Innocence Project documents mistaken identification as a factor in the wrongful conviction of 61 of the first 70 (87 pc) of a total (by August 2005) of 161 cases of exoneration by D.N.A. evidence which had not been available at trial: ., cited by Valentine, T., ‘Forensic Facial Identification,’ in Heaton-Armstrong, A., Shepherd, E., Gudjonsson, G., and Wolchover, D. (eds.) Witness Testimony: Psychological, Investigative and Evidential Perspectives, Oxford: OUP, 2006, chap 17. Similar projects have been instituted in Australia (see . html and the UK (see ). See also Scheck, B., Neufield, P. and Dwyer, J., Actual Innocence, New York: Doubleday, 2000.
[11] Coles and Pringle, cited above at n.4.
[12] McKenzie and Dunk, supra, p.179.
[13] Bull, R., Bustin, B., Evans, P., and Gahagan, D. (eds.), Psychology for Police Officers, Chichester: Wiley, 1983, cited in ibid.
[14] Ibid.
[15] Bruce, V., “Changing faces: visual and non-visual coding processes in face recognition,” British Journal of Psychology (1982) 73, pp.105-116.
[16] See Valentine, “Forensic Facial Identification,” cited above at n.10, paras 17.18 and 17.19, citing Wells, G.L., “What do we know about eyewitness identification?” (1993) 48 American Psychologist 553-571. See also Valentine, paras 17.29 and 17.30, and Wright, D.B. and McDaid, A.T., “Comparing system and estimator variables using data from real lineups” (1996) 10 Applied Cognitive Psychology 75-84.
[17] There has been a wealth of research on the degree to which identification accuracy may be impaired by stress levels occasioned by the incident in question. For a leading account and comprehensive bibliography see generally Loftus, E. F., Eyewitness Testimony, Cambridge, Mass: Harvard University Press, 1996 (2nd ed; 1st ed, 1979). For a good summary and more recent select sources see Loftus, E. F., Wolchover, D., and Page D., “General Review of the Psychology of Witness Testimony,” in Heaton-Armstrong, A., Shepherd, E., Gudjonsson, G., and Wolchover, D. (eds.) Witness Testimony: Psychological, Investigative and Evidential Perspectives, above n.10, chap. 1, para 1.41. Most recently see Valentine, T., and Mesout, J., “Eyewitness identification under stress in the London Dungeon,” (2009) Applied Cognitive Psychology, 23, 151-161 (witnesses who experience high stress show a catastrophic failure of witness ID).
[18] H.O. 109/1978.
[19] See Police and Criminal Evidence Act 1984 (Codes of Practice)(Temporary Modifications to Code D) Order 2002 (S.I. 2002, No.615).
[20] The Police and Criminal Evidence Act 1984 (Codes of Practice) Order 2005 (S.I. 2005 No. 3503), in force from 1.01.06; The Police and Criminal Evidence Act 1984 (Codes of Practice) (Revision of Codes A, B and D) Order 2011 (S.I. 2011 No. 412), in force from 14.02.11; The Police and Criminal Evidence Act 1984 (Codes of Practice) (Revision of Codes C, D and H) Order 2017 (S.I. 2017 No. 103) in force from 23.02.17.
[21] D3.28 in the 2006 edition, and amended and expanded in the 2013 revision. The paragraph was removed for the 2017 edition and is now not used.
[22] Report, para. 8.12.
[23] [2001] 1 Cr.App.R. 430, at para. 26.
[24] PACE Act 1984, s.78.
[25] Under ibid. s.67(11). As to these issues see Chapter 5, below.
[26] A notorious exception is the case of Peter Hain, the then leader of the Young Liberals, an internationally prominent campaigner against Apartheid, and later a senior cabinet minister in the Labour Government, who in 1975 was ludicrously arrested and subsequently tried for bank theft: see his account Mistaken Identity: The Wrong Face of the Law: London: Quartet, 1976.
[27] The decisions in point are considered further at pp.76-78, below.
[28] Indeed, no distinction is recognised: Grimer [1982] Crim.L.R. 674, C.A.
[29] Blenkinsop [1995] 1 Cr.App.R. 7, C.A. It has been suggested that the same rationale would allow a solicitor’s representative to exculpate his client on the basis of sitting in on thes client’s short interview at the police station: Bogan, op. cit., p.178.
[30] Ibid.
[31] See further below at p.40-42.
[32]Attorney-General’s Reference (No. 2 of 2002) [2003] 1 Cr.App.R. 21, C.A.
[33] [1995] 2 Cr.App.R. 333, C.A.
[34] Supra.
[35] This may be the result of the phenomenon termed “unconscious identity transference”: see further pp.43-44and 78-79, below. For further consideration of the possible implications of exposure to social media images see below, Chapter 2, at p.39.
[36] [2015] Crim.L.R. 534, C.A., considering Forbes [2001] 1 A.C. 473, H.L. (CLW/00/45/2); Att.-Gen.’s Reference (No. 2 of 2002) [2003] 1 Cr.App.R. 321(21), C.A.; Smith (Dean Martin) and others [2009] 1 Cr.App.R. 521 (36), C.A.; Moss [2011] Crim.L.R. 560, C.A.
[37] The revision had come into force in March 2011. Although the incident had occurred in April the formal identification procedures were not undertaken until April 2012, a year later.
[38] D3.1A (formerly 3.4 of the 2013 edition). The definition dates from the 1995 revision in which its status was merely that of Note for Guidance 2E. It is consonant with the legal justification for an arrest by a police constable under the PACE Act, requiring the existence of reasonable grounds for suspecting a person to be guilty of an offence: 24(5)(b).
[39] Chapman v DPP (1988) 89 Cr.App. R. 190, C.A.; Castorina v Chief Constable of Surrey [1996] LGRR 241, C.A.; O’Hara v Chief Constable of the Royal Ulster Constabulary [1997] 1 Cr.App.R. 447, H.L.
[40] Kitchen [1994] Crim.L.R. 684, C.A. (no breach in showing photographs because officer’s suspicion held to have been based on instinct rather than reasonable grounds); El Hannachi and others [1998] 2 Cr.App.R. 226, C.A. (street identification of four youths standing with officers, one of whom stated that they would have been arrested regardless of an identification and it was submitted therefore that they were known; held, no breach because in spite of the officer’s frankness the information was plainly insufficient to justify arrest); K v DPP [2003] EWHC 351 (Admin); [2003] 4 Archbold News 1 (street identification held to be in breach where the accused matched a description and had been tracked by a police dog from the scene of a robbery in the early hours).
[41] To the existence of reasonable grounds for suspicion it was proposed to add the words “and the officer who for the time being is in charge of the investigation of the offence in question believes that it is necessary for the proper and effective conduct of the investigation of that offence to invite them to take part in at least of the [formal] eyewitness identification procedures in [Part (A)]”.
[42] [1997] C.O.D. 91, D.C.
[43] [2001] EWCA Crim 2283; 10 Archbold News 1, C.A.
[44] See also, e.g., K. v DPP [2003] EWHC 351 (Admin); [2003] 4 Archbold News 1 (facts at n.40, above); Bush (1997) unreported, 27 January, C.A., cited by Bogan, op. cit., para.3.22, n.21 (assailant was known to be inside a nightclub which had been sealed off by the police; appellant matched the description of long hair and tattoos and on being asked to stand aside from the crowd was identified by] the complainant). Cf. Rogers [1993] Crim.L.R.386, C.A. (no breach where, shortly after a man with slurred speech was observed damaging cars on an industrial estate, the police found the appellant in a warehouse near by and although his speech was slurred, he was not wearing the pullover described by eyewitnesses; the arrest criterion was applied although the second edition of the Code which was in force at the time did not then include the definition of “known” suspects; for commentary on the case see Bogan, op. cit., para. 3.22, n.20). In Gornall 149 S.J. 300, C.A. (judgment 3 March 2005) a householder disturbed a man who then fled with one or more others into the garden of an adjacent house. She told the police the man was wearing a white t-shirt and in the adjacent house the police found twin brothers, one of whom had on a white t-shirt. The police first brought the other brother to the householder who said he was not the man she had seen. The other brother was then brought to her and she identified him. It was held that before the identification the identity of the suspect was not known and that it would have been extraordinary if the police had not done as they did. In commentary on the decision the editor of Criminal LawWeek points out that the very brief details in the report do not disclose whether the brothers lived in the adjacent house or were merely found hiding there, whether any other candidates were found, the time between the escape and the brothers being found by the police or whether they were arrested before the identification: see CLW/05/09/4. Nevertheless he rightly suggests that as the accused fitted the description and was found where the offenders had been seen to go and as the brother had been eliminated these facts alone ought to have been found to provide the necessary reasonable grounds for suspicion to require a formal procedure.
[45] In Williams [2003] EWCA 3200; (2003) SJ 1305, the appellant’s detention clearly amounted to an arrest but the Court of Appeal declined to rule on whether the grounds were sufficient to make him a known suspect.
[46] Cumming and others v Chief Constable of Northumbia Police [2003] EWCA Civ 1884 99 (action for wrongful arrest).
[47] See Bogan, op. cit. para. 3.26.
[48] Evidence (General), Cmnd. 4991, June 1972, para. 196.
[49] Report, paras. 5.15 and 8.10.
[50] (1977) 63 Cr.App.R. 132.
[51] Report, Cm. 2263, July 1993, Chapter 2, para. 10, recommendation 4, responding to submissions based on various academic treatments. See, e.g., Heaton-Armstrong, A., “Identification(description of suspects” [1986] Crim.L.R. 215; Heaton-Armstrong, A., and Wolchover, D., “Exorcising Dougherty’s Ghost,” (1991) 141 New Law Journal 137; “Recording Witness Statements” [1992] Crim.L.R. 160, at p.165. On the other hand, a senior officer appointed to draft policy on identification procedures for the Northern force to which he is attached has informed the author that failure to record early first descriptions among officers of his own force is endemic.
[52] The semantic problems arising from inadequately distinguishing between the competing terms “culprit” and “suspect” are further considered below, at pp.62-63, in commentary on D3.12.
[53] (2013) unreported, July 5, C.A. For commentary see [2013] Law Soc. Gaz., December 2, p.21.
[54] Moreover, it was held, a tactical decision not to deploy the note at the trial was not unreasonable. If the contents had been used, while some further inconsistencies in the witness’s description would have emerged, there would have been a correspondingly increased concentration on the assailant’s face, on which, at the time, she was actually focusing her attention. The reality was that the correctness of the identification of the appellant was given powerful support by the remaining evidence. Without the benefit of the contents of the note the jury entertained no reservations about it.
[55] Home Office, August 1994, D2.0.
[56] Wolchover, D., and Heaton-Armstrong, A., “Questioning and Identification: Changes under P.A.C.E. ’95” [1995] Crim.L.R. 356, at p.367.
[57] Eg Kent Constabulary: see McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, A., Shepherd, E., and Wolchover D. (eds.), Analysing Witness Testimony, London: Blackstone Press 1999, p.182.
[58] See Wolchover and Heaton-Armstrong, [1995] Crim.L.R., at p.368.
[59] [2015] Crim.L.R.631, C.A.
[60] D3.29
[61] See Fisher, R.P., and Cutler, B.L., “The relationship between consistency and accuracy of eyewitness testimony,” in Davies, G., Lloyd-Bostock, S., McMurran, M., amd Wilson, C. (eds.) Psychology and Law: Advances in Research, Berlin: De Gruyter, 1995 (participants viewed a mock theft and were rated for consistency in a series of descriptions given to the researchers).
[62] Cutler, B.L., Penrod, S.D., and Martens, T.K., “The reliability of eyewitness identification: the role of system and and estimator variables,” Law and Human Behaviour (1987) 11, 233-258.
[63] Farrington, D.P., and Lambert, S., Offender Profiling: Theory, Research and Practice, Chichester: Wiley, 1997 (chapter entitled “Predicting offender profiles from victim and witness descriptions”; 1,069 descriptions of burglary or violence in Nottinghamshire); Tollestrup, P.A., Turtle, J.W. and Yuille, J.C. “Actual victims and witnesses to robbery and fraud: an archival analysis,” in Ross, D.F., Read, J.D., and Toglia, M.P. (eds.) Adult Eyewitness Testimony: Current Trends and Developments, Cambridge: Cambridge University Press, 1994, pp.144-162. A good deal of research on general description appears to indicate that these are the categories under which witnesses usually provide information: see McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, A., Shepherd, E., and Wolchover D. (eds.), Analysing Witness Testimony, London: Blackstone Press 1999, p.182.
[64] See Kebbell, M.R., and Wagstaff, G.F., “Face Value? Evaluating the Accuracy of Eyewitness Information,” Police Research Series Paper 102, Policing and Reducing Crime Unit, Home Office 1999, p.27. By contrast with descriptive ability, cross-racial identification seems to be more difficult than same-race identification, studies showing that people from one racial background have more difficulty identifying individual members of another race than do members of their own race: see eg Valentine, T., Pickering, A. and Darling, S., “Characteristics of eyewitness identification that predict the outcome of real lineups” (2003) 17 Applied Cognitive Psychology 969-993. Cf. Pike, G., Brace, N. and Kyman, S., The visual identification of suspects: procedures and practice, Briefing note 2/02, Policing and Reducing Crime Unit, Home Office Research Development and Statistics Directorate, 2002, reporting no effect of cross-race identification on the outcome of live parades.
[65] See Lindsay, S., “Memory source monitoring and eyewitness testimony,” in Ross, D., Ceci, S.J., Dunning, D., and Toglia, M., (eds.) Adult eyewitness Testimony, Cambridge: Cambridge University Press, 1994, p.27, at pp.45-46. See also Valentine, T., and Maras, K. (in press), “The effect of cross-examination on the accuracy of adult eyewitness testimony,” in Applied Cognitive Psychology (showing that cross-examination can reduce the accuracy of eyewitness testimony.)
[66] For a good general summary with comprehensive bibliography see B.R. Clifford, and A. Memon, “Obtaining detailed testimony: the cognitive interview,” in Heaton-Armstrong, A., Shepherd, S., and Wolchover, D. (eds.) Analysing Witness Testimony, London: Blackstone Press, 1999, pp.146-161.
[67] McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in ibid., at pp.183-184.
[68] Ibid., p.185.
[69] See Bogan, op. cit., p.179.
[70] Reiterated in D5.12. Provision is also made by s.64A and D5.12 for the photographing of persons who have been made to wait with a police community support officer (PCSO), given various kinds of penalty notice or fixed penalty notice by a constable in uniform, PCSO or other accredited person.
[71] D.5.12A, condition (a). Note for Guidance 5E states that examples of when it would not be practicable to obtain the person’s consent include (a) when a person is drunk or otherwise unfit to give consent, (b) where there are reasonable grounds to suspect that if the person became aware a photograph, suitable to be used or disclosed for the use and disclosure in the prevention or detection of crime, the investigation of offences or the conduct of prosecutions, they would take steps to prevent it being taken, eg by violently resisting, covering or distorting their face, etc, and it would not otherwise by possible to take a suitable photograph, (c) when, in order to to obtain a suitable photograph, it is necessary to take it covertly, and (d) in the case of a juvenile, if the parent or guardian cannot be contacted in sufficient time to allow the photograph to be taken.
[72] D5.13.
[73] D5.14. Note for Guidance 5F states that the use of reasonable force to take the photograph of a suspect elsewhere than at a police station must be carefully considered. In order to obtain a suspect’s consent and co-operation to remove an item of religious headware to take their photograph, a constable should consider whether in the circumstances of the situation the removal and the taking of the photograph should be by an officer of the same sex as the person. It would be appropriate for these actions to be conducted out of public view.
[74] D5.18.
[75] D5.15.
[76] D5.16.
[77] Ibid.
[78] D5.17.
[79] D5.18.
[80] D5.19 and 5.20.
[81] D5.21(c).
[82] Section 54A(1) and D5.1 and 5.4.
[83] D5.5.
[84] D5.4. Note for Guidance 5D states that examples of when it would not be practicable to obtain a detainee’s consent to the taking of a photograph of an identifying mark include (a) when the person is drunk or otherwise unfit to give consent, (b) when there are reasonable grounds to suspect that if the person became aware that an identifying mark was to be photographed, he or she would take steps to prevent this happening, e.g. by violently resisting, covering or cbncealing the mark and it would not otherwise be possible to photograph the identifying mark, (c) in the case of a juvenile, if the parent or guardian cannot be contacted in sufficient time to allow the photograph to be taken.
[85] D5.21(d).
[86] PACE Act, s.64A and D.12A, condition (b). Note for Guidance 5B sets out a long list of examples of purposes related to the prevention or detection of crime, the investigation of offences or the conduct of prosecutions.
[87] D5.6.
[88] D5.22. The uses to which retained photographs may be put are restricted to those given in D5.6 (identifying marks) and D5.12A (photographs of the person): ibid. The person must be given the opportunity to witness the destruction or to have a certificate confirming the destruction if they request one within five days of being informed that the destruction is required: D5.23.
[89] Those set out in D3.5 to 3.10, 3.21, and 3.23, ie, video identification format using still images, live parade, group identification and confrontation. The reference to confrontation is anomalous: in contrast with live parades and group identifications Annex D makes no provision for any photographs to be made or used in connection with confrontations. Nothing in D3.31 affects any separate requirement under the Criminal Procedure and Investigations Act 1996 to retain material in connectrion with criminal investigations: D3.33. D3.32 replicates the requirements of D5.23 for notice of destruction.
[90] Tolson (1864) 4 F. & F. 103.
[91] Taylor v Chief Constable of Cheshire (1987) 84 Cr.App.R. 191, C.A.
[92] Fowden and White [1982] Crim.L.R. 589, C.A. (video); Maqsud Ali [1966] 1Q.B. 688, C.A. (audio).
[93] Kajala v Noble (1982) 75 Cr.App.R. 149, D.C. Queries as to provenance and therefore admissibility may be avoided by adducing continuity evidence and maintaining a comprehensive record of any copying, alteration or enhancement, particularly involving expert evidence: see Bogan, op. cit., pp176-177.
[94] Taylor v Chief Constable of Cheshire, supra.(CCTV recording of shop premises accidentally erased after being viewed by police officers).
[95] PACE Act, s.78.
[96] See above, p. 19.
[97] As a general rule, the Crown should neither adduce evidence, nor refer to, police photographs: Lamb 71 Cr.App.R. 198, C.A. The prejudicial effect of adducing such evidence of bad character usually outweighs any probative value. However, the crown might, unusually, lead such evidence where the accused’s criminal record has been adduced by the defence: Allen [1996] Crim.L.R. 426, C.A.
[98] Bogan, op. cit., paras. 3.10, p.24.
[99] Ibid., and paras. 3.12, p.24, citing Kitchen [1994] Crim.L.R. 684 where the suspect had escaped from custody and the equivalent provision under the 1991 edition (D2.18) was held to permit an identification by photographs: see further below.
[100] See Lindsay, R.C.L., Nosworthy, G.J., Martin, R., and Martynuck, C., “Using mug shots to find suspects,” Journal of Applied Psychology (1994) 79, 121-130.
[101] Gorenstein, G.W., and Ellsworth, P., “Effect of choosing an incorrect photograph and later identification by an eyewitness,” Journal of Applied Psychology (1980) 65, 616-622.
[102] With Home Office funding the University of Aberdeen developed a system of coding witness descriptions to compile a relatively smaller number of photographs for inspection, leading to more accurate performance than was possible using the full albums: see Shepherd, J., “An interactive computer system for retrieving faces,” in Ellis, H.D., Jeeves, M.A., Newcombe, F., and Young, A., Aspects of face processing, Dordrecht: Martinus Nijhoff, 1986, pp.398-409.
[103] R v Governor of Pentonville Prison ex p Voets [1986] 1W.L.R. 470, D.C.
[104] Bogan, op. cit., at p.186.
[105] (1980) 71 Cr.App.R. 198, C.A. It was held that exceptionally it might be led where the suspect has refused to co-operate in a subsequent identification procedure but in Williams (2000) unreported, February 22, C.A., it was held that the availability of covert procedures precluded any need for recourse to a police photograph. See also Bleakley [1993] Crim.L.R. 203, C.A. where the defendant tried to speak to the witness the evening before the parade at which she identified him. It was suggested by the defence that that was the reason why she had identified him at the parade and, in order to avoid their being misled, the prosecution were allowed to refer to her original identification of him from a police photograph.
[106] Bruce, V., and Young, A., In the Eye of the Beholder, Oxford University Press, 1998, note at pp.180-181, that the construct from an assemblage of isolated features is reminiscent of Leonardo da Vinci’s attempt to construct an inventory of all possible facial features in order to provide instruction on the drawing of portraits from a single glance at a face: see his Trattato, quoted by Gombrich, E.H., The image and the eye: Further studies in the psychology of pictorial representation, Oxford: Phaidon Press, 1976.
[107] E-fit was developed jointly by researchers at the University of Aberdeen, the Home Office and in the private sector: Bruce and Young, cited above at n.11, p.181.
[108] See Valentine, “Forensic Facial Identification,” cited in Chapter 1 above at n.10, para 17.31.
[109] D3.3.
[110] For a review of the relevant research see Davies, G.M. and Valentine, T., “Facial composites: forensic utility and psychological research,” in Lindsay, R.C.L., Ross, D.F., Read, J.D. and Toglia, M.P., Handbook of eyewitness psychology, Mahwah: LEA, in press, vol 2, “Memory for people,” cited in ibid.
[111] See Kovera, M.B., Penrod, S.D., Pappas, C., and Thill, D.L. “Identification of computer-generated facial composites,” Journal of Applied Psychology (1997) 82, 235-246. See also Ellis, H.D., Davies, G.M., and Shepherd, J.W., “A critical examination of the Photofit system for recalling faces,” Ergonomics (1978) 21, pp.297-307; Ellis, H.D., “Face recall: A psychological perspective,” Human Learning (1986) 5, pp.189-196. Bruce and Young, op cit., at p.181, have suggested that the poverty of performance with Photofit likenesses may be as poor as it is because the method is based on the assumption that a face representation can be deconstructed into its component parts, whereas face recognition is probably holistic (as to which see below, Chapter 4, p.142). Furthermore, relationships between different features are at least as important as the features themselves, but Photofit offers only limited manipulation of features. Although E-fit is designed around a basic kit of parts of faces if affords much more opportunity for global manipulations, blending and elaboration than with photofit and the resulting images are more realistic.
[112] See Kebbell, M.R., and Wagstaff, G.F., supra, p.25.
[113] “Forensic Facial Identification” (above, Chap 1, at n.10), paras 17.14 and 17.15, citing Frowd, C., Hancock, P. J. B. and Carson, D., ‘EvoFIT: A holistic evolutionary facial imaging technique for creating composites’ (2004) 1 Transactions on Applied Psychology 1-21 (Association for Computing Machinery); Frowd, C., Carson, D., Ness, H., McQuiston-Surrett, D., Richardson, J., Baldwin, H. and Hancock, P., ‘Contemporary composite techniques: The impact of a forensically-relevant target delay’ (2005) 10 Legal and Criminological Psychology 63-91 (in press); Gibson, S., Pallares Bejarano, A. and Solomon, C., ‘Synthesis of photographic quality facial composites using evolutionary algorithms,’ in Harvey, R. and Bangham, J. A., (eds.) Proceedings of the British Machine Vision Conference 2003, London: British Machine Vision Association, 2003, 221-230; Valentine, T., Davis, J. P., Thorner, K., Solomon, C., and Gibson, S., “Evolving and combining facial composites: Between-witness and within-witness morphs compared,” (2010) Journal of Experimental Psychology: Applied, 61, 72-86 (showing that the new system of Efit (Efit V) works well especially when multiple composite from different witnesses or even the same witness are merged.)
[114] See The Sunday Times, 25 November 2007. The research was published by Frowd, C.D., Bruce, V., Ross, D., McIntyre, A. and Hancock, P.J.B., “An application of caricature: how to improve the recognition of facial composites” (2007) 15 Visual Cognition, 1-31 (forthcoming), available in microsoft word at 3.doc.
[115] According to The Sunday Times report (above) the system involves distorting and arranging facial features and does so “by marking the different elements of the face – from its overall shape to the size of the nose and separation of the eyes – with 200 points. Each point is then compared with the position of its counterpart on an average face of the same sex, age and racial group, produced by merging large numbers of different images. Any differences from the norm are then exaggerated to generate a caricature. Different versions of the face are produced, ranging from an extreme caricature to a blander version with all the exaggerations removed. Conventional static Photofits often include mistakes that can prompt the brain to reject a particular match. The animated caricature system overcomes this by including the blander version of the face, allowing poorly recalled features to be ignored. By morphing from the blander version to the extreme caricature, the brain is able to detect a possible match more often.” According to the researchers, “the identity of the face seems to ‘pop out’ for one frame in the sequence.”
[116] See Chapter 1, p.23, above.
[117] Research by Cornish, referred to without full citation by McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, A., Shepherd, E., and Wolchover, D. (eds.), Analysing Witness Testimony, London: Blackstone Press, 1999, pp 178-193, at p.184, note.
[118] Cook (1987) 84 Cr.App.R. 369. Its admissibility may survive a subsequent identification which is less than satisfactory (ibid.), the failure of the witness to pick out the suspect at an identification parade (Okorudu [1982] Crim.L.R. 747, CCC, but cf. O’Brien et al [1982] Crim LR 746, CCC) or the exclusion of a later positive identification as inadmissible (Constantinou (1990) 91 Cr.App.R.74, C.A.). See Bogan, op. cit., p.176.
[119] See below pp.44-45.
[120] Annex E.1, which makes reference to D3.11 on the various duties of identification officers.
[121] Where it is proposed to show photographs to a witness in accorrdance with Annex E it is the responsibility of the officer in charge of the investigation to confirm to the officer responsible for superivsing and directing the showing, that the first description of the suspect given by that witness has been recorded: Note for guidance 3E. If this description has not been recorded, the procedure under Annex E must be postponed: ibid.
[122] Annex E.3.
[123] Annex E.3.
[124] Annex E.3.
[125] Annex E.4.
[126] Annex E.5.
[127] Annex E.5. The phrase formerly used was “positive identification”: see infra, p.138, et seq.
[128] Annex E.5.
[129] Annex E.5.
[130] Annex E.6.
[131] Annex E.6.
[132] Annex E.7.
[133] Annex E.8.
[134] Annex E.9.
[135] Annex E.9.
[136] Annex E.10.
[137] Annex E.10.
[138] Annex E.11.
[139] Annex E.11.
[140] Annex E.12.
[141] (2012) 177 J.P. 73, C.A.
[142] For an account of the evidential issues involved in the use of such images see Valentine, “Forensic facial identification,” cited above at Chapter 1, n.10, section H. For a recent study see Davis, J. P., and Valentine, T., “CCTV on trial: Matching video images with the defendant in the dock,” (2009) 23 Applied Cognitive Psychology, 482-505 (reporting error rates in human face matching from CCTV images; average is about 20%, but for some particular faces can be as high as 45 per cent).
[143] [2009] 1 Cr.App.R. 521 (36). For a further case of recognition following internal police circulation of CCTV footage see Lariba [2015] Crim.L.R. 534, C.A., cited above at p.13.
[144] [2011] Crim.L.R. 560, C.A. (commentary by Andrew Roberts).
[145] See eg Fowden and White [1982] Crim.L.R. 588, C.A. (conviction quashed where a store detective and a police officer viewing a video-recording of a shop-lifting incident had been allowed to give evidence of recognition of the suspect in relation to another alleged incident a week later).
[146] As to which see below, pp.44-46, under paragraph (3), “Use of CCTV footage or stills in court.”
[147] D3.34 to D.3.37.
[148] D3.38.
[149] Ross, D., Ceci, S.J., Dunning, A. and Toglia, M., “Unconscious transference and line up identification, toward a memory blending approach,” in Ross, et al. (eds) Adult eyewitness Testimony, Cambridge: Cambridge University Press, 1994, pp.80-100. For a graphic illustration of the phenomenon in which an image of the man later tried (though never more than obliquely “identified”) was published under the banner “bomber”: see Wolchover, D., Culprits of Lockerbie, online, DavidWolchover.co.uk.
[150] See McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, A., Shepherd, E., and Wolchover, D. (eds.), Analysing Witness Testimony, London: Blackstone Press, 1999, pp 178-193, at p.185.
[151] D3.39. The provision does not affect any separate requirement under the Criminal Procedure and Investigations Act 1996 to retain material in connection with criminal investigations that might apply to sub-paras (a) and (b) of D3.38: ibid. It is noteworthy that the paragraph is silent about the retention of material shown to police officers under D3.38.
[152]D3.41. The record must be made in accordance with the provisions of D3.38 insofar as they can be applied to the viewing in question: ibid. Note for Guidance stresses the importance of making the record as soon as practicable after the viewing and whilst it is fresh in the mind of the individual who makes the recognition.
[153] D3.40.
[154] Ibid. Subject to the eye-witness’s recollection, the record required by D3.41 should be completed: D3.40.
[155] Attorney-General’s Reference (No. 2 of 2002) [2003] 1 Cr.App.R. 21, C.A. But cf Morrisey (2014) 78 J.C.L. 460, C.A. (appellant convicted of shoplifting three times on same day; CCTV images of insufficient quality to permit an identification showed the thief to be a woman matching her description; appellant admitted to police she had visited the shop three times on that day but denied theft and later claimed that she had lied in order to protect her twin sister who had previous convictions and adduced photographs of the sister; after retiring jury asked to see CCTV footage again; held, they should have been warned against “performing an impermissible detection task” of determining the identity of the thief based on a comparison of the CCTV images, the sister’s photographs and the appearance of the appellant in the dock, but appeal nonetheless dismissed); for trenchant criticism of the judgment see Criminal Law Week commentary, CLW/15/1/2).
[156] [1995] 2 Cr.App.R. 333, C.A. (above, p. 10).
[157] Dodson and Williams (1984) 79 Cr.App.R. 220, C.A.; Downey [1995] 1 Cr.App.R. 547, C.A.
[158] However, a defendant may not be compelled to stand up in the dock for an inspection by the jury. No adverse inference is permissible from his refusal to do so: R v McNamara [1996] Crim.L.R. 750, C.A. To avoid prejudice and the need for a post hoc corrective direction the prosecution ought to give warning in the absence of the jury that they would seek to invite an inspection.
[159] [2015] 2 Cr.App.R.215(14), C.A, preferring Downey, supra, to Blenkinsop [1995] 1 Cr.App.R. 7, C.A.
[160] CLW/15/31/2; Turnbull [1977] Q.B. 224, C.A. As the editor points out, in Shanmugarajah it was unclear as to exactly what the jury were being asked to scrutinise. According to the judgment the judge began his directions “on the photograph. He directed them that they should very much consider the quality and the clarity of the image before they went any further. He also directed them that it would be ‘completely wrong’ for the jury . . . to do nothing more than start comparing the photograph and the still from the CCTV footage.” The passage relates to a comparison between images – that of the crime scene and one indisputably of the appellant – not between images and the defendant in the dock.
[161] See, eg, Brown [1991] Crim.L.R. 368, C.A.
[162] See, eg, Coulman v D.P.P. [1997] C.O.D. 91 Q.B.D. In the leading case of Forbes [2001] 1 A.C., at p.487, the witness spotted the assailant minutes after the offence, called the police and was then driven around by them until he pointed out the appellant as his assailant.
[163] Noted by Devlin, para 4.16. In Wait [1998] Crim. L.R. 68, C.A., the complanaint saw two of his assailants some days after the attack and flagged down a police vehicle whereupon they were arrested.
[164] In Popat [1998] 2 Cr.App.R. 208, C.A., the victim of an attack subsequently saw her assailant in the same vicinity and then, over a number of days, kept observation there with a police officer until identifying the assailant.
[165] Jones and others (1994) 158 J.P. 293, C.A.
[166] “In cases when the suspect’s identity is not known, a witness may be taken to a particular neighbourhood or place to see whether they can identify the person they saw on a previous occasion.”
[167] See Bogan, op.cit., paras. 3.27 and 10.65.
[168] In one case of memory the complainant was robbed of money in a dimly lit street during a weekend neighbourhood carnival. It emerged in cross-examination that the fleeting view he had of the face of his black assailant was when, lying on the tarmac, he looked up at the man, who was looking down at him with his head silhouetted against a distant street light behind him. On agreed evidence the only street light in front of the robber was a considerable distance away and it was strongly argued that the contours of his face must have been difficult to distinguish. Minutes later the complainant reported the offence at a mobile police post located at the junction of two roads on the carnival route and as he left the post the defendant was passing by chance across the junction among a large and milling crowd. The complainant professed to recognise him as the robber and pointed him out to two police officers who immediately arrested the defendant. No identification parade was held but the case went to the jury anyway on account of alleged admissions by the defendant. The defence called a member of their professional team who gave evidence that the normal lighting at the junction was too poor to illuminate the faces of black people passing by at the time of the experiment. The police evidence was tainted by discrepancies and the defendant was happily acquitted.
[169] See Vaughan (1997) The Independent 12 May, C.A. (judgment 30 April). Such a preconception may have been reinforced as a result of overhearing police conversation: see Bogan, op.cit., para. 5.14, p.65.
[170] See Malashev [1997] Crim.L.R. 587.
[171] See Roberts, A. and Clover, S., “Managerialism and Myopia: The Government’s Consultation Draft on PACE(Code D,” [2002] Crim.L.R. 873, at p.888; Clover, S. and Roberts, A., “Short-sighted or forward-looking?” (2002) New Law Journal 870, at p.872 (June 7).
[172] Based on guidance provided by Mitchell J in Hickin [1996] Crim.L.R. 584, C.A. In her commentary at p.586, Professor Birch endorsed the view that further attention needed to be given in the Code to informal identification procedures.
[173] Hickin [1996] Crim.L.R. 584.
[174] (1997) The Independent 12 May, C.A. (judgment 30 April). See also Miah [2001] EWCA Crim 2281.
[175] Bogan, op. cit., para 5.17, p.66.
[176] Ryan v D.P.P. (2000) unreported, 10 October, C.A.
[177] Vaughan (1997) The Independent 12 May, C.A. (judgment 30 April).
[178] McMath [1997] Crim.L.R. 586, C.A.
[179] See, eg, Crowe, unreported C.A. 6 March 1998, and El Hannachi [1998] Crim.L.R. 881, C.A.
[180] Duggan (1998) unreported 16 June, C.A. It has been noted of the decision that the “admitted failure to take a first description was compensated for by the description available from the 999 call [and so] the breach caused no prejudice”: Bogan, op. cit., para 5.20, p.67. However, it is difficult to see why any code breach should have been regarded as having occurred when the tape-recording was clearly a record made by the police.
[181] D3.12.
[182] Bogan, op. cit., para 5.25, p.68, citing at n.27, McMath [1997] Crim.L.R. 586, C.A.: “In an ideal world, there would only be one potential witness in each car to preserve the integrity of each and every identification.”
[183] D3.2(e).
[184] The provision was applied in Fuller (2000) unreported, 22 May, C.A. Until it was excised in the 2011 edition the sub-paragraph continued: “Subject to paragraphs 3.12 and 3.13, it is not necessary for the witness who makes such a positive identification to take part in a further procedure.” See further, chapter 3, section B.(2) (“Continuation of the strict approach under the code regime after 2002”).
[185] [1996] Crim.L.R. 584, C.A.
[186] (1994) unreported, 10 March, C.A. See Bogan, op. cit., para 5.23, p.67.
[187] Emphasis supplied. Note for Guidance 3F warns that the admissibility and value of any street identification may be compromised if, before a person is identified, the witness’s attention is specifically drawn to that person.
[188] See McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, A., Shepherd, E., and Wolchover, D. (eds.), Analysing Witness Testimony, London: Blackstone Press 1999, p.185.
[189] Roberts and Clover, [2002] Crim.L.R., at p.888.
[190] Ibid.
[191] Coulman v D.P.P. [1997] C.O.D. 91, D.C.
[192] See eg Nunes [2001] EWCA Crim 2283; [2001] 10 Archbold News 1, C.A.
[193] In June 2016, when the previous edition of the present work was posted, no statistics had been published on the frequency or efficacy of street identifications in England and Wales. Estimates suggest that “showups,” as they are known in the United States, account for between 30 and 77 pc of all identification procedures in that country: see Stebley, N., Dysart, J., Fulero, S. And Lindsay, R.C.L., “Eyewitness accuracy rates in police showup and lineup presentations: A meta-analystic comparison” (2003) 27 Law and Human Behaviour 523-540 In a sample of 224 identification procedures recorded by a police officer 172 were showups and 52 were photospreads, with an identification being made in 22 pc of the showups and 56 pc of the photospreads and in laboratory studies the identification rate from showups was also lower than from photographic lineups: Gonzales, R., Ellesworth, P. and Pembroke, M., “Response biases in lineups and showups” (1993) 64 Journal of Personality and Social Psychology 525-537. By contrast, in another, archival, American study the identification rates fom showups were higher than from photospreads, with 48 pc of witnesses making an identification in 284 photographic lineups and 77 pc of witnesses making an identification in 258 showups: Behrman, B.W., and Davey, s.I. “Eyewitness identification in actual criminal cases: an archival analysis” (2001) 25 Law and Human Behavious 475-491. With the exception of the study by Gonzales et al the laboratory studies involved presentation of a photograph in the showup. See also Valentine, “Forensic Facial Identification,” cited above in Chapter 1, n.10, para 17.41. For a general critique based on empirical research see now Roberts, A., Davis, J., Valentine, T. and Memon, A., “Should We Be Concerned About Street Identifications?” [2014] Crim. L. Rev. 633.
[194] [1996] Crim.L.R. 584, C.A.
[195] See Bogan, op. cit., para 3.25.
[196] The acknowledgement of such an exception would not necessarily have been inconsistent with the structure of Code D as it was then formulated, the definition of “known suspect” in the 1995 edition being contained in an advisory Note for Guidance and not an imperative code provision: see ibid., para. 3.31. The actual decision in Hickin may be contrasted with that of other decisions in which the Court of Appeal, while sympathising with the predicament of the police and expressing reservations as to whether Code D could have been intended to restrict their investigative freedom, nevertheless held that they were in breach in conducting a street identification: see, eg, Vaughan (1997) The Independent 12 May (judgment 30 April); Nunes [2001] EWCA Crim 2283; 10 Archbold News 1 (see above, at p.4 for facts); Lennon (1999) 63 J.C.L. 459 (facts closely analogous to those in Nunes; breach described as “flagrant and gross”).
[197] The comparative delays in arranging live and video parades are reviewed below.
[198] Bogan, op. cit., para 3.25, chapter 1, note 34.
[199] D3.4.
[200] D3.23.
[201] See Graham [1994] Crim. L.R. 414, C.A.
[202] On the other hand, science is to some extent now deployed to support representations for the improvement and refinement of the methods of controlled selection.
[203] D2.1 (“In a case which involves disputed identification evidence a parade must be held if the suspect asks for one and it is practicable to hold one. A parade may also be held if the officer in charge of the investigation considers that it would be useful.”) The provision was reproduced unchanged in the second, 1991, edition.
[204] D2.3. The requirement was subject to certain practical exemptions contained in D2.4, D.2.7 or D2.10. The second limb of D2.1 of the first edition was perpetuated. The requirement for the suspect’s consent was omitted in the fourth edition, an omission which it has been pointed out will have the effect of making an identification procedure obligatory even though neither the police nor the suspect seeks one: see Bogan, op. cit., para. 3.56.
[205] Ibid., para. 3.35.
[206] I am grateful to Daniel Stevenson, barrister, for some assistance in the preparation of this section.
[207] Rutherford and Palmer (1994) 98 Cr.App.R. 191.
[208] Nicholson [2000] Cr.App.R. 182, C.A.; Harris [2003] EWCA Crim 174.
[209] Byron (1999) unreported, 12 February, C.A., where it was held that no parade was required although the witness expressed an ability to recognise the offender. Had the court taken heed of Rutherford and Palmer the decision would presumably have gone the other way.
[210] D3.13, the precursor paragraph of D3.12 of the final version of 2003; proposed additions shown in emphasis.
[211] Roberts, A., and Clover, S., “Managerialism and Myopia: The Government’s Draft Consultation on PACE(Code D,” [2002] Crim.L.R., 873, at p.879. In Popat [1998] 2 Cr.App.R. 208, at p.215, the second limb was taken to cater for the situation where “come the trial, a witness is not going to be relied upon as an identifying witness.”
[212] In terms which adopt those proposed at ibid.
[213] The eye-witness will therefore be expressing no ability to identify someone whom he may never have seen before and will require no prophetic qualities, as Bogan sardonically points out: op.cit., para. 3.39, n.49.
[214] McKenzie and Dunk noted the fact that although psychologists had developed the practice of referring to the person who actually committed the offence as the “culprit” and the person later arrested and put up for identification as the “suspect,” the regulatory documents made no such distinction: McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, A., Shepherd, E., and Wolchover, D. (eds.), Analysing Witness Testimony, London: Blackstone Press, 1999, pp 178-193, at p.179. See also Roberts and Clover, [2002] Crim.L.R. 873.
[215] Editorial commentary in Criminal Law Week (CLW/15/31/2) on Shanmugaragah and Liberna [2015] 2 Cr.App.R. 215 (14), C.A. In the 2011 edition of Code D the words “eye witness” replaced the word “witness” in condition (i) of D3.12 but that clearly did not justify the CLW editor’s interpretation.
[216] This corrects the author’s view to the contrary expressed in previous editions of the present work. Even if this be wrong D3.12(b) clearly imposes no prohibition of a procedure in these circumstances. It may still be held if the officer in charge of the investigation considers it would be useful: D3.13, as to which see infra, p.97, et seq.
[217] Gayle [1999] 2 Cr.App.R. 130, C.A.
[218] See also below, Chapter 4, pp.124-125.
[219] See pp.61-62, above.
[220] See Bogan, op. cit., para. 3.45.
[221] Ibid., para. 3.47.
[222] Meredith and Cowan [2001] EWCA Crim 1415; McCartney and others [2001] EWCA Crim 2283; [2003] 6 Archbold News 2; Lambert [2004] EWCA Crim 154; (2004) 68 J.C.L. 285, C.A., 13 January (commentary by Andrew Roberts); Shanmugarajah and Liberna [2015] 2 Cr.App.R. 215 (14), C.A. (trial judge held not wrong to agree with the assumption of the officer in the case that a “non comment” interview did not engage the need for an identification procedure).
[223] Commentary on ibid, CLW/15/31/2. All three complainants had thought that they would be able to identify their attackers and in the editor’s opinion “[t]his was a case in which it was blindingly obvious that an identification procedure should have been held. If the defendant had been picked out, the prosecution would have come to court armed with one or more positive identifications, plus the images from the CCTV, plus their other circumstantial evidence, instead of with a tenuous case (that resulted in an initial jury disagreement); and if there had been no identification of the defendant, he would have had the opportunity to make the legitimate point that witnesses, who had thought they would recognise their attackers, had failed to pick him out in a properly conducted identification procedure. Thus the police and the prosecution had adopted an approach that equated ‘no comment’ with admission, and had thereby deprived themselves of potentially highly probative testimony and the defendant of a safeguard for his liberty, and the Crown Court and the Court of Appeal had connived in this by denying him even the benefit of the jury being told that there had been a breach of Code D, the provisions of which existed for good reason, and that they should take this into account.” The comment is justly made but a simple riposte might be that the police ought in their discretion to have held an identification procedure under D3.13, as to which see below section D., at p.96, et seq.
[224] [2004] EWCA Crim 154. See also Emiku [2003] EWCA Crim 2237, where the complainant alleged that a man had tried to force her into his car and the defendant, whose vehicle had been identified, denied the allegation in interview but admitted that two hours earlier he had stopped at the location of the incident because of mechanical trouble and had spoken to a woman there. Held, that the judge had been wrong to rule that the admission amounted to that of some form of association with the complainant leaving to the jury only the question of what had happened. In essence the Appellant was disputing identity and there should have been a parade.
[225] A very lengthy lapse of time may mean that the witness feels no longer able to make an identification.
[226] See Bogan, op. cit., para. 3.47.
[227] Lambert (2004) 68 J.C.L. 285.
[228] [2011] Crim.L.R. 311, C.A.
[229] The editor of Criminal Law Week (CLW/11/12/04, March 28, 2011) observes that the decision to quash the conviction was surprising in view of the court’s opinion that the supporting evidence was “abundant” and having regard to the court’s robust rejection of the ground of appeal based on the judge’s failure to stop the case at the end of the prosecution case. He cautions that defence advocates should beware of reading too much into the decision, which is highly likely to be treated as depending on its own particular set of facts.
[230] In relation to the pre-2011 edition it was suggested that D3.12 should be interpreted as “the suspect disputes being the person the witness claims to have seen doing what is alleged ”: see Bogan, op. cit., para. 3.48. With at least equal succinctness it might be suggested that the same result could have been achieved by intepreting the phrase “the person” as “a specified person”.
[231] Ibid., citing McMath [1997] Crim.L.R. 586, C.A; K v DPP [2003] EWHC 351 (Admin). See also O’Leary and Lloyd-Evans 67 J.C.L. 115, C.A.; CLW/03/16/5; Muhidinz (2005) 70 J.C.L. 197, C.A. (victim of a street robbery immediately flagged down a police car and pointed out the three robbers whom the police pursued; the appellant, one of the two men arrested, claimed he was an innocent passer-by who had himself witnessed the robbery but had panicked and himself started to flee the scene; held, trial judge had been wrong to treat the issue as one of participation instead of identification; the police should have conducted a formal identification procedure and the jury should have been directed that the failure to do so deprived the appellant of the safeguard of having the street identification tested and that they should take account of this in their assessment of the case as a whole; for commentary see CLW/06/22/5, noting the silence of the judgment on whether the complainant had kept the two arrested person continually in his sight and on the timing of the appellant´s claim that he had been swept up in the events). See the difficult case of Hope, Limburn and Bleasdale [1994] Crim.L.R. 118, C.A., cogently discussed in Bogan, op. cit., para. 3.51.
[232] Slater [1995] 1 Cr.App.R. 584 (appeal against conviction for inflicting grievous bodily harm in a club where there was no suggestion of anyone else being present who was anything like as large as the appellant, who admitted being present; issue concerned the analagous identification issue of whether a Turnbull direction was appropriate; held, there was no evidential basis for the possibility of mistake ).
[233] [2001] 5 Archbold News 3, C.A.
[234] Annex A.12.
[235] For an exhaustive analysis of the theoretical basis for applying the code to resolve issues of participation see Roberts, A., “Questions of ‘Who was there?’ and ‘Who did what?’: The Application of Code D in Cases of Dispute as to Participation but not Presence” [2003] Crim.L.R. 709.
[236] Thus, alternatives had to be considered if the unusual appearance of the suspect made it impracticable to assemble sufficient people who resembled him, or where the witness was in fear, or where the suspect was refusing to co-operate: see D2.4 to 2.16, 1995 edition.
[237] [1998] 2 Cr.App.R. 208, C.A., followed in El Hannachi and others [1998] 2 Cr.App.R. 226, C.A., and Bell [1998] Crim.L.R. 879, C.A.
[238] Report, para. 5.77. The description was used by the court in Anastasiou [1998] Crim.L.R. 67, C.A.
[239] [2001] 1 A.C. 473; [2001] 1 Cr.App.R. 430. As Bogan points out (op. cit., at para. 3.54, n.71), this “extraordinary jurisprudential saga” was not concluded before, first, the Court of Appeal in Forbes [1999] 2 Cr.App.R. 501 rejected the approach of the Court of Appeal in Popat, and then another constitution of the Court of Appeal in Popat (No 2) [2000] 1 Cr.App.R. 387 held that the attempt by the Court of Appeal in Forbes to undermine the decision in Popat was “misguided and unfortunate” and ought not to be followed.
[240] At para. 20.
[241] This declaration did not stop the Law Lords nonetheless coming to the somewhat puzzling conclusion that on the facts of the case the street identification following what had been a comparatively fleeting intervening encounter after the initial incident had been both “compelling and untainted.”
[242] See Brown [1991] Crim.L.R. 368, C.A.; Wait [1998] Crim.L.R. 68, C.A.; McMath [1997] Crim.L.R. 586; O’Leary and Lloyd Evans 67 J.C.L. 115, C.A.; Nunes [2001] 10 Archbold News 1, C.A. For a decision subsequent to Forbes see Harris [2003] EWCA Crim 174.
[243] This is now accepted by the present author despite previously contending to the contrary in the course of participating in a lively debate on the topic: see Wolchover D., and Heaton-Armstrong A., “Farewell to Forbes,” [2003] 7 Archbold News 4; Bogan, P., “Forbes Alive and Well,” [2003] 8 Archbold News 5; reply by Wolchover and Heaton-Armstrong in ibid.; Robert English, letter, [2004] 1 Archbold News, 9. Prof. M. Zander QC in The Police and Criminal Evidence Act 1984, 4th ed., London: Sweet and Maxwell 2003, at p.279 predicted that under the regime after 2002 a body of case law would be likely to unfold on the circumstances covered by the phrase “no useful purpose.” It had also been observed that “the extent of the discretion conferred in the [provision] is so ill-defined as to give rise to the danger . . . that the police or prosecution will conclude that a procedure would serve no useful purpose in circumstances where there is, in fact, a real risk of erroneous identification”: Roberts and Clover, [2002] Crim.L.R., p.878.
[244] D3.2(d) has now been amended in the 2017 edition to read: “once there is sufficient information to establish, in accordance with [D]3.1(a) that the suspect is ‘known,’ e.g. after the eye-witness makes an identification, the provisions set out from [D]3.4 onwards shall apply for that and any other eye-witnesses in relation that individual”
[245] [1998] 9 Archbold News 1, C.A.
[246] Roberts and Clover, [2002] Crim.L.R., at p.878.
[247] Annex A.11 and Annex B.16.
[248] Cf the caution recommended by the Technical Working Group for Eyewitness Evidence of the U.S. Department of Justice Office of Justice Programmes, Eyewitness Evidence: A Guide for Law Enforcement, National Institute of Justice, Washington D.C., 1999, cited by Roberts and Clover, [2003] Crim.L.R., at p.887, as to which see infra.
[249] [2003] EWCA Crim 174, per Potter L.J., at para. 33 (for the relevant passage see infra., p.78). In Callie [2009] EWCA Crim 283 it was noted that as a matter of construction the phrase “no useful purpose” was strong and did not admit of a “proportionality exercise” permitting an exception to the mandatory rule to apply where only some, very limited, purpose, might be served.
[250] Valentine, T., Davis, J.P., Memon, A. and Roberts, A., “Live Showups and Their Influence on a Subsquent Video Line-up,” (2012) 26 Applied Cognitive Psychology, 1-23; Davis, J.P., Valentine, T., Memon, A. and Roberts, A.J., “Identification on the Street: A field comparison of police street identifications and video lineups in England,” (2014) 21 Psychology, Crime & Law, 9-17, DOI: 10.1080/1068316X.2014.915322 (originally unveiled at the Nuffield Foundation workshop “Eyewitness Identification Evidence”, January 30, 2009; cited in Wolchover, D. and Heaton-Armstrong, A., “Street Identification,” (2014) 178 CJ&JW, No 10, 135-137, March 8); Roberts, A., Davis, J., Valentine, T. and Memon, A. “Should we be concerned about street identifications?” [2012] CrimL.R. 633-654.
[251] Ibid., pp.651-3.
[252] At para. 21.
[253] An expressed ability to recognise the clothing worn by the person seen on the relevant occasion would plainly form no basis for holding an identification procedure: see Oscar [1991] Crim.L.R. 778; Forbes, supra, and Haynes [2004] EWCA Crim 390.
[254] Harris [2003] EWCA Crim 174, per Potter L.J., at para. 28, citing D2.15 of the 2002 transitional provisions.
[255] See, eg, Conway (1990) 91 Cr.App.R. 142, C.A., where the suspect in his interview denied any acquaintance with two witnesses who were claiming to know him and it was held that an identification parade should have been conducted to “test the crucial issue of whether or not the witnesses knew him.” In Williamson [2002] EWCA Crim 1809, the C.A. doubted whether the fact of the witness having seen a man casually on a number of occasions over the three years prior to the offence of arson but without knowing his name and where her description of him did not entirely match the appellant, could properly be characterised as rendering the man “well known” to her, so as to justify a finding that a parade would be pointless. In a case at first instance cited by Robert English, letter, [2004] 1 Archbold News, 9, prosecuting counsel attempted unsuccessfully to argue that no useful purpose would have been served by a parade because the suspect, having been identified in the street as the perpetrator, had for that reason become well-known to the witness! Where a special constable had known a person 14 years beforehand, but had only seen him once since 1990, her identification of him could not be said to be the recognition of someone “well known” to her, and thus an identification procedure should have been held under D3.12, which imposed a mandatory obligation on the police: McKenna v. DPP [2005] 5 Archbold News 2, QBD (Newman J.).
[256] Supra.
[257] Indeed Potter L.J., at para. 40, “observe[d] the irony that, whilst the judge’s ruling was given on the basis that the case was one of recognition of someone well known to the witness, he also informed the jury that the appellant was not someone whom either witness knew well.”
[258] See above Chapter 2, p.43.
[259] [2003] EWCA Crim 3869.
[260] See the Devlin report, at para. 4.69, discussing the impact of “hot pursuit” on identification, and Anastasiou [1998] Crim.L.R. 67, C.A., a classic hot pursuit case where the police saw the offence, gave chase, briefly lost sight of the fugitive and then found the defendant hiding on a rooftop. Additionally, where the suspect is arrested after the police have sealed off the area, as in Nunes [2001] EWCA Crim 2283; [2001] 10 Archbold News 1, C.A., this may reinforce still further an inference that he is the culprit who was pursued in the first instance. For the facts see above p.5. In that case it was held that a street identification by the eyewitness officer was in breach of the code.
[261] D3.1A(b). The definition dates from the 1995 revision in which it appeared as Note for guidance 2D. The definition was moved from D3.4 in the 2013 edition.
[262] Bogan, op. cit., para. 4.11.
[263] Ibid. A suspect whose arrest was imminent might well be regarded as available: Kitchen [1994] Crim.L.R. 684, C.A., cited in ibid.
[264] In D3.4.
[265] D3.17(iv). Consent may not be compelled by force or threats: Jones and Nelson, The Times 21 April, 1999, C.A.; Harley (2000) unreported, 22 February, C.A.
[266] See Bogan, op. cit., para. 4.14.
[267] The author is grateful to Mr R. Chandrapala, H.R.A., for pointing out that although placing the onus of consent on the parent or guardian seems sensible enough in theory there may a number of potential drawbacks in permitting responsibility for the decision to lie in effect on the shoulders of a lay adult in the absence of qualified legal advice.
[268] For an example of the prosecution successfully arguing at trial that for the defendant to be guilty he must have changed his appearance, see Amin [2015] Crim.L.R. 631, cited above, Chapter 1, F(1), p.21. Conviction quashed when police photographs were found showing he had not done so.
[269] D3.17(ix).
[270] In George [2003] Crim L.R. 282, C.A., the accused grew a full beard in the thirteen months which passed between his arrest and video identification procedures. For an account and commentary on the decision see Bogan, op. cit., paras. 10.51-10.56.
[271] Within the meaning of D3.17(ix).
[272] See Bogan, op. cit., para. 4.13.
[273] D3.5. See, infra, p.97 for the exemptions.
[274] See p.88, below.
[275] Although the principle may no longer be directly relevant to the regime under 5th edition it has been stated that the mere fact that insufficient volunteers are available on the day of an arranged parade does not of itself render a parade on another day impracticable: Penny (1992) 94 Cr.AppR. 345; [1992] Crim.L.R. C.A.
[276] Bogan, op. cit., para. 4.21, n.32.
[277] It will also be the only method for use in cases where the suspect is not available: see below.
[278] Cm. 5563, July 2002, para. 3.14.
[279] The figures for delay are broadly in line with those given on a variety of websites about VIPER consulted during the preparation of this paper, to which reference is made later.
[280] See eg Tinsley, Y., “Even better than the real thing? The case for reform of identification procedures” [2001] 5 Journal of Ev. and Proof, pp.99-110. For a critique of the resource-led pressure for change see Roberts, A., and Clover, S., “Managerialism and Myopia(The Government’s Consultation Draft on PACE Code D” [2002] Crim.L.R. 873).
[281] Of now limited relevance is the observation that even if an identification parade is more suitable than an alternative method significant delay caused by difficulties in assembling volunteers of similar appearance can render a parade impracticable: Jamel [1993] CrimL.R. 52, CA. As the factor of delay is now the principal test of suitability under D3.14, the potential delay caused by such difficulties will of itself exclude from the equation other aspects of suitability favouring a parade.
[282] See Bogan, op. cit., para. 4.21.
[283] Ibid. For a decision involving the use of electronic masking see Marcus (Ruel) [2004] EWCA Crim 3387; [2004] All ER (D) 351 (Nov); case number 0400198 D3, 23 November 2004. See further below p.126.
[284] Bogan, op. cit., para. 4.21. In George [2003] Crim. L.R. 282, C.A., a standard video format was employed and witnesses reported that their ability to make a positive identification was impeded by not seeing the whole body of the persons on the array who struck them as resembling the man observed on the relevant occasion. See Bogan, op. cit., para. 10.52.
[285] The equal ranking formula was removed for the 2006 edition. The Home Office schedule of changes is available on-line.
[286] See Bogan, op. cit., para 4.20, n.31.
[287] Bogan’s suggestion at ibid, paras. 4.21-4.22, that an identification parade or group identification may be more suitable than video in cases where whole body appearance, posture, gait and movement are important, with the clear implication that there will therefore be a significant albeit minority use for them in the management of evidence gathering, is hardly supported by other correct statements in his work. Thus, in the Preface he observes that “video identification is now used in all but the rarest cases. The identification parade has become a relic of the past, likely only to be seen in old movies.” At para. 4.15 he states that “for some police forces the identification parade is a creature now almost extinct.”
[288] The paragraph states that a record should be made of the reasons for for refusal and any representations made.
[289] D.3.15. If the officer decides it is not suitable and practicable to offer an alternative identification procedure, the reasons for that decision must be recorded: ibid.
[290] Gasper [2002] EWCA Crim 1764, C.A.
[291] Professor Michael Zander Q.C. in his work The Police and Criminal Evidence Act, 1984, 4th ed., 2003, para. 6-164, states, at p.279, that “[t]he decision as to which method to use is made by the officer in charge of the case after consulting with the I.O. as to which ‘is the most suitable and practicable in the particular case’.” This is incorrect and Professor Zander was citing the June 2002 consultation draft plainly on the assumption, when he was preparing the 4th edition, that the provision would be adopted for the draft to be laid. (He made the same error when describing D3.13 as, conversely, providing that an identification procedure can also be held if the I.O. considers that it would be useful. In fact, the 2003/2004 version states that it may be held if the officer in charge of the investigation considers it would be useful.)
[292] Suspects being “available” are those immediately available or those who will be within a reasonably short time and willing to take an effective part: D3.4.
[293] Hence the title of the author’s lecture “No Dodging Viper” (with apologies to the DaimlerChrysler Corporation) given at 7 Bell Yard, London WC2 on 26 November 2003.
[294] Alternatively, arrangements may be made for a group identification.These provisions may also be applied to jurveniles where the consent of their parent or guardian is either refused or reasonable efforts to obtain that consent have failed: D3.21. Note for Guidance 3D states that D3.21 would apply when a known suspect becomes “unavailable” and thereby delays or frustrates arrangements for obtaining identification evidence and when a suspect refuses of fails to take part in a video identification or other identification procedure. D3.21 also applies when a suspect refuses or fails to take part in a video identification, an identification parade or a group identification, or refuses or fails to take part in the only practicable options from that list. It enables any suitable images of the suspect, moving or still, which are available or can be obtained, to be used in an identification procedure: ibid. In the Code revision which came into force on 1 January 2006 the note was expanded to include such examples as images from custody and other CCTV systems and from visually recorded interview records. In Folan [2003] EWCA Crim 908, see below at p.98, the appearance of the defendant had changed significantly in the 20 years between his wife´s disappearance and the discovery of her remains during demolition work at a hospital and the police showed his old passport photograph to witnesses who had been involved in building works at the site during the period the wife had gone missing and who recognised the defendant as having worked at the hospital at that time. It was held that in the unusual circumstances application of the Code was inappropriate but that nevertheless the witnesses should have been shown a series of photographs. The current edition of the Code would have been applicable on the facts but for the rule that still photographs of the suspect and foils may only be shown in a video format where the suspect is known but unavailable.
[295] For consideration of the possible sources of still photographs see below, pp.94-95 and Chapter 4, p.103.
[296] Annex A.4.
[297] Under D3.20, addressed below.
[298] D3.18.
[299] Ibid. The signed copy must be reained by the identification officer: ibid.
[300] D3.17(x).
[301] D3.17(iv).
[302] D3.17(v).
[303] Ibid.
[304] D3.20.
[305] With some justice Roberts and Clover note the lack of any criteria in forming a reasonable belief that the suspect will not co-operate and they note the irony of the apparent permission to withhold the notice where the officers reasonably believe that the suspect will fail to co-operate as a consequence of receiving it, as opposed to believing that the suspect is likely to be unco-operative in any event: [2003] Crim.L.R., at p.881. Presumably, however, there would have to be some history or other extraneous indicator of non-coöperation before the notice could properly be withheld.
[306] Ibid., at pp. 881-882.
[307] D3.20.
[308] See the definition of “available” in D3.4. Cf D3.11, providing that where an identification procedure is required, in the interest of fairness to suspects and witnesses, it must be held as soon as practicable.
[309] (1992) unreported, C.A. 20 March, 90/02296/Y4.
[310] The PACE Act, s.64(4)(a), specifically permits the use of a photograph of the suspect in the investigation of an offence. An important source now available under the PROMAT system is the enormous national database of images provided by the Custody Suite Imaging System: see below Chapter 4, p.103.
[311] [2003] Crim.L.R., at p.882. Article 8(1) of the Convention states: “Everyone has the right to respect for his private and family life, his home and his correspondence.” Article 8(2) states: There shall be no interference by a public authority with the exercise of this right except such as is in accordance with the law and is necessary in a democratic society in the interests of national security, public safety or the economic well-being of the country, for the prevention of disorder or crime, for the protection of health or morals, or for the protection of the rights and freedoms of others.”
[312] (2003) E.C.H.R. No. 63737/00, 17 July.
[313] See commentary at [2003] 7 Archbold News 2.
[314] See Loveridge et al. [2001] 2 Cr.App.R. 29, C.A.
[315] D3.21.
[316] D3.23.
[317] Subject to the proviso in Annex A3(b) that any differences in the conditions would not direct a witness’s attention to any individual image: see supra, p.92 and p.120, infra.
[318] Thus, in Shanmugarajah and Liberna [2015] 2 Cr.App.R. 215(14), C.A. it ought to have been required under D3.13: see commentary on the decision in CLW/15/31/02, considered at n.25, p.65, above.
[319] See Bogan op. cit., para. 4.17
[320] As they were in earlier versions, see notes 5 and 6, p.60, supra.
[321] [2003] EWCA Crim 908.
[322] See above p.91.
[323] Bruck, M., Cavanagh, P., and Ceci, S., “Fortysomething: Recognizing faces at one’s 25th reunion,” Memory and Cognition, 1991, 19, pp.221-228, discussed in Bruce, V., and Young, A., In the Eye of the Beholder, Oxford: Oxford University Press, 1998, pp.98-99, referring also to the case of identification across the years of John Demjanjuk (see below, Chapter 4, B.(17), pp.133-134.
[324] This section is based on a digest of information gleaned from consulting several websites, not all of which is consistent.
[325] Quoted from release 072/2003 on the website .uk.
[326] Bogan suggests that not all systems in use can accommodate selection from a larger number or substitution to meet objections: op. cit., para 5.38, p.72, n.52, or sequence alteration: ibid., para. 5.39, n.54. Whether or not this is correct the Wakefield Viper centre can certainly deal with any such requests.
[327] See Valentine, “Forensic facial identification,” cited above in Chapter 1, n.10, para 17.35. See also text under (vi) below on the system’s technical specification.
[328] The account of the PROMAT system given here was compiled from material published on imagered.co.uk, the website of manufacturer Image Red Ltd, Videoid.co.uk, and .uk, and information kindly furnished by Inspector R. Tucker, formerly officer in command of the Metropolitan Police Identification Suite at Kilburn. Another system in use in some forces is “Parade Manager”: see Bogan, op. cit., para. 5.36, p.71, n.47.
[329] Valentine, “Forensic facial identification,” cited above in Chapter 1, n.10, para 17.35.
[330] See “Mobile ID cuts cost of parades,” The Job, vol. 36, issue 899, March 7, 2003, reporting on the five Metropolitan Mobile Identification Units led by Chief Inspector Jim Davis, believed then to be the only fully mobile identification process in the country. Cf. Bogan, op. cit., at para. 4.15, n.26, who acknowledged C.I. Davies as one of his sources of information, refers to the existence of a single Metropolitan Police mobile unit which “in an urgent case can, in the same day, attend a police station in which an arrested person has been detained, set up its equipment and video the suspect, complete a line up from its database and confirm its content with the suspect or solicitor and then attend the witness’s home or a hospital to conduct the procedure.”
[331] See proposal discussed below at B.(27), pp.147-148.
[332] I am grateful to Mr. Peter Burton, Head of Imaging Unit with the West Yorkshire Police Scientific Support Department, who in a personal communication helpfully furnished information about regional differences in practice.
[333] See infra, p.131 for the additional provisions.
[334] See Tinsley, Y., “Even better than the real thing? The case for reform of identification procedures” [2001] 5 Journal of Ev. and Proof, pp.99, at n.26.
[335] See Roberts and Clover, [2002] Crim.L.R., at p.884.
[336] See ibid.
[337] For a review of the relevant research see Cutler, B. and Penrod, S., Mistaken Identification: The Eyewitness, Psychology and the Law, Cambridge: Cambridge University Press 1995, pp.105-106.
[338] See Roberts and Clover, [2002] Crim.L.R., at p.884. See also George [2002] The Times, August 30, C.A. (sanctioning examination of the degree to which a qualified identification could be properly attributed by the Crown to delay).
[339] Cm. 5563, July 2002, para. 3.14. The 50 per cent cancellation figure was first identified in a survey by Slater, A., “Saving money on ID parades,” (1995) 11 Policing, pp.203-231. In an analysis of 1083 parades in West Yorkshire between 1997 and 1999, Pike and others found that only 5.2 per cent of the 403 VIPER parades were cancelled compared to 46.4 per cent of the 680 live parades: Pike, G.E., Kemp, R., Brace, N., Allen, J. and Rowlands, G., “The effectiveness of video identification parades,” (2000) 8 Proceedings of the British Psychological Society, 8(1), p.44.
[340] The requirement goes back at least as far as the Consolidated Home Office Circular to police issued on January 1, 1925, the extract from which on conducting identification parades is set out at Appendix 6 to the Report of the Royal Commission on Police Powers and Procedure, Cmd. 3297, 1929.
[341] In Slater’s 1995 survey 50 per cent of witnesses thought that less than half of the parade participants resembled each other: “Saving money on ID parades,” (1995) 11 Policing, pp.203-231.
[342] The Home Office Consolidated Circular of January 1, 1925, supra, cautioned against “the witnesses’ attention being directed to the suspected person in particular instead of indifferently to all the persons paraded.” Curiously, no such enjoinder is contained in the 2003 code in relation to live parades. Perhaps it is so obvious that it never occurred to the draftsman to include one. In the Report of the Royal Commission on Police Powers and Procedure, para. 129, it was acknowledged that the provision in the rules which allowed the accused to change position after each witness, though intended to prevent unfairness, actually made it impossible to conceal his identity from the others on the parade. (The suggested antidote was to require all on the parade to change places or to require them to walk about, the latter proposal of which had the added advantage of affording the witnesses an opportunity to observe the gait of the suspect and volunteers.)
[343] See Report of the Tribunal of Inquiry on the Arrest of Major R. O. Sheppard, D.S.O., R.A.O.C. (ch. The Right Hon. J.F.P. Rawlinson, K.C.) Cmd. 2497, 1925. For a lively account of the Sheppard case see C.H. Rolph, op. cit., ch. 3.
[344] The behaviour of the police and volunteers should have been prevented by the Home Office Consolidated Circular of January 1, 1925, referred to above.
[345] The technique is known as “pixellation.” Bogan reports that its development and use is not presently widespread: op. cit., para 5.36, p.71. It is understood that the VIPER centre at Wakefield has full facilities, and the service is available nationwide, even if not all local centres are suitably equipped. PROMAT allows sophisticated electronic adaptation of imagery, including changes in appearance of the background in order to assimilate images of the suspect with that of the foils. For Annex A provisions on electronic alteration of images see below, pp.125-126.
[346] See Annex A.9 and see generally section C.(4), pp.112 and 150, et seq, below. The issue of witness stress elimination in the context of removing the suspect’s absolute right to have his or her solicitor present at the identification has been raised in the Consultation Paper Modernising Police Powers: Review of the Police and Criminal Evidence Act (PACE) 1984, Home Office, March 2007, paras. 3.39 to 3.41, considered below at pp.154-155.
[347] Roberts and Clover, [2002] Crim.L.R., at p.885.
[348] See Kemp, R., Pike, G., and Brace, N., “Video-based identification procedures: Combining best practice and practical requirements when designing identification systems,” Psychology, Public Policy, and Law, December 2001, for a discussion on the contribution which psychologists might make to the design of investigative procedures.
[349] Dent, H., and Gray, F., “Identification on Parade” (1975) New Behaviour 366-369, cited in the Devlin report, para. 5.56.
[350] The Law Commission of New Zealand, Total Recall? The Reliability of Witness Testimony Evidence, Miscellaneous Paper 13, Wellington 1999.
[351] See, e.g., Cutler B.L., et al., “Eyewitness Identification Test Media,” in Ross, D.F., Read, J.D., and Toglia, M.P. (eds.) Adult Eyewitness Testimony:Current Trends and Developments, Cambridge: Cambridge Univ. Press, 1994.
[352] Valentine T. and Heaton, P., “An evaluation of the fairness of police lineups and video identifications,” 13 Applied Cognitive Psychology (1999) 13 (special issue), S59-S72. See also Lindsay, R., and Wells, G., “Improving Eyewitness Identifications From Lineups: Simultaneous Versus Sequential Lineup Presentation,” (1985) 70 Journal of Applied Psychology 556, and see Online Conference Report on the British Association Annual Festival of Science, September 2001, reported on the New Scientist web site.
[353] Valentine, T., Harris, N., Colom Piera, A. and Darling, S., “Are police video identifications fair to African-Caribbean suspects? (2003) 17 Applied Cognitive Pxychology 459-476.
[354] Wright, D.B. and McDaid, A.T., “Comparing system and estimator variables using data from real lineups” (1996) 10 Applied Cognitive Psychology 75-84.
[355] D3.6.
[356] D3.11.
[357] Recommendation by Pike, G., Brace, N. and Kyan, S, The Visual Identification of Suspects: Procedures and Practice, Home Office Briefing Note, 2/02, March 2002, p.6.
[358] D2.21. Such staff must have regard to any relevant provisions of the PACE Codes of Practice: D2.22.
[359] D3.11.
[360] D2.21. The 2017 edition of the code further adds to D3.11 the clarification that where any action referred to in the paragraph is taken by another officer or police staff at the direction of the identification officer, the outcomes must, as soon as practicable, be reported to the identification officer who, for the purposes of the identification procedures in D3.5 to 3.10, retains overall responsibility for ensuring that the procedure complies with Code D and in addition, in the case of detained suspects, their care and treatment until returned to the custody officer.
[361] D3.19. The information to be given is that set out in D3.17. The written notice is required by D3.18.
[362] D3.19.
[363] Note 3C.
[364] See below B.(27), pp.147-148. See Levi, A., and Lindsay, R., “Issues concerning policy recommendations: The example of lineups and photospreads,” Psychology, Public Policy, and Law, December 2001; Wells, G.L., Small, M., Penrod, S., Malpass, R., Fulero, S. and Brinacombe, C., “Eyewitness identification procedures: Recommendations for lineups and photospreads,” (1998) 22 Law and Human Behaviour, 603-647. It has been noted that this might be rather difficult to achieve with live parades in practice but slightly more manageable with video identification. Kemp, R., Pike, G., and Brace, N., “Video-based identification procedures: Combining best practice and practical requirements when designing identification systems,” Psychology, Public Policy, and Law, December 2001.
[365] Also called Modernising Police Powers: Review of the Police and Criminal Evidence Act (PACE) 1984, issued September 2006 (hereafter “NIF paper”).
[366] I am grateful to Police Inspector Martin Griffiths, of West Yorkshire Police, whose insights have inspired a number of the prognostications which follow in the text. These are included in a submission to the Home Office by the present author and Anthony Heaton-Armstrong, formulated in response to the Consultation Paper. The submission was subsequently incorporated into articles: see Wolchover, D. and Heaton-Armstrong, A., “PACE Code D Review on video ID parades: The innocuous cloaking the insidious,” [2007] 6 Archbold News 7; “VIPER disappointments in the PACE review,” [2008] 10 Archbold News 4; “Viper and the Vandalizing of PACE,” (2009) 173 Criminal Law and Justice Weekly 5, January 3 and 10.
[367] Para. 7.5.
[368] Para 7.6.
[369] D3.14. See above, p.46, et seq.
[370] Para 7.6.
[371] Para 7.7.
[372] D3.11.
[373] See e.g., Gall [1989] Crim.L.R. 745, C.A. (investigating officer permissibly escorted the witness to a parade but then entered the parade room and had the opportunity of speaking to the witnesses before the parade took place; held, although there was nothing to suggest anything untoward had occurred the evidence ought to have been excluded); Quinn [1995] 1 Cr.App.R. 480, C.A., (investigating officer wrong to have conducted a demonstration of an identification parade to witnesses before the actual one was held); Ryan [1992] Crim.L.R. 187, C.A. (“substantial breach” of the code where investigating officers attended a robbery victim at the police station where a confrontation was held). Cf. Walters [2001] EWCA Crim 1261 (driving the suspect to the police station for a parade held not part of the arrangements).
[374] D3.11. Such consultation is required by D3.14, discussed supra in C.(2)(iii), at p.85-88.
[375] Annex A.15.
[376] D3.17(i).
[377] D3.17(ii).
[378] D3.17(iii), subject to Annex A.9 (see further infra, pp.150-156).
[379] D3.17(vii).
[380] D3.17(viii).
[381] D3.18.
[382] Ibid.
[383] D3.1, D3.17(xiii), and Annex A.8.
[384] See Bogan, op. cit., at para. 5.32, p.70.
[385] Ibid., para. 5.37.
[386] D3.3 further provides that if the suspect’s identity is not known, the showing of any such images to an eye-witness to see if they can identify a person whose image they are shown as the person they saw on a previous occasion must be done in accordance with Annex E to the code.
[387] Annex E.9. The fact must also be included in the Notice to Suspect: D3.17(xi). Annex e.9 does not apply where an image was created and based on the witness’s own description: Wright [1994] Crim.L.R. 131, C.A.
[388] Annex E.9 and see Note for Guidance 3B.
[389] Noted by Bogan, op.cit., at para. 5.33. Contrast may be drawn with D3.29 and Annex E.9, infra, which expressly grant suspects or their solicitors the right to view media material before a video identification procedure.
[390] See above, section (6), at p.117, for the arguments for and against requesting the distinctive feature to be added electronically to the foils where it had not been mentioned by the witness in the first or any subsequent description.
[391] Supra, pp.42-43.
[392] D.3.39 and Annex A.8. Although the wording is disjunctive both the suspect and their solicitor are presumably permitted to view the material. C3..39 further provides that the paragraph does not affect any separate requirement under the Criminal Procedure and Investigations Act 1996 to retain material in connection with criminal investigations that might apply to both limbs of C3.38. Such a rationale would be to overlook the potency of unconscious memory. It is noteworthy that the paragraph is silent about the retention of material shown to police officers under D3.38.
[393] D3.39 and Annex A14. D3.39 states that it does not affect any separate requirement under the Criminal Procedure and Investigations Act 1996 to retain material in connection with criminal investigations.
[394] Annex A.3. See above Chapter 3, C.(3)(i)(c), p.91.
[395] Annex A.3. The reasons why identical conditions are not practicable must be recorded on forms provided for the purpose Annex A.4.
[396] For the evidential use of video monitoring recordings of the police station precincts, see Wolchover, D., and Heaton-Armstrong, A., Confession Evidence, London: Sweet and Maxwell, 1996, pp.460-463.
[397] As in the cases of Kennedy (1992) unreported, C.A. 20 March, 90/02296/Y4, Perry (2000) unreported, 3 April, C.A., and Perry v. United Kingdom (2003) E.C.H.R. No. 63737/00, 17 July; [2003] Crim.L.R. 281.
[398] D3.21.
[399] See above Chapter 3, C.(3), pp.88-96.
[400] See Bogan, op. cit., para. 5.50.
[401] See this chapter, A.(4)(v) supra, at p.105-106.
[402] Annex A.2..
[403] The change was proposed by the CPS: see schedule of proposals and HO responses.
[404] Originally argued by Luus, C.A.E. and Wells, G.L., “Eyewitness identification and the selection of distractors for lineups” (1991) 15 Law and Human Behaviour 43-57. See also Wells, G.L., Small, M., Penrod, S., Malpass, R., Fulero, S. and Brimacombe, C., “Eyewitness identification procedures: Recommendations for lineups and photospreads,” (1998) 22 Law and Human Behaviour 603-647. See further below p.123.
[405] “Forensic facial identification,” cited above in Chapter 1, at n.10, para 17.25, original emphasis. Valentine cautions against placing foils of different sex or ethnic background on the same array where such “default values” were not mentioned in a description, but he advises that the problem of missing “descriptors” may be eliminated by police officers taking a careful and comprehensive description: ibid, at para 17.26, citing a study by Lindsay, R.C.L., Martin, R. and Webber, L., “Default values in eyewitness descriptions: A problem for the match-to-description lineup foil selection strategy” (1994) 18 Law and Human Behaviour 527-541).
[406] McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, Shepherd and Wolchover (eds.), Analysing Witness Testimony, London: Blackstone Press 1999, pp.185-186.
[407] “Forensic facial identification,” para 17.27, citing Wells, G.L., Small, M., Penrod, S., Malpass, R., Fulero, S. and Brinacombe, C., “Eyewitness identification procedures: Recommendations for lineups and photospreads,” (1998) 22 Law and Human Behaviour, 603-647 (cited above, this chapter, n.41). Valentine cites the report of an experimental study supporting the argument that match to description selection produces more reliable identification evidence: Wells, G.L., Rydell, S.M. and Seelau, E., “The selection of distractors for eyewitness lineups” (1993) 78 Journal of Appplied Psychology 835-844. More correct identifications were made from culprit-description arrays (67 pc) than from suspect-resemblance arrays (22 pc) when students witnessed a live staged theft and were immediately asked to identify the perpetrator from an array of photographs. However, when the culprit was absent from the array the difference between mistaken identification of foils from culprit description arrays (32 pc) and those from suspect-resemblance arrays (47 pc) was not regarded as statistically significant. Moreover, in two other studies no statistically significant difference was found in the rate of correct or mistaken identifications between culprit-description and suspect-resemblance arrays: Lindsay, Martin and Webber, cited above, and Tunnicliff, J.L. and Clark, S.E., “Selecting foils for identification lineups: Matching suspects or descriptions?” (2000) 24 Law and Human Behaviour 231-258.
[408] See Valentine, “Forensic facial identification,” para 17.24.
[409] In any event, obvious differences may well warrant a judgment under D3.12 that there being no reasonable chance of an identification of the suspect no identification procedure is required.
[410] McKenzie, I., “Psychology and legal practice, fairness and accuracy in identification parades” [1995] Crim.L.R. 200; see also McKenzie and Dunk, supra, p.187.
[411] Bogan, op. cit., para. 5.37.
[412] Marcus (Ruel) [2004] EWCA Crim 3387; [2004] All ER (D) 351 (Nov); case number 0400198 D3, 23 November 2004. However, because of local concerns among the police that masking was impeding the ability of witnesses to make identifications it was decided at the initiative of the CPS that an alternative compilation using the same foils should be made, but without masking, and that the unmasked array should be used if the witnesses were unable to make an identification from the masked array. The exercise was a novelty and apparently undertaken as a test case. Neither the suspect nor his solicitor were informed of the intention to resort to the procedure until immediately before the video identification procedure took place, and the solicitor´s objection to the unfairness of the procedure was overruled by the identification officer, in spite of the lack of proper notice. In cross-examination a police inspector (not the identification officer who had presided over the identification) conceded that it was “blatantly unfair” as it would mean that the Appellant would “blatantly stand out” from the other images. However, the trial judge allowed in the evidence because, as he said, whatever its flaws it at least had the virtue of transparency. The Court of Appeal disagreed and quashed the relevant counts of robbery.
[413] See Bogan op.cit., para. 1.32, drawing attention to the risk of latent unexpressed awareness of a particular facial feature.
[414] See McKenzie and Dunk, supra, p.182.
[415] Ellis, H.D., “Practical aspects of face memory,” in Wells, G.L. and Loftus, E. (eds.), Eyewitness Testimony(Psychological Perspectives, Cambridge University Press, 1984, pp.12-37. In one case described by McKenzie and Dunk, supra, at p. 183, a witness observed a robber under reasonably good conditions and gave a very detailed description of scarring to the culprits’s face. However, apart from the scars the only description of the face was that it was “chubby.”
[416] Supra, at p.183.
[417] Annex B.9.
[418] Bogan, op. cit., para 5.41, p.73. Doubling up pales into sweet insignificance compared with the practice in some parts of the United States referred to above in Chapter 2 at p. 54 where the whole line-up may consist of suspects.
[419] Bogan, op. cit., para. 5.69, p.82.
[420] Annex B.11.
[421] Annex B.9.
[422] Annex A.10.
[423] Annex A.5. See below Chapter 5 for the admissibility of words and acts signifying an identification.
[424] Annex A.6.
[425] Annex B.4.
[426] As in Kennedy (1992) unreported, C.A. 20 March, 90/02296/Y4.
[427] Annex B.5.
[428] Annex A.6.
[429] Footage taken on a portable machine, e.g. when the suspect is in prison, may appear different from the other images and the solicitor should be looking for consistency in image quality, colour, lighting, camera position, background and duration: see Corre, N., and Davies, J., “Identification on the Move,” London Criminal Courts Solicitors Association lecture, 21 January, 2003 (lecture notes obtainable from Neil Corre, e-mail, corre@corre.demon.co.uk).
[430] In Quinn [1995] 1 Cr.App.R. 480 it was held that there had been no code breach where the suspect and his solicitor had played an active role in assembling an identification parade and voiced no complaint about its composition until the trial. It has been observed that the suspect may be caught in a dilemma between, on the one hand, refusing to co-operate with what is perceived to be an unacceptable array and suffering in consequence the resort to a less desirable procedure, and, on the other, facing the risk that even “a grudging” acquiescence may be taken as approval: Bogan, op.cit., para 5.39, citing Mendili [2001] EWCA Crim 757. This may be a valid concern where a traditional live parade is in issue but the commentator here was referring to video identification, the suspect’s withdrawal of co-operation from which in protest could not conceivably justify the police resorting to one of the less desirable statutory alternatives.
[431] See Peter Burton (Head of the Imaging Unit, West Yorkshire Police Scientific Support Department) “VIPER(the future of identification parades?” Public Service Review, Information Technology section, Home Office, Spring 2002, p.34. For an update see “The Viper Revolution,” ibid., Autumn 2003.
[432] Wagenaar, W.A., Identifying Ivan: A case study in legal psychology, Hemel-Hempstead: Harvester-Wheatsheaf, 1988, discussed in Bruce, V., and Young, A., In the Eye of the Beholder: the science of face perception, Oxford: Oxford University Press, 1998, at pp.100-102. Demjanjuk’s conviction was overturned on appeal and he was allowed to return to the U.S. from where he had been extradited. Comparable research was conducted by a distinguished forensic psychologist commissioned by the defence for the purposes of the subsequently abandoned appeal in the case of Abdel Baset Ali al-Megrahi, the man convicted of planting a bomb on Pan American flight 103, which exploded over Lockerbie in 1988: see David Canter, “Lockerbie: was this man’s memory reliable?” The Times, October 26, 2009, and Canter, D., Youngs, D., and Hammond, L., Relating to the Appeal of Abdelbaset Ali Mohmed Al Megrahi: With particular attention to the evidence of Mr Anthony Gauci, Huddersfield: International Research Centre for Investigative Psychology, The University of Huddersfield, 2009 (available at ).
[433] In Elson (1994) unreported, C.A. 94/0547/Y3, 27 May, two witnesses who attended an identification had sat together in the waiting room and communicated with each other before the parade, in breach of the code. See also Finley [1993] Crim.L.R. 50, C.A.
[434] Annex A9.
[435] Annex A.9. See further, infra.
[436] There is no explicit provision in Annex A granting the entitelement to representation but A.7, A.8 and A.9 assume a solicitor’s involvement: see Bogan, op, cit , para. 5.31, n.33. A.9 sets out the circumstances in which the solicitor may be allowed to observe the process of showing images to the eye-witness: see infra.
[437] Annex A.9. This is in contrast with the defending solicitor’s entitlement to notice. Presumably this is a precaution against the suspect or a confederate waiting nearby for the eye-witness to arrive at the venue in order to convey some sort of threat.
[438] Annex A.13.
[439] This would seem to be confirmed by reference to the contrasting provision in Annex A.14, below.
[440] For C3.38(a) see supra, at pp.42-43). It has been pointed out that under the 1995 edition the responsibility for asking this question was curiously the responsibility of the investigating officer: see McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, A., Shepherd, E., and Wolchover, D. (eds.), Analysing Witness Testimony, London: Blackstone Press, 1999, pp 178-193, at p.184.
[441] See eg ibid, p.187.
[442] Report, paras. 5.63 and 8.16.
[443] (1977) 63 Cr.App.R. 132.
[444] Annex A.11.
[445] Ibid., citing Wells, G., “How Adequate is Human Intuition for Judging Eyewitness Testimony?” in G. Wells and E. Loftus (eds.) Eyewitness Testimony: Psychological Perspectives, Cambridge Univ. Press 1984.
[446] Report, para. 129, referring to evidence in response to Q.6322, et seq. The Commissioners judged there were “certain obvious objections, particularly the difficulties already experienced in persuading suitable persons to attend a parade.” C.H. Rolph commented that this in fact seemed the only objection that could be called “obvious”: op. cit., p.43. See also Law Reform Commission of Canada, Pretrial Eyewitness Identification Procedures, Criminal Law Series Paper, Law Reform Commission of Canada, Ottawa, 1983, at pp.150-153.
[447] Report, para. 5.57.
[448] Roberts and Clover, [2002] Crim.L.R., p.886.
[449] Devlin report, para. 5.57.
[450] See Lindsay, D. S., Read, J. D. and Sharma, K., “Accuracy and confidence in person identification: The relationship is strong when witnessing conditions vary widely” (1998) 9 Psychological Science 215-218, and Sporer, S., Penrod, S., Read, D. and Cutler, B. L., “Choosing, confidence and accuracy: A meta-analysis of the confidence-accuracy relations in eyewitness identification studies” (1995) 118 Psychological Bulletin 315-327, both cited in Valentine, T., “Forensic Facial Identification,” in Witness Testimony: Psychological, Investigative and Evidential Perspectives, Oxford: OUP forthcoming, para 17.53.
[451] For a review of pre-identification confidence and accuracy see Cutler, B. L. & Penrod, S. D., “Forensically-relevant moderators of the relationship between eyewitness identification accuracy and confidence” (1989) 74 Journal of Applied Psychology 650-652; for a review of post-identification confidences and accuracy see Bothwell, R. K., Deffenbacher, K. A., and Brigham, J.C., “Correlations of eyewitness accuracy and confidence: Optimality hypothesis revisited” (1987) 72 Journal of Applied Psychology 691-695, all cited in Valentine “Forensic facial identification,” cited above Chapter 1, n.10, para 17.54.
[452] Ibid.
[453] Annex A.11 of the 2013 edition, emphasis supplied.
[454] See, eg, Bogan, op. cit., para 5.56. The topic was originally discussed in depth in the Devlin report, paras 5.58 to 5.62.
[455] Bogan, para 5.56, noting that the procedure for group identification and showing photographs make provision for a selection by a witness who is unable to confirm the identification, when the witness is then asked how sure he or she is (Annexes C.23 and E.9, respectively).
[456] Bogan, para 5.57.
[457] Proposed by West Yorkshire Police and the Crown Prosecution Service: see schedule of recommendations and responses posted on the relevant Home Office webpage. The gradation scale idea is an adaptation of the scheme employed by scientific experts and may have been inspired by Bogan’s treatment.
[458] Report, paras. 5.58-5.62.
[459] Proposed by Karl Burn, Identification Officer.
[460] Bogan, op. cit., para. 5.58
[461] Roberts and Clover, [2003] Crim.L.R., at p.887, citing a recommendation by the Technical Working Group for Eyewitness Evidence of the U.S. Department of Justice Office of Justice Programmes, Eyewitness Evidence: A Guide for Law Enforcement, National Institute of Justice, Washington D.C., 1999.
[462] See 2016_CodeD-Draft__tracked_22-03-16_
[463] In Sloan (Brandon), unreported, Inner London Crown Court, T2019 1017, ruling October 25, 2021, per HHJ Newberry, an eye-witness had failed to identify the suspect in a video identification but subsequently believed that she could, after, all identify him as present on the array and had contacted the officer in the case to say so. It was unsuccessfully submitted that the opportunity she had enjoyed to speak to “other people,” that is to say, the officer, precluded the proper exercise of the discretion to allow her a second attempt. It was held that the the phrase was restricted to other potential eye-witnesses rather than to investigating officers. Otherwise, the mere approach by an eyewitness to her police contact in order to seek another go would implicitly negate the discretion to permit it.
[464] Annex A.11.
[465] Annex A.12.
[466] Annex A.11
[467] McKenzie and Dunk, cited above Chapter 1, n.6, at pp.187-188.
[468] See eg Stern, L.B., and Dunning, D., “Distinguishing accurate from inaccurate eyewitness identifications: a reality monitoring approach,” in Ross et al (eds.), Adult Eyewitness Testimony, cited above Chapter 1, at n.65, pp.144-160; McKenzie and Dunk, cited above Chapter 1, n.6, pp.187-188; Deffenbacher, K.A., Forensic Facial Memory: Time is of the Essence, New York: Elsevier Science Publishers, 1989. For a general study on the science of face perception with a comprehensive bibliography see Bruce, V. and Young, A., In the Eye of the Beholder, Oxford University Press, 1998, and see also e.g. Laughery, K., Alexander, J. and Lane, A. “Recognition of human faces: Effects of target exposure, target position, pose position and type of photograph,” (1971) 55 Journal of Applied Psychology 1159-1164; Bruce, V., “Changing faces: Visual and non-visual coding processes in facial recognition,” (1982) 12 British Journal of Psychology 284-309; Valentine, T. and Bruce V., “The effects of distinctiveness in recognising and classifying faces,” (1986) 15 Perception 525-535; Valentine, T., “A unified account of the effects of distinctiveness, inversion and race in face recognition,” (1991) 43A Quarterly Journal of Experimental Psychology, 161-204; Shepherd, J.W., Gibling, E. and Ellis, H.D. ‘The effects of distinctiveness, presentation time and delay on face recognition,” (1991) 3 European Journal of Cognitive Psychology 137-145.
[469] See in particular Valentine, T., Pickering, A. and Darling, S., “Characteristics of eyewitness identification that predict the outcome of real lineups” (2003) 17 Applied Cognitive Psychology 969-993.
[470] Galton, F., Inquiries into human faculty and its development, London: Macmillan, 1883, p.3, cited by Bruce and Young, supra, this chapter, at n.145, p.142. For experimental research strongly suggesting that facial perception is based on the whole rather than on individual unconnected features see Young, A.W., Hellawell, D.L. and Hay, D.C., “Configural information in face perception,” (1987) 16 Perception, 747-759, and Tanaka, J.W. and Farah, M.J., “Parts and and wholes in face recognition,” (1993) 46A Quarterly Journal of Experimental Psychology 225-245, both summarised by Bruce and Young, cited above at n.139, p.145.
[471] See Lindsay, R. and Wells, G., “Improving Eyewitness Identifications From Lineups: Simultaneous Versus Sequential Lineup Presentation,” (1985) 70 Journal of Applied Psychology 556; Shapiro, P.N., and Penrod, S., “Meta-analysis of facial identification studies,” (1986) 100 Psychological Bulletin, 139-156; Cutler, B., and Penrod, S., “Improving the Reliability of Eyewitness Identification: Lineup Construction and Presentation,” (1988) 73 Journal of Applied Psychology 281; Slater, A., Identification parades: A scientific evaluation, London: Home Office, 1994; Cutler, B. and Penrod, S., Mistaken Identification: The Eyewitness, Psychology and the Law, Cambridge: Cambridge Univ. Press, 1995; Pike, G., Kemp, R., Brace, N., Allen, J. and Rowlands, G., “The effectiveness of video identification parades’ (2000) 8(1) Proceedings of the British Psychological Society, 44; Online Conference Report on the British Association Annual Festival of Science, September 2001, reported on the New Scientist web site; Stebley, N., et al., “Eyewitness Accuracy Rates in Sequential and Simultaneous Lineup Presentations: A Meta-Analytical Comparison” (2001) 25 Law and Human Behaviour 459. Pike, G., Brace, N. and Kynan, S., The Visual Identification of Suspects: Procedures and Practice, Briefing Note 2/02, London: Home Office, 2002; Roberts and Clover, cited above in Chapter 2, n.75.
[472] See Kneller, W., Memon, A. And Stevenage, S., “Simultaneous and sequential lineups: decision processes of accurate and inaccurate witnesses,” (2001) 15 Applied Cognitive Psychology 659-671; Gronland, S.D., “Sequential lineups: shift in criterion or decision strategy?” (2004) 89 Journal of Applied Psychology 362-368, both cited by Valentine, “Forensic Facial Identification,” cited above in Chapter 1, n.10, chap. 17.
[473] Rolph, C.H., Personal Identity, London: Michael Joseph, 1957, p.42. As early as 1929 it was recognised by the Royal Commission on Police Powers and Procedure that some witnesses might tend “unconsciously to identify the person who most resembles their recollection of the culprit, disregarding apparently the alternative that he may not be present at all”: Report, Cmd. 3297, para.124.
[474] Valentine, “Forensic Facial Identification,” (above, Chapter 1, n.10) cites a study of 23 papers comparing sequential and simultaneous presentation of photograph lineups: Stebley, N., Dysart, J., Fulero, S. and Lindsay, R.C.L., “Eyewitness Accuracy Rates in Sequential and Simultaneous Lineup Presentations: A Meta-Analytical Comparison” (2001) 25 Law and Human Behaviour 459. In the aggregate, more witnesses identified the culprit from simultaneously presented arrays than from sequential ones (50 to 35 pc) but fewer made incorrect rejections of culprit-present simultaenous ones than of sequential ones (26 to 46 pc). Culprit-absent arrays: more witnesses made correct rejections of sequential than of simultaneous ones (79 to 49 pc) but fewer made incorrect identifications of a foil in sequential than in simultaneous arrays (28 to 51 pc). The overall effect of sequential presentation was to provide more protection against mistaken identification from culprit-absent arrays but at a cost to the sensitivity of culprit-present procedure, a finding supported by subsequent studies cited by Valentine: Memon, A. and Bartlett, J.C., “The effects of verbalisation on face recognition” (2002) 16 Applied Cognitive Psychology 635-650; Memon, A. and Gabbert, F., “Unravelling the effects of a sequential lineup” (2003) 6 Applied Cognitive Psychology 703-714; Memon, A. and Gabbert, F. “Improving the identification accuracy of senior witnesses: Do pre-lineup questions and sequential testing help?” (2003) 88 Journal of Applied Psychology 341-347. See also Valentine T. and Heaton, P., “An evaluation of the fairness of police lineups and video identifications” (1999) 13 Applied Cognitive Psychology S59-S72 (mock witnesses who had read the witness’s description of the suspect misidentified the suspect significantly less frequently in a video array than in a simulated live parade when the true culprit was not present; on average they picked out the suspect from a picture of a live parade 25 per cent of the time, a figure more than double the 11 per cent expected by chance, whereas in a video line-up the suspect was chosen in only 15 per cent of the sample); Lindsay, R. and Wells, G., “Improving Eyewitness Identifications From Lineups: Simultaneous Versus Sequential Lineup Presentation,” (1985) 70 Journal of Applied Psychology 556, and see Online Conference Report on the British Association Annual Festival of Science, September 2001, reported on the New Scientist web site. For comparable accuracy results in culprit-present arrays see, e.g., Cutler B.L., et al., “Eyewitness Identification Test Media,” in Ross, D.F., Read, J.D. and Toglia, M.P. (eds.) Adult Eyewitness Testimony: Current Trends and Developments, Cambridge: Cambridge Univ. Press, 1994.
[475] A review of research findings led the Law Commission of New Zealand to conclude that procedures using live subjects produced more reliable results than those using video recordings: Total Recall? The Reliability of Witness Testimony Evidence, Miscellaneous Paper 13, Wellington: LCNZ, 1999.
[476] See Roberts and Clover (cited in Chapter 2, n.75) at p.886. See also Lindsay, R., Lee, J., and Fulford J., “Sequential Lineup Presentation: Technique Matters” (1991) 76 Journal of Applied Psychology 741.
[477] Valentine, T., Darling, S. and Memon, A. “Do sequential viewing instructions and moving images increase the reliability of police video identification procedures” (in preparation at time of citation by Valentine, “Forensic Facial Identification,” above, Chapter 1, n.10, para. 16-40).
[478] Pike, G., Rowlands, G., Towell, N. and Kemp, R. “Video identification and simultaneous vs. sequential lineups,” paper presented at the first joint conference in Dublin in 1999 of the American Psychology-Law Society and the European Association of Psychology and Law.
[479] Lindsay, et al (1991) cited above at n.147, pp.741-745.
[480] Kemp, R., Pike, G., and Brace, N., “Video-based identification procedures: Combining best practice and practical requirements when designing identification systems,” Psychology, Public Policy, and Law, December 2001. See also Thomson, D.M., “Eyewitness testimony and identification tests,” in Brewer, N., and Wilson, C. (eds.) Psychology and Policing, Hillsdale, New Jersey: Lawrence Erlbaum Associates, 1995, pp. 119-154.
[481] Roberts and Clover (cited in Chapter 2, n.76) at p.886.
[482] Ibid.
[483] Originally described by Loftus, E., Eyewitness Testimony, Harvard, Conn: Harvard University Press, 1979, chap. 1, instancing in particular the notorious Sacco and Vanzetti case in the United States in 1920.
[484] Wells, G.L. and Bradfield, A.L., “ ‘Good you identified the suspect’: feedback to eyewitnesses distort their reports of the witnessing experience” (1998) 66 Journal of Applied Psychology, 688-696, cited para 17.50 above.
[485] Semmler, C., Brewer, N. And Wells, G.L., “Effects of postidentificaiton feedback on eyewitness identification and nonidentification confidence” (2004) Journal of Applied Psychology, 334-346, cited ibid.
[486] Bogan, op. cit., para 5.47.
[487] R v Willoughby [1999] 2 Cr. App.R. 82.
[488] See Richardson, P.J., commentary on ibid in Criminal Law Week, CLW/99/25/04.
[489] See McKenzie, I., and Dunk, P., “Identification parades: psychological and practical realities,” in Heaton-Armstrong, A., Shepherd, E., and Wolchover, D. (eds.), Analysing Witness Testimony, London: Blackstone Press, 1999, 178-193, at p.190.
[490] See Valentine, “Forensic facial identification,”, cited above in Chapter 1, n.10, para 17.79.
[491] See pp.110-115, above.
[492] D3.24 and Annex A.18. A specimen form is set out in Bogan, op. cit., Appendix C.
[493] Annex A.17.
[494] Annex A.18.
[495] Annex A.4.
[496] Annex A.7.
[497] Annex A.15.
[498] Ibid.
[499] Annex A.16, and see D3.30 to D3.33. The detail is beyond the scope of this monograph.
[500] Report, para. 5.47.
[501] ibid.
[502] Report, Cm. 2263, July 1993, Recommendation No. 6, para. 10.
[503] Annex B.23.
[504] It was arguably consonant with the absence in Annex A of any express provision for legal representation corresponding to that in Annex B.1 for identification parades. On the other hand, D3.17(iii) required notification to the suspect of the right to legal advice and, as suggested by Bogan, op. cit., at para. 5.31 (n.33), Annex A.7, A.8 and A.9 therefore assumed (and assume) a solicitor’s involvement as by right. Since the suspect has no right to be present when the images are shown to the witness the purpose of notifying the unrepresented suspect of the time and place must be to enable the suspect to instruct a solicitor to attend: see Bogan, at para. 5.31. If (as may be assumed) prior to 2017 defence solicitors enjoyed the right to be present as an observer during the eye-witness viewing of the video images, the requirement (under the pre-2017 regime) to maintain a video-recording of the process only in the absence of a solicitor suggests that the absence of a solicitor was treated in practice as an exceptional circumstance requiring a video-recording of the viewing. For example, a suspect might have shown an utter disregard for their own potential peril and insouciant about seeking legal support. An example postulated by Bogan, ibid., is that of a critically ill witness where the urgency of conducting the video identification made it impracticable to give the solicitor reasonable notification. Given, however, the speed with which a video identification could be arranged (as to which see infra), it is difficult to envisage a situation in which no duty solicitor could be found in the time it would take to set up a video identification even though the solicitor’s role would have included advising on the selection choice of images.
[505] Bogan, op. cit., para. 5.54.
[506] Ibid.
[507] In Elson (1994) unreported C.A. 94/0547/Y3, 27 May, the foils bore so little resemblance to the defendant that the solicitor’s clerk took the view that a photograph was desirable and he stayed away from the parade in order to ensure that a photograph was taken in accordance with the code. In the event no camera was forthcoming.
[508] Wells, G.L., Small, M., Penrod, S., Malpass, R., Fulero, S. and Brimacombe, C., “Eyewitness identification procedures: Recommendations for lineups and photospreads,” (1998) 22 Law and Human Behaviour, pp.603-647. For a view favouring video validation see Kassin, S.M., “Eyewitness identification procedures: The fifth rule,” ibid., pp.649-653.
[509] PACE Code C11.1.
[510] It is assumed he was thinking of Birch, Bryant and Crowley (1992) unreported C.A. 90/0947/Z3, judgment March 10; see Butterworth’s PACE casebook, VIII, para. [930], in which it was held that the qualification of an identification by the words “I think” was a question for the jury. For the admissibility of qualified identifications see infra, Chapter 5.
[511] However, this was not before the defendant had gone on trial for another very serious indictment in which the fact that he had been wanted on warrant in the present case featured as a prominent aspect of the evidence in that trial. Because he was yet to be tried on the present indictment he was not permitted by the judge to state that he intended pleading Not Guilty in respect of it. In the event, he was acquitted on that other matter. The case was Maughan (Bernard), Wood Green Crown Court Indictment No. T98 0680. The present author has retained the video tape and the case papers, including the correspondence with the CPS complaining of the unconscionable delay before the prosecution was finally dropped. In Willoughby [1999] 2 Cr.App.R. 82, C.A.; CLW/99/25/4 (5 July, 1999), decided some months after the Wood Green case the witness on an identification parade said “It’s number 4, I think, it might be number 3” but subsequently made a statement saying she was sure that it was number 3. Held, the judge had been entitled to exercise his discretion to admit the evidence, having conducted a voir dire and having been satisfied that nothing had been said or done to prompt or encourage the change. However, it was desirable that wherever there was any significant difference in what was said on the parade and what is said in a subsequent statement, the suspect and his solicitor should be informed of this as soon as possible. Further, it was desirable that a witness should not be told whether the person picked out was the suspect until after the witness had reduced to writing his or her account of the parade.
[512] See above, B.(28), p.148, paragraph captioned “Preventing non-vocal cues during the identification process.”
[513] The face of any eye-witnesses and, in the case of enquiries linked to terrorism or to serious crime, any police officers or police staff present, may be concealed if the identification officer considers that this is justified: A.9. Note for Guidance A.1 states: “The purpose of allowing the faces to be concealed is to protect witnesses, police officers, or police staff where there is reliable information that suspects in serious organized crime or terrorism investigations or their associates, may threaten or cause harm to witnesses, police officers, or police staff.” This may mean that in such a case the defence will be deprived of any opportunity to observe facial gestures which may be inconsistent with, or at least tend to offset, an apparently unequivocal vocal identification.
[514] It is not now the practice in England and Wales to seek or permit dock identifications, but there is no absolute rule against them: see Tido v. The Queen, unreported, June 15, 2011, P.C., where it was said that the admission of such evidence is not to be regarded as permissible in only the most exceptional circumstances but a trial judge will always need to consider whether the admission of such testimony, particularly where it is the first occasion on which the accused is purportedly identified, should be permitted on the basis that its admission might imperil the fair trial of the accused. The discretion to admit the evidence must be exercised in light of the particular circumstances of the individual case, including consideration as to why an identification parade was not held. Where a dock identification is admitted, it will always be necessary to give the jury careful directions as to the dangers of relying on that evidence and, in particular, to warn them of the disadvantages to the accused of having been denied the opportunity of participating in an identification parade, if indeed he has been deprived of that opportunity. In such circumstances, the judge should draw directly to the attention of the jury that the possibility of an inconclusive result to an identification parade, if it had materialised, could have been deployed on the accused’s behalf to cast doubt on the accuracy of any subsequent identification. The jury should also be reminded of the obvious danger that a defendant occupying the dock might automatically be assumed by even a well-intentioned eye-witness to be the person who had committed the crime with which he was charged (considering Aurelio Pop (2003) 147 S.J. 692, P.C. (CLW/03/24/6), Holland v. H.M. Advocate, The Times, June 1, 2005, P.C. (CLW/05/21/5), Edwards (2006) 150 S.J. 570, P.C. (CLW/06/18/1), and saying that, in Edwards, the board went further than was warranted by the earlier authority). In Neilly v. R. [2012] 2 Cr.App.R. 248(20), P.C. it was again stressed that the decision whether to permit a dock identification was one for the trial judge to be exercised in the light of all the relevant circumstances. Ultimately, the question was one of fairness, bearing in mind the judge’s ability and duty to give appropriate directions in summing up, as indicated by the authorities. Where there has been no identification parade and whether there is any good reason (and if so what) for there not having been one, is a material circumstance. Where the uncontroversial evidence is that the defendant was well-known to the witness before the offence, and the witness has previously identified him, a dock identification may be no more than a formality.
[515] The topic is comprehensively discussed in Bogan, op. cit., at paras. 10.38 to 10.43. As the latter commentator rightly points out at para 10.38, n.52, five different approaches were taken in the earlier leading authority Christie [1914] AC 545, H.L. and more recent authorities, notably McKay (1900) 91 Cr.App.R. 84; [1990] Crim.L.R. 338, C.A., and Osborne and Virtue [1973] 57 Cr.App.R. 297; [1973] Crim.L.R. 178, C.A., have attracted academic criticism.
[516] Supra.
[517] Supra.
[518] Supra.
[519] Bogan, op. cit., para. 10.41.
[520] See above Chapter 4, B. (24), pp.138-140, on statements of confidence.
[521] [2003] Crim.L.R. 282, C.A., discussed in depth by Bogan, op. cit., paras. 10.51-10.56.
[522] Popat [1998] 2 Cr.App.R. 208, at pp.212 and 224, C.A.
[523] Allen [1999] Crim.L.R. 643, C.A.
[524] The appellate court may sometimes be confronted with the duty to unravel such curiosities as the rulings in Harris [2003] EWCA Crim 174, discussed above at Chapter 3, pp.75-77.
[525] Where a defendant is unreasonably denied an identification procedure, a court should be slow to permit the prosecution to call evidence, such as facial mapping, to fill a gap in the identification evidence: Walker (1994) unreported C.A. 14 November.
[526] [2001] 1 A.C., at p.487.
[527] Ibid., at p.489. See also Harris, supra.
[528] Such a failure has been likened to the dangers of a dock identification: see Maynard (1979) 69 Cr.App.R. 309, C.A.
[529] See above, Chapter 3, pp.75-77.
[530] See Allen [1995] Crim. L.R. 643, C.A. (fairness of reception of the complainant’s description of the perpetrator of a purse-snatch held vitiated by the failure to hold a parade).
[531] See e.g., Powell v D.P.P [1992] RTR 270, DC; Joseph [1994] Crim.L.R. 48, C.A. (exclusion of confrontation in court custody area held at request of defence after a ruling that all other procedures were impracticable). Cf. O’Brien [2003] EWCA Crim 1370 (no code breach or unfairness where appellants insisted on exonerating themselves after a public house fight).
[532] See Jones and Nelson (1999) unreported, 26 march, C.A., cited by Bogan, op. cit., para. 10.64, (exclusion where appellant compelled by force to participate in confrontation).
[533] See Nagah (1991) 92 Cr.App.R. 344, C.A. (Code held completely flouted where the appellant was released from the police station while the complainant waited outside in a car and identified him as the man who had raped her); Fuller (2000) unreported, 22 May, C.A. (street identification a “gross violation” where suspect known and failure to record first description); K v DPP (2003) unreported, 11 February, C.A. (known suspect, and also no formal controlled procedure).
[534] See Hickin [1996] Crim.L.R. 584, C.A. (evidence excluded where witnesses made identifications in each others presence, no prior descriptions were taken athough this was not then a code requirement, and no records were kept as to which witness had identified which suspect); Vaughan (1997) The Independent 12 May, C.A. (judgment 30 April, no first description recorded).
[535] Gall [1989] Crim.L.R. 745, C.A. (officer escorted a witness to the parade and saw the parade before the viewing began; held the evidence ought to have been excluded)
[536] Finley [1993] Crim.L.R. 50, C.A. (one witness to a photographic identification certain to have overheard remarks by another when identifying the defendant; risk of contamination between four witnesses either before or when attending a parade together; conviction quashed for these and other reasons).
[537] See ibid., in which the defendant was the sole blond skinhead on display in a photograpic array in contravention of Annex E.4.
[538] See Quinn [1995] 1 Cr.App.R. 480, C.A.; Mendili [2001] EWCA Crim 757.
[539] (1977) 98 Cr.App.R. 313, C.A. The topic is extensive and is considered beyond the scope of this treatise.
[540] [2001] 1 Cr.App.R. 430, H.L., followed in Harris [2003] EWCA Crim 174 (supra., at pp.73-77).
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.