harp graphic.

THE COURT OF APPEAL

APPROVED                                                                  Appeal Record Number: 2022/022

                                                                                       Neutral Citation: [2023] IECA 330

Bill SCDP0009/2019

Birmingham P.

McCarthy J.                                                                                    

Ní Raifeartaigh J.

 

BETWEEN/

 

THE PEOPLE AT THE SUIT OF THE DIRECTOR OF PUBLIC PROSECUTIONS

RESPONDENT

-AND-

 

DARREN MURPHY

 

APPELLANT

JUDGMENT of the Court delivered on the 19^th day of December 2023 by Ms. Justice Ní Raifeartaigh

Introduction

1.                  This is an appeal in respect of the appellant's conviction for murder following a trial in the Special Criminal Court in June and July 2021. Judgment was delivered by the court on the 15^th of November 2021. The appellant's co-accused DK was found not guilty of murder and not guilty of possession of firearms. The appellant was found guilty of murder and not guilty of possession of firearms.

2.                   Most of the issues raised by the appellant relate to evidence of identification/recognition of him by members of An Garda Síochána from CCTV footage. The appellant also challenges the trial court's conclusions concerning common design, and its reliance upon CCTV footage despite a number of discrepancies within it in terms of timing.

3.                  There was originally a ground of appeal relating to the search of a vehicle but counsel on behalf of the appellant indicated at the appeal hearing that this was not being pursued because nothing obtained from that search was deployed by the prosecution to establish the guilt of the appellant. The identification issue was the one most strongly pressed at the hearing of the appeal.

Background

4.                  Mr. Eamon Kelly was shot and killed at 16.00 hours on the afternoon of the 4th of December 2012 at Furry Park Road, Killester, in the city of Dublin. A male emerged from a Lexus motor car, ran after him and shot him a number of times with a handgun, fatally wounding him. He was shot four times and a Glock handgun was used. The gunman made his escape in the Lexus driven by a second male. The Lexus was set on fire in Stiles Court, Clontarf, approximately 1.1 km from the murder scene, at 16.04 hours. The two occupants fled on foot down an alley which leads to a vehicular entrance to Clontarf rugby and cricket clubs and then onto Castle Avenue, Clontarf. The two occupants of the car were the shooter and his getaway driver. Sean Connolly was arrested very shortly afterwards, on nearby Castle Avenue. He was not in possession of the firearm but was wearing two tracksuits and latex gloves. His clothing was found to hold particles characteristic of gunshot residue, and this had peculiarities which matched the spent shells ejected from the gun used to kill Eamon Kelly. He subsequently pleaded guilty to the murder. The gun was subsequently recovered by police from a location in Northern Ireland.

5.                  At the trial of the appellant and his co-accused DK, the prosecution case was based on a number of strands including a number of pieces of CCTV evidence concerning the movements of certain vehicles and persons. It included evidence that in the days preceding the murder, a car associated with the appellant (an Opel Meriva) had been repeatedly visiting the area where the murder ultimately took place; that this car was in the area shortly before and at the time of the murder; and that at particular times this car was travelling in convoy with two other vehicles one of which was the Lexus from which the gunman emerged. The linkage between the appellant and this car and the other men depended on CCTV footage.  For present purposes, the following pieces of CCTV footage should be highlighted.

6.                The Howth Road footage-There was CCTV footage which showed, inter alia, that at approximately 14:46 three vehicles were driving in convoy on the Howth Road northbound towards Killester Village (i) a green or turquoise Opel Meriva, (ii) a black Lexus (which was the car from which the gunman Sean Connolly was later to emerge), and (iii) a black Toyota Landcruiser. The Opel Meriva had distinctive features, one of which was the absence of hubcaps. The driver had white stripes on dark sleeves.

7.                  The Opel Meriva took up position parked on The Demesne close to the junction with the Howth Road at approximately 15:28, with a clear view onto the Howth Road. A man (who the prosecution contended was the appellant) exited the vehicle at 15:49:12 and turned right onto the Howth Road, and then returned to his vehicle at 15:50:35. Eamon Kelly walked past the Bank of Ireland on the Howth Road, very close to the junction, at 15:55:14 and the Opel Meriva departed and drove northbound on the Howth Road towards Raheny at 15.55.39. As already noted, the shooting took place at 16.00 hours. Immediately after the shooting, the Opel Meriva was travelling southbound on Howth Road at Killester Village at 16:01:32, followed by the black Lexus only 8 seconds later, in the direction of Stiles Court where the Lexus vehicle was shortly thereafter burnt out.

8.                  Inner city footage- There was footage of three men in the inner city which the prosecution contended were the appellant, Sean Connolly and DK. This footage was however of relatively poor quality. The prosecution therefore relied on a different piece of footage, which will here be called "the Temple Street footage", in which (the prosecution contended) the appellant was clearly shown, to connect him to the man interacting with Sean Connolly and DK in the inner city. The timing, location and appearance of the man was said to connect the two pieces of footage.

9.                  The Temple Street footage- There was CCTV footage which showed a man in an Adidas top walking a distinctive dog at Temple Street Hospital, Spar Temple Street and various locations in Eccles Street and walking from and into St. George's Place between 12.13 and 12.50 hours on the date of the murder. The distinctive dog was a Kerry Blue and was on a lead with an orange or red spool handle. The man appeared to be the same man (from his clothing, general physical appearance) who got out of the Opel Meriva in Killester, and the prosecution relied upon the Temple Street footage to say that the man at the Opel Meriva in Killester was the appellant. By reason of the importance of this footage in connecting the appellant with Sean Connolly, the identification of the appellant as the man walking the dog at Temple Street assumed particular importance at the trial. Seven Gardaí, none of whom were involved in the murder investigation, said they recognised the appellant from that footage in 2018. Another twenty-five Gardaí who were shown the footage did not identify the appellant; some said they recognised no one while others named different individuals.

10.              Before turning to the process and procedures relating to the recognition/identifications of the appellant, it may be noted that the evidence also established that the Opel Meriva was parked in the communal car park outside the appellant's flat at St. George's Place between 1^st and 8^th December and was registered at that address to his partner who was also living there at the time. It was parked at the back of the car park when the Gardaí arrived to arrest him four days after the murder. The court was separately satisfied that it was the same car in other CCTV evidence showing that on the 1^st, 3^rd and 4^th December 2012 it left from the car park for Killester and returned to that location from Killester. It repeatedly emerged from the Dunseverick Road to the demesne on to Howth Road. Further, the Toyota Land Cruiser travelling in convoy on the afternoon of the murder had been in the car park of the St. George's Place flats the previous day.

The procedures in place concerning identification.

11.              The process by which the Temple Street footage was shown to Gardaí was as follows. In the first instance, a list was compiled of Gardaí who knew the appellant. Detective Superintendent Paul Scott then put in place a process by which Gardaí were to be invited to view the footage to see if they could identify the person on it. He gave Inspector Brian Downey the task of overseeing the job, who in turn selected two members to carry it out.

12.              Inspector Downey said that he wanted to ensure that each occasion of identification would be carried out "in isolation" from other members and that the place where the CCTV footage was shown was "a sterile environment", preferably in an interview room, where the viewing and identification process could be recorded. He suggested that contemporaneous notes would be taken by members involved and that statements should be taken or provided. Sergeant Farrelly was allocated the task of managing the identification process in respect of the appellant. Inspector Downey himself did not know what the identification or what the investigation was about, and neither did the Sergeants: this was described as a "double blind" to ensure the integrity of the process.

13.              Inspector Downey gave evidence that the process involved the Sergeant showing the footage and asking the member viewing the footage the following: 'I am now going to show you CCTV footage. You may stop the footage at any time, rewind or go forward. If you identify anybody on the CCTV footage, I would like you to identify that person and state the person's name'", or equivalent words. The Sergeant was to note down if the member identified or did not identify a person from the CCTV footage. The Sergeant would take a statement or invite the member to write a statement and the taking or provision of the statement by the member was to be recorded on the CCTV in the interview room itself. The member who was invited in to identify a person from the DVD would be advised not to communicate his identification to any other member or other person. The recorded CCTV footage from the interview room would be retained by the Sergeant conducting the independent identification process should it be required in court. The Sergeant conducting the independent identification process as outlined above would note down the time and date each listed member attends Balbriggan Garda Station so that the member's arrival and time spent in the interview room would correlate with the independent station CCTV system should footage from the system be required in court on a future date.

14.              The Temple Street footage bore a date stamp of 4 December 2012 which was the date of the murder. It carried no other information as to what investigation was taking place.

15.              The process conducted by Sergeant Farrelly was commenced and completed in 2018, some six years after the murder. The seven Gardaí who identified the appellant were from a pool of thirty-two Gardaí who viewed the footage.

16.              At the trial, there was considerable focus upon matters such as whether the members who identified the appellant on the "Temple Street" footage had foreknowledge of what the investigation was about, whether they made a connection from the date stamp with the Eamon Kelly murder, whether they knew the appellant was a suspect in the Eamon Kelly murder, and whether they had spoken to other members about the process before engaging in it. In those circumstances, it is necessary to summarise the evidence concerning each of the seven identifying members.

The evidence of the Gardaí who positively identified/recognised the appellant from the footage

17.              The footage started with a clip of a man walking past Temple Street Children's Hospital entrance on the footpath. He was walking with a distinctive dog on a lead with an orange or red spool handle. This clip was followed by three clips of a man wearing a baseball-type cap walking from an Opel Meriva car which was parked in the demesne in Killester facing out onto Howth Road. The footage concluded with four still frames which had been taken from the clips already shown. The location of the CCTV footage was not identified to the witnesses. The period of time when a man appeared on the CCTV clips was short: approximately 9 seconds, 5 seconds, 7 seconds and 5 seconds. It appears that the footage in Killester showed less of the man's face, but it showed the clothes and footwear well. The man wearing the baseball cap in the later footage was wearing the same clothing and was, as the trial court said, "exactly the same build, shape of head and hairline" as the man in the Temple Street footage.

18.              During the trial, and upon request made by the appellant, the prosecution compiled information about the Gardaí who had identified/recognised the appellant from the footage in terms of their previous entries to, or inquiries made of, the Garda PULSE system concerning either or both of the appellant and the Eamon Kelly murder. This information was then deployed in cross-examination of each of the relevant Gardaí.

19.              Garda Laura Hayden was stationed in Mountjoy Garda station at the time the appellant was living at George's Place, Dublin 1. On 30^th August 2018, she attended Balbriggan Garda station and said she recognized the appellant from CCTV footage. In cross-examination she said she was not aware that the appellant had been arrested on 8^th December 2012 in connection with the murder. The appellant maintains that this denial is not credible in light of the following facts:-

·         On the 5^th of December 2012, the day after the murder, she accessed the Pulse record in relation to the murder.

·         On the 11^th of December 2012, seven days after the murder, she accessed the Pulse record in relation to the murder.

·         She made Pulse entries about the appellant on three occasions in 2013.

·         Prior to attending at Balbriggan, she discussed her attendance there with her immediate superior, Sergeant Aoife Reilly.

·         Garda Hayden is married to Garda Mark Deasy who also identified the appellant. (She gave evidence that she knew he had to go out to Balbriggan but that was as much as she knew).

20.              On the 24^th of July 2018, Sergeant Caitriona Brody attended Balbriggan Garda station to view CCTV footage and said she recognied the appellant from it. She was attached to community policing in the north inner city (Mountjoy and Fitzgibbon Street stations) from 2006 to 2014. She said she met him walking his dog when she was on duty in the area (before 2014) and knew he lived between Rory O'Connor House and George's Place. She said the date on the footage meant nothing to her and denied knowing that the appellant was a suspect in the Eamon Kelly murder. She said she did not know that the appellant appeared before the Special Criminal Court on 5^th April 2013. The appellant contends her denials were not credible in light of the following facts:

·         She made Pulse inquiries regarding the Eamon Kelly murder on the day after the murder, the 5^th December 2013.

·         She made Pulse entries regarding the appellant on 8^th December 2012, the date of his arrest.

·         She made further inquiries concerning the appellant on 5^th April 2013, 16^th April 2013 and 20^th April 2013.

·         She made an entry on 5^th April 2013 regarding the appellant's admission to bail on charges before the Special Criminal Court, and further entries on 16^th and 20^th April 2013.

21.              Garda Aoife Farrelly, who was stationed to Mountjoy station in 2011 and was still there in 2018, attended Balbriggan on 31^st July 2018 and said she recognied the appellant in the footage but only by his first name. She said she knew him from signing on in Mountjoy station and that he was from the "George's Pocket" area, as it was locally known. She made the identification from the man with the dog in the CCTV clip and not any of the other clips. She denied that she knew that the appellant was a suspect in the murder of Eamon Kelly when she went out to do the viewing and said she did not make the connection when she saw the date stamp. The appellant contends that this denial is not credible in light of the following facts:-

·         On the 10^th of December 2012, she accessed Pulse concerning the appellant, which was two days after his arrest for the Eamon Kelly murder.

·         On 25^th July 2015 she made an entry on Pulse regarding the appellant at 1.14pm and four minutes later she accessed Pulse in relation to the Eamon Kelly murder. (When asked about this in cross-examination and whether she had made the connection at that time, she said she accepted she had made these entries/inquiries but had no recollection of it and could not speculate on something she could not recall. She insisted the recognition/identification was genuine).

22.              Garda Shane Darcy, of Mountjoy station, attended at Balbriggan on the 13^th of August 2018 and said he recognised the appellant. He said he recognised the man on the clips as the appellant who was wearing the Adidas hoodie. He said he had been attached to Mountjoy since 2007 and had encountered the appellant on numerous occasions, possibly over 100 times. He knew he lived between Rory O'Connor House and George's Place but did not know the exact address. In cross-examination he accepted that he made the identification after the footage was played a second time. He gave evidence that he did not know that the appellant was a suspect in the Eamon Kelly murder. His dealings with him were when he was in uniform and the appellant was signing on in the station on a daily basis. He said that "this happened in 2018 and I wouldn't think back to a murder six years ago when I was put on the spot with CCTV in Balbriggan at that time, but I did know who I was looking at". He said he did not put any weight on the date stamp. The appellant contends that his denial that they knew the appellant was a suspect for the Eamon Kelly murder is not credible for the following reasons:

·         He made six Pulse entries concerning the appellant between May 2013 and August 2016.

·         He made a Pulse inquiry in relation to the Eamonn Kelly murder on 4^th July 2020 (although it is difficult to see the relevance of this, given that it was some two years after his identification of the appellant from the CCTV footage).

·         He discussed his attendance at Balbriggan with his Sergeant, Hugh McInerny, in advance of attending (although he said this was purely to tell him about taking the time off so that he would get paid).

23.              Garda Mark Deasy viewed CCTV footage at Balbriggan Garda station on the 23^rd August 2018 and said he recognized the appellant from the CCTV footage. He said that he thought he recognised the appellant from all the clips. He said he knew him from frequenting Mountjoy station when the appellant was signing on. He denied knowing that appellant was a suspect in the Eamon Kelly murder. The appellant contends this is not credible in light of the following facts:-

·          He is married to Garda Laura Hayden who also made an identification of the appellant and he discussed going to Balbriggan with her. (He said in evidence was all that he told her was that he was going to Balbriggan. He worked in a different office to her).

·         He accessed Pulse regarding the appellant on the 30^th May 2013, 12^th August 2013 and 14^th May 2017.

24.              Garda Emer Cantwell, who was in Mountjoy from 2008 and still there in 2018, attended Balbriggan Garda station on the 28th of August 2018, and said she recognized the appellant from CCTV footage. She was attached to the community policing unit in the station from 2016 and knew him from that and knew where he lived. She referred to the clip with the dog, and said she remembered that he was a lover of animals and went everywhere with the Kerry Blue dog shown in the clip. In cross-examination she said that his fondness for the dog was something that stuck in her head. She also associated him with a car shown in the clip. She noted the date stamp but gave evidence that it was of no significance to her. She said she had no knowledge at all regarding the Eamon Kelly murder either before the 2018 trip to Balbriggan or at the time of the identification. She said that she made many entries and would be checking things on Pulse regularly because of her job in community policing. The appellant contends her evidence is not credible in light of the following facts:

·         Garda Cantwell made 25 Pulse entries concerning the appellant between January 2015 and December 2018.

·         She made inquiries about the Eamon Kelly murder on the date of the murder itself.

·         She made further inquiries about the Eamon Kelly murder on four further occasions: 9^th October 2016 (on which date she also accessed Pulse in relation to the appellant), 21^st January 2017 (on which date she also accessed Pulse in relation to the appellant), 8^th May 2019 (twice) and 3^rd April 2021.

·         She discussed going to Balbriggan with Sergeant Paul Madden.

·         She ultimately agreed in cross-examination that she must have been aware on the 9^th October 2016 there was a connection between the appellant and the Eamon Kelly murder

25.              Garda Robert Tonkin, who was attached to Mountjoy station in 2018 and had been there since 2009, attended at Balbriggan on 13^th August 2018 and said he recognized the appellant from the CCTV footage. He said he identified him as a man resident in their district and he had a rough idea of where he lived. He had interactions with him previously in the public office, from signing on. When he came in to sign on he had a dog with him. He viewed the footage again in September 2019 to clarify his earlier statement. He said he recognised the appellant from the Temple Street footage. He said he did not know that the appellant was a suspect in the murder of Eamon Kelly. He said he was not aware of the date of the Eamon Kelly murder as he was not in their district and that the date stamp did not "ring any bells" for him. The appellant maintains this denial is not credible in light of the following facts:-

·         Garda Tonkin accessed Pulse regarding the appellant on 27^th May 2013 at 13.29. He accessed Pulse regarding the Eamon Kelly murder at 13.33 on the same day. (In cross-examination he said he had no recollection of this and did not recall any discussion about a connection between the two at the time. He added, when questioned by the presiding judge, that if he did make the connection in 2013, he would have forgotten it by 2018. In any event, he said, he did not make the connection when he viewed the footage).

·         He accessed Pulse regarding the appellant on 16 June 2013 and accessed the Pulse record regarding the Eamon Kelly murder twice on the same day (12.50 and 13.00).

·         He accessed Pulse regarding the appellant on 2 March 2014.

·         He accessed Pulse regarding the appellant on 3 September 2016.

The admissibility challenge

26.              At the trial, as is the usual procedure in the Special Criminal Court which sits without a jury, there was no separate voir dire on admissibility prior to hearing the evidence in front of the decision-maker. The evidence was heard and its admissibility ruled upon in July 2021.

27.              At trial, the appellant made many criticisms of the process, including that the CCTV was of a single individual and did not amount to an identification parade as there were no "foils"; that the footage was overprinted with the date of the offence being '04/12/12'; and that the Gardaí had accessed PULSE in respect of the appellant and the Eamon Kelly murder, and in three instances had done so at the same time on the same day, showing that they must have known of the connection. In the circumstances, it was submitted, the identification procedure did not amount to a forensically "sterile" process because it was was contaminated with the prior knowledge and actions of the identifying Gardaí.

28.              The trial court found the identification evidence admissible in a ruling on the 12^th of July 2021. Having described the compilation of the list of Gardaí who knew the appellant, the court went on to describe the essential features of the process adopted:

"Sergeant Farrelly was unaware that the footage related to the Eamon Kelly murder, or the potential identity of any person shown on the footage. He did not know what the footage related to. He knew that his role was to summon each of those on the list to attend separately at Balbriggan Garda Station and to show the footage to each name and ask if that person could identify anybody on the footage. Following each viewing, he instructed each attendee not to discuss the matter with anybody. He took notes of his interactions with each attendee during this exercise. He e-mailed each proposed attendee separately.

The process of showing the footage was video recorded. This recording included and also audio recorded. This recording included any comments of those shown the footage, and in the event that a person shown the footage was in the position to make an identification, the person was required to make a written statement immediately. The instructions given did not extend to any specific request to those asked to look at the look at the footage to identify any point in the timing of the footage at which the viewer came to a conclusion that a specific person could be recognised, or to indicate any specific features leading to the recognition.

The viewers were not asked whether they could identify the location shown in the CCTV footage. Each viewer of the footage would not know in advance of attendance at Balbriggan Garda Station what footage would be shown or what offence or investigation that footage might relate to. They knew that they were going to Balbriggan to look at CCTV footage to see if they could recognise a person on the footage and nothing more. The text of the e-mail said to each potential participant is to the found in the transcript for 6^th July 2021 at page 21. Some of them were aware that others in their station had also been asked to go to Balbriggan.

 

The viewers were not told why they were being asked to view the footage, beyond that they were asked to look at it, and they were then asked when viewing the footage to say whether they recognised any person on it. The purpose of Detective Superintendent Scott in putting this process in place is obvious, it was to ensure that the circumstances in which any potential witness who purported to have recognise Darren Murphy from the CCTV footage shown were fully recorded, and strengthen the quality of the recognition by excluding the possibility that the witness would know in advance that the exercise concerned either Darren Murphy or the Eamon Kelly murder."

29.              The trial court observed that this jurisdiction does not have "any set list of procedures governing the controlled showing of films, photographs or images with a view to obtaining evidence of recognition" and that it was clear from People Director of Public Prosecutions v. Tynan [2017] IECA 202 that procedures of the type embodied in code D part B of the Police and Criminal Evidence Act in England and Wales ("PACE"), as discussed in R v. Smith (Dean Martin) [2008] EWCA Crim 1342; [2009] Cr App R 36, were neither mandatory nor directory procedures in Irish law.

30.              The trial court went on to say that that there was " no great mystery about the process of recognising a person from CCTV footage" and that "[i]n principle, any person who knows the appearance of a person sufficiently well to recognise that person and who is shown photograph imagery of that person of sufficient quality to identify that other person on the footage and who recognises that other person may give evidence of that recognition, whether the footage is sufficient in quality or quantity to make the recognition, and whether the evidence of the witness is reliable or matters which go to the probative value of the evidence".

31.              The court then referred to the danger of "confirmation bias", saying:-

"Another matter which goes to the weight which should be attached to recognition evidence of persons or vehicles on CCTV is the possibility of confirmation bias. If a person expects to see something, that person may be confirmed in an expectation. There's no rule that persons who are not involved in an investigation and who are familiar with a suspect because of interaction with that suspect during an investigation, may not look at CCTV footage, and give evidence that they recognise the person shown on the footage as the suspect. In many cases the CCTV footage will be of good quality, and it'll be perfectly obvious that the footage shows a particular person or the persons viewing the footage will have had a very recent opportunity to fully assess the suspect's appearance from recent interactions with that person, which will enable them to positively recognise a person on footage as one and the same person as the suspect."

32.              The trial court contrasted identification evidence with recognition evidence:

"There are differences between identification by witnesses of a person who is not known by those witnesses and recognition of a person who is known from CCTV footage. The recognition process is not a casual process of the sort which we may engage in when passing a person on the street. The use of CCTV gives the viewer the ability to look at the footage on a number of occasions and to look at any stills and the possibility to pause the footage, taking details and come to a conclusion. The rationale for identification parades when they can be arranged is to test the ability of a witness to make an identification of a suspect which may be fleeting. The value of the test depends on the appropriate safeguards relating to identification parades being in place. And if these are not in place, the evidence may be excluded.

 

The concern of the Court is that the evidence of identification may be -- in other words, should be sufficiently reliable to be left to the jury. This is a rare area in which it is necessary for judges to involve themselves in an assessment of the weight to be given to a particular item of evidence which arises from the special risks which may attach to identification evidence. If the identification parade is conducted in a manner which makes it unreliable, or if the witness is given advanced help in making an application, a court will exclude such evidence as inherently unreliable.

While the courts do appreciate that a recognition by a witness from CCTV may be mistaken and that factors such as confirmation bias may enter into an assessment of recognition where a person appearing on screen looks like or has similar appearances to a person who may be known to be a suspect for a particular offence or for a range of offences, or to be likely to be involved in crime. The stage has not yet been reached when this type of evidence is considered inherently unreliable. In general, evidence which is relevant is also admissible, subject to the rule that the evidence may be excluded as inadmissible, where its probative value is outweighed by its prejudicial effect. This would happen if a piece of evidence of some characteristics which might prove little, but which would disproportionately introduce other material unfavourable to an accused or where evidence is inherently unreliable as a result perhaps of a flawed identification procedure."

33.              Having made the above observations, the court expressed the conclusion that it was not satisfied that there was any basis on which the evidence of recognition should be ruled inadmissible:

"No basis has been identified on which we could exclude the evidence of the witnesses who have given evidence of identification of Darren Murphy on the CCTV footage shown to them. There was nothing unfair about how the process was conducted, the recognition process was recorded on sound and CCTV, and the defence had reasonable opportunity to assess the reactions of the witnesses when confronted with the footage.

The witnesses were not told in advance about the offence being investigated or about where the footage was taken. It was inevitable that the witnesses would know that they were selected because of a possibility that they would be in a position to assist because of knowledge of the appearance of somebody in their area. There's always a prospect that a garda member in a particular area, even if not directly involved with an investigation may become incidentally aware that a particular person is associated with a particular crime. But that does not render this type of recognition exercise inadmissible.

Gardaí consult PULSE which holds information for all sorts of reasons, and they may be expected to know something about those who live in their area who are suspected of criminal involvement. To exclude on the basis of such actual or potential knowledge, recognition evidence from a witness who has encountered a suspect in daily affairs and is familiar with the appearance of that suspect from work as a garda would give an unfair advantage to those suspected of offences, as it might well preclude any diligent garda from being called to give recognition evidence."

34.              The court returned to Code D, Part B of PACE which it had mentioned earlier, commenting that "broadly speaking", the procedure used here was similar and designed to reduce the possibility of mistaken recognition and avoid any possibility that the witnesses would collude. It accepted that some of the Code D instructions were not replicated, such as (i) the Gardaí being asked to view the material twice before making a decision; (ii) being asked whether they had recognised the location shown in the footage; or (iii) asked about the reasons for their conclusion or features of the image which led to the recognition. With regard to the latter point, the trial court said:

"Recognition is often a matter of overall assessment which in many cases may not involve isolation of particular features. The reason for a recognition is that the person asked to study the footage already knows the person shown in the footage and can put a name on the person shown on that footage."

35.              The trial court also mentioned the date stamp issue. It concluded that all of these matters were "issues which are relevant to the weight which we may attach to recognition evidence" but that nothing had been identified that made the evidence "so inherently unreliable either by reason of some unfair gathering or record keeping process or because of some indication of prompting or poor record keeping or lack of transparency or lack of real opportunity of the defence to test it, that [the court] should reject it as inadmissible".

36.              The trial court went on to say that the information provided to the appellant by way of disclosure concerning logging on to Pulse records gave the appellant an adequate opportunity to challenge the reliability of this evidence, and issues identified by the defence could be fully assessed as part of the court's evaluation of the weight of the evidence.

The Trial Court's conclusions with regard to the identification evidence

37.    As to the weight to be afforded to the evidence, in its ultimate judgment on the 15^th of November, the trial court said as follows:-

"The evidence establishes to our satisfaction that the man in the Adidas top walking a distinctive dog who can be seen in the CCTV footage at Temple Street Children's Hospital, Spar Temple Street and at various locations in Eccles Street and walking out from and into St George's Place in the period between 12:13 hours and 12:50 hours on the afternoon of the murder was Darren Murphy. There can be no possible doubt of that.

This Court accepts the evidence of the gardaí who looked at CCTV footage and stills from that footage and identified Darren Murphy from this material as a person who they knew and recognised when they attended Balbriggan Garda Station in 2018. This Court also accepts their evidence that they were unaware that they were being asked to make an identification which related to the murder of Eamon Kelly when they were confronted with the CCTV footage and images at Balbriggan Garda Station in 2018. This Court has taken into consideration that some of these witnesses accessed PULSE records in early years in circumstances which showed that when they made these inquiries they knew or were likely to have known from PULSE information and from the nature of their inquiries that Darren Murphy was suspected of having a role in the murder of Eamon Kelly. These witnesses were firm in their recognition evidence. We accepted the evidence of these witnesses that they were not aware that they had made these PULSE inquiries or made a connection between Darren Murphy and the Eamon Kelly murder when they made their recognition identifications of Darren Murphy from the CCTV footage. These witnesses were not part of the investigation team for the murder. They had encountered Darren Murphy in the course of community policing duties or as station gardaí when he was signing on at garda stations and two of the witnesses recalled him walking a dog on a lead.

[The trial court then repeated the procedures which had been followed, already set out above and continued]:

The witnesses did not notice where the CCTV clips were taken from at the time they made their identifications of Darren Murphy in Balbriggan Garda Station. Three of the clips are from a suburban setting which gardaí operating in the Mountjoy Station and Fitzgibbon Street Station areas of Dublin would be unlikely to recognise. The first clip is taken from a camera mounted on the wall of the front of Temple Street Children's Hospital. There is nothing in the footage which would enable a person who was not told of this fact or who was not familiar with the spot to immediately recognise this location. The witnesses were not shown other available footage attributed to the same man at other city centre locations which might more readily have prompted recognition of the location of a CCTV camera.

This Court has taken into account the points made on behalf of the defence about the delay between the murder in 2012 and the exercise carried out at the behest of Detective Inspector Scott, as he then was, at Balbriggan Garda Station in 2018, the dangers of confirmation by witnesses who are familiar with the appearance of a person or with an incident or asked to look at footage and to see if they recognise the person shown or where two pieces of different footage may or may not show the same person or thing. It is necessary to be cautious in drawing conclusions as to the identity of persons or vehicles from recognitions made by witnesses viewing CCTV footage because it is possible to make errors and jump to conclusions which may not be warranted by the footage. And this Court has taken these points into consideration in assessing the evidence. This Court has had regard to both pointers within the evidence of individual witnesses who gave evidence that they encountered Darren Murphy and remembered him walking a dog and that they were aware of where he resided and external supporting evidence that Darren Murphy did, indeed, reside at the location identified by them and had an association with the turquoise Meriva car parked in the communal carpark at that location around the time of the murder.

Independent evidence associating Darren Murphy with this car might strengthen the reliability of any evidence of recognition of Darren Murphy in the suburban setting. Matters which may support the reliability of evidence of recognition may include circumstantial evidence from other sources which may connect the person recognised in the footage to a particular location or object, such as a dog, or a place or a vehicle. In assessing the reliability of the evidence of the gardaí who identified Darren Murphy as the man in the first clip of footage at Temple Street, this Court has taken into account supporting evidence. This includes the evidence from sources other than the identifying witnesses that Darren Murphy did, indeed, reside in the flats complex at St George's Place during the relevant period. It is also clear from the CCTV footage played in court that the movements of the man walking the dog -- identified as Darren Murphy -- on the day of the murder started and ended in the flats complex at St George's Place.

40. In a passage which the appellant contends amounted to a "dock identification" by the court itself, the court said:-

In evaluating recognition evidence or any other evidence which tends to identify or associate a person shown on CCTV footage as being a particular person it is necessary to use -- to be realistic and use common-sense. This Court considers that the footage in the Temple Street Children's Hospital to be of good quality. This Court was not invited to examine the CCTV footage and carry out a recognition between Darren Murphy as he appeared in court and the man shown on the footage. This Court has had an opportunity to observe Darren Murphy during the trial. This includes periods when he was standing up and walking. This trial has taken place some eight and a half years after the events of the 4th of December 2012. This Court would be most reluctant to perform such an exercise of making a decision based on a comparison between the footage and the person present in court in this trial. This Court does not know what Darren Murphy looked like in 2012. His appearance may have changed somewhat over the years. At the same time it is unrealistic to engage in an assessment of the evidence from which a recognition has been made and which may also in this case be accompanied by other evidence of a circumstantial nature which also tends to support the reliability of that recognition which is totally divorced from what is visible in court for all to see over a number of weeks of a trial...This Court cannot pretend that Darren Murphy was not present during the trial and that his present appearance is not known. The appearance of Darren Murphy in court during the trial does have a relevance. Where CCTV footage shows various aspects of the appearance of a person, similarities and dissimilarities with the person visible in court can be taken into account in assessing the reliability of the recognition evidence. This Court can say from viewing Darren Murphy during the trial that there is no inconsistency between his features as observed in this court and those which were visible in the CCTV footage of the man identified by the witnesses as Darren Murphy which would lead this Court to doubt the correctness of the recognition evidence or undermine in any way the other circumstantial evidence which supports this Court's conclusions that Darren Murphy is on the Temple Street Children's Hospital CCTV footage and the demesne CCTV footage shown to the witnesses and the other footage of him, indeed, at Eccles Street and at Spar, Temple Street." (Emphasis added)

38.  A little further on, the court discussed recognition evidence in general:-

"Recognition of a person who may be known to the viewer invited to look at CCTV footage is not an expert or scientific process. Either the person on the CCTV footage is recognised by the viewer or no person is recognised. The strength of the evidence of a recognition, as we have -- as the Court has indicated, depends on all of the relevant circumstances. These include the quality of the footage shown to the witness, the foreknowledge of the witness of the person who may be recognised, the matter being investigated and other matters which the witness mentions which may be independently proven and which point to the reliability, or unreliability indeed, of the identification. The exercise conducted at Balbriggan Garda Station in 2018 was controlled and cannot be compared with the sort of incidents where strangers are mistakenly confused with a person known to another person, as sometimes happens when a stranger is mistakenly greeted as a friend or acquaintance on the street. It is apparent that some of the witnesses were not prepared to make an identification by reference to the demesne footage, presumably because of absence of sufficient detail of facial features, and that others took an overall view. The recognition of a person on CCTV is not just a matter of looking at the face of the person shown on the footage. Build, height, hairline, shape of head and the sort of clothes which the person recognised was remembered as wearing may all be factors. A striking feature of the recognition evidence of some of the witnesses at the trial was the memory that Darren Murphy walked the dog and those witnesses encountered him in the course of their work.

We accept that other officers who were asked to perform the same exercise as the gardaí and -- exercise of the gardaí who gave evidence made incorrect identifications. Other officers may have known Darren Murphy and may have been in a position to -- who may have known Darren Murphy and may have been in a position to recognise him from the clips were unable to do so. There may have been many reasons for this which we're not prepared to speculate on.

We consider the footage of the man at the Temple Street Children's Hospital as reasonably good and that the person familiar with the appearance of that man shown in that footage during the years between 2012 and 2018 could potentially recognise him from that footage. The risk in error in recognition might have been greater if the witnesses had been shown footage also of Darren Murphy emerging from the gateway of the flats complex in George's Place and this might prompt them to connect the place with the person. But the date appearing on the CCTV footage might have alerted the Garda officer with a very good memory to the fact that this was the day of the murder of Eamon Kelly, we accept that this did not happen in the case of any of the Garda officers who gave evidence of recognising Darren Murphy from the footage shown to them at Balbriggan Garda Station. This fortifies the reliability of their evidence of recognition which is offered at the additional level of assurance that it comes from the controls adopted in accordance with the directions of Detective Inspector Scott."

39.  The court then went to link the Temple Street footage with the Howth road footage, saying that the man in the Temple street footage was "in all respects identical in appearance in terms of height, build, clothing, shape of head, cut of hair around the head, ears and footwear to the man who appears in the footage of CCTV cameras showing the domain (sic) on Howth Road". It also observed that "[t]here is good quality CCTV footage of him at the Spar in Temple Street where he's walking the dog and he's accompanied by Sean Connolly." The court then discussed the appellant's association with the Opel Meriva and commented:

"It's beyond credible responsibility that some doppelgänger of Darren Murphy who looked exactly like him and was dressed in identical clothes as those which he was wearing at Temple Street Children's Hospital and Spar Temple Street was driving the turquoise Opel Meriva around Killester area and getting in and out of it and walking around at the domain and Howth Road on the afternoon of the murder. Even if this Court were to discount all of the evidence of the gardaí who stated that they recognised Darren Murphy on the Force 10 footage at the domain, the only credible conclusion open is that the footage at Temple Street locations and at the domain is of the same man. This is because of the very obvious sameness of the person whom the footage is analysed. This conclusion is supported by the CCTV evidence of 4th of December 2012 in Temple Street and Eccles Street areas of Dublin which connect Darren Murphy to Sean Connolly and to the getaway driver of the Lexus. These two men and the Lexus are also circumstantially linked by evidence to the turquoise Opel Meriva as is Darren Murphy."

40.              The court went on to discuss those circumstantial links in further, one might say meticulous, detail, paying close attention to the details of each piece of evidence. Ultimately it was satisfied that the evidence was consistent only with the appellant's participation in a murder scheme and that the events on the footage concerning the appellant and the car associated with him could not be "cannot be explained away as coincidences".

The Notice of Appeal

41.              By Notice of Appeal filed on the 2nd of February 2022, the Appellant set out 27 grounds of appeal against conviction. Those which were pursued may be broadly grouped into three categories:

                                                              i.      Issues relating to recognition and identification;

                                                            ii.      Common Design/Joint Enterprise;

                                                          iii.      Discrepancies within the montage of CCTV evidence

The First issue: Identification

The appellant's submissions

42.              The appellant submits that the recognition evidence of the seven Gardaí who identified the appellant from the Temple Street footage should not have been admitted as evidence. He submits in the alternative that if it was correct to admit the evidence, the trial court afforded too much weight to the evidence. It was weak and should have had minimal impact on the ultimate finding of the court, whereas the trial court in fact used the identification evidence as the "keystone" of its finding of guilt.

43.              The appellant submits that the identification evidence presents multiple problems and was not "forensically sterile" as claimed by the prosecution. He questions the credibility of the identifying Gardaí when they said they had not made the connection between the appellant and Eamon Kelly murder, despite having earlier accessed Pulse records concerning both. He notes the five-and-a- half years between the murder and viewings, and he observes that twenty-five other Gardaí did not identify the appellant from the same footage. He says that the Gardaí who identified Darren Murphy as the man in the CCTV clips and stills were not disinterested members of the public giving evidence of recognising an acquaintance from an image but rather serving police officers whose business it is to educate themselves regarding any person who may be living in their police area who is suspected of having committed serious criminal offences. He points to the fact that the murder date was stamped on the footage.  The appellant submits that upon being asked to identify a person from images overprinted with the date of a notorious murder, the educated viewer could not fail to realise the images have something to do with that murder. The appellant submits that insofar as the court dismissed the defence concerns concerning the date stamp, it did so on the basis of speculation to the effect that no Garda in fact remembered the date in this case.

44.              The appellant submits that the following further safeguards should have been employed: (i) the witnesses should have been asked not to proceed to a conclusion until they had viewed the footage at least twice; (ii) they should have been asked did they recognise the location; (iii) they should have been asked to include reasons for the recognitions such as features in the subject that aided the recognition.

45.              The appellant further complains that although viewings were recorded, the positioning of cameras was such that the images being shown on the laptop could not be viewed by the appellant and his legal team.

46.              The appellant also contends that a "worrying" aspect of the process was that the court itself engaged in a dock identification despite never having been asked to do so by the prosecution. Although the trial court said that it was not engaging in that exercise, it was in fact doing precisely that. The appellant does not know what similarities or dissimilarities were taken into account, nor was any warning given to the defendant that his presence in court could be used in identifying him as the person in the footage. He cites People (DPP) v. Gruchacz [2019] IESC 45 and says that while identification by the jury of an accused person in court was not something the prosecution must notify to the accused before his trial, the distinction here is that the prosecution did not even invite the court to enter upon its own identification exercise. Further, the defence team had no opportunity to address the court on the issue. He also refers to the comment in Gruchacz by O'Malley J that a trial judge should give the jury a strong warning that they were liable to the same weaknesses as an eyewitness.

47.              The appellant also submits that there were timing discrepancies within the overall prosecution montage of CCTV clips and stills which the prosecution relied upon to connect the appellant with the murder. Various examples of discrepancies are set out in the written submissions but were not elaborated upon in oral submissions.

48.              As to the risks inherent in visual identification, the appellant referenced People (DPP) v. Mekonnen [2011] IECCA 74; [2012] 1 IR 210 while recognising that recognition is treated with slightly less caution. He cites People (DPP) v. Stafford [1983] IR 165 for the proposition that the jury should be reminded that mistakes in recognition of relatives and friends has sometimes been made, and people People (DPP) v. Smith [1999] 2 ILRM 161 where it was said that while there is a distinction in degree between recognition and visual identification evidence of a person not previously known to the witness, this does not justify abandoning or neglecting the rules in People (Attorney General) v Casey (No 2) [1963] IR 33.

49.              As to the procedures necessary in respect of pre-trial identification of a suspect, the appellant cites R v. Smith (Dean Martin) [2008] EWCA Crim 1342; [2009] Cr App R 36 for the proposition that it is necessary to ensure a record of the initial reaction of the witness and the basis on which the recognition occurred; People (DPP) v Maguire [1995] 2 IR 286 for the necessity for a witness to indicate the capacity in which and how well they knew the accused and what feature(s) identifies the accused with the person in the image. The appellant acknowledges the decisions in People (DPP) v. Kirwan [2015] IECA 228, People (DPP) v. Tynan [2017] IECA 202 and People (DPP) v Power [2019] IECA 74 do not exclude that members of the Garda Síochána may in principle give recognition evidence from CCTV footage, and that it may be appropriate to admit recognition evidence in the absence of contemporaneous notes. However, he submits that the factual issue here is that none of the gardaí accepted that they knew the appellant was a suspect to the murder prior to the identification, despite the evidence regarding the access to Pulse Records, and that this claim undermined their credibility. With the exception of Garda Deasy, each of the identifying gardaí had demonstrable background knowledge which would connect the appellant to the Eamon Kelly murder.

50.              The appellant cites People (DPP) v Lawless & Touhy (27 June 2019, Portlaoise Circuit Court, Keenan Johnson J) where the Circuit Court expressed concerns about the absence of a paper trail and suggested certain procedures in identification and recognition cases that might be followed.

51.              The appellant refers to Code D of the Police and Criminal Evidence Act 1984, as amended, in the United Kingdom and submits that the procedures therein are more robust than has been required by case law in Ireland thus far and are more in line with the requirements of the appellant's fundamental right to a fair trial under Article 38.1 of our Constitution.

Respondent's Submissions

52.              The Director refers to the complaint made in the appellant's written submissions that CCTV of a single individual did not amount to an identification parade and that no foils were presented. She describes this as "a somewhat unusual contention" given the qualitative difference between the two types of identification process. The use of foils in an identification parade is to ensure that the person is not simply pointing to the only person present with a particular look. With CCTV footage, the identifier is being asked to identify and name a person out of all the people they know in the world. The procedures are so completely different as to be incapable of true comparison.

53.              As to the complaint that the footage was overprinted with the date, the Director submits that no context was provided to any of the identifying gardaí in relation to the footage and that the footage did not relate to the locus of the murder itself.

54.              The Director refers to paras 31-33 of Kirwan where the Court discussed the absence of contemporaneous records and said that while the records ought to have been kept, the significance of the absence had to be seen against the fact that the relevant garda had indicated that he made his identification on careful analysis and having watched it a number of times.

55.              The prosecution draws a contrast between the present case and the facts in Kirwan, where the Court upheld the conviction despite the fact that the identification(s) took place in an incident room with the appellant's name visible, simultaneous viewings, and no records of the process. Even then, the Court upheld conviction. Here the process had a number of safeguards in the process including the exercise involving a "double-blind". The Director submits that the defence could cross-examine the gardaí without any fear of prejudice to the fact-finding process because there was no jury, and they were provided with significant disclosure material (the Pulse access information) by way of a spreadsheet. The information on the spreadsheet was collated over the weekend during the trial but qualitative information had not been  sought in relation to matters such as how long the identifiers knew the appellant or how regularly they had interacted with him.

56.              The prosecution cites People (DPP) v. Power [2019] IECA 74 where the Court (judgment delivered by McCarthy J.) said that the principles regarding recognition were as set out in People (DPP) v. Larkin [2008] IECCA 138; [2009] 2 IR 381.

57.              The Director contends that the appellant's suggestion that the trial court engaged in its own identification relies on two lines in the judgment which have been taken out of context and therefore have "the capacity to grossly misrepresent the analysis that the Court actually engaged in." The passage when read as a whole makes it clear that the court was not substituting its own view of the identity of the person in the CCTV but rather that this was an additional factor which allowed it to be satisfied of the reliability of the recognition evidence. They were simply saying that there were no obvious inconsistencies, and no more than that.

58.              The Director also submits that the CCTV was of high quality and an exhibit in the case, and the appellant clearly visible, at times with his distinctive Kerry blue dog. She cites Gruchacz for the proposition that there is nothing in principle impermissible in a fact-finder making a decision as to whether an accused is identifiable in the footage and cites paras 98-100 thereof in particular.

59.              The Director submits that it was clear from the evidence that the timings on the stills were not to be regarded as definitive and/or may have related to the end of the clip of CCTV footage as opposed to the start time of same. However, what was relevant was the continuity of the sequence with one camera leading onto another which permitted the sequencing of the stills as outlined to the Court. Further it was the totality of the evidence which allowed the Court to be satisfied of the journeys embarked upon by the vehicles in question and their proximity to each other on the relevant occasions. None of the alleged discrepancies related to the key pieces of footage of the green Opel Meriva in convoy with the black Lexus before or after the murder or its positioning on The Demesne with the clear view of the Howth Road and the perambulating Eamon Kelly at approximately 4:00 pm on the 4th December 2012.

Discussion and conclusion in relation to the first issue: the recognition/identification evidence

 

Applicable legal principles

60.              Recognition evidence and identification evidence are closely related, but in the present case we are primarily concerned with recognition evidence. What is normally referred to by lawyers as "identification" is where X sees an individual committing an offence who is unknown to X, and later picks out Y as the person X saw committing the offence. The identification is usually conducted by means of identification parade but there are other methods (including the use of photographs), with greater or lesser risks of misidentification. "Recognition" is where X, who already knows Y to a greater or lesser degree (e.g. neighbour, local, friend), views an image of a person committing a crime (for example, on CCTV), and makes the connection between that person and Y.

61.              The processes are different but related. Let us contrast two scenarios. The first is where X fleetingly sees the face of a person who robs him in a dark alleyway, and picks out Y from an identity parade some months later as being the robber. The second scenario is where X knows Y well for many years and picks Y out from footage of a crime being committed which he is shown in a Garda station. These are examples taken from opposite ends of the spectrum to illustrate the differences between the two processes, although many cases fall somewhere in-between.

62.               Despite the dissimilarities between the two processes, there are also some similarities  which have led to courts being cautious before accepting not only identification but also recognition evidence. Both processes involve our brains making comparisons between recalled images or sightings and currently viewed images. Our brains are being asked whether we can make a connection between something we saw on a previous occasion(s) and something we are now being shown: there are always two dates. There is always and inevitably a lapse of time between the two dates in question. Relevant matters which affect the reliability of what the brain sees on each of those two dates are numerous. They include matters such as -

·         The length of exposure to the image or person on either date (Was it a fleeting crime lasting seconds or a lengthy experience? Was there a single viewing of a CCTV image or two viewings? Had the neighbour just moved in next door or had he been there for years?)

·         The length of time between the two dates (When did you last see the local you picked out on the footage? What is the time lapse between the robbery and the identification parade?)

·         The opportunity to view (was it bright or dark on the occasion of the robbery? How good is the quality of the CCTV footage?)

·          Were there physical obstructions preventing a clear view (such as parked cars blocking  the view, the attacker being disguised and so on)?

·         The witness' ability to make facial recognitions in general (which is highly personal; some people are exceptionally good, most are average, and some are exceptionally poor);

·         Other matters personal to the witness such as their eyesight, their state of concentration (Was the driver distracted by children in the back of the car when the robber ran across the street?), or whether the experience was traumatic (Were you concentrating on the gun facing you at the time rather than the gunman's face)?

·         Whether the identified person's appearance has physically changed between the two dates (beard, hairstyle, age etc.).

·         Potentially contaminating factors, such as whether the recognition/identifying witnesses have been tipped off or given hints (innocently or deliberately) that a particular person appearing in an identification parade, in the line of photographs, or in CCTV footage, is the suspect.

The list could be extended indefinitely. We draw attention to the above merely to point out that while recognition does differ from identification in an important way (that the process involves picking out a person already known to the witness), there are many factors which can influence the reliability of recognition evidence just they can influence identification evidence.

63.              The question of recognising an individual from CCTV footage has been considered in a number of authorities. A recurring theme has been whether and when a garda witness may offer such evidence to a jury.  In principle, it is permissible for a Garda witness to give evidence that he or she recognised a person on footage as the accused person, but this is subject to the condition that the evidence must be more probative than prejudicial. In People (DPP) v. Larkin [2008] IECCA 138; [2009] 2 IR 381, the Court of Criminal Appeal (judgment delivered by Kearns J) reviewed the previous authorities and said:

"If identification evidence is available from police officers and the same can be given in circumstances where the probative value of the evidence outweighs the prejudicial effect, the court sees no reason why such evidence should not be given. It is difficult to conceive of a greater affront to the community's interest in the prosecution of crime than to deny the prosecution the opportunity of calling such evidence ... That is not to say that an endless stream of police witnesses should be paraded through court, each of whom would successfully provide recognition evidence of an accused person from a video or photograph. As the cases indicate, there is a balancing exercise to be performed in this sort of situation by the trial judge. It is important that witnesses who are called should, wherever possible, be able to point to some non-criminal background context whereby the identification or recognition has been made."

64.              In the present case the appellant does not complain of the fact that the identifying witnesses were gardaí who previously knew the appellant. This is hardly surprising in circumstances where the fact-finder was the Special Criminal Court and not a jury, and where the interactions between those gardaí and the appellant were relatively innocuous (seeing him out in the area with his dog, signing on at Mountjoy station in relation to other matters, and so on). Instead, the appellant contends that various aspects of the process by which the identifications were made were unsatisfactory, to which we will return. Before doing so, it is worth noting that the Court has on a number of occasions acknowledged that the nature of the exercise of recognising a person from CCTV footage is qualitatively different from other identification exercises, such as conducting an identity parade, or presenting a number of Garda photographs of suspects, for example.

65.              In this regard we note that in People (DPP) v Kirwan [2015] IECA 228, the Court (judgment delivered by Birmingham J. as he then was) contrasted recognition evidence from CCTV footage with other exercises of identification, such as identifying a person from 'mug shots' held by the Gardaí, saying at para 28:-

"The present situation is quite different. The CCTV footage that was available to be viewed showed the culprits at or near the crime scene. Whether any of the culprits are identifiable or recognisable at the crime scene is enormously significant. It is not simply a question of identifying a suspect or offering confirmation in relation to a person already nominated. The question is whether the suspect is actually shown on the CCTV at the crime scene at the moment when the crime was committed."

66.              The question of appropriate procedures in the context of an exercise in CCTV-recognition/identification has also been considered in the authorities. One of the grounds of challenge in Kirwan was that there had been a failure to keep contemporaneous notes in relation to the single Garda identification from CCTV footage, as well as in relation the other gardaí who had looked at the footage but failed to make an identification. The Court did not consider this fatal to the admissibility of evidence. It also commented:-

"...[T]he fact that individuals failed to make an identification is of limited relevance. There might be any number of reasons why an identification was not made. The gardaí involved may not have known Mr. Kirwan or may never have had dealings with him or they may have known him, but despite knowing him did not feel confident to make an identification based on the extent and quality of the footage that was available. The defence's interest in all of this was that it established that Garda Culhane was alone in purporting to make an identification, and that he was in that sense an outlier."

67.              In People (DPP) v Tynan [2017] IECA 202, the Court again described the qualitatively different nature of the exercise from other identification exercises and said at para 12:

"It must be appreciated that what occurred in the present case differs from what is usually at issue in cases involving visual identification or indeed voice identification. What ordinarily happens is that a witness, at a time after a crime was committed is asked to view an identification parade or view a number of people at an identification opportunity with a view to seeing whether they can pick out someone and say that the person they are viewing was the same person as they saw commit the crime on the earlier occasion. Here, what is happening is quite different. Here, there is footage actually showing the crime being committed in the sense that it shows those involved in the incident present at the crime scene and the person viewing the footage is doing so in order to see whether he can identify anyone on screen as someone who was previously known to him. This is a fundamental difference and it may mean that procedures suitable for a visual identification parade or voice recognition opportunity may not be readily transferable."

68.              In the same case, in relation to the complaint that there was a failure to keep contemporaneous notes, the Court said at para 10:

"There was no statutory obligation to take contemporaneous notes and the Court does not feel that there was any obligation on the trial judge to exclude the evidence by reason of the absence of notes. Those who had participated in the viewing that was organised at Kevin Street Garda Station were in a position to give their evidence as to what occurred and did in fact give their accounts."

69.              The above constitutes the context in which the Court must consider the appellant's complaints as to the process adopted with regard to the identification exercise in the present case. Having regard to the appellant's submissions, those complaints may be divided into two categories; (1) complaints relating to the procedures adopted, and (2) complaints relating to the credibility of individual gardaí in connection with the identification/recognition exercise.

Analysis of the procedures adopted

70.              The Court is of the view that the procedures adopted in the present case were excellent in many respects, including the fact that the members in charge of the exercise did not themselves know what the investigation in question was; the fact that the individual members conducted the exercise in isolation from other members; the fact that the introductory words inviting the members to look at the CCTV footage were entirely neutral and not in any way 'leading'; that statements were taken from the individual gardaí after they had viewed the footage (with follow-up in one case for clarification by way of a second statement); and that the whole process was video-recorded. Indeed, it might be said that the failure of numerous gardaí (twenty-five out of thirty-two) to identify the appellant tends to support the conclusion that the process was not pre-loaded in some way that would tilt a garda towards a positive identification of the appellant. If it had been so pre-loaded, one might have expected that far more gardaí would have identified the appellant.

71.              As a matter of procedure, we do not consider that the presence of the date stamp of the murder date on the footage (which was presumably embedded in the footage and could not be removed) was, in and of itself, a matter which would have tipped off the identifying gardaí as to what investigation was in issue. The impact of such a date stamp might vary from case to case; for example, if there were a notorious event, such as a bombing, or the recent Dublin city protests, where the date had entered public consciousness as being associated with that event, the position might be different. It might also be different if a notorious event had taken place recently. However, in the present case, the murder of Eamon Kelly had taken place some six years before the exercise was conducted and none of the identifying gardaí were involved in the investigation. In those circumstances, we do not consider that the presence of the date stamp on the footage in and of itself introduced an inherently unfair or prejudicial element into the recognition/identification exercise.

72.              The appellant  points to other issues, such as the fact that the angle of the video-recording did not permit of seeing precisely what was on screen at the time the member was making a positive identification; that the witnesses were not asked to proceed to a conclusion in viewing the CCTV until they had viewed the footage at least twice; and that they were not asked to include reasons for the recognitions such as features in the subject that aided the recognition. He contends that these omissions were problematic and should have affected both the admissibility and weight of the evidence.

73.              The test to be applied to the Court in terms of the admissibility of the evidence, in terms of the procedural aspects of the exercise, is whether, when taken in the round, the procedures were sufficiently robust to ensure a fair process or whether they fell below a threshold such that the evidence could be considered more probative than prejudicial. The Court is of the view that taking all of the matters into account, the procedures did not fall below the threshold of safety such that the evidence should have been excluded as inadmissible.

74.              It bears repeating that the procedures set out in the UK Code D of PACE do not apply in this jurisdiction. While the procedures set out in the Code may well represent best practice, and it might be a counsel of perfection to follow all of them despite the fact that they have no legal status in this jurisdiction, it is certainly not the case that a failure to follow those procedures should necessarily result in the exclusion of the evidence. All depends on the particular facts of the case and whether the procedures are sufficiently fair to admit the evidence, or so unfair that it should be excluded. As we have said above, we consider that the procedures employed in the present case were sufficiently robust to pass the threshold for admissibility.

75.              For the same reasons, the Court does not accept the appellant's submissions that little or no weight should have been given to the identification/recognition evidence by reason of the procedural aspects of the process.

Analysis of the Credibility issue

76.              As noted above, the second category of complaints by the appellant within the 'recognition/identification' issue related to the credibility of the relevant gardaí. This arose in connection with the issue of whether (most of) the gardaí who had made a positive identification of the appellant from the CCTV footage had in fact made the connection between the appellant and the Eamon Kelly murder at the time of the viewing in 2018, even though they had not been told or tipped off about the connection.

77.               It was accepted by the appellant that knowledge of the connection would not by itself necessarily taint the recognition evidence, and indeed this must be correct. If Garda A knows an accused very well from multiple interactions over the years, but also knew that he was a suspect in a particular investigation, this would usually not in and of itself disqualify Garda A's identification/recognition evidence from being admissible; it would go to weight. In any event, what was put forward was something slightly different; not that the existence of knowledge of such a connection was necessarily problematic, but rather that the gardaí lacked credibility when they said they did not know of the connection.

78.              The appellant relied for this proposition on the PULSE records which showed that (most of) the relevant Garda members had on previous occasions accessed PULSE in connection with both the appellant and the Eamon Kelly murder, and that the inquiries in respect of each had, in some cases, been made within minutes of each other on the same day. Each and every relevant garda denied knowing of the connection at the time he or she made the identification. Thus, the trial court faced an issue of whether the gardaí who made the latter claim were credible, particularly those who had previously accessed PULSE in connection with both the appellant and the Eamon Kelly murder. It should be observed, however, that the lapse of time (years) between those PULSE inquiries and the identification date (2018) was highly relevant to the trial court's assessment.

79.              The appellant accepts that credibility determinations are a matter for the trier of fact but contends that the trial court failed to engage with the problems identified with regard to credibility.

80.              We are not satisfied that this is a fair characterisation of the trial court's judgment, which proceeded as follows. The trial court expressly said that it was taking into consideration that some of these witnesses had accessed PULSE records in earlier years in circumstances which showed that when they made these inquiries they knew or were likely to have known from PULSE information and from the nature of their inquiries that Darren Murphy was suspected of having a role in the murder of Eamon Kelly. Thus, the trial court took the position most favourable to the appellant in terms of the evidence on this issue, even though the gardaí never made the concession themselves (apart from one, and it is even debateable as to whether the concession was made by the garda in question).

81.              Nonetheless the court went on to accept the evidence of these witnesses that they were not aware that they had made these PULSE inquiries or made a connection between the appellant and the Eamon Kelly murder at the time when they made their recognition/identifications of the appellant from the CCTV footage in 2018. In other words, the court was drawing a clear distinction between the knowledge of gardaí some years before, and the knowledge of the same gardaí in 2018.

82.              In this regard, the court pointed out that the witnesses were not part of the investigation team for the murder and had encountered the appellant in the course of community policing duties or as station gardaí when he was signing on at garda stations, and that two of the witnesses recalled him walking a dog on a lead.

83.              The court observed that they were "firm" in their identification evidence.

84.              The court noted, as the appellant had pointed out, that the witnesses did not recognise the locations in the Temple Street footage, but observed that the footage from the hospital camera contained nothing which would enable a person who was not told of this or who was not familiar with the spot to recognise the location.

85.              The court also took into account the delay between the murder in 2012 and the exercise carried out at Balbriggan Garda station in 2018.

86.              The court said it was necessary to be cautious in drawing conclusions as to the identity of persons from footage because it is possible to make errors and jump to conclusions which may not be warranted by the footage. The court found it helpful that there were "pointers" within the evidence of witnesses, mentioning witnesses who remembered the appellant walking a dog, and some who remembered where he lived and had an association with the turquoise Meriva car. The court took into account that the movements of the man walking the dog on the day of the murder started and ended in the flats complex at St. George's Place.

87.              The court also accepted that the date stamp on the CCTV footage might have alerted a Garda officer "with a very good memory" that this was the day of the murder of Eamon Kelly, but accepted the evidence of the gardaí this did not happen in fact in the case of any of the garda officers who gave evidence of recognition. The appellant complains that the latter finding was "speculation". We do not agree. It was a finding of fact that the trial court was entitled to make, having observed and heard the witnesses; it was better placed to make a finding of fact on this issue than this Court.

88.              We are of the view that it is clear from the above that the trial court engaged with each and every point raised by the appellant in connection with the recognition/identification process, and was overall highly aware of the need to proceed with caution for those reasons and because of the dangers inherent in identification/recognition evidence. The court was nonetheless persuaded that the witnesses had made firm and untainted identifications when they viewed the footage in 2018. The court had the opportunity to closely observe the witnesses giving evidence, and it is manifest from the transcript that there was not only such close observation but active questioning from the Bench on matters of detail.

89.              We are of the view that the trial court's conclusion was well within the remit of its role in making findings of fact and credibility determinations and that it did indeed "engage" with the matters raised by the defence. Nor were those findings of fact perverse such that an appellate court would be entitled to interfere with them.

90.              Insofar as the appellant now complains that he did not receive adequate information about the PULSE access details concerning the gardaí who failed to make a positive identification, we note that the appellant's legal team did not seek this information/seek adjournment prior to or during the trial. Accordingly, it is not necessary to deal with the question of whether they would be entitled to such information in any event as it is a point raised on appeal which was not raised at the trial. The appellant maintained that an accused person would be entitled to such information because it would be helpful for him, for example, to point out (if it were the case) that a particular garda or gardaí knew the person very well from numerous interactions but had failed to make a positive identification from the footage. We do not propose to express any opinion on this except to note that, as Birmingham P. has said (see above discussion of the authorities) there are many reasons why individuals may not make a positive identification.

91.              To summarise, we do not accept the appellant's submissions to the effect that the court did not engage with the alleged weakness of the recognition/identification evidence. We are of the view that the trial court was not incorrect in admitting the evidence, and further, that it was entitled to place the weight upon it that it ultimately did, despite the credibility issues that had been raised by the appellant. These issues were carefully addressed by the trial court and taken into account before reaching its conclusions.

 

The alleged dock identification

92.              We were not persuaded that the trial court had in fact conducted a 'dock identification' as alleged. The trial court noted in its judgment that it had not been invited to do so and observed that it would be "most reluctant to perform such an exercise", given the delay of over 8 years between the murder and the trial. It pointed out that the appellant's appearance might have changed over the years and it did not know what he looked like in 2012. What the court went on to say after that was simply that there was "no inconsistency between his features as observed in court and those which were visible in the CCTV footage" which would lead the court to doubt the correctness of the recognition evidence or undermine in any way the other circumstantial evidence which supported the conclusion that he was the person shown on the footage. We are not persuaded that this amounted to a dock identification. The trial court was relying upon the Garda recognition/identification evidence and merely making a general observation that from what it could see on the footage from 2012 and the physical appearance of the man before them in 2018, there was no great discrepancy that might cause alarm bells to go off.

The Second Issue: Common Design/Joint Enterprise

The appellant's submissions

93.              The appellant submits that the trial court erred in concluding that the appellant was guilty on the basis of joint enterprise or common design when the evidence led by the prosecution was insufficient to establish this. The appellant cites authorities such as People (DPP) v Cumberton (unreported, Court of Criminal Appeal, 5th December 1994) and People (DPP) v Doohan [2002] 4 IR 463, which decisions approved the following dicta in R v Anderson and Morris [1966] 2 QB 110:

"Where two persons embark on a joint enterprise, each is liable criminally for acts done in pursuit of the joint enterprise, including unusual consequences arising from the execution of the joint enterprise; but if one of them goes beyond what has been tacitly agreed as part of the joint enterprise, the other is not liable for the consequences of the unauthorised act."

94.              The appellant refers to People (DPP) v Dekker [2015] IESC 107; [2017] 2 IR 1 (Judgment delivered by Dunne J.) at para 74:

"[i]t will be relevant in a case of joint enterprise to consider what the participants in the crime had agreed to do, tacitly or expressly. To that extent issues of foresight are very relevant. Thus if an individual participates in an assault which they know is intended to cause serious injury and death occurs, given that the principal would be guilty of murder if death ensured, it seems to me that, having regard to the provisions of s.4 [of the Criminal Justice Act, 1964], it follows as a matter of logic that the accessory should also be guilty of murder, even if the accessory did not realise that the principal actually assaulted the victim with intent to kill."

95.              The appellant also relies on Dekker for the point that there must be a proper evidential foundation for asserting the possibility that the accused foresaw an intent to commit the actual offence. In Dekker, the Court stated at para 76:

"As was pointed out in R v A [2010] EWCA Crim 1622, [2011] QB 841, if an accused knows that a co-accused is carrying a weapon then ordinarily that will mean that they realised (foresaw) that the principal might act with intent to kill or do really serious injury but it will be a matter for a jury to decide whether or not there is a proper evidential basis for asserting the possibility that the accused foresaw an intent to inflict no or minor harm and if, in fact, that was the situation."

96.              The appellant also cites People (DPP) v Cummins & Davy [2021] IECA 198, where the Court concluded there was insufficient evidence from which to infer the existence of an agreement.

97.              The appellant contends that the trial court erred in making inferences that were not justified by the facts actually proven. He contends that there was no evidence that the appellant realised or foresaw that a murder was to take place; that there was no evidence that the appellant had a murderous intent or was aware of a murderous intent; and that there was insufficient evidence to ground the conclusion that the appellant's role was "meeting the gunman Sean Connolly and the driver of the getaway car on the day of the murder and also to facilitate the escape."

98.              He contends that the trial court failed to give adequate consideration to the fact that there was no statistical analysis regarding Darren Murphy's DNA allegedly recovered from the Opel Meriva vehicle.

99.              He also contends that the conclusion that he was guilty of murder on the basis of common design is at odds and inconsistent with the finding of the court that the appellant was not guilty of possession of the firearm used in the murder.

100.          Finally, he contends that the trial court paid insufficient regard to the fact that his co-accused was acquitted and the implication of that for the appellant's case, particularly in circumstances where the trial court considered that the evidence that KD was present in the Opel Meriva with the appellant was not sufficient to prove that KD had a role as a perpetrator of the murder.

The respondent's submissions

101.          The Director agrees that the authorities cited by the appellant set out the relevant principles concerning joint enterprise and common design. As to the application of the relevant principles to the facts, she contends that it was always the prosecution's position that the appellant was not the gunman but rather that his role was that of a "spotter in advance of the murder" and someone who would "render assistance as required in the aftermath". She contends that the trial court was correct in concluding that the evidence established beyond a reasonable doubt that the appellant was one of a "murder gang" who went out to kill the victim on the 4^th of December 2012 and that the evidence admitted of "no other credible explanation for [his] activities on that day consistent with innocence or with some involvement short of intention to participate in a murder".

102.          The Director points out that the court concluded that the evidence in respect of KD admitted of inferences of illegal and discreditable activities and associations in the period leading up to and around and after the death of Eamon Kelly which might not add up to deliberate involvement in murder, but contrasts this with the court's view that there was "no room for such doubt in the case of the evidence relating to Darren Murphy." The court was satisfied that the appellant was using the car as a spotter and that the sightings of this car in the Killester area of Dublin close to the scene of the murder on the afternoon of that day, along with the other vehicles used in the planning and execution of the murder, were consistent only with participation in the murder scheme and could not be explained away as coincidences or as serving any purpose other than as activities related to the commission of the murder and steps taken by the murder gang in its immediate aftermath.

103.          In the written submissions, the Director sets out the individual pieces of evidence upon which the trial court relies in reaching the conclusion it did concerning the appellant.

Discussion and Conclusion on Common Design

104.          There is no dispute between the parties as to the principles to be applied and the Court accepts that they are as set out in the appellant's submissions and summarised above. However, the Court does not accept the arguments of the appellant insofar as it concerns the application of those principles to the facts.

105.          The trial court set out in meticulous detail the individual items of circumstantial evidence concerning the appellant. It clearly distinguished between the evidence which concerned KD and the appellant respectively. It is not necessary to summarise all the evidence here, which occupied a significant portion of the trial court's judgment, and which we have taken into consideration. We agree entirely with the court's conclusion that the evidence as a whole made it clear that this was not a spontaneous endeavour but rather involved careful pre-planning, advance surveillance, practice runs, counter forensic measures and a number of vehicles operating at times in convoy, and that a crucial part of the plan was the "spotting" exercise by the appellant so that the gunman to be in position to ambush Eamon Kelly at the point in time when he was making his way home down the alleyway to Furry Park Road from the Howth Road direction, and to assist the occupants of the Lexus following the burning out of same. Having regard to the totality of the evidence as described in the trial court's detailed judgment, we see no reason to disturb the trial court's conclusion that the evidence pointed beyond a reasonable doubt to the appellant's role in the enterprise.

106.          We do not accept that there is any inconsistency between the verdict of guilty on the murder count and the verdict of not guilty on the possession of firearms count. It is entirely possible for an individual to assist another in carrying out a shooting without in any way having possession of the firearm itself, even having regard to the legal concept of possession which encompasses more than mere physical possession. That is what the trial court concluded in this case and there is no legal inconsistency as between the verdicts in respect of the appellant either on the facts or the law.

107.          Likewise, we do not accept that there is any inconsistency between the verdicts in respect of KD and the appellant. It frequently happens that the evidence in respect of one accused is more compelling than that in respect of another. Indeed, the acquittal of KD is testament to the care and caution exercised by the trial in respect of the evidence. It is manifest from the transcript that the court was attentive to the minute details of the evidence both as it was being given and in its judgment, and we have no hesitation in rejecting this submission.

The Third Issue: The time discrepancies on the footage

108.          The appellant submits that an analysis of the various CCTV clips and stills reveals that, in addition to the usual discrepancies in times recorded on CCTV and real time, the sequence of the clips and stills, although presented as being in order of time, was not always in such order. However, the trial court in its judgment relied on the accuracy of the times in the clips presented in the trial as being the times of travel of various vehicles. In the written submissions delivered on behalf of the appellant, the appellant instances a number of examples of what he describes as problematic footage. He submits the evidence lacked consistency to the extent that the court acting as a jury was obliged to speculate as to the accuracy of timing in order to fill the gaps in timing and the images purporting to show the planning and commission of the murder of Eamon Kelly

109.          In written submission filed on behalf of the Director, it is said that it was clear from the cross-examination and re-examination of the witness in question that the stills were ordered not merely by reference to times but also by reference to the continuity of the footage from one camera to another and, on occasion, the witness was prepared to accept his error in the calculation of time periods. The discrepancies, such as they were, did not undermine the integrity of the analysis of CCTV carried out. The Director observes that thousands of hours of footage were recovered and was available to be analysed by the defence to support any justified substantive attack on the integrity of the analysis. She also points out there were approximately 240 stills in booklets relating to three different dates (1st, 3rd and 4th December 2012) and showed travel over large geographic areas.

110.          The Director then addresses each of the alleged problems identified by the appellant in the written submissions, with reference to detailed evidence from the transcript from the Garda member who had compiled the montage, including his cross-examination.

111.          The Director submits that it was the totality of the evidence which allowed the Court to be satisfied of the journeys embarked upon by the vehicles in question and their proximity to each other on the relevant occasions. She points that none of the alleged discrepancies related to the key pieces of footage of the green Opel Meriva in convoy with the black Lexus before or after the murder or its positioning on The Demesne with the clear view of the Howth Road and the perambulating Eamon Kelly at approximately 4:00 pm on the 4th December 2012.

112.          At the hearing of the appeal, counsel on behalf of the appellant did not respond to the points made by the Director in her written submissions and was content to rest on his written submissions. The Court notes that the appellant at the appeal hearing did not point to any particular or individual piece of footage as being problematic. Nor did he dispute that none of the alleged discrepancies related to the key pieces of footage of the green Opel Meriva in convoy with the black Lexus before or after the murder or its positioning on The Demesne with the clear view of the Howth Road and the perambulating Eamon Kelly at approximately 4:00 pm on the 4th December 2012.

113.          Having considered the written submissions of both parties carefully, the Court does not consider it necessary to set out the lengthy and detailed evidence in relation to particular stills. It suffices to say that, having regard to the particular stills identified, and the evidence concerning those stills, as well as the trial court's judgment concerning the overall movements of persons and vehicles, the Court is not satisfied that the trial court was in any way misled by the discrepancies adverted to by the appellant. This ground of appeal is refused.

Outcome

114.          For all the above reasons, the Court dismisses the appeal.