THE SUPREME COURT
[Appeal No: 54/2016]
Denham C.J.
O’Donnell J.
Clarke J.
Dunne J.
O'Malley J.
The People at the Suit of the Director of Public Prosecutions
Prosecutor/Respondent
Accused/Appellant
Joint Judgment of Mr. Justice Clarke, Ms. Justice Dunne and Ms. Justice O'Malley delivered the 19th July, 2017.
1. Introduction
1.1 There has been a significant growth in the use of forensic evidence in the criminal process over recent decades. This case principally involves issues arising out of the circumstances in which certain DNA evidence was collected and also the extent to which it is possible to establish the identity of the perpetrator of a crime to the criminal standard on the basis of DNA evidence alone. Questions concerning the manner in which such issues should be explained to the jury by the trial judge also arise.
1.2 In circumstances which it will be necessary to outline in greater detail in the course of this judgment, the accused/appellant (“Mr. Wilson”) was convicted of murder. Mr. Wilson appealed to the Court of Criminal Appeal who upheld his conviction (see The People (Director of Public Prosecutions) v. Keith Wilson [2014] IECCA 48). However, the Court of Criminal Appeal acceded to an application, under s.29 of the Courts of Justice Act, 1924 as amended, for a certificate that its decision involved a point of law of exceptional public importance and that it was desirable in the public interest that an appeal should be taken to this Court (see The People (Director of Public Prosecutions) v. Keith Wilson [2016] IECCA 2). In addition, this Court permitted certain further issues to be raised on the appeal. However, in order to understand the precise issues which require to be addressed it is appropriate to set out, in brief terms, those aspects of the background to Mr. Wilson’s trial and conviction which are material to the issues with which this Court is now concerned. We therefore turn first to the background.
2. Background
2.1 On the 14th August, 2010, Mr. Daniel Gaynor was shot dead on a public road in North Dublin by a lone gunman who approached him and fired a handgun a number of times. According to the prosecution evidence in the trial of Mr. Wilson the gunman was seen to be wearing a peaked cap and a white glove. He was also seen to discard items as he ran away and a subsequent search of the area turned up a revolver, a cotton glove, a cream knitted glove, a baseball cap and a hoodie.
2.2 These items were subjected to forensic analysis. Firearm residue was found on the cotton glove, the baseball cap and the front and cuffs of the hoodie. This was of a type described by the relevant forensic witness as providing extremely strong support for the view that these items had been in contact with a source of firearm residue.
2.3 DNA analysis succeeded in generating a full male profile from the cotton glove. Mixed profiles were obtained from the baseball cap and the hoodie, meaning that there was DNA present from more than one person. In both cases the major profile matched the one found on the cotton glove.
2.4 Swabs from the grips of the gun were analysed and a mixed profile was generated. Again, the major profile matched that on the cotton glove.
2.5 Mr. Wilson was arrested on suspicion of the murder in November 2010. It appears that the arrest was grounded upon confidential information rather than on any form of admissible evidence. While he was in Garda custody he refused to cooperate with requests for a bodily sample for the purpose of forensic testing. The Gardaí decided therefore to obtain samples by means of collecting items that came into contact with his mouth while he was in the station. Mr. Wilson was a smoker and a number of cigarette butts discarded by him in the station yard were collected and analysed. Tests revealed a match with the DNA found on the gun, glove, cap and hoodie.
2.6 The prosecution adduced expert evidence concerning DNA which was to the effect that there was only an extremely remote chance of a random match with an unrelated member of the general public. It will be necessary to turn to the evidence in that regard in due course.
2.7 There was no other evidence implicating Mr. Wilson and at his trial for the murder of Mr. Gaynor he challenged both the admissibility and the weight of the DNA evidence. The admissibility issue was concerned with the question whether Mr. Wilson’s constitutional rights were breached by the taking and testing of the cigarette butts, while the application for a direction was based on the argument that the statistical analysis offered by the scientist who carried out the DNA testing meant that the DNA profiles could not be said to have been matched beyond reasonable doubt. Having failed to secure either the exclusion of the evidence or a directed acquittal, Mr. Wilson argued unsuccessfully that the jury should be warned that it would be dangerous to convict on the basis of uncorroborated DNA evidence.
3. The Issues in the Appeal
3.1 The certified question related to the manner in which the samples of Mr. Wilson’s DNA had been obtained and is in the following terms:
“Is evidence of DNA samples taken from cigarette butts used and discarded by the detained person whilst in custody admissible evidence at his trial?”
3.2 In the process of case-management, this Court granted leave to Mr. Wilson, pursuant to the procedure for the inclusion of additional grounds where there is a certificate from the Court of Criminal Appeal, to argue the issues raised by him in relation to the strength of the prosecution case and the charge to the jury. The questions as framed are:
“When the sole evidence against an accused person is DNA evidence, is such sufficient to convict an accused or upon the prosecution case being closed, should a judge withdraw a case from the jury upon an application of the defence that there was no case to answer?
Should it be mandatory for a trial judge to warn a jury of the dangers of convicting an accused in circumstances where the sole evidence against the accused is DNA evidence alone?”
3.3 We turn first to the admissibility issue which principally arises out of an alleged breach of Mr. Wilson’s constitutional right to privacy.
4. The Privacy/Admissibility Issue
(a) The method by which the DNA match was obtained
4.1 Mr. Wilson was arrested on suspicion of the murder of Mr. Gaynor late on the evening of the 7th November, 2010. He was taken to Finglas Garda Station, where he was detained under the provisions of s. 50 of the Criminal Justice Act 2007. This section permits detention for a period of up to seven days for the purpose of the proper investigation of the offence. No issue arises in this appeal in relation to the lawfulness of the arrest or detention.
4.2 A standard notice of his rights as a person in custody was provided and explained to Mr. Wilson as part of the process of receiving him into custody in the station. The notice contained a reference to the taking of personal samples for forensic testing and it was recorded that at that point Mr. Wilson responded by saying that there was “not a chance” of him consenting to the taking of any sample from his body. Later, on the morning of the 8th November, he formally declined to sign a consent form for the taking of such a sample. In the afternoon of that day he refused a further request.
4.3 The evidence of the detective inspector in charge of the investigation was that he was aware that a DNA profile had been generated from the items found near the scene. He was present when Mr. Wilson said that he would not give a sample voluntarily. The detective inspector’s evidence was to the effect that he did not believe that he had a statutory power to use force to take the usual type of sample, which would normally be done by means of a buccal swab (that is, a sample taken from the inside of the person’s cheek). Accordingly, he directed that possession should be taken of any items discarded by Mr. Wilson and that such items should be retained for investigation. In preparation for this the station yard was swept, photographed, and videoed.
4.4 Mr. Wilson smoked a number of cigarettes in the yard during the course of his detention. The cigarette butts discarded by him were collected by the Gardaí and preserved in appropriate fashion for DNA testing. Other items were also retained including a toothbrush, a towel, a tissue and a dental implement used when he was attended for toothache during the course of his custody.
4.5 Three cigarette butts were analysed and a full male DNA profile was generated from each. The three profiles matched each other and matched the profile on the cotton glove, the cap, the hoodie, and the grips of the gun.
4.6 The expert evidence as to the significance of these findings is considered later in this judgment. However, the first issue is whether the evidence was admissible.
(b) The statutory power to take bodily samples
4.7 The statutory regime relating to the taking of and dealings with bodily samples has recently been radically altered with the coming into force of the Criminal Justice (Forensic Evidence and DNA Database System) Act 2014. However, the relevant legislative provision at the time was s. 2 of the Criminal Justice (Forensic Evidence) Act 1990 (as amended by the Criminal Justice (Drug Trafficking) Act 1996, the Criminal Justice Act 2006 and the Criminal Justice Act 2007) (“the 1990 Act”). A conveniently consolidated version is available in O’Sullivan’s Criminal Legislation in Ireland (2nd ed., Bloomsbury Professional) and reads in relevant part as follows:
(1) Subject to the provisions of subsections (4) to (8A) of this section, where a person is in custody under the provisions of section 30 of the Offences Against the State Act, 1939, section 4 of the Criminal Justice Act, 1984, section 2 of the Criminal Justice (Drug Trafficking) Act, 1996 or section 50 of the Criminal Justice Act 2007, a member of the Garda Síochána may take, or cause to be taken, from that person for the purpose of forensic testing all or any of the following samples, namely -
(a) a sample of -
(i) blood,
(ii) pubic hair,
(iii) urine,
(iv) saliva,
(v) hair other than pubic hair,
(vi) a nail,
(vii) any material found under a nail,
(b) a swab from any part of the body including the mouth but not from any other body orifice or a genital region,
(c) a swab from a body orifice, other than the mouth, or a genital region,
(d) a dental impression,
(e) a footprint or similar impression of any part of the person’s body other than a part of his hand
(1A) A reference in subsection (1) of this section to the mouth shall be read as including a reference to the inside of the mouth.
(2) [Omitted.]
(3) [Omitted.]
(4) A sample may be taken under this section only if -
(a) a member of the Garda Síochána not below the rank of superintendent authorises it to be taken, and
(b) in the case of a sample mentioned in subparagraph (i), (ii) or (iii) of paragraph (a) of subsection (1) of this section, or in paragraph (c) or (d) of the said subsection (1), the appropriate consent has been given in writing.
(5) An authorisation to take a sample under this section shall not be given unless the member of the Garda Síochána giving it has reasonable grounds -
(a) for suspecting the involvement of the person from whom the sample is to be taken -
(i) in a case where the person is in custody, in the offence in respect of which he is in custody, or
(ii) [Omitted]
and
(b) for believing that the sample will tend to confirm or disprove the involvement of the person from whom the sample is to be taken in the said offence.
(6) Before a member of the Garda Síochána takes, or causes to be taken, a sample under subsection (1) of this section, or seeks the consent of the person from whom the sample is required to the taking of such a sample, the member shall inform the person -
(a) of the nature of the offence in which it is suspected that that person has been involved,
(b) that an authorisation has been given under subsection (4)(a) of this section and of the grounds on which it has been given, and
(c) that the results of any tests on the sample may be given in evidence in any proceedings.
(7) An authorisation under subsection (4)(a) of this section may be given orally but, if it is given orally, it shall be confirmed in writing as soon as is practicable.
(8) A sample of a kind specified in subparagraph (i) or (ii) of paragraph (a) of subsection (1) of this section or in paragraph (c) of the said subsection (1) may be taken only by a registered medical practitioner and a dental impression may be taken only by a registered dentist or a registered medical practitioner.
(8A) [Omitted]
(9) A person who obstructs or attempts to obstruct any member of the Garda Síochána or any other person acting under the powers conferred by subsection (1) of this section shall be guilty of an offence and shall be liable on summary conviction to a fine not exceeding €3,000 or to imprisonment for a term not exceeding 12 months or both.
(10) [omitted]
(11) The powers conferred by this section are without prejudice to any other powers exercisable by a member of the Garda Síochána.
4.8 Section 4 of the 1990 Act provided for the mandatory destruction of samples taken pursuant to s. 2 if proceedings were not instituted against the person, or were discontinued, or if he or she was acquitted.
4.9 It is clear that, under the 1990 Act, samples of blood, pubic hair or urine, or swabs from a body orifice (not including the mouth) or a genital region, could only be taken with the appropriate authorisation of a senior officer and with the consent of the individual in question. However consent was not required in the case of a swab from the mouth.
4.10 It is, therefore, equally clear that, although the evidence was that the investigating Gardaí were not conscious of it, they had a statutory power to take a buccal swab without the consent of Mr. Wilson. In such circumstances, it would appear implicit that there would be an entitlement to use reasonable force and that Mr. Wilson would have committed a criminal offence if he had obstructed them.
4.11 Section 7 of the Criminal Justice Act 2006 may also be of some significance. It provides in relevant part as follows:
(1) Where a member of the Garda Síochána who is in -
(a) a public place, or
(b) any other place under a power of entry authorised by law or to which or in which he or she was expressly or impliedly invited or permitted to be,
finds or comes into possession of any thing, and he or she has reasonable grounds for believing that it is evidence of, or relating to, the commission of an arrestable offence, he or she may seize and retain the thing for use as evidence in any criminal proceedings for such period from the date of seizure as is reasonable or, if proceedings are commenced in which the thing so seized is required for use in evidence, until the conclusion of the proceedings…
(c) The judgment of the Court of Criminal Appeal
4.12 In both its substantive judgment on the appeal and in the ruling on the application for a certificate, the Court of Criminal Appeal expressed concern for the privacy rights of persons held in investigative detention. It noted and agreed with the judgment in Director of Public Prosecutions v Scanlon (an ex tempore decision of that court, considered below, given on the 20th April, 2007,) where it was said that everything happening in a Garda station should be scrutinised carefully and that evidence should not be obtained by “a stratagem, trickery or deception”.
4.13 The ratio of the decision of the Court of Criminal Appeal on the admissibility issue in this case can be found at paragraphs 81 to 83 of the judgment, which are set out here in full:
“81. As stated, this Court recognises the gross restrictions on personal liberty imposed upon incarcerated persons. Those restrictions amount to more than mere confinement to a particular place or restraints on the detainee’s freedom of movement. Persons so detained are in fact entirely dependent on the gardaí for necessaries such as food and refreshments, utensils to eat with, furniture to sit on, bedding to rest upon, and washing and toilet facilities. Into this category would fall other items that might be deemed essential to an individual detainee: for example, an inhaler for an asthmatic person or a syringe to administer insulin to a diabetic. The same could also be said of any medical or, as in this case, dental equipment or supplies which were used to treat the detainee whilst in custody. These and all similarly considered items, as determined by reference to the reasons herein given, can conveniently be referred to as ‘necessaries’, being items essential to the realisation or preservation of the constitutional rights of a detainee whilst in custody, as distinct from ‘non-essential’ items, which do not bear upon the constitutional rights of such detainee.
82. The making available of such items is essential to life, to bodily integrity, and to privacy and dignity, to name but some of the constitutional values involved, which, subject to the overriding restriction on liberty, continue to apply to detained persons and manifestly must be respected. Further, such are basic requirements of the humane way in which the State must treat people in custody. This Court is therefore acutely concerned with the potential use for the purpose of obtaining a DNA sample of any such item(s), properly described as ‘necessaries’. Consequently, it would raise a major concern for the court if the DPP should have attempted to use any such material of the type described as a basis for obtaining a DNA profile, whether or not the 1990 Act was invoked: but in particular where its provisions had not been seen through to finality. Any such item, which the gardaí, prison or detaining authorities, as the case may be, were obliged to provide to the incarcerated individual pursuant to the requirement to uphold his/her constitutional rights while in custody, should not be available for the purpose of creating a DNA profile, unless the detainee consents to same or unless there is some other identifiable basis in law, consistent with due respect for such rights, by which such an item may be lawfully used for that purpose.
83. By contrast, the cigarette butts at issue in the instant case can be distinguished from those items which can be deemed necessary to the preservation of such constitutional rights. It is undoubtedly the case that detainees continue to enjoy all of their legal rights, except to the extent that those rights have been removed, curtailed or otherwise circumscribed by legislation. However, whilst a smoker may be inconvenienced, or even made uncomfortable, by an absence of cigarettes, the provision (or non-provision) of cigarettes does not engage the constitutional or legal rights of the detainee. A detainee is no more entitled to have a cigarette than he is to have alcohol brought to him from a local establishment.”
(d) Submissions in this appeal
4.14 Before embarking on a consideration of the submissions and authorities, it is helpful to bear in mind that the purpose of DNA testing in the context of criminal investigation is almost invariably concerned with comparing a profile found in biological material associated with the crime to the profile of a suspect. The DNA profile is derived from examination of a small number of specified locations in the non-coding regions of DNA molecules and not any of the rest of the 3.3 billion pieces of code to be found there. The important point about this is that while the original biological material contains a very large amount of information about the individual from whom it came, a DNA profile does not. As a tool in criminal investigation, its main purpose is to establish a match between a suspect and biological material associated with the crime.
4.15 Mr. Wilson relies in the first instance on the reasoning of the Court of Criminal Appeal in The People (Director of Public Prosecutions) v Scanlon. The issue there was the admissibility of DNA evidence obtained from a piece of chewing gum discarded by the accused in the Garda Station. It was argued on behalf of the accused that samples could only be taken from a suspect pursuant to the statutory regime and that the manner in which the prosecution had obtained this evidence was unfair. According to the approved note of the judgment, Finnegan J. said on behalf of the Court:
“The court is ever mindful that anything which occurs in a Garda Station should be scrutinised carefully in order to ensure that there is no interference with the rights of a person in custody. But having applied that stringent degree of scrutiny we can we can see nothing wrong with the manner in which the Gardaí acted. There was no trick. There was no stratagem. There was no deception. There was nothing oppressive in the manner in which this sample came into the possession of the Gardaí.”
4.16 In the instant case it is argued on behalf of Mr. Wilson that, in circumstances where he had made it clear that he was not prepared to give a sample, he was subjected to “a clear and orchestrated stratagem” which involved, in effect, deception on the part of the Gardaí, if not a trick, to obtain items which might be discarded by him. He had exercised his right to “free will” to protect his constitutional right to privacy, bodily integrity and against self-incrimination, and the Gardaí had put in place a stratagem to circumvent that choice. The only legitimate option open to them in the circumstances was to prosecute him for obstruction.
4.17 It is submitted that by taking the evidence outside of the statutory regime the Gardaí were in a position to disregard the statutory safeguards in respect of data retention and thereby gave themselves an indiscriminate power to maintain an unregulated suspect database.
(e) Discussion and conclusion on the issue of admissibility
4.22 The constitutional right to privacy is well-established and it is not controversial to hold that it encompasses the intimate information about an individual contained in DNA. It is equally well-established that the right to privacy is not absolute and may be outweighed by the exigencies of the public good (see, for example, the judgment of this Court in Haughey v Moriarty [1999] 3 IR 1.)
4.23 The law has always permitted a degree of invasion of the right to privacy where it comes into conflict with the compelling public interest that exists in relation to the investigation of serious crimes. Thus, a person suspected of having committed a serious offence may be arrested and detained in accordance with law for the purpose of the proper investigation of that offence. As the Court of Criminal Appeal observed, such detention of itself necessarily involves a significant invasion of privacy and personal autonomy as well as the right to liberty. Similarly, things found in the possession of the person at the time of arrest may be seized and examined for potential evidence. A search warrant may authorise incursion into private premises and the seizure and examination of articles or documents that are undeniably private. These powers are conferred in the public interest but are of course subject to the requirement that they be exercised in accordance with law and in a manner proportionate to the purpose for which they are conferred. Thus, an arrest and extended detention for an ulterior motive, not being the proper investigation of a serious crime, would amount to a breach of constitutional rights.
4.24 Mr. Wilson was in lawful custody for the purpose of the proper investigation of the offence of which he was suspected. In that context, the 1990 Act conferred a power on investigating Gardaí to take bodily samples and have them analysed, in circumstances where the statutorily-prescribed form of authorisation was given for the purposes set out in s. 2. The statutory power is not under challenge in this case and must be presumed to represent a legitimate curtailment of the right to privacy. The situation, therefore, was that the Gardaí were entitled to take a swab from Mr. Wilson’s mouth and he was not entitled, under the 1990 Act, to refuse. Had they attempted to take one by the use of reasonable force and he had resisted, that would have constituted a criminal offence on his part.
4.25 Counsel has argued that, in refusing to cooperate, Mr. Wilson was exercising a right to protect his constitutional rights to privacy and bodily integrity and his privilege against self-incrimination. His conduct might give rise to a right on the part of the authorities to prosecute him, but not to obtain a sample without his consent except with the use of force.
4.26 That analysis does not accord with the statute (which requires consent in some instances and not in others) and cannot, in our view, be correct as a matter of principle. To begin with, if a person has a constitutional right to refuse to cooperate with a particular process, then legislation criminalising a refusal would itself be unconstitutional. The fact is that Mr. Wilson had no legal right, whatever about his “right to free will”, to obstruct the taking of a sample.
4.27 As it happens, by reason of the procedure ultimately adopted by the Gardaí, matters do not seem to have come to a point where he could be said to have committed an act of obstruction. The question, then, is whether that procedure breached any right of Mr. Wilson’s.
4.28 One path towards a conclusion that a detainee’s rights are breached in circumstances such as this is Mr. Wilson’s argument that the DNA material was obtained by a stratagem involving deception or trickery. There is no clear Irish authority as to the circumstances in which deception of a suspect could bring the exclusionary principle into play. One might hazard the speculation that, for example, deception which misleads a suspect as to his or her legal rights might do so. However, this is not an appropriate case in which to embark upon the discussion since there was, as a matter of fact, no deception or trickery on the part of the Gardaí. Firstly, the mere fact that a particular course of action was planned in advance does not per se make it illegitimate. Secondly, it was never represented to Mr. Wilson that his refusal to cooperate with the taking of a sample from his body meant that no other means of taking biological material would be adopted, or that cigarettes or other items handled by him would not be examined. In any event, one would have to ask what the result would have been if the Gardaí had clearly told him that such items would be tested. After all, he could not have exercised a “choice” not to shed cells containing DNA in the station. It is an aspect of human biology that humans shed cells containing DNA. Cells are rubbed off and deposited in clothing, or fall off the skin onto a chair. Saliva may be deposited on drinking vessels or (as in this case) cigarettes or by licking an envelope. A hand holding any item may leave DNA in sweat. These are not matters of choice and do not depend on the freedom or detention of the individual. It may be that persons carrying out a crime can protect themselves from depositing DNA for some relatively short period of time by wearing and then destroying protective clothing, but it is not a viable prospect for any greater length of time.
4.29 This leads to what is really the core argument made by Mr. Wilson grounded on the proposition that, because he was a person in detention, his privacy rights required greater protection than if he had been at liberty. Thus, it is accepted by counsel that if the Gardaí see a suspect drop a cigarette butt in the street they are entitled to seize it and analyse it; or if they see him drink a cup of coffee in a café they may take the cup and test it. That, it is said, follows from the fact that a person at liberty has the freedom to make choices and to protect himself. By contrast, a detained person has no option but to be present in the Garda station and to use facilities and items provided for his use.
4.30 We accept that a person in detention is, for a variety of reasons, in a particularly vulnerable position. Apart from other potential issues, the Court of Criminal Appeal was correct in pointing out that he or she will be dependent on the Gardaí for the supply of all necessary things and facilities for the duration of the detention. It is partly for that reason that the courts will give extra scrutiny to events occurring during detention, to ensure that the rights of a vulnerable person are not breached.
4.31 However, the requirement for enhanced scrutiny, and careful examination of the question whether a right has been breached, cannot create a situation where the actual substantive content of the right differs according to whether a person is at liberty or in custody. It cannot be that a person who is in custody for the purpose of investigation has a more extensive privacy protection than a person at liberty. The issue has to be the same in both cases - was the constitutional right to privacy breached by the manner in which the sample of biological material was obtained and analysed? The answer cannot vary simply because one is in custody.
4.32 The first part of the question is easily answered in this case. The Gardaí must be entitled to pick up items discarded by persons in detention in a Garda station in the same way that they would in a more public place. There is no question of any property right being interfered with. The cigarette butts were therefore lawfully in the hands of the Gardaí. We would accept that, while he had relinquished all interest in the physical cigarette butts, Mr. Wilson continued to retain a privacy interest in the information contained in the DNA deposited on them. However his rights in this regards were, as already noted, subject to the public interest in the proper investigation of the offence. The lawfulness of carrying out an analysis of the DNA material therefore depended on whether it was properly related to that objective. In our view the generation of a DNA profile, for the purpose of discovering whether or not it matched the profile associated with the crime, was a justifiable intrusion into Mr. Wilson’s privacy.
4.33 The argument made on behalf of Mr. Wilson (and to a certain extent accepted in the Court of Criminal Appeal) would lead to the conclusion that a person in custody could exercise a veto on the investigation of DNA material, unless the sample was taken by the use of force in exercise of the powers under the 1990 Act. Under the analysis adopted by the Court of Criminal Appeal that veto could arise in any event in the case of “necessary” items handled by the prisoner.
4.34 In our view this analysis is misconceived and fails to take sufficient account of the relevance of the unchallenged legislation conferring on the Gardaí the power to take bodily samples, if necessary by force. The statute, in the terms in which it conferred that power, must be taken as having embodied the policy of the legislature in respect of the parameters of the privacy rights of an individual in DNA material. It restricted the taking of bodily samples to cases where an authorisation was given for stated reasons and the material was taken only for the purposes of the investigation in question. That proportionate invasion of privacy was justified by the need to investigate serious crime, and the legal obligation on the detained person was to refrain from obstructing it.
4.35 Where the detained person indicated an intention to resist the taking of the sample, and the Gardaí had an alternative source lawfully in their hands, it would not accord with principle to elevate the privacy rights of the person in custody - the whole purpose of which is investigatory - beyond those of either a person who complied with the statutory regime or a person at liberty. Since it is accepted that the latter would have no cause for complaint if his cigarette butts were picked up in a public place, or in an authorised search of his premises, it is impossible to hold that the rights of the detained person are breached by the same procedure in respect of things that he discards. Equally, it would clearly be contrary to public policy to hold that the Gardaí were in the circumstances constrained to use force, thereby risking injury to both the suspect and themselves, and that a failure to use force rendered the picking up of the discarded items unlawful.
4.36 In circumstances where the biological material is obtained without recourse to the statutory power to take it by force, the right to privacy would however arguably entail a right that it not be tested or used for purposes outside the investigation of the offence. That question does not however arise in this case, since there is no evidence that the sample was subjected to any process beyond the generation of the DNA profile or that the Gardaí retained it in an unauthorised databank.
4.37 We would therefore reject the ground of appeal based on the contended for inadmissibility of the DNA evidence due to breach of the constitutional right to privacy. We next turn to the issue concerning whether DNA evidence alone can be sufficient to establish identity to the criminal standard.
5. Can DNA and Statistical Evidence be sufficient?
5.1 This second issue arises from a contention put forward on behalf of Mr. Wilson to the effect that DNA evidence, without more, and no matter what the statistical probabilities, cannot be sufficient to establish proof beyond reasonable doubt.
5.2 It is perhaps appropriate to start by taking a more general view of the circumstances in which DNA evidence can be proffered as evidence of guilt. The following analysis, which identifies three significant elements of the evidential chain, is not intended to be exhaustive for there will undoubtedly be unusual cases where different variations on that analysis may require to be considered.
5.3 The first issue concerns the potential connection between a DNA sample found at what one might loosely call a crime scene and the potential guilt of the accused. The term “crime scene” is used loosely for it may involve the actual scene of the alleged crime but also may involve some other location where there is evidence of a connection with the crime. DNA may, for example, be found in a car which can be associated with a crime.
5.4 For the purposes of this aspect of the analysis it is appropriate to assume, but for the purposes of the argument only, that it can be shown that the DNA found in a particular location is in fact DNA which comes from the accused. But even if that is so to what extent can it be said that this tends to establish the guilt of the accused? Here there may be a range of circumstances stretching from those where there is a very high likelihood that the DNA found at a particular aspect of the crime scene must be that of the perpetrator of a crime to those which only demonstrate, at least in and of themselves, a tangential connection. For example, the sole DNA found on a murder weapon where there was evidence that the accused was not wearing a glove and used his or her hands to deploy the weapon concerned falls into a very different category from the finding of DNA on, for example, a cigarette butt located at the scene of a crime but where there is no direct evidence that the perpetrator discarded a cigarette at the time of the crime and no evidence to suggest that the cigarette butt on which the relevant DNA sample was found was necessarily discarded in the course of the commission of the relevant crime and by a participant in that crime. Doubtless a whole range of intermediate examples could be given where the connection between the relevant DNA sample and the identity of the perpetrator of the crime might be established to a greater or lesser degree.
5.5 In passing it is appropriate to emphasise that the Court is not here concerned with the admissibility of such evidence. Clearly even evidence of a tangential connection may be relevant in the overall context of a particular case but would be unlikely to provide sufficient evidence, without more, to allow for a safe conviction. The extent to which the DNA evidence concerned provides evidence of guilt will depend on all the circumstances of the case. It follows that there will undoubtedly be cases where, even should it prove possible to establish that the DNA sample concerned is that of the accused, the connection of the DNA sample to the perpetrator, having regard to the circumstances in which the sample concerned was found, may fall short, or, indeed, a long way short, of providing sufficient evidence, certainly in and of itself, to establish guilt.
5.6 However, there may be other cases where there is an inextricable logic in connecting the sample found at the crime scene with the perpetrator. Whether that is so will depend initially on the assessment by the trial judge as to whether the evidence is sufficient to go to the jury and, if so, ultimately on the view which the jury takes.
5.7 It is, however, important to emphasise that the starting point in any analysis of DNA evidence must be an assessment of the extent to which it can be said that the identification of a sample found at a crime scene as potentially that of the accused can, in all the circumstances, be probative of guilt of the offence as charged and in particular, in a case where there is no other evidence, whether that association is capable of establishing proof beyond reasonable doubt.
5.8 The second set of issues concern the extent to which it can be said that the evidence tendered does not give rise to any reasonable doubt concerning the integrity of the samples taken both at the crime scene and from the accused together with the methodology and practise of retention and analysis of those samples and comparison of the results of that analysis. Courts are well aware that many of the cases where there have been demonstrated miscarriages of justice arising from convictions based on DNA evidence have involved problems at a practical level in the way in which the respective samples were obtained, maintained, analysed, or compared. Of course no system is infallible. But that comment applies to the gathering, maintenance, and analysis of any evidence and in particular any forensic evidence. It is for that reason that the defence must always be entitled to the opportunity to explore the methodology actually applied and to investigate any possible deficiency which might raise a reasonable doubt as to whether the evidence truly supports the view that the crime scene sample and a sample obtained from the accused properly establish whatever degree of similarity the scientific evidence suggests.
5.9 But it is the third element of the equation which comes into particular focus in the circumstances of this case. In that context it is appropriate to assume, for the purposes of the argument, that the crime scene sample is highly probative of it coming from the perpetrator. Furthermore, it is appropriate to assume that no infirmity in the collection, maintenance, or analysis of both the crime scene sample and the sample said to come from the accused has emerged in the evidence which is sufficient to cast doubt on the scientific probative value of the testimony. In such a case the evidence relating to the connection of the accused with the crime may come down to the degree of certainty which may be derived from the statistical evidence in favour of the proposition that the accused was indeed the person whose DNA was found at the crime scene.
5.10 It is also appropriate to draw attention to one other possibility which may have to be taken into account in an appropriate case. One of the significant advances in forensic science in recent years has been the ability to obtain DNA samples capable of analysis from a very small amount of human material. However, even allowing for those advances, it is not always possible, in respect of any particular sample, to conduct an analysis at as many points on the DNA sequence as might be considered ideal. In that context it is worth recording that a very great deal of the DNA of any individual (at least 99.9%) is likely to be identical to that of any other person. All that such DNA shows is that the individual concerned is a human being rather than some other animal. It is, therefore, only a relatively small portion of the DNA of any individual which shows a variation from one person to another. Clearly the greater number of those aspects of the DNA profile which differ from person to person which can, in the practical circumstances of the case in question, be analysed then the greater degree of likelihood that two persons could not have the same profile at the relevant locations without those persons being closely related.
5.11 It is in that context appropriate to refer to the evidence of Dr. Clara O’Sullivan, who was called on behalf of the prosecution. Dr. O'Sullivan holds a PhD from University College Dublin and had, at the time of the trial, worked in the forensic science laboratory in Garda Headquarters for approximately eight and a half years. Her expertise developed in the forensic science laboratory is in DNA profiling.
5.12 Dr. O'Sullivan gave very practical evidence to the jury concerning the random picking of a card from a pack of cards. She drew attention to the fact that the chances of drawing any particular named card was, of course, one in 52. However, as she explained, the odds grow very rapidly where a sequence of cards requires to be drawn. In passing, in that context, it is worth noting that predicting the first five cards to be drawn from a pack correctly in the right order would require bringing home odds of less than one in 300,000,000 (one in 311,875,200 to be precise). However, predicting the first three cards (as opposed to five) in correct order gives rise to odds of “only” one in 132,600 (as opposed to one in 311,879,200). Therefore, a corollary of the analysis of Dr. O'Sullivan is, of course, that, if the relevant samples only prove capable of analysis at a small number of locations, then the chances of a random match may be much greater.
5.13 It may be appropriate at this stage to be clear about the terminology used in the course of this judgment. Dr. O'Sullivan spoke, in her evidence, in terms of the chances of a particular occurrence (such as a DNA match with an unrelated member of the public) being one in a particular number. For ease of reference we will use relatively small numbers to explain the point we are making. Using Dr. O'Sullivan’s terminology it might be possible to describe the chances of something happening as being one in 100. Another way of putting that may be that for every one chance that it will happen there are 99 chances that it will not. The same proposition might be described as the probability or likelihood of the event occurring being one divided by 100 or 1/100. These means of expression all record the same thing. There is one chance out of 100 of the event occurring with it following that there are 99 chances out of the same 100 that it will not occur.
5.14 However, there is one further aspect of the use of that terminology which it is necessary to note. Where the chances or probabilities are expressed in the manner used by Dr. O'Sullivan in the course of her evidence, the likelihood of the event under discussion becomes all the less when the number given gets all the greater. To understand why this is so it is, perhaps, useful to take a very simple example using small numbers. Everyone knows that a half is bigger than a quarter. A half may be expressed as 1/2 and a quarter may be expressed as 1/4. However, when using the terminology adopted by Dr. O'Sullivan a half might be expressed as one chance in two and a quarter might be expressed as one chance in four. While four is obviously greater than two, 1/4 is smaller than 1/2. Therefore, when considering Dr. O'Sullivan’s evidence, it is important to keep in mind that as the number she identifies as representing the likelihood of any particular occurrence gets greater then the chances of that occurrence actually happening gets proportionately smaller.
5.15 In addition, it should be noted that the evidence in this case suggests that there has been no recorded case to date, on the basis of the current level of sophistication of DNA testing, where two unrelated persons have been found to have the same DNA code at each of the points which are the subject of typical analysis. The evidence suggests that the figures given for the likelihood of a random match are more in the nature of numbers indicating the scale or order of magnitude of the lack of probability of two unrelated persons having a complete match rather than representing a precise mathematical calculation of that likelihood.
5.16 It is also important to emphasise that the evidence of Dr. O'Sullivan noted that the reason why it could not be scientifically asserted that every individual’s DNA profile, at the current level of sophistication of testing, was unique was that it would not be possible so to assert without profiling each individual on the planet and assessing whether there was, in fact, a random match to be found somewhere in the world. It is also true that the figure given for the theoretical possibility of a random match involving unrelated persons was somewhat differently expressed at different points in Dr. O'Sullivan’s evidence starting with the suggestion that it was “considerably less than one in a thousand million” and finishing, under cross-examination, with a figure of “one in a thousand billion”.
5.17 The principal evidence in that regard was given on Day 5 of the trial. Much of Dr. O'Sullivan’s evidence was concerned with establishing the integrity of the testing of the samples presented to her from, respectively, the crime scene and that said to come from Mr. Wilson himself. It is worth noting that the DNA samples developed from the crime scene materials already referred to in this judgment were capable of producing a so-called full male profile. Thus it was possible to analyse those samples to the full level recommended in international best practice. Modern techniques do allow extremely small pieces of human material to be capable of analysis but it is not always possible to be able to identify the profile at as many points on the DNA string as might be regarded as ideal. However, this was not, on Dr. O'Sullivan’s evidence, such a case. This is a point to which it will be necessary to return later in this judgment.
5.18 In any event Dr. O'Sullivan first turned to the question of likelihood or probability in a passage of her evidence in chief recorded on the transcript for Day 5, p.60 line 11 as follows:
“Q. And when you say baseball caps, the hoodie top and the grips of the gun, this is the complete profile you’re talking about?
A. This is the complete or the major part of the profile, yes. And then in order to assess the significance of this match, the fact that they matched, I would estimate the chance that a person unrelated to the smoker of the cigarettes or i.e. Keith Wilson would share this profile is considerably less than one in a thousand million, or that’s another way of saying one billion.
Q. So, it may be that somebody else, or may six of them or seven in the world possibly are around the place somewhere?
A. There’s a possibility, because we can’t profile everybody in the world. The only way to determine whether everybody in the world would have a different DNA profile would be to actually DNA profile everybody in the world. Now, I don’t know the figures of the number of people being born per minute or per second in the world but it’s an unfeasible task and that’s why we would never say that it’s a unique, but we give an estimation of the chance of somebody randomly unrelated in the population, having the same profile, given that Keith Wilson would have the profile, is one in a thousand million.
Q. You’ve given evidence that the possibilities are shorter, for instances, if there’s a relation, perhaps a brother?
A. That’s correct. Yes.
Q. Are any figures in relation to that, in statistical terms?
A. In - with regard to this, it could be one in a few thousand. So, the chance that a brother would have the same profile would be one in a few thousand. I can look up my statistics.
Q. Yes?
A. It would be just under one in three thousand chance of a brother having a similar profile.
Q. Do you mean similar - you mean the exact same?
A. The same - sorry, I beg your pardon, yes, the same profile as this person, because they would be related. If there was a suggestion of a brother having a profile, this could be checked and the information I have on file was -
Q. Well never mind about that?
A. Yes, okay.”
5.19 On Day 6, Dr. O'Sullivan, before being cross examined, corrected the last part of her evidence given on Day 5 concerning the possibility of a sibling match when she said, at page 3 line 31, the following:-
“A. My apologies to the Court for the inconvenience. In my evidence, I gave a statistic for the chance that a brother would share the same profile as was seen on the crime stains, for example, the cotton glove, would be less than one in three thousand, I wish to correct that, that that is, in fact, one in 27400. There was a factor of 10 and the reason was the annotation was done in scientific annotation, by 10 to the power of four and I just misread it.”
5.20 Under cross-examination on the same day there then followed an exchange between counsel for Mr. Wilson and Dr. O'Sullivan starting at p.7 line 6 of the transcript in these terms:-
“Q. All right, okay. So, there’s basically equations of some kind that you apply to the numbers to come up with that answer that you’ve given, where you say this profile is considerably less than one in a thousand million in terms of an unrelated person matching it?
A. That’s correct.
Q. All right. And again, just forgive me just I’m having difficulty with the use of language here and it may that I'm - I've just got the wrong end of the stick, but when you say considerably less than one in a thousand million, how much less than one in a thousand million?
A. Well, one in a thousand million is 10 to the power of nine.
Q. Yes?
A. Which is a billion.
Q. Yes?
A. The figure that I have actually calculated for this is 10 to the power of 12.
Q. Yes?
A. So, it’s a million million - it’s a thousand million.
Q. I see.
A. So, it’s considerably -
Q. It’s actually more -
A. No, it’s the chance of this matching -
Q. Yes?
A. -- somebody -
Q. Yes, is -
A. -- is considerably less than a billion times.
Q. I'm sorry, I'm -
A. There’s -
Q. I'm still lost -
A. The way I have -
Q. -- but go on?
A. The way I have written it there it’s - the opposite way of sating it is there’s a one in a billion chance that somebody other than Keith Wilson would share this profile.
Q. That’s an unrelated person?
A. An unrelated person, yes.
Q. That’s important, yes?
A. That’s what the one in a thousand million.
Q. Yes.
A. The considerably less than is actually more, it’s a thousand, thousand million.
Q. I know that’s what you’ve said?
A. It’s a very, very small chance. It’s a very large figure and it is one in a very, very large figure.
Q. I understand we’re dealing with large figures and I suppose in this country we’ve got used to dealing with billions in recent times as opposed to millions, but what I hear you saying is that, in fact, it’s greater than one in a thousand million, the chances of something unrelated to Keith Wilson sharing this profile. But what I read and what you’ve told that jury is that it’s considerably less than one in a thousand million. That’s what’s in your report and that’s what I've difficulty, is it my own command of the English language seems to me to suggest that when you say that something is considerably less than one in a thousand million, then it’s some figure that you can’t say that’ less than one in a thousand million?
A. No, if is say it is less than one in a thousand million.
Q. Yes?
A. There’s - the chance that somebody would share that profile is a one in a billion chance.
Q. All right, so -
A. If - for example, if it was saying it’s more than one in a thousand million, as I think you're trying to say.
Q. Yes, I'm trying to understand, yes?
A. More chance would be one in a hundred, you have a greater chance of getting a match. That would be more.
Q. Yes?
A. Less is less that a billion could be less than 10 billion, less than a hundred billion and so the factual figure for this is considerably less than one in a thousand million.
Q. Yes. I mean, if we took the billions and the thousand millions out of it and you were expressing a view, for example, in relation to one in 10, would you still just e saying it’s considerably less than one in 10. Is that the way you would still express it, of a person unrelated to the smoker of the cigarette sharing the profile?
A. It depends on what the actual figure was.
Q. Yes, I see?
A. If the figure was one in 10, I would have said one in 10.
Q. Well, I'm - as I say, maybe I'm just getting nowhere with this. It’s just the way it’s expressed, considerably less than one in a thousand million?
A. As that’s the chance of it matching -
Q. Yes?
A. -- is less than one in a thousand chance.”
5.21 Under further examination Dr. O'Sullivan, at line 14, did describe the likelihood of a random match as being “very improbable, given the scientific …” There then followed a discussion concerning the increased possibility of a random match with a closely related person. Having reiterated her clarification that the actual figure for a random match for a sibling was one in 27,400, the following final exchange under cross-examination occurred starting at p.13, line 16:-
“A. The figure was 27400.
Q. Is that an exact figure?
A. That’s an exact figure.
Q. And, again, that’s based on statistical analysis; is that right?
A. Yes, that’s correct.
Q. Of the database that you work from with the 300 Irish persons on it?
A. That’s correct, yes.
Q. I see. So, whichever way you look at it, it goes from greater than - it goes from one in a - did you say a million million or a thousand million?
A. A thousand billion.
Q. A thousand billion to one in 27400?
A. Yes, that’s correct.”
5.22 As we have noted earlier the greater the number, in the formulation used by Dr. O'Sullivan, the smaller the likelihood of the event occurring. Thus there is no inconsistency between Dr. O'Sullivan starting by saying that the likelihood of a random match with an unrelated person is “considerably less than one in a thousand million” and finally putting the figure in relation to a possible match with an unrelated person at one in a thousand billion. A thousand billion is clearly a thousand times larger than a thousand million. But it follows that a chance of one in a thousand billion is likewise a thousand times smaller than a chance of one in a thousand million.
5.23 Furthermore, it is important to emphasise that the cumulative chance of a random match would involve the chance of there being a sibling with the same DNA profile (on the evidence one in 27,400) together with the chance associated with some other reasonably close relative having a match together with the chance associated with an entirely unrelated individual. This is a point to which it will be necessary to return. However, it does not seem to us that there is anything in the analysis of that evidence to suggest that Dr. O'Sullivan was not entitled to assert that the chance was, as she put it “very improbable”.
5.24 That being said, it is clear that there remains a theoretical, however slight, mathematical possibility that two unconnected persons could have the same DNA. The real question of principle which arises for this Court to decide on this appeal is whether that possibility must give rise to the sort of reasonable doubt which would require a trial judge not to allow the case to go to the jury in a case where there is no other evidence of identity beyond DNA profiling.
5.25 Perhaps an appropriate starting point for that analysis is to address what has sometimes been called the “prosecutor’s fallacy”. Most people’s eyes glaze over at the mention of very large numbers such as a billion. Furthermore, an assertion that the odds of something happening are one in a billion or less has the potential to create a false impression unless properly analysed and, in the context of a jury trial, explained to the jury. The question of explanation to the jury is a matter to which it will be necessary to return in due course.
5.26 The prosecutor’s fallacy involves confusing two different concepts. One is the chance of any two individuals randomly sharing a set of characteristics. The second is the chance of there being some individual out there amongst the entire population who happens to share the given characteristic. The two concepts may superficially appear to be the same or similar but they are, in fact, quite different. In certain circumstances, giving the figure for the chances of a random match and presenting it as being the same as the figure for the likelihood of there being someone out there who shares the relevant characteristic can be apt to mislead.
5.27 However, to make the issue somewhat less complex and involving less astronomical figures that are actually present in this case, it might be useful to consider a situation where it possible to identify that one in (say) a thousand people had a particular forensic characteristic evidence of which was found at a crime scene. In such circumstances it might nonetheless have been easy to suggest to a jury (or, perhaps, even to a judge) that the chances of the relevant sample found at the crime scene not being that of the accused (assuming that the sample and one from the accused were a match) was a thousand to one and invite the decider of fact to take the view that this was such a remote possibility that it could safely be disregarded even on the basis of proof to the criminal standard. Such a suggestion has, of course, a superficial attraction. It might very well be correct to say that only one in one thousand persons had the same characteristic. But it is a very different thing to say that it follows that it is a thousand to one against the accused not being the person whose sample was found at the crime scene. To suggest that both those assertions amount to the same thing is to fall into the so-called “prosecutor’s fallacy”.
5.28 To explain why that is so one might, for example, take a simple case where the perpetrator was, for whatever reason, highly likely to be a male adult from Dublin. Depending on where one drew the boundary of “Dublin” for those purposes it might follow that there were approximately 400,000 relevant male adults. If it truly was the case that one in every one thousand persons was likely to have the identified characteristic then it would follow that there were, most probably, something of the order of 400 male adults in Dublin who would be a match. More rigorously it is likely that the probability of any particular number of male adults sharing those characteristics being present in the male Dublin adult population would be represented by a distribution curve centring around approximately 400.
5.29 However, for practical purposes, if there were no other evidence at all to link the accused with the crime (other than the identified common characteristic) then there would in truth be no reason to believe that the accused was any more likely to have committed the crime than any of the other 399 or so male adults in Dublin who would have been likely to share the same characteristic. When looked at in that way it is abundantly clear that evidence of that sort could not have gone anywhere near providing, without a lot more, proof beyond reasonable doubt. An analysis of this type is to be found in the unreported UK Court of Appeal case of R. v. Lashley in a passage cited in R. v. FNC [2016] 1 Cr App R 12.
5.30 In passing it does require to be said that some of the earlier decisions from other common law countries, such as Australia, to which the Court was referred in argument do seem to have stemmed from a time when the level of sophistication of DNA testing was not as advanced as that which can be achieved today. Using the rather crude analysis adopted in the preceding paragraphs it is clear that, even if the incidence in the male adult population of a particular DNA sequence could be put at one in ten thousand, there would still be likely to be something of the order of forty persons in the Dublin male adult population who would share those characteristics and, in the absence of any other evidence, no reason to believe that the accused was any more likely to be the perpetrator than any of the other thirty nine.
5.31 The real issue with which courts are now confronted, however, stems from the fact that the sort of probabilities which can now be identified are not in the order of one in a thousand or one in ten thousand but rather, as previously indicated, of the order of much less than one in a billion (i.e. for the reasons already analysed, one in a number which is much greater than a billion).
5.32 It is, of course, long established that the prosecution, in order to sustain a conviction, must exclude any reasonable possibility of innocence. In the context of a case where there was undoubtedly a serious crime but where the issue turns on whether the accused can be sufficiently identified as the perpetrator of that crime, then that issue may turn into a question as to whether the prosecution has established that there is no reasonable possibility of the true perpetrator being a different person than the accused. In the context of a case maintained solely on the basis of DNA evidence it follows that it is necessary to address the question of whether the possibility that there might be another person with the same DNA profile is a reasonable possibility for the purposes of the test. This question has been much debated in what has come to be described as the “trial - by - mathematics” literature. It raises the important question of whether, for the purposes of proof beyond reasonable doubt, or its corollary the reasonable possibility of innocence, there are levels of mathematical probability which are so small that they cannot be said to give rise to a reasonable doubt.
5.33 The relevant literature can, perhaps, best be explained by reference to an oft quoted analysis of a hypothetical, if slightly unrealistic, prison yard assault on prison officers. It is assumed that there is evidence which establishes that all but one of the 25 prisoners in a particular block were engaged in a collective assault on a prison officer but that it is equally clear that one prisoner kept away from the scene and had no involvement. It is also assumed, perhaps not entirely realistically, that there is no evidence at all which tends to identify any particular prisoner as either being the one who remained aloof or as one of those who were involved in the assault. In any event, in such circumstances, it is quite clear that there is a 96% chance that any one of the 25 prisoners was guilty but a 4% chance of innocence. Would it be possible to establish guilt beyond reasonable doubt in those circumstances? One of the disquieting aspects of the possibility of a conviction based on that evidence is that, if it should prove possible to establish guilt to the criminal standard on the basis of that evidence and no more, then there is no reason to believe that any of the 25 prisoners would not properly be convicted. The evidence against each of them would be identical. There is no reason to acquit some and not acquit others. But if each were to be convicted then there is not only a risk but a certainty that one of them will have been wrongly convicted.
5.34 However, the issues which arise in so-called “cold hit” DNA cases (being cases where there is no evidence other than what is said to be a DNA match) are different. The question which must be assessed is as to whether there is a reasonable risk that the one person who is on trial might, notwithstanding the match concerned, nonetheless not be the person whose DNA was found at the crime scene.
5.35 Put another way, does the forensic and statistical evidence leave a reasonable doubt that the relevant DNA found at the crime scene was that of the accused or can it be said that the prosecution have failed to establish that there is no reasonable possibility that the DNA concerned is that of another person.
5.36 In that context it is, perhaps, appropriate to look at two different potential scenarios both of which were canvassed in the course of this appeal. The first concerns the possibility of a DNA match (to the extent of the level of sophistication of the methods of current testing) with a random unconnected member of the general public. The second concerns the possibility of such a match involving a close relative (such as a sibling) who shares more of the same DNA pattern with the accused on the basis that it is clear that close relatives share a much greater degree of DNA than total strangers. It is proposed to deal with the latter circumstance first.
5.37 The evidence in this case suggested that the possibility of two siblings having an exact DNA match under the current conditions of DNA testing was of the order of less than one in twenty five thousand (as noted Dr. O’Sullivan gave the figure as one in twenty seven thousand four hundred). As it happens the evidence was also to the effect that Mr. Wilson had two male siblings. On that basis it was argued that the prosecution had not excluded the reasonable possibility that the DNA found at the crime scene might have been that of one or other of the two siblings concerned. Doubtless similar arguments could be raised in many cases by reference to siblings or other close relatives. In passing it should be recorded that the case of identical twins is, of course, entirely different for identical twins have identical DNA. However, that possibility, which might arise in a very limited number of cases, has no application in this case.
5.38 It might be thought, therefore, that the risk of a false match (being a match with the accused where the DNA was actually that of someone else with the same DNA profile) involving a sibling was much greater than the risk of a false match involving a member of the general unrelated public. There is a certain superficial attraction to that argument. There is no doubt but that, as the evidence in this case demonstrated, it is much more likely that siblings will have the same DNA as opposed to unrelated members of the general public. Using the figures given in evidence in this case, it is very many times less likely that a random member of the public might give rise to a false match compared to the case of a sibling (see the evidence of Dr. O'Sullivan, at Day 7 p.13 lines 25 - 26 cited earlier). That much is undoubtedly correct. However, the problem is that there are a lot more members of the general public than there are siblings. To demonstrate why that superficially attractive conclusion would itself be a fallacy, it is perhaps appropriate to again avoid the sort of astronomical numbers which tend to mislead those but the most numerate.
5.39 An example may explain. One might imagine a lottery for which there were 100 tickets. The organisers’ siblings (two of them as it happens) might be given ten tickets each. The remaining 80 tickets (making a total of 100) might be given one each to general members of the public. Both intuitively and as a matter of mathematical analysis the chances of either sibling winning the raffle will be significantly greater than the chances of any one member of the general public. Each sibling has a one in ten chance of winning. Each member of the general public has a one in 100 chance. However, at the end of the day, it remains much more likely that a member of the general public will win (for those members have collectively 80 tickets) rather than the siblings (who collectively have 20). It is, thus, four times more likely that a member of the general public will win rather than a sibling. While it is nonetheless a fact that any individual sibling may be a lot more likely to win than any individual member of the public, this does not prevent it being the case that it is a lot more likely that a member of the public, rather than a sibling, will be victorious. The reason, of course, is that there are only two siblings and 80 members of the public. The combined chance of any one of those members of the public winning is greater than the combined chance of the two siblings even though the individual chance of any one sibling is greater than the individual chance of a single member of the public.
5.40 A similar analysis, at least in very general terms, potentially applies to the possibility of a false DNA match in the sense in which that term is used in this judgment. The risk of a false match involving a sibling is much greater that the risk of a false match with a member of the general public. But, even in the case of a large family, there are a great deal more members of the general public than siblings so that, unless the potential number of relevant members of the public were small for some reason or other, it would potentially be more likely that there could be a false match involving some random member of the general public rather than a sibling.
5.41 For that reason it does not seem to us that the emphasis on the risk of a false match with a sibling, which was argued on behalf of Mr. Wilson, really stands up to close analysis. Despite the fact that the risk of a false match with a sibling is more likely (by a significant margin) than a false match with a member of the general public, that factor may be outweighed by the fact that there are a great deal more members of the general public who might theoretically be the subject of a false match compared with the number of siblings. The cumulative possibility of any one of those members of the general public being a match could well exceed the cumulative possibility of a sibling match simply because there are so many more members of the public than siblings. For those reasons, we are afraid that we cannot fully agree with the analysis of the Court of Appeal for England and Wales in R. v. Watters [2000] All ER D 1469 which placed particular relevance on the sibling issue.
5.42 We have used slightly conditional language in the preceding paragraphs for there may always be a question, in the circumstances of any particular case, about the total number of members of the general public who might theoretically be said to be the potential perpetrator of the crime in question. Where, as here, the crime is physically committed in Dublin then it follows that the only persons who could have committed the crime are those who are either based in Dublin or who had the potentiality to travel to Dublin on the occasion in question. Other forms of similar analysis might have to be conducted in the circumstances of different cases. However, given that it is the large number of members of the general public concerned which affects the cumulative risk of a false match with a random member of the general public, it follows that the greater the number of persons who might theoretically have committed the crime, the greater the cumulative possibility of a false match with one of those members of the general public and the greater the likelihood that the cumulative risk of a false match with a member of the general public will exceed the cumulative risk of a false match with a sibling.
5.43 Before leaving Watters it should be said that there were a number of complications in that case. First, it would appear that the prosecution sought, at the appeal, to tender revised (and, from the prosecution point of view, more favourable) evidence concerning the likelihood of a false match with a sibling. Second, it would appear that there was evidence that the sibling concerned was actually a suspect in the case in question. It may well be that these factors played some role in the overall conclusions reached. Nonetheless it has been argued that Watters might be considered as persuasive authority for the proposition that the existence of a sibling and evidence of an agreed likelihood that there could be a false match with a sibling at a probability of the order of one in 25,000 leads to the conclusion that a conviction based on DNA evidence in such a case is unsafe. To the extent that Watters might be considered to adopt such a proposition, we are not persuaded that it is correct.
5.44 The evidence by the time the appeal in Watters was heard appeared to suggest that the estimation of “the chance that a brother of Robert Watters would share the same DNA profile as him is about one in twenty nine thousand”. The Court of Appeal took the view that this meant that those odds “do nothing to eliminate the possible brother. They certainly make it unlikely, perhaps unlikely in the extreme, that it was the brother, but they are not sufficient, taken on their own, to enable one to be sure that it could not be the brother in the circumstances of this case”. On that basis a conviction was quashed and no order in respect of a retrial made.
5.45 However, on the basis of the analysis conducted above, it is by no means clear that a possibility of one in twenty five thousand that a sibling might share all of the DNA profile, at the analysed markers, with the accused creates any greater risk of a “false positive” than the theoretical possibility that any one of the very many more members of the general public might have the same profile on an entirely random basis. This is because there are so many more members of the public than siblings. If placing the odds against a false positive in the case of a sibling at less than one in twenty five thousand still gives rise to a reasonable doubt by means of having failed to exclude the reasonable possibility that it was the sibling who was guilty and not the accused then it is hard to see how a similar failure might not well also arise, having regard to the very large number of members of the ordinary population who might theoretically be involved, at the very much lower level of probability applying to the population in general. For example, at a probability of one in one hundred billion but with a population of ten million, a crude calculation would suggest that the probability of a false positive would at least be of the order of magnitude of one in ten thousand (one hundred billion divided by ten million) thus less than the probability identified in the evidence in this case and in the evidence in Watters for a sibling match. The reason, of course, is that in Watters there was one sibling and this case there were two but, on the assumption made in the example given above, the relevant population is ten million.
5.46 Before leaving this aspect of the case it should, of course, be recorded that the actual risk of a false positive is itself dependent on the family profile of the person concerned. The existence of an identical twin would, in and of itself, create a 50% risk of wrong identification. While the incidence of identical twins is relatively small an analysis of the risk of a false positive without knowing anything about the family of the person concerned would have to factor in the possibility of there being an identical twin. On the other hand if it was known that there was not an identical twin then that factor would not need to be taken into account. Given that, as we have seen, the risk of a false positive in the case of a sibling is relatively small but is of the order of one in twenty five thousand, the overall risk of a false positive would have to factor in the possibility of siblings (and their number of the same gender if known) along with the risk of a false positive involving an entirely unrelated person. These and doubtless other factors could potentially play a part. However, the purpose of this analysis is simply to demonstrate that the mere fact that the risk of a false positive involving a sibling may be much higher than the risk of a false positive involving a random unrelated member of the public does not necessarily mean that the overall risk derived from there being a sibling whose DNA was not tested is necessarily any greater than the cumulative risk associated with all other members of the public including those who may be somewhat more remotely related to the subject than a sibling.
5.47 In the light of those general observations it is necessary to consider the issue of principle which arises being as to whether it is possible to establish proof beyond reasonable where the only evidence of identity is DNA evidence. If it is not then clearly this case could not properly have been allowed to go to the jury and the appeal would have to be allowed.
5.48 Perhaps it might be appropriate to start by looking at other, more traditional, areas of forensic identification such as fingerprinting. While there have been challenges on a case by case basis to the probative value of fingerprint evidence in the circumstances of the case in question, nonetheless it has been broadly accepted for a very long period of time that fingerprint evidence can be sufficient to establish guilt. However, at least in very general terms, similar questions can be asked concerning fingerprint evidence as may arise in the context of DNA profiling. While it is said that fingerprints may be unique, the same questions can arise in relation to that issue as arise in relation to the potential uniqueness of DNA. As Dr. O'Sullivan pointed out, the reason why scientists will not assert that it has been established that an individual’s DNA profile is unique is that not everyone has been tested. But not everyone’s fingerprints have been taken either. If it is appropriate to approach DNA evidence on the basis that there is at least a remote possibility that there might be, at the current level sophistication of analysis, a match between unrelated persons then the same potential risk at least arguably arises in relation to fingerprinting.
5.49 It is true, of course, that there are many distinctions between fingerprint evidence and the sort of DNA profile evidence with which this Court is concerned on this appeal. In a sense fingerprint evidence involves expert opinion on whether the respective samples (being one taken from the crime scene and one taken from the accused) can be said to establish the identity of the person who left the fingerprint at the crime scene. The opinion evidence is normally based on the identification of a sufficient number of points on the respective fingerprints which are said to be the same (and, of course, the absence of any points of difference). There can often be debate as to the clarity of the prints most particularly the print obtained from the crime scene which may often not be in perfect condition for analysis purposes.
5.50 But at the end of the day courts have been prepared to allow cases to go to the jury (and juries have been prepared to convict) where the only evidence identifying the accused has been fingerprint evidence. There does not seem to have been any great level of statistical analysis of the possibility of a false identification by fingerprinting even in cases where the crime scene fingerprint does not suffer from any absence of clarity. But it is difficult to see how, at the level of principle, it can be said that evidence which presents a risk which has not been the subject of precise mathematical calculation can properly be assessed by a jury as the only evidence of identity but evidence which presents a risk, which may indeed be a smaller risk, but which is capable of some degree of mathematical precision, can not go to the jury as the sole evidence of identity justifying a conviction.
5.51 Of course fingerprint evidence and DNA evidence are dependent on experts giving their view. As in all such cases it is important that the jury understand the basis on which the expert evidence is tendered and that the jury be informed that they are to assess expert evidence in the same way as they assess all other evidence and should not treat an expert view as being binding on them despite the expertise of the witness concerned. Furthermore, in addition to the duty of the judge to endeavour to ensure that the jury properly understand any expert evidence tendered, it is always open to the accused and the accused’s advisers to consider whether alternative expert evidence might be tendered for the defence so as to dispel any perceived misapprehension which the jury might be under as a result of the evidence tendered by the prosecution. This applies just as much to the statistical analysis which lies at the heart of the inferences which might be drawn from a DNA match in a case such as this as it does in relation to any other expert testimony. While we will return to the question of the role which the trial judge should play in endeavouring to ensure that the jury fully understand the issues involved in due course, it also needs to be noted that it will always be open to the defence, should they feel that the evidence tendered by the prosecution gives a wrong impression of the statistics involved, to tender their own evidence for the purposes of seeking to persuade the jury that the prosecution case on DNA is not as probative as the prosecution might have sought to suggest.
5.52 Furthermore, all with experience of the criminal courts are aware that many forms of evidence, which are frequently regarded as a sufficient basis to allow a case to go to the jury, can be fallible. Evidence can be false, whether deliberately so or due to mistaken recollection. Particular types of evidence, such as identification, have been shown to run risks. It is, of course, the case that jurors can at least bring their ordinary every day experience to the assessment of such evidence. They can observe witnesses and form a view as to whether they are telling the truth or lying. They can watch the witness being cross examined and assess whether the witness’s recollection is likely to be accurate. But all know that there is some, perhaps remote, risk that the assessment will be incorrect but that a jury will, nonetheless, form the view, having been properly directed by the trial judge, that the accused is guilty beyond reasonable doubt. The fact that it may not be possible to do a precise mathematical calculation as to the risk of that not being so (and, indeed, the fact that the risk probably varies quite significantly from one case to another) does not mean that the remote risk is not there but that such cases still properly go to the jury.
5.53 In that context it is, perhaps, appropriate to refer to the decision of this Court in Director of Public Prosecutions v. Connolly [2011] IESC 6. That case involved a prosecution under s.15A of the Misuse of Drugs Act, 1977 as amended. The value of the relevant drugs was, therefore, of particular importance, for a conviction could not be secured unless the jury were satisfied beyond reasonable doubt that the drugs in question had a market value over the relevant threshold. The evidence in the case was that an analysis had been conducted of a sample of five packs of the drugs in question. Each of the bags tested positive for amphetamines. The forensic scientist called to give evidence in that case gave evidence that, having analysed five packs, it was possible to say with 99% certainty that at least seven of the packs contained amphetamine. This evidence was clearly given on the basis of the same sort of statistical analysis of probability that lies at the heart of the DNA profiling evidence in this case. Five packs had been analysed and there was clear scientific evidence that each of them contained amphetamines. However, there was statistical evidence that there was only a 1% likelihood (or less) that none or only one of the other packs (which were not tested) did contain amphetamines. This was because the five packs taken for testing were randomly selected so that it would have been improbable that none, or only one, of the other packs contained amphetamine so that only those packs containing amphetamine would have been chosen at random.
5.54 In analysing that aspect of the case, Fennelly J., speaking for this Court, indicated, at para. 25 of his judgment, that the Court must proceed on the assumption that at least seven of the packs contained amphetamine. It is clear, therefore, that the Court accepted that a statistical analysis which placed the probability of there being less than seven packs which contained amphetamine at 1% was sufficient to allow the Court to proceed on the basis that it had been established beyond reasonable doubt that there were seven packs containing amphetamines. The reason why, of course, the number of packs which had been shown beyond reasonable doubt to contain amphetamines was relevant was that the number of packs impacted directly on the value.
5.55 In passing it should be noted that the appeal was ultimately allowed on the basis that there was insufficient evidence to establish the amount of amphetamine in each pack and thus insufficient evidence to establish that the total amount of amphetamine was sufficient to bring the value of the drugs over the threshold. However, it is clear from the judgment that, had there been sufficient evidence to establish a value over the threshold with seven packs but not sufficient evidence to establish the same value with five packs, then the appeal would have been dismissed. Thus the Court was prepared to accept that proof beyond reasonable doubt could be established with the aid of statistical evidence where there remained a small possibility (in that case 1% or one in 100) that the inference which might be sought to be drawn from the statistical evidence was not correct.
5.56 A precise calculation of the risk of a false positive in this case was not the subject of evidence. It would, of course, be open to counsel on behalf of an accused to seek to identify, to the greatest level of precision possible, the extent of that risk in all of the circumstances of a particular case. In that context there are a number of international publications which give assistance both to expert witnesses, to practitioners representing parties in criminal trials and judges conducting such trials, on the proper use and understanding of the sort of statistical evidence which lies at the heart of this issue. One such are the series of publications under the title “Communicating and Interpreting Statistical Evidence in the Administration of Criminal Justice” published by the United Kingdom Royal Statistical Society. It should be emphasised that if, as a result of the examination of any relevant expert witness, it appeared that the statistical risk of a false positive in all the circumstances of the case in question was significant, then the trial judge would have to consider whether the evidence was sufficient to be allowed go to the jury. A range of factors might well be material in such a case including the number of members of close family and the total population who could realistically be considered to have potentially been the perpetrator. However, where the overall risk of a false positive can be said to be extremely small, we are satisfied that, provided the remainder of the chain of evidence relating to the DNA profiling in the case in question is sufficiently robust, the case can properly be permitted to go to the jury.
5.57 For those reasons we are not satisfied that it can, in principle, be said that proof of guilt to the criminal standard cannot be established even though the only evidence of identity is by means of DNA profiling at the current level of sophistication. As already noted however, it is important to emphasise that factors, such as, but not confined to, those identified in this judgment may also need to be taken into account. The extent to which the DNA sample taken from the crime scene may be probative of the involvement in the crime as charged of the person leaving that DNA may vary significantly from case to case. Likewise the circumstances in which a DNA sample was obtained from the accused may be more or less conclusive that the DNA analysed in the laboratory is in fact that of the accused. The size of the samples, most particularly those from the crime scene but also, potentially, that of the accused, may not allow for as comprehensive a test as might be desired thus increasing, and often very significantly increasing in appropriate cases, the risk of a false positive. There may be issues concerning possible contamination or deterioration of samples. In certain cases the ethnic mix of relevant populations may be relevant in particular to the assumptions on which the statistical analysis of risk of false positives may be based.
5.58 The English case of Raymond Easton from 1999 is illustrative of some of those risks. On the basis of a less rigorous analysis of samples, Mr. Easton was identified with DNA found at a crime scene. However, it transpired that he was suffering from a disease which made it improbable that he was the perpetrator, and more sophisticated tests were carried out revealing that the DNA found at the crime scene was not his. The analysis referred to earlier, by reference to taking cards from a deck, about how quickly the numbers go up or down depending on how many cards are involved provides the backdrop to that problem. Where only a limited number of points are analysed then the odds of a false positive are much greater.
5.59 We would wish to emphasise, therefore, that there are a whole range of factors which need to be assessed before determining whether a so-called “cold hit” DNA case, where there is no other evidence of identity beyond the DNA profiling evidence, can properly provide sufficient evidence to prove identity to the criminal standard. However, where the other elements of the equation are robust, it does not seem to us that the fact that there may be a highly theoretical and tiny mathematical possibility of a false positive can, in and of itself, require that the case be withdrawn from the jury. To take that view would be to suggest that we should be happy to take remote risks with evidence which is not capable of detailed statistical analysis but not happy to take potentially even more remote risks with evidence which may, in fact, be a lot more probative but is capable of statistical analysis to demonstrate an extremely small and very remote possibility of a false positive.
5.60 What the criminal standard requires is proof beyond reasonable doubt or alternatively that the prosecution exclude any reasonable possibility of the accused being innocent. That test, in whichever of those ways it is expressed, makes reference to the doubt or possibility of innocence being “reasonable”. It is implicit in the fact that the test is formulated in that way that there may always be highly theoretical possibilities of innocence which a jury can be satisfied are so remote that they do not give rise to a reasonable doubt or a reasonable possibility of innocence. A witness may identify the perpetrator in circumstances where they have a very good view and actually know the individual quite well. A jury is entitled to conclude that the risk that the witness concerned might be mistaken is so remote that it can be discounted in assessing whether there is a reasonable doubt. The same applies to forensic evidence. A mere theoretical possibility of a false positive or some equivalent phenomenon may be so remote that a jury is nonetheless entitled to regard it as being insufficient to create a reasonable doubt or to establish that the prosecution has failed to exclude a reasonable possibility of innocence.
5.61 Where, unlike here, DNA evidence is corroborative, the same issues may not arise at all. Clearly if there is some evidence, other than DNA profiling, pointing to the guilt of the accused then the fact that there is also a DNA match may simply demonstrate that it would be too great a coincidence to accept that a person who had, for example, both motive and opportunity was nonetheless innocent because of a false match.
5.62 However, even where DNA evidence is the sole evidence of identity, it is capable, in an appropriate case, of establishing proof beyond reasonable doubt. However, before that can be so it is necessary that all of the elements which go into the scientific conclusion are examined. In a case involving only DNA evidence, a significant problem arising in any element of the process (such as, for example, the extent to which it can be said that the DNA sample found at the crime scene must be associated with the perpetrator) might in itself be enough to lead to a conclusion that there was insufficient evidence to go to the jury. Furthermore, it may, again in an appropriate case, be that a trial judge could properly take the view that the cumulative effect of a number of aspects of the process which fell short of being at the height of robustness, might leave the evidence as a whole in a state where it would be unsafe to allow it to go to the jury. Just as an accused can be convicted on the basis of circumstantial evidence where the chances of an innocent explanation for each one of a series of “coincidences” might be material but where the cumulative possibility of all of those “coincidences” occurring might be so remote as not to give rise to a reasonable doubt, so also a series of small flaws in the DNA process, any one of which on its own might not be sufficient to lead to a situation where that evidence would not be sufficient to go to the jury, might nonetheless persuade a trial judge that the overall cumulative effect of the evidence was such that the DNA evidence was not sufficiently reliable to enable it to represent sufficient evidence of identity so as to form the only basis for a proper and sustainable conviction. But whether that is so will, primarily, be a matter for the trial judge in assessing all of the evidence then before the Court.
5.63 It follows that, where the evidence surrounding the DNA profiling in the case is sufficiently far from the robust end of the spectrum, there may well be cases where the trial judge could take the view that it would be unsafe for the jury to convict and, therefore, direct the jury to acquit. Against that backdrop it is necessary to consider the facts of this case.
6. The Facts of this Case
6.1 Having identified the proper approach at the level of principle it is necessary to assess the evidence in this case. There was clear evidence that the perpetrator of the crime discarded items of clothing as he fled from the scene and that those items were quickly identified, secured, and kept for testing. There was nothing in the evidence to suggest that the samples had been contaminated or degraded in any way. We have also described, in an earlier part of this judgment, the circumstances in which a sample of Mr. Wilson's DNA was obtained. While this judgment has already dealt with legal issues concerning the manner in which that sample came into the possession of the authorities, there was nothing in the evidence to suggest that there was any risk that the DNA on the cigarette butt in question was not that of Mr. Wilson, and further there was nothing to suggest that the sample which came to be analysed in the forensic science laboratory had been contaminated or degraded in any way. It follows that the evidence linking the sample obtained from the clothing found at the crime scene with the perpetrator of the crime and the evidence suggesting that the comparator sample was indeed that of Mr. Wilson was very much at the robust end of the spectrum.
6.2 Furthermore, the evidence given by Dr. O'Sullivan as to the manner in which the various DNA samples which were obtained from the materials presented to the forensic science laboratory were tested, analysed, and compared demonstrated that the case in this regard was again at the robust end of the spectrum. This was not a case where, for example, it proved difficult to carry out an appropriate test on one or other sample to the level of sophistication which accords with best international practice.
6.3 It follows, therefore, that the comparison made by the forensic science laboratory of the samples in question was conducted in circumstances where there was extremely strong evidence to suggest that the comparison was being made between a sample taken from Mr. Wilson, on the one hand, and a sample of the perpetrator of the crime, on the other. As noted earlier, such a conclusion may not always be capable of being reached to the same high level as was possible in this case. This may be so in the circumstances of other cases for any one of a whole range of reasons some of which we have sought to analyse earlier. But no such difficulties arose in this case.
6.4 The statistical evidence presented in this case was, as already noted, given in the form of indicating the very small risk indeed of an individual unrelated to Mr. Wilson having the same DNA profile. It would, of course, have been open to Mr. Wilson to seek to have that evidence presented in a different way. For example, having regard to the profile of his family (including the fact that he had siblings) and having regard to the total number of persons who might theoretically form the population from whom the perpetrator was drawn together with any other factors which might be considered relevant, in the form of an estimate of the risk that there could have been another individual who could realistically have been the perpetrator but who would have had the same DNA profile. It is unnecessary, for the purposes of this case, to express any concluded view on the level of such risk which might create a reasonable doubt or a reasonable possibility of innocence which has not been excluded by the prosecution. Suffice it to say that the 1% risk identified in Connolly was not regarded as sufficiently significant to create such a doubt. In a future case where the evidence establishes an estimate of the particular level of risk in all the circumstances of the case in question it may be necessary to revisit this issue. However, for the purposes of this case, it is sufficient to indicate that the forensic evidence as a whole was such that it was open to the trial judge to allow it to go to the jury and it was also open to the jury to reach the conclusion which it did on the basis of that evidence.
6.5 We would, therefore, reject that aspect of the appeal which was based on the assertion that the trial judge should have directed the jury to acquit. We, therefore, finally turn to the issue of whether the trial judge should have given the jury a warning concerning the dangers of convicting where the only evidence of identity was DNA profiling evidence.
7. A Warning to the Jury?
7.1 This third issue concerns whether or not the trial judge erred in law in failing to warn the jury of the dangers of convicting Mr. Wilson in circumstances where the sole evidence of identity against him was DNA evidence. It is contended on behalf of Mr. Wilson that a trial judge should be obliged to give a warning to the jury on the dangers of convicting an accused with regard to DNA evidence if the prosecution sought to rely on DNA evidence alone
7.2 The starting point for the arguments made in this regard on behalf of Mr. Wilson was a Law Reform Commission Report on the Establishment of a DNA database (LRC 78 - 2005). Interestingly, the Law Reform Commission in a consultation paper prior to this report was of the view that a mandatory warning to the jury should be given where the prosecution seeks to rely on DNA evidence alone and where statistical evidence is relied upon in support of that evidence. That recommendation did not form part of the final report which instead proposed that there should be discretion on the part of the trial judge as to whether a warning should or should not be given. In that context it is useful to refer to paragraphs 5.33, 5.34 and 5.35 of the Report which are as follows:
"5.33 … A DNA evidential hearing could provide an adequate means of ensuring that the evidence is sufficiently reliable before being introduced. Whether corroboration is necessary should be left to the judge to decide on the basis of the facts in each individual case. Consequently, the Commission does not recommend that there should be a prohibition on convicting on DNA evidence alone.
5.34 However, the Commission considers that due to the factors that can impact on the probative value of a DNA match and the perceived infallibility of DNA evidence, it may be appropriate that a warning should be given of the dangers of convicting on DNA evidence alone. However, the Commission considers that any warning should be left as a matter for the general discretion of the trial judge as already suggested.
(3) Report Recommendation
5.35 The Commission does not recommend that there should be a prohibition on convicting on DNA evidence alone. The Commission recommends that in all cases where it is sought to rely on DNA evidence alone, it should remain a matter of discretion for the trial judge whether the jury should be warned of the dangers of convicting on this evidence in the absence of other supporting evidence."
7.3 The provisions of the Criminal Justice (Forensic Evidence and DNA Database System) Act 2014 are silent as to the recommendation contained in the Law Reform Commission report.
7.4 Apart from placing reliance on the Law Reform Commission Report, reference was also made on behalf of Mr. Wilson to the decision of the Court of Criminal Appeal in The People (Director of Public Prosecutions) v. Allen [2003] 4 I.R. 295 to argue that a warning should be given to the jury as to the limitations of and the risk of error in respect of DNA evidence. Particular reliance was placed on a passage from the judgment of the Court delivered by McCracken J. at pg. 299 where it was stated:
"Expert evidence comparing D.N.A. profiles is a comparatively recent scientific technique and indeed it would appear that it is still being perfected. As in many scientific advances, the jury have to rely entirely on expert evidence. One of the primary dangers involved in such circumstances is that, the matter being so technical, a jury could jump to the conclusion that the evidence is infallible. That, of course, is not so in the case of D.N.A. evidence, at least in the present state of knowledge."
7.5 Thus it was submitted that the warnings associated with the infallibility of DNA evidence as adumbrated in People (DPP) v. Allen need to be solidified into a warning for cases where the prosecution seek to rely solely on DNA evidence or alternatively that this case was one in which the trial judge in his discretion should have given a warning. It should be noted in passing that the decision of the Court of Criminal Appeal in Allen was delivered on the 18th December, 2003, and I think it would be fair to say that, whilst accepting that such evidence is not infallible, the scientific techniques in relation to matching a DNA sample with a DNA profile have improved in the intervening period as pointed out previously in the course of this judgment.
7.6 Allen highlighted some of the difficulties that could occur in the use of DNA evidence, and it was found in that particular case that there had been a failure to elicit either in direct or in cross-examination the actual statistics used regarding the comparison of DNA profiles which had the potential to confuse or mislead the jury into believing that, even among brothers, an increased probability, starting from a base of one in one thousand million, would be so improbable that they could still disregard it. As the Court in that case observed, the problem in the case was not the evidence given by the expert witness but "rather the evidence which she did not give". The case did not, however, touch on the charge to the jury or the need for any warning. McKechnie J. in his judgment of the Court of Criminal Appeal in this case stated at para. 99:
"The case of The People (DPP) v. Allen [2003] 4 I.R. 295 does not support the proposition that either a specific warning or a warning in general terms is required where the sole evidence offered on behalf of the prosecution is DNA based. In that case the court was not concerned with the question of a warning but rather with the quality of the evidence put before the jury, in particular having regard to what the forensic scientist called on behalf of the prosecution had - and had not - said. The case is therefore fact specific."
This Court agrees with that observation.
7.7 Nevertheless, it is contended on behalf of Mr. Wilson that a mandatory warning should be given to the jury in relation to the dangers of convicting an accused solely on the basis of DNA evidence. The argument is made by reference to other areas in which mandatory warnings are required, particularly the area of visual identification.
7.8 Much reference has been made in the written submissions on behalf of Mr. Wilson to the requirement for a warning in the case of visual identification, relying on an English authority, R. v. Turnbull [1977] QB 224, which has been adapted in that jurisdiction to deal with identification by means of voice recognition. It is suggested that the warning could be likewise modified in the case of identification by means of DNA evidence. The law on visual identification in this jurisdiction has been clear for a very long time and was set out in The People (Attorney General) v. Casey (No. 2) [1963] I.R. 33. It is instructive to look at a passage from the judgment of this Court in that case as it helps to explain why a warning is considered necessary in the case of visual identification. Kingsmill Moore J. at pg. 37 of the judgment stated as follows:
"It is the function of a judge in his charge to give to the jury such direction and warnings as may in his opinion be necessary to avoid the danger of an innocent man being convicted, and the nature of such directions and warnings must depend on the facts of the particular case. But, apart from the directions and warnings suggested by the facts of an individual case, judicial experience has shown that certain general directions and warnings are necessary in every case and that particular types of warnings are necessary in particular types of case.
Such accumulated judicial experience eventually tends to crystallise into established rules of judicial practice, accepted rules of law and statutory provisions. Thus the general directions which must be given in every case as to the onus of proof and the necessity of establishing guilt beyond reasonable doubt have arisen from experience of the fallibility of human testimony in general, whether due to mendacity, imperfect observation, auto-suggestion or other causes. The suggestibility and lack of responsibility of children of tender age find recognition in the statutory provision that their unsworn evidence shall not be sufficient to convict of an offence, unless corroborated by other material evidence implicating the accused, and even when such evidence is received under oath it is customary for judges to tell juries that they should not convict unless they have weighed the evidence with the most extreme care. Similarly the opportunities for giving false evidence afforded to an accomplice and to a person who alleges that a sexual offence has been committed against him or her, coupled with the extreme temptation to give false evidence frequently present in such cases, have given rise to the rule that a judge must warn the jury that it is always dangerous to convict on the evidence of such persons unless it is corroborated in some material particular implicating the accused.
The category of circumstances and special types of case which call for special directions and warnings from the trial judge cannot be considered as closed. Increased judicial experience, and indeed further psychological research, may extend it."
7.9 It was therefore contended in that case on behalf of the accused that the time had come for making an extension to the type of case in which a special direction and warning from the trial judge was required. Thus it was contended that:
". . . accumulated experience has demonstrated the necessity for warning a jury as to the mistakes which can be made, and which have been made, in the identification by witnesses of persons accused and, in particular, that a jury should be told that an identification parade, though the best available method of confirming identification, is very far from infallible."
7.10 Having considered that submission, this Court accepted the proposition put forward in that case and Kingsmill Moore J. went on at pg. 39 to say as follows:
"We are of opinion that juries as a whole may not be fully aware of the dangers involved in visual identification nor of the considerable number of cases in which such identification has been proved to be erroneous; (sic) and also that they may be inclined to attribute too much probative effect to the test of an identification parade. In our opinion it is desirable that in all cases, where the verdict depends substantially on the correctness of an identification, their attention should be called in general terms to the fact that in a number of instances such identification has proved erroneous, to the possibilities of mistake in the case before them and to the necessity of caution. Nor do we think that such warning should be confined to cases where the identification is that of only one witness. Experience has shown that mistakes can occur where two or more witnesses have made positive identifications. We consider juries in cases where the correctness of an identification is challenged should be directed on the following lines, namely, that if their verdict as to the guilt of the prisoner is to depend wholly or substantially on the correctness of such identification, they should bear in mind that there have been a number of instances where responsible witnesses, whose honesty was not in question and whose opportunities for observation had been adequate, made positive identifications on a parade or otherwise, which identifications were subsequently proved to be erroneous; and accordingly that they should be specially cautious before accepting such evidence of identification as correct; but that if after careful examination of such evidence in the light of all the circumstances, and with due regard to all the other evidence in the case, they feel satisfied beyond reasonable doubt of the correctness of the identification they are at liberty to act upon it."
7.11 What jumps out of the passages referred to above is the fact that the need for a warning was based on experience, experience that showed that honest, responsible witnesses could make mistakes and did make mistakes in relation to visual identification. It was for that reason that the Court was of the view that, in the light of the fact that witnesses genuinely made mistakes as to identification, juries were required to be told of that fact and also told to exercise caution in relation to such evidence.
7.12 Accepting the general principle that the category of circumstances and special types of case which may call for special directions and warnings from the trial judge cannot be considered as closed, as was stated by Kingsmill Moore J., the issue as to whether such a special warning or direction in the case of DNA evidence arises. In the course of the written submissions, counsel on behalf of Mr. Wilson made suggestions as to the issues which should be dealt with by the trial judge in relation to DNA evidence. It was argued that the trial judge in this case should have indicated that, where the only evidence of identity against an accused is DNA evidence and particularly in circumstances where it has been established that he has brothers who have not been eliminated from the case, there was a special need for caution before convicting the accused in reliance on such DNA evidence. A number of other points were made as to issues that should have been pointed out to the jury in relation to the nature of DNA evidence, for example where a partial DNA profile had been obtained. However, as was pointed out on behalf of the D.P.P., this was not a case where a partial DNA profile was obtained. It was suggested on behalf of Mr. Wilson that the jury should be told that there were risks of inaccuracy because of contamination of forensic samples or laboratory error and again it was pointed out that this was not a case in which there had been a suggestion that there was any inaccuracy in laboratory analysis of the items submitted for profiling. As was pointed out earlier, the evidence in this case did not touch on any issue as to the integrity of these samples, the methodology in relation to the retention of the samples or the analysis and comparison of the analysis of the samples. In such circumstances a warning to the jury which made reference to such issues would be otiose. The purpose of a judge's charge is not to give directions on issues that do not arise but to deal with the issues of law and fact that arise in the particular case before the Court.
7.13 The reason why a warning is required in relation to visual identification evidence and in respect of the evidence of an accomplice is straightforward and obvious. As Kingsmill Moore J. pointed out as long ago as 1963, even honest and responsible witnesses make mistakes about visual identification. In the case of an accomplice giving evidence against a co-accused, one can see how such evidence could be tainted by self-interest and in such circumstances, again, the need for a warning is clear. Notwithstanding the fact that there may be mistakes made in relation to DNA evidence having regard to the circumstances in which it is obtained, collected, retained, analysed, etc., it cannot be said that there is any inherent unreliability in DNA evidence per se which would require a special warning as contended for on behalf of Mr. Wilson. There may be cases in which the evidence led before the Court in relation to DNA raises issues be it in relation to the collection of the samples at the crime scene, the retention of the samples, the possibility of contamination of the samples, and so on. In such cases, it will be necessary for the trial judge to give a direction to the jury in relation to the relevant issue.
7.14 McKechnie J. in the course of his judgment in the Court of Criminal Appeal in this case at para. 101 stated:
"The probative value of DNA evidence depends on the nature and quality of such evidence as given in a particular case. It is of course largely opinion based, but it does not fall into any recognised category of evidence which it is thought requires the trial judge to mandatorily warn the jury about the dangers of convicting in reliance thereon. That being the situation, it remains a matter for the discretion of the trial judge: a discretion which however should be more acceded to than refused. This point is quite separate from the judge’s obligation, when his overall charge is considered, to ensure that the same can be said as having adequately explained to the jury what the available evidence is and how they should approach its evaluation."
7.15 While a number of useful points are made in that passage, it does not seem to this Court that it is necessary to go so far as to require that the trial judge have a discretion to give a jury a warning about the dangers of convicting an accused in reliance on DNA evidence alone. It will, of course, be necessary to deal with any specific issues that arise from the evidence but that would always be the case.
7.16 A further observation may be made in relation to the use of DNA evidence. Obviously evidence in relation to DNA is given by expert witnesses and in this case the learned trial judge made a number of observations about the evidence of the experts. As McKechnie J. pointed out in the Court of Criminal Appeal judgment at para. 102:
“… the jury was expressly told that it was a matter for them to assess ‘the reliability and the credibility of the witnesses and it is for you to decide what weight you attach to the evidence that you hear, be it expert witnesses or any other witnesses. In fact it is for you to decide on the credibility and reliability of all of the witnesses in the case. This is entirely a matter within your province.’”
7.17 McKechnie J. went on to conclude that the direction to the jury in respect of witnesses viewed in the context of the overall charge was sufficient to ensure that the jury knew their role with regard to the evidence. It is useful to note that further directions were given to the jury towards the end of the charge in relation to expert witnesses by the trial judge who expressly pointed out that the case rested on the jury's view of expert evidence. He went on to say:
"In deciding whether you believe all or part or nothing of what an expert says, it may be helpful to ask yourself some questions.
(1) Are you satisfied from the witness's qualifications that he or she is competent to give the opinion evidence he or she is giving?
(2) Are the facts on which the opinion is based true?
(3) Is the witness expressing an opinion on evidence that you accept and finally, has the witness appeared to have been fair or does he or she intentionally or unintentionally appear to have shown bias in favour of the party who called them, in this case the prosecution?
These are questions it may be helpful to ask yourselves. But, as I say, how you go about assessing the reliability and the weight to be attached to the evidence is entirely a matter for you."
7.18 The directions given to the jury as to the approach they were to take in relation to assessing the evidence given by the expert witnesses could not have been clearer.
7.19 In those circumstances we are not persuaded that the general criteria, by reference to which it has been decided in other cases that a judge must give a warning to the jury, apply in the case of a trial where the only evidence of identity of the accused derives from DNA profiling. We would, therefore, reject the ground of appeal which suggests that the verdict in this case should be overturned because of the absence of such a warning. However, as in any case, it is important that the trial judge ensure that the jury fully understands all of the issues which arise in such a case. In the context of a trial where DNA evidence looms large, it follows that it is important that the jury fully understands the role of the expert evidence relating to DNA and, depending on the issues concerning the DNA evidence which may have actually arisen during the trial, also understand any questions which may have arisen as to the probative value of the DNA evidence concerned. In that regard, we propose to make a number of observations concerning general guidelines as to how a trial judge might consider directing the jury in cases involving DNA evidence.
8. General Guidelines
8.1 When it comes to giving directions to the jury in relation to DNA evidence, the trial judge should point out that DNA evidence is given by an expert and as such the evidence is opinion evidence. They should be reminded of the approach to be taken by the jury in assessing and weighing expert evidence. The jury should be told that DNA evidence, whilst it is technical and scientific by nature, is not infallible. Depending on the facts of the case it may be necessary to point out to the jury any issues that arise from the circumstances of the case. The fact that a sample of DNA is found at a crime scene which is a match for the DNA profile of the accused is not necessarily probative of the guilt of the accused. For example, one can look at the case of The People (DPP) v. Michael O'Callaghan [2013] IECCA 46, a judgment of the Court of Criminal Appeal. That case concerned DNA of the accused found on a homemade balaclava fashioned from the sleeve of a pullover. The balaclava had been worn by one of two men who had carried out a robbery at a post office. One of the men took off the balaclava and threw it into a canal. The balaclava was subsequently recovered and a sample of DNA was obtained from the material on the inside of the balaclava. In the course of evidence it was noted that the DNA sample consisted of a mixture of DNA, that is to say DNA from more than two people. There was a major male component and a minor component. The major profile obtained from the sample matched the profile of the appellant before the Court. The expert giving evidence was able to say that the DNA sample was present on the sleeve when it had been recovered but she was not in a position to say when the DNA material might have been deposited on the sleeve. The Court of Criminal Appeal allowed the appeal of the accused in circumstances where the material in question contained DNA from more than two people. There was no evidence as to when any of the DNA samples were deposited on the material. In its conclusion, Murray J. in giving the judgment of the Court of Criminal Appeal commented:
"The forensic evidence therefore went no further than establishing that the wearer of the balaclava could have been the applicant, a person with the same DNA profile, or any of these two or more persons. The finding of a lot more DNA matching the DNA profile of the applicant in the nose and mouth area of this homemade balaclava than DNA which did not match that profile was not shown to be evidence from which a jury, properly directed, could conclude beyond reasonable doubt that the applicant rather than one of the others had worn the balaclava after it had been fabricated from the severed sleeve."
8.2 The Court went on to express the view that the evidential value of the DNA in that case was that the appellant could have been one of three people who had been in contact with the balaclava or sleeve at some unknown point but probably since it was last washed. The Court went on to note that there was no forensic evidence which would entitle the jury to differentiate between the various people who had been in contact with the material for the purpose of determining which one of them was wearing it at the time the robbery was committed. On that basis the Court was satisfied that there was no evidence on which a jury properly directed could rationally find beyond reasonable doubt that one of those persons rather than another was the person wearing it at the time of the robbery. Accordingly, the appeal was allowed. This case highlights the fact that there was a DNA sample relating to the accused person found on a balaclava, the balaclava was found near the crime scene, there was also evidence that other persons had worn the balaclava but there was no evidence to connect the accused person directly with the crime scene. There was, of course, evidence to connect him with the balaclava but that does not mean that he had been wearing it at the particular time when the robbery was committed. Thus, although there was a DNA match, it did not amount to proof of guilt.
8.3 Equally if an issue arises as to a particular feature of the DNA evidence in a case concerning the way in which the relevant sample was obtained, maintained, analysed, or compared, the trial judge should explain the issue to the jury and outline the relevant evidence in that regard, from both the prosecution and the defense. This should usually be a straightforward process arising from the evidence. Thus, it may be that an issue arises in relation to the possibility of cross-contamination or an issue may arise as to the quality of the sample of DNA obtained at the crime scene. It is not necessary for a trial judge to advise the jury as to potential difficulties that may arise in some cases if those issues have not in fact arisen in the case at hand. Accordingly, if there is no evidence of contamination of forensic samples or laboratory error there is no need to warn the jury that there may be a risk of inaccuracy in assessing whether the suspect's DNA profile matches the relevant sample despite what was contended in this respect by counsel on behalf of Mr. Wilson. Directions given to the jury should be relevant to the particular case before the Court and to the issues that arise in that case.
8.4 Perhaps the area of greatest potential for confusion in terms of the instructions to be given to a jury in relation to DNA evidence concerns the statistical analysis which underpins DNA evidence. Care should be taken by the trial judge that the jury are not led into any confusion in relation to the statistics and in particular that the evidence is not presented in such a way as to involve what has been described previously in the course of this judgment as the prosecutor's fallacy. The Law Reform Commission in its consultation paper on the establishment of a DNA database (LRC CP 29 - 2004) had the following suggestions for a trial judge in relation to informing the jury in respect of DNA evidence at pg. 257:
"However, it is useful at this point to outline some of the issues on which the judge could consider informing the jury if he deemed it desirable in the particular case in hand. In summing up, the judge could give a direction as to the probative value of the DNA match. The contents of this direction could highlight to the jury that DNA evidence is not infallible and should be approached with caution. It could explain that a match does not conclusively prove the guilt of the accused. It could also outline the problems that can arise with a DNA match. These difficulties include the risk of laboratory error, coincidental or chance matches and an innocent explanation being present for the match. . . . The judge could also advise the jury to consider the DNA evidence in the context of all the other evidence in the case. In addition, where there are additional circumstances which may impact on the probative value of the DNA match, these circumstances should also be included in any warning. Careful directions could be given in respect of the expert evidence that arises in each particular case. Attention could also be drawn to the extraneous evidence which provides the context which gives the ratio its significance. Problems with the crime scene sample, for example degradation or contamination, could be addressed in any warning. This would ensure that the jury have a clear understanding of the probative value of the match."
8.5 On the subject of the presentation of evidence on statistics, the Law Reform Commission had this to say at para. 5.16 of its Report:
“It must be made sufficiently clear to a jury that the estimates are not intended to be precise, that they are the products of mathematical and scientific theory, not concrete facts, that they do not purport to define the likelihood of guilt, that they should only be used to form a notion of the rarity of the genetic profile of the accused and most importantly, that the DNA evidence must be considered along with all the evidence in the case relating to the issue of identification."
It is difficult to improve on those suggestions from the Law Reform Commission which was of the view that the trial judge should be required to give guidance on the statistical presentation of a DNA match due to the inherent complexities of that area.
8.6 Insofar as particular questions concerning the statistical evidence may arise in the circumstances of an individual case a trial judge might usefully have regard to international publications, such as those referred earlier in this judgment, which may provide assistance in identifying the best way to explain particular questions of a statistical nature to a jury. However, it must always remain the case that the explanations given by the trial judge to the jury must be rooted in the evidence. Furthermore, trial judges should endeavour to ensure that any statistical issues arising are clarified to the greatest extent possible so as to enable the jury to understand the issues and to avoid being misled by the manner in which statistical evidence is presented.
8.7 For completeness, it may be helpful to refer to a possible direction designed to avoid the dangers of the prosecutor’s fallacy which can be found in sample directions used in the U.K., perhaps adapted for Irish circumstances and taking into account the actual statistical evidence given in the case concerned. The UK guidance is expressed in the following terms:
"The direction should be expressed in terms of probability: for example
'. . . if you accept the scientific evidence called by the Crown there are probably only four or five white males in the U.K. from whom the semen stain could have come. You must look at that scientific evidence and all the other evidence in order to decide whether it was D who left that stain or whether it is possible that it was left by another of the small group of men who share the same DNA characteristics.'"
8.8 It should hopefully be possible, having regard to the general guidance set out above, to formulate an appropriate direction to the jury without over complicating a confusing and technical area of evidence.
9. Conclusions
9.1 As noted earlier in this judgment, there are three issues which require to be decided having regard to the issue certified by the Court of Criminal Appeal together with the additional issues which this Court allowed to be advanced.
9.2 The first issue concerned whether the trial judge was incorrect to allow evidence to be given of the DNA profile of Mr. Wilson having regard to the circumstances in which the cigarette butt, from which the DNA profile concerned was ultimately generated, was obtained. For the reasons set out in this judgment, we do not consider that there was any breach of Mr. Wilson’s rights and in particular his right to privacy. We would, therefore, reject that ground of appeal.
9.3 The second issue concerned the question of whether the trial judge should have allowed the case to go to the jury in circumstances where the only evidence identifying Mr. Wilson as the perpetrator of the murder as charged was DNA evidence. For the reasons set out in this judgment, we are satisfied that the trial judge was correct to allow this evidence to go to the jury. There is no reason in principle why a jury may not be satisfied to the criminal standard of the identity of the perpetrator of a crime where the only evidence of such identity derives from DNA profiling. However, we make clear that, in our view, it is necessary that the evidence concerning the obtaining of a DNA profile from a crime scene, the probative connection between the presence of the DNA at a crime scene and participation in the crime as charged, the preservation, analysis and comparison of any DNA found at the crime scene with any DNA which can be established to be that of the accused together with evidence concerning the likelihood that the DNA of the perpetrator might not, despite a match, be that of the accused, must be sufficiently robust to warrant it being possible for the jury properly to conclude that proof beyond reasonable doubt has been established. Where such evidence is insufficiently robust to that end it is, of course, appropriate for the trial judge to direct the jury to acquit. We, therefore, reject this second ground of appeal.
9.4 The third issue concerned the contention of Mr. Wilson that the trial judge should have given a warning to the effect that there were dangers in convicting in circumstances where the only evidence of identity stemmed from DNA profiling. For the reasons set out earlier, we do not consider that this ground is well made out for we do not consider, having regard to the general principles by reference to which it has been determined that warnings are required in other cases, that a warning is required in a case where the only evidence of identity is confined to DNA profiling. However, we have taken the opportunity to give some general guidance as to the manner in which a trial judge might consider directing the jury in such cases. We, therefore, reject the appeal based on this third ground as well.
9.5 It follows that, each of the grounds of appeal having been rejected, the appeal must be dismissed.
