7.2.1 The requirement for corroboration was cited in McLean as one of several elements in the criminal justice system that contributed to the overall fairness of the trial procedure. The Review's Terms of Reference expressly set out that this is an area which must be examined. It is one of the issues that has attracted most comment in the consultation responses, workshops and other meetings. It has also been the focus of considerable media attention. More than any other feature of the criminal justice system, it is seen by many as a defining and distinctive characteristic of the Scots Law of evidence in criminal cases.
7.2.2 The Review has no doubt that a majority of persons, especially lawyers, practising in criminal law, regard corroboration of testimony as an important aspect of their professional lives . Many see it as an important tool, which helps them to make decisions at various stages in the investigative and trial processes. The police officer has regard to it in deciding whether to report a case to the procurator fiscal. The procurator fiscal looks to see if there is corroborated evidence before drafting a complaint or moving for full committal on a petition. The defence solicitor or advocate has to examine the prosecution case in order to form a view on whether there is sufficient corroborated evidence before advising his/her client on how to plead. He/she will look at the evidence at trial and an absence of corroboration will justify a "no case to answer" submission at the conclusion of the Crown case. The judge will respectively either acquit the accused or direct the jury to do so where no corroboration exists at the stage of the submission or at the conclusion of all the evidence. The whole matter may be reviewed at appeal, when a lack of corroboration will inevitably result in the quashing of a conviction. It is so much part of the daily decision making process that it has come to be seen as a, if not the, pillar of the system.
7.2.3 The necessity of having corroborated evidence has, as will be clear from the previous chapter, lain at the heart of the criminal justice system since time immemorial and has been, and still is, regarded by many as an "invaluable safeguard" against the occurrence of miscarriages of justice . Its stated purpose, in criminal cases, is to prevent an accused from being wrongly convicted on the basis of a single witness, who may be either fallible or dishonest. This is, as noted above, the reasoning of Hume . Burnett states that the rule is founded "both in reason and humanity" . Professor Gordon  neatly sums up the point, describing it as an example of "rule-utilitarianism", as follows :
"we accept that sometimes a single witness can be reliable and that by refusing to believe him we may be doing injustice in the particular case; but we cannot always be sure about our judgments of reliability, and indeed we are so likely to be wrong, and the results of our error are likely to be so serious, that it is better to make it a rule that we shall never rely on only one witness, because, on the whole, that will lead to less injustice than will reliance on our ability to detect unreliability".
7.2.4 It must be recognised that criminal law, including the rules of evidence and procedure, has advanced some distance since the days of Hume and Burnett. Capital punishment no longer exists in time of peace. Prosecutors and judges are, for the most part, well educated in the law and procedure and, so far as is known, not readily susceptible to corrupt practices. There is a legal aid system providing effective representation for the defence. Scientific proof is a regular feature in trials. Persons convicted in the High Court have a right of appeal. Almost all persons, including the accused, are acceptable as competent witnesses. It is an altogether different legal and social world from that of the early or late eighteenth century.
7.2.5 The question, which the Review has asked itself, is whether the requirement is a useful tool for achieving Professor Gordon's stated purpose in the modern world or whether it is an artificial construct that actually contributes to miscarriages of justice in the broad, rather than appellate, sense. Is corroboration merely a comfort blanket for decision makers; that is, something which does not really assist in making a decision the correct one, but which can be used to justify that decision in objective terms?
7.2.6 According to the requirement, there must first be at least one source of evidence (i.e. the testimony of one witness) that points to the guilt of the accused as the perpetrator of the crime. That evidence may be direct  or circumstantial . Secondly, each "essential" or "crucial" fact , requiring to be proved, must be corroborated by other direct or circumstantial evidence (i.e. the testimony of at least one other witness).
7.2.7 Generally, there are two crucial facts requiring proof in every crime: (1) that the offence was committed; and (2) that the accused committed it. Dicta to the effect that " mens rea" is a fact which requires to be corroborated are widely regarded as erroneous  even if this view continues to be advanced, especially in sexual offences cases . Intention is a fact which may be inferred from proof by corroborated evidence of the crucial facts. In relation to proof of the crime, not every element requires corroboration. Thus, in an assault, if one part of the attack is established, that may be sufficient, at least where the assault is all of one type .
7.2.8 There are some limited statutory exceptions to the requirement for evidence to be corroborated . These exceptions, which tend to relate to minor crimes, do not attract any substantial adverse criticism.
7.2.9 Where then has the law reached? What was, at least until recently, regarded as the classic statement on the requirement for corroboration in modern terms came from a civil case, namely O'Hara v Central SMT , which pre-dates Gillespie v MacMillan. The dictum is from the then Lord President (Normand) and is as follows :
"Corroboration may be by facts and circumstances proved by other evidence than that of a single witness who is to be corroborated. There is sufficient corroboration if the facts and circumstances proved are not only consistent with the evidence of the single witness, but more consistent with it than with any competing account of the events spoken to by him. Accordingly, if the facts and circumstances proved by other witnesses fit in to his narrative so as to make it the most probable account of the events, the requirements of legal proof are satisfied".
7.2.10 It can be seen immediately that what is being described as corroborative evidence is not evidence which, of itself, points towards guilt. When one witness speaks to seeing a person committing a particular crime, the potential corroboration is not testimony that necessarily also describes the crime. Rather it is testimony which covers facts which are consistent with the first witness's account and appears to confirm that testimony, as distinct from confirming the facts, or events spoken to, as accurate or true. This is, as described in the last chapter, the law bending in the face of a requirement which, if strictly applied, cannot operate satisfactorily in practical terms in the modern world of criminal justice, since it would result in very few convictions despite the existence of strong circumstantial, or even direct, evidence.
7.2.11 After what might be described as a period during which the requirement appeared to be strengthened, or perhaps changed, following Mackie v HM Advocate , it was re-affirmed in the criminal case of Fox v HM Advocate in the following, rather different, terms :
"Corroborative evidence is… evidence which supports or confirms the direct evidence of a witness… The starting-point is that the jury have accepted the evidence of the direct witness as credible and reliable. The law requires that, even when they have reached that stage, they must still find confirmation of the direct evidence from other independent direct or circumstantial evidence… The evidence is properly described as being corroborative because of its relation to the direct evidence: it is corroborative because it confirms or supports the direct evidence. The starting point is the direct evidence. So long as the circumstantial evidence is independent and confirms or supports the direct evidence on the crucial facts, it provides corroboration and the requirements of legal proof are met".
7.2.12 Evidence can thus be corroborative even if, taken on its own, it does not point conclusively, or even at all, towards an accused's guilt. In a case where identification is in issue, a positive identification by one witness may be corroborated by a resemblance identification by another  which, on its own, would not have been sufficient for what might be called the first source. It is capable of corroborating a first source but, itself, would not amount to sufficient evidence if supported only by another resemblance identification.
7.2.13 Put another way, the law does not require two witnesses in the original Romano-canonical law sense of two testimonies, each confirming guilt. It requires one such testimony and another witness speaking to facts which make the truth or accuracy of the first witness's evidence more likely. Whether these facts do make the first witness's evidence more likely to be true is a matter of judgment, upon which there can be disagreement, even amongst the judiciary. Even then, it is not every fact which the law imbues with this corroborative characteristic. Although a statement made by a witness to a third party immediately after an incident  may, if proved, lend support to that witness's veracity, a jury would be directed that, even so, it does not provide corroboration as it emanates from the same source (i.e. the same witness).
7.2.14 The requirement for corroboration permeates the whole criminal justice process from suspicion to conviction. It is not just a test adopted in court. Its existence means that, in respect of all crimes, however trivial, the prosecution, and hence the police, need to find corroboration unless the requirement is excluded by statute. If corroboration is not found, except where the case is a particularly serious one, it will not normally be reported to the procurator fiscal and the case will therefore not be progressed. Of course, even when corroboration is found, the investigative process does not suddenly stop. It is accepted that, within reason and depending to a degree on the seriousness of the offence, the police will, or at least ought to, seek to find further relevant evidence. Once the case is reported and it is agreed that there is sufficient corroboration in a technical sense, the procurator fiscal or Advocate Depute still requires to consider whether it is in the public interest to proceed. This will involve an additional element, notably the question of whether there is a realistic prospect of a conviction.
7.2.15 How does the requirement work in practice? How do these rules within rules actually operate in the courts? The best clue to understanding how corroboration works is to grasp that, however the requirement is phrased in theory, it is about the number of witnesses speaking to a fact or facts and not about the number of different, or separate, facts necessary to prove a case. When reference is made, in the context of the requirement, to a "source of evidence", it is to the testimony of a single witness and not to a particular fact. Thus, as an easy example, two closely related witnesses standing at a street corner may each identify a particular individual as someone who assaulted a victim. They will each corroborate the other, even if they are brother and sister, husband and wife or members of the same gang. Another example is where there is CCTV involved. Provided that the recovered recording is proved by two witnesses to show the scene of a crime, two witnesses each identifying a person from the recording as the assailant will suffice, even if they are both police officers viewing the recording later. The CCTV recording is not the "source of evidence". The corroborating sources are the two witnesses speaking to the recording and the two (possibly the same) witnesses identifying the person shown.
7.2.16 Similar considerations apply to forensic evidence, such as a fingerprint recovered or found at the scene or the discovery of a person's DNA on an object at the locus. The finding of a fingerprint or DNA at the scene of a crime does not constitute a single source of evidence which requires to be corroborated for guilt to be established. It is a solitary fact, but one which, if proved by corroborated testimony, may of itself be sufficient to prove guilt. This may occur where, for example, the print or DNA is found at the scene of a housebreaking or on a murder weapon. In that event, the fact of finding the print or DNA will be sufficient for a conviction if it, and its successful comparison to samples taken from the accused, are each spoken to by two witnesses.
7.2.17 It is in this type of area that misunderstandings can occur especially, but by no means exclusively, by those not trained in the criminal law. If the finding of a fingerprint or DNA sample is one fact amongst several others in a case pointing towards guilt, there is no need for that finding to be spoken to by two witnesses. One is sufficient where there is other testimony proving the other facts. It is only where the finding is the only fact incriminating the accused that the finding of the print or DNA, the provenance of the sample taken from the accused and the comparison of the two, require to be proved by corroborated evidence; i.e. the testimony of two witnesses. The problem with this is that it may well not be known, at the time of the finding or the comparison of the finding, whether this is the only piece of incriminating material. Thus, it is the common and accepted practice for there to be two witnesses  to speak to the finding of any print or DNA, two witnesses  to speak to the taking of samples from the then suspect and two witnesses  to speak to the comparison. There is little doubt that this leads to considerable, unnecessary and costly duplication of effort and sometimes testimony .
7.2.18 It may seem immediately apparent that, were modern legislators seeking to devise a model for sufficiency of proof in the today's sophisticated world of information and communications technology, corroboration would not be at the forefront of their thinking. If a crime is shown happening on a CCTV recording, it hardly seems necessary for proof that the recording be spoken to by two witnesses, even if the camera does sometimes "lie". If a person's body is examined by a recognised pathologist and he assures the court that the person was dead, it hardly seems necessary to insist that another doctor give evidence to prove that death. There are many more examples, even if there are also situations, including certain instances of eye-witness identification, where a fact finder may be advised not to rely on the evidence of one witness alone. However, in order to deal with the problems of the requirement in the twentieth and twenty first centuries, the courts developed a series of subsidiary rules all of which, like the dicta in Gillespie v Macmillan, were ultimately designed to create a sufficiency of proof, where none would exist were the requirement to be enforced in its traditional Romano-canonical guise.
7.2.19 The most complex aspect of the law of corroboration in modern times is mutual corroboration, or corroboration by similar facts, highlighted in, although not strictly created by, the celebrated case of Moorov  in 1930. Moorov itself was concerned with the sufficiency of identification evidence where each of a number of complainers, who were all assistants in a draper's shop, spoke to the appellant as the perpetrator of indecent assaults on her. No single act of indecency was corroborated by the testimony of any other witness, yet it was held that each separate act, spoken to by one complainer, could be corroborated by the testimony of another complainer speaking to another such act. The use of the principle was expanded over time to cover other aspects of proof, including the facts of the crime itself and situations where there was no eye-witness identification at all of the perpetrator in respect of one or more of several charges . This form of corroboration was always available at least in theory, whereby one complainer's testimony about a particular attack could corroborate another complainer's testimony of a different attack; provided both incidents were sufficiently closely connected in time, character and circumstances. In that situation, what are separate acts are treated as a single course of conduct. It is that course of conduct, if demonstrably perpetrated by the same person, that requires to be proved by corroborated evidence and not each separate incident .
7.2.20 This area of the law is fraught with dangers in relation to both what amounts to a sufficiency of evidence and how to direct juries in the event of one or more complainers, in respect of several charges, being believed or disbelieved . There are frequent arguments about what might, or might not, amount to corroboration especially where two or more incidents are separated by significant time gaps .
7.2.21 Similar considerations apply to the use of a complainer's distress, as observed by a third party after an alleged sexual offence, to corroborate the use of force , or lack of consent in the course of that offence , even where there is an alternative explanation for the particular emotional response. The reasoning behind this is that the complainer's distress, spoken to by a third party, is a physical manifestation or reaction, akin to but not the same as a de recenti statement. Independent proof of the reaction can lead to an inference that whatever happened did so against the will of the complainer and was therefore something brought about by violence or, now, at least without the complainer's consent. This development has in turn led to new problems, since the courts have also held that, whilst distress can corroborate lack of consent , it cannot be used as corroboration of the facts in the libel, such as intercourse or particular acts of violence or indecency . There is, of course, little problem with using this type of distress as a bolster to a complainer's credibility. It is its use, as something emanating from a single source (i.e. the complainer in the form of her oral testimony and demonstrable distress) that concerns corroboration purists on the issue of sufficiency.
7.2.22 A confession, whether corroborated or not, is regarded as just one source of evidence pointing towards guilt. It too, like the testimony of an eye witness (of which it is a variant) needs to be corroborated . How is that to be achieved? On a strict view, there ought either to be evidence from another source which points to the guilt of the accused or at least testimony confirming the truth of a fact or facts in the confession, such as the location of the body, which could only have been known to a guilty person (a "special knowledge" confession) . However, so far as the former is concerned, the general view is that the corroborating testimony need only be such as to make the truth of the confession more likely , not that it points to a fact demonstrating the accused's guilt. In relation to the latter, the Crown do not need to prove that the accused was the only person who knew of the fact spoken to. Proof that other persons, such as the police, were aware of the fact does not prevent it being used as corroboration . As an addendum to this area, where a "special knowledge" confession is the only evidence against the accused, it is necessary for the confession itself to be proved by corroborated evidence. This means that, even if the confession is recorded on video or audio tape, two persons, usually police officers, require to speak to the interview and the special knowledge adminicles. This is not the case with an admission, the truth of which is supported by other testimony. In that event, the evidence about the admission need not be corroborated. However, intricate problems may then occur when directing a jury, where the other testimony may be rejected by a jury as incredible or unreliable.
7.2.23 Article 6 of the Convention does not require a state to insist that evidence be corroborated in order for a person to be convicted. It does, of course, provide a right to a fair trial. There have been cases in which the European Court has expressed some willingness to consider the way in which evidence has been obtained in determining whether a trial as a whole has been fair . In the context of hearsay, for example, it has made some pronouncements which have had a bearing on what evidence might be regarded as sufficient, even if this area is approached from the angle of fairness, in the sense of the ability of a person to test whatever evidence is adduced against him/her. Since the requirement for corroboration does not exist in any other jurisdiction, it is reasonable to assume that, were it to be removed, there would be no basis for arguing that this would result in the trial process as a whole infringing Article 6. However, the fact that the removal of the requirement for corroboration would not result in an infringement of Article 6 is not, of itself, a sufficient reason for its abolition. It would, of course, be ironic if Scotland were to remove the requirement for corroboration just as the European Court decided that, following the hearsay example , it was unfair for a person to be convicted solely on the evidence of one witness. There appears to be no immediate prospect of that becoming its jurisprudence .
7.2.24 Common law jurisdictions may have a requirement for corroboration in certain cases. In England and Wales, for example, by statute, it is still required in cases of perjury  and, curiously, speeding . But generally it is not an absolute requirement. As one legal historian put it :
"…English law did not adopt the general principle of merely counting witnesses".
However, in jury trials it became obligatory to give the jury a special warning  about the dangers of accepting uncorroborated evidence in certain classes of case. These were trials where the evidence against an accused came solely from a child , the complainer  in a sexual offences case or an accomplice . The warning was to the effect that it was dangerous to convict without corroboration, but that the jury could do so if they were nevertheless satisfied of the truth of the testimony . There are other situations where warnings of a similar type could be given, including where a witness might have an improper motive for lying or where the evidence against an accused was in the form of a fleeting eye witness identification.
7.2.25 The rules concerning the need to give warnings have largely been abolished in England and Wales  and elsewhere . The reason given for abolition in England and Wales was that the rules concerning corroboration were deemed to be inflexible, complex and anomalous . This was so despite firm pronouncements at the highest level in earlier cases about the utility and desirability of their application . The thinking was that the rules had been developed :
"for the protection of accused persons against the possibly unsophisticated reasoning processes of juries and lay magistrates--against the failure of lay people to take account of the motives which accomplices may have for lying, or the alleged propensity of the victims of sex offences to fantasise and fabricate".
But, as outlined above, there were many cases, outwith the rigid categories described, where a judge might be expected to give a warning and many within these categories where no warning was appropriate. It is now a matter for the discretion of the trial judge whether to give a warning to a jury in respect of unsupported evidence . The terms of any warning will depend on the circumstances of each case, the issues raised and the content and quality of the witness's evidence. For a warning to be given, there has to be evidence that the witness may be unreliable. Assertion or suggestion in cross-examination is not sufficient.
7.2.26 Some other common law jurisdictions also continue to allow warnings to be given if justified in the circumstances. On the other hand, in Australia and Canada, judges are specifically prohibited in sexual offence cases from warning juries of the dangers of convicting an accused on the basis of the uncorroborated evidence of the complainer . Otherwise, where the prosecution is relying on a single witness, the judge may warn the jury to scrutinise the evidence of that witness with great care .
7.2.27 The position in Ireland is similar to that in England and Wales and other common law jurisdictions. McGrath explains that the drawback of a corroboration requirement in certain categories of case was thought to be that :
"it had the potential to lead to unmeritorious acquittals in circumstances where the evidence of a potentially unreliable witness was considered to be reliable in the particular case by the tribunal of fact but no corroborative evidence could be adduced".
In emphasising that the common law eschews any quantitative requirements in favour of qualitative assessments, he quotes from Weiller v United States where Black J said :
"Our system of justice rests on the general assumption that the truth is not to be determined merely by the number of witnesses on each side of a controversy. In gauging the truth of conflicting evidence, a jury has no simple formulation of weights and measures upon which to rely. The touchstone is always credibility; the ultimate measure of testimonial worth is quality and not quantity. Triers of fact in our fact-finding tribunals are, with rare exceptions, free in the exercise of their honest judgement, to prefer the testimony of a single witness to that of many".
7.2.28 Ireland requires corroboration in cases of perjury, procuration of females  and speeding . There is no rule requiring corroboration in cases of accomplices, but a warning must be given of the dangers of accepting their evidence uncritically . The jurisprudence on the nature of the warning is extensive . A warning has been required in sexual offences cases, but this came to be criticised as not only technical and complex but also "misogynistic in conception and application" . It has been abolished  in favour of permitting the judge a discretion to give a warning. The position in relation to children is not dissimilar. Although originally corroboration of the unsworn testimony of children was mandatory on the basis of fears concerning their reliability, research demonstrated these fears to be "based upon untested and unfounded assumptions" . The requirement was therefore also abolished  in favour of a discretion in the judge to give a warning.
7.2.29 In February 2011 the Crown Office announced  that 867 live cases were no longer to be progressed because the police interview was not Cadder compliant. Fifty-one of these cases had been indicted in the Sheriff Court and nine in the High Court. The vast majority were therefore summary prosecutions. In addition, the Crown Office had to concede 3 summary appeals.
7.2.30 It is the Review's perception that, although these abandonments occurred as a direct result of Cadder, it was the requirement for corroboration that prevented the majority of the cases proceeding rather than the Article 6 issue. Without the requirement for corroboration, a significant number of these cases may very well still have had a sufficient quality of evidence to have justified continued proceedings, even though evidence of the police interview would be inadmissible. They would have met the test of a "realistic prospect of conviction". Although, when compared to the totality of prosecutions in Scotland, this number may be perceived as small, it is still numerically significant. On one view it means that, in the broad sense, miscarriages of justice may have occurred in a number of these cases because of the requirement for corroboration. With no prosecution, evidence suggesting that a crime had been committed was not tested and witnesses, including victims, may have been left seeing the person, whom they regarded as perpetrating a significant crime, go free. This, of course, is in the context of a sample of cases which had been instructed for prosecution because corroboration existed in the form of non Cadder compliant statements to the police. A far greater number can be assumed to exist where no such statements existed and the complaints made are not reported to procurator fiscal in the first place.
7.2.31 In order to explore this issue further and to give some empirical underpinning to the arguments being advanced, the Review commissioned research from COPFS. The research looked at two sets of cases: (1) for the entire calendar year 2010, all cases that were put on petition, but marked as 'no further proceedings due to insufficient evidence' (458 cases identified); and (2) for the six month period July to December 2010, all cases reported to the National Sexual Crimes Unit where the accused was not placed on petition due to lack of evidence (141 cases identified) .
7.2.32 These cases were examined by experienced procurators fiscal using a two stage test. First, consideration was given to whether there was sufficient evidence to proceed if the requirement for corroboration did not apply. Secondly, of those cases that passed this threshold, a calculation was made of the number which would have further survived the test applied by the Crown Prosecution Service in England and Wales in deciding whether to prosecute. That test is whether there is a reasonable prospect of conviction.
7.2.33 The results of this work were quite striking, raising serious questions as the extent to which the requirement for corroboration may be preventing the prosecution of cases that could result in convictions. For the first set, it was judged that 268 of the 458 cases considered (58.5%) would have been prosecuted on the two-stage test applied. For the second set, it was considered that 95 of the 141 cases examined (67%) would have been prosecuted. It should also be noted that the cases considered in the first set covered a number of categories of crime. Most predominant were crimes of violence, but crimes of dishonesty and drugs offences also featured. Across the categories, a substantial proportion of the cases considered were judged to have had a reasonable prospect of conviction if prosecuted. The full research report can be seen at Annex A.
7.2.34 The Review has no reason to believe that these figures are not reasonably indicative of what might happen in the absence of the requirement for corroboration. They suggest that a substantial proportion of cases, which are currently not prosecuted because they fail the corroboration test, could be prosecuted with the reasonable prospect of securing a conviction. If that is the case, it is hard to avoid the conclusion that the requirement for corroboration is an impediment to justice, rather than a safeguard, in a significant number of cases.
7.2.35 In considering the utility of the requirement, the Review is not starting from a blank sheet of paper. The Review is not considering whether or not to recommend introducing such a requirement. Rather, it is acutely conscious that the requirement not only exists but is also perceived as a key element in the criminal justice system. Removing the requirement would require significant changes in perception for the police, the Crown, defence agents and counsel and the courts . Such a change should only be recommended if it is demonstrated to have a positive aspect.
The case for retaining corroboration
7.2.36 The principal justification for retaining the requirement for corroboration is clear. It is the same as that stated by Hume ; that it provides some protection against miscarriages of justice. The requirement, it is argued, reduces the prospect of miscarriages of justice occurring as a result of a judge or jury convicting an accused or the basis of a single piece of testimony that is untrue or unreliable. Evidence can be fallible in a number of ways. A witness may be deliberately lying. Contrary to some popular belief, or at least expectation, witnesses do frequently lie in the criminal courts. Their motives for doing so will vary but at least some will be anxious, for reasons perceived by them to be sound, to see that the accused is convicted. That may be because of malice or, more likely, because they are convinced, from information received or otherwise, that the accused committed the crime. Such witnesses may therefore come to court prepared to lie, either on their own or in combination with others. Despite skilful cross-examination, the lies may remain undetected by the judge or jury.
7.2.37 Even more problematic can be the witness who is attempting to tell the truth, and perhaps trying to be as helpful as possible to the court, but is simply wrong. A witness can be unreliable, even if manifestly credible. In particular, eye-witness identification evidence is often regarded as being potentially unreliable . It must therefore be recognised immediately and clearly that it is quite possible for a judge or jury simply to err in the assessment of a single witness .
7.2.38 The requirement for corroboration can only prevent a miscarriage of justice occurring in cases where there is only a single witness providing incriminating evidence of the commission of the crime or the identity of the perpetrator. It cannot do so in the many cases where there is a technical sufficiency. Theoretically, all miscarriage cases identified after the appeal process has been completed in Scotland ought to have had such a technical sufficiency since otherwise the trial judge ought to have sustained a no case to answer submission or, if he/she erred in that regard, the conviction ought to have been quashed on appeal.
7.2.39 The requirement must, at least in theory, prevent miscarriages occurring in the single witness situation as, almost by definition, it eliminates that category of case. As Lord Morris said in the context of a case requiring corroboration of child evidence in England :
"Any risk of the conviction of an innocent person is lessened if conviction is based upon the testimony of more than one acceptable witness. Corroborative evidence in the sense of some other material evidence in support implicating the accused furnishes a safeguard which makes a conclusion more sure than it would be without such evidence".
The more difficult issue, however, is the measure of protection that corroboration provides.
7.2.40 If the requirement were removed, prosecutors would need to decide whether cases should be prosecuted on the evidence of one witness alone. It is not disputed that prosecutors could make such decisions. They do so everywhere else in the United Kingdom, Europe and the Commonwealth. However, if a broad test such as "likelihood of conviction" or "interests of justice" were used, this could lead to inconsistencies in decision making, at least in cases lacking corroborated evidence. Similarly, Justices of the Peace and sheriffs may, whilst continuing to adhere to the universal standard of beyond reasonable doubt, tend to apply their personal views, perhaps even more divergent, on what constitutes adequate evidence to merit a finding of guilt 'beyond reasonable doubt'.
The case for abolishing corroboration
7.2.41 A principal argument for abolishing the requirement for corroboration is that it does not, in practice, serve its stated purpose of preventing miscarriages of justice. The real protection against miscarriages of justice at first instance is the standard of proof required; that the judge or jury must not convict unless convinced of guilt beyond reasonable doubt. It is the need to satisfy this test that makes the existence of supporting evidence, whether currently classified as corroboration or not, important. Removing the formal requirement for corroboration may not result in significant changes to conviction rates, at least in cases of the type currently prosecuted. It is also highly unlikely that, at least in serious cases, the abolition of the requirement would reduce the standard of police investigations. It is, after all, not enough for the police to find some evidence. Rather, they need to find evidence that will be sufficient to convince first the prosecutor to proceed and secondly a judge or jury to convict.
7.2.42 A second, quite separate, argument is that abolishing the requirement may prevent miscarriages of justice occurring, if by that it is also meant that convictions could be achieved in cases where the evidence of perpetration or identification comes from only one witness. If there is a single witness to an offence, and that witness is credible and reliable, why should the suspect not be prosecuted? If the judge or jury is satisfied beyond reasonable doubt of his/her guilt, why should a conviction not follow? A requirement that prevents such a conviction is creating an injustice. It is preventing a person, who may in fact be guilty of a serious or a minor offence, from being prosecuted and convicted. In the modern world, judges and juries ought to be regarded as quite capable of deciding what weight to give to a witness's evidence. An absence of corroboration, especially in circumstances where it would be unlikely that there could ever have been corroboration, should not prevent a judge or jury from deciding that the evidence of a complainer or other witness is believable and sufficient to establish guilt.
7.2.43 Corroboration concerns the quantity and not the quality of testimony. Yet, there are many circumstances where the testimony of a single witness is much more persuasive than a multiplicity of witnesses. The obvious example is when the judge or jury is considering the testimony of a single independent or impartial eye-witness, whose character cannot be impugned, as distinct from listening to members of a rival gang in a street fight or feuding neighbours in a public order debacle.
7.2.44 A third argument is that requirement for corroboration is frequently misunderstood by lay persons and lawyers, not least judges. It is an artificial bar to prosecution and conviction because of its restrictive method of looking at the quantity of the evidence without reference to its quality. The system is skewed by prioritising quantity over quality.
7.2.45 As outlined above, elaborate legal theories, unique to Scotland, have been devised over recent years in an attempt to fit an archaic requirement into today's reality. These have included the division of facts into different categories, viz.: crucial or essential, evidential and procedural so that only the former require to be corroborated. They include the idea that distress is corroborative of the use of force or a lack of consent in sexual offences. This is not only difficult to explain clearly to a jury in the context of a real case, it is highly doubtful whether a jury, even with its collective intelligence, can fully grasp the legal nuances of the judge's directions.
7.2.46 The application of the principle in Moorov is yet another example of where the law has become stretched. It becomes highly artificial where the events are years apart . Furthermore, especially where there are several different incidents involving different complainers and, quite possibly, different categories of crime over prolonged periods, it is very difficult to give a jury directions which are both comprehensible and comprehensive. This is particularly so where an acquittal on one or more charges must result, as a matter of law, in acquittal on some but not all others.
7.2.47 Different judges have different approaches to corroboration. This problem may have become particularly acute since the abolition of corroboration in civil cases. Since most judges and sheriffs are appointed after many years of practice as advocates or solicitors, they may, if they practised solely in the civil law, be applying a rule of law of which they have had little or no experience and little knowledge at all beyond the content of long forgotten lectures at University or newly read academic texts and precedents.
7.2.48 It is by no means apparent that the requirement provides any more consistency than an alternative approach, based on quality, would bring. Although corroboration can provide an objective minimum "baseline" for judges considering whether to sustain a "no case to answer" submission, at the risk of unnecessary repetition, it remains the position that judges can have very different views on what constitutes corroboration or sufficiency in general in a particular case .
7.2.49 Corroboration is more likely to exist in relation to some offences than others. This is exactly the same consideration as prompted reform of the civil law. Where crimes are usually committed in private, the only potential evidence may be from the testimony of a complainer. There may be nothing else, or very little, in the absence of statements made by suspects at interview. Equally, with minor assaults or thefts, where there may also be little evidence other than that of the complainer, it may simply not be cost effective to put the level of effort into investigating such cases, and thereby finding corroboration, as is appropriate to more serious offences.
7.2.50 Finally, it is of some note to recognise the role which the requirement of corroboration has in the advice given to suspects. There is little doubt that in Scotland it plays a major part in the solicitor's decision to advise his/her client to say nothing for fear of his/her inadvertently corroborating other evidence and thereby creating a sufficiency, which would otherwise not exist. The situation is, in all cases and in itself, an oddity because it may mean in practice that whether a person is prosecuted for and convicted of an offence, which would be inevitable in other jurisdictions, is actually entirely dependent on whether he/she elects to respond to questioning. Be that as it may, the advice tendered can place a suspect in a difficult position. It may be felt that a judge or jury would be more likely to accept his/her account as credible if it were raised at the earliest opportunity. Yet, the suspect would almost always be well advised not to speak, at least in situations where there was no obvious sufficiency of evidence.
Corroboration and the civil law
7.2.51 Until relatively recently, the requirement for corroboration was an equally significant cornerstone of the civil law. It was one of many rules of evidence, including the requirement of proof by writ or oath, hearsay and various prescriptions concerning the need for writing, which have almost all been swept away in recent years. Its demise began with its abolition in personal injuries litigation . In that class of case, the requirement came to be regarded, at least by some, as a bar to justice, especially in cases where an employee had been alone, or in darkness, when he/she suffered his/her injury . Abolition was met with considerable resistance within the senior judiciary . This took the form of pronouncements concerning the importance of corroboration in the establishment of fact, despite the statutory abolition. Thus, it was said, a fact could only be found on the basis of uncorroborated testimony if it were demonstrated that no corroboration could have been made available .
7.2.52 Abolition was ultimately, and this time successfully, extended to all civil cases in 1988 . At that time, the Civil Evidence (Scotland) Act eradicated many of the other prescriptive rules and the Requirements of Writing (Scotland) Act 1995 abolished several others. The Scottish Law Commission had expressed two "guiding principles" which it considered ought to be followed in relation to the reform of the law of evidence :
"(1) The law should be simplified to the greatest degree consistent with the proper functioning of a law of evidence;
(2) As a general rule all evidence should be admissible unless there is good reason for it to be treated as inadmissible".
7.2.53 This view is consistent with that of this Review and with modern thinking that it is a hindrance, rather than an advantage, to have to apply rules which prohibit a judge or jury from reaching a just conclusion based upon a liberal consideration of all relevant testimony. Were that thinking to be transferred to criminal cases, it would amount to an endorsement of Professor Wilkinson's opinion that the need for corroboration :
"is at odds with the rejection of other safeguards formerly employed and goes against the modern emphasis on the free assessment of evidence unencumbered by restrictive rules".
7.2.54 It is worthy of comment that the ultimate abolition of the requirement for corroboration in all civil cases, along with the general admission of hearsay evidence quantum valeat, has not met with any substantial adverse comment.
7.2.55 The Review is in no doubt that the requirement of corroboration should be entirely abolished for all categories of crime. It is an archaic rule that has no place in a modern legal system where judges and juries should be free to consider all relevant evidence and to answer the single question of whether they are satisfied beyond reasonable doubt that the accused person committed the offence libelled. The argument is not that such a reform would bring Scotland into line with the rest of Europe and the Commonwealth. It is that it would bring Scots law into line with modern, and almost universal, thinking on how to approach evidence in criminal, and indeed all other, cases.
7.2.56 The particular arguments in favour of abolition are set out above and it is not necessary to repeat them. However, there are certain highlights. First, in practice, there is no evidence or even anecdote to support the idea that the formal requirement for corroboration reduces miscarriages of justice. In particular, there is nothing to suggest that Scotland has a lower miscarriage of justice rate than any other jurisdiction in the civilised world. Secondly, the requirement creates miscarriages of justice by preventing cases, where there is only one witness to speak to the crime being committed or the suspect committing it, from being prosecuted. This is unjust. If a person is the victim of a criminal act and is capable of identifying the perpetrator, that person's case should be judged on the quality of the testimony and its progress should not depend on a formal requirement, compliance with which is often matter of chance. Similarly, it may be that the complainer can speak to none of these matters by reason of the effects of the crime upon him/her, yet there is a single credible and reliable witness who can. A system of justice, which does not permit the prosecution of the alleged offender in such circumstances is, putting the matter mildly, an imperfect one. Thirdly, the requirement has developed into a series of rules which, realistically, are not capable of being understood by many outside the world of criminal legal practice. They are applied differently by courts, depending upon their own experience in that practice.
7.2.57 If corroboration were to be abolished, it might reasonably be anticipated that there will be some judicial resistance in line with that experienced when it ceased to be a requirement in personal injuries cases. This may take the form of the court taking the view that, in certain categories of case, or in all cases where there is no corroboration, a special warning requires to be given to juries about the dangers of conviction. Such a warning would be similar to that currently given in cases of eye witness identification. It will be important, following especially the experience in the other jurisdictions referred to, to make it clear that, although a trial judge may, at his/her discretion in a particular case, give a jury such assistance, by way of warning or otherwise, as is appropriate in relation to the assessment of the credibility and reliability of witnesses , the law does not require that a warning be given in any case simply on the basis that there is a lack of corroboration.
I therefore recommend that:
- the current requirement for corroboration in criminal cases be abolished; and
- in solemn prosecutions where there is no corroboration of testimony, there should be no requirement on the judge to warn the jury of any dangers perceived purely as a consequence of the absence of such corroboration.
 see Crowe: A case for the abolition of corroboration in criminal cases 2011 SLT (news) 179
Morton v HM Advocate 1938 JC 50, LJ-C (Aitchison) at 55; MacPhail: "Evidence", Stair Memorial Encyclopaedia para 766
 Hume ii, 383
 Criminal Law (1811) p 509; quoted by the LJG (Rodger) in Fox v HM Advocate 1998 JC 94 at 98
 now Sir Gerald Gordon
 Gordon: "At the mouth of two witnesses…" ( supra) p 35
 e.g. eye witness evidence identifying the accused as the perpetrator of the offence
 otherwise known as "indirect", i.e. evidence of a fact (e.g. fingerprint) or facts from which another fact (e.g. presence of accused at the scene) may be inferred
 Walker & Walker: Evidence (1st ed) para 380, p 402 et seq; (3rd ed) para 5.2.2
Spendiff v HM Advocate 2005 JC 338 analysing Lord Advocate's Reference (No 1 of 2001) 2002 SCCR 435; Mackintosh v HM Advocate 2010 SCCR 168; see generally Chalmers: Distress as corroboration of Mens Rea 2004 SLT (news) 141
Adamson v HM Advocate 2011 SCCR 271, although this now refers to "corroborated proof of distress", see LJG (Hamilton) at para 26
Campbell v Vannet 1998 SCCR 207; it may be different if proof of the use of a weapon is required
 e.g. Road Traffic Offenders Act 1988 s 21, parking and related offences
 1941 SC 363
 at 379
 1995 SLT 110, LJG (Hope) at 119
 1998 JC 94, LJG (Rodger) at 100-101
Ralston v HM Advocate 1987 SCCR 467
 " de recenti"
 e.g. scenes of crime officers
 e.g. police officers or a combination of doctor and police officer
 e.g. fingerprint experts or forensic scientists
 dual testimony is not normally required; 1995 Act ss 280-281
Moorov v HM Advocate 1930 JC 68
Howden v HM Advocate 2009 JC 308 and Townsley v Lees 1996 SLT 1182
 see The Scottish Law Commission "Similar Fact Evidence and the Moorov Doctrine"
Dodds v HM Advocate 2003 JC 8
 see even in the last two years: M v HM Advocate  HCJAC 6 doubting Cannell v HM Advocate 2009 SCCR 207; Pringle v PF (Dumbarton) 2011 SCCR 97; ANM v HM Advocate 2011 SCCR 47; Hussain v HM Advocate 2010 SCCR 124 and CAB v HM Advocate 2009 SCCR 216
 as the law stood pre Lord Advocate's Reference (No 1 of 2001) 2002 SCCR 435
Yates v HM Advocate 1990 JC 378n, 1977 SLT (notes) 42; Smith v Lees 1997 JC 73
 Which a de recenti statement cannot
Smith v Lees ( supra)
 cf the Romano-canonical law supra
Manuel v HM Advocate 1958 JC 41 following Alison ( supra), p 580
Hartley v HM Advocate 1979 SLT 26, Lord Dunpark at 33; see also the dicta in O'Hara v Central SMT and Fox v HM Advocate (supra)
McAvoy v HM Advocate 1982 SCCR 263
 e.g. Edwards and Lewis v United Kingdom (2005) 40 EHRR 24 on entrapment and disclosure
N v HM Advocate 2003 JC 140
Gafgen at para 164
 Perjury Act 1911 s 13
 Road Traffic Regulation Act 1984 s 89; cf Gillespie v McMillan ( supra)
 Plucknett: Concise History of the Common Law p 387
 a "care warning", or what would be described as a " cum nota" warning in Scotland
 whose testimony would not have been on oath
 whether male or female
 see generally Archbold 2011 para 4-404b; Blackstone 2011 para F5.1 et seq
 Criminal Justice and Public Order Act 1994 s 32
 see recently in India : State of Madhya Pradesh v Ramesh and Anr, Supreme Court, 18 March 2011, Dr BS Chaunun J at paras 11 and 13 on the lack of any rule of law requiring corroboration of a child's evidence
 Law Commission Report: Corroboration of Evidence in Criminal Trials (Cmnd 1620) (1991)
 see e.g. R v Spencer  1 AC 128, Lord Ackner at 141; R v Hester  AC 296, Lord Morris at 315
 Ashworth: "Corroboration": in place of formalism  Crim L R 1
R v Muncaster  Crim LR 409 and R v Makanjuola  1 WLR 1348, Lord Taylor CJ at 1351
 Australia's Criminal Procedure Act 1986 s 294AA; Canada: Criminal Code RSC 1985 s 274
 this is referred to as a "Murray" direction, R v Murray (1987) 11 NSWLR 12
 McGrath: Evidence para 4-02
 (1945) 323 US 606 at 608
 Criminal Law Amendment Act 1885 s 2 (as amended)
 Road Traffic Act 1961 s 105
Dental Board v O'Callaghan  IR 181, Butler J at 183
 McGrath: Evidence at para 4-18 to 4-109
 McGrath: Evidence at para 4-112 under reference to Temkin: "Towards a Modern Law of Rape" (1982) 45 MLR 399
 Criminal Law (Rape) (Amendment) Act 1990 s 7
 McGrath: Evidence at para 4-130 under reference to the Report of the Committee on Sexual Offences Against Children and Youths published in Canada (1984)
 Criminal Evidence Act 1992 s 28
 see Crowe, supra
 see para 7.1.2
 Identification Procedure under Scottish Criminal Law (the Bryden Report) Cmnd 7096 (1978)
 Macphail: Revised Research Paper on Evidence para 23.02 under reference to "Anon: "Corroboration of Evidence in Scottish Criminal Law" 1958 SLT (news) 137
DPP v Hester  AC 296 at 315
AK v HM Advocate  HCJAC 52
 hence the introduction of Crown appeals in the Criminal Procedure (Scotland) Act 1995 ss 97A et seq and 107A et seq following the "World's End" murder trial ( HM Advocate v Sinclair, unreported, High Court, 10 September 2007; on the criticism of which see Di Rollo: "Legitimate Inference v Illegitimate Speculation" 2008 SCL 151) and Mackintosh v HM Advocate 2010 SCCR 168 and O'Hara v HM Advocate 2009 SCCR 629
 Law Reform (Miscellaneous Provisions) (Scotland) Act 1968 s 9
 Scottish Law Commission: "Proposals for Reform of the Law of Evidence Relating to Corroboration" para 5; this paper recommended abolition in all civil cases but that recommendation was not then adopted
Morrison v J Kelly and Sons 1970 SC 65; See Macphail: Revised Research Paper on Evidence para 23.13 et seq
McGowan v Lord Advocate 1972 SC 68; cf McLaren v Caldwell's Paper Mill Co 1973 SLT 158
 Civil Evidence (Scotland) Act 1988 s 1
 100th Report para 1.3
 Wilkinson: Evidence 204 referring to the precept " testimonia ponderanda non numeranda sunt"; and see Gordon: "At the mouth of two witnesses…" ( supra) p 39
 see Practice Note, 18 February 1977, issued by LJG (Emslie)