2.1 By taking into account the defining characteristics of the digital world, the use and admissibility of evidence in digital form have largely been accomplished through the definition and redefinition of legal concepts in the malleable rules of evidence. This chapter sets out to review the rules of evidence in the categorization of, means of proof of, treatment of and weight given to electronic evidence. An overview of the issues of hearsay, the treatment of software code as the witness, the presumption that computers are ‘reliable’ and authentication of electronic evidence are covered in detail in other chapters.

Direct and indirect evidence

2.2 The purpose of adducing evidence in legal proceedings is to prove the facts in issue. Facts in issue can also be properly inferred from any facts that are presented. Where evidence is used to prove the facts in issue, it is direct evidence. Where evidence is used to prove facts from which the facts in issue may be inferred, it is indirect evidence. If the facts in issue involve proving the existence of an electronic record, the electronic record itself constitutes direct evidence. Direct evidence refers to evidence which proves the facts in issue, and indirect evidence, or circumstantial evidence, is defined as evidence which proves facts which are relevant to the facts in issue.

2.3 Unless the existence, character or circumstance of the generation or storage of an electronic record is itself a fact in issue, it is more frequently the case that electronic evidence is used as indirect evidence to prove certain facts from which the facts in issue may be inferred. For instance, if an electronic record is adduced in evidence to show that A owes B a debt, the electronic record as indirect evidence only proves that there is a record that A owes B a debt, and it is necessary to make the additional inference that A actually owes B a debt.

Evidence in both digital and analogue form

2.4 Although there are differences in form and format between the analogue or non-electronic version of an item of evidence and its electronic equivalent, if the differences are not material, courts will not reject electronic evidence in favour of other forms of evidence.

2.5 The differences may be material depending on the facts in issue: the alternative representations of data in digital form and in human-readable form, on a screen or printed piece of paper, may become significant. In the Tasmanian case of Maynard,¹ the trial magistrate declined to admit a printout purporting to indicate the dates and times when the accused obtained access to data stored in the computer on the basis that not all of the data that were evident on the computer screen were fully replicated on the printout. In a motion to review, Wright J upheld the magistrate’s decision. The judge observed that if all the data were relevant, the prosecution could have recorded the data on the screen by video. In this case, it was demonstrated that the information recorded on the printout was incomplete and not an accurate rendition of the data, and it did not involve only minor format changes, as the prosecution sought to contend.

2.6 In contrast, in People of the State of New York v Rose¹ Morse J in City Court, City of Rochester, New York had to consider the use and admissibility of ‘computer generated simplified traffic information tickets’ or ‘e-tickets’. The defendants moved for dismissal of the charges for driving while intoxicated because the State Police had issued the charges in the computer-generated simplified information form rather than the multi-copy handwritten simplified traffic information form used across New York State. In a carefully reasoned judgment, Morse J set out how the system worked, and determined that the computer terminal used by the police generated each e-ticket with simplified traffic information for the defendants, printed duplicate originals of the e-ticket and affixed the arresting officer’s electronic signature to the e-ticket. Although there were minor format differences, such as the colour and the number of sides on which the e-tickets were printed, these differences were not sufficient to persuade the judge that the e-tickets conformed substantially to a paper ticket. Thus, the motion for dismissal was denied.

2.7 A similar consideration arose in Griffiths v DPP,¹ where photographs taken with a speed camera on photographic film were admitted as evidence of a vehicle being driven at a speed greater than the speed limit. The evidence was also available in digital form, and the defence argued that the digital data should have been disclosed as well as the printed photographs. It was revealed that the camera technician had carried out a secondary check to confirm the speed of the vehicle on the digital files of the photographs. The judge indicated that the photographs were real evidence – they showed the times at which the vehicle crossed a number of pre-measured lines painted on the road – and that by using all this information it was possible to carry out the secondary check from the photographs themselves; it was not necessary to carry out the secondary check on the digital files. For this reason, it was held that whether the digital data were disclosed to the defendant or not was irrelevant.²

Metadata and electronic evidence

2.8 There is a distinction between a document in digital form (and the content of the digital document as a printout) and the metadata logically associated with the document in digital form. The metadata may be relevant, either as indirect evidence in relation to the document in digital form, or it may itself be relevant as direct evidence. For instance, when there are multiple versions of a digital document, the metadata as indirect evidence will enable the parties to identify the most relevant version of the document. On the other hand, where there is an allegation that the user manipulated the metadata of the file, such as its date–time stamp, to his own advantage, the correct date and time of the file becomes the fact in issue and the metadata is the direct evidence. In such a case the metadata may need to be rendered into human-readable form.

Means of proof

2.9 All direct and indirect evidence used to prove a fact in issue or a relevant fact takes one (or more) of the following forms: testimony, hearsay, documents and real evidence.

Testimony and hearsay

2.10 Testimony is the declaration (which must be admissible) in court of a person who actually perceived the fact in issue or facts from which facts in issue may properly be inferred. Thus the human perception of a computer display as narrated via oral testimony is admissible as evidence that a counterfeit computer game was being played in breach of copyright.¹

2.11 If, however, the best that a witness can do is to depose as to what someone else said on the fact in issue, it will be hearsay, because it is ‘an assertion other than one made by a person while giving oral evidence in the proceedings … as evidence of any fact asserted’.¹ In the context of digital evidence, what someone else said is typically recorded electronically. Hearsay is generally inadmissible unless it falls within one of the exceptions to the rule against hearsay. (A further treatment of this subject is found in Chapter 3.)

Real evidence

2.12 The term ‘real evidence’ has been used in three divergent senses: (i) evidence from things as distinct from persons; (ii) material objects produced for the inspection of the court; and (iii) perception by the court (or its results) as any perceptions by the court (or its results) as distinct from the facts perceived,¹ and is best described as ‘Material objects other than documents, produced for inspection of the court’.² Veronica Cowan considered that there is no authoritative definition of ‘real evidence’, and suggested that ‘where a document is tendered simply to prove the fact that a statement was made (and not to prove a fact stated therein), it is not properly described as “real evidence”’.³

2.13 To highlight the difference between real evidence and hearsay in electronic evidence, Professor Daniel Seng and Sriram S. Chakravarthi formulated the following categorizations: digital data that is stored on a device, a device that processes data, and a device that processes and stores data.¹ The digital data is hearsay, because the device contains a record of human assertions. As for the second and third cases, where the data is produced without human intervention, it is real evidence. If the data is a record of human assertions, it is hearsay. Although the distinction is a clear one, it can be difficult to apply in practice,² as the cases discussed below illustrate.

Evidence in analogue form

2.14 The treatment of evidence in analogue form (which preceded the use and acceptance of digital computers) first received detailed treatment in the case of R v Ali (Maqsud), R v Hussain (Ashiq),¹ where the issue was the admissibility of a tape recording. In admitting the evidence, Marshall J analogized tape recordings with photographs, and noted that just as evidence of things seen through telescopes or binoculars which otherwise could not be picked up by the naked eye had been admitted, the same would apply to devices for picking up, transmitting and recording conversations, noted, at [701] (emphasis added):

2.15 Shortly thereafter, Sir Jocelyn Simon P determined in The Statue of Liberty, Sapporo Maru M/S (Owners) v Steam Tanker Statue of Liberty (Owners)¹ that the film recording on a radar set of echoes of ships within its range was real evidence, even though it was recorded from a mechanical instrument.² The judge considered there was no distinction in the manual operation of a camera by a photographer or the observations of a barometer operator and the equivalent operation by a trip switch, a clock or a dial-recording mechanism. The judge held, at [196], that ‘[t]‌he law is bound these days to take cognisance of the fact that mechanical means replace human effort’, and accepted that the film comprised real evidence because it recorded the information given out by the radar set, rejecting the submission that the evidence was hearsay.

Evidence in digital form

2.16 The characterization of evidence as real evidence or as hearsay becomes more complicated with evidence in digital form, especially when some computational processing is carried out. In R v Pettigrew (Stewart Douglas), R v Newark (John)¹ the Court of Appeal held that the printout from a computer operated by an employee of the Bank of England was a hearsay statement. The operator fed bundles of bank notes with consecutive serial numbers into the machine, and the machine automatically rejected any notes that were defective. The machine recorded the first and last serial numbers of each bundle of 100 notes, and the operator also noted the first serial numbers in the bundle on a card as he fed them into the machine. It was the printout from this machine that was sought to be admitted in evidence. The purpose of adducing the evidence was to permit the prosecution to trace the issuance of the notes, and to link bank notes found in the possession of Pettigrew to a particular bundle of notes that had been stolen in a burglary. Counsel for the prosecution argued that the printout was admissible as a business record under the provisions of the Criminal Evidence Act 1965.² However, s 1(1)(a) required that for such a record to be admissible as evidence of the truth of any matter dealt with in the record, the information would have to be supplied by a person who had, or may reasonably be supposed to have, personal knowledge of the matters. The members of the Court of Appeal reached the conclusion that the operator did not have personal knowledge of the numbers of the notes that were rejected, because the machine automatically compiled the list.

2.17 While this was an accurate application of the hearsay rule, the analysis omitted any consideration that the printout might be considered real evidence.¹ Professor J. C. Smith has noted that ‘the operator had personal knowledge of the first number of each bundle which he fed into the machine because he recorded that number on a card’,² and suggested that because the operator had knowledge of the number at a given point in time, it was not material that he forgot it afterwards. Once the first number could be established, it could then be inferred that the new notes bore consecutive serial numbers.³ Professor Smith considered that this is not hearsay but direct evidence, because there was an absence of human intervention.⁴ On the other hand, Professor Colin Tapper took the view that the printout was partly hearsay and partly non-hearsay – the first number is hearsay and the last number and the numbers of the notes that were rejected were not hearsay because they were the output of the device.⁵

2.18 Professor Seng considered that the views of Professors Smith and Tapper were both plausible: ‘The difference lies in whether the operator fed the first number into the machine, and whether the machine processed this number.’¹ Seng continued:

2.19 As computers are designed to undertake a wide range of tasks, this means that the evidence available as output of a computer is equally as varied. A review of cases shows that whether electronic evidence is real evidence or hearsay turns on whether the evidence is characterized as being a product of a device’s processing functions or of its storage functions.

2.20 In R v Wood (Stanley William),¹ the software in a computer that made calculations of chemical compositions based on spectrometer readings was considered to be a tool, and the printout was an item of real evidence. The basis of admitting a printout of an output as an item of real evidence was explained by Professor Tapper:

2.21 The same distinction was drawn by Professor Smith as regards the computer printout in R v Ewing (Terence Patrick),¹ between its use as evidence to prove that a thing was done (money had been credited to a bank account) and as evidence that something was recorded as being done (the bank clerk records a payment, as opposed to creating the credit).²

2.22 The admissibility of more complex electronic evidence is illustrated in a case about the breath alcohol printout from a portable breath analyser device, the Intoximeter 3000. In Castle v Cross,¹ it was determined that the printout was an item of real evidence and not hearsay.² The judge compared the device to a speedometer, a calculator or a sophisticated tool. In this instance, the breath alcohol value in the printout comprised information that was produced by the Intoximeter because the data had not passed through a human mind. On the other hand, Kennedy J also remarked that ‘where a computer is used in respect of its memory function, it is possible to envisage where it might fall foul of the rule against hearsay’.³

2.23 In R v Spiby (John Eric),¹ Taylor LJ held that there was a distinction between a printout as real evidence and as hearsay. Professor Smith² noted the difference between the content of a printout as a mere recording of a fact, such as when data are processed by a computer without any human input of any description,³ and the content of a printout as being processed in some way by a human being. The printout in this case was a log of telephone calls made, which was generated by a computerized machine called a ‘Norex’ that monitored the telephone calls of hotel guests in order to work out how much to charge for telephone use. It was held to be real evidence.

2.24 In R v Robson (Kenneth), R v Mitchell (Bernard), R v Richards (Alan),¹ a printout of telephone calls made on a mobile telephone was adduced as evidence of the calls made and received in association with the number. The defence’s challenge that the evidence was documentary hearsay failed. Orde J held that ‘where a machine observes a fact and records it, that record states a fact. It is evidence of what the machine recorded and this was printed out … The record was not the fact, but evidence of the fact.’²

2.25 In the business context, two popular uses of computers are for the formation of records and the recording of the credits and debits of an account. In the latter, the records of computer payment transactions are considered real evidence, as Their Lordships made clear in R v Governor of Brixton Prison, ex p Levin.¹ In this appeal against extradition, it was alleged that Levin had used a computer terminal in St Petersburg to gain unauthorized access to a Citibank terminal in Parsipanny, New Jersey to make 40 fraudulent transfers of funds from the accounts of clients of the bank to accounts which he or his associates controlled. Printouts of screen displays of the historical records of computer payment transactions were adduced, and a witness gave evidence as to how the records were created. Lord Hoffmann took the opportunity to make clear the difference between a hearsay statement and evidence of a record of a transaction:

Documents and disclosure or discovery

2.26 In evidentiary discovery (or disclosure as it is called in England and Wales) a ‘document’ has been construed widely in both criminal and civil proceedings.¹ While the emphasis is on the recording of the content by the application of (usually) text on to (usually) paper, early decisions such as the Court of Appeal in Lyell v Kennedy (No. 3)² have admitted photographs of tombstones and houses as documents for the purposes of discovery. In R v Daye (Arthur John),³ Darling J suggested, at [340], that the meaning of ‘document’ should not be defined in a narrow way:

2.27 In Hill v R,¹ Humphreys J held, at [332–333], ‘that a document must be something which teaches you something … To constitute a document, the form which it takes seems to me to be immaterial; it may be anything on which the information is written or inscribed – paper, parchment, stone or metal.’ Likewise, statutes adopt a similarly broad definition of a ‘document’. Section 13 of the Civil Evidence Act 1995 defines a ‘document’ as ‘anything in which information of any description is recorded’. The same definition is provided in s 20D(3) of the Taxes Management Act 1970. Schedule 1 of the Interpretation Act 1978 does not define ‘document’, but defines writing as follows:

2.28 Audio tapes were accepted by Walton J as a discoverable document in Grant v Southwestern and Country Properties Ltd,¹ where a ‘document’ was defined as having the quality of conveying information. Television film is also considered a document,² as are the output of facsimile transmissions³ and a label on a bottle containing a specimen of blood provided by the accused.⁴

2.29 In Derby & Co Ltd v Weldon (No. 9),¹ one of the earliest modern decisions on the point, it was held that data stored on a computer in the form of an online database constitutes a document for the purposes of the obligation to discover under the provisions of Order 24 of the Rules of the Supreme Court. In analysing this point, Vinelott J referred to the Australian case of Beneficial Finance Corp Co Ltd v Conway,² in which McInerney J held that a tape recording was not a document because the information is not capable of being visually inspected. Vinelott J, however, preferred the opposing view in Grant v Southwestern and County Properties Ltd,³ in which Walton J pointed out that there is no difference between recording a conversation on a tape recorder and in shorthand. Both are methods of recording the same conversation. Vinelott J quoted Walton J with approval as follows:

2.30 Thus the interposition of a computer to enable the retrieval of data stored in the online database did not disqualify the data from being considered a document. A similar issue for discovery as to the meaning of ‘document’ in the context of data stored on a computer was discussed in Alliance & Leicester Building Society v Ghahremani¹ on a motion to commit Naresh Chopra, a solicitor, to prison for contempt of court. Mr Chopra was alleged to have deliberately deleted part of a file that showed crucial transaction details stored on his computer in contempt of court, when investigations into possible mortgage fraud and negligence were being conducted into his affairs. A court order had directed Chopra to restrain from destroying or altering any document relating to the transaction, and required him to deliver up all such documents in his control. In the contempt proceedings, counsel argued that the word ‘document’ required there to be some form of visible writing on paper or other material, and because there was no physical document, the order had not been breached. Hoffmann J noted the comments of Vinelott J in Derby & Co Ltd v Weldon (No. 9),² and held that ‘document’ would bear the same meaning in the discovery order. Taking into account the expert evidence, Hoffmann J concluded that it was proved beyond reasonable doubt that Chopra did alter or destroy part of a document³ and granted the motion, although Chopra was eventually fined instead.⁴

2.31 There is judicial recognition that the acceptance and use of technology will increase the range of objects that fall within the definition of ‘document’. In R v McMullen,¹ Linden J held that a current account ledger card printed from a computer was a document within the meaning of s 29(2) of the Canada Evidence Act. The judge commented that: ‘It is merely a new type of copy made from a new type of record. Though the technology changes, the underlying principles are the same.’² On appeal, citing this comment, Morden JA observed that the ‘section should be considered as “always speaking” and “be applied to the circumstances as they arise”’.³ The same view was emphasized by Buxton LJ in Victor Chandler International v Customs and Excise Commissioners,⁴ where he observed, at [55], that ‘the word “document” is not constrained by the physical nature that documents took in 1952, so we are entitled, and indeed bound, to consider the appropriate application of the concept of circulation, etc, of a document in the light of current practice and technology’. In this case, an advertisement contained in a teletext transmission was held to be a document for the purposes of the Betting and Gaming Act 1981. This view was reinforced by Pumfrey J in Marlton v Tectronix UK Holdings,⁵ when the judge held that a computer database, in as far as it forms part of the business records of a company, is a document for the purposes of the Civil Procedure Rules, and therefore can be disclosed. Calvert Smith J also concluded, in Kennedy v Information Commissioner,⁶ that the word ‘document’ in s 32 of the Freedom of Information Act 2000 included information recorded in an electronic medium. The judge said, at [79]:

2.32 As such, a ‘document’ is a medium upon which information is stored. The medium may sometimes determine the admissibility of the evidence, but the definition of ‘document’ is considered wide enough to bring any medium into its ambit without causing difficulties.¹ This must be correct, because if information is not stored on a medium, the content is not available without the medium, and therefore the information remains oral evidence. As Lord Milligan in Rollo (William) v HM Advocate² said, when he indicated that the information stored in a Sharp Memomaster 500 hand-held device was a document:

Visual reading of a document

2.33 Although the meaning of ‘document’ has been construed widely, nevertheless it was held by the court in Darby (Yvonne Beatrice) v DPP¹ that a visual reading cannot be a document. This must be correct. Unless the reading is stored in some way that enables it to be read at a later date, the reading is merely a transitory phenomenon that can only be captured by a person who provides original testimony by giving evidence about her perception.²

2.34 Oral testimony may be provided in lieu of documentary evidence. In a number of breath specimen cases, the defendants’ counsel have submitted that it is necessary to provide the printout as documentary evidence of the output recorded by the machine, and that substitute evidence given by a police officer as to the machine output is not admissible.¹ In Thom v DPP,² the printout from an Intoximeter was not produced, and the defence objected to testimony by a police officer as to what he had seen on the printout. Clarke J addressed this point as follows, at [14G]:

2.35 Likewise, in Sneyd v DPP,¹ when the printout from an Intoximeter was not produced, the court accepted the police officer’s testimony of what he had seen on the printout provided by the device, rather than what he had seen on the screen. Rejecting the objection on the basis that the testimony was secondary evidence, Richards LJ held, at [32], that ‘it is well established that evidence both as to the results of the analysis and as to the reliability of the machine can be given either in the form of a written print-out or orally by the officer who carried out the procedure’. He determined that there was no difference between the oral evidence of the results shown on the printout and oral evidence of the results on the screen of the machine – both were not inadmissible hearsay. In R (on the application of Leong) v DPP,² Silber J applied the analysis of Richards LJ, holding admissible the oral evidence of the police officer’s reading from a printout, at [14]: ‘Where, as in the present case, there is evidence that the machine is working properly, there is no reason why the police officer concerned cannot give admissible evidence of what he saw in the print-out.’

Authentication

2.36 When a document is tendered as evidence of its contents, it is often accompanied by proof that the document ‘has some specific connection to a person or organization, whether through authorship or some other relation’.¹ As noted by Austin J: ‘Authentication is about showing that the document is what it is claimed to be, not about assessing, at the point of the adducing of the evidence, whether the document proves what the tendering party claims it proves.’² Similarly, where any object is tendered in evidence, an adequate foundation for admission will require testimony first that the object offered is the object which was involved in the incident, and further that the condition of the object is substantially unchanged.³

2.37 As is the case for any other form of evidence, electronic evidence must be authenticated. The authentication process for electronic evidence is even more critical,¹ and can occasionally be challenging.² Undoubtedly the use of technology has afforded us convenience and efficiency. But if parties and investigative authorities choose to use the fruits of technology, they must also accept the need to prove the authenticity and integrity of the evidence produced by technology, even though the cost of such proof might be considered to be high. This is particularly the case where authentication evidence will shed light on the latent assumptions and hidden errors inherent in electronic evidence, which could affect the accuracy of the electronic evidence itself.

2.38 Authentication evidence may also demonstrate that the errors in question will not have an adverse effect on the evidence itself. For instance, in DPP v McKeown (Sharon), DPP v Jones (Christopher),¹ the clocks on the Intoximeter 3000 used to measure the breath alcohol values of the defendants were not accurate. For this reason, the defendants challenged the admissibility of the printouts from the device. In addressing whether the accuracy of the clocks was relevant to the accuracy of the printout readings, Lord Hoffmann examined the functioning of these devices and concluded that, for the purposes of s 69 of the Police and Criminal Evidence Act 1984,² a malfunction is irrelevant unless it affects the way in which the computer processes, stores or retrieves the information used to generate the statement.³ On the facts, the clock was not part of the processing mechanism of the Intoximeter, and the convictions of the defendants based on the printout readings were upheld.

2.39 This does not mean that authentication evidence will always have to be supplied for each item of evidence. In civil proceedings in England and Wales, a party is deemed to admit the authenticity of a document disclosed under the provisions of Civil Procedure Rule (CPR) 31 unless notice is served that the party wishes the document to be proved at trial. Thus where the authenticity of a document is questioned, the party raising the issue is required to do so at an early stage of the proceedings, thereby providing the party submitting the document the opportunity of gathering evidence to prove the veracity of the document.

2.40 See Chapter 6 on authentication for a more detailed discussion.

Best evidence

2.41 The best evidence rule can be considered from two points of view. It can be regarded as an inclusionary rule under which whatever is the best evidence is admissible, thus overcoming exclusionary rules such as the hearsay rule; alternatively, it can be regarded as an exclusionary rule, so that anything which is not the best evidence is inadmissible. Since Omychund v Barker,¹ the majority of the cases have used the rule in an exclusionary way to deny the use of copies of documents when the absence of the original was not satisfactorily accounted for.

2.42 Reaction against this rule began in the nineteenth century,¹ and by the latter part of the twentieth century it was recognized that the best evidence rule was no longer as relevant as it once was. In Kajala v Noble,² Ackner LJ held that the rule is now confined to written documents in the strictest sense of the term. Echoing the robust comments of Lord Denning MR in Garton v Hunter (Valuation Officer),³ His Lordship said:

2.43 By 1990, Lloyd LJ in R v Governor Ex p Osman (No 1) observed that the best evidence rule had become a rule of practice or procedure.¹ He also made the following remarks about the rule:

2.44 The best evidence rule has been effectively limited to requiring a party having possession of an original document who is relying on it for the statements recorded on the document (primary evidence) not to wilfully refuse to produce the original document as primary evidence, and instead produce copies or substitutes (secondary evidence) in its place.¹

2.45 Where good reasons exist for the failure to produce the original document, secondary evidence, even in the form of oral testimony, is permissible. This may be illustrated by the case of Taylor v Chief Constable of Cheshire,¹ a case involving the inadvertent destruction of evidence. In this case, video images of the accused allegedly committing theft from a store were recorded on the store video recorder, and the manager of the store, three police officers and the lawyer for the accused later saw these recordings. When the case was heard, it transpired that new security officers had erased the recording of the video images. The magistrates permitted the witnesses to give evidence of what they saw on the video recording. An appeal was made that the best evidence – the video recording – could not be admitted because it had been destroyed, and that testimonial evidence of the recording was not the best evidence. This was rejected. Although the best evidence in this instance was the video recording, its unavailability did not preclude the admission into evidence of the testimony of those witnesses who viewed the recording. The recollections of the witnesses ought not to be precluded because the best evidence was not available. The evidence offered by the witnesses was, as pointed out by Ralph Gibson LJ, ‘direct evidence of what was seen to be happening in a particular place at a particular time’, and it was for the trier of the facts to assess its weight, credibility and reliability.²

2.46 Since the statutory intercession of the Civil Evidence Act 1995 and the Criminal Justice Act 2003, the best evidence rule has further taken a simplified, statutory form. The judgment of the Court of Appeal in Masquerade Music Ltd v Springsteen¹ suggests that the best evidence rule is hardly of any relevance. After considering the best evidence rule in detail and reviewing the case law extensively,² Jonathan Parker LJ outlined the position with respect to the best evidence rule in the twenty-first century, at [85]:

2.47 Waller and Laws LJJ concurred. In other words, there is no automatic bar to the failure to admit the original document as primary evidence. Instead, when the original document is no longer available, a copy of the original evidence is admissible, but an adjudicator must consider its weight as secondary evidence.

2.48 The modern application of this rule is illustrated by Post Office Counters Ltd v Mahida.¹ In this case, the Post Office sought to claim an alleged deficiency of social security benefits paid out against the defendant, the sub-postmaster general. The deficiency was set out in a schedule prepared by investigators of the Post Office based on checks conducted against the underlying dockets and foils. Subsequently, the dockets and foils were destroyed as part of a routine process. The trial judge accepted the schedule as secondary evidence and found against the defendant. On appeal, the Court of Appeal was concerned that the secondary evidence was of insufficient weight to prove the precise amount of the debt claimed against the defendant. In particular, the Post Office as an institution could not readily be said to have discharged the burden of proving the precise amount of the debt when it was alleged that the defendant had been responsible for this loss, and had denied the defendant the opportunity to check those figures.² For this reason, the very basic unfairness should have led the trial judge to consider that the amount of the debt was not proved, and the defendant’s appeal was allowed.

Analogue evidence

2.49 Although the best evidence rule is now tightly confined, it applies to both civil and criminal proceedings.¹ But as the statutory formulations of the rule in s 8 of the Civil Evidence Act 1995 and s 133 of the Criminal Justice Act 2003 retain the differentiation between primary and secondary evidence, the ramifications are different, depending on whether the evidence is in analogue or in electronic form.

2.50 In the physical world, primary evidence is an original document, and secondary evidence is in the form of copies of the original. The best evidence rule will require the production of the original document to prove the content in question, and the submission of copies is considered inferior evidence. But the fact that copies were made, for instance, by a reprographic process such as photocopying, will not prevent the copies themselves from being originals. In Miller-Foulds v Secretary of State for Constitutional Affairs¹ regarding orders issued by Brentford County Court, Pelling J noted the following, at [26]:

2.51 The concepts of ‘primary’ and ‘secondary’ evidence take a different shape when applied to material objects that must be processed to be viewed. Consider, for instance, a photograph taken with a camera containing film, or a plate. The negative or the plate comprises the only copy of the image in reverse.¹ It is the negative or plate that is the material upon which the primary evidence is recorded. However, few people will be satisfied by looking at the primary image, if only because it is not easy to view, and is not intended to be viewed in this form unless by means of a projector, if the primary image is a negative. This means that the printed image is secondary evidence. Any number of copies of the primary object can be made, although no printed copy will be an exact copy of the film or plate. This is because the processes applied and the mix of chemicals used in transforming the negative into a print will determine how accurately the photograph reflects the image, in particular the degree of contrast (that is the range of grey tones) captured on the negative. For example, the degree of contrast will affect how bruising is reproduced on the photograph: a high contrast makes the bruising appear darker and more dramatic, while a low contrast will lessen the effect of the visual image, making the bruise seem somewhat less consequential.

Digital evidence

2.52 In contrast to the discussion above, the range of evidence in digital form is vast, and it comprises not just printouts of what might be termed conventional files, such as copies of letters, contracts or spreadsheets. Other forms of digital documents include reports from computer databases, the electronic records of transactions and the digital store and reproduction of images, such as the scanned image of an original paper document. The treatment of evidence in digital form calls for different and occasionally difficult considerations.

2.53 First, there may be issues identifying the primary evidence of a digital document. In Derby & Co Ltd v Weldon (No. 9), Vinelott J considered the memory or database of a word processor or computer to be the ‘original document’,¹ presumably on the basis that these are components ‘on which material fed into a simple word processor is stored’.² However, Professor Tapper disagrees, and takes the view that the printout from the word-processed electronic document is the original and the document in computer memory is the copy.³ Both views are possible. Vinelott J’s analysis is plausible – where the printout is generated as a physical draft to aid in the editing of the word-processed document. But Professor Tapper’s view could also be justified where the object behind the use of the word processor is the generation of the printout as the final, definitive version of the document. In such a case, the authentic printout may be a better form of evidence than the state of the document in internal memory at a later time. This inversion provides a good illustration of the danger of assuming that the printout may not be the best evidence in any given situation.

2.54 In addition, the use of a digital device need not always produce an ‘original document’. Where the ‘original document’ is created in digital form but is never stored in a more permanent, non-ephemeral manner, the ‘original’ digital ‘document’ ceases to exist for all practical purposes. Instant messaging is an example of evidence that might not be stored, which makes it analogous to an oral conversation.

2.55 The issues may be further considered with the following extended illustration. For instance, the original of a physical document, such as a commercial contract between two parties, signed by the authorized representatives of both parties and acknowledged as the original, is primary evidence of the content of the contract. Even if the contract was created on a computer, the physical document will still be the original document as it was signed and adopted by both parties.¹ However, should the contract, which is subsequently acted upon by both parties, exist only in digital form on a computer, the primary evidence of the document will be the digital contract residing on an identified computer storage device such as the hard drive of a computer. Printing out the document on paper will provide copies in a human-readable form, which will in turn comprise secondary evidence of the document.²

2.56 Now let us take the matter one stage further. Assume the original digital file is accessed multiple times after the contract is executed, but its file contents are not altered: perhaps particular clauses are copied for other reasons. The metadata for the digital file may have been changed to record the action of opening and closing the file, even if no substantive changes are made. Although the metadata might have been altered, the content of the file in question has not been affected. In these circumstances, it might be considered that the integrity of the original digital data is compromised. But as the content (rather than the metadata) of the digital document is unchanged, the digital document remains the primary evidence, and a printout of that document a faithful copy of the original.¹ The metadata merely records when the file was opened and viewed. Metadata can be compared to a file register in the physical world that records the name of the person to whom the physical file was given, the date and time the person obtained the file, and the date and time it was returned: the register information does not alter the content of the statements made in the file (unless the person obtaining access to the file alters its contents). In such circumstances, the metadata does not affect the integrity of the digital data, which also means the secondary evidence of the file in the form of the printout remains a reliable reproduction of the digital file.

2.57 Consider another example: the drafting of a contract by an external lawyer for a multinational company. This task will comprise a number of stages, including: liaising with a number of people internally with different responsibilities to produce an initial draft of the contract; passing it to the other contracting party for comments; and producing a final version to the satisfaction of both parties, after a substantial period of negotiation. In all probability, various versions of the draft contract will exist in storage devices on computers, hand-held devices and backup devices belonging to several companies and their employees, perhaps across different jurisdictions. If the contract is then printed and signed by the authorized representatives of the two parties, the original document will be the printed version. If there is an issue regarding a particular version of the contract at a particular point in the negotiations, the draft digital version of the contract will be original evidence because that electronic copy is the best evidence of that version of the contract, and a printout of that version will be secondary evidence.

2.58 In addition, digital documents may themselves be stored, changed, compiled and collated into new documents, and the new documents may be original documents in themselves. The Canadian case of R v Bell¹ is instructional in this regard. In this case, the bank’s computer software processed the various transactions of its customers’ chequing accounts into a monthly statement for each account. Two identical copies of the monthly statement were printed, one for the customer and one for the bank. The bank retained its copy of the monthly statement, but did not retain a record of the transactions. The trial judge held that a copy of the statement was not admissible because the transaction information stored on a computer was the record, and the original ‘record’ as a record of the dealings of a financial institution (and its subsequent copy) no longer existed. On appeal, this analysis was rejected. Weatherston JA noted that the form in which information is recorded may change from time to time, and a new form in which information is recorded, such as a compilation or collection of other records, is equally a record of that kind of information. The court found the monthly statement to be such a ‘record’ that consolidated the transactions of a financial institution and allowed the appeal.²

Civil proceedings

2.59 The admissibility of secondary evidence in civil proceedings is governed by of s 8 of the Civil Evidence Act 1995, which permits the introduction of copies of documents into evidence for the purpose of proving the statement contained in the document:

2.60 A ‘document’ is in turn defined in s 13 as ‘anything in which information of any description is recorded’, and a ‘copy’ of a document as ‘anything onto which information recorded in the document has been copied, by whatever means and whether directly or indirectly’. There are two operative parts to s 8. Section 8(1)(a) provides that an admissible statement contained in a document may be proved by the production of the original document. Section 8(1)(b) provides that the same document may be proved by the production of a copy of that document or a material part of it, with the expression ‘whether or not that [primary] document is still in existence’ completely eviscerating the common law best evidence rule. And although s 8(1) uses the language of ‘a statement contained in a document’, suggesting that the statutory version of the best evidence rule only applies to documentary evidence used in a testimonial sense, a better reading is that s 8 applies to documentary evidence both as testimonial evidence and as real evidence. This means that s 8 will apply to the analogue record of the measurements of a device (the measurement constitutes the statement of the document)¹ or the printout from an Intoximeter.

2.61 The admissibility of the copied document as secondary evidence is subject to one condition and one qualification. The condition is that, as set out in the proviso to s 8(1), the copied document must be ‘authenticated in such manner as the court may approve’, just as the primary document must be authenticated. In other words, where the credibility of the digital data is in question, foundation evidence, typically in the form of testimony, will have to be introduced and tested to determine whether the secondary evidence can be accepted as ‘a copy’ of the original document. The residual judicial control over the admissibility of secondary evidence takes the form of judicial prescription of the requisite authentication evidence to prove that it is an accurate and reliable copy of the whole or a material part of the original document.

2.62 The qualification is that, by s 8(2), the number of removes between the copy and the original document is statutorily deemed to be irrelevant. This detracts from the judicial control role as explained above, and also undermines the judicial assessment of the authentication evidence as to the true accuracy and reliability of the secondary evidence.¹

Criminal proceedings

2.63 The starting point for the application of the best evidence rule in criminal proceedings is s 133 of the Criminal Justice Act 2003:

2.64 The s 133 provisions are identical to those for civil proceedings in the Civil Evidence Act 1995, save for the fact that there is no mention of the number of times a copy is removed from the original in s 133 in the Criminal Justice Act. (It is suggested that the elimination of the number of removes qualification in s 133 is an improvement over the equivalent formulation of the best evidence rule in the Civil Evidence Act, in getting rid of the judicial handicap on assessment of the authentication evidence.) The other difference is that proof in criminal proceedings must rise to the appropriate standard, which is proof beyond reasonable doubt in the case of the prosecution, and proof on the balance of probabilities in the defence case. Otherwise, it should also be noted that, notwithstanding the reference to ‘a statement in a document’, for the same reasons as outlined above in relation to the Civil Evidence Act 1995, the best evidence provisions should apply equally to a document as real evidence and to a document as testimonial evidence.¹ In other words, as in civil proceedings, secondary evidence of an electronic document is admissible subject to authentication evidence.

2.65 The effect is that while the original electronic document, if available, should be adduced into evidence, in practice a copy of the document tends to be adduced as secondary evidence. The copy may be at least one, if not two removes¹ from the original. This should not matter, provided the digital copy has been produced in a way that captures the file in its entirety, including all its attributes, such as the metadata, without altering the original data. (On this point, please see the detailed discussion in Chapter 6 on authentication.)

2.66 To a certain extent, rather than question whether a document in digital form is an original or a copy, it might be more useful and relevant to refer to the proof of authenticity, or provenance, or reliability of a digital file. This is required under both s 133 of the Criminal Justice Act 2003 and s 8 of the Civil Evidence Act 1995. This in turn encapsulates proof of the integrity of the content of the data. Because of the ease with which a digital document may be migrated from one storage device to another, and thereby undergo formatting and other changes, including content and metadata changes, it is vital to require any such changes to be documented in such a way as to preserve the integrity and authenticity of the copy. Thus it might be more relevant, when referring to digital data, to concentrate on establishing which version of the data is required, particularly whether the making of copies of the digital document is properly documented.

Admissibility

2.67 That evidence takes electronic form has not been an impediment to its admissibility. Judges have admitted digital records of the product of mechanical devices and automatic recordings, photographs,¹ tape recordings,² automated film recordings of the movements of a ship as traced by radar,³ microfilm,⁴ printouts of test results undertaken on a breath test machine,⁵ video recordings,⁶ computer printouts⁷ and a recording of an oral statement of testamentary intentions on a DVD.⁸ The types and categories of electronic evidence are not closed.

2.68 Evidence is admitted into legal proceedings if it is relevant to an issue in dispute, subject to a number of exceptions.¹ It is a matter of law for a judge to determine whether evidence is admissible. Generally, judges are required to determine whether evidence is to be excluded in criminal trials far more frequently than in civil matters, especially where admitting the evidence might not be in the interests of justice.² For instance, in R v Fowden and White³ the Court of Appeal held that a video film showing activities that were consistent with the acts of theft had been improperly admitted.⁴ The prejudicial value outweighed its probative effect, because the witnesses who identified the accused knew them from a similar case of theft that occurred a week after the events recorded in the video film, and the defence was therefore not able to test the accuracy of the identification without causing prejudice and embarrassment.⁵

2.69 In civil proceedings, evidence that is admissible can be excluded in accordance with the provisions of CPR 32.1(2), which provides a judge with the explicit general power to exclude evidence when in the role of managing a case:

2.70 However, in adopting the argument of the appellants in Great Future International Ltd v Sealand Housing Corporation, Arden LJ pointed out, at [24], that the power ‘must be used with great circumspection for the purpose of achieving the overriding objective’.¹ The modern tendency is to admit evidence, and then consider its weight, as illustrated by the comment of Cockburn CJ in The Queen v Churchwardens, Overseers and Guardians of the Poor of the Parish of Birmingham: ‘People were formerly frightened out of their wits about admitting evidence lest juries should go wrong. In modern times we admit the evidence and discuss its weight.’²

Weight

2.71 The questions of weight, credibility and sufficiency of the evidence are decisions for the members of a jury, and for the judge where a case is tried without a jury. There are no fixed rules to determine what weight to give to any item of evidence. In R v Madhub Chunder Giri Mohunt, Birch J observed: ‘For weighing evidence and drawing inferences from it, there can be no canon. Each case represents its own peculiarities and in each common sense and shrewdness must be brought to bear upon the facts elicited’,¹ and Lord Blackburn commented in Lord Advocate v Blantyre that ‘[t]he weight of evidence depends on rules of common sense’.²

2.72 When conducting a trial with members of a jury, the judge may withdraw an issue because the proponent has failed to adduce sufficient evidence in support of the claim. Furthermore, in summing up to the members of the jury at the end of the trial, the judge is required to provide directions on a range of issues, including, but not limited to: who has the burden of proof; what presumptions, if any, apply; when supporting evidence should be considered before putting weight on certain types of evidence; and to offer comments on matters including the weight of the evidence, although it must be made explicit that such comments are meant only to help the members of the jury in reaching their own decision.¹ In addition, there are a number of factors set out in s 114(2) of the Criminal Justice Act 2003 that deal with the assessment of the weight of hearsay in criminal proceedings.

Video and audio evidence

Testimonial use in legal proceedings

2.73 In exceptional instances, video-recorded and tape-recorded evidence may be used in lieu of testimonial evidence. In civil proceedings, evidence may be given by means of a video link or any other means, subject to leave being obtained from the court.¹ In criminal matters, it is possible to record the initial interview with children² and admit the recording in evidence, subject to leave of the court and any editing that the court decides is necessary.³ Leave is required to adduce a video recording of the testimony of a witness in accordance with the provisions of s 27 of the Youth Justice and Criminal Evidence Act 1999.⁴

2.74 Video-conferencing and web-conferencing technology have also made it possible to provide testimonial evidence from outside the court.

Identification and recognition evidence

2.75 Surveillance cameras are very much part of life in the twenty-first century, ever since the foundations of their use were laid in the latter decades of the twentieth century. Evidence of images from security cameras can be very helpful in identifying the perpetrators of crimes. Such evidence has been admitted in English courts, mainly in criminal cases.¹ The widespread availability of video-recorded and tape-recorded evidence has opened up the possibility that such evidence may be augmented with more advanced techniques, and the enhancement of the sounds or images, together with the use of techniques such as aural identification and facial mapping, can help to identify the parties in a recording.

2.76 Before such evidence is used, there should be a careful examination¹ of the technology in question. A good example of this judicial scrutiny was that done by Steyn LJ in R v Clarke (Robert Lee),² where His Lordship analysed the technique of facial mapping³ by video superimposition. The court carefully considered the reliability of the underlying scientific techniques, noting that the techniques themselves could be fit for debate, and their improper use by an expert in the particular case could in turn affect the probative value of such evidence. It was only after the court was satisfied on these two grounds that the identification evidence from the application of the technique was admitted.

2.77 Issues regarding the reliability and application of these techniques are very much for expert evidence, depending on the nature and sophistication of each technique. But some guidance may be sought that stems from the best practices for handling electronic evidence. For instance, for evidential techniques that involve manipulating and enhancing digital imagery, Gregory Joseph has noted that the following steps must be taken before enhanced digital imagery can usefully be used:¹

2.78 Important lessons were also spelt out regarding the use of voice recognition technologies and techniques for identification purposes in R v Flynn and St John.¹ In this case, the prosecution sought to identify the two appellants as conspirators of a robbery through voice recognition techniques. Before the robbery, the police secretly fitted a listening and transmitting device to one of the vehicles it was assumed (correctly) that the conspirators would use for the robbery. Four police officers testified that they recognized the appellants’ voices from the 60 minutes of covert recording made by the device. The trial judge ruled admissible the evidence of the police officers and the transcripts of the recording, and placed the evidence before the jury. The appellants challenged the decision of the trial judge to admit the voice recognition evidence of the officers and the judge’s failure to give an appropriate direction to this evidence.

2.79 In giving judgment on appeal, Gage LJ noted that there are two categories of voice recognition evidence: expert evidence using either auditory analysis or acoustic/spectrographic analysis, or lay listener evidence, where the lay listener as a witness is required to possess some special knowledge of the suspect that enables him to recognize the suspect’s voice. Such witnesses may be close relatives or friends, but they may also be persons who acquire such familiarity by the frequency of their contact with the suspect. Gage LJ also noted that suspect identification by voice recognition is more difficult than visual identification, that voice identification by experts using sophisticated auditory, acoustic and spectrographic techniques are likely to be more reliable than identification by a lay listener, and that the quality of identification by a lay listener is highly variable. In addition, research has shown that a confident recognition by a lay listener of a familiar voice may nevertheless be wrong, because while an expert is able to draw up an overall profile of the individual’s speech patterns, in combination with instrumental analysis and reference research, a lay listener’s response is fundamentally opaque because he cannot know and has no way of explaining which aspects of the speaker’s speech patterns he is responding to, and has no way of assessing the significance of the individually observed features relative to the overall speech profile. This makes it more difficult to challenge the accuracy of his evidence.

2.80 For all these reasons, the Court of Appeal allowed the appeal, holding that the police officers as lay listeners had a limited opportunity to acquire familiarity with the appellants’ voices, and that the quality of the covert recording was poor. In contrast, both experts, one representing the prosecution and the other representing the appellants, were unable to identify the voices as being those of the appellants, further casting doubt on the officers’ voice recognition evidence.

2.81 While R v Flynn and St John did not close the door on voice recognition evidence, Gary Edmond, Kristy Martire and Mehera San Roquem suggest the following minimal safeguards be required before the prosecution can seek to admit voice recognition evidence from lay listeners:¹

2.82 These safeguards are certainly in line with the issues raised by Gage LJ in R v Flynn and St John, and highlight the care with which both the parties and the courts must proceed when seeking to admit computer-generated and computer-augmented evidence, in order to safeguard the evidential process.

Computer-generated animations and simulations

2.83 Digital visual evidence presentation systems (including digital displays, computer-generated graphical presentations, animated graphics and immersive virtual environment technology) have been used in legal proceedings in many jurisdictions. Such tools can be used to present evidence and illustrate hypotheses based on scientific data, or to depict the perception of a witness, or to illustrate what may have occurred (as seen from a specific viewpoint) during a particular incident. Digital reconstruction technology may also be applied in a court to explore and illustrate ‘what if’ scenarios and questions, to test competing hypotheses and to expose any possible inconsistencies and discrepancies within the evidence.

2.84 Computer animations and interactive virtual simulations are potentially unparalleled in their capabilities for presenting complex evidence.¹ The use of such enabling visualization technologies can affect the manner in which evidence is assimilated and correlated by the viewer. In many instances, visual media can potentially help make the evidence more relevant and easier to understand.² In other cases it may be seen to be unfairly prejudicing the members of a jury.

2.85 At first glance, these computer-generated graphical reconstructions may be seen as potentially useful in any court, and they are often treated like any other form of digital evidence regarding their admissibility. In particular, they are admitted as part of expert testimonial evidence or as a special type of real evidence.¹ However, this specific form of digital media warrants special care and attention due to its inherently persuasive nature, and the undue reliance that the viewer may place on evidence presented through a (potentially photorealistic) visualization medium such as this, often to the exclusion of the underlying evidence and the assumptions made in generating these graphical representations. This is often referred to as the ‘seeing is believing’ tendency.²

2.86 As courts begin increasingly to use multimedia and cinematic displays, this has profound implications for the legal processes taking place that are intrinsically tied to the application of such technology. It must be questioned whether the decisions made in courts using such technologies are adversely affected by the manner in which the evidence is presented.¹

Computer-generated evidence in England and Wales: civil proceedings

2.87 An early occurrence of the use of computer-generated evidence is seen in the civil case of The Owners of the Ship Pelopidas v The Owners of the Ship TRSL Concord.¹ In 1996 a collision took place in the Access Channel to Buenos Aires between two vessels: the Pelopidas and TRSL Concord. The issue for the court to decide was the liability for the collision and the apportionment of that liability. The items of computer-generated evidence submitted were two-dimensional computer-generated simulations of the trajectories of both vessels; these were, in effect, animated maps. A ‘black box’ on the Concord recorded various positioning, speed and heading data at 15-second intervals for the relevant collision time period. Both sides accepted the accuracy of the plot. David Steel J concluded that a fair apportionment of liability was 60:40 in favour of the Pelopidas, and stated:

2.88 In stating the above, David Steel J was remarking on his accumulated experience of the usefulness of computer-generated reconstruction evidence.¹ Computer simulations are now more frequently used in civil proceedings.²

Computer-generated evidence in England and Wales: criminal proceedings

2.89 The Court of Appeal has indicated that it favours the use of digital images in criminal proceedings, as indicated in R v Smith (Peter Kenneth)¹ by Thomas LJ, at [61]:²

2.90 Because of the critical nature of criminal trials, it is crucial that any computer-generated evidence that is put forward be thoroughly examined.¹ The use of a jury in criminal cases is another important reason for assessing the relevance, accuracy and possible prejudicial effect of computer-generated evidence carefully. For this reason, it is important for defence counsel to be aware of the issues that can arise and be suitably prepared to test the evidence. In R. v Gardner (Trevor Elton),² a person was killed during a fire in a block of flats. One of the experts who gave identification evidence for the prosecution used a new technique that deployed computer software to provide an analysis of video surveillance footage, as described by Waller LJ at [34]:³

2.91 Even though the defence did not have any material in relation to which they could cross-examine the expert witness and enable the jury to judge the accuracy of the expert’s analysis and assessment that the person identified in the surveillance footage was the defendant, the court guardedly accepted the admissibility of this evidence. In doing so, Waller LJ also sounded a note of caution regarding new techniques relating to identification. The judge quoted, at [45], the following statement of Lord Hope in Hopes and Lavery v HM Advocate:

2.92 But even if juries are to be enabled to benefit from the full spectrum of technological evidence, they are particularly vulnerable, often more so than judges and coroners, to any prejudicial effect and inaccuracy of scientific animations. Perhaps this is because juries do not have the same level of cynicism that years of experience of analysing evidence has given judges and, to a lesser degree, coroners. In the case of R v Ore,¹ Tucker J stated the defence’s apprehension regarding the admissibility of a computer-generated animation:

2.93 The concerns stated above are highly relevant and illustrate real fears about any computer-generated evidence. This is especially true for forensic reconstructions. Hence, any computer-generated reconstructions should be made as precisely and in as unbiased a way as possible, and their use has to be shown to be necessary.¹ Their probative value should outweigh any potential prejudicial effect.

2.94 These lessons may be illustrated by the case of R v Ore, which introduced one of the first forensic computer-generated animations to an English criminal trial. The Crash Investigation and Training Unit of the West Midlands Police Service produced the animation. The case involved a collision between two vehicles at a junction; one of the drivers was killed as he pulled out in front of an oncoming vehicle. The views of both drivers were partially obscured by large hedges and walls around the junction.¹ Tucker J, who presided over this case, further stated in his ruling on 25 November 1998:²

2.95 A well-known example from Northern Ireland is the computer-generated evidence that was extensively used during the Bloody Sunday Inquiry.¹ In 1972, 13 people were killed during a peaceful demonstration. The original inquiry produced a report within 11 weeks of the incident, and acquitted the soldiers involved. In 1998, a Tribunal of Inquiry was established to reassess the events.² Lord Saville, the chair of the tribunal, took full advantage of ensuing improvements in technology, and used a computer software system designed especially for use in the Inquiry to amplify the testimony of witnesses. The Northern Ireland Centre for Learning and Resources produced the computer-generated virtual models, which reconstructed a large area of Londonderry that had been extensively altered since 1972. The user was able to compare the same scene as it appeared at the time of the Inquiry and as it was in 1972. There were 80 locations stored in the system that could be explored, with specific points of view being recalled when switching between the representations. The system could also store oral evidence about location and movement, and export scenes to a mark-up system so that witnesses could draw on top of images. The computer software system that was admitted was deemed to be unbiased and accurate.

2.96 The Bloody Sunday Inquiry computer system was not interactive in three-dimensions. Virtual reality or VR, by definition, is an interactive computer-generated simulated environment with which users can interact using a computer monitor or specialized hardware. The computer system used for the Bloody Sunday Inquiry was interactive in the sense that viewers were able to view images of different scenes at varied times. However, the viewer was not able to move around a full three-dimensional virtual environment of Londonderry itself, since such a three-dimensional virtual model of the area did not exist. But since then, courts in England and Wales have begun to introduce interactive three-dimensional VR crime scene environments for a number of high-profile criminal cases.¹ There is little doubt that, with the increasing complexity of criminal investigations, we will see more use of virtual environments and immersive virtual environments in legal proceedings.

2.97 Virtual environments possess the potential to sway juries and decision makers, even more so than computer animations in general. Creating an environment that allows viewers to take different perspectives and manipulate objects in that environment do indeed allow for ‘what if’ scenarios to be played out, and could lead to more robust decisions. But the reconstructions of scenes in these environments are based on various assumptions and premises, not all of which can be elucidated or are transparent, or easily accessible for review by opposing experts and by decision makers. Analyses of computer-generated displays show that they can be extremely advantageous in the court, provided they are used appropriately. The consequences of a failure to investigate these issues cannot be underestimated, since errors, inaccuracies, misuse, tampering or biases within visualizations are capable of leading to miscarriages of justice.¹