R v Bain
[2009] NZSC 16
•6 March 2009
For a Court ready (fee required) version please follow this link
IN THE SUPREME COURT OF NEW ZEALAND
SC 13/2009
[2009] NZSC 16DAVID CULLEN BAIN
v
THE QUEEN
Hearing:2 March 2009
Court:Elias CJ, Blanchard, McGrath, Wilson and Gault JJ
Counsel:H A Cull QC and P A Morten for Appellant
K Raftery and C L Mander for Crown
Judgment:6 March 2009
Reasons:18 March 2009
JUDGMENT OF THE COURT
AThe appeal is allowed.
BThe disputed evidence is excluded and the material in issue is to be excised.
REASONS
Para No
Elias CJ and Blanchard J [1]
McGrath J [69]
Wilson J [90]
Gault J [97]
ELIAS CJ AND BLANCHARD J
(Given by Elias CJ)[1] On 6 March 2009 the Court allowed the appeal and indicated it would give reasons at a later date. The members of the Court now give their reasons.
[2] The appellant, David Cullen Bain, was convicted in 1995 of the murders in 1994 of five members of his family. His convictions were eventually quashed in 2007 by the Privy Council, which ordered a new trial.[1] In preparation for the trial, which has now started before Panckhurst J and a jury, the Judge has been obliged to determine a number of applications. Some have been appealed to the Court of Appeal. One concerned the High Court ruling that a digital recording said by the Crown to be a true copy of a taped 111 call is provisionally admissible at the trial, subject to proof of the chain of custody of the recording. [2]The 111 call was made by the appellant when he reported the deaths of his family members. On 30 January, the Court of Appeal dismissed the appellant’s appeal against the ruling as to admissibility. [3] Mr Bain has been granted leave to appeal further to this Court on the admissibility of part of the recording. The part of the recording to which objection is taken contains what the Crown says is an admission: “I shot the prick” or “I shot that prick”.
[1]Bain v R [2007] 23 CRNZ 71 (PC).
[2]R v Bain (High Court, Christchurch, CRI-1994-012-217294, 13 November 2008, Panckhurst J).
[3][2009] NZCA 1 (William Young P, Chambers and O’Regan JJ).
[3] Although the 111 call formed part of the evidence at the appellant’s earlier trial in 1995, the words were not recognised and did not form part of the transcript of the
call. The presence of the disputed admission was first found by Detective Ward, when he reviewed the recording, in preparation for the retrial, in July 2007. At that time he listened to the recording at a commercial sound studio in Dunedin. Mr Dempsey, the ambulance officer who took the call, had not heard the disputed words and did not suggest their inclusion in the transcript originally prepared in 1994. In October 2007 he was asked to recheck the recording after being told of the words which had been discovered. In listening to the recording with that knowledge, he heard the words “I shot the prick I shot” and was “stunned that I hadn’t heard the words previously”.
[4] Because of the discovery, the recording was sent for analysis to the United Kingdom to forensic consultants, expert in analysing recordings of speech. It was also analysed by experts for the defence. There is very little difference between the experts as to their findings. The opinions constitute evidence extrinsic to the recording itself upon which the decision to admit the recording was based.
[5] Although it is necessary to refer to the forensic opinions at some length later in these reasons, they may be briefly summarised for present purposes. None of the experts is able to say that the sounds relied upon in the recording are words, rather than meaningless exhalation of breath. If they are words, none of the experts is able to say that they amount to the words the Crown wishes to rely upon as evidence. Some consider such words can be heard in the recording, with effort. But all experts caution as to the dangers of hearing something that may not in fact be there, because of accident in arrangements of sounds. The principal Crown expert uses the analogy of an image glimpsed in a cloud formation to illustrate the dangers. All experts suggest that it is not possible for experts or lay people to resolve the question whether the sounds are speech by listening to the recording. And all express the view that, if the recording is admitted, the jury should not be told what the sounds are thought to amount to. All are agreed that such “priming” entails a high risk of suggestibility, which would be difficult to counter. They differ about whether the recording can be left for assessment by the jury and, if so, whether the jury will be assisted by expert evidence.
[6] Panckhurst J ruled that the recording should be admitted on the basis that its interpretation was a “jury question”.[4] He indicated that the jury would not be provided with a transcript, but considered that they should have the assistance of expert evidence as to the interpretation of the disputed part of the recording. The Court of Appeal agreed. It held that the interpretation of items of real evidence such as the recording was for the jury to assess. The recording should first be played to the jury without their being told of the interpretation put on the disputed sounds by the Crown, but with a request for them to listen to it carefully and with the warning that they will hear some speech “produced on breath”.[5] The Court of Appeal agreed with Panckhurst J that after the jury had heard the recording “unprimed”, expert evidence should be given concerning its interpretation.[6] The Court envisaged that at the end of the trial the Judge, in summing up, would warn the jury of the dangers of suggestibility.[7]
[4]At para [57].
[5]At para [258].
[6]At para [259].
[7]At paras [257] – [260].
[7] On 23 February this Court granted leave to appeal in respect of the question of admissibility and the appropriateness of the procedure suggested by the Court of Appeal to address the risk of suggestibility. Panckhurst J had also dismissed objections to the authenticity of the digital recording based on its being a true record of the call, and provisionally admitted it under s 14 of the Evidence Act 2006,[8] subject to further evidence at trial establishing the chain of custody. The Court of Appeal dismissed an appeal based on this point,[9] and this Court declined leave to appeal further.[10]
Expert assessment of the recording
[8]Which permits evidence to be admitted subject to later evidence to establish its admissibility.
[9][2009] NZCA 1 (William Young P, Chambers and O’Regan JJ).
[10][2009] NZSC 12.
[8] Opinion evidence bearing on identification of the claimed admission in the 111 recording was given by four experts on a voir dire before Panckhurst J. Professor
French and Mr Harrison gave evidence for the Crown, Dr Foulkes and Dr Innes gave evidence for the defence. In addition the working papers of Ms Cauley, an experienced researcher working with Professor French, were part of the materials before the Court. These experts had available to them sophisticated equipment to block out background noise and to analyse speech patterns. There is little disagreement between them.
[9] Professor French and Mr Harrison are both independent forensic consultants attached to the Department of Language and Linguistic Science at the University of York. They are specialists in the analysis of digital and magnetic recordings and speech and language samples. Both are highly qualified and experienced in giving evidence about the analysis of speech and audio recordings. They prepared a report, completed in July 2008, of their examination of the 111 audio recording. As part of the work preparatory to the report, the 111 recording was also listened to by their assistant, Ms Cauley, a researcher at the Department of Language and Linguistic Science at the University of York. She too is acknowledged to be experienced in analysing speech on recordings.
[10] Professor French and Mr Harrison prepared a transcript of the 111 call. In their report they describe how the transcript was “prepared over several revisions by replaying the material within computer software, repeatedly, one short section at a time, whilst listening via high-quality headphones.” The initial draft prepared by Mr Harrison was finalised by Professor French and was appended to their report. It reads:
Transcript of 111 Call
AO Ambulance, can I help you?
C Help.
AO Yeah.
C They’re all dead.
AO What’s the matter?
C They’re all dead. I came home and they’re all dead.
AO Whereabouts are you?
C Erm- erm Every Street.
AO W-w- Every Street?
C Sixty-five Every Street. They’re all dead.
AO Who’s all dead?
C My f- my family. They’re all dead. Hurry up.
AO It’s okay. Every Street and it runs off s- off Summerville Street?
C (Yeah.) Yes.>>> [POSITION OF QUESTIONED UTTERANCE]
AO And what phone number you’re calling from?
C Four-four five four.
AO Mm-hm.
C Two five two seven. [Whispered on out-breath]
AO Four five four?
C Two five two seven.
AO Two five two seven. And your last name?
C (Ig-er) Bain.
AO Bain. Okay. We’re on our way. Okay Mister Bain.
C Please hurry up.
AO Yeah, we’ll be there very shortly.
[11] As appears from the transcript, neither Mr Harrison nor Professor French was able to say that the sounds Detective Ward thought he had heard when playing the recording in July 2007 in Dunedin were words and, if so, what they were. That was so even though they had been advised of the Crown contention and so were looking for the words. The place at which the disputed words is located is identified in the transcript, although the words themselves are not recorded by these experts. The passage of interest is said in the report to consist of “an unvoiced out-breath – ie, the vocal folds are not vibrating during it”. The sounds in the questioned material were:
not of the type that one would normally consider suitable when carrying out our forensic speaker comparisons ... . The tests one normally would apply ... could not be applied to the material in question.
The report describes the caller as appearing distressed and out of breath throughout the recording and indicates “many audible breath sounds”:
At one point at least, he produces speech on an out-breath without there being vocal fold vibration – phonetically, the process and effect are aligned with whispering. This is at 40.5 seconds into the recording where the caller first gives the last four digits of his telephone number: ‘two five two seven’ (transcript page 3, line 7). This utterance provides a clear indication that Mr Bain could – and did – during the course of the call speak on exhaled air without vocal fold vibration.
[12] The material part of the recording was described in the report as being produced on “exhaled air”, with “no voicing and modification of the airflow in the oral tract”. While the “nature and timings of the modification” were “such that the material could be heard as ‘I shot the prick’ or ‘I shot that prick’”, neither Professor French nor Mr Harrison could say that the sounds were in fact words:
However, whilst we cannot discount the possibility that the material amounts to speech – particularly as he did utter ‘two five two seven’ in this way – it also remains entirely possible that it is not speech. Rather, it could be no more than an audible out-breath that has, in the distress of the moment, been modified by a random and unfortunately-sequenced series of movements of the tongue and lips so as to create a series of sounds that could – albeit with a little effort – be heard as ‘I shot the/that prick’. If one were to draw an analogy with visual stimuli, we all have had the experience of seeing pictures in the clouds that arise from no more than the co-incidental arrangement of water particles. In respect of the questioned material, there is, unfortunately, no way of ‘reading backwards’ from the sounds themselves to the psychological processes that gave rise to their production, and therefore no way of resolving the issue of whether Mr Bain did, in fact, whisper the words suggested.
[13] In order to demonstrate the dangers of misinterpreting audible exhalations of speech, Professor French and Mr Harrison prepared a compact disc which, while sounding like speech, had been fabricated by editing together parts of out-breaths from different areas of the call. They had no linguistic content or meaning.
[14] The conclusion of the report prepared by Professor French and Mr Harrison was:
In summary, and having given this issue extended consideration, it is our view that it would be dangerous to put before a jury an interpretation of the questioned material as ‘I shot the/that prick’ ... . While it was not possible to discount the questioned sounds being ‘I shot the/that prick’, they could equally be simply an out-of-breath exhalation that happens to resemble those words, but which, in fact, is empty of linguistic content or meaning. We would consider it unsafe to place the interpretation ‘I shot the/that prick’ before a jury.
[15] In his evidence at the voir dire Professor French expressed the view that if the recording were played to the jury it would be dangerous for them to be “primed” by being told what words to listen for. He confirmed his view in the report that it was impossible to resolve whether the sounds heard at that stage in the recording were audible breathing or actual words and his opinion that there was no justification to put the interpretation “I shot the prick” to the jury. Professor French made the point that no “speech oratist” could determine whether the sounds were breathing or speech and suggested that whether they were one or the other in fact depended upon the intent with which the sound was made (a matter that could not be resolved by listening to the recording). Professor French affirmed his earlier view that it was not possible for an expert or a layman to resolve “either way” the matter of whether the sounds were speech or breath by listening to the recording. While the same reservation applied to the “two five two seven” whisper, Professor French thought it to be “rather clearer”. What it did show was that, in his opinion, Mr Bain was “capable of speaking on whispered out-breath and in fact does so in the context of this particular call”.
[16] Mr Harrison similarly confirmed in his evidence at the voir dire that he was “not able to say 100 per cent either way whether it is speech or it isn’t speech”. He too acknowledged the heightened risk of misinterpreting the words if an interpretation was suggested before they were heard.
[17] The working notes produced by Ms Cauley show that she was first asked to do a “‘blind’ first draft transcript” with no case information. She says in her report that there were certain sections where speech is “whispered with heavy breath”. Although she transcribed possible interpretations she noted that she would not include these in the final version because “my level of confidence is not high enough”. In the first draft she described the sounds which are disputed as “yeah, I can’t [touch it] I can’t - [whispered]”. Her notation indicates that the words “touch it” were ones she would not include in a final transcript because her level of confidence about them was not high enough.
[18] Ms Cauley’s first draft transcript of the call, with round brackets around words in respect of which she had a “lower level of confidence”, dots in respect of “unintelligible speech”, and square brackets in green ink around words she would exclude from any transcript as “level of confidence too low”, was as follows:
1st Draft Transcript of 1st call on Track 1 of CD
TO … can I help you?
M Help.
TO (Yeah.)
M They’re all dead.
TO What’s (the matter)?
M [Sobbing] They’re all dead. I came home and they’re all dead.
TO Whereabouts are you?
M (Erm, erm/I’m, I’m j-) Everey Street.
TO What- Everey Street?
M Sixty-five Everey Street. They’re all dead.
TO Who’s all dead?
M My (ch-/p-), my family, they’re all dead. Hurry up.
TO It’s ok. Everey Street that runs off S- off (Somerville) Street.
M Y-yes (yeah, I can’t [touch it]. I can’t)- [whispered]
TO (Do you) know what phone number you’re calling from?
M F- four, four, five, four.
TO Mhm.
M Two, five, two, seven. [whispered]
TO Four, five, four.
M Two, five, two, seven.
TO Two, five, two, seven.
M …
TO And your last name?
M (lk- er) Bain.
TO Bain, Okay, we’re on our way.
M [(Please come back.)]
TO Okay, Mr Bain?
M [(I can’t do it.)]
TO (Yeah), we’ll be there very shortly.
M (Okay.)
[19] After being advised that a possible interpretation of the relevant sounds was “I shot the prick”, Ms Cauley noted in her report that she could then “hear how these words can be heard”. Her comments on the suggestion were that although she herself heard “what they consider” the “sh” of “shot” as “caused by breathiness”, the “sh” sound was a “viable interpretation”. Once primed, she could also hear the presence of an “r” in the sound “thought to be ‘prick’ along with strong burst at beginning of word”. She had transcribed this as “touch” in the first draft but indicated that she had also considered “p” or “r” as “feasible possibilities”. She noted that she considered that the vowel in the sound suggested to be “shot” was “too un-rounded and more ‘a’-like” and that it “doesn’t sound like correct vowel for ‘shot’”. She noted she transcribed this as “can’t” because of the vowel. Ms Cauley noted that the word she “presumed” was heard by others as “the”, “is not speech”.
[20] The defence obtained a report from Dr Foulkes, also of the Department of Language and Linguistic Science at the University of York. He too is a professional academic linguist and independent forensic consultant who has worked closely with the Crown expert witnesses. Dr Foulkes confirmed that transcription of Mr Bain’s speech in the 111 call was complicated by “his obvious state of distress”:
He is clearly in severe difficulty with his speech in general which is reflected in the fact that he is speaking on breath and whisper throughout the recording in different sections. As experts are having difficulty in resolving that because the acoustic evidence that we have is very difficult to resolve the speech and if it is speech it is impossible to decide precisely what the words and sounds are [sic].
[21] In the transcript he made without prompt as to the words heard in the disputed section, Dr Foulkes put the tentative suggestion “I can’t breathe” into the portion of the recording concerned. Dr Foulkes was in full agreement with the report of Professor French and Mr Harrison that it would be dangerous to put the interpretation “I shot the/that prick” to a jury although he agreed with their conclusion that the material could be heard as “I shot the/that prick”. It was however equally possible that the section was not speech at all. He remained of the view that “I can’t breathe” was a possible interpretation.
[22] Dr Foulkes agreed with Professor French’s view that it was impossible to resolve whether the sounds were speech or other audible sounds. He explained the position by comparison with fingerprint evidence:
In my view this is a situation that is in essence parallel to having some potential fingerprint evidence. We have an evidential mark that may or may not be a fingerprint and if it is a fingerprint it may or may not match the fingerprint of a suspect. As experts we are unable to decide whether or not it is a fingerprint and whether or not it matches. On that basis I wouldn’t have thought it was appropriate to expect a jury to make a decision when the experts can’t make that decision. I take that view to be applicable in this case where in a situation where the experts cannot agree whether or not it is speech, can’t agree the specific contents of the section if it is speech. I would not consider it is appropriate to expect a jury to be able to resolve that material.
[23] Dr Foulkes expressed the view that it would be “very dangerous” if the jury were told that a possible interpretation was “I shot the prick” because “objectively and impartially we cannot provide evidence to confirm or disconfirm that interpretation”. He thought it too dangerous to put that interpretation to the jury. There was “no objective scientific way to resolve the issue of whether or not that is speech and if it was speech what the specific words are”. Therefore he agreed “in principle” with the suggestion that excision of the disputed noises would not remove any identifiable words from the recording. Although other passages in the recording were unclear, Dr Foulkes thought that they were actually resolved by the context. That was the case in relation to the telephone number:
[T]he context alone gives us an indication that that section was indeed speech, even though hearing it on its own it is not necessarily easy to resolve. There is no such contextual information from the section we are particularly interested in.
[24] Dr Foulkes was questioned by the Judge about whether his analogy with fingerprints (where expert evidence is generally accepted by the courts) was appropriately applied to listening to a recording (where assessment is generally left to the trier of fact). He expressed the view that the material on this recording was “so difficult that it would be inappropriate to ask an untrained listener to interpret the words”:
Whereas a scientist approaches the material objectively and with top quality equipment and the capacity to listen and re-listen to the section, we are all of the view that it is impossible to resolve exactly what is going on at this particular point …
[25] Dr Foulkes indicated that in a number of such cases in the United Kingdom, expert evidence had been admitted and treated as important but he was “not aware of any other cases exactly of this type where the material has been so difficult”. When asked to comment on the different ways in which experts had heard the disputed sounds, he again referred to the section of the recording as being “very difficult” because of “the particular circumstances in which Mr Bain is talking”.
[26] The defence also obtained a report from Dr Innes, a New Zealand consultant linguist. She transcribed the recording of the 111 call using computer software to facilitate the process. The software allowed small portions to be selected and replayed many times for analysis and also provided a spectrographic analysis of the sounds, representing all sounds in wave form to permit detailed analysis of timing and sounds. Her transcript attempted to represent orthographically all sounds on the recording. Some were clearly words. Others were not. Her transcript indicated emphasis. In the disputed passage she was unable to discern words and sequences. When subsequently asked to listen for the words “I shot the prick” she considered there was insufficient linguistic evidence to support that interpretation and expressed the view:
[M]ost people would not hear the words you have mentioned to me unless they were somehow expecting to hear them or had been told they were (or might be) there.
[27] Dr Innes made the point that mistakes in interpretation often occur on the basis of such expectations. She too expressed the view that there was a danger in leaving the disputed portion to be played to the jury and suggested its excision on the basis that it would “not remove any identifiable words from the recording”. In evidence, Dr Innes expressed the view that the disputed passage did not contain words but heavy breathing or panting. Certain sounds in the passage had been produced which “perhaps coincidentally are sounds that are used in language as well”, but she did not consider that they were words.
The decisions in the High Court and Court of Appeal
[28] Panckhurst J ruled that the 111 recording was admissible in his judgment of 13 November 2008. The ruling as to admissibility was subject to fulfilment of conditions under s 14 of the Evidence Act as to authenticity in respect of the custody of the recording. Because of the recognised risks of suggestibility, Panckhurst J indicated that the jury would not be provided with a transcript of the 111 call, but considered they should have the assistance of expert evidence. The arrangements and final directions for that evidence were reserved for consideration at the trial.[11]
[11]At para [57].
[29] In concluding that the recording of the suggested admission was admissible, Panckhurst J described the expert evidence already traversed above. He considered it to be well settled that the interpretation of real evidence, such as a tape recording, is a jury question, citing R v Wickramasinghe[12] and R v Taylor.[13] The Judge’s analysis was made under s 8 of the Evidence Act (requiring the exclusion of relevant evidence the probative value of which is outweighed by its prejudice), under which he thought the question fell to be determined. He considered that the words plainly have “great probative value”, if the Crown contention as to their meaning is accepted.[14] The Judge took the view that, while the interpretation of the disputed words would prolong the trial, such extension was necessary. Any potential prejudice was, he thought, answered “with reference to the directions and safeguards which would need to accompany the admission of the evidence”[15]:
[54] The conversation is about one minute in length. In the main the words spoken are clearly audible. I do not consider that the jury would require the assistance of a transcript. Firm directions would be required considering the need for particular care in evaluating the disputed segment, including reference to the danger of suggestibility.
[55] Given the contents of the reports I am in no doubt that expert evidence should be received concerning the interpretation of the disputed words. This is not a situation where words are difficult to discern simply on account of background noise, over-speaking or the like. Rather, the issue is whether words intended to be sotto voce were spoken in the course of an exhalation of breath. This issue, I think, is one in relation to which the jury “is likely to obtain substantial help ….” from experts in order to determine both whether there are spoken words in the disputed segment and, if so, their interpretation: s 25(1) of the Act.
[56] In reaching this opinion I have listened to the recorded conversation on a compact disc. My view as to the interpretation of the disputed section is of no moment. However, the exercise of listening to the disputed segment did assist me in resolving the s 8 enquiry.
[57] In the result I am of the view that the 111 tape is admissible and that the interpretation of the disputed words is a jury question. The jury will not be provided with a transcript, but should have the assistance of expert evidence. The further arrangements to be made and the final content of directions to be given at trial are reserved for consideration at that time.
[12](1992) 8 CRNZ 478 at pp 480 – 481 per Hardie Boys J for the Court (CA).
[13][1993] 1 NZLR 647 at pp 650 – 651 per Thomas J for the Court (CA).
[14]At para [53].
[15]At para [53].
[30] The Court of Appeal dismissed the appeal. It, too, listened to the recording in the knowledge that the disputed sounds were “there to be heard”:[16]
With that prior knowledge, the disputed sounds can undoubtedly be heard as an inculpatory sentence. But what is very interesting is that, with the exception of Detective Ward (and/or the staff at Strawberry Sound), no one who has listened to the tape “unprimed” would appear to have heard and construed the disputed sounds in this way.
[16]At para [221].
[31] In dealing with the appeal point (which is no longer live before us) as to authenticity (in the sense of originality) of the physical recording itself, the Court of Appeal took the view that it was for the Judge to screen for admissibility “in terms of being satisfied that there is a ‘prima facie case of originality’ before allowing the exhibit to be produced”.[17] This test was to be applied on the balance of probabilities. That is not however the approach the Court took to the question of identification of the sounds as an admission. The Court of Appeal, as had the High Court Judge, addressed that question on the basis of s 8 of the Evidence Act by balancing probative value and prejudicial effect.
[17]At para [232].
[32] The Court of Appeal considered that there were two issues to be resolved: whether the disputed sounds represented speech; and, if so, what was said. It regarded the first question as more important because of the view taken that “it is reasonably obvious that the disputed sounds can be heard as the inculpatory sentence which the Crown wishes to have before the jury”.[18] The competing hypotheses on the first issue (that the disputed sounds can be properly characterised as either an audible out-breath or speech) were not ones which could be resolved by scientific evidence and “both must be regarded as open”.[19]
[18]At para [251].
[19]At para [252].
[33] The Court of Appeal remarked that none of the experts appeared to have considered whether the random movements of tongue and lips which could operate to produce the effect of speech on an audible out-breath made “reasonable contextual sense”.[20] The Court expressed its own view that the disputed sounds did make reasonable contextual sense, if construed as the admission. It was critical of the expressions of opinion from the experts that the interpretation of the disputed sounds should not go to the jury, saying that they had “strayed from their proper roles”.[21] Professor French, the Court pointed out, had correctly said that the question was one for the Judge.
[20]At para [253].
[21]At para [254].
[34] The Court of Appeal took the view that it would be “open to the jury to conclude that the disputed sounds should be construed as speech for two reasons”:[22]
(a)This hypothesis is open on the expert evidence focussing on the sounds themselves and is consistent with the way in which the appellant later gave his telephone number (when doing this, he initially expressed the numbers by the same form of exhalation of air: it was clear that this was deliberate speech, because the ambulance officer asked him to repeat the numbers and he confirmed those which had been given in the course of the exhalation); and
(b)The jury may regard as implausible the theory that these sounds, which can be construed as a sentence making contextual sense, were created by random movements of the tongue and lips.
On this basis we see the evidence as plainly relevant.
[22]At para [255].
[35] In dealing with prejudice, the Court saw the risk principally in terms of suggestibility:
[256] As to prejudice, the primary risk is that the jury may wrongly construe the disputed sounds as an inculpatory sentence – in other words, may simply get the facts wrong. But risks of this sort – that the trier of fact may get the facts wrong – are an inescapable part of the trial process and do not in themselves usually represent the sort of prejudicial effect which warrants evidence exclusion. It is, of course, the responsibility of the judge to guard against obvious risk (and particularly one that will be more apparent to a professional judge than lay jurors) of misunderstanding. In this case, there is an obvious risk, namely suggestibility, which must be addressed. But providing this happens, we see no reason why the evidence should not be admitted.
[36] The Court considered the “mechanics” of production. After noting that the interpretation of items of real evidence was a jury question and agreeing with the Judge that a transcript was not required, it expressed the view that it was best if the jury first heard the recording without being primed “except perhaps with a request that they listen to it carefully and possibly advice (as recommended by Professor French) that they will hear some speech produced on breath”:[23]
If they initially do not hear the disputed sounds as an inculpatory sentence (which we think is likely given past history) but, once primed, subsequently do hear the disputed sounds in this way, this should provide a graphic indication of the power of suggestion.
The Court of Appeal agreed that after the recording had been played to the jury “unprimed”, expert evidence should be given concerning its interpretation.[24] It expressed confidence that the Judge would warn the jury of the dangers of suggestibility in his summing up. Finally and “standing back” the Court of Appeal concluded:
[261] Before we leave this topic, and standing back from the detailed arguments we have heard and addressed, it is right [to] recognise that it would be quite extraordinary for this Court (or the Judge) to deny the jury the opportunity to listen in full to what the Crown can credibly claim is a recording of the account given by the appellant, within 25 minutes or so of the completion of his paper round, of what he found when he returned to the house.
[23]At para [258].
[24]At para [259].
Argument on Appeal
[37] For the appellant, Ms Cull contends that the disputed sounds should be excluded under s 7 of the Evidence Act because they do not have a tendency to prove anything and are irrelevant. In addition, the appellant contends that the sounds, if admitted, would have an unfairly prejudicial effect which could not be cured by judicial direction and should be excluded in application of s 8 of the Act. A further ground of exclusion is advanced based on s 28, by which unreliable statements are not to be admitted in evidence. Ms Cull submits that the issue of relevance was not addressed in the Court of Appeal or the High Court, despite being a threshold requirement for admissibility. If the sounds are not words at all, the passage has no probative value and is irrelevant. It is submitted that the Court of Appeal erred in treating “an open hypothesis” expressed by the experts (that the disputed sounds could possibly be words and could equally possibly not be words) as sufficient basis to admit the evidence as relevant. In relation to the application of s 8, the appellant submits that where inferences are equally open and where experts are agreed it would be dangerous to suggest the “disputed words”, the prejudicial effect outweighs the probative value of the evidence. It should have been excluded by the Judge under s 8(1). The risk of suggestibility cannot be overcome by direction, creating a risk of jury bias that affects the right of the defendant to offer an effective defence (a consideration under s 8(2)). Indeed, Ms Cull argues that the expert evidence should not be admitted because it would itself create the risk of suggestibility and fail the substantial helpfulness test under s 25 of the Act. Associated with this submission, Ms Cull submits that the procedure suggested by the Court of Appeal for admission of the evidence of disputed sounds was inadequate. It would entail the jury being primed by the expert opinion evidence, contrary to the advice of the experts. Warnings as to suggestibility would be an inadequate response. The contention for the appellant is therefore that the disputed passage should be excised from the recording.
[38] For the Crown, Mr Raftery submits that expert opinion evidence cannot resolve the question whether the disputed sounds are in fact words and, if so, what they are. Both are questions for the jury, whose task it is to interpret real evidence. The role of the judge is to consider “whether a properly instructed reasonable jury could place any weight on the evidence (if accepted)”. If evidence is unreliable then s 8 of the Act may be engaged, rather than s 7. The Crown maintains that the evidence is that the disputed sounds can be interpreted as either speech or breathing and, if as speech, can be interpreted as “I shot the/that prick”. Ultimately the interpretation of the disputed passage depends not only on analysis of the sounds but on the appellant’s intent at the time they were made. The Crown considers that it would be reasonable for the jury to place some weight on the evidence since the hypothesis that the words are “I shot the prick” is available on the expert evidence and elsewhere in the 111 recording the appellant had used audible exhalation in deliberate speech. In addition “the jury could find it unlikely that random lip and tongue movements would make contextual sense”. The evidence goes to the central fact in issue and on one interpretation available to the jury amounts to an admission of guilt. On this basis, Mr Raftery submits that it is not appropriate for the trial Judge to resolve the issue on a threshold basis. The task of assessing the material falls “squarely in the domain of the jury” and it does not matter that competing interpretations are available. Even if the issue is incapable of final resolution it does not follow that the evidence should be excluded under s 8. The jury may ascribe to it such weight as is appropriate in the circumstances. The real issue, it is suggested, under s 8 is the risk of suggestibility. That, the Crown submits, can be countered by the expert evidence and a suitable direction from the Judge, such as was suggested by the Court of Appeal.
Relevance
[39] Pivotal provisions in the Evidence Act are s 7, making relevance an essential condition of admission of evidence, and s 8, providing a general rule of exclusion of relevant evidence when its probative value is outweighed by its prejudicial effect. They reflect the modern treatment of evidence, building on the work of Thayer[25] and Wigmore,[26] which draws a distinction between rejection of evidence because it is not relevant and the rejection of relevant evidence because of legal policy based on the experience that some relevant and material evidence is an unsafe foundation for fact‑finding. Thus the Australian Law Reform Commission in its Interim Report on Evidence in 1985 distinguished between relevance and exclusion of relevant evidence, on the basis that relevance determines “whether evidence could relate to an issue” while exclusion of relevant evidence is concerned with whether the importance of the relationship with proof is “worth the price to be paid by admitting it in evidence”.[27] This is the approach followed also by the New Zealand Law Commission[28] and adopted in the Evidence Act 2006.
[25]Thayer, Preliminary Treatise on Evidence at the Common Law (1898).
[26]Wigmore, Evidence (Chadbourn rev, 1970).
[27]Australian Law Reform Commission, Interim Report on Evidence (ALRC 26, 1985), para [315].
[28]New Zealand Law Commission, Evidence Code and Commentary (NZLC R55(2), 1999).
[40] Section 7 states the “fundamental principle” that relevant evidence is admissible and its converse, that “evidence that is not relevant is not admissible in a proceeding”. It also provides that “relevance” is defined by “tendency to prove or disprove anything that is of consequence to the determination of the proceeding”. Section 7 reads:
7 Fundamental principle that relevant evidence admissible
(1)All relevant evidence is admissible in a proceeding except evidence that is-
(a)inadmissible under this Act or any other Act; or
(b)excluded under this Act or any other Act.
(2)Evidence that is not relevant is not admissible in a proceeding.
(3)Evidence is relevant in a proceeding if it has a tendency to prove or disprove anything that is of consequence to the determination of the proceeding.
[41] Section 8 provides that relevant evidence is to be excluded if its probative value is outweighed by unfairly prejudicial effect. It serves the interests of accuracy, fairness and efficiency in fact-finding:
8 General exclusion
(1)In any proceeding, the Judge must exclude evidence if its probative value is outweighed by the risk that the evidence will—
(a)have an unfairly prejudicial effect on the proceeding; or
(b)needlessly prolong the proceeding.
(2)In determining whether the probative value of evidence is outweighed by the risk that the evidence will have an unfairly prejudicial effect on a criminal proceeding, the Judge must take into account the right of the defendant to offer an effective defence.
[42] The 111 recording is tendered by the Crown to prove an admission by the appellant. Before it is admissible as having “a tendency to prove or disprove” that admission (itself of evident “consequence to the determination of the proceeding”), the connection or relationship of the recording to the fact sought to be proved (the admission) must be established. That connection or relationship is not a matter of assessing the probative weight of the evidence but of accepting its logical connection to the fact it is said to prove. In the case of real evidence such as a recording, that connection or relationship is usually established by facts as to authenticity and identity. These terms, which appear in the Evidence Act in s 13 as aspects of relevance, are not terms of art and are sometimes used interchangeably in the case‑law and commentaries.[29] They are concerned with the preliminary facts by which an item of real evidence is shown to tend to prove or disprove a fact in issue. Authenticity and identity in respect of a voice recording said to be confessional commonly turns both on proof of the making of the recording and its subsequent custody (to negate tampering),[30] in addition to proof that the voice recorded is the voice of the accused.[31] It may also turn on the accuracy of the recording. In the United States it has been held that where the material portions of a recording are inaudible, it may be excluded.[32] Such determination is highly contextual. It depends in particular on the purpose for which the particular proof is offered and the extent to which extrinsic evidence may provide context or assist in understanding meaning.[33]The exclusion applies a more general principle that material proffered as evidence is not probative where its connection with the fact to be proved is conjectural only.[34] This more general principle is invoked also when determining whether there is a prima facie case to go forward[35] or, on appeal, whether a necessary inference of fact is supported by any evidence at all.[36] Speculation is not enough. In the case of real evidence, authentication and identification is “the logical condition” of its admissibility to proof.[37] If such connection is speculative, the evidence is irrelevant because it does not tend to establish the disputed fact.[38] Wigmore referred to the need for authentication and identity as resting on “inherent logical necessity”.[39]
[29]See, for example, Wigmore, §2128.
[30]Such questions of provenance were considered in Butera v Director of Public Prosecutions(Vic) (1987) 164 CLR 180 at p 186 per Mason CJ, Brennan and Deane JJ.
[31]See Mathieson (ed), Cross on Evidence (New Zealand looseleaf ed, 2008), para [1.11]. See also s 46 of the Evidence Act.
[32]Wright v State 79 So2d 66 (1954) discussed by Auld in “Admissibility of Tape Recordings” [1961] Criminal Law Review 598. See also People v Stephens 256 P2d 1033 (1953); State of New Jersey v Driver 183 A2d 655 (1962); State of New York v Sacchitella 295 NYS2d 880 (1968); US vScaife 749 F2d 338 (6th Circ 1984).
[33]See, for example, US v Aisenberg 120 FSupp2d 1345 (MDFla, 2000).
[34]Hawkins v Powells Tillery Steam Coal Company [1911] 1 KB 988 (CA) and Jones v Dunkel (1959) 101 CLR 298 at pp 304 – 305 per Dixon CJ.
[35]See, for example, Metropolitan Railway v Jackson (1877) 3 App Cas 193 and Jones v Dunkel (1959) 101 CLR 298.
[36]Hawkins v Powells Tillery Steam Coal Company [1911] 1 KB 988 (CA).
[37]Michael and Adler, “Real Proof: 1” (1952) 5 Vanderbilt Law Review 344, p 362.
[38]See Australian Law Reform Commission, para [979].
[39]Wigmore, §2129.
[43] Where relevance to a fact is questioned, the judge must determine whether the evidence “is reasonably capable of supporting the fact”.[40] If it is, then the evidence is relevant and, subject to rules of exclusion on policy grounds, must be left to the jury to evaluate. If it is not, the evidence must not be admitted. Applying the same approach to determination of the relevance of confessional evidence, Cross on Evidence takes the view that while the issue of whether a confession was made by the accused is a question for the jury, the judge must first determine whether there is a “prima facie case”.[41] In adoption of this test, the Australian Law Reform Commission considered that the standard for admission was whether it was “reasonably open” to make the finding which establishes relevance.[42] The Australian reforms, in turn, influenced the shape of the New Zealand Evidence Act.
[40]This is the test suggested in R v Thomas [1970] VR 674 at p 679, where exclusion for irrelevance was contrasted with exclusion where the voluntariness of a confession was not established.
[41]Mathieson (ed), para [1.11]. See also Tapper, Cross and Tapper on Evidence (11th ed, 2007), p 197. And see Cleland v R (1982) 151 CLR 1, discussed at para [51].
[42]Australian Law Reform Commission, para [646].
[44] The approach that it is for the judge to determine relevance by considering whether it would be “reasonably open” to the jury to rely on the proffered evidence as tending to prove a fact in issue in the case was the approach taken by the Court of Appeal to the question of authenticity of the physical 111 recording. We think it should also have been the approach taken to relevance of the disputed sounds as tending to prove an admission. The judge must conclude that the inference could reasonably be drawn if the evidence is to be admitted as relevant.
[45] In the present case, counsel for the appellant were correct to suggest that the Judge and the Court of Appeal, while referring to the materiality (or probative value, if accepted) of the evidence, did not address its relevance.[43] Indeed, there are passages in both judgments which suggest that the materiality of an admission was treated itself as establishing its relevance of the evidence. And the Crown submissions seemed to proceed on the same basis. An admission by the appellant would certainly be highly material to the issue before the Court. But the relevance of the disputed sounds depends upon their identification and authentication as an admission of the appellant. If the jury could not reasonably rely on the disputed sounds as evidence of admission, they do not tend to prove or disprove something of consequence to the proceeding. On that basis they must be excluded as irrelevant, however material they might have been if reasonably capable of being taken as an admission.
[43]The difference between materiality (connection with a fact in issue) and relevance (relationship to the facts sought to be established) was emphasised by Wigmore, although it has not generally been adopted by the case-law. See discussion in Montrose, “Basic Concepts of the Law of Evidence” (1954) 70 Law Quarterly Review 527, p 532. See also Heydon, Cross on Evidence (Australian looseleaf ed, 2009), para [1580] and Tapper, p 78.
[46] In many cases, where the relevance of evidence depends on whether it can reasonably be identified or authenticated, additional evidence to establish identity or authenticity may be anticipated. Section 14 provides for the provisional admission of evidence, “subject to evidence being later offered which establishes its admissibility”. This procedure was proposed by the New Zealand Law Commission to recognise that “the practicalities of court proceedings are such that, at the time evidence is adduced, other evidence may not have established its admissibility”.[44] The procedure under s 14 was followed by the Judge in admitting the physical recording on a provisional basis, subject to further anticipated evidence about the custody of the 111 tape and the recordings taken from it going to authenticity. The procedure under s 14 is not appropriate in circumstances where, as here, there is no suggestion that further evidence is likely to bear on the question of relevance. Nor indeed do we think it would be appropriate to use the discretion conferred by s 14 where there is real risk that the condition will not be fulfilled and the evidence is prejudicial. In such cases, the Judge must make the preliminary determination of relevance before admitting the evidence at the trial. What does appear from s 14, however, is the legislative assumption that the threshold question of relevance (whether it is reasonably open to the jury to rely on the evidence as tending to proof or disproof) is not a “jury question” but a question for the judge.
[44]New Zealand Law Commission, para [C71].
[47] That the judge must be satisfied that it is reasonably open for the jury to find the authenticity or identity which makes the evidence relevant is also the assumption on which s 13 of the Act is based. It provides that the judge may determine the relevance of documents including by drawing any “reasonable inference” as to their “authenticity and identity”:
13 Establishment of relevance of document
If a question arises concerning the relevance of a document, the Judge may examine it and draw any reasonable inference from it, including an inference as to its authenticity and identity.
[48] As s 13 indicates, inferences as to authenticity and identity go to the relevance of a document. The section makes it clear that whether reasonable inferences as to authenticity and identity (which establish relevance) can be drawn are matters for the
judge. As the New Zealand Law Commission said in its commentary to what has become s 13:[45]
Authenticity is in the first place an aspect of relevance, and therefore of admissibility. Unless a document is authentic – that is, the document is what it purports to be – it is irrelevant and inadmissible.
[45]New Zealand Law Commission, para [C69]. The commentary indicates that the purpose of s 13 is to abrogate the common law rule requiring authenticity of a document to be proved by evidence extrinsic to the document. The judge is therefore empowered to examine the document and draw any reasonable inferences about authenticity from the document itself. The Commission comments at para [C70]: “Thus a document that contains the necessary information can be self-authenticating.”
[49] In the case of the disputed sounds, it is not suggested that any additional relevant external evidence is available to identify the sounds as an admission. The recording in those circumstances is a document that has to be assessed for what it is in its own terms, subject to the assistance able to be provided by the expert opinion evidence as to its internal qualities or any help available from the context. As has already been described, the expert opinion evidence is unable to provide such identification or authentication because the experts are able to say only that they cannot exclude the possibility that the sounds are the suggested words. And they say that the hypothesis that they are not words at all is equally possible.
[50] In Hollingham v Head Willes J expressed the view that relevance ends when “speculation begins”: [46]
I am of opinion that the evidence was properly disallowed, as not being relevant to the issue. It is not easy in all cases to draw the line, and to define with accuracy where probability ceases and speculation begins: but we are bound to lay down the rule to the best of our ability. No doubt, the rule as to confining the evidence to that which is relevant and pertinent to the issue, is one of great importance, not only as regards the particular case, but also with reference to saving the time of the court, and preventing the minds of the jury from being drawn away from the real point they have to decide.
Similar thinking is behind case-law in the United States that recordings must be rejected if so inaudible and indistinct that they invite the jury to speculate as to the contents.[47] The same, more general, approach is to be seen in the reasons expressed by Dixon CJ in Jones v Dunkel,[48] quoting the Full Court of the High Court of Australia in Bradshaw v McEwans Pty Ltd,[49] that relevant evidence “… must do more than give rise to conflicting inferences of equal degrees of probability so that the choice between them is mere matter of conjecture”:[50]
…[T]he law … does not authorise a court to choose between guesses, where the possibilities are not unlimited, on the ground that one guess seems more likely than another or the others.
[46](1858) 4 CB NS 388 at p 391; 140 ER 1135 at p 1136.
[47]See “Omission or inaudibility of portions of sound recording as affecting its admissibility in evidence” 57 ALR3d 746.
[48](1959) 101 CLR 298 at pp 304 – 305.
[49](1951) 217 ALR 1 at p 5.
[50]Jones v Dunkel at p 305.
[51] Clearly, there is a continuum, as Willes J suggests. Unreliability is itself a ground for exclusion of relevant evidence under s 8 or the more specific ground for exclusion to be found in s 28 of the Act. In some cases, however, evidence may be so unreliable that its identification or authenticity cannot be resolved.
[52] In Cleland v R the issue was whether a confession denied by the accused should have been admitted.[51] The other members of the High Court of Australia decided the question on the basis of voluntariness.[52] Murphy J however considered that there was a prior question to voluntariness: whether the statement had been made. He took the view that if the judge was not satisfied that a confession had been made, he “should not have to decide voluntariness on the hypothesis it was made”:[53]
Such artificiality should not be part of criminal justice. Neither should he have to exercise on an artificial basis any discretion to exclude on grounds of unfairness or unlawful or improper conduct.
Therefore the question whether a confession or admission was made is for the purposes of admissibility to be decided by the trial judge (along with the question of voluntariness, if voluntariness is suspect). A finding against its making requires exclusion, a finding that it was made is not binding on the jury.
[51](1982) 151 CLR 1.
[52]Gibbs CJ, Wilson, Deane and Dawson JJ.
[53]At para [5].
This view, with which we agree, seems to us to apply equally to the hypothetical basis on which the Court of Appeal would have invited the jury to consider the disputed admission. It also suggests that, once identity or authenticity cannot reasonably be inferred, it is “artificial” to deal with challenges to the admission of the evidence for unreliability or prejudice as though the evidence were relevant, when it is not. That is why we consider that the question of admissibility in this unusual case is preferably considered first under s 7, rather than s 8.
[53] In summary, whether the evidence is accepted as tending to prove or disprove the matter in issue, is ultimately a matter for the jury or trier of fact. If such acceptance is not reasonably available to the jury, however, the statement is irrelevant and must be excluded. Evidence is not a “jury question” until its relevance has been accepted by the judge. If the disputed sounds cannot reasonably be taken to be speech (an admission), they are not relevant because they do not tend to the proof of the fact of the appellant’s responsibility for the deaths, the purpose for which they are tendered.
Evaluation of relevance
[54] Three reasons weighed principally with the Court of Appeal in concluding that the evidence of the disputed words on the recording should be admitted. They were the fact that the “hypothesis” was “open” on the expert evidence; the fact that the appellant spoke on an exhaled breath without vibration of the vocal folds when giving his telephone number; and the “contextual sense” of the sentence, which left it open to the jury to decline to accept that the sounds had been made by random movements of the tongue and lips in an exhaled breath.[54]
[54]At para [255].
[55] The whispered telephone number does not strike me as support for identification (and therefore relevance) of the disputed admission. The experts declined to rely on this comparison because the disputed admission lacked any contextual confirmation within the recording such as was available in respect of the telephone number (which was repeated after questioning from the operator). None of the experts was willing to say that the sounds constituted speech, nor what any such words were, beyond acknowledging that, with effort, they could be heard by someone listening for them.
[56] In some cases, evidence external to a document such as the recording may be available to show its identity or authenticity. That has not been suggested here, except perhaps by the Court of Appeal’s view that the admission makes contextual sense. That reference to context cannot be to the internal context of the recording, because there is nothing in the telephone conversation itself which provides context for any such statement. Indeed, a reference to having “shot the prick” is inconsistent with the immediate context in which the appellant is saying that all his family is dead. Nor do the questions of the ambulance officer elicit a response dealing with culpability. They are concerned with locating the street address and obtaining a telephone number. The reference in the Court of Appeal judgment to “context” seems, rather, to look to the wider trial context. If the wider context is indeed to be looked at (a proposition of some doubt since it seems improperly question-begging if dependent on the nature of the charge), the alleged admission appears in fact to be entirely a-contextual. The suggested admission was as to one shooting, when five people had died. The suggested statement was whispered under breath by the appellant during a 111 call when the Crown case is that he was feigning distress as part of a calculated plan entailing fabrication of alibi. If made, it is the only confessional statement made by the appellant. These matters of context are quite at odds with the claimed admission and provide no safe external check upon it.
[57] We do not think the Court of Appeal was right to consider that, in a case where relevance was dependent on a double contingency as to the identity of the sounds as speech and their content as an admission, it was sufficient discharge of the responsibility of the Judge to admit the evidence if of the view that these were “hypothes[es] … open on the expert evidence”. We proceed on the basis that the words described can be heard with a little effort, as Professor French put it. Certainly, Panckhurst J and the Court of Appeal Judges seem to have been of the view that they were there to be heard. Panckhurst J said that listening to the recording had been of assistance to him in the s 8 consideration.[55] The Court of Appeal seems to have been similarly assisted in coming to the conclusion that the hypothesis was “open” by hearing the recording, while remaining concerned about suggestibility in relation to the jury. For the purposes of further appeal to this Court, we have thought it appropriate to accept these assessments of the High Court and Court of Appeal that the disputed sounds can be heard as the admission contended for by the Crown, without further reviewing the recording. The assessment of the High Court and Court of Appeal is not inconsistent with the evidence of witnesses at the voir dire that the sounds can be heard as the suggested words. But we do not consider that such impression could reasonably overcome the thrust of the evidence of the experts that drawing such an inference would amount to conjecture.
[55]At para [56].
[58] In our view, counsel for the appellant were right in the submission that Panckhurst J and the Court of Appeal failed properly to address the threshold question of relevance. The question of relevance is fundamental, as s 7 makes clear. It is an aspect of trial fairness that, while the issue of guilt is for the jury, a fair trial requires that the jury should not be invited to act on evidence which is irrelevant. It is necessary for the judge to determine whether tendered evidence is relevant or irrelevant. The test here to be applied by the judge is whether it would be open to the jury, acting reasonably, to rely on the disputed words as an admission. This threshold question is not properly addressed by pointing to expert evidence which cannot exclude one hypothesis or the other. If the judge is not satisfied that the jury can reasonably conclude that the sounds are an admission, then in our view they must be excluded as irrelevant. This is the test as to authenticity applied by the Court of Appeal in relation to the physical recording. It equally applies to authentication or identification of the disputed words. If the material is admitted, the jury will have to speculate about its content. That is clear on the voir dire evidence of the experts who listened to the recording under ideal conditions. They thought it impossible for expert or lay listeners to resolve the disputed words. No contextual external evidence that might provide assistance in resolving the dispute was suggested in argument. It could only be the context provided by the ultimate issue at trial, which cannot be a permissible check on the authenticity of the disputed sounds as an admission. If the Judge cannot be satisfied that the jury could reasonably conclude that the sounds are the admission contended for, he is obliged to exclude them from the evidence. This is not a case where error in fact-finding as to authenticity or identity of real evidence can be regarded with equanimity. If taken to be an admission, the disputed sounds go directly to the ultimate issue of guilt. It would be open to the jury to convict on the basis of the admission alone.
[59] As Willes J suggested in Hollingham v Head, relevance ends when “speculation begins”. The line between relevance and speculation, as he acknowledged, is not an easy line to draw in all cases, but is one that must be attempted. Here, we do not think it was attempted by the High Court or Court of Appeal. The conclusion as to relevance is different from one of materiality. The experts who have heard and reheard the recording and who are entirely objective are unable to exclude either the hypothesis that the sounds are speech or that they are not. They acknowledge that analysis of the recording presents particular difficulty. Ms Cauley, Dr Foukes and Dr Innes, who listened unprimed to the recording to begin with did not pick up the disputed words despite the ideal conditions in which they were able to hear and rehear the material. Indeed, their tentative assessments are remarkably similar in hearing the word the Crown suggests to be “shot” (if it was a word) as “can’t”. The principal expert for the Crown, Professor French, believed that it was not possible for either a layperson or an expert to resolve whether the sounds were speech or meaningless breath “either way”. In his view there was no justification to put the interpretation “I shot the prick” to the jury. On that basis, we consider that the disputed sounds are not reasonably capable of being taken by the jury to be an inculpatory statement. We would therefore exclude them as irrelevant and inadmissible under s 7(2) of the Evidence Act.
Relevance of recording as to state of accused at the time
[60] Evidence may be relevant to prove one fact, but not another. In the present case, the principal focus has been on the admission of the disputed sounds as evidence of an admission. They are also however directly relevant to the state of the accused at the time, a purpose in respect of which their identification or authenticity (save as to proof of the physical recording) is not in issue. The admission in evidence of the disputed sounds for this purpose must be addressed under s 8.
[61] The recording is relevant to the question of the appellant’s condition at the time of the call. That is itself a subsidiary issue in the trial from which inferences may properly be drawn in relation to the question of culpability. On this basis, there is no preliminary fact to be considered in relation to relevance, once questions of authenticity of the physical recording have been concluded since there is no issue as to identity of the caller. Admission for this purpose, given the risk that the jury will speculate as to the disputed sounds, would however be contrary to s 8. The probative value of the 111 call in relation to the state of the accused at the time is not affected by excision of the disputed sounds. The prejudice which would be caused by speculation as to what they contain could be substantial (as is discussed below in relation to s 8). Nor did the Crown suggest that the disputed sounds should be left in the recording other than for the purposes of supporting the contention that they amounted to an inculpatory statement. We would exclude them by excision of that portion of the recording.
Section 8 exclusion
[62] While relevance determines “whether evidence could relate to an issue”, exclusion under s 8 is concerned with whether the connection between the evidence and proof is “worth the price to be paid by admitting it in evidence”.[56] Under s 8 the judge must exclude evidence if its probative value is outweighed by the risk that it will have an unfairly prejudicial effect on the proceeding.
[56]Australian Law Reform Commission, para [315].
[63] Sections 28, 45 and 46 provide particular examples of evidence which may be excluded and are helpful analogies in applying s 8. Although counsel for the appellant argued that the disputed sounds should also be excluded in application of s 28, the exclusionary rule provided by s 28(2) does not readily fit the case here where it is not the circumstances in which the disputed admission is made that gives rise to concerns about its reliability, but rather its interpretation.[57] Once prima facie unreliability is shown, the judge must exclude the statement unless satisfied on the balance of probabilities that “the circumstances in which the statement was made” were not likely to have affected its reliability.
[57]It is not necessary to determine whether s 28(2) could apply to unreliability for lack of intelligibility. It should be noted however that the definition of “circumstances” contained in s 16 are broad and include the nature of the statement and its contents.
[64] Similar analogy is provided by s 46, dealing with the admissibility of voice identification evidence. Such evidence offered by the prosecution in criminal proceedings is inadmissible “unless the prosecution proves on the balance of probabilities that the circumstances in which the identification was made have produced a reliable identification”. The New Zealand Law Commission commentary to what became s 46 refers to recent research suggesting that voice identification is “generally even less reliable than visual identification”.[58] The section therefore provides a presumption of unreliability.[59]
[58]New Zealand Law Commission, para [C229].
[59]New Zealand Law Commission, para [C229].
[65] A comparable approach is taken in relation to the admission of visual identification evidence by s 45. If a formal procedure for obtaining visual identification is followed, the evidence is admissible unless the defendant proves on the balance of probabilities that it is unreliable. If such formal procedure is not followed, the evidence is inadmissible unless the prosecution proves beyond reasonable doubt that the circumstances in which it was made have produced a reliable identification. This higher standard in the case of informal visual identification is statutory adoption of the common law suspicion of such identification evidence.
[66] Even had our conclusion on relevance been that the evidence was reasonably capable of supporting proof of the existence of an inculpatory statement, we would be of the view that the disputed sounds should properly have been excluded in application of s 8(1).[60] But, in considering the application of s 8(1), it is clear that the probative value of the evidence, when contrasted with its prejudicial effect, is slight. In our view the evidence falls short of being sufficiently probative for reasons already traversed in relation to relevance.
[60]I agree with the submission for the Crown that s 8(2), relied on by Ms Cull, is not in point. As the New Zealand Law Commission commentary at para [C61] makes clear, it is concerned to reform the law contained in the line of authority culminating in Lobban v R [1995] 1 WLR 877 (PC). Under the Evidence Act the right of the accused to present evidence as part of his case is not absolute but is simply a factor to be considered in balancing probative value against unfairly prejudicial effect in excluding relevant evidence. The admission of Crown evidence which is relevant and not unfairly prejudicial does not affect the ability of the accused to present relevant evidence as part of his case, subject to s 8(1) and (2) and the other provisions for exclusion of relevant evidence.
[67] The prejudicial effect on the proceeding could be profound. The jury would, as Mr Raftery acknowledged, be entitled to find the accused guilty simply on the basis of an inculpatory statement unable to be proved to the satisfaction of experts or, in their estimation, of lay people. The case is unusual, as Professor French emphasised. There is no contextual explanation for the admission such as would give confidence in its going to the jury for their assessment. Even if rejected by the jury, the admission of the disputed sounds in evidence could nevertheless impermissibly affect the jury’s consideration of other evidence and would necessitate firm directions from the trial judge. The consequences of error cannot simply be shrugged off as part and parcel of fact-finding at trial, as the Court of Appeal seems to have been inclined to do. The rules of evidence, as the Act makes clear, seek to secure the “just determination of proceedings” consistently with fairness to parties and witnesses and the New Zealand Bill of Rights Act 1990 entitlement to fair trial.[61] The substantial risk of suggestibility, acknowledged by the High Court and the Court of Appeal and demonstrated by the principal Crown witness’s compilation of fabricated but meaningless apparent “words” from the same recording, compounds the problem. The Judge would be bound to direct the jury that it was unsafe to rely on the disputed sounds as a confession. In those circumstances, the admission of the disputed sounds could only operate to the prejudice of the defence.
Result
[61]Section 25.
[68] For these reasons, we consider that the disputed sounds are not admissible. The risks of jury speculation as to the content of the sounds, and the risk of the contentions put forward being available to them, make it necessary to excise the portion of the recording.
McGRATH J
[69] My reasons for joining the judgment of the Court delivered on 6 March 2009 differ from those of the Chief Justice, Blanchard and Wilson JJ. I have concluded that the sounds on the disputed portion of the tape recording of the 111 call made by the appellant amounted to relevant evidence in terms of s 7 of the Evidence Act 2006. I have decided that, nevertheless, the relevant portion of the tape should be excluded because, applying s 8 of the Act, the probative value of the material is outweighed by the risk that its treatment by the jury would have an unfairly prejudicial effect at the appellant’s trial.
[70] Section 7 of the Evidence Act provides:
7 Fundamental principle that relevant evidence admissible
(1)All relevant evidence is admissible in a proceeding except evidence that is –
(a) inadmissible under this Act or any other Act; or
(b) excluded under this Act or any other Act.
(2)Evidence that is not relevant is not admissible in a proceeding.
(3)Evidence is relevant in a proceeding if it has a tendency to prove or disprove anything that is of consequence to the determination of the proceeding.
[71] As its heading indicates, s 7 states the fundamental principle of the Act: relevant evidence is admissible. The Law Commission points out that all modern evidence codes commence with a provision in similar terms.[62] Section 7(1) provides that all relevant evidence is prima facie admissible. The only exceptions to admissibility are where the Act provides to the contrary. The section thus contemplates that evidence will be rejected only if it is not relevant or for reasons of policy stipulated in the Act.[63]
[62]New Zealand Law Commission, Evidence: Reform of the Law (NZLC R55(1), 1999), para [23].
[63]As Thayer put it in A Preliminary Treatise on the Law of Evidence at Common Law (1898), p 530:
The two leading principles should be brought into conspicuous relief, (1) that nothing is to be received which is not logically probative of some matter requiring to be proved; and (2) that everything which is thus probative should come in, unless a clear ground of policy or law excludes it.
[72] The internal definition of “relevant evidence” in s 7 is broadly expressed. As the Law Commission put it:[64]
“Relevant” evidence is defined as evidence that according to logic and common sense has a tendency to prove or disprove anything that needs to be decided in order to determine a proceeding …
[64]New Zealand Law Commission, Evidence Code and Commentary (NZLC R55 (2), 1999), para [C53].
[73] The test of “tendency to prove or disprove anything of consequence to the determination of the proceeding” requires no more than that, if accepted, it could rationally affect the probability of existence of a fact in the proceeding. Accordingly, there is no qualitative requirement for admissibility, in terms of probative force, once it is established that the material has the stipulated tendency. This reflects the common law position as discussed in cases such as DPP v Kilbourne[65] and Wilson v R.[66] The position is well expressed by the Australian Law Reform Commission in explaining what is now s 55(a) of the Evidence Act 1995 (Commonwealth):[67]
The definition requires a minimal logical connection between the evidence and the “fact in issue”. In terms of probability, relevant evidence need not render a “fact in issue” probable, or “sufficiently probable” – it is enough if it only makes the fact in issue more probable or less probable than it would be without the evidence – ie it “affects the probability”. The definition requires the judge to ask “could” the evidence, if accepted, affect the probabilities. Thus, where a judge is in doubt whether a logical connection exists between a fact asserted by evidence and a “fact in issue”, he should hold that the evidence is relevant if satisfied that a reasonable jury could properly find such a logical connection. An indirect connection with a matter in issue is sufficient …
[65][1973] AC 729 at p 757 (HL).
[66](1970) 123 CLR 334 at p 344 (HCA) per Menzies J.
[67]Australian Law Reform Commission, Evidence (ALRC No 26(1), 1985) (footnotes excluded from original).
[74] The question for this Court in relation to s 7 accordingly was whether a reasonable jury could find that there is a logical connection between the sounds recorded on the tape and whether Mr Bain committed the murders, that being the central fact in issue at the trial. If that connection were shown, the evidence will be relevant as its tendency will be to support the Crown’s case.
[75] I do not consider that s 13 of the Act assists in resolving the question of relevancy of the disputed position of the tape. Section 13 provides:
13 Establishment of relevance of document
If a question arises concerning the relevance of a document, the Judge may examine it and draw any reasonable inference from it, including an inference as to its authenticity and identity.
[76] What is in dispute is the content of a portion of the tape and whether the portion is relevant evidence. The dispute is not concerned with whether the portion of the tape is or is not genuine. Section 13 would be of assistance if the issue were whether the recorded passage on the 111 tape was authentic, that is a genuine recording of sounds emanating from the appellant. The provision would enable inferences as to authenticity and identity to be drawn from the tape recording itself without having to rely totally on extrinsic evidence. Section 13, in other words, allows the tape to be self-authenticating in relation to the authenticity and identity aspects of its relevance.[68] There being no doubt on these matters, there is no need to resort to s 13.
[68]The Law Commission took the view that the common law rule requiring authenticating of a document to be established by evidence extrinsic to the document no longer serves any useful purpose: New Zealand Law Commission, Documentary Evidence (NZLC PP22, 1994), para [39].
[77] The issue faced by the Court concerns the contents of the tape and whether it “tends to prove the relevant fact in dispute”. The disputed fact comes down to whether the sounds were words and, if so, the words the Crown asserts.
[78] The Crown experts, Professor French and Mr Harrison, say that the sounds have the characteristics of exhaled air without voicing or modification of airflow. In their opinion they can be heard as the words the Crown says they are. The possibility that they are speech, rather than simply audible out-breath, modified by a random and unfortunately sequenced series of tongue and lip movements, cannot in their opinion be discounted, particularly because at a later point during the 111 call Mr Bain uttered what are clearly words as he exhaled air. The defence expert, Professor Foulkes, likewise said that the section could comprise those words but might constitute no more than audible out-breath. He added that in his opinion it was not possible to resolve which it was. Dr Innes, in a report, said that most people listening to the tape would not hear the words unless expecting to hear them or told that they were there. She thought it would be dangerous to play the tape to the jury because technical knowledge was required to assess it.
[79] A tape is real evidence, the contents of which, as a general proposition, are proved by playing it to the jury. It is well established that deficiencies or indistinctness in a recording may make it necessary and desirable to play tapes more than once to get an understanding of their contents.[69] In general that will also be the case where inaudibility arises from the nature of sounds themselves rather than the recording. The jury at times will benefit from the assistance of expert evidence. That is the position in the unusual circumstances of this case. The expert evidence supports the proposition that the sounds may be an exhalation of breath containing the incriminating words as speech intended to be sotto voce. A jury would find contextual assistance in determining if this was so from another section of the tape, which records what are plainly words spoken in conjunction with an audible exhalation. The wider context, being the coincidence that a short time prior to the 111 call five people had been shot at the house where the call came from, would also be a legitimate source of assistance for the jury. So would the jury’s view of the likelihood of random tongue and lip movements making such contextual sense. In this context I am satisfied that the disputed material is reasonably capable of influencing a jury’s assessment of the probability of the existence of a disputed fact at the trial and that it meets the test of relevance in terms of s 7 of the Evidence Act. It was accordingly admissible, subject to being excluded on the policy grounds foreshadowed.
[69]R v Menzies [1982] 1 NZLR 40 at p 49 (CA); Butera v Director of Public Prosecutions (Vic) (1987) 164 CLR 180 at p 187 (HCA) per Mason CJ, Brennan and Deane JJ.
[80] I turn to s 8 of the Act which relevantly provides:
8 General exclusion
(1)In any proceeding, the Judge must exclude evidence if its probative value is outweighed by the risk that the evidence will –
(a)have an unfairly prejudicial effect on the proceeding; or
(b)needlessly prolong the proceeding.
(2)In determining whether the probative value of evidence is outweighed by the risk that the evidence will have an unfairly prejudicial effect on a criminal proceeding, the Judge must take into account the right of the defendant to offer an effective defence.
[81] Section 8(1)(a) requires that a judge undertake a balancing exercise in which the probative value of the disputed relevant evidence is weighed against its prejudicial effect. The Court must assess the importance and impact of each of the conflicting values so they are each given due weight in the scales. If the risk of unfair prejudice outweighs probative value, the Court is required to exclude the evidence.
[82] Probative value of evidence is concerned with its force rather than its tendency as under s 7(3). As the Law Commission said:[70]
Probative value will depend on such matters as how strongly the evidence points to the inference it is said to support, and how important the evidence is to the ultimate issues in the trial.
[70]NZLC R55(2), para [C58].
[83] The sounds in question are indistinct and the experts have had difficulty hearing them. Two different possible interpretations are placed on the sounds by each of the expert witnesses for Crown and defence who have listened to them. The two Crown experts were not able to say from their technical analysis that they were words rather than simply sounds of exhaled breath. On the other hand, a Detective who played the tape in July 2007 recognised inculpating words in the sounds. I am, however, satisfied that the uncertainty of the conclusions of the expert witnesses, which are reached because of the limited audibility and indistinctness of the sounds themselves, impacts on their probative value and lessens the weight of the recordings as evidence of an admission in the balancing exercise.
[84] Turning to prejudice, it is important to bear in mind that s 8(1)(a) is concerned with unfair prejudice. As the Law Commission said:[71]
There must be an undue tendency to influence a decision on an improper or illogical basis, commonly an emotional one; for instance, graphic photographs of a murder victim when the nature of the injuries is not in issue. Evidence will also be unfairly prejudicial if it is likely to mislead the jury; for example, if it appears far more persuasive than it really is, as is occasionally the case with some types of expert and statistical evidence. The judge will need to consider whether any misleading tendency can be countered by other evidence that is likely to be available, or by a suitable direction to the jury. Whether evidence has an unfairly prejudicial effect must be considered in terms of the proceeding as a whole, and not just from the point of view of a particular party or a defendant.
[71]NZLC R55(2), para [C59].
[85] The principal consideration in relation to unfair prejudice arises from concern over the basis on which the jury will, if the evidence is admitted, overcome the problems of inaudibility and determine whether there is an admission on the recording. The Crown experts both thought it dangerous for the jury to be prompted (“primed”) when the recording was played as to the words they were listening for. Any suggestion of an interpretation of the words heightened the risk of a misinterpretation.
[86] The Court of Appeal accepted that suggestibility was an “obvious” risk of the jury wrongly construing the sounds. It set out a procedure by which it considered that risk would be adequately addressed:[72]
[258] We think it would be best if the jury first heard the tape without being primed, except perhaps with a request that they listen to it carefully and possibly advice (as recommended by Professor French) that they will hear some speech produced on breath. If they initially do not hear the disputed sounds as an inculpatory sentence (which we think is likely given past history) but, once primed, subsequently do hear the disputed sounds in this way, this should provide a graphic indication of the power of suggestion.
[259] The Judge concluded that after the tape has been played to the jury “unprimed”, expert evidence should be received concerning the interpretation of the disputed sounds. We agree.
[260] We have no doubt that in his summing up the Judge will warn the jury of the dangers of suggestibility.
[72][2009] NZCA 1, CA 571/2008 (William Young P and Chambers and O’Regan JJ).
[87] The prejudice, however, lies in the risk that in this unusual case, even with appropriate direction, the jury will not assess the evidence of whether there is an admission recorded on the 111 call with due caution, and will give what they hear on the tape more weight than it deserves. That risk is associated with the particular difficulty of suggestibility that arises from following the usual course approved in R v Menzies.[73] The sounds on the recording may not be treated as simply one aspect of the Crown case to be weighed with other Crown evidence, including contextual evidence. The impact of a superficially reached perception that there is an admission is likely to be seen as more persuasive than the taped evidence actually is in the context of a Crown case which is otherwise almost entirely a circumstantial one.
[73][1982] 1 NZLR 40.
[88] The risk in the end is that the consequence of the indistinct and inaudible nature of the sounds will cause the jury to leap to a conclusion that the sounds which, with effort, they can hear are inculpatory words without that being a conclusion that is logically based on all the evidence relevant to the preliminary question, but the result of superficial reasoning to which they have been diverted. The likely prejudicial effect is substantial and, to my mind, outweighs the probative effect of the disputed part of the recording as evidence.
[89] For these reasons I joined the Court’s judgment in this appeal excluding the evidence from the trial.
WILSON J
[90] I am in general agreement with the reasons of Elias CJ and Blanchard J and add the following observations.
[91] The requirement of relevance in s 7 is a very low threshold for admissibility. It is satisfied if the evidence, either on its own or (as will usually be the position) in conjunction with other evidence, is reasonably capable of proving or disproving a proposition which is in issue. Such reasonable capability will almost always be self‑evident. If not, it must be determined by the Judge, whether sitting alone or with a jury.
[92] The disputed sounds are a discrete item of evidence. The meaning now sought to be given to them was not identified, even as a possible meaning, by any witness at the first trial. Neither the “internal” context of the telephone call nor the “external” context of the killings assists in their interpretation.[74] The Crown and
defence experts agree that, even with the benefits of their expertise and experience and their repeated resort to the latest technology, they cannot on the probabilities attribute to the disputed sounds the interpretation for which the Crown contends. The experts also agree that, if the jury are “primed” with the suggested interpretation, there is a risk that they will therefore give that meaning to the sounds. Given the expert evidence, it cannot be said that the disputed sounds are reasonably capable of proving that the appellant said what the Crown claims he said. They therefore cannot satisfy the relevance requirement of s 7.
[74]For the reasons discussed by the Chief Justice at [56].
[93] Even if the evidence were admissible under s 7, it must be excluded under s 8(1)(a) if “its probative value is outweighed by the risk that the evidence will … have an unfairly prejudicial effect on the proceeding”. The word “outweighed” evokes the image of a set of scales. Both probative value and unfair prejudice must be assessed to enable their placement on these notional scales. It can then be seen whether the prejudicial effect outweighs the probative value.
[94] The disputed sounds may well not be words at all. Even if they are, it is uncertain what those words are. Suggesting an interpretation may cause the jury to reach a view which they would not reach “unprimed”, whether the “priming” occurs before or after the tape is heard for the first time.The sounds therefore have, at best, very limited probative value. Conversely, the prejudicial effect of admitting the evidence, when the Crown case is otherwise a circumstantial one, would be very great. As Mr Raftery properly acknowledged in argument and as the Chief Justice has noted,[75] the jury would be entitled to find the appellant guilty simply on the basis of the inculpatory statement which the disputed sounds are said by the Crown to represent.
[75]At para [67].
[95] But that is not the end of the matter. As Mr Raftery correctly submitted, the question is not simply whether the evidence is prejudicial; evidence is thought to be prejudicial to the interests of the opposing party or there is no point in calling it. Rather the test is whether the evidence is unfairly prejudicial. Counsel submitted that, for a number of reasons, admission of the evidence of the disputed sounds would not be unfairly prejudicial. First, cross-examination and directions by the trial Judge would protect against suggestibility. Secondly, the evidence of Professor French at the voir dire demonstrated that he is a cautious expert. Thirdly, the Court could rely on the jury to perform its role responsibly. Fourthly, Detective Wood had heard the alleged admission when listening to the recording “unprimed”. Fifthly, the trial Judge and the Court of Appeal had also heard it, albeit after being “primed”. These are all valid points. But even collectively they cannot obviate the unfairness of requiring Mr Bain to accept the risk of the jury thinking that they had heard the alleged admission in the disputed sounds, after being “primed” to do so.
[96] In delivering the judgment of the Court of Appeal in R v During,[76] Turner P said that “it is only when the probative weight of the evidence is small compared with the prejudice against the accused which it may excite in the minds of the jury, that this ground of rejection can be invoked.”[77] His observation is equally applicable to s 8(1)(a), and is particularly apt in its application to the present appeal. The probative value of the disputed sounds is very modest, but the risk of prejudice resulting from their introduction into the trial is very great. There is more than a minor imbalance; the scales come down firmly on the side of exclusion of the evidence. Even if the disputed sounds were admissible under s 7, they should plainly be excluded under s 8.
GAULT J
[76][1973] 1 NZLR 366.
[77]At p 375.
[97] The tape recording of the 111 telephone call made by the appellant to the ambulance service includes sounds which the Crown contends constitute a statement against interest. It may be however that they do not constitute speech at all and are breath exhalations reflecting, or intended to reflect, the appellant’s state of agitation at the time.
[98] Section 7 of the Evidence Act 2006 contains the fundamental rule of evidence that only relevant evidence is admissible. It provides:[78]
Evidence is relevant in a proceeding if it has a tendency to prove or disprove anything that is of consequence to the determination of the proceeding.
[78]Section 7(3).
[99] I consider the application of this provision calls for a judicial assessment of whether the statutory test is met and is not helped by the imposition of a standard of persuasion. Further, I do not consider anything is gained by postulating what assessment a jury might make.
[100] There may be an issue of whether the sounds in issue, insofar as it is proposed to rely on them as a statement, amount to evidence at all. But the recorded conversation is real evidence and the sounds have significance either as speech or breathlessness so that I do not consider they are easily ruled inadmissible as irrelevant.
[101] I am not persuaded that evidence that the sounds constitute speech does not exist at all. The evidence of the experts is that the sounds would not be heard as speech by a lay person not “primed”. But at least some of them accept that, once told of it, listeners may discern the claimed interpretation. And the sounds were so interpreted without prior suggestion by the police officers that first raised the issue. That speech is not identified without prior suggestion does not mean it is not there.
[102] Accepting relevance by having a tendency to prove the character of the sounds indicates little about the probative force of the evidence. That must be assessed on the evidence before this Court and calls for consideration of s 8 of the Act. That section contains the long-standing rule that: “In any proceeding, the Judge must exclude evidence if its probative value is outweighed by the risk that the evidence will – (a) have an unfairly prejudicial effect on the proceeding …”.
[103] The evidence is that lay persons, including jurors, would not be able to determine without assistance whether the disputed sounds constitute speech or mere exhalations of breath. The expert witnesses are unanimous, that they are unable to determine that issue and therefore they could not assist a jury to do so. Therefore, I cannot see how the tendency of the evidence to prove speech could be manifested. Members of the jury, aided or unaided, might hear and interpret for themselves the sounds. That would reflect some probative value in the evidence. There is a real risk however that a jury would accept what is suggested to them and find that the sounds constitute an admission rather than breath exhalation when the evidence before us is that that cannot be done. Such a finding would be speculation unfairly prejudicial to the accused. That risk clearly outweighs the probative value of the evidence.
Solicitors:
Duncan Cotterill, Auckland for Appellant
Crown Law Office, Wellington
18
5
0