DAVIES.html

The report of the Working Group on Criminal Trial Procedure did not discuss the right to silence in any detail or arrive at any conclusion about it. There were, it appears, two reasons for this. First it would have required considerable research and time and, as the report in effect remarks, the members did not have the luxury of either. Secondly, at the time this report was published, reports from the New South Wales Law Reform Commission and the Western Australian Law Reform Commission on this topic were expected. The former has not yet appeared; the second is referred to below.

Whilst I therefore make no criticism of the Working Group's failure to deal with this topic I think that that failure is, in retrospect, unfortunate in two respects. The first is that this is an area of the law which is in need of substantial reform and nothing which has yet emerged from any law reform body shows any real understanding of this. The second is that a true appreciation of the need to reform the present law in this respect would have led, in turn, to a recommendation of the most useful, perhaps even the only effective, means of implementing that part of a pre-trial regime which provides for disclosure by a defendant, namely providing for a jury to be capable of drawing an adverse inference from a failure to disclose and for a defendant to be told that a failure to disclose may result in such an inference being drawn. Both of these are discussed in what follows.

No rule has been the subject of as much confused thinking as the right to silence. That confusion results in and from a failure to accurately define the right and has, I believe, been in substantial part, a source of conclusions by law reform bodies in New South Wales,^[1] Victoria^[2] and Western Australia^[3] that it should not be substantially changed. Once the true nature of the so-called right is revealed it can be seen that it has no logical, common sense or historical rationale, that its precise ambit is uncertain and that, consequently, it is in need of reform. Let me start, then, by saying what it is and then I shall say what it is not.

The so-called right to silence is, and is no more than, an immunity of a criminal defendant from adverse inferences being drawn from his or her silence, an immunity which may arise at one of two stages; when being interviewed by a police officer or other person in authority and in court.

It is not, and is not part of, an immunity from compulsion to answer questions or give evidence. That immunity has two legitimate rationales. The first is respect for human dignity; to force someone to speak is arguably to infringe a basic human right. The second is a more pragmatic one, the unreliability of a forced confession.

That these immunities are antithetical can be seen from the fact that, if juries were permitted to draw an adverse inference from a criminal defendant's silence in answer to a suggestion or accusation there would be less incentive for the interviewer to resort to compulsion or trickery to obtain a confession. Conversely, refusing to permit the drawing of such inferences tends to encourage such behaviour.

That they have no legitimate historical connection can be seen from their respective histories which I trace briefly in Part 4.^[4] In short, whereas the immunity from compulsion is of ancient origin, existing in the courts of Chancery long before the time of the Star Chamber, the immunity from adverse inferences is a 20th century invention which appears to owe little, if anything, to the privilege against self-incrimination. It would be none the worse for that if it had some logical or common sense rationale but it doesn't, a point which I take up under the next heading.

Nor is the immunity from adverse inferences a right not to speak or an incident of any such right for there is no such right. There is a tendency to slip into the error that thet immunity from compulsion to speak is a right not to speak and then to say that the immunity from adverse inferences from silence is an incident of that right. But an immunity from compulsion to speak is not a right not to speak. The difference between them may be seen by the application to each of the doctrine of waiver. It would be nonsensical to speak of waiving an immunity from compulsion; no-one would sensibly waive such immunity. But, if there were a right not to speak, it would be sensible to speak of waiver of that right; it could be waived simply by speaking. The failure to distinguish between these two, an immunity from compulsion to speak which has a legitimate rationale and a long history, and a right not to speak which has no rationale and has never existed, has been a major source of the confusion.

The so-called right to silence is therefore no more than an immunity from adverse inferences being drawn from silence; and to describe it as a right to remain silent or a right to silence is misleading. Such misdescriptions have, I believe, contributed to the confusion to which I have referred. Once it is accepted that the so-called right to silence is no more than an immunity from adverse inferences from silence and that it has no logical, common sense or legitimate historical connection with the immunity from compulsion to speak the rest, I think, falls into place.

Bear in mind here that I am not saying that an adverse inference should always be drawn, or should even be drawn in most cases from a defendant's failure to speak in response to an incriminating accusation or question or piece of evidence. On the contrary, I accept that such an inference should be drawn only where a denial, explanation or answer would be reasonably expected. My point is simply that there should be no immunity against such an inference ever being drawn.^[5]

In almost every sphere of human endeavour, other than in a criminal trial, it is thought to be common sense to draw an adverse inference from a failure to answer an incriminating question or accusation or incriminating evidence when a denial, explanation or answer would be reasonably expected.^[6] Unsurprisingly that sensible approach has been taken in civil cases where it is common to draw an adverse inference from a failure of a party to answer an accusation or evidence which he or she is capable of rebutting or explaining and might reasonably be expected to do.

So also in criminal trials, before the advent of police forces, the failure of a defendant to answer an incriminating question or accusation was commonly the subject of adverse inferences. And this has survived where the accuser or questioner is not a person in authority.

How does it come about then that where the questioner is a person in authority no adverse inference can be drawn from silence and, subject to a limited exception,^[7] no adverse inference can be drawn from a failure of a defendant to give evidence? The answer lies in a mixture of historical accident and distrust of juries.

There is no doubt that the immunity of a criminal defendant from compulsion to answer questions or give evidence derives historically from the privilege from self-incrimination which existed in ecclesiastical courts in England long before the Reformation, though it was honoured more in the breach than in the observance by the High Commission and the Star Chamber and there were exceptions to it. To the extent that it did exist, however, it had the rationales which I have already mentioned.

Even though, under the privilege against self-incrimination, a person could not be compelled to testify, his or her failure to do so could justify an adverse inference being drawn. That view probably prevails today in the witness privilege against self-incrimination; in many cases an adverse inference may be drawn from a refusal by a witness to give evidence in reliance on the privilege.^[8]

The immunity against adverse inferences, by contrast, emerged, for the first time, in the first half of this century, not as part of the privilege against self-incrimination, but because of the wording of the standard police caution which had been part of the Judges' Rules in England since 1912. The caution did not have as a purpose the prevention of adverse inferences being drawn from silence. Its purpose was twofold; to convey to the defendant that he or she was under no compulsion to speak and to satisfy the court that he or she knew that. Nevertheless the terms of this caution, it was said by the 1930's, were, in effect, an invitation to say nothing. Consequently to permit an adverse inference to be drawn from a failure to speak in answer to that invitation would make the caution a trap. So it was held that a failure to answer a police question or accusation, after a caution had been given, could not give rise to an adverse inference. However what began as an unintended interpretation of the caution became, in 1970, a principle of the common law, whether a caution was given or not. So the origin of the immunity from adverse inferences from silence in police questioning derives, illogically, from the terms of the caution which was never intended to have that effect. That is the historical accident to which I referred.

Nor was it until the 1950's that judges, generally, evidenced reluctance to comment on the failure of defendants to give evidence. Before that time it was quite common to do so. What caused this change of heart (there was no logical reason for the change) is by no means clear. It was no doubt, in part, influenced by the emergence of the immunity to answer questions or accusations, already referred to. But underlying the emergence of both of these immunities was an increasing distrust, by judges, of the capacity of juries to draw sensible unprejudiced inferences, even with guidance from trial judges. I shall return to this distrust a little later. But what I have said so far shows that the emergence of this immunity was not the result of any logical process or even of any incremental, easily explicable historical development.

They are numerous. I shall mention only those which have at least some superficial credibility. They are the burden of proof and the presumption of innocence, that removal of the immunity would involve, albeit indirectly, a form of compulsion and some general notion of fairness.

These cannot possibly justify the immunity from adverse inferences from silence at the first of its two stages, that is in response to police questioning, because the evidence from which the jury would be asked to draw the adverse inference would be adduced as part of the prosecution case. It is in this respect like evidence of an admission. Its strength may depend on whether the defendant gives evidence but its admissibility should not be in doubt.

Nor can these justify the immunity from adverse inferences from a failure to give evidence. Because of the onus and standard of proof no such inference could be drawn until the prosecution case is capable of supporting an inference of guilt. But a case which is not so capable would almost never proceed to the point where a defendant must decide whether or not to give evidence. And once it is so capable there must be cases where it would be reasonable to expect a denial, explanation or answer to be forthcoming from an innocent defendant.^[9]

This is a variant of the argument just discussed. It is that, to permit an adverse inference to be drawn from the failure to speak is, in effect, to compel the defendant to speak. But once it is accepted that an adverse inference may be drawn from such failure only where facts have been proved which call for a denial, explanation or answer, that is, only where such denial, explanation or answer would be reasonably expected from an innocent defendant, the argument that that expectation is a form of compulsion is seen to be absurd.

There is, of course, some substance in an unfairness argument, in respect of a police interview, where the terms of a caution are, in effect, an invitation to say nothing. But the terms of the caution may be put on one side for they may easily be changed and should be changed if reform is to take place.

There may also be an argument, in particular cases, that the way in which a police interview has been conducted makes the drawing of an adverse inference from silence unsafe and unfair. In recent times, since police interviews with suspects have almost invariably been electronically recorded, the incidence of complaints about impropriety in the form of interview have dropped dramatically. No doubt they may still occur but two points should be made about the relevance of such complaints here. The first is that, as mentioned earlier, we are not speaking here about whether in every case, or even in most cases, it is appropriate to draw an adverse inference but whether there should be an immunity from the drawing of such inferences in all cases. Secondly, it seems plain enough that police impropriety is very much more likely to result in an unreliable confession than in an unreliable inference from silence. So it would be more logical to exclude all confessions than to exclude all evidence about silence and that, of course, would be absurd.

A common weakness in these arguments, or in the way in which they are presented, is that they tend to start from particular cases. But my point is not, I repeat, that an adverse inference should always be drawn from silence; it is that there should be no absolute immunity against the drawing of such an inference in all cases. It is to the factors relevant to whether, in any case, an adverse inference should be drawn from silence that I now turn.

I have now said perhaps too many times, that the removal of the immunity does not mean that an adverse inference should or even could be drawn from silence in every case, or even most cases, of a question, accusation or evidence. On the contrary it should be drawn only when it would be reasonable to expect that an innocent person would have denied, explained or answered the allegation or evidence against him or her. And it could be drawn only when a reasonable jury could have that expectation.

Three factors will almost always affect whether, in a particular case, an adverse inference can be drawn and the strength of that inference. The first is the strength of the case against the defendant; the stronger the inference of guilt from other evidence, the more readily an adverse inference can and should be drawn from the defendant's silence. In the case of silence at court that inference should never be drawn unless there is already a prima facie case established. The second is the strength of the evidence that the defendant knew or must have known facts said to call for a denial, explanation or answer at the time when he or she was given the opportunity to respond. And the third is the credibility of any innocent explanation for the silence.

There are a number of other factors which, in particular cases, may affect this question. Some of these are the extent to which the case against the defendant was explained before he or she was called on to answer and the extent to which he or she must have been aware of that case from other sources; and whether the defendant had the opportunity of obtaining legal advice before being called on to answer and the nature of that advice.

None of these are matters which it is beyond the capacity of a modern jury to understand and apply to the consideration of the question whether an adverse inference should be drawn from silence. For if the immunity were removed, the question whether, in any case, an adverse inference should be drawn from a failure by a defendant to answer a question or accusation in a police interview or to give evidence would be a question for the jury. It would be for the jury to say, in any particular case, whether there was an explanation, other than guilt, reasonably open for the defendant's failure to answer a police accusation or question; or whether, having regard to the onus and standard of proof and the strength of the prosecution case, there was some explanation, other than guilt, reasonably open for the defendant's failure to give evidence.

Moreover it is easy to see why trial judges, steeped in the mystique of the so-called right to silence, and with a somewhat cynical view of the capacity of juries to draw only adverse inferences which are safe and unprejudiced, have been reluctant to leave such questions to juries. That, it seems to me, is the real reason for the immunity and for its perpetuation; in short, a distrust of juries.

But is this distrust justified? If it is we should be reconsidering the jury system, not perpetuating an absurd fiction that adverse inferences should never be drawn from silence in the face of incriminating questions, accusations or evidence. On the assumption that we are content to retain the jury system and juries as the sole judges of fact, I suggest in the next part of this paper how the law may be easily reformed to give effect to logic and common sense.

It follows from what I have said that the immunity should be removed. The common law position is now such that this can be achieved only by legislation. But the necessary legislation is simple and short. The law should permit an adverse inference to be drawn from silence either at police interview or in court when it would be reasonable to expect a denial, explanation or answer from an innocent defendant.

Two consequential changes necessarily follow. The first of these is that the terms of the police caution must be changed. It will still need to assure suspects that they are under no compulsion to answer questions but it will also need to warn them that, if they fail to answer an accusation or question which, if they are innocent, it would be reasonable to expect them to deny, explain or answer, that failure may be taken as an admission.

Secondly at the end of the prosecution case the judge will need to ensure that the defendant knows that he or she is under no compulsion to give evidence. But the judge will also have to ensure that the defendant knows that a failure to deny, explain or answer evidence against him or her, where the jury thinks that a denial, explanation or answer should reasonably have been given, may result in their accepting that failure as an admission of guilt or in their more readily accepting the prosecution case.

Some guidance should also be given to trial judges so as to ensure some consistency in directions about when it is reasonable to draw an adverse inference from a defendant's silence. In the first place juries should be told that they cannot draw any adverse inferences unless it would be reasonable to expect the defendant to deny, explain or answer the question, answer or evidence. But they should also be told that this will depend on the strength of the prosecution case apart from any such inference, and on the extent to which facts which are said to call for a denial, explanation or answer were within the personal knowledge of the defendant when given an opportunity to respond. And they should be told about any possible innocent explanation for the silence. But they must be told that, in the end, it is for them to decide whether, having regard to any possible explanation, they think it reasonable to draw an adverse inference from the silence.^[10]

The failure of the Working Group to consider the right to silence has resulted in the absence of any recommendation for its reform. That is unfortunate. The law should reflect underlying community values. In erecting an immunity from adverse inferences from silence, which is what people call the right to silence, the common law has departed from those values. There is a conflict in this respect between the law and the ordinary understanding and practice of the community.

Moreover the reality mocks the rule. Judges solemnly tell juries not to draw adverse inferences from a defendant's silence, knowing that they will. Juries draw such inferences notwithstanding instructions to the contrary because to do so is plainly common sense. And the rule as it presently exists is inconsistent and uncertain in its operation. There is inconsistency between its application at the first stage, the police interview, and the second, in court. And there appears now to be uncertainty as to the extent of its operation in that latter situation.^[11]

It is also unfortunate that, as it seems to me, the failure by the Working Group to consider reform of the right to silence resulted in the apparent rejection by it of a legislative solution to the question of pre-trial disclosure which included permitting the drawing of an adverse inference from a failure by a defendant to disclose his or her defence. But in practice the same principles should apply. No such inference should be drawn unless it would be reasonable to expect an innocent defendant to say what his or her defence is; and in considering that question the jury would have to consider any explanation for the failure to do so. It seems to me that to provide for the possibility that such an inference may be drawn is the only effective way of achieving a result in which mutual rather than unilateral disclosure is the norm rather than the exception.

* The views expressed in this paper are, in summary form, views expressed by me at much greater length in "The Prohibition Against Adverse Inferences from Silence: A Rule Without Reason?" Part I (2000) 74 ALJ 26 and Part II (2000) 74 ALJ 99. For ease of reading I have, for the most part in this paper, eschewed references to authorities. Those who are seeking authority for any of the propositions which follow should look for them in the above article.

¹ The Right to Silence, Discussion Paper 41, New South Wales Law Reform Commission, May 1998; to the extent that it defined the so-called right it did so, at least by implication, in terms of or as part of an immunity from compulsion to give evidence; see eg par 2.12, par 2.16, par 3.1, par 3.11; and see par 3.80.

⁵There is at present a complete immunity from the drawing of any such inference from silence when questioned by a person in authority (Petty and Maiden v The Queen (1991) 173 CLR 95) and an immunity from any such inference being drawn from a failure to give evidence except where the jury is capable of drawing such an inference from proved facts, evidence to contradict or explain which must be within the knowledge of the accused (Weissensteiner v The Queen (1993) 178 CLR 217; RPS v The Queen [2000] HCA 3).

⁸ That would seem to follow logically from the fact that, before a court will grant a claim of privilege, it must be satisfied that there is reasonable ground to apprehend incrimination, usually for perjury, if the witness is compelled to answer: Sorby v The Commonwealth (1983) 152 CLR 281 at 289.