Privacy, Self-Incrimination and the Constitution – III: “Volition”

In the last post, we saw that in Kathi Kalu Oghad, the Supreme Court held that what distinguished a compelled fingerprint scan from compelled self-incriminatory testimony was the issue of volition: one’s fingerprint pattern was independent of one’s volition, while testimony (spoken or written) was not. In linking volition with changeability, the Court seemed to adopt the crime-control model as the basis of Article 20(3). The reason why non-volition brought fingerprints out of the scope of Article 20(3) was because the accused was powerless to change or affect them in any way; in other words, fingerprints had an objective, verifiable existence beyond the control of the accused, and consequently, were not amenable to deception or falsification. Thus, because of their accuracy (as opposed to spoken or written testimony), even producing them through compulsion passed constitutional scrutiny.

In linking voluntariness with veracity (via changeability), the Court was tapping into an established common-and-statute law tradition. Consider Section 24 of the Evidence Act which, in popular parlance, is understood to exclude involuntary confessions. Section 24 renders irrelevant confessions that are obtained through “inducement, threat or promise… proceeding from a person in authority and sufficient… to give the accused person grounds… for supposing that by making it he would gain any advantage or avoid any evil of a temporal nature in reference to the proceedings against him.” The purpose is clear: there is a good chance that statements made in order to obtain a benefit or avoid an evil are false. More specifically, Halsbury’s Laws of England define a person’s statement to be “voluntary” as long as “it has not been obtained from him by fear of prejudice or hope of advantage, exercised or held out by a person in author­ity, or by oppression.” This rule – in almost exactly the same wording – was also laid down by the Privy Council in the 1914 case of Ibrahim v The Queen. Again, the concern seems to be – following the crime-control model – that fear of prejudice and hope of advantage are likely to taint the accuracy of the statement in question.

It is in this context that the 1979 Canadian Supreme Court decision of Horvath v The Queen assumes tremendous importance. In Horvath, the appellant’s mother was murdered in her bed one morning. The appellant, who was seventeen years old, was arrested, and subjected to a three-hour interrogation. He then confessed, but soon retracted. The next day, he underwent a four-hour interview with a polygraph operator, during which he made certain self-incriminatory statements, and at the end of which he wrote a signed confession. At the trial, a psychiatrist testified that for certain periods during the interview, the appellant had been in a “light hypnotic state“. The question before the Court was whether statements made under and immediately after light hypnosis were “voluntary” or not. The trial judge held that the statements were inadmissible; the Appeals Court reversed, on the basis of the narrow definition of “voluntary” in Ibrahim v The Queen. The case went up in appeal before the Supreme Court. By a narrow, 4 – 3 majority, the Supreme Court allowed the appeal, reversed the decision of the lower court, and restored the trial court’s order of acquittal. The dissenting judges based their decision on a strict understanding of “voluntary”, as per Ibrahim. They held that Ibrahim referred to a closed set of categories that could not be judicially expanded, as per certain other Canadian decisions.

However, it is the majority opinion of Beetz and Pratte JJ that is of considerable interest to us. Analysing the effect of the “interview”, the two judges concluded that there was a similarity between light hypnosis and narco-analysis, in the sense that “emotional controls are less inhibited than they are in a fully conscious state and the subject is assisted to say what he was not able to say voluntarily because of the emotional intensity of it.” The use of the word “voluntary” here is crucial, because it suggests a different understanding of the term, one which goes beyond fear of prejudice or hope of advantage. Immediately afterwards, the two judges went on to develop their theory of “voluntariness”, arguing that the “principle” underlying the Ibrahim rule necessitated its expansion when required:

“… if a mere threat or a mere promise are considered to have such an impact on mind and will as to render involuntary the confession of an accused who is not under induced hypnosis, whose state of conscious­ness has not been altered, who is in full and voluntary control and possession of his faculties and who is therefore quite capable to resist sugges­tion, then Horvath’s statement made under hypno­sis in the circumstances described above was a fortiori involuntary even though the cause of involuntariness was neither hope of advantage nor fear of prejudice.”

In other words, something was “voluntary” only if it was done under full control and possession of one’s faculties. What is the basis for this understanding? The Court held that:

“… through the use of an interrogation technique involving certain physical elements such as an hypnotic quality of voice and manner, a police officer has gained unconsented access to what in a human being is of the utmost privacy, the privacy of his own mind… it is my view that this was a form of violence or intrusion of a moral or mental nature, more subtle than visible violence, but not less efficient in the result than an amytal injection administered by force.”

Here is a new and powerful understanding of the interests protected by the voluntariness requirement. It is not, after all, about accuracy, but about privacy: the right of every person to choose whether or not to reveal what is in their minds, to keep their mental maps hidden from view, or to reveal it to the world, as they wish. Unsurprisingly, then, towards the end of its judgment, the Court expressly abandoned the veracity basis of the voluntariness requirement:

“Apart from the untrustworthiness of confessions extorted by threats or promises, other policy reasons have also been advanced to explain the rejec­tion of confessions improperly obtained. But the basic reason is the accused’s absolute right to remain silent either completely or partially and not to incriminate himself unless he wants to. This is why it is important that the accused understand what is at stake in the procedure. In a voir dire, voluntariness not veracity governs admissibility. Dr. Stephenson’s expert opinion as to the truthful­ness of Horvath’s confessions is accordingly not determinative of the issue of admissibility… I wish to add that the accused’s absolute right to remain silent includes not only the right to censor any information which is on his conscious mind but also the right to control the administration of artificial processes whereby unconsciously held information might emerge into consciousness. In this sense, the accused has the absolute right to censor his unconscious mind as well as his con­scious one.”

Let us return to Kathi Kalu Oghad. We can now see that the volitional requirement in that case could be read in one of two ways. Volition as non-changeability conforms to the narrow definition of “voluntariness” in the common law, based upon the crime-control model, and permissive of (arguably) narco-analysis, brain-mapping and other such intrusive tests that escape unconstitutionality by being neither threatening some prejudice, nor promising some advantage. On the other hand, volition as understood by the Canadian Supreme Court in Horvath, based upon the due-process model, is broader, and focuses on the right of the accused to mental privacy, to choose precisely which part of his mental landscape he or she wishes to reveal to the world. The tension between these two approaches must be kept in mind when, in the next post, we discuss the most recent Supreme Court decision on 20(3), one that was directly about narco-analysis and brain-mapping: the 2010 case of Selvi vs State.