Essentially therefore, the Court was holding two sets of ideas, one legal and one factual: volition as defined as choice (legal), and the distinction between physiological and mental faculties as the factor that set apart BEAP and lie-detector tests from fingerprints, blood samples, and DNA (factual). To complete the argument, it would have to find a constitutional bridge between the two, an argument for why the issue of choice became particularly salient in cases where what was at stake was interference with mental processes per se.
The Court built this bridge upon R.C. Cooper’s theory of intersecting rights under the Constitution. Following Cooper and Maneka Gandhi, it held that the nature, scope, and interpretation of Article 20(3) could not be considered in isolation from Article 21, which guaranteed the right to life and personal liberty. One aspect of personal liberty was the right to privacy:
In conceptualising the ‘right to privacy’ we must highlight the distinction between privacy in a physical sense and the privacy of one’s mental processes … We must recognise the importance of personal autonomy in aspects such as the choice between remaining silent and speaking. An individual’s decision to make a statement is the product of a private choice and there should be no scope for any other individual to interfere with such autonomy, especially in circumstances where the person faces exposure to criminal charges or penalties. Therefore, it is our considered opinion that subjecting a person to the impugned techniques in an involuntary manner violates the prescribed boundaries of privacy.68
The constitutional bridge between volition and interference with mental processes, therefore, was Article 21’s conception of mental privacy and the right to informational self-determination. As the Canadian Supreme Court had noted in Horvath, the right to remain silent ‘includes not only the right to censor any information which is on [the] conscious mind but also the right to control the administration of artificial processes whereby unconsciously held information might emerge into consciousness’.69 In other words, any individual, including an accused, had the right to informational self-determination, that is, the right to control information about herself.70
We are now in a position to understand the argument in Selvi, and the break that it made with the crime control logic of M.P. Sharma and Kathi Kalu. In Kathi Kalu, volition, defined in terms of changeability, had played a crucial role, and its definition was derived from the crime control model: information that you had no control over, or power to change, could not possibly be fabricated. In Selvi, on the other hand (although the Court did initially embark upon a discussion of the reliability of the investigative techniques), the Court framed volition in terms of choice. It then drew a distinction between physical privacy (blood samples, fingerprints, DNA) and mental privacy.71 Linking the word ‘witness’ to testimony, and understanding testimony as the impartation of information present within the mind of a person, the Court placed the concept of mental privacy—understood as autonomous mental processes and informational self-determination—at the heart of the guarantee against self-incrimination. And with this, it was able to undermine M.P. Sharma’s formalistic view of compulsion as well: it was not simply about whether a person was being compelled, in the narrow sense, at the time of the testimonial act, but about the entire set of circumstances under which the testimonial act came to be.72
At the heart of this reading of Article 20(3) was the understanding that the sparse phrase ‘no person shall be compelled to be a witness against himself’ had to be interpreted in the context of the Fundamental Rights chapter as a whole, and especially Article 21 of the Constitution. Through this, the due process model—which placed strong emphasis on human dignity and individual rights as being non-negotiable in the criminal process—came to shape the scope and meaning of the self-incrimination clause. The distinction with M.P. Sharma and Kathi Kalu could not be starker: crime control and colonial law on the one hand, due process and the Constitution on the other.
The judgement in Selvi, ultimately, was a vindication of the ideals expressed, perhaps most eloquently, by Justice Claire L’Heureux-Dubé, of the Canadian Supreme Court:
Although the search of an individual’s home is an invasion of privacy, and although the taking of fingerprints, breath samples or bodily fluids are even more private, there is no doubt that the mind is the individual’s most private sanctum. Although the state may legitimately invade many of these spheres for valid and justifiable investigatory purposes vis-à-vis the accused, it is fundamental to justice that the state not be able to invade the sanctum of the mind for the purpose of incriminating that individual. This fundamental tenet is preserved, in its entirety, by the principle against self-incrimination.73
IV. Due Process and the Transformative Constitution
Was Selvi correct in introducing the due process model as a constituent element of Article 20(3) of the Constitution? I argue here that it was. While elements of the guarantee against self-incrimination existed in the colonial era (as noticed by the Court in M.P. Sharma), they were partial in character, and subservient to the overall colonial imperative of law and order. This was not just in terms of their substantive content, but in the very structure of the criminal justice system in colonial India. This was a structure that vested concentrated power in the hands of the executive, regularly undermined the separation of powers, and intentionally denied to Indians the rights and liberties that were available to individuals in Britain. The clearest example of this was provided by the Lieutenant-Governor during the debates around the Criminal Procedure Code of 1873. While defending provisions that authorized sweeping police searches (of the kind that would never have been allowed in Britain), the Lieutenant-Governor noted that:
[such rights] were especially out of place in a country where it was not pretended that the subject enjoyed that liberty which was the birthright of an Englishman, and it was not intended to introduce rules into the criminal law which were designed with the object of securing the liberties of the people. That being so, His Honour thought they might fairly get rid of some of the rules, the object of which was to secure for the people that jealous protection which the English law gave to the accused. It seemed to him that they were not bound to protect the criminal according to any Code of fair play, but that their object should be to get at the truth, and anything which would tend to elicit the truth was regarded by the Committee to be desirable for the interests of the accused if he was innocent, for those of the public if he was guilty … for instance, His Honour did not see why they should not get a man to criminate himself if they could; why they should not do all which they could to get the truth from him; why they should not cross-question him, and adopt every other means, short of absolute torture, to get at the truth. They had already done a good deal in the direction of clearing away English prejudices, and the Committee proposed to make further concessions to common sense in the present Bill.74
This ‘colonial common sense’ was partially quoted by Justice Shah in his dissenting opinion in Shyamlal Mohanlal, soon after he had pithily summarized the state of colonial law:
The Evidence Act and the Code of Criminal Procedure were enacted at a time when the primary aim of the Government was to maintain law and order. The Legislature was merely a branch of the executive government, and was not in the very nature of things concerned with the liberty of the individual.75
Of course, for Justice Shah, this was a reason not to read in an expansive right against self-incrimination into Article 20(3). However, this was because he and, for that matter, the majorities in M.P. Sharma and Kathi Kalu continued to be in thrall to the unarticulated assumption that the Constitution changed nothing. This assumption, of course, is at entirely at odds with the argument of this section: that the framing of the Constitution was a transformative moment that elevated what had been a heavily diluted aspect of criminal procedure into a fundamental right, and one which was organized around ideas of freedom and autonomy.
To understand why this is so, we need to go beyond the text of the Ar
ticle. In stipulating that no person shall be compelled to be a witness against himself, the framers were using words that had a long, almost ancient, vintage. While the most familiar predecessor was the Fifth Amendment of the US Constitution,76 the framing of the self-incrimination clause in the language of being a ‘witness’ against oneself had its origins in common and canon law.77 Over the years, it had been invoked to justify a number of competing rationales78 (two of these being the crime control and due process models), without a decisive verdict in any direction. As the Canadian Supreme Court summed it up, ‘The principle against self-incrimination may mean different things at different times and in different contexts.’79
The austere text of Article 20(3), therefore, tells us nothing about its underlying values. Nor do the Constituent Assembly Debates help to answer the question. The guarantee against self-incrimination found its way into the Constitution almost absent-mindedly, without any debate.80 That, however, does not exhaust the interpretive resources open to us. If the text of Article 20(3) does not speak, its placement in the Fundamental Rights chapter certainly does.
Alongside Articles 19, 21, 20(1), 20(2), and 22, Article 20(3) is located in the sub-section of Part III titled ‘Rights to Freedom’. Neither the placement of Article 20(3) nor the company it keeps are accidental, and nor were they inevitable.81 The first, a conscious move by the framers, tells us that liberty is, at least, one of the values that constitutes the Indian Constitution’s guarantee against self-incrimination. The case for due process being at the foundation of the self-incrimination guarantee grows stronger if we fan out still further. The provisions that accompany Article 20(3) under the heading ‘Rights to Freedom’ include all the basic civil rights, viz., speech, assembly, association, movement, life, and personal liberty, that constitute the basis of human dignity and autonomy. And indeed, it was by reading Article 20(3) alongside 21 that the Court in Selvi gave the guarantee against self-incrimination its foundations in mental privacy and informational self-determination. In this family of rights that must be read together and in harmony (as R.C. Cooper correctly taught us), the crime control model strikes a discordant note.82 Article 20(3), therefore, is illuminated by the constellation of freedoms that surround it, all of which point to due process being at the heart of the guarantee against self-incrimination.83
The difference is not merely formal, or linguistic. As Ian Dennis points out, it makes all the difference whether the guarantee against self-incrimination exists for utilitarian reasons (ensuring reliability), or whether it is treated as ‘an adjectival right in the criminal process’.84 The reason being:
Once the privilege is granted the status of a procedural right, one which is an essential component of a fair trial, it acquires much greater force. It is then capable of applying broadly in a range of different contexts, and it becomes much more difficult to negotiate, in the sense that acute problems arise over the conditions, if any, under which it can be overridden by legislation or judicial decision.85
It is this distinction that the theorists of colonial continuity failed to grasp. And it is in understanding it that we are also now in a position to understand the fundamental error in M.P. Sharma’s case. In M.P. Sharma, the Court surveyed the pre-Independence legal landscape, found a number of legal provisions that chipped away at the scope of the self-incrimination guarantee, and concluded from this that the constitutional right against self-incrimination had, therefore, to be reconciled with the law as it stood. This was a mistake, because it ignored the transformative character of the Constitution. It ignored, in particular, the fact that whatever minimal procedural guarantees that existed in the colonial period existed within the framework of law and order. In stark contrast, on the other hand, the self-incrimination clause had been consciously articulated in the language of freedom, and nested within a group of rights that unambiguously spoke the vocabulary of civil liberties.86
It is important to note that this argument does not seek to prove too much. It is not my case that due process and individual freedom are the only values underlying Article 20(3).87 The values of reliability and accuracy, under the crime control model, remain relevant and important. They are found, for example, in Section 27 of the Evidence Act, which—in specific cases of recovery of evidence pursuant to a statement—allows that statement to be admissible in evidence even if it is an otherwise barred confession to a police officer. The argument, however, is that in interpreting the scope of Article 20(3), and, in particular, the meanings of words such as ‘compelled’ and ‘to be a witness’, the individual freedom-oriented values of the due process model have to be given a central and framing role. Selvi was the first judgement under Article 20(3) that understood this, and, therefore, introduced the concept of mental privacy as a core organizing principle emerging out of a combined reading of Articles 20(3) and 21.88 For this reason, it is a transformative judgement.
V. Conclusion
The relationship between the individual and the State is at its most unequal when an accused person is brought face to face with the machinery of official power. Constitutional rights are actualized only when they are applied meaningfully to such relationships. This requires infusing constitutional norms into the legal framework that structures the relationship between the accused and the forces of law and order at the point of greatest vulnerability: the interrogation room.
This legal framework is constituted in the first instance by the Code of Criminal Procedure and the Evidence Act and, secondly, by multiple special laws that lay down the procedure for dealing with specific categories of offences. Selvi reversed a decades-old logic that had come to be a staple feature of the Supreme Court’s jurisprudence: that a system of law instituted by a colonial regime, with the expressed purpose of controlling a subject population, could be applied without alteration to the relationship between a constitutional democracy and its citizens. It did so by placing the individual, and individual rights, at the heart of its interpretation of the guarantee against self-incrimination, and making privacy the organizing principle to determine the interplay between the Constitution and the criminal process.
The actual impact of Selvi, however, depends on whether future judgements take its transformative logic forward. While the correctness of Selvi was affirmed by a nine-judge bench of the Supreme Court in Puttaswamy,89 some of its insights remain untapped. For example, it is an accepted part of Indian law, driven by imperatives of crime control, that illegally obtained evidence remains admissible in Court, as long as it is relevant. Selvi, however, drew a distinction between illegal and unconstitutional evidence, and held that exclusionary principle would apply to the latter.90 While this distinction was also made in previous cases such as Kathi Kalu, Selvi’s clear reiteration—drawing from its own application of the due process model—placed the issue back in the spotlight, this time within the due process framework. For example, Section 27 of the Evidence Act stipulates that police confessions, otherwise inadmissible under Section 25, become admissible to the extent that they lead to recovery of material evidence. But to the extent that a police confession is compelled, it cannot be admitted in evidence, even if it leads to recoveries. This can, potentially, check some of the more abusive police practices that are presently staple features of investigations, such as disclosure and pointing out memos, the contents of which are the result of compulsion, but where the compulsion is of no relevance to their admissibility.91
Furthermore, the actual holding in Selvi is of only limited reach. While the Court made due process the foundation of interpreting the phrases ‘compelled’ and to be a ‘witness against himself’, Article 20(3) has a third, crucial element: it applies only to an accused person. In M.P. Sharma, the Court resisted an attempt to confine Article 20(3) to the courtroom, and observed that the ambit of the Article covered any ‘person against whom a formal accusation relating to the commission of an offence has been levelled which in the normal course may result in prosecution’.92 While the Court also clarified that it wa
s not deciding whether or not the right would apply in other cases, subsequent judgements invoked ‘formal accusation’ as a necessary requirement to trigger Article 20(3).93 The criminal legal scholar, Abhinav Sekhri, points out that this limitation became significant with the passage of a number of laws, dealing with socio-economic offences, that were enacted or modified in the 1960s. Laws such as the Sea Customs Act and, subsequently, the Foreign Exchange Management Act and the Narcotics, Drugs and Psychotropic Substances Act, institute a detailed investigatory regime before the stage of a formal accusation. Although the investigating officers have the same powers as police officers, including search, seizure and arrest, issuing summons to individuals, who are bound to answer on pain of perjury, and imposing detention, they are deemed, by legal fiction, not to be police officers.94 However, the Supreme Court has held that because, under these investigatory regimes, there is no ‘formal accusation’, Article 20(3) does not apply.95
If the heart of the due process model is a concern for protecting the rights of individuals who are placed in a special position of vulnerability vis-à-vis the State, then the Court’s exclusion of entire legal regimes from the ambit of Article 20(3), purely on the basis of formalistic reasoning, is indefensible. After Selvi, therefore, we have an anomalous situation, where Article 20(3) appears to have been partially interpreted from the perspective of the due process model, and partially from a restrictive perspective that can only be traced back to the imperatives of crime control. As we have argued, if Selvi is correctly decided, then the Court’s judgements interpreting the word accused perhaps call for reconsideration. To extend the meaning of ‘accused’ to situations where an individual is subjected to ‘detention in custody by any authority during an investigation’,96 as Sekhri suggests, would be a true vindication of Selvi’s transformative potential.
The Transformative Constitution Page 37