Article 20(3) embodies the principle of protection against compulsion of self-incrimination which is one of the fundamental canons of the British system of criminal jurisprudence and which has been adopted by the American system and incorporated as an article of its Constitution. It has also, to a substantial extent, been recognized in the Anglo-Indian administration of criminal justice in this country by incorporation into various statutory provisions. In order, therefore, to arrive at a correct appraisal of the scope and content of the doctrine and to judge to what extent that was intended to be recognized by our Constitution-makers in article 20(3), it is necessary to have a cursory view of the origin and scope of this doctrine and the implications thereof as understood in English law and in American law and as recognized in the Indian law.30
Using this interpretive framework, the Court was able to hold that there was no warrant to read the self-incrimination clause broadly, because the principle had been repeatedly contested and criticized in English common law.31 Furthermore, the Court observed that in successive criminal procedure codes, enacted in India in 1861, 1872, and 1882, there was a clear distinction between a search and seizure, and compelled production of documents. It then noted that because the framers of the Constitution had consciously declined to insert into the Constitution a right akin to the US Fourth Amendment (that prohibited unreasonable searches and seizures), they implicitly intended this position to remain unchanged.32
M.P. Sharma did two things. It accorded a broad reading to the phrase ‘to be a witness’ in Article 20(3), including both oral testimony and documentary evidence within its scope (and also expanding its scope beyond the courtroom, to include police interrogations). However, it then employed the crime control model and colonial law to accord a narrow reading to the word ‘compelled’, holding that it only applied to forced production of documents, and not to a search and seizure under the Criminal Procedure Code (even though, in most cases, the result would be the same).33
The breadth of the first part of the Court’s holding—a radical departure from colonial practice in placing constraints upon police interrogation—caused significant churning in criminal legal proceedings across the country.34 Unsurprisingly, therefore, the broad reading of witness came under serious challenge seven years later, in State of Bombay v. Kathi Kalu Oghad, when the Court set up an eleven-judge bench to reconsider the verdict of eight judges in M.P. Sharma.35 If to be a witness meant to furnish evidence, it was argued, then compelled fingerprinting, or the taking of bodily measurements (as part of criminal investigations, under the Identification of Prisoners Act of 1920), was clearly unconstitutional. So were the compelled taking of handwriting samples for comparison36 and the admissibility of (otherwise inadmissible) confessions to the police that led to the recovery of evidence.37 High courts across the country returned conflicting findings on the issue, and the case came up to the Supreme Court for resolution.
In Kathi Kalu, the Court evidently felt that fidelity to M.P. Sharma’s definition of witness would outlaw some important criminal investigation procedures, and invalidate long-standing provisions of the Evidence Act, as well as other colonial laws such as the Identification of Prisoners Act. Its response was to hastily backtrack. It did so, as we shall see, by extending to the rest of Article 20(3) the logic that M.P. Sharma had used to narrow the meaning of compelled.
The majority began by holding that M.P. Sharma had been incorrect to equate to be a witness with to furnish evidence. To be a witness meant something narrower. It meant to communicate knowledge of a relevant fact. Providing a fingerprint, or a handwriting sample, did not involve communicating knowledge, and therefore did not amount to being a witness.38 Notice, however, that in comparison with how we use the word ‘witness’ in everyday language, this seems a strained and artificial interpretation. There is no invariable association of witness with conscious, testimonial act: the very existence of the colloquialism dumb witness (as noted by the Court in M.P. Sharma) is evidence of this. So, also, is our reflexive attribution of the role of witness to decidedly unconscious entities such as history, as well as the fact that the sentence ‘your fingerprints at the scene of the crime speak of your guilt’ sounds entirely natural.
The majority, therefore, had to dig deeper. It did so by taking from M.P. Sharma the phrase ‘a positive, volitional act’ (of furnishing evidence), and equating ‘volition’ with the ability to alter a state of affairs. In other words, the Court held that (unlike oral or written testimony), a person had no ability to change their fingerprints, or their bodily measurements, or (subject to certain exceptions) the character of their handwriting. The giving of fingerprints or a blood sample, therefore, did amount to furnishing evidence, but not to ‘being a witness’, because the character of the evidence was not within the ‘volition’ or control of the person.39
For this reason, the Court noted, fingerprints or handwriting samples did not themselves incriminate an accused. It was only when they were matched against, say, a fingerprint found at the scene of crime or a forged document with the accused person’s signature that they became salient as evidence. But taken by themselves, they remained ‘innocuous’40 (once again, the Court equated innocuous with unchangeable). The concurring and partially dissenting opinion in Kathi Kalu deepened this argument by noting that even if it could be said that fingerprints or handwriting samples fell within the definition of being a witness, the individual was not being compelled to become a witness against himself. This was because, as reasoned above, it was not the sample that incriminated him, but the fact that the sample was successfully matched with another piece of evidence.41
At the heart of the judgement in Kathi Kalu was the understanding that one is a witness only about a fact, or a state of affairs, which one has the power to change or alter (volition). A moment’s reflection should make clear that the foundation of this analysis is the crime control model. That which the accused cannot change has an independent character, a character that does not depend upon the will of the accused (which, presumably, can be undermined or overcome in interrogation). To put it simply, the evidentiary character of fingerprints, or bodily measurements, or handwriting samples (or even blood samples), is not affected by the nature of the interrogation that the accused was subjected to. Therefore, interpreting Article 20(3) as drawing a line between evidence that the accused can change and evidence that she can’t demonstrates that the overriding concern is with the character of the evidence, its accuracy, and its reliability.42
Indeed, this concern was evident in both the majority and the concurrence. The majority noted that in protecting the right of the accused against self-incrimination, ‘lawmakers could not have intended to put obstacles in the way of efficient investigation’.43 The concurrence counselled that because they had no evidentiary value in themselves, allowing the taking of fingerprints or handwriting samples would not tempt law-enforcement officials to become lazy or to take short-cuts44 (for example, like allowing police confessions would). To buttress this interpretation, the majority, in particular, relied extensively upon the colonial landscape, citing both ‘English law on the subject’45 and colonial statutory provisions. Taking note of the provisions of the Evidence Act (1872), the Code of Criminal Procedure (1898), and the Identification of Prisoners Act (1920), the majority observed that the ‘framers were aware’ of this legal landscape, and consequently the constitutional right under Article 20(3) was to be interpreted within this framework.46
At first glance, this seems decidedly odd. The natural method would be to subject existing laws to constitutional standards, instead of deriving constitutional standards from existing laws that preceded the Constitution itself. It is not so odd, however, if we understand that, in the view of the Court, the function of the Constitution was not so much as to transform the relationship between the State and the individual, but to preserve it. This is visible most starkly in a judgement (Shyamlal Mohanlal Choksi) delivered soon after, which also considered the issue of compelled
production of documents under the Code of Criminal Procedure. The majority held that, in view of Article 20(3), the provisions requiring compelled production could not apply to a person accused of an offence. But in his dissenting opinion, Justice Shah observed that the Court’s interpretation went against the previous ‘eighty years’ of jurisprudence.47 The fact that sixty-three out of these eighty years were spent under a colonial regime was considered to be of no consequence. This interpretation could make sense only if the framing of the Constitution was seen as a moment of conservation and continuity, rather than a moment of transformation and change.48
Let us sum up. Article 20(3) guarantees the right against self-incrimination, stipulating that ‘no person shall be compelled to be a witness against himself’. In M.P. Sharma v. Satish Chandra, the Supreme Court held that a search and seizure by the police did not amount to compulsion under Article 20(3), because the accused herself was not being forced to do anything. In State of Bombay v. Kathi Kalu Oghad, the Court held that to be a witness only included a conscious testimonial or communicative act (knowledge of a relevant fact), and was limited to furnishing only that type of evidence that a person had control over, or ability, to alter. The Court did this by taking the word ‘volitional’ from M.P. Sharma, and equating it with ‘changeable’. In M.P. Sharma, the clauses on search warrants were upheld. In Kathi Kalu, the provisions of the Evidence Act and the Identification of Prisoners Act were upheld. Both judgements rested upon two conceptual premises: first, that Article 20(3) embodied the crime control model, and the content of the phrase ‘no person shall be compelled to be a witness against himself’ was to be determined by considering the impact upon efficient, accurate, and reliable criminal investigation; and second, that judicial interpretation of Article 20(3) must be such that the constitutional provision was brought in line with the existing legal framework of criminal investigation and procedure, a substantial part of which was of colonial origin.49
III. Selvi and the Due Process Model
The judgements in M.P. Sharma and Kathi Kalu—delivered by benches of eight and eleven judges, respectively—settled the issue for a few decades. In the meantime, however, with advances in technology and in interrogation techniques, the bright lines drawn in those cases began to blur, not just in India, but all over the world. In Schmerber v. California, the US Supreme Court maintained the difference between ‘physical’ and ‘testimonial’ evidence (on the lines of Kathi Kalu), but acknowledged that there existed a wide range of situations in which the distinction would break down.50 And ten years later, in Horvath v. The Queen, the Canadian Supreme Court was compelled to reconsider the entire common law jurisprudence on the meaning of ‘voluntariness’. Horvath involved a murder case in which the accused was subjected to a lengthy period of interrogation, during which he confessed to the murder, and provided the location of the murder weapon. It was later found that, as a result of the intensity of the interrogation, he had been in a ‘light hypnotic state’51 during this time. It was argued that this would have no bearing on the issue of compelled self-incrimination, because, as per the classic common law definition of voluntariness, laid down in Halsbury’s Laws of England and by the 1914 case of Ibrahim v. The Queen, a statement was voluntary if ‘it has not been obtained from [a person] by fear of prejudice or hope of advantage, exercised or held out by a person in authority, or by oppression’.52 However, a closely divided Supreme Court discarded this definition, and held that what mattered was whether the individual was ‘in full and voluntary control and possession of his faculties’53 when making the statement. The Court grounded this definition within the right to privacy, noting that the individual’s will was overborne,54 and in the process ‘a police officer has gained unconsented access to what in a human being is of the utmost privacy, the privacy of his own mind’.55 Consequently, the statements were held to be inadmissible, notwithstanding a psychiatrist’s testimony that the accused had been probably telling the truth ‘as he saw it’.
It was in this transformed context—both technological and constitutional—that, almost fifty years after Kathi Kalu, the Supreme Court was called upon to re-examine the meaning of Article 20(3). Selvi v. State of Karnataka involved a constitutional challenge to the forcible use of three popular interrogation techniques. In narco-analysis, an individual was injected with sodium pentothal56, ‘a drug which lowers inhibitions … and induces the person to talk freely’.57 She was then questioned while under the effect of the drug. Under the polygraph, or lie-detector, test, a person’s physiological responses (such as ‘respiration, blood pressure, blood flow, pulse and galvanic skin resistance’58) were measured to determine whether she was likely to be lying or telling the truth. The BEAP (brain electrical activation profile or brain-mapping) test was designed to detect ‘whether an individual is familiar with certain information by way of measuring activity in the brain that is triggered by exposure to selected stimuli’.59
Selvi was a difficult case. While the judgements in M.P. Sharma and Kathi Kalu had dealt with very different issues, the analytical framework that they had employed remained controlling. In narco-analysis, the administration of sodium pentothal could be compared to the seizure of documents in M.P. Sharma: at the point of compulsion, there was no testimonial act, and, therefore, no violation of Article 20(3). Under Ibrahim’s definition of voluntary, the individual was not compelled to answer questions simply because she was under the influence of the truth serum. Lie-detector tests and brain mapping bore a family resemblance to Kathi Kalu. Like fingerprints, they involved physiological responses that were not in themselves incriminatory, but only of value when put alongside other pieces of substantive evidence.60
And yet, the analogies were not quite so clear-cut. Was it possible to draw a bright-line distinction between the administration of sodium pentothal (compelled) and the subsequent interrogation (voluntary)? Didn’t the results of a lie-detector or a BEAP test determine what was in the individual’s mind, and were, therefore, of a nature different from fingerprints or blood samples, or even DNA? Selvi, thus, gave the Court a chance to reconsider not just the meaning of Article 20(3), but also its underlying values.
However, the State argued that the case presented no trouble at all. Section 53 of the Code of Criminal Procedure already allowed for the examination of ‘blood, bloodstains, semen, sputum, sweat’ through the use of ‘modern and scientific techniques including DNA profiling and other such tests’.61 Consequently, it was contended that the three interrogation techniques at issue ought to simply be read into Section 53.
The Court rejected this argument, noting that Section 53 referred only to ‘bodily substances’ and not to ‘testimonial responses’.62 However, the facts in Selvi were precisely those in which the seemingly straightforward distinction between testimonial/communicative and physical/real evidence—as pointed out in Schmerber v. California—came undone. While an interview under the effects of sodium pentothal evidently involved a ‘testimonial response’, what was it about physiological reactions (in lie-detector tests and brain-mapping) that brought them within the category of testimony and not physical evidence?
The Court answered this question by going back to Kathi Kalu, and holding that a testimonial act involved ‘the imparting of knowledge by a person who has personal knowledge of the facts that are in issue’.63 It then applied this definition to the facts:
Even though the actual process of undergoing a polygraph examination or a BEAP test is not the same as that of making an oral or written statement, the consequences are similar. By making inferences from the results of these tests, the examiner is able to derive knowledge from the subject’s mind which otherwise would not have become available to the investigators. These two tests are different from medical examination and the analysis of bodily substances such as blood, semen, and hair samples, since the test subject’s physiological responses are directly correlated to mental faculties. Through lie-detection or gauging a subject’s familiarity with the stimuli,
personal knowledge is conveyed in respect of a relevant fact.64
Therefore, the key distinction between the BEAP and lie-detector tests and the facts of Kathi Kalu and, by extension, the ingredients of Section 53 of the CrPC was the involvement of the mind: while both involved physiological responses, it was only in the former case that these responses were ‘directly correlated to … mental faculties’.
But that then raised the further question: what was it about the involvement of the mind (a physical part of the body, after all)65 that made the distinction salient? The Court answered this by noting that:
The compulsory administration of the impugned tests impedes the subject’s right to choose between remaining silent and offering substantive information. The requirement of a ‘positive volitional act’ becomes irrelevant since the subject is compelled to convey personal knowledge irrespective of his/her own volition.66
In this paragraph, the Court made a fundamental interpretive shift from the foundations of Kathi Kalu. In Kathi Kalu, M.P. Sharma’s articulation of the ‘positive volitional act’ had been understood in terms of reliability of the evidence. Volition had been interpreted to mean having the power to alter a piece of evidence (such as oral testimony), a definition that flowed from the imperatives of the crime control model. In Selvi, however, the Court understood volition (or voluntariness) not in the language of the changeable and the unchangeable, but in the language of choice, a choice whether or not to convey personal information.
Of course, in a trivial way, being forced to provide a blood sample or a fingerprint also involves conveying personal knowledge (of your blood group or your fingerprint pattern) irrespective of your volition.67 In earlier cases, the Court had gotten around that by arguing that a blood sample or a fingerprint was innocuous in itself. Only when subsequently corroborated with another piece of evidence (fingerprints at the scene of the crime) did it become incriminatory. However, that reasoning was not open to the Court in Selvi, because physiological responses to stimuli are also innocuous in themselves. The responses to stimuli in a lie-detector test or a BEAP do not possess incriminatory content, like a confession or a signed document does. So, even though mental processes were involved, these mental processes lacked the essential attribute that set apart oral or documentary testimony, and indeed, appeared closer to physical evidence taken in the form of blood or DNA.
The Transformative Constitution Page 36