Furthermore, our legal system is riddled with legislation that has nothing to do with public or national security, but nonetheless incorporates elements of the state of exception. At the forefront stands the Bombay Prevention of Beggary Act, another brutal colonial holdover now existing in twenty Indian states. This law, as its name suggests, effectively criminalizes poverty and creates ‘status offenders’: that is, it punishes people not for what they have done, but for who they are.170
Quite apart from the constitutional problems with criminalizing the act of begging in the first place, the Act goes much further: it constructs a jurisdiction of suspicion by criminalizing individuals for ‘having no visible means of subsistence and wandering about or remaining in any public place in such condition or manner as makes it likely that the person doing so exists [by] soliciting or receiving alms’;171 it authorizes detention for up to one year for ‘crimes’ of this kind;172 and it endorses guilt by association by allowing courts to detain dependants of ‘beggars’.173 It is one thing to normalize the state of exception in the context of anti-terrorism. It is quite another to enshrine it as just another method of ‘social control’.174
Justice Thipsay’s judgement in Jyoti Chorge, consequently, is not simply about bail standards under the UAPA. Its transformative reach extends much further and much deeper. It has the potential to dismantle some of the most suffocating features of India’s security-State legal apparatus, as well as its social control legal apparatus, and erect in its place a judicial philosophy that places personal liberty and the freedom of speech and association at the heart of the constitutional order. Its rejection of the permanent state of exception, and its unravelling of the complex web that has bound together normalcy and exception into a repressive mesh, is true to the Constitution’s text, structure, and philosophy, and true to the civil liberties tradition that animated the framing of the Fundamental Rights chapter. But it is a lonely truth, not yet accepted or validated by the accumulated force of judicial opinions.
Whether the Court will choose to adopt the transformative constitutional vision of Jyoti Chorge in any significant manner, so that the world’s largest democracy sees no more Chandrashekhar Azads condemned to jail without trial and without bail, is a question that only the future can answer.
9
Privacy and the Criminal Process: Selvi v. State of Karnataka
What happens after a person is arrested or detained? His troubles begin then. When he is detained or arrested and he is in the clutches of the police, he is alone in the world, and the forces of the Police, the forces of the Crown and all other forces combine against him and he is helpless.
Thakur Das Bhargava, Constituent Assembly Debates (September 1949)
IN THE EARLY 2000s, the Karnataka government’s Forensic Science Laboratory (FSL) at Bengaluru unexpectedly shot to fame. Notorious counterfeiter Abdul Karim Telgi, accused terrorists, alleged Naxalites, and run-of-the-mill criminals were all subjected to a ‘truth serum’, under the influence of which they apparently ‘sang like canaries’,1 condemning both themselves and their accomplices. Important leads were supposedly revealed and knotty crimes were cracked. The assistant director of the FSL even earned the sobriquet of Doctor Narco or Narco Queen (after ‘narco-analysis’, the technical term for the process of using this ‘truth serum’).2
By the end of the decade, fame had turned to infamy. In a letter written from Nagpur Central Jail, in December 2008, Arun Ferreira, an accused Naxalite, publicly accused Doctor Narco of fabricating his test results to achieve the outcome that the government wanted.3 Dismissing his objections, an unnamed expert at the FSL confidently proclaimed, ‘During narco-analysis the subject goes through phases of trance … [when] he speaks the truth’.4 An unnamed police officer proclaimed, with equal confidence, that Ferreira had confessed to his crimes, and provided the investigation with vital evidence of further terrorist activities.5
After five years in jail, Arun Ferreira was acquitted of all charges.6
In the meantime, doubts and concerns grew over narco-analysis, both for its unreliability, and for the near-third-degree manner in which it was administered. Constitutional challenges wound their way through multiple high courts, until finally Selvi v. State of Karnataka ended up before the Supreme Court. It was contended that three of the most prominent interrogation techniques, viz., narco-analysis, the lie-detector test, and brain-mapping, violated an accused person’s right against self-incrimination under Article 20(3), and her right to life and personal liberty under Article 21 of the Constitution.
The case seemed to have scant chance of succeeding. The colonial logic that had privileged the maintenance of law and order over the rights of the accused, and that had seen no problem with conscripting the individual’s body (and her private spaces) for the purposes of law enforcement and crime detection, had been largely endorsed by the post-Independence Supreme Court.7 The Code of Criminal Procedure allowed general warrants and searches,8 and recent amendments had expanded the extent to which the body could be requisitioned in a criminal investigation, bringing in DNA profiling, among other things.9 In this context, the mere administration (even if forced) of a drug that sent the subject into a hypnotic trance (narco-analysis), or the measurement of physiological (lie-detector) or cognitive (brain-mapping) responses to questions, seemed relatively minor infractions, if at all.
In Selvi, the Supreme Court disagreed. It declared the forced application of all three interrogation techniques to be unconstitutional. In this chapter, I argue that Selvi was a transformative judgement because of the way in which it understood and articulated the relationship between the individual and the State in the context in which the imbalance of power between the two is at its highest: when the individual is accused and the State is the interrogator.
I begin by describing two possible competing philosophies that underlie criminal legal regimes: the crime control model, which views the accurate solving of crime to be the highest goal of the law, and the due process model, which holds that even for the detection of crime, there are lines that the State cannot cross (I); I then demonstrate how the Supreme Court’s early judgements on Article 20(3) were delivered within the framework of the crime control model (II), and how Selvi departed from this understanding in its interpretation of the words ‘voluntary’ and ‘compelled’ (III). In conclusion I contend that Selvi’s embrace of the due process model is correct in light of our pre-constitutional experience of an authoritarian State and the decision of the framers to elevate the guarantee of self-incrimination from a procedural safeguard to a constitutional right (IV). Furthermore, it has important implications for our understanding of the limits that the Constitution and the Code of Criminal Procedure impose upon the police and upon the criminal investigation process (V).
I. Two Models of the Criminal Process
According to the dominant understanding, the practices and procedures of criminal investigation, evidence, and trial are structured around determining truth. A crime scene is scoured for fingerprints. A suspect’s call data records are accessed to know whom she might have conspired with. Witnesses are cross-examined in court to test the credibility of their narrative. It is, of course, understood that in a fallible world, even the best procedures will yield, at the highest, an approximation of the truth. For this reason, the standard for conviction in criminal cases is ‘beyond all reasonable doubt’, instead of ‘beyond all doubt’. And whether in science fiction novels10 or in the euphemistic dubbing of Doctor Narco’s procedure as a ‘truth serum’, the effort has always been to reduce to a vanishing point the gap between the truth and the determination of truth by the criminal legal procedure.
But that is not all. We recognize, as well, that there are limits to what the State can do to individuals in the quest for truth. In the words of the US Supreme Court, ‘too high a price may be paid even for the unhampered enforcement of the criminal law, and that, in its attainment, other social objects of a free society should not be sacrificed
’.11 This is reflected most starkly in the debates around torture. Some argue that torture should be outlawed because it doesn’t work. It is unreliable, encourages shoddy investigations, and all other things being equal, a system with institutionalized torture will lead to worse outcomes in terms of controlling crime. Others argue that this framing is a mistake. Torture should be outlawed because it violates human dignity. It should be outlawed even if it could be demonstrated that it worked perfectly.12 To test both positions, there exist a number of well-worn hypotheticals. Is torture permissible in a ‘ticking time-bomb’ scenario? Can you retrieve evidence from a person’s body by subjecting him to stomach-pumping?13 What about the forceful administration of a truth serum that actually worked perfectly, if such a thing was invented one day?
In 1964, in a famous but controversial article titled ‘Two Models of the Criminal Process’, the US jurist Herbert Packer argued that the answers to such questions would depend upon the values that underlay the criminal legal system. Packer proposed two possible models: the ‘crime control model’ and the ‘due process model’. Under the crime control model, the ultimate goal of the criminal process is the ‘repression of criminal conduct’.14 He wrote, ‘In order to achieve this high purpose the Crime Control Model requires that primary attention be paid to the efficiency with which the criminal process operates to screen suspects, determine guilt, and secure appropriate dispositions of persons convicted of crime.’15 Simplifying greatly, therefore, ‘the criminal process … must put special weight on the quality of administrative fact-finding. It becomes important, then, to place as few restrictions as possible on the character of the administrative fact-finding processes and to limit restrictions to such as enhance reliability, excluding those designed for other purposes.’16 Under the crime control model, the question of whether torture was to be allowed or not would have to be answered after conducting empirical studies to determine whether, on the whole, evidence obtained under torture is more or less reliable than evidence obtained without. In simpler language, whether torture works.
In contrast with the crime control model, the due process model holds that ‘the combination of stigma and loss of liberty that is embodied in the end result of the criminal process is the heaviest deprivation that government can inflict on the individual’.17 Therefore, ‘because of its potency in subjecting the individual to the coercive power of the state, the criminal process must … be subjected to controls that prevent it from operating with maximal efficiency’.18 In other words, under the due process model, the coercive edge of the criminal process must be tempered in a manner that, on some occasions, will have the inevitable effect of retarding efficiency, in service of other values such as human dignity and individual freedom.19
Consider, for example, the notorious ‘exclusionary rule’ in evidence law. The exclusionary rule stipulates that evidence obtained through unconstitutional means (e.g., a warrantless search, or through torture) is inadmissible in court. Naturally, this includes situations in which the quality and relevance of evidence is entirely unrelated to the manner in which it was obtained. Consequently, the crime control model, which is concerned with accurate determination of guilt, will only judge evidence with regard to its relevance, and will see little reason to deny admitting illegally obtained evidence. On the other hand, the due process model, which is concerned with protecting citizens’ rights (including against warrantless searches and torture) might well determine that an effective way of protecting these rights is to ensure that evidence obtained in their violation, cannot be used against the accused. Upon the spectrum, for example, the United Kingdom and India veer more towards considering evidence based primarily on its reliability, while the United States remains more sensitive to the manner in which it is collected.20
Criminal legal systems attempt to strike a balance between the two models, i.e., to allow the State enough power to detect and punish crime, but not so much power that the individual is entirely effaced before the State.21 As is evident with the fate of the exclusionary rule, different countries strike the balance differently. Furthermore, the balance is a contested one, subject to constant battle between the proponents of crime control and the supporters of due process. This battle assumes particular salience in circumstances where an aspect of criminal procedure can be justified under both models, but the degree of protection it affords to individuals will change depending upon which model is endorsed.
For example, Article 20(3) of the Constitution stipulates that ‘no person accused of an offence shall be compelled to be a witness against himself’. It codifies the long-standing common-law ‘privilege against self-incrimination’, which, on some reflection, can fit within both the crime control and the due process models. Compelled testimony—simply by virtue of being compelled—will often be flawed. Reasonable criminals are unlikely to willingly give evidence incriminating themselves, and history bears witness to the fact that coerced, false confessions have been the norm rather than the exception. Consequently, the crime control model, with its focus on accurate determination of guilt and innocence, has good reason to guarantee a right against self-incrimination.22 The due process model, of course, is concerned with the rights violations that invariably accompany compelled, self-incriminatory testimony, and has equal, if different, reason to guarantee that right as well.23 It is quite possible, therefore, that as the Canadian Supreme Court observed, a self-incrimination clause may be driven by both elements: ‘to protect against unreliable confessions … and … to protect against the abuse of power by the state’.24
However, these two rationales are not always complementary. It is when they clash that the difference becomes particularly significant. To go back to our earlier example, if an infallible truth serum is invented, then its use under the crime control model would no longer be problematic, since the central objection to compelled testimony—the likelihood of inaccuracy—has been removed. However, a very different set of concerns would be articulated by the proponents of due process: in particular, concerns about the right to privacy.
Article 20(3), therefore, is not self-interpreting. Its protective scope, and the extent to which it can check the State’s power as an interrogator, depends upon which philosophy it is underpinned by. The sparse wording of the text offers us no assistance to decide that question. We must therefore turn to how, over the years, the Indian Supreme Court has understood this right against self-incrimination.
II. Let the Body Speak: Self-Incrimination and the Early Supreme Court
Soon after Independence, the Supreme Court was called upon to decide its first case dealing with the scope of the right against self-incrimination. In M.P. Sharma v. Satish Chandra, certain persons were suspected of committing large-scale corporate fraud, through a network of companies. Under the Code of Criminal Procedure (as it then stood), the police applied to the district magistrate for a warrant to search the offices of the various companies. The magistrate issued the warrant. Searches were carried out at thirty-four places, and documents seized. The petitioners (who held various positions in the companies) came to Court and asked that the warrants be quashed. They argued that the search and seizure of their private documents violated their rights under Article 20(3) of the Constitution. In case the documents contained something incriminating, their forced seizure would amount to compelling the petitioners to be ‘witnesses against themselves’.
The first question before the Court, therefore, was whether the words ‘to be a witness’ (as part of the phrase ‘no person shall be compelled to be a witness against himself’) included self-incrimination through documentary evidence. The Court held that it did. It defined a ‘witness’ as any person who furnished (or was required to furnish) evidence. ‘To be a witness’, therefore, meant ‘any positive volitional act that furnishes evidence’.25 There was no reason this could not include documents.
The next question, then, was whether a search and seizure amounted to compelling a person to be a witness against herself. Here,
the Court held that it did not:
… there is no basis … for the assumption that a search or seizure of a thing or document is in itself to be treated as compelled production of the same … A notice to produce is addressed to the party, concerned and his production in compliance therewith constitutes a testimonial act by him within the meaning of Article 20(3) as above explained. But search warrant is addressed to an officer of the Government, generally a police officer. Neither the search nor the seizure are acts [sic] of the occupier of the searched premises. They are acts of another to which he is obliged to submit and are, therefore, not his testimonial acts in any sense.26
In other words, to be a witness indicated that the person in question was performing a testimonial act of some kind (being compelled to speak, or being compelled to produce documents). However, when the forces of law and order searched premises and seized documents, there need not be any involvement of the person herself. Article 20(3), therefore, was violated if a person was forced to produce incriminating documents (by the police, or by a court order), but not if those documents were recovered in a police raid.27
This distinction seems somewhat thin and formalistic.28 How did the Court justify it? It employed two arguments. First, it made a direct reference to the crime control model, noting, ‘[The] power of search and seizure is in any system of jurisprudence an overriding power of the State for the protection of social security and that power is necessarily regulated by law’.29 The scope of compulsion under the self-incrimination clause, therefore, was to be defined and limited within the framework of the crime control model. And second, the Court made extensive reference to the colonial legal landscape. First, it observed:
The Transformative Constitution Page 35