Fruits Of The Poisonous Tree - How Indian Evidence Law Makes Custodial Violence Rational
Author: LawSphereX
Category: General Law
Published:
Updated:
FRUITS OF THE POISONOUS TREE - HOW INDIAN
EVIDENCE LAW MAKES CUSTODIAL VIOLENCE RATIONAL Aarefa Bootwala, BBA LLB, BITS Law School I.
INTRODUCTION
As stated in the Status of Policing in India Report 2025, the National Crime Report Bureau reported that between 2011 and 2022, there were 1,107 deaths in police custody and not a single office was convicted. The same report further notes that out of 394 custodial deaths between
2018 and 2022, only 41 cases were registered against the police, chargesheets were filed for
only 5 of those and there were zero convictions. 1 As per the SPIR 2025 survey, more than 70% of the police personnel believed that the police should be able to use force without any punishment2 and 63% in Gujarat agreed torture is necessary to gain information. 3 These are not statistics of a system that has failed to deter custodial violence but of a system that has never tried. This article asserts that the most neglected and significant structural reason for this failure lies in the law of evidence. When a legal system excludes evidence derived through torture or unlawful means, it eliminates the investigative utility of such methods as the police cannot rely on the resulting evidence in court.
However, the opposite is achieved when a legal system admits such evidence irrespective of the manner in which it was obtained, it effectively rewards constitutional violation with convictions. India, unfortunately, in practice consistently follows the latter. Drawing on Bentham’s deterrence calculus and Bruinsma’s classical theory, it contends that only an exclusionary theory can remove it.
1Common Cause & Lokniti – Centre for the Study Developing Societies (CSDS) and others, Status of Policing in India Report 2025 (Vipul Mudgal and others, Common Cause & Lokniti – Centre for the Study Developing Societies (CSDS) 2025) page 19 2 Common Cause & Lokniti – Centre for the Study Developing Societies (CSDS) and others, Status of Policing in India Report 2025 (Vipul Mudgal and others, Common Cause & Lokniti – Centre for the Study Developing Societies (CSDS) 2025) page 72 3Common Cause & Lokniti – Centre for the Study Developing Societies (CSDS) and others, Status of Policing in India Report 2025 (Vipul Mudgal and others, Common Cause & Lokniti – Centre for the Study Developing Societies (CSDS) 2025) page 112 The Indian Evidence Act, 1872 (“IEA”) which is now replaced by the Bharatiya Sakshya Adhiniyam, 2023 (“BSA”) recognizes relevancy as the only criteria for admissibility.
The source from which the evidence emerges is legally inconsequential however violative it is to the constitutional rights of the suspect. A confession given under torture is inadmissible but the weapon which a suspect may reveal under torture is admissible. 4 This distinction incorporated in Section 27 of the Indian Evidence Act 1872 and now reproduced without any amendment as Section 23 of the BSA is the explicit apparatus through which law in India converts custodial violence into admissible evidence. II.
SECTION 23 AND THE INVERSION OF DETERRENCE
The continued admissibility of evidence illegally obtained in Indian law cannot be simply categorized to be a lingering influence of colonialism. Instead, it reflects a contradiction consciously maintained by the legislature, which has repeatedly chosen not to address it and most recently reaffirmed it in the BSA 2023. To understand the sustenance of this contradiction and its consequences, it is important to examine not only what the law allows but also the incentives it creates. When examined through the lens of the classical deterrence theory, the picture is deeply concerning. In the late eighteenth century, Jeremy Bentham argued that human behavior is governed by a rational calculation of pleasure and pain. As per him, the role of a legislator is to engineer conditions where the expected punishment for a wrongdoing is greater than the benefit a person may gain for committing it.
Building on this idea, criminologist Garben Bruinsma explains that deterrence depends on three factors which are certainty, severity and the speed of punishment, and for deterrence to work all three factors must work together. If a punishment is severe but is unlikely to occur, it won’t have an effect and a mild punishment which is certain will also be insignificant. 5 When this theory is applied to the issue of custodial violence in India, it reveals not only a failure of deterrence but an inversion of it. The proviso of Section 23 of the Bharatiya Sakshya Adhiniyam does not simply fail to discourage police misconduct, but it actually creates an incentive for it. Section 23(1) and 23(2) make confessions to police officers or confessions 4 Hirday Virdi & Ishani Mukherjee, 'Rethinking Doctrine of 'Fruits of Poisonous Tree' into Indian Jurisprudence' (2021) 4 Int'l JL Mgmt & Human 220
5 Daniel S Nagin, Francis T Cullen and Cheryl Lero Jonson (eds), Deterrence, Choice, and Crime:
Contemporary Perspectives, 23 (Francis T. Cullen and others, Routledge 2018) made while in police custody in absence of a Magistrate inadmissible in court. The legislature recognized as early as in 1855 that if confessions to police officers was made admissible, they might obtain it through unlawful means and thus, imposed a strict bar. The policy rationale behind Section 25 of the Indian Evidence Act, which is now Section 23(1) of BSA, is the enactment of a disincentive to obtain confessions using torture, coercion, or other unlawful and involuntary means, this has been stated by James Stephen himself. 6 However, the proviso introduces an exception.
If a statement made in custody leads to a discovery of a material fact or object, the part of the statement relating to that discovery and the discovered object becomes admissible in court irrespective of the manner in which that statement was obtained. In practice, this means that police may use the confession not as evidence in court but as a guide to the physical evidence that can be produced in trial. This proviso effectively provides the police with a loophole, a large enough one to undermine the exclusion established in the sub sections. 7 From a Benthamite perspective, the issue becomes clear. If the purpose of excluding custodial confessions was to remove the incentive for the police to obtain them through coercion, Section
23 undermines that purpose by indirectly restoring the same incentive. A police officer who
uses unlawful means such as torture to extract a confession and later “discovers” a physical object suffers no evidentiary consequence for the illegal conduct. As a result, the certainty of punishment which is one of the main elements of deterrence as identified by Bruinsma is practically nonexistent. The officer who resorts to torture and recovers evidence is rewarded with admissible proof, while the officer that follows correct procedures but finds nothing gains no comparable advantage. As per the theory proposed by Bentham, a rational actor does not need cruel intent to resort to custodial violence, the incentive structure itself makes it a logical choice.8 III.
THE CALCULUS IN PRACTICE
The real-world consequence of this legal system is not merely theoretical. A 2009 investigation by Human Rights Watch, which was based on interviews of more than eighty police officers and sixty victims of abuse across several states in India, revealed that the use of physical force and torture to obtain confessions was not an unusual practice with the police force but a 6 James Fitzjames Stephen, “A History of the Criminal Law of England” 1 (Cambridge University Press) 442. 7 Vijay Pal Singh, “Poison Tree Principle: It’s Applicability in India,” 3 (International Journal of Advanced Research and Development, 2018)
8 Ibid 5
common investigative process. 9 During the study, a sub-inspector posted at Varanasi explained that force was often used in high-profile cases as officers lacked access to scientific investigative tools. He described that police officers face immense pressure to solve cases swiftly and fear punishment or suspension for failure to produce results. As a result, they feel forced to resolve cases “by any means necessary”. 10 This statement does not depict a police officer acting irrationally or immorally, rather it shows how officers react to the incentive structure created by the legal system. The Law Commission of India acknowledged this issue in its 152nd Report on custodial crimes, observing that pressure for rapid results and heavy workloads often lead to officers abandoning scientific methods of investigation to resort to physical force to extract information.
11 The officer under this pressure is given no reason to stop using unlawful means by the legal system. No officer interviewed believed in seeking a voluntary confession, they had the assumption that suspects would not confess unless beaten. An inspector posted in Bangalore stated, “There is no possibility of solving a crime without using some tough measures. If the police were to follow the rules, we would bring a suspect in, ask him politely if he has committed the crime. If he denies it, we will let him go so he can commit more crimes. "12 What the courts and the Law Commission have not fully acknowledged is that this pattern of behavior cannot be accredited only to the flaws in police culture but also the poorly designed institutions. The likelihood of punishment for the illegal misconduct by the police is very low.
Even when sanctions exist, such as departmental inquiries, in practice they rarely result in consequences like dismissal. Moreover, the process of imposing such penalties is also extremely slow. 13Thus, across all three factors identified by Bruinsma which are certainty, severity and celerity, the deterrent effect of Indian law on custodial violence is negligible. IV.
THE COMPARITIVE DIMENSION
In the United States and United Kingdom, individual rights are given more importance. Even though there is no absolute prohibition on admitting confessions to police officers, these jurisdictions have developed procedural safeguards such as recording requirements, discretion
9 Naureen Shah and Meenakshi Ganguly, Broken System
Dysfunction, Abuse, and Impunity in the Indian Police (New York: Human Rights Watch, August 2009)
10 Ibid 9
11 Law Commission of India, “152ND Report on Custodial Crimes” 1994,
https://delhihighcourt.nic.in/files/reports/152.pdf (accessed on 10 March 2026), para. 13.3
12 Ibid 9
13 Piyush Khandelwal, 'Custodial Death in India
Misuse of Power' (2022) 2 Indian J Integrated Rsch L to exclude improperly obtained evidence, judicial oversight. These give real effect to the individual rights promised. However, even after realizing the widespread abuse and custodial violence cases, the practice remains institutionally advantageous.14 The Law Commission of
2023 also suggested the elimination of the distinction but the creators of BSA did not take the
suggestion into consideration. There was a new rule suggested by the Law Commission Report No. 113 of 1983 which would ensure that police officers would be held responsible if a suspect in custody gets hurt. The court would inquire on several factors such as the statement of the victim about the injuries, the duration of police custody, medical reports. However, there still remains a lack of accountability for police actions as this suggestion was not included in Section 22 or 23 of BSA. There was also a proposition by the Malimath Committee in 2003 to remove certain sections of the IEA, particularly Section 25 to 29 which are on confessions made to the police. Even the Law Commission Report No. 185 of 2003 recommended to make evidence obtained by unlawful means inadmissible.
This reflects the awareness of the commissions and committees about the consequences of the proviso of Section 23 of BSA and the growing cases of custodial violence. However, it was still ignored during the drafting of the BSA. 15 V.
THE COURT’S CONTRADICTION
The apex court of India has not been silent on custodial violence. In the case of Sheela Borse vs State of Maharashtra, the court issued directives while confronting torture and ill-treatment in police lock-ups which aimed at protecting detainees from the same officers that are responsible for their custody. 16 In another case, Nilabati Behra vs State of Orissa, the court held the state responsible for the custodial deaths by recognising that when a person is in custody of the police, they do not surrender their right to life. 17 And in the case of Munshi Singh Gautam v.
State of Madhya Pradesh, the judge highlighted with evident agony about how custodial violence is so extensive and unrestricted in a democracy governed by the rule of law,, he condemned those who regard themselves as defenders of democracy while allowing the men in uniform to be free without any repercussions for their unlawful methods. 18 These observation
14 Bhardwaj, N. (2024). "Confessions and Admissions
A Comparative Study of the BSA and the Indian Evidence Act." Journal of Criminal Law and Evidence, 28(2), 89-101.
15 Md. Imran Wahab, ‘Shortcomings of the Bharatiya Sakshya Adhiniyam, 2023
Challenges for Courts, Prisons, and Police’ (2024) 6(5) IJFMR accessed on 10 March 2026
16 Sheela Barse v State of Maharashtra AIR 1983 SC 378
17 Nilabati Behara v. State of Orissa AIR 1993 SC 1960.
18 Munshi Singh Gautam & Others v. State of Madhya Pradesh [2005] 9 SCC 631.
are not peripheral, they are the court’s own observation of custodial violence as an institutionally entrenched, systematic and constitutionally violative practice. However, still the same court that issued such condemnations has through cases like the Pooran Mal vs Director of Inspection and those that follow it, refused to implement the one legal instrument that would give those convictions structural effect. The court has held definitively and repeatedly that as long as the evidence is relevant, even if it is obtained through unconstitutional means it remains admissible. 19 It held custodial violence as alarming in one courtroom while made it legally profitable in another. The pomposity of rights and the logic of admissibility coexist in the Indian jurisprudence without any confrontation.
It is this very unexamined gap between judicial inaction and judicial condemnation that custodial torture continues to thrive. VI.
CONCLUSION
The continuance of custodial violence in India is sometime accredited to police culture, or to poverty or to failure of political will. Each reason captures something real, however none are sufficient. This article argues a more precise reason which is that custodial violence continues because it is made rational by the law. The proposed alternatives to an exclusionary rule fail on identifiable grounds. Criminal prosecution of individual officers is ineffective as statutory immunities are so broad that courts have upheld the quashing of complaints where the illegality of the officer’s actions was not seriously disputed. Departmental liability, on the other hand, reaches the institution and not the officer. Civil suits also face major obstacles such as good-faith exemptions and the nearimpossibility of establishing malice.
20 These failures attempt to address police misconduct outside the criminal trial while leaving the evidentiary benefit of that misconduct intact, a feature identified by the classical deterrence theory. An exclusionary rule is effective because it operates in a fundamentally different way from other remedies. It does not punish the violation after it occurs, but removes the investigative benefit gained from the illegal act the moment the state seeks to rely on that evidence in court.
19 Poorna Mal v. Director of Inspection AIR 1974 SC 348
20 Prema, Dr. E and Angamuthu, Shanmuga Sundaram, ‘Fruits of the Poisonous Tree – Exclusionary Rule and its Application in India’ (May 11, 2023 SSRN accessed on 11 March 2026 Only by eliminating the investigative advantage can the law remove the incentive that Section
23 of BSA has maintained for over a century.21
The recodification in 2023 was an opportunity to address a contradiction that was created by the legislature with full awareness, unfortunately it wasn’t taken. Until exclusion extends to the fruits of illegal conduct, and not only to the unlawful act, investigators will continue to treat torture as a workable investigative method as the law itself continues to signal that such methods work.
21 Ibid 20