Guest Post: Reconsidering the Expulsion Requirement in The Offence of Causing Miscarriage

 (This is a guest post by Ananya Kumar)

Under the Indian Penal Code 1860 (IPC), causing the death of a foetus was a criminal offence, even if done with the consent of the mother. The adoption of the Medical Termination of Pregnancy Act (MTPA) in 1971 created an exception to this. It allowed registered medical practitioners to terminate pregnancies with the mother’s consent, i.e., to carry out abortions. The IPC continued to penalise acts done with criminal intent to harm the foetus, or the mother, or both, i.e., the offence of causing miscarriage. Now, Sections 88 to 92 of the Bharatiya Nyaya Sanhita 2023 (BNS) prescribe punishment for the offence of causing a woman to miscarry without her consent in its varying forms in the same manner as the IPC.

Despite this statutory framework, a fundamental issue remains unresolved. Neither the MTPA, nor the BNS define ‘miscarriage’, and how it differs from ‘abortion’. Consequently, courts assume that both abortion and miscarriage entail the expulsion of a foetus from the womb before its term with the intent of causing its destruction, the only difference being that the former is done with the mother’s consent (see para 4).

However, miscarriage and abortion have significant differences and implications, beyond mere consent. For instance, miscarriages happen spontaneously and can sometimes lead to the death of a foetus inside the womb without it getting expelled. Yet, as things stand, courts have interpreted ‘miscarriage’ under the IPC to necessarily require expulsion. Consequently, courts have incorrectly dismissed charges of the offence of causing a foetus’ death on the grounds that it was not expelled from the womb. 

In this blog, I highlight how the law has wrongly assumed that expulsion of the foetus is a necessary element for the offence of causing miscarriage. I then point out cases where this has led to the acquittal of probable offenders for causing miscarriage. Accordingly, I suggest two interpretations through which courts can correct their approach towards interpreting the offences of causing miscarriage, in the absence of legislative action.

The Difference between Miscarriage and Abortion

In medical sciences, miscarriage is also referred to as ‘spontaneous abortion’. By contrast, what we ordinarily call abortion is called ‘induced abortion’. It is perhaps this terminology that has led courts to conclude that principally, the only difference between miscarriage and abortion is the non-consensual and spontaneous nature of miscarriage. 

For instance, Modi’s Medical Jurisprudence And Toxicology (10th edn, 1949) explains that when a woman consents to the expulsion of the foetus from her body, it is abortion. On the other hand, when a foetus is made to be expelled from a woman’s womb without her consent, it is termed as miscarriage (see page 319). In light of this, even when Parliament has defined miscarriage, in social security legislations, it has described it as the ‘expulsion of the contents of a … uterus’ (see, Maternity Benefit Act 1961 and the Code on Social Security 2020). This definition replicates the same error courts have been interpreting under criminal law.

The Supreme Court in Dr. Jacob George v State of Kerala (1994) relied on this explanation to state that the distinction between abortion and miscarriage is ‘not material’ for the purposes of the IPC so long as there is an intent to cause miscarriage (para 5). However, the difference between spontaneous and induced abortion cannot be reduced to the presence or absence of consent alone. 

There are many circumstances and complications that lead to different kinds of spontaneous abortions that do not have similar characteristics to abortions. Sometimes, complications in miscarriages, can lead to death of the foetus in-utero, without the expulsion of the products of conception. This type of miscarriage is called an ‘incomplete’ or a ‘missed’ miscarriage. This is why, equating miscarriages to ‘abortions with consent’ can miss the complicated reality of miscarriages. 

Another distinction between the two is the intent behind the act. Abortions are done with the deliberate intent of terminating and removing the foetus from the womb. However, it is not necessary that an offence is committed against a woman with the specific intent of bringing about the expulsion of the foetus. The intention may simply be to cause harm or death to the mother, or the foetus. In some cases, the child may die in-utero due to extraneous circumstances like shock, stage of foetal development, or death of the mother. Therefore, courts should recognise that miscarriages can also take place without the expulsion of the foetus.

Case Dismissals Due to the Expulsion Requirement

Courts have previously dealt with cases of incomplete or missed miscarriages under Sections 312 and 314 of the IPC. The interpretation by the courts in these cases shows flawed reasoning that has led to the acquittal of probable offenders, only due to non-expulsion of foetus from the mother’s body. 

For example, in Prasoon Kumar Jaiman (2017) before the High Court of Rajasthan, the petitioner was accused of causing miscarriage (Sec 313) where there was no expulsion of foetus from the mother’s womb. However, the Court held that no offence was made out due to a ‘lack of medical evidence’ proving miscarriage. The ‘lack’ here referred to the fact that the foetus had not been expelled from the mother’s body, which had been termed by the doctor as a ‘missed abortion’ (page 3).

Along similar lines, the judgment of the High Court of Chhattisgarh in Rajesh Kumar (2024) was another case where the expulsion requirement was misused by the accused. There the accused had allegedly administered a poisonous liquid to the mother with the intent to kill the foetus conceived outside marriage. The mother died soon after. In the autopsy, it was found that the foetus had died inside the womb before the mother, without getting expelled. The Court observed that causing miscarriage necessarily requires the expulsion of foetus (para 14). Since no such expulsion had occurred, the accused were acquitted on the charges under Section 312 and 314 of the IPC (Sections 88 and 90 of BNS). These cases demonstrate how the absence of a definition of miscarriage, as well as the judicial insertion of the expulsion requirement, allows perpetrators to evade culpability.

Avoiding Miscarriage of Justice

The most straightforward resolution to this issue would be for the legislature to introduce a statutory definition of ‘miscarriage’ in the BNS, one that does not make expulsion of the foetus a necessary requirement. A definition that instead focuses on the death of the foetus before term, whether or not followed by expulsion, would bring penal law in line with medical reality.

Alternatively, so long as there is legislative silence on the definition of miscarriage under the BNS, courts can prevent its abuse through two approaches. First, by interpreting the offence in a manner consistent with its medical meaning. Alternatively, by proceeding against perpetrators under other offences.

Firstly, as established above, miscarriages differ from induced abortions for more reasons than just consent. Keeping this in mind, interpretation of ‘miscarriage’ for the specific purpose of these crimes must be in line with the medical realities of their outcomes. Thus, the focus of the definition of miscarriage and abortion should not be on the expulsion of the foetus. Both aim to destroy the foetus inside the womb of the mother before the completion of the gestation period. Thus, regardless of whether there has been an expulsion of the foetus or not is not relevant for establishing the crime of causing miscarriage. The death of the foetus before term should be enough to meet that threshold.

Secondly, pending an authoritative reinterpretation of miscarriage by a larger bench, two alternative charges may be available to prosecutors in cases where there has been no expulsion. The foremost is Section 62 of BNS, which criminalises attempt to commit a crime. Thus, even if an offence of causing miscarriage is not made out on grounds of non-expulsion, offenders may still be tried for attempting to cause miscarriage with criminal intent if evidence is found.

The other alternative is Section 91, which punishes the act of ‘preventing a child from being born alive’ amongst others. Unlike the miscarriage provisions, Section 91 makes no reference to miscarriage or expulsion. While no court has applied Section 91 to situations of missed miscarriages yet, a textual interpretation of this Section would be wide enough to cover cases of intrauterine foetal death. 

Conclusion

By equating miscarriage with abortion for legal purposes, legislators and courts have overlooked the medical realities of miscarriage and in doing so, unreasonably narrowed the scope of criminal liability. This approach has serious consequences, as it allows individuals who have caused the death of a foetus with mala fide intent to evade culpability simply because the foetus was not expelled from the womb. Prior precedents demonstrate how absence of a statutory definition of miscarriage has become a loophole that undermines the objective of the law.

An interpretation of ‘miscarriage’ that focuses on the death of the foetus in-utero rather than its expulsion from the womb would resolve this inconsistency. Pending such a reinterpretation by a larger bench, the alternative routes available under Sections 62 and 91 of the BNS offer prosecutors a meaningful, even if not entirely ideal, path forward. Change in this area will require courts to look to the harm the provisions were designed to prevent.

Update - IBC Moratoriums, Personal Insolvency, and Cheque Bounce Cases

About a year ago, the blog covered a decision of the Supreme Court [Rakesh Bhanot v. Gurdas Agro, 2025 INSC 445] dealing with the issue of interim moratoriums under the Insolvency and Bankruptcy Code, 2016 [IBC] and prosecutions for offences under the Negotiable Instruments Act, 1881 [NI Act]. It took the view that the decision in Rakesh Bhanot did not sufficiently explain its conclusions, especially in light of existing precedent in the field. At the same time, the post also noted that the concerns animating the decision were, arguably, originating in poor legislative design of the IBC.

Over the past year, it appears that the correctness of the view taken in Rakesh Bhanot has come under a cloud. The most recent doubts were raised when in May 2025 a bench of Three Justices — headed by the Chief Justice — issued notice and granted interim reliefs in a petition which has, at least going by the news article, directly challenged the correctness of Rakesh Bhanot [Karan Bhatia, W.P. (Crl.) 131/2026]. 

There are tons of such fact scenarios playing out causing much grief to (some) litigants. A Three Justices Bench granting an interim relief in the face of the existing judgment, is a decent sign of which way the wind may blow. Additionally, it would be nice if, this time, the issue can also get the clarity that it requires. As the previous post discussed, there are some thorny law and policy issues involved here. It requires careful reading of the Indian Contract Act, the IBC, as well as precedent (such as P. Mohanraj). 

One can only hope that the issue thus gets finally resolved soon, one way or another. That is eminently more desirable than a Vijay Madanlal kind of scenario where the Court is kicking the proverbil can down the road which clogs the docket, as more and more persons move for interim reliefs, and also continues the confusion in courts below.

  

Course Correction on Delays and Bail under the UAPA?

[This post was updated at 6PM on 19.05.2026 to add reference to this news article and on 23.05.2026 to add reference to the SC Order dated 22.05.2026]

On 18.05.2026, a Division Bench of the Supreme Court handed down its verdict in Syed Iftikhar Andrabi [2026 INSC 503]. The bail application was allowed, which if we see the facts is not entirely surprising: the accused had been in jail for more than five years, there were many co-accused already on bail, some of them quite similarly placed, the incriminating material was largely confessional, and the trial was nowhere near completion. 

It is not because of the factual analysis that the judgment runs into 102 pages, which is quite unusual for a bail petition. The length was necessitated as the Bench felt it had to clear the legal position on a key issue affecting bail adjudication in the UAPA context, and the broader context of special crimes more generally. This, of course, was the issue of what impact does lengthy undertrial incarceration have in bail petitions under the UAPA with its additional restrictions for grant of bail under Section 43-D.

According to the Bench, a position had been set out by the Bench of Three Justices in K.A. Najeeb [(2021) 3 SCC 713], holding that the rigours of merits-based statutory restrictions for granting bail would, in the case of grave delays, be trumped by the concerns for an accused's liberty. The value of liberty and to a speedy trial was seen as grounded in Article 21 of the Constitution; lengthy undertrial detention would, in case of grave delay, be seen as causing detention in a manner contrary to these fundamental rights.

This position seemed to have been doubted and departed from in two subsequent judgments delivered by Division Benches; the first was Gurwinder Singh [(2024) 5 SCC 403], and the second Gulfisha Fatima [2026 SCC OnLine SC 10] (both were opinions of the same author). These subsequent decisions had, seemingly, held that bail was not the rule in UAPA cases owing to the restriction placed by Section 43-D UAPA which evidenced a legislative preference for jail as the default, and no relaxation or mollification of this statutory restriction was caused by delays or by the lengthy undertrial detention of an accused. 

The validity of this line of reasoning from both these opinions has been expressly doubted in Syed Iftikhar Andrabi [Paras 26 to 41]. The simplest line of critique is that the Division Benches could not have doubted or disagreed with the judgment of Three Justices in Najeeb. But the Court does not rest there, perhaps due to the manner in which both Gurwinder Singh and Gulfisha Fatima made departures. Neither said that the law laid down in Najeeb was bad; instead, the fight was about what law, if any, did Najeeb lay down in the first place. 

Gurwinder Singh and Gulfisha Fatima departed from the opinion in Najeeb by reading the latter as tightly wound to its facts. In that case there was no prospect of speedy trial whereas in Gurwinder Singh notable progress had already been made, and in Gulfisha Fatima the delays were 'natural' owing to the complexity of the case. In other words, both opinions tried to denude the opinion in Najeeb of having laid down any law at all. Having done so, both opinions offered their reading of the law, which saw the statutory bar of Section 43-D UAPA remaining unaffected by any length of delay. Even if there was delay, the detention of an accused may yet be justified in the overall facts of a case.

Syed Iftikhar Andrabi has specifically disapproved of this interpretation of Najeeb, as well as the holding of both Gurwinder Singh and Gulfisha Fatima that the statutory bar is adamantine and remains unaffected by delays. It reminds us of the very clear observation in Najeeb, that in appropriate cases the statutory bar would 'melt' in the face of the pressure created by the prolonged and unconstitutional deprivation of an accused's liberty. The statutory bar is, and remains, subservient to constitutional interests.

The verdict is a helpful course correction by the Supreme Court in the domain of bail on grounds of delay which has seen a renaissance of sorts in the past five years across multiple statutory contexts containing a clause restricting grant of bail. Whether the course correction will hold or not is something to be seen. All it takes is another Bench of Three Justices to come and doubt Najeeb, after all. And it took only a day for the State to raise the possibility. Think about this for a moment. It is not a right to release on bail, but the mere possibility for considering bail pleas slightly more favourably for those who have spent inordinately long periods in custody pending trial, is what the fight is about. And even this possibility is precarious, dependent upon how some words in a statute are read. For now though, the possibility shines brighter than it did six months ago.

At the same time, let us not overstate the usefulness of this jurisprudence of apology created by our courts. Where, after depriving individuals of their liberty for years, at some unspecified point in the future their custodial detention mysteriously burnishes their claims to liberty with greater light. In fact, Syed Iftikhar Andrabi has ironically reinforced the sheer unfairness of it all quite drastically. At paragraph 42, the Court cites data suggesting extremely low conviction rates for UAPA crimes, and rhetorically asks how can the prolonged detention be justified merely because of serious allegations if there are such high possibilities of a future acquittal. If this reasoning is good for prolonged detention and delay on grounds of bail, it is also good for demanding greater accountability for deciding bail applications on merits and seriously assailing the constitutionality of the restrictive bail clauses under the UAPA and other laws.

The allegations against the accused were serious in Syed Iftikhar Andrabi, but they were based on hardly any credible material. So why did the courts below continue to shy away from granting bail on merits for the five years that it took for this judgment to pass? That, really, is the question which deserves a lot more scrutiny going forward in bail jurisprudence under the UAPA. 

The vigour with which the Court has sought to resolve the creases within the legal fabric on delay-based bail for UAPA and other such crimes requires urgent reciprocation in the realm of merits-based bail. Only then will the proclamation of the Court in Syed Iftikhar Andrabi carry any real force: that bail remains the rule, and jail the exception, in all circumstances in our law.

UPDATE: It did not take long for the issue to get referred to a bench of larger strength. On 22.05.2026, while granting interim bail to the applicants, a bench of the Supreme Court took note of the observations in Syed Iftikhar Andrabi and requested the Chief Justice to constitute a bench of appropriate strength to "clarify or expound the position of law laid down in K.A. Najeeb's case, particularly in the backdrop of the rigour of 43D(5) which imposes restriction consciously and has received the assent of the Parliament, which obviously was brought in keeping in mind the valuable right enshrined in Article 21 of the Constitution of India." (Para 24). All eyes on the Court, then.     

[On reading of restrictive bail clauses, readers may find the six-part series 'Jail is Not the Rule' helpful]

Snippet - The Prison Statistics India 2024 Report

This past week, the Prison Statistics India Report for 2024 ['PSI Report'] was published by the National Crime Records Bureau ['NCRB]. I could not spot any massively interesting data points as such, especially if we read this Report together with the 2023 PSI Report. Indian prisons continue to be filled overcapacity, mostly with undertrial prisoners, who are usually young and poorly educated. While we do know that the prisons were collecting both religion and caste demographics, these indicators continue to be out of sight in these Reports. Economic profiles of prisoners are also not provided. If we did have that data provided in the Reports, one suspects it would further confirm that our prisoners are, disproportionately, coming from marginalised backgrounds.

The five-year charts in the 2024 Report show us that there has been a tapering of the prisoner population. The highs seen in the 2022 Report have, thankfully, not been witnessed again. Any positivity generated by this data is offset when we consider that that reasons for that tapering is not systemic improvement. One big reason are the efforts made in 2023 by the Supreme Court to push for release of prisoners through undertrial review committees. This is coupled by consistent decrease in prisoner population seen in Uttar Pradesh over the past couple of years. Anybody can see that these are not long-term solutions.

The 2024 Report is going to be interesting historically, as hereon the legal framework will no longer speak to the Criminal Procedure Code of 1973 but the Bharatiya Nagarik Suraksha Sanhita of 2023. 1973 saw the introduction of a new Code with an impetus on ensuring speedy justice and provisions meant to reduce unjust undertrial incarceration. It went out in 2024 with an unpleasant legacy on these fronts. The new law has built upon the same rotten foundations but with some tweaks on aspects of arrest and bail, so it would be interesting to see what changes it brings on the ground. 

A case in point will be the impact on default bail under what was Section 436-A of the 1973 Code, which has now become Section 479 of the 2023 Sanhita. This scheme of default bail under the old Code had enabled release of prisoners who had served half their prison term. The revised scheme allows for release on serving one-third of the sentence, provided it is a first-time offender (with some more caveats which we needn't elaborate here). 

Implementation of 436-A is seen by some as a chokepoint resulting in unjust incarceration. However, the PSI data has consistently shown that this may not be the case, as there was hardly a sizeable number of the prison population eligible for release in the first place. The 2024 Report reaffirms this trend. As we have seen, 479 BNSS will change the eligibility criteria, potentially expanding the reach of the clause. It will be fascinating to see the impact it has. If the half-year's worth of data is anything to go by, the future of this species of default bail may not be very different from its past.

 

Thinking about Section 27 (and 8) of the Evidence Act

Come July 2026, it would be two years since the three new criminal codes were brought into force. That is hardly long enough for the heavy, centuries-long, shadows of the old codes to disappear from the scene. At some points, you even feel sad for the new laws, which nobody can get around to either understanding or applying without referring to that familiar array of old clauses and provisions. Take Section 23 of the new Bharatiya Sakshya Adhiniyam 2023 ['BSA'] for example. Will Section 23 ever have an identity or an existence independent of its illustrious ancestors, Sections 25, 26, and 27 of the Indian Evidence Act 1872 ['IEA']? I highly doubt it.    

Ruminating about the identity crisis that will plague the new criminal codes is not the point of this post, as fun a topic as that may be. My focus is on the proviso to Section 23 of the BSA, or more specifically, its predecessor in Section 27 of the IEA. It is prompted by a sobering realisation, that there has been no post on this topic so far on this Blog.

The Statutory Scheme

Sections 24 to 26 IEA placed different kinds of barriers on evidence of confessions. Section 24 rendered any confession given due to inducement or threat from a person in authority 'irrelevant'. Section 25 stated that no confession made to a police officer by a person accused of an offence 'shall be proved' as against him. Section 26 stated that no confession made while in police custody shall be proved against a person unless it was made in the 'immediate presence' of a magistrate. Then came the proverbial backdoor in the form of Section 27, providing that where any fact is discovered pursuant to information given by a person in the custody of police, so much of the information which relates distinctly to the fact discovered. may be proved. Note that this scheme has been condensed into two provisions under the BSA; Sections 22 and 23. Section 22 is the successor to Section 24 IEA, and Section 23 of the BSA contains the erstwhile Sections 25 to 27 IEA.   

Logic and Theory

It can be argued that there are two dominant rationales guiding the legal scheme on confessions: first, respect for the 'best evidence' rule, and second, a distrust of acts done or statements made by a person who is under police control.

The best evidence rule is a hallmark of modern evidence law theory, which has two components. Most of us are familiar with its court-facing part: courts ought to decide cases based on the best evidence available to them. There is a legislative corollary, too, which commands the legislature to tear down technical rules which stand in the way of courts accepting the best evidence. For its time, the Indian Evidence Act made a lot of progress on the legislative side of things by removing older common-law hearsay objections. Today, though, the scheme of the Indian Evidence Act again appears antiquated when tested on this anvil, as can be seen with the approaches taken in other jurisdictions over the 20th Century. And, with blind repetition of this antiquated 1872 scheme in 2023, the process of law on the whole continues to suffer (yet another digression). 

In India's context, the pursuit of best evidence for courts under the adopted English system has coexisted with a general distrust of evidence purportedly given by persons under control of the police. A legal bar on using any statements, including of ordinary witnesses, recorded by police during investigation came in place since the start of codification in the 1860s. Such a step, no doubt, reduced the evidence potentially available before a court to decide a case; but the decision to remove such material from the adjudicative exercise was consciously made owing to fears about the police creating the evidence in question, which rendered it highly unreliable. Such fears were at their highest when it came to evidence of confessions taken by police from the mouth of an accused, prompting the total bar on evidence of confessions made to police.

I have offered a very generic sketch of a fascinating historical development in the mid 19th Century, but it serves our purpose for now, which is to highlight the two rationales in play guiding the legal scheme on the evidence of confessions. Understanding these two rationales, we can now proceed to appreciate the compromise arrived at by Section 27 IEA adopting the logic of confirmation by subsequent events. On its own, a statement by an accused may be worthless; but what if it leads to discovery of facts? Excluding the evidence of the statement even where it leads to discovery was seen as a step too far. As a result, Section 27 IEA allows proof of only so much of the information given by an accused which relates distinctly to the fact discovered.  

Interpretive Questions

One may be forgiven to think that the existence of strict rules barring admissibility of confessions, except in limited cases, would have been enough for the Indian criminal process to gradually move away from trying to use an accused as the primary source of gathering incriminatory material. The shift never really took place, apart from one of nomenclature where confessions are now increasingly called disclosures. Even today, a classic criminal investigation will almost always involve the accused making a disclosure, which may or may not directly lead to any recovery. The continued reliance on confessions for gathering evidence by police has naturally meant that courts are repeatedly faced with questions about admissibility of such statements. Which, in turn, means interpreting Section 27 IEA and figuring out of any part of the confession fits its contours.

I want to focus here on three interpretive questions which courts have faced and answered over time: (i) the meaning of 'fact discovered' for the clause, (ii) scope of the 'information' that is rendered admissible, and (iii) relevance of the conduct of an accused in such scenarios. There is a surprising level of coherence in judicial interpretation around these aspects, with of course existence of the odd incorrect view taken by courts from time to time.  

Fact Discovered

Since the logical premise underlying Section 27 IEA is that the discovery of a fact ensures reliability of the information given by an accused in custody of police, it is perhaps unsurprising to note that courts are strict about reading 'discovery'. If the police already knew about something, then there is no discovery of that fact made pursuant to the information given by an accused. Where this causal link is snapped, courts will not invoke Section 27 IEA at all.

This is all fairly straightforward, until you travel into the realm of determining what is a 'fact' that stands to be discovered pursuant to a statement given by an accused. A person's state of mind is also a fact, so can we argue that this is a fact discovered from the information given by an accused? The answer is no. It is not difficult to reason why this argument does not, and should not succeed. In this case, the information by an accused is the fact discovered — there is no independent confirmation of the reliability of information by any subsequent event. To apply Section 27 IEA this way would negate the very purpose of the clause and allow admission of confessions wholesale. Thankfully, this is one of the areas where the life of the law has remained closely tethered to logic, and courts have consistently insisted that the fact discovered must be a tangible object having an existence in the real world independent of the information by an accused. A great place to go and read about the development of law in this context remains the judgment in Navjot Sandhu [(2005) 11 SCC 600] 

'So Much' of the Information     

A common mistake is to think that discovery of a fact renders the evidence of a confession admissible in its entirety. The clause is much narrower in scope, admitting only evidence of so much of the information by an accused which distinctly relates to the discovery, and nothing more. Faithful application of Section 27 IEA thus requires to things in turn: first, clear and cogent material to show what was the information given by an accused, as the Supreme Court recently reminded police agencies in Boby v. State of Kerala [2023 INSC 13]. Second, courts to run a scalpel over such information to excise everything other than so much of the information which relates distinctly to the fact discovered. A good example of this approach of wielding a scalpel in action is Mohd. Inayatullah [(1976) 1 SCC 828]. 

What about Conduct?

It is quite often that an accused person does not merely give a statement, but either points out a spot from where an object is found or, even better, actually produces it himself. At the same time, an accused may straightaway point to a spot or produce an object without making any clear statement about the facts. Such scenarios throw up a different kind of evidence for a court to consider: the evidence of conduct, which is separately relevant under Section 8 of the IEA. How to sort such situations out?  

Conduct mixed with statements is a problem catered to by the statute. Section 8 itself clarifies that conduct does not include statements, except those which explain conduct. So it cannot be that by using conduct we can somehow broaden the scope of what is permitted through the narrow confines of Section 27 IEA, and say that the statements themselves are independently made relevant. There is another reason for this; any statements to police are unusable in evidence by virtue of Section 162 of the Criminal Procedure Code 1973. Section 27 is a statutory exception to this bar, but Section 8 is not.

When it comes to conduct which is not accompanied by any statements, we are faced with the same issue of the best evidence rule tussling with the distrust of evidence gathered from an accused by the police. Intuitively, an accused leading the police to material objects connected to the case is a relevant fact, which law should allow evidence for. But the risks of police creating evidence, which we saw associated with the evidence of statements made to police officers, do not magically disappear if we replace statements with conduct. 

Courts have held that these tests remain as important where police claim that the accused showed us a spot, or led to a recovery, although there is no comparable statutory prohibition in this scenario. As a rule, we see courts insist on some discovery of fact to render evidence of conduct relevant and admissible, and analyse whether it was truly the conduct which led to the discovery or was it a fact already known to the police. Additionally, courts also repeatedly insist that conduct evidence alone is quite a poor indicator for proof, and requires solid corroboration for it to further along the chain of circumstances being created in any case.                  

Constitutional Conundrums?

After the introduction of the Constitution in 1950, the evidentiary scheme of confessions has come within the constitutional crosshairs on a two notable occasions before the Supreme Court (there have been some High Court decisions as well). 

The first of these was Deoman Upadhyay [AIR 1960 SC 1125] where the validity of Section 27 was questioned on grounds of arbitrariness. The clause only applies to an accused person in police custody, prompting the unsuccessful argument that this invidiously discriminates between those in the custody of police and others. The second, was Kathi Kalu Oghad [AIR 1961 SC 1808]. Unlike Deoman Upadhyay, the constitutional question was only indirectly raised, as the Supreme Court was asked to decide whether there was a presumption of coercion accompanying statements made in police custody, thereby rendering Section 27 IEA contrary to the right against compelled self-incrimination. The Court unanimously rejected this argument and insisted that mere interactions with the police could not be deemed coercive [Note, though, that 15 years later a different albeit smaller bench in Nandini Satpathy did consider such issues more favourably].

The Future

It is quite likely that the current state of affairs will remain undisturbed in the short-term future; for no reason other than the legislative call to stick with the antiquated IEA approach even in the 21st Century. It will be interesting, though, to see what happens in the next 10-15 years as more interactions with the police have the benefit of audio-video recordings. Will a contemporaneous recording of the interaction prove to be the antidote for centuries of distrust regarding police affairs? With sufficient safeguards in place, there is no reason for it not to. Once this hurdle is crossed, the underlying basis of Section 27 will also be open to contestation. But not before that.

Snippet: Bellin's Critique of Criminal Procedure doctrine as a Law of Cooperation

I came across a new article posted to SSRN by Prof. Jeffrey Bellin titled 'Criminal Procedure as the Law of Cooperation' (forthcoming in the Georgia Law Review, 2026). The paper is focused on the development of U.S. criminal procedure by their Supreme Court, and so is not directly dealing with things that this blog talks about usually. But I thought that the argument made in the paper resonated quite well with critiques of how the Indian judiciary has developed criminal procedure as well.

His claim is straightforward - it is maybe a mistake to think that the main path taken by American courts on developing criminal procedure law is a rights-affirming one, which imagines restrictive interpretations as anomalies. He argues that, if anything, the dominant path has been imagining criminal procedure as a law for securing cooperation with law enforcement. Cooperation is the primary goal, and the enforcement of rights is subservient to ensuring that this primary goal is not unduly compromised. Decisions actively affirming rights while truncating cooperation-based interests are few and are outliers; these have often been whittled down over time because of the dominant cooperation-based framework (Miranda being the example he offers). 

This critique connects with existing critiques of Indian Supreme Court jurisprudence. For instance, Mrinal Satish and Aparna Chandra had made a similar claim while surveying Supreme Court decisions on the fundamental rights linked most closely to criminal procedure in their 2016 essay 'Criminal Law and the Constitution' which was part of the Oxford Handbook. 

In the context of Indian law, one feature which adds to the complexity of the argument, is that unlike the U.S. where criminal procedure is largely a creation of judicial interpretation of the relevant clauses in the Bill of Rights, in India the law on criminal procedure remains rooted in statute, and this statute influences how the courts develop doctrine. A key focus of statutes governing criminal procedure would naturally be on maximising cooperation with agencies. This places it in conflict, somewhat, with other laws seeking to secure individual rights even at the cost of cooperation. Whenever there is a conflict, it is no surprise that Indian courts tend to read the law in a way which ensures maximal cooperation, with some outliers over the years. 

Those interested may find the piece engaging and as a useful resource to turn to for developing arguments about the doctrinal development of Indian criminal procedure.

Guest Post: Criminal Status as a Constitutional Filter - What the Uttarakhand High Court's Order in Deepak Kumar Gets Wrong

(This is a guest post by Ammar Shahid and Syed Raiyyan)

The order passed by the Uttarakhand High Court (“The Court”) on 19th March 2026 in Deepak Kumar @ Akki v. State of Uttarakhand is a tight disposal of a petition that was asking for considerably more than it deserved. Three of four reliefs were turned down, and rightly so. However, the reasoning deployed against a prayer for police protection to Deepak Kumar is receiving significant criticism. It is not that the Court reached the wrong result. It is the path it chose to get there that rests on a doctrinal footing that does not hold up under scrutiny. When a Court ties the availability of relief under Article 21 to a Petitioner’s status as a “suspected accused,” it makes a statement about constitutional entitlement that no precedent supports.

Factual Backdrop

On Republic Day, 2026, Deepak Kumar (“Petitioner”), a gym owner from Kotdwar, Uttarakhand, intervened when a group entered the shop of one Vakeel Ahmed. He was a seventy-one-year-old shopkeeper and was pressured into removing the word “Baba” from the shop’s name. When the crowd demanded to know who he was, Petitioner identified himself as “Mohammad Deepak.” Days later, a First Information Report (“FIR”) was registered against him and his associate on charges including misconduct, snatching a mobile phone, and criminal intimidation. A mob later gathered near his gym and shouted slogans against him. Petitioner filed his own complaints, and his gym membership dropped from approximately 150 to nearly 15 in the weeks that followed. It was in this backdrop that he approached the High Court under Article 226.

The Petition and its Reliefs

The Petitioners were accused persons in FIR No. 19 of 2026, registered under sections 115(2), 191(1), 351(2) and 352 of the Bhartiya Nyaya Sanhita, 2023. They approached the Court with four prayers: quashing of the impugned FIR; directions to register FIRs on their own complaint against alleged perpetrators of hate speech; police protection for themselves and their families; and a departmental inquiry against certain police officials for alleged partisan conduct.

The Court dismissed reliefs (ii), (iii), and (iv) at this stage. Relief (i), the quashing of the FIR, was listed for further arguments. On reliefs (ii) and (iv), the reasoning is difficult to fault. On relief (ii), the petitioner had suppressed the registration of two FIRs on his own complaint. Both were registered at the same police station as the impugned FIR, and both predated the writ petition by more than a fortnight. Additionally, the statutory remedy under Section 175(3) of Bharatiya Nagarik Suraksha Sanhita, 2023 (“BNSS”), 2023 was available and unused. On relief (iv), a prayer for departmental proceedings against investigating officers while one’s own investigation is still pending practically invites the inference of an attempt to complicate the inquiry. The Court’s scepticism there is entirely well-founded and supported by reasoning. It is relief (iii), the prayer for police protection, that demands a harder look.

The Court’s Reasoning on Relief (iii)

The Court refused police protection primarily on the ground that a “suspected accused” who is under investigation cannot pray for adequate police protection, and that the police administration is “competent enough to keep on watch for his safety.” The Court also noted that the concerned officer had conveyed by telephone that there was no credible threat perception. The last ground of there being no credible threat was a perfectly adequate basis to refuse the relief. The first ground, on the other hand, is constitutionally tenuous.

Entitlement to Relief vs. Merits of Relief

There is a distinction in adjudication that has been overlooked in the order: the distinction between a petitioner’s entitlement to seek a particular relief and whether the petitioner has made out a case for that relief on the merits. These are entirely separate questions, and treating them as one produces reasoning that goes further than any given case requires.

Entitlement asks: Is this person, as a matter of constitutional or legal principle, capable of invoking this relief at all? Merits ask: has this person placed sufficient material before the Court to justify granting it? A Court can answer “yes” to the first and “no” to the second, and proceed to deny the relief without any doctrinal baggage. What a Court cannot do is answer “no” to the first question when the relief in question flows from Part III of the Constitution. Article 21 has long been understood to encompass an affirmative obligation on the State to protect it. In Akansha v. State of U.P., the Allahabad High Court held precisely this. The right to protection of life is a constitutional obligation of the State, and this obligation runs independently of the petitioner’s conduct or legal status.

Consider how this plays out in practice. When the Court says that a “suspected accused” cannot pray for police protection, it is not just deciding this case. It is articulating a principle. That principle, taken at face value, would mean that any person against whom an FIR is registered loses the ability to approach a Court for protection of their life until the investigation concludes. The logical endpoint of that proposition is uncomfortable: a person facing genuine, credible, life-threatening danger from a third party would be turned away at the door of the writ Court simply because he is simultaneously accused in another matter. Yet it is what the reasoning in this order, read literally, imports.

The merits dimension, by contrast, was entirely workable on the facts of this case. The petitioner placed no concrete material before the Court demonstrating a real threat. The investigating officer himself reported no credible threat perception. In those circumstances, the Court had everything it needed to deny the relief on the merits: there was simply nothing on the record to justify invoking the Court’s extraordinary writ jurisdiction for police deployment. The entitlement question did not need to be answered, and answering it the way the Court did created a proposition that the facts of the case did not require. This distinction is not novel, and the Courts have been asked to apply it in precisely this context before.

What Constitutional Precedent Actually Establishes

The Supreme Court’s Article 21 jurisprudence has maintained one consistent thread: the right to life does not shrink based on a person’s status in criminal proceedings. Several cases make this unavoidably clear.

In Arun Gavli v. State of Maharashtra, the Bombay High Court addressed this question of entitlement in the context of a person with a formidable criminal history. The Court’s approach was that the State’s obligation to provide protection under Article 21 is not neutralised by the fact that the person threatened carries FIRs or pending cases against him. The inquiry must remain limited to the threat, not the character of the person threatened. The Court held that “if a person like the petitioner, makes out a specific case of a genuine and imminent threat to his life on a particular occasion or at a particular place, the State Government would not be justified in refusing to grant armed police protection to him qua that particular occasion or place. This would, however, be subject to occasional review by the Threat Perception Committee constituted by the State Government.”

The Madras High Court’s position in K. Kalaivanan v. Superintendent of Police, Thiruvarur District is to a similar effect. The petitioner was involved in criminal cases and sought protection from rival group threats. The Court, in directing appropriate protection, reiterated that the right to life does not stand diminished by the existence of criminal proceedings against a person. Further, it is for the police to assess the credibility of the threat on the ground and not for the Court to bar the gateway to relief based on the petitioner’s legal status.

In Sunil Batra v. Delhi Administration, the Supreme Court held that even a convict retains fundamental rights. The Court drew a clear line: lawful custody restricts only those rights that cannot be exercised because of confinement itself. All other rights remain intact. If a convicted prisoner retains rights under Article 21, the argument that a “suspected accused,” who benefits from the full constitutional presumption of innocence, has a diminished entitlement to seek protection of his life would be legally untenable.

Taken together, these decisions convey what the Uttarakhand High Court’s order does not engage with. Courts have been asked this question of entitlement several times and have answered it in the affirmative. The identity, criminal status, or legal record of the person seeking protection goes, at most, to the credibility of the threat claim. It does not operate as a threshold disqualification.

Conclusion

To be fair to the Court, the outcome on relief (iii) is defensible. The petitioner placed no material before the Court establishing a genuine, immediate, or concrete danger. The police’s own assessment was that no credible threat existed. The relief simply was not made out on the facts. These were sufficient, clean, and principled grounds to deny the prayer.

However, the path the Court took to get there is a path that constitutional precedent does not sanction. Precedents converge on the same point: criminal status is not a constitutional filter through which the right to life must pass.

Courts often deal with petitions that are, as this one appears to have been, partly instrumental. The instinct to reject them firmly is sound. However, the legally sound path is in rejecting them on the ground that corresponds to the actual problem of insufficient merits. Formulating a broader principle that could one day be applied against someone whose threat is entirely real and whose only misfortune is the existence of an FIR bearing his name is untenable. One ground was enough here. Two grounds, when one is constitutionally unsound, is one ground too many.