Codification of Criminal Law (Part I): Four Themes in the IPC and BNS

July 1 will mark two years since the repeal of the colonial criminal codes and their replacement with their decolonised doppelgangers, making it the perfect time to talk about them. By now, it is beyond debate that the core architecture of the new laws mimics that of the old codes, so I need not spend time on rehashing the point and what it entails for a politics of decolonisation.

In a three-part series, I will discuss some themes in these three criminal codes. The themes being identifies are, as the title suggests, more interested in the idea of codification rather than the nuts and bolts of what the three codes do. So, for instance, while there is copious literature debating the merits of how the Indian Penal Code, 1860 [IPC] draws the line between murder and culpable homicide, there is comparably far lesser out there on how the IPC draws a line on criminalisation per se, and its enduring impact on the Indian legal system. I hope that gives a sense of the kind of discussion we have in store.      

This post looks at the IPC, with the next two posts addressing the Criminal Procedure Code of 1973 [Cr.P.C.] and Indian Evidence Act of 1872 — and by extension, their 2023 counterparts.

Theme 1 — Does the IPC Overcriminalise?

The Draft Penal Code for India prepared by the First Law Commission was published in 1837, on which the IPC was largely based, contained a dizzying array of crimes; both, in terms of categories—bodily offences to forgery to mischief—and classes within the categories—separate offences of simple hurt, assault, and criminal force all stemming from engaging in certain physical acts.

Being comprehensive was one of the key objectives behind the science of codification as it had emerged in the 18th and 19th Century, and if we judge the IPC by that metric, it is an obviously good effort at codification. But what about a different metric of the code's content. Did all of what the IPC render as a crime, merit criminalisation in the first place? 

The 1837 Report and its reviews by subsequent Law Commissioners, show us that there was an acknowledgment of this issue of line-drawing, especially because the IPC was not digesting the existing law but admittedly creating a new system of law. Viewed from this lens, the IPC was a politically and morally significant legislative exercise. A decision was being made by the ruling class to treat some forms of acts as warranting interference by the state, in the form of allowing prosecutions and then imposing sanctions in the form of imprisonment. 

Besides politics, there is also a question of what impact criminalisation has from a legal process perspective. In denoting some conduct a crime, it may be that there was already a recognition of that conduct as wrong or illegal. For instance, think of assault and battery. Even if it was designated a crime, there was legal recourse available to sue persons in tort and claim damages. Criminalising this conduct did not evaporate the pre-existing civil remedy but gave a new option for the same grievance. Both sets of recourse came with different aims, costs to litigate, procedural routes, burdens of proof, and outcomes. 

This is the point I want to drive home. In its framework, the IPC subsumed swathes of conduct that would also fall within the realm of private law. But because criminal law was a distinct legal process, it meant that the private law remedy continued to exist in tandem for those who wished to exercise it. Nothing was done in the Code to address this synergy between civil and criminal remedies for the same conduct, although the phenomenon was indeed acknowledged in the Draft Report of 1837 where Macaulay discussed cheating. The sense of overlap is also made clear when we look at a different source — The Draft Civil Wrongs Bill which was prepared by Frederick Pollock later on in the 19th Century (it did not come to pass). In many areas, the bill simply turned to the IPC, and reiterated that bringing a suit would not preclude criminal actions.

Understanding that the creation of the IPC as a political choice should make it obvious that it is influenced heavily by the times that one lives in, and requires periodic revisits to assess whether the choices of the 19th Century remain valid. But even jurisprudentially, the justifications offered by Macaulay in 1837 for criminalising vast swathes of purely private conduct — minor assault, battery, damage to property, insult, defamation, denying creditors, to name a few — were also a product of the times he lived in. Those choices were avidly contested back in the 19th Century, so it stands to reason that they should, if anything, be subjected to even greater scrutiny in the 21st.

Theme 2 — Widening, and not Questioning, State Authority       

Continuing with the theme of the times one lived in, comes the argument that a key purpose for the IPC was to solidify, legitimate, and maximise authority over a still-nascent dominion held by a colonial power. Understandably, the document does not spend much time fleshing out principles on which state power may be questioned or tested. 

To put it bluntly, there is no principle of criminalisation afforded within the document itself. As we saw earlier, this was a question which concerned the drafters (and reviewers), and their personal preferences about what should or should not be a crime informed the IPC. But there is no talk or discussion about codifying a set of principles to guide future legislators on how to principally go about the line-drawing exercise to identify crimes from torts. Presumably, because there could be no restraining the power of the state to unleash criminal sanctions on colonial subjects for whatever ends it saw fit.   

A logic of widening, and not questioning, state authority may make sense for a document that is being created by a young colonial power unsure of how to grapple with its increasing legislative functions. It does not make any sense for a sovereign democracy which professes to respect the idea of a culture of justification in theory and practice. Absence of any general norms on creating crimes has enabled the proliferation of an array of legislation utilising penal sanctions for purely political purposes. In failing to address this gap within the Bharatiya Nyaya Sanhita 2023 (the counterpart to the IPC), India's parliament has ensured a colonial continuity of the worst kinds.    

Theme 3 — The General and the Special Parts of Criminal Law

The proliferation of special laws without any attempt at consolidating the general law is an issue flagged throughout the course of the IPC's life after the 1970s, as more laws came to be passed creating crimes to address all kinds of social problems. In the 2005 edition of the wonderful Essays on the Indian Penal Code published by the Indian Law Institute, a chapter on reform specifically flagged this gap between the general criminal law within the IPC and the multitude of special criminal laws as a problem. 

This is an interesting theme to unravel, because we can clearly see that the start of special laws begins during colonial rule itself. A focused study may be able to point out what prompted the move to create crimes outside of the Code. From my limited, and very preliminary study, I would guess the answer was not as much considerations of substantive law, but more to do with the procedural framework. The creation of special criminal laws was, I would argue, to avoid giving those accused persons the full extent procedural guarantees that the Criminal Procedure Code would confer. Instead, these special laws prescribed their own procedure, which deviated quite substantially from the ordinary law. A good example here would be the Rowlatt Act. One might say that the same logic continues to inform special laws today. Would the UAPA, PMLA or NDPS carry the same weight without their truly draconian procedural regimes? 

If we view the proliferation of special laws as a procedural issue, then it becomes difficult to imagine how there can be any neat or proper consolidation of the general and special laws. This is where I remain skeptical of the patchy attempts at this consolidation made in the 2023 BNS, which has introduced organised crime and terrorism related offences. These offences in the BNS co-exist with the special laws with their special procedural regimes, and one wonders why any investigating agency would want to rely on them.            

Theme 4 — Punishments and Prejudices

Lastly, we come to how the IPC punishes crime. This issue is actually two issues rolled into one: the is first about judicial discretion to decide and inflict punishments, and the second about what the punishments are. The points made in the previous sections all blend in here. Given that the IPC was a document designed to consolidate and maximise state power, it did not wish to create any principles for what kind of crime may warrant imprisonment, even though there was a clear recognition by the drafters that all kinds of crime should not warrant a jail term.

In this regard, there has been some recognition of changing attitudes over time on the modes of punishment, as can be seen with the abolition of whipping and transportation as punishments in the first decade of Indian independence. But there remains little or no reflection on the scheme of unfettered judicial discretion to impose punishments and the arbitrariness that it entails. This, despite there being a telling set of passages in the 1837 Report about how judicial discretion to awarding sentences is an imperfect solution which can beget unfair outcomes.     

The BNS has addressed very little of this issue of judicial discretion as well, minimally altering the scheme of discretion present within the IPC. As for what punishments can be awarded, it has a new category of punishment — community service — which it doesn't use much. Arguably the only consistent reform on this front has been raising the amount of fines which can be imposed upon conviction. 

Concluding Remarks

These are overlapping and broad themes. One wonders what role they have played in shaping the legal system and their impact on how the system of criminal law operates even today. The absence of any guiding principles for criminalisation and levying of imprisonment means that anything, like dishonour of a cheque, can be made a crime punishable by imprisonment if the state thinks so. Or, for that matter, offences can be given terribly broad definitions to practically permit the prosecution, or persecution, of mere thought.

Then there is the issue of dockets. The exuberant coverage of purely private wrongs through the penal route gives litigants options. Many of them find the criminal process preferable, being a comparatively low cost route to civil proceedings, while at the time carrying far sterner bite with the prospect of imprisonment. This often prompts parties to settle their disputes, raising doubts as to whether the objective of the claimant was simply to use the cheapest means possible to get damages rather than to punish the defendant. 

What would have happened if the BNS shed some of this load and left parties to their own devices with civil remedies for many of the private wrongs that the IPC covered? No more assault and battery cases, no more defamation, and definitely no cheating cases where there is a contract, and you go file your tort claims before the Civil Judge. Would it help make a fairer, and more robust, criminal process? It would certainly reduce the use of criminal law for collateral purposes. 

Some things for the next review to think about, I guess.

Guest Post: On Promise to Marry and its Complexities

(This is a guest post by Shashank Sinha)

An interesting pattern can be observed in recent judgements which virtually creates an exception for the offence of rape by false promise of marriage. These judgements, which will be presently discussed, deal with the question of whether a complaint for this offence (which now stands codified with s. 69 of the Bharatiya Nyaya Sanhita 2023) is maintainable when the complainant knew that the accused was married at the time she had intercourse with him. Of course, this offence itself and its existence is hardly uncontroversial but the scope of this brief post is far too limited to be able to properly delve into those murky waters.

I first briefly go through some of the judgements which discuss the question broached above. Secondly, it will argue that these judgements are flawed as they fundamentally misunderstand the offence as it was created by the Supreme Court in Uday v. State of Karnataka [(2003) 4 SCC 46] and discuss certain judgements which have, directly or implicitly, held to the opposite effect. Finally, I offer my take on the correct position of law on this point.

Flawed Grounds

A spate of recent judgements have treated knowledge of the fact that the accused was married at the time of the intercourse on the part of the complainant as either completely exculpatory or as a contributory factor in the favour of the accused. The first of this line of judgements is that of the Kerala High Court in Sayooj S v. State of Kerala [2025 SCC OnLine Ker 4414; Decided on 02.07.2025] where the High Court, while considering a bail application, treated the complainant’s prior knowledge about the accused’s marital status as a favourable factor and granted bail [Para 9]. The Court also observed that it was possible that in the facts of the case, the complainant would be precluded from even raising a complaint under s.69 BNS. [Para 6]

One of the judgements the Court cited to take this view was its decision in Ranjith v. State of Kerala [2021 SCC OnLine Ker 5116; Decided on 15.12.2021], where the appellant was acquitted citing the fact of knowledge regarding the marriage of the appellant at the relevant time. Of course, there were other factors too on which the Court arrived at the conclusion that it did in Ranjith. However, what I do question is its holding on the point germane here. In the end, the Court in Ranjith, despite the many factors pointing in favour of the accused, still chose to rely, to some extent, on the fact that the complainant knew the marital status of the appellant and incorrectly held that it was a relevant factor in ascertaining whether the complainant’s consent was based on a false promise of marriage.

Another judgement in this line is that of the Supreme Court in Naim Ahmed v. State, [(2023) 15 SCC 385; Decided on 30.01.2023] where the fact that both the complainant and the appellant were married at the time of their relationship, which fact was known to the complainant at least for part of their relationship, was virtually treated as a complete defence even while the Court itself observed that one would be tempted to hold the appellant guilty if one remained limited to the relevant statutory provision and precedent laid down by it including its judgement in Uday. [Paras 18 and 21] The Supreme Court has hardly held so once. Recently, the Supreme Court in Pramod Kumar Navratna v. State of Chhattisgarh, [2026 INSC 124; Decided on 05.02.2026] explicitly ruled that an allegation of rape on the basis of false promise to marry would be unsustainable if one of the parties was married (as the complainant was here) and the same was within the knowledge of the other party. The Court ruled so despite the fact that the complainant’s divorce petition was pending adjudication. [Para 19] Again, in Jaspal Singh Kaural v. State (NCT of Delhi), [2025 INSC 457; Decided on 07.04.2025] one of the factors on which the Court relied on to discharge the appellant was the fact that the prosecutrix already knew that the appellant was a married person. [Para 13]

A Fundamentally Flawed Understanding of the Offence

One need say no more in the matter than the fact that in Indian law it has been the position for a very long time now that marriage is no longer sacrosanct in the sense that it may be legally ended. It is very much possible for a person to reasonably anticipate being married to another person while still being part of a subsisting marriage by hoping to end the latter first. In this view, it is quite inexplicable why courts recently seem to have taken a view that it is not possible for an individual to have this reasonable expectation when it is fully supported by the legal system. It would be a different thing to use knowledge of this fact to assess conduct of the parties but the judgements cited above use them as a factor in the offence itself as if its absence is an integral constituent of the offence.

It would be apposite, therefore, to refer to what actually are integral constituents of this offence. When the offence was, seemingly, created by the Supreme Court in Uday, it was only to curb the supposed menace of men inducing a false belief in women that they will marry them and obtaining their consent for sexual intercourse even when they had no intention of doing so. The crucial ingredients of the offence, therefore, are -

• False representation made by the accused to the prosecutrix to promise future marriage
• Sexual intercourse between the accused and the prosecutrix
• Consent given under the belief that the accused intends to marry her in the future

Nothing requires the accused to not be a married person or even makes it a relevant fact. The fact that the accused is married does not preclude him from making a promise of marriage in the future for two reasons. Firstly, the belief that he may induce might be such that he will marry the prosecutrix by legally ending his marriage by divorce or any other means available to him by the law. Secondly, the ability of the accused to marry the prosecutrix is not a relevant consideration for the simple reason that the ingredient of the offence is not whether the belief of the prosecutrix is tenable in law but whether it is bona fide and reasonable. While there is a fine line between the two, there is a difference nonetheless. The new section on this offence does state this does not amount to rape. However, the ingredients of the offence still remain the same and therefore these judgements cited above will unfortunately remain good law.

Conclusion

In conclusion, the only enquiry relevant to this offence is whether at the time of the intercourse, the accused made a promise to marry the complainant, whether the promise was genuine or fraudulent, and whether this gave rise to a reasonable expectation on the part of the complainant which, in turn, played a factor in her consenting to the intercourse. The marital status of either party should not be a relevant enquiry. One hopes, therefore, that these judgements do not hold the field for long and if this offence is to continue on the statute books, then it should, in the very least, be consistent.

Guest Post: Attachment without Safeguards? Section 107 of the BNSS

(This is a guest post by Dhruvika P)

Recently various High Courts across the country have clarified that the Bharatiya Nagarik Suraksha Sanhita (“BNSS”) does not permit unilateral police action under Section 107 of BNSS. In Headstar Global Pvt Ltd v. State of Kerala, the Kerala High Court held that the attachment of property cannot circumvent procedural safeguards and judicial scrutiny. In this case the move to freeze bank accounts under Section 107 of the BNSS without referring it to a magistrate was held to be clearly impermissible. The High Court also distinguished between Section 106 (seizing of property by police) and Section 107, affirming that under Section 107 attachment can only be ordered by a magistrate. Similar concerns have been addressed by the Delhi and Bombay High Courts. Following the reasoning of the Kerala High Court observing “blanket freezing” of bank accounts violates fundamental rights under Article 19 and 21. Seemingly, these rulings appear to reinforce procedural safeguards by judicial oversight. However, they raise a deeper and more troubling question: does involvement of a magistrate in attachment reasonably constraint exercise of arbitrary authority, or does it merely formalise a weakly regulated framework?

This post argues that Section 107 of the BNSS, despite requiring Magistrate approval, continues to enable broad and less regulated powers of attachment. In fact, compared to the much criticised Prevention of Money Laundering Act, 2002 (“PMLA”), the BNSS attachment regime lacks comparable safeguards and accountability mechanisms. Further leading to a paradox of how general procedural law is permitting attachment with fewer safeguards than a specialised penal statute. By examining Section 107 through the lens of the PMLA, this piece contends that judicial involvement alone does not cure the underlying defects in the BNSS framework. Instead, the statute allows exercise of excessive discretion thereby raising serious concerns about constitutional guarantees of fairness, proportionality, and due process.

Attachment: From CrPC to BNSS

The Code of Criminal Procedure (“CrPC”) contained limited provisions for property attachment. Section 83 dealt with attachment of property of proclaimed offenders and Section 102 allowed for seizures of movable property and bank accounts. As compared to CrPC provisions, Section 107 of BNSS introduces broader powers for attachment of proceeds of crime.

However, the BNSS does not clearly define the term "proceeds of crime" within the operative provision itself. Instead, reference must be made to Section 111 which appears in a different chapter dealing with reciprocal arrangements and procedures for attachment and forfeiture of property. According to this, “proceeds of crime” encompasses property derived directly or indirectly from criminal activities. This structural displacement creates interpretive ambiguity, as it is unclear whether the definition is intended to apply uniformly across the statute. The absence of an explicit and clear definition under Section 107 creates gaps in the definition and authorises for a broader application of this definition by investigating officers.

This ambiguity becomes significant when contrasted with the PMLA definition. PMLA is a special statute for serious economic offences and provides clear definition under Section 2(1)(u) which defines proceeds of crime comprehensively to include property derived from or generated from criminal activity related to scheduled offences (and includes property of equivalent value). 

Despite their different purposes, both the BNSS and the PMLA empower the State to attach property at the pre-trial stage. While the underlying objectives may differ in both the statutes, the means through which this power is exercised, that is the procedural framework governing attachment are structurally similar. This similarity of provisions has also been argued in Pay 10 Services Private Limited v. Union of India to prove consequences of attachment under BNSS. It is this convergence in procedure that provides a room for comparative analysis between BNSS and the PMLA provisions on attachment.

Procedural Safeguards: A Comparative Analysis

According to Section 5(1) of the PMLA property can be attached where the Director or any other officer not below the rank of Deputy Director authorised by the Director believes it to be proceeds of crime. The reason to believe has to be based on material in their possession. In contrast, according to Section 107(1) of the BNSS a police officer investigating with the approval of the Superintendent of Police or Commissioner of Police, may make an application to the Court or the Magistrate exercising jurisdiction to take cognizance of the offence or commit for trial or try the case, for attachment of proceeds of crime. Further according to Section 107(2) if the Court or Magistrate has reasons to believe, whether before or after taking evidence, that all or any of such properties are proceeds of crime, it may issue a notice upon such person calling upon him to show cause within a period of fourteen days as to why an order of attachment shall not be made.

Procedurally, Section 5 (1) PMLA provides more safeguards as compared to BNSS 107(1) & (2). According Section 5(1) of PMLA, the reasons for attachment have to be in writing and only an officer not below the rank of deputy director can attach. BNSS provision lacks explicit mention of written explanation and allows the magistrate to attach if they have reasons to believe. There are two concerns in BNSS. First, absence of explicit mention of written explanation. The Supreme Court in Kranti Associates Private Limited v. Masood Ahmed Khan has reiterated that “reason is the soul of justice”. It was also noted that reasoning should not be done merely for the sake of procedure but it must be done to serve the wider principles of justice. Second, the procedure to seek approval is unclear. It merely states that approval must be sought from the Superintendent of Police or Commissioner of Police but does not prescribe the manner, standard, or material basis for such approval. The vagueness of this requirement may result in approvals being treated as a mere formality.

Concerns with Adjudication of Attached Property

With regards to adjudication of attachment, PMLA under Section 8 empowers an adjudicating authority to decide these issues. BNSS empowers the court and the magistrate to decide such matters. Under PMLA there is a period of provisional attachment of 180 days. If the adjudication is not completed under 180 days the provisional attachment order becomes void. However in BNSS, there is only a period of 14 days to reply to the show cause notice if issued by the court or the magistrate, post the 14 days period an ex parte attachment or seizure of such property can follow, and it shall remain in force till a distribution order under 107(6) is passed.

Both the statutes deprive an individual’s right to enjoy property. In PMLA it is temporarily for a period of 180 days without any justification. The BNSS does not specify any time limits or mandatory review mechanisms. It is unclear and arbitrary in BNSS. PMLA’s framework includes specific provisions for modification or revocation of attachment orders. Section 5(5) allows affected persons to apply for order modification. BNSS lacks at least similar provisions, leaving remedies to general appellate mechanisms under the code.

In this light, the BNSS attachment provisions seem relatively ambiguous and underdeveloped in terms of procedural safeguards when contrasted with the PMLA framework. This procedural ambiguity regarding adjudication under BNSS also hampers judicial decisions that have established practical guidelines for the custody and disposal of property seized during criminal proceedings under CrPC regime.

Constitutional Standards and the Draconian Risk

The constitutional validity of Section 107 must be seen against the guarantees of Article 14 and 21 along with constitutional right to property. As established in EP Royappa v. State of Tamil Nadu, arbitrariness is antithetical to equality, and any arbitrary state action necessarily violates Article 14. This principle was subsequently strengthened in Maneka Gandhi and its just fair and reasonable standard. Section 107 BNSS, when tested against this standard, raises three at least concerns. First, absence of a mandatory requirement to record reasons at the investigative stage creates a risk of arbitrariness. Second, lack of clear thresholds guiding approvals, resulting in an open-ended and weakly-structured check. Third, lack of review mechanisms allowing for prolonged and disproportionate deprivation of property. Hence, these deficiencies also dilute the quality of judicial scrutiny itself.

The decision in Vijay Madanlal Choudhary v. Union of India further strengthens the case against Section 107. The Supreme Court’s reasoning to uphold the constitutionality of the PMLA attachment provisions, despite its severity rested on the structured safeguards within the PMLA framework. Section 107 BNSS, by contrast, replicates the coercive effects of attachment without incorporating equivalent safeguards even that of PMLA. As a general procedural provision, it cannot rely on the same exceptional rationale that underpinned the Court’s reasoning in Vijay Madanlal.

Conclusion

A more balanced attachment framework under Section 107 of the BNSS may be developed that requires states to ensure attachment of proceeds of crimes are complemented by adequate procedural safeguards. International scholarship suggests that these concerns are not unique to India. Modern attachment regimes struggle to harmonise between asset recovery and guarantees under rule of law. In AGOSI v. UK, the European Court of Human Rights held that in the confiscatory process, authorities must maintain a “fair balance” between public interest in combating crimes and protection of individual property rights. 

Legal developments in India indicate a need for attachment frameworks to retain sufficient procedural safeguards along with their effectiveness in attaching “proceeds of crime”. By explicitly mentioning recorded reasons, ensuring adequate judicial review, and providing meaningful opportunities to challenge attachment, the new BNSS attachment regime can avoid structural weaknesses that render a general criminal procedural law draconian.

Statutes, Courts, and the Constant Creating & Plugging of Gaps

Three decisions caught my eye over the past couple of months. Two by different benches of the Indian Supreme Court — Pune Bar Association v. UOI [Decided on 22.05.2026] and Dineshchand Surana [2026 INSC 579] — and one by a Two Justices' bench of the Kerala High Court — CMRL & Ors. v. UOI & Ors. [Decided 26.05.2026]. 

Pune Bar Association dealt with the new Bharatiya Sakshya Adhiniyam 2023 [BSA] regime for the certification of "electronic evidence" such as printouts, and the issue was whether its requirements were unconstitutional. This minefield of an issue was dealt with in two paragraphs, reasoning that since electronic evidence was more likely to be tampered or manipulated with, the imposition of stringent certification requirements was valid. No engagement with the change in law, with the fact that most other countries don't impose certification, and with whether adding a new certification requirement is, in fact, undermining the very claim of the BSA to eliminate the difference between traditional documents and electronic records. Just a lazy recitation of an old logic of tampering which has prompted judicially created barriers of entry for otherwise valuable evidence [discussed at length a long time ago on this blog]. 

At the other end of the spectrum was Dineshchand Surana, which is more than 150 pages long and a mini double-take on an issue which one member of the Bench had dealt with just over a year ago [discussed here]. Are prosecutions for cheque dishonour something which should be covered by the statutory bankruptcy moratorium for individuals? Every possible aspect of how to read the text of the cheque dishonour offence, text of the moratorium clause, and that of the criminal procedure regime, is covered across those 150 pages. The result is that the court ties us in knots by the end of the judgment, because the question it set out to answer — whether an act punishable with imprisonment is really a crime or not — is not something which the text of any law caters to or provides.

Then, there is CMRL, which may yet find its way before the Supreme Court given the stakes of the case and of the issue involved. A Two Justices' Bench of the Kerala High Court has held that (i) there is no relief of 'quashing' an Enforcement Case Information Report (ECIR) that can be sought (but will a writ remedy lie against this action? the court doesn't really answer), and (ii) the Enforcement Directorate can record an ECIR and use PMLA provisions for summoning persons to give statements even before any underlying scheduled offence is linked to the proceedings. A key component of the court's reasoning was, again, the statute, but in an inverse manner. What mattered was the absence of any statutory existence of the ECIR, and its non-statutory basis is the foundation for the judgment.

All three are significant because of the issues that they dealt with and they will surely generate a lot of critique. What I am interested in is the picture we see when we step back a little and look at these three decisions as illustrative of a much broader and commonplace theme within our legal system of how courts engage with statutes, and vice versa. 

For example, it isn't by accident or oversight that the PMLA (or for that matter the Customs Act before it) does not give any statutory basis to an ECIR. It is the result of parliament wisening up to judicial decisions back in the first two decades after independence on wielding of powers by the central statutory agencies dealing with issues such as smuggling and tax avoidance. The Supreme Court decided that a fuzzy statutory basis for conducting investigations allowed these central agencies to wield powers that regular state police do not have — such as taking sworn statements from accused persons, for instance. The message which Parliament took away was that lack of a clear statutory basis would insulate these powers from scrutiny. And we find that it has since become a hallmark of how most central agencies operate.

If CMRL is an example of what happens with Parliament keenly pays attention, one can say that Pune Bar Association and Dineshchand Surana are good examples of what happens when no attention is paid. Since Dineshchand Surana is sort of new, one can perhaps excuse legislators for not fixing the issues which first became obvious when the Supreme Court touched upon this issue in P. Mohanraj more than five years ago. If you are not feeling so charitable, like me, then you can clearly spot how the ongoing fiasco could have been so easily avoided by amending the statute in a timely manner. Now, we have more confusion on hand, in one of the largest arenas of litigation around — cheque dishonour cases. 

Certification is a problem far older than moratoriums, and which is why the BSA provisions and now their treatment in Pune Bar Association are perhaps the most upsetting parts of this story. It took nearly two full decades for law to be settled on how to treat the certification requirement under the old Indian Evidence Act 1872. Even so, many recognised how this requirement was, arguably, unhelpful, pointing famously to the fact that one of the inspirations for the requirement — similar clauses in UK Legislation — had since been repealed. Seemingly aware of this saga, we saw the BSA declare that electronic evidence will be at par with documentary evidence. So far so good, right? Well, not really, because along side this declaration, was the retention of an even more restrictive certification requirement which had to be met to introduce printouts and other obvious items into evidence — the clause under challenge in Pune Bar Association.

What is the point, really? I guess it is to write peculiar judicial opinions, to cite them, to make arguments using them, and pat ourselves on the back for our acumen being the only people who can access this arcane and byzantine domain. It certainly is not to ensure that the law is simple enough for everyone to understand it, use it, and hope for some results within one's lifetime.

Guest Post: Chaturvedi and the Remission Standard the Supreme Court Declined to Write

(This is a guest post by Kshitij Saruparia)

On 15 May 2026, in Rohit Chaturvedi v. State of Uttarakhand, a bench of Justices B.V. Nagarathna and Ujjal Bhuyan quashed the Ministry of Home Affairs’ refusal to concur in the premature release of a convict in the 2003 Madhumita Shukla murder case. The refusal was a three-line non-speaking order, and on that ground alone it was indefensible. The Court did not stop there. It held in paragraph 7 that the heinousness of an offence “stand[s] exhausted at the stage of sentencing” and cannot be revived to deny remission, and that to refuse remission on that ground “is to reconvert remission into a retrospective reaffirmation of guilt”.

The order quashing the MHA letter is plainly correct. The "exhausted at sentencing" formulation is not, and the judgment itself supplies the proof: the same judgment rests on two authorities that contradict it. The result will not be litigated. That sentence will be.

A Statutory Veto, and the First Test of how it may be Used

The case is decided within a frame the judgment sets out with care. The crime was investigated by the CBI, so by the Court’s own order of 31 January 2025 the State of Uttarakhand could exercise the remission power under Section 473(1) of the Bharatiya Nagarik Suraksha Sanhita only after the concurrence of the Central Government under Section 477(1). The textual change from Section 435 of the old Code appears slight, “consultation” has become “concurrence”, but the operational change is slighter still. The Constitution Bench in Union of India v. V Sriharan Murugan [(2016) 7 SCC 1] had already held, in 2015, that “consultation” under Section 435(1) meant “concurrence”, with the result that the State could not effect remission in CBI-investigated cases without the Centre’s agreement. That reading had operated as settled law for nearly a decade before the Sanhita was enacted. Section 477 does not create the Union veto; it codifies it, entrenching the stricter construction beyond judicial dispute and insulating it from future legislative attempts to restore the provision's literal meaning.

The Standing Committee on Home Affairs, which examined the Bill across twelve sittings in 2023 and submitted the 247th Report, made no observation on what became Section 477 in its clause-by-clause recommendations. The Home Secretary’s presentation of “important changes” to the Committee listed a Central Government concurrence requirement only in the context of sentences awarded by military courts, which is the subject of Section 476, a separate provision. The change at Section 477 converting a Constitution Bench interpretation of a federal power into statutory text drew no recorded scrutiny. Parliament codified Sriharan without Parliament deliberating on whether codifying it was the right choice.

For every centrally-investigated case, the Centre therefore holds a statutory veto over release that is now unambiguous on the face of the legislation and immune to the challenge that a future bench might read “consultation” more narrowly. Chaturvedi is the first judgment to test how that veto may lawfully be exercised, which is why its reasoning, and not only its outcome, matters beyond the petitioner.

The Court’s organising distinction is between sentencing and remission. Sentencing looks back and fixes punishment by the gravity of the act. Remission looks forward to conduct in custody, reform, and reintegration. Allowing the gravity of the offence to defeat remission lets the executive re-impose, through an administrative refusal, a punishment the sentencing court did not pronounce. That objection is sound. It is also not new. The Court itself cites Satish v. State of UP (2021), which held that neither the length of the sentence nor the gravity of the offence can, by themselves, be the sole basis for refusing premature release, and that the risk of reoffending must be assessed from the prisoner’s antecedents and conduct in custody.

The Contradiction is Inside the Judgment

Satish is the careful statement of the rule. Gravity is not the sole basis; it is one input, and it feeds the forward-looking question of recurrence. Chaturvedi adopts something larger. To say gravity is “exhausted” at sentencing is to say it has no residue at the remission stage at all.

That cannot be squared with the test the same judgment adopts a few paragraphs earlier. At paragraph 5.3 the Court sets out the Laxman Naskar factors as the governing criteria for premature release noting that those factors were also reproduced in Rohit Chaturvedi in paragraph 5.3.. The first factor is whether the offence is an individual act of crime without affecting society at large. That is, unavoidably, an inquiry into the nature and gravity of the offence, addressed to the remission authority.

So, the judgment instructs the executive to weigh the character of the offence and, in the same document, declares that character spent. Read literally, the two directions cannot both be operative: an authority cannot apply a factor that turns on the gravity of the offence while simultaneously treating that gravity as having no residue. A remission authority handed both has been given a contradiction, not a standard.

The way to read Chaturvedi without breaking it is to treat “exhausted” as rhetoric and the Satish rule as the holding. Heinousness cannot be a freestanding and conclusive ground of refusal, because used that way it does convert remission into re-sentencing. But the nature of the offence does not disappear. It re-enters as one input into the prognostic question the Court itself frames, the risk of recurrence and the safety of the public, exactly as Satish and Laxman Naskar contemplate.

The Court’s instinct is defensible, and worth stating at its strongest. If offence gravity is freely available at the remission stage, executive review collapses into a standing reconsideration of desert, and the sentence imposed by the court is no longer treated as settled, but is repeatedly revisited by the Ministry through successive remission decisions. Holding the line between punishment and post-conviction reform is a serious constitutional purpose, and it is plainly what the bench was reaching for. The difficulty is that the bench already had the instrument for the task. Satish keeps gravity out as a sole or dominant ground while letting it inform the prognosis, and that protects the distinction without pretending the offence has vanished. On that reading the case decides nothing the law did not already hold. “Exhausted” does no more work than Satish; it does the same work less accurately, and leaves behind a sentence the next litigant will weaponise.

The Standard the Court did not Write

The judgment’s nearest approach to a positive standard is paragraph 7.1, where the decision on remission “must emerge from a holistic assessment of the prisoner and after balancing societal interests with the prisoner’s right to be considered for release on fair and reasonable criteria”. This is not a standard. It does not tell the Ministry what a lawful refusal must contain, what evidentiary threshold attaches to a finding of dangerousness, or, most importantly, through which doorway the seriousness of the offence may legitimately enter the “societal interests” side of the very balance the Court prescribes. Having narrowed the admissible grounds on one side, the Court left the content of a permissible refusal undefined on the other.

It then demonstrated the gap instead of closing it. Rather than remit to the Ministry with directions on how to decide, the Court declined to remand at all, called a remand “an empty formality,” and performed the remission assessment itself, on parity with the released co-accused, time served, conduct in custody, and the State’s recommendation. The executive’s function was not corrected. It was substituted. That may be defensible on these facts, where the petitioner had served twenty-two years and a co-accused convicted in the same case was already free. It is no guide at all to the Ministry in the next case, where the facts are worse and the offence graver.

What is settled after Chaturvedi is what was already settled after Satish: gravity alone will not sustain a refusal. What is open, and now sharper, is the question the Centre’s Section 477 veto makes unavoidable. What must a reasoned, non-arbitrary concurrence refusal actually say, on what material, and through which doorway, if any, may the seriousness of the offence return as part of “societal interests”? The “exhausted at sentencing” line will sit at the centre of that fight, pressed by convicts to shut the offence out entirely and distinguished by the Union as loose language unsupported by the judgment’s own authorities. The Court had the materials to write the standard. It offered a philosophy of punishment where a workable standard was needed, and left that standard for another bench and another prisoner. The door it shut needed shutting. The one it left open is the one that decides cases.

Guest Post: Renumbering as Repeal? The Allahabad High Court on the interplay of BNSS with the SC/ST Act

(This is a guest post by Ammar Shahid)

Two orders from the Allahabad High Court raise a peculiar question: does the bar on anticipatory bail under Section 18 of the Scheduled Castes and Scheduled Tribes (Prevention of Atrocities) Act, 1989 (“SC/ST Act“) continue to operate now that the Bharatiya Nagarik Suraksha Sanhita, 2023 (“BNSS“) has replaced the Code of Criminal Procedure, 1973 (“CrPC“)?

In Dinesh Kumar Srivastava v. State of UP (December, 2025), the applicant faced an FIR under Sections 376, 506, 406, 323, and 504 IPC along with Section 3(2)(v) of the SC/ST Act, arising from allegations of a relationship formed on the matrimonial website jeevansathi.com allegedly on a false promise of marriage. He sought anticipatory bail under Section 482 of the BNSS. The informant’s counsel opposed this on the ground that Section 18 of the SC/ST Act barred it. The Court disagreed, holding that firstly, Section 18 specifically refers to “Section 438 of the Code” and anticipatory bail applications are now filed under Section 482 BNSS, the bar had no application. Secondly, the Court held that prima facie the material on record did not establish an offence under Section 376 IPC, given that the relationship appeared consensual. The petition was allowed, and bail granted.

Next, Sushil Kumar v. State of UP (May, 2026) arose from rejection of an anticipatory bail application solely on the ground of Section 18. The alleged facts involved a commercial dispute over unpaid sale consideration for food grains, accompanied by an alleged assault and caste-based insult. The High Court set aside the trial Court’s order on two grounds. The first was the trial Court’s failure to consider the principle laid down in Prathvi Raj Chauhan v. Union of India, where the Supreme Court held that courts retain the power to grant pre-arrest bail in SC/ST Act cases where prima facie the ingredients of the Act are not made out. The second was the BNSS transition argument drawn from Dinesh Kumar Srivastava. The High Court held that these two factors had not been considered by the trial court, and sent the matter case for fresh consideration.

Section 8 of the General Clauses Act

Section 8 of the General Clauses Act, 1897 provides that where a Central Act is repealed and re-enacted with or without modification, references in any other Central Act to the repealed enactment shall, unless a different intention appears, be construed as references to the corresponding provisions of the re-enacted legislation.

Section 482 BNSS corresponds with Section 438 CrPC. Both carry the heading of anticipatory bail and provide the same relief to the same class of persons in the same procedural context. When Parliament enacted the BNSS, it was re-enacting an existing one under a different statutory number. By operation of Section 8 of the General Clauses Act, Section 18’s reference to Section 438 CrPC must now be read as a reference to Section 482 BNSS.

Neither Dinesh Kumar Srivastava nor Sushil Kumar consider Section 8, which is a serious flaw. The legal question, at its foundation, was one of statutory interpretation across a legislative transition. Section 8 of the General Clauses Act is the rule Parliament has itself put in place for resolving exactly this question. Its absence from the reasoning of both decisions is a significant gap.

The only caveat Section 8 carries is the phrase “unless a different intention appears.” One would need to find, within the SC/ST Act or Section 18 itself, some indication that Parliament intended the bar to be tied specifically and permanently to Section 438 CrPC as a provision, rather than to the institution of anticipatory bail as such. No such intention is apparent. Section 18 was aimed at the relief of anticipatory bail, not at a particular section number. The number was simply the address at which that relief resided at the time.

What the 2018 Amendment Tells Us

The legislative history of Section 18 adds important context. In Subhash Kashinath Mahajan v. State of Maharashtra, the Supreme Court permitted anticipatory bail in SC/ST Act cases subject to certain conditions. Parliament responded within months by enacting the SC/ST Amendment Act, 2018, inserting Section 18-A which, among other things, reinforced the bar on anticipatory bail. Parliament actively closed a space that the Supreme Court had opened (which it has, since, reopened - more on that below).

The suggestion that this same Parliament, in 2023, inadvertently allowed the Section 18 bar to lapse through the BNSS transition without any express amendment is difficult to accept. When a legislature has demonstrated such attentiveness to a particular provision, the more reasonable inference is that it relied on established principles of statutory interpretation, specifically Section 8 of the General Clauses Act, to ensure continuity. If Parliament had wanted to lift the bar, it could have done so expressly in the BNSS. It did not.

The Prathvi Raj Chauhan Exception

Both decisions also place reliance on the principle formulated by Prathvi Raj Chauhan, where the Supreme Court held that in cases where no prima facie material exists to make out an SC/ST Act offence, Courts may grant pre-arrest bail through inherent jurisdiction under Section 482 CrPC. This reading of Prathvi Raj Chauhan, however, is broader than what the judgment actually supports.

The inherent jurisdiction of the High Court in criminal matters vested in Section 482 CrPC, which corresponds to Section 528 of the BNSS, not Section 482 BNSS, which is the anticipatory bail provision. Section 18’s bar has always applied to the anticipatory bail route, not to the exercise of inherent jurisdiction in exceptional cases. These are two separate heads of power, and conflating them misrepresents both.

Beyond this structural issue, the Prathvi Raj Chauhan exception is not a straightforward gateway to routine merits review. The Supreme Court was clear that this inherent power is available “sparingly” in “very exceptional cases” to prevent a “miscarriage of justice or abuse of process of law,” and was explicit that “a liberal use of the power to grant pre-arrest bail would defeat the intention of Parliament.” The Court in Sushil Kumar, having identified Prathvi Raj Chauhan as a primary ground for remand, does not specify how the trial court is to apply this demanding threshold on reconsideration. Without any guidance, the remand functions as an implicit invitation to conduct the kind of routine merits review that Prathvi Raj Chauhan cautioned against.

This issue is also visible in Dinesh Kumar Srivastava. The Court's own prima facie analysis found no offence under Section 376 IPC on the basis that the relationship appeared consensual and no force was alleged. But the Prathvi Raj Chauhan threshold is directed at the absence of SC/ST Act ingredients from the face of the FIR, not at the tenability of a connected IPC charge. Where an FIR contains specific allegations of caste-based insult or harassment, as it does in both the decisions under discussion, it is not obvious that the threshold for invoking the inherent jurisdiction exception has been crossed. Hitesh Verma v. State of Uttarakhand, which Sushil Kumar itself cites, reaffirms that the bar under Section 18 is absolute where SC/ST Act offences are prima facie made out. Citing Hitesh Verma while simultaneously remitting for fresh consideration of anticipatory bail, without a definitive holding on whether SC/ST Act ingredients are absent from the FIR, creates a real tension that the judgment leaves unresolved.

Conclusion

The question raised by these two decisions is ultimately not a complicated one. Section 8 of the General Clauses Act provides a clear and direct answer: the bar under Section 18 travels with the legislative transition from CrPC to BNSS because Section 482 BNSS is the corresponding provision to Section 438 CrPC. The legislative history of Section 18, including the 2018 amendment, makes it implausible that Parliament intended this bar to lapse silently. The Prathvi Raj Chauhan exception operates through inherent jurisdiction in exceptional cases, not through anticipatory bail as a matter of routine.

The SC/ST Act is protective legislation enacted for a specific purpose. Its provisions are intended to give effect to that purpose, not to be read in ways that drain them of content through procedural technicality. Courts interpreting special protective statutes are expected to favour readings that advance the statute’s objects. A reading that allows the bail bar to lapse because a provision number changed does the opposite.

[Note: at the time of writing, it is not clear whether either decision has been challenged in the Supreme Court, or has been questioned by a bench of larger strength in the High Court itself]

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.