The Other Origin Story of Preventive Detention in Independent India - Article 373 and the Extension of Duration Order

Preventive Detention, the practice of arresting and detaining a person in custody without a trial in court, has a long and storied history on the Indian subcontinent. That it was not only retained by the founders of independent India, but was given place in the Constitution of India, continues to raise eyebrows and lead the curious down the alleyways of history to learn about the peculiar set of circumstances that led to insertion of Article 22 of the Constitution of India. 

On this history of Article 22 — originally Article 15-A in the Draft Constitution that was deliberated in the Constituent Assembly — there is today a lot of writing [For instance, Granville Austin's books, Rohan Alva's book, Hallie Ludsin's book, and even old posts on this blog form 2016!]. Studying this history helps retain a sense of faith (for the faithful), as it shows that Article 22 was never meant to cement or legalise preventive detention (which was done when the Assembly passed the legislative lists) but to try and make sure that there is a respectable set of safeguards that no future government can avoid as it exercises powers of preventive detention.

There is another story, though, which has not been written about as much or at all. This is about Article 373 of the Constitution and promulgation of the Preventive Detention (Extension of Duration) Order on January 26, 1950, the day India became a Republic [pg. 8 here]. Let's get to it.   

'A Disastrous Situation'

Draft Article 15-A was discussed in September 1949 in the Constituent Assembly. As mentioned above, it was intended to introduce safeguards that could bind governments while exercising preventive detention powers. The key guarantee here, arguably, was the prescription to limit the period of detention by government to three months, and permit lengthier detention only if it was confirmed by an advisory board consisting of a judicial member. 

Either knowingly or unknowingly, the Constituent Assembly had just thrown up a curveball for the mandarins responsible for administering the affairs of state. A big part of the efforts to maintain law and order in post-partition India was to use preventive detention powers and detain persons for varying lengths of time. These powers were exercised at the level of the provinces through a motley bunch of public order statutes, almost all of which permitted preventive detention but did so by prescribing differing procedures. Not all of these statutes required an advisory board with judicial members to keep persons detained. Would all persons detained under such laws have to be released at the expiry of three years when the Constitution came into force? 

It appears that the chief mandarin in the Ministry of Home Affairs, HVR Iengar, was vexed with this question. He discussed it with R. Gupta, Home Secretary in the West Bengal Government while on a visit there and brought this "difficulty" to the attention of the Drafting Committee. As he recounted to Gupta in a confidential letter dated November 12, 1949, Iengar had pointed out to the Drafting Committee that "if the constitutional position was that persons who had completed more than three months of detention would have to be released on the 26th January, 1950, then the situation, from the security point of view, would be disastrous." 

The Drafting Committee, evidently, paid heed to Iengar's concerns. It "now recommended" insertion of Article 373 to the Constitution which sought to confer a power upon the President — in short, the Executive — to pass an order which would temporarily prevent the enforcement of Article 22 in full and allow the three-month limit to be extended. The Draft Order so prepared by the Drafting Committee provided that detention of persons detained (for any litany of reasons) "may be extended" beyond three months without orders of an Advisory Board, up to a maximum period of one full year from the date of the Order.  

The disastrous situation could thus be averted, and Iengar informed Gupta that the Government was taking "necessary steps" to pass this draft order on January 26.

Extending All Detentions For a Year?

HVR Iengar's confidential letter of November 12 was sent out to all provincial governments and the States Ministry of the Government of India. There were seventy-five days left till January 26, and it made for a frantic last two months across governments to ensure that no loopholes were left which accidentally allowed detenus to be freed. To this end, the Government of India sent out instructions on December 31. 1949, to all the provinces asking them to "forward immediately by telegram a statement indicating in exact terms the grounds on which such (preventive detention) prisoners have been detained.". A separate letter of the same date shared a draft of the order to be passed, and sought comments "not later than" January 14, 1950. (All emphasis in original). A stream of telegrams followed, but very few offered substantive comments. The proposed draft of the order shared by the Government seemed comprehensive enough to them.

This process threw up a different problem for consideration too. Replying to the request for comments, the Chief Secretary to the Government of the United Provinces flagged concerns on December 19, 1949 that, while an Advisory Board already existed, there were other parts of their preventive detention law [The United Provinces Maintenance of Public Order (Temporary Act, 1947] that were inconsistent with Article 19 and 22 of the soon-to-be passed Constitution. 

We get a glimpse of how the Draft Order was understood by the Government of India in the reply letter dated January 20, 1950 issued to the United Provinces. Interestingly, this letter referred to not one but two Executive Orders that would be issued by the President for preventive detention. The first of these, to be issued on January 26, "will permit detention under Provincial Acts up to a period of one year ... and will also make it unnecessary during this period to submit the cases for review by the Advisory Boards in order to get over the limit of three months" and, critically, as a result "the extension will be automatically granted to all detentions under provincial acts by the terms of the proposed order." (Emphasis mine). In other words, while the draft had said that the period of detention "may" be extended, the Government understood this to mean that detentions shall be extended without exception. Since all detentions would be extended, there was no need felt to amend any preventive detention laws for that immediately, and "it will be sufficient if these acts are brought into accord with Article 22 of the Constitution at any time before the expiry of one year's period" (Emphasis mine) for which the Central Government would issue a second order by April 1950 detailing what changes are required in all provincial laws.

The Curious Shift in Stance

On January 26, 1950, the Gazette of India (Extraordinary) published order C.O.8, titled the "Preventive Detention (Extension of Duration) Order, 1950" issued in exercise of powers conferred by Article 373. It did not contain any exhaustive list of cases for which detention could be extended, and permitted extending the time period of detention for only three further months. This was a big change from the proposed extension of an entire year, which we know that the Government had committed to in writing less than a week prior. The seeds of this change were, presumably, sown already in the idea of passing two Presidential Orders instead of one. Since only three months were seen as being required to bring the existing legal machinery in tune with the Constitutional scheme on preventive detention, there was no need to detain persons for an entire year.

The archives suggest that not every province received copies in time or in sufficient numbers, as frantic telegrams were sent to the Central Government asking for additional copies to be sent at the earliest. It also appears that the original order issued in the Gazette had an embarrassing typographical error, having gotten the date wrongly written as January 20 (incidentally, the date of the letter sent to United Provinces), and not January 26, 1950! This was not the only problem in the Order, though, which the Home Ministry admitted on February 13, 1950 was "revised and issued at very short notice" because of which "the full implications of its provisions could not be examined in detail". The implication being referred to was this: the Order potentially created a three month upper limit even for those persons detained after January 26, 1950 under laws fully compliant with the constitutional scheme. These and other issues were "being further examined" at the time. 

Post-Script

As it turned out, the Preventive Detention (Extension of Duration) Order 1950 did not even tide over the three months as had been planned. 

The problem flagged by the United Provinces Government in December 1949 of provincial laws being inconsistent with the Constitution and therefore void was seized upon by enterprising lawyers in cases already pending at the time of the Constitution's coming into force. It led to the Patna High Court holding detentions invalid on February 14, 1950 in Brahmeshwar Prasad [AIR 1950 Patna 265] concluding that the Order could not save the provincial law itself from the vice of constitutionality (The judgment came one day after the Home Ministry's letter and interestingly also suggests that the Government Lawyer did not have access to proper copies of the Gazette with the Extension of Duration Order). 

Much before this, though, the Government of India had already begun to draw up plans to render the Presidential Order redundant by passing a Central Law on the subject. In a letter of January 31, 1950 sent by the Home Ministry to all provinces, we find reference to this plan. Perhaps the judgment of the Patna High Court, coupled with the many problems thrown up by the language of the Presidential Order, are what led to the ungodly speed with which the Preventive Detention Act was tabled and cleared on February 25, 1950 by the Provisional Parliament. And the rest, as they say, is history.

[The post is based on a file being F.No.10(95)-P/49 titled "Protection against arrests and detention in certain cases - Clause (7) of Article 22 and 373 of the New Constitution. Issue by the President of the Preventive Detention (Extension of Duration) Order, 1950". It can be viewed, for free, on the National Archives of India portal.]

Guest Post: The Trajectory of Credibility - Forensics, Suspicion, and the Future of Circumstantial Justice in India

(This is a Guest Post by Sumedha Edara)

Could the Supreme Court’s decision in Vaibhav v. State of Maharashtra (2025 INSC 800) mark a pivotal shift in Indian criminal law jurisprudence? The context is as such – Mangesh, a medical student, passed away from a gunshot fired by the service pistol of his friend, Vaibhav’s father who was serving as a police officer. Both the Trial Court and the Bombay High Court convicted Vaibhav of murder by relying on his conduct post the incident—removing the body, cleaning bloodstains, and feigning ignorance—which they treated as incriminating “subsequent conduct” under Section 8 of the Indian Evidence Act 1872.

While the judgment itself confines its reasoning to established doctrines (Section 8’s evidentiary logic, forensic analysis, and prosecutorial burden) this article advances an interpretive framework, reading Vaibhav as signalling a transition from “suspicion justice” to “forensic rationalism” in criminal adjudication. What follows is a conceptual structure: the “Trajectory of Credibility,” organising judicial reasoning around forensic coherence, behavioural plausibility, and probability of motive. This analytic extension draws on comparative jurisprudence and psychological insights, situating the case as both doctrinal correction and a node for broader institutional debates.

The Judgement: Three Transformative Moves

The Court’s reasoning can be distilled into three moves which, while based purely on evidentiary doctrine, benefit from a closer reading between the lines to highlight broader themes that are implied:

i. The Limits of Conduct

The first move was to restrict conduct-based inferences under Section 8 of the IEA. The Trial Court and High Court had treated the acts of concealment as determinative of guilt. Their reasoning was circular – because the conduct was incriminating, it proved guilt; but because guilt was presumed, the conduct appeared incriminating. The Supreme Court broke this cycle by reaffirming that Section 8 only establishes relevance and not sufficiency. Conduct, while it may corroborate existing proof, cannot fill evidentiary gaps.

Importantly, the Court explained that conduct is “equivocal”, meaning that the same act may admit multiple readings. A frightened teenager may conceal evidence out of panic rather than guilt too. By ignoring this duality, the High Court arguably failed to fully adhere to the time-tested principle that for any conviction on circumstantial evidence there must be one hypothesis only, and that must be one of conclusive guilt. In Vaibhav, there was no such unequivocal conclusion: no ballistic confirmation, no motive, no eyewitness, and so no singular conclusion.

In clarifying the distinction between relevance and sufficiency, the Supreme Court rejected reliance only on conduct to establish guilt. I interpret this move as a step away from "suspicion justice"- where behavioural shortcuts substitute for concrete evidence as required by the law - using the judgment as a basis for further conceptual reflection. Such an interpretation repositions Section 8 as an evidentiary aid and not a judicial crutch.

ii. Forensic Trajectory as the Decisive Narrative

The second move was to foreground the factual significance of ballistic trajectory, rather than the accused’s conduct. This fact had not been considered below, by virtue of the pistol belonging to the accused’s father, chances of it being an “accident” are irrelevant. This conflated ownership of the weapon with the authorship of the act itself. The Supreme Court treated this as a fundamental flaw and held that the witnesses’ testimony regarding the bullet trajectory was clear and trustworthy, because only such an account led to any coherence in determining the chain of events that took place. In doing so, the Court effectively established that a trajectory analysis cannot be reduced to entry and exit wounds, and that the HC’s failure to grasp this was not a technical error but one that revealed a collapse of reliance of forensic literacy. Although the Court stopped short of theorising this as a broader principle, its reasoning implicitly restores the role of scientific coherence as central to criminal adjudication.

iii. The Re-embedding of Burden of Proof

The third and perhaps most constitutionally significant move was assertion of prosecutorial burden. The High Court had held that the burden of proof lay with the accused in explaining how the pistol had reached victim's hand and why he had concealed the evidence. This approach effectively inverted the doctrine of presumption of innocence. The Supreme Court restored the doctrinal balance in its judgement, reaffirming that the burden lies on the prosecution to first establish guilt beyond reasonable doubt before one turns to the accused for explanations. The accused’s obligation is always one of plausibility, not certainty. We are reminded here that silence or implausibility on part of the accused cannot rescue a weak case for the prosecution. Proof is the State’s burden and doubt is the citizen’s entitlement.

Together, these three reasonings by the Supreme Court successfully righted the wrongs that have been committed by the lower courts in adjudicating this case, and consequently, their broader implications have also reaffirmed evidentiary law’s true nature – conduct is only corroborative, science is decisive, and burden is constitutional.

The Test of "Trajectory of Credibility"

Drawing upon the judgment, I introduce here a conceptual “Trajectory of Credibility” framework, meant as a tool for analysing circumstantial evidence cases. It demands that the credibility of evidence must be followed along several axes until it reaches its destination, much like the path of a bullet.

The first is forensic coherence. It demands that courts ask whether scientific traces like ballistics, pathology, DNA, etc., can sustain guilt independently. In this case, the prosecution story was refuted by forensic evidence which showed that the bullet's upward trajectory indicated accidental discharge rather than homicide, proving that doubt endures where coherence fails.

Plausibility of behaviour is the second axis. Human psychology, cultural background, and stress response must all be taken into consideration when interpreting post-crime behaviour. Concealment may signal guilt, but it may also reflect panic. In this case, the accused’s conduct was reinterpreted as acts of a frightened student fearing his father who was a police officer. Behaviour of such sort is punishable under Section 201 of IPC, but not probative of murder.

The third axis is probability of motive. Courts must determine if the record reveals any animosity or a legitimate motive. The cases of Anwar Ali v. State of HP (2020) and Nandu Singh v. State of MP (2022) demonstrate that while lack of motive is not always fatal per se, its complete absence in cases based on circumstantial evidence tilts the scales in favour of acquittal. The same happened in this case, where a lack of animosity between the victim and accused weakened the hypothesis of guilt pointed towards the accused.

Only when these axes converge does the judicial trajectory close around guilt. Where they diverge, as in this case, an acquittal is constitutionally mandated.

Structural Gaps Exposed

The case also exposes the fragility of India’s judicial infrastructure. The judgement revealed not only prosecutorial shortcomings but systemic deficits that simply cannot be ignored.

The most glaring one is the country’s forensic deficit. India’s Forensic Science Laboratories (FSLs) are chronically understaffed and underfunded. Reports of DNA samples lying unexamined for months or even years are common; ballistic reports are often unavailable, as in this case. The Malimath Committee Report (2003) and 239th Law Commission Report (2012) both flagged this weakness, yet capacity remains stagnant. Without timely forensic reports, cases tend to lean only on circumstantial suspicion, further raising risks of wrongful conviction.

A second gap lies in the overbreadth of application Section 8 of the IEA. While the Supreme Court in its judgement insisted on caution, trial courts remain free to overvalue conduct. The risk is amplified in a society where fear of police is pervasive, and acts like panic driven concealment or tampering of evidence is common.

Third, judicial trust in science arguably remains low. The High Court’s refusal to give due weight to the bullet's trajectory despite clear expert testimony on the same, exemplifies how courts can undervalue or misunderstand scientific evidence. This is not unique to this case. In the Aarushi Talwar case too, conflicting forensic reports left trial courts confused, producing faulty inferences. And in the Uphaar fire case, lack of technical expertise hampered fact-finding. Without systemic judicial training, scientific ambiguity, which is possibly a result of the lack of investment in the science out of the distrust as discussed, will continue to be resolved through intuition rather than expertise.

Finally, this case highlights the absence of accountability for investigative failures. No sanction or consequence followed the failure to obtain ballistic verification. Without any mechanisms to discipline such lapses, investigators, obviously, face little incentive to meet forensic standards.

Together, these gaps explain why suspicion has long substituted for science in India’s criminal justice system. The implications of this judgment can mark an affirming point in jurisprudence, but without institutional reform, its jurisprudential promise risks remaining aspirational.

Broader Implications

The resonance of this case lies not just in doctrine but also in its potential ripple effects across institutions.

For police, it may signal towards the end of shortcut investigations. In a system often driven by narrative first, science later policing, the judgment redirects energy towards scene preservation, evidence collection, and expert consultation.

For prosecutors, it may raise the evidentiary threshold. Plausible narratives are no longer sufficient without forensic ballast. Prosecutions must now secure scientific backing to their claims - a burden, but also a liberation from reliance on conjecture.

For defence counsel, this precedent may be a powerful shield. It provides a doctrinal justification to dismantle prosecutions hinging on conduct or silence. It also supplies and supports a vocabulary - “trajectory,” “probability,” “alternate narratives” - that empowers defense advocacy.

For the judiciary, the case may redefine judicial satisfaction. No longer can suspicion or silence tilt the scales. Conviction now demands for the convergence of forensic coherence, plausibility of conduct, and probability of motive. Judges are required to think less as moral arbiters and more as epistemic adjudicators.

The ripple effects may extend further. Law schools may finally learn to integrate criminal law with forensic science and psychology. Forensic institutions gain visibility and pressure to expand. Civil society and rights advocates gain a powerful precedent affirming that wrongful conviction is as grave as wrongful acquittal.

Conclusion

The judgment in Vaibhav v. State of Maharashtra reconstructs the logic of evidence. The Court restored forensic centrality, reaffirmed prosecutorial burden, and curtailed excessive reliance on Section 8 in order to re-ground criminal law in constitutional principles. It enforces epistemic discipline through its framework, which tests consistency across science, conduct, and motive: divergence forces acquittal, while alignment supports conviction. Its acknowledgment of suspicion bias - that concealment may be motivated by fear rather than guilt - is equally noteworthy. However, the ruling also revealed institutional flaws, including the misuse of Section 8, judicial illiteracy, a fragile forensic system, and an impunity for investigative errors. Without reform, suspicion risks continuing to masquerade as proof.

Guest Post: A Sorry State of Affairs, which is the Law Governing 'Dowry Deaths' in India

(This is a guest post by Shikhar Aggarwal)

The dowry-related ‘murder’ of Nikki Bhati, nine years after her marriage, sent shockwaves throughout society. More so because this incident broke the illusion of dowry harassment being a thing of the past. It also somewhat shattered the narrative surrounding an ever-increasing number of ‘false’ cases involving ‘misuse’ of provisions meant to safeguard women and to deter men and their families from indulging in such harassment.

Upon closer examination, I realised the inadequacy of these laws in tackling the menace, especially due to the inherent structural issues surrounding their application and interpretation. This post examines the substantive criminal law provisions attracted in such cases and the concerns therein. I argue that the need of the hour is to overhaul the criteria governing applicability of these provisions (i.e., the threshold of a woman’s death occurring within seven years of marriage).

Glaring loopholes surrounding these provisions ex facie

Section 304B, introduced in the Indian Penal Code (‘IPC’) in 1986, defined ‘dowry death’ as follows:

(1) Where the death of a woman is caused by any burns or bodily injury or occurs otherwise than under normal circumstances within seven years of her marriage and it is shown that soon before her death she was subjected to cruelty or harassment by her husband or any relative of her husband for, or in connection with, any demand for dowry, such death shall be called "dowry death", and such husband or relative shall be deemed to have caused her death.

This deeming provision is pari materia with Section 80 of the Bharatiya Nyaya Sanhita, 2023 (‘BNS’). It holds husbands and/or their relatives liable for any unnatural death of a woman within seven years of marriage, if ‘soon before her death’, they subjected her to cruelty or harassment for/in connection with dowry-related demands. A gaping hole in this provision regarding what constitutes ‘cruelty’ was resolved judicially in Shanti v. State of Haryana, by importing the definition of ‘cruelty’ from Section 498A, which was introduced three years before Section 304B separately punished the specific act of causing bridal deaths within seven years of marriage for dowry.

The Explanation to Section 498A, however, defines cruelty specifically “for the purposes of this Section”. It reads as:

(a) any wilful conduct which is of such a nature as is likely to drive the woman to commit suicide or to cause grave injury or danger to life, limb or health (whether mental or physical) of the woman; or (b) harassment of the woman where such harassment is with a view to coercing her or any person related to her to meet any unlawful demand for any property or valuable security or is on account of failure by her or any person related to her to meet such demand.

The BNS has now split Section 498A IPC into Sections 85 and 86, the latter defining ‘cruelty’ “for the purposes of Section 85”. Hence, the term ‘cruelty’ in Section 80 BNS remains undefined, relying on judicial interpretation that seemingly contradicts the bare text of Section 498A IPC/86 BNS, which restricts the definition of this term “for the purposes of” just one Section.

It is unlikely that this omission of a definition in Section 80 BNS/304B IPC was mere oversight, given that Parliament was careful enough to explicitly define ‘dowry’ as under Section 2 of the Dowry Prohibition Act, 1961. In Shanti, the Court noted that Sections 304B and 498A are not mutually exclusive – yet, despite an opportunity to correct this obvious anomaly, the BNS continues to perpetuate this confusion apparent within the text of the law.

How this plays out practically

To understand the significance of this, take the case of Ms. Bhati, who's unfortunate demise took place more than seven years after her marriage. What offences would be attracted then? The settled position is that such dowry-related deaths occurring after seven years of marriage are not covered by Section 304B IPC/80 BNS (carrying a maximum punishment of life imprisonment), but rather by Section 498A IPC/85 BNS (carrying a maximum of three years’ imprisonment). This stark difference in the quantum of sentence occurs only because of one factor, i.e., the number of years since marriage.

It is worth mentioning here that the Criminal Law (Second Amendment) Act, 1983, had introduced Section 498A to the IPC as a solution for dowry deaths as well as other instances of cruelty inflicted on married women by their husbands and in-laws. However, it was specifically the inadequacy of this provision owing to which Section 304B was separately introduced in 1986, with a stricter punishment for dowry deaths as defined therein.

While dowry demands may be the most persistent during the first few years of marriage, that does not mean there can never be such scenarios after the first seven years. The fact remains that Section 304B IPC/80 BNS imposes an arbitrary and unreasonable criteria for classifying ‘dowry deaths’. There is no specific rationale for this threshold of seven years. After seven years of marriage, the evidence law presumptions that can reduce the burden of proof for establishing the factum of dowry death (and to thereby ensure ‘easier’ convictions) no longer apply.

The statute ends up creating even procedural distinctions. There is no discernible reason why the Police is required to conduct an inquest into every unnatural death of a married woman occurring within seven years of marriage, but not those after seven years of marriage. This came to fore in Vimala Devi’s case, where the husband is said to have intentionally waited for the seven years’ period to end before getting his wife killed.

In cases such as those of Bhati, some have advocated for the State to invoke murder charges against the accused under Section 302 IPC/103(1) BNS, which is also punishable with death. It should be noted that doing so would mean that the evidence law presumptions applicable to dowry-related provisions would no longer be available, and would instead carry the thresholds (of evidence and causation) as applicable to a regular murder trial.

What is the point of these provisions?

Three pertinent questions arise, in my opinion, from this discussion:

1. Why does Section 304B IPC/80 BNS carry an artificial threshold of criminalising only those deaths which occur within ‘seven’ years of marriage?

2. What is the point of having a separate provision criminalising dowry deaths, when there is no option other than charging the accused with murder if the said death occurs after seven years of marriage?

3. Does Section 498A IPC/85 BNS present an adequate remedy, if the ingredients of Section 302 IPC/103 BNS are not met in such cases?

This predicament raises questions about the very utility of these provisions in combating the menace they had sought to address, especially when they were introduced with a gap of three years (in the years 1983 and 1986, and finally re-promulgated in 2023 in the BNS). Clearly, the solution to this problem does not lie in having an artificial numerical threshold with universal applicability. Rather, it is better to determine such dowry-related harassment as a question of fact having relevance to the death of a wife/bride caused by either suicide or homicide, whether culpable as murder or otherwise. The 243rd Report of the Law Commission of India did not delve into these issues affecting the very structure of the law, keeping itself limited to managing the ‘misuse’ of Section 498A. In this context, even the introduction of the BNS seems to have only continued the extant problems without really reforming the situation at hand.

In this light, the 2022 statistics of the National Crime Records Bureau reveal a gruelling tale too. India loses nearly 18 of its women to ‘dowry deaths’ (as defined in Section 304B IPC/80 BNS) every day, with the actual numbers for subsequent years likely to be higher. Table 2A.2 within these statistics reveals that beyond these numbers, 951 murder cases were additionally motivated by dowry: whether these murders occurred within or after seven years of marriage is anyone’s guess. Further, Table 3A.5, showcasing how the Police dealt with a total of 10261 cases of dowry deaths (those from 2022 as well as those carried forward from previous years), shows how the final reports in 239 of them showed that the case therein was ‘false’. It is difficult to explain how a case involving the passing away of a woman can be called ‘false’ in any official statistics. 253 final reports showcased either a civil dispute or a ‘mistake of fact’. In 359 cases, despite the case being true, evidence therein was insufficient or untraced, or that the Police simply had ‘no clue’ thereof (owing to which a closure report had to be filed).

This alarming picture arises out of a multiplicity of factors, including the difficulty in meeting evidentiary standards (despite the presumptions in evidence law), as well as the inherently poor and overly patriarchal state of Police machinery in our country. Nonetheless, the attention which Section 498A/85-86 BNS gets in our country, for all the right and wrong reasons, is a pale shadow of what Section 304B IPC/80 BNS deserves to get.

The need of the hour is to do away with this arbitrary threshold of ‘seven years’ as a universally tumultuous period for married women – rather, the law should firstly recognise that tumult is a natural part of married women’s lives in India, for reasons connected with dowry or otherwise. To this end, the deeming provision contained in Section 80 BNS should be effectuated by inquiring into acts of cruelty and dowry leading to the death of a woman, whether in the form of suicide or otherwise, regardless of the time-period for which a woman has been married.

Concluding remarks: on society’s attitude towards dowry and crime

The current state of criminal law keeps the meaning of ‘cruelty’ contingent on the social standing and socio-economic background of the parties in a case. While it laudably covers both physical and mental aspects, its effects are dependent on a number of subjective factors. This essentially keeps these provisions strangulated within a realm of subjectivity, which partly explains the brouhaha over the ‘misuse’ of Section 498A.

At the same time, the State’s efforts at safeguarding the sanctity of a marriage, as even the Supreme Court encouraged/endorsed in Preeti Gupta and Rajesh Sharma, place an onerous burden on an affected woman to treat her husband like a 'Pati Parmeshwa'r (as per a Bombay High Court judgment concerning a movie so named). These, instead, ought to be redirected towards equipping the criminal justice system with the requisite means to ensure speedy investigation and trial into such cases: strengthening ‘Crimes Against Women’ cells across States, ensuring access to proper legal aid, and proper awareness campaigns on women’s rights are the barest minimum required herein.

Clarity and a sound legislative policy would be the first step in this rather arduous journey.

Does Criminal Law Matter?

For practitioners in Delhi, courts resume regular functioning after a summer break (the length of which is determining entirely by how high you are up the judicial ladder). The start of term in July is a mix of emotions: happiness at being able to earn some money, sadness at having to work to earn that money, and a sigh of relief to be able to mingle about in familiar surroundings with familiar faces. 

For a small section of the legal community, comprising those concerned with criminal law and procedure, the predominant feeling when term started in July 2024 was neither relief or sadness, but uncertainty. The triumvirate which had defined our criminal law for more than a century — the criminal codes that were neatly combined and kept in the Criminal Manual — were being replaced by new laws. That first month was defined by conversations and discussions about what was good, what was bad, and just how much of it would apply to the cases already in motion. 

By the end of the month, most of us had realised that we weren't being asked to learn a new language in middle age, but essentially revise what we have learnt. Nevertheless, in many cases, people suffered due to the uncertainty that the introduction of new laws brought. In a system where dates of hearing are given at an interval of months and not days, losing out on a date simply because nobody knows what law applies is, well, pathetic. Yet this was the situation faced by many a litigant, and in fact, it continues to be so.

We tend to forget that not everything was smooth sailing in the aftermath. There was a strong protest by a section of the public to two changes (hiking punishments for death by negligence) and the government relented. But even this protest did not care about anything more than the specific, limited, interest, of this interest group. It did not care about the process which would affect all persons potentially ensnared within the criminal law which was being made more muscular and police-friendly. 

Why? Is it because people were not mobilised by political actors around these issues? Is it because it all happened so fast? Or was it because criminal law, honestly, simply, just does not matter within the public consciousness. It matters as gossip, as a lurid news item, as something to create a public-villain figure — at that distance, one level of remove. It does not matter because none of us ever want anything to do with it, and know that if we are sucked into that black hole of police stations, then law is not what will matter most. The overbearing discretionary nature of police powers is one way in which the relevance of criminal law is negated. This point applies to all executive power and is a broader critique of Indian law in general.

There is another critique more specific to criminal law. At its core, criminal law seeks to govern society's moral code through a combination of influences. The simple and obvious truth is that for a large part of the public, a much larger part than which stays in cities surrounded by courts and prisons and police stations, the influences of criminal law are so distant that they may as well be phantasms. The shadow of law barely manages to cover a fraction of the lives of this public, and most interpersonal relations here take place on the basis of a code that operates entirely outside the language of the sanhitas and the codes that came before. The occasions on which the state swoops in to intervene in these interpersonal relations through the criminal law are so few as to render the criminal process almost irrelevant from their calculus of concerns.

The peculiarity of criminal law means that the state knows that the elites won't speak up and the plebians can't. It means that the state can pretty much do as it wants because while criminal law matters for everyone it simultaneously does not matter at all.

Sanjabij Tari and More Guidelines for Speedy Disposal of Cheque Bounce Cases

Since the criminalisation of defaulting of cheque payments in 1988 by amendment of the Negotiable Instruments Act 1881, the Supreme Court has made many attempts to contain and reduce mounting arrears created by such cases. Much of that good-natured effort, it would seem going by statistics, has proved to be in vain. Nevertheless, the Supreme Court cannot be faulted for not making an effort. This indefatigable spirit is perhaps what animated a Division Bench in Sanjabij Tari v. Kishore S. Borcar & Anr. [2025 INSC 1158 (“Sanjabij Tari”)] to pass a fresh series of guidelines / directives, to curb arrears in cheque bounce cases.

Getting the Right Pendency Data

There can be no doubting the general observation that there is a deluge of cheque bounce cases in the system. The Court gives a sense of perspective to this by citing data from the National Judicial Data Grid (NJDG) to state that the figures for Delhi alone are 6.5 Lakh cases as of 01.09.2025 (Para 33).

The tenor of the judgment suggests that it is this sky-high pendency figure which has prompted the Court into giving directions. This shows us the importance of having data at hand. Unfortunately, it also shows us the difficulty in getting the “right” data which we face in India. One would think that using NJDG is as good as it can get to make a statement about caseloads. But as Prashant Reddy and Chitrakshi Jain explain in Tareekh Pe Justice, NJDG cannot be seen as that gold-standard. The website itself accepts this, and carries a disclaimer on the figures being given and requires visitors to cross check the correctness of the information.

The ordinary citizen access for NJDG does not seem to allow searching for statute-specific data and perhaps this is a unique feature that is available to those with special access, so we do not know exactly what the Court saw (it would be nice for everyone to have this access!). It is safe to assume, though, that this data was also not wholly accurate, given how the NJDG data presentation leads to double counting. In its gross figures, the NJDG often adds smaller miscellaneous applications filed within cases, besides the cases themselves, leading to an inflation of the total figures.

Given that pendency is anyway quite high in cheque bounce cases, overestimation here perhaps does not cause huge problems. There may be other situations though where uncritical reliance on NJDG data by a Bench does cause serious trouble. One hopes that day does not come to pass.

What to make of the Guidelines?

Having got the data part out of the way, we can turn our attention to the guidelines / directions itself that are issued in Paragraph 36 of Sanjabij Tari. A part of this is the Court restating / tweaking earlier directives, perhaps with a hope that their enforcement with some renewed vigour will lead to better results. The old is complemented by the new, where emphasis appears to be on pushing the system to (1) secure swift appearance of accused parties and, linked to this, (2) encourage settlements once they have appeared. Within these, I only want to deal with one part viz. the Court’s move to remove application of Section 223 of the Bharatiya Nagarik Suraksha Sanhita 2023 ['BNSS'] from these proceedings [Guideline E].

Section 223 BNSS is a new clause (unlike a lot of the statute) by which a court is now required to give the accused an opportunity to be heard before the court takes cognizance of an offence. The Court has approved the view taken by a Karnataka High Court judgment without adding any explanation. This view is that the Section 223 pre-cognizance notice part will not apply to cheque bounce cases since (i) these case are prosecuted under a special law (Negotiable Instruments Act 1881) and (ii) the offence is quasi-criminal where speed in the trial is key.

It is unfortunate that the Court has dealt with a legal issue in such a summary fashion, because on closer scrutiny neither of these reasons hold water. A special criminal law must prescribe a different method of proceeding for it to override the general procedural law in the BNSS, and there is nothing within Section 142 of the Negotiable Instruments Act 1881 which overrides Section 223 BNSS. This is why the Supreme Court itself has extended Section 223 BNSS to other special criminal laws, such as the Prevention of Money Laundering Act 2002.

Which naturally brings up the other point — you need this pre-cognizance hearing for serious crimes like money laundering, but not ‘quasi-crimes’ like cheque bouncing. The nature of the cheque bouncing crime is an issue which came up in another context recently on the blog — the right of a victim to file an appeal in such cases. While on the hand courts see the cheque bounce crime as 'quasi-criminal', a bench of the Supreme Court in Celestium Financial emphasised on the very clearly 'criminal' nature of this offence to hold that the complainant in such cases was at par with victims in other, traditional, offences. Interestingly, Sanjabij Tari neither cites nor refers to Celestium Financial. 

One may disagree with the approach in Celestium Financial and want to support the view that a cheque bounce case is really 'civil sheep in wolf's clothing'. Unfortunately though, this view is just not supported by law. There is no legally defined category of 'quasi-crime' in Indian law and when the legislature decides to deal with offending conduct by means other than criminal proceedings, it has prescribed punishments other than jail time. No matter how innocuous we make cheque bounce cases appear by likening them to civil disputes, it does not change the reality that a conviction for this offence can result in imprisonment and loss of liberty — the core component of proper crime. 

Conclusion

There is no disputing the fact that cheque bounce cases contribute hugely to arrears within the legal system, making it a natural target for attention by the Supreme Court. Prior to Sanjabij Tari, multiple attempts have been made by issuing directives and guidelines but without lasting or measurable success, and whether this most recent round of efforts will meet a different fate is something only time can tell. 

At the same time though, if we look deeper, perhaps there is a different lesson to be learnt here, especially with the repeated reference by the Supreme Court to the idea that cheque bounce are not real crimes but 'quasi-criminal' in nature. Why not actually take that step and conjure a set of remedies that deliver on this premise by changing the infraction in question from being a crime with possible jail time to something which can be dealt with outside of the proper legal system by means of penalties and other, non-penal, sanctions. The judiciary cannot take that step, as much as it may want to. 

It is soon going to be four decades since Parliament decided to experiment with levying penal sanctions for dishonour of cheques. The experience of these past four decades suggest that whatever benefits the move has brought to financial transactions, it is significantly outweighed by the harm that it has caused to the integrity and stability of the judicial system. Maybe it is time to call it a day on the offence of cheque bouncing altogether? 

On Conspiracy — A Brief Historical Interlude

A long time ago, the Blog began a series (yet incomplete) on the blog discussing the Conspiracy offence in Indian criminal law. A couple of those posts looked at the immediate context leading to the introduction of the substantive offence — the attempt on the Viceroy's life in 1910 — and the 1913 debates in Council where the amendment bill sailed through with considerable ease.

Again, thanks to the digitisation of the Archives, one can get a glimpse of what went on behind the scenes within the machinery of colonial government. The file appears to be titled: "Enquiry into the Possibility of Strengthening the Hands of the Authorities in Dealing with Anarchical Conspiracy: Act amending the Indian Penal Code and the Code of Criminal Procedure in the matter of the law of Conspiracy". Full marks to whoever decided on this for summing up what the conspiracy offence entails.

The initial memo tabling the proposal gives a glimpse of the legal oppression wrought globally by the colonial system, for a key primary reference for introducing a law in India was introduction of a similar law in Egypt in 1910 (later notes on file also cite Ireland as another example). The Egyptian law punished as "criminal association" the act of persons who "associated together" to commit crimes, even where perpetration of crime had "only been contemplated". 

The deficit in Indian law as of 1912 / 1913, was its inability to permit prosecuting persons for only "contemplating" the commission of crime. In other words, what was needed to 'strengthen' the hands of authorities, was giving them the ability to round up people pure and simple, because what one contemplates is rarely written upon one's face. Interestingly, even the Egyptian Law did not go far enough and had to be blended the Egyptian law of 1910 with older English law in his proposed draft, as the latter allowed punishing persons doing legal acts, by illegal means.

The government file is a great place to look for understanding the process of criminalisation. It is not about principles, or justice. It is, especially when a sovereign feels under threat, about what is going to be useful to contain that threat. Exchanging notes on file, the officials from the Home Department and Legislative Department speak of bringing Indian Law in line with English Law, but this is not because they are jurisprudes seeking consistency in the definition of laws. They are wanting this move, at this particular time, because having such a crime "is likely to be useful" to deal with the problem at hand.

How was a conspiracy offence proper going to be useful? Because it rid the need for finding any overt act, allowed enhanced punishments, and perhaps most importantly cheapened the kind of evidence required to prove a case. A.P. Muddiman could not have said it better when he notes: "The vagueness of the English law has its advantages ... In matters of sedition it is rather a question of policy than of jurisprudence what conduct should be punishable, and some degree of vagueness is a necessary condition of efficiency." [Pp. 2-3, Note dated 08.01.1913]. His superior in the Department, W.H. Vincent, noted that merely adding this law is not going to be enough without sufficient procedural powers to take meaningful advantage of this vagueness that the law would bring [Pp. 3-5, Note dated 14.01.1913]. 

Usefulness of having a dedicated conspiracy law is made plain by the Home Department. As RH Craddock records: "The proposed additions to the law will not only embrace anarchist plots to murder, but conspiracies to commit crimes of all kinds, including criminal intimidation, mischief, wrongful restraint, etc., which are the usual accompaniments of a boycott movement." [Pp. 6-7, Note dated 17.01.1913]. He really lets the cat out of the bag, in a sense. Once you are willing to take a step back, it becomes obvious that rarely will an actual revolutionary conspiracy be prosecuted without overt acts being visible. What you really need this new offence for, is to pre-emptively quell even the far lesser obvious kinds of politically undesirable acts challenging British rule.  

As the proposed law goes outside the four-walls of government, the conversation loses some of its candour. We see this best expressed in the Draft Statement of Objects and Reasons that is being discussed on file. This statement is a brief description of the reasons that prompted government to propose a legislation, and will be consumed by the general public. Naturally, such a document cannot say that we need vague criminal laws to efficiently stem the rising anti-British feeling. Instead, the statement mainly gives jurisprudential reasons (there is a gap in Indian law which must be filled) and addresses the political element by simply stating that "The existing law has been found inadequate to deal with modern conditions in India ..." [Pp. 9-12]. 

This sanitising criminal law from the contest of politics is a sleight of hand which only history and the archives can help uncover, and give us a truer picture. It was not merely to fill some gap in law that Section 120-A was added to the Indian Penal Code in 1913, and has been retained in the Bharatiya Nyaya Sanhita a century later. It is because the offence of conspiracy contains the necessary vagueness allowing it to be used by the government in situations which it deems are undesirable. Prosecuting such conspiracies has never been about law, but have always been about pursuing government policies, and no amount of clothing in statutory or legalistic verbiage can somehow make them appear apolitical. 

Guest Post: Extending the Conversation on Law and a 'Legal Science' in India

(This is a guest post by Rajyavardhan Singh)

The editor of this Blog recently sketched a provocative thought. His argument, stated simply, was that Indian law is a landscape without a lodestar, lacking any “legal science” that could guide coherence in lawmaking or adjudication. What we confront instead is a dissonant landscape, where legislatures, courts, and enforcement agencies operate at cross-purposes. This is so, for their approach seems to be shaped more by immediate contingencies rather than a settled conception of law.

By tracing the Constitution back through colonial codification, company rule, and pre-colonial inheritances, the earlier post showed how Indian law is a layered historical construct, but one that lacks an organising framework to make its elements cohere. In this piece, I seek to continue that conversation by asking what it might mean to imagine coherence in Indian law today. To pose the question in this way is to recognise that law and order without coherence risks devolving into administration alone, where outcomes are dictated more by contingency and power rather than by reason.

Towards a Grammar of Coherence

To imagine coherence in Indian law is not to posit a single, all-encompassing theory that tidily resolves every institutional conflict. Rather, it is to ask whether there exist conceptual anchors that can discipline the otherwise unstructured interplay between legislation, adjudication, and enforcement. What such anchors might look like, however, is a contested question. Comparative experience offers some possibilities — Hart and Sacks’ Legal Process school in the United States, or Dubber’s project of a New Legal Science, both sought to reconstruct law on the basis of shared principles that could orient lawmaking and adjudication without erasing pluralism.

The Indian context, however, is complicated by the absence of any comparable intellectual tradition. As was also argued in the earlier post, our legal system is an inherited mosaic of colonial codification, postcolonial constitutionalism, and precolonial practices — none of which has been systematically interrogated for the principles that might underwrite a coherent order.

This absence, I suggest, has consequences. Take, for example, the Courts oscillating between the established maxim that bail is the rule and the contrary practice that treats custody as the default; with little more than case-specific discretion to explain the difference. What results, as Gautam Bhatia has recently argued, is a kind of judicial double-standard. For when a fact is potentially favourable to the individual (say, in a UAPA matter), courts invoke the Watali judgement to refuse engagement, shutting their eyes to material that might support bail. But when the fact is adverse, the same courts are happy to wade deep into the evidence, even drawing inferences that go beyond the record.

Legislatures, meanwhile, frequently add new offences or enhance punishments in response to episodic crises (as with the post-Nirbhaya amendments to the erstwhile Indian Penal Code in 2013, their further expansion after Unnao and Kathua in 2018, or, most recently, the Aparajita Bill introduced in West Bengal following RG Kar) without any reference to a settled framework of penal theory. As scholars of penal populism observe, they embody a politics of symbolic criminalisation, where the severity of punishment substitutes for any substantive justification of criminal law.

Enforcement agencies then, for their part, exploit ambiguities in the law by exercising vast discretion in arrest, investigation, and prosecution. Despite repeated judicial reminders that arrest must remain an exception, police and specialised agencies continue to use it as the default. Besides, investigations themselves have long been marred by political misuse and lack of independence, while prosecutions operate less as trials of guilt than as punishment by process. The resulting system, hence, is one where liberty turns on happenstance rather than principle.

Thus, coherence (in this sense) would not mean the elimination of judicial discretion or the homogenisation of legislative policy. It would mean that discretion and policy are exercised within an intelligible grammar — such that it allows citizens, institutions, and scholars alike to evaluate whether state action is consistent with the law’s core commitments. What, however, would such a grammar consist of?

At its simplest, such a grammar would rest on a set of conceptual commitments that organise the otherwise scattered practices of lawmaking, judging, and enforcement. These commitments are not difficult to name. To name a few — Culpability as a precondition for punishment. Proportionality as the measure of how much punishment is justified. Due process as the assurance that the road to punishment is neither arbitrary nor degrading.

None of these, I submit, are alien imports. They exist in fragments across case law and statutory text. Yet, because they are rarely articulated as organising principles, they fail to discipline practice. Proportionality, for instance, is invoked in sentencing but abandoned when legislatures escalate punishments to death or life terms. Due process, again, is hailed in theory but hollowed out in practice by a set of procedural checklists. Seen thus, this grammar is not just a code to be written once and for all. It is a discipline of justification.

It demands that legislatures explain why a new offence requires criminalisation rather than civil liability (as with the criminalisation of Triple Talaq), that courts justify why liberty is denied in one case but granted in another (for instance, Gulfisha Fatima’s treatment in the Delhi riots bail hearings, contrasted with the release of similarly placed accused), that enforcement agencies demonstrate why an arrest was necessary beyond the fact that it was possible and used as a reflexive tool of control.

A coherent grammar, however, would not guarantee agreement — scholars, judges, and lawmakers will still disagree about the scope of culpability or the limits of proportionality. But it would guarantee that such disagreement occurs within an intelligible frame, one that allows citizens to assess state action against principles rather than personalities, and to demand fidelity to the law’s core commitments rather than its most immediate conveniences. Without such a grammar, Indian law risks remaining what it too often is today — a set of rules and practices that work until they do not, lacking any deeper account of why they should endure.

To Conclude

If coherence in Indian law, indeed, is to be imagined as a grammar, then the challenge is in cultivating a culture that insists on returning to them. The law does not fail for a want of lofty ideals. It falters when those ideals are not treated as binding coordinates. To that extent, coherence is less a destination than a discipline of thought, one that requires institutions to account for themselves in ways that transcend immediate expediency. For too long, the discourse has been content with doctrinal bricolage — stitching together precedent here, statutory text there, and leaving the seams to show. What a grammar of coherence demands, instead, is sustained reflection on the why of law, and not merely the what.

Now whether Indian law can rise to this challenge remains open. But to ask the question is, at the very least, to resist the comfort of dissonance.