Guest Post: Illegally arrested, Legally Detained - The Paradox of Habeas Corpus

(This is a guest post by Bipul Kumar and Vaneet Kumar)

Recently, the petitioner, an accused under murder case, in Neeraj v. State of UP (“Neeraj”), filed a writ of habeas corpus in Allahabad High Court arguing that his remand is illegal since he was not informed of the grounds of his arrest. However, Neeraj vide its order dated 27 May 2026, rejected the petition by holding that the post-cognizance remand cannot be vitiated because of illegal arrest (para 88). Moreover, it also declared Mihir Rajesh Shah v. State of Maharashtra line of judgments that declared arrest and subsequent remand illegal on non-communication of grounds as per incuriam, for conflicting with Kanu Sanyal v. DM, Darjeeling line of judgments (paras 87). The Court, while reaching these conclusions, also introduced two limitation ‘points’ on the availability of habeas corpus petition against arrest and remand: (a) taking of cognizance, and (b) the rejection of bail application. This article upon critically engaging with Neeraj argues that while the High Court correctly limits the remedy to pre-cognizance detention, its holding of per incuriam and its second limitation on habeas corpus are legally unsustainable.

Is Mihir Rajesh Shah Really Per Incuriam?

Neeraj concluded that Mihir Rajesh Shah and like “are not binding precedents and hit by the principles of stare decisis” (that is, per incuriam), without providing any reason provided alongside this conclusion (paras 86-87). This conclusion invites two lines of enquiry: (a) Whether the High Court can declare a precedent as per incuriam? And (b) Whether the conclusion reached by Neeraj is correct, irrespective of its competence? On the first line of enquiry, the answer can be found in Sundeep Kumar Bafna v. State of Maharashtra, where the Supreme Court held that the all courts, including High Courts, can declare a decision to be per incuriam; however, such declaration should be made with extreme caution and circumspection (para 16). Therefore, Allahabad High Court, via Neeraj, is competent to declare Mihir Rajesh Shah per incuriam.

As per Sundeep Kumar Bafna, “a decision or judgment can also be per incuriam if it is not possible to reconcile its ratio with that of a previously pronounced judgment” (para 15). There should be a “glaring case of obtrusive omission”, as per Bajaj Alliance General Insurance Company v. Rambha Devi. In simple terms, if the overlooked binding precedent alters the conclusion of the judgment, such judgment is per incuriam. It is the assertion of Neeraj that Mihir Rajesh Shah line of decisions “do not appear to have considered the aforesaid earlier pronouncements of the Apex Court” (that is, Kanu Sanyal line of judgments) (para 77). However, Neeraj proceed to declare the former line of judgments as per incuriam without so much as asserting the irreconcilability of the two lines of judgments. This renders its conclusion doubly flawed, falling short of both the legal test enunciated in Bajaj Alliance and the salutary caution prescribed in Sundeep Kumar Bafna. Unlike Neeraj, it is our assertion that these two lines of judgments are not competing at all and, even if deemed to be so, are readily reconcilable. To enunciate, while the Mihir Rajesh Shah line of judgments deal with the post-arrest remand orders issued under section 187 of the Bharatiya Nagarik Suraksha Sanhita (“BNSS”) (or pari materia section 167 of the Code of Criminal Procedure, 1973), the Kanu Sanyal line of judgments deal with the preventive detention or remand orders issued under sections other than section 187 of the BNSS. Since the basis of remand is different under the two lines of judgments, they were never at loggerheads.

First Limitation ‘Point’

Although there is no limitation ‘period’ for filing a writ petition for enforcing the fundamental rights, especially under Article 22(1) of the Constitution (it mandates communication of grounds of arrest), Neeraj introduces limitation ‘points’ for this remedy. We shall discuss these two ‘points’ at length in this and following section. First, as mentioned above, Neeraj postulates that cognizance is a cut-off point for filing a habeas corpus writ under Mihir Rajesh Shah regime. This conclusion, in our opinion, is but obvious, and flows from multiple reasons which we proceed to address in turn. But before turning to these reasons, it is important to distinguish between the pre-cognizance and post-cognizance remand orders. While the remand orders under section 187 of the BNSS are pre-cognizance orders, the remand orders under sections 232 and 346 of the BNSS (or pari materia sections 209 and 309 of the Code of Criminal Procedure, 1973, respectively) are made after cognizance and hence, called post-cognizance orders. Neeraj postulate the reason as follows: “The order taking cognizance is a judicial order and stands on a higher footing than an order of remand passed under [section 187 of the BNSS], which is merely a pre-investigation judicial order” (para 75).

This quote provides us with first reason that the post-cognizance remand is based on separate considerations (that is, investigative material or chargesheet) that pre-cognizance remand (that is, pre-investigation material), making it legally untenable to adjudge the legality of the former detention on the basis of legality of the latter detention. The second reason, albeit absent from Neeraj, is straightforward: while pre-cognizance remand cannot be made but for arrest, the post-cognizance remand can be made independent of arrest. Therefore, the legality of arrest cannot impose as a litmus test for post-cognizance test. Thirdly, doing it otherwise would have violated the well-established and accepted ‘rule of return’. Fourthly, the factual matrix underlying Mihir Rajesh Shah regime concerned itself exclusively with pre-cognizance remand orders. The extension of that regime to post-cognizance remand orders would therefore be contrary to the doctrine of stare decisis, which demands that precedents be applied only to cases arising from analogous facts and circumstances. Therefore, cognisance as the ‘point’ of limitation for writ of habeas corpus can be founded well-established in law.

Before moving to the second ‘point’, it is pertinent to discuss the murkiness that lies beneath the superficially clear limitation ‘point’ of cognizance. Suppose X is illegally arrested and remanded under section 187 of the BNSS. The police subsequently file a charge-sheet under Section 193 of the BNSS; however, as is invariably the case in practice, there ensues a period of delay before the Court takes cognizance. Now, the writ of habeas corpus filed by X is heard at this stage. Should X be released? Now, as per the discussion up till now, the answer might appear yes. However, the answer is not as straightforward as it seems. Surprisingly and unfortunately, there is no provision that governs the detention during the period between charge-sheet and cognizance because while section 187 of the BNSS, as per Sanjay Dutt v. State, ceases to operate upon filing of chargesheet, sections 232 and 346 of the BNSS, as per State of UP v. Laxmi Brahman, cannot be applied until cognizance is taken.

Despite this peculiar situation, Neeraj, without providing any reason, declares that remand under section 187 of the BNSS “remains operative only till cognizance of the offence” (para 74). This inference despite treating X’s remand and pre-cognizance alike is hit by the doctrine of per incuriam as it ignores Sanjay Dutt. Contrarily, the Delhi High Court in Nitin Nagpal v. State declared X’s remand patently illegal (para 18). Therefore, the X’s writ of habeas corpus under Mihir Rajesh Shah regime will succeed in either scenario; however, X does not even need to allege ignorance of grounds, if Nitin Nagpal is considered. Conversely, the Allahabad High Court in Sudhakar Rastogi v. State of UP, by ruling that X’s remand is governed by section 346 of the BNSS, ends up classifying it as post-cognizance remand. This classification renders the X’s writ infructuous due to first limitation ‘point’ of Neeraj. Therefore, writ filed by an accused, who is in position similar to that of X, will receive different answers by different benches or Courts, even though we believe that Delhi High Court’s opinion is correct because the Courts, as per Natabar Parida v. State Of Orissa, do not have inherent power to detain.

Second Limitation ‘Point’

On the second limitation ‘point’, Neeraj reasons that subsequent bench should not entertain habeas corpus writ if the High Court or Supreme Court has rejected the bail application because doing otherwise “will amount to appeal/review of the judgment passed by the bench deciding bail application” (para 88(ii)). Even if this reasoning seems palatable, we contend that it is incorrect representation of law because while bail application contends that the accused satisfies the triple test and poses no risk of flight, tampering, or reoffending, a writ of habeas corpus challenges the legal validity of the detention at its very root. Furthermore, the distinction also lies in the judicial approach: while bail is discretionary and admits of no straitjacket formula, the writ of habeas corpus is, in a sense, compulsory. This may appear an extreme proposition; however, what is meant is simply that under the Mihir Rajesh Shah regime, the writ operates on a near-mechanical basis — failure to communicate the grounds of arrest must necessarily result in release from pre-cognizance custody. If the grounds, purposes, and operation of two are inherently different, Neeraj ought not to consider one as an appeal or review of another, simply because both ends up in same outcome (that is, liberty of the accused).

Rather, the Courts have even entertained both the applications simultaneously and one after another. For example, Arvind Kejriwal (in CBI matter) filed both writ of habeas corpus and bail application, and the Delhi High Court vide two separate orders decided these two petitions (see Arvind Kejriwal v. CBI (I) and Arvind Kejriwal v. CBI (II)). For another example, Kejriwal (in ED matter) filed a bail application in the Special Court, after the Supreme Court in Arvind Kejriwal v. ED reserved the matter involving writ of habeas corpus. While it is conceded that neither of the examples exhibit a bail application filed after habeas corpus writ; however, these examples exemplify that the courts have treated both the petitions differently. Critics, like Neeraj, might resort to Eastern Coalfields Ltd. v. Dugal Kumar to argue that the accused should exercise writ at the earliest reasonable opportunity. Though we concede that this allows the accused to avail relief at the earliest opportunity, this in itself cannot act as a bar for a writ petition filed after the bail application is rejected, especially when first limitation ‘point’ exists.

Conclusion

Upon culminating all the arguments, it can be concluded that the Neeraj not only wrongfully declared the Mihir Rajesh Shah line of judgments as per incuriam but also provided bail rejection as a legally untenable limitation ‘point’ for filing writ petition of habeas corpus. Upon removing these observations from Neeraj, the correct legal position can be stated as follows: the accused can file a writ petition of habeas corpus for illegal arrest until the court takes cognizance and detains the accused under post-cognizance remand orders.

Furthermore, it is important to note that this analysis of Allahabad High Court’s judgment comes at a crucial time, especially when the Supreme Court in State of Meghalaya v. Sonam Raghuvanshi on 23 July 2026 created a distinction between non-service of grounds and insufficiency of reasons stated therein. In our opinion, it effectively limited the scope of Mihir Rajesh Shah to form (that is, service of grounds) and hinged the vitiation of arrest on substance of grounds (that is, prejudice caused due to insufficiency of reasons). This might give rise to mechanical act of furnishing grounds of arrest by the police, leaving the accused to prove the prejudice before the courts — ultimately, becoming a brewing ground for upcoming jurisprudence on the grounds of arrest and its intersection with Mihir Rajesh Shah. Interestingly, the apex Court while hearing this very matter, like Neeraj, orally mulled referring Mihir Rajesh Shah to a larger bench for being per incuriam. However, the Court, without delving into reasons, relied on Shah, leaving the question of per incuriam for a future occasion. This constant attack on Mihir Rajesh Shah also exposes the possibility of potential judicial conflict. Hence, this piece should not be confined to the vision of Neeraj, but rather be considered against the ever-expanding jurisprudence on ‘grounds of arrest’.

Lastly, as a parting note, it is important to recollect that, as per Vihaan Kumar v. State of Haryana (one of the judgments in Mihir Rajesh Shah regime), the illegal arrest violates fundamental rights enshrined under both Article 21 and 22. The gravity of violation is well reflected by Supreme Court’s statement: “Deprivation of liberty even for a single day is one day too many” in Arnab Manoranjan Goswami v. The State of Maharashtra. Considering the gravity of violation, we opine that the shift to post-cognizance remand should not leave the arrestee remediless; rather, the accused should be entitled to compensation for illegal detention. The Supreme Court in Bhim Singh v. State of J & K, a case dealing with illegal detention, provided compensation even though illegal detention has ended. Therefore, the Constitutional Courts like Bhim Singh should push for compensation, in addition to release, as a relief under the writ of habeas corpus to deter the state machinery from violating the accused’s fundamental rights.

On the Bombay High Court granting bail to the Kochhars

The Bombay High Court has re-affirmed the most elementary legal position — the mere power to do something does not act as a justification for doing it. That its order of 09.01.2023 directing the release of Deepak Kochhar and Chanda Kochhar by granting them interim bail on this very premise has generated such a buzz is not indicative of any pathbreaking findings made, but on how distant even these elementary legal principles appear when it comes to our criminal process that every such order warrants celebrations within the community of defence lawyers akin to India winning a cricket match. 

Chanda and Deepak Kochhar were named as accused persons by the Central Bureau of Investigation (CBI) in an FIR lodged in 2018. They were not arrested then, and appeared numerous times before the agency (as well as other agencies investigating them) [a useful list of dates begins at page 8 of the order]. On 23.12.2022, they were again present at the office of the CBI for questioning, only this time they were both arrested. The grounds disclosed for their arrest in the Arrest Memo were that (i) they were named as accused persons, and (ii) they were not cooperating with the ongoing probe. 

Without seeming to engage with the sufficiency of these reasons for arrest, the trial court first remanded them to police custody and then to judicial custody. On 27.12.2022, it appears that the FIR, the arrests, remand orders were all challenged before the Bombay High Court. As we already know, going by the premise that the mere power to do something—arrest—does not act as a justification for it, the High Court held that the justifications for the Kochhars' arrests required scrutiny, and when scrutinised these came up horribly short of passing muster under the law as present under Section 41 of the Criminal Procedure Code 1973 and as had been clarified by the Supreme Court most recently in Arnesh Kumar (2014) and Satender Antil (2022). 

It is likely that the High Court's specific findings on the insufficiency of justifications behind the arrests will garner most attention. Being named in a case in and of itself was never grounds for arrest, even more so where four years had lapsed between the lodging of a case and the arrest. And in respect of the 'non-cooperation' — an idea incidentally also critical to the Satendar Antil judgment — the High Court agreed with the view that cooperation could not be a substitute for 'confession', such that when an accused did not sing the tune of an agency she would be labelled as being non-cooperative. On closer scrutiny, arguably more important conclusions become visible that are not headlined in the same manner, as well as some cracks in the reasoning and the obvious limits to this order. 

First, to the conclusions that hopefully other courts can build upon to restore a wall to protect personal liberty. The High Court order does not only advance the basic premise that there must be sufficient justifications for an arrest, but it also indirectly indicates that these justifications must be captured in a public facing document in the form of the arrest memo, and not hidden away in the case diaries where an accused will not have access to them. This is critical, because it is only the arrest memo that is shared with the family upon arrest and, as extracted in the order, the reasons for arrest that are disclosed therein are often unintelligible or party excuses. The High Court's order stands for the position that given the gravity of the measure in question, full and adequate reasons ought to be recorded and disclosed for the process to be in consonance with the letter and spirit of the law. 

Besides the casting of clearer duties on police, the order also restates a requirement for trial courts to engage with the sufficiency of an arrest when a remand to custody beyond 24 hours is sought. In this case, the High Court could not discern from the remand orders whether the trial court had even considered the issue of the legality of arrest, and it severely reprimanded the trial court for failing to do so. It is not the first order to do so, and the High Court invokes earlier cases to support its view, but the importance of our trial courts discharging their role as the first bastions for protecting our personal liberty — as the late K.G. Kannabiran had said — cannot be overstated. Not only from a broader, socio-political standpoint, but also from a legal remedies standpoint. Where an illegal arrest is sanctified by a remand to custody, the writ of habeas corpus against such an arrest is practically extinguished. 

It is possibly why the Kochhars also did not file a writ only challenging their arrests, but filed an omnibus writ petition challenging the very registration of the case and all executive / judicial action emanating therefrom (a strategy that had also been taken when Arnab Goswami was arrested in an old case, where incidentally the senior counsel for Ms Chanda Kochhar was on the opposite side). Moving an interim relief in such a petition, as opposed to a regular habeas or even seeking regular bail before the trial court, was a strategic call that clearly paid great dividend.   

Then there are the tiny cracks. While the High Court comes down heavily on agencies treating cooperation synonymous with confessions, that happens by the High Court accepting the argument that this was the only possible cooperation left. The High Court did not engage with the idea of cooperation on its own terms. If it had done so, it would have had to square up to the language of Section 41 of the Cr.P.C. which justifies an arrest in cases where it is necessary for 'proper investigation of the offence'. What is the scope and ambit of this 'proper investigation' thus remains unclear, even as it remains clear that it certainly is not the practice of taking custody in the hope of extracting confessions. One imagines that this order will not be the slaying of the non-cooperation dragon after all 

Lastly, the limitations. The offences alleged against the Kochhars were not of the very serious variety — a maximum seven years punishment was prescribed. Within the statutory universe of Section 41, and cases that have interpreted the power to arrest such as Arnesh Kumar, there is a different treatment for these offences. The enumerated list of justifications provided for arrest in such cases does not cover the more serious class of offences punishable with more than seven years imprisonment. One can only hope that the findings on sufficiency of recorded justifications and engagement with them by the trial courts at time of remand is not sought to be limited by future courts using this classification. 

To sum up, The Kochhars were arrested four years after a case was filed against them without any real basis recorded in writing. They could afford to move the High Court during vacations, get an urgent listing and hearing, and get the High Court to remind everyone that personal liberty ought not to be trifled with. What they did is not a route that many similarly placed accused persons can hope to follow, but hopefully the fruits of their labour can be enjoyed and built upon to narrow the gap between the promise of Article 21 and the practices of our law enforcement agencies.       

Guiding Police Discretion: Police Circulars and The Limits of Top-Down Reform

Let's do a thought experiment. Think of the criminal justice system — or criminal process as I like to call it — as a building. Now, the basic architecture of this building is fairly well-known to us, and exists in the form of three statutes: the Indian Penal Code 1860, Indian Evidence Act 1872, and the Criminal Procedure Code 1973. The interplay and intersections between these statutes, and the many judicial decisions that have interpreted their text, gives us a four-walled structure. Add to these the Police Act 1861, and the different state laws on policing, and now you get a real building with doors, floors, and windows. 

But what's inside these structures? To get a glimpse of that, we need to think about the various actors who are at the heart of the process. Go to any criminal court and you will find a judge, a prosecutor, a defendant, her lawyer, probably a victim, and the police. The incentives guiding these actors are what mark the ebbs and flows of the criminal process. But if I had to pick one actor out of these as being the most critical to the process, it has to be the Police.   

Police officers are at the heart of our criminal process. Like many other jurisdictions, police play a critical role in funnelling disputes, as suggested below:

Total Incidents of Crime = 100% → Total Incidents Reported / Noted by Police = 70%   → Total Cases Investigated = 50% → Total Cases Prosecuted in Court = 30%      

Let's take a moment to appreciate the dimensions of this suggestion. Even though we might forget it sometimes, at some level we all know that every incident of crime does not end up in court. What we often don't ask, is whether this selection is by design or by accident: Is it purported incompetence which leads to cases not being pursued, or the outcome of the police being aware of the limitations of the criminal process? 

The honest answer is that it is a little bit of both. But what matters for us is the fact that there is a very real design element to the funnelling process. This element tends to remain hidden if we only look at the statutes, or even at the judicial decisions. It only comes to light if we look at the many rules that govern the police as an organisation. The problem is that a large part of these rules and circulars are hidden in a more literal sense. Despite being public documents, they are often not published in a Gazette, or are not readily available in a bookshop or on the internet. 

In Delhi (where I mostly operate) we are lucky, insofar as the Punjab Police Rules of 1934 are not very difficult to find in court libraries or bookshops (and online). These Rules tell us the manner in which the Police are supposed to exercise the powers they are granted under law. But again, looking at the Rules to figure out how the police work means you are only looking at half the picture, and are bound to end up coming to incorrect conclusions. This is because there are a host of internal circulars that are issued under the offices of senior police officers to control different aspects of police powers and guide their exercise of discretion. 

Thanks to the Right to Information Act, I was able to get my hands on some of these police circulars that are relevant for the investigative process. These circulars, which are now available online, span a decade (from 2008 to 2018) and shed light on many different aspects. Some musings are below:

• The circulars show that the Delhi Police (and the Delhi High Court) was quite concerned with arrests in matrimonial cases, and created internal mechanisms that were designed to limit the possibility of arrest in such cases. For instance, by requiring that prior permission for arrest be sought from Deputy Commissioner of Police level officers, even though this is not required in the Cr.P.C. itself  (Circular No. 330 of 2008, Circular No. 48 of 2011). This is very interesting, especially if we consider that the Supreme Court itself proceeded to adopt a policy of curbing the powers of arrest generally in a matrimonial case as well (Arnesh Kumar v State of Bihar).
• On the subject of the Arnesh Kumar judgment, one finds that this is also the basis of a circular (Circular No. 38 of 2014). Importantly, while this judgment curbing police powers came out in July 2014, the corresponding Circular is only issued on 17 November 2014. This gap, between the rendering of a judgment and issuing a corresponding Circular, helps to grasp the idea of a lag in our systems of communication between the different branches of state. Perhaps a study can be constructed which looks at such internal circulars to determine the many factors which affect the time-lag, and thus help us improve our systems of governance (see here, for a related study in context of Section 66-A of the IT Act). 
• The General Guidelines for investigation (Circular 15 of 2015) shows that the police agencies are clearly receptive of the objections being faced in trial courts, and are using this feedback to issue appropriate instructions to officers in a bid to prevent future objections of the same kind from arising. 

There will be many more of these circulars guiding different aspects of the swathes of discretionary power that police officers are vested with. At the same time, this multiplicity of internal circulars — especially those affecting our right to personal liberty — force me to ask: what is the point? 

For instance, consider Circular No. 16 of 2017, which requires that the "calling of any person to the police station should be properly documented." This Circular was issued pursuant to the orders of the Delhi High Court, and is effectively guiding the statutory powers of police to examine suspects and witnesses, and arrest persons. So, we have all three sources of law guiding police conduct here: rules, statutes, and judicial decisions. All of them are curbing police powers that interfere with personal liberty. But how often are these rules, laws, and decisions not followed? Let's just say the probability is closer to 1 and not 0. And what are the odds of officers being punished for misconduct? Exactly the opposite.  

Should we think of our system of laws as a pointless paper tiger then? I wouldn't go so far. Rather, I'd say that these Circulars offer more proof for the view that top-down reform efforts, that focus on the Supreme Court and hope to make practices change on the ground, are securing very limited impact. Without a concerted reform effort from the bottom-up, by focusing on police stations and trial courts, the core values of liberty and freedom that we want our system of laws to protect will continue to remain illusory.

Kanojia Gets Bail: A Strange Kind of Justice

Prashant Kanojia was arrested for posting a tweet about a political personage. He did not abuse, use names, or anything of that sort. He posted a link to another story that was carried in the news about the claims of another person, about having an affair with the said personage. People were not up in arms at the suggestion. There was no public order disturbance. And yet, Prashant Kanojia was taken into custody by the police on trumped up charges, contrary to law, some rules, and judicial guidelines as well (along with him, other persons were also arrested in connection with the same incident). Why? Because the personage also happens to be the one who, indirectly, controls the police. 

For this unbelievably horrific act, the police not only kept him in custody for the 24 hours that the Constitution allows. But subsequently, he was taken before the magistrate on Sunday and remanded to eleven days in judicial custody. Not once did the magistrate seem to question the irregularities that seemed to accompany the arrest. For instance, the police invoking Sections 66 and / or 67 of the Information Technology Act 2000 when neither provision had anything do with the case. But I guess it was expected. After all, the personage also happens to be someone who, indirectly, can derail the career of the Magistrate. 

Kanojia's wife is then forced to take this issue all the way up to the Supreme Court — lucky for her that she stays in New Delhi and does not have to travel too far. The country's highest court is required to intervene in a simple case of bail? And once the Court does intervene and does its job by ordering that Kanojia be released on bail because it is patently ridiculous to keep a man locked up for tweeting a news piece, we are supposed to be happy and be thanking our stars that the Supreme Court exists?  

This is not the time for being happy, but for being outraged at just how pathetic the system is. And how pathetic we are for doing ever so little to change it.

Kanojia's arrest and the Supreme Court's intervention are very painful reminders of just how deep and pervasive the rot runs in our legal system. Where the executive controls the police, what else do you expect will happen if someone decides to challenge a Chief Minister? Of course it shouldn't happen, but let's not kid ourselves. And since the legislature and executive are joined at the hip in our system of government, no wonder we are happy to retain / enact broadly worded criminal laws that confer police with wide powers of arrest.

Technically, an independent judiciary should be serving as the corrective here. Magistrates can, and should, be questioning the arrest and why it was necessary. But that is a rare occurrence generally, let alone in a case involving political personages. Why? Because, again, political interference is a thing. Magistrates are working on a long-term career path that can be derailed by an adverse comment in their service record, or the evil-eye of an influential person. Can we really expect each person to be fearless and not care about making sure of food in the house? Of course not. 

Which then brings me to the Supreme Court itself. Time and again, the Court is presented with cases like Kanojia's, some far worse. This was made very clear when PUCL moved the Supreme Court in January 2019 citing cases of people being arrested under the now-defunct Section 66-A of the IT Act — something that seems to have played a role even in Kanojia's case. This has led the Court to do a number of things: issue a powerful rebuke to authorities, pass guidelines regulating powers of arrest, and even go ahead and push through new legislation for insulating police from political interference. Things brings a good measure of praise for the Court, but manages to secure precious little in terms of actual change in police behaviour. 

Again, to expect any other outcome is foolish, as the Court does not control the police. All it can do is talk by passing orders, and threaten states with contempt in case of non-compliance. The hollowness of that threat is evident if anyone takes stock of how many states continue in contempt of the court's orders of police reforms passed way back in 2006. 

What is my point here? That the outrage which we experience while hearing news about politically motivated policing and prosecution, such as the one in Kanojia's case, should not fizzle out because the Supreme Court agreed to intervene and grant bail. Keep rewinding the tape and you will find that many legislators have made promises about breaking this unholy nexus between the government and police, right from the Constituent Assembly, and even today we have not managed to make any real headway in achieving this goal. The Supreme Court is not an ever-vigilant sentinel. It cannot step in to save every Kanojia who is arrested for a tweet, or a Pehlu Khan who is lynched for practising his trade, or the many young tribal men picked up by virtue of who they are and not what they've done. 

It is time that we, as participants in a democratic setup, up the ante and continually push for reforms in the criminal justice system. If governments can listen to business lobbies of a few hundreds and get working on tax reforms, they will surely have to listen to the many thousands whose lives are made worse by a criminal system which is only "just" in name and very rarely in action.

Update (12.06.2019): Further arrests have taken place in the case after the Supreme Court passed its order. Persons arrested in Karnataka on similar allegations (relating to a different political personage) have not been released on bail yet. And we are still a nation that has a Constitution.

Hiding in Plain Sight: The Relevance of Character Evidence in Indian Criminal Law

Can our past behaviour serve as a guide for predicting future conduct? The intuitive answer, is yes. Cutting-edge behavioural science research adds substance to this intuition and demonstrates that there are identifiable links between our past and future choices, helping policymakers to design better systems for ease of living (and companies to make more money through better advertising). This notion of prediction is at work almost everywhere, but not in the law of crimes and evidence in most Common Law countries. Instead, here we find that the law has strict rules that greatly limit the ability to rely on this link for prosecuting crime. Indian law is no exception: Sections 14, 15, 52-55 of the Indian Evidence Act 1872 [IEA] show that evidence of a defendant’s bad character is not generally relevant to prove that she was guilty of the offence in question. 

The Shift from Character to Responsibility in Criminal Law

In Making the Modern Criminal Law, Professor Farmer argues it was not always like this, and until the mid-19th Century it was routine for evidence of bad character to be the basis for court decisions. Things took a turn between 1870-1920 due to a multitude of factors, and the relevance of character was slowly replaced by a focus on responsibility in criminal trials. Two factors were arguably more important than others. 

First, responsibility came to fore as it became impossible for juries to arrive at judgments of character with any certainty for it to be used as a basis to render judgments. A fixed local context in a village meant everyone was well-known and judgments of character had resonance. But industrialisation gradually uprooted this idea of a fixed local context, and meant that the law had to look away from character for criminal trials to keep running. 

Second, the question "Why criminal law?" began getting different answers in this period. Till now, criminal law was seen as the tool to help ruling classes exclude certain undesirable types of persons from regular society. In this period between 1870-1920, this determinism in criminal types gradually lost out (as did the notion of ruling classes) and new opportunities were seen in the signalling effect of criminal law. Subjecting persons to censure and sanctions for what they intentionally or knowingly did, as opposed to who they were or what they did accidentally, was seen as a better way to send the message to persons on what kind of conduct was acceptable to maintain a desirable civic order.

Sir James Fitzjames Stephen was a prominent part of the intellectual movement that supported these changes to the Common Law of crimes. He was also the main draftsman of the Indian Evidence Act of 1872, which explains why the statute adopted some of these ideas rather early. In the years since, Indian law has not moved away from these precepts. Much like the rest of the Common Law world, evidence of bad character continues to be generally irrelevant towards proving guilt or innocence in India, but it can become relevant in certain exceptional cases specified by statutes. Today it is difficult to find many people disagreeing with the idea that nobody should be sent to prison only based on the bad reputation they harbour, for what if it might happen to you? 

Note, that the above discussion is about proving guilt or innocence through character evidence. If we find proof of guilt on completing trial, character evidence comes right back into the spotlight when courts are deciding punishment. Prior convictions can mean higher punishments, and having none can entitle a defendant to probation instead of jail time. This shifting of focus from the offence back to the offender at this stage is standard practice, especially so in a system like India, where trial judges have vast discretion at the sentencing stage and can award sentences anywhere between zero and the prescribed maximum. 

The Hidden Salience of Character Evidence Today

On paper, this is perhaps the end of the matter when it comes to evidence of bad character, but in practice it is not so. Consider a phenomenon that has been demonstrated in India, in context of the limits placed on use of character evidence to question the credibility of victims testifying in rape trials [Proviso to Section 146, IEA]. The rule was designed to curb decision-making based on ideal stereotypes of rape victims: it can’t be rape since the victim had an active sex life, etc.

In his book, Professor Satish demonstrates that, in practice, the rule only displaced such inquiries about a victim’s character from the guilt-determination phase to the sentencing phase. His empirical study demonstrated that sentences were higher in cases where the victims fit the stereotype of vulnerability, as against cases where they did not. Admittedly, the text of the rule does not bar the judge from considering the evidence of a victim’s character at the sentencing phase. But surely this indirectly perpetuates the same stereotypes that the rule was intended to eliminate.

This example nicely illustrates the hidden salience of character evidence today. Although the last century saw limits being placed on the relevance of character for questions of guilt or innocence, it is becoming increasingly apparent that notions of character — who you are — are still very pertinent in how criminal law is administered. While the example above was about victims’ interests, the same issue also exists when we consider the position of defendants: character evidence plays a big role in how persons are targeted as suspects and defendants in the criminal process. 

The argument is simple: None of the changes introduced to restrict the use of character evidence puts a handbrake on how crimes were investigated. To understand this better, imagine criminal law as a funnelling exercise. It is impossible to discover all instances of crime. Out of reported instances, it is again impossible to prosecute each of them for want of resources. Enforcement of law has a necessary selectiveness; an exercise of discretion by the police decides which cases merit the use of precious government resources.

How does the police exercise this discretion? They do so by employing a strategy of “rounding up the usual suspects”, i.e., on the basis of the very notions of bad character that the law has deemed improper for use at the trial. As a result, across the world, the data on persons arrested and prosecuted for crime appears skewed. Persons who are poor or from minority and / or immigrant groups are far more likely to be targeted by the police as suspects while choosing which cases to prosecute. In India, this has been witnessed in the disproportionate numbers of Muslim men being arrested in custody.

Character Evidence and Prosecuting Terror 

A lack of effective checks on the discretion of law enforcement agents is at the heart of the hidden salience of character evidence in the criminal law today. This is true for India as well. There is no set of rules or standards that guides which offences are classed as cognizable (offences allowing arrests without judicial warrants) leading to an unprincipled growth in cognizable crimes over time. A wide array of cognizable crimes gives legal sanction to the police for pretextual arrests and prosecutions and an opportunity to enforce character-based prejudices rather than the pursuit of genuine threats or suspects.

Broadly defined criminal laws also give wide, unchecked discretion to the police. These can be very deceptive as they often have a mental element like “knowledge” or “intention” clearly set out, as if the lawmakers are telling us that this is a proper crime and not a regulatory offence that criminalises without any need of showing culpability. But once we move past this, it is quite common to find a broad definition of the conduct that is being criminalised. Since the police cannot be expected to prove existence of the culpable mental element when they initiate the criminal process by arrest, effectively, through broadly defined criminal laws the law confers a license on the police to arrest anyone they desire. 

To see this in play, imagine a statute that makes it a crime to knowingly offer financial support to an organisation banned by the government. Now, imagine another law that makes it a crime to “associate” with a banned organisation with an “intention to further its activities”. Remember that the police will normally initiate the process based on observable conduct and not the mental element that accompanied it. Keeping this is mind, it is evident that the first example offers a degree of clarity: the criminal conduct is specified, and consequently helps confine the compass of suspicion for the police to justify arrest and investigation. The second is like clay putty in a child’s hands, as permitting police to arrest for suspected “associations” means, according to the dictionary, they can arrest for suspicions of any connection between two entities. 

These are not imaginary crimes but part of India’s main anti-terror statute, the Unlawful Activities Prevention Act 1967 (UAPA) [Sections 40 and 38, respectively]. Another example of an offence that has a clearly stated mental element, coupled with a broad range of conduct that can trigger the crime is Section 15 of the UAPA. This provision criminalises "Terrorist Acts" by first specifying the mental state, but then expands the range of proscribed conduct to include damage to any property. What this translates to is a vast ocean of unregulated discretion with the police to target persons who “look like” trouble, or are “anti-national types”, or “urban maoists” through the UAPA without any hard proof of conduct that justifies the kind of fear that the label of terrorism generates. 

Is this why an overwhelmingly large number of prosecutions under the UAPA not bring convictions? Perhaps. But the problem is that the trial stage, where an improper and possibly a character based prosecution can be screened out after being tested for proof and reasonable doubt, comes after an excruciatingly long wait in UAPA cases. Trials under the UAPA involve a long waiting period that is usually spent by defendants in custody, since the UAPA adopts a much harsher set of rules for granting bail and also allows for longer pre-trial custody during investigations [Section 43-D, UAPA]. Thus, even though the rule of law mandates that individuals must not be targeted and tried simply because of who they are, the prosecution of anti-terror crimes in India actively perpetuates these consequences.

Conclusion

Criminal trials counteract the intuitive notion that our past acts can serve as a predictor of future conduct. But it would be a mistake to assume that the rules of criminal trials spread consistently throughout the legal system. A cursory look confirms that character evidence remains relevant in the stages of sentencing with express legal support. But a closer look reveals the existence of a hidden salience of character evidence, often working to undermine the very goals that the exclusion of such evidence at trials was designed to achieve. 

Ending this haphazard nature of regulation for character evidence will certainly help to bring any legal system in alignment with its professed aims. Thus, if Indian legislators really do think that the evidence of character is useful for determining guilt or innocence, then they should effect such changes. The limited legislative engagement with these issues in recent decades suggests that India’s professed values support the opposite conclusion, however, which then means that steps must be taken to reduce the role character evidence plays through the criminal process.

An obvious point of reform would be placing stricter checks on the enforcement of laws by police by arrests. It is something that the Indian Supreme Court has supported, albeit in a haphazard manner itself through cases like Joginder Kumar, DK Basu and Arnesh Kumar, and so is arguably a point of intervention that can gather broad consensus. But this will take time, and in that time, hundreds of persons will remain in custody awaiting a trial in their cases, having been prosecuted under vague offences such as the UAPA. 

Perhaps the first point of intervention, then, is to reform the process of pre-trial custody in this context. Given the high probability of character-based prosecutions which do not withstand the test of "beyond reasonable doubt", should the law stick to a process which not only allows for upto six-months of custodial detention during an investigation, but also makes it virtually impossible to grant bail till conclusion of the case? If the trial cannot follow in a conviction, we must ensure that the process is not a proxy for the same punishment.