LLM Group 5 Criminal Law and Criminal Administration Criminal Jurisprudence 2024-25 Question Paper with Solutions

Intention: the law will not act, and why

A bare intention is never punishable, and two reasons are given.

Proof. An unexpressed intention leaves no trace. It could be established only by the accused's own confession or by conjecture from conduct, and conduct that reveals an intention has usually passed beyond intention already. A law punishing intention would in practice punish suspicion.

Principle. The criminal law punishes conduct, not character. A person who forms an intention and abandons it has done nothing, and punishing him would remove the incentive to abandon it, which defeats the preventive purpose the rule is supposed to serve.

But once the intention is shared, the law acts at once. Criminal conspiracy under section 120A of the Penal Code, now section 61 of the Bharatiya Nyaya Sanhita, 2023, makes an agreement to commit an offence punishable in itself, without any act in pursuance. The reason is that the two objections fall away: an agreement between persons leaves traces, in meetings, in money and in the conduct of the parties; and a person who has committed himself to others is far less likely to abandon the design.

Illustration. A buys poison intending to kill B. He may throw it away. Until he does something towards administering it he is not liable for an attempt, though he may be liable for a separate offence if the purchase itself was unlawful.

Attempt: where the law acts

Section 511 of the Code, now section 62 of the Sanhita, is the residuary provision. Specific attempts are separately punished: attempt to murder was section 307 and is now section 109; attempt to commit culpable homicide was section 308 and is now section 110.

Four tests compete for the boundary.

The last act or proximity test of R v. Eagleton, (1855) Dears CC 515, where Parke B. said that acts remotely leading towards the commission of the offence are not to be considered as attempts, but acts immediately connected with it are. Its merit is certainty; its defect is that it makes the law wait until intervention is least useful.

The equivocality test, asking whether the act, standing alone, is unequivocally referable to the offence. Its defect is that very few acts are unequivocal.

Om Parkash v. State of Punjab, AIR 1961 SC 1782, applied that in the most striking way. The appellant confined his wife and deprived her of food over a long period, in what the High Court found to be a scheme of regular starvation designed to accelerate her death; she escaped in a state of extreme emaciation. He argued that no single act of his could have caused death. The Supreme Court upheld the conviction under section 307, holding that a person attempts murder when, with the intention to commit murder, he does an act towards its commission, irrespective of whether that act is the penultimate act. A continuing course of omission can therefore be an attempt.

State of Maharashtra v. Mohd. Yakub, (1980) 3 SCC 57, extended it to smuggling. Silver ingots were carried by jeep to a lonely place near a creek and unloaded near the sea, and the sound of a mechanised vessel was heard; no silver had left India. The Supreme Court held it an attempt, requiring an act proximate to the offence and not the penultimate one.

Malkiat Singh v. State of Punjab, AIR 1970 SC 713, is the case on the other side, and the contrast is what makes both memorable. A truck carrying paddy was stopped thirty-two miles from the Delhi border in breach of an order restricting export. The Supreme Court held this preparation only: the driver could still have been dissuaded or could have turned back.

How much an attempt is punished, and why less

Section 511, now section 62, punishes an attempt with imprisonment of any description provided for the offence, for a term extending to one-half of imprisonment for life or one-half of the longest term provided for the offence, or with the fine provided for it, or with both.

Two theories explain the halving and its exceptions. The subjectivist view is that the attemptor showed the same guilty mind and the same willingness to act, and that whether the bullet struck is a matter of luck; on that view the punishment should be the same. The objectivist view is that the criminal law responds to harm, that no harm occurred, and that society's alarm is correspondingly smaller.

Indian law took the objectivist side in the residuary section and the subjectivist side in the specific ones: section 307, now section 109, punishes attempt to murder with imprisonment up to ten years, and with imprisonment for life if hurt is caused. The more serious the offence attempted, the closer the punishment for the attempt comes to that for the completed crime.

Where preparation is itself punished, and why

The Code departs from the general rule wherever the three reasons for it fail. The list should be given.

State of Madhya Pradesh v. Narayan Singh, (1989) 3 SCC 596, shows how fine the margin is where there is no such special provision. Lorries carrying fertiliser without the permit required by the Fertilisers (Movement Control) Order, 1973 were intercepted before reaching the Maharashtra border. The trial court held this mere preparation to export; the Supreme Court reversed, holding that the conduct had passed into an attempt.

Commission, and abandonment

When the actus reus is complete and accompanied by the required mental element, the offence is made out and the earlier stages are absorbed into it. An attempt is not charged in addition to the completed offence.

A question that follows directly from locus poenitentiae is whether an accused who has crossed into attempt may escape by desisting. The Indian answer is that he may not: once the offence of attempt is complete it is complete, and a change of heart afterwards goes to sentence and not to liability. The doctrine operates before the line, marking the accused as still preparing, and has no application after it. That is the practical significance of Abhayanand Mishra, and the contrast with Malkiat Singh, where the possibility of turning back was itself the reason the Court held the accused at preparation.

For full marks, cover: the question is about the Court's role, not merely its structure, so the answer must say what the Court is for, what it has actually achieved, and where it has failed. Give the road to Rome briefly, the four crimes, complementarity as the Court's real method of control, the organs because they are how it works, worked cases, and then an honest assessment. India's position belongs here too, because a court India has not joined controls nothing in India except through the Security Council.

What the Court was created to control

International criminal law before 1998 was retrospective and ad hoc, and its control of atrocity was accordingly accidental.

The International Military Tribunals at Nuremberg and Tokyo established the two propositions on which all later control rests: that international law imposes duties on individuals and not only on States, and that official position is no defence. But they tried only the vanquished and were created after the acts.

Crimes against humanity, Article 7: enumerated acts including murder, extermination, enslavement, deportation, torture, rape and other sexual violence, persecution, enforced disappearance and apartheid, committed as part of a widespread or systematic attack directed against any civilian population, with knowledge of the attack. No armed conflict is required, which is what makes this the widest of the four.

War crimes, Article 8: grave breaches of the Geneva Conventions of 1949 and other serious violations of the laws and customs of war, extending expressly to armed conflict not of an international character.

The crime of aggression, Article 8 bis, defined by the Kampala amendments of 2010 and activated in 2018 on narrower terms than the other three.

The Court prosecutes individuals, never States. Article 27 removes official immunities, including that of a serving Head of State; Article 26 excludes persons under eighteen at the time of the conduct; and jurisdiction is prospective only, from 1 July 2002 or a State's later accession.

The machinery: how a case moves

Triggers, Article 13: referral by a State Party; referral by the Security Council under Chapter VII, the only route reaching a non-party; and the Prosecutor acting on his own motion under Article 15, with the authorisation of a Pre-Trial Chamber.

The organs, Article 34, are four. The Presidency, responsible for the administration of the Court other than the Office of the Prosecutor and for relations with States on the enforcement of sentences. The Judicial Divisions, being the Pre-Trial Division of not less than six judges, which authorises investigations, issues warrants under Article 58 and confirms charges under Article 61; the Trial Division of not less than six; and the Appeals Division, composed of the President and four other judges. The Office of the Prosecutor, which acts independently as a separate organ and whose members neither seek nor act on instructions from any external source, and which is obliged by Article 54 to investigate incriminating and exonerating circumstances equally. And the Registry, which carries the Victims and Witnesses Unit, victim participation under Article 68(3), and the Trust Fund through which reparations ordered under Article 75 are administered.

Prosecutor v. Dominic Ongwen, in 2021, held that the accused's own abduction as a child by the Lord's Resistance Army did not excuse the crimes he later committed as a commander, though it was relevant to sentence. Its importance is that it refused to allow a real and terrible mitigating history to defeat responsibility altogether.

Beyond convictions, the Court's role in control has three further aspects worth naming. It has produced a settled body of definitions, in Articles 6 to 8 and in the Elements of Crimes, which national legislatures and courts now use. It has driven domestic implementing legislation in many States Parties, which is complementarity working as intended. And its victim participation and reparations regime, which no earlier tribunal had, has shifted international criminal justice away from being purely about the accused.

Where it has failed

Enforcement is the structural weakness and it defeats everything else. The Court has no police force. A Pre-Trial Chamber may issue a warrant, but only a State can execute it; the Presidency may designate a State of enforcement, but only if one volunteers.

India and the Court

India participated at Rome and voted against the Statute. It has not signed and is not a State Party, so the Court has no jurisdiction over Indian territory or nationals save on a Security Council referral.

India's objections, unchanged since 1998, are five: the Security Council's power to refer and to defer; the omission of the use of nuclear weapons from the list of war crimes while conventional weapons are listed; the omission of terrorism; the inclusion of non-international armed conflicts, which reaches internal security; and the Prosecutor's proprio motu power.

India relies on complementarity, its own courts having jurisdiction over the same conduct. The weakness in that answer is that India has enacted no domestic legislation defining genocide, crimes against humanity or war crimes as such, so the conduct is prosecuted, if at all, as ordinary offences under the Sanhita. The conduct in Lubanga, the recruitment of children, is criminalised by section 83 of the Juvenile Justice (Care and Protection of Children) Act, 2015; genocide has no Indian counterpart at all.

Conclusion. The International Criminal Court controls international crimes by defining them, by prosecuting individuals for genocide, crimes against humanity, war crimes and aggression, and above all by making national failure the condition of its own jurisdiction, which is what complementarity under Article 17 means. It works through the four organs of Article 34, with an independent Prosecutor bound to investigate exonerating circumstances and a Registry that gives victims a place no earlier tribunal gave them. Lubanga, Al Mahdi and Ongwen show what it achieves; Al Bashir shows what it cannot, because the Court has no police and depends entirely on States to execute what it decides. India's absence, and its five stated objections, mean that for India the Court's role is at present confined to the normative one.

Section 96, now 34: nothing done in the exercise of the right is an offence. Section 97, now 35: the right runs against any offence affecting the human body, and against theft, robbery, mischief and criminal trespass in relation to property. Section 98, now 36: it exists even against a person immune by youth, immaturity, unsoundness of mind or misconception. Section 99, now 37: the four restrictions. Section 100, now 38: the seven occasions on which the defence of the body extends to causing death. Section 101, now 39: otherwise, harm short of death. Sections 102 and 105, now 40 and 43: commencement and continuance. Section 103, now 41: the four occasions on which the defence of property extends to causing death. Section 104, now 42: otherwise, harm short of death. Section 106, now 44: the deadly assault with risk to an innocent person.

Two characteristics govern everything: the right is preventive and not punitive, and it runs against an aggressor only.

The first criticism: the right is too narrow for the person it protects

The lists are closed. Section 100 permits killing on seven descriptions and no others; section 103 on four and no others. A threat that is grave but unlisted can never justify killing, however it appeared at the time. The list was extended once, in 2013, to add acid attacks, which is itself an admission that the closed form is inadequate: for a century and a half the throwing of acid was not among the occasions on which a person might defend her life by taking one.

The property list is narrower still. Section 97 answers only theft, robbery, mischief and criminal trespass. There is no right of private defence of property against cheating or criminal breach of trust, however dishonest.

Section 99's third restriction assumes a State that answers. There is no right where there is time to have recourse to the protection of the public authorities. That presupposes public authorities who can be reached and who will come, which in much of the country is not the position, and it is the restriction most likely to defeat a genuine plea.

Amjad Khan v. State, AIR 1952 SC 165, illustrates how far apprehension can carry. During communal rioting at Katni the appellant, inside his shop with his family, heard a mob break down the door of the adjoining shop and begin battering his own, and fired through his own door, killing a man he never saw. The Supreme Court set aside the conviction for murder. The result is defensible; the method, which is to accept a lethal response to a danger the accused inferred from sound, shows how little the section requires.

The burden of proof is very light. Section 105 of the Evidence Act, 1872, now section 108 of the Bharatiya Sakshya Adhiniyam, 2023, puts the burden on the accused, but Salim Zia v. State of Uttar Pradesh, (1979) 2 SCC 648, holds that it is discharged on a preponderance of probabilities and may be made out from the prosecution's own material without the accused leading any evidence or entering the witness box. And Darshan Singh holds that the plea need not even be expressly taken. A defence which need not be pleaded, need not be supported by evidence, and need only be made probable is a substantial concession.

The right also assumes a defender who can judge, in the instant, whether the assault falls within one of seven descriptions and whether more harm than necessary is being inflicted. Deo Narain concedes the unreality of that by refusing to weigh the force in golden scales. The concession is necessary and it is also an admission that the statutory scheme describes a deliberative process which does not occur.

Where the criticisms are answered

On the closed lists. The alternative to enumeration is a general standard of proportionality, which would give courts no more guidance than the words "reasonable" and "necessary", and would make outcomes turn on the tribunal rather than on the law. Enumeration buys predictability at the cost of the occasional hard case, and Parliament can and does add to the list, as it did in 2013.

A position

The right of private defence is correctly framed and imperfectly administered.

Its structure is sound: an enumerated right, preventive not punitive, available against an aggressor, beginning with a reasonable apprehension and ending with it, limited by an express requirement of necessity, and graduated at the point of excess. Every element of that structure answers a real problem.

Its weaknesses are of two kinds and they pull in opposite directions, which is itself evidence that the balance is roughly right. It is too narrow where the State is absent, because the third restriction in section 99 assumes an availability of public authority that does not exist, and because the closed lists cannot anticipate every grave threat. It is too wide where the accused is the only surviving witness, because apprehension is asserted rather than proved and the burden on him is light.

The reform that would do most is not doctrinal. It is evidentiary: prompt investigation, proper recording of injuries on both sides, and reliable determination of who began the incident. Those are what actually decide these cases, and Deo Narain, Amjad Khan and James Martin are all, in substance, decisions about evidence.

For full marks, cover: five theories, and for each the question it answers, its philosophical source, its Indian statutory expression, its leading case and its criticism. That structure turns a list into a 25-mark answer. Then trace the historical movement, from private vengeance to restorative justice, because the theories are stages of that movement and not merely alternatives. Close by saying which theory Indian law actually follows, supported from the sentencing decisions.

The question each theory answers

Punishment is the deliberate infliction of suffering by the State on a person, and it therefore requires justification. Each theory answers a different question.

Deterrence answers what will punishment achieve? Retribution answers what does the offender deserve? Prevention answers how do we stop him doing it again? Reformation answers what does the offender need? Compensation answers what does the victim need?

Because they answer different questions, they yield different measures of how much may be inflicted, and that is why a system cannot simply adopt all five.

Criticism. Deterrence has no internal limit: if severity deters, the theory justifies disproportionate punishment, and taken to its end it would justify punishing an innocent person where the public believed him guilty. Empirically it is weak. The 262nd Report of the Law Commission of India, 2015, found the deterrent effect of capital punishment unproved and recommended abolition for all offences except terrorism-related offences and waging war. Certainty of detection deters more than severity of sentence, and the system delivers little certainty.

The retributive theory

The idea. Punishment is deserved because the offender has done wrong. It looks backward to the offence rather than forward to consequences, and its measure is proportionality. Kant is the classical exponent: a person must be punished because he has committed a crime and never merely as a means to some other good, since to use him so is to treat him as a thing.

In Indian law. Imprisonment for life; under section 104 of the Sanhita imprisonment for the remainder of natural life where a life convict commits murder; disqualification from office; cancellation of licences; and at its extreme the death penalty, which incapacitates absolutely.

Criticism. Prevention justifies detaining a person for what he might do rather than for what he has done, which is preventive detention rather than punishment, and it has no natural limit either, since the safest course is always to detain longer. It assumes that dangerousness can be predicted, and the record of prediction is poor.

The reformative theory

The idea. The offender is a person capable of change, and the sentence is an occasion for rehabilitation. Crime is treated as having causes, in circumstance and in psychology, which the sentence should address.

Criticism. Reformation gives no measure of severity: a sentence fixed by what the offender needs may be longer or shorter than what he deserves, and indeterminate sentences fixed by administrators invite abuse. It also assumes a capacity and a machinery for reform which Indian prisons largely lack.

The expiatory or compensatory theory

The idea. In its older form, punishment is the means by which the offender expiates his wrong and is restored to the community. Its modern form directs attention to repairing the harm to the victim rather than to the offender's suffering.

In Indian law. Sections 395 and 396 of the Bharatiya Nagarik Suraksha Sanhita, 2023, formerly sections 357 and 357A of the Code of Criminal Procedure, on compensation out of a fine and on the State victim compensation scheme, the latter operating even where there is no conviction and even where the offender is not traced. The compounding of offences. The power to order compensation as a condition of release under the Probation of Offenders Act.

The third was the State's appropriation of the wrong, when the King's Peace made an offence against a subject an offence against the Crown. Prosecution ceased to belong to the victim, and with it the sanction ceased to be a transfer to him. Deterrence and prevention are the theories of a State that punishes in its own name.

The fourth, and the most recent, is reformation, which begins from the offender rather than from the offence, and restorative justice, which brings the victim back in without returning to composition.

Which theory does Indian law follow?

No single theory does, and the marks lie in showing which governs which part.

Capital sentencing is retributive at its threshold and reformative in its mitigation. Bachan Singh v. State of Punjab, (1980) 2 SCC 684, upheld the death penalty but confined it to the rarest of rare cases where the alternative of life imprisonment is unquestionably foreclosed, required aggravating and mitigating circumstances to be weighed together, gave the mitigating circumstances a liberal construction, treated the probability of reformation as itself a mitigating circumstance, and placed on the State the burden of showing that the accused is beyond reform.

Conclusion. The theories of punishment are deterrent, retributive, preventive, reformative and expiatory or compensatory, and each answers a different question and yields a different measure of severity. They are also stages of a historical movement from private vengeance through limitation and composition to State prosecution and, most recently, to reform and restoration. Indian law contains all five: retribution supplies the proportionality that caps the sentence; deterrence supplies the minimum sentences and much of the rhetoric, though the Law Commission's 262nd Report found its empirical basis wanting; prevention supplies imprisonment for the natural life; compensation has become a mandatory consideration after Ankush Shivaji Gaikwad; and reformation supplies the operative principle within the retributive ceiling, as Bachan Singh, Bariyar, Manoj, Mohammad Giasuddin and Sunil Batra together show.

The Penal Code never uses the phrase. Macaulay's method was to write the required mental element into the definition of each offence, and then to supply in Chapter IV a set of general exceptions which negative fault. Section 6 directed that every definition be read subject to that chapter. So mens rea in Indian law is found by reading the section charged.

It is not one state of mind but several. Intention, the aim or purpose, as in section 300 of the Code, now section 101 of the Bharatiya Nyaya Sanhita, 2023. Knowledge, awareness of a consequence as likely, as in the third limb of section 299, now section 100. Rashness, conscious risk-taking, and negligence, inadvertence to a risk that ought to have been seen, as in section 304A, now section 106. Dishonesty and fraudulently, defined in sections 24 and 25, now sections 2(7) and 2(9). And good faith, defined in section 52, now section 2(11), as requiring due care and attention.

It is not motive. Intention is the mental element the definition requires; motive is the reason that moved the accused to form it. A man who kills his terminally ill mother to end her suffering has a compassionate motive and a guilty intention. Motive is irrelevant to guilt, relevant to proof where the case is circumstantial, and always relevant to sentence.

Sherras v. De Rutzen, [1895] 1 QB 918, sets the presumption against it and names the classes where displacement is usual: acts not criminal in any real sense but prohibited in the public interest, public nuisances, and proceedings criminal in form but really a summary means of enforcing a civil right. Sweet v. Parsley, [1970] AC 132, restored the presumption after a period of drift.

State of Maharashtra v. Mayer Hans George, AIR 1965 SC 722, is the Indian case for displacement. A German national carrying about thirty-four kilograms of gold in a specially made jacket flew from Zurich towards Manila and the aircraft touched down at Bombay, in breach of a Reserve Bank notification made shortly before, of which he knew nothing. The Supreme Court held by majority that the offence under the Foreign Exchange Regulation Act, 1947 was one of strict liability, because the object of the Act was to conserve foreign exchange and prevent smuggling and would be defeated if knowledge had to be proved.

Noor Aga v. State of Punjab, (2008) 16 SCC 417, is the controlling decision. The appellant, an Afghan national, was convicted under the NDPS Act on a recovery of heroin, and the reverse-onus provisions were challenged. The Supreme Court upheld them, describing the presumption of innocence as a human right which cannot per se be equated with the fundamental right under Article 21, but held that a reverse burden does not absolve the prosecution of anything: it must first establish the foundational facts and make out a prima facie case, and only then does the legal burden shift, the accused's burden thereafter being a preponderance of probabilities.

Evaluation. This exception is the most defensible in form and the most dangerous in practice. In form it does not abolish the mental element; it reallocates the burden of establishing it, and Noor Aga keeps the prosecution to proving the facts on which the presumption operates. In practice the foundational facts are often easy to prove and the presumption then does the whole work, and the accused who cannot afford evidence cannot displace it. The safeguard is real but it depends on courts insisting on the foundational facts, which is a matter of vigilance rather than of doctrine.

Evaluation. The imputation is a fiction, and it operates only against the accused. Its justification is that the accused chose to drink and cannot rely on a condition he created. The objection is that the choice to drink is not a choice to kill, and that the fiction convicts on a mental state the accused demonstrably did not have. English law met the same problem in Director of Public Prosecutions v. Majewski, [1977] AC 443, by holding self-induced intoxication no answer at all to a crime of basic intent, which is a cleaner solution because it states the rule rather than pretending a fact. India's version is less honest and, thanks to Basdev, narrower in effect, because it leaves intention to be proved.

The fourth exception: offences drafted without a mental element

Section 304A, now section 106, punishes causing death by a rash or negligent act. Negligence is not a state of mind in the way intention and knowledge are; the accused is liable precisely because he did not advert to the risk.

The general burden never shifts. Dahyabhai Chhaganbhai Thakker v. State of Gujarat, AIR 1964 SC 1563, holds that even where an accused fails to establish a general exception, the material he has led may still raise a reasonable doubt about the intention the offence requires, in which case he must be acquitted on the prosecution's own burden. That proposition survives every one of the exceptions above.

It has constitutional weight. The presumption of innocence has been read into the fair-trial content of Article 21, which is why every reverse-onus clause is read down to require the foundational facts first.

For full marks, cover: this answer is built as a justification and a critique, because the paper says "explain in detail" and the interesting thing about constructive liability is not its content but whether it can be defended. Set out the five devices with the Sanhita numbering, work the cases, and then ask of each device whether the person convicted is genuinely an author of the act. Note the trap that common intention no longer has a section of its own.

The problem, and the price of solving it

Criminal law is built for the individual offender. It asks what this accused did and with what mind, and it punishes him for that.

Serious crime is collective. A dacoity, a riot or a planned killing is committed by a group in which one pair of hands strikes the blow while others plan it, arm it, watch the door and carry away the proceeds. To convict only the hands misdescribes the crime. To acquit the rest because no individual act can be traced to them makes organised crime unpunishable.

Barendra Kumar Ghose v. King Emperor, AIR 1925 PC 1, the Postmaster case, establishes what the section attributes. On 3 August 1923 several men entered the sub-post office at Shankaritola in Bengal. The sub-postmaster was counting money when they demanded it; shots were fired and he was killed. Barendra Kumar Ghose was caught fleeing and said he had stood at the door, fired nothing and done nothing. The Privy Council upheld his conviction of murder read with section 34, holding that the section deals with the doing of separate acts, similar or diverse, by several persons, and that if all are done in furtherance of a common intention each is liable for the result of them all as if he had done them himself.

Mahbub Shah v. Emperor, AIR 1945 PC 118, supplies the limit. Allah Dad and others took a boat down the Indus to cut reeds from disputed land; on their return Ghulam Quasim Shah tried to stop them and was struck; Wali Shah and Mahbub Shah came up armed and Wali Shah shot Allah Dad dead. The Privy Council set aside Mahbub Shah's conviction: there was no evidence of a pre-arranged plan, and the two men had at most the same intention formed on the spur of the moment. Care must be taken, the Board said, not to confuse same or similar intention with common intention; the partition dividing them is often very thin, and the distinction is real and substantial.

The second device: common object

Section 149 of the Code, now section 190 of the Sanhita, makes every member of an unlawful assembly guilty of an offence committed by any member in prosecution of the common object, or of an offence the members knew to be likely to be committed in prosecution of that object. It depends on section 141, now section 189(1), defining an unlawful assembly as five or more persons whose common object is one of five listed objects.

Three differences from section 34, and each raises the price. Five persons instead of two. No prior meeting of minds is required: membership is the basis. And the second limb reaches offences the members merely knew to be likely.

Mizaji v. State of Uttar Pradesh, AIR 1959 SC 572, works that limb. A party set out to take forcible possession of a field; one member carried a pistol and used it when the party was resisted, killing a man. The Supreme Court held the others liable: where men set out to take possession by force and one of them is armed with a deadly weapon, a killing is an offence the members knew to be likely in prosecution of the common object. No agreement to kill existed and none was required.

Three things are said in defence. First, he chose to join an assembly whose object was itself unlawful, and a person who joins an armed party bent on forcing possession accepts what such parties do. Second, the alternative is impunity, because in a riot no individual act can be traced and a rule requiring proof of individual acts would acquit everyone. Third, the courts have narrowed it: Baladin and Masalti on membership, the rule of prudence on proof, and the requirement that the offence be committed in prosecution of the common object, which excludes an act done by one member on a private impulse.

The defence is partial. The first argument works where the assembly's object was violent and fails where it was not; the second is an argument from necessity rather than from justice; and the third relies on judicial vigilance rather than on the section's own terms. Section 149 remains the weakest link in the doctrine, and the Sanhita reproduced it as section 190 without change.

Are they justified? Conspiracy is defensible on the ground that the agreement is itself the dangerous thing, since a person who has bound himself to others is far less likely to withdraw. Its danger is evidentiary rather than doctrinal: the actus reus is invisible and the evidence is usually accomplice testimony or the conduct of others, and the co-conspirator rule in section 10 of the Evidence Act, now section 8 of the Bharatiya Sakshya Adhiniyam, 2023, admits material against a person who was not present. Kehar Singh's insistence on agreement rather than association is the whole safeguard.

Abetment is the most easily justified of all, because the abettor has done a positive act directed at the commission of the offence.

The fifth device: offences defined by the group

Dacoity is robbery by five or more persons, section 391 and now section 310, so the number is an element of the offence and not a route to liability. Section 396, now section 310(3), makes every one of the five liable to death or imprisonment for life where any one commits murder in the course of it.

The threshold: what is charged

Nanak Chand v. State of Punjab, AIR 1955 SC 274, holds that section 149 creates a specific offence and section 34 does not. A charge under section 302 read with section 149 is therefore a distinct charge from one under section 302 alone or read with section 34, and a person charged with the one cannot be convicted of the other without a specific charge. Conversion is permitted where no prejudice results, but prejudice must be decided on the record.

The devices ranked by how well they can be defended

For full marks, cover: the paper offers four and asks for two, so all four are written out below and the choice is yours. At 25 marks for two, each note is worth about twelve or thirteen: definition, statutory or doctrinal basis, at least one worked case and a criticism. Do not write four in the hall.

(a) Jus necessitatis

Jus necessitatis, the right of necessity. It is the principle: the principle that a person may do what would otherwise be an offence in order to avert a greater harm. The maxims are necessitas non habet legem, necessity knows no law, and salus populi suprema lex, the safety of the people is the supreme law.

The limit. R v. Dudley and Stephens, (1884) 14 QBD 273. The yacht Mignonette was wrecked in the South Atlantic and four men, Dudley, Stephens, Brooks and the cabin boy Richard Parker, were adrift in an open boat with almost no food and no water. After about twenty days, the boy being weakest and probably dying, Dudley and Stephens killed him and the three survivors fed on his body; they were rescued four days later. The Queen's Bench Division held that necessity is no defence to murder. Lord Coleridge C.J. reasoned that to allow it would compel a court to measure the comparative value of lives and to decide who should be selected to die, and that an extreme temptation is not a legal excuse. They were convicted and sentenced to death, and the sentence was commuted to six months' imprisonment.

The case explains section 81 rather than contradicting it: the deliberate killing of an innocent person is done with the intention of causing the harm, and falls outside a section that excuses only harm caused with knowledge.

Criticism. Necessity is the exception courts distrust most, because it invites the argument that law must yield to circumstances, and its boundaries have been fixed by hard cases rather than by principle. The reply is that a law with no room at all for the lesser evil would criminalise the ship's captain in the Code's own illustration.

(b) Decriminalisation

Definition. Decriminalisation is the withdrawal of conduct from the catalogue of offences, so that what the law once punished ceases to be punishable at all.

Three distinctions. It is not legalisation, which creates a lawful regulated activity; decriminalised conduct may remain a civil wrong. It is not non-enforcement, where the offence survives on the statute book and continues to support arrest and blackmail. And it is not the same as reading down, where a court narrows a provision that remains in force.

Section 497, adultery. Joseph Shine v. Union of India, (2019) 3 SCC 39, struck it down, holding that it treated a wife as her husband's property because the offence turned on whether the husband had consented. Adultery remains a ground for divorce, which is the cleanest demonstration that decriminalisation is not legalisation.

Section 309, attempt to suicide. Displaced by section 115 of the Mental Healthcare Act, 2017, and the Sanhita retains it only in the narrow form of section 226, an attempt made to compel a public servant.

The counter-example. Triple talaq was criminalised, not decriminalised. Shayara Bano v. Union of India, (2017) 9 SCC 1, held the pronouncement void; Parliament then made it an offence by the Muslim Women (Protection of Rights on Marriage) Act, 2019.

Criticism. Judicial decriminalisation is counter-majoritarian and can only read a provision down, leaving the text in force. Legislative decriminalisation by omission can leave gaps, as the repeal of section 377 did.

Death. Section 393(3) of the Bharatiya Nagarik Suraksha Sanhita, 2023, formerly section 354(3) of the Code of Criminal Procedure, requires the court to record special reasons for a death sentence, so life imprisonment is the rule and death the exception. Bachan Singh v. State of Punjab, (1980) 2 SCC 684, confined it to the rarest of rare cases where the alternative of life imprisonment is unquestionably foreclosed, on a weighing of aggravating and mitigating circumstances in which the circumstances of the offender count equally. Machhi Singh v. State of Punjab, (1983) 3 SCC 470, grouped the categories. Mithu v. State of Punjab, (1983) 2 SCC 277, struck down section 303, which made death mandatory for a life convict who committed murder, because a mandatory sentence leaves no room for mitigation.

Imprisonment for life means imprisonment for the whole of the remainder of the convict's natural life unless remitted, and section 6 of the Sanhita provides that in calculating fractions of terms it is reckoned as equivalent to twenty years.

Fine. Section 8 governs the amount, liability in default and the term of imprisonment in default.

(d) Presumption of innocence

What it asserts. A person accused of an offence is to be treated as innocent until the prosecution proves his guilt beyond reasonable doubt. It is not a prediction about accused persons; it is a decision about who bears the risk of a mistake. Every trial can err by convicting the innocent or by acquitting the guilty, and the presumption instructs the court to prefer the second error to the first. Blackstone gave it its arithmetic: better that ten guilty persons escape than that one innocent suffer.

Its classical statement. Woolmington v. Director of Public Prosecutions, [1935] AC 462. The accused shot his estranged wife and said the gun had gone off accidentally. The trial judge directed the jury that once the killing was proved malice was presumed and the accused must show excuse. The House of Lords quashed the conviction, Viscount Sankey LC saying that throughout the web of the English criminal law one golden thread is always to be seen, that it is the duty of the prosecution to prove the prisoner's guilt, subject to the defence of insanity and to any statutory exception.

Its limits. Noor Aga v. State of Punjab, (2008) 16 SCC 417, described the presumption as a human right which cannot per se be equated with the fundamental right under Article 21, and upheld the reverse-onus provisions of the Narcotic Drugs and Psychotropic Substances Act, 1985, while holding that the prosecution must first prove the foundational facts and that the accused's burden thereafter is only a preponderance of probabilities. Reverse burdens now appear in the NDPS Act, in section 29 of the Protection of Children from Sexual Offences Act, 2012, in section 24 of the Prevention of Money-Laundering Act, 2002, and in the dowry-death provisions.

Criticism. The presumption is under steady legislative pressure, and the Committee on Reforms of the Criminal Justice System chaired by Justice V. S. Malimath recommended in 2003 that the standard of proof be lowered from beyond reasonable doubt to "clear and convincing". That recommendation was not accepted, and it drew heavy criticism for borrowing an inquisitorial remedy without the inquisitorial safeguards. What has preserved the presumption is not its constitutional rank, which Noor Aga deliberately left modest, but the courts' insistence that the State prove the foundational facts before any presumption operates against a man at all.