A judiciary mains answer is marked on four moves: the issue named, the provision cited, the rule applied to the facts, and a conclusion that decides the question. A 10-mark answer gets roughly nine minutes to make all four.
Those four moves have stayed the same while the law inside them changed. Parliament replaced the Indian Penal Code, the Code of Criminal Procedure and the Indian Evidence Act with effect from 1 July 2024, and State Public Service Commissions and High Courts have since revised their syllabi onto the three new codes. A candidate writing a criminal law answer in 2026 is therefore citing section numbers that did not exist when much of the published preparation material was written.
Section numbers are the smallest part of what the paper asks for. A mains paper exists to find out whether a candidate can decide a matter, which is the work a civil judge does from the first week of the posting, and deciding a matter means applying a provision to a set of facts and stating an outcome. That is why the Delhi Judicial Service paper prints an instruction above its questions requiring candidates to support each answer with reasons, relevant legal provisions and principles, and relevant case law.
The instruction says what to put into an answer and not what a marked answer looks like. No High Court publishes a model answer, no notification states how long a 10-mark answer should run, and serving judicial officers evaluate the scripts to standards candidates only ever learn second-hand. Most aspirants therefore calibrate against other aspirants, which is how a habit such as listing six case names with no principle attached to any of them spreads through a coaching batch and survives.
The gap between knowing the law and writing it in a marked shape shows up most clearly when two candidates answer the same question. Take the 15-mark question on basic structure set in the Uttar Pradesh mains paper, which puts the proposition that the doctrine is a judicial creation mentioned nowhere in the Constitution and asks the candidate to explain it. One candidate opens on the facts of Kesavananda Bharati and spends four paragraphs narrating the reference, the bench strength and the split.
The second candidate concedes the textual half of the proposition in a single line, that Article 368 speaks of amendment and says nothing about a basic structure, and then spends the rest of the answer on why that concession does not carry the conclusion the examiner has invited. Both candidates know the same law and can cite the same cases.
The second answer scores higher, and the reason is that the question was never asking what the doctrine is. It puts a charge and asks the candidate to test it, so an answer that narrates the leading case has described the doctrine without ever meeting the charge. What separates the two scripts is not knowledge of Kesavananda Bharati but the order in which the material arrives.
Marking scheme and time allocation in the mains paper
A judiciary mains paper tests whether a candidate can turn law into a decision inside a fixed number of minutes, and the papers say so on their own face. The Delhi Judicial Service Examination (Written), 2022 prints this above the Civil Law questions: “Support each of your answers with reasons, relevant legal provisions and principles and also relevant case laws.” It follows that with an instruction most candidates read past: “Even if you do not know the answer, it is advisable to attempt as much, as the test is not only of the knowledge of law but also of analytical reasoning.”
Those two sentences are the mark scheme, printed on the paper. Reasoning is being marked separately from recall, which means a candidate who reaches a defensible conclusion through a stated provision and a stated principle collects marks that a candidate with better recall and no visible reasoning does not.
The Uttar Pradesh paper carries an instruction of its own that decides how a script is counted rather than how it is marked. Law Paper I (Substantive Law) in 2022 ran three hours for 200 marks, with Questions 1 and 2 compulsory and any four of the remaining seven, six questions in all. Its note reads that attempts are counted in sequential order, that an attempt counts even if the question is attempted only partly unless it is struck off, and that any page or part of a page left blank in the question-cum-answer booklet must be clearly struck off. A candidate who starts a seventh question loses the marks of the sixth, and a candidate who abandons a half-written answer without striking it off has spent that attempt anyway.
That counting rule sits on top of an arithmetic that governs everything else. Two hundred marks across 180 minutes leaves slightly under a minute a mark, so a 10-mark sub-part is a nine-minute answer and a 15-mark sub-part is a fourteen-minute one. Delhi runs 10-mark short notes and 20-mark problems in the same Civil Law paper, with fifteen minutes of reading time before the booklet may be opened and a fresh page for each answer. A candidate who has never timed an answer at home is discovering these numbers in the hall.
What the examiner looks for
The examiner is looking for a determination, and finds it or fails to find it in the first two lines under the question. Scripts are marked by serving judicial officers who read at speed, in volume, against an answer they have already formed. An answer that states its conclusion at the top and then justifies it is being checked; an answer that builds towards a conclusion across a page is being searched.
That changes where the best sentence in an answer goes. A candidate who saves the sharp point for the closing line has written it for a reader who may already have scored the answer, and the same sentence placed first turns the rest of the page into support for a position rather than a hunt for one. Anyone comparing what different states demand before writing a single answer should start with the calendar of upcoming judicial services exams, because the paper pattern differs enough between states to change how the time is divided.
Structure of a mains answer
A mains answer runs in five moves, namely the issue, the provision, the rule, the application and the conclusion, and while the weight given to each move changes from question to question the order does not. Naming the issue costs one sentence and buys the examiner a frame for everything after it. Setting out the provision, quoted where it is short and paraphrased where it is long, does the work that a bare case citation cannot.
What the coaching listicles leave out is that the skeleton flexes by question type, and a candidate who runs the same five paragraphs at every question loses marks at most of them. Four types appear across these papers, and each puts its weight somewhere different.
A short note of 10 marks is provision-heavy and conclusion-light. The reader wants the definition, the statutory source, the two or three distinctions that matter and one authority, and a candidate who spends three of nine minutes on an introduction has spent a third of the answer saying nothing.
A problem question puts the facts first and the issue second, framed the way a court would frame it. The Delhi paper of 2022 set one in Civil Law-I about a car stolen from a paid parking space where the slip carried an owner’s risk clause, and then asked whether the answer would differ had the car been handed to hotel valet staff against a slip reading “guest’s own risk”. Both limbs turn on a single fact, which is who took possession of the car, and a candidate who treats the two limbs alike has answered neither. Framing the issue as possession rather than as liability is most of the work.
A statement-based question gives the candidate a proposition and asks for it to be explained or discussed, and the proposition itself is the issue. These dominate the substantive law papers, and they are where the marks leak, because the instinct is to restate the proposition in fuller words rather than test it. Three of the four questions worked through below are of this kind.
Judgment writing takes the same five moves and renames them: the issues as framed, the finding on each issue, the reasons for each finding, and an operative order that grants or dismisses. Chhattisgarh set a 40-mark judgment-writing question in its 2024 mains paper requiring the candidate to frame the necessary issues before writing, and framing them wrongly costs the whole answer rather than part of it.
Use of case law in an answer
One landmark, named with its citation and stated as a principle, earns more than five case names in a row. The name-dropping habit is the commonest way a candidate believes the case-law marks have been collected when they have not, because a citation with no proposition attached to it tells the examiner nothing about whether the candidate knows what the case decided.
The landmark belongs after the provision and before the application, in the position where it converts statutory words into a usable test. Two sentences do it: what the court held, and the rule that holding produces. Every one of the four answers below is built that way, and each runs on a single landmark with the rest of the line kept subordinate to it.
Supporting cases still have a use, which is to show a line rather than a point. A candidate who writes that one case was decided one way, a later one reversed it and a third settled the position has demonstrated command of an area in three sentences. A candidate who writes the same three names separated by commas has demonstrated a memorised list.
Length and time per answer
A 10-mark answer runs roughly 250 to 300 words, which is a page and a half in most handwriting and nine minutes of writing time. A 15-mark answer runs 400 to 450 words and a 20-mark answer 550 to 600. These are working figures rather than rules, and their value is that a candidate who has written to them at home knows, without looking at a watch, when an answer has run long.
Overrunning is not a neutral choice, because the minutes come out of a later answer rather than out of nowhere. An answer taken from 9 marks to 10 by three extra paragraphs has cost the time in which a later question could have moved from 0 to 6.
“The principle of basic structure is judicial creation”
Uttar Pradesh set this as part of its compulsory constitutional law question in 2022:
“The principle of basic structure is judicial creation. It has no where been mentioned in the Constitution.” Explain with the help of decided cases. Discuss the principle of basic structure.
Uttar Pradesh PCS (J) 2022, LAW (PAPER-I) Substantive Law, Question 1(c), 15 marks
The proposition here is half correct, and the fifteen marks sit in identifying which half. Its first sentence is largely true and its second is entirely true, and neither carries the conclusion the wording invites, which is that a doctrine unmentioned in the text lacks constitutional authority.
That is what makes this a statement-based question rather than a topic. A candidate who reads it as an instruction to write everything known about basic structure will produce a competent account of Kesavananda Bharati and score in the middle, because nothing in that account engages the charge. The answer has to concede what is true, and then explain why the concession does not do the work the examiner has offered it for.
Model answer on the basic structure doctrine
The doctrine of basic structure is judicial in origin, and it is not for that reason extra-constitutional, because it was arrived at by interpreting a word that Article 368 actually uses. The statement is therefore accepted in part and denied in part.
The textual half of the statement is correct and should be conceded at once. Article 368 confers on Parliament the power to amend the Constitution and prescribes the procedure, and the expression “basic structure” appears neither in that Article nor anywhere else in the text. No provision lists the features that cannot be amended, and no provision says that any limit on the amending power exists at all.
The constitutional question is what the word “amend” in Article 368 permits. The Supreme Court answered it in Kesavananda Bharati v. State of Kerala, (1973) 4 SCC 225, where a bench of thirteen judges held by a majority of seven to six that the power to amend does not extend to altering the basic structure or framework of the Constitution so as to change its identity. The limit was located in the amending power itself: a power to amend a Constitution is a power to change it while it remains the same Constitution, and a change that destroys its identity is not an amendment of it but a replacement of it. On that reasoning the doctrine is an interpretation of Article 368 rather than an addition to it.
The route to that holding shows why an express textual limit was never available. The Court had held in Shankari Prasad Singh Deo v. Union of India, AIR 1951 SC 458 and again in Sajjan Singh v. State of Rajasthan, AIR 1965 SC 845 that a constitutional amendment is not “law” within Article 13(2) and so cannot be tested against Part III. It reversed that position in I.C. Golaknath v. State of Punjab, AIR 1967 SC 1643, holding fundamental rights unamendable altogether, which left Parliament unable to amend Part III at all. Kesavananda Bharati overruled Golaknath and replaced an absolute bar with a structural one, restoring the power to amend Part III while denying the power to destroy the Constitution’s identity.
The doctrine has since been applied rather than merely stated, which is the strongest answer to the suggestion that it is a judicial invention without content. The Court struck down clause (4) of Article 329A in Indira Nehru Gandhi v. Raj Narain, 1975 Supp SCC 1, for removing judicial review of an election dispute. It struck down Sections 4 and 55 of the Constitution (Forty-second Amendment) Act, 1976 in Minerva Mills Ltd. v. Union of India, (1980) 3 SCC 625, holding that a limited amending power and the balance between Parts III and IV are themselves basic features. It struck down the Constitution (Ninety-ninth Amendment) Act, 2014 in Supreme Court Advocates-on-Record Association v. Union of India, (2016) 5 SCC 1, holding that the National Judicial Appointments Commission damaged the independence of the judiciary.
The reach of the doctrine has been extended as well as applied. The Court held in I.R. Coelho v. State of Tamil Nadu, (2007) 2 SCC 1 that laws placed in the Ninth Schedule after 24 April 1973 remain open to challenge for damaging the basic structure, so Article 31B cannot be used to put a law beyond review.
The statement is accordingly correct in its literal content and wrong in what it implies. The doctrine is judicial in origin and textual in method, and it has been used to strike down constitutional amendments on three separate occasions rather than left standing as a formula. No bench has reopened it since 1973.
When does a debt become Avyavaharik?
The same paper set this as part of its compulsory Hindu law question:
What is Avyavaharik debt? When debt becomes Avyavaharik? Discuss the law regarding such debts.
Uttar Pradesh PCS (J) 2022, LAW (PAPER-I) Substantive Law, Question 2(b), 10 marks
An avyavaharik debt is a debt contracted for a cause repugnant to good morals, and it matters because it is the one exception that releases a son from the pious obligation to pay his father’s debts out of coparcenary property.
The structural point here is the one candidates miss. Three questions have been asked, each carrying roughly a third of the ten marks, and an answer that runs them together as a single essay on pious obligation will be marked as having answered the first and gestured at the other two. Answering in three visible parts, in the order the question puts them, costs nothing and secures the marks allotted to each.
Model answer on avyavaharik debt
What it is: under Mitakshara law a son, grandson and great-grandson carry a pious obligation to discharge the debts of the father, grandfather and great-grandfather out of the coparcenary property. The obligation is religious in origin, and the civil courts enforce it as a liability attaching to their interest in that property. It does not attach at all where the debt is avyavaharik.
Colebrooke rendered that term as a debt for a cause repugnant to good morals. The courts have preferred his rendering to the wider translations, which would have swept in any debt a prudent man would have avoided, and have treated it as the nearest approach to the true conception.
When a debt becomes avyavaharik: the character of the debt is fixed when it is contracted, and the use later made of the money does not decide it. The Supreme Court applied that test in S.M. Jakati v. S.M. Borkar, AIR 1959 SC 282. A father who had served as managing director of a co-operative bank was held liable in misfeasance proceedings for negligence in discharging his duties, and the decree was executed against the family property.
The Court held that a liability so incurred is not avyavaharik, because negligence in performing a duty for which a person draws remuneration cannot be called repugnant to good morals. Mere negligence therefore falls short of the exception, and so does a decree founded on it. What the exception requires is a taint of illegality or immorality in the transaction itself, such as a debt arising from criminal breach of trust, from a fine, or from a liability for a wrong done wilfully.
The law regarding such debts: where the debt is not avyavaharik, a creditor may proceed against the whole of the coparcenary property including the sons’ interest in it. A partition effected after the debt was contracted does not defeat that right, which S.M. Jakati also settled.
That position changed prospectively in 2005. Sub-section (4) of Section 6 of the Hindu Succession Act, 1956 was substituted by the Hindu Succession (Amendment) Act, 2005 with effect from 9 September 2005. It directs that no court shall recognise any right to proceed against a son, grandson or great-grandson for the recovery of a debt due from his father, grandfather or great-grandfather on the ground of the pious obligation alone. Its proviso preserves the creditor’s existing rights, and any alienation made in respect of such a debt, where the debt was contracted before that date.
The doctrine therefore now operates only on debts contracted before 9 September 2005, and within that shrinking class the avyavaharik exception continues to decide real disputes. A son resisting execution against family property for a pre-2005 debt must still show that the debt was tainted when it was incurred, and after S.M. Jakati he cannot do so by showing only that his father was careless.
“Proof of loss is a sine qua non for claiming liquidated damages”
Delhi put the proposition and the qualification it needs into a single twenty-mark question:
“Proof of loss is a sine qua non for claiming ‘liquidated damages’. On proof of damages, the person complaining of breach is entitled, whether or not actual loss is proved, only reasonable compensation not exceeding the amount so named in the contract as liquidated damages or penalty.” Explain with the help of relevant case laws. Whether proof of actual loss is necessary.
Delhi Judicial Service Examination (Written), 2022, Civil Law-I, Part II, Question 7, 20 marks
The proposition is correct as far as it goes, and the twenty marks are in the qualification that the last sentence of the question asks for. A second and narrower question has been tacked on, and an answer that explains the proposition and stops has left roughly a quarter of the marks unclaimed.
Reading the question closely is worth a minute here. The phrase “whether or not actual loss is proved” is lifted from the statute itself, so the examiner has put the statutory words and the proposition side by side and is asking the candidate to reconcile them. A candidate who notices that has found the answer before writing a line.
Model answer on Section 74 of the Indian Contract Act
Loss is a condition of any award under Section 74 of the Indian Contract Act, 1872, and the statutory words “whether or not actual damage or loss is proved” relax the manner of proving loss rather than dispensing with loss itself. The proposition in the question is accordingly correct, and the answer to the further question is that proof of actual loss is necessary wherever loss is capable of being proved.
Section 74 provides that where a contract has been broken and a sum is named in it as the amount to be paid on breach, the party complaining of the breach is entitled, whether or not actual damage or loss is proved to have been caused, to receive reasonable compensation not exceeding the amount so named. The section abolished the English distinction between a genuine pre-estimate of damages and a penalty, so the label the parties used does not decide the claim.
The Supreme Court established the governing approach in Fateh Chand v. Balkishan Dass, AIR 1963 SC 1405, a Constitution Bench decision. It held that the named sum operates as a ceiling on what a court may award rather than as an entitlement to it, and that the court must itself assess what reasonable compensation is. The effect is that naming a figure in the contract settles nothing except the upper limit of the claim.
The landmark on the precise question asked is Kailash Nath Associates v. Delhi Development Authority, (2015) 4 SCC 136, where a bidder’s earnest money was forfeited by the authority after a delay in which the authority itself was partly at fault. Since Section 74 speaks of compensation for loss or damage caused by a breach, the Court held that damage or loss is a sine qua non for the section to apply at all. It read the words “whether or not actual damage or loss is proved” as meaning that proof is dispensed with only where damage or loss is difficult or impossible to prove, and that in such a case the named sum may be awarded if it is a genuine pre-estimate. Where loss is capable of proof, the Court held, proof of it is not dispensed with.
That reading also places the earlier decision in Oil and Natural Gas Corporation Ltd. v. Saw Pipes Ltd., (2003) 5 SCC 705, which had allowed compensation on the named sum without specific proof of loss in a contract where the terms were clear and the loss was difficult to establish. The two decisions sit together once Saw Pipes is confined to that situation, which is the situation the statutory words were written for, and a candidate who presents them as a conflict rather than as a line has misread both.
Applying that to the proposition, a claimant who can prove loss must prove it, and a claimant who genuinely cannot must still show that the sum named was a genuine pre-estimate rather than a figure inserted to intimidate. Forfeiture of earnest money stands apart only where the sum is genuinely earnest and the forfeiture is not compensation at all, and in Kailash Nath Associates the forfeiture failed because no breach by the bidder had been established. The proposition in the question is therefore sound, and proof of actual loss is necessary in every case in which loss admits of proof, which covers most commercial claims. A fuller treatment of the reasoning sits in this analysis of contractual damages under Section 74.
“The test of sovereign immunity has disappeared in determining the tortious liability of State”
Uttar Pradesh set this among its optional questions, with an instruction attached to it:
“The test of sovereign immunity has now been disappeared in determining the tortious liability of State.” Explain this statement with the help of latest judicial pronouncements.
Uttar Pradesh PCS (J) 2022, LAW (PAPER-I) Substantive Law, Question 5(b), 15 marks
The statement overstates the position, and the fifteen marks are in saying by how much. An immunity that has been confined to a narrow residue is not the same as an immunity that has gone, and the difference matters because the decision that established it in India has never been overruled by a larger bench.
The words “latest judicial pronouncements” are an instruction and not decoration. The examiner is asking for the line of authority after 1965 and specifically for what has happened to Kasturi Lal, so an answer that reaches 1965 and stops has failed the instruction even if everything in it is accurate.
Model answer on the tortious liability of the State
Sovereign immunity survives in Indian tort law as a formal defence and has been reduced in operation to a residue that almost no claim now touches, so the statement is correct in substance and inaccurate in its terms. The practical position is that a claimant injured by the negligence of a public servant will recover, and the route by which recovery is allowed has changed more than once.
Article 300(1) of the Constitution provides that the Government of India may sue and be sued, subject to any provision made by Parliament, in the same cases as the Dominion of India and the corresponding Provinces might have sued or been sued had the Constitution not been enacted. Liability is therefore traced back to the pre-Constitution position rather than defined afresh, and that is why a judgment of 1861 still has to be dealt with.
The Supreme Court of Bombay drew the original distinction in P&O Steam Navigation Co. v. Secretary of State for India, (1861) 5 Bom HCR App 1. It held the East India Company liable for acts done in the conduct of undertakings that a private person might also carry on, and not liable for acts done in the exercise of sovereign powers.
Both halves of that distinction were applied within three years of each other. The Supreme Court held the State vicariously liable in State of Rajasthan v. Vidhyawati, AIR 1962 SC 933, where a government jeep driven negligently by a driver taking it back from repairs killed a pedestrian. It reached the opposite result in Kasturi Lal Ralia Ram Jain v. State of Uttar Pradesh, AIR 1965 SC 1039, on facts where gold seized by the police was misappropriated by a head constable who then absconded. The claim failed because the seizure had been made under statutory powers referable to the sovereign power of the State.
What the question means by latest pronouncements is a decision of 1994, which dismantled the Kasturi Lal position rather than distinguishing it away. The Supreme Court decided N. Nagendra Rao & Co. v. State of Andhra Pradesh, (1994) 6 SCC 205 on facts in which stocks of fertiliser and foodgrain seized under the Essential Commodities Act, 1955 had been allowed to deteriorate in the custody of the State.
The Court held that barring the primary and inalienable functions of a constitutional government, such as the administration of justice, the maintenance of law and order and the repression of crime, the State cannot claim immunity. A seizure under a welfare statute is not one of those functions. It added that no rational basis exists for holding an officer personally liable for negligence while denying a remedy against the State that employed him.
Two later decisions complete the line of authority the question asks for. The Court recorded in Common Cause, A Registered Society v. Union of India, (1999) 6 SCC 667 that Kasturi Lal has lost its relevance, without formally overruling it. Alongside the tort claim there runs a public law remedy in which the defence is unavailable altogether, established in Rudal Sah v. State of Bihar, (1983) 4 SCC 141 and stated in terms in Nilabati Behera v. State of Orissa, (1993) 2 SCC 746, where compensation for a custodial death was awarded under Article 32 on the footing that sovereign immunity has no application to a claim for enforcement of a fundamental right. The Court applied that reasoning to a non-citizen in Chairman, Railway Board v. Chandrima Das, (2000) 2 SCC 465.
The accurate conclusion is that the test has not disappeared but has been narrowed to acts of State properly so called, and that in any event a claimant can frame the claim as a constitutional tort under Articles 32 or 226, where the defence does not arise. Because Kasturi Lal has not been overruled by a bench of larger strength, it remains formally binding on its own facts, and stating that qualification is what distinguishes a full answer from a confident one.
Answering a partly correct proposition
Two of the four questions worked through here put a proposition that is true in part, which is not a coincidence, because that is how examiners test judgment rather than recall. The method is the same in both. Say in the first line which part is accepted and which is disputed, give the accepted part its due in two or three sentences, and spend the remaining time on the disputed part with authority attached.
What loses marks is agreeing with the proposition wholesale in order to avoid taking a position. An answer that says sovereign immunity has disappeared, cites Nagendra Rao and stops has taken the examiner’s bait, because the question was set precisely to see whether the candidate knows that Kasturi Lal still stands unreversed.
Citation of law after the 2024 criminal codes
Cite the provision that was in force on the date of the offence, lead with the new section, and put the old equivalent in brackets on its first appearance. That single habit answers the question a candidate cannot otherwise resolve in the hall, which is whether a 2026 paper wants Bharatiya Nyaya Sanhita numbering or Indian Penal Code numbering, and the answer is that it depends on the facts the question gives.
The reason is that the changeover was not retrospective. Offences committed before 1 July 2024 continue to be tried under the Indian Penal Code, and proceedings already pending on that date continue under the Code of Criminal Procedure by virtue of the repeal and savings provision in the Bharatiya Nagrik Suraksha Sanhita. A problem question dated 2019 that is answered entirely in Sanhita numbering is as wrong as a 2025 fact pattern answered entirely in Penal Code numbering, and the examiner is a serving officer running both codes in the same courtroom on the same morning.
One sourcing habit costs marks invisibly. Several bare-act aggregator websites still serve the text of the 2023 Bills rather than the enacted provisions, and the numbering and wording differ between the two. A candidate who memorised section numbers from such a page has memorised a draft that was never law, so section numbers are worth checking against the gazette text or a court’s own citation of the provision. For the procedural sequence a question is likely to be built on, this account of the stages of a criminal trial under the BNSS sets out the order the new Sanhita follows.
Provisions in both numbers
A working candidate does not need the whole concordance and does need the provisions that questions are actually built on. Murder is Section 103 of the Bharatiya Nyaya Sanhita, 2023, which replaces Section 302 of the Indian Penal Code, and culpable homicide not amounting to murder is Section 105, replacing Section 304. Theft moves from Sections 378 and 379 to Section 303, and cheating from Sections 415 and 420 to Section 318. A fuller comparison of the BNS against the IPC covers the provisions that changed in substance rather than only in numbering.
On the procedural side, the registration of a first information report moves from Section 154 of the Code of Criminal Procedure to Section 173 of the Bharatiya Nagrik Suraksha Sanhita, 2023. A statement to the police moves from Section 161 to Section 180, and the examination of the accused from Section 313 to Section 351. In the law of evidence, the general burden of proof moves from Section 101 of the Indian Evidence Act, 1872 to Section 104 of the Bharatiya Sakshya Adhiniyam, 2023, and Section 63 of that Adhiniyam replaces Section 65B on the proof of electronic records, with Sections 61 and 62 placed in front of it as enabling provisions.
One renumbering converts a familiar number into a wrong answer. Anticipatory bail now sits at Section 482 of the Sanhita, in place of Section 438 of the old Code, while the inherent powers of the High Court that every candidate knows as Section 482 of the Code have moved to Section 528. A script that cites Section 482 without naming the statute it belongs to is therefore ambiguous between two unrelated provisions, and writing “Section 482 of the BNSS” rather than “Section 482” costs three words.
Presentation, time management and common errors
The marks most candidates lose have nothing to do with how much law they know. Six errors account for most of it, and each of them is fixable in practice before the paper rather than in the hall during it.
Not reading the instruction line is the first, and the Uttar Pradesh rule that attempts are counted in sequential order punishes it directly. A candidate who answers seven questions where six are required has donated one answer, because the extra one is not marked in place of a weaker attempt. The blank pages that were never struck off do the same damage more quietly.
The long introduction on a short question is the second. Three of nine minutes spent restating the question in longer words leaves six minutes for the definition, the provision, the authority and the conclusion, and the answer then runs out of time at exactly the point where the marks were. The third is the case name with nothing attached to it, which remains the most common single defect in these scripts.
The fourth is a conclusion that summarises rather than decides. A problem question asks who wins, and an answer closing with a survey of the arguments has declined to answer it. The fifth is overrunning, which converts a strong answer into a weak later one. The sixth is failing to number sub-parts where the question was numbered, because an examiner who cannot find the answer to part (b) will not hunt for it inside a continuous page.
Use of the reading time
Both papers give reading time before the booklet may be opened, and it is not spare time. Read every question first and mark the ones that will be attempted, because the choice structure means the selection itself carries marks. A candidate who starts writing on the first question and discovers the fifth is stronger has already spent the time that would have been used on it.
Within the marked questions, note the command words. “Decide” requires an outcome, “discuss” requires both sides, “explain” requires the mechanism, and “distinguish” requires a stated basis of comparison rather than two descriptions side by side. Note also where a question has hidden a second question in its last sentence, as the Delhi liquidated damages question does.
Attempting a question you do not know
Attempt it anyway, and the Delhi paper says so in terms, because the test is of analytical reasoning as much as of recall. A candidate who identifies the issue correctly, names the statute even without the section number, states a principle in their own words and reaches a defensible conclusion will collect a working proportion of the marks on a question they could not have answered from memory.
What not to do is leave a blank page or write an answer that hedges every line. An examiner can award marks for reasoning that reaches the wrong destination and cannot award anything for a page that declines to travel.
Frequently asked questions
How long should a 10-mark judiciary mains answer be?
Roughly 250 to 300 words, which is about a page and a half in average handwriting and nine minutes of writing time in a paper of 200 marks over three hours. The figure matters less than having written to it in practice, because the point of a target is to recognise an overrunning answer without checking a watch.
Do I have to cite case law in every mains answer?
Not in every answer, but in every answer where a decided case supplies the rule. The Delhi Judicial Service paper instructs candidates to support each answer with reasons, relevant legal provisions and principles and also relevant case law, and one landmark stated with its principle and applied to the facts is worth more than a list of names.
Should I write in points or in paragraphs?
Paragraphs for reasoning and points for anything genuinely enumerable, such as the conditions of a test or the ingredients of an offence. An answer written entirely in bullet points gives the examiner no reasoning to mark, which is half of what the paper is testing.
What happens if I do not know the answer to a question?
Attempt it anyway, because the paper awards marks for analytical reasoning independently of recall. Identify the issue, name the statute, state the principle in your own words and reach a conclusion on it. A blank page earns nothing, and a reasoned answer that lands in the wrong place usually earns something.
Do I cite IPC or BNS sections in a 2026 mains paper?
Cite the law in force on the date of the offence in the question, lead with the new section and give the old equivalent in brackets on first use. Offences before 1 July 2024 are still tried under the Indian Penal Code, and proceedings pending on that date continue under the Code of Criminal Procedure.
Is judgment writing marked differently from a problem question?
It is marked on the same reasoning but in a fixed judicial form, being the issues as framed, the finding on each, the reasons and an operative order. Framing the issues wrongly costs the whole answer rather than part of it, which is why Chhattisgarh’s 2024 paper asked candidates to frame the necessary issues before writing the judgment.
References
- Uttar Pradesh Public Service Commission, LAW (PAPER-I) Substantive Law, PCS (J) Main Examination 2022, Everest-54, Serial No. 2500203 (sections cited: Questions 1(c), 2(b) and 5(b), and the Specific Instructions and Note).
- Delhi Judicial Service Examination (Written), 2022, Civil Law-I, Parts I to III, and General Legal Knowledge and Language paper (Important Instructions).
- Chhattisgarh Judicial Service Examination (Mains) 2024, Question 1.
- Kesavananda Bharati v. State of Kerala, (1973) 4 SCC 225.
- Minerva Mills Ltd. v. Union of India, (1980) 3 SCC 625.
- I.R. Coelho v. State of Tamil Nadu, (2007) 2 SCC 1.
- Supreme Court Advocates-on-Record Association v. Union of India, (2016) 5 SCC 1.
- S.M. Jakati v. S.M. Borkar, AIR 1959 SC 282.
- Section 6(4), Hindu Succession Act, 1956, as substituted by the Hindu Succession (Amendment) Act, 2005.
- Fateh Chand v. Balkishan Dass, AIR 1963 SC 1405.
- Kailash Nath Associates v. Delhi Development Authority, (2015) 4 SCC 136.
- Oil and Natural Gas Corporation Ltd. v. Saw Pipes Ltd., (2003) 5 SCC 705.
- Kasturi Lal Ralia Ram Jain v. State of Uttar Pradesh, AIR 1965 SC 1039.
- N. Nagendra Rao & Co. v. State of Andhra Pradesh, (1994) 6 SCC 205.
- Nilabati Behera v. State of Orissa, (1993) 2 SCC 746.
- Ministry of Home Affairs notifications bringing the Bharatiya Nyaya Sanhita, 2023, the Bharatiya Nagrik Suraksha Sanhita, 2023 and the Bharatiya Sakshya Adhiniyam, 2023 into force on 1 July 2024.
This article is for informational and educational purposes only and does not constitute legal advice. Examination patterns, syllabi and marking schemes are set by the relevant High Court or State Public Service Commission and change between cycles, so candidates should verify the current notification for the state they are applying to.