The Bharatiya Sakshya Adhiniyam 2023 is the statute that replaced the Indian Evidence Act, 1872, and the first thing to understand about it is how little of the law of evidence it actually changed. On 1 July 2024 the Adhiniyam came into force as Act 47 of 2023, and the Evidence Act, which had governed what Indian courts may receive as proof and how they must reason from it for a century and a half, stood repealed. Yet a lawyer who opens the new statute looking for a new law of evidence will mostly find the old one, renumbered. The relevancy scheme is the scheme of 1872. The confession rules are the confession rules of 1872, compressed into fewer sections. The dying declaration, the burden of proof, the presumptions, the law of witnesses and their examination all crossed over in substance. What genuinely changed is concentrated in one place, the treatment of electronic records, and even there the change is double-edged: electronic records were written into the heart of the Act as documents in their own right, and at the same time the certificate requirement that had generated a decade of appellate litigation was carried across, given a prescribed form, and made in one respect harder to satisfy than before.

This article separates the three things the Adhiniyam did, because accounts that collapse them into a single claim misdescribe the statute. The first was an act of retention on a very large scale, in which the doctrines worked out under the 1872 Act were carried forward under new section numbers. The second was a small set of genuine changes, of which the electronic records provisions are the largest and a handful of others, on absconding co-accused, on accomplice corroboration, and on the definition of secondary evidence, are worth knowing precisely. The third was an act of renumbering, and renumbering is not free. Where a successor provision reproduces its predecessor in substance, every existing judgment remains good law, but every citation has to be translated, so the immediate cost of the new evidence statute is a decade of double citation in pleadings, in textbooks, and in examination answers. That cost, which this article calls the renumbering tax, is paid by the profession whether or not the reform delivers anything in return. The honest assessment of the Adhiniyam turns on whether the electronic records changes are worth that tax, and on whether the certificate problem that the old law bequeathed has been solved or merely re-enacted in a new number.

The Act That Ended: What the Indian Evidence Act 1872 Did

The Indian Evidence Act, 1872 was drafted principally by James Fitzjames Stephen and brought into force on 1 September 1872. It was an unusual statute even by the standards of nineteenth-century codification, because it did not merely collect rules about proof; it attempted to state, in 167 sections arranged in eleven chapters and three parts, a complete theory of how a court should move from evidence to conclusion. Its first part dealt with the relevancy of facts, opening with the proposition that evidence may be given of facts in issue and of facts declared relevant, and then working through the connected facts, admissions, confessions, statements by persons who cannot be called as witnesses, opinions, and character. Its second part dealt with proof: facts that need not be proved, oral evidence, documentary evidence, and the exclusion of oral evidence by documentary evidence. Its third part dealt with the production and effect of evidence: the burden of proof, estoppel, witnesses, the examination of witnesses, and the consequences of improper admission or rejection. That three-part architecture, relevancy, proof, and production, is the architecture the Adhiniyam inherited, and it survived the repeal almost intact.

The design had two features that explain its longevity. The first is that Stephen codified principles rather than instances. The relevancy provisions state tests, about connection, about motive and conduct, about the circumstances of a transaction, that courts could apply to kinds of evidence the drafters never imagined, and the statute absorbed photography, sound recording, and eventually computer output without structural amendment for well over a century. The second is that the Act was a rulebook for reasoning under conditions of distrust. Its confession provisions embody a settled suspicion of the police custodial environment. Its presumption provisions allocate the risk of missing proof where experience said the risk should fall. Its burden provisions state, with unusual clarity, who must prove what and what happens when nobody does. A statute of that kind does not date in the way a statute of offences dates, because human proof problems change slowly even when technology changes quickly.

What the 1872 Act never contained, in its original form, was any treatment of electronic records, for the sufficient reason that there were none. The definitions with which it worked, document, evidence, proved, disproved, were framed for a world of paper, of oral testimony, and of things produced for the inspection of the court. When electronic records arrived in litigation, the Act had to be adapted to them, and the adaptation took the form, in the year 2000, of amendments carried by the Information Technology Act, 2000, which inserted a special regime for electronic records into a statute that had been complete without one. The full account of the old Act, its structure and its long judicial life belongs to the guide to the Indian Evidence Act 1872, and is summarised here only so far as the replacement story requires it. The point that matters for that story is the shape of the inheritance. The Adhiniyam did not replace a statute that had failed across the board. It replaced a statute whose central doctrines were in daily working order, and whose one visibly strained region was the annexe built for electronic records in the year 2000.

The Strain That Showed: Electronic Records Under the Old Act

The amendments of the year 2000 did two things. They adjusted the definitions so that electronic records could count as documents and as evidence, and they inserted two provisions, Section 65A and Section 65B, which created a special route by which the contents of electronic records could be proved. Section 65A provided that the contents of electronic records could be proved in accordance with Section 65B, and Section 65B laid down conditions under which a computer output, a printout or a copy stored on a medium, would be deemed a document and admitted without production of the original, provided the computer and the process satisfied stated conditions about regular use, regular feeding of information, proper operation, and faithful reproduction, and provided a certificate identified the record and the device and spoke to those conditions. The scheme was bolted on rather than integrated: electronic records were documents for most purposes, yet their proof was governed by a self-contained code sitting inside the documentary evidence chapter, and the boundary between the special route and the ordinary law of primary and secondary evidence was never drawn cleanly in the text. The wider statutory setting of those amendments is traced in the guide to the Information Technology Act 2000, which owns that statute for this series.

The drafting choice might have mattered little if the certificate requirement had been applied with a steady hand. It was not, and the resulting decade of appellate litigation is the practical pressure that produced the corresponding provisions of the Adhiniyam. The instability began before the strict reading took hold. In the early years of the special regime, the Supreme Court, in the Parliament attack case, State (NCT of Delhi) v Navjot Sandhu, treated electronic records as documents receivable under the ordinary provisions, and a line of decisions followed that treated the certificate as one mode of proof among others rather than as a gate. That permissive period is what Anvar brought to an end in 2014, and the swing between the two positions, permissive and strict, is the pattern that repeated when Shafhi Mohammad relaxed the requirement in 2018 and Arjun Panditrao restored it in 2020. Three positions in two decades, each stated by the final court, is the litigation history the Adhiniyam inherited, and it is the reason the certificate question cannot be discussed as though the old provision had simply worked or simply failed. It did both, in sequence, at the hands of its interpreters. In Anvar P.V. v P.K. Basheer, decided in 2014, the Supreme Court held that Section 65B was a complete and special code for electronic evidence, that its conditions were mandatory, and that an electronic record produced as secondary evidence without the certificate was inadmissible, departing from an earlier and looser approach that had treated electronic records as ordinary documents. In Shafhi Mohammad v State of Himachal Pradesh, decided in 2018, a smaller bench relaxed the requirement for a party who was not in possession of the device from which the record came, reasoning that a litigant cannot certify the working of a machine it does not control. In Arjun Panditrao Khotkar v Kailash Kushanrao Gorantyal, decided in 2020 on a reference, the Court restored the strict position: the certificate was a condition precedent to admissibility, the contrary relaxation was overruled, and the judgment in Anvar was reaffirmed, while the Court also held that a defective or missing certificate was a curable defect, that the certificate could be produced at a later stage of the proceedings, and that a party genuinely unable to obtain one could seek the assistance of the court, which could summon the record and the person able to certify it. Those holdings belong, in their full analysis, to the article on the 1872 Act linked above; they are stated here, in compressed form, because they define the problem the drafters of the Adhiniyam inherited. By the early years of the decade in which the replacement was drafted, the law on the commonest kind of evidence in modern litigation was a strict statutory condition, softened by judge-made safety valves, whose exact operation a practitioner could state only by reciting three Supreme Court decisions in sequence.

The stakes of that instability were not academic. Electronic records are rarely decorative evidence; they are characteristically the evidence on which the contested fact turns, the call record that places a phone, the recording that captures the demand, the log that fixes the transaction. A rule whose breach excludes the record therefore does not trim a case at its edges; it decides it, and a rule whose interpretation swings between benches converts the merits of prosecutions and claims into a lottery on admissibility. Trials were lost and convictions were reopened on certificate defects in cases where the genuineness of the record was not seriously in issue, and other records were received without certificates in cases where it was, and both outcomes were available in the reports within a few years of each other. That is the record the drafters of the Adhiniyam confronted, and it explains both the decision to legislate on electronic records at all and the hardest question about what they enacted, which is whether re-enacting the condition in a stricter form addresses a failure that was, on the evidence of the decade, a failure of application rather than of text.

Two conclusions follow, and the assessment of the Adhiniyam depends on keeping them apart. The first is that the difficulty was at least as much adjudicative as statutory. Section 65B said, in terms, that the certificate was required; the instability came from courts alternately enforcing and relaxing that requirement as its hardship in particular cases became apparent. A legislature that replaced Section 65B with a provision repeating its substance would therefore not, without more, remove the instability, because the hardship cases would recur and would press for the same relaxations. The second conclusion is that there were nevertheless genuine statutory defects to repair. The old Act treated the electronic record as an annexed special case a century and a quarter after its enactment, its definitions did not place electronic records inside the concept of a document without argument, it prescribed no form for the certificate, leaving its contents to be reconstructed from the conditions in the section, and it said nothing systematic about records stored across multiple files and devices, which is how records are in fact kept. A replacement statute could plausibly fix those four defects. Whether the Adhiniyam did so, and what it did instead, is the subject of the sections that follow.

From Introduction to Commencement: The Passage of the Evidence Bill

The Adhiniyam was the third of the three criminal law codes replaced in a single exercise, and its passage record is the same record as that of its companions, stated here in brief because the detail belongs to the companion articles. The Bharatiya Sakshya Bill, 2023 was introduced in the Lok Sabha on 11 August 2023 by the Home Minister, Amit Shah, together with the bills that became the Bharatiya Nyaya Sanhita and the Bharatiya Nagarik Suraksha Sanhita. The three bills were referred to the Parliamentary Standing Committee on Home Affairs, which reported on the evidence bill on 10 November 2023. The first set of bills was then withdrawn, and second bills, incorporating some of the changes the committee stage had produced, were introduced on 12 December 2023. The Bharatiya Sakshya (Second) Bill was passed by the Lok Sabha on 20 December 2023 and by the Rajya Sabha on 21 December 2023, and it received the assent of the President on 25 December 2023, becoming Act 47 of 2023. Its provisions were brought into force on 1 July 2024, the commencement date shared by all three codes, and on that date the Indian Evidence Act, 1872 stood repealed.

The government’s stated objects ran on two tracks that the rest of this article keeps separate, because the statute answers them very differently. The first was presentational and constitutional in character: the three codes were described, in the Home Minister’s presentation of them, as a casting off of laws made in the colonial period for the governance of subjects, and their replacement by laws made by the Parliament of a republic for its citizens. The second was technical: the law of evidence was to be modernised for a digital era, in which the typical exhibit is no longer a paper deed but a phone, a server log, a message thread, or a recording held by an intermediary. The first object explains the new name, the new numbering, and the determination to replace rather than amend. The second object is the one by which the text should be judged, because it identifies the region of the old Act that was actually under strain.

The process drew criticism that belongs in the record of this statute as much as in the record of its companions. The drafting was done within government, the bills were introduced without a prior public draft, and the December 2023 passage of the three codes took place while 146 members of the opposition stood suspended from the two Houses, a figure recorded in the companion articles of this series, so that the debate on the floor was conducted in the absence of much of the membership elected to conduct it. Defenders of the process answer that the bills had been examined by the Standing Committee, that the second bills incorporated committee-stage changes, and that the codes had been in gestation since the constitution of a committee for reforms in criminal laws in 2020. Critics answer that a committee’s examination of a bill it did not draft, against a deadline it did not set, is not a substitute for pre-legislative consultation on codes that govern every courtroom in the country. Both positions are stated here because the passage record is part of what a replacement statute is: the Adhiniyam’s text is the product of that process, and its drafting faults and its strengths are equally its products.

What did the Parliamentary Standing Committee say about the evidence bill?

The Committee examined the bill and reported on 10 November 2023. It recommended that secure handling and a proper chain of custody be mandated for electronic records gathered in investigation, it found that Part A of the certificate form in the Schedule did not fully track the conditions for admissibility stated in the bill, and it advised improving that form.

The committee stage mattered for this statute in a more specific way than for its companions, because the Committee engaged with the central drafting difficulty of the electronic records provisions. It recorded the overlap between two ideas the bill contained at once: that electronic records produced from proper custody are primary evidence, and that electronic records are admissible on the certificate route. Two routes to the same exhibit, stated in different chapters without a stated priority, invite the argument that the certificate is optional whenever the record can be characterised as primary. The Committee observed that overlap and recommended adherence to the admissibility provision, the certificate route, for electronic records. It also heard, and endorsed, the submission that the bill said nothing about the chain of custody of digital evidence, and it recommended that a provision mandating secure handling and proper chain of custody be inserted, pointing to the procedure code as its proper home. On the certificate form, its criticism was concrete: Part A of the Schedule, the part to be completed by the person in charge of the device, did not call for a declaration covering all the conditions that the admissibility provision itself laid down, including the condition of the device. The second bill revised the certificate arrangements, and the enacted Section 63 requires a certificate signed both by the person in charge of the device or activity and by an expert. How far that revision answered the Committee, and how far it raised the cost of the certificate instead, is examined in the section on the certificate question below.

The Shape of the Bharatiya Sakshya Adhiniyam: One Hundred and Seventy Sections and a Schedule

The Bharatiya Sakshya Adhiniyam contains 170 sections, against the 167 of the Act it replaced, arranged in twelve chapters and four parts, and followed by a Schedule, which the 1872 Act never had. The parts preserve Stephen’s architecture under new labels: Part I is preliminary, Part II covers the relevancy of facts in Sections 3 to 50, Part III covers proof in Sections 51 to 103, and Part IV covers the production and effect of evidence in Sections 104 to 170. Within that familiar frame the internal reordering is modest but real. The definitions, which the old Act had placed in an interpretation clause at Section 3 after the relevancy provisions had already begun, were moved to the front as Section 2, so that the statute opens with its concepts, as a modern code is expected to do. The burden of proof chapter occupies Sections 104 to 120, estoppel Sections 121 to 123, witnesses Sections 124 to 139, and the examination of witnesses Sections 140 to 168, before the closing provisions on improper admission and rejection in Section 169 and on repeal and savings in Section 170.

Two of the opening provisions carry more weight than their position suggests. Section 1 applies the Adhiniyam to all judicial proceedings in or before any court, which is the same field the old Act occupied, and it preserves the two traditional exclusions: the Act does not apply to affidavits presented to a court or officer, and it does not apply to proceedings before an arbitrator. Section 2 then does the definitional work on which the electronic records changes rest. A document, by Section 2(1)(d), means any matter expressed, described or otherwise recorded on any substance by means of letters, figures or marks, or by more than one of those means, and the definition states expressly that it includes electronic and digital records. Evidence, in turn, is defined to cover documents including electronic records produced for the inspection of the court. The consequence is structural rather than cosmetic. Under the old Act, the electronic record had been brought within the definition of a document by amendment and then fenced off in a special proof regime; under the Adhiniyam, the electronic record is a document from the first page, and the special provisions that remain, Sections 61 to 63, operate inside the documentary evidence chapter rather than as an annexe to it.

Does the Bharatiya Sakshya Adhiniyam apply to affidavits and arbitration?

No. Section 1 of the Adhiniyam applies it to judicial proceedings in or before any court, and preserves the exclusions the old Act carried: affidavits presented to a court or an officer sit outside it, and proceedings before an arbitrator sit outside it, exactly as they did under the 1872 Act.

A chapter walk fixes the map in the mind. Chapter I, Sections 1 and 2, is preliminary: application and definitions. Chapter II, Sections 3 to 50, is the relevancy chapter, the longest in the Act, carrying the connected facts, admissions, confessions, the statements of persons who cannot be called, statements under special circumstances, the relevance of judgments, opinions, and character. Chapter III, Sections 51 to 53, states the facts that need not be proved, the facts judicially noticeable and the facts admitted. Chapter IV, Sections 54 and 55, is oral evidence in two sections, the proof of facts by oral evidence and the requirement that oral evidence be direct. Chapter V, Sections 56 to 93, is documentary evidence, and it is the chapter the replacement worked on hardest: proof of contents, primary and secondary evidence, the electronic records provisions at Sections 61 to 63, public documents, and the presumptions as to documents. Chapter VI, Sections 94 to 103, excludes oral evidence where documentary evidence governs, carrying the rule that the terms of a contract, grant or disposition reduced to writing are proved by the document itself. Chapter VII, Sections 104 to 120, is the burden of proof. Chapter VIII, Sections 121 to 123, is estoppel. Chapter IX, Sections 124 to 139, is witnesses. Chapter X, Sections 140 to 168, is the examination of witnesses. Chapter XI is a single section, Section 169, on improper admission and rejection of evidence, and Chapter XII is Section 170, repeal and savings. A reader who holds that skeleton holds the Act: twelve chapters, of which one, the fifth, contains nearly all the novelty, and one, the second, contains nearly all the doctrine.

The definitions in Section 2 also carry, undisturbed, the statute’s vocabulary of proof, and it is worth pausing on that vocabulary because the whole Act speaks in it. A fact is proved when, after considering the matters before it, the court either believes it to exist or considers its existence so probable that a prudent person ought, under the circumstances of the particular case, to act upon the supposition that it exists. A fact is disproved when the court believes it does not exist or considers its non-existence so probable that a prudent person ought to act on that supposition, and a fact is not proved when it is neither proved nor disproved. Around those three states the definitions arrange the presumptions: where the Act provides that the court may presume a fact, the court may either regard the fact as proved unless and until it is disproved, or call for proof of it; where it provides that the court shall presume a fact, the court shall regard it as proved unless and until it is disproved; and where the Act directs that one fact is conclusive proof of another, the court shall, on proof of the one, regard the other as proved and shall not allow evidence to disprove it. That machinery, which the old interpretation clause stated and the Adhiniyam moved forward as definitions, is the grammar in which every later provision is written, and its survival verbatim in substance is the deepest layer of the continuity this article describes: the replacement did not alter what it means, in an Indian court, for a fact to be proved.

The exclusion of affidavits is older than it looks and frequently misunderstood. Affidavit evidence is received in Indian courts under the procedure codes and under specific statutory powers, and its reception has never depended on the evidence statute; the exclusion simply records that the Adhiniyam is a code for the proof of facts in judicial proceedings, not a code governing every sworn paper in the legal system. The arbitral exclusion matters more in practice, because commercial parties sometimes assume that a tribunal is bound by the technical rules of evidence. It is not, and the Adhiniyam preserves the position: a tribunal may apply the statute’s principles by analogy or by agreement, but the Act does not bind it of its own force. Both exclusions are continuity, not change, and they are flagged here because commencement publicity for the three codes encouraged the assumption that every corner of proof had been rewritten. In this corner, nothing was.

The First Move: What the New Act Kept

The retention is best grasped doctrine by doctrine, because the claim that most of the law of evidence survived is either a precise claim about particular provisions or it is nothing. The relevancy scheme survived entire. Section 3 opens the subject as Section 5 of the old Act did, by confining evidence to facts in issue and to facts the Act declares relevant, and the closely connected facts follow in the same order and to the same effect: the facts forming part of the same transaction in Section 4, which is the res gestae provision of the old Section 6; the facts that are the occasion, cause or effect of facts in issue; motive, preparation and conduct in Section 6, the successor of the old Section 8; the facts necessary to explain or introduce relevant facts; and the statements and acts of conspirators, which close the group. Admissions occupy Sections 15 to 21 in the new numbering, reproducing the scheme of the old Sections 17 to 23, with the same distinction between admissions, which are relevant and may be proved against their maker, and confessions, which are admissions by an accused person and are then fenced about by the protective provisions that follow.

The confession provisions survived in substance and were compressed in form, a combination examined in detail in a later section; what matters for the retention account is that every operative rule a practitioner knows is still there. A confession caused by inducement, threat, coercion or promise from a person in authority, in circumstances of the kind the provision describes, is irrelevant. A confession made to a police officer may not be proved against the accused. A confession made in police custody may not be proved except in the immediate presence of a Magistrate. The discovery exception survives, so that so much of the information given by an accused in custody as relates distinctly to a fact discovered in consequence of it may be proved. And the distinct regime for recording confessions, which lives in the procedure code and sits at Section 183 of the Bharatiya Nagarik Suraksha Sanhita, was left where it was. The doctrinal superstructure the courts built on these provisions, including the line of authority on the discovery exception from Pulukuri Kottaya onwards, was therefore carried into the new Act undisturbed, because the text on which it rests was carried in undisturbed.

The statements of persons who cannot be called as witnesses survived in Section 26, the successor of the old Section 32, and with them the dying declaration in the opening clause of that section: a statement by a person as to the cause of that person’s death, or as to the circumstances of the transaction that resulted in death, is relevant when the cause of death comes into question, whether or not the maker expected death when making it. The law on judgments of courts when relevant, on opinions of third persons, and on character, occupies the same relative ground in Sections 34 to 50 as it did in the old Sections 40 to 55. Expert opinion sits at Section 39, the successor of the old Section 45, so that the opinion of a person specially skilled in foreign law, science, art, or the identity of handwriting or finger impressions remains a relevant fact, and Section 39(2) carries forward the opinion of the Examiner of Electronic Evidence, the office created under Section 79A of the Information Technology Act, 2000, which the old Section 45A had recognised. The treatment of proof survived in the same way: facts that need not be proved, the directness requirement for oral evidence, the proof of documents by primary and secondary evidence, the public documents provisions, the presumptions as to documents in Sections 78 to 93, and the exclusion of oral evidence by documentary evidence in Sections 94 to 103 all reproduce the old scheme, with the electronic records provisions, the one region of genuine change, embedded inside it.

The opinion provisions repay a closer look, because they show the retention operating at the level of the Act’s internal economy. Opinions of third persons occupy Sections 39 to 45, as they occupied Sections 45 to 51 of the old Act, and the sequence within them is unchanged: the expert’s opinion at Section 39; the facts bearing on that opinion; the opinion as to handwriting and signature; the opinion as to the existence of a general custom or right; the opinion on usages and tenets; the opinion on relationship; and the grounds of opinion where the opinion is relevant. Character evidence follows at Sections 46 to 50, with the old asymmetry preserved between civil cases, where character is relevant only so far as it appears from facts otherwise relevant and for the assessment of damages, and criminal cases, where the good character of the accused is relevant and previous bad character is relevant only in the stated circumstances, ordinarily in reply. The relevance of judgments, Sections 34 to 38, preserves the careful distinctions the old Sections 40 to 44 drew: a judgment that bars a second proceeding operates as conclusive proof of what it bars; judgments on matters of a public nature are relevant but not conclusive; judgments are otherwise relevant only where they constitute a fact in issue or fall within the relevancy provisions; and a judgment obtained by fraud or collusion may be shown not to have been delivered by a competent court or to have been obtained without authority. None of this was rethought, and none of it needed to be: these are the provisions through which the Act disciplines second-hand material, and their intricate settlement is precisely the kind of accumulated doctrine that a replacement preserves by carrying the text.

The burden and presumption provisions survived as a block. Chapter VII of the Adhiniyam, Sections 104 to 120, opens at Section 104 with the general rule of the old Section 101, that whoever desires a court to give judgment as to a legal right or liability dependent on facts must prove that those facts exist, and it proceeds through the particular allocations the old Act made: the burden as to a particular fact, the burden of proving a fact that makes evidence admissible, the burden in particular cases including the fact especially within a person’s knowledge, the presumptions about a person alive or dead, the presumption of legitimacy in Section 116, and the two presumptions the criminal law relies on most heavily, the presumption as to abetment of suicide by a married woman in Section 117 and the presumption as to dowry death in Section 118, which are the old Sections 113A and 113B under new numbers. Section 119 carries the old Section 114, the general power of the court to presume the existence of facts it thinks likely to have happened, with its illustrations, and Section 120 carries the old Section 114A on the absence of consent in prosecutions for specified sexual offences. The documentary presumptions survived in the same fashion, including the presumption for documents thirty years old produced from proper custody in Section 92, the successor of the old Section 90, and the presumption for electronic records five years old in Section 93, the successor of the old Section 90A. Estoppel sits at Section 121, witnesses and their compellability, competence, and privilege in Sections 124 to 139, and the examination of witnesses, including the Judge’s power to put questions in Section 168, closes the substantive text.

The burden chapter also deserves its working statement, because its sections are the ones a trial court recites daily and their numbering is the numbering practitioners will convert most often. Section 104 states the general rule. Section 107 places the burden of proving any fact necessary to be proved in order to enable a party to give evidence of another fact on the party who wishes to give it, which is the provision that governs the proof of the conditions of admissibility themselves, including, in an Adhiniyam trial, the conditions of Section 63. Section 108 places on an accused the burden of proving that the case falls within a general or special exception, or within a proviso, of the penal law relied on, with the court presuming the absence of the excepting circumstances; it is the old Section 105, and it is the load-bearing provision of every plea of private defence, insanity, or grave and sudden provocation. Section 109, the old Section 106, places the burden of proving a fact especially within a person’s knowledge on that person, the provision that decides what a householder must explain about a death in the house and what a traveller must explain about a journey taken without a ticket. Around those four, the chapter arranges the presumptions and the particular burdens, on proof of life and death, on ownership, on the good faith of transactions between parties in a position of active confidence, and on the remaining allocations the old Act made. Every one of those allocations survived, and the litigation they generate, which is considerable, will be conducted in the new numbering on the old authorities without a break. A reader who knows the old Act can find every one of these doctrines in the Adhiniyam within minutes. That is the retention, and it is the largest single fact about the statute.

The Certificate Question: What Happened to Section 65B

The centre of the Adhiniyam, and the provision by which its claim to have modernised the law of evidence will stand or fall, is the pair of sections that replaced Sections 65A and 65B of the old Act. Section 62 provides, as the old Section 65A did, that the contents of electronic records may be proved in accordance with Section 63. Section 63 then states the admissibility rule. Information contained in an electronic record which is printed on paper, stored, recorded or copied in optical or magnetic media, or in semiconductor memory, and produced by a computer or a communication device, is deemed to be a document, and is admissible without further proof or production of the original, if the conditions in the section are satisfied. The conditions, stated in Section 63(2), track the old conditions closely: the computer output was produced by a computer or communication device regularly used to store or process information for regularly performed activities, by the person having lawful control over the device; the information was regularly fed into the device in the ordinary course of those activities; the device was operating properly throughout the material part of the relevant period, or any respect in which it was not did not affect the record or the accuracy of its contents; and the information contained in the record reproduces or is derived from the information so fed in. Two expansions of reach are built into the text. The references to communication devices bring the phone, and the records held through intermediaries and in semiconductor memory, expressly inside a provision that had been drafted around computers and optical and magnetic media. And Section 63(3) addresses output produced by a combination or succession of computers and devices, as records in practice are, rather than by a single machine.

The certificate sits at Section 63(4), and its text answers the first of the seeded questions this article must resolve, whether the requirement survived. It did more than survive. Where a statement is to be given in evidence under the section, a certificate must be submitted along with the electronic record at each instance where the record is submitted for admission, doing the things the old certificate did: identifying the electronic record containing the statement and describing the manner in which it was produced, giving the particulars of the device involved, and dealing with the conditions in Section 63(2) as appropriate. The certificate must be signed by two persons where the old law required one signature from a person in a responsible official position in relation to the device or the activity: it must be signed by the person in charge of the computer or communication device or the management of the relevant activity, and by an expert. And it must be given in the form prescribed in the Schedule to the Adhiniyam, which sets out the certificate in two parts. Part A is completed by the person in charge of the device, identifying the record and the device, and it requires the hash value of the electronic record to be stated. Part B is completed by the expert, and it addresses the record from the technical side, including the algorithm by which the hash value was produced. The old Act prescribed no form at all; practitioners reconstructed the certificate from the conditions in Section 65B and usually swore it as an affidavit. The Adhiniyam replaces that reconstruction exercise with a form, and replaces a self-declared statement of genuineness with a certified fingerprint of the file.

Three separate judgments are needed on this, and the professional discussion goes wrong when they are merged. The first concerns the form. A prescribed form is a genuine practical gain of a limited kind. Under the old law, certificates were rejected for omissions their drafters could not have anticipated, because the statute never said what a complete certificate contained; a form that states the fields removes that species of failure and standardises what registries and courts receive. The second judgment concerns the hash value, which is the most defensible innovation in the provision. The mischief a certificate exists to meet is that a printout or a copied file can be altered without visible trace; a hash value, being a fixed-length value produced from the record by an algorithm, changes if the record changes, so certifying the hash gives the opposing party and the court a way to test whether the record tendered is the record certified. That is a real evidentiary improvement, stated in the statute rather than left to forensic practice. The third judgment concerns the second signature, and it runs the other way. Requiring an expert to sign Part B in every case in which an electronic record is tendered as secondary evidence adds a professional engagement, a fee, and a potential point of failure to the proof of records that are, in the great mass of litigation, mundane: a bank statement, a call record, a message thread, a closed-circuit recording. The expert’s certificate addresses the integrity of the record, which the hash value already addresses mechanically. Whether courts will insist on the expert signature as a condition precedent with the strictness Arjun Panditrao applied to the old certificate, or will treat defects in Part B as curable in the manner that judgment allowed for the old certificate, is the first substantial question the new provision sets for litigation, and it cannot be answered from the text alone.

One phrase in Section 63(4) will do quiet work in the litigation and is worth isolating: the certificate must be submitted along with the electronic record at each instance where it is submitted for admission. On the words of the provision, the requirement is not satisfied once for the life of the case; it attaches at each instance of submission, which covers the tendering of the same record at a later stage, in a connected proceeding, or in appeal, where records are re-tendered from the trial file. Practitioners under the old law had treated the certificate as a document filed once and read thereafter from the record; the new phrase gives an objector a textual argument that a record carried into a second proceeding without its certificate is a record submitted without the statutory accompaniment. Whether the appellate courts read the phrase with that literal force, or as describing the point of first admission, is a third question the provision sets, and it interacts with the transition rule: a record certified under the old Act in a pending trial that concludes, and then travels in appeal into a posture the courts treat as governed by the Adhiniyam, may have to be certified again, in a form the old proceeding never used. None of this is a reason the phrase should not have been written; certainty about the point of submission was one of the old law’s genuine defects. It is a reason to count the phrase among the provision’s operative changes rather than its cosmetics.

Two further provisions frame the certificate and must be read with it. Section 61 provides that nothing in the Adhiniyam denies the admissibility of an electronic or digital record on the ground that it is an electronic or digital record, and that such a record has the same legal effect, validity and enforceability as any other document. That provision does parity work: it forecloses the argument, which surfaced repeatedly under the old law, that an electronic record is inherently a lesser species of document. It does not, on its face, displace the specific admissibility conditions in Section 63, and the relationship between the general parity of Section 61 and the specific conditions of Section 63 is the second question the provision sets for litigation. Section 57, meanwhile, carries the primary evidence changes. Its opening explanations reproduce the old Section 62, on documents executed in several parts, in counterpart, and by a uniform process. Its Explanations 4 to 7 are new, and they speak to the way electronic records exist. Where an electronic or digital record is created or stored, and the storage occurs simultaneously or sequentially in multiple files, each such file is primary evidence. Where an electronic or digital record is produced from proper custody, it is primary evidence unless it is disputed. Where a video recording is simultaneously stored in electronic form and transmitted or broadcast or transferred to another, each of the stored recordings is primary evidence. And where an electronic or digital record is stored in multiple storage spaces in a computer resource, each such automated storage, including temporary files, is primary evidence. These explanations answer, in the statute, a question the old law had left to argument: in a world where the same record lives in the originating device, in a backup, in a cache, and on a server, the hunt for the one original was misconceived, and the Adhiniyam says so.

The conditions in Section 63(2) repay an applied reading, because they decide real objections and they translate unevenly across the three kinds of record that dominate practice. For a record taken from the phone of a party, the conditions work as the drafters would have expected: the device is identified, its regular use is within the knowledge of the person in charge of it, and the certificate speaks from first-hand control. For a record held by an intermediary, a call detail record held by a telecommunications provider or a log held by a platform, the conditions work, but the certifier changes: the person in charge is an officer of the provider, the party’s control is nil, and the certificate arrives, if it arrives, through the provider’s compliance machinery, which is why the summons route the courts developed under the old law remains the working answer for such records. For a recording from a system nobody in the case controls, a municipal camera or a shop’s recorder overwritten in a cycle, the conditions strain at their edges, because regular use and proper operation must be spoken to by whoever maintained the system, and the expert’s Part B must address a record whose original environment may no longer exist by the time of trial. The provision does not distinguish these cases, and the early case law on Section 63 will in large part be the courts deciding, case type by case type, how much of the Section 63(2) account is enough. That is the same work the courts did under Section 65B, which is the continuity underneath the novelty: the conditions are old, their application to devices is the standing difficulty, and the Adhiniyam’s contribution to that difficulty is a better form and a fingerprint, not a new test.

Primary Evidence Status and the Certificate Are Different Questions

The single most productive confusion in the discussion of the Adhiniyam is the treatment of primary evidence status and the certificate as though they were one question, and the statute’s structure explains both why the confusion arises and why it is wrong. Primary evidence is a classification: it states what the record is, the document itself produced for the inspection of the court, as against secondary evidence, which is evidence of the contents of a document produced from something else. The certificate is a condition: it states what must accompany a computer output before the output may be received without production of the original. A record can be primary evidence in the sense of Explanation 5 to Section 57, produced from proper custody, and the party tendering a printout of it can still face the question whether the printout, as a computer output tendered in place of the record in its native form, must travel through Section 63. The Standing Committee noticed exactly this overlap, as the passage section recorded, and recommended adherence to the certificate route, and the enacted text keeps both provisions without stating a priority between them. The practical consequence is that the prudent litigant treats the two as cumulative until the courts rule otherwise: the record is characterised under Section 57, and the output is certified under Section 63, and neither step is assumed to excuse the other.

This is also the point at which the hardest question in the old law reappears in new dress, and the article must frame it rather than pretend to settle it. Arjun Panditrao identified the hardship case precisely: the party who needs the record but does not control the device, the subscriber whose call records sit with a service provider, the litigant whose evidence is a recording on a stranger’s phone. The Court’s answer within the old law was procedural: the certificate remained mandatory, but the court could be moved to summon the record and the person who could certify it, and the defect could be cured as the trial proceeded. The Adhiniyam does not restate that answer. Its Section 63(4) presupposes a person in charge of the device and an expert, and says nothing about the party who has neither. Defenders of the drafting, including the position taken in the Committee’s recommendation of adherence to the certificate route, would say the procedural answer survives because it was an exercise of the court’s ordinary powers to summon evidence, which no provision of the Adhiniyam removes. Critics of the drafting answer that a legislature re-enacting a provision after a decade of litigation about its hardship cases had the opportunity to write the safety valve into the text, and chose instead to add a second signature. Both readings of the same section are available, and which of them the appellate courts adopt will decide whether the certificate reform is remembered as a tidying of the old law or as its hardening. The transition position described in this article states the general law as enacted; which statute governs a particular pending matter, and what a court will require in it, turns on that matter’s procedural posture, and questions of that kind repay advice taken on the individual case rather than reliance on any general account, this one included.

The Second and Third Moves: What Changed, What Was Dropped, What Was Added

Outside the electronic records provisions, the genuine changes are few enough to be listed with precision, and precision matters here, because loose claims that the Adhiniyam transformed this or that doctrine usually dissolve on inspection. The confession block was restructured. Sections 24 to 30 of the old Act, seven provisions, became Sections 22 to 24 of the Adhiniyam, three provisions with provisos. Section 22 gathers the rule on confessions caused by inducement, threat or promise, which was the old Section 24, together with the two qualifications the old Act had stated separately, the confession made after the impression of the inducement has been removed, which was the old Section 28, and the confession that is otherwise relevant despite secrecy, deception, drunkenness, or the absence of a warning, which was the old Section 29; and Section 22 adds the word coercion to the inducement, threat or promise of the old text, making express a ground that argument had previously had to locate inside the older words. Section 23 gathers the police provisions: sub-section (1) is the old Section 25, the bar on proving a confession made to a police officer; sub-section (2) is the old Section 26, the bar on proving a confession made in police custody otherwise than in the immediate presence of a Magistrate; and the proviso to sub-section (2) is the old Section 27, the discovery exception, preserved with its limiting words, so that only so much of the information as relates distinctly to the fact discovered may be proved. Nothing in that restructuring changes a single outcome, and the case law on the old sections, from Pulukuri Kottaya on the discovery exception to the decisions on the voluntariness rule, reads across without translation of substance.

One change inside the confession block is substantive. Section 24, which carries the old Section 30 on the confession of a co-accused, provides that when more persons than one are being tried jointly for the same offence and a confession by one of them, affecting himself and another, is proved, the court may take the confession into consideration against the other as well as against its maker. An Explanation added by the Adhiniyam provides that a trial of more persons than one, held in the absence of an accused who has absconded or who fails to comply with a proclamation issued under Section 84 of the Bharatiya Nagarik Suraksha Sanhita, is deemed to be a joint trial for the purposes of the section. Under the old Act, the prosecution faced a real difficulty where one of several accused absconded: the trial that proceeded against those present was arguably not a joint trial of the absconder, and the confession of a present accused could not be considered against the absent one. The Explanation removes that argument. It is a small provision with a definite direction, extending the reach of a co-accused confession to a person who is not in the dock to answer it, and it deserves to be noticed as the one place in the confession block where the replacement moved the law rather than rearranging it.

The restructuring of the confession block also deserves a worked statement, because the discovery exception is the provision in the Adhiniyam most likely to be used in an ordinary trial and the one whose operation is most often misdescribed. An accused in police custody tells the investigating officer where a weapon is buried. The statement is a confession made in custody, and Section 23(2) bars proving it, the recording machinery for confessions before a Magistrate in Section 183 of the Bharatiya Nagarik Suraksha Sanhita being the route by which a custodial confession may enter the record at all. The proviso to Section 23(2) then admits so much of the information as relates distinctly to the fact discovered in consequence of it, the buried weapon, and no more. The prosecution may prove that the accused gave information and that the weapon was found where the information pointed; it may not prove the accompanying narrative of how the weapon came to be there. That discipline, worked out under the old Section 27 in Pulukuri Kottaya and its successors, survives the renumbering word for word in substance, and a court applying the proviso under the Adhiniyam is applying the old learning to materially identical words. The interface with the Magistrate-recording provisions is likewise undisturbed: the evidence statute states which confessions are relevant, and the procedure code states how a confession is to be recorded with its safeguards of voluntariness, warning, and reflection time, and neither statute’s replacement disturbed the other’s half of the arrangement.

The illustrations of the old Act were carried across with the sections they illustrate, which is a quieter continuity than any other in the statute and one with practical consequences. Stephen’s illustrations were never decoration; they were worked examples of the principles, cited in argument and in judgments for a century and a half, and their retention means that the stock of canonical examples, the placards printed from one original in the illustration to Section 57 among them, remains available to courts and candidates without translation. The modernisation of the illustrations was partial, updating archaic names and terms in some while leaving the older vocabulary standing in others, as the committee-stage scrutiny recorded. A reader comparing the two texts illustration by illustration will find the same pattern as in the sections themselves: continuity as the rule, modernisation at the edges, and occasional inconsistency where the two meet.

The second substantive change sits at Section 138, on the accomplice. The old Act stated the position in two places: Section 133 provided that an accomplice is a competent witness against an accused person and that a conviction is not illegal merely because it proceeds on the uncorroborated testimony of an accomplice, while Illustration (b) to Section 114 permitted the court to presume that an accomplice is unworthy of credit unless corroborated in material particulars. The two provisions were read together for a century and a half, producing the settled practice that accomplice evidence is legally sufficient and practically insufficient without corroboration. The Adhiniyam states the whole position in Section 138 itself: the competence rule and the legality rule are carried in, and a proviso then provides that the court may presume that an accomplice is unworthy of credit unless he is corroborated in material particulars. The presumption has moved from an illustration of a general power into the text of the governing section. Whether that relocation stiffens the corroboration requirement, converting a practice into something nearer a rule, or merely collects in one place what the courts already did, is a question the appellate courts will have to work out; the drafting change itself is undeniable, and it is the kind of change a renumbering exercise was not supposed to contain.

The remaining changes are of a third kind, tidying with real but modest effects. Section 58 widened the definition of secondary evidence, bringing oral admissions of the contents of a document, and written admissions of those contents, within the categories the old Section 63 had stated more narrowly, and closing a gap through which a party’s own account of a document’s contents could be kept out on a technicality. Section 122, the estoppel of tenant and licensee that was the old Section 116, added the words or any time thereafter, extending the estoppel across the whole period of the tenancy’s continuance and beyond its beginning. The cross-references throughout the Act were rewritten to the new codes, so that provisions that had referred to the Indian Penal Code and the Code of Criminal Procedure refer instead to the Bharatiya Nyaya Sanhita and the Bharatiya Nagarik Suraksha Sanhita. Archaic vocabulary was replaced in the body and the illustrations, the vocabulary of vakils, pleaders and attorneys giving way to advocates, and the older term for a person of unsound mind replaced by modern equivalents, though the committee-stage scrutiny of the bill, and the analysis published by Pranav Verma and Anupama Sharma of the National Law School, both recorded places where the substitution was incompletely carried through, leaving the older vocabulary standing in some illustrations. And one provision was dropped outright that can be named with confidence: the old Section 166, which empowered jurors and assessors to put questions to witnesses, was omitted, jury trial having disappeared from Indian criminal practice generations before the repeal. Its omission changes nothing in any living courtroom, which is precisely why it is instructive: it is the model of a defensible deletion, a provision whose world had ended.

Why does the Bharatiya Sakshya Adhiniyam have a Schedule when the old Act had none?

Because the certificate for electronic records needed a statutory form. The old Act stated the conditions a certificate had to speak to and prescribed no form, so certificates were drafted by reconstruction and sworn as affidavits. The Schedule prescribes the two-part certificate, Part A for the person in charge and Part B for the expert.

The Schedule is easy to overlook and should not be, because it is the only place where the Adhiniyam adds an apparatus rather than a rule. A schedule in an Indian statute is part of the Act, amendable by Parliament like any other part, and its presence here means that the content of the Section 63 certificate is a legislative text, open to the Standing Committee’s criticism that Part A did not track the conditions of Section 63(2), and open to legislative repair without reopening the section itself. It also means that a certificate which departs from the form raises a question of statutory compliance rather than a question of practice, which is one more reason the early litigation on Section 63 will be watched closely by anyone who tenders electronic records for a living.

The Evidence Continuity Map

The table that follows is the working core of this article. It gives, for each major doctrine, the provision of the 1872 Act, the provision of the Adhiniyam that carries it, whether the substance changed, and whether the case law built on the old provision survives. The survival column follows the ordinary principle that governs every replacement in this series: where the successor provision reproduces the predecessor in substance, the authority construing the predecessor continues to govern the successor, because the words being construed are, in the respects that matter, the same words.

Doctrine	Indian Evidence Act 1872	Bharatiya Sakshya Adhiniyam 2023	Substance changed?	Prior case law survives?
Relevancy of facts	Section 5	Section 3	No	Yes
Facts forming part of the same transaction	Section 6	Section 4	No	Yes
Motive, preparation and conduct	Section 8	Section 6	No	Yes
Admissions	Sections 17 to 23	Sections 15 to 21	No	Yes
Confession caused by inducement, threat or promise	Sections 24, 28 and 29	Section 22	Wording widened to include coercion	Yes
Confession to a police officer	Section 25	Section 23(1)	No	Yes
Confession in police custody	Section 26	Section 23(2)	No	Yes
Discovery of fact from information in custody	Section 27	Proviso to Section 23(2)	No	Yes
Confession of a co-accused	Section 30	Section 24	Yes, extended to trials where an accused has absconded	Yes, on the shared text
Dying declaration and statements of persons who cannot be called	Section 32	Section 26	No	Yes
Expert opinion	Section 45	Section 39	No	Yes
Opinion of the Examiner of Electronic Evidence	Section 45A	Section 39(2)	No	Yes
Primary evidence	Section 62	Section 57	Yes, four explanations added for electronic and digital records	Yes, on the shared text
Secondary evidence	Section 63	Section 58	Yes, definition widened to oral and written admissions	Yes, on the shared text
Cases in which secondary evidence may be given	Section 65	Section 60	No	Yes
Parity of electronic and digital records	No equivalent	Section 61	New provision	Not applicable
Special provisions for electronic records	Section 65A	Section 62	No	Yes
Admissibility of electronic records and the certificate	Section 65B	Section 63	Yes, certificate put in a scheduled form, hash value and expert signature added	Yes, subject to the new certificate text
Presumption for documents thirty years old	Section 90	Section 92	No	Yes
Presumption for electronic records five years old	Section 90A	Section 93	No	Yes
Burden of proof, general rule	Section 101	Section 104	No	Yes
Presumption of legitimacy	Section 112	Section 116	No	Yes
Presumption as to abetment of suicide by a married woman	Section 113A	Section 117	No	Yes
Presumption as to dowry death	Section 113B	Section 118	No	Yes
General power to presume facts	Section 114	Section 119	No	Yes
Presumption of absence of consent in specified prosecutions	Section 114A	Section 120	No	Yes
Estoppel	Section 115	Section 121	No	Yes
Accomplice evidence	Section 133 with Illustration (b) to Section 114	Section 138	Partly, the corroboration presumption moved into the section as a proviso	Yes, subject to the proviso
Judge’s power to put questions	Section 165	Section 168	No	Yes
Improper admission or rejection of evidence	Section 167	Section 169	No	Yes
Power of jury or assessors to put questions	Section 166	Omitted	Provision dropped	Not applicable

The Presumptions as to Documents in Practice

Sections 78 to 93, the presumptions as to documents, are the least discussed part of the documentary chapter and the part a busy court uses most, because presumptions are what allow routine proof to proceed without a witness for every paper. The scheme carried across from the old Sections 79 to 90A is a catalogue of situations in which the court shall presume, or may presume, a document’s genuineness or a fact about its origin: the genuineness of certified copies of judicial records and of the records of evidence taken in accordance with law; the genuineness of documents produced as records of evidence; the official character of the Gazette and of documents directed by law to be kept; the accuracy of maps and plans made by the authority of government; the due execution of powers of attorney; and, in the group the amendments of the year 2000 added and the Adhiniyam retained, the presumptions concerning electronic agreements, electronic records and signatures, and secure electronic records and signatures. Two of the catalogue’s entries do disproportionate work. The presumption for a document thirty years old, produced from custody the court considers proper, in Section 92, relieves parties of proving the execution of old deeds, wills, and grants whose witnesses are dead, and it is the provision on which a large share of title litigation quietly proceeds. The presumption for an electronic record five years old, produced from proper custody, in Section 93, is its digital counterpart, and it sits in an instructive tension with Section 63: the old record in proper custody attracts a presumption of genuineness, while the fresh printout of a record attracts a certificate. Both provisions survived the replacement untouched in substance, and their survival illustrates the retention principle at its most practical: the daily shortcuts by which courts receive paper and files were too well worn to justify disturbance, and the drafters, whatever their presentation, left them alone.

The admissions block contains a closing rule of the same practical character, which the renumbering carried without change: admissions are not conclusive proof of the matters admitted, but they may operate as estoppels against the person who made them or those claiming under that person. The distinction between an admission, which is evidence to be weighed, and an estoppel, which is a bar to denial, is one of the old Act’s most examined subtleties, and its survival in the closing provision of the admissions group in the Adhiniyam means the whole jurisprudence of admissions, who may make them, whose statements count as admissions, when an admission binds a successor in interest, and when it does not, continues to state the law. A replacement that had disturbed any of this would have generated litigation out of all proportion to any gain; this one did not.

Reading the Two Statutes Side by Side

Because the two regimes will run together for years, the skill the replacement demands most often is conversion: taking a proposition stated under the old Act and stating it correctly under the new one. The conversion has three steps, and skipping any of them is how errors enter opinions and pleadings. The first step is numerical: find the successor provision, using a concordance of the kind the continuity map in this article provides. The second step is textual, and it is the step most often skipped: read the successor provision against the predecessor, because the Adhiniyam’s changes are exactly the kind that a number-only conversion misses, an added word such as coercion in Section 22, an added phrase such as or any time thereafter in Section 122, a proviso that has moved house, as the accomplice presumption moved from an illustration into Section 138, an explanation that has no predecessor, as the absconder Explanation to Section 24 and Explanations 4 to 7 of Section 57 have none. The third step is precedential: ask whether the authority being cited construed words the successor kept, or words the successor changed. Authority of the first kind converts freely. Authority of the second kind converts only so far as its reasoning attaches to the shared text, and on the certificate, where the successor changed the signature, the form, and the contents of the certification, the old authority is a starting point for argument and not a statement of the new law.

Two habits protect against the commonest conversion errors. The first is to cite the statute with the section in every transition-year document, Bharatiya Sakshya Adhiniyam Section 63 and Indian Evidence Act Section 63 being different provisions on different subjects, secondary evidence in the old Act and the electronic records certificate in the new one, and the same collision occurring at other numbers, including Section 58, secondary evidence in the Adhiniyam and admissions in the old Act’s neighbourhood, and Section 26, statements of persons who cannot be called in the Adhiniyam against confessions in custody in the old Act. The second habit is to distrust any secondary source, including examination material and practice notes, that states a BSA position without identifying the provision it rests on, because the renumbering has made confident miscitation cheap: a proposition can be sound law and a wrong citation at once, and in an appellate record the wrong citation is the part that will be remembered. The researcher’s version of the same discipline is to treat the pair of statutes as one text with a concordance, which is what, in substance, the replacement made them.

Repeal and Savings: Section 170 and the Two-Regime Years

Section 170 is the whole of Chapter XII of the Adhiniyam and it does two things, in two sub-sections, with unusual clarity by the standards of transition provisions. Section 170(1) repeals the Indian Evidence Act, 1872. Section 170(2) then provides that, notwithstanding the repeal, if immediately before the date on which the Adhiniyam comes into force there is any application, trial, inquiry, investigation, proceeding or appeal pending, that application, trial, inquiry, investigation, proceeding or appeal shall be dealt with under the provisions of the Indian Evidence Act, 1872, as in force immediately before the commencement, as if the Adhiniyam had not come into force. The unit of transition is therefore the proceeding, not the case in the abstract and not the individual item of evidence. A trial that was pending on 30 June 2024 is an Evidence Act trial to its end, conducted under the old numbering, the old certificate provision, and the old case law as a living code rather than as history. A proceeding begun on or after 1 July 2024 is an Adhiniyam proceeding throughout. The statute book consequently carries two evidence regimes at once, and will carry them for as long as the longest pending proceeding lasts, which in Indian litigation is a horizon measured in years and, in some appellate matters, comfortably in decades.

The pattern is not peculiar to this statute. The companion codes adopted the same transitional philosophy through their own savings provisions, Section 358 of the Bharatiya Nyaya Sanhita for offences and proceedings under the penal code it replaced, and Section 531 of the Bharatiya Nagarik Suraksha Sanhita for the procedure it replaced, so that across the criminal system the governing code for a pending matter is the code under which the matter stood when the new codes commenced, and the new codes govern what begins after commencement. The consequence for evidence is a doubled working library in every criminal court: the old Act for the pending stock, the Adhiniyam for the new filings, and a bar and bench that must hold both numberings at once. The savings language, as if this Adhiniyam had not come into force, is strong language, and it has a consequence worth noting for the precedent discussion that follows: judgments delivered in pending matters after 1 July 2024 will continue to state Evidence Act law as governing law, not as persuasive history, so the old Act’s case law will go on growing, in its own line, alongside the first authorities on the Adhiniyam. The two lines will converge wherever the texts are the same, and the profession’s task is to know where they are not.

What happens to evidence already recorded when a case crosses the changeover date?

It stays where it was recorded. Section 170(2) keeps any application, trial, inquiry, investigation, proceeding or appeal that was pending immediately before 1 July 2024 under the Indian Evidence Act as if the Adhiniyam had not come into force, so evidence in a pending matter is neither reopened nor re-proved under the new numbering.

The clarity of the proceeding-based test should not conceal its edges, and the edges are where the transition litigation will live. An appeal filed after commencement against a decree or conviction in a trial that was pending before it raises the question whether the appeal is itself a pending proceeding within the saving, or a new proceeding governed by the Adhiniyam, and the statutory language, which saves an appeal that was itself pending immediately before commencement, points to the second answer while the forensic reality, that the appeal reopens the record of an Evidence Act trial, pulls toward the first. An investigation that straddles the date raises a different version of the same question for each act done in it. And a proceeding that was pending but in which evidence begins only after commencement is governed by the old Act on the words of the saving, however odd it may feel to open a witness box in a later year with a certificate under a repealed section. These are questions to be worked out on the language of Section 170 in the courts, and the honest position at the level of a general article is that the test is clear and its application to hybrid postures is not. What can be said without qualification is that the drafters chose the more conservative of the available transitions: nothing already in the system was disturbed, at the price of running two regimes in parallel for a long tail of years.

The Continuity of Precedent and the Renumbering Tax

The most practically useful proposition in this article is also the one most often misstated in commencement publicity, so it is worth stating in its exact form. The repeal of the 1872 Act did not repeal its case law. Where the Adhiniyam reproduces a provision of the old Act in substance, a judgment construing the old provision continues to state the law of the new one, because a court construing the successor is construing materially the same words, and the established reading of those words is the reading the successor inherits. This is not a courtesy extended to old judgments; it is the ordinary consequence of legislative continuity, and it is why the great evidence authorities, on res gestae, on dying declarations, on the discovery exception, on burden and presumptions, remain the authorities to cite in an Adhiniyam courtroom, with the new section number set beside the old. The exception proves the boundary of the rule. Where the Adhiniyam changed the text, in the certificate provisions, in the explanations to Section 57, in the Explanation to Section 24, and in the proviso to Section 138, the old authority governs only so far as the shared text carries it, and the changed words are open ground on which new authority will have to be built. A citation practice that ignores that boundary, citing a Section 65B judgment as though Section 63 were its twin, will mislead in exactly the region where the statutes differ.

Against that continuity stands the cost, and the cost deserves its name. Where a successor provision reproduces its predecessor in substance, every existing judgment remains good law but every citation has to be translated, so the immediate price of the new evidence statute is not doctrinal change but a decade of double citation. Pleadings will carry both numbers, because the pending half of the system runs on the old ones. Textbooks will carry both numbers, because their readers practise in both regimes. Headnotes, digests, and examination syllabuses will carry both numbers, because the questions set in the transition years are answered by candidates who must know which regime a problem belongs to before they can answer it at all. None of this is an argument that the replacement should not have happened; the electronic records integration, if it works, is a benefit that recurs in every case involving a phone or a server, while the translation cost is paid once per citation and falls away as the old-regime tail shortens. It is, however, an argument for honesty in the accounting. The benefit side of the Adhiniyam is narrow and specific, and the cost side is broad and certain, and an assessment that counts only the benefit is not an assessment.

What is the renumbering tax under the Bharatiya Sakshya Adhiniyam?

It is the translation cost of a replacement that changed the numbering of provisions it left substantively intact. Every judgment under the old Act remains good law where the text is shared, but every citation must be converted to the new number, so pleadings, textbooks and digests carry both numbers for years.

The tax is heaviest where the old Act was most heavily glossed, which is to say in the confession provisions and the presumptions, and lightest where the doctrine was thin. It also falls unevenly across the profession. A practitioner whose work is in the trial courts of one district will convert the twenty provisions used weekly and be done; a practitioner in appellate work, or a teacher, or an author of a commentary, must maintain the full concordance as a permanent feature of working life for the duration of the two-regime years. The continuity map in this article is offered as one instalment of that conversion work, and the study treatment of the three replacement codes together is taken up in the criminal law reform examination guide, which owns the revision architecture for candidates.

How a Document Is Proved Under the Adhiniyam

The documentary evidence chapter, Sections 56 to 93, is where the retention and the change meet most concretely, and it rewards a walk through its working sequence, because the sequence is what a trial lawyer actually performs. Section 56 states that the contents of documents may be proved either by primary or by secondary evidence. Section 57 defines primary evidence, as set out earlier, and Section 58 defines secondary evidence, the certified copies, the copies made from the original by mechanical processes, the counterparts, and, in the Adhiniyam’s widened version, the oral and written admissions of contents. Section 59 states the default rule, that documents must be proved by primary evidence except in the cases Section 60 permits, and Section 60 states those cases, reproducing the old Section 65: the original is in the possession of the adversary or of a person out of reach of the court’s process or not bound to produce it, the original is lost or destroyed or cannot be produced in reasonable time, the original is a public document, the original consists of numerous accounts, and the remaining cases the section has always carried. A document that passes through that gate is then proved according to its kind: public documents under Sections 74 to 77, by certified copies whose genuineness the court presumes in the circumstances the presumptions chapter states; private documents by the evidence of execution and attestation the following sections require. None of that choreography changed. What changed is that the electronic record performs the same steps as a document of any other kind, entering at Section 56 rather than at a special annexe, and meeting its one special condition, the certificate, at Section 63, within the sequence rather than outside it.

The sections that complete the documentary chapter govern how execution is proved once a document is before the court, and they too crossed over without disturbance. A document required by law to be attested, a will, a mortgage, a gift of immovable property, is not used in evidence until at least one attesting witness has been called to prove its execution, where such a witness is alive, subject to the process of the court, and capable of giving evidence, with stated relaxations where no attesting witness can be found or where the document is registered and its execution is not specifically denied. Signature and handwriting are proved by the evidence of the writer, of a person who saw the writing or signing, of a person acquainted with the hand, by comparison of the disputed writing with an admitted or proved specimen by the court or by an expert, and, for electronic signatures, by the certificate framework the information technology statute supplies. The proof of a document’s contents, its execution, and its genuineness are three separate operations under the chapter, and confusing them is the commonest documentary error in trial practice; the Adhiniyam preserved the separation exactly, which is why the documentary learning of the old Act, on attestation, on comparison of handwriting, on the proof of thirty-year-old deeds, and on the treatment of documents withheld by an adversary after notice, continues to be the learning of the new one without translation of anything except the citation.

The practical difference that integration makes is easiest to see at the point of objection. Under the old Act, an objection to a printout characteristically opened a threshold dispute about which code governed the exhibit at all: the special code of Section 65B, on the Anvar view, or the ordinary documentary provisions, on the view it displaced, with the further question whether the document before the court was the record itself or secondary evidence of it. Under the Adhiniyam, the questions are the same in number but better ordered. The exhibit is a document by definition. Its contents are proved by primary or secondary evidence under Sections 56 to 60. If it is tendered as a computer output in place of the record in native form, Section 63 states the conditions and the certificate. The ordering does not remove the disputes, because the disputes were about hardship and reliability rather than about ordering, but it removes the preliminary dispute about the map, and map disputes are the purest waste in litigation.

Witnesses, Examination, and the Untouched Machinery

The third and fourth parts of the Adhiniyam contain the machinery by which oral evidence is taken, and the replacement left that machinery essentially as it found it, which is itself a finding worth recording, because the trial’s oral phase is where most evidence law is lived. Competence is stated at Section 124: all persons are competent to testify unless the court considers that they are prevented, by tender years, extreme old age, disease of body or mind, or a like cause, from understanding the questions put or giving rational answers to them. Compellability and privilege occupy the sections that follow, reproducing the old scheme: the parties and their spouses in civil and criminal proceedings, the privileges of judges and Magistrates for their official conduct, the protection of communications during marriage, of official communications and of information as to the commission of offences, and the privilege for professional communications between a client and a legal adviser, all carried across with the substance and the internal balance the courts had settled. The accomplice provision, Section 138, sits at the end of the witnesses chapter, and the witness who cannot be compelled to answer a self-incriminating question, the production of title deeds, and the remaining protections follow the old order.

Examination occupies Sections 140 to 168, and its rules are the rules of every Indian trial: the order of examinations, examination in chief, cross-examination, and re-examination; the discretion of the court to permit a party to cross-examine its own witness, the hostile witness situation, which sits at Section 157 as it sat at the old Section 154; the rules on leading questions, on evidence of oral admissions as to contents, on the writing that may be used to refresh memory, on the documents a witness summoned to produce must bring, and on the consequences of refusing to answer; and at Section 168 the Judge’s power, carried from the old Section 165, to put any question at any time and to order the production of any thing, for the purpose of discovering or obtaining proper proof of relevant facts. The judge’s question power is the provision practitioners most often feel in the room, and its survival unamended means that the inquisitorial element Stephen built into an otherwise adversarial code remains exactly as strong, and exactly as bounded by the requirement that judgment rest on facts declared relevant and duly proved, as it was. A replacement statute that had wanted to change the character of the Indian trial would have had to begin in these chapters. It did not touch them, and the character of the trial did not change.

What the Adhiniyam Did Not Do

A replacement is defined as much by its omissions of reform as by its reforms, and four roads not taken mark the boundaries of this one. The first is chain of custody. The authenticity problem in digital evidence is only partly a problem of the exhibit tendered in court; it is at least as much a problem of what happened to the device and the data between seizure and production, and the Standing Committee recommended a statutory mandate for secure handling and proper chain of custody of electronic records gathered in investigation. The Adhiniyam contains no such mandate; the Committee itself pointed to the procedure code as the mandate’s proper home, and the evidence statute confines itself to the court’s side of the question, the conditions on which a record will be received. The integrity of the journey from seizure to courtroom therefore continues to rest on practice, on the directions of courts in particular cases, and on the certificate’s account of the device, rather than on a rule of the evidence code.

The second road not taken is hearsay. The Adhiniyam reproduces Stephen’s relevancy scheme, under which hearsay is excluded by structure rather than by a single prohibition, together with its exceptions, the dying declaration, the statements in the ordinary course of business, the entries in public records, and the rest, without revisiting any of the standing criticisms of that structure or any of the proposals, periodically made, to widen the exceptions for reliable documentary hearsay in civil cases. The third is expert evidence. The law of expert opinion, Section 39, was carried across without amendment at a time when the quality and the courtroom use of forensic expertise are under sustained criticism in adjacent debates; the Adhiniyam neither widened the categories of expertise nor added any gatekeeping discipline of the kind other systems have adopted, and the Examiner of Electronic Evidence at Section 39(2) remains, as under the old Act, a relevant voice rather than a standard. The fourth is the presumptions. The presumption provisions, including the two that allocate proof in prosecutions for abetment of suicide and dowry death and the general power in Section 119, were carried across without re-examination of their fit with the offences as the Sanhita redefined them. Each of these omissions is defensible in the language of continuity: a replacement is entitled to leave settled law settled. Together they fix the scale of the exercise. The Adhiniyam is a statute that decided its questions in advance, and the questions it decided were numbering, integration, and the certificate.

One Record, Two Regimes: A Worked Example

The operation of the replacement is clearest in a single worked example. Take a message thread on a phone, exported as a printout by the complainant in a prosecution for cheating, and consider it in two courtrooms. In the first, the trial was pending immediately before 1 July 2024. By Section 170(2), the proceeding is dealt with under the Indian Evidence Act as if the Adhiniyam had not come into force. The printout is a computer output tendered as secondary evidence of the record on the phone. Section 65B governs. The prosecution files a certificate identifying the record and the device and speaking to the conditions, signed by a person in a responsible official position in relation to the phone or the activity. If the certificate is defective or missing when the printout is first tendered, the Arjun Panditrao discipline applies: the defect is curable, the certificate may be produced at a later stage, and if the complainant cannot certify a device in another’s control, the court may be moved to summon what is needed. The objection, when the defence takes it, is argued in the vocabulary of Anvar and Arjun Panditrao, and the judgment, when it comes, will cite the old sections, because they are the governing text.

In the second courtroom, the trial began on or after 1 July 2024, and everything about the exhibit is the same except the governing text. The thread is a document by the definition in Section 2(1)(d), because an electronic record is a document. Its contents may be proved by primary or secondary evidence under Section 56. If the phone itself is produced and the thread displayed from it, the record is produced from proper custody and is primary evidence under Explanation 5 to Section 57, unless disputed. If the printout is tendered instead, Section 63 governs its admission: the conditions in Section 63(2) must be shown for the device that produced it, and the certificate under Section 63(4) must accompany the record at the instance of its submission, in the form in the Schedule, Part A completed by the person in charge of the device with the hash value of the record stated, and Part B completed by an expert. The defence objection has new material to work with, the absence of the expert signature, a hash value that does not match the record as tendered, a Part A that omits the device particulars, and the trial court must decide, without the shelter of an exactly applicable precedent, whether such defects go to admissibility or to weight, and whether they may be cured and when. The example shows the replacement in miniature. The doctrine is continuous, the governing considerations are recognisable to any lawyer trained under the old Act, and the points of genuine novelty are exactly where the text changed: the form, the fingerprint, and the second signature.

The Argument About the Adhiniyam: Modernisation, or Renumbering With One Real Change

The case for the replacement, in its strongest form, is the case the government made for the three codes together, applied to the statute where it fits least neatly and therefore most instructively. Amit Shah, introducing the bills, framed the exercise as the substitution of laws made by a republic for its citizens in place of laws made by a colonial power for its subjects, and that framing, whatever its force for a penal code, has a real if narrower purchase on evidence: the 1872 Act was drafted by an English lawyer for a system in which the judiciary distrusted the police of a colonial state, and a republic is entitled to ask whether its proof rules should still be organised around that inheritance. The technical limb of the government’s case is stronger than the presentational limb. Electronic records are the typical exhibit of modern litigation, and the old Act’s treatment of them was an amendment-era annexe whose boundary with the main text had to be litigated for a decade. A code that defines the document to include the electronic record from its second section, states the primary evidence position of multiply stored records in its text, prescribes the certificate in a schedule, and requires the record’s hash value, is, the government can fairly say, better drafted for the litigation that exists than the code it replaced, and drafting quality of that kind compounds quietly in every registry and every trial.

The case against, in its strongest form, was made early and precisely. Pranav Verma and Anupama Sharma, analysing the bill for the National Law School’s forum in December 2023, concluded that the majority of its changes merely amounted to the renumbering or restructuring of existing provisions, that those changes could have been achieved by amending the existing law instead of enacting a new one, and that the bill was marked by drafting errors and misplaced provisos capable of causing confusion on substantive law in places. That analysis, written before passage, reads after passage like a description of the enacted text: the relevancy scheme, the confession rules, the burden chapter and the presumptions all survived, as this article has documented section by section, and the changes that are substantive, the absconder Explanation and the accomplice proviso among them, are changes of a size that an amending Act of a dozen clauses could have carried without disturbing a single citation elsewhere in the system. The critics’ deeper point goes to the certificate. The one region of the old law under genuine strain was re-enacted with its strictness intact and a second signature added, so that on the critics’ account the replacement took the provision most in need of a rethink and gave it a form and a fingerprint instead. The committee-stage record lends that point some support: the Standing Committee’s interventions on the evidence bill were directed overwhelmingly at the electronic records provisions, at chain of custody, and at the certificate form, which is where the Committee, like the critics, located the statute’s actual difficulties.

Between those cases, the evidence supports a measured verdict on the framing question. The claim that a new evidence law was needed because the old one could not handle digital records is, as a statement about the old text, too strong: the old Act handled digital records from the year 2000 onwards, imperfectly and through litigation, but the failures of that period were failures of adjudicative consistency in applying a workable provision, not of statutory silence. The claim that nothing was gained is also too strong. Integration is worth having, the hash value is worth having, the prescribed form is worth having, and the four explanations to Section 57 settle a question about multiply stored records that the old law left genuinely open. What the record does not support is the proposition that these gains required the renumbering of the entire law of evidence, with the translation cost that followed. They required an amending Act on electronic records and accomplice evidence and a short schedule. The rest of the Adhiniyam is the presentation of continuity as change, and the profession will pay the renumbering tax on the presentation for as long as the two regimes run in parallel.

The Working Life of the Replacement

For the practitioner, the replacement resolves into a small set of habits. The first is concordance discipline: no evidence provision should be cited by number alone in the transition years without identifying the statute, because the same number names different provisions in the two Acts, as the conversion discussion in this article shows. The second is regime discipline: the first question in any matter touching evidence is which statute governs the proceeding, and that question is answered by the pendency test in Section 170(2), not by the date of the facts, the date of the offence, or the date of the document. The third is certificate discipline in new proceedings: the record must be identified, the device described, the Section 63(2) conditions addressed, the hash value stated, and both signatures obtained, in the scheduled form, at each instance the record is submitted for admission. Chambers and teaching institutions preparing training material on the new certificate have taken to circulating the Schedule and the section maps as slide decks, and a deck of that kind can be opened and read on any machine through the ReportMedic presentation viewer, which keeps the file in the browser and off the download folder of whichever device is at hand. The fourth habit is preservation of the old learning: the authorities under the 1872 Act are not a closed historical subject but the operating commentary on the Adhiniyam’s shared text, and a library that discards them in favour of new commentaries discards the substance for the numbering.

The registries and the courts carry a version of the same burden. Cause lists, filing counters, and defect memos must work in both numberings, because the file that arrives in the morning may belong to either regime, and a defect memo that cites the wrong statute’s section sends the practitioner away to correct the wrong text. Courts have met the position in the manner courts usually meet a renumbering: by citing both numbers in the early judgments, the new provision first and the old in parenthesis, a practice that costs a line and saves an appeal. Textbook publishers and the editors of the standard commentaries face the largest version of the task, since a commentary on the Adhiniyam that ignored the old authorities would be useless and one that reproduced them without conversion would be dangerous, and the concordance tables that have begun to circulate in chambers and classrooms are the profession’s own answer to the renumbering tax: privately produced, quickly standardised in practice, and nowhere provided for in the statute that made them necessary.

For the student and the candidate, the working position is the mirror image. The doctrines are the doctrines of the old Act, and preparation that treats the Adhiniyam as a new subject wastes the largest asset the candidate has, which is the accumulated commentary and question bank of a century and a half. What must be learned fresh is the concordance, the electronic records provisions in their new form, and the transition rule, and what must be unlearned is the habit of citing the old numbers without the statute’s name. For the researcher, the replacement poses a cleaner question than its companions do, because the evidence statute changed so much less than the penal and procedure codes: it is a controlled experiment in what renumbering alone does to a body of law, and its litigation over the coming years, on the certificate, on the absconder Explanation, and on the edges of Section 170, will show which parts of a code’s life live in its text and which live in the practice that surrounds the text.

The Honest Assessment

The Bharatiya Sakshya Adhiniyam is a conservative statute wearing the dress of a reforming one, and it is best valued when both halves of that description are kept in view. As an act of retention, it is nearly complete: the relevancy scheme, the confession protections, the dying declaration, expert opinion, the burden chapter, the presumptions, estoppel, witnesses and their examination all crossed over, and the case law that gives those provisions their working meaning crossed with them. As an act of change, it is narrow and specific: electronic records were integrated into the definition of a document, multiply stored records were given a primary evidence rule, the certificate was retained and put into a scheduled two-part form with a hash value and an expert signature, a confession was made usable against an absconding co-accused in a deemed joint trial, the accomplice corroboration presumption was moved into its governing section, and secondary evidence was modestly widened. As an act of renumbering, it imposed a translation cost on every citation in the system for the duration of the two-regime years, in exchange for a coherence that is real in the electronic records chapter and largely presentational elsewhere. The question the statute leaves open is the question it inherited: whether the certificate regime, re-enacted with greater formality, will be applied with greater consistency than the regime it replaced. The text cannot answer that. The courts, working through Section 63 case by case, will.

Three watchpoints will tell the story early. The first is the expert signature: whether a certificate in the scheduled form but without Part B completed by an expert is treated as no certificate, as a curable defect on the Arjun Panditrao pattern, or as a matter going to weight, will decide whether Section 63 is a gate or a guideline in ordinary litigation. The second is the relationship between Section 57 and Section 63: whether a record produced from proper custody as primary evidence must also travel the certificate route when tendered as an output, the overlap the Standing Committee flagged, will decide how much of the primary evidence reform is real. The third is the edge of Section 170: whether appeals and investigations that straddle the commencement date are held inside the saving or outside it will decide how long the two regimes run in parallel in practice. On each of these, the profession’s interest is the same: an early, clear, appellate answer, because uncertainty in the law of proof is not an abstraction; it is a retrial ordered years later for a defect nobody could price at the time. The Adhiniyam has given the courts a cleaner text in the one region where the old text was strained, and has left them, in that same region, the hardest questions to answer. That is a fair summary of the whole replacement: a statute that solved the problems drafting could solve, and disclosed, by what it re-enacted, the problems only adjudication can.

Frequently Asked Questions

Q: What does the Bharatiya Sakshya Adhiniyam replace and cover?

The Bharatiya Sakshya Adhiniyam, 2023, Act 47 of 2023, repealed and replaced the Indian Evidence Act, 1872 with effect from 1 July 2024. It covers the whole law of evidence for judicial proceedings in or before any court in India: which facts are relevant, how facts are proved, who bears the burden of proof, what presumptions apply, and how witnesses are examined. It sits beside the other two replacement codes, the Bharatiya Nyaya Sanhita on offences and the Bharatiya Nagarik Suraksha Sanhita on procedure. Proceedings that were pending immediately before 1 July 2024 continue under the repealed Act by the savings rule in Section 170(2), so the Adhiniyam governs proceedings begun on or after its commencement.

Q: Are electronic records now primary evidence?

In defined circumstances, yes. Section 57 of the Adhiniyam, which defines primary evidence, adds four explanations for electronic and digital records. Where a record is created or stored simultaneously or sequentially in multiple files, each file is primary evidence. A record produced from proper custody is primary evidence unless it is disputed. Each stored copy of a video recording that is simultaneously stored and transmitted or transferred is primary evidence. Each automated storage of a record in a computer resource, including temporary files, is primary evidence. That settles a question the old Act left open. Primary evidence status is separate from the certificate question: a computer output tendered in place of the record may still have to satisfy Section 63.

Q: Did the certificate requirement for electronic evidence survive?

It survived and was strengthened. Section 63 of the Adhiniyam is the successor of Section 65B of the Indian Evidence Act, and Section 63(4) requires a certificate to be submitted along with the electronic record at each instance where the record is submitted for admission. The certificate must identify the record, describe how it was produced, give the particulars of the device, and deal with the conditions in Section 63(2). It must be signed by two persons, the person in charge of the computer or communication device or the relevant activity, and an expert, where the old law required one signature. It must follow the form prescribed in the Schedule to the Adhiniyam, and it must state the hash value of the record.

Q: How many sections does the new evidence law have?

The Bharatiya Sakshya Adhiniyam has 170 sections, against the 167 sections of the Indian Evidence Act, 1872. They are arranged in twelve chapters and four parts, and the Act carries one Schedule, which the old Act never had, prescribing the form of the certificate for electronic records. Part I is preliminary, Part II covers the relevancy of facts in Sections 3 to 50, Part III covers proof in Sections 51 to 103, and Part IV covers the production and effect of evidence in Sections 104 to 170. The small increase in the count conceals extensive internal merging: the seven confession sections of the old Act, Sections 24 to 30, became three sections, Sections 22 to 24, in the new one.

Q: What changed about confessions under the Bharatiya Sakshya Adhiniyam?

The substance of the confession law did not change; its arrangement and one rule did. Sections 24 to 30 of the old Act were compressed into Sections 22 to 24. Section 22 gathers the rule on confessions caused by inducement, threat or promise, with the qualifications the old Sections 28 and 29 stated separately, and adds the word coercion. Section 23 carries the bar on confessions to police officers and on confessions in police custody, and its proviso carries the discovery exception of the old Section 27. Section 24 carries the co-accused confession rule, and adds an Explanation deeming a trial a joint trial even where one accused has absconded, which is the one extension of substance in the block.

Q: Does case law under the Indian Evidence Act still apply to the Bharatiya Sakshya Adhiniyam?

Yes, wherever the Adhiniyam reproduces the old provision in substance, which is most of the Act. A judgment construing words that the successor statute kept is authority on the successor, because the words being construed are materially the same. The dying declaration cases, the discovery exception cases from Pulukuri Kottaya onwards, the burden and presumption authorities, and the examination rules all continue to state the law, cited with the new section number beside the old. The boundary is the changed text: the certificate provisions, the new explanations to Section 57, the absconder Explanation to Section 24, and the proviso to Section 138 must be argued on their own words, with the old case law as a starting point rather than a settlement.

Q: Why was a new evidence law needed at all?

The government’s case had two limbs: that the criminal codes of the colonial period should be replaced by codes enacted by the Parliament of the republic, and that the law of evidence needed modernising for a digital era in which the typical exhibit is a phone, a log, or a recording. The pressure point was real: the electronic records provisions inserted into the 1872 Act in the year 2000 had produced a decade of conflicting appellate rulings on the certificate requirement. Critics answered that the old Act’s difficulty was adjudicative inconsistency rather than statutory silence, and that the genuine changes could have been made by amendment. The enacted statute is best read as a renumbering with one region, electronic records, genuinely reworked.

Q: How does the new evidence law treat accomplices?

Section 138 of the Adhiniyam states the whole position in one provision. An accomplice is a competent witness against an accused person, and a conviction is not illegal merely because it proceeds upon the uncorroborated testimony of an accomplice. The old Act stated that much at Section 133, and placed the counterweight in Illustration (b) to Section 114, which allowed the court to presume that an accomplice is unworthy of credit unless corroborated in material particulars. The Adhiniyam moves that counterweight into Section 138 itself as a proviso. The legal sufficiency of accomplice evidence and the practical demand for corroboration therefore sit in a single section, and whether the move stiffens the corroboration expectation is a question for the appellate courts.

Q: How is a document defined under the Bharatiya Sakshya Adhiniyam?

Section 2(1)(d) of the Adhiniyam defines a document as any matter expressed, described or otherwise recorded upon any substance by means of letters, figures or marks, or by more than one of those means, intended to be used or capable of being used as evidence of that matter, and it states expressly that the definition includes electronic and digital records. That express inclusion is a structural change from the old Act, under which electronic records had been brought within the definition by the amendments of the year 2000 and then fenced off in a special regime. Under the Adhiniyam the electronic record is a document from the first page, and the special provisions on electronic records operate inside the documentary evidence chapter.

Q: What is the hash value in a Section 63 certificate and why is it required?

A hash value is a fixed-length value produced from an electronic record by applying an algorithm to its contents; if the record is altered, the value produced changes. Part A of the certificate in the Schedule to the Adhiniyam requires the hash value of the record to be stated, and Part B, completed by the expert, addresses the record from the technical side, including the algorithm used. The requirement serves the purpose the certificate itself serves: a printout or copied file can be altered without visible trace, so the statute asks for a fingerprint by which the record tendered in court can be tested against the record certified. A mismatch between the stated hash value and the record is a concrete ground of objection.

Q: Who must sign the certificate for an electronic record under Section 63?

Two people. Under Section 63(4) of the Adhiniyam, the certificate must be signed by the person in charge of the computer or communication device, or of the management of the relevant activity to which the record relates, and by an expert. The certificate is in the form prescribed in the Schedule: Part A is completed by the person in charge, identifying the electronic record, describing how it was produced, and giving the particulars of the device and the hash value; Part B is completed by the expert. This is a change from Section 65B of the old Act, which required a single certificate signed by a person occupying a responsible official position in relation to the device or the activity.

Q: What can a party do when it cannot obtain a certificate for an electronic record?

The difficulty is the standing hardship case of this branch of the law: the record is held by a service provider, a platform, or a stranger, and the party cannot certify a device it does not control. Under the old law, the Supreme Court in Arjun Panditrao held that the certificate remained mandatory but that the defect could be cured, the certificate could be produced at a later stage, and the court could be moved to summon the record and the person able to certify it. Section 63 of the Adhiniyam does not restate that answer in terms. The practical course remains to apply to the court for the summoning of the record and of the person in charge of the device, and to file the certificate when obtained.

Q: What happened to dying declarations under the Bharatiya Sakshya Adhiniyam?

They were carried across without change of substance. Section 26 of the Adhiniyam is the successor of Section 32 of the Indian Evidence Act, and its opening clause preserves the dying declaration: a statement made by a person as to the cause of that person’s death, or as to the circumstances of the transaction which resulted in death, is relevant in a case in which the cause of that person’s death comes into question, whether or not the person who made the statement was under expectation of death. The rest of Section 26 carries the remaining statements of persons who are dead, cannot be found, have become incapable of giving evidence, or whose attendance cannot be procured without unreasonable delay or expense, as the old section did.

Q: Can a confession made by one accused be used against a co-accused who has absconded?

Yes, in the circumstances Section 24 of the Adhiniyam covers expressly. The section carries the old rule that where more persons than one are being tried jointly for the same offence, and a confession by one of them affecting himself and another is proved, the court may take the confession into consideration against the other as well. An Explanation added by the Adhiniyam provides that a trial of more persons than one, held in the absence of an accused who has absconded or who fails to comply with a proclamation issued under Section 84 of the Bharatiya Nagarik Suraksha Sanhita, is deemed to be a joint trial for the purposes of the section. Under the old Act that deemed joint trial was arguable; the Adhiniyam settles it in the prosecution’s favour.

Q: What does Section 61 of the Bharatiya Sakshya Adhiniyam say about electronic records?

Section 61 is the parity provision for electronic and digital records. It provides that nothing in the Adhiniyam denies the admissibility of an electronic or digital record on the ground that it is an electronic or digital record, and that such a record has the same legal effect, validity and enforceability as any other document. Its work is to foreclose the argument that an electronic record is inherently a lesser species of document, an argument that surfaced repeatedly under the old law. Section 61 does not itself displace the specific conditions for admitting computer output in Section 63, and the relationship between the general parity of Section 61 and the specific certificate regime of Section 63 is one of the questions the new Act sets for the courts.

Q: Did the Bharatiya Sakshya Adhiniyam change the rules on burden of proof?

No. Chapter VII of the Adhiniyam, Sections 104 to 120, reproduces the burden of proof scheme of the old Sections 101 to 114A without change of substance. Section 104 states the general rule that a party who wants the court to give judgment on a legal right or liability dependent on facts must prove those facts exist. Section 108 preserves the burden on an accused to bring the case within an exception, Section 109 preserves the burden of proving a fact especially within a person’s knowledge, and the presumptions follow in the old order, including legitimacy in Section 116, abetment of suicide by a married woman in Section 117, dowry death in Section 118, and the general power to presume facts in Section 119. The old authorities govern unchanged.

Q: What is Section 62 of the Bharatiya Sakshya Adhiniyam for?

Section 62 is the gateway provision for the proof of electronic records, and the successor of Section 65A of the Indian Evidence Act. It provides that the contents of electronic records may be proved in accordance with the provisions of Section 63. Its function is to direct every question about proving the contents of an electronic record into the single admissibility provision that follows it, where the conditions for receiving a computer output and the certificate requirement are stated. The pairing of Sections 62 and 63 preserves the architecture of the old Sections 65A and 65B: a short directing provision, and a long operative provision. What changed is the operative provision’s content, its certificate form, and its place inside an Act whose definitions already treat the electronic record as a document.

Q: Did the new evidence law change how expert opinion on electronic evidence is taken?

The provision was carried across rather than changed. Under the old Act, Section 45A made the opinion of the Examiner of Electronic Evidence, the office referred to in Section 79A of the Information Technology Act, 2000, a relevant fact when the court had to form an opinion on a matter relating to information transmitted or stored in any computer resource or in any other electronic or digital form. Section 39(2) of the Adhiniyam carries that provision into the expert opinion section itself, immediately after the general rule on experts in Section 39(1). The general law of expert opinion, on foreign law, science, art, and the identity of handwriting or finger impressions, is likewise unchanged, and the authorities on the weight of expert evidence continue to apply.

Q: What happened to the illustrations in the Indian Evidence Act?

They were carried across with the sections they illustrate. The Adhiniyam retains the illustration as a drafting device throughout, so the worked examples that accompanied the old provisions, the placards printed from one original that illustrate primary evidence among them, remain available to courts, practitioners, and candidates in the new numbering. The modernisation of the illustrations was partial: archaic terms and names were updated in some, while the older vocabulary was left standing in others, an inconsistency the committee-stage scrutiny of the bill recorded. Because generations of judgments cite the illustrations in reasoning about the sections, their retention is a substantive continuity and not a cosmetic one: the examples through which the doctrines were taught and applied survived the repeal.

Q: What should a litigant check first when citing an evidence provision after the changeover?

Three things, in order. First, the regime: establish whether the proceeding was pending immediately before 1 July 2024, because Section 170(2) keeps a pending application, trial, inquiry, investigation, proceeding or appeal under the Indian Evidence Act, and the old number is then the governing citation. Second, the successor: where the Adhiniyam governs, convert the citation using a concordance, and name the statute with the section, because several numbers mean different provisions in the two Acts, Section 63 being secondary evidence in the old Act and the electronic records certificate in the new one. Third, the text: compare the successor provision against the old one for added words, moved provisos, and new explanations before relying on an old authority.