Discussion of Evidence — Charge-wise Analysis

Every part of a criminal judgment before this point is preparation. The cause-title identifies the contest, the statement of the prosecution case narrates the accusation, the charge crystallises the legal allegation, and the plea records the accused's denial. But it is in the discussion of evidence that the court actually decides — sifting testimony, mapping proof against statutory ingredients, and recording the reasons that Section 354 of the Code of Criminal Procedure, 1973 (now Section 393 of the Bharatiya Nagarik Suraksha Sanhita, 2023) makes mandatory. This chapter explains how a judiciary aspirant should structure that discussion charge by charge, accused by accused, so that the eventual order of conviction or acquittal flows visibly from reasons and not from assertion.

Why the Discussion of Evidence Is the Heart of the Judgment

A criminal judgment is not a chronicle; it is an act of reasoning. The discussion of evidence is the portion in which the court demonstrates how it reached its conclusion, and it is the portion an appellate court scrutinises first. Sections 354(1)(b) of the Code of Criminal Procedure, 1973 and its successor, Section 393(1)(b) of the Bharatiya Nagarik Suraksha Sanhita, 2023, require that every judgment "contain the point or points for determination, the decision thereon and the reasons for the decision." Those three limbs — points, decision, reasons — are the skeleton of the discussion. A judgment that records a decision without reasons, or reasons that do not engage with the actual evidence, is liable to be set aside as non-speaking.

The discussion also performs a constitutional function. A reasoned judgment is the visible guarantee that the presumption of innocence has been respected and that conviction rests on proof beyond reasonable doubt rather than suspicion. For the aspirant, mastering this section is therefore not a matter of drafting elegance but of judicial discipline: it is where the law of evidence is applied, not merely cited. This chapter assumes you have already framed the charge correctly and recorded the plea; the discussion now tests whether the charge is made out.

The Statutory Mandate: Points, Decision and Reasons

Both Section 354 of the old Code and Section 393 of the new Sanhita are identically structured on the essentials. The judgment must be in the language of the court; it must set out the point or points for determination; it must record the decision on each; and it must give the reasons for that decision. Where there is a conviction, the judgment must specify the offence and the section of the penal law (the Indian Penal Code, 1860, or now the Bharatiya Nyaya Sanhita, 2023) under which the accused is convicted, together with the sentence. Where there is an acquittal, the judgment must state the offence of which the accused is acquitted and direct that he be set at liberty.

A subtle but examined feature of both provisions is the alternative-conviction clause: where it is doubtful under which of two sections, or which of two parts of the same section, the offence falls, the court must distinctly express the doubt and pass judgment in the alternative. This is a drafting discipline the discussion must anticipate — for instance, where the evidence is equally consistent with culpable homicide amounting to murder and culpable homicide not amounting to murder, the court frames the point on both and reasons towards the part that the proof actually supports. The statutory architecture, in short, demands that the discussion be organised around points, and the most natural unit of a "point" in a criminal trial is the charge.

Framing Points for Determination, Charge by Charge

The cardinal organising principle of the discussion is this: one point of determination for each charge, against each accused. If three accused are charged under Section 302 read with Section 34 of the Indian Penal Code, and one of them additionally under Section 27 of the Arms Act, the court should not write a single undifferentiated narrative of "who did what." It should pose discrete questions — whether the prosecution has proved that the deceased died a homicidal death; whether the accused shared a common intention; whether each named accused participated; and whether the firearm offence is independently established. Each question is then answered by reference to the evidence that bears on it.

This charge-wise discipline guards against a recurring error: convicting on a general impression of guilt rather than on proof of the specific offence charged. The framing flows directly from the statement of the prosecution case and the charge sheet. A useful habit is to reproduce, at the head of the discussion, the points for determination in numbered form, mirroring the charges, and then to address them seriatim. This is not mere formalism: in Masalti v. State of Uttar Pradesh (AIR 1965 SC 202), where forty persons were tried on a charge under Section 302 read with Section 149, the Supreme Court stressed that the court must satisfy itself, charge by charge and accused by accused, that the participation of each is independently established, particularly where the evidence is partisan.

Mapping Evidence to Statutory Ingredients

Once the point is framed, the discussion must break the offence into its statutory ingredients and test the evidence against each. A charge of murder under Section 302 of the Indian Penal Code (Section 103 of the Bharatiya Nyaya Sanhita, 2023) requires proof that a human being is dead; that the death was caused by the accused; and that the act was done with the intention or knowledge specified in Section 300. The discussion should walk through these one at a time: the fact and cause of death are usually established by the post-mortem report and the medical witness; the authorship of the act by ocular or circumstantial evidence; and the mental element by the nature of the weapon, the seat of injury and the surrounding circumstances.

Ingredient-mapping is what separates a competent judgment from a mere summary of depositions. It also makes the reasoning auditable: if the prosecution proves death and authorship but the medical evidence shows a single blow in a sudden quarrel, the discussion can reason its way to Exception 4 of Section 300 and a conviction under Section 304 Part I instead of Section 302 — precisely the alternative-finding contemplated by the statute. The aspirant should resist the temptation to discuss evidence in the order the witnesses were examined; the better method is to discuss it in the order the ingredients must be proved.

Appreciation of Oral Evidence: Quality Over Quantity

Indian law does not count witnesses; it weighs them. Section 134 of the Indian Evidence Act, 1872 (Section 139 of the Bharatiya Sakshya Adhiniyam, 2023) declares that no particular number of witnesses is required for the proof of any fact. In Vadivelu Thevar v. State of Madras (AIR 1957 SC 614), the Supreme Court upheld a conviction resting on a single eyewitness and laid down the enduring proposition that "it is not the number, the quantity, but the quality that is material" — evidence is to be weighed, not counted. A solitary, wholly reliable witness can sustain a conviction; a crowd of unreliable ones cannot.

The discussion should therefore classify each material witness and state, with reasons, the category into which the court places that witness — wholly reliable, wholly unreliable, or neither wholly reliable nor wholly unreliable. For the third, intermediate category, the court must look for independent corroboration on material particulars before acting on the testimony. This classification, drawn from Vadivelu Thevar, is the analytical spine of oral-evidence appraisal and should be applied transparently rather than assumed. The court's task, as the same judgment puts it, is to assess whether the testimony has a "ring of truth" and is cogent, credible and trustworthy.

Handling Contradictions, Omissions and Improvements

No two truthful witnesses describe an event identically, and a judgment that discards testimony for trivial inconsistency is as defective as one that overlooks a fatal one. The governing authority is State of Rajasthan v. Smt. Kalki ((1981) 2 SCC 752), in which the Supreme Court distinguished between "normal" and "material" discrepancies. The Court observed that in the depositions of witnesses there are always some normal discrepancies, however honest and truthful they may be, arising from normal errors of observation, errors of memory due to lapse of time, and mental disposition such as shock at the time of occurrence. Such discrepancies do not corrode credibility; only material discrepancies that touch the core of the prosecution case do.

The discussion must therefore distinguish, with reasons, between a contradiction that goes to the root (for example, the witness placing the accused elsewhere at the time of the act) and one that is peripheral (whether the deceased was standing or lying when struck — the very kind of detail the Court in Kalki treated as immaterial). Omissions amounting to contradictions under the proviso to Section 162 of the Code of Criminal Procedure must be confronted in cross-examination and put to the investigating officer before the court can act on them; the discussion should record whether that procedural requirement was satisfied. The court should read the testimony as a whole and ask whether, despite imperfections, it carries conviction.

Falsus in Uno and the Grain-and-Chaff Approach

A frequent appellate complaint is that the trial court rejected a witness wholesale because part of his evidence was false. Indian law does not permit that shortcut. The maxim falsus in uno, falsus in omnibus — false in one thing, false in everything — is not a rule of law in India but, at most, a rule of caution. In Leela Ram (D) Through Duli Chand v. State of Haryana (AIR 1999 SC 3717), the Supreme Court reiterated that the maxim has no application in this country and that the duty of the court is to disengage the truth from falsehood, to sift the grain from the chaff. Where the residue, after discarding the embellished or false portion, is sufficient to establish guilt, conviction can rest on that residue.

The discussion must therefore reason explicitly: it identifies the part of a witness's account it disbelieves, gives the reason, and then asks whether what remains is reliable and adequate. The same approach governs partly hostile witnesses. In Sucha Singh v. State of Punjab ((2003) 7 SCC 643), the Court held that the evidence of a witness declared hostile is not effaced or rendered unworthy of consideration altogether; the portion that is creditworthy and corroborated may be relied upon by the prosecution as well as the defence. A judgment that treats a hostile declaration as automatically erasing the deposition misapplies the law and is vulnerable on appeal.

Related, Interested and Chance Witnesses

In most violent crimes the natural witnesses are relatives or close associates of the victim, and their evidence cannot be discarded merely on that ground. The settled position, traceable to Masalti v. State of Uttar Pradesh (AIR 1965 SC 202), is that it is unreasonable to reject the evidence of a witness only because he is partisan or interested; relationship is often a guarantee of truth rather than a reason for suspicion. What the court must guard against is the tendency of an interested witness to implicate the innocent along with the guilty. The proper judicial response is closer scrutiny, not automatic rejection.

In Masalti the Court approved, as a rule of prudence in cases of large unlawful assemblies, the practice of requiring the overt act of each accused to be spoken to by a consistent and reliable plurality of witnesses before convicting on the basis of a common-object charge — a caution, not an inflexible mandate. The discussion should record the relationship of each witness to the parties, note the test of careful scrutiny it has applied, and explain why the interested testimony is, or is not, worthy of acceptance. A chance witness — one who happens to be present fortuitously — is not for that reason untruthful, but the court should examine whether his presence at the spot is naturally and convincingly explained.

Medical and Expert Evidence: Corroboration and Conflict

Medical evidence ordinarily corroborates and does not, by itself, identify the assailant. The discussion should align the ocular account with the post-mortem findings: do the number, nature and dimensions of the injuries match the weapon and the manner of assault deposed to? Where there is a genuine and irreconcilable conflict between credible eyewitness testimony and the medical opinion, the law generally prefers the direct evidence, because medical opinion is an opinion and rarely so definitive as to exclude the eyewitness account altogether. But where the medical evidence completely rules out the possibility of the injury having been caused in the manner alleged, the ocular version is rendered doubtful and the benefit must go to the accused.

The same caution attends other expert evidence. The court is not bound to accept an expert's opinion mechanically; it must apply its own mind to the data on which the opinion rests. In State of Uttar Pradesh v. Krishna Gopal (AIR 1988 SC 2154), the Supreme Court warned against converting forensic probability into mathematical certainty and emphasised that the evaluation of evidence ultimately rests on robust common sense and the trained intuition of the judge rather than on the ipse dixit of an expert. The discussion should therefore treat expert testimony as one strand to be woven with the rest, never as a self-executing verdict.

Circumstantial Evidence: The Panchsheel of Proof

Where there is no eyewitness, the discussion must marshal the chain of circumstances with special rigour. The foundational test was laid down in Hanumant v. State of Madhya Pradesh (AIR 1952 SC 343), where the Supreme Court held that the circumstances from which guilt is inferred must be fully established and must be consistent only with the hypothesis of guilt, excluding every other reasonable hypothesis. That test was refined into the celebrated five "golden principles" — the panchsheel of circumstantial proof — in Sharad Birdhichand Sarda v. State of Maharashtra (AIR 1984 SC 1622).

Under Sharad Sarda, (i) the circumstances from which guilt is drawn must be fully established; (ii) the facts so established must be consistent only with the hypothesis of guilt; (iii) the circumstances must be of a conclusive nature and tendency; (iv) they must exclude every possible hypothesis except guilt; and (v) the chain of evidence must be so complete as to leave no reasonable ground for any conclusion consistent with innocence and must show that in all human probability the act was done by the accused. The discussion of a circumstantial case should be written as a chain: each link (last seen together, motive, recovery, conduct) is established separately, and then the court demonstrates that the links join up to exclude innocence. A missing or broken link entitles the accused to acquittal, however strong the suspicion.

Last Seen, Recovery and the Limited Burden Under Section 106

Two recurring circumstances deserve careful handling. The "last seen together" theory, explained in State of Rajasthan v. Raja Ram ((2003) 8 SCC 180), applies where the time gap between the accused and the deceased being last seen alive together and the recovery of the body is so small that the possibility of any other person being the author of the crime is practically ruled out. Where the gap is wide, the circumstance loses its force, because others could have intervened. The discussion must therefore fix the time of last sighting and the time of death with care, and reason about the gap rather than asserting last-seen as if it were self-proving.

Section 106 of the Indian Evidence Act, 1872 (Section 109 of the Bharatiya Sakshya Adhiniyam, 2023) supplies the complementary principle: when a fact is especially within the knowledge of the accused, the burden of proving it lies on him. So, once the prosecution proves that the deceased was last seen alive in the company of the accused, an obligation arises on the accused to explain what became of the victim; an absent or false explanation becomes an additional link in the chain. The discussion must, however, deploy Section 106 carefully: it does not relieve the prosecution of its primary burden of proof, and a conviction cannot be founded on the accused's silence alone. The recovery of a weapon under Section 27 of the Evidence Act (Section 23 of the Bharatiya Sakshya Adhiniyam) is admissible only to the extent of the information distinctly leading to discovery, and the discussion should confine itself to that admissible portion.

The Standard of Proof and the Benefit of Reasonable Doubt

The discussion must keep the correct standard of proof in view throughout, and articulate it when recording findings. The prosecution must prove its case beyond reasonable doubt; the accused is not required to prove his innocence. But reasonable doubt is not vague conjecture. In State of Uttar Pradesh v. Krishna Gopal (AIR 1988 SC 2154), the Supreme Court held that a reasonable doubt is a fair doubt grounded in reason and common sense, growing out of the evidence, and not an imaginary, trivial or merely possible doubt; the law cannot afford any favourite other than truth. The discussion must avoid both extremes — convicting on suspicion and acquitting on fanciful doubt.

Where, after a full appraisal, two views are reasonably possible on the evidence, the discussion must adopt the one favourable to the accused. This is the "golden thread" running through criminal law, affirmed in Kali Ram v. State of Himachal Pradesh (AIR 1973 SC 2773): if the evidence is capable of two interpretations, one consistent with guilt and the other with innocence, the accused is entitled to the benefit of doubt and to acquittal. The discussion should reach this conclusion explicitly and with reasons, identifying the competing inferences and explaining why the prosecution's has not displaced the defence's. This reasoning is the formal bridge between the appraisal of evidence and the operative order.

Structuring Findings for Multiple Accused and Multiple Charges

When the trial involves several accused and several charges, the discussion must be tabulated in the writer's mind, if not on paper, so that no charge against any accused is left undecided. The disciplined sequence is: take the first point (charge) for determination; appraise all the evidence bearing on it; record a clear finding of "proved" or "not proved" against each accused individually; and then move to the next point. The court must never reach a composite verdict that leaves it unclear which accused has been convicted of which offence. This is the practical meaning of the requirement, in the structure of a criminal judgment, that points, decision and reasons be separately recorded.

Constructive liability charges demand extra care. For Section 34 (common intention) the discussion must find a prior meeting of minds and participation; for Section 149 (common object) it must find membership of an unlawful assembly and that the offence was committed in prosecution of, or as a known likelihood of, the common object. Masalti remains the touchstone for testing such evidence against each accused. Where the court convicts under one section and acquits under another, or convicts in the alternative, it must say so distinctly, as the statute commands. A clean charge-wise table of findings is the surest protection against an appellate remand for a non-speaking judgment.

From Findings to the Operative Order

The discussion ends by gathering its charge-wise findings into the operative conclusion that the rest of the judgment will formalise. If the court finds the offence proved, it records a conviction, specifying the exact offence and the section of the penal law, and proceeds to hear the accused on sentence — under Section 235(2) of the Code of Criminal Procedure, 1973 (Section 258(2) of the Bharatiya Nagarik Suraksha Sanhita, 2023) in a sessions trial, the accused has a statutory right to be heard on the question of sentence, and the discussion's findings on aggravating and mitigating circumstances feed directly into that stage. If the court finds the offence not proved, it records an acquittal, states the offence acquitted of, and directs that the accused be set at liberty.

What links the discussion to the order is the chain of reasons. An appellate court reading the judgment should be able to trace, without guesswork, how the framed point became a finding and how the finding became the order. This is why the discussion of evidence is the most consequential section of the entire judgment, and why aspirants should treat it as the culmination of everything taught in the criminal judgment writing series — from the statutory basis of judgment writing to the framing of the charge. A judgment whose discussion is sound rarely fails on appeal; one whose discussion is conclusory rarely survives it.