Supreme Court judgments and legal records

Rewritten judgments arranged for legal reading and reference.

Jnanendra Nath Ghose vs The State Of West Bengal

Rewritten Version Notice: This is a rewritten version of the original judgment.

Court: supreme-court

Case Number: Criminal Appeal No. 101 of 1958

Decision Date: 8 May, 1959

Coram: Syed Jaffer Imam, J.L. Kapur

In the matter of Jnanendra Nath Ghose versus the State of West Bengal, the Supreme Court of India delivered its judgment on 8 May 1959. The case was reported in 1959 AIR 1199 and 1960 SCR (1) 126, and the bench consisted of Justice Syed Jaffer Imam and Justice J. L. Kapur. The dispute concerned an accusation of murder tried before a Sessions Judge in Birbhum, with the assistance of a jury, under the provisions of the Indian Evidence Act, 1872, specifically section 133, relating to the admissibility and reliability of an approver’s testimony and the requirement of corroboration. The appellant, Jnanendra Nath Ghose, was charged with murder, and the prosecution’s case rested primarily on the testimony of an approver together with circumstantial evidence that was said to corroborate the approver’s statements and to connect the appellant to the crime. The jury returned a verdict of guilt, and the Sessions Judge, accepting that verdict, sentenced the appellant to life imprisonment. The appellant appealed the conviction to the Calcutta High Court, which dismissed the appeal after finding no misdirection in the charge given to the jury. The appellant then raised, before this Court, the contention that the jury had not been instructed, as required by the precedent set in Sarwan Singh v. State of Punjab [1957] SCR 953, to apply a double test to the approver’s evidence—namely, that the approver must be a reliable witness and that his testimony must be sufficiently corroborated. The appellant argued that the corroborative evidence presented was insufficient to link him to the offence. The Court held that no misdirection occurred. It observed that the requirement in Sarwan Singh concerning the reliability of the approver was invoked only because, in that case, the approver’s testimony was found to be wholly unreliable due to serious inconsistencies. In the present proceeding, there was no indication that the approver’s evidence lacked reliability, and therefore the observations in Sarwan Singh were distinguished. The Sessions Judge had correctly directed the jury that corroboration must relate not merely to the occurrence of the crime but also to the connection of the accused with that crime. The evidence of material circumstances in this case was found to corroborate the approver’s testimony and to link the appellant with the murder. Having established such corroboration, the issue of sufficiency was left to the jury’s assessment. The appeal, numbered Criminal Appeal No. 101 of 1958, was filed by special leave from the Calcutta High Court’s order dated 12 September 1956, which itself arose from the Sessions Judge’s order dated 8 December 1955 in Sessions Trial No. 1 of November 1955. Counsel for the appellant appeared on behalf of the appellant, while counsel for the respondent represented the State of West Bengal.

In this case the Court recorded that the appellant had been sentenced to life imprisonment under section 302 by the Sessions Judge of Birbhum. The Sessions Judge had concurred with the majority verdict of the jury that the appellant was guilty of murder. The appellant challenged that conviction by filing an appeal in the Calcutta High Court. The High Court examined whether the Sessions Judge had misdirected the jury in his charge and concluded that there was no misdirection; consequently it dismissed the appellant’s appeal. In the same trial two other persons, Jagdish Gorain and Sudhir Gorain, were tried together with the appellant. The jury acquitted both of them and the Sessions Judge accepted that acquittal. The appellant then sought from the High Court a certificate that would permit him to appeal to the Supreme Court, but the certificate was refused. The present proceedings therefore arise by way of special leave to appeal. According to the prosecution, the deceased Sibapada Hati was married to a woman named Lila. Approximately one month before the incident, the appellant had proposed to Lila that she should live with him; Lila rejected the proposal. The appellant is alleged to have believed that eliminating Sibapada Hati would remove an obstacle and would increase his chances of winning Lila’s favour. On 26 May 1955 the appellant is said to have murdered Sibapada Hati. The prosecution narrated that the murder was carried out with the assistance of Jagdish Gorain, Sudhir Gorain and an approver, Sastipada Ghose. The conviction of the appellant therefore rested on two sources of proof: the testimony of the approver and circumstantial evidence that was said to corroborate the approver’s account and to connect the appellant with the killing of Sibapada Hati. The Court reiterated that, unless the charge given to the jury contained a misdirection or a failure to give necessary direction that resulted in a miscarriage of justice, the jury’s verdict must stand and cannot be disturbed. The High Court had held that the charge was correct, and this Court agreed with that assessment after reviewing the charge. The Sessions Judge, when addressing the jury about the approver’s evidence, made the following statements: “Before proceeding, it is necessary to explain some established legal principles concerning the evidence of an approver and the confessions on which the prosecution relies in this case. An approver is a competent witness against an accused person, and although his testimony is strictly admissible and a conviction based solely on such testimony is not illegal, established practice dictates that a person should not be convicted on approver evidence except in very rare and exceptional cases, and usually substantial corroboration is required. Accordingly, I warn you that it is highly dangerous to convict solely on the basis of an approver’s evidence. While it is legally possible to convict on uncorroborated accomplice testimony, and such testimony is admissible, experience shows that an accomplice, being always an unreliable person, must be treated with extreme caution and his evidence must be materially corroborated before a conviction can be justified.”

The judge warned that the approver was an infamous individual who had abandoned his associates and friends in order to protect himself, and although he was a criminal who had secured his freedom through betrayal, his testimony had to be received with extreme caution and should not be acted upon unless it was materially corroborated. He explained that the rule requiring corroboration had become a settled practice of such universal application that it now possessed almost the force and reverence of law. The judge further clarified that corroboration must relate both to the commission of the crime and to the identification of each accused person, and that the required corroboration must consist of independent and reliable evidence of a different kind. He emphasized that such corroborative evidence must be independent testimony that connects, or at least tends to connect, the accused with the offence. In other words, the evidence must implicate the accused by matching, in material particulars, not only the fact that the crime occurred but also that the accused was the perpetrator. The judge stressed that corroborative evidence is evidence that shows, or tends to show, that the accomplice’s story that the accused committed the crime is true. He noted that the corroboration need not be direct proof of the accused’s participation; it could be sufficient if it were merely circumstantial evidence indicating a connection between the accused and the crime. The corroboration in material particulars must be such as to connect or identify each accused with the offence. Regarding the present matter, the judge pointed out that a previous statement of the approver, namely a confession, had been exhibited, but that this earlier statement could not be used to corroborate the later statements made by the approver before the Court. He instructed the jury that, in determining the amount of corroboration required, they must exercise careful discrimination and consider all surrounding circumstances to decide whether the facts stated by the approver Sastipada were supported by those circumstances. Counsel for the appellant, Mr. Umrigar, argued that the direction given by the Sessions Judge to the jury was inadequate. He contended that the jury should have been instructed, in accordance with the decision of this Court in Sarwan Singh v. The State of Punjab (1) [1957] S.C.R. 953, that the approver’s evidence must satisfy a double test: the approver must be shown to be a reliable witness and his evidence must receive sufficient corroboration. He further asserted that the evidence of an approver must be confirmed not only regarding the circumstances of the crime but also regarding the identity of the prisoner. The required corroboration, he said, ought to consist of circumstances affecting the identity of the accused, relying on The King v. Baskerville (1). Finally, counsel maintained that the circumstantial evidence offered to corroborate the approver was insufficient to link the appellant with the murder of the deceased.

It was observed that the murder of the deceased had likewise been connected to an approver’s testimony, and that in a similar situation the accused Jagdish Gorain had been acquitted on the basis of comparable corroboration of the approver’s evidence. The Court held that there was no genuine distinction between the circumstances of Jagdish Gorain’s case and those of the present appellant. The Court recalled the ruling in Sarwan Singh’s case, where it was declared that the assessment of an approver’s testimony must satisfy a double test. First, the approver must be shown to be a reliable witness, and second, his testimony must receive sufficient corroboration; the requirement of reliability is common to all witnesses, whereas the need for corroboration is a special condition applicable to weak or tainted evidence such as that of an approver. These observations were made in the special factual context in which the Court was addressing the Sarwan Singh matter.

The Court further observed that the contention that the High Court had failed to consider the character of the approver’s evidence could not be treated as merely academic or theoretical. The Court noted that the approver’s evidence in the present case was so thoroughly discrepant that it was difficult to resist concluding that the approver was a wholly unreliable witness. It was pointed out that the learned judges of the High Court themselves had criticized the approver’s evidence in the prosecution case against Gurdial Singh, ultimately finding the approver’s account unreliable. Although there was circumstantial evidence that raised some suspicion against Gurdial Singh, the Court held that such suspicion was insufficient to sustain his conviction. The Court added that if an approver’s account against one accused is found to be wholly discrepant, that finding should inevitably lead the court to scrutinise the approver’s evidence regarding the other accused with greater caution, citing the decision in The King v. Baskerville (1916) 2 K.B.D. 658.

Consequently, the Court emphasized that in the special circumstances of Sarwan Singh’s case the approver had been declared a wholly unreliable witness. It was stressed that the Court had previously stated that an approver’s evidence must demonstrate the reliability of the witness, a test common to all witnesses. In the present matter, however, nothing had been shown to indicate that the approver’s testimony was unreliable in the manner established in Sarwan Singh’s case. The only assertion was that the approver’s testimony was regarded as tainted evidence, but no evidence was presented to prove that the testimony was intrinsically unreliable.

The Sessions Judge, recognizing the admissibility of approver evidence, instructed the jury that although a conviction based solely on such evidence was not illegal, the settled practice was to avoid convicting a person on approver testimony except in very rare and exceptional circumstances where substantial corroboration was present. The Judge explicitly warned the jury of the danger of relying on an approver’s evidence without sufficient corroboration, thereby directing that conviction could not be based merely on the approver’s testimony unless it was substantively supported by other evidence.

In this case the Court observed that a conviction based solely on the testimony of an approver could be sustained only in rare and exceptional circumstances and that, as a general rule, substantial corroboration of the approver’s evidence was required. The Sessions Judge had warned the jury very clearly about the danger of relying on an approver’s testimony and had instructed the jurors that they should not return a guilty verdict unless the approver’s evidence was supported by independent, substantial corroboration. Apart from the issue of whether such corroboration existed, the Court found that nothing had been submitted to it or to the High Court that would suggest the approver’s evidence was unreliable after it had been examined in the same manner as the evidence of any other prosecution witness. The Court therefore held that the decision in Sarwan Singh’s case was distinguishable because, unlike that earlier case, there was no allegation that the approver’s evidence in the present matter was wholly unreliable when tested according to ordinary standards. Consequently, the Court could not accept the appellant’s contention that the charge to the jury was invalidated by the precedent set in Sarwan Singh.

The Court then turned to the second submission made by counsel for the appellant and recalled the authority from Baskerville’s case, decided by the Court of Criminal Appeal in England. In that decision the Court had examined a number of authorities and articulated a definition of corroboration as independent testimony that affects the accused by connecting, or tending to connect, him with the crime. In other words, the corroborative evidence must not only confirm that the offence was committed but also establish, in some material particular, that the accused was the person who carried it out. The Court emphasized that this test applied equally in situations governed by the common‑law rule of practice and in those where a statute expressly required corroboration, because the statutory language incorporates the common‑law principle. While the exact form of corroboration will vary according to the specific facts of each case, the Court warned against attempting to prescribe a rigid formula. Instead, corroborative evidence should be understood as evidence that shows or tends to show that the accomplice’s account—that the accused committed the crime—is true, rather than merely establishing that the crime occurred. The Court noted that such corroboration need not be direct proof of the accused’s guilt; it is sufficient if it is circumstantial evidence that indicates a connection between the accused and the offence. Accordingly, the Sessions Judge had directed the jury that any corroborative evidence must implicate the accused, confirming both that the crime took place and that the appellant was the one who perpetrated it.

The Sessions Judge instructed the jury that “Corroborative evidence, you should bear in mind, is evidence which shows or tends to show that the story of the accomplice that the accused committed the crime is true. The corroboration need not be direct evidence that the accused committed the crime. It is sufficient if it is merely circumstantial evidence of his connection with the crime. The corroboration in material particulars must be such as to connect or identify each of the accused with the offence.” The Court observed that this direction was consistent with the principle laid down in the Baskerville case and that no serious objection could be taken to the manner in which the Sessions Judge had addressed the jury on this point. The Court further explained that once there is corroborative evidence that connects, or tends to connect, an accused with the crime, that evidence relates to the identity of the accused in relation to that crime. It noted that the approver’s testimony constituted the direct evidence of the commission of the crime, but that the prosecution also required corroboration in material particulars not only concerning the crime itself but also corroboration of the approver’s story by evidence that connects or tends to connect the accused with the offence. According to the Court, it is this corroborative evidence that shapes the mind of a jury to accept the approver’s narrative that the accused committed the crime.

With regard to the third submission advanced on behalf of the appellant, the Court recorded that the jury had accepted the following circumstances as established by the evidence: first, that the appellant possessed a motive to commit the offence because he had made an immoral proposal to Lila, the wife of the deceased; second, that on 25 May 1955 the appellant had visited Lila’s house and conversed with the deceased; third, that on the morning of 26 May 1955 the appellant again came to the house and spoke with the deceased, and later that same day, a little after sunset, the appellant had asked the deceased to accompany him for a walk, to which the deceased acquiesced; fourth, that after the walk the appellant was observed proceeding with Jagdish Gorain, the deceased, Brojeswari and Lila toward the north of the village after five in the evening while they were bathing in the Talbona tank; fifth, that according to the approver, at the time the deceased was stabbed by the appellant, the appellant sustained an injury on the dorsum of his left palm, a fact corroborated by medical evidence showing an almost healed ulcer one inch in length on the posterior surface of the left side of the palm, one inch below the wrist joint, and another healed ulcer of one‑third inch on the posterior surface of the left thumb, both injuries being consistent with an instrument such as a knife; and sixth, that after night fell the deceased had not returned, prompting Lila’s mother to search for him.

Brojeswari and her uncle Radharaman Sadhu began searching for the missing person. They proceeded to the club‑house where the appellant regularly met with his two co‑accused, Jagdish Gorain and Sudhir Gorain, as well as the approver. When Brojeswari questioned the appellant, he initially denied that the deceased had accompanied him and claimed to have no knowledge of the deceased’s whereabouts. After being reminded that it was he who had taken the deceased for a walk, a claim he continued to reject, the appellant altered his answer, stating that the deceased had indeed walked with him up to a canal on the northern side of the village, but that the appellant had left the scene because he suffered a headache and therefore could not provide any further information about the deceased’s location. The Court observed that each of these statements, considered singly or together, might be insufficient to establish that the appellant was responsible for the deceased’s murder. The High Court had previously expressed the view that, apart from the direct testimony of the approver, these surrounding facts would not persuade a reasonable person that the appellant had committed the crime. Nonetheless, the approver’s testimony was the direct evidence asserting that the appellant murdered the deceased, and the jury was tasked with determining whether the ancillary circumstances were adequate to corroborate the approver’s claim. Counsel for the appellant argued that the surrounding facts did not constitute material corroboration linking the appellant to the offence. In the Court’s assessment, at least one of the circumstances did provide material corroboration connecting the appellant to the crime. Specifically, the approver’s account that the appellant sustained an injury to the dorsum of his left palm while committing the murder was supported by medical evidence. Although it was contended that the medical report did not explicitly identify the injury as being on the dorsum of the left palm, the Court found this argument unsubstantial because the first injury described in the report was located on the posterior surface of the left palm, one inch below the wrist joint—precisely the dorsum of the left palm. The second injury, situated on the posterior surface of the left thumb, likewise concurred with the approver’s description of a wound on the dorsum of the left palm. Consequently, the jury was justified in accepting this medical evidence as sufficient corroboration of a material point that connected, or at least tended to connect, the appellant with the murder. Moreover, the fact that the appellant was the individual who summoned the deceased shortly after sunset, escorted him away, and was subsequently observed moving northward in the village together with the deceased—after which the deceased was never seen again—constituted another circumstance upon which the jury could reasonably rely to infer a connection between the appellant and the alleged crime.

The evidence that the appellant was alive was a factor that the jury could rely upon in concluding that it linked or tended to link him with the crime. The appellant’s claimed ignorance about the deceased’s whereabouts that night, together with his later admission that he had taken the deceased toward the north of the village, provided another circumstance the jury could consider as contrary to his claim of innocence. In our opinion, all of the circumstances previously mentioned offered sufficient corroboration of the approver’s testimony that connected or tended to connect the appellant with the crime, and therefore the approver’s assertion that the appellant committed the offence was established as true. The fourth submission argued that, although there was similar corroboration of the approver’s evidence against Jagdish Gorain, he had been acquitted by the jury and no distinction existed between his case and that of the appellant. We consider that argument unsound because, in our view, the two matters are not comparable due to significant differences in motive, circumstances, and the nature of the evidence presented. First, Jagdish Gorain had no motive to commit the murder, which distinguishes his situation from that of the appellant. Second, the injury suffered by Jagdish occurred while he seized the appellant’s knife and questioned, “what have you done?”, indicating that he was attempting to stop further stabbing. Consequently, the approver’s testimony tended to demonstrate that Jagdish was trying to prevent the appellant from inflicting additional wounds on the deceased. These facts could have led the jury to draw a distinction between the case of Jagdish Gorain and that of the appellant. It was for the jury to determine whether the circumstantial evidence was sufficient to connect or tend to connect Jagdish with the crime. On the basis of the approver’s evidence, the jury may have found the circumstances regarding Jagdish to be insufficient corroboration to link him to the offence. In our opinion, there is no valid reason to hold that the charge delivered to the jury by the Sessions Judge was defective in a manner that would invalidate the jury’s verdict. Accordingly, the court dismissed the appeal, finding no error in the trial proceedings that would warrant a reversal of the jury’s finding. The appeal is therefore dismissed, and the judgment of the lower court stands affirmed without any modification or remand. All considerations demonstrate that the jury’s verdict was reached on a sound evidentiary foundation and no ground exists for interference by this court. Consequently, the matter stands concluded, and no further relief shall be granted to either party.