Landmark Judgments on Expert Evidence

Under the Bharatiya Sakshya Adhiniyam 2023, expert opinion becomes relevant under section 39 when the court must form an opinion on a point of science or art, a question of foreign law, or the authenticity of handwriting or finger impressions. The word 'relevant' is precise in Indian evidence law: it means the opinion may be placed before the court, not that it is conclusive or automatically credited. This reflects a deliberate legislative choice to keep experts in an advisory role.

The Supreme Court of India has elaborated on this in a line of decisions. In Ramesh Chandra Agrawal v. Regency Hospitals Ltd (2009), the Court held that expert evidence is opinion evidence of an advisory character and the court is not bound by it. The court must assess the expert's qualifications, the basis of the opinion expressed, and whether the facts on which the opinion rests have themselves been proved. An opinion offered on an unproved foundation carries no weight. In Murari Lal v. State of MP (1980), the Court applied this reasoning to handwriting evidence specifically, holding that the opinion of a handwriting expert is a weak type of evidence and a conviction based solely on it, without corroboration, is unsafe.

Indian courts also scrutinise the expert's institutional standing. An opinion from a government-accredited laboratory carries more initial credibility than one from a private consultant, though both are subject to cross-examination on methodology. The Bharatiya Nagarik Suraksha Sanhita 2023 (which replaced the Code of Criminal Procedure) contains provisions for sending exhibits to designated government laboratories, and reports from those laboratories are admissible without the scientist necessarily appearing in person unless the court or a party requires it. This shifts the default but does not remove the court's right to demand live testimony and cross-examination.

The US Supreme Court decided three cases that together define the federal standard for expert admissibility. Daubert v. Merrell Dow Pharmaceuticals (1993) held that Federal Rule of Evidence 702 displaced the Frye general-acceptance test. The trial judge must determine whether the proposed testimony rests on sufficient facts or data, is the product of reliable principles and methods, and has been applied reliably to the facts of the case. The Court listed four non-exhaustive factors to assess reliability: whether the theory can be or has been tested; whether it has been subject to peer review and publication; the known or potential error rate; and general acceptance within the relevant scientific community.

General Electric Co v. Joiner (1997) added that appellate courts review a district judge's gatekeeping decisions under the deferential abuse-of-discretion standard, not de novo. This gives trial judges wide latitude to exclude experts whose methodologies they find unreliable, and makes such exclusions difficult to overturn on appeal. The practical effect was to make the gatekeeping hearing consequential: losing that hearing often ends the case.

Kumho Tire Co v. Carmichael (1999) extended the Daubert framework beyond scientific testimony to all expert testimony governed by Rule 702, including engineers, economists, and forensic practitioners whose work may be more technical or experiential than strictly scientific. The Court held that a district judge has the same gatekeeping obligation for any kind of expert testimony and may adapt the Daubert factors or apply others appropriate to the discipline. Kumho matters for forensic disciplines such as fingerprint comparison, toolmark analysis, and handwriting examination that do not fit neatly into a scientific hypothesis-testing framework.

The English Court of Appeal addressed the threshold for expert admissibility in R v. Turner (1975). The Court held that expert evidence is admissible when it provides information likely to be outside the knowledge and experience of a judge or jury. On the facts, psychiatric evidence about normal human emotions and reactions was not admissible because the jury could form its own judgment without expert assistance. Turner established the necessity principle: expert opinion is allowed only where lay persons genuinely need it, not as a convenient way to dress up a party's argument in scientific clothing.

R v. Bonython (1984), decided by the Full Court of South Australia, articulated a two-limb test widely adopted in English courts: first, is the subject matter of such a kind that the opinion of an expert can assist the court in forming a conclusion on it; and second, does the witness have sufficient knowledge or experience to render an opinion on the subject? Bonython is cited regularly in English voir dire arguments even though it originates from Australian jurisdiction, because English courts found its formulation more precise than Turner alone.

Part 19 of the Criminal Procedure Rules 2020 now codifies the expert witness's duties in England and Wales. An expert's duty is to the court, not to the party who instructs them. The report must state the expert's qualifications, the facts and assumptions relied on, any range of opinion on the relevant issue, and a summary accessible to non-specialists. The Court of Appeal has repeatedly excluded expert evidence where the report failed these requirements, most notably in cases involving novel forensic techniques such as cell-site analysis, where the underlying methodology was either insufficiently disclosed or not adequately validated.

Fingerprint evidence was once treated by courts as near-infallible. The Scottish case of McKie v. Chief Constable of Strathclyde (Scottish High Court, 1999) shook that assumption. Shirley McKie was prosecuted partly on the basis of a fingerprint identification later shown to be wrong. The wrongful identification and the institutional resistance to accepting the error prompted a government inquiry and forced courts and practitioners to acknowledge that fingerprint comparison involves human judgment, not algorithmic certainty, and that examiners can reach different conclusions from the same mark.

In the United States, several district courts post-Daubert examined whether fingerprint comparison satisfied the reliability requirements of Rule 702. In United States v. Llera Plaza (3d Cir. 2002), the Third Circuit allowed fingerprint testimony but noted that the ACE-V method (Analysis, Comparison, Evaluation, Verification) had not been subjected to the kind of empirical validation that would allow computation of a known error rate. The court permitted examiners to testify to their conclusions while acknowledging this methodological limitation. Subsequent National Academy of Sciences and PCAST reports in 2009 and 2016 formally documented the absence of validated error rates for several pattern-based forensic disciplines, including fingerprints and toolmarks.

DNA profile evidence generates a match probability: the probability that a random unrelated person from the relevant population would share the same profile. Courts have accepted this evidence in every major jurisdiction, but the case law is filled with warnings about how the probability figure is communicated and understood. The prosecutor's fallacy, also called the transposition of the conditional, occurs when the probability of the evidence given innocence is stated as though it were the probability of innocence given the evidence. In R v. Adams (No 2) (1998), the English Court of Appeal discouraged Bayesian updating by juries using formal calculations, on the basis that introducing numerical probability reasoning into the deliberation room would distort rather than assist the process.

In India, the DNA Profiling Bill has been in legislative process for several years; as of 2024, there is no standalone DNA evidence statute, and the admissibility of DNA evidence is handled under the general provisions of the Bharatiya Sakshya Adhiniyam 2023 covering scientific opinion, read with constitutional right to privacy principles from Justice K.S. Puttaswamy v. Union of India (2017). The Supreme Court in Sharda v. Dharmpal (2003) addressed DNA testing in the matrimonial context and acknowledged that courts can direct DNA tests in appropriate cases, though the right against self-incrimination constrains compelled testing in criminal cases.

The European Court of Human Rights addressed the retention of DNA profiles in S. and Marper v. United Kingdom (2008), holding that indefinite retention of DNA data of persons not convicted violated the right to private life under Article 8 of the European Convention. This decision shaped UK legislation through the Protection of Freedoms Act 2012 and has influenced debates about DNA databases in other Council of Europe member states. The judgment illustrates that expert evidence jurisprudence extends beyond the courtroom to the policies that govern how forensic data is collected and stored.

Every generation of forensic science produces techniques that arrive in court before the scientific community has fully validated them. Digital forensics provides contemporary examples: courts have had to assess the reliability of cell-site location analysis, social media account attribution, and file metadata as forensic evidence without the benefit of decades of peer-reviewed literature. The judicial response across jurisdictions has been to require the proponent to demonstrate the foundation of the technique before it is accepted, and to treat the absence of error rate data as a factor reducing weight even when it does not trigger exclusion.

In the UK, the Court of Appeal addressed cell-site evidence in R v. Jama (2008) and subsequent cases, noting that cell-site analysis tells a court where a phone was likely connected to the network, not where the phone's user was. Experts who overstated the positional precision of cell-site evidence were criticised. The court required experts to disclose the coverage areas of cells, the signal strength assumptions used, and the degree of uncertainty in any locational inference.

The common thread across India, the US, and England and Wales is that courts do not automatically defer to novelty. A new technique carries a higher burden of methodological disclosure precisely because it has not yet accumulated the track record of peer review and adversarial testing that older techniques possess. Forensic practitioners introducing new methods should expect more searching voir dire examination, not less, and should prepare methodological transparency as the first line of defence.