Opinion Evidence and Its Limits

The foundational rule in adversarial systems is that witnesses testify to facts within their direct knowledge: what they observed, heard, smelled, or measured. They do not state conclusions from those facts. The inference from facts to conclusion is the court's work, assisted by advocates arguing from the evidence. Opinion evidence short-circuits that process by having the witness do part of the court's reasoning for it.

The exception for experts rests on a frank admission of the court's limitations. Where the subject matter requires specialised skill or knowledge that ordinary experience does not supply, the court cannot reliably draw the inference on its own. A jury asked to determine whether a blood-spatter pattern is consistent with the victim standing or lying at the time of the wound needs the help of someone trained in bloodstain pattern analysis. A court asked to evaluate whether a chemical compound is toxic at a given concentration needs a toxicologist. In those situations, excluding the expert's inference would not protect the fact-finding process; it would impair it.

The exception is therefore admitted reluctantly and fenced. Every legal system that admits expert opinion also imposes conditions: the witness must be qualified, the subject must genuinely require expertise, the opinion must be relevant, and the methodology must be reliable. These conditions are not decorative. In cases where they have not been enforced, miscarriages of justice have followed. The wrongful convictions overturned after review of hair microscopy evidence in the United States, and the cases reconsidered after expert evidence on shaken-baby syndrome was reassessed in England, illustrate the cost of admitting unreliable expert opinion without rigorous testing.

The ultimate issue in a criminal case is whether the accused is guilty of the offence charged. The ultimate issue in a civil case is whether the defendant is liable. The traditional rule prohibited experts from giving an opinion directly on these questions. The concern was that a jury, on hearing a respected expert say "in my opinion the defendant caused the victim's death", might simply defer to that opinion rather than evaluate the evidence independently. The expert's view on the ultimate issue, the argument went, invades the province of the jury.

The strict prohibition has retreated in most systems. The US Federal Rule of Evidence Rule 704(a) expressly allows opinion testimony that embraces an ultimate issue, provided the opinion is not about whether the defendant had the required mental state in a criminal case, which Rule 704(b) still bars. English courts adopt a practical approach: the question is whether the expert opinion assists the court, and opinions on ultimate issues are not barred outright but are scrutinised for whether they actually assist or merely assert a conclusion. Indian courts have applied a similar functional test, treating the ultimate-issue concern as a caution about weight rather than an absolute bar on admissibility.

Regardless of formal rules, the concern about ultimate-issue opinions has practical force. An expert who says "this was arson" is doing something different from one who says "the burn patterns and the location of origin are consistent with a flammable liquid having been poured along the floor and ignited." The second formulation states the finding and leaves the inference to the court. Many experienced forensic scientists prefer the second approach because it is more precise, harder to challenge as exceeding expertise, and more useful: it tells the court what the science shows rather than what the expert thinks the verdict should be.

An expert opinion is only as strong as the facts on which it rests. Courts require that the basis of an expert opinion be disclosed, and they evaluate the opinion in light of that basis. If the facts assumed by the expert are not established in evidence, the opinion built on those facts may carry little weight. If the assumed facts are disputed, the opinion is contingent: it is relevant only if the court accepts the underlying facts.

The basis of opinion has two layers. The first is factual: what data, samples, measurements, or observations did the expert work from? The second is methodological: what technique, protocol, or interpretive framework did the expert apply to those inputs? Both layers must be transparent. An expert who presents a conclusion without stating the factual basis gives the court nothing to evaluate. An expert who states the factual basis but not the methodology invites the objection that a different methodology applied to the same facts might yield a different conclusion.

In India, the Supreme Court has consistently held that the opinion of an expert must be based on identified facts and that the court must scrutinise both the qualifications of the expert and the soundness of the basis before attaching weight to the opinion: see Ramesh Chandra Agrawal v Regency Hospital Ltd (2009) 9 SCC 709. In England and Wales, the Criminal Procedure Rules require the expert's report to state the facts on which each opinion is based and to identify whether any of those facts are within the expert's own knowledge or are assumed. The US Daubert framework requires the trial judge to assess, at a threshold level, whether the methodology is sound before the jury ever hears the opinion.

The principle that courts are not bound by expert opinion, even uncontradicted expert opinion, is stated in every major jurisdiction. It is not a formality. Courts have rejected expert evidence on firearms identification, handwriting comparison, and forensic pathology after determining that the underlying reasoning was inadequate, the methodology was not sufficiently validated, or the conclusion did not follow from the data.

The reasons why a court might reject an expert opinion fall into several categories. The expert may lack the specific qualifications relevant to the question at hand. The methodology may not be accepted in the relevant field or may not have been reliably applied to the facts of the case. The opinion may be internally inconsistent, or the expert may have changed their position under cross-examination in a way that undermines confidence in the conclusion. The opinion may rest on assumed facts that the court finds unproven. Or the court may simply find that the expert has overstated the strength of the scientific support for their conclusion.

Indian courts have repeatedly affirmed this principle. In State of Himachal Pradesh v Jai Lal (1999) 7 SCC 280, the Supreme Court held that the opinion of an expert is not conclusive and is to be taken only as a piece of evidence. The court retains full power to evaluate it alongside other evidence and to accept or reject it. The same position is taken in English case law and is embedded in the US system through the gatekeeping function of the trial judge and the jury's right to assess the weight of admitted expert evidence.

• Qualification gap: The witness's expertise does not cover the specific question. A general chemist opining on pharmacokinetics may fall outside their competence.
• Methodology not validated: The technique has not been peer-reviewed, lacks a known error rate, or is not accepted in the field.
• Unreliable basis: The data the expert relied on is contaminated, incomplete, or not authenticated.
• Overstated certainty: The expert expresses more confidence than the science supports; cross-examination reveals that the conclusion is probabilistic, not definitive.
• Conflicting evidence outweighs: Other evidence in the case, including witness testimony or documentary evidence, contradicts the expert's opinion and the court prefers it.

The question of whether expert opinion should be admitted at all, separate from the question of what weight it deserves, is handled differently across systems. The two most influential models come from the United States, but their underlying logic has influenced courts globally.

The Frye standard, from Frye v United States (D.C. Cir. 1923), asks whether the scientific principle from which the expert's deduction is drawn has gained general acceptance in the relevant scientific community. Frye is binary and conservative: if the technique is not yet generally accepted, the opinion based on it cannot go to the jury. Several US states continue to apply Frye.

The Daubert standard, from Daubert v Merrell Dow Pharmaceuticals (US Supreme Court, 1993), replaced Frye in federal courts under Federal Rule of Evidence 702. Daubert requires the trial judge to assess four non-exclusive factors: whether the theory can be and has been tested; whether it has been peer-reviewed and published; the known or potential error rate; and whether it is generally accepted. Unlike Frye, Daubert permits the admission of emerging science that has been validated but not yet universally accepted, and it focuses attention on the reliability of the specific application of the methodology, not just the methodology in principle. Two subsequent Supreme Court cases, General Electric Co v Joiner (1997) and Kumho Tire Co v Carmichael (1999), extended Daubert to non-scientific technical expertise and clarified the abuse-of-discretion standard for appellate review of gatekeeping decisions.

England and Wales do not have a formal equivalent of Daubert gatekeeping. Admissibility is governed by whether the evidence is relevant and whether it will assist the court, and reliability concerns are largely left to cross-examination and judicial directions on weight. However, the Law Commission's 2011 report on expert evidence recommended a reliability test broadly analogous to Daubert, and the courts have moved in that direction in practice. India's BSA 2023 preserves the admissibility gate of "persons specially skilled": the court determines at the threshold whether the witness qualifies, and the weight of the opinion is assessed separately. See also the discussion in Admissibility Standards Around the World for a broader comparative survey.

A structural feature of the adversarial system is that experts are typically instructed and paid by one side. That creates an obvious tension: the expert's commercial relationship is with the party whose case the opinion supports, but the expert's duty is to tell the truth to the court. Courts across jurisdictions have developed rules to manage this tension, and the rules share a common core: the expert's duty to the court is higher than the duty to the instructing party.

In England and Wales, this duty is explicit in the Criminal Procedure Rules Part 19 and the Civil Procedure Rules Part 35. The expert must state in the report that they understand their duty to the court, that they have complied with it, and that their report contains only matters within their expertise. In the United States, expert witnesses do not have a codified duty of this kind in the same terms, but courts and bar associations enforce ethics rules that prohibit giving opinions the expert does not honestly hold. The Indian Supreme Court has endorsed the principle that an expert witness is a friend of the court, not a partisan, citing the need for objectivity in evidence that laypersons cannot independently evaluate.

The limits of opinion also include the scope of the expert's discipline. A forensic pathologist may give an opinion on cause of death but not on the credibility of an eyewitness account. A handwriting examiner may give an opinion on the authorship of a document but not on the accused's psychological state. Opinions that stray outside the field in which the expert is qualified are vulnerable to exclusion or to being given no weight, and they expose the expert to professional consequences. Staying within scope is both a legal requirement and a professional obligation. For a fuller discussion of the expert's reporting duties, see The Expert Report: Structure and Duties.