Physical Evidence: Types, Classification, Probative Value

Physical evidence has to satisfy two conditions to be useful in an Indian court. It has to be tangible, in the sense that it can be lifted, packaged, sealed and produced as an exhibit. And it has to be amenable to scientific examination, in the sense that an FSL analyst can do something with it that a witness's testimony alone couldn't. A confession is testimonial, not physical. A bloodstain is physical. A CCTV recording is physical evidence in its storage medium and electronic evidence in its content; Section 65B of the Indian Evidence Act treats the second category specially.

Three broad questions decide whether a recovered item is worth processing:

• Is it relevant to a fact in issue? A loose nail from a homicide kitchen is rarely relevant. A nail from the suspect's shoe tread that matches a nail recovered from the victim's clothing is highly relevant. Relevance is set by the case theory, not by the item.
• Has it been recognised, documented and packaged correctly? The work covered in Processing Physical Evidence at the Scene decides whether what you collect survives to trial.
• Can it be linked back to the source through an unbroken chain? Without the chain laid out in Chain of Custody, even the most beautifully analysed sample is inadmissible.

Edmond Locard, working in Lyon around 1910, proposed that every contact between two surfaces produces a mutual transfer of material. Some material moves from surface A to surface B, and some moves the other way. The principle is the philosophical foundation of trace-evidence work and the reason hair, fibre, soil, glass and gunshot residue are even worth collecting.

The qualification that modern practice adds is this: Locard's principle says transfer happens, not that the transfer survives. A single fibre transferred from a sweater to a car seat may persist for minutes or for months, depending on the substrate, the activity after transfer, and the environment. Modern Indian SOCO training treats Locard as a starting hypothesis (look for the transfer) rather than a guarantee (the transfer will be there to find). When defence counsel argues that the absence of trace evidence proves no contact occurred, the correct rebuttal is that Locard predicts transfer occurs, not that transfer persists to the time of search.

A short non-exhaustive list of what survives well and what doesn't:

• Survives well: glass fragments embedded in skin or clothing, gunshot residue on hands within 6 hours, dried bloodstains on porous fabric, paint chips on rigid substrates.
• Survives poorly: loose surface fibres after vigorous activity, touch DNA after handwashing, latent prints on rough or porous surfaces in humid conditions, gunshot residue after 6 hours or after handwashing.

A class characteristic places an item in a category. Type O blood, .32 calibre, blue cotton fibre, the make and model of a tyre. An individual characteristic identifies one specific source. A specific STR DNA profile, the friction-ridge pattern on a fingerprint, the unique wear marks on a tool's working surface, the breech-face impression a specific firearm leaves on a cartridge case.

Almost all routine forensic work generates class evidence. Individual identification is rare, expensive and tightly defined. The Indian courts treat this distinction explicitly: a class match is corroborative, an individual match can be substantive on its own.

The non-obvious point is that the same item can carry both kinds of characteristic. A cartridge case carries class characteristics (calibre, manufacturer, base stamp) and individual characteristics (breech-face impression, firing-pin indentation, extractor mark). A blood sample carries class characteristics (ABO group, Rh factor) and individual characteristics (STR profile). The forensic report should list both layers, and Indian appellate courts increasingly expect both to be reported even where only one is asked for.

Textbooks vary in how many categories they list, some collapsing to four and some expanding to ten. The taxonomy used here sits at seven classes, mapped to how the FSL routes samples on intake.

A brief note on each class:

• Biological. Blood, semen, saliva, hair root, tissue, bone, urine. ABO grouping is class; STR profiling is individual. Wet samples need air-drying before packaging, and the rules are covered in Processing Physical Evidence at the Scene.
• Trace. Hair shaft (no root), fibre, glass, paint, soil, gunshot residue. Almost entirely class evidence. Glass refractive index, fibre weave, paint layer sequence are the workhorses.
• Impression. Latent and patent prints (friction ridge), shoe and tyre impressions, bite marks, toolmark casts. Individual when ridge minutiae or wear patterns are clear; class otherwise.
• Firearm and toolmark. Cartridge cases, bullets, gunpowder residue, tool wear impressions. Class for calibre and make; individual for breech-face, firing-pin and toolmark wear.
• Document and digital. Questioned documents, ink and paper analysis, handwriting comparison, electronic records, mobile phone forensics, hard-disk imaging. Section 65B IEA governs the digital side.
• Drugs and toxicology. Bulk narcotics (Section 8 NDPS), viscera in poisoning deaths, alcohol estimation, drug-facilitated assault samples. Quantitative chemistry plus chromatography.
• Controlled hazardous substances. Explosives (pre- and post-blast), IED components, accelerants, military or industrial chemicals. NSG and explosive wings handle the most sensitive samples.

The seven-class taxonomy isn't universal. Saferstein collapses to a different five-or-six list; the FBI Handbook uses a different breakdown organised around lab divisions. The seven-class version is what NFSU and most state syllabi test on, and it's the version the CSI tools and mobile lab topic maps kit contents against.

Probative value is the degree to which a piece of evidence makes a contested fact more or less likely. It is a courtroom concept, not a forensic one. The FSL analyst reports the finding; the trial judge decides what weight to attach to it under the Indian Evidence Act, taking into account the chain of custody, the qualifications of the analyst, the methodology used, and any contradictory evidence.

The following ladder reflects roughly how Indian appellate courts have treated each tier of evidence:

1. DNA STR profile (individual)
   Highest probative weight in the absence of contamination claims. The match probability is usually quoted at 1 in a quintillion or higher for full 13-loci profiles. Treated as substantive on its own where the chain holds.
2. Fingerprint with 8+ minutiae (individual)
   Long-standing high weight in Indian courts. The 8-point standard is convention rather than statute; modern practice quotes more minutiae where available.
3. Toolmark or ballistics individual match
   High weight when the breech-face or wear-pattern comparison is independently reviewed. Lower when the comparison rests on a single examiner's opinion.
4. Class evidence (blood group, fibre, glass RI)
   Corroborative. Weight rises when multiple class matches converge on the same suspect; falls when only one class match is offered and a witness is contradicting it.
5. Electronic record under Section 65B
   Admissibility depends on the 65B certificate, not on the content. A perfectly authentic record without a 65B certificate is inadmissible (Anvar P.V., Arjun Panditrao). With a certificate, weight is then judged on content.
6. Negative findings
   Absence of expected trace evidence is rarely a strong defence point unless paired with an active reason transfer should have occurred. Courts treat 'no fibre was found' as a weak negative.

A non-obvious point on probative value: Indian courts have repeatedly held that no single forensic finding is conclusive on its own, including DNA. The reason is procedural, not scientific. A DNA match plus a broken chain of custody is weaker than a DNA match plus an intact chain plus an eyewitness, even though the underlying science is identical. The chain is doing the evidentiary work, not the lab.

The 2023 criminal-law reforms changed the trial-stage handling of physical and digital evidence in three material ways.

• BNSS 2023, Section 176(3): mandatory FSL visit for serious offences. For any offence punishable with 7+ years' imprisonment, an FSL team must visit the scene. This means the FSL is part of the chain of custody from the scene itself, not from the moment the IO transports the packet to the lab.
• BNSS 2023, Section 105: mandatory videography of search and seizure. The video runs in parallel to the seizure memo, and any divergence between the two weakens both. Defence counsel routinely demand the video at trial.
• Indian Evidence Act, Section 65B (now mirrored under the Bharatiya Sakshya Adhiniyam 2023): the certificate regime for electronic records. Anvar P.V. v Basheer (2014) and Arjun Panditrao Khotkar v Kailash Kushanrao Gorantyal (2020) settled that a 65B certificate is mandatory for electronic records, and the absence of the certificate makes the record inadmissible even where its authenticity is otherwise unchallenged.

The Bharatiya Nyaya Sanhita 2023 (BNS) replaced the IPC but didn't materially change how physical evidence is treated; the offence-definition changes don't disturb the evidentiary architecture. The changes you'll be tested on sit in BNSS (procedure) and the Sakshya Adhiniyam (evidence), not in BNS itself.