Forensic Science: Definition, Principles, History and Ethics

The working definition most commonly cited in Indian forensic textbooks is some variant of Saferstein: forensic science is the application of science to those criminal and civil laws that are enforced by police agencies in a criminal justice system. The Indian statute equivalent is found in Sections 39 to 49 of the Bharatiya Sakshya Adhiniyam 2023, which set out when expert opinion is admissible and what weight it carries.

The scope is wider than most beginners assume. Forensic biology covers DNA, serology and entomology. Forensic chemistry covers drugs, explosives, fire debris and toxicology. Forensic physics covers ballistics, tool marks and glass fractography. Forensic documents covers handwriting, ink dating and counterfeit currency. Cyber forensics, audio-video authentication, forensic psychology, forensic anthropology and forensic odontology fill out the rest. A CFSL like the one at Hyderabad runs all of these under one roof; a District RFSL might run only two or three.

Three arguments consistently appear in NHRC and MHA position papers on forensic investment. First, scientific evidence is harder to fabricate than witness testimony, raising the quality floor on convictions. Second, front-loading forensic work compresses trial timelines, relevant in a system where the average case runs five-plus years. Third, forensic records such as photographs, spectra, and DNA profiles are replayable on appeal in a way that human memory is not. The flip side, the part the NAS 2009 report stressed, is that bad forensic science is worse than no forensic science because juries and judges over-trust it. The procedural side of how this evidence enters trial is in BNSS 2023: Investigation, FIR, Search and Seizure; the admissibility rules are in Bharatiya Sakshya Adhiniyam: Forensic Evidence in Court.

Every textbook lists slightly different principles, but the six below are the consensus set across Indian forensic curricula. Learn them as a hierarchy, not a flat list.

Locard is the load-bearing one. Edmond Locard, a French criminologist working out of Lyon between 1910 and 1928, ran the first police laboratory and articulated the rule that became the foundation of every trace-evidence case ever brought. The full original formulation goes longer than the famous one-liner, but the workable summary is: every contact leaves a trace, and every trace, in principle, can be detected.

The difference between individuality and uniqueness is worth holding clearly. Individuality is a working assumption used to license identification, not a proven mathematical certainty. Saferstein is explicit about this, and the NAS 2009 report hammered the point. We act as if two fingerprints from different people are never identical because the probability of a match is vanishingly small, but the strict mathematical claim is weaker than the rhetorical one.

The international history of forensic science is six names worth knowing. The year, the country and the one-line contribution.

1. Alphonse Bertillon · France · 1879
   Invented anthropometry: eleven body measurements (head length, head width, ear length, forearm, foot, middle finger and so on) used to identify repeat offenders. The 1903 Will West case at Leavenworth Penitentiary, where two men with near-identical Bertillon measurements were distinguished only by fingerprints, killed the system.
2. Francis Galton · United Kingdom · 1892
   Published 'Finger Prints', which proved fingerprints are permanent, unique and classifiable. His three classes (loops, whorls, arches) and the minutiae (Galton points) still anchor every fingerprint syllabus.
3. Juan Vucetich · Argentina · 1892
   Made the first conviction by fingerprint evidence in human history (the Francisca Rojas case, Necochea). Built the first national fingerprint classification system, which is still the dominant one in Latin America.
4. Edward Henry · United Kingdom and India · 1897
   As Inspector General of Police in Bengal, worked with Indian sub-inspectors Azizul Haque and Hem Chandra Bose to design the Henry Classification System. Adopted by Scotland Yard in 1901 and exported worldwide.
5. Edmond Locard · France · 1910
   Founded the first police laboratory at Lyon in 1910 and articulated the exchange principle. Trained an entire generation of European criminalists.
6. Albert S. Osborn · United States · 1910
   Published 'Questioned Documents', the foundational textbook on forensic handwriting and document examination. Still cited in Indian QD cases.

A detail worth knowing: the Henry system was not really Henry's. The mathematical work was done by Azizul Haque and Hem Chandra Bose, two Bengali sub-inspectors at the Calcutta Anthropometric Bureau. Haque received a Khan Bahadur title and a cash award in 1924 after a long campaign. Bose got the King's Police Medal. The system carries the British administrator's name; the underlying classification was Indian work.

India's forensic history is unusually rich because the country was the global testbed for fingerprinting between 1858 and 1900. Six dates anchor that timeline.

The world's first fingerprint bureau was founded in Calcutta in 1897, four years before Scotland Yard established its own bureau in 1901.

For the wider institutional map of CFSL, SFSL and RFSL labs, read the next topic in this module: Indian Forensic Laboratories: CFSL, SFSL, RFSL. For the international and BNSS-era institutions, see Crime Scene Management: National and International.

The funding case for forensic science in India rests on three planks, all of which show up in NHRC and MHA position papers.

• Conviction-rate quality, not just quantity. India's IPC-era conviction rate hovered around 40 percent. Scientific evidence raises both the conviction rate and, more importantly, the rate of correct convictions. The Nirbhaya 2012 case and the Talwar 2008 case are the two routinely cited as positive and negative reference points.
• Trial-time compression. BNSS Section 176(3) makes FSL visits compulsory for offences punishable with seven-plus years, which front-loads the scientific work and shortens the trial.
• Objective record. A photograph, an FTIR spectrum, a DNA STR profile and a ballistic comparison are all replayable on appeal. Witness memory is not. This is the single biggest argument the Justice Malimath Committee 2003 made for forensic expansion.

The criticism, again from NHRC and a series of Supreme Court observations across the 2010s, runs in two directions. First, capacity. Even after BNSS, the average state SFSL has a 12 to 24 month backlog on DNA cases. Forensic visits to scenes are aspirational outside the metros. Second, calibre. The 2009 NAS report's critique applies to India with extra force: training is inconsistent, accreditation patchy, and the science behind some traditional disciplines (hair comparison, bite-mark analysis, certain types of handwriting opinion) is empirically thin. This is the Indian version of the Daubert problem, and we come back to it in the ethics section below.

Forensic ethics is as technically demanding as any analytical protocol. The five rules below are the working code used by ENFSI, the AAFS and increasingly the DFSS.

1. Objectivity
   The expert serves the truth, not the side that paid for the report. In India, this is doctrinally settled by Murari Lal v State of MP (1980) and reinforced by State of HP v Jai Lal (1999): the expert is an aid to the court, not an advocate.
2. Disclosure of method and limits
   Every report must state the method used, the validation behind it, and the known error rates. NAS 2009 made this a baseline requirement that Indian SOPs are slowly catching up to.
3. Conflict-of-interest declaration
   If the analyst has any prior relationship with the investigation or its outcome, this is declared up front. The Aarushi Talwar 2008 controversy turned partly on whether the AIIMS panel had been pre-briefed.
4. Chain-of-custody discipline
   Every break in custody is recorded. The expert refuses to issue an opinion on evidence with a broken chain, even when pressured by the IO. We cover this end to end in [Chain of Custody](/topics/crime-scene-management/chain-of-custody).
5. Duty to court above duty to client
   In any conflict between what the prosecution wants and what the data say, the data wins. This is the Indian equivalent of the duty-to-court principle in English expert-witness law.

The two scandals worth knowing. The FBI hair-microscopy debacle (2015) showed that of 268 cases reviewed, 257 contained erroneous testimony in favour of the prosecution. The Annie Dookhan case (Massachusetts, 2013) saw a single drug analyst falsify results on 24,000 cases, leading to the largest mass exoneration in US history. Both are now standard reading in any forensic ethics module.

The open Indian problem is what the literature calls the Indian Daubert problem. The US uses the Daubert standard (Daubert v Merrell Dow Pharmaceuticals, 1993) to gatekeep junk science out of court: testability, peer review, known error rate, general acceptance. Indian courts use a much looser standard based on the credentials of the expert rather than the validity of the method. Section 39 of the BSA 2023 did not reform this. The result is that bite-mark analysis, certain hair comparison work, and ear-print identification all continue to be admitted in Indian courts despite being criticised in the US and UK.