Crime, Criminology and the Classification of Offences

The legal definition of crime follows Salmond: a crime is a wrong whose remedy is a sanction imposed by the state, in contrast to a private remedy sought by an individual. The sociological definition, associated with Sutherland, treats crime as conduct that violates a criminal law and is punishable by the state. Both formulations share a common requirement: the state must have named the conduct as punishable. That is the line separating a crime from a sin or a moral failing.

The split between offence and civil wrong is the classification every working textbook leads with. A defamation suit can be filed as a civil action for damages and also prosecuted under BNS Sections 356 and 357 as a criminal offence. The same conduct, two parallel tracks. Forensic students see this in road-traffic deaths: the family sues the driver under the Motor Vehicles Act for compensation, the State prosecutes him under BNS Section 106 for causing death by negligence, and the post-mortem report goes into both files.

This is also where the forensic professional first meets the section number. A scene-of-crime officer arriving at a homicide must already be thinking about which BNS section the investigator will invoke, because that choice determines what evidence the FSL will be asked to examine. We work through the section list in detail in our companion piece on the Bharatiya Nyaya Sanhita for forensic students. For now, treat the legal definition as the gate and the sociological one as the map.

Indian criminal law inherits from English common law the rule that every crime requires two elements: a guilty act and a guilty mind. The Latin maxim is actus non facit reum nisi mens sit rea: the act does not make a person guilty unless the mind is also guilty. BNS 2023 preserves this structure. The general explanations in Chapter II define intention, knowledge, voluntarily, dishonestly, fraudulently, and reason to believe, each representing a distinct flavour of mens rea.

The actus reus has three layers worth separating in answers:

• Conduct. The physical act itself, whether positive (firing a weapon) or omission (a parent failing to feed a child).
• Consequence. Where the section requires a result, such as death in BNS Section 103 murder, the consequence is part of the actus reus.
• Circumstances. The surrounding facts the law cares about, such as the victim being under eighteen in BNS Section 65 rape provisions.

1. Identify the section
   Read the section charged; BNS Section 103 (murder), Section 117 (grievous hurt), Section 303 (theft) each demand different conduct and different mental states.
2. Map the conduct
   What did the accused physically do or fail to do? A voluntary act is required; reflex movements and unconscious acts are excluded under the general defences in Chapter III.
3. Locate the result
   If the section requires a consequence, prove causation. The 'but-for' test plus the 'substantial cause' test from R v. Smith are routinely cited by Indian courts.
4. Establish the mental state
   Show intention, knowledge, recklessness, or the negligence standard the section uses. The FSL's blood-spatter or trajectory report often does this circumstantially.
5. Check defences
   Insanity (BNS Section 22), minority (Sections 20-21), private defence (Sections 34-44), and accident (Section 23) can negate either element.

Bloodstain pattern analysis, autopsy findings on wound number and depth, and the seized weapon allow the prosecution to argue intention rather than mere knowledge. A single stab to the chest with a screwdriver is consistent with intention to cause that specific injury; thirty stabs to the same region supports an inference of intention to cause death. The FSL records the physical findings; the court reads the mental state into them.

There is no single official taxonomy of crime. The eight-category scheme below is the one used in the NCRB's Crime in India statistical volume and in most Indian criminology textbooks. Each category maps to one or more BNS chapters and to a different evidence profile at the scene.

A few clarifications worth holding precisely:

• Dacoity vs robbery. Robbery (BNS Section 309) becomes dacoity (Section 310) the moment five or more persons commit it jointly. Not four. Five.
• Theft vs extortion. Theft is taking without consent. Extortion is taking through induced consent obtained by putting the victim in fear of injury.
• Cyber crime is not a separate silo. A phishing scam that drains a bank account is cyber in method but property in nature, and is now chargeable under BNS Section 318 read with IT Act Section 66D.

Edwin Sutherland coined the term "white-collar crime" in his 1939 American Sociological Society address to describe offences committed by persons of respectability and high social status in the course of their occupation. Major Indian cases, including the Harshad Mehta securities fraud (1992), the Satyam accounting scandal (2009), and the PNB-Nirav Modi fraud (2018), all fit the Sutherland profile: exploitation of occupational trust, leaving a paper trail rather than a physical scene. In these cases the forensic professional's work shifts from bloodstain pattern analysis to document examination and digital forensics.

Criminology has developed four principal answers to the question of why people offend. The classical school treats the offender as a rational actor who weighs expected gains against expected costs. The positivist school locates the cause in biological or psychological determinants. The sociological school attributes crime to social structure, anomie, or learned association. Modern integrated approaches combine these explanations with different empirical weights. These schools underlie much of the inferential work the forensic professional does at a scene.

Three sub-theories within the sociological school recur most frequently in Indian criminology curricula:

• Durkheim's anomie. Anomie is a state of normlessness during rapid social change. Crime rises when shared moral regulation weakens, as Durkheim argued for industrialising France in De la Division du Travail Social (1893).
• Merton's strain theory (1938). Society sets culturally approved goals (wealth, status) but unevenly distributes legitimate means. Individuals adapt through conformity, innovation (crime), ritualism, retreatism, or rebellion. The Mumbai chit-fund frauds of the 2010s are textbook innovation.
• Sutherland's differential association (1939). Criminal behaviour is learned through intimate personal groups, the same way any other behaviour is learned. Nine propositions; the most quoted is the sixth, that a person becomes delinquent because of an excess of definitions favourable to violation of law over definitions unfavourable.

Travis Hirschi's social bond theory (1969) inverts the standard question: rather than asking why people offend, it asks why most people do not. Hirschi's answer rests on four bonds: attachment to others, commitment to conventional goals, involvement in legitimate activities, and belief in shared norms. When these bonds weaken, offending becomes more likely. Indian juvenile justice boards draw on this framework when assessing children in conflict with law under the JJ Act 2015.

Lombroso's atavism, the claim that criminals are evolutionary throwbacks identifiable by physical stigmata, has not survived empirical scrutiny. Skull measurements, ear shape, and jaw projection do not predict offending. More recent biological research, including work on the MAOA-low-activity allele and antisocial behaviour and on the XYY karyotype and impulsive aggression, is methodologically more rigorous. The current scientific consensus is that biological variables raise risk modestly in interaction with environmental factors; they do not determine criminal outcomes.

Criminological knowledge has three direct operational payoffs for the forensic professional:

• Scene interpretation. A burglary scene with stacked, sorted goods and no vandalism reads differently from one with smashed property and faeces on the floor. The first suggests instrumental, rational offending (classical school's expected-utility actor); the second suggests expressive offending tied to personality variables (Eysenck) or to a subcultural display. The kit you call for differs.
• Offender profiling. Behavioural evidence analysis, covered in our criminal motivation and typologies topic, leans on the psychological and sociological schools to convert physical findings into inferences about the offender. A sexual-homicide scene with extensive post-mortem mutilation is read against psychiatric typologies; a gang-related shooting is read against organised-crime literature.
• Report framing. When the FSL report references intent, knowledge, or recklessness, it is implicitly invoking the legal-doctrinal frame from Section 2 of this topic. A report that says 'the trajectory is consistent with a single discharge from approximately 1.5 metres' is feeding directly into the prosecution's argument on mens rea.

Consider a practical sequence. A body is found in a one-room kholi in Dharavi: ligature around the neck, no signs of forced entry, household goods undisturbed, the victim's phone unlocked on a WhatsApp thread with the husband. The legal frame determines section: BNS 103 (murder) if intention is established, BNS 101 (culpable homicide) if only knowledge is shown. The criminological frame characterises the offending: domestic, instrumental, committed by someone with access. Together, both frames direct the scene-of-crime officer to prioritise the phone for digital forensics, the ligature for fibre comparison, the husband's clothing for fibre and skin cells, and the neighbours for time-of-last-contact statements. The theory directed the search; the physical evidence answers it.

For the chain that turns these scene decisions into court-ready exhibits, walk through our chain of custody topic next.

The classification below consolidates the key distinctions for quick reference.

Two distinctions to lock in before any working summary is complete:

• Cognisable vs non-cognisable. A cognisable offence is one for which a police officer may arrest without warrant; non-cognisable requires a magistrate's permission. The BNSS 2023 First Schedule lists which is which.
• Compoundable vs non-compoundable. Compoundable offences (BNSS Section 359) can be settled between the parties with court permission; non-compoundable offences (murder, rape, terrorism) cannot.

For a deeper map of how individual BNS sections sit inside this scheme, read the companion piece on the Bharatiya Nyaya Sanhita for forensic students. For how scene evidence is actually collected against each category, see processing physical evidence at the scene.