History of Criminal Profiling: Dr Bond to BSU to Indian Practice

The 19th century gave the discipline two threads. One is the typology thread: the attempt to define what kind of person a criminal is. The other is the case-analysis thread: the attempt to reason backward from a specific crime to a specific kind of offender. Modern profiling is the case-analysis thread; the typology thread mostly died, but it shows up in syllabi because the history matters.

Cesare Lombroso and atavism (1876)

Lombroso, an Italian physician, published L'uomo delinquente in 1876. His central claim was that criminals were evolutionary throwbacks, biologically distinguishable by features like a sloping forehead, asymmetric face, large ears, or excessive body hair. The theory is now thoroughly discredited; it conflated correlation with causation, ignored class and selection effects, and slid easily into eugenics.

Why does Lombroso still appear in most Indian forensic-science programmes? Two reasons. First, he was the first person to claim that the offender could be inferred from observable features, which is a structural ancestor of profiling even though the features are wrong. Second, the typology thread he started, suitably cleaned up, became the modern offender-typology literature you'll meet in Criminal Motivation and Behavioural Typologies.

Hans Gross and the investigator's manual (1893)

Hans Gross, an Austrian magistrate, published Handbuch fur Untersuchungsrichter (Criminal Investigation: A Practical Handbook) in 1893. The book is often called the founding text of criminalistics. Gross argued that the investigator had to think like a scientist, building inferences from physical traces, witness accounts, and crime patterns. He didn't call it profiling, but the analytical posture he described, "reason from the scene to the offender", is the case-analysis thread that the FBI BSU would pick up in the 1970s.

The Whitechapel murders of 1888 produced what most historians call the first criminal profile. Dr Thomas Bond, a police surgeon, was asked by Scotland Yard to review the autopsy findings of the five canonical Jack the Ripper victims and to offer an opinion on the offender. His report, dated 10 November 1888, is short, careful, and recognisably modern in its structure.

Bond inferred that the killer:

• Worked alone, not as part of a gang.
• Possessed a degree of anatomical knowledge, though not necessarily medical training.
• Was physically strong, cool, and daring.
• Was likely solitary, eccentric, and "subject to periodical attacks of homicidal and erotic mania".
• Probably appeared respectable and "middle-aged".

You'll notice what's not there. Bond didn't claim to name the offender or to predict where he lived. He didn't reach beyond the evidence. The report is, by modern standards, a competent and conservative behavioural inference. The Ripper case was never solved, so Bond's accuracy can't be tested, but the report's structure (behavioural inferences tied to wound pattern evidence) is the template that the FBI would formalise eighty years later.

If Bond is the structural ancestor, Dr James Brussel is the publicity ancestor. Brussel was a New York psychiatrist consulted in 1956 on the "Mad Bomber" case, in which an offender had been planting bombs across New York City for 16 years. From the bombs, the threatening letters, and the locations, Brussel produced a now-famous profile:

• Male, 40 to 50 years old.
• Eastern European, probably Slavic.
• Roman Catholic.
• Single, living with an older female relative (a mother or aunt).
• A neat, methodical worker.
• Suffering from paranoia, probably with a grievance against Consolidated Edison.
• When arrested, would be wearing a double-breasted suit, buttoned.

George Metesky was arrested in January 1957. He was male, 53, Slavic, Roman Catholic, single, living with his two older sisters, a former Con Ed employee with a workplace grievance. He came to the door in pyjamas, changed at the officers' request, and yes, put on a double-breasted suit and buttoned it.

The double-breasted-suit detail entered profiling folklore. Brussel was lionised, profile-as-prediction looked like real science, and the FBI started paying attention. Later historians have pointed out that Brussel's profile included a number of misses that didn't make it into the popular retelling (he was wrong about Metesky's age, marital living arrangement details, and several other points), and that the public retelling cherry-picked the hits. The methodological lesson, which the BSU would re-learn the hard way, is that confirmation bias loves a profile.

The FBI's Behavioural Science Unit was set up at Quantico in 1972. The early architects were Howard Teten and Pat Mullany, instructors who started teaching what they called "applied criminology" to FBI new-agent classes. Teten had been quietly consulting on cases since the late 1960s, sometimes with Brussel's input. The unit's profile rose through the 1970s as agents took the training back to field offices and started using it on active cases.

The 1979 to 1983 Criminal Personality Research Project, run by Robert Ressler, John Douglas and Ann Burgess, was the empirical work the BSU is most known for. Ressler and Douglas interviewed 36 incarcerated sexual murderers, including Edmund Kemper, David Berkowitz, Ted Bundy, Charles Manson and others. The interviews produced the Crime Classification Manual (covered in the next topic, Behavioural Evidence Analysis) and the organised vs disorganised dichotomy.

The 36-offender study was groundbreaking but also methodologically thin by today's standards. It was a convenience sample (only offenders who agreed to be interviewed), it relied on offender self-report (notoriously unreliable), and the organised/disorganised split was abstracted from a small number of cases. The BSU has spent the decades since refining and, in places, walking back the original framework.

The 2000s brought a sustained methodological challenge to profiling. The most cited paper is Brent Snook and colleagues' 2008 review, The Criminal Profiling Illusion, which argued that there was no convincing empirical evidence that profilers outperformed untrained controls at inferring offender characteristics, and that the discipline had survived on plausibility rather than predictive validity. Other reviews followed: Alison and colleagues on the Barnum effect in profiles (vague statements that feel specific), the FBI's own internal reviews after the Brent Turvey criticisms, and the steady rebranding from "profiling" to "criminal investigative analysis" inside the Bureau itself.

Where this leaves the field in 2026:

• Profiling exists as a useful narrowing tool, not as an identifier. A profile can help prioritise suspects and shape interview strategy. It cannot, on its own, name an offender.
• The organised/disorganised dichotomy is still taught, still examined, but increasingly treated as a starting point rather than a finding. Most real cases sit on a continuum or shift between modes within a single offence.
• The strongest contemporary profiling work is integrated rather than stand-alone. It combines behavioural inference with geographical profiling (see Geographical Profiling), digital forensics, and victimology, rather than treating any one input as decisive.

India has no statutory equivalent to the FBI BAU. What it has is a quieter, ad-hoc tradition that runs through three institutions and a handful of high-profile cases.

• CFSL Delhi's behavioural science consults. The Central Forensic Science Laboratory at Lodhi Road has, since the 1990s, provided behavioural-analysis inputs in serial cases and high-profile homicides, usually at the request of the CBI or a state CID. The work isn't always called profiling and isn't always public.
• NIMHANS, Bengaluru. The National Institute of Mental Health and Neurosciences runs the country's most credible forensic psychology programmes and has provided clinical-psychological inputs in cases ranging from suicide-vs-homicide differentiation (psychological autopsy, covered separately) to competency evaluations.
• NFSU, Gandhinagar. Since the National Forensic Sciences University was constituted in 2020 (succeeding the earlier Gujarat Forensic Sciences University), behavioural and investigative-analysis training has become part of the formal curriculum, including a Master's track in Behavioural Forensics. DCP-rank officers from several state police forces have been trained at NFSU on profiling and interview techniques.

Three Indian cases that are usually discussed in profiling lectures, with the caveat that none of them produced a published, formal profile:

• The Tandoor Murder (1995). Sushil Sharma's killing of Naina Sahni in Delhi. Profiling commentary was retrospective and informal, but the case became a teaching example for staging-and-disposal behaviour.
• The Nithari killings (2005 to 2006). Moninder Singh Pandher and Surinder Koli, Noida. The case raised questions about whether structured profiling might have shortened the investigation; it didn't get one in real time.
• The Aarushi Talwar case (2008). Ad-hoc profiling commentary circulated in the press throughout the investigation, much of it speculative. The case became a cautionary tale about media-led profiling untethered from evidence.

The honest summary, for a 2026 NFSU candidate: Indian profiling is real but quiet, mostly clinical-psychology-led, mostly consult-driven, and not yet institutionalised. Expect the next decade to formalise it, starting with the NFSU pipeline.