From Whitechapel To Behavioural Evidence - What Forensic Profiling Can—And Cannot— Infer About Jack The Ripper

Author: LawSphereX

Category: Evidence & Procedure

Published: 2026-08-23T19:55:54.548Z

Updated: 2026-08-30T07:11:40.108Z

FROM WHITECHAPEL TO BEHAVIOURAL EVIDENCE:

WHAT FORENSIC PROFILING CAN—AND CANNOT— INFER ABOUT JACK THE RIPPER Ishaan D. Joshi, Consulting Forensic and Criminal Intelligence Expert, Visiting Faculty, MIT-WPU School of Law

ABSTRACT

The Jack the Ripper case remains a foundational reference point in criminal psychology because it lies at the intersection of behavioural evidence and cultural myth. This paper examines what forensic profiling can responsibly infer from the surviving late-Victorian record, and where such inference becomes speculative. Rather than attempting to identify the offender, the study treats the case as a methodological stress-test for contemporary investigative psychology: incomplete documentation, inconsistent witness and inquest accounts, uncertain timelines, and pervasive media contamination mirror problems that continue to distort modern investigations.

Using a conservative evidence-to-inference approach, the paper synthesises three evidential domains—crime-scene behaviour (including the distinction between modus operandi and signature), victimology and opportunity structure, and profiling under evidential uncertainty. Multiple plausible motivational formulations are comparatively framed (power/control, anger/revenge, sexualised aggression, and fantasydriven repetition), while situational and instrumental explanations are treated as live alternatives, limiting diagnostic overreach. Two competing profile variants are evaluated—a locally familiar, socially functional opportunist and a mobile, marginal situational predator—demonstrating why both remain plausible given the record’s constraints.

The paper concludes that modern forensic psychology’s principal contribution would be tighter inference controls via linkage analysis, behavioural consistency testing, and contamination safeguards. Overall, the central thesis is restrained: profiling is most defensible when it generates constrained, testable hypotheses rather than claims of identification. Keywords: Offender profiling, Behavioural evidence analysis, Victimology, Geographic profiling, Media contamination.

Introduction

The Jack the Ripper case continues to matter to criminal psychology because it sits at the boundary between behavioural evidence and myth-making: it is one of the earliest modern examples in which crime-scene actions, victim selection, and public narrative became intertwined, shaping how later generations think about “profiling” (Walkowitz, 1992; Sugden, 2002). Forensic and investigative psychology still return to such historical cases because they force a disciplined question that remains central in contemporary practice: how far can we move from traceable behaviour to defensible psychological inference without smuggling in speculation (Alison, Bennell, Mokros and Ormerod, 2013; Canter, 2004).

The Ripper file is therefore not merely a Victorian curiosity; it is a stress-test of method— highlighting problems that also occur in present-day investigations, such as incomplete records, biased witness accounts, media contamination, and the temptation to over-interpret rare or shocking behaviours (Alison et al. , 2013; Douglas, Burgess, Burgess and Ressler, 1992). This paper asks: What can—and what cannot—be inferred about offender psychology from limited, noisy historical evidence? The aim is not to “solve” the case, but to evaluate the psychological propositions that are often asserted about the offender (e. g. , motivation, control needs, interpersonal style) and to separate reasonable inference from narrative excess.

The scope is restricted to the canonical evidential domains used in offender assessment: • behavioural evidence analysis (distinguishing modus operandi from psychologically meaningful patterns), • victimology and opportunity structure (who was targeted, under what situational constraints), • crime-scene behaviour (sequencing, risk, and behavioural consistency) (Turvey, 2012; Canter, 2004). Methodologically, the paper follows a conservative “evidence-to-inference ladder”: begin with what the record can support, treat psychological labels as hypotheses rather than conclusions, and explicitly state uncertainty where the historical data does not warrant clinical-level claims (Alison et al. , 2013; Canter, 2004).

Case Snapshot and Evidential Constraints: What the Record Can (and Cannot) Support Any psychological reading of Jack the Ripper must begin with a sober case snapshot and, more importantly, with the constraints of the surviving record. The “canonical” sequence commonly treated as the best-supported cluster involves a small set of late-1888 killings in the Whitechapel/Spitalfields area of London, occurring within a narrow time window and linked primarily by geography, timing, and certain broad behavioural consistencies as reconstructed from contemporary investigative materials (Sugden, 2002; Begg, 2003).

At a high level, these offences are typically described as opportunistic street encounters culminating in rapid lethal violence, with the offender operating in public spaces that nonetheless offered momentary concealment and escape routes—conditions that matter for any inference about offender confidence, local familiarity, and risk tolerance (Sugden, 2002; Canter, 2004). Importantly, the “canonical” label is itself an evidential judgment, not a certainty: linkage is probabilistic, and any profile built on a presumed series inherits that uncertainty (Begg, 2003; Alison, Bennell, Mokros and Ormerod, 2013). The data available for analysis are uneven and mediated.

What counts as evidence includes: • witness reports and statements (often inconsistent, filtered through fear, prejudice, and retrospective reconstruction), • medical and inquest notes that describe injuries and estimated time-of-death (subject to the limits of nineteenth-century practice and documentation), • policing records and investigative memoranda that reflect both genuine leads and institutional pressures (Sugden, 2002; Begg, 2003). A further category—highly salient but methodologically hazardous—is the corpus of letters and communications attributed to “Jack the Ripper. ” Most are disputed, and many scholars treat them as likely hoaxes or at least as contaminated by public notoriety, making them weak foundations for stable psychological inference (Begg, 2003; Sugden, 2002). These constraints impose clear methodological limits.

Timelines are frequently uncertain: witnesses estimate times under poor visibility and stress; medical estimates are broad; and later retellings compress ambiguity into false precision (Sugden, 2002). The case is also unusually vulnerable to contamination—not only from sensational journalism and reward-driven tip-offs at the time, but from subsequent myth-making that retrofits motive, psychopathology, and “signature” into a narrative of inevitability (Walkowitz, 1992; Alison et al. , 2013). Accordingly, this paper treats the historical record as a set of imperfect behavioural traces: useful for generating bounded hypotheses, but insufficient for clinical diagnosis, definitive motive attribution, or confident claims about stable personality structure.

Victimology and Opportunity Structure: Selection, Access, and Situational Control Victimology is central to behavioural inference because it anchors psychological hypotheses to opportunity patterns rather than to narrative speculation. In the Jack the Ripper case, the victims commonly associated with the canonical series were encountered in a late-Victorian urban environment shaped by poverty, precarious housing, and intense street-level exposure, conditions that affect routine activity patterns and the availability of suitable targets (Walkowitz, 1992; Sugden, 2002). A routine activity perspective is useful here: offending becomes more likely when a motivated offender converges with a suitable target in the absence of capable guardianship (Cohen and Felson, 1979).

Applied cautiously, this lens suggests that victim selection may have reflected not a “type” in a psychological sense, but a situationally produced vulnerability—persons whose nightly routines increased accessibility and reduced guardianship, particularly in crowded but poorly policed micro-locations where brief isolation could occur (Cohen and Felson, 1979; Sugden, 2002). Victim selection, in this view, turns on three interacting factors. First, routine activities: latenight movement through streets, courts, and lodging-house zones increased contact opportunities and reduced the predictability of protective companions (Walkowitz, 1992).

Second, vulnerability: economic insecurity and transient accommodation can increase exposure to coercion, reduce the likelihood of immediate search, and constrain victims’ ability to avoid risky spaces—factors that offender decision-making often exploits even without explicit “targeted hatred” (Turvey, 2012). Third, accessibility: dense neighbourhoods with narrow passages and variable lighting create pockets where rapid offending is possible while still permitting quick disengagement (Canter, 2004; Sugden, 2002). “Guardianship failures” should be understood broadly: not only the limited presence of formal policing, but also the structural absence of reliable social protectors and the difficulty of sustained surveillance in labyrinthine streets (Cohen and Felson, 1979; Walkowitz, 1992). Opportunity also clarifies risk management, a key bridge between victimology and profiling.

Offender choices about time and location can reflect situational rationality rather than extraordinary daring. Research on criminal decision-making emphasises that offenders frequently balance reward against perceived risk, selecting contexts that offer control, speed, and plausible escape (Cornish and Clarke, 1986). In the canonical Ripper geography, the concentration within a relatively tight area is consistent with a “comfort zone” logic: local familiarity increases navigation efficiency, reduces the cognitive load of escape planning, and enables selection of micro-sites that offer transient concealment (Canter, 2004). Such spatial patterning does not prove residence or employment, but it supports a cautious inference of environmental competence—the ability to move without attracting attention and to exploit short windows of low guardianship (Canter, 2004; Cornish and Clarke, 1986).

Finally, victimology suggests something about the offender’s social navigation and predation strategy. If the offender repeatedly approached victims in public without immediate disruption, this implies a capacity to initiate contact without triggering rapid alarm—whether through ordinary appearance, situational plausibility, or learned interactional tactics (Douglas, Burgess, Burgess and Ressler, 1992). This is not proof of charm or “psychopathy”; rather, it indicates functional social presentation sufficient to close distance in a high-risk environment.

The strongest conclusion victimology supports, therefore, is modest but useful: the offender likely leveraged predictable vulnerabilities created by routine activities and weak guardianship, and operated within a familiar opportunity field that reduced risk while enabling quick, controlled attacks (Cohen and Felson, 1979; Canter, 2004; Sugden, 2002). Behavioural Crime-Scene Analysis: Modus Operandi, Signature, and the Limits of Inference Behavioural crime-scene analysis attempts to translate observable offence behaviours into defensible hypotheses about offender decision-making, skill, and psychological needs. In the Jack the Ripper case, this translation is complicated by the fragmentary nature of the record and by the tendency of later narratives to treat contested details as settled facts.

A rigorous approach therefore begins by separating what is plausibly reconstructable from what is merely repeated, and then applying a conservative distinction between modus operandi (MO) and signature (Douglas, Burgess, Burgess and Ressler, 1992; Turvey, 2012). MO refers to the practical behaviours that enable an offender to locate a target, complete the offence, and reduce the likelihood of detection—choices shaped by opportunity, situational constraints, and perceived risk (Douglas et al. , 1992; Turvey, 2012). MO includes selection of time and place, approach methods, the speed of offending, and post-offence movement. Critically, MO is expected to evolve: offenders learn what works, adjust to policing changes, and refine strategies to minimise exposure (Douglas et al. , 1992). In a historical series such as this, apparent consistencies (e. g.

, a relatively circumscribed geography, night-time offending, rapid completion) may reflect situational efficiency rather than deep psychology—behaviours that are “good tactics” for avoiding capture in that environment (Canter, 2004; Turvey, 2012). Likewise, any perceived changes across incidents should not be automatically read as “escalation of pathology”; they can reflect adaptive learning, fluctuating environmental pressure, or inconsistent documentation (Alison, Bennell, Mokros and Ormerod, 2013; Turvey, 2012). By contrast, signature is theorised as the set of behaviours that are not strictly necessary to complete the offence, but which express an offender’s underlying psychological needs—such as control, domination, or ritualised meaning (Douglas et al. , 1992).

Signature behaviours are often described as relatively more stable than MO because they gratify internal motives rather than practical goals (Douglas et al. , 1992; Turvey, 2012). However, stability is a tendency, not a guarantee. In the Ripper record, the principal methodological hazard is that “signature” is frequently inferred from details that may be uncertain, variably reported, or later amplified by sensational retellings. Overconfident signature claims risk turning sparse behavioural traces into a complete personality portrait, which is precisely the inferential overreach that critics of profiling warn against (Alison et al. , 2013; Canter, 2004). Within those limits, several behavioural dimensions are still analytically useful when framed as hypotheses rather than conclusions.

First, control: offence sequences that imply swift incapacitation and rapid disengagement can be interpreted as prioritising control of the situation, but they can also be explained by time pressure and fear of interruption in public spaces (Turvey, 2012). Second, escalation: where later incidents appear more audacious or complex, this might indicate growing confidence, compulsive drive, or a need for increased stimulation; equally, it may reflect changes in opportunity, reduced guardianship, or distortions in record-keeping (Canter, 2004; Alison et al. , 2013). Third, confidence and rehearsal/learning: repeated offending within a limited area can suggest environmental familiarity and practice—yet “practice” may be as banal as knowing routes, lighting patterns, and patrol rhythms, rather than indicating exceptional criminal sophistication (Canter, 2004; Cornish and Clarke, 1986).

A rational-choice perspective reminds us that many offenders behave as bounded decision-makers who seek contexts that minimise effort and maximise escape probability, without requiring exotic psychopathology to explain their choices (Cornish and Clarke, 1986). The overarching caution is therefore straightforward: many behaviours that look “psychological” may be situational. In cramped, poorly lit nineteenth-century streets with variable policing, speed and concealment are environmental imperatives. Where the evidence is contested or thin, the most defensible outputs of behavioural analysis are constrained statements—e. g.

, that the offender likely exploited low-guardianship micro-locations and operated with sufficient local competence to approach and withdraw efficiently—rather than definitive claims about diagnosis, fantasy life, or stable personality structure (Canter, 2004; Alison et al. , 2013; Turvey, 2012). Psychological Formulations: Motivation and Personality Hypotheses (Competing Models, Not a Single Story) Psychological formulation in historical cases should be treated as a disciplined exercise in hypothesis-generation rather than diagnosis. Contemporary critiques of profiling repeatedly warn against the “narrative fallacy”: the tendency to convert sparse behavioural traces into a coherent personality story that feels explanatory but exceeds the data (Alison, Bennell, Mokros and Ormerod, 2013; Canter, 2004).

In the Jack the Ripper record, uncertainty about linkage, timing, and the reliability of secondary materials further requires that psychological labels remain conditional (“consistent with”) rather than assertive (“proves”) (Sugden, 2002; Turvey, 2012). Within those constraints, it is still possible to frame multiple plausible motivational models, each anchored to different behavioural indicators. A first family of formulations centres on power and control. In this model, the offence is interpreted as a means of asserting dominance—reducing a victim to an object, controlling the scene, and exiting before interruption. Behavioural indicators typically invoked for power/control include rapid incapacitation, apparent situational management, and behavioural choices that suggest prioritising command over the interaction (Douglas, Burgess, Burgess and Ressler, 1992).

A competing model emphasises anger or revenge, where violence is an expression of hostility directed at a category of victim, an interpersonal grievance, or a broader ideological resentment. Here, the analyst looks for cues of expressive aggression—overkill relative to instrumental goals, repeated targeting patterns, and behaviours that convey contempt (Turvey, 2012). A third model is sexualised aggression, which need not require conventional sexual activity at the scene; rather, it focuses on violence as a sexualised act of domination, arousal, or fantasy enactment, inferred cautiously from patterns that are not necessary for completion of the offence but may indicate psychologically motivated acts (Douglas et al. , 1992). Importantly, these models can overlap: power motives and sexualised aggression are not mutually exclusive, and anger can be embedded within either.

The methodological requirement is to treat overlap as an empirical possibility, not as a licence to assert everything at once (Alison et al. , 2013; Canter, 2004). A second formulation concerns fantasy-driven offending and compulsivity. Some theories of serial violence propose that repeated offending is sustained by internal fantasy reinforcement and escalating need for stimulation, producing a cycle in which the offender seeks increasing psychological payoff (Douglas et al. , 1992). In behavioural terms, analysts look for repetition across offences, apparent “ritualised” elements beyond MO, and temporal clustering suggestive of mounting drive.

Yet in this case, temporal clustering and apparent repetition are also compatible with situational factors—policing patterns, opportunity availability, and historical record distortions—so “compulsivity” must remain a cautious inference rather than a conclusion (Turvey, 2012; Sugden, 2002). A third set of hypotheses addresses psychopathy/antisocial traits versus alternative explanations. It is tempting to infer psychopathy from the severity of violence alone, but clinical and forensic literature distinguishes between behavioural cruelty and the full construct of psychopathy, which involves persistent interpersonal and affective traits (e. g. , shallow affect, manipulativeness) not reliably inferable from crime-scene behaviour by itself (Hare, 2003; Cooke, Michie and Hart, 2006).

Alternative explanations include instrumental violence (violence used as a means to an end, including rapid control and escape) and situational drivers that shape apparent coldness, such as time pressure, fear of interruption, and the need for efficient offence completion in public environments (Cornish and Clarke, 1986; Turvey, 2012). A responsible formulation therefore treats psychopathy as a possibility that cannot be confirmed from the available evidence, and weighs it against simpler situational accounts (Cooke et al. , 2006; Alison et al. , 2013). Finally, analysts sometimes propose cognitive-style hypotheses—entitlement, dehumanisation, thrill seeking, or paranoid ideation—because these constructs can plausibly relate to predatory harm.

However, such inferences are only warranted when specific behavioural or communicative evidence supports them; otherwise they risk becoming descriptive labels rather than explanatory mechanisms (Canter, 2004; Turvey, 2012). The most defensible position is thus plural and conditional: multiple motivational and personality hypotheses can be articulated, but each must remain bounded by evidential limits, explicit uncertainty, and competing non-psychological explanations. Profiling Synthesis: Offender Characteristics and Competing Profiles Under Evidential Uncertainty A profiling synthesis should be read as a bounded set of probabilistic propositions, not as identification.

Modern investigative psychology emphasises that the strongest profiles derive from consistent behavioural and spatial patterns, while explicitly marking where the evidence is ambiguous, incomplete, or vulnerable to contamination (Alison, Bennell, Mokros and Ormerod, 2013; Turvey, 2012). For this case, any “offender characteristics” must therefore be framed with uncertainty markers (e. g. , likely, possible, cannot be excluded), because linkage assumptions, time-of-offence estimates, and the quality of witness and inquest documentation remain contested (Sugden, 2002; Begg, 2003). Demographics and practical inferences. A cautious demographic inference is that the offender was likely an adult male (because the offences require the physical capacity to rapidly control a victim and disengage), with an age range most plausibly spanning early adulthood to middle age (e. g.

, roughly 20s–40s), though this range remains speculative given the absence of reliable offender sightings and the potential for misdescription in stressful conditions (Begg, 2003; Sugden, 2002). Occupational or skills inference is even more fragile. Certain accounts have historically encouraged speculation about anatomical knowledge or occupational familiarity with cutting instruments; however, forensic profiling literature warns that “skill” is frequently over-attributed from injury patterns that may be produced by ordinary tools and opportunistic circumstances rather than specialist training (Turvey, 2012; Alison et al. , 2013). At most, the record supports a limited claim: the offender likely possessed functional competence with a blade and the situational confidence to act quickly. Geographic profiling logic.

Spatial patterning is one of the more defensible inferential domains because it relies on repeated location choice rather than on contested psychological narratives. Geographic profiling frameworks suggest that many offenders operate within a “comfort zone” shaped by routine activities, familiarity, and perceived escape feasibility, often near an “anchor point” such as home, work, or a frequented locale (Canter, 2004; Rossmo, 2000). Where offences cluster within a relatively tight area, this can be consistent with a locally familiar offender; yet it does not prove residence, because a commuting offender can also select a target-rich district that offers anonymity and escape routes (Rossmo, 2000).

The key point is comparative: the tighter and more behaviourally consistent the spatial pattern, the more plausible local environmental competence becomes—while still admitting alternative explanations (Canter, 2004). Social functioning: blends-in vs marginal drifter. Competing hypotheses often polarise around social presentation. One model proposes a “blends in” offender—someone capable of approaching victims without immediate alarm, dressing and behaving in a manner that attracts little attention, and moving through the neighbourhood with plausible purpose (Douglas, Burgess, Burgess and Ressler, 1992). An alternative model is the marginal or drifting offender, socially unstable, intermittently employed, and operating in the area because it is a lowguardianship environment rather than because it is a stable home base (Sugden, 2002; Begg, 2003).

These yield two profile variants that remain simultaneously plausible: Variant A: • Local, socially functional opportunist. • Likely familiar with the streets and escape routes; presents as ordinary; offences reflect rapid situational control and efficient withdrawal. • This fits geographic comfort-zone logic and the practical need to approach targets without disturbance (Canter, 2004; Rossmo, 2000).

Variant B: • Mobile, marginal, situational predator. • Less stable residence or employment; uses the district for anonymity and opportunity; may loiter without attracting sustained attention because the environment normalises transience. • This aligns with the possibility that clustering reflects target availability rather than anchoring (Sugden, 2002; Rossmo, 2000). Both remain plausible because the record does not decisively discriminate between anchored familiarity and strategic commuting, and because social presentation can be ambiguous in crowded urban settings where many individuals—local or transient—appear unremarkable (Alison et al. , 2013; Walkowitz, 1992).

The most defensible synthesis, therefore, is not a singular portrait but a constrained comparison: a locally competent offender is likely, but whether that competence reflects stable anchoring or opportunistic mobility cannot be resolved from the evidence alone. Forensic and Investigative Implications and Conclusion: Profiling as Constrained Hypothesis-Generation If the Jack the Ripper offences were investigated today, modern forensic psychology would add value primarily through structured inference controls rather than through “intuition-led” profiling. First, contemporary linkage analysis would treat case connection as an empirical question, testing behavioural similarity and distinctiveness across incidents (e. g. , whether observed patterns exceed what would be expected from opportunity and environment) (Woodhams, Hollin and Bull, 2008; Bennell, Jones, Taylor and Snook, 2014).

Second, behavioural consistency testing would explicitly separate stable elements from those likely driven by context or learning, reducing the risk of mistaking situational necessity for psychological “signature” (Bennell et al. , 2014; Alison, Bennell, Mokros and Ormerod, 2013). Third, current best practice would implement stronger media contamination safeguards—tight evidence control, careful release strategies, and analytic firewalls to prevent public narratives and hoaxes from shaping investigative hypotheses (Innes, 2004; Alison et al. , 2013). These measures matter because the historical case demonstrates how publicity can create false coherence: it encourages copycat communications, distorts witness memory, and pressures investigators toward dramatic explanations. The case’s central lesson for profiling is therefore methodological.

Profiling can be useful when it converts behavioural and spatial information into testable investigative propositions (e. g. , likely operating area, approach feasibility, risk-management style). Yet the case also demonstrates its limits: incomplete records, disputed materials, and retrospectively “cleaned” timelines make confident psychological claims—especially about diagnosis, motive certainty, or stable personality structure—scientifically and ethically unsafe (Alison et al. , 2013; Turvey, 2012). The responsible conclusion is that profiling is not a substitute for identification; it is a disciplined framework for narrowing possibilities.

Accordingly, this paper’s thesis is simple: profiling can generate constrained hypotheses, not identifications—and its legitimacy depends on transparent uncertainty, competing explanations, and strict separation between observation and interpretation (Alison et al. , 2013; Bennell et al. , 2014).

References

Alison, L. , Bennell, C. , Mokros, A. and Ormerod, D. (2013) ‘The personality paradox in offender profiling: A theoretical review of the processes involved in deriving a profile’, Psychology, Public Policy, and Law, 19(1), pp. 48–62. Begg, P. (2003) Jack the Ripper: The Definitive History. London: Pearson. Bennell, C. , Jones, N. J. , Taylor, P. J. and Snook, B. (2014) ‘Linking crimes using behavioural clues: Current levels of linking accuracy and strategies for moving forward’, Journal of Investigative Psychology and Offender Profiling, 11(1), pp. 29–56. Canter, D. (2004) Criminal Shadows: Inside the Mind of the Serial Killer. London: HarperCollins. Cohen, L. E. and Felson, M. (1979) ‘Social change and crime rate trends: A routine activity approach’, American Sociological Review, 44(4), pp. 588–608. Cooke, D. J. , Michie, C. and Hart, S. D.

(2006) ‘Facets of clinical psychopathy: Toward clearer measurement’, Personality and Individual Differences, 40(3), pp. 613–627. Cornish, D. B. and Clarke, R. V. (1986) The Reasoning Criminal: Rational Choice Perspectives on Offending. New York: Springer. Douglas, J. E. , Burgess, A. W. , Burgess, A. G. and Ressler, R. K. (1992) Crime Classification Manual: A Standard System for Investigating and Classifying Violent Crimes. New York: Lexington Books. Hare, R. D. (2003) The Hare Psychopathy Checklist–Revised (PCL-R). 2nd edn. Toronto: Multi-Health Systems. Innes, M. (2004) ‘Signal crimes and signal disorders: Notes on deviance as communicative action’, The British Journal of Sociology, 55(3), pp. 335–355. Rossmo, D. K. (2000) Geographic Profiling. Boca Raton, FL: CRC Press. Sugden, P. (2002) The Complete History of Jack the Ripper. New York: Carroll & Graf.

Turvey, B.E. (2012) Criminal Profiling: An Introduction to Behavioral Evidence Analysis. 4th edn. London: Academic Press. Walkowitz, J.R. (1992) City of Dreadful Delight: Narratives of Sexual Danger in Late- Victorian London. Chicago: University of Chicago Press. Woodhams, J., Hollin, C.R. and Bull, R. (2008) ‘Behavioural linkage analysis: A review of the literature and future directions’, Legal and Criminological Psychology, 13(1), pp. 87–105.