Assessment of the ProblemThis section of the policy analysis framework will assess the theoretical and academic principles behind profiling, highlighting the cross-fertilization it has received from within the field of psychology (Link to a general history of personality profiling in psychology). It will then argue that before profiling can be simply discarded due to a lack of empirical validity or utility, researchers (and in particular, forensic psychologists) need to methodologically study its principles, and either refute or build upon them. Just because a field is within its early stages of development does not necessarily mean that it is not useful. Similar arguments were lanched against psychologists and psychiatrists when they too were fledgling fields without any empirical weight behind them.
The psychological profile as an investigative tool helps to focus investigators on individuals with similar personality traits that parallel the traits of others who have committed very similar offenses in the past (Pinnizzotto, 1984). As an investigative strategy, a criminal profile allows investigators to narrow the fields of options and generate educated guesses about the perpatrator (Douglas, Burgess, Burgess, and Ressler; Rossi, 1982).
The profile is generated based upon what can really be termed a psychological assessment of the crime scene (Ault and Reese, 1980). The profiler identifies and interprets evidence at the crime scene which might be indicative of the personality type of the individual(s) committing the crime. Depending upon how much evidence is available at the crime scene, Ault and Reese (1980) note that the elements a profile may include are:
These items make up the social and psychological core variables of the offender's personality, and will aid the investigation by allowing the police to narrow their focus to a more manageable framework (Holmes and Holmes, 1996). As items 8 and 9 suggest, a carefully prepared profile may also be able to help connect past unsolved crimes to the current offender (an aid to linkage blindness), and provide an indication as to whether or not future attacks are likely. The profile can also suggest possible items to add to a search warrant such as souvenirs, pornography, photos, etc. Finally, the profile should include possible interrogation strategies designed to elicit information from the suspect.
The ability of profilers to retrospectively produce the above psychosocial information seems ridiculous to many psychologists, who find crime scene "reconstruction" beyond the capacity of behavioral science research and theory alone. Others, such as Godwin (1978) make the point that profilers:
"Play a blindman's bluff, groping in all directions in the hope of touching a sleeve. Occasionally they do, but not firmly enough to seize it, for the behavior producing them must necessarily deal in generalities and types. But policemen can't arrest a type. They require hard data:names, faces, fingerprints, locations, times, dates (emphasis added). None of which psychiatrists can offer" (Godwin,pg.276).
Two widely published academics on the topics of serial homicide and mass murder, Levin and Fox (1985), also have unkind words to say about the use of criminal profiling in violent crime investigation. In their book, Mass Murder: America's Growing Menace, they state that psychological profiles are essentially vague and general, and thus virtually useless in identifying a killer.
The "general and non-specific descriptors" inherent in profiling are also viewed as problematic by Campbell (1976) who believes that profiles will cause investigators to focus all of their attention on searching for a suspect and corroborating evidence to match the profile. If the profile is wrong, the investigation can thus easily be led astray.
Finally, Liebert (1995), a clinical psychologist, labels the profiling process as "superficial, phenomenological, and perhaps, even worse, distracting" (Liebert, pg.198). Instead of rigidly reducing a crime scene such as serial murder down into a few observable parameters, he believes that each case should be distinguished by its uniqueness. From this, a profile can be developed "against the psychopathology of (severe forms of) narcissistic and borderline personality disorders" (Liebert, pg.199).
He argues that the mutilation and disembowelment of victims common in the lust murderer shares similar paraphilic roots to the Borderline Personality Disorder, or latent psychotic who is defending himself from psychological disintegration (Liebert, 1985). These primitive motivations arise out of a fusion of the destructiv and sexual impulses of the preoedipal matrix, with the incapacity for empathic bonding of the psychopath.
The psychiatrist can help the investigator see that lust murder represents an extreme of the sadomasochistic end of the Borderlinbe-Narcissistic spectrum (Ibid). In this manner the psychiatrist can give investigators a sense of the "type" of suspect they are dealing with. How Liebert's "type", mixed in with a lot of psychological jargon, will be of more practical use to investigators than a criminal profile is not clearly articulated in the article.
Additionally, the concentration of prophilers on the fetishistic and paraphilic sexual behaviors is further invalidated in Liebert's view because, he argues, the vast majority of disordered individuals manifesting sadomasochistic and perverse behaviors are unlikely to commit serial murder. What Liebert fails to recognize is that the classifications upon which profiling theory has developed are not based on all individuals who manifest the same DSM IV diagnosisas the offender. For in that he is quite correct in noting that there is a wide continuum of behaviors and severity even within one single DSM-directed diagnosis. The paraphilia of Sexual Sadism (302.84), for example can range from sadistic sexual urges with a consenting partner to torture, mutiliation, and even killing.
Recognizing the constraints of traditional methods of offender classification in providing leads to investigators, an interdisciplinary team from the FBI's Behavioral Science Unit (BSU) set out to empirically examine the correlation between offender characteristics and offense charcteristics in one of the first efforts to advance profiling theories and methods ever conducted (Ressler, et al, 1980; Ressler, Burgess, Douglas, 1988). Based upon these, and later findings highlighting the correlation between certain crime scene indicators and offender characteristics, profilers develop a classification of the likely offender, rather than classifying according to psychological symptoms which may be the basis for the official diagnosis the offenders might fall under.
In this sense, profiling is really a form of retroclassification in that it works backwards (Douglas et al,1992). Unlike in the traditional clinical evaluation, profilers in a homicide investigation have neither the offender nor the victim to rely on for sources of information. They thus have to rely solely on crime scene-related data to make their classification (Ibid.).
The Diagnostic and Statistical Manual of Mental Disorders (DSM IV) (APA, 1995) is an attempt by mental health professionals to organize behavior into neat categories, clustered by symptoms. Continued revisions and analysis since the first edition was put out in 1952, has led to a multiaxial system containing explicit diagnostic criteria and descriptions that is now very reliable, especially amongst those who are well trained in its use.
However, despite its increased statistical reliability and facility of use, at its core the DSM IV is really just a convenience by which psychologists and psychiatrists are better able to communicate. Behaviors do not always fit neatly into fixed sets of categories. This is one reason why the DSM IV also allows for a single individual to have several different diagnoses across the multiaxial system.
What is ultimately the most important to the psychologist conducting the evaluation are the various symptoms of his or her client, or in other words, the particular behaviors and responses that are of interest to the professional. Similarly, the profiler sees the crime scene, particularly a bizarre one, as a reflection of the behavior or symptoms of the perpatrator.
Holmes and Holmes (1996) claim that because the violence inherent in the types of crimes profiled is so difficult to comprehend, and that because such offenders do not share in the common values of society, they are thought to reflecft a pathological personality condition. Just as clinicians expect the pathologies of variously classified disorders to be apparent in the day-to-day lives and homes of the individual, so too here the profiler expects the crime scene to reflect the symptoms of such a pathology.
Given these central assumptions, it is clear that profiling is less suited for those cases where the "normal" motives for killing are not present (Holmes and Holmes, 1996; Douglas and Burgess, 1986; Palmiotto, 1994; Sears, 1991). Both research and practical experience indicates that there must be an underlying psychopathology to the crime in order for profiling to offer any "value added" to the investigation (Gerberth, 1981). Examples of such offenses are: sadistic torture in sexual assaults, evisceration, postmotem slashing and cutting, motiveless fire setting, lust and mutilation murder, rape, satanic and ritualistic crimes, and pedophilia (Gerberth, 1993). However, future empirical study may extend some of profiling's utility to other crimes as well. For example, the choice of motive or expression may be indicative of the offender's personality even in crimes with less overt psychopathology. This point will be elaborated upon in the evaluation section of the policy analysis framework.
Using the criminal investigative analysis techniques of profiling as a base, a Crime Classification Manual (Douglas,et al., 1992) has been developed that serves as the investigator's own version of the DSM. This classification system, using the DSM as a guide and drawing upon the expertise of law enforcement personnel, provides the first attempt to operationalize these concepts in a format that is both comprehensive and practical. Like the DSM itself once was, the authors note that "although (the system) is in a rudimentary stage of model development, (it has) progressed to the stage of advancing a testable system" (Ibid, pg.22). Empirical validity and reliability will be established as social scientists heed the call for much needed research and development in this area.
The main underlying assumptions of profiling rely heavily on elements of trait theory discussed pervasively in the psychological literature. It is argued that "the 'person variable' repeatedly shows a greater consistency and weight when we move from normal to pathological populations"(Pinnizzotto and Finkel, pg.217). This is largely due to the consistent findings correlating fantasy to repetitive, serial offending.[see Simon or Anderson, 1994].
A distinction is made in profiling between the traditional methods of operation (MO) discussed in the investigation of any crime, and the signature left behind at the crime scene of a pathological offender. The MO involves any action or behavior which is necessary for the offender to commit the crime (Gerberth, 1995). While this can be an important element in linking separate offenses committed by the same individual, sole reliance on this can be dangerous in that with many violent crimes, especially those involving sadism (as noted in the DSM IV), the MO can evolve over time.
Douglas and Munn (1992) liken the signature to the "calling card" of the offender. Like the MO, this too is a learned behavior, but it goes beyond simply those actions necessary to commit the offense (MO) in that "it composes a unique and integral part of the offender's behavior while he is committing the offense" (Douglas and Munn, pg.261). The signature is what the violent, repetitive offender's behavior requires in order to express his violent fantasies.
For instance, interviews with a surviving rape victim can tell the investigator the types of verbal activity the rapist demanded of the victim (Hazelwood, 1983). A rapist who requires the victim to say such things as "I love you" suggests a need for ego-building, whereas the offender who demands that the victim pleads or screams tells us we are dealing with a sexual sadist (Hazelwood, 1983). Such elements are obviously not necessary to simply fulfill the act of rape, but rather provide the profiler with a picture of the "type" of personality the offender is most likely to be.
With repeatedly violent offenders, the paraphilic fantasy element involved in the offenses becomes a driving compulsion, and thus we can also assume the offender will likely repeat these crimes in a very similar fasion, i.e. according to the ritual necessary to satisfy his needs (Prentky, Burgess, Rokous, Lee, Hartman, Ressler, and Douglas, 1989). Such a need for repetition, which a profiler can help the investigator to recognize in crime scene characteristics, is one of the central diagnostic features of the paraphilias in the DSM IV (APA, 1994). Gerberth (1995) notes that the signature can evolve over time, but such change usually "involves a progression of violence and sexual mutilation which is consistent with the paraphilia Sexual Sadism most often seen in lust murders" (Gerberth, 195, pg.2). For example, Ted Bundy began as a neighborhood voyeur, and then his fantasies gradually evolved to requiring the torture and ultimate death of his victims.
The above discussion of MO and signature highlights the close match between the central assumptions of profiling theory and the progression and consistencies assumed to be characteristic of the various disorders documented within the DSM IV. Both the escalation and repetition which are central concepts in our understanding of profiling, also match the patterns and trends illustrated in the DSM, particularly the paraphilia Sexual Sadism.
Early work on serial sexual homicide by the FBI, involving the systematic structured interviewing of convicted offenders (Ressler, Burgess, Douglas, 1988) developed the organized and disorganized categories which provide one of the guiding frameworks for criminal investigative analysis or profiling. Although originally hypothesized to group empirical findings in studies of sexual homicide, its utility has since been extended to other crimes as well, such as rape, sexual assault, mutilation, necrophilia, and piquerism (Holmes and Holmes, 1996).
Holmes and Holmes (1996) discuss how in their book Sexual Homicide, Ressler, Burgess, and Douglas (1988) delete the labels asocial which was originally attached to disorganized offenders, and nonsocial which was attached to organized offenders. This author agrees with Holmes and Holmes (1996) that this was an unfortunate mistake because the words offer very valuable information that helps clarify the distinctions between the two categories, as we will discuss below. Therefore, throughout our discussion, these secondary labels will remain.
The previous link outlines the personal characteristics, postoffense behavior, suggested interview techniques, and common crime scene characteristics for the disorganized asocial and organized nonsocial offenders. The disorganized asocial offender is characteristically a socially inept individual, with below average intelligence. The FBI's research (Ressler et al, 1988) also indicates that this personality tends to be a non-athletic white male. There is a lack of planning in the crimes that is manifested in a crime scene which leaves behind a lot of physical evidence. The fact that he usually lives or works near the scene of the crime also demonstrates a lack of foresight into how to avoid detection.
The main diagnostic features of the Borderline Personality Disorder (301.83) seem to correspond with the disorganized asocial classification. Borderline personalities demonstrate "a pervasive pattern of instability of interpersonal relationships, self-image, and affects, and marked impulsivity that begins in early adulthood and is present in a variety of contexts" (APA, pg.650). The chaotic nature of the crime scene suggests loosened associations and disturbed cognitive processes at the time of the offense. Its spontaneous nature could be symptomatic of the borderline's regression into a Axis I psychotic state. That the victim is usually not raped or mutilated after death highlights the intense feelings of inadequacy of the disorganized asocial offender.
In contrast, the organized nonsocial offender, also most often a white male, is socially adept, charming, and able to manipulate his victims. The high degree of planning involved in the offenses, and the lack of evidence left at the crime scene indicates to the profiler that we are not dealing with a psychotic asocial offender type. The crime scene reveals little hard evidence, and the body is often dumped at a different location than where the killing actually took place.
Although the organized nonsocial offender often appears socially adept and charming, he is really only capable of manipulating others to believe he is genuine, when in reality he cares about nothing but his own immediate gratification. The scariest component of the organized offender's profile is that he is in enough control of his actions to take great pains to avoid detection. Successful apprehension of this personality type is very difficult, and often requires a slip up on the part of the offender (James, 1991; Keppel, 1995).
The patterns of the organized nonsocial fit clearly into the Axis II diagnostic criterion for the Antisocial Personality Disorder (301.7). Often these offenders have exhibited precursors to the later sexual homicides in their early childhood behaviors which, if caught early enough, could have been diagnosed as Conduct Disorder (312.8). Childhood precursors common amongst the sexual murderers included cruelty towards animals and violent acting out behaviors (Ressler et al, 1988).
The torture during the commission of the crime, use of restraints etc. reflects the extreme end of the paraphilia Sexual Sadism (see also American Journal of Psychotherapy, January, 1990). Although this too is an Axis I diagnosis, it lacks the psychotic elements found in the disorganized/borderline personality offender. With the disorganized offender the sexual act is usually after death, wheras with the organized offender the rape often occurs prior to and only very rarely postmortem. Humiliation and personalization are key elements of the organized killer's act, supporting a diagnosis of Sexual Sadism.
Yarvis (1995) also found that rapist/murderers displayed an extraordinarily high prevalence of sexual sadism. Psychotic psychopathology in his sample was most prevalent among murderers than rapists and rapist/murderers, whose crimes require more functional capacity, as is illustrated in the case of the organized offender's category (Ibid.,pg.418). Arguments can thus be made for a useful overlap between the psychologist's and profiler's classication systems.
Clinical risk assessment by psychologists is used extensively throughout every area of the criminal justice system, touching on civil commitment proceedings, treatment needs, predictions of recidivism, probation and parole decisions, insanity trials, death penalty sentencing, and personnel evaluations. The risk assessment process usually involves balancing those predictor variables associated with increased or decreased risk, looking at the base rates for violent behavior of similar offenders, and establishing the time frame and situational factors to which the risk assessment can safely apply. After an extensive interview process and examination of the totality of the circumstances and available data, the psychologist reaches his or her recommendations.
Although the data available is somewhat different, profiling resembles the process of clinical reasoning used by psychologists conducting risk assessments. "Both the profiler and the clinician assimilate available data, reconstruct the sequence of events, hypothesize the causal processes, assign the morbidity to typological categories, and from these derive expectations about associated features, complications, and future course" (Dietz, pg.218).
Douglas, Ressler, Burgess, and Hartman (1986) provide the most elaborate detail of the profiling process in their article "Criminal Profiling From Crime Scene Analysis". The basic operatingpremise of the decision model illustrated below, is that the way a person thinks (such as his cognitions and fantasies) will direct behavior, and thus be reflected in the crime scene.
Figure 1 consists of six stages, of which the last is the apprehension of the offender. The first stage takes a complete synopsis of all of the crime scene environment, victim, and forensic pathology characteristics. This information is then placed into the context of decision process models which organize and classify this material into underlying patterns and trends. THe crime assessment stage subsequently uses these models to reconstruct the sequence of events, as well as the behavior of both offender and victim. Here such things as the organized/disorganized typology discussed above are brought into the model. The fourth stage is the actual generation of the profile in which the "type of person who committed the crime and that individual's behavioral organization with relation to the crime" (Douglas, et al, 1986) is outlined.
As with risk assessments (Monahan, 1994), the use of correlations and established crime base rates is a key component of profiling. Pinnizzotto and Finkel (1990) demonstrate that a specific detail of the crime, such as the race of the victim, allows the profiler to immediately skip the motivational aspect of the crime, and suggest the same race for the offender based upon what is known about most crimes (Pinnizzotto and Finkel, pg.229).
Psychology's Trait Theory and the Modus Operandi(MO)/Signature Distinction
Profiling's Organized Vs. Disorganized Classifications: An Axis I and Axis II Dichotomy
Clinical Risk Assessment and the Profiling Decision Process
Go To:
Emergence of the Problem
Decision-making
Implementation
Evaluation
References