Skip to Main Content

Sexual Homicide: Patterns and Motives- Paperback

About The Book

Who are the men committing the rising number of serial homicides in the U.S. -- and why do they kill? The increase in these violent crimes over the past decade has created an urgent need for more and better information about these men: their crime scene patterns, violent acts, and above all, their motivations for committing these shocking and repetitive murders.

This authoritative book represents the data, findings, and implications of a long-term F.B.I.-sponsored study of serial sex killers. Specially trained F.B.I. agents examined thirty-six convicted, incarcerated sexual murderers to build a valuable new bank of information which reveals the world of the serial sexual killer in both quantitative and qualitative detail. Data was obtained from official psychiatric and criminal records, court transcripts, and prison reports, as well as from extensive interviews with the offenders themselves.

Featured in this book is detailed information on the F.B.I.'s recently developed Violent Criminal Apprehension Program (VICAP) and a sample of an actual VICAP Crime Analysis Report Form.


Chapter 1

The Study of Murder

Interpersonal violence spans a wide range of human behaviors of which murder represents one of the terminal disruptions in the equilibrium of a society. Sexual homicide, a crime of increasing concern in our society, is the killing of a person in the context of power, sexuality, and brutality. Such crimes often receive widespread publicity, and they create a great deal of fear because of their apparent random and motiveless nature. Particularly in cases of serial sexual homicide, law enforcement officials feel public pressure to apprehend the perpetrator as quickly as possible.

Apprehension of the sexual murderer is one of law enforcement's most difficult challenges. Because sexual killings often appear motiveless and random, they offer few clues about why the murder occurred or, consequently, about the identity of the murderer. Even the sexual nature of these murders is not always immediately obvious, for conventional evidence of a sexual crime may be absent from a crime scene.

The Scope of the Problem

The number of sexual homicides occurring in a given year is difficult to assess, partially because of the manner in which crimes are investigated. In obvious cases of sexual assault and murder, the crime most often is reported as a homicide, not as a rape assault (Brownmiller 1975; MacDonald 1971). In other cases, conclusive evidence of sexual assault may be inadequate or lacking (Groth and Burgess 1977). Investigators may not recognize the underlying sexual dynamics of what appear to be either ordinary or motiveless murders (Cormier and Simons 1969; Revitch 1965). In addition, investigators often fail to share their findings, limiting the collective pool of knowledge on the subject (Ressler et al. 1980).

The difficulty in determining the scope of the problem of sexual homicide is reflected in official reports on murder. The FBI's Uniform Crime Reports (UCR) publication, Crime in the United States, presents statistics for crimes committed in the United States within a given year. For example, a survey of the statistics in the UCR for all murders committed in the period 1976 through 1986 shows that the number of murders in the United States has fluctuated from 16,605 in 1976 to a peak of 21,860 in 1980, dropping to 20,613 in 1986 (Crime in the United States 1986). The UCR also gives information about age, race, and sex of victims and offenders, about types of weapons used, and about situations in which killings took place. However, there is no differentiation in these official reports between homicides that included initially undetected sexual assault and those that did not.

The dearth of incidence statistics for sexual homicide promotes other approaches to evaluating this crime's magnitude. One of these approaches is to examine statistics for murders with unknown motives as indicators of the sexual homicide. Murder with unknown motive is one of five homicide categories for which the UCR annually provides data. These categories are as follows:

1. Felony murder (occurs during the commission of a felony).

2. Suspected felony murder (elements of felony are present).

3. Argument-motivated murder (noncriminally motivated).

4. Other motives or circumstances (any known motivation that is not included in previous categories).

5. Unknown motives (motive fits into none of the above categories).

Because sexual homicide so often appears random, motiveless, and with its sexual aspects obscured, some experts argue that many sexual homicides may be reported in the unknown motive category.

Figures for most categories of motives, except the unknown category, have remained relatively stable over the past decade. For example, felony-connected murders represented 17.7 percent of all murders in 1976, 17.2 percent in 1981, 18.0 percent in 1984, and 19.4 percent in 1986. The percentages for those murders placed in the "other" category are as follows: 1976, 18.6 percent; 1981, 17.1 percent; 1984, 17.6 percent, and 1986, 18.6 percent.

However, the number of murders committed for unknown motives has risen dramatically. These murders represented 8.5 percent of all murders in 1976, 17.8 percent in 1981, 22.1 percent in 1984, and 22.5 percent in 1986. This trend is of particular interest in the study of sexual homicide. Opinion has varied on the importance of this problem. Sociologists Wolfgang and Ferracuti proposed that the vast majority of murders are committed for a specific reason and that "probably less than 5 percent of all known killings are premeditated, planned, and intentional" (1982, 189). On the other hand, in a 1969 article psychiatrists Cormier and Simons cite studies observing that dangerous sexual offenses are both rare phenomena and nonescalating in aggression; in contrast, their data suggest that sexual murderers may be more common than we wish to acknowledge and that they do in fact have progressively violent records. The rapid increase in the unknown motive category, while most categories have held relatively steady, is an indication that the Wolfgang and Ferracuti proposition, first noted in 1967, may no longer be valid. Reporting methods for the unknown motive category have remained constant since 1976 and thus do not account for the dramatic increase in motiveless murder.

Although the incidence of sexual homicide may be difficult to ascertain, its effects are less unclear. Many sexual homicides are difficult to solve precisely because of the lack of clues. Law enforcement investigators note that solution rates may be misleading and should be scrutinized closely, since some crimes reported as unsolved are, in fact, solved after the figures have been reported. It is the belief of both investigators and clinicians that the majority of serial murders are sexual in nature (Lunde 1976; Ressler et al. 1985; Revitch 1965).

Veteran investigators say that sexual homicide is not a new phenomenon, although its apparent increase is. Public awareness of the problem has also increased, largely because of news media attention to these often bizarre, sensational crimes. People fear these crimes precisely because of the seemingly random nature of the murders.

Studies of Murder

Because of society's view of murder as both abhorrent and pervasive, the many aspects of murder have been studied by various disciplines. Law enforcement, the newcomer to the research arena, has added over the past decade its growing expertise in crime scene analysis.

Studies of murder have offered different perspectives on this violent crime. Epidemiological studies report on demographic data concerning victims and perpetrators (Constantino et al. 1977) and patterns of homicide (Rushforth et al. 1977; Wolfgang 1958). All disciplines have tried to categorize murderers. They have been typed in terms of motive (Revitch 1965), intent (Kahn 1971), number of victims (Frazier 1974), and type of victim (Cormier and Simons 1969). However, the majority of studies do not specifically address sexual homicide but rather study murder without differentiating between sexual and nonsexual crimes (Perdue and Lester 1974).

Studies examining murder divide into the various disciplines focusing on the subject: (1) individual case reports by journalists, (2) studies of the psychological aspects of the individual who committed the murder, (3) studies investigating murder and murderers within a sociological context, (4) legal aspects of the crime, and (5) crime scene patterns.

The Journalist's Perspective

The first category represents a mass communication mechanism for informing the public of crimes being committed. As such, the possible mechanisms include newspapers, magazines, and books. The writers are usually journalists, and on occasion they may have additional training like Joseph Wambaugh, formerly a police detective, who wrote The Onion Field, which depicts the murder of a police officer.

The news of a murder in general and a sexual homicide in particular arouses strong emotion in the public. Newspaper headlines highlight the bizarre aspects ("Body-hack Case," "Vampire Killer's Trial Begins"), the victim ("The Victim's Last Hours"), the serial aspect ("Killer Strikes Again"), the outcome ("Next Stop: The Electric Chair") and the twenty-year anniversary ("July 14, 1966").

Detective magazines are another mechanism for communication and often contain the facts learned at trial or during a journalist's investigation. In addition, journalists (Keyes 1976), as well as mental health professionals (Brussel 1968; Schreiber 1983), write books aimed at the general population.

The Psychological View

The second category includes studies that examine the individual murderer in terms of psychiatric diagnoses, childhood antecedents to criminal behavior, and criminal behavior as the result of learned responses to particular stimuli. Also in this category are studies designed to determine techniques for treating incarcerated violent offenders.

Study focus on the characteristics of individual murderers was evident as many as two hundred years ago. In the late eighteenth century, Swiss mystic Johann Kaspar Lavatar developed the art of physiognomy, whereby an individual's facial features were held to reveal his or her character. At about the same time, Franz Josef Gall contended that phrenology, the study of the formation of an individual's skull, could disclose both character and mental capacities. Physiognomy and phrenology were both used to study criminals, and death masks of murderers' heads were examined in an attempt to explain criminal acts (Attick 1970).

Among more recent theories that seek to explain violent human behavior through examination of individual physical characteristics is the approach based on the finding that certain violent criminals have a rare chromosomal abnormality called XYY (Jacobs, Brunton, and Melville 1965). According to this theory, the presence of an extra male, or Y, chromosome results in a greater likelihood that the individual will engage in aggressive and violent behavior. This idea received a great deal of attention in 1968, when it was reported that Richard Speck, the man who one night murdered eight student nurses, had the XYY abnormality. It was later found that Speck did not have the extra Y chromosome. Subsequent research demonstrated that although the XYY abnormality is four times more likely to show up in the male criminal population than in the general male population, most individuals with this condition display no abnormally violent behavior (Bootzin and Acocella 1980).

In the last decade, the field of psychiatry has paid increasing attention to brain function in all types of psychological problems. Neurological, genetic, and biophysiological components are being studied in the murderer (Lewis et al. 1986; Morrison 1981).ar

Psychological theories and explanations of violent behavior are substantially more common than those based on physical characteristics. As do studies based on individual physical attributes, these psychological studies focus on the characteristics of the individual murderer. According to Bootzin and Acocella (1980), psychological theories can be categorized as having either (1) a psychodynamic perspective (that unconscious childhood conflicts give rise to abnormal behavior), (2) a behavioral perspective (that inappropriate conditioning results in abnormal behavior), or (3) a humanistic-existential perspective (that abnormal behavior), results from a sense of personal failure).

Although individual-focused studies of violent behavior may be widely divergent in terms of their theoretical bases, they often have one thing in common: they group individuals into various categories. By producing taxonomic schemes to aid in understanding the causes of human violence, researchers seek to clarify methods of treating violent individuals and eliminating violent behavior.

Among recent general studies of human violence, several propose typologies having many categories. For example, Wille (1974) groups murderers into ten types: (1) depressives, (2) psychotics, (3) murderers with organic brain dysfunction, (4) psychopaths, (5) passive-aggressive individuals, (6) alcoholics, (7) hysterical personalities, (8) child killers, (9) mentally retarded murderers, and (10) murderers who kill for sexual satisfaction. In contrast, Guttmacher (1973) suggests six categories of murderers: (1) "average" murderers free from prominent psychopathology but lacking societal values, (2) sociopaths desiring revenge on the population at large for maltreatment during childhood, (3) alcoholics who kill their wives over fear of losing them, (4) murderers avenging a lover's loss of interest, (5) schizophrenics responding to hallucinations and delusions, and (6) sadists who kill to achieve sexual pleasure.

Other researchers have suggested fewer types of murderers. Megargee (1966) claims that there are only two types of extremely assaultive individuals (1) undercontrolled persons, who respond with aggression when frustrated or provoked, and (2) overcontrolled persons, who, despite rigid inhibitions against the expression of violence, build up an aggression that eventually results in violence. Simon (1977) studied thirty murderers and classified them into three groups: type A murderers had a tendency to murder impulsively, often with a weapon present and under the influence of alcohol. Type B murderers were more likely to be involved in a victim-induced homicide, with the victim usually a woman with whom the murderer had formed a relationship. Type AB murderers were able to sustain enduring, though sadistic, relationships with their victims and posed the greatest risk of future homicidal acts. Tanay (1972) has classified homicidal behavior into three categories: (1) ego-syntonic homicide, compatible with the offender's thinking and fulfills a consciously acceptable wish for the perpetrator; (2) ego-dystonic homicide, noncompatible with the offender's thinking and occurs against the conscious wishes of the perpetrator (who is in an altered state of consciousness); and (3) psychotic homicide, which occurs while the murderer is in a delusional state.

The few studies that specifically address sexual homicide suggest the existence of two types of sex murderers: the rape or displaced anger murderer (Cohen et al. 1971; Groth, Burgess, and Holmstrom, 1977; Prentky, Burgess, and Carter 1986; Rada 1978), and the sadistic, or lust, murderer (Becker and Abel 1978; Bromberg and Coyle 1974; Cohen et al. 1971; Groth, Burgess, and Holmstrom 1977; Guttmacher and Weihofen 1952; Podolsky 1966; Prentky, Burgess, and Carter 1986; Rada 1978; Ressler 1985; and Scully and Marolla 1985). Podolsky notes that rapists who murder kill after raping their victims, primarily to escape detection. These murderers, according to Rada, rarely report sexual satisfaction from their murders or perform postmortem sexual acts with their victims. In contrast, the sadistic murderer kills as part of a ritualized, sadistic fantasy (Groth, Burgess, and Holmstrom 1977). For this murderer, aggression and sexuality become fused into a single psychological experience -- sadism -- in which aggression is eroticized. According to Brittain (1970), subjugation of the victim is of importance to this type of sexual killer; cruelty and infliction of pain are merely the means to effect subjugation.

The various systems for classifying murderers have themselves been categorized. According to Athens (1980), the first system type assumes there is a uniformity, or pattern, to the characteristics of a particular violent crime. In theory, once such a pattern is discovered it will reveal the cause of the crime. The second type of system assumes that violent crimes are committed by individuals with psychopathic personality makeup; according to the third classification type, a combination of social and psychological factors results in violent behavior.

These studies represent a small sample of the many different classification systems of violent individuals. Nevertheless, they demonstrate that there is little relationship between the many different taxonomic schemes. Thus, as Megargee and Bohn (1979) have stated, they have limited use in providing information to those whose task it is to deal with the problem of violence and are instead designed more for theoretical discussion. A similar criticism has been voiced by Tanay (1972), who expresses concern that significant innovations in homicide control have not been developed, despite the many descriptions of the crime. Studies attempting to provide both a new description and a new solution for criminal behavior are based in learning theory (Samenow 1984). Researchers (Yochelson and Samenow 1976) spent many years studying criminals committed to a mental institution and found that specific thinking patterns separate hard-core criminals from society at large, regardless of background. In other words, criminal behavior is the result of the way criminals think, not of their environment or of other factors that are commonly used to explain this behavior. Thus, say the researchers, changing violent behavior must involve changing the way an individual thinks rather than changing that person's environment.

The Sociological View

In contrast to studies and classification systems based on the individual murderer, other research has examined homicide as a social phenomenon. Included in this third category are sociological studies examining violent crime within the context of the society in which it occurs. These studies consider cultural factors and social characteristics that might contribute to aggressive behavior. Like sociologists, historians have also studied murder with the belief that an analysis of the patterns of violence within a given society at a given time is a unique method of demonstrating many characteristics of that society (Buchanan 1977).

One such approach to understanding violent crime has been offered by Wolfgang and Ferracuti (1982). These researchers suggest that because the sociological approach to criminology, with its emphasis on cultural aspects, and the psychological approach, with its emphasis upon the individual, are apparently divergent, criminology must integrate the different ways of studying a homicide. Although an understanding of the social structure within which a murderer acts is as important as an understanding of the individual murderer, the valuable information to be gleaned from sociological research is not readily available for practical application. In addition, individuals who deal with the practical matters of homicide are unable to confirm researchers' theories.

With this need for integration as their starting point, Wolfgang and Ferracuti studied homicides that result from passion or from an intent to harm but not to kill. They theorized that within many cultures a subculture of violence exists and contended that homicide rates are highest among a homogenous subcultural group. In a subculture of violence, the use of force disrupts the values of society at large. Although the idea that values are actually motives for specific behaviors can be criticized as an oversimplification, it may be true that specific acts can reflect a basic value system.

The sociological literature has also looked at the victim. One of the most pervasive ways of analyzing victims has been through the concept of victim precipitation and victim participation. The victim is one of the causes of a crime, suggests Hans von Hentig. In 1948 he stated, "In a sense the victim shapes and molds the criminal...To know one we must be acquainted with the complementary partner." Mendelsohn (1963, 239-241), in writing of the biopsychosocial personality of the accused and of the victim, elaborated on the doctrine of victimology while preparing for the trial of a man who, had it not been for "the perversity of his former wife," would never have been found guilty of murdering her and her lover. Wolfgang (1958, 252) has used the concept of victim precipitation in his well-known studies of criminal homicide, applying it to those cases in which the "role of the victim is characterized by his having been the first in the homicide drama to use physical force directed against his subsequent slayer." An example is the husband who attacked his wife with a milk bottle, a brick, and a piece of concrete block while she was making breakfast. Having a butcher knife in her hand, she stabbed him. Wolfgang (1958) found victim-precipitated homicides represented 26 percent of a total of 588 homicides studied through police reports in Philadelphia. Adding to this concept, Schafer (1968, 152) concluded that "it is far from true that all crimes 'happen' to be committed; often the victim's negligence, precipitative action, or provocation contributes to the genesis or performance of a crime."

Modern researchers have stated the importance of sociological examinations of murder not only to understanding homicide, but also to understanding society's interrelationships, particularly those of the past. In a study of several sensational murders that took place during the Victorian era, Attick (1970) focuses on the social detail revealed through these cases. Statistics quoted by Attick indicate that although the characters of an individual murder may change from one historical period to another, patterns of murder do not. For instance, during the period 1837 through 1901, Attick analyzed that most murders were domestic in nature, while the next largest group was homicides committed during the course of a felony. A comparison of these statistics with the 1986 Crime in the United States figures shows that descriptive categories are fairly consistent with the addition of the unknown motive: arguments (37.5 percent); unknown motive (22.5 percent); felony (19.4 percent); miscellaneous nonfelonies (18.6 percent); and suspected felonies (2 percent).

The Legal View

The fourth category includes the mechanisms through which the legal community communicates its work, often in professional journals and occasionally through textbooks (Heymann and Kenety 1985).

The problem of murder has been described from the legal perspective, most often written from the trial perspective. In fact, law reviews often focus on specific cases. Emphasis is usually given to trial strategies from the prosecution or defense perspectives.

Law Enforcement's View

The last category of law enforcement's perspective looks at the investigation, suspect identification, and suspect apprehension as critical to homicide study. Law enforcement, whose responsibility it is to combat violent crime, has looked carefully at homicide analyses. Unlike other disciplines concerned with human violence, law enforcement does not, as a primary objective, seek to explain the actions of a murderer. Instead, its task is to ascertain the identity of the offender based on what is known of his actions. Described by one author as an emitter of signals during the commission of a crime (Willmer 1970), the criminal must be identified as quickly as possible to prevent further violence. While studies explaining why certain individuals commit violent crimes may aid them in their search, law enforcement investigators must adapt study findings to suit their own particular needs. Criminal profiling and criminal personality assessment are ways in which law enforcement has sought to combine the results of studies in other disciplines with more traditional investigative techniques in an effort to combat violent criminal behavior.

Criminal profiling has been used by law enforcement with success in many areas and is viewed as a way in which the investigating officer can narrow the field of investigation. This assessment does not provide the identity of the offender. Instead, it indicates the kind of person most likely to have committed a crime having certain characteristics.

Airport antihijacking measures and drug courier apprehension have been aided through law enforcement's use of basic profile characteristics. Screening techniques apply groups of characteristics to differentiate between the flying public and potential hijackers and drug traffickers.

Criminal profiling techniques have also been used in identifying anonymous letter writers (Casey-Owens 1984) and persons who make written or spoken threats of violence (Miron and Douglas 1979). In cases of the latter, psycholinguistic techniques have been used to compose a "threat dictionary," whereby every word in a message is assigned, by computer, to a specific category. Words as they are used in the message are then compared to those words as they are used in ordinary speech or writings, and the vocabulary usage of a particular author or speaker may yield "signature" words unique to that individual. In this way, police may not only be able to determine that several letters were written by the same individual, but they may also be able to learn about the background and psychology of the offender.

Rapists and arsonists also lend themselves to criminal profiling techniques. Through careful interview of the rape victim about the rapist's behavior, law enforcement personnel may be able to build a criminal personality profile of the offender (Hazelwood 1983). The theory behind this approach is that behavior reflects personality, and by examining behavior the investigator may be able to determine what type of person is responsible for the offense. Common characteristics of arsonists have been derived from an analysis of the Uniform Crime Reports data (Rider 1980). Knowledge of the arsonist's psychodynamics can aid the investigator in identifying possible suspects and in developing techniques and strategies for interviewing these suspects.

Profiling has been found to be particularly useful in investigating sexual homicide because, as was discussed earlier, many of these crimes appear motiveless and thus offer few obvious clues about the killer's identity. In murders that result from jealousy, a family quarrel, or commission of a felony, the readily identifiable motive generally provides vital information about the identity of the killer. Because many sexual homicides fail to provide this information, investigators must look to methods that supplement conventional investigative techniques in order to identify the perpetrator.

Criminal profiling uses the behavioral characteristics of the offender as its basis; sexual homicide crime scenes yield much information about the killer's behavior. A new dimension of information is available to the investigator through profiling techniques. For example, the apparent lack of motive in many sexual homicides is illustrated by the murderer's seemingly random selection of victims. Yet as Lunde (1976) points out, selection is based on the murderer's perception of certain victim attributes that are of symbolic significance to him. Although the victims are probably unaware of this significance, the murderer may assume the victims do indeed know of their place in his delusional scheme. Thus, an analysis of similarities and differences among victims of a particular murderer may produce important information concerning the motive in apparently motiveless crime. This in turn may provide information about the perpetrator himself.

The idea of constructing a composite of a murderer is not new. In 1960 Palmer published results of a three-year study of fifty-one murderers who were serving sentences in New England. Palmer's "typical murderer" was twenty-three years old when he committed murder. With a gun, this typical killer murdered a male stranger during an argument. He came from a low social class and achieved little in terms of education or occupation. He had a well-meaning but maladjusted mother, and he experienced physical and psychological frustrations during childhood.

Similarly, Rizzo (1982) studied thirty-one accused murderers during the course of routine referrals for psychiatric examination at a court clinic. He concluded that the average murderer was a twenty-six-year-old male who most likely knew his victim and for whom monetary gain was the most probable motivation for the crime.

Criminal profiling as used by law enforcement today seeks to do more than describe the typical murderer, if in fact there ever is such a person. Law enforcement investigators must assess the crime scene in terms of what it may reveal about the type of person who committed the crime. Although Lunde has stated that the murders of fiction bear no resemblance to the murders of reality (1976), a connection between fictional detective techniques and modern profiling methods may indeed exist. For example, it is an attention to detail that is the hallmark of famous fictional detectives; the smallest item at a crime scene does not escape their notice. As stated by the famous Sergeant Cuff in Wilkie Collin's 1868 novel The Moonstone, widely acknowledged as the first full-length detective story,

At one end of the inquiry there was a murder, and at the other end there was a spot of ink on a tablecloth that nobody could account for. In all my experience along the dirtiest ways of this dirty little world I have never met with such a thing as a trifle yet.

This attention to detail is equally as essential to present-day profiling. No piece of information is too small; each detail is scrutinized for its contribution to a profile of the killer.

In contrast, criminal profiling methods do not share with fictional techniques the reliance on one specific clue to solve a case. The broken matchstick that leads police to the killer's door simply does not exist. Instead, experienced profilers look at the overall picture of the crime. Obvious as well as implied clues are used together, not individually, to provide information about the murder.

Criminal profiling is best viewed as a way for law enforcement to focus its efforts in a particular area. Profiling has been described as a collection of leads (Rossi 1982), as an educated attempt to provide specific information about a certain type of criminal (Geberth 1981), and as a biographical sketch of behavioral patterns, trends, and tendencies (Vorpagel 1982). Geberth has also described the psychological profile as particularly useful when the criminal demonstrates some form of psychopathology, and has stated that the crime scene often reflects the murderer's behavior and personality in much the same way as furnishings reveal the homeowner's character. Criminal personality assessment helps investigators identify those patterns of behavior and personality.

Copyright © 1992 by The Free Press

About The Authors

John Douglas is the legendary FBI criminal profiler and former Chief of the FBI’s Behavioral Sciences Unit where he researched, investigated, and conducted interviews of some of America’s most violent criminals. Over the past half-century, he either directly worked on or had overall supervision in over 5,000 violent crime cases. He is one of the foremost experts of the criminal mind, its methods, and motivations. Douglas is a veteran of the United States Air Force and holds a doctorate degree in Education, and lives with his wife, Pamela, in the Washington, DC, area.


Product Details

  • Publisher: Free Press (June 1, 1995)
  • Length: 256 pages
  • ISBN13: 9780028740638

Browse Related Books

Resources and Downloads

High Resolution Images

More books from this author: John E. Douglas