Despite what you are about to read I am a great fan of Cracker and Val McDemid’s Dr Tony Hill. However profiling in real life is a pale shadow of its fictional counterpart. The theory of profiling was first developed in Britain by Dr David Canter, who built on the approach of the FBI’s Behavioural Analysis Unit. That unit was staffed by investigators rather than psychologists and, according to Dr Canter, used a pragmatic rather than academic approach to profiling, relying on their intuitive interpretation of every case based on their experience rather than developing a set of principles that could be applied to any case.
The attraction of profiling is that it gives the investigator a picture of the criminal thus enabling the search for a suspect to be limited in some way – e.g. where he or she lives and works, any criminal history, or whether he or she is married or single. The difficulty is that, for a number of reasons, it is not at all reliable and its use can bring with it significant dangers.
I found it difficult to discern clear principles from Dr Canter’s work and he appears now to be focusing more on the issue of geographic (where the suspect is likely to live or work) rather than psychological profiling. The few that I can find from the literature appear to be as follows.
Commuter / Marauder
This principle is more successful when considering a series of rapes / murders which occur in clusters in bigger towns and cities, as it lies in developing a relationship between the locations of the crimes and where the criminal lives (or sometimes works). Commuters live outside the area of the crimes and focus their offending in one or two specific areas where they feel comfortable, perhaps because they have lived or worked there at some time, or believe that it is where they are more likely to find the type of victim they seek. Marauders live in the area where the offences occur. The problem for the detective is that until he or she catches the criminal, it is unclear whether they are a commuter or a marauder, as this factor is not predictably associated with any other, for example if the marauder always attacked their victim in a public area it would be possible to associate these two factors. All the investigator has are clusters of similar crimes probably committed by the same person who feels comfortable working in that area – and he or she knows that already.
Three Types of Rapist / Murderer
Dr Canter describes three different types of rapists according to how they treat the victim: as an object, as a victim or as a person. He believes that this can be useful as each has distinctive features which may assist the investigator to identify potential suspects.
The victim as an object. There the rapist is only interested in the victim as something to control. He has no feelings towards them, not even anger, and only uses them to fulfil his inner fantasies. Features of this type of offender are:
- the assault is made in public and the victim is usually a target of opportunity
- he will tend to attack the same types of person – young girls, older women, gay men
- the control may continue after death – necrophilia, retention of the body or body parts, cannibalism
- he is unlikely to be living with a partner.
The victim as vehicle. This person has much in common with the first. The key difference is that he sees the role that empathy can play although he has never or rarely felt it for anyone, certainly not his victim. She is seen as someone to be exploited so that the killer can somehow restore himself to being the person in control. Features of this type are:
- he may have a history of unsuccessful relationships
- the initial contact may not be threatening to the victim
- the location may carry some meaning for the attacker
- emotional events in the killer’s life may provide the trigger for the assaults to begin or temporarily stop
- he will tend to be a commuter
- he will be from the older end of the attacker age range, i.e. late twenties or early thirties.
The victim as a person. This attacker come as close to ‘normal’ as is possible in that he tries to empathize with the victim and succeeds in his own distorted way. During the course of the assault he will attempt to develop a relationship with the victim, perhaps by asking about boyfriends and other personal details. In this way he confuses the rape with a real sexual relationship. Features of this type are:
- he moves to rape from other crime, especially burglary
- he may stalk the victim first
- he may break into the victim’s house and wait for her to return
- he may be living with a partner although it is probable that the relationship will be a difficult or uneasy one in which he continually seeks to dominate.
Other Descriptive Factors
The literature describes the following factors as being common to serial killers. It does not give any clues about how many of the seven factors you are likely to find, or the likelihood of any particular one being present. Serial killers tend to:
- be working-class, in unskilled jobs providing a low income and little job satisfaction
- come from families that were dysfunctional in serious ways – rejection by one or both parents is a feature
- be poor at forming relationships
- be the runts of the family, small in stature
- have been poor achievers at school
- have become involved in minor crime at an early age
- have ‘psychopathic’ personalities (whatever that may mean – my psychopath may be your sane but obsessed collector).
Unfortunately most of these factors apply to most of the criminals I have come across so their use as a reliable predictor is questionable, and any study of serial killers will provide examples to which none, other than sexual fantasies and psychopathic personalities, apply – look at Dennis Nilsen and Ted Bundy. This list look impressive at first sight but from an investigator’s point of view the more factors there are the less useful they become as some or most of them could apply to anyone. In Wire in the Blood Val McDermid lists thirty factors, then adds ‘a psychopathic personality’. Anyone satisfying most of the factors would need to be either locked up or a saint.
Problems with Profiling
From a police perspective I have two major concerns about profiling. First, the methodology demands that it can only be based on criminals who have been caught; secondly, it can tempt the SIO and his or her team to become too narrowly focused too quickly, and keeping an open mind is a key skill for an SIO.
Only the Ones We’ve Caught
Profiling is based on the analysis of crimes and people who have been arrested for them. This means that the profile is based on the least successful or most unlucky criminals. It is not surprising, therefore, to find that they were poor achievers at school. Dr Harold Shipman, intelligent, with a stable family background, the father of three sons and apparently a good husband and father, is reckoned to have killed around 250 people, mostly older women. There is some evidence that he began killing people in the early 1970s and he was not arrested until 1998. Even his arrest was triggered by the fact that he had forged a will for his latest victim, and some interpret this as a desire on his part to be caught. While it is probable that his case is an unusual one and his position gave him fairly unique access to potential victims, the fact that he got away with his crimes for so long and does not fit any of the factors described in the profile powerfully shows that there could be an unknown number of Shipmans out there, knowledge about whom would fundamentally change the principles upon which profiling is based.
Keeping an Open Mind
Closing down the options too quickly is the greatest danger that profiling presents. As I have said, the SIO must keep an open mind, especially at the beginning of an investigation or when it has run into the buffers. The dangers of failing to do so are only too clear, as the cases of Colin Stagg and the Yorkshire Ripper show.
In 1992 Rachel Nickell was found brutally murdered on Wimbledon Common with her young son clutching her bloodied body. The case created a predictable public outrage and the pressure on the inquiry team to get a result must have been immense. Colin Stagg came into the frame as he was seen frequently on the common and he fitted the profile developed by Paul Britton, a criminal psychologist used by the SIO. As a result Mr Stagg was made the target of a ‘honey pot’ operation and was induced into making a number of compromising statements, none of which amounted to an admission. After spending fourteen months in prison and having his life destroyed by the way the murder and the trial were reported he was acquitted. Another man, Robert Napper, has since been charged with the murder and was due to face trial in November 2008.
The Yorkshire Ripper case shows how focusing on one line of inquiry to the exclusion of all others can cripple it. The murder team there received tape recordings and letters from a man claiming to be The Ripper. The voice on the tapes had a Geordie accent and the SIO, Detective Chief Superintendent George Oldfield, decided to focus the inquiry on this lead. A detective who interviewed Sutcliffe during this time put him into the ‘strong suspect’ category but was overruled as Sutcliffe did not have a Geordie accent.
The most infamous case in British police history showing the danger of having a closed mind was the mass murder of the Bamber family in Essex in 1985. Jeremy Bamber and his sister Sheila married and had young twin sons. On the break-up of her marriage she became depressed and was diagnosed as paranoid schizophrenic. At the time of the killings it is known that she had stopped taking her medication and there were concerns for the twins’ well-being. At 3.26 on the morning of the 7th of August Jeremy Bamber telephoned the local police to tell them that his father had just telephoned him saying, ‘Sheila’s got the gun, she’s gone crazy, come over here quickly.’ The call had then been ended and when he phoned back he could only get an engaged tone. He told the police that his sister had a history of mental illness and he arranged to meet them at his parents’ house at White House Farm.
Bamber arrived at the farmhouse shortly after the first police and by 5.30 they were joined by armed police. They discussed Sheila’s history with Bamber and eventually, at 7.30, broke into the farmhouse. They found Ralph Bamber’s body in the kitchen, which was in disarray and looked as if a struggle had taken place; June Bamber’s body was in the main bedroom; the twins were found shot in the head, lying in their beds, and Shelia’s body lay by the side of her parents’ bed. They had all been shot with the .22 rifle that Shelia was found clutching to her chest. She had two gunshot wounds to the throat. The rifle was bolt action, which meant that the bolt had to be worked between shots to eject the used round, reload the chamber and cock the trigger. The SIO immediately took the view that he was dealing with a mass murder of the family by Sheila, after which she took her own life.
The police not only allowed Bamber to have the house cleaned very soon after the killing but actually assisted him in the process, thus removing almost all the evidence and compromising what was left.
Following the inquest Jeremy Bamber’s behaviour began to raise the suspicions of both the police and Ralph and June’s relatives. When one of them, David Boutflour, was clearing some items from the farmhouse he found the .22’s noise moderator (silencer) in the farmhouse gun cupboard. He saw that it had red paint marks which looked as if they had come from some paint damage in the kitchen and that there was a grey hair stuck to the end. He brought this to the attention of the police and they had the silencer examined forensically. That examination found that the paint was the same as that in the kitchen and was consistent with damage seen there, and that there were spots of blood inside the silencer which were the same blood type as Sheila’s (DNA was not used at that time). With the silencer mounted it would have been impossible for Sheila to shoot herself and she could not have shot herself then removed the silencer to put it back in the gun cupboard. A body of circumstantial evidence was then developed pointing to Jeremy Bamber as the killer which included the following facts:
- he appeared to have telephoned his girlfriend telling her about the murder before his phone call to the police
- he was in debt
- his girlfriend at the time of the killings said that he had talked of murdering his parents.
He was arrested, tried and convicted. He has appealed three times unsuccessfully, and still protests his innocence. I have not seen enough of the original evidence to make a judgement, and most of the accounts of the murder are given from a partial viewpoint. But the key issue for me is the fact that the SIO’s decision that Sheila had killed the others, then committed suicide, was made without enough consideration of the evidence available. Given the scale of the killings it is certain that there would have been enough forensic evidence at the scene, Jeremy Bamber’s house and on the route between the two locations to provide definite proof either way. The fact that the house was cleaned and that evidence was not collected until several weeks later effectively deprived the investigators of the bulk of the forensic, and thus objective, evidence, leaving them to rely on circumstantial, therefore interpretative, evidence.
Dr Canter does not appear to appreciate this danger. In his book Criminal Shadows he describes how the FBI agent Roy Hazelwood of the Behaviour Science Unit ‘knew’ on entering a crime scene that ‘a black assailant who lived in the vicinity had done it’. If I were an SIO and a profiler said this to me I’d kick him off the team immediately because of the danger that he would then build a profile where the factors that supported his ‘knowledge’ would be recognized and maybe even enhanced, and those that didn’t would be ignored.
This certainly still appears to be part of Dr Canter’s approach. In an article in the Guardian (27 December 2007) he discussed the Jill Dando shooting and said that he been consulted he could have told the investigators that over 80 per cent of killers who escape on foot live within 525 yards of the crime. It took police over a year to question and arrest Barry George, a man with a record of stalking celebrities. He lives within the 525-yeard radius. He went on to say, ‘The point is not that it was Barry George and that I turned out to be right. The point is that, with a focused investigation, the police could have got to him a great deal earlier than they did and eliminated him if necessary.’
In fact Barry George had been treated as a suspect for some time, but a profiling approach based on more than distance would appear to exclude him. He is a fantasist, adopting a number of pseudonyms including Paul Gadd (Gary Glitter) and Barry Bulsara (Freddie Mercury). He is described as inept and disorganized and has suffered from learning difficulties all his life. Jill Dando’s killer on the other hand was either very organized or incredibly lucky. He shot her at her front door in the middle of the day in a central London street. The front gardens in this part of London are tiny, only a few steps separating the garden gate from the front door. The killer forced Miss Dando to the ground and effectively used her head as silencer. Only two people saw him walk away and their descriptions do not appear to readily fit Barry George. The streets in this part of London are narrow, with cars parked on both sides. The danger of being blocked in by a double-parked car is ever present so the safest exit from the scene would be on a motorcycle or on foot to a rendezvous point to be picked up. On 15 November 2007 the Appeal Court found the conviction of Barry George to be unsafe and ordered a retrial. He was acquitted on 1 August 2008.
Catching Serial Killers
If profiling is of such limited use, how are serial killers caught? Usually it is a mixture of arrogance and bad luck on their part and solid police work. Their arrogance shows in trying one more time, even when they know the police have mounted a large operation to find them; bad luck occurs in that a member of the public or a police officer sees something suspicious and solid policing makes sure that this break is exploited. To see how these factors work together we will look at some of the more notorious British serial killers:
- Dennis Nilsen and Peter Sutcliffe because of the number of people they killed before being caught
- Robert Black to look at the limitations of geographic profiling
- since their case involved the first use of profiling, the ‘railway murderers’ John Duffy and David Mulcahy.
Between 1979 and 1987 Dennis Nilsen killed sixteen men, and five others survived his attacks. His modus operandi was to pick up men in gay bars and clubs and take them back to his flat. He usually selected men who were homeless or whose absence would not be notices. He was discovered only because he overloaded his drains with bits of body. A drain-clearance operator was called to the blocked sewer outside Nilsen’s home. In the inspection chamber he found what he thought were bits of human flesh. He reported this to his supervisor who advised him to ‘sleep on it’ and call the police in the morning if he was still worried about it. This he duly did and DCI Peter Jay went with him to inspect the drain. In the meantime Nilsen had been told about his visit and had done his best to ensure that the sewer was clear by the time the police arrived.
However DCI Jay still found some bones that he thought were knuckle bones. He took them to Professor Bower of Charing Cross Hospital, a highly experienced Home Office pathologist. He identified the bones as being neck bones and recognized marks on them as being consistent with ligature marks, indicating that the person had been strangled. So the police had two pieces of luck: the discovery of body parts, and the fact that they were examined by a highly experienced pathologist, as not every pathologist may have recognized the significance of the marks.
DCI Jay returned to Nilsen’s flat where there took place one of the shortest and easiest interrogations of a murder suspect in history.
DCI Jay: Mr Nilsen, your drains were blocked with human remains.
Nilsen: How awful.
DCI Jay: Now don’t mess me about, where is the rest of the body?
Nilsen: they’re in two plastic bags in the other room. I’ll show you.
DCI Jay: I thought so. What’s been going on here?
Nilsen: It’s a long story. I’ll tell you everything.
In fact there were the parts of two bodies in the bedroom and two heads in the fridge.
Profiling would not have helped with Nilsen. He killed all his victims in his flat and disposed of the remains through the drains. There could be no geographic pattern to follow. His motivation did not appear to be sexual as consensual sex had already taken place and at least ten men passed through his flat for every one who was attacked. Even if there had been a pattern of deaths it is not clear that Nilsen would have fitted any profile. He was described as a ‘grey man’. He was reasonably well educated and at the time of the killings had a responsible job in the Manpower Services Commission. He had no difficulty in finding worthwhile employment, he had been a probationary police officer and had trained as a chef.
The Yorkshire Ripper
Peter Sutcliffe attacked twenty-one women between 1975 and 1981, killing thirteen of them. Most were prostitutes. Most were killed using a hammer blow to the head followed by knife slashing to the stomach and groin area. The distinctive m.o. and selection of victims made it clear to investigators that they were looking for one killer from a very early stage. The inquiry was complicated by the fact that the attacks occurred in three police force areas and the police had not yet developed protocols for dealing with this nor a computer system to support it. The spread of the attacks in both time and geography made it impossible for police to mount an effective patrol strategy and at that time there was not the CCTV coverage which was critical to the success of the recent Suffolk case where Steve Wright killed five women over a three-week period.
On 2 January 1981 Sutcliffe was stopped by police driving a car with false number plates. He was accompanied by a prostitute. He was arrested and at the police station was questioned at length because of his similarity to the description of the Ripper. The arresting officer, a Sergeant Ring, decided to go back and search the place where he had arrested Sutcliffe as the latter had asked to be allowed to go behind an oil storage tank to urinate. There he found a hammer, a knife and a rope. After two days of questioning Sutcliffe suddenly broke and admitted that he was the Ripper. His confession took over sixteen hours and filled thirty-four pages. Some weeks later he claimed that a voice from God had told him to attack women.
In theory profiling should have worked in the Ripper case, had it been available. Sutcliffe not only had a very consistent m.o. – only five of the murdered woman were not prostitutes, all bar one was attacked outside, and a ballpein hammer and knife were used in most of the assaults. He had also developed the ‘signature’ in that he mutilated the victim by slashing the stomach and groin areas, even when they were already dead. Yet if he is examined against a list of profiling characteristics the likelihood of a profile helping to identify him looks poor:
- although working-class he used redundancy money from one job to become qualified as an HGV driver and had a fairly good employment record
- he was described as a quiet, shy boy
- his family were dysfunctional in that the father was a bully and the mother a doormat. Both had fairly public affairs. This does fit the profile but is not the sort of information that is available to detectives who are trying to prioritize a list of thousands
- he had met his wife in 1967, married her in 1974 and their marriage had lasted through a series of miscarriages
- he is of average height
- he left school at fifteen but that was normal for working-class boys of that age at that time and this needs to be seen in the light of his HGV qualification
- he had no previous criminal record.
If profiling had been used it is likely that Sutcliffe would have been closer to the bottom than the top of any list. His stable employment background, apparently settled marriage and lack of criminal record would probably have been enough to ensure that he never made a ‘strong suspect’ list because he didn’t fit the profile, just as had the fact that he did not have a Geordie accent.
Although the inquiry itself could never be considered a success it is worth considering the effort that the three forces put into the inquiry:
- 260,000 people interviewed
- 32,000 statements taken
- 5.4 million car registration numbers checked
- 40 tonnes of paperwork accumulated
- 250 detectives engaged in the inquiry over three years.
Sutcliffe was interviewed nine times and as we have seen one detective at least thought that he should be treated as a ‘strong suspect’. At the time of his arrest a line of inquiry was being pursued which might have identified him. A new five-pound note was found in the discarded handbag of one of the victims, Jean Jordan. It had been tracked back to a wages payment which produced a list of 300 employees. The sequence of notes containing the five-pound note would have reduced that list to thirty and Sutcliffe was in this group.
I have included Black as his case shows the difficulties created when a killer is mobile and moves across not just police areas like Sutcliffe, but even jurisdictions (this is a more common problem in the USA – the issues this raises are discovered in Chapter 6).
On 30 July 1982 Susan Maxwell, aged five, was abducted from Portobello just outside Edinburgh. Her naked body was found in a ditch in Leicestershire. Decomposition of both bodies made it impossible to determine the exact cause of death, or whether they had been raped.
On 26 March 1986 Sarah Harper, aged ten, was abducted from Leeds. Her naked and strangled body was found dumped in the River Trent near Nottingham a month later.
All three bodies were found within thirty miles of one another and the common m.o. led to the inquiries being linked. By the time of Sarah Harper’s murder the HOLMES system had been developed and the inquiry had computer support. The pattern of the killings and bodies indicated that the killer probably travelled as part of his occupation – perhaps as a salesman or a lorry driver. An enormous police investigation was launched and every lead was followed up; for example in Sarah Maxwell’s case a maroon Triumph 2000 was suspect and 19,000 drivers were seen, interviewed and eliminated; similarly in Caroline Hogg’s case a blue Cortina was suspect and this led to 20,000 drivers being seen, interviewed and eliminated.
On 14 July 1990 near Stow in Scotland, about thirty miles from the border with England, Mandy Wilson, aged six, was seen by a member of the public being bundled into the back of a van. He called police and was able to give them a registration number. Miraculously the van passed the witness when he was still talking to the police and it was successfully stopped and Black was arrested. One of the officers at the scene of the arrest was Mandy’s father. She was found bound and gagged, stuffed into a sleeping bag in the back of the van. Had Black not been stopped she would probably have suffocated in a relatively short time.
Black pleaded guilty to the abduction of Mandy Wilson but has never admitted abducting and killing the other three. He was tried and found guilty of their murder in May 1994. He was sentenced to life imprisonment and the judge recommended that he served at least thirty-five years, making him eighty-two before he would be released.
Unlike Sutcliffe, Black fitted almost all the academic criteria.
- he was working-class, a poor time-keeper and lucky to find and keep a job where he could work unsupervised
- there is evidence that he suffered physical abuse at the first home where he was fostered and that he was sexually abused in one of the children’s homes where he lived
- although he appeared sociable, was a regular player on the amateur darts circuit and liked football, he never formed any friendships and had few girlfriends
- he was clever enough to get in a grammar school but made nothing of it
- his record of indecent assaults from an early age (twelve) showed that he regularly indulged in sexual fantasies
- he was convicted of a serious sexual assault at the age of seventeen although it was recorded as a relatively minor one. Paedophile killers are more likely to have a criminal record and some of their convictions may be for sex offences, although they may not be particularly serious – simply being a peeping Tom.
Would profiling have helped to find Black sooner? Geographic profiling certainly would not. The range of locations would have defeated any attempts to focus enquiries. Black lived in and worked from London while the girls were abducted from the north and their bodies were dumped in the Midlands. Any such focus would only have misled the inquiry.
A psychological profile, however, might have helped to identify him. After his conviction the SIO was criticized for restricting the possible suspects to people with a criminal record for more serious offences. This meant that Black was never in the system. The SIO justified this on the understandable grounds that the inquiry could not have handled the bigger list a wider net would have created. The effort required to eliminate suspects cannot be underestimated, as all the facts that they give need to be corroborated before they can be eliminated. A profile might have enabled the inquiry to broaden the net to include minor sex offences as it could then have been used as a filter to prioritize these ‘possibles’. This is easy to say with the benefit of hindsight knowing that Black did it. I am not sure that any SIO today would be willing to put the level of faith in the profile that such an approach would demand.
The Railway Murders
Finally we come to the case that introduced the use of profiling in Britain. Between 1982 and 1985 John Duffy and David Mulcahy abducted nineteen women from or near railway stations in north London and raped them. Duffy matched the description given by some victims and he was put on an identification (ID) parade but was not identified. Mulcahy was only routinely questioned and released.
On 29 December 1985 Alison Day was abducted at Hackney railway station, repeatedly raped her and then strangled with a piece of string.
On 17 April 1986 Maartje Tamboezer, aged fifteen, was attacked near West Horsley railway station in Surrey. She was raped, strangled with string and an attempt was made to set her body on fire.
On 18 May 1986 Anne Locke was abducted from the Brookmans Park railway station in Hertfordshire then raped and murdered.
The Metropolitan Police (Met) had already linked the murder of Alison Day with the previous rapes and the SIO of the inquiry recognized similarities with the murder of Maartje Tamboezer. Both murders were linked and the joint inquiry was now led by Detective Chief Superintendent Vince McFadden of Surrey. Dr Canter, then of Surrey University, had already been contacted by the Met to discuss the possibility of developing a profile. This process was continued by Vince McFadden, who gave Dr Canter a couple of detectives to assist him in order to speed up the work. The profile that Dr Canter developed fitted Duffy in thirteen of the seventeen characteristics.
Duffy was in the list of over 2,000 ‘possibles’, and had been close enough to the description given by one of the victims to have been put in an ID parade. It was also known that he had raped his wife, from whom he was separated, at knifepoint and she had complained that he had forced her to have sex while they were together by tying her up. After he became known to the inquiry his behaviour took a bizarre turn; he appeared battered and bruised at a police station claiming to have been attacked and as a result having lost his memory. He was admitted to a psychiatric hospital for treatment for this ‘amnesia’. Later an associate of Duffy’s came forward to say that he had beaten Duffy up at his request so that Duffy could convince the police that he had amnesia.
Duffy was then targeted and put under surveillance. He was arrested in the act of stalking a woman in a park. A search of his house revealed the same unusual string that had been used to bind the murder victims. Forensic evidence of fibres on his clothing, blood-typing and a positive identification on an ID parade built up a very strong case against him. He was found guilty of two of the three murders and five rapes and was given life imprisonment with a recommendation that he serve at least thirty years. He did not admit to any of the offences, so did not give any information about his accomplice, Mulcahy.
In 1997, while in prison, Duffy admitted the murders andhis role in the murder of Anne Locke to a counsellor. He also implicated Mulcahy, who had been a friend since childhood, claiming that Mulcahy was the instigator and major planner of the killings. The cases were reopened and DNA recovered which corroborated Duffy’s accusations. Mulcahy was arrested in 1999 and found guilty of the murder and rape of the three women and of seven more rapes. He was given three life sentences for the murders and twenty-four years for the rapes.
At first sight the fact that Duffy hit Dr Canter’s profile in thirteen of the seventeen characteristics looks impressive, and it was treated as such at the time. A closer analysis in the light of Duffy’s admissions and Mulcahy’s convictions, however, significantly changes its impact. Despite the fact that the police were certain that there were two assailants in most of the rapes and were pretty sure that Duffy had an accomplice in the murders, no profile was developed for a possible second assailant.
Although there was a high probability that two were involved no allowance appeared to be made for the fact that theprofile could have been a composite of two men. In Criminal Shadows, Dr Canter describes his reasoning to support the profile: ‘He [the killer] had become a lone attacker.’ But Duffy was raping with Malcahy in July 1985, and Alison Day’s murder was in December of that year. The pattern did not indicate that either rapist had begun to work alone and police had always suspected there was a second killer.
Dr Canter continued: ‘It would be strange indeed that a man could keep a deep, caring relationship with his wife and still go out frequently with a friend or on his own to rape and then murder.’ But Mulcahy was married and stayed married until his arrest in 1999.
Dr Canter says: ‘The viciousness of Duffy’s assault on the 15-year-old Maartje Tamboezer seemed to me very difficult to associate with the actions of a man who had been involved in bringing up children himself.’ Mulcahy has four children.
You can see from all that I have written that I am sceptical about the usefulness of profiling. I have found no cases in the UK where it has played a significant role in the investigation. Even in the USA where there is a greater need and the technique has a longer history it would not have helped to identify that country’s most ruthless serial killer, Ted Bundy, as he did not come anywhere near fitting the classic background – he was a graduate, was involved in politics and changed his m.o. at least three times, whereas his treatment of victims as an object would, in profiling terms, suggest a poor achiever making disorganized attacks.
So can a crime writer use profiling? It is a question of fact and degree as the lawyers say. Cracker was phenomenally popular, as is Val McDermid’s Dr Tony Hill. If the characters are believable, the plot is good and the text persuasive it can obviously be a useful crime-writing device. But remember that in Wire in the Blood the killer was initially identified by a straightforward analysis of missing persons reports together with a trawl of local newspapers to see what they coincided with. The psychological element came later – quite like real life actually.