by Sue Black
Recognizing a face is one skill; being able to describe it is another. We are all familiar with the facial composites produced by the police from witness descriptions to help with the identification of criminals. The different areas of the face are considered individually and then brought together to build the final face: forehead, eyebrows, eyes, nose, cheeks, mouth and chin.
Originally, likenesses were drawn by artists. Identikit, the first trademarked system using templates of separate features, was introduced in the United States in 1959. Subsequent methods such as Photofits and e-fits, involving photographs and electronic software, may produce more polished-looking results but they still rely on each feature being selected singly from a database of possibilities and then overlaid to construct the final composite image.
Nobody would claim that this can create a perfect replica of the subject. If you put together a face from, say, Angelina Jolie’s eyes, Stephen Fry’s nose and Eartha Kitt’s mouth, you are bound to come up with something of a dog’s dinner. The aim is to produce an image that has sufficient resonance with the viewer to elicit some intelligence that can be followed up by investigators. It is accepted that the accuracy of a composite face may be less than 50 per cent, which might not be considered terribly encouraging, but we must remember that sometimes it might be all an investigation has to go on. There is a tendency for the human eye to focus on, and the brain to remember, the unusual or abnormal. This can cut both ways. If an anatomical anomaly is present, and correctly described, it can prove to be an enormous help in the identification process. However, if it is wrong, it can send an inquiry severely off track.
Our recognition skills are, of course, normally only called upon in the context of identifying fellow living human beings. When it comes to recognizing the dead, our perceptions can be very different. Those of us who have sat with a loved one through their dying and death, or who have paid our respects to their bodies prior to their funeral, will be aware of how, when the essence of a person, and the animation and expression of their face, has been lost, the outer shell in which it resided often does not look quite like the person we remember. It is usually very much smaller and somehow empty.
Those facing the horrendous task of trying to identify someone who has suffered a violent or catastrophic death, or an individual who has been dead for some time, are going to find recognizing the person they loved more challenging. In the wake of the 2002 Bali bombings, around half of the bodies were incorrectly identified by their families, who had walked up and down rows of bloating, decomposed and fragmented corpses looking for their missing relative.
It is no surprise that, in such traumatic circumstances, many got it wrong. Their distress, the daunting mortuary environment and their urgent psychological need either to find, or not to find, their loved ones would all have contributed to their confusion. It is difficult to suggest to a close relation who insists they are 100 per cent certain, either for or against an identification, that they might be mistaken. It is for this reason that Interpol standards for disaster victim identification (DVI) stipulate that bodies should not be returned to their families on the basis of facial recognition alone. For a body to be released with scientific certainty, one of the three major identifiers is required: DNA, fingerprints or dental information.
When the face of a deceased person becomes unrecognizable, through decomposition or damage, we can reconstruct it in an attempt to identify them. Reconstruction is a tool in our forensic arsenal to which we frequently turn when all other avenues have been explored, and it requires a special set of skills that combines both art and science. The basic premise of facial reconstruction is the close relationship between our appearance and the morphology of the underlying bone and its covering of muscle, fat and skin.
Facial reconstruction can be achieved either by producing clay models or through three-dimensional computer modelling. The Manchester method, accepted today as the gold standard, and which I believe to be the most rigorous, requires us to have a skull, or at the very least a good cast or three-dimensional scan of the skull. Wooden pegs are glued, either physically or virtually, to the skull to indicate the thickness of the soft tissue that covers the bones at all points. This will vary according to the sex, age and ethnic origin of the individual.
Next, each of the forty-three or so muscles is added, one by one, layer by layer, to build up the underlying soft tissue scaffold as accurately as possible. The parotid glands, the major saliva glands, are also included at the side of the face, as are the buccal pads of fat in the cheeks. Then the skin is laid over everything and moulded into the contours of the face, much as you would lay a sheet of icing across the surface of a cake.
How the more cosmetic elements of the reconstruction are approached will depend on its purpose. Sometimes we do facial reconstructions for display, for example, with archaeological remains for a museum exhibit. With those models, the artist can be given reasonable leeway to add skin tone, eye colour, hair colour and style, facial hair and so on.
If it is to be issued to the press in the hope that it will help to identify a body, a greyscale illustration may be produced. Forensically, we may not be sure of skin colouration, and we do not want to guess at hair or eye colour as this could unduly influence the viewer to include or exclude a possible candidate.
The current research on DNA phenotyping may before long consign this uncertainty to the past. It is believed that we can now identify natural hair colour or eye colour from DNA. Other more complex features, such as eye shape, nose length or mouth width, may also have a genetic predisposition. They are more difficult to interpret, but it may well be possible, some time in the future, to partially reconstruct a face that looks like the living person from DNA alone.
Often, a simpler depiction may be sufficient to create a likeness that renders a disfigured or decomposing face acceptable for general circulation. This was the direction about to be taken by North Yorkshire police when I was approached to assist with the identification of a young female who had been found in the most unusual circumstances.
A couple of young lads had been out driving in the countryside when they spotted a silver suitcase dumped in a ditch at the side of a remote lane. Naturally, they stopped to have a good look. It was very heavy, and when they noticed it was starting to ooze a pungent brown liquid, they very wisely decided not to open it and to phone the local constabulary instead.
The suitcase was bagged and tagged and taken to the mortuary unopened because the police had a sneaking suspicion as to what might be inside. Their fears were well founded. At the mortuary, the police and the pathologist unfastened the case and discovered the body of a virtually naked young woman, curled up in a fetal position with her hips and knees bent to squeeze them into the tight space. Her face and head were wrapped in plastic tape. Her visible facial features indicated that she was of Asian origin.
Her DNA and fingerprints were run through the various databases but there was no match, and nothing seemed to tally with anyone listed on the UK missing persons register. Decomposition was not overly advanced, and the pathologist had determined that she had been dead only a matter of weeks. The cause of death was most likely asphyxiation.
The stage at which a forensic anthropologist enters the picture is often once an initial postmortem examination has been completed, the ensuing police investigations have not resulted in any solid new leads and progress is starting to falter. At this point we may be asked to perform a second PM to establish whether there is more information to be extracted from the body, which is what happened in this case.
The first examination is usually a hive of activity but by the time you get to the second PM, there is less bustle. I prefer it that way: the atmosphere is calmer and there is less pressure on you to perform. You might have a police photographer turn up, or you might not. The pathologist probably won’t do more than pop in to say hello. It is usually just you and the mortuary technician. As a result, we get to form quite close working relationships w
ith the anatomical pathology technologists (APTs), or mort techs. One piece of advice we always give our students is that you will never go wrong if you turn up at the mortuary bearing gifts. Biscuits are good (I bring biscuits with me wherever I go), chocolates are better, but jam doughnuts open all doors and melt the iciest of hearts. Trust me, you always want the APT on your side, and they never forget a kindness.
The disruption left by a first PM can take a bit of getting used to. The scalp will have been peeled back to gain access to the skull and the skull cap sawn off to excise the brain. The cavity is then usually filled with cotton wool and the scalp drawn back into place and stitched. The torso will show a T- or Y-shaped sutured incision that runs horizontally across the collar bones and vertically down to the pubic region.
When this is unstitched, there is generally a plastic bag inside the body cavity containing the brain and viscera, removed previously for examination or sampling for further laboratory testing. There is little reason for the anthropologist to open the bag of viscera: our interest lies in the external form and internal skeleton. It is quite common to find the back and the upper and lower limbs still intact, unless there is a trauma or pathology that has attracted interest and attention to those areas of the body.
Radiographs or even CT scans of the whole body may have been taken before the first PM and these, along with photographs from both the mortuary and the scene, give us as complete a background picture to the second examination as we could hope for.
The body may have been stored in a freezer, in which case it is usually taken out the day before the forensic anthropologist’s postmortem to allow it to defrost. Mortuaries are not known for being cosy at the best of times, and working on a cold, wet, semi-defrosted body leaves your hands stinging with pain. This is where the doughnuts come into their own. The favour will be reciprocated with a hot cup of tea when you take a break and it is the best mug of warmth in the world.
What the police were keen to know from my PM on the young woman in the suitcase was how old she was and her ethnic origin. From the X-rays and my examination, I was able to establish that she was around twenty to twenty-five years of age when she died. Among other parts of the skeleton that provided me with this information were some small areas of bone around the edges of the breastbone (which we’ll look at in detail in Chapter 4), and developmental changes we could see in her pelvis and skull.
I believed, from my assessment of her face and skull, that her ethnic origin was likely to be in the region of Vietnam, Korea, Taiwan, Japan or China. I didn’t think she had the facial characteristics that would take her south into Malaysia or Indonesia. This deduction was based on the shape of her face, nose, eyes and teeth and on her hair type and colour. The suitcase was later established to have been manufactured in either South Korea or Lebanon.
However, none of this led us any further in terms of potential matches with missing persons and neither DNA nor fingerprint evidence was moving the case forward, either. We advised that the police should issue a black INTERPOL notice, the official international notification that a body has been found and not identified. The police had previously brought in a forensic artist, someone trained to produce a likeness of the human face that is accessible and palatable to the public—even when the real thing is showing signs of discolouration, decomposition and bloating, as this young lady’s was—with the intention of releasing an image to the media.
Unfortunately, on this occasion, the melding of realistic portraiture and artistic interpretation had not produced a harmonious result.
While the talent of the artist was not in question, the upshot was a faithful reproduction of the face as she actually saw it, minus the decomposition, obviously. Bear in mind that the dead woman’s face had been bound tightly by plastic tape and decomposition gases had caused bloating. The expanding face had therefore been constrained by the tension in the plastic. So although the final image was technically accurate, the impression it created was very odd. The victim’s lips were ballooned in the midline and the crenulation of the top lip, where it had been pushed against the teeth, gave it a scalloped appearance. It didn’t look like any mouth I had ever seen. My strong advice was that this drawing should not be released.
A more experienced artist would have made allowances for facial alteration due to decomposition. As it was, all people would see would be the postmortem consequences of the way the victim’s body had been treated and it was highly unlikely that this image would lead to her being identified. Indeed, it might well prove more of a hindrance than a help. Fortunately, the police agreed.
Thankfully, as it turned out, the image was not crucial to giving her back her name: INTERPOL were able to confirm that they held a yellow notice, the notification issued for a missing person, relating to a twenty-one-year-old student from South Korea whose disappearance had been reported by her university in France. A communication with the South Korean embassy, a transfer of fingerprints from her ID card and her identity was swiftly confirmed.
Jin Hyo Jung had been visiting the UK as a tourist and had rented a room in a London flat owned by a Korean man. In the flat the police found a roll of plastic “Gilbert and George” gift tape, belonging to his girlfriend, which matched the tape wound around the face of the young student. Probably only 850 rolls of this tape had been sold by Tate Gallery outlets in the UK, and this one had his blood on it. Jin Hyo Jung’s blood was also found in the flat and in the landlord’s car, and her bank account had been emptied.
Often it is only during a trial that the full story of a crime emerges, and so it was in this case, when the landlord, Kim Kyu Soo, appeared at the Old Bailey. I was not required to give evidence on this occasion as her identity had been confirmed. During the trial we learned that a few weeks after the discovery of the body, the Metropolitan Police had become aware of a second South Korean student who had gone missing. A joint investigation had duly been launched by North Yorkshire Police and the Met.
The second student was eventually discovered, bound and gagged with the same tape, boarded up in a wardrobe in another of Kim’s properties. He was found guilty of the murder of both women, theft from their bank accounts and perverting the course of justice by concealing their bodies. He received two life sentences.
I have used this case as a cautionary tale in teaching forensic art students about the importance of interpretation and of understanding the impact of the circumstances of a death on a facial depiction. When I show them a photo of Jin Hyo Jung alive alongside the postmortem drawing of her face, the response of over 90 per cent of students is that they would not consider them to be a match. They would therefore have rejected the possibility that this may be one and the same person.
I cannot say why the artist took the approach she did. Perhaps, as I supposed at the time, it was a lack of experience. Perhaps she was focusing on accuracy. Whatever the reason, it could have been a lead wasted, and it is a salutary reminder that even when a body is relatively fresh, we cannot rely solely on a likeness or facial recognition to bring us the information we need.
When a facial depiction is combined with a reconstruction in the hands of someone with experience and skill, the result can be eerily accurate. In a case from 2013, which readers of All That Remains may recall, it was a computer-generated image of the face of a murder victim, based on a CT scan of the skull, that led to the identification of a missing woman.
The discovery of human remains was made in a woodland clearing on Corstorphine Hill, on the outskirts of Edinburgh, by a ski instructor out cycling who had stopped for a breather. When he looked down at his feet, he thought he saw a dirt-covered face staring up at him from the ground. He recounted how he recoiled and had to take a second look, thinking that perhaps all he had seen was a root system that looked like a face, but he had been right the first time. He had stumbled upon a shallow grave that concealed the beheaded and dismembered body of a woman.
Analysis of the remains established her age, sex and height, blunt-for
ce trauma injuries and the constriction of her throat by means unknown. But a careless comment at the scene from a scientist with no specialist knowledge of anthropology initially sent the police on a wild-goose chase in their quest to identify her. The “non-anthro” expert suggested that the woman looked “eastern European,” perhaps Lithuanian, her cosmetic dentistry “vaguely Hungarian” and that she might be a migrant. This was a lesson in the inadvisability of relying on the uncorroborated hunches of armchair experts venturing outside their area of expertise. We have learned over time that it is best to say nothing at all until you are sure you cannot be overheard.
The assistance was then sought of more suitably qualified scientists from my department at Dundee University, who were involved in the analysis of tool marks associated with the dismemberment, in examining the remains for further information that might help to establish the victim’s identity and in the depiction of the face, undertaken by my colleague Professor Caroline Wilkinson.
Using the computer method, the muscles and soft tissue were layered one by one over a CT scan of the skull and the skin stretched over the anatomical framework. Using the age assigned to the woman by the team, and available hair as a guide to length and style, Caroline produced a striking representation for circulation in the media that felt incredibly real.
Some jewellery, including a Claddagh ring—the traditional Irish ring featuring hands clasping a crowned heart—found on the body suggested a possible Celtic origin, so the police were advised to ensure that the facial image was also distributed across Ireland. And it was indeed in Dublin, not Lithuania, that the victim’s family saw the reconstructed face on the news. The likeness was uncanny, and they contacted the Scottish police immediately.
The woman had been in Edinburgh visiting her son who, after DNA confirmed her identity, was arrested for her murder. The charge was reduced to culpable homicide on the basis of diminished responsibility. He was found guilty and given a nine-year custodial sentence for homicide and for the dismemberment and concealment of his mother’s remains. The judge and three psychiatrists did not uphold his psychiatric plea for leniency on the grounds that he suspected his mother of being a reptile and wanted to look inside her to see if she was impersonating a human. As to why he cut her head and limbs off, dug a hole and buried her, he offered no explanation. Doubtless his reasons were more banal. Having mutilated her body, he decided to use a suitcase to carry it to the disposal site. Most murderers find it easier to do this if they separate the body into smaller parts.
