Crime Library: Criminal Minds and Methods

Forensic Art

The Reconstruction of a Face

Karen Taylor, Forensic Artist
Karen Taylor, Forensic Artist (Karen T. Taylor)

For over eighteen years, Karen T. Taylor, author of the definitive Forensic Art and Illustration, was a jack-of-all-trades for the Texas Department of Public Safety, at least where art was concerned.  Previously, she'd spent three years in London where she did freelance portrait sculpture for Madame Tussaud's Wax Museum, but in 1981 she decided to find a more secure job.  Her first assignment for the Texas DPS was to illustrate posters for road safety, but that soon evolved into other projects.

"A Texas Ranger approached our office and said that he'd heard a lady here could really draw faces.  I said, 'I guess you're talking about me.'  He asked if I could draw a face from a description and I said I'd be willing to try." 

She interviewed an eight-year-old girl who'd witnessed a hit-and-run incident that had taken the life of her younger cousin.  The girl had spotted the driver and was able to describe what she'd seen.  Taylor worked with her to make a composite drawing, and this proved to be her first success story with this form of forensic art.  From her drawing, the Rangers identified and apprehended the man.

As Taylor took on more cases, she noticed that interviewing witnesses, especially those who had suffered some trauma, required a special technique.  She needed to learn how that kind of memory was processed and develop ways to delicately enhance recall. 

"The work that forensic artists do is complex," she says, "and it's easy to underestimate, particularly composite drawing.  Interviewing the victim of a traumatic event and trying to retrieve the memory that's associated with the trauma is quite difficult.  It must be done with great care.  You have to know how to handle a person who's been through something like that, so artists need to be trained in how to do this.  A face gets encoded and stored in someone's mind, and ideally we perform the retrieval in a way that corresponds to the encoding process."

Composite drawing and photo
Composite drawing and photo (Karen T. Taylor)

First, she gathers information.  Before talking with the witness, Taylor wants to know from the police if he or she was a victim rather than just an observer, and if so, the nature and degree of the trauma.  Also, was the suspect a member of the witness' own racial group (which can make a difference in how well they discern certain features)?  And is the witness apprehensive about making an  identification?  In short, each interview is situation-specific, and Taylor carefully evaluates the types of questions she will ask the witness for their balance of intrusiveness and effectiveness.

First, she uses a cognitive style of interviewing which allows the witnesses to describe the event in their own words, and then with carefully crafted questions, she gains additional information, such as:

  • How long did the witness see the subject?
  • What were the lighting conditions?
  • How far away was the witness from the subject or incident?
  • How long ago did the witness observe the incident or subject?
  • What was the witness's vantage point?
  • Were there any obstacles to the witness's view?

Piece by piece, she draws a composite.

While there are software programs that purport to accurately duplicate this process, eliminating the need to draw, Taylor is skeptical.  "It doesn't work well," she states,

 "when software is designed to pull up a whole screen of pairs of eyes, noses or mouths to choose from.  People don't relate to that.  They need to see the parts in context.  One thing that's often underestimated is the importance of proportion.  I know of a high-profile case in which a computer program for facial composites was used, and it was far off from the suspect's actual appearance, because the system's user had no knowledge of facial anatomy and the spatial arrangement of facial features.  I refer to this as the gestalt of the face or the holistic grasp of the features.  That's what gets encoded about a face in our minds and that's what we want to retrieve to document that likeness.  We have to ask the right questions to get the right proportions."

It wasn't long before Taylor took on another artistic task: drawings of unidentified corpses.

"Some officers asked if I could assist them with identifying homicide victims from morgue photos," she recalls, "so I began doing postmortem drawings.  We never had any want for cases, and there are many reasons why these people need to be identified.  You can't easily commence a death investigation without first knowing the victim's identity."

From photos of the deceased (or from a viewing), the artist draws a reasonable likeness, which includes identifying marks such as scars, tattoos or moles.  The primary purpose is to put out a facial image to the media, but it may also provide something for people to look at who have reported a missing person, in cases where the body itself is badly injured or decomposed.

Post-mortem drawing compared to photo
Post-mortem drawing compared to photo (Karen T. Taylor)

To acquire skill at this, Taylor did plenty of research.

"I've extensively explored many books on forensic pathology," she says, "and I've talked with pathologists to try to find specific subtle things that happen on faces after death.  People are found in all different conditions, from soon after a homicide to semi-skeletal to fully skeletal, depending on things like time of year and the type of climate they're in.  Then there are other taphonomic factors to take into account, like insect and animal activity.  The best candidates for postmortem drawing have photographs available. I've occasionally gone to the morgue and viewed a body directly, but it has to have enough details intact for a sketch.  Sometimes you simply cannot accept a case from morgue photos alone, because there's too much damage to the face."

The types of cases in which these drawings are used include:

  • Suicides and homicides
  • Trauma fatalities
  • Drowning or boating accidents

Since the actual morgue photos are inappropriate for media distribution, drawings help to bridge the gap, allowing publication of a good likeness of the deceased as he or she might have been in life. If someone sees the drawing and feels that there's a sufficiently close resemblance, then the missing person's medical records and dental records can be compared to the dead person, along with fingerprints and/or DNA.

One of the advantages of drawings over morgue photos is that the artist can add facial expression.  It's often the case that the deceased lacks similarity to the living person because the face is missing its characteristic vitality, but adding "the look of life" is complicated.

"Postmortem drawings pose a particular challenge," Taylor explains, "in that so much of what we recognize in the faces of people we know in life is associated not only with the arrangement of the facial features themselves but with the animation of the face.  The real challenge is to reanimate the face.  It's perhaps, for me, the most difficult type of forensic art." 

Yet forensic art is more than just drawing.  The next step for Taylor was to learn to do three-dimensional sculpture from skulls, and for that, she took a course with Betty Pat. Gatliff in 1982. 

For eight years, Gatliff worked as a technical illustrator for the navy, and for twenty years as a medical illustrator for the Federal Aviation Administration.  She did artwork for anthropologist, Dr. Clyde Snow, and through that connection, she eventually became the country's foremost forensic sculptor.  Her first attempt at three-dimensional facial reconstruction had been in 1967, working with Snow on the case of an unidentified Native American man. Although she struggled with the piece, it was good enough to lead to a positive identification and the case was solved.  Over the next three decades, Gatliff went on to work over two hundred cases for law enforcement, including creating sculptures based on the remains of nine of the thirty-three victims of serial killer John Wayne Gacy.  From her work, one victim was identified as a young man from Nebraska---he was the first one Gacy killed---and five more were tentatively identified.

Successful identification using 3-D facial reconstruction by Betty Pat. Gatliff
Successful identification using 3-D facial reconstruction by Betty Pat. Gatliff (Karen T. Taylor)

"That case was a big job," she recalls.  "I did seven of them here [in her studio], which took about six weeks, and then I went to Chicago and did the last two in the medical examiner's office so the press could cover it to stir some interest."

Her first step in working with these skulls was to establish the facial contours based on the combined thickness of the muscle, fatty and connective tissue, and skin.  For that, she used tissue-depth tables developed by anthropologists.  Once she knew how far out from the skull she needed to build the tissues, she glued the right length of eraser-type markers at various strategic points (morphological landmarks).  Then she applied modeling clay to those depths.

There are various formulae for the development of the individual features.  For example, with no soft tissue or cartilage left, she has to rely on certain measurements for the nose.

"For the width of the nose," she says, "we measure across the nasal aperture, which is the hole where the nose was.  That gives us the total width, and we increase it by ten millimeters, which would be five on either side.  The projection is calculated by measuring the nasal spine, which is the sharp bone right under the nose at the septum.  Then we multiply that by three.  Then we make a measurement to project the length of the nose, so if the measurement is 8, the projection is 24.  So we use two measurements, the width and projection, and when you connect it to the bridge, that establishes the nose.  That's the way I've done it for thirty-five years, and it works."

Then she inserted prosthetic eyes.  If any hair had been found with the body, she used a wig of the appropriate color.  Finally, her reconstructions were ready for the photographer.

Successful identification using 3-D facial reconstruction by Betty Pat. Gatliff
Successful identification using 3-D facial reconstruction by Betty Pat. Gatliff (Karent T. Taylor)

Gatliff went on to develop workshops to teach these methods to other forensic artists.  As Taylor learned various techniques, she started running workshops, too. 

"To my mind," Taylor says, "facial reconstruction on the skull, is a two-phase process.  The first half is technical in nature, and the second is artistic.  I can teach things about the various muscles as they lie over the skull, and the soft tissue correspondences, but so much of what is incorporated into the process is artistic instinct based on experience from drawing thousands of faces.  There are certain things that tend to go together about a face.  So I think of it as art/science, and ideally you'll have an artist working hand-in-hand with a scientist, like a pathologist, odontologist or anthropologist." 

For three years, Taylor worked with three-dimensional sculpture, but soon pioneered a new technique.

"I'm credited with the two-dimensional method of putting a face on a skull," she says.  "What I did was to incorporate the available tissue depth data and the various formulas for the development of facial features.  In the past, several people had tried taking the photo of a skull and putting tracing paper over it to develop a face, but because I learned Betty's method, I did it differently.  It was a simple concept, yet added so much to the accuracy of the process.  I began doing drawings over skulls.  I applied all the rubber markers to the skull as if I were going to do a three-dimensional reconstruction, and then photographed it.  I developed prints of the frontal and lateral views that were life-size and corresponded one-to-one with the skull.  I then did life-size drawings based on photos of the skull.  By doing it actual size, the same information applies as in the three-dimensional method."

Two-dimensional facial reconstruction
Two-dimensional facial reconstruction (Karen T. Taylor)

When this method proved successful with identifications, Taylor was invited to present it at an international conference at the FBI in 1990.  She then went to Quantico on a semi-annual basis to teach classes, because while both two and three-dimensional methods work equally well, there are certain cases that lend themselves to the former. "Cases in which the skull is very fragile and can't hold the clay would do better with two-dimensional reconstruction."

Successful identification using 2-D facial reconstruction
Successful identification using 2-D facial reconstruction

In addition to drawings and reconstructions, Taylor also does computer-enhanced imaging with age progression programs for missing children and for updating fugitive photos.  In her opinion, this is the area where computers have made a big contribution. "Computer programs are best for cases of kids and fugitives, in which good original photographs exist.  Where good quality photos don't exist, the old fashioned method of drawing comes back into play."  Yet even with computer sophistication, to get accurate results the artist must still understand both the subject and the concepts of facial anatomy.

Taylor has been invited to work on several cases for America's Most Wanted, and one was particularly memorable: a fugitive update of Virgilio Paz Romero, a conspirator in the assassinations of the Chilean ambassador to the United States and his aide. Taylor was given only some bad photocopies of old photos, so she went to work researching the man.  From what she could tell, he'd have watched his weight and would still be a fairly flashy person.  "So I drew him with a red shirt," she recalls. 

Her image was aired on the show, and within three days, Paz was in custody.  Affirming Taylor's instinct, he was wearing a red shirt.

These days, she has resigned from the DPS to concentrate on other types of art.  She takes occasional freelance forensic assignments through her company, Facial Images, but she's looking forward to doing more portraits and sculptures "of the good guys."  Her next book, Understanding the Human Face, uses information from what she's learned as a forensic artist to teach fine artists the nuances of head and neck anatomy, and attaining likeness in portraiture.  It's likely that her book on forensic art will become the standard text in the field.

Another task that a forensic artist might accept is to work with the photo of someone who's been missing for a long time.  That's called age progression, which may involve either the developmental stages of facial-cranial growth in a child or the adult aging process. So, let's take a look at how that's done.

We're Following
Slender Man stabbing, Waukesha, Wisconsin
Gilberto Valle 'Cannibal Cop'