Resurrecting the Dead: what did Jane Austen look like?
Linda Rodriguez McRobbie on how art, history and software help us make educated guesses about unknown faces.
Jane Austen is a pretty woman, with curly brown hair hidden under a rather dowdy mobcap, brown eyes, the long, straight, “Austen” nose, and a clear, rosy complexion. She looks friendly, clever, bright, like what you’d think Jane Austen would look like.
Except that we don’t really know what Jane Austen looked like. And we don’t know for sure that the waxwork described above and unveiled at the Jane Austen Centre in Bath, UK on July 9 is her. But it’s the best, most educated guess anyone has made so far – a guess informed by techniques honed catching rapists and murders, reconstructing the dead, and trying to locate the lost. The Jane Austen waxwork is nothing less than a fancy, fleshy composite sketch, the kind of portrait familiar to anyone who’s ever watched the nightly news.
It took a forensic artist to peel back the layers of history and memory, and uncover what Jane Austen probably looked like.
Surprisingly, there is no definitive portrait of the author whose genteel novels launched a thousand fan fiction writers: The only existing contemporary image of her was made by her sister, Cassandra, in 1817, the year Jane died. In it, she looks pinched and sickly. Members of Austen’s family didn’t remember their beloved (and famous) relative as this pallid harridan and declared the pencil and watercolor sketch a bad likeness. In 1869, more than 50 years after her death, her nephew commissioned artist James Andrews to paint a more flattering portrait based on the sketch and fading family recollection.
For more than 130 years, the Andrews portrait, which was sold at auction for £164,500 in December 2013, was considered by most Janeites to be the best image of the author. Most, but not all. In 2002, David Baldock, director of the Jane Austen Centre, a permanent historical exhibition devoted to all things Jane (and, by the looks of the gift shop, all things Colin Firth as Mr. Darcy, too), wanted to create a portrait that he hoped would capture the real Jane Austen.
“We were always asked what Jane Austen looked like and in a way the only reference point we had was Jane Austen’s sister, Cassandra’s, very poor watercolor,” he explained. “And we didn’t feel it did her justice.” Baldock wanted someone with proven skill in drawing real images out of recollection, of resurrecting the dead in pencil and paper.
He found Melissa Dring after reading about her speculative likeness of Antonio Vivaldi, the 18th century Venetian composer, for a production company making a film about his life; she’d been recommended to the production company by the Metropolitan Police Service, better known as Scotland Yard. Dring, a freelance forensic artist who frequently worked with Scotland Yard, brought with her more than a decade’s experience in composite sketching and forensic facial reconstruction, and even longer in fine portrait painting.
Dring was similarly unimpressed by Cassandra’s sketch – “It makes her look a little bit as if she’d been sucking lemons and it’s so totally unlike the feeling that you get from reading her books.” – but nevertheless used it as a starting place. She then spent a year consulting the many eyewitness accounts that described Jane; scouring portraits of members of Austen’s family for shared traits; even consulting a graphologist who examined Austen’s tight, cramped hand, and highlighted the writer’s private, secretive nature, her practicality, and her right-handedness. Drawing on all the available information – which, she said, was more than she normally has to go on – Dring created a composite portrait of younger Jane Austen that she felt captured the author’s physical appearance as well as her character.
“I have to admit I was quite pleased with it. What was really nice when I had letters from people, people saying that’s my Jane Austen. They recognized her,” she said, noting that Austen fans can be fairly “possessive” of their patron writer.
But though the portrait was fairly successful – it certainly generated national interest in the Centre – Baldock felt there was still more to do. So in 2011, and with a bit more money to spend, the Jane Austen Centre commissioned sculptor Mark Richards, onetime senior sculptor at Madame Tussauds, to work with Dring to recreate her portrait in wax. (Baldock refused to disclose how much the waxwork cost, but did say that the Centre would be insuring the one-of-a-kind piece for £1 million.) It took nearly three years to finish the waxwork; finally, Jane was dressed in authentic period costume by BAFTA and Emmy Award-winning designer Andrea Galer, and the finishing touches applied by hair and color artist Nell Clarke, also formerly of Madame Tussaud’s. She now stands in pride of place at the Centre.
“Everybody really loves it, and I’m very, very confident when someone asks me, ‘How do you know what she looks like?’ We’ve put in the work,” said Baldock. “I don’t think anyone can say it’s not like Jane Austen ... This is as good as we can get.”
Jane Austen is hardly the first dead celebrity to be resurrected by forensic artists; it’s a practice that has become more popular in the last 15 years. Production companies working for edutainment outlets like the Discovery Channel and the History Channel routinely tap forensic artists to create likenesses of dead famous people for programs. So too do groups with a vested interest in the branded image of a famous person; for example, in 2005, the Mount Vernon Estate, then building a new museum and visitor centre dedicated to the life of George Washington, were frustrated that the only existing images of the first American president were created when he was a lumpen over-40 suffering from tooth rot and jawbone decay. They contracted a team of experts, including age regression forensic artists and a forensic anthropologist, to create accurate full-length figures of the man at 19, 45 and 57, in an effort to portray the founding father as the strapping action man he really was.
In February 2013, a team of forensic artists from Dundee University unveiled its lifelike bust of much-maligned king Richard III, whose bones had been lately discovered under a Leicester parking lot. The bust, which was commissioned by the Richard III Society, a group dedicated to rehabilitating the king’s tarnished reputation, was based on CT scans of his skull and created using computer imaging. How Richard III actually looked is hugely important to his legacy – the Wars of the Roses king died at the Battle of Bosworth on August 22, 1485, poleaxed in the back of the head and posthumously stabbed in the buttocks. He was later resurrected as the murderous humpbacked villain in Shakespeare’s history play. That he was meant to look the part of a villain was a major set piece of the play; that his bust reveals that he probably looked fairly average may go some way in reclaiming his reputation.
But the work that goes into resurrecting a long dead king or president or author, for whatever reason, is no more astounding than the quotidian work of being a forensic artist. It’s also probably a bit easier and almost definitely a lot cheerier.
A quick primer on the grim art of forensic imagery
Forensic art is essentially any visual aid used during criminal proceedings. There are four major areas of concentration: Composite sketches, those drawings of suspects that are usually created from a combination of eyewitness accounts, or sometimes available photographic evidence, such as blurry CCTV footage; age progression or regression, the kinds of images used to find long-lost children; demonstrative evidence, used in court as trial displays; and reconstructed likenesses of the nameless dead for identification. It’s also not a terribly new thing; Karen Taylor, a veteran forensic artist, noted in her seminal 2001 manual on forensic art, Forensic Art and Illustration, that all major forms of forensic art were in use before 1915.
The growing use of forensic artistry throughout the 19th century coincided with the increasing regularity of police forces and policing practices over the same period. Which makes sense: Devoted police forces, paid and maintained by the government, as opposed to night watchmen or private security forces, have the time and the resources to actively investigate crime. Artists’ skills were instrumental in advancing forensic medicine, drawing cadavers in various state of decomposition, for example, or, such as in the case of the investigations into Jack the Ripper’s grisly 1888 killings, post-mortem drawings of the victims’ bodies illustrating the extent of their wounds.
Taylor also pointed out that culturally, people in the late 19th and early 20th centuries had a passion for cataloging, measuring, and classifying, efforts to order the world rationally and make it more manageable; consider the curious obsession with spurious fields like phrenology, the claim that formations of the skull corresponded to character traits. This interest in cataloging the known world had the side effect of entrenching forensic art in policing. In the 1880s, French anthropologist and Parisian police records clerk Alphonse Bertillon designed a system of identification based on measurements of parts of the face and body, inventorying physical characteristics, such as tattoos or scars, and, perhaps most importantly, standardized full face and profile photographs of incoming prisoners. Bertillon’s system was intended as a way to identify and track prisoners, and it revolutionized the way police departments kept records. Though it flourished for 20 to 30 years across the globe – after Bertillon presented his system at the 1893 Chicago World’s Fair, it became popular in America – it was replaced by the less labor-intensive fingerprint classification.
But even as prisoners were no longer prodded with calipers, elements of Bertillon’s system lived on. Firstly, the standardized photographs taken of incoming prisoners are better known now as mug shots, and the inventorying of identifying characteristics is still standard procedure. More importantly to forensic art, Bertillon’s system also involved the detailed comparison of individual faces and facial features; as such, it became the basis for 20th century composite kits, facial catalogues, and computer-generated recall systems. More than 50 years later, law enforcement was using a direct descendent of Bertillon’s system, an assemblage system of composite sketch production called Identi-KIT, a product patented by gun manufacturer Smith & Wesson. Identi-KIT used an index of mix-and-matchable facial features that allowed the witness, under direction, to piece together a face that closely matched the suspect. Several similar systems followed, including Britain’s PHOTO-FIT in the 1970s; each relied on the idea that witnesses could recognize better than they could recall.
By this time, forensic art, in its various forms, had long been a major weapon in the policing arsenal; in 1938, the FBI had established a dedicated forensic art and graphics unit, then called the Cartographic Section. At the same time, advancements were continually being made in the discipline, from age progression techniques, 3d sculptural modeling, and facial reconstruction from human remains, the last driven in part by the tragic post-war need to identify concentration camp victims. In the 1970s, forensic art, specifically composite art, went through what Taylor described as a “renaissance” after a series of high profile cases were solved using hand-drawn composite sketches.
However, several cases of ‘70s also revealed some of the weaknesses of composite sketches, weaknesses that were more and more evident the higher forensic art’s profile became. For example, multiple composite drawings of serial rapist and murderer Ted Bundy produced wholly different impressions, due to Bundy’s ability to change his outward appearance and to the unreliability of human memory. In the case of the “Hillside Strangler”, a series of rapes and murders of young women in California between 1977 and 1978, attempts to create composite sketches of the suspect were confounded by the fact that the killer actually turned out to be two men.
Stephen Mancusi spent 24 years on the NYPD force as a detective and forensic artist, joining the force in 1984. Forensic art, he said, is only as good as the information it has to go on: “Sometimes the information given to us is really great information and we can come up with a really strong image, other times the information is really weak.”
So though forensic art is a hugely important tool, it’s also a weak one in some ways. “It can never be a hard piece of evidence, like fingerprints or DNA. It’s actually weaker than a one-witness identification case, and that’s a very weak case … Witness identification is not a strong, strong thing, and sketches, we’re a notch lower,” Mancusi continued. (That’s fairly low: Analysis of cases of DNA exoneration since 1992 show that the “vast majority” of them involved mistaken eyewitness testimony, according to a 2006 article from Psychological Science in the Public Interest.)
Even so, in the 1980s, forensic art grew in stature and importance, first with the establishment of the FBI’s police artist training program in 1984, and then in 1986, with its inclusion as a discipline under the auspices of the International Association for Identification, the world’s largest and oldest forensic professional organization. In 1987, the hugely popular America’s Most Wanted premiered, offering forensic art a nationally televised platform; by the end of its 23-season run on Fox in 2011, it had led to the capture of more than 1,100 fugitives. (The show was created by its host, John Walsh, for a very personal and horrifying reason – the 1981 abduction and murder of his 6-year-old son, Adam.) By the close of the 1980s, for better or worse, regardless of its success or reliability, forensic art was a deeply entrenched part of the American perception of crime and justice.
The most significant change to forensic art in the last 20 years has been, perhaps unsurprisingly, the increased use of computer software programs in forensic art, in all areas. Some of these programs allow regular police officers to guide witnesses and victims through creating a composite sketch; others allow authorities to digitally age progress a subject in a photograph; and still others use complicated modeling algorithms to reconstruct facial likenesses from bare bones.
But the increasing use of computer generated forensic art means that human forensic artists are on the decline. In December 2013, the Washington Post reported that there were only around 100 full-time forensic artists working in America, a decrease largely due to budget cuts – it’s generally cheaper to use a computer than a trained human. Indeed, despite the growing call for historical simulated likenesses, Mancusi cautioned that forensic art is not a particularly lucrative or open field. “When I talk to students, let’s say, someone asks, ‘Hey, is forensic art a viable career path?’ I say, ‘No, it’s a viable career path,’” he laughed.
However, computer-aided composite creation has been criticized for being ineffective and, in some cases, producing simply ridiculous results (like this e-fit image of a Hampshire, UK burglary suspect with what appears to be a cabbage on his head from September 2010). In 2011, Gary Wells and Lisa Hasels of Iowa State University examined several studies of the effectiveness of mechanized facial composite assemblage systems and found that overall, these systems tend to do a very poor job creating a fitting likeness of the intended face – one study they cited found that only 2.8 percent of participants correctly identified a well-known celebrity based on a likeness created by composite software. That, said Wells and Hasels, indicates that people simply don’t remember faces by their individual components. In an article published in the February issue of Current Directions in Psychological Science, they explained, “Numerous lines of evidence converge on the view that faces are generally processed, stored and retrieved at a holistic level rather than at the level of individual facial features.”
Some manufacturers recognize that: Evo-Fit is a new facial composite construction system that works on this holistic level; its makers claim that it has a 60 percent recognition rate, making it far more reliable than the 5 percent recognition rate they cite for other composite generation systems. Unlike other systems, Evo-Fit presents witnesses with dozens of randomly chosen faces, asking them to choose the six that most resemble the suspect. From there, dozens more are generated, each time asking the witness to choose the best six; the composite “evolves” over time to ultimately resemble the witness’s best recollection. The technology has already been used to solve several crimes: In September 2011, a serial rapist was caught and convicted in Manchester, UK, after his startlingly accurate Evo-Fit image was recognized.
But even as technology continues to improve and offer more realistic images, as the officers quoted in the Washington Post piece noted, a trained forensic artist can bring something that no computer software program could: Sensitivity, imagination, and intuition.
“Technology and machinery is cold,” Wayne Promisel, a detective at the Loudoun County Sheriff’s Office told the Washington Post. “It is also missing the ability to ask the questions in a certain way in an interview while having a sense of compassion.”
Mancusi agreed. When interviewing a victim or a witness, he said, the experience often turns on the language he or she is using and trying to figure out what that means. “Every descriptive word you get from the victim, what you’re trying to do is try to find a solution to those words. If she says he was ‘mean looking’, you’re trying to find a solution, a strategy,” he said. The interview is itself an “investigative duty”, but it’s a delicate one that requires both compassion and control.
“There’s a sensitivity to the session when you’re drawing with pencil… it’s really the art of interaction,” he said. “You can’t just pick out eyes and a nose and a mouth and slap it together on the computer.”
Making artistic magic
Melissa Dring has, by now, mastered the art of interaction, having sat at “a lot of bedsides” in her career as a freelance forensic artist.
Dring, who lives in Northampton, UK, fell into police work when the local constabulary called up the art school she taught at in need of a portrait artist who could draw freehand. She was thrown into the deep end on her first case: A woman was raped at knifepoint by a man who broke into her home; it was dark, but she caught glimpses of her attacker when he paused to light his cigarette as he stood at the end of her bed. That first sketch led to others and, about a year later, in 1988, Dring spent two weeks at Quantico, completing the FBI’s Composite Art training program.
She now regularly works with law enforcement across England, creating composite sketches of kidnappers, murderers, and rapists from other people’s memories, and reconstructing lifelike images from nearly destroyed bodies (leaving out, she says, “all the horrific bits”). She spends a minimum of two hours with a witness, often longer, pouring over her FBI facial catalog, which shows real faces with some features blurred out. “It takes a huge amount of concentration and focus from both myself and the person who’s trying to remember the person. It’s a process of very gradually building this image and fine tuning it until the person says, ‘That’s it, that’s him, I can’t bear to look at it,’” she explained. “I can’t tell you how many people who have burst into tears or gone suddenly very red in the face or bleach white, or you see beads of sweat burst out on their forehead. One poor girl had to run off and be very violently sick… I always say, ‘I’m quite glad to see you cry because it means that the police can put quite a bit of faith in this image.’”
Dring is good at what she does – once, a rapist who recognized himself from a composite sketch she’d done actually turned himself in. “He walked into the local police station and they more or less fell over in surprise because he said, ‘I think you’re looking for me,’” she said. “It’s very satisfying, I was pretty chuffed about that.”
Dring’s stories, delivered in a cheery, even jolly British accent, are chilling, like the one about the “poor chap” whose body was found burnt beyond recognition by the side of the M45 in 2008. She spent an afternoon with him in the morgue, trying to use what was left of his face to come up with an impression of what he might have looked like (the victim was later identified as 60-year-old Sher Khan; his flatmate was found guilty of his murder in 2009). Recreating Jane Austen, who was neither suspected of a heinous crime nor missing, nor being put back together from grisly remains, was far less harrowing. But Dring’s work on Jane, like her work on the “poor chap” by the M45, highlights something vital about forensic art: The necessary marriage of fact and speculation, tempered by experience.
“When forensic science falls short, they call in the artist and that sums it up. When the sciences can’t do it, you call in the artist and we work a little artistic magic,” Mancusi said. A forensic artist, he says, doesn’t have to get it perfectly right – just close enough to shake out some new clues, jostle some memories. “We live in the world of almost, and it’s a great place to be. To do your job, you only have to succeed almost.”
We will never know if the wax Jane Austen standing in the Jane Austen Centre in Bath really looks like the Regency novelist. But all things considered, she’s probably pretty damned close.
David Robinson used the data from the 28,657 people who self-selected to take the Stack Overflow survey to investigate the relationship between programmer pay and the conventions of using either tabs or spaces to mark indents, and found a persistent, significant correlation between using spaces and bringing home higher pay.
It’s the end of an era, sort of: Fraunhofer IIS, the developers of the MP3 audio compression format, announced that they are ceasing their licensing program. In a blog post, spokesman Matthias Rose says that it’s had a good 20-year run and is obsolete. But it’s also true that the decoding patents expired last year, […]
Freddy deBoer writes that he’s been telling the same joke for years about Silicon Valley’s only product, which might be universalized as “At last, a way to verb with nouns on the internet!” But the social-media techopoly is stable, now, and so the venture capitalists have moved on to the three terrible trends that will […]