by Mike Pitts
Which is part of their eternal draw. Most of the time archaeologists study anonymous people from the past – faceless people. When Jo Appleby was excavating Skeleton I, part of her saw a man, an individual with a name and an identity, a man with a face. Yet the other, perhaps the bigger part, saw a skeleton, much like the many other remains she had worked with, once a living being but now a relic, which would tell her about aspects of life and death around the time the man had died, but not the intimate details of his unique personality. He had no face.
But what if we could faithfully reconstruct it? There would be historical value in so doing. Critics have questioned the likeness in the many portraits of Richard III, so – on the face of it, we might say – comparing Skeleton I’s appearance with the paintings would not help us assess his identity. Perhaps, however, he really is Richard (it’s certainly looking promising), in which case we could then assess the portraits, a distinctly unusual opportunity in historic representation. Yet who among us would stop at historical value? What would we see in the real face of Richard III? Might we do better than Hastings – and perhaps Shakespeare, and generations of leading actors – and know his heart?
A human skull, says Caroline Wilkinson, is as unique as a human face. We are programmed to read only faces, however, so if we wish to recognize an individual from their skull – of obvious use in forensic investigation, from murder cases to war crimes and natural disasters – we need to put a face on the bones. The principle behind this was established by German scientists in the late 19th century: after measuring soft tissue thickness at fixed points on heads in the laboratory, average statistics could be applied to bare skulls to model flesh with clay. In this way were revealed the apparent faces of J. S. Bach, Dante, Schiller and Raphael – and that of a prehistoric woman.
The idea caught on, but soon lost respectability after different artists working with the same skulls produced entirely different faces, reflecting the then novelty of the technique, but also the inextricable mix of science and artistry behind a complete facial reconstruction. Research continued, however, and craniofacial identification is now a proven, established international practice. Prominent techniques include the Russian method, in which muscles are modelled on to a skull, and the American method, in which soft-tissue thickness is the exclusive guide. Richard Neave, a medical artist at the University of Manchester, pioneered what became known as the Manchester method, which combines the two previous techniques – fixing pegs around the skull cut to the depth of flesh, and then building up the face by layering on muscles and other soft parts. It is a painstaking process that has been revolutionized by digital technology, both in refining facial understanding (particularly important has been clinical imaging on living people) and in creating the faces.14
Wilkinson began her career on Neave’s team in Manchester, and her distinctive short blond hair and strong jaw became familiar to many through her appearances on archaeological television series such as Meet the Ancestors. She is now an award-winning anthropologist and artist and Professor of Craniofacial Identification at the University of Dundee. Skeleton I’s skull could not have been in better hands.
She and Chris Rynn, a Dundee colleague, worked with high-resolution photos and the CT data from Leicester. Because the university was hoping to resolve the skeleton’s identity (or not) early in the new year, just four or five months after the bones had been found, she began immediately with the first hospital scans, so that the skull revolving on her computer bore curious flecks of digital dirt; she had to correctly assemble on screen the different skull parts, and replace a missing front tooth. Later she would confirm some of the hidden details with the micro-CT scans.15
Her initial analysis was good news for Jo: the skull was undoubtedly male, with a wide palate, square jaw and prominent chin, although with gracile elements such as a more feminine brow. Drawing on an extensive database of soft tissue thicknesses, muscles and facial features, she worked around the bones, adding eyeballs and digital pegs, and layering up the flesh with the aid of a haptic arm, a mechanism that transferred the movements of her right arm and hand into on-screen directions – allowing her to throw digital balls of clay and model them on to the skull, feeling her way as an experienced anatomical artist, a 21st-century George Stubbs of human portraiture.16
The creation of every part had its basis in scientific observation, as she chose average statistics for a contemporary white European male aged 30 to 40, selecting the thinner end of the range to reflect a physique likely to be more appropriate to the 1480s than the 1980s. She guided average eyeballs into their correct places in the sockets. She determined eyebrow shape from the character of the eyes, the brow and the upper orbital margin – Skeleton I’s eyebrows were low and straight – and crafted the mouth from key details of the anatomy. As she worked round the skull bit by bit, a flayed, muscled face emerged, on which she could build up virtual skin as she would have done with a clay head.
The chin was very prominent, protruding slightly beyond the mouth, whose corners were turned down a little. The lips were quite thin, the lower a touch more prominent than the upper, with a wide philtrum rising up to the nose. Clinical imaging has particularly improved understanding of noses, which used to be a weak area in this field: ‘Noses are our best feature now,’ said Wilkinson. Skeleton I’s – as related by exquisite details of bone structure – had a relatively sharp tip, a straight base and a subtle rise along the ridge. Ears, however, remain difficult. All that can be said about Skeleton I’s is that they were quite large, and like most people’s, they had lobes. Using standard imagery, Wilkinson then added a neck (adjusted for width as determined by features on the skull), and shoulders, the right one a little higher to match the rest of the skeleton.
It looked like Richard III. At least, it looked uncannily like the portraits, which she remembered seeing as a child, but only now had returned to for a proper viewing. At this point, no one knew how the DNA research was progressing, and she had no idea whether or not Skeleton I’s identity would be confirmed. A little rattled, and, she told me, knowing that she would have to justify her results to a critical audience, she checked the procedures twice. She was happy that the face she had built was ‘a scientifically and anatomically accurate’ outcome of analysing the skull.
The bald, expressionless bust she had on her screen should closely resemble Skeleton I’s face. Wilkinson had conducted research at Dundee in which she compared reconstructed faces to known originals in blind tests, proving the success and objectivity of the process; on average, 70 per cent of the surface of a recreated face was less than 2 mm out. Yet this was not the end of the story. There is more to a face than shape, and after shape, in Wilkinson’s words, ‘we’re guessing’.
Type of hair and eyebrows; hair, eye and skin colour; facial hair, skin texture, blemishes and wrinkles; and other superficial details so important to individual identity, including ‘face fat’, are not determined by bone structure, but by a mix of inheritance, lifestyle and cultural choice. In forensics, where Wilkinson does most of her work, recognition is key. Research has shown that a bare head, though correct as far as it goes, is not always recognized – her own 12-year-old daughter did not know the unadorned shape of her mother’s face. So adding humanizing details can help.
With nothing to go on, as often happens in forensic work and is clearly the case with Skeleton I, the best that can be done is to draw on banks of average statistics for the appropriate ethnicity, age and gender. People, though, are unique, and average eyebrows, skin texture and so on can also make a reconstructed face unrecognizable to people who knew the subject well. Between the bust and the person lies a critical gap that can mostly be filled only by art and subjectivity.
Wilkinson presented Janice Aitken, a digital artist at Dundee University, with a life-sized 3D printout of Skeleton I’s head, made in Cardiff at PDR, the National Centre for Product Design and Development Research. It arrived on Aitken’s desk with prosthetic eyes, and – in
a chance nod to the skeleton’s provenance – sprayed grey with a can of car paint. They knew the skull dated from around 1500, give or take 50 years. Aitken would study relevant contemporary portraits for dress and styling, including those of Richard III, and craft an individual on the plain, acrylic plastic face.
In her forensic work, Wilkinson will sometimes superimpose an image of a skull on a portrait photo, rotating it to the correct angle, to see if it’s even possible that the two could be the same person. Now she thought she would try the same approach with Skeleton I. She selected three portraits of Richard III. One of them, a fine example in the National Portrait Gallery and very similar to one in the Royal Collection – thought to be one of the oldest to survive – gave an astonishing match.17 ‘It’s incredible how well the skull fits with this portrait’, she said, adding that even in modern cases she very rarely sees such a good fit. It was enough to convince her that Skeleton I was Richard III.
Which was odd, because historians had been saying that the portraits didn’t look like him.
Why might they not? Could it be that at that time portraits were more conventional than real? Or perhaps there was a more sinister reason? Certainly, by the time Richard came to the throne in 1483, there was a growing tradition of European painting which aimed to capture personality as well as physical realism. The Dutch artist Jan van Eyck, who died ten years before Richard was born, had created some stunning portraits; among well-known paintings by Leonardo da Vinci, his Portrait of a Young Man dates from a year or two after Bosworth, and the Mona Lisa from the early 1500s – but for Bosworth, Richard III might have been in his 50s as it was being painted. We have a right to seek the real king in early portraits.
There are plenty of them, too, approaching 20, largely thanks to the popularity of historical series of royal portraits to hang in great halls. Unfortunately none are by a known artist, nor is there proof that any were painted when the king was alive. With one exception (a piece now in the collection of Sir Tim Rice, which is a little fuller) all show just head and shoulders, with arms across the chest. The three oldest are different from each other: one (apparently painted soon after 1510 and in the collection of the London Society of Antiquaries) faces left, another (in the Royal Collection, apparently painted around 1500–20) faces right, and the third (also owned by the Society of Antiquaries and with attributed dates ranging from soon after 1523 to after 1550) faces left and holds a broken sword. In all others he holds a ring on his smallest right finger in his left finger and thumb, and the remaining dozen probably painted before 1620 copy this and other details from the Royal portrait, or a similar missing early version.
As Caroline Wilkinson found, there is a striking similarity between the faces in these portraits and the one she built from the skull of Skeleton I. Most have the thin mouth with downturned ends, prominent chin, straight nose and relatively thin eyes below the straight, smooth brow of the model. Indeed, a small line drawing in the Beauchamp pageant, an illustrated biography thought to have been made for the Earl of Warwick’s daughter Anne around 1485 when Richard was still alive, shows the same features.18 There is a strong case for saying that all these portraits show us the real Skeleton I.
Richard III in an early portrait (left; National Portrait Gallery, London) with the face on Skeleton I’s skull built by Caroline Wilkinson and interpreted by artist Janice Aitken (right; Mike Pitts).
They also show a raised shoulder. This is on the king’s right in all but the two Society of Antiquaries’ paintings – posing the question whether these might be inverted copies (the Beauchamp sketch, perhaps diplomatically, hides shoulders under a cloak). So the portraits show not just the face of Skeleton I, but the raised right shoulder too.
These asymmetric shoulders have been much discussed. X-ray images revealed that in two of the earliest paintings (the one with the sword, examined in the 1954 and again in 2007, and the Royal portrait, examined before 1973), the original shoulders appeared even. It was immediately assumed they had been altered after Richard’s death to match the view of antagonistic Tudor historians, elaborated by Shakespeare, that Richard III was deformed (albeit Thomas More said it was his left shoulder that was higher).19
Here was objective evidence that the histories written after Richard’s death could not be trusted, of a Tudor need ‘to make Richard physically resemble the nature of his crimes … crude, if effective propaganda’.20 One writer goes to some trouble to distinguish between ‘commentaries’ written when Richard was alive, ‘personal reminiscences’ of people who knew him writing after his death, and ‘Tudor “histories”’ – with the last word pointedly in quotation marks.21 Evidence of later distortion becomes demonstration of earlier truth. The logic is applied to the paintings, too. In her excellent catalogue to an exhibition about Richard III held by the National Portrait Gallery in 1973, Pamela Tudor-Craig argues that the added raised shoulder and ‘slit-like’ eyes suggest the original painting owned by the Queen dates from Richard’s lifetime – or those features would have been put in at the start.22
Anne Neville (centre) is admired by her first husband, Edward Prince of Wales (left) and her second, Richard III (right), in the Beauchamp Pageant. (British Library, London)
There was always an alternative explanation. Richard III could have had a raised right shoulder, but either successfully concealed it, or – quite reasonably – did not wish it portrayed. Once he was dead, and his naked body had been seen by two armies and everyone in Leicester, there was neither means nor need to hide his real appearance. There is evidence that history portraits were researched for verisimilitude.23 Perhaps later artists and their clients simply wanted to get it right.
So if, as it seems, the portraits of Richard III illustrate Skeleton I so well, what of the written descriptions? How do they compare? The answer is again very well. We can dismiss extremes, such as the notion that Richard was born feet-first with all his teeth after a two-year pregnancy. More’s history is imaginative and rhetorical, and can be ignored for this exercise – thus we lose the ‘shrivelled, withered and small arm’, and the limp, both absent before More.24
Richard, says John Rous around the time of the king’s death, was ‘small of stature’, and ‘small in body and physically weak’. Nicolas von Poppelau, a visiting German knight, compares Richard to himself (whom he unfortunately does not describe), as ‘three fingers taller, but a little slimmer and not as bulky, also very much more lean; he had very fine-boned arms and legs’.
Rous describes ‘unequal shoulders, the right higher and the left lower’. Polydore Vergil (who reaches England from Italy in 1502, and pronounces Richard of ‘petty stature’) says one shoulder was higher than the other.25
As we saw in the last chapter, all these features – short, fine-boned and a higher right shoulder – are characteristic of Skeleton I. We cannot from the paintings tell Richard’s height, but otherwise the descriptions fit those too. It all matches so well, it seems reasonable to listen to what the records say about aspects of Richard’s appearance we can’t see in the bones.
On the morning of Bosworth, says the Croyland Chronicle (1486), Richard’s ‘countenance, which was always attenuated, was … more livid and ghastly than usual’. Vergil describes a ‘short and sour countenance, which seemed to savour of mischief’, and a ‘little and fierce face’. ‘He used constantly to chew his lower lip,’ he adds, ‘as if the savage nature in that tiny body was raging against itself’. And he fiddled: ‘he was wont to be ever with his right hand pulling out of the sheath, and putting in again, the dagger which he did always wear.’26
Having come this far, it does not seem too much to see these traits in the portrait faces, which show a ‘man of care’ old for his age, with ‘evident intelligence’.27 He is ‘a not uncomely man, despite the obvious lines of anxiety on his brow’, but bearing ‘a gaunt, bony, tight-lipped face’, and looking ‘much older than his true age’.28 Perhaps here are the clues to filling that gap between Skeleton I’s smooth bust – a
nd suggesting a yet thinner face – and a truly recognizable individual.
Maybe he bites his projecting bottom lip, and twists his finger ring. A compelling case is emerging. It starts with the grave, which both archaeology and contemporary record describe as being in the choir of the Greyfriars church in Leicester, and whose stark functionality is consistent with a lack of ‘pompe or solemne funeral’. It continues with the scientific evidence of the remains, carbon dated to the right era, identified as a man of the right age and status, who died violently in battle – through targeted assassination rather than random killing – and who matches the portraits and the historical descriptions. Skeleton I, the sitter in the paintings and Richard III could all be the same man. There remained just one line of evidence to consider, that could prove or demolish the argument: DNA.
Philippa Langley was right to be excited, back in 2006, when John Ashdown-Hill announced that he had found Richard III’s mitochondrial DNA ‘alive and well in Canada’.29 It meant that if remains were found that might be the king’s, there would be the possibility of testing the theory. Archaeologists had found a likely grave. After intensive study, everything seemed to support the idea that it might have been Richard III’s. Could DNA now conclude the case?
We are all more or less related, and share a great deal of genetic material. Most of our DNA cannot be used to determine precise ancestry, as it consists of a random combination from each parent (which is why we do not look exactly like either of them). However, two types of DNA are not mixed as they are passed on. During millions of years of evolution, sufficient changes have accumulated in these forms to enable us to distinguish individuals who are not related in a particular way. But for those who are, going back just a few centuries, the DNA is likely to be identical.
