This week there’s been much excitement about Professor Amit Roy Chowdhury’s scientific proof that this portrait, from Nidd Hall in Yorkshire, is a true likeness of Queen Anne Boleyn.

(c) NPG Publications

Professor Chowdhury compared it with a lead portrait medallion now in the British Museum, which shows Anne Boleyn ‘moost happy’ when she was pregnant in 1533. As you can see from this image from the Guardian, the medal is rather battered and is not the best guide to her appearance. Nonetheless, Professor Chowdhury believes that his mathematical algorithms establish a perfect match between the two.

(c) British Museum

And up to a point I agree with him. I am sure that the Nidd Hall portrait is indeed Anne Boleyn, and provided it is a contemporary likeness – because I’ve never seen it in the flesh – then it may well be the earliest surviving painted portrait of her.

I am sceptical that the computer can really make allowance for the way that different artists’ style can produce distinct interpretations of a single face. Professor Chowdhury certainly says it can. His algorithms also discovered that Veroccchio’s Lady with flowers and Botticelli’s Lady at a window are both portraits of Smerelda Brandini, but this rings a bell from Renaissance Florence at the National Gallery in 1999. I could be wrong, but I hope that art historian is still one of the last jobs in the world which can’t be replaced by a machine. Sir Roy Strong published the Nidd Hall painting as Anne Boleyn forty-five years ago (Tudor and Stuart Portraits National Portrait Gallery 1967). He notes the AB jewel at her breast, and the fact that Elstrack engraved the Nidd Hall portrait as Anne Boleyn in the early Seventeenth Century; Holland used the image in some copies of his Baziliologia in 1618. Fascinatingly Strong mentions a full-length portrait, in Lord Lumley’s collection by the late Sixteenth Century before being cut down in 1733. Lumley’s collection included paintings bought as a job lot from Nonsuch Palace.

Professor Chowdhury’s findings lead him to an unexpected conclusion. His experts further suggest that the familiar image of Anne, represented by the portraits at Hever Castle, in the NPG and by corridor-portraits up and down the country, might be an impostor. They suggest instead that this is a portrait of an unknown woman, pressed into service in the later Sixteenth Century after the real Anne’s portraits had all been destroyed.

(c) Hever Castle

(c) National Portrait Gallery

This surprises me: first, this portrait type was popularised during the reign of Anne’s daughter, Queen Elizabeth I. True, Elizabeth was three when her mother died, but I don’t believe she would have promoted a random woman in her place. The familiar Hever/NPG-type of Anne may be a later interpretation, rather than a copy of a lost original – and I stress may be since you never know in this business, and an early version of this portrait could easily surface. But I am certain that it would have been recognisable as Anne to people at the time, some at least of whom would still remember her, ‘good-looking, of a dark complexion’ according to Simon Gryné in 1531 (David Starkey and Bendor Grosvenor 2007 Lost Faces Philip Mould Ltd 2007).

The second reason is that – to me – these portraits agree about Anne’s appearance. Give or take the distinct artistic styles, they look like the same woman. And there’s a fourth portrait which clinches it for me. Holbein’s drawing, below, was labelled as Anne in the Eighteenth Century, following Sir John Cheke’s inscription written for King Edward VI. Scholars used to believe it showed one of the Wyatt family. Starkey and Grosvenor successfully argued for it to be reinstated as an authentic life portrait. Holbein’s sitter is a lady in her chamber, wearing her cap but not her hood and a furred dressing gown. Who but Anne, they ask, could appear at Court so informally?

(c) Royal Collection

So which of all these portraits is the real Anne? The short answer is: they all are.