Photo credit: National Trust for Scotland, Alloa Tower
If this is by Cornelius Johnson (Cornelis Janssens van Cuelen), the identity of the sitter is arguably then not Anne of Denmark, as the Queen would have been dead at the approximate time of this portrait (circa 1620–30?). There are strong stylistic similarities to other portraits of Johnson to support this suggestion, such as the artist’s portrait of Lady Hester Mainwaring. https://bit.ly/2xWb1cA
According to the collection, a 2016 valuation suggested that the artist is Paul van Somer, with the estimated date 1590–1610. It certainly does bear similarity to Van Somer’s work as well, however there is physically quite a difference in this portrait when contrasted to the other works of Anne by Van Somer. Also considering that, according to H. L. Meakin, Van Somer didn’t settle in England until 1616, it would make this portrait of a youthful-looking Anne difficult to realise. Consequently, I believe either the sitter or both artist and sitter may need reassessment.
The collection comments: ‘There are no notes on our records that indicate Van Somer's signature is inscribed on the work. The painting has been checked up close and on the back (although it is not taken out of the frame) and the pendant has no inscription on it, it looks like clear glass. There are no signatures visible (again with the frame in place) and no inscriptions on the back of the painting.’
Although it is difficult to speculate without more detailed photos, the portrait appears to bear a greater stylistic resemblance to the work of Van Somer, or even - more loosely - William Larkin or his circle. The suggestion that the sitter is Anne of Denmark is not particularly convincing, but I cannot suggest an alternative at present. A higher resolution image of the face would help in any further assessment of authorship.
Andrew, please find the PCF image attached.
This is a woman in her early twenties, possibly younger. It cannot be Anne of Denmark if van Somer painted it after his arrival in England, when she was in her forties. The same applies to Cornelius Johnson, who was active as a painter in England from ca. 1618. Also, compared to portraits of an older Anne, this face is a little too soft and dreamy. This could, however, be Anne's daughter Elizabeth (born 1596), perhaps in mourning for Anne's death in 1619 (the lace collar fits a date ca. 1620 but not so late as 1630).
Princess Elizabeth Stuart, of course, became the "Winter Queen" of Bohemia, having married Frederick V of the Palatinate in 1613. Here's a portrait of her by Mytens ca. 1626:
Interestingly, the gold medallion worn in our picture may be this one, made ca. 1616:
Here's a 1623 portrait of Elizabeth Stuart by Miereveld:
Elizabeth, by the way, was also the grandmother of George I.
There is a potential problem with the colour of the hair, as later portraits of Elizabeth Stuart show darker hair, but portraits of her as a child show hair of similar colour to our picture:
In any case, based on the age of the sitter, and regardless of who painted her, this cannot be Anne of Denmark, since that would mean the picture dates from the 1590s, which is too early for the collar.
It is noted above that as an adult the Winter Queen was shown as having dark hair, but her hair was shown as being fairer when she was young. This looks as if it might be a painting of Elizabeth as a younger girl before her marriage. Could the dark clothes relate for the death of her brother Henry Prince of Wales? She would have been 16 yrs when her brother died in 1612. She married Frederick V, Elector of Palatine (or Palatinate) early in 1613.
In later pictures and paintings of Elizabeth the Winter Queen she is shown with an extended downward tip to her nose. While such features can become more pronounced with age here the artist shows this young lady with her head tilted to one side, This creates an impression that the tip is less prominent than it would appear if her face was upright. Other features such as the relatively narrow, rather than wide set, eyes and low setting of the ears and lobes are also appropriate for the Winter Queen.
I can tell you with absolute confidence that it is not Anne of Denmark, as the sitter doesn’t have the face of Anne of Denmark. That would probably be the minimum requirement. Saying that, some famous portraits that noted people in the industry think are of AoD are not of AoD, so there is massive confusion. So, in that climate, I forgive.
Every female portrait would be called AoD if they could be, just like any was of EI, as they are the most prestigious and valuable sitters.
It is not Elizabeth Stuart, Queen of Bohemia, either.
It does look a bit like a few portraits of ESQB, at a distance, however looks nothing like her in most ESQB portraits, and if you zoom in on the faces of the portraits that do look like her because of the light hair, it’s not the sitter, it’s ESQB.
The sitter is not ESQB.
I believe the sitter is Lady Elizabeth Cecil nee Egerton, Countess of Exeter.
It does look like some works said to be Paul Van Somer, and could well be Cornelius Johnson. It is an oval, and CJ did ovals from 1619 onward, but he’d paint the oval in, on a rectangular standard canvas, and often sign and date them on the bottom right of the oval. This has been cut away from the oval, perhaps to remove the CJ signature and date which would have indicated that this can’t be AoD, as the sitter is clearly too young to be AoD in 1619, the year she died and that CJ started using ovals.
I don’t think there is a single Paul Van Somer oval painting that anyone has ever said is ‘by PVS’ it is always ‘after PVS’ or ‘circle of PVS.’ That suggests PVS didn’t do ovals, but someone in his style did. Cornelius Johnson did ovals.
The sitter is high on the canvas for a CJ, who is known for sitters’ heads being lower. The ‘PVS’ oval paintings sit very low in the frame for PVS – so I think they’re CJ.
The portrait could well be mid or late 1619.
On Lady Elizabeth Cecil nee Egerton, Countess of Exeter:
Her parents, John Egerton and Lady Frances Stanley married on 27 June 1602, her date of birth might be 1604, making her about fifteen in this 1619 portrait, which is marriage age at the time. She is their eldest child.
Her father was Lord President and Lord Lieutenant of Wales and the Marches.
Her mother was the second daughter of Ferdinando Stanley, 5th Earl of Derby, and her aunt Anne Brydges nee Stanley, Countess of Castlehaven, was legally Queen according to H8’s will, the legality of Katherine Grey’s sons not having been proven.
In 1826 no more descendents of Anne are recorded as living, so the claim passes to Frances Stanley’s line, so we are retroactively looking at a Princess, according to law, according to some, and called Elizabeth, so it’s curious that people were thinking it is Elizabeth Stuart, Queen of Bohemia.
Here is a portrait of her father, John Egerton, 1st Earl of Bridgewater, from ‘1617-19’ by ‘PVS or CJ’. I think it is another 1619 oval by CJ. It is a clear companion to the piece in question, same artist probably.
Here is a portrait said to be of ‘Lady Frances Stanley,’ her mother, c. 1619 by PVS. The resemblance is massive, it could be the same person as our sitter, if not for the receding hair indicating age:
Daughter, father, mother. If you whacked the three in one of those computer programs that can determine family relation via face analysis, it’d come out right.
Here is her son, John Cecil, 4th Earl of Exeter, of which there is quite a resemblance to our sitter, too:
Our sitter’s brother, through which a royal claim passes:
Here is our sitter and her daughter, Lady Frances Cecil, who married Anthony Ashley Cooper, 1st Earl of Shaftesbury, or else a close relative of her, heavy resemblance to the sitter:
Is this a relative of our sitter also, ‘c.1619’?:
Her cousins are the Egerton sisters:
This may well be Vere Booth nee Egerton, the eldest of the sitter’s three Egerton cousins from the above portrait:
This sure looks like our sitter or a relative also, rather than being Elizabeth Knollys, née Howard (1586–1658), Viscountess Wallingford, Later Countess of Banbury, as purported:
It is likely this is one of her relatives if not her, and not Lady Ann Morton as stated:
Such a sumptuous portrait, that dress like treacle.
In terms of provenance, there is a connection between my sitter and the Earl of Mar and Kellie.
Her only daughter, Frances Cooper nee Cecil, was Countess of Shaftesbury.
Susan Violet Ashley-Cooper (1868- 1938), daughter of Anthony Ashley-Cooper, 8th Earl of Shaftesbury (1831- 1886), married Walter John Francis Erskine, 12th Earl of Mar and Kellie (1865- 1955) in 1892.
Maybe this lady, descended from the Earls of Shaftsbury, marrying into Scottish aristocracy, the Earls of Mar and Kellie, brought this portrait, as, if thought to be AoD, it was of the Queen of Scotland, wife of James VI and I, the Queen at the pivotal point of the union of the crowns.
If that’s not plausible, nothing is.
To conclude: the artist is very probably Cornelius Johnson, mid to late 1619 perhaps, and it probably got cut out and away from the signature and date, and the sitter is Elizabeth Cecil, Countess of Exeter, although there are a group of relatives who look alike. Her daughter married into the Earls of Shaftsbury, of which a late-Victorian descendent married into the Earl’s of Mar and Kellie, of which the current Earl is the owner of the work.
Hope this is helpful.
Hope everyone is doing ok regarding the pandemics and lockdowns. Such a hard time for so many.
I agree with Luke's post today that this portrait is not of Anne of Denmark, as he explains. However, I am confident that it is not the work of Cornelius Johnson on grounds of style. The portrait dates to about 1615 or soon after.
@Jacob, thanks for that. In a discussion on a previous work https://www.artuk.org/artdetective/discussions/discussions/who-might-be-the-artist-of-this-portrait-and-the-sitter-portrayed I go into massive discussion like you wouldn’t believe, on how, judging by the work attributed to him, CJ didn’t start off painting like the mid-to-later CJ.
If he could imitate Van Dyck later, he can imitate other people earlier, and if he didn’t sign them, then we may not know, as we weren’t supposed to know, that’s the point, he can imitate.
He did do ovals, however. We have signed ovals, and a progression of style from what’s said to be early CJ, to signed ovals by him, where legacy of that style remains, into later style CJ ovals. Then he stops signing, and, it is said, imitates Van Dyck later, so that some Van Dycks may not actually be by Van Dyck.
If from 1615 however, CJ wasn’t supposed to be around then, so if that’s provably correct, it makes early unsigned CJs unsafe, and suggests CJ started doing ovals because of imitating this other person.
If from 1615, I revise my assessment of the sitter to be Lady Frances Egerton nee Stanley, Countess of Bridgewater, who is the lady in this ‘c. 1619 by PVS’ work:
She would be 32 in 1615.
If it’s her, the portrait of her husband is a companion to it: ‘John Egerton, 1st Earl of Bridgewater,’ from ‘1617-19’ by ‘PVS or CJ’:
However the piece in question is oval, CJ did ovals, and no earlier than 1619 that we know, and it is very similar to other CJ 1619 ovals:
Here is a portrait said to be of the ‘Countess of Exeter’ c. 1620, by ‘Cornelius Janssen van Ceulen’ (Cornelius Johnson) at the Milwaukee Art Museum. In reality, it is one of the previous Countess of Exeter’s daughters, and Elizabeth Cecil nee Edgerton was the next Countess of Exeter, so this sitter here could be her sister-in-law:
If the work in question is 1619 or 1620, it is Elizabeth Cecil nee Egerton, and this other work above is a companion piece, sister-in-laws, the daughter of the earlier Countess of Exeter.
If this work in question is 1615, it is Frances Egerton nee Stanley, Countess Bridgewater, the mother of my first suggestion, and it calls into question what we know about other works attributed to early Cornelius Johnson, and who started doing oval portraits and when.
As to the sitter's identity, I think an approach through the history of the portrait, which is still in the collection of the Earl of Mar and Kellie, is worth considering, as Luke has already suggested. Trying to identify portraits of this period thru' likeness is problematic unless a portrait is near identical in almost every feature, both facial and costume. Resemblance is not enough.
@ Thanks Jacob. Resemblance seemed to be enough for many people when they got the resemblance clearly wrong. When I tell you correctly, suddenly it is not enough.
I never said resemblance was necessarily enough or the only thing, but it gets you in a very close ballpark, and sure obliterates obvious mistakes of misidentification.
I said it looked so much like the mother it could have been the mother. It is one or the other. If it looks like portraits of a person’s father, mother, cousins, brother, son, then, hey, I do my best. But then suddenly ‘resemblance is not enough.’ Before I showed you this, misidentification of the resemblance seemed to be enough.
And, to correct you, resemblance may well be enough, if you put the faces in one of these computer programs to identify relations. Obviously the faces are not photographs in a passport-photo type position, but you get an expert to put the computer dots on the face from what can be seen, that then analyses the spacing and relationship between. If it comes back 80% likelihood that x is the daughter of y and 80% that it is the mother of z, it’s her. That is a 96% chance of being her, by an unbiased AI face analysis that has been taught what is correct in familial relationships. Sorry to go all 21st Century, but that is where we are. I guess I’m talking about a revolution that hasn’t happened yet.
I’ve always had a bit of a problem with provenance, as if for example a painting is burgled out of one house then sold to another collection, that doesn’t mean the artwork is of someone else, just because of mistaken or lying provenance, when most people can’t even see with their eyes that the sitter is not Anne of Denmark or Elizabeth Stuart, Queen Bohemia.
I thought my link from family to family was rather spiffing, so perhaps someone can wake up the great Earl, and get him to tell his people to ask their people. I’m sure he’d be interested in a portrait of his relative—either the mother, who the royal claim to the Earls of Jersey goes through, or her daughter, his ancestor who married into the Cecils of Exeter.
To be frank I consider talk of the artist of a work a total moot point, but can’t have failed to notice how people in the industry, highly-educated, in the standard mindset, are crazy about it, to the point of not having an eye for who the sitter actually is or sometimes seemingly not hardly even actually caring.
It is a moot point as, artists imitated others, wanted you to think it was by another, customers would have wanted that also. As Van Dyck was known to do, it is likely earlier major artists had studios where others did the basic painting, then the master came in and finished them off. They would have understudies, apprentices, that is how later masters and names learnt. If there was a rush, and the lesser man finished one off, is that by the same, or different, or both, or does it really matter?
And big names, with styles, weren’t born into the world with that later mature style. Sometimes you will know your attribution is correct, because everyone else tells you you’re wrong, and you are talking about a less mature style, or an imitation of another style. What really does it then matter? You have to be officially called wrong to be right, so why bother?
I know there is money in ‘who painted it,’ but there is therefore money in false authentications. I’ve heard respected people saying they know that big name big money authenticators lie for the money. So who, exactly, are we wasting braincells in trying to please? Liars?
There is money in actually knowing the sitter, I feel, but that takes eyeballs.
I am concerned about the identity of the sitter, and if people are related they look alike, I don’t want to shock anyone.
I think the sitters would want their true likeliness to be known.
There is a spiritual component in finding and revealing to the world the true likeliness of a previous human
I appreciate where Luke is coming from. However, the difficulty with putting faces into a computer programme to identify relations is that such a process ignores two fundamentals. Firstly, period likeness, since at a particular point in time hair styles, make up and costume give individuals a period resemblance. Secondly, each artist brings their own vision to a portrait, meaning that there is sometimes more likeness between portraits by a particular artist of different sitters than between portraits by different artists of the same sitter. I've yet to see a reliable programme for comparing painted images as opposed to photographs from the life.
@Jacob. I hear what you say.
I feel I have an insight that artists are not making up facial features. I am aware of the fashions. I never look at makeup and costume. Hair tends to stay in a ballpark, however, I have noticed, generally speaking, with some exceptions.
I have noticed portraits of EI and MI were painted to look more like H8 in the face, but aside from that, can’t say I’ve seen anyone being painted more like anyone for any reason. No-one else is important enough to paint anyone like. Being a true-ish likeness, so resembling your parents or grandparents and your children, is quite important, wouldn’t you say? The hair does copy the Queens, however. EI, and AoD. Suddenly many people have hair like the current queen.
I’ve said this often: expensive likenesses are a likeness. I didn’t get taught otherwise, and I’m not having it, I eyeball it.
Even if there is an ‘Instagram filter’ on the portrait, smoothing it out in a disorientating way, there is a sitter behind, as with Instagram.
I think that where you think people are unrecognizable because of different artists, I say people are always recognizable, but there are a raft of misidentified artworks.
To illustrate: people were honestly signing off publicly on this portrait in question being AoD or ESQD! But that’s nothing compared to some.
I’m recognizing faces like I would in real life. I feel that if I can see relatives, and the computers can do it with a high % accuracy on photos of humans, then it is inevitable for portraits, eventually. I think the vibe will clearly be that the industry doesn’t want it!
From my experience facial featured are identifiable. So much so that I ask people – so, did this artist use this other random person as a model in how to mispaint someone, or are they both portraits of the same person, perchance? Occam’s razor. It’s like hearing perfect pitch in music. Not everyone can do it. I never said I had perfect pitch, however.
Perhaps it is truly rare for someone to come in and tell art historians who have been taught they can’t tell by looks, and probably can’t tell by looks, that I can tell by looks, and then be brave enough to say what’s what. It’s a real benefit of these message boards and the art detective concept, crowd-sourcing it. In real life I’d never be let near.
No comment, Luke.
I agree with Luke. I have not checked this listing for a while so had missed this disagreement as to whether resemblance is important in identifying pictures and portraits.
For 100 years Egyptian art experts claimed ancient artist had no concept of portraiture in the sense of producing an impression of a living likeness. Since facial reconstructions have been obtained for royal Mummies dating from around 1200 BCE and earlier it is clear that the art experts were completely wrong. However the dogmatic beliefs of these experts is so ingrained they cannot bring themselves to admit they were hopelessly wrong.
The same line of teaching has been pursued for European art, we are even told that checking facial features is 'dangerous'. This has often resulted in identification which are clearly wrong being linked to antique portraits. It has encouraged many art historians to pay minimal attention and interest to the features portrayed and identity of the sitter.
How many art courses teach the subject of facial recognition. European art is obsessed with the identification of the artist and provenance but provenance can also just be a box to tick and often be misleading. As Luke notes above Museums have a substantial number of older portrait paintings (and this includes Royal portraits) that are incorrectly identified.
I admit that there are some people who can make wild identifications based on hasty perceived resemblances but facial recognition is a basic human skill and yes it applies for photos and paintings particularly if the artist is competent. They can not all be as good as Holbein, who must have had a photographic memory, but even in medieval times or from the reign of Henry III onwards English artists were making recognisable portraits instead of the imaginary models of royal perfection that art historians insist we must be looking for.
To be fair to the art experts here (of which I am clearly not one) facial recognition is both a very new science and one which appears to be entirely focussed on photographs and CCTV - and sufficiently successfully so for it to accepted as expert evidence in criminal trials.
There may be an interesting opportunity for a collaboration between university art history and forensic science faculties, working from securely identified painted and photographic portraits of modern and historical figures by different artists to work out whether the photographic identification techniques transfer to paintings or whether new techniques specific to paintings could be used. If so then it could be of great interest to Art UK. Does anyone here have the connections with academia to make a proposal along these lines?
Alison, please see this link to Art UK's experimental work with the University of Oxford's Visual Geometry Group. https://bit.ly/2Q85exg
Some attempts have already been made to use artificial intelligence to discover the identity of early portraits.
They chose Anne Boleyn as one of the first subjects and made limited progress. Anne Boleyn was a difficult place to make a start, 'fools rush in where angels fear to tread'. Boleyn's portraiture is problematic because so many portraits linked to Anne were provided with incorrect titles so soon after her death,
However in theory such systems should help to find interesting portraits and even make important discoveries in the near future. I believe work is being carried out on algorithms to identify the artist as well as the sitter.
Something which might help for the Art Detective investigation into the Trinity College painting of a young girl by Herman Verelst would be an algorithm or to get specialist to age the young ladies face to show how she would look when older. There is even a programme available that claims to be able to rotate 2D images and paintings to generate 3D views but I found its performance of limited benefit.
I think that generally the analysis provide by Art Detective contributors is extremely insightful, but it is a pity, though for understandable reasons, that investigations are limited to studying pictures that must be located at sites within the UK.
Marion, thank you for that fascinating link - which does seem to confirm that "resemblance" can be a proper consideration in identifying a sitter.
The Oxford facial recognition experiment was interesting and resemblance can be a consideration in the right context, but can't be relied on here for reasons that Jacob has already given (27/3/21, 12:43)