Quinten Massys
NG5769 
An Old Woman (‘The Ugly Duchess’)

Emma Capron, 2024

With technical contributions by Rachel Billinge, Britta New and Marika Spring, 2024

Extracted from:
The National Gallery; with entries by Emma Capron, Dillian Gordon, Sarah Herring, Mary McMahon, Letizia Treves and Francesca Whitlum‐Cooper; with technical contributions by Paul Ackroyd, Rachel Billinge, Lynne Harrison, Catherine Higgitt, Helen Howard, Larry Keith, Marta Melchiorre, Britta New, David Peggie, Marika Spring and Hayley Tomlinson, Online Entries (London: The National Gallery, 2024).

About 1513–19

Oil on oak panel, 64.1 × 45.2 cm

Provenance

An Old Woman (‘The Ugly Duchess’) probably remained in Antwerp until at least about 1645, when Wenceslaus Hollar (1607–1677), who was then residing in the city, published an etching reproducing it and its pendant An Old Man (or copies after those originals; see Copies, Engravings and Versions).¹ NG5769 and An Old Man may have travelled together to France soon after the print was made, but they seem to have gone their separate ways before the end of the seventeenth century. A copy after NG5769 appears to have been executed for Anne‐Marie‐Louise d’Orléans, duchesse de Montpensier (1627–1693), Louis XIV’s wealthy unmarried cousin. It may have formed part of a gallery of ancestral portraits at her Château d’Eu in Normandy, where it is first documented in 1839.² NG5769’s pendant, An Old Man, is also documented in Normandy, in the collection of a councillor at the parliament of Rouen named Boisguiraud, by the early eighteenth century.³ It then passed by descent to the comte de Saint‐Priest in Evreux before reappearing on the Parisian art market in 1954.⁴ NG5769 was probably sold in Paris on 22 November 1809 as part of the estate sale of Baron Pieter‐Nicolaas Van Hoorn van Vlooswyck (1743–1809), an affluent Dutchman who had settled in the French capital in 1797 and amassed a vast collection of antiquities, furniture and decorative arts, as well as a lesser group of paintings including some ‘Tableaux gothiques’.⁵ Among the latter featured lot 668, described as: ‘Quintin Mathsys. Maultasch, duchesse de Carinthie & de Tirol, reine des Romains, figure de grandeur naturelle, vue à mi‐corps, vêtue dans le costume du temps. Haut[eur]. 30 po[uces]., larg[eur]. 20. B[ois]’ (‘Quinten Massys, Mautasch, Duchess of Carinthia and Tyrol, Queen of the Romans, life‐size figure, seen at half‐length, dressed in the costume of her day. Height 30 inches, Width 20. On wood’; these dimensions, larger than the painting, may include the frame and the panel’s later extensions; on the latter, see Technical Notes). NG5769 was purchased by the dealer Maurice, who was acquainted with Henry Seymour (1776–1849) of Knoyle House, Wiltshire, and probably sold it to him. His son Henry Danby Seymour (1820–1877) showed it to Gustav Waagen before 1854 in the drawing room of his London home.⁶ It then passed to his brother Alfred Seymour (1824–1888) and by descent to his daughter Jane Margaret Seymour (1873–1943). She consigned NG5769 at Christie’s on 23 January 1920, causing a sensation on the London art market.⁷ The painting sold for a considerable £880 to the dealer Williams, from whom it was acquired within a few months by the painter and dealer Hugh Blaker (1873–1936). Manifestly fond of the picture (‘My Duchess of Carinthia and Tyrol by Quentin Matsys [sic] is a most important example of the master. Rejoices in the title of the ugliest portrait in the world. But what a masterpiece. I simply love her’, he wrote in September 1921),⁸ Blaker decided not to part with it and instead offered it twice on loan to the National Gallery in September 1920 and February 1922. The Gallery declined on both occasions, citing as reasons the lack of display space and the fear of seeing the painting sold off its walls.⁹ Also at issue was a matter of taste, for the Director and Trustees were unsettled by the figure’s appearance and saw NG5769 as little more than a ‘curiosity’ rather than a superlative work of art.¹⁰ Upon Blaker’s death the painting was inherited by his sister Jenny Louisa Roberta Blaker (1869–1947), who again offered it on loan to the Gallery in vain in 1938.¹¹ She bequeathed NG5769 to the National Gallery in memory of her brother and the painting entered the collection in 1947.

Exhibitions

London 1928 (X31); London 1932 (650); Brussels 1963 (166); London NG 1975 (25); London NG 2008–9 (70); London 2021 (19); London NG 2023 (1); Antwerp 2023–4 (15).

Pendant

A painting showing an old man (about 1513–19; Private collection, New York) was identified as the pendant to NG5769 in 1954 by Luis Reis‐Santos and published as such in 1977 by Larry Silver (fig. 1).¹² Seen in strict profile, the man has heavy eyelids, arched eyebrows, an aquiline nose and a fleshy lower lip. Puffy bags show below his eyes, white hair grows inside his ear and black stubble speckles his cheek. The wrinkles at his temples are picked out by white highlights. He has a double chin and creases at the neck. He is dressed in a deep purple‐red velvet bonnet and a dark blue coat with a wide fur collar and fur‐trimmed sleeves. Looking sternly to the left, he rests one beringed hand on a marble parapet while raising his other towards someone outside the picture plane. The marble ledge before him matches the one in NG5769 and the two figures’ left hands resting on the ledges closely resemble each other. While An Old Man bears all the unflattering signs of advancing age, in contrast to An Old Woman his features are not outside the ordinary and his more conventional profile never achieved fame. This disparity in appearance might make the pairing of NG5769 and An Old Man seem incongruous, a point that is addressed below.

An Old Man is in good overall condition.¹³ The paint surface is abraded in areas, most prominently on the sitter’s fur collar, and retouching is visible in the green background, which is nonetheless mostly original. At an unknown date the panel was thinned, marouflaged and cradled, and wooden strips were attached to all four edges of the panel. During these interventions, An Old Man may have been trimmed down slightly, but the panel seems to have been preserved at more or less its original dimensions: although there is no sign of barbes, examination with infrared reflectography revealed an underdrawn rectangle located about 2 mm from the current edge of the painted panel, seemingly made by the artist to mark out the dimensions of the composition. The current size of An Old Man, 64.3 × 45.8 cm (including 1 mm wood veneers on all edges), is nearly identical to that of NG5769. Materially, the two works show the same panel construction: they are each made of two oak boards, the left one (when seen from the front) being wider than the one on the right in both cases, and the join being about 30 cm from the left edge.¹⁴ The boards are slightly tapered so that the joins are not vertical but at an angle, and were perhaps cut from the same two planks, since if the panel for An Old Man is positioned above An Old Woman, the joins form a continuous line. This hypothesis could not be verified by dendrochronology given that oak veneers were applied to the back and edges of an An Old Man. However, if correct, it would seem to place the panels in the studio together, suggesting that they were conceived at the same time. These observations also appear to discount the possibility that NG5769 and An Old Man were ever painted on a single panel that was sawn into two at a later date.

Although intended as pendants, Quinten Massys may have worked out some key elements of the design on An Old Man first. For instance, it is unclear whether he planned to include the figure’s hands resting on a marble ledge from the start, as he painted the fur collar of his robe all the way down to the bottom edge of the panel, without keeping a reserve for the parapet. The figure’s left hand, however, was reserved. One explanation could be that the artist had already begun painting the collar, or at least blocking it in, when he decided to include the marble ledge and left hand. He then sketched it in and proceeded with the painting. By contrast, in NG5769 the ledge was underdrawn and reserved from the onset. The paintings’ green backgrounds now appear slightly different in colour. This is in part because of their differing condition and conservation histories, but close observation also suggests that the paint layers in the background of each picture were built up differently (see Technical Notes).¹⁵ Both NG5769 and An Old Man have lost their original frames, making it impossible to determine whether they were ever hinged together as a diptych. The paintings have been reunited twice, both times at the National Gallery, first during the Renaissance Faces exhibition in 2008 and then for The Ugly Duchess: Beauty and Satire in the Renaissance in 2023.

A painting in oil on paper measuring 48.4 × 37 cm and showing exactly the same profile head and bust as in An Old Man, minus the hands and ledge, is now in the Musée Jacquemart‐André in Paris (fig. 2).¹⁶ It is first documented at the estate sale of Joanna‐Aldegonis de Proli (died 1779), widow of Joseph Pelgrom (died 1769) of Antwerp, from a family already present in the city in Massys’ days, with members in the service of Emperor Maximilian.¹⁷ The work is painted on one single sheet of uncoloured paper.¹⁸ No watermark has been detected. The underdrawing, as revealed by infrared reflectography, consists of minimal delineation of the contours of the figure, without any indication of shading or signs of transfer, and showing minor adjustments.¹⁹ The painting is far finer in execution and in better condition than Massys’ other small heads on paper. Sometime after 1902, the work was transferred from its wooden support to canvas and a caramel‐coloured glaze was applied on its white background.²⁰ This tinted layer was removed during conservation treatment by the Centre de recherche et de restauration des musées de France in 2008–14.²¹ The oil on paper is signed and dated at the upper left ‘QUINTINUS METSYS PINGEBAT ANNO 1513’ (‘Quinten Massys was painting [this work] in the year 1513). Several tears run through the paper, most prominently in the background, along the figure’s temple and in the inscription. Consequently, the letters ‘TIN’ of the painter’s first name, ‘T’ of ‘PINGEBAT’ and the digit ‘3’ of the date have been reconstructed.²² When this damage and restoration took place is unknown. They may date from after 1902, during the transfer of the sheet from panel to canvas, but may also have taken place during an earlier, undocumented transfer.

The heads of the New York and Paris old men are one to one in scale, which implies either that they derive from the same cartoon or that one served as a template for the other. The underdrawing for the New York painting, made visible using infrared imaging, suggests that the Paris oil on paper was probably preparatory for it. In the New York underdrawing the figure’s head is neatly delineated, implying that Massys worked from an established pattern; by contrast, he seems to have devised the rest of the composition straight onto the panel itself, with free, searching lines for the hands. The colouring, modelling, repartition of highlights and other details that only occur in paint, such as creases in the figure’s upper lip, are visible on both the New York and Paris paintings, suggesting that the artist did not rely on a drawing alone to repeat the design but that he had a coloured model at hand – in all likelihood the old man in oil on paper – when at work on the panel. The Paris Old Man might have then been kept in the workshop for reuse in other compositions, as a similar face reappears with some variations on a smaller scale in two religious and secular paintings by and after Massys: Christ among the Doctors (a panel from the Madre de Deus Altarpiece, 1509–13; Museu Nacional de Arte Antiga, Lisbon) and A Grotesque Betrothal (about 1525–35; Museu de Arte de São Paulo Assis Chateaubriand, São Paulo, inv. MASP 00652).²³ This is one of the only instances in early Netherlandish painting where both a work in oil on paper and a related design on panel have survived.²⁴ The existence of such a head on paper for An Old Man suggests that an equivalent work might have existed for An Old Woman.

Related Drawings

Two drawings believed to reproduce a lost original by Leonardo da Vinci show the same face and bust as An Old Woman:

(1)

Windsor Castle, Royal Library, inv. 12492, red chalk, 17.2 × 14.3 cm, about 1510–20 (no watermark) (fig. 3). It is attributed to Francesco Melzi (1491/3–1570), Leonardo’s associate and heir, and may be one of the so‐called ‘replacement copies’ Melzi is believed to have produced to keep a record of original designs that had left the studio.²⁵

(2)

New York Public Library, Spencer Collection, Lenox and Tilden Foundation (b16832711), black chalk, 13.5 × 8.3 cm (fig. 4). It is one of 104 copies after Leonardo’s grotesque drawings that were bound at an unknown date to a 1669 edition of François Rabelais’s works.²⁶ The watermark indicates that this suite of drawings was made in Lombardy in the third quarter of the sixteenth century. This particular drawing is one of the largest and most refined in the series, its dimensions requiring a flap.

The two drawings are neat and legible and replicate Leonardo’s left‐handed hatching, yet lack his boldness and verve.

An engraving made by Philippe de Soye and published by Hans Liefrinck I in Antwerp between 1539 and 1575 reproduces Leonardo’s design.²⁷ A version of this drawing must have circulated in Antwerp by the 1510s for Massys to use (see Patronage, Function and Date) and remained available there for several decades. De Soye’s engraving belongs to a series of four prints depicting pairs of grotesque figures modelled partly after Leonardo’s compositions, and partly after pastiches of his work. The old woman is shown alongside an equally unattractive and elderly female partner. Although they are not a couple, the engraving is captioned: ‘Sordida deformis sic est coniuncta marito / Foemina, quo quaerat quisque sibi similem’ (‘The dirty and ugly woman resembles her husband. So it is that everyone seeks someone like themselves’).²⁸

Copies, Engravings and Versions

(1)

Wenceslaus Hollar, about 1645, etching, 6.7 × 12.4 cm, inscribed ‘Rex and Regina de Tunis / Leonardo da Vinci inv[enit]. WHollar fecit’ (‘The King and Queen of Tunis / Invented by Leonardo, made by W. Hollar’) (fig. 5).²⁹ This is a free reversed copy after NG5769 and An Old Man. It is the only instance in which the pair is reproduced together, and in fact the only known engraving or copy after An Old Man. Hollar was in Antwerp from 1642, seeking refuge from the English Civil War. This print belongs to the series Varie figuræ et probæ, which he published to support himself. It comprises a title page and 13 etchings after drawings by Leonardo da Vinci that Hollar had copied in the Arundel collection during his time in England. That collection included the drawing of a grotesque old woman now in Windsor discussed above (fig. 3; see Related Drawings (1)). Unlike De Soye’s engraving, however (see above), Hollar’s etching reproduces Massys’ paintings (or copies after them) rather than Leonardo’s composition. This is made clear by the inclusion of the old man, by details of costume (such as the trellis of flowers in the woman’s headdress and the gauffered edges of her veil) and by the incorporation of both a ledge and hand gestures – all features absent from the Leonardesque drawings. In his caption, Hollar ascribed the originals to Leonardo rather than Massys, probably because he recognised the old woman he had previously seen in the Arundel collection. Where the identification of the figures as the ‘King and Queen of Tunis’ may have come from is more difficult to explain (see The Identity of An Old Woman).

(2)

Anonymous copy, oil on canvas, 79.5 × 60.5 cm, Château d’Eu, Musée Louis‐Philippe, Eu. This copy is said to have come from the collection of Anne‐Marie‐Louise d’Orléans (1627–1693), duchesse de Montpensier.³⁰ It has a black background and extensions to all four edges, suggesting that modifications were carried out on the original before 1693 (see Technical Notes), with the additions at the left and bottom still in place when NG5769 was acquired by the National Gallery.³¹ The figure is on the same scale as the original, suggesting that the latter was traced.³² An inscription in French on the reverse of the lining canvas identifies the figure as Margaret ‘Maultasch’, Countess of Tyrol: ‘Marguerite surnommée Maultasche (“gueule de Sac”) / Comtesse de Tyrol, fille de Henry X Comte de Tyrol / et Roi de Bohème et d’Anne fille de Veneslas IV / née en 1300 / Mariée en premières noces en 1329 / à Jean Remy Duc de Moravie et en seconds noces / à Louis de Bavière morte en 1366’ (‘Margaret, nicknamed Maultasch (“pocket mouth”), Countess of Tyrol, daughter of Henry X Count of Tyrol and King of Bohemia and of Anne daughter of Wenceslaus IV. Born in 1300, first married in 1329 to Jean Remy Duke of Moravia and then to Louis of Bavaria, died in 1366’). It is the first instance of the figure being identified as the Duchess of Tyrol and Countess of Carinthia, Margaret ‘Maultasch’ (1318–1369).

(a)

This copy was engraved in reverse from 1787 by Gilles‐Antoine Demarteau the Younger, after a drawing of 1777 by ‘G[irolamo?] Paris’.³³ The engraving bears a caption identical to the inscription that appears on the reverse of the Château d’Eu copy. It is unlikely that the charcoal drawing now in the Tiroler Landesmuseum Ferdinandeum at Innsbruck is the drawing on which the painting is based.³⁴ Rather, it seems to derive from the print and faces the same way. The Demarteau print was later reproduced in reverse as a lithograph.³⁵

(b)

The Château d’Eu copy was itself copied by Louis‐Edouard Rioult at the request of King Louis‐Philippe for the historical museum of Versailles in 1839: oil on canvas, 21 × 18 cm, Musée national des châteaux de Versailles et de Trianon, Versailles (MV 4003; INV 7620; LP 3619).³⁶ This copy was later engraved by Ephraim Conquy, for Charles Gavard’s Galeries historiques de Versailles (1838–46).³⁷

(3)

An anonymous copy dating from the late seventeenth or early eighteenth century (oil on canvas, dimensions unknown) was formerly in the collection of Léon Armengol, Béziers.³⁸ It was inscribed: ‘MARGUERITE DAUTRICHE DUCHESSE DE BRABANT FILLE DE / MAXIMILIEN ET DE MARIE DE BOURGOGNE FIANCÉE A CHARLES VIII R / DE FRANCE EPOUSE EN I NOCES DE JEAN FILS UNIQUE DE FERDINAND / ET DISABELLE ET EN 2 NOCES DE PHILIBERT II DUC DE SAVOYE’ (‘Margaret of Austria Duchess of Brabant, daughter of Maximilian and Mary of Burgundy, engaged to Charles VIII King of France, first married to Jean only son of Ferdinand and Isabella, and then to Philibert II Duke of Savoy’). Close comparison suggests that this copy was probably also made after the original rather than the Château d’Eu copy.³⁹ It is the only known instance of the figure being identified as Margaret of Austria, but documentary evidence hints that more copies of this sort may have been circulating in French aristocratic collections in the eighteenth century: ‘a hideous portrait of a duchess of Brabant’ is described by Antoine‐Nicolas Dézallier d’Argenville in 1755 as hanging among some ‘old paintings’ in the cabinet of the king’s chamber in the Château de Saint‐Ange de Villecerf, near Fontainebleau;⁴⁰ an oil portrait depicting ‘the Cruel [woman] of Brabant’ and described as ‘a curious monstrosity’ was among the goods seized in 1794 from the Count of Artois, brother to deposed King Louis XVI and future King Charles X.⁴¹

(4)

An anonymous copy (oil on panel, 63.5 × 45 cm, Private collection, Vienna) was formerly in the collection of the Princes of Schönburg‐Hartenstein at the castle of Rothen‐Lhota in Bohemia.⁴² The brooch pinned to the woman’s headdress shows the figure of Venus and the letters C and B. The copy has a black background but does not reproduce the left and bottom extensions added to the work in the seventeenth century. Its dimensions are almost identical to the original.

(5)

Bayard, ‘Bulle d’excommunication du pape, et portrait d’après nature, de la princesse Porcia, sa soeur’ (‘Excommunication Bull from the Pope, and Portrait from Life of His Sister, Princess Porcia’), 1797, woodcut, 48 × 38.5 cm, Bibliothèque nationale de France, Paris (FRBNF40253576) (fig. 6).⁴³ This political broadsheet, published at the height of the French Revolution, borrows the features of NG5769 to depict the evil Porcia, fictional sister of Pope Pious VI, sent by the deposed pontiff to overthrow the French Republic. It was most likely made after the Demarteau print, rather than the original picture or one of its painted copies.

(6)

In two illustrations for the first edition of Lewis Carroll’s Alice’s Adventures in Wonderland (1865), John Tenniel took direct or indirect inspiration from NG5769 for the character of the Duchess, described in the book as ‘very ugly’ and with ‘an uncomfortably sharp chin’ (fig. 7).⁴⁴ Details such as the lattice of flowers on her headdress and the scalloped edges of her veil demonstrate that Tenniel did not use the Leonardesque drawings for his work but instead based his design on either NG5769 or one of the many engravings made after it. Whether he actually saw NG5769 is hard to ascertain, but the painting was in Britain by that date, in the collection of Henry Danby Seymour (1820–1877), who kept it in his London home. Tenniel may have met and visited Seymour through Carroll’s friend John Ruskin, who knew the collector well.⁴⁵

(7)

The German and English publishers of Lion Feuchtwanger’s Die häſzliche Herzogin (The Ugly Duchess), a historical novel fictionalising the life of Margaret of Tyrol and Carinthia, which first appeared in Germany in 1923 and in the UK in 1927, used the painting or versions after it for the cover of numerous editions of the book. Testifying to the power of this image to disturb, the dustjacket for its first English ‘cheap’ edition was banned from parts of the British book trade for being too provocative and indecent.⁴⁶

Technical Notes⁴⁷

Conservation History and Condition

The painting was cleaned in 1947–8 by Messrs Drown, when a discoloured varnish was removed, as was black overpaint across the background and other layers beneath it (tiny islands of this are still present – see further below).⁴⁸ An addition at the lower edge was also removed in 1947–8 as its grain was perpendicular to that of the original panel, which was affecting its stability. This, and another addition of beech on the left,⁴⁹ 3.5 cm wide, probably date from when the background was overpainted, since on the side addition the black paint lay directly on the wood. The Eu copy seems to have been made after NG5769 had been enlarged and overpainted, because it includes the additions and black background. The copy is thought to have been painted for Marie‐Louise d’Orléans, duchesse de Montpensier, who died in 1693, so the additions and black overpaint of NG5769 are likely to date from before that year. Rebates on the back at the top and bottom edges are probably evidence of half‐lap joins made when enlarging the panel; cross‐grain timber of the same type as the additions remained in place after removal of the addition at the bottom in 1947, and glue had been left in the channel of the upper rebate, implying there had at one point also been an addition at the top, already removed by the time the painting was acquired by the Gallery. A digital overlay of images of NG5769 and the copy at Eu seems to confirm that there was indeed an addition at the top when the latter was made, and that the panel had also been enlarged at the right side. Before 1947 the top and right additions had been removed and the panel had been thinned and cradled. The cradle, the addition at the left and the wood and glue in the rebates were removed in 2021–2, when the painting was again cleaned and restored. Most of the picture is well preserved. There is some damage, however, to the background, including erosion and abrasion and some areas of more extensive paint loss. On top of the opaque apple‐green paint, traces of a very thin dark green translucent layer are visible in places; this layer was probably originally across the whole background, although it is not possible to say whether it might have varied in thickness to indicate light and shade. When the painting was cleaned and restored in 2022 the losses in the opaque green paint of the background were filled and retouched and the finish of the restoration was taken to a level that matched the traces of dark green glaze.

Support

The original part of the panel, which measures 64.1 × 45.2 cm, consists of two boards of oak, vertical in grain. The join, between 13.3 cm and 12.9 cm from the right edge, is at an angle, as the boards are slightly tapered. It is secured with two dowels, exposed when the panel was thinned. It has been established that the oak is from the Baltic region, that the 274 growth rings, grown from left to right, of the first board were formed between 1191 and 1464, and that the 97 rings of the second board, grown from right to left, were formed between 1393 and 1489. The first board is from the same tree as the first board of the Crucifixion (about 1515; NG715), also attributed to Quinten Massys, and the first board of the Salvator Mundi, now given to the Master of the Mansi Magdalen (1505–30; Suermondt‐Ludwig‐Museum, Aachen, inv. GK295), previously attributed to Quinten Massys.⁵⁰

Materials and Technique

The panel has been prepared with a chalk ground (confirmed by SEM‐EDX analysis) onto which a very thin light grey priming composed of lead white and a little black has been applied. The ground and paint extend up to the very edge of the panel at all four sides, although the final paint layers do not quite reach the edges. No traces of barbes have been found around the edges. This is likely to indicate that it was not painted in an engaged frame, rather than that the picture has been cut down. There is an incised line very close to the bottom edge that probably marked the outer boundary of the composition (see also Pendant).

Extensive linear underdrawing in a liquid medium is revealed with infrared reflectography (fig. 8). The underdrawing for the face consists of relatively fine lines, whereas that for the headdress, clothes and parapet is bolder and probably only intended as a rough guide rather than exact placement: for example, the fluted edge of the veil over the woman’s right shoulder is indicated by a wavy line. A number of changes were made to the underdrawing. In the face, both eyes were moved up slightly, with corresponding repositioning of the nose and chin, but not apparently the mouth. Her right ear was enlarged. More dramatically, the angle of her body to the picture plane was changed. An underdrawn line at a steeper angle can be seen extending from the woman’s neck to the top of her right arm lower than where the shoulder was painted, while above the left shoulder there are lines visible in the reflectogram that could be for the shoulder in a higher position. The decision to reposition her body was made before painting started, since there is underdrawing that relates to the figure as finally painted. Her right breast was also repositioned. The free nature of the underdrawing in the veil makes it difficult to be sure whether this was adjusted as part of the repositioning or just made neater, tightening up the loose drawing. Similarly, free underdrawing can be seen for the hands, both of which have been changed. Unfortunately, the dark underpaint in the deep purple central part of the woman’s bodice makes it impossible to see parts of this drawing so that interpretation is difficult. There are several horizontal lines suggesting the top edge of the parapet. These are freehand and not very straight, but the final position has been reinforced by ruling an incised line along one of the underdrawn lines.

The changes mentioned above were made before painting began, but the infrared reflectogram also reveals a change made when the painting was at quite an advanced stage. The carefully painted, geometric design of the coloured flowers on the horned headdress shows clearly on the horn over the woman’s right ear, but on the other side, in addition to the flowers now visible, some other flowers are revealed in what must have been their first painted positions (figs 9, 10). The first positions of these flowers are also visible in several of the XRF maps. It is also evident from the XRF maps that the brighter red flowers on the headdress in the more well‐lit area on the left are painted in only vermilion, while in the shadowed area on the right this pigment has been mixed with a duller red earth.⁵¹

The blue dress and deep purple underdress both have a greyish‐pink modelled underpaint consisting of lead white, red lake and black in varying proportions. In the blue dress the upper paint layers are based on azurite, mixed with some lead white in the highlights and mid‐tones, or with a little red lake in the shadows. The underpaint gives the drapery a subtle greyish‐blue hue. In the purple underdress, visible in the centre of the bodice through the lacing, the uppermost layer contains only red lake (fig. 11). It seems that the artist, rather than using a conventional mixture of red lake and a blue pigment, has relied on the optical effect of the underpaint (a darker pinkish‐grey than in the blue dress) showing through the translucent upper layer to give the purple colour. Scumbles of white or pink have been painted wet‐in‐wet into the red glaze to indicate the highlights and convey the texture of the fabric.

The opaque green paint in the background contains a mixture of a green mineral copper pigment composed mainly of copper sulphate and a small amount of malachite, mixed with lead white and lead‐tin yellow. This seems originally to have been covered by a further thin translucent layer containing the same copper mineral pigment, of which only traces remain; it was initially present across the whole background, giving it a deeper hue, although it is impossible to say whether it also provided some indication of shadows.⁵² A paint cross‐section of one of the tiny surviving islands of the black overpaint and other older layers that had been present before the 1947–8 cleaning indicated that on top of the green layers just described there had been a substantial layer of lead white followed by a fairly thin deep green, again containing a copper mineral pigment composed mostly of copper sulphate, presumably applied relatively early, since in Netherlandish paintings this type of pigment has been found mainly in paintings dating from the late fifteenth century to about 1550.⁵³ This was followed by a layer of discoloured varnish and then finally the black paint, which we know from the copy at Eu is likely to be from before 1693 (see above).

In the only sample taken for analysis of the medium, which was from the white veil, the medium was identified as walnut oil.⁵⁴ The artist, in his distinctive manner,⁵⁵ has taken full advantage of the ways in which oil paint can be manipulated, handling the paint in a vigorous but assured and precise way. He feathered the paint to soften transitions between light and shade, for example in the face, where it also creates a sense of less than perfect skin (fig. 12), or in the darkest grey shadows of the veil. He incised into the wet paint with the end of a small brush or other tool to create texture or pattern (as in parts of the veil, some of the flowers in the headdress and some of the hair just above her right ear)(figs 13, 14, 15). He also exploited the buttery properties of paint containing lead‐tin yellow or lead white, for example in the tiny yellow dots and fine strokes that highlight the jewellery (fig. 16) or the lively touches of thick white paint around the fluting at the edges of the veil.

Description

The woman has small, lively, hazelnut eyes. Each animated by two catchlights (fig. 17), they are set deep in their sockets and crowned by untidy brows. A fine network of radiating wrinkles emanates from them. She has a large domed forehead and a triangular, snub nose flanked by wide nostrils. The distance between her nose and mouth is long. Her sunken lips signal that she has lost all her teeth. High cheekbones and a powerful square chin give structure to her face. Her complexion is ruddy, her skin flecked with small pimples. Her right cheek bears a hairy mole (fig. 18). The tips of her forehead, nose and chin glisten. Tufts of red hair peppered with white escape from either side of her imposing horned headdress, a black V‐shaped structure decorated with a trellis of flowers and covered by a white linen veil with gauffered edges that cascades down her shoulders. A bejewelled brooch secures the veil to the headdress. The woman wears a blue dress with a low neckline over a purple undergarment and a white chemise. The dress is laced tightly at the front with a green ribbon to prop up her wrinkled bosom. Her shoulders and chest are broad. The cuffs of her sleeves are embellished with a band of gold embroidery. Rings adorn her fingers, while dirt is lodged underneath her fingernails. She rests her left hand on a marbled parapet while her other hand projects outward, holding a droopy rosebud whose tip suggestively aligns with the edge of her cleavage. Slightly larger than life‐size, commanding and confident, she presents herself to the viewer. She defies all the canons of beauty and rules of propriety that defined female decorum during the Renaissance.

The old woman’s appearance fits within the well‐known definition of the grotesque body given by the Russian literary scholar Mikhail Bakhtin in his seminal book Rabelais and His World, first published in 1965: Of all the features of the human face, the nose and the mouth play the most important part of the grotesque image of the body; the head, ears, and nose also acquire a grotesque character when they adopt the animal form or that of inanimate objects … [The grotesque] is looking for what protrudes from the body, all that seeks to go out beyond the body’s confines … The grotesque body is a body in the act of becoming. It is never finished, never completed; it is continually built, created, and builds and creates another body … Thus, the artistic logic of the grotesque image ignores the closed, smooth, and impenetrable surface of the body and retains only its excrescences (sprouts, buds) and orifices, only that which leads beyond the body’s limited space and into the body’s depths. Mountains and abysses, such is the relief of the grotesque body.⁵⁶

Bakhtin’s description of the grotesque body as unsteady terrain – ‘mountains and abysses’ – holds especially true for An Old Woman: her extraordinary, volcanic face leads us from the crest of her vertiginous headdress down the mound of her forehead, up the hill of her nose, through to the plains of her philtrum, down the canyons of her wrinkles, and into the deeper crevasse of her cleavage. She also perfectly answers to Bakhtin’s notion of a body that ‘transgresses its own limits’: her tightly laced corset compresses and constrains her bosom, which bulges forward in reaction, and seems about to break out of the picture plane.

Attribution

Like all but four surviving paintings by Massys, the panel is neither signed nor dated.⁵⁷ The same applies to its pendant, An Old Man. However, from the first appearance of An Old Woman on the Paris art market in 1809, it has been unanimously attributed to Quinten Massys.⁵⁸ The attribution was only dismissed by Martin Davies, who considered the painting’s facture too weak to be an autograph work and catalogued it as a copy after a lost original dating from the 1510s.⁵⁹ His opinion was not accepted by subsequent scholars, including Andrée de Bosque, Larry Silver and more recently Lorne Campbell, who continued to catalogue this work as a fully attributed painting by Massys.⁶⁰ As noted by Campbell, the presence of many adjustments in the underdrawing of the face and changes in the woman’s headdress made during painting seems to discount the possibility that this is a copy, as does the free, exploratory nature of the underdrawing for her bust and garment.⁶¹ This is further confirmed by the superlative quality of execution, with such finely rendered, egregious details as the calligraphic white and brown hair sprouting from her mole. Many of Massys’ ‘signature traits’ are present in NG5769. They include the profusion of feathering to soften tonal transitions in the flesh, particularly visible in the face, and the exploitation of the buttery properties of lead‐tin yellow paint for the tiny strokes and dots that describe metalwork; the evidence of scratching into the wet paint to create texture, especially prominent in the linen veil (compare with the Lamentation Altarpiece – see below); the refined yet economical rendition of marble surfaces, with the characteristic cloudy veins on top of speckles; and the grey modelling in the whites and the wet‐in‐wet application of the paint on the edge of the figure’s white chemise (compare with Pieter Gillis and see Technical Notes).

Massys and Leonardo

Gustav Waagen first connected NG5769 to the drawing now in Windsor.⁶² It has since generally been assumed that Massys knew this composition and drew inspiration from it for his painting. This was not the first time that the artist had engaged with Leonardo’s grotesque designs. Those drawings, or more likely copies thereof, had reached the Netherlands by the first decade of the sixteenth century and Massys was a key figure in their introduction in Antwerp.⁶³ Quite how he encountered them is unclear. As noted by Martin Clayton, in the absence of an established trade in drawings at that period, it is probable that originals or copies of Leonardo’s design were brought to the Netherlands by a Northern artist who had travelled to Italy.⁶⁴ Jan Gossaert made such a journey in 1508–9, stopping in Florence and Rome but not Milan.⁶⁵ The possibility that Massys went to Italy himself has been raised but cannot be substantiated, especially given how little we know of his biography in the first two decades of his activity, between the time he is first documented as a master craftsman in Antwerp in 1491 and his first securely datable work, the Lamentation Altarpiece (1508–11, Koninklijk Museum voor Schone Kunsten, Antwerp, inv. 245–249).⁶⁶ This triptych already shows knowledge of Leonardo’s A Man tricked by Gypsies (about 1493, Royal Library, Windsor, RCIN 912495), whose figures were repurposed as Saint John the Evangelist’s tormentors on the right wing. Massys appropriated motifs invented by Leonardo on at least seven other occasions during his career.⁶⁷ A knowledge of Leonardo’s drawing is also evident in the work of Massys’ younger Antwerp contemporary Joos van Cleve, and has been detected in paintings by Hieronymus Bosch and Albrecht Dürer too.⁶⁸ In this context, it is notable that no drawings after Leonardo’s grotesques made by a Northern hand before the second half of the sixteenth century seem to have survived.⁶⁹ Importantly, these drawings’ main appeal to Northern artists seems, at least initially, to have resided in their formal qualities – as a prodigious repertory of types and expressions – rather than their narrative content or Leonardo’s name, which had not yet achieved widespread fame by that date.⁷⁰

The figure of the old woman with a horned headdress and low neckline was a mainstay of Leonardo’s grotesque imagery: she haunts his drawings under endless guises, from small profile heads to more complex compositions.⁷¹ As discussed more fully below, this particular figure may reflect a folkloric type drawn from the world of carnival, as testified by a Florentine print probably published during Leonardo’s youth (see A Comic Type). The Windsor and Spencer drawings stand out as the largest and most elaborate renditions of this favoured type.⁷² The three‐quarter view likens the composition to a portrait, comparable to – and perhaps satirising – those that Leonardo was painting around the same time for the Duke of Milan, Ludovico Sforza.⁷³ While some of Leonardo’s grotesques are no more than doodles, his most ambitious drawings of the type, such as A Man tricked by Gypsies, seem to have been intended for an audience beyond the confines of the studio, perhaps as court entertainment.⁷⁴ Jokes travel fast: if the original for the Windsor and Spencer drawings served a similar public function, this would explain its early dissemination towards northern Europe. Whether an equivalent Leonardesque prototype existed for An Old Man remains unclear.⁷⁵ No drawing featuring a comparable figure has been identified. Although Leonardo’s grotesques display varying degrees of exaggeration, the old man’s features and outfit are rather tame in comparison to them.

Recently, Lorne Campbell has called attention to An Old Woman’s strikingly lifelike quality and the fact that her costume looks more coherent in Massys’ painting than in the Leonardesque drawings. He concluded that in the particular case of the Old Woman, Leonardo or his followers copied an original by Massys, which they may have known through drawings that the two artists would have exchanged on account of their shared interest in the grotesque.⁷⁶ Leonardo was certainly alert to northern European innovations in the realm of portraiture during his Florentine youth,⁷⁷ and instances of such trade between Italian and Northern artists are recorded for the early sixteenth century, famously between Raphael and Albrecht Dürer.⁷⁸ Yet there is no surviving evidence that Massys and Leonardo knew each other, nor any extant drawing by Massys. Importantly, this proposition does not address the date of Leonardo’s grotesque drawings, which scholars unanimously agree were primarily made in the 1490s, several decades before Massys’ painting.⁷⁹ The grotesque print designed by Philippe de Soye and published by Hans Liefrinck that clearly replicates Leonardo’s design further demonstrates at least one drawing of this particular face was circulating in Antwerp by the mid‐sixteenth century.⁸⁰ It is therefore safe to conclude that the motif for An Old Woman originated with Leonardo, although not without inspiration from northern Europe, as discussed below.

Perhaps as significant as the question of who came up with the design, however, is the issue of what each artist did with it. Massys was anything but a subservient copyist, as other instances of his borrowing from Leonardo attest. He typically would appropriate, manipulate and sometimes obscure his source material freely. This creative engagement is visible, for instance, in Head of a Fool (1520s; The Phoebus Foundation, Antwerp), in which a study of facial expression by Leonardo – the bald head of an irate man – is transformed into a shrieking jester with a full head of hair and brandishing a spoon.⁸¹ Leonardo confined his visi mostruosi (‘monstrous faces’, as Giorgio Vasari called them) to small scale and paper.⁸² Massys, by contrast, elevated the grotesque to panel and translated the intimately sized drawing into an impressively large composition that straddles the realms of genre and portraiture. He further added vivid additional details – hair, mole, pimples – and a sense of texture, a vibrancy of colour and a tactile appeal absent from the drawing.⁸³ This required him to expand the design substantially (adding arms, hands, flower attribute and marble parapet), as well as clarify and correct motifs that were only sketched out in the drawing or misunderstood by copyists of Leonardo’s work (adorning the headdress with a patterned trellis, pinning the veil securely to it and giving the linen fabric fluted edges, as was customary in northern Europe). The underdrawing for An Old Woman confirms this scenario: it contrasts careful lines showing few changes in her face, which suggests that he was working from a guiding drawing, with freer exploratory marks for the costume and the hands, where he was inventing something new.⁸⁴ The fact that the composition is more resolved in Massys’ painting than in the Windsor and Spencer drawings does not so much prove that the former was the prototype for the latter, but rather emphasises the different nature of a finished panel relative to a more informal work on paper.

The Identity of the Old Woman

An Old Woman has taken on many identities over the centuries. Hollar inscribed his etching of the painting and its pendant An Old Man as ‘Rex and Regina de Tunis’, a puzzling caption given that the artist, who made a career of engraving earlier works of art, would have been unlikely to mistake the figures’ European garb for North African dress.⁸⁵ This fanciful title, no doubt a seventeenth‐century fabrication, might have originated with Hollar himself, but the panels may also have been known as such in Antwerp at the time.⁸⁶ Kings and Queens of Tunis occur in plays and novels of the day, from William Shakespeare’s The Tempest (written in 1610–11; first printed in 1623) to Sébastien Grenadine’s Homaïs, Queen of Tunis (1670s), although none of these literary characters are notable for their unusual appearance.⁸⁷ Erica Tietze‐Conrat posited that Hollar’s print depicted two real‐life court dwarfs, whom she argued were routinely given such mock titles.⁸⁸ This proposition can be dismissed, however, as nothing in the figures’ appearance hints at this condition and nothing in their costume suggests the apparel usually worn by court jesters.⁸⁹ Although Hollar’s intentions cannot be securely recovered, it is possible that his label intended to other ugliness by making it an extra‐European trait, lending the print a sinister racial dimension.⁹⁰

Copies after NG5769 painted in seventeenth‐ and eighteenth‐century France offered alternative identifications of the figure. This included Emperor Charles V’s aunt, the regent of the Netherlands Margaret of Austria (1480–1530), but NG5769 does not look anything like her well‐known portraits by Bernard van Orley (see Copies, Engravings and Versions). This erroneous identification might be explained by the anti‐Habsburg sentiment pervading early modern France as well as the dynasty’s notoriously prognathous jaw.⁹¹

While the identification to Margaret of Austria did not hold, another copy enshrined a more enduring – if equally specious – identity for the old woman, eventually earning her the nickname ‘The Ugly Duchess’. A painting presumably made after NG5769 for Louis XIV’s cousin, Anne‐Marie‐Louise d’Orléans, duchess de Montpensier, and hanging in her gallery of dynastic portraits at her Château d’Eu, identified the figure as an infamous medieval ruler, Margaret, Countess of Tyrol and Duchess of Carinthia (1318–1369) (see Copies, Engravings and Versions). Margaret was the daughter and heiress of the Count of Tyrol and Duke of Carinthia Henry VI and his second wife Adelheid of Brunswick‐Lüneburg‐Grubenhagen.⁹² Margaret’s rule was tumultuous and marked by her wilful if ultimately fruitless resistance to her neighbours’ ambitions over her lands. Upon her father’s death in 1335, Carinthia was annexed by the Habsburgs, but Margaret held on to Tyrol, which she ruled with her first husband John Henry, a brother of Holy Roman Emperor Charles IV. John Henry was rumoured to be impotent, and returning from a hunt in 1341 he found the gates of Tyrol Castle closed. Although their marriage was not officially annulled by the Pope until 1349, from 1342 Margaret had taken a new husband, Louis of Bavaria, Elector of Brandenburg, and she bore him a son, Meinhard. However, the passings of both husband and son left her without an heir and in 1363, six years before her own death, she relinquished Tyrol to Rudolf IV of Habsburg and retired to Vienna.

The only contemporary depiction of Duchess Margaret, the seal on the 1363 charter that transferred her territorial possessions to the Habsburgs, presents an utterly generic, stylised and slender figure.⁹³ A contemporary chronicler describes her as very beautiful.⁹⁴ However, as a result of her scandalous divorce and unwomanly claim to political rule, a legend portraying Margaret as evil and promiscuous began taking hold during her own lifetime. Her enemies nicknamed her ‘Maultasch’ (‘pocketmouth’), the German slang for vagina, also used as a derogatory term for a sex worker.⁹⁵ The moniker was probably later taken literally as a physical trait reflective of Margaret’s moral turpitude, and the tale of the extraordinarily ugly duchess was born.⁹⁶ Quite how the memory of this relatively obscure German medieval noblewoman was alive enough in seventeenth‐century France to attach itself to the Eu copy of NG5769 can only be speculated upon. Perhaps it offered a warning to its owner, the rich, unmarried duchesse de Montpensier, known as ‘La Grande Mademoiselle’, a powerful woman who was herself no stranger to sexual scandal and political miscalculations.⁹⁷ In any case, this identification was circulated and became entrenched via several prints (see Copies, Engravings and Versions). When NG5769 sold at auction in London in 1920 under Margaret’s name, the British press christened its sitter ‘The Ugly Duchess’, a moniker later cemented when Margaret’s life story was fictionalised by the novelist Lion Feuchtwanger in 1923 as Die häſzliche Herzogin, which appeared in English in 1927 as The Ugly Duchess, with NG5769 as its cover image.⁹⁸

More recently, it has been proposed that NG5769 portrays someone with Paget’s disease, a rare illness causing bone hypertrophy that frequently affects the skull and more rarely the jaws and facial bone; a particularly advanced stage of the condition could perhaps result in such atypical features.⁹⁹ This would fit with the morbid, voyeuristic fascination some early modern painters had with disability and non‐normative bodies, and recalls Vasari’s romantic account of Leonardo, sketchbook in hand, roaming the Florentine streets in search of unusual faces.¹⁰⁰ Diagnoses established on the basis of a painting should be reached with caution, however, for not only are they fraught from a medical standpoint, but they also tend to overlook the contingencies of style and undermine artistic licence, confining painters to the literal replication of the realities around them and underplaying their power of invention. Although some grotesques by both Massys and Leonardo may well be rooted in direct observation, rather than the faithful portrayal of specific pathologies, they are probably best understood as exaggerations and imaginative explorations of the comic, expressive and subversive potential of distorting the human face.

Several identities have also been proposed for An Old Man. Silver saw it as a satirical portrait of Cosimo de’Medici, personifying greed, based on comparison with a medal depicting the Florentine ruler and a later portrait by Pontormo.¹⁰¹ More persuasively, Karel Boon noted similarities between An Old Man and a sixteenth‐century copy after a Portrait of Duke Philip the Bold from about 1400 (fig. 19).¹⁰² As discussed below, parallels with earlier portraits are more likely to do with Massys’ source material than any deliberate attempts at either posthumous portraiture or political caricature. The oil on paper version of An Old Man (see Pendant) has been described in an early auction record as the artist’s self portrait on no apparent grounds whatsoever.¹⁰³

A Comic Type

Despite Massys adopting the visual conventions and fine execution of double portraiture – showing his figures at bust length, before a marble parapet and against a monochrome background – elements of costume and decorum immediately locate NG5769 and its pendant within the comic realm and reveal them as parodies rather than genuine likenesses. The Old Woman’s flamboyant and revealing garments have nothing to do with early sixteenth‐century female attire. Instead, they consist of a fanciful miscellany of fifteenth‐century courtly dress from both sides of the Alps.¹⁰⁴ This type of bodice laced at the front was popular in the 1470s and 1480s both in the Netherlands and in Italy, but it was invariably covered by a linen fichu or gauze undercollar, which the old woman has brazenly dispensed with, immodestly revealing her wrinkled cleavage.¹⁰⁵ The exposure of such ‘shameful parts’ as breasts was ranked first among laughable actions by the sixteenth‐century French physician Laurent Joubert in his famous Treatise on Laughter (1579), providing a sense of how this outfit would have been received by contemporary viewers.¹⁰⁶ The old woman’s high hairline and plucked forehead further emulate mid‐fifteenth‐century ideals of courtly beauty.¹⁰⁷ Her outlandish horned headdress recalls the monumental headpieces worn by aristocratic women in France and the Burgundian Netherlands around 1430–40. Already denounced as sartorial excess in their own day, by Massys’ time such headgear had become a ubiquitous iconographic shorthand for female vanity and evil.¹⁰⁸ Although the Old Man’s attire is less extravagant than his counterpart’s, it does not reflect male fashion from the 1510s, especially his puffy bonnet. The medallic profile Massys adopted would also have looked deeply antiquated in the Netherlands by that date.¹⁰⁹

Massys’ contemporary and two‐times patron, the great humanist Desiderius Erasmus (1466/9–1536), touched upon the ridicule associated with the type of outdated dress seen in NG5769 in The Ciceronian, a fictive dialogue published in 1528. In a passage in which Erasmus seeks to demonstrate that literary style, like fashion, is historically contingent, one of his characters enlists earlier works of art: Just look at pictures that aren’t that old, painted, say, sixty years ago and see what was being worn by those of the fair sex belonging to prominent families or living at court. If a woman went out in public dressed like that now, the village idiots and street‐urchins would pelt her with rotten fruit. His interlocutor concurs, going on to describe an outfit strikingly similar to the Old Woman’s as the epitome of the outmoded and ridiculous: ‘Who would put up now with a decent married woman wearing those huge horns and pyramids and cones sticking out from the top of the head and having her brows and temples plucked so that nearly half her head is bald?’¹¹⁰ Erasmus’s verdict about such an outfit is unequivocal: ‘Everyone would laugh and boo’.

This dialogue is telling on several accounts. Firstly, it may provide an insight into Massys’ process of creation. It is likely, indeed, that the artist did look at ‘pictures that aren’t that old’ in preparation for these panels that deride fifteenth‐century courtly portraits: such works as the likeness of Isabella of Portugal by the workshop of Rogier van der Weyden (fig. 20); the crypto‐portrait of Agnès Sorel in Jean Fouquet’s Virgin and Child in the case of NG5769; and the profile portrait of Duke Philip the Good in the case of An Old Man.

Secondly, Erasmus’s passage allows us to get a sense of the derision such historical dress would have attracted by the 1510s. In a similar spirit, in his Treatise on Painting Leonardo cautioned against incorporating contemporary fashion in paintings so that ‘our successors will not be driven to laughter by the mad inventions of men’.¹¹¹ Departure from this rule was one of the principal comic devices of his visi mostruosi and it is no coincidence that he made the horned headdress the defining attribute of his grotesque old women. Similarly attuned to the semantics of historical costume, Massys too exploited the stigma then attached to fifteenth‐century dress both as a cutting allusion to his figure’s advanced age, and as an effective means to heighten her comic potential.¹¹²

Another breach of decorum makes Massys’ satirical purposes in these panels clear. In a carnivalesque inversion, the artist depicted the old woman on the right (the viewer’s left), the more prestigious side reserved for men in court etiquette, heraldry and by extension in double portraits of spouses.¹¹³ Although this departure from conventional placement might have simply resulted from the sources Massys had at hand when conceiving his pair (the left‐facing oil on paper and the right‐facing Leonardesque drawing), the resulting subversion of gender hierarchies would not have been lost on contemporary viewers. This locates NG5769 and its pendant within a rich strand of Northern secular imagery that revolved around the battle of the sexes and the power of women, in which alluring maidens, disobedient wives and fearsome witches all stood as potent incarnations of social disorder.¹¹⁴ In further transgressions, An Old Woman upends the ideals of youth, beauty and propriety that defined female portraiture at the time.¹¹⁵ Her ailing rosebud plays on such symbols of engagement as carnations and thistles, which were commonly held by prospective brides in double portraits. Roses, by contrast, had carnal associations and the woman’s proffering of it to her companion as a love token would have read as a shocking, unwomanly sexual overture, while the fact that the bud has yet to (or might never) bloom functions as a bruising commentary on her age.¹¹⁶ The old man raises his hand in response, a gesture that has been read alternatively as a salutation or rebuke and was perhaps meant to elicit debate among viewers.¹¹⁷ The painting takes aim at the presumed folly, vanity and lust of elderly women: a frequent target for Renaissance jokes that marshalled social stigma, physiological misconceptions around the menopause and a misogynistic tradition going back to classical writers such as Horace.¹¹⁸

Representative of this satirical vein, Erwin Panofsky first connected NG5769 to a vitriolic passage from Erasmus’s most famous work, Praise of Folly, a lashing critique of the excesses of his time, first published in 1511.¹¹⁹ After inveighing against the ‘amorous silliness’ of those ‘old drybones with a foot in the grave [who] take some tender young girl for a wife’, Erasmus continues: Yet it’s even more fun to see the old women who can scarcely carry their weight in years and look like corpses that seem to have risen from the dead. They still go around saying ‘Life is good’, still on heat, ‘longing for a mate’ as the Greeks say, and hiring some young Phaon by paying out large sums of money. They’re forever smearing their faces with make‐up, always looking in the mirror, and taking tweezers to their pubic hairs, exposing their sagging withered breasts and trying to rouse failing desire with their quavery whining voices, while they drink, dance among the girls, and scribble their little love letters.¹²⁰

This passage, now invariably cited in conjunction to the painting, rightly crystallised a view of NG5769 as a piece of social satire. Yet, rather than assuming a direct connection between the text and the painting, it seems more likely that both Massys and Erasmus were drawing on a common visual tradition already well established in the Netherlands by that date: the theme of unequal love, in which acquisitive youngsters are shown consorting with lecherous old people for their money.¹²¹ Massys’ paintings break with both this imagery and Erasmus’s text in that, far from looking to attract some youth, the Old Woman seeks someone her own age – a point revisited below.

While a variety of promiscuous older women populate the literature and visual arts of the period, evidence suggests that NG5769 depicts a specific folkloric type. Indeed, a similar figure reappears across a number of contemporary prints, carvings and manuscripts. These include a Florentine engraving by the Monogrammist SE (active 1480s) that depicts a Morris dance or moresca, a frenzied dance performed during carnival, civic festivals and as court entertainment (fig. 21).¹²² The choreography saw a group of young male performers strike acrobatic poses around a young woman – often played by a man – holding up a prize ring or apple.¹²³ In a grotesque parody of the moresca, the engraving swapped the athletic youths for bulky cavorters, the maiden for a hag, and the prize for a pig’s trotter and a skewer of sausages (phallic symbols from which she gained her nickname of ‘Sausage Woman’).¹²⁴ This figure is strikingly reminiscent of both Leonardo’s Windsor and Spencer drawings, and by extension of NG5769: she wears a similarly extravagant gown with a low neckline and a nearly identical horned headdress with a trellis, covered by a trailing veil pinned by a brooch. The Florentine engraving, which probably dates from the 1480s, has been connected to a series of drawings by Andrea del Verrocchio featuring grotesque dancers that may have been known to the young Leonardo.¹²⁵ As a type, the ‘Sausage Woman’ therefore appears to be a distant cousin, or rather an ancestor, of the ‘Ugly Duchess’.¹²⁶

Like a wider group of secular prints made in Florence around the same time, the moresca engraving has long been thought to reflect Netherlandish works of art present in Florence in the 1470s and 1480s.¹²⁷ The same Northern prototype probably served as the basis for a very similar composition, Morris Dancers (about 1470–1536) by the German artist Daniel Hopfer, in which the ‘Sausage Woman’ also features front and centre (fig. 22).¹²⁸ As first noted by Panofsky, the same figure pops up again in an English misericord dating from about 1447 in the parish church of St Laurence’s in Ludlow (fig. 23).¹²⁹ It shows a sneering old woman sporting an impressively large horned headdress adorned with a lattice of flowers and a flowing veil. She is flanked by two male supporters whose striding poses resemble the characteristic leaps of Morris dancers; it may be that here, once more, we are in the world of the moresca and in the presence of its mock May Queen.¹³⁰ The supporters’ attributes are hard to identify: a tabor, shield or mirror have been suggested for the figure on the right; the figure on the left may once have held a sword.¹³¹ It could be a variation on the moresca whereby the men fight off the old woman rather than court her. Like many such English carvings of the period, the misericord probably derives from a Netherlandish model, thereby confirming the Northern origins of this folkloric type. This figure seems to have been a versatile character: she reappears around 1496–8 under the guise of the queen of hell, Proserpina – described as bearing a ‘terrible face’ – in a miniature by Robinet Testard from Evrart de Conty’s Livre des échecs amoureux moralisés (fig. 13).¹³² With its central mouth of hell framing the chthonian gods and its two side openings that seem to lead backstage, the space recalls the set of a mystery play, linking this particular figure to the world of popular theatre and public performance.

Therefore, swinging like a pendulum across the Alps, a Northern folkloric type appears to have made its way to Florence by the 1470s before bouncing back to Antwerp in the 1510s via the intermediary of Leonardo. Today, only sparse echoes of this vernacular character survive, reverberating across a few disparate images. Unlike such figures as the hell raider ‘Dulle Griet’ (‘Mad Meg’), the fidgety housewife ‘Metz Unmuss’ (‘Metz the Restless’) or the fool Marcolf and his wife Polikana, the original name of the ‘Ugly Duchess’ in vernacular tradition – if she ever had any – has not come down to us. Yet there can be little doubt that Massys and his contemporaries would have recognised this folkloric figure when confronted with Leonardo’s design.¹³³ Perhaps it was Leonardo’s novel and ambitious take on a familiar type that drew Massys to this particular drawing. Massys’ acquaintance with a figure from the world of civic festivals may also have been conditioned by the fact that the Antwerp painters’ guild was associated with the city’s rhetorician’s chamber De Violieren, responsible for putting together farces and morality plays in the city throughout the year.¹³⁴ In NG5769, Massys therefore borrowed an established comic figure and dramatically refashioned her, jokingly elevating this obscene folkloric character to the rank of dignified portrait sitter.

In a visual culture that had so far confined the grotesque to the sidelines of representation, it is hard to overstate how novel this irruption of popular culture onto the rarefied scene of panel painting would have felt. It is worth noting, however, that at least one prominent precedent of a painting similarly blurring the boundaries between genre and portraiture is documented in the fifteenth century: a lost painting by Antonello da Messina depicting two elderly people laughing at each other, described in early modern sources as portraits – a tantalising record, given Antonello’s debt to Northern art.¹³⁵ Lurking behind these parodies seems to be a critique of the pretension inherent in sitting for any portrait.

Finally, an important feature of the ‘Sausage Woman/Ugly Duchess’ type was its gender ambivalence: as mentioned above, in carnival dances a man would perform the character. The related prints and paintings depicting this folkloric figure appear to have alluded to this by lending her masculine traits. Thus, the old woman in NG5769 has powerful facial features, a strong chin and jaw, and broad shoulders, elements that contradict her generous bosom. This observation by no means implies that this is the portrait of a man wearing women’s clothes. As discussed, NG5769 parades as a likeness but is in fact the depiction of an imaginary, vernacular figure channelled through Leonardo’s pen. Rather, what the sexual indeterminacy of the ‘Sausage Woman/Ugly Duchess’ type reveals, is that the ‘ugly’ was partly understood in the early modern period as whatever resisted and transgressed the sealed categories of male and female.¹³⁶

Patronage, Function and Date

Although nothing of their genesis and date is known, it is probably safe to assume that panels of this scale, quality and ambition were not made on speculation but resulted from a commission. This is all the more evident when comparing NG5769 to the several other grotesque heads that Massys painted over the course of his career.¹³⁷ Smaller, more loosely painted, and featuring characters from low society, they are straightforward tronies that do not create the fiction of portraiture the way NG5769 and its pendant do.¹³⁸ These more modest heads seem destined for the open market. By contrast, An Old Woman stands out, monumental, commanding and painted with consummate attention. As discussed above, the panel demonstrates Massys’ full mastery of his medium, with virtuoso techniques and myriad unflattering details captured with relish (see Technical Notes and Description above). Despite the lack of documents, it therefore seems worth asking what kind of patron such novel, brazenly indecorous yet surprisingly sophisticated paintings might have appealed to, and what kind of intellectual environment might have fostered such unusual creations.

Ever since the studies conducted by Georges Marlier and Erwin Panofsky, NG5769 and her companion have been interpreted in light of Massys’ documented relationship with Erasmus, whom he portrayed twice, first in a double portrait commissioned in 1517 with the Antwerp town clerk Pieter Gillis to be presented as a gift to their common friend Thomas More, and secondly in a portrait medal commissioned two years later.¹³⁹ As discussed above, Erasmus’s critique of lustful old women in Praise of Folly and dismissal of historical dress in The Ciceronian can be productively brought to bear with NG5769 (see A Comic Type). More than merely topical, however, the connection between NG5769 and Erasmus’s texts is also formal. In treating figures of fun with the serious trappings and refined execution of portraiture, Massys created a powerful tension between dignified form and obscene content. Such glorification of a subject deemed unworthy, known as ‘paradoxical encomium’ (‘paradoxical praise’), was a foundational rhetorical device of both ancient and early modern satire.¹⁴⁰ Massys’ association with De Violieren would explain his familiarity with paradoxical encomium, but importantly, and perhaps not coincidentally, it is also the rhetorical mode that entirely underpins Erasmus’s Praise of Folly.¹⁴¹ This was certainly not the only option available to Massys when engaging with the grotesque: the genre saw otherwise skilful artists such as Daniel Hopfer and Bartolomeo Passarotti adopting a coarse facture that mirrored their subject matter.¹⁴²

The care that Massys lavished on NG5769 naturally reflects his meticulous style but may also have to do with humanistic debates around the specific pictorial genre that he was parodying: portraiture. By that date, Netherlandish painters were renowned all over Europe for their ability to convey likeness through the most minute detail.¹⁴³ Erasmus famously derided such tireless imitation in The Ciceronian. In a passage believed to refer to his own numerous, lengthy sittings for Massys,¹⁴⁴ he told of an unnamed painter who, unable to capture the essence of his sitter (that is, his spirit), resorted to documenting the most anecdotal change in his appearance, leading to a never‐ending and pointless enterprise:A certain painter of this kind was recently a source of amusement to us. He had engaged to paint our friend, Murius, and since he could not paint the true form of the man, he looked about if he had anything noteworthy on his body or clothes. He began in the summer and had already for the most part finished the picture, had painted a ring which he was wearing, a purse and girdle, and had carefully copied the felt cap on his head. Then he noticed that on the finger of his left hand there was a scar; this he represented with studied care. Then on the right wrist he found a large swelling and painted that. On the left eyebrow he put some hairs in different directions. On the cheek too he put a scar, the mark of a wound. Time went on and he had many sittings. If, when he came, he saw that the beard had been shaved, he painted a new chin; again if he saw that the beard had grown out he changed the chin, because that pleased him more. Between times a slight fever seized Murius which, as is usual, left a sore on his lip; the painter portrayed this. At length winter came, another cap was put on; he changed the picture. Winter clothing of furs was put on, he painted a new dress. Cold changed the complexion and as usual shrunk the skin, he changed the entire skin. Murius caught a chill which affected his left eye and made his nose somewhat larger and very much redder, he painted a new eye and a new nose. If ever he saw him uncombed, he ruffled up the hair. If perchance Murius was sleeping while he painted, he represented him sleeping. If he had taken medicine which made him look and feel years older, he changed the face. If he could have painted the true and native form of the man, he would not have taken refuge in these incidental things.¹⁴⁵

In this passage, Erasmus turned the great strength of painting – its mimetic abilities – into a weakness, articulating a masterful indictment of the failure of images to convey reality in anything but a superficial way. In a manner reminiscent of Erasmus’s painter in The Ciceronian, in NG5769 Massys pushed the indexing of physical traits to the brink of absurdity. Deploying a dazzling range of techniques, he meticulously recorded the faintest irregularity, refusing to spare the viewer a single wrinkle, pimple or white hair. On this account, Max J. Friedländer condemned the painter’s facture in NG5769 as ‘punctilious’ and ‘pedantic’.¹⁴⁶ Yet Massys’ excessive, obsessive illusionism in NG5769 rather seems deliberate and pointed: in exploiting the possibilities of oil paint to produce a disturbingly vivid figure, the artist may have intended to proclaim the power of images in the face of humanistic critique. As a demonstration of wit, he did so by parodying the medium of painting itself, with its double‐edged ambition to create a convincing rendition of reality.¹⁴⁷

That a self‐conscious reflection on the power and pitfalls of mimetic representation is at work in NG5769 is confirmed by an anecdote that Erasmus included in his 1518 edition of the Adages, just a year after he was first portrayed by Massys. The Adages were an erudite and hugely influential compilation and explanation of ancient proverbs, first published in 1508 and continuously expanded until Erasmus’s death, which have been credited for introducing classical texts to a wide readership during the Renaissance.¹⁴⁸ In the 1518 edition, in order to define the ancient notion of ‘sardonic laughter’ Erasmus drew on a dictionary compiled by the Roman grammarian Festus, which recounted the demise of the ancient Greek painter Zeuxis, who, like Apelles, was a tutelary figure for Renaissance artists:¹⁴⁹ Erasmus writes that Zeuxis died of laughter while beholding the portrait of an old woman he had painted.¹⁵⁰ Although more obscure than other episodes of Zeuxis’ life, this story was included by Karel van Mander in his Schilder‐boeck (1604) and later inspired Rembrandt’s famous Self Portrait as Zeuxis (about 1668; Wallraf‐Richartz Museum & Fondation Corboud, Cologne, inv. WRM 2526), as well as a self portrait by his pupil Aert de Gelder (1685; Städel Museum, Frankfurt, inv. 1015) in which the old woman is shown holding a golden apple and thus impersonating the goddess of love Aphrodite – her self‐delusion akin to that of Massys’ old woman.¹⁵¹

Massys probably knew the tale of Zeuxis’ tragicomic end, as his work elsewhere certainly hints at his awareness of Apelles’ and Zeuxis’ other legendary feats of illusionism.¹⁵² In painting the fictive portrait of a risible old woman, Massys was therefore declaring himself Zeuxis’ successor, but not without irony, perhaps mocking, via this sinister tale, the pompous tendency of his learned contemporaries to claim the antique legacy by comparing modern painters to their ancient counterparts – exactly as Thomas More had done when, enthused by the double portraits of Erasmus and Gillis he had just received, he called Massys the ‘renovator of Ancient art (…) no less great than the grand Apelles’.¹⁵³ This antique precedent further endowed Massys’ obscene vernacular figure, the beleaguered hag from the world of carnival, with a reputable classical aura, collapsing high culture and low humour into one, in a manner typical of humanistic jest.

In a sort of anti‐praise of painting, the Zeuxis story offered a reflection – and perhaps a warning – on the power of images to create unsettling (and even lethal) illusions of reality, calling attention to Massys’/Zeuxis’ astonishing skills in the process.¹⁵⁴ Through the legend of Zeuxis’ artful demise, the portrayal of older women might have become during that time a measure of artistic achievement and a locus of competition for ambitious painters, for instance for Dürer and Giorgione around 1507 (fig. 25).¹⁵⁵ Unlike beauty, which is always constrained by strict norms, the possibilities offered by ugliness – and the shapes it can take – are infinite and unpredictable, thereby affording artists a privileged space for play, experimentation and demonstration.¹⁵⁶ This might provide a clue to NG5769’s original function: Massys may have conceived this work as a formidable showpiece intended to impress and amuse prospective clients.

Another credible hypothesis, given the classical reference, is that the panels were made for a humanist patron, possibly around the time Massys was involved with Erasmus, about 1517–19, which is also the moment when the Zeuxis anecdote first appeared in The Adages. That contravenes the traditional dating of NG5769 and its pendant to about 1513, based on the date inscribed on the preparatory oil on paper for An Old Man.¹⁵⁷ As discussed above (see Pendant), a recent conservation treatment has revealed the last numeral ‘3’ to be a later restoration necessitated after a tear in the paper rendered the original completely illegible. The oil on paper was transferred from panel to canvas after 1902, and the damage and subsequent treatment might have happened then.¹⁵⁸ In this scenario, the reconstruction would simply reproduce faithfully the lost original digit, as auction catalogues as early as 1791 already record the date to be 1513.¹⁵⁹ If the restoration is older than 1791, however, it opens the possibility for this work, and by extension NG5769, to have been created at a later date. Given the humanistic context laid out above, pushing the creation of NG5769 and its pendant towards the end of the decade, when Massys’ involvement with Erasmus and his acolytes was at its most intense, would be sensible. Rather than considering NG5769 as mere illustration of Erasmus’s satires, as has traditionally been the case, the work might best be seen as Massys’ playful if pointed response to the humanist’s critique of images. Such a dating also fits more appropriately within what is understood of Massys’ artistic development. The current assumption that An Old Woman and An Old Man predate The Money Changer and his Wife, signed and dated 1514, a genre scene that seems less resolved compositionally and more archaic stylistically, is difficult to receive. However, matters of dating in Massys’ oeuvre remain notoriously prickly: only five of his paintings bear a date and a few more can be dated thanks to documents;¹⁶⁰ his style is remarkably consistent throughout his two decades of activity and attempts at periodisation are perilous. Therefore, caution should remain and NG5769 and its pendant can provisionally be dated to 1513–19.

However one settles on the panels’ dates, placing the creation of NG5769 and its pendant within Massys’ humanist milieu raises the possibility of An Old Man – who unlike his partner is not a grotesque confection – being a satirical portrait, maybe as part of a humanistic inside joke: the cheeky likeness of a friend or the bruising portrayal of an intellectual rival. This would help explain the lingering incongruity of their mismatched appearance – grotesque versus conventionally old – and would fit neatly with Maximiliaan Martens and Jan Van der Stocks’s recent proposal that Massys’ Laughing Democritus (about 1525–30; Private collection) is a crypto‐portrait of Thomas More.¹⁶¹

Although conclusions must remain tentative, works like NG5769 and its pendant would undoubtedly have been warmly received within such circles. The poet and humanist Oswald Myconius tells us of Erasmus’s glee upon seeing Hans and Ambrosius Holbein’s lively pen illustrations in his copy of the Praise of Folly and Erasmus himself was an amateur caricaturist, as shown in the satirical self portraits that dot the margins of one of his manuscripts.¹⁶² Laurent Joubert further tells us of the element of stagecraft that sometimes accompanied the presentation of images such NG5769:when we are promised the sight of a beautiful young woman, and just as we are aroused, we are shown a wrinkled old lady with one eye, a runny nose, a thick and kinky beard and underslung buttocks, dirty, smelly, drooling, toothless, flat‐nosed, bandy‐legged, humpy, bumpy, stinking, twisted, filthy, knotty, full of lice and more deformed than ugliness itself. Here there is something truly to laugh at, seeing ourselves being made fun of in this way.¹⁶³ It is tantalising to imagine that An Old Woman may have been unveiled in such theatrical fashion, through the use of a curtain or the dramatic opening of the diptych it may have once formed with An Old Man, for instance.

What both Myconius’s and Joubert’s anecdotes reveal is that looking at images such as NG5769 was a profoundly social – and more precisely homosocial – business in the early sixteenth century. Massys’ pair was not conceived to hang silently on a wall. Rather, its ability to elicit the viewers’ astonishment, and afford them a chance to show off their learning and wit, would have been a measure of its success.¹⁶⁴ Tellingly, in Joubert’s account, it is the fooled viewer rather than the hag who is the ultimate object of laughter. Similarly, and despite the era’s pervading misogyny, recent studies have shown that deficient men rather than women were more often the butt of obscene jokes in the sixteenth century.¹⁶⁵ Likewise, Renaissance viewers may have laughed at Massys’ old woman, but perhaps also with her, at the pedantic‐looking old man who found himself matched with her, and at the societal norms, gender hierarchies and beauty standards she proudly defied – a breach that beholders were invited to partake in vicariously. In this freedom and irreverence perhaps lies one explanation for An Old Woman’s extraordinarily enduring appeal.

Notes