Herodias with the Head of St. John the Baptist (‘Raczynski Herodias’)

Titian’s earliest portable treatment of Judith is the painting of 1515-16 in the Doria Pamphili collection; it was no doubt presented to Alfonso I during Titian’s stay in the Duchy of Ferrara in the early months of 1516 (fig.1). Copies exist but, it seems, no autograph variants. [2] The painting is widely called Salomebecause the severed head rests upon a trencher – which could, of course, serve as a symbolic halo – whereas Holofernes’ head is usually deposited in a sack or basket. But trenchers had already been transferred to paintings that unequivocally depict Judith, by Botticelli, Ghirlandaio, and Michelangelo, among others – and such transfers were to continue. What can only be Titian’s painting is described in the Este inventories in 1530 as a Judith and this identification fits the moral and physical demeanour of the two women. [3] The leading lady’s déshabille and restrained clothing is appropriate for Judith rather than Salome, usually lavishly dressed, as is her pensive expression; so is her age: she is the young widow of the Bible, not the adolescent cat’s paw who cavorted for Herod. And her maidservant, Abra, here characterised as a girl rather than – as is more usual – an older woman, looks up at her with admiration, not complicity or horror. [4] The action takes place at dawn, partly out of doors, not in a palace, as the two women make their escape from the Assyrian bivouac. Nevertheless, the absence of a sword and the inclusion of a trencher introduce an ambiguity which will continue to the end of Titian’s career. A simple halo above the severed head would have been sufficient to anchor the subject as a Salome but there is no halo on any of Titian’s severed heads; nor do any of them resemble at all closely those of his Baptists; unlike, say, Solario and Luini, Titian never ennobles a severed head.

Half-length treatments of Judith were produced by several Venetian painters in the first two decades of the 16^th century. The earliest is probably the picture dated 1510 in the National Gallery, painted by Sebastiano del Piombo. It is usually identified as a Salome because the severed head rests upon a trencher, but the composition employs a portrait formula to bring the viewer into direct contact with the heroine, whose firm, strong and unadorned characterisation identifies her as Judith.[5] A somewhat later composition by Catena once existed in two versions of which only that in Galleria Querini-Stampalia is known today.[6] It too shows Judith in a portrait-like pose, holding a sword with Holofernes’ severed head placed on a parapet before of her; it displays an obvious debt to the Self-Portrait as David by Giorgione, with whom Catena for a time shared a studio.

A different approach was taken by Cariani who, in a painting generally dated in the later 1510s in a private collection, placed Judith and Abra together, in the open but not obviously in flight; Judith holds a carving-knife in her right hand and Holofernes’ head in her left, while Abra claps her left hand over his mouth, as though still smothering his screams. [7]  Venetian treatments of Salome, however, are virtually non-existent, except in narrative paintings: she was rather a speciality of Milanese artists. The most famous treatments of the subject, at half-length, by Andrea Solario and Bernardino Luini, show Salome receiving the Baptist’s head from the executioner –sometimes reduced just to a muscular arm suspending John’s head above Salome’s cup – and their somewhat perverse psychology strongly suggests the inspiration of Leonardo. But Luini also showed Salome looking directly at the viewer, holding out the trencher with the Baptist’s head and accompanied by a shadowy executioner on the left, in a painting now in the Kunsthistorisches Museum, Vienna, which seems to exist in only one version (fig.2). [8]

So far as we know Titian represented neither Judith nor Salome again for some years and when he returned to the subject(s), it was with a new composition in which the protagonist, whichever role she is playing, gazes directly at the viewer and presents him or her with a head on a trencher or in a sack, held near the lower edge of the picture. It seems inescapable that Titian was aware of Luini’s Vienna prototype and that it is another instance in his work of inspiration found in Milanese painting.  Of course, Titian had frequently included in his compositions figures looking outwards but this seems to have been a new departure for him. Its effect is ambiguous: on the one hand it might involve the viewer as a participant in the action, enrolling him or her as the recipient of the head; on the other it can be seen as a distancing device which reduces the status of the action to that of an attribute of the actress playing the role. The latter is the kind of conception with which we are familiar at a later date: Mr or Mrs or Miss X as Y or Z, but it was not adopted by Titian for other subjects. [9]

The earliest painting of this type—or the earliest of which visual record survives—was a now-lost canvas once in the collection of the Archduke Leopold William. [10] While it does not appear in any of David Teniers’ paintings of the Archduke’s Gallery we have his small painted copy, which records the picture’s colours, as well as Lucas Vorsterman II’s reversed engraving after it (figs. 3 and 4). [11] The action is set in an interior and concerns three figures; the leading lady displays on a trencher the severed head. At the left is a woman, whose age is hard to judge, clad in a yellow cloak which she lifts from her face with one hand below and one above her head. Lower right is a black page-boy for whom there is no warrant in the Book of Judith, or in visual tradition. However, Mantegna had represented Abra as a black woman and perhaps this prompted Titian. [12]  Titian placed black pages in comparable positions in the Laura dei Dianti of ca. 1524 and in the Fabrizio Salvaresio of 1557 and he liked to contrast skin colours; he did so again, with a black woman, in his Diana and Actaeon. This boy, who is repeated in later versions and variants, may have been an insertion without specific meaning; here he is little more than a bystander.

Judging from Teniers’ copy the Archduke’s painting was a finished work, and the figure-type and costume, notably the ruff, suggest a date in the early to mid-1540s. [13] If this is correct, Titian soon returned to this composition. In 1548 he undertook, at Charles V’s request, a posthumous portrait of Isabella of Portugal (117 x 98 cm.). [14] He painted Isabella not on a fresh canvas but over a variant of Leopold William’s painting (figs. 5 and 6). The main head is angled more pronouncedly, her features seem a little pulpier – conforming to the fuller female types that Titian came to employ towards 1550, but the arrangement is the same. That Titian submerged this lay-in implies that another was available and, indeed, this revised type is known in two variants. One, the main focus of this essay, is the ex-Raczynski canvas, now privately owned in New York (fig.7). The other is in the Museum of Western Art in Tokyo (fig. 8). Both paintings, which are to a lesser or greater extent unfinished, are generally dated by critics to Titian’s later years – but this judgement, while reasonable for the Tokyo canvas, which shows the ‘broken’ all-over approach of Titian ca. 1570, needs to be nuanced for the ex-Raczynski canvas, which was probably laid in  in the mid-to-late 1540s at the same time as the painting beneath Isabella. Its surface is not equally unfinished: the bodice is quite smoothly painted, as are her arms.  The head and the platter, however, are treated much more in Titian’s later manner as are the page boy and the woman on the left. It is evident that the picture remained as a work in progress in a Titian’s studio for many years, and that he never fully resolved what he wished to do with it: the heroine’s hands, for example, were left undefined.

The ex-Raczynski and Tokyo paintings are related to a third, now in Detroit (fig. 9). Containing only two figures, it unambiguously represents Judith (112 x 93), shown holding a sword and dropping into a sack the head of Holofernes. The head is large enough to suggest that of Goliath and Titian was no doubt paralleling Judith with David. [15] Abra’s role is now taken by the black boy who now holds open the sack – the single painting in the group in which he plays an active part. Judith, clad in a luxurious shift appropriate to attract the Assyrian general, is minimally bejewelled and wears only a single string of pearls; her forearms are bared, a wise precaution since, as we know from Caravaggio and Artemisia, severing heads is messy. But bared forearms are a common feature of all the paintings under discussion and not necessarily specific to Judith. X-ray examination establishes that this Judith is also a superimposition, upon a portrait which seems to be of an ecclesiastic (fig. 10), with a mitre, not of Charles V, so unconnected with the portrait of Isabella, which covers the related composition.  But there is no evidence from the x-ray that the composition of the Detroit Judith was modified in any significant way while Titian was working on it, so it is possible that it followed an earlier version, now lost. It is taken further than the other two paintings and is more homogenous in surface effect; it is generally and probably correctly dated ca.1570.

The ex-Raczynski and Tokyo canvases follow the Leopold William and ‘sub-Isabella’ prototypes, but are closer to the latter in the main figure’s fuller forms; the figures are about the same size and the black page is now a little more involved, but the former is more finished. It was first published in 1960 by Antonio Morassi when it was in a private collection in Zurich. [16]  With a label on the back in Polish giving the Raczynski name, it seems first to be recorded in that family’s possession in the early 19^thcentury. The Tokyo painting (oil on canvas; 90 x 82.5 cm.),  is more obviously unfinished and, as remarked above, later in style; it was probably laid in from the ex-Raczynski canvas: x-ray reveals no significant pentimenti and no other underlying composition. The protagonist’s dress is simultaneously more seductive and lavish and her movement more vigorous. The smaller format and tighter framing give it greater immediacy and it is notable that the trencher and head are placed higher in the picture surface and thrust forward more emphatically. The Tokyo version has a longer history.[17]  Once owned by Charles I and bearing his brand, it was offered in the commonwealth sale of 1650, at the high price of £150 but was withheld by Oliver Cromwell. Recovered in 1660 it remained in the Royal Collection until 1732/36, after which all trace of it was lost until its reappearance in 1994. At an unknown date after 1667 it was enlarged to the same dimensions (124 x 99 cm) as, and no doubt in conjunction with, a Type I Magdalen by Titian or from his studio, still in the Royal Collection, which was expanded from 93 x 78 to 125 x 99 cm.; [18] presumably they were displayed together but the Magdalen, a finished picture, was not re-worked. The enlargement and over-painting of the Tokyo picture have now been removed.

The four paintings – ‘Leopold William’, ‘sub-Prado’, ‘ex-Raczynski’, ‘Tokyo’ – contain the same three figures, but in each the head on the trencher is displayed at a different angle. But whose head is it: the Baptist’s or Holofernes’? The issue of subject has been deferred hitherto but now becomes pressing. Do all four represent the same event? Or did Titian employ the same composition for different ones? Can their histories enlighten us? There is, of course, none for the sub-Prado canvas and no commentary on the ex-Raczynski painting earlier than Morassi’s publication has yet been found, except the versoinscription which identifies the subject as Judith and Holofernes and which must shortly precede 1960. But there are early, if laconic, interpretations of the other two. Leopold William’s canvas was successively described in the Della Nave Inventory as Judith or perhaps Salome; in the Hamilton inventory as, simply, Salome, and then, in that of Leopold William, as Herodias; in the Prodromos of 1735, pl.18, in which it became a Palma, it was again called Salome. And when Charles I’s painting was put up for sale in 1650 it was as ‘Herod holding St John’s head in a platter by Tytsyan’ – Herod, obviously, Herodias abbreviated. But its reservation by the Lord Protector is directly relevant to its interpretation: alert to examples of divine authorisation for a beheading, Cromwell must have believed the painting to represent Judith. Once it had regained the Royal collection, the 1650 description was repeated, slightly amplified: ‘A Herodias with A John Baptist head in A platter and her maide by her’. Thereafter it was recorded consistently as Herodias until, in 1732/36, it once again became Judith; it is worth underlining that it was never called Salome.

When we first hear of them, therefore, ca. 1640, the subject-matter of these two very similar paintings was uncertain, varying between one heroine – Judith – and two villainesses – Salome or Herodias. How can we identify them? In Leopold William’s picture the sober demeanour of the central figure and as far as one can make out, her expression, suggests Judith and the woman on the left would therefore be Abra, although her gesture is hard to interpret: for what it is worth, Abra is also dressed in yellow by Lotto as well as by Botticelli and Michelangelo. If this is so, the moment shown would presumably be Judith’s display of Holofernes’ head after her return to Bethulia; a subject represented in sculpture by Lorenzo Ghiberti and, more relevantly for Venice, by Tullio Lombardo on the Vendramin Tomb. Yet the arrangement could also suggest Salome or Herodias, offering the Baptist’s head to Herod. The plainness of the costume, of course, might seem to favour Judith, shown soberly clad by such as Cariani and Catena. [19] But Judith describes dressing herself in her gladdest clothes and jewels in the book of Judith and Veronese and his followers show her richly clad. In short an elaborate costume might be worn either by Salome or Judith; but Salome is less likely to have a plain one.

The ex-Raczynski canvas is the closest to Leopold William’s version in colour and form but the central woman seems somewhat older and more matronly, although the re-working of her face compromises our reading of her age. When he laid in the composition, roughly contemporaneously with the lay-in beneath Isabella as suggested above, Titian probably intended it to represent Judith. For, as Dianne Modestini has observed, a little below the second woman’s head is the face of an older woman, who bends forward to look down at the severed head. Such a positioning and type suggest Abra. Titian then moved this head higher up the picture-surface to its present position. It is thus possible that the ex-Raczynski lay-in preceded the composition seen in Leopold William’s painting, for the lower placing of the head appears in no other version and could in principle have been a first attempt. But on balance it seems more probable that it was a speculative revision of the Leopold William arrangement, with which Titian decided not to proceed.

Titian’s latest intentions for the ex-Raczynski canvas are clarified by the male head inserted on the right; this, although imprecisely defined, is sombre and seems to be elderly; it can only be that of Herod. Therefore, what Titian is now showing is Herodias displaying the head of the Baptist in triumph, having taken it from her daughter Salome, who would be the veiled woman at the left and who, all viewers seem to agree, is the younger of the two. Herod’s introduction turns the picture into what Freud called ‘a family romance’ and naturally focuses our attention on the personality of Herodias. There may be a reflection of the perverse psychology that so fascinated Leonardo and his followers, which would be taken up again at the end of the century in Milan, notably by Francesco Cairo.

In the Tokyo painting, however, which, we may remember, was called Herodias but which Cromwell must have interpreted as Judith, it does seem more likely that the protagonist is Salome. While her face too has been re-worked, she is more youthful than the woman in the ex-Raczynski canvas, and her lavish jewellery, the chain around her neck and jewels and pearls in on her shoulder, her necklace and her hair, increase her erotic allure, as does her décolleté. It is also true that alone of these pictures, the head in the trencher bears at least a slight resemblance to that of the Baptist in other paintings by Titian. In this case the woman on the left would be Herodias and she does, indeed, seem somewhat older than the same figure in the ex-Raczynski version. But, it must be stressed, these observations do not constitute proof of the subject, and while it seems reasonable to question Cromwell’s judgement and call the Tokyo painting Salome any interpretation must come with another caveat: that until we have a description of the subject dating from Titian’s lifetime, any interpretation can only be provisional. [20]

Paul Joannides

notes

Updated in November 2018