The Artist: For a biography of Dieric Bouts, see the Catalogue Entry for
Virgin and Child (
30.95.280).
The Painting and its Subject: The Nursing Madonna, or
Maria lactans, was one of the most popular subjects in Early Netherlandish painting, as well as another type, the so-called
Glykophilousa type (or Affectionate Virgin; see
30.95.280). The representation of the nursing Virgin originates with Early Christian Coptic Egyptian and Byzantine
Theotokos Galaktotrophousa, of which the iconographic origin may ultimately be traced back to Ptolemaic
Isis Lactans (see, for example,
45.4.4;
2021.41.39).[1] The
Maria Lactans became an especially favored subject in the late medieval Catholic world, on both sides of the Alps, in the wake of private devotional practices. Numerous examples from the Burgundian Netherlands still survive, which enables us to trace the genealogy of compositional types and variations since the generation of Robert Campin (ca. 1375–1444) and Jan van Eyck (d. 1441).[2]
Following the basic format common both for the
Maria Lactans and the
Glykophlousa, The Met’s painting shows the figures in a spatially ambiguous setting closed off by a cloth of honor that covers the entire background. Since the original frame is lost, it is uncertain whether The Met’s
Virgin and Child was originally conceived as an independent panel or as part of a devotional diptych paired with the portrait of a donor in prayer (see, for example, the Catalogue Entry for
14.40.626–27).
In comparison with the numerous surviving images of the
Maria lactans, the iconographic oddity of The Met panel is the red-beaded necklace worn by the Christ Child. The translucent red beads of equal size exclude the possibility that the necklace functions as a paternoster for prayer (as in
1975.1.130) or an apotropaic coral necklace (as in
1982.60.47). This type of purely decorative red necklace worn by the Child is not found in any other fifteenth-century Netherlandish paintings, whereas there are several fifteenth-century Italian examples (see
65.234). Close observation under the microscope suggested that the necklace of the Child is probably a later addition (see Technical Notes).
Furthermore, there remains a trace of a pearl fillet on the Virgin’s forehead which is now painted out yet can still be seen in the x-radiograph (see Technical Notes and fig. 1 above). The state before the overpainting is clearly visible in the photo reproduction of 1924 (fig. 2; Rieffel 1924, pp 56–57, pl. XV.). Although it was quite common to represent the Virgin with a headband in early Netherlandish painting, its arc was always depicted higher on the forehead (see, for example,
30.95.280). In contrast, in the 1924 photo, the arc circles downward with lavish pearls, reminiscent of contemporary Italian examples, such as the
Virgin and Child by Carlo Crivelli (
49.7.5). Max J. Friedländer wrote about our panel, on which there are “strange decorative chains on the Virgin's forehead and the child's neck” (Friedländer 1925, p. 127; Friedländer 1968, p. 73.), whereas the exhibition catalogue of 1929, a year after the painting’s purchase by Jules S. Bache from art dealer Joseph Duveen, shows The Met’s painting with the pearl fillet already painted out as it is now (Sperling 1929, no. 14). The motif was clearly overpainted due to Friedländer’s advice to the dealer. In the correspondence dated November 2, 1928, he stated “I believe that the beads on the forehead of the Madonna are not genuine and spoil the picture. Perhaps you will have a cautious restorer try to see whether this unsuitable ornament can be wiped out” (Duveen 1927; Friedländer 1937, p.91; Friedländer 1968, pp. 88–89).[3] In this correspondence, Friedländer only mentioned the fillet, which might explain the reason why the necklace remained untouched before the sale of the painting to Bache in 1928. Although The Met’s technical investigation provided neither a specific date for the addition of the necklace, nor concrete evidence showing the fillet as a later addition (see Technical Notes), the two decorative motifs may have been added together at some point by an owner with Italianate taste.
The Attribution and Date: In the previous scholarship (Schöne 1938, p. 215, no. 146; Campbell 1998, p. 56; Sprinson de Jesús 1998, pp. 230–31), The Met’s
Virgin and Child has generally been regarded as executed by a less-skilled assistant in the workshop, or a follower of Dieric Bouts, modelled on the
Virgin and Child at the National Gallery, London (fig. 3). In the London painting, unanimously considered as by Dieric Bouts himself (Campbell 1998, pp. 56–59), the
lactans group is located in a domestic setting with an open window view. By contrast, The Met panel shows the figures before a brocade cloth of honor covering the entire background. Without the motif of the cushion on the windowsill of the London panel, the Christ Child in The Met panel is shown in a spatially ambiguous position, as if his naked body is hovering in the air rather than sitting securely on the lap of the Virgin. This oddity is often regarded as the result of the omission of the cushion motif in the copying process, thereby being an indication that The Met Virgin was made as a reduced copy based on the London prototype by an artist of inferior skill (Sprinson de Jesús 1998, p. 230).[4] Still, The Met’s painting shares some prominent features with the London panel such as the idealized facial type of the Virgin, the positions of her hands, and the ‘pomegranate’-patterned embroidered cloth of honor (Monnas 2008, p. 355).
Nonetheless, The Met’s painting seems to be more of a hybrid of the London prototype and a Rogerian type. The lactating Virgin with an embroidered cloth of honor was influential already in the generation before Bouts, as seen in the Master of Flémalle’s
Madonna and Child, Van Eyck’s
Lucca Madonna, or Rogier’s
Saint Luke Drawing the Virgin (fig. 4). This iconographic combination of the
Maria lactans and the embroidered cloth of honor was adopted quickly into the traditional half-length format of the Virgin and Child image, especially in the circle of Rogier.[5] In Rogerian lactating images, the Christ Child’s position is often ambiguous in the pictorial space, including the position of a cushion on Mary’s lap.[6] These panels also feature, in common with The Met’s Virgin, the white headscarf, one end of which runs over the hair and the bosom, together with the breastfeeding arm in an embroidered sleeve of her gown. These details are not found in the London prototype, indicating that The Met’s panel was modelled also on this type of Rogerian nursing image, which had probably originated from Rogier’s Boston
Saint Luke. Bouts’s contact with Rogier is open to question, unlike the case of Hans Memling, who shows, especially in his early works, a close stylistic affinity with the underdrawing techniques of Rogier (see
17.190.7).
The hybrid origin does not, however, designate The Met Virgin a pastiche status. For example, the resemblance of the Virgin’s face to the London prototype deserves scrutiny. When the two panels are overlapped with each other, despite the different measurements of the two panels, it becomes clear that the Virgins’ heads are painted roughly in the same placement. Some features, such as the proper right eye, bridge of the nose, and mouth, match well, but not exactly, in size and position with each other (fig. 5), whereas the other parts do not overlap, being in different sizes and positions. The infrared reflectography of The Met panel reveals that the underdrawing was likely to have been executed freehand, showing some spontaneity in the contour lines (fig. 6; see Technical Notes). It appears different from the rigid and schematic underdrawing that is typical of a transfer (in contrast to, for example,
71.156–57),[7] suggesting that the near-identical facial features in The Met’s painting were the result of freehand copying based on the London prototype or a common pattern.
Although condition issues make it difficult to assess the painting technique, especially in the flesh tones (see Technical Notes),[8] The Met’s panel seems to lack the attention to detail that is typical in the autograph works of Bouts. For example, the comparison of the cloth of honor demonstrates different hands in the two panels. The London panel exhibits meticulous cross-hatching in varied colors in the embroidery on the cloth, whereas our panel mostly shows parallel hatching in yellow (fig. 7). The latter, therefore, fails to convey the detailed effects of light reflecting on the embroidery threads. These observations provide additional evidence that The Met’s
Virgin and Child was not executed by Dieric Bouts himself.
For further discussions on the attribution of this panel, it is helpful to study the x-radiography of The Met’s panel. The autograph works of Bouts, such as The Met’s
30.95.280, exhibit his careful, selective build-up of lead white paint to establish the volume of forms, especially in the flesh tones,[9] and the x-radiograph of The Met’s
lactans panel shows, under the abraded and retouched paint surface, a similar distribution of lead white (see fig. 1 and Technical Notes). This, together with the observations above, indeed suggests that The Met Virgin was executed in the workshop of Dieric Bouts, although it is impossible to say if this was before or after the death of the master in 1475. In the Burgundian Netherlands, many
Maria lactans images were produced in the second half of the fifteenth century. This small panel eloquently demonstrates the enterprise of the Bouts workshop in reaction to the popular demand for this intimate devotional image in the time of the Devotio Moderna.
Sumihiro Oki 2022
[1] Sabrine Higgins, “Divine Mothers: The Influence of Isis on the Virgin Mary in Egyptian Lactans-Iconography.”
Journal of the Canadian Society for Coptic Studies 3–4 (2012), pp. 71–90.
[2] Dirk de Vos, “De Madonna-en-Kindtypologie bij Rogier van der Weyden en enkele minder gekende Flemalleske voorlopers.”
Jahrbuch der Berliner Museen 13 (1971), pp. 60–161.
[3] Duveen 1927. In the correspondence, November 8, 1928, Friedländer wrote, “I studied this picture [49.7.18] very carefully some weeks ago at München and am quite sure now, that it is characteristic work by Dieric Bouts.” It is noteworthy that this comment postdates the publication of Volume III of
Die Altniederländische Malerei, in which the author attributes The Met panel more negatively as a close replica of a copy of the
Virgin and Child, the autograph work now in the National Gallery in London. See Friedländer 1925, p. 127; Friedländer 1968, p. 73. Later, he changed the attribution of The Met’s painting to “an original by Dieric Bouts” since the “disfiguring restorative work has been removed.” See Friedländer 1937, p. 91; Friedländer 1968, pp. 88–89.
[4] Sprinson de Jesús 1998, p. 230. The composition of the London panel was certainly influential. Many variations still survive including the German example at The Met (
22.96), or even the
Madonna and Child by Giovanni Bellini shows a distant echo (
08.183.1).
[5] See, De Vos 1971.
[6] See, for example: Rogier van der Weyden,
Virgin and Child, left wing of the
Diptych of Jean Gros, Musée des Beaux-Arts, Tournai, inv. 481; Workshop of Rogier van der Weyden,
Virgin and Child, Art Institute of Chicago, inv. 1933.1052; Follower of Rogier van der Weyden (Master of the Saint Ursula Legend Group),
Virgin and Child (The Met,
17.190.16.
[7] Maryan W. Ainsworth,
Facsimile in Early Netherlandish Painting: Dieric Bouts's "Virgin and Child", Exh. cat., New York, 1993, pp. 9–10. The matching of the eye, nose, and mouth with the London panel is also found in the
Maria lactans panel in the Städel Museum in Frankfurt (inv. 1217), of which x-ray shows the typical Boutsian handling of the lead white in the flesh tones. See, Jochen Sander,
Niederländische Gemälde im Städel,1400–1550, Mainz 1993, pp. 47–58.
[8] Note that The Met’s panel is quite abraded and has a discolored varnish layer. The current matte appearance does not necessarily reflect the original appearance of the hanging (see Technical Notes).
[9] Ainsworth 1993. esp. pp. 12–14. Her attributions are unanimously accepted. For example, see Périer-d'Ieteren 2006, pp. 126–33.
