The painting belongs to an uncommon genre of Byzantine origin of devotional icons painted on both sides. It is a simplified version of the portable diptych in which, as in the example discussed here, the Madonna and Child was usually represented on the obverse, and Christ on the Cross on the reverse. The peculiarity here, however, is the presentation of the Virgin according to the iconographic type of the Madonna of Humility: Mary is humbly seated on the ground instead of on a throne. Yet at the same time she is venerated as Queen of Heaven by two angels who flank her in flight, and blessed from above by the half-length figure of Christ, who appears in a trefoil, surrounded by seraphim. On the reverse of the panel is an isolated image of Christ on the Cross, unattended by the usual figures of mourners, soldiers, or onlookers who allude to the event of the Crucifixion. That the image was intended as a panel for private devotion is underlined by the presence of the minuscule figure of the female donor kneeling in front of the Madonna to the right; some scholars have identified her as a nun.
At the time of the panel’s first emergence in Florence in the 1920s, art historians expressed rather disparate views about it. Roberto Longhi, in a manuscript expertise probably dating to the years 1925 – 1930, considered the image of the Madonna likely a work of a close follower of Simone Martini, identifiable with Lippo Memmi or with Simone’s brother Donato, whereas that of Christ on the Cross on the reverse seemed to him a later addition by a painter close to Paolo di Giovanni Fei dating to the final years of the Trecento. In 1934, similar expertises were sought from other leading art historians of the time. F. Mason Perkins, followed by Giuseppe Fiocco and Wilhelm Suida, attributed it to Andrea di Bartolo; Bernard Berenson also accepted this attribution. For his part Adolfo Venturi came to the conclusion that it was an autograph work of Simone Martini, while Raimond van Marle spoke of a “close follower” of Simone, probably identifiable with his brother Donato. The catalogs of the National Gallery of Art (1941, 1959) accepted Longhi’s proposal. In a polemical article (unpublished), Richard Offner contradicted this, preferring to leave the panel in anonymity. George Martin Richter (1941) also rejected Lippo Memmi’s hand, arguing for an attribution to Andrea di Vanni in a youthful phase, when he was still working in Lippo’s shop. Charles Seymour and Hanns Swarzenski (1946) also placed the attribution to Lippo in doubt. Nonetheless, Andrea di Bartolo’s responsibility for the execution of both sides of the painting was gradually recognized, beginning with Millard Meiss (1936, 1951), followed by the catalogs of the Gallery (1965, 1985) and Fern Rusk Shapley (1966). All the more recent contributions to the literature accepted this attribution, with the exception of Gaudenz Freuler (2009), who considered the image of the Madonna of Humility as executed by Bartolo di Fredi, possibly assisted by his son Andrea di Bartolo.
The date of the painting, however, has given rise to considerable divergence of opinion. It was thought to have been executed by Memmi as early as c. 1330–1340 in the first catalog of the Gallery (1941). But in the later catalogs of 1965 and 1975 and in Shapley 1966 it was dated to c. 1415. In the meantime Meiss (1951) supported a dating to c. 1400, while Hendrik W. van Os (1969, 1974) offered the view that it must have been painted before the end of the fourteenth century. Shapley (1979), returning to the question, concluded that “the date may be in the 1380s.” For his part Creighton E. Gilbert (1984) seems to have favored a later dating: around 1394. Freuler (1987), van Os (1989, 1990), and Daniele Benati (1999) accepted this hypothesis, as well as Gilbert’s suggestion that the painting was commissioned by the Dominicans of the monastery of Corpus Domini in Venice, consecrated in 1394, and connected its execution to that year. In 1994, however, van Os seemed to have abandoned this position, preferring a dating to c. 1415. Bearing in mind that the connection of the panel with the Dominican nuns of Venice is purely conjectural and seems contradicted by the fact that the donor in our panel is not dressed in the habit of that order, the terminus a quo of 1394 should now be excluded from discussion of our panel.
Some help in establishing the panel’s date can, however, be derived from an analysis of its punched decoration, undertaken by Mojmir S. Frinta (1998). This scholar identified the presence in our panel of some punches already used in paintings produced in the shop of Bartolo di Fredi, as well as in youthful paintings by Andrea: I refer in particular to the polyptych in the museum at Buonconvento, probably dating to 1397, and various paintings that art historians have unanimously assigned to Andrea’s initial phase. But especially significant to me are the stylistic affinities with works by Bartolo dating to the mid-1380s or shortly after, in which signs of Andrea’s assistance can, I believe, be glimpsed: Adoration of the Magi, now in the Pinacoteca Nazionale in Siena, to which the Adoration of the Cross now divided between the museums of Altenburg and Charlottesville formerly belonged, and Massacre of the Innocents in the Walters Art Museum in Baltimore, probably part of the polyptych executed by Bartolo’s shop for the church of Sant’Agostino at San Gimignano in 1388. In these paintings the harsher features of Bartolo’s style are softened, the swirling calligraphy is attenuated, and the emotional tensions give way to a rather somnolent tranquility which — in conjunction with the close morphological affinities — suggests Andrea’s participation. The Christ on the Cross of the reverse, though more hasty in execution, recalls the similar passage in the Adoration of the Cross, no. 50 in the Lindenau-Museum in Altenburg, of which it seems indeed a simplified version. The Madonna of Humility of the obverse, on the other hand, with its minute, exquisitely chased detailing, reveals close affinities especially with the figures of the mothers in the Massacre in Baltimore, suggesting it was a work realized by Andrea, when he was still working in his father’s shop.
It may be added that the composition of the Madonna of Humility evidently enjoyed considerable success. The artist replicated it many times. The example in the Gallery is likely to be one of the earliest, together with the signed version published by Berenson, its whereabouts now unknown. In the later versions the figure of Mary seems to expand to fill the painted surface, which is enclosed within an arch decorated on the inside with cusping and sometimes with figures of the Angel of the Annunciation and the Virgin Annunciate placed in medallions in the two upper corners. At the same time, the gold-tooled carpet that covers the floor is replaced by a flowering meadow; the design becomes more simplified, while the line of the hem of the Virgin’s cloak is wavier; and an increased number of angels surround the protagonists. Such late Gothic developments, however, are characteristic of far later phases in the artist’s career. Here he is still strongly influenced by the figurative formulae of his father. This is evident both in the painting’s ornamental decoration and in its stylistic features, both indebted to Trecento models.
The support (contrary to Shapley 1966, 1979) is a single piece of wood, with a vertical grain and about 1 cm thick, painted on both sides. The panel has a slight convex warp relative to the obverse. It has been cut down along its upper edge. The top of the original gable is truncated, and in order to make the outer shape rectangular, triangular insets were added on both sides (approximately 1.5 cm on the top and 3 cm on the vertical edge on the left, and 1 cm along the top and 2 cm on the vertical edge on the right, as seen from the front). The original engaged frame has been lost and replaced by a modern one.
The painting was executed on the usual gesso ground, over which a thin red bole was applied in the gilded areas of the obverse. The reverse was silver gilt. An old photograph shows that the painted surface on the obverse ended approximately 1 cm from the lateral and lower edges and that the unpainted area at the sides of the gable was at that time gessoed. A later photo, probably from c. 1940/1941, proves that in the meantime the upper corners had been regilded. On the reverse the painted area extends to the edges of the panel. The x-radiograph shows impressions of three roundels on each side of the gable, the upper pair cropped in half at the top of the panel. This is evidence that the panel functioned as a reliquary, containing in its now lost engaged frame seven circular cavities (allowing for one cut off at the top of the gable) to receive relics. No underdrawing was found during an infrared examination (Vidicon), but incised lines mark the placement of the principal figures. A green undermodeling is visible beneath the flesh of the figures on both the obverse and the reverse. The painted surface of the obverse is fairly well preserved, with some inpainting in the donor’s robe; some small, scattered paint losses; and some abrasion in the Madonna’s robe. On the reverse the silver gilded area is heavily rubbed and the image itself is damaged by scratches, paint losses, and wormholes.
