Bartolo di Fredi, Virgin Annunciate

on the Virgin’s book, ECCE VIRGO CONCIPIET E PARIET FILIUM ET VO[CABITUR NOMEN EIUS EMMANUEL] (Isaiah 7:14: “Behold a virgin shall conceive and bear a son [and his name shall be called Emmanuel]”)

Charles Butler (1821–1910), Warren Wood, Hatfield, England (not Mass., as asserted in Seymour 1970); Robert Langton Douglas (1864–1951), London, by 1912; Dan Fellows Platt (1873–1937), Englewood, N.J.; Edward Hutton (1875–1969), London, by 1923; Maitland Fuller Griggs (1872–1943), New York, 1923

The panel support, of a vertical wood grain, retains its original thickness of 2.5 centimeters. The beveled lateral frame moldings were originally silver gilt, but at some modern date, they and most of the other frame moldings and crockets were covered with a gold paint that was not removed in the radical cleaning of 1958. The reverse of the panel had at one point been reinforced by having glued onto it a square oak support of two vertical and three horizontal slats; the uppermost horizontal slat remains attached. Tension from this support may have been responsible for a small vertical split in the panel passing through the nose and left eye of the Virgin; this has generated very little paint loss. A larger split, running from her throat through her right hand and the cuff of her left wrist, was provoked by two nails that originally secured a vertical batten on the reverse. The clinched ends of these nails are visible where they have caused areas of total loss directly beneath the Virgin’s right hand and in the folds of the blue robe beneath her left wrist. The flesh tones of the Virgin’s face and hands have been abraded to the terra verde underpaint in areas of shadow but retain their highlights of pink and white; the drawing of the features is unimpaired. Extensive flaking losses and abrasion interrupt the blue of the Virgin’s robe, most notably in the area of her left shoulder and along her left arm, while her red dress is disfigured by deeper gouges and abrasion along the cupped edges of the (predominantly horizontal) craquelure. The book held in the Virgin’s left hand with the inscription is largely undamaged. The gold ground is irregularly abraded, exposing broad areas of bolus preparation, and three large gouges at the left edge have removed everything above the gesso layer. At present, the paint surface is coated with an opaque synthetic varnish, further dulling the palette and exaggerating the effects of solvent damage.

Although the first recorded reference to the Virgin Annunciate as a work by Bartolo di Fredi appears to be Richard Offner’s manuscript opinion of 1924 (preserved in the Frick Art Reference Library, New York), the painting had been sold the preceding year by Edward Hutton to Maitland Griggs with that name already attached to it. According to correspondence in the Yale University Art Gallery’s archives, Hutton urged Griggs to consult with Offner for confirmation of the attribution but not to contact Robert Langton Douglas or F. Mason Perkins—the then-acknowledged experts on Sienese painting—as either would likely demand a commission from Hutton for their opinion. Bernard Berenson listed the painting as by Paolo di Giovanni Fei and as in the Platt collection, in Englewood, New Jersey, even though it had by then belonged to Griggs for nearly a decade.¹ Surprisingly, the first published identification of the painting as by Bartolo di Fredi occurs in Berenson’s Central Italian lists of 1968.² Charles Seymour, Jr., inexplicably qualified the painting as “attributed to Bartolo di Fredi” and equally inexplicably suggested it could have been the right wing of a diptych, notwithstanding its original, engaged chamfered moldings and the evidence of two nails securing a vertical batten to its reverse.³ He revised this assessment in 1972 on the advice of his former students, Michael Mallory and Gordon Moran, who that same year published the painting as a lateral pinnacle from an altarpiece but also as probably executed with workshop assistance.⁴ Discussion of the painting since then has focused entirely on differing proposals for identifying the altarpiece of which it might have formed part. For Keith Christiansen, followed by Patricia Harpring and, apparently, Wolfgang Loseries, this would have been the altarpiece of which the Adoration of the Shepherds now in the Cloisters Collection at the Metropolitan Museum of Art, New York (fig. 1), once formed the center panel.⁵ For Mallory and Moran, it would more likely have been the triptych by Bartolo di Fredi currently in the Museo Civico in Lucignano (fig. 2). Gaudenz Freuler, in his comprehensive monograph on the artist, rejected both of these suggestions—without, however, proposing an alternative.⁶

A consensus of scholars identifies the Cloisters Adoration of the Shepherds with an altarpiece recorded by Ettore Romagnoli (ca. 1835) in the church of San Domenico at San Gimignano bearing the signature of Bartolo di Fredi and the date 1374. Romagnoli described the painting as accompanied by figures of the four Evangelists, the Annunciation, the Baptism of Christ, and the Coronation of the Virgin. Several authors, as has been mentioned, assumed that the Yale Annunciate could be one of these ancillary panels. Freuler proposed identifying two of the lateral full-length standing saints with panels of Saints John the Baptist and John the Evangelist now in the Alana Collection, Newark, Delaware, neither of which is large enough, however, to have accommodated the Yale panel as a pinnacle above it. Freuler’s reconstruction is questionable on a number of counts. First, although Romagnoli did mention a Baptism of Christ, he did not mention a figure of the Baptist, referring instead to the four Evangelists. Second, the pastiglia moldings and punched borders of the Alana panels imply a system of framing incompatible with that partially preserved on the Cloisters panel. And third, the Alana panels are to be attributed to Andrea di Bartolo and recognized as works of a considerably later date. It may therefore be said that there is no physical impediment to accepting Christiansen’s proposal for associating the Yale Annunciate with the Cloisters Adoration, but several arguments incline to preferring the reconstruction offered by Mallory and Moran. These writers pointed out that the punched decoration along the margins of the Yale Annunciate is identical to that in the Lucignano triptych (see fig. 2)—which shows the Virgin and Child with Saints John the Baptist and John the Evangelist—and that the width of the former is compatible with the upper portion of the Lucignano laterals. Freuler contested this observation, reporting that the present frame on the Lucignano triptych is modern but that the width of the top of the two lateral panels is 27 centimeters. This is more than adequate to support the Yale Annunciate, whose painted surface measures 26.4 centimeters wide.

Following the inscription preserved on the riser of the Virgin’s throne in the Lucignano triptych—“DNA [LI]NA FILIA OLI[M] PETRI CE[IA]TI JAKOB OLI[M] D[OMINI] GRIFFI FECIT FIERI I[N] CAPEELLA PER ANIMA SUA” (Dona Lina, daughter of Pietro Cacciati, deceased, [and widow of] Don Jacopo Grifi made this chapel for the salvation of her soul)⁷—Freuler identified its patrons as the Cacciati and Griffi families and established its original provenance as the chapel of Saint Peter in the church of San Francesco in Montalcino. As the dedication of this chapel does not correspond to the identity of either of the saints in the lateral panels of the triptych, he reasonably assumed that the altarpiece was originally a pentaptych and posited that two further lateral panels portraying Saints Peter and Lucy are missing today. The assumption that Saint Lucy was one of the missing saints derived from his inclusion of a predella now in the Pinacoteca Nazionale, Siena,⁸ showing the Adoration of the Magi with scenes from the lives of Saints Peter, John the Baptist, John the Evangelist, and Lucy. This predella, however, was painted not by Bartolo di Fredi but by Andrea di Bartolo. It is not clear whether Andrea was yet an active member of Bartolo’s studio at the date of this altarpiece, and the iconographic evidence for including the predella with this reconstruction is subject to alternative interpretations. Freuler assumed that the scene of the martyrdom of Saint Lucy at the right of the predella was included as a reference to another of the Cacciati family charities, the Ospedale della Misericordia di Santa Lucia. Lacking further evidence, this suggestion is plausible, but his argument for the choice of the unusual Petrine scene of the healing of Saint Petronilla is less convincing. An alternative proposal to identify the predella as a later work by Andrea di Bartolo and as having stood beneath an altarpiece originally in the church of San Petronilla in Siena, four lateral panels of which are now preserved in the basilica of the Osservanza there, has the merits of accounting for this singularly unusual iconography but requires the presumption that at least two other scenes are missing from the predella in its current configuration.⁹ Neither of these contentions can at present be confirmed. Freuler’s other proposal, to identify a much-damaged panel portraying Saint Mary Magdalen now in the WL-Museum für Kunst und Kultur (formerly the Westfälisches Landesmuseum), Münster, Germany,¹⁰ as the probable pinnacle completing the missing panel of Saint Peter at the extreme left of the altarpiece, also cannot be verified on physical or stylistic grounds. Although it is similar in style to the Yale Annunciate and vaguely comparable in size—the Magdalen has been cut laterally and at the top, it currently measures 38 by 25.5 centimeters—it differs considerably in the system of punch tooling visible along its intact right margin, where the cusping of the picture surface begins only at the midpoint of the saint’s halo rather than at the level of her shoulder, as in the Yale panel. There is little likelihood that these two pinnacles originated from the same altarpiece. Which of them, if either, might have been part of the triptych in Lucignano cannot be established with confidence.

Opinions expressed on the dating of the Yale Annunciate have varied widely but necessarily follow from arguments for reconstructing its original context. Christiansen assumed the panel should be dated close to 1374. Freuler accepted this general dating, objecting to Christiansen’s reconstruction only on the grounds that the Yale panel is too large to have accompanied the lateral panels he wished to include with the 1374 Adoration altarpiece. Mallory and Moran believed the Yale panel could be dated between 1382 and 1385. If anything, the harshly cleaned condition of the Yale pinnacle leaves the impression that it is all but contemporary to two other Franciscan altarpieces painted by Bartolo di Fredi in Montalcino: the Beato Filippo Ciardelli altarpiece of 1382¹¹ and the Fraternità di San Francesco altarpiece probably of 1381–82.¹² Although he omitted the Yale panel from his discussion, Freuler’s contention that the Cacciati/Griffi altarpiece is most likely datable ca. 1380, at the beginning of an extended sojourn at Montalcino undertaken by Bartolo di Fredi, is entirely convincing, and a range of dates for the Yale panel, whatever its reconstruction, between 1380 and 1382 seems prudent. —LK

1. Berenson 1932a, 183. ↩︎

2. Berenson 1968, 1:30. ↩︎

3. Seymour 1970, 71. ↩︎

4. Mallory and Moran 1972, 10–15. ↩︎

5. Christiansen 1982, 38, fig. 35; Harpring 1993, 151, no. 8; and Loseries 2012, 60, 63n23. ↩︎

6. Freuler 1994, 141, 446–47, no. 17, fig. 363. ↩︎

7. Although Freuler, in Freuler 1985, interpreted the abbreviation before “Capeella” as “in,” it might better be read as “istud” (this) and as referring to the foundation and endowment of the entire chapel, not just the provision of the altarpiece. ↩︎

8. Inv. no. 103. ↩︎

9. Kanter 1986, 28n14. ↩︎

10. Inv. no. 911 BM. ↩︎

11. Museo Civico Montalcino, inv. nos. 25, 31, 41. ↩︎

12. Museo Civico Montalcino, inv. nos. 37, 40; Museo Diocesano d’Arte Sacra, Montalcino. ↩︎