The Sassetta label: dating a small gold-ground panel
A small panel painting, 31.5 × 24 cm, on poplar, entered a Maltese collection in 1907 as 'Follower of Sassetta, c. 1450'. The gold ground is punched with a rosette and diamond pattern; the flesh tones are modelled with green earth underpaint. Infra-red reflectography reveals fluid, freehand underdrawing with long parallel hatching, and no evidence of a transferred cartoon. Pigment analysis finds natural azurite, verdigris, lead white and red lake, but no smalt and no lead-tin yellow. Dendrochronology on eight growth rings gives a terminus post quem of 1462; sapwood is absent. A shield with a lion, painted over the Virgin's mantle, was added later. The panel has a vertical split repaired with animal glue and a 2 cm strip added at the left edge. The museum must decide whether to label it 'Sienese, c. 1450', 'Florentine, c. 1470' or 'unknown Italian, later addition', and whether the shield is part of the original commission.
1. Build the strongest dating argument available from these data, stating which evidence is secure and which is inferential.
Read the worked answer
Secure: the panel is poplar and the pigment range is consistent with a 15th-century Italian workshop. Azurite without smalt does not fix a date, because smalt is an unreliable marker before c. 1500 and is absent from many workshop inventories. The dendrochronological terminus post quem of 1462 is a hard lower bound only if the rings are correctly sequenced; with eight rings and no sapwood, it cannot narrow the date by itself. The underdrawing is the strongest stylistic discriminator: fluid freehand hatching is common in Florentine and north Italian workshops but rare in surviving Sienese panels of the 1450s. The punchwork is generic and may have been added or reworked. Weighing these, a date c. 1465–80 is defensible; 'c. 1450' is not, because it precedes the dendrochronological bound. The attribution should be phrased as 'Italian, probably Florentine' rather than 'Follower of Sassetta', because the technical evidence actively weakens the Sienese link.
2. How should the shield be interpreted and catalogued?
Read the worked answer
The shield is later overpaint: it sits above the Virgin's mantle, its paint structure differs, and it has no underdrawing related to the original design. It should not be used to date the panel or to identify the original patron. Two interpretations remain open: a later owner added arms to claim the image, or a patron's arms were added after delivery when a confraternity changed affiliation. Without documentary evidence, the catalogue should describe it as a later addition, not as part of the commission. Conservation should not remove it solely to 'restore' the image; removal is justified only if it is unstable or obscures a secure original surface, and should be documented and reversible. Its presence should be recorded in a diagram. If the museum interprets it, label it as 'later heraldic addition, date unknown', avoiding a story that treats the arms as original evidence.
3. Recommend the final label and justify it.
Read the worked answer
Recommended label: 'Italian, probably Florentine, c. 1465–80. The Virgin and Child with a later heraldic shield.' Justification: the date range is bounded below by dendrochronology and supported by the underdrawing; 'probably Florentine' is a calibrated hedge. Avoid 'Follower of Sassetta' and 'Sienese' because the technical evidence points away. State the later addition in the label so viewers do not read the arms as part of the original. Add a technical note: 'Poplar panel; azurite, verdigris, lead white, red lake; dendrochronological terminus post quem 1462; later overpaint.' If the museum requires a single attribution, 'Unknown Italian, 15th century' is more honest than a named follower. The label should mention the repaired split and added strip, because condition affects interpretation of the composition.