Reading a Roman Portrait Without Its Inscription
A regional museum owns a marble portrait head, slightly over life-size, with deep drilled pupils, a short cropped beard, and hair rendered as flat incised locks rather than carved volume. The inscription on its lost pedestal is gone. The registrar's file says 'Roman, 1st–2nd century CE.' A visiting scholar proposes redating it to the 4th century CE and reattributing it to a Late Antique workshop. The museum must decide whether to change the label before a loan, and the lender requires a written justification. The head has no known findspot, and its marble has not been isotopically tested. Two comparanda are cited: a bearded emperor portrait with drilled pupils from the later 2nd century, and a colossal 4th-century portrait with similarly flattened hair and enlarged eyes.
1. What visual evidence actually distinguishes the two proposed dates, and how strong is it?
Read the worked answer
The decisive features are the drill work and the treatment of hair and eyes, not the beard alone.
- Deep drilled pupils indicate that the sculptor was working after the widespread adoption of the drill for eyes, which is characteristic of the later 2nd century onward. That supports the registrar's range but does not exclude the 4th century.
- Flat, incised locks with little undercutting and enlarged, staring eyes are hallmarks of Late Antique portraiture, where carving becomes more graphic and less volumetric. If the hair is genuinely flattened rather than merely weathered, that leans 4th century.
- Slight over-life-size scale fits imperial or high-status portraiture in both periods.
The honest assessment: the features are consistent with a 4th-century date but not diagnostic of it, because drilled pupils and stylized hair also appear in later 2nd-century work. Visual style alone cannot settle a two-century gap when the object is unprovenanced.
2. What should the museum do before the loan, and what wording should the label use?
Read the worked answer
The museum should not silently adopt either date. Recommended steps:
- Isotopic/marble sourcing - stable isotope analysis can at least place the stone in a quarry region, which may exclude or support a workshop attribution. Without it, the findspot-based argument is unavailable.
- Technical imaging - raking light and microscopy can distinguish original flattened carving from surface loss or recutting, which is the crux of the stylistic argument.
- Comparative dating - cite the two comparanda explicitly and state which features align with which.
Label wording should present a range with the reasoning exposed: 'Roman, 2nd–4th century CE (date debated); drilled pupils and flattened hair have been read as Late Antique, but the portrait cannot be securely dated without provenance or stone analysis.'
A student who simply asserts 'it is 4th century because the hair is flat' has treated a stylistic tendency as proof; the exam rewards acknowledging that unprovenanced portraits carry irreducible uncertainty and that the label should encode it.