A Marble Head, Two Labels, and a Ten-Day Deadline
You are the curatorial assistant at a provincial museum in Málaga. A marble head, 32 cm high, is due to go on display in ten days. The acquisition file (1962) has no provenance before 1930. The conservation report notes: a smooth horizontal join at the neck secured with an iron dowel; deep, channel-like drill work in the hair; a highly polished face; the back only roughly finished; two rectangular marble patches on the crown; and surface tool marks that differ between face and hair. Isotopic analysis matches Carrara. The draft label reads 'Greek original, 2nd century BC, Hellenistic'. The director prefers that wording because 'Roman copy' would reduce the insurance valuation by roughly 40 per cent. A local scholar has emailed arguing for 'Roman copy, 1st–2nd century AD'. You must recommend one label and justify it in terms a PAU student could reproduce, distinguishing strong from weak evidence.
1. Which label is better supported by the physical evidence, and which features are diagnostically strong, weak or ambiguous? Show your reasoning.
Read the worked answer
The defensible label is 'Roman, probably a copy after a Hellenistic original, 1st–2nd century AD'. The reasoning has to be cumulative, because no single feature proves a copy.
- Carrara isotope match (strong, not conclusive): Carrara was exploited on a large scale from the late 1st century BC; a 2nd-century BC Greek original on Carrara is therefore unlikely, though not impossible.
- Neck join with iron dowel (moderate): common in Roman copy production and in ancient repairs; it shows the head was made separately, not who made it.
- Deep drill channels in the hair (weak): a stylistic technique used in Hellenistic Pergamene work and in Roman classicizing sculpture; it dates a style, not an origin.
- Rough back (weak): typical of works set in niches or against walls, both copies and originals.
- Patches and high polish (weak): condition and finish, not attribution.
- Different tool marks (moderate): suggest separate campaigns or hands, consistent with workshop production.
The combination - Carrara, a neck join, an unfinished back, repairs and Roman-period tool marks - makes a Roman copy more probable than a Greek original. But the label should be hedged: 'probably a copy' is more honest than either 'Greek original' or a flat 'Roman copy'.
2. The director asks for a 60-word label that is defensible and readable. Draft it and explain each choice.
Read the worked answer
Draft: 'Head of a youth, probably Roman, 1st–2nd century AD, after a Hellenistic Greek original. Deep drill work in the hair and a polished face recall Hellenistic style; the neck join, unfinished back and Carrara marble point to Roman workshop practice. Acquired 1962; provenance before 1930 unknown.'
Choices: 'probably Roman' avoids false certainty; 'after a Hellenistic Greek original' identifies the stylistic model without claiming the work is that original; the evidence clause lets a visitor see why the label is hedged; the provenance sentence is ethically necessary because the early history is missing. The word 'copy' is avoided in the public text, but the concept is preserved by 'after' and 'workshop practice'. This is not evasion: it states the argument in plain language.
3. What one further non-destructive or minimally invasive analysis would you request, and what would it change?
Read the worked answer
Request X-ray computed tomography of the neck join and dowel, plus isotope analysis of the marble used for the two patches. CT would show whether the dowel is ancient or modern, whether the join was cut in antiquity, and whether the head and neck were carved from one block or assembled from two. Isotope analysis of the patches would show whether they are ancient repairs using the same marble source or modern restorations. Either result could change the label: a modern dowel and modern patches would weaken the case for a coherent Roman workshop object and raise questions about restoration; an ancient dowel and ancient patches would strengthen the Roman-period context. Neither test would prove authorship or origin on its own, but both would make the condition history explicit, which is what the label currently lacks.