Reconstructing a pediment from 23 fragments
A museum in Valletta holds 23 fragments of Parian marble said to come from a temple pediment. All were recovered from a late-antique wall and have no surviving stratigraphy. The pieces include a right forearm gripping a snake, a striding leg wearing a sandal with a strap, a torso wearing a cuirass with a shoulder strap, and a horse’s hoof. Traces of Egyptian blue and red ochre survive in recesses. Deep drill channels run behind the snake arm. There is a lead clamp channel and a single mason’s mark, a letter delta. The fragments vary in thickness from 12 to 28 cm and in surface preservation from crisp to heavily weathered. The board proposes a full-size plaster reconstruction of the entire pediment for the new galleries, with the original fragments set into it so that “visitors can see the whole picture”. The opening is in eight months; the excavator’s archive is un catalogued and no fragment coordinates were published.
1. What can be inferred with confidence from these fragments, and what cannot? Answer as the curatorial adviser.
Read the worked answer
Inferable with reasonable confidence.
- Polychromy: Egyptian blue and red ochre in recesses proves the sculpture was painted, which is worth saying plainly because the public still expects white marble.
- Technique and broad date: deep drill channels behind the snake arm point to a period of heavy drill use, broadly Hellenistic rather than Archaic; the clamp channel shows standard multi-block construction.
- Subject type: a cuirassed figure, a striding leg and a snake together fit a combat or gigantomachy-type scene, and a horse’s hoof indicates at least one equestrian or harnessed element. That is a scene type, not an identification.
Not inferable.
- That these 23 pieces belong to one pediment. They come from a late-antique wall: they are spolia, redeposited, possibly from more than one building. Thickness varying from 12 to 28 cm is exactly what you would expect if pieces from different structures were bundled into a lime wall - or from a pediment’s varying depths. The evidence cannot yet choose.
- Which deity or myth. A snake arm could belong to Athena, to a giant, or to a river personification.
- A date tighter than “probably Hellenistic”. Drill alone does not date a sculpture; it dates a technique in fashion over centuries.
- Any reconstruction of the whole. You cannot rebuild a composition from four body parts and a hoof without inventing most of it.
2. The board wants a full-size plaster reconstruction with the originals set into it. Give your recommendation and the conditions attached to it.
Read the worked answer
Recommendation: do not set the originals into a plaster reconstruction; do build a reconstruction, but as a separate, clearly hypothetical object.
Reasons. Embedding authentic fragments in a plaster whole converts a guess into furniture. The plaster supplies the joints, and the joints are precisely what is unknown: because the pieces are spolia with no recorded coordinates, the reconstruction would silently assert that all 23 belong together and that their positions are known. Visitors would read a nineteenth-century-style certainty the evidence cannot carry. There is also a conservation argument: plaster in contact with marble creates salt and moisture problems and makes future study of joins harder.
Conditions if the board proceeds:
- Display the fragments as found, on a mount that shows their broken edges, with the redeposited context stated in the label.
- Build the reconstruction in a visibly different material and colour, in a separate case, labelled “Hypothetical reconstruction - not a restoration”.
- Show a digital model instead, with a toggle that removes everything not attested, so the invented percentage is visible.
- Publish the reasoning and the fragment photographs before opening.
- Rewrite the label from “the pediment of…” to “fragments probably from a temple pediment, reused in a late-antique wall”.
Eight months is enough for a digital model and an honest label. It is not enough for a defensible physical reconstruction.