I · THE OBJECT · BRITISH MUSEUM

The Sutton Hoo Shoulder-Clasps

Anglo-Saxon · Early 7th century · Gold, garnet, millefiori glass

▶ Listen · narrated

Gold can be mined in Britain. Garnet cannot. These clasps, buried in Suffolk soil for thirteen centuries, record a supply chain that reached across continents.

At a glance

Date: Early 7th century AD
Findspot: Sutton Hoo, Suffolk, England
Material: Gold, garnet, millefiori glass
Technique: Cloisonné with stepped garnets

Look closer

Stepped garnet cells
The garnet is not laid flat. Each piece sits in a gold cell lined with patterned gold foil, and the stone itself is cut with a flat top and bevelled sides — a technique called stepped cloisonné. The bevel reflects light back through the stone, intensifying the red. The gold foil beneath is stamped with hatching or chequerboard patterns that shimmer through the translucent garnet. At arm's length, the individual cells disappear and the surface reads as continuous red punctuated by gold.
Interlace animals
The outer borders of each clasp carry interlaced animals rendered in garnet against gold. The bodies loop and bite, necks crossing over haunches, limbs tucked under torsos. This is not decoration applied to a functional object; the entire clasp is a vehicle for the ornament. The animals are stylised to the edge of legibility — you read them as much by convention as by observation. Similar interlace appears on other Sutton Hoo objects and in Insular manuscripts, suggesting a shared visual grammar across media.
Hinge and pin mechanism
Each clasp is actually two halves. One half carried a pin, the other a hinge loop, so they could be fastened and unfastened. They would have sat on the shoulders, joining the front and back panels of a garment or cuirass. The mechanism is concealed on the reverse, leaving the visible surface uninterrupted. The precision required to make the two halves align, while also accommodating thousands of inlaid garnets, suggests a workshop with sustained access to materials and highly trained hands.

The story

The shoulder-clasps were found in 1939 in the ship burial at Sutton Hoo, part of a suite of gold and garnet dress fittings that included a belt buckle, strap mounts, and purse lid. They are hollow gold shells, each approximately twelve centimetres long, divided into panels filled with garnet cloisonné and millefiori glass. The garnets are Indian almandine, a species that does not occur naturally in Britain. Geological analysis points to sources in Rajasthan or Sri Lanka. By the time these clasps were made, the stones had travelled thousands of miles, probably through multiple hands.

The technique is cloisonné: thin gold walls soldered to a gold baseplate to form cells, each cell then filled with a precisely cut piece of garnet. The garnet is not simply sliced flat. Each piece is shaped with a flat top surface and angled sides — a stepped or bevelled profile — so that light entering the stone is reflected back by the gold foil beneath. The foil itself is patterned, stamped with fine hatching or chequerboard designs that create texture under the translucent stone. The effect is a surface that glows rather than merely shines.

The precision is difficult to overstate. Each clasp contains more than three thousand individual garnet pieces. The cells are tiny, often less than a millimetre across, and the garnet must be cut to fit each cell exactly. The joins between adjacent stones are so fine that they become invisible at a short distance, and the surface reads as continuous fields of red divided by gold lines. No adhesive has been identified; the stones appear to be held by friction and the slight inward pressure of the cell walls.

The garnets arrived as rough stones, probably traded along routes that connected the Indian Ocean to the Mediterranean and from there to the Atlantic seaboard. How they reached East Anglia is less clear. They may have come through Frankish intermediaries, or through direct contacts with merchants operating out of Byzantine or Coptic ports. The ship burial also contained silver bowls from the Eastern Roman Empire, so the workshop that made these clasps had access to materials from multiple distant sources.

The design is structured and symmetrical. Each clasp is divided into rectangular and curved panels, filled with either garnet cloisonné or chequerboard millefiori glass. The outer borders carry interlaced animals in garnet and gold. The style is consistent with other Anglo-Saxon metalwork of the period, particularly objects from East Anglia and Kent, but the technical execution is exceptional even within that tradition. The workshop must have had sustained access to materials, tools, and trained craftspeople. This was not a one-off commission.

Who wore them is uncertain. The burial is usually associated with Rædwald, king of the East Angles, who died around 624, though the identification is not conclusive. The clasps would have fastened a garment or cuirass at the shoulders, visible and prominent. They were made to be seen, and to display access to materials and skills that few could command.

Why it mattered then

Garnet was a prestige material in early medieval Europe, valued precisely because it had to be imported. Wearing it announced participation in long-distance networks, access to goods that moved along trade routes stretching from South Asia to the Atlantic. The shoulder-clasps were not simply ornaments; they were legible statements of connection and resource control. The technical skill required to make them was itself a form of wealth. Cloisonné work of this quality required specialised training, sustained practice, and access to tools and materials that were expensive and difficult to obtain. The workshop that produced these clasps was operating at a level of sophistication comparable to contemporary Byzantine or Frankish ateliers. That such a workshop existed in seventh-century East Anglia, and that it had reliable access to Indian garnet, suggests a more connected and technically accomplished society than the term 'Dark Ages' would imply. The objects buried at Sutton Hoo were not everyday possessions. They were selected for display in death, and their selection tells us what mattered. The shoulder-clasps sat alongside a helmet, a sword, a shield, Byzantine silver, and a lyre. Together, they present a particular image of authority: martial, wealthy, connected to distant places, and visually literate in a shared ornamental language.

Why it matters now

The shoulder-clasps are evidence of a supply chain. They require us to imagine not just a workshop in Suffolk, but a network that moved raw garnet from South Asia to the Atlantic coast, and that did so reliably enough for a craftsperson to plan work around its availability. Early medieval Europe is often imagined as fragmented and isolated, but objects like these suggest the opposite: sustained contact, repeated transactions, and the infrastructure to support both. They also raise questions about how technical knowledge moved. The stepped garnet technique appears in Anglo-Saxon, Frankish, and Lombardic metalwork during the same period. Did craftspeople travel? Were objects copied? Did workshops train apprentices who then migrated? The clasps do not answer these questions, but they make clear that the knowledge was shared across regions, and that it was valued enough to be transmitted with care. Finally, they are a reminder that ornament is information. The interlace animals, the symmetry, the colour contrasts — these were not arbitrary choices. They were part of a visual language that an Anglo-Saxon audience could read, and that signalled identity, allegiance, and status. We have lost most of the grammar of that language, but the objects remain, dense with meaning we can only partially recover.

The surprising detail

The garnets were not glued. Modern analysis has found no trace of adhesive in the cells. The stones are held in place by the precision of the fit and the slight inward spring of the gold walls. This means that if a stone was cut even fractionally too small, it would fall out, and if it was cut too large, it could not be inserted. The margin for error was effectively zero, and the craftsperson was working with a material that cannot be reshaped once cut. Every piece had to be correct on the first attempt.

What is disputed

The identification of the burial with Rædwald is widely accepted but not certain. No inscription names the occupant, and the dating relies on coin evidence and stylistic comparison. Some scholars have proposed alternative candidates, including Rædwald's son or successor. The question remains open.

How it got here

The shoulder-clasps were excavated in 1939 by Basil Brown from a ship burial on land owned by Edith Pretty. In 1942, Pretty donated the entire assemblage to the British Museum. An inquest held shortly after the discovery ruled that the objects were not treasure trove, as they had been deliberately buried rather than hidden with intent to recover, and therefore belonged to the landowner. Pretty's donation was made during the Second World War, and the objects were moved to safety shortly afterward. They have remained in the British Museum's collection since. There is no active repatriation claim.

Remember this

Three thousand pieces of Indian garnet, cut to fit gold cells so precisely that no glue was needed. A supply chain reaching across continents, and a workshop that could not afford a single miscalculation.

Test yourself

Why would a seventh-century East Anglian workshop go to the trouble of using stepped garnets with patterned foil beneath, rather than simply laying flat stones on plain gold?

Go deeper

Image: Anglo Saxon burial mound at Sutton Hoo by Neil Theasby. Licence: CC BY-SA 2.0. Source.

← Back to day 44