The Staffordshire Hoard

In July 2009 the metal detectorist Terry Herbert found the largest hoard of Anglo-Saxon gold and silver metalwork ever recovered, in a field near Hammerwich, Staffordshire, on land farmed by Fred Johnson. Unlike a grave find such as Sutton Hoo, there was no body, no ship, and, so far as excavation could establish, no burial pit at all: around 4,600 pieces, some 5.1 kg of gold and 1.4 kg of silver, had been dismantled from swords, at least one helmet and a handful of Christian objects, and left in the ground with no one ever returning for them.

Valued at 3.285 million GBP by the Treasure Valuation Committee and bought for the nation in 2010 after a public campaign led by the Art Fund, the hoard is now jointly held by Birmingham Museums Trust and the Potteries Museum & Art Gallery, Stoke-on-Trent. A second find in the same field in 2012 added a further 81 Treasure-declared pieces. What the hoard is actually for remains disputed: battlefield loot, a votive offering, a war leader's bullion reserve and collected tribute have all been argued by named specialists, and none has settled the question. Even the burial date is not fixed beyond a commonly cited but contested window of about 650 to 675 AD.

This record treats the hoard as the dispersed, still partly unresolved find that it is: not the Sutton Hoo helmet, a different object from a different county and a different kind of site, and not the Cheapside Hoard, a wholly unrelated Jacobean jewellers' cache. Every claim below is tied, where the sources allow, to a specific catalogued piece.

What it is

The Staffordshire Hoard is not one object but around 4,600 pieces and fragments of gold and silver metalwork, overwhelmingly stripped from war-gear: sword hilt fittings, an estimated one-third of the whole by mass from a single dismantled helmet, and almost nothing else, aside from a small group of Christian objects. In aggregate it holds about 5.1 kg of gold and 1.4 kg of silver, decorated with roughly 3,500 separate pieces of cloisonne garnet. Almost none of it is whole: swords are represented by their gold and garnet hilt fittings but hardly ever by an iron blade, and the helmet survives only as a large but incomplete set of decorated silver panels with no trace of the iron or leather structure that would have held them together.

There is no coinage, no brooch, no item of female dress and, so far as the excavators and later researchers could establish, no grave, no body and no ship. That absence is as important as anything present: this is not a burial in the way Sutton Hoo is a burial. It is a deliberately assembled and then deliberately buried collection of dismantled, mostly martial, metalwork, and the two questions every subsequent piece of research keeps returning to are why it was dismantled and why it was buried.

How it was found

Terry Herbert, a 55-year-old detectorist who had been unemployed because of a disability and had detected for more than eighteen years with the Bloxwich Research and Metal Detecting Club, was searching a field in the parish of Ogley Hay, near the village of Hammerwich in south Staffordshire, on Sunday 5 July 2009. He had written permission from the landowner, farmer Fred Johnson, and a prior written agreement that any reward would be split equally between them.

The findspot lies on the north-western spur of a low ridge about 3.5 km west of Wall, the Roman settlement of Letocetum, directly beside the A5, which follows the line of the Roman Watling Street. Over the following five days Herbert recovered around 500 items lying loose in the plough soil and on the surface, ranging from the cheek-piece of a helmet and a folded gold cross down to strips of gold weighing less than a gram. He correctly judged that the material was not coming from any surviving buried feature, and reported the find to Duncan Slarke, the Portable Antiquities Scheme's Finds Liaison Officer for Staffordshire.

The excavation

A meeting at Birmingham Museum and Art Gallery, attended by Bill Klemperer and Lisa Moffett of English Heritage, agreed a strategy for recovering the rest of the material. Stephen Dean, Principal Archaeologist of Staffordshire County Council, and a colleague, Ian Wykes, first dug a single 1 x 1 m test pit, which alone produced about a hundred further items of gold, garnet-inlaid gold and silver-gilt, all recovered from the topsoil; no artefact was found in the subsoil, and no feature could be identified.

Birmingham Archaeology, the commercial fieldwork unit of the University of Birmingham's Institute of Archaeology and Antiquity, was then commissioned to continue, with Alex Jones directing the work on the ground. English Heritage and Staffordshire County Council jointly funded the excavation, the council also paying for night security because of the risk of nighthawking on a highly visible site bounded by roads on three sides. A total of 152 sq m was hand-excavated in 1 m squares, every spit of soil scanned with a metal detector, recovering around 800 further objects; a magnetometer survey covered 5.3 ha around the findspot. A second stage of fieldwork in March 2010, comprising a resistivity survey and trial trenching, was intended to place the hoard in its immediate landscape context.

No archaeological feature that could be securely linked to the hoard was ever found. The one substantial cut recorded, a shallow ditch aligned north-west to south-east in the south-western corner of the excavated area, proved on excavation to be of natural, clay-filled origin, and two nearby palisade trenches produced no datable material and were judged to belong, if anything, to an undated farmstead unconnected with the hoard. Whatever pit the objects were originally placed in had been entirely destroyed by later ploughing; the excavators' own conclusion was that the objects were recovered essentially where the plough had spread them, most densely around the original findspot, with heavier pieces dragged further by repeated cultivation than lighter ones.

A second find in the same field, 2012

In November 2012, after the field had been ploughed again, archaeologists from Archaeology Warwickshire returned with metal detectorists and recovered a further 91 items, including additional gold and silver fragments and, among them, a further large piece of gold cross and a second helmet cheek-piece. At a coroner's inquest in January 2013, 81 of the 91 new finds were declared Treasure; the remaining ten were assessed as non-ancient wastage rather than genuine hoard material.

The 2012 pieces were added to the collection already held jointly by Birmingham Museums Trust and the Potteries Museum, bringing the assemblage to its present total of almost 4,600 pieces. The fact that further material turned up three years after the original, thoroughly fieldwalked and magnetometer-surveyed excavation is itself informative: it shows how completely the original burial context had been destroyed by ploughing, and how widely a single dispersal event, or several, had scattered even quite substantial pieces through the topsoil.

What is actually in it

Counts vary slightly between publications as cataloguing has continued, but the shape of the assemblage is consistent. There are somewhere between 86 and 94 richly decorated sword pommel caps, by far the largest number ever recovered from a single context, together with numerous gold hilt plates and hilt collars, several with fine filigree ornament. An estimated one-third of the hoard's total mass, by fragment count, belongs to a single dismantled helmet.

Only two objects in the entire assemblage are complete or close to complete. One is a large folded gold cross, catalogued as no. 539 in the 2019 Society of Antiquaries volume, missing some of its original garnet settings and possibly made to serve as an altar or processional cross, or as decoration for the cover of a gospel book. The other is a smaller, equal-armed gold pectoral cross (Portable Antiquities Scheme Find-ID 28631), set with cloisonne garnets that originally numbered about 95, of which 58 survive, around a central cabochon garnet. A third Christian object, a bent gold strip inscribed with a Latin quotation from the Book of Numbers, and a unique gold mount combining cloisonne garnet work with millefiori glass, catalogued as no. 541 and informally nicknamed the mystery object by researchers because its original function is still not agreed, complete the small group of pieces that stand apart from the martial material. Five small gold snake or serpent fittings, some fitted with pins, have also been recovered; whether these were worn, carried, or served some other purpose is not established.

What is absent is as telling as what survives. There are no coins, no brooches and nothing identifiable as female dress equipment, and remarkably few iron sword blades or the belt and scabbard fittings that would ordinarily accompany a hilt of this quality, as if the hilts had been deliberately separated from everything else that once went with them. One further item recovered from the same plough soil and formally processed through the Treasure Act inquest alongside the ancient material was a twentieth-century key ring, catalogued as no. 890: a mundane, entirely modern object that happened to be lying in the same field and had to be identified and excluded from the ancient assemblage rather than assumed to belong to it.

How it was made

The metalworking on display is technically accomplished and repeatedly deceptive about its own material. Scanning electron microscopy carried out by Dr Eleanor Blakelock of the British Museum, on a pilot sample of sixteen objects later expanded to more than 130, found that many gold-silver alloy surfaces had been artificially depleted of silver, so that a piece with a core of 25 to 50 percent silver could present a surface reading of only 6 to 10 percent: enough to look like far purer gold than the metal underneath actually was. As Blakelock has put it, for all we know it was the colour of the gold that was the priority, not its purity. On filigree pieces such as the pectoral cross, the fine gold wire ornament and its backing sheet were enriched to different degrees, producing a deliberate colour contrast between the pattern and the ground it sits on.

The helmet's decorative silver sheet was made by die-stamping, pressing a design cut into a hard die into thin silver, and was then fire-gilded so that the finished object would have read as solid gold rather than gilded silver. The same basic method, though not the same physical dies, decorates the Sutton Hoo helmet and the boat-grave helmets excavated at Vendel and Valsgarde in Sweden, evidence of a shared North Sea craft tradition among Anglo-Saxon and Scandinavian smiths rather than proof that any of these objects came from the same workshop.

No maker's mark survives on any piece, and the raw material's own origin is debated. By the mid-seventh century, Anglo-Saxon gold coinage was becoming progressively debased with silver, giving some of it a silvery tinge, and it has been suggested, though not proven in the sources consulted, that recycled coinage of this kind supplied some of the hoard's gold. David Symons, a curator involved in the hoard's research, has put the scale of the raw material in context: by his estimate the fragments contain about 11 lb of gold, the equivalent of more than 3,000 contemporary gold coins and, in his words, certainly whoever owned the Hoard was very high status, there's three times as much gold as at Sutton Hoo, and it came from elite warriors.

The dating debate: what the objects date to is not what the burial dates to

Two different questions get run together in most short accounts of the hoard, and they have different answers. When were the objects made, and when were they buried? On the first question there is broad agreement across a wide range: stylistically, individual pieces run from the later sixth century to the earlier eighth. A silver sword pommel decorated in a Scandinavian or Germanic style has been placed at around 550 AD, a century before the hoard's newest pommels, decorated with paired ring-knobs and datable to about 630-660 AD, were made. Whatever else is uncertain, the objects buried together were not all made together; some were already old when the newest pieces were new.

On the second question, when the burial itself took place, there is no settled consensus, only a conventional and widely repeated window of about 650 to 675 AD, tied loosely to the reign and death of the Mercian king Penda. That window rests substantially on the stylistic dating of the objects themselves plus an assumption about historical context rather than on any independent, agreed method of dating the deposit. The clearest evidence of real scholarly disagreement is the inscribed gold strip. Its lettering has been read by Michelle Brown, a specialist in medieval manuscript studies, as consistent with a seventh or earlier eighth century date, on the basis of its use of half-uncial forms; Elisabeth Okasha, working from the same physical evidence, has identified features, including the use of insular majuscule forms, that in her reading point to an eighth or ninth century date instead. Nicholas Brooks has noted explicitly that using the dating of manuscript handwriting to date an inscribed object is a hazardous typological exercise, precisely because Brown and Okasha have reached different conclusions from the same script. Beyond this, other researchers have proposed considerably later deposition dates altogether, associating the burial with renewed Mercian conflict around 716 or with the civil war between King Offa and the usurper Beornred in 757, rather than with Penda's mid-seventh-century wars. This record does not treat any single date as settled. The manufacture range is wide and largely agreed; the burial date is neither.

Why was it buried: the unresolved dispute

No theory of why the hoard was buried commands anything like consensus, and the genuine, still-open state of the argument is itself the most important thing to report accurately about this object.

The battlefield loot or trophy theory has the most support among the specialists who worked on the Historic England-funded research project, Chris Fern among them: the objects were violently dismantled, chopped up with little regard for their own decoration, the assemblage is overwhelmingly martial and entirely male in character, and stripping an enemy's sword hilt or helmet of its precious fittings would have been both a practical way to render a defeated man's war-gear worthless to him and a way of stripping away his identity as its owner. On this reading the hoard may represent booty gathered across several battles with different peoples over a period of decades, rather than the spoils of one engagement, which would also explain the wide date-range among the pieces.

A second, related but distinct reading treats the fragments as a portable store of bullion rather than trophies as such. David Symons has argued that reducing precious sword and helmet fittings to small, easily divided pieces would have given a war leader a form of currency he could use to pay his followers, in a kingdom that was not yet minting gold coinage in any comparable quantity. Nicholas Brooks, in a comment reported by the excavators, proposed a further alternative: that the objects could represent tribute collected by a Mercian king rather than spoils taken personally in battle, although their damaged and dismembered state sits awkwardly with the idea of tribute, which would usually have been handed over intact.

A fourth reading treats the burial itself, not just the dismantling, as the meaningful act: a votive or ritual deposit. Leslie Webster has compared the hoard directly to the barrow-treasure scene in the Old English poem Beowulf, thought to have been composed in Mercia, in which golden objects taken earlier from a hoard are deliberately placed back in a barrow and left to the earth, gold in the dust, where it still remains, as useless to men as it was before. Webster describes the poem's narrative as led by the need to consign the ill-gotten treasure back to earth, and the comparison raises the possibility that burying the Staffordshire material was itself understood as a kind of ritual disposal of tainted or dangerous wealth, rather than a temporary hiding place intended for recovery. A further variant of this reading, reported by Tom Shippey from the full academic publication, treats the destruction as ideologically driven: a deliberate act of destroying the symbols of a defeated lord's authority, to prevent his heirs or followers from reclaiming legitimacy through his old regalia.

None of these has been established over the others, and they are not all mutually exclusive; a collection assembled as trophies over years could still have been finally buried for votive reasons. What can be said plainly is that whoever buried the material never came back for it, and an unmarked field beside a busy Roman road, chosen (on one reading) precisely because it was easy to find again, was in the event never revisited. As one review of the case has put it, that in itself is a bad sign.

One event, or several, over decades

A closely related and equally unresolved question is whether the hoard was buried in a single episode or accumulated in the ground across more than one occasion. The excavators themselves noted, from the character of the material recovered in 2009, that the presence of around ninety sword pommels of consistently high quality gave the impression of a careful selection built up over a period of time, rather than the spoils of a single battle, describing the assemblage as having the feel of a trophy hoard.

The wide stylistic date-range among the objects, from a mid-sixth-century heirloom pommel to pieces made only shortly before probable deposition, is consistent with either reading: a single burial of wealth accumulated gradually before that one event, or several separate depositions added to the same spot over years. Archaeology cannot settle the question from the ground itself. No stratified sequence, no separate pits, and no datable feature of any kind survived the destruction wrought by centuries of ploughing, so any answer has to come from the objects and their find-relationships to each other rather than from excavated context.

The Christian objects, and what their company says

The folded gold cross, the smaller pectoral cross and the inscribed gold strip are a small minority within an assemblage that is otherwise overwhelmingly martial and, in its imagery, largely in a pre-Christian idiom. The inscribed strip's own text, drawn from Numbers 10:35, is itself a war-cry rather than a devotional formula: rise up, Lord, and let your enemies be scattered, and let those who hate you flee from your face. Its presence alongside sword fittings is therefore not, on its own, evidence of any tension between Christian and martial identity; the same words could have accompanied an army with an entirely orthodox Christian self-understanding.

What is genuinely disputed is why the large cross, in particular, had been folded and damaged before burial, alongside the strip, which was also bent. One reading treats the folding as purely practical, a way of fitting a large liturgical object into whatever container or space the depositor had available. A second treats it as an act of desecration, consistent with looting Christian material during a period, associated with Penda's reign, when the Mercian kingdom's ruling elite was still substantially pagan; Bede records Penda himself as remaining a pagan even while tolerating Christianity among his own family. A third reading, harder to distinguish from the second on the physical evidence alone, treats the folding as an act of reverence rather than desecration: a way of ritually retiring a damaged sacred object rather than discarding it carelessly. None of these readings is established as correct in the sources consulted, and this record does not choose between them.

Penda, Mercia and the country the hoard sat in

The findspot sits inside the historical heartland of the kingdom of Mercia during a period when Mercia was aggressively expanding at the expense of its neighbours. The Mercian prince and later king Penda fought a string of major battles across the second quarter of the seventh century: at Hatfield Chase in 633, in alliance with Cadwallon of Gwynedd, where the Northumbrian king Edwin was killed; at Maserfield, near Oswestry, in 642, in alliance with Cynddylan of Powys, where Edwin's successor Oswald was killed; against the East Angles around 635, where Bede records the deaths of the retired king Sigeberht and his successor Ecgric, both descendants of Raedwald, who is himself the leading candidate for the person buried under the Sutton Hoo ship; and finally at the unlocated battle of the Winwaed in 655, where Penda was killed or fatally wounded by the Northumbrian king Oswiu.

Welsh poetic sources add a further, more local episode. The elegy Marwnad Cynddylan describes a raid on a place called Caer Lwytgoed by men of Powys under a leader named Morfael, in which, according to the poem, they took fifteen hundred head of cattle, eighty horses and harnesses besides: extreme booty, in the poem's own phrase. Caer Lwytgoed has been identified by several scholars with the Roman site of Letocetum at Wall, close to the hoard's findspot, and the raid has traditionally been read as an attack on a monastic community established in the vicinity of Lichfield, though the historian Jenny Rowland has argued instead that the men of Powys were fighting in support of Cadwallon and Penda against a Northumbrian-aligned target, which would make the monks served by that target's own religious community rather than a target in themselves.

Local place-names hint, without proving anything about the hoard directly, at a landscape where pre-Christian belief still had a presence: Wednesbury and Wednesfield both derive from the name of the god Woden, and the place-name Weeford, from the Old English weoh, meaning a pagan shrine, has been read as evidence for a heathen temple somewhere along Watling Street to the east of the findspot. Two nearby settlements are both called Hammerwich, from Old English hamor and wic, the hammer-working or trading place, which could indicate a metalworking association with the district; the excavators are explicit, however, that no direct archaeological link between either place-name and the hoard itself has been established, and the finding should be read as suggestive landscape context rather than as evidence bearing directly on the object.

The helmet, and its unfinished reconstruction

An estimated one-third of the roughly 4,600 recovered pieces, by one account well over a thousand individual fragments, belongs to a single dismantled helmet, making it only the sixth substantially represented Anglo-Saxon helmet known, after Benty Grange (found 1848), Sutton Hoo (1939), Coppergate in York (1982), Wollaston or Pioneer in Northamptonshire (1997) and Shorwell on the Isle of Wight (2004). Each of those is a separate object from a separate site and should never be confused with the Staffordshire material or with each other.

The surviving decoration is silver sheet, die-stamped with panels showing marching and kneeling warriors, mounted and priestly figures, and zoomorphic ornament, then fire-gilded so that the finished helmet would have looked entirely gold rather than silver. Reeded silver strips and rivets held the decorated sheet to the helmet's structural base, and U-sectioned edging strips probably clamped the sheet and an underlying leather layer together. At least two cheek-pieces are known, one recovered during the 2009 excavation and a second only identified among the additional material found in November 2012.

What is entirely missing is the helmet's own skeleton. No iron cap, and no trace of the leather or other organic structure that would once have supported the decorated sheet, survives; unlike Sutton Hoo, where an iron cap was recovered and a physical helmet could eventually be reassembled around it, the Staffordshire helmet's decorated panels have nothing solid left to be reassembled onto. Because physical reassembly of an actual object was therefore not possible, a project led by the School of Jewellery at Birmingham City University, working from 2014 with laser-scanned and CAD-modelled data taken from the surviving fragments, produced two full-scale replica reconstructions, using computer-milled and hand-finished dies to re-strike new silver sheet matching the original panels, with a blacksmith and a leather-worker supplying the missing iron and leather elements based on archaeological inference rather than surviving physical evidence. The two replicas, one for each owning museum, were unveiled together on 3 November 2018.

Both institutions involved are explicit that this is an informed interpretation of how the helmet probably looked, not a certain recreation of it. Tom Shippey, reviewing the full academic publication that followed the research project, notes a further, more fundamental uncertainty: given how much of the reconstruction rests on comparing and pattern-matching a large number of small, damaged fragments rather than on any surviving overall shape, it is not impossible that the fragments actually come from more than one helmet. The identification of a second cheek-piece only in the 2012 finds, three years after the main excavation, is a reminder that the fragment set used to build the 2018 replicas may itself still be incomplete.

Valuation and the fight to keep it in the region

The Treasure Valuation Committee, an independent panel chaired by Professor Norman Palmer, met at the British Museum on 25 November 2009 and valued the hoard at 3.285 million GBP under the Treasure Act 1996. Both museums hoping to acquire it, Birmingham Museum and Art Gallery and the Potteries Museum and Art Gallery, and both Terry Herbert and Fred Johnson, accepted the figure. Under the Act's reward provisions, finder and landowner were each entitled to half.

Raising that sum for two regional museums, rather than allowing the hoard to be sold on the open market, required a dedicated public campaign. The Art Fund launched the appeal on 13 January 2010 with an initial grant of 300,000 GBP, assisted by the Wolfson Foundation, and pledges of 100,000 GBP each from Birmingham and Stoke-on-Trent City Councils. The National Heritage Memorial Fund contributed 1.285 million GBP, and the public gave a further 900,000 GBP, of which 640,000 GBP was raised through the Art Fund's own members and supporters, with further trusts making up the balance. The full 3.285 million GBP was secured several weeks ahead of the campaign's own deadline, and the hoard was formally acquired jointly by Birmingham Museums Trust and the Potteries Museum and Art Gallery in 2010, becoming, by the Art Fund's own description, the most valuable Treasure find ever made on British soil.

The human aftermath of that sum is less well documented but worth recording plainly. A year after the discovery, Herbert, still using the same battered metal detector that had found the hoard, said he remained exactly the same person and felt a bit bewildered by what had happened. According to at least one account from within the detecting community, the windfall caused a lasting rift between Herbert and Johnson that had not been repaired by the time that account was written; this record notes the claim but treats it cautiously, since it rests on a single source not corroborated elsewhere in the material consulted.

After acquisition: research, redisplay, and what is still unresolved

Historic England and the two owning museums funded a major research project running from 2014, involving specialists including Chris Fern, Tania Dickinson and Leslie Webster, which physically rejoined many fragments, produced the helmet reconstructions unveiled in November 2018, and culminated in the 2019 publication of The Staffordshire Hoard: An Anglo-Saxon Treasure, a research report of the Society of Antiquaries of London running to more than 580 pages, followed by a shorter book for general readers by Fern and Jenni Butterworth in 2022. It is this project's renumbered catalogue, rather than the temporary numbers first assigned during initial recording in 2009-2010, that is now the closest thing to a standard reference for individual pieces.

The Potteries Museum's own Staffordshire Hoard gallery in Stoke-on-Trent was substantially redisplayed in 2023, removing an earlier reproduction helmet in favour of clearer interpretive material and adding regional early medieval comparanda from nearby sites at Wychnor, Stapenhill and Catholme, to place the hoard within its wider Mercian landscape rather than presenting it in isolation. In the summer of 2022, a group of Staffordshire Hoard pieces, including a gold and garnet sword pommel, a filigree hilt collar, a mount with horse-head terminals, a zoomorphic seax collar and a gilded silver pommel cap, travelled to Sutton Hoo in Suffolk for the exhibition Swords of Kingdoms: The Staffordshire Hoard at Sutton Hoo, which ran from 19 May to 30 October 2022 after a two-year pandemic delay from its original planned date. Original Sutton Hoo material on loan from the British Museum, including a gold and garnet shoulder clasp, a gold and garnet sword pyramid and a gold coin, was shown alongside it: the first time objects from the two collections had been displayed together. Curators involved, including Sutton Hoo's Laura Howarth and the Staffordshire Hoard specialist Chris Fern, presented the pairing explicitly as a meeting of two of the most important, but distinct, Anglo-Saxon discoveries in England, and argued, on stylistic grounds, that some Staffordshire pieces may have been made in the same East Anglian workshops that produced material found at Sutton Hoo. That connection remains an argued proposal rather than a documented fact.

The two finds should not be run together. Sutton Hoo is a 1939 excavation of an intact, furnished ship burial in a different county, Suffolk, with a body-shaped void, grave goods still in their original relationships, and a named leading candidate for its occupant, Raedwald. The Staffordshire Hoard is a 2009 metal-detector find in Staffordshire of dismantled, dispersed war-gear fragments with no burial, no body and no grave goods in the conventional sense at all. Nor should either be confused with the wholly separate Cheapside Hoard, a cache of roughly 500 pieces of Elizabethan and Jacobean jewellers' working stock found in a London cellar in 1912 and generally thought to have been buried around the time of the English Civil War in the 1640s, now held mainly by the Museum of London: a different period, a different country's material culture, and a different reason for burial, a jeweller's stock hidden for safekeeping rather than war-gear disposed of after battle. All three are major, genuinely important English hoards; none of the three is the other.