Friday, 9 December 2016

Rewriting history: the fascinating coins of the Watlington hoard - part 1

In 2015 an important Viking silver/ gold hoard was discovered near Watlington, Oxfordshire. The hoard appears to have been in Viking hands when it was buried in c. 879/80 AD. It contains a characteristically Scandinavian mix of coins, ingots and both complete and hacked jewellery. But its standout feature is its c. 200 silver coins. All but a handful were minted by the Anglo-Saxon kings Alfred (the Great) of Wessex (ruled 871-899) and his much less well-known contemporary Ceolwulf II of Mercia (ruled c. 874-879). Surviving coins of the 870s are rare. As explained in this new mini-book, the hoard casts vital new light on one of the most formative periods of medieval British history.

The Watlington hoard included a typically-Viking mix of coins, ingots, arm-rings and hack-silver, in addition to a tiny piece of hack-gold. (Copyright Ashmolean Museum)

Ceolwulf II is barely mentioned in the main documentary source for the period, the Anglo-Saxon Chronicle. This pro-Wessex source, first written down in c. 892, describes Ceolwulf as a ‘foolish king’s thegn’: a puppet ruler put in place by the Vikings to rule over western Mercia on their behalf in the late 870s, while they were busy attempting to subdue other Anglo-Saxon kingdoms.

But the coins tell a different story. In the Watlington hoard, coins issued by both Alfred and Ceolwulf (individually, rather than jointly) share the same design, and in some cases the same moneyer and mint place, suggesting some kind of monetary alliance on a grand scale: only possible if Alfred accepted the legitimacy of Ceolwulf and was willing to partner up with him in public. One of these is called the ‘Two Emperors type” because of the image of – you guessed it – two emperors on the reverse of the coin. Only two were known before this new discovery, with the current total now 15. Clearly, the coins were far more numerous than previously thought. The silver content of the coinage also seems to have uniformly increased around this time, further suggesting joint action.

A 'Two Emperors' type coin. The design was copied from earlier Roman coinage, but the message conveyed by the image of joint rulers was probably not lost on contemporaries. The winged figure above might be an angel, or a figure of Victory. If the latter, it's possible that the coinage was issued to celebrate the success of an otherwise unknown military coalition against the Vikings. (Copyright Ashmolean Museum)

Despite this monetary alliance, the relationship between Alfred and Ceolwulf quickly soured as Alfred moved to annex the kingdom of Mercia in around 880, and the shadowy Ceolwulf disappears from the historical record altogether (perhaps having been invited for a refreshing cliff-top walk). The fact that the coin types represented within the hoard have so far proved rare may indicate that Alfred deliberately melted them down following his takeover of Mercia, perhaps wanting to erase all reminders of Ceolwulf and their (now-embarrassing) former relationship. Indeed, Alfred began to issue a new coin type (the ‘Two-Line type', so called because…….yes, it has two lines of writing on it) at exactly this time.

A 'Two-Line type' coin of Alfred, found in Wiltshire. The two lines on the reverse of the coin read 'WLF RED' ie. Wilfred, the moneyer who produced the coin. (Copyright PAS). 

One of these new issues is included in the hoard - but just one. It seems that the Watlington hoard was assembled and deposited at a key moment of transition, when the Alfred-Ceolwulf coins were being melted down, but before Alfred had time to produce lots of his new 'Two-Line type'. Were is not for this hoard, we might well have consigned Ceolwulf to the margins of history. But the coins suggest that Ceolwulf was a legitimate ruler whilst painting Alfred was a shrewd manipulator of the facts.  

Monday, 8 August 2016

PODCAST: The Means of Exchange in Viking England

A 50-minute research talk from October 2015 about my current project: the bullion economy in Viking England. Here!

Wednesday, 13 April 2016

An unexpected absence: Anglo-Saxon silver in Viking hoards

The recent excavation of a Carolingian silver cup found in 2014 as part of the Galloway hoard revealed a number of extraordinary objects. Among them was a series of large Anglo-Saxon silver disc brooches, in styles that date them to the ninth century.  Such objects, worn by the Anglo-Saxon elite (possibly in pairs) seem like an obvious choice for inclusion in a Viking hoard.  Yet rarely are Anglo-Saxon silver objects found in such contexts.

One of the 9th century Anglo-Saxon silver brooches, here with preserved textile,
from the Galloway hoard (copyright Historic Scotland) 
Anglo-Saxon silver items such as brooches, mounts and strap-ends, ought to have been highly prized by the Vikings. Surviving ninth-century silver is usually good quality (over 90% precious metal), an important feature for the Vikings who carefully screened for debased metal. It is also highly decorative, frequently depicting lively, semi-naturalistic animals in the so-called Trewhiddle style (as seen on the brooch below). These are given added emphasis by niello inlay, a black copper/ silver sulphide, which provides a striking contrast with a silver background.  In Viking hands, such silver could be worn as high-status dress items, or broken up as a form of hack-silver currency.

A fragmented Anglo-Saxon silver brooch with niello inlay, depicting 9th century animal ornament.
The brooch formed part of an Anglo-Saxon hoard from Norfolk.  We know from the date of coins found with the brooch that it was buried in or around 869 AD, the year the Viking Great Army invaded East Anglia. The brooch is 95% silver. (Copyright Portable Antiquities Scheme) 
But the Anglo-Saxon brooches from the Galloway hoard are unique in their context, and other Anglo-Saxon items in Viking hoards notably few and far between. The largest Viking hoard on record from Britain, from Cuerdale, Lancashire, contains over 1,100 silver artefacts, but just a handful of Anglo-Saxon pieces, including fragments from a disc brooch, a brooch pin, and fragments of a strap-end and mount.  Anglo-Saxon silver jewellery is also rare within Scandinavia, suggesting that the Vikings only occasionally transported it home. Rare examples include two broken and cut Anglo-Saxon disc brooches from Sweden (from the Sturkö and Igelösa hoards), which likely functioned as bullion currency.

Why this unexpected absence? The Galloway hoard demonstrates that Anglo-Saxon silver objects must have passed through Viking hands, and Anglo-Saxon silver coins are a common component of hoards from Britain. Clearly, it is not a question about access to silver. 

The answer may lie in the form of the silver itself. Large disc brooches were not part of Scandinavian dress, while the distinctive animal and vegetal decoration of precious metalwork set it apart from indigenous Viking styles.  I think it’s likely that Anglo-Saxon silver objects were routinely melted down into forms that the Vikings found more ‘acceptable’, such as ingots and rings.  Silver is these forms fit more readily with the Viking aesthetic and could be easily assimilated into their currency systems.

In principle, then, the Scandinavian bullion economy made use of all silver, regardless of its form. But in reality, it preferenced silver in a limited range of homegrown and instantly recognisable shapes. 

Friday, 16 October 2015

Where were dirhams fragmented?

A good way of learning about the uses of silver within a bullion economy is to measure the extent of silver fragmentation. Whereas the frequent occurrence of highly fragmented silver suggests that silver actively circulated as money (that is, it often changed hands in commercial settings), the dominance of large, complete silver items suggests that silver was mainly stored as wealth.

One source of Viking silver that is often found in a fragmentary condition is Arabic dirhams – the critical source of silver fuelling the Viking bullion economy. Patterns of dirham fragmentation have the potential to reveal important insights into the local use of silver.  But, first, we have to establish where dirham fragmentation took place.

A big question is: were dirhams broken up near their source, in the Caliphate, or following their import into the Baltic and Scandinavian lands? The answer is, most likely, in both areas.  From c. 850 AD Near Eastern mints stopped producing coins to fixed weight standards, and a couple of decades later, broken dirhams start to appear in local hoards. Such fragments were probably required to ‘top up’ weighed payments to the required sum. Some hoards contain upwards of 40% fragments, suggesting that these pieces formed an important part of local monetary circulation. So some dirhams arriving in Scandinavia after c. 850 will have been fragmented already.

But it’s also clear that dirhams were further fragmented within Scandinavia. This is shown by the fact that the degree of dirham fragmentation across Scandinavia varies region by region. Writing in the 1980s, the Swedish numismatist Brita Malmer found that just 35% of hoarded dirhams from the Swedish island of Gotland were fragmented, compared to c. 66% of dirhams from the Swedish mainland provinces of Södermanland and Småland and 91% of dirhams from Östergötland. What’s more, the fragments from Östergötland were much more finely divided than those on Gotland. 

These patterns may partly reflect differences in local hoarding practices rather than the overall balance of fragmentary and whole dirhams in the local currency, but it seems clear that some dirham fragmentation took place locally.  In addition, whereas dirham fragments from Near Eastern hoards are usually roughly broken, many fragmented dirhams found within Scandinavia have been cut to sharp edges, probably with a chisel, and this provides a strong clue that division took place locally. Some dirhams found in Scandinavia have both broken and cut sides, suggesting at least two different locations of fragmentation (for examples from Kaupang, Norway, see the chapter by Mark Blackburn in this publication).
A complete dirham from Northamptonshire. Half of dirhams recorded singly in England are complete. (Image PAS)

Exactly where dirhams found in England were cut is difficult to say, and the answer will no doubt vary from coin to coin. Dirhams travelling as far west as England probably changed hands many times along their journey from the Caliphate, and thus stood a high chance of being cut at some point. Among the c. 60 single finds of dirhams I’ve recorded from England, half are whole (as the example shown above) and half are cut. The cut coins have an average weight of just 0.59g: less than a quarter of a complete coin (weighing c. 3g).

A very thin dirham fragment from Torksey, Lincolnshire, weighing just 0.38 grams. The edges are roughly broken, rather than cut, perhaps suggesting that it was fragmented in the caliphate. (Image PAS)

Interestingly, dirhams from Viking military camp sites in England are even more finely divided. At Torksey, Lincolnshire, occupied by the Viking Great Army in 872/3, over 90 dirhams have been found, but none are complete – they are all fragments with an average weight of just 0.44g: less than 1/6th of a complete coin. Notably, even ‘young’ dirhams that had only recently been minted, are very fragmentary.  Assuming that extremely fragmented dirhams were unlikely to have travelled over long distances in a short amount of time (simply because they would likely be lost in the process), this suggests that these coins were at least partly fragmented at Torksey. However, the sliver of a dirham shown above is roughly broken, rather than cut.  It may have been entirely fragmented in the Near East. 

In sum, the evidence suggests that at least the final stages of dirham fragmentation were carried out locally, rather than from a central source. Patterns of dirham fragmentation do, then, seem to be a good guide to regional bullion needs and economic structures. 

Friday, 24 July 2015

The art of the silversmith: stamped decoration on Viking silver

Stamped decoration is ubiquitous on Viking silver. This blog takes a quick look at how it was executed.

Within Scandinavia, the two most common shapes for stamps to take were ring-stamps and triangles, often containing 1 or 3 pellets. Often, two triangles were placed end-on-end creating an hourglass shape.

Such simple shapes were relatively easy to produce. In most cases, a punch was applied to the surface of the object soon after casting. To avoid a double impression, the punch had to be made in a single blow. Actual examples of punches used to create the ornament are rare, but one example comes from a late Viking-Age tool chest found at Mästermyr on Gotland. This is essentially a square iron rod, which originally had an hourglass shape on the striking face (it’s now very worn). A lead pad from the same tool chest seems to have been a testing piece for this stamp and others: it is covered on both sides with stamped hourglasses and rings.

Lead trial piece for triangular stamps,
from Torksey, Lincolnshire
(my photo, taken at the Fitzwilliam Museum)
But although the stamps themselves are rare, there are a few examples of other trial pieces, always made of lead. Three come from the Viking winter camp of Torksey, Lincolnshire, which the Vikings occupied in 872/3, at the height of their raiding activity in Britain. One of these, pictured to the right, has about twenty punches of a triangle with three dots – a form of decoration that could have been applied to Scandinavian neck-, arm- or finger-rings. When not busy taking over Anglo-Saxon kingdoms, it seems the Viking army was producing decorated silver ornaments (among other items). 

Despite the practice that the trial pieces imply, surviving ornaments show that the application of stamps was not an exact science – stamps were not always evenly applied, and some even contain cracks (see blog post below). This is the case with some unusual trefoil-shaped stamps on a silver arm-ring from a Viking-Age hoard from Yorkshire, which I studied at the Yorkshire Museum. The first stamps were applied in the middle of the ring, and are intact, but as the silversmith moved towards the outer edges of the ring, a crack appeared in the punch. This is clearly visible where the lozenge ornament has split, especially on the upper level of the ring.

Silver arm-ring from Yorkshire with stamped trefoils. The punch cracked partway through, resulting in splits in the deception. My photo, taken at Yorkshire Museum.

Interestingly, it wasn’t always necessary to punch the silver after casting. Sometimes, the ornament could be stamped into a beeswax model, and that model used to create a clay mould from which silver objects could be cast. This is clearly demonstrated by a terrific find from the harbour site of Fröjel, Gotland: it’s a Viking-Age clay mould which preserves stamped hourglass decoration, neatly picked out in the drawing (see here for other examples). This mould would have been used to cast decorated silver arm-rings.

A mould fragment (right) shown together with a modern clay imprint, with stamped hourglass decoration.
Image © Anders Söderberg
An illustrated database of stamps - building on James Graham-Campbell's recent work in this book - would be of great help in establishing workshop traditions and links between different objects.

Thursday, 9 April 2015

A modern fake exposed

Modern reproductions of Viking-Age artefacts are often of a really high quality. Indeed, sometimes they are so good that even the experts mistake them for genuine items. One such ‘fake’ is this Thor’s hammer pendant, said to have been found south of Carlisle. It was acquired by the British Museum in 1990.

A modern-day forgery of a Thor's hammer pendant
(image © British Museum)

Thor’s hammers are interpreted as miniature representations of Mjolnir, the mythical hammer of the pagan Norse God Thor.  In the Viking Age, both men and women wore them as pendants, probably for protective purposes. At the time of the ‘discovery’ of the Carlisle pendant, only one other Thor’s hammer was known from England (from the Cuerdale hoard). But since then, many more have come to light. In the context of these new finds, a few features of the Carlisle Thor’s hammer stand out. As James Graham-Campbell shows in a recent book, it can now be identified as a modern fake.

There are 3 main clues: 2 stylistic, and 1 scientific.

The first is its decoration. Triangular stamps are very common on Viking-Age artefacts.  But these typically contain just one or three pellets: the use of six pellets, as on this hammer, is extremely unusual. Zooming in to see these triangular stamps up close, it's clear that they have been applied very precisely (although not symmetrically). Indeed, it looks like they were cast in the mould, rather than applied to the pendant after casting. Such precision is unusual. Normally, such stamps are applied casually - they overlap, are unevenly executed and, in general, are a bit wonky. The image on the right shows the stamps on a silver arm-ring from the Vale of York hoard (including triangles with three pellets). These give a much more accurate impression of the true character of Viking stamped decoration.

The triangular stamps are too precise,
and the use of 6 pellets is atypical
(image © British Museum)
The overlapping and lop-sided stamps on this arm-ring are more
representative of Viking-Age decoration (image © British Museum)

The second feature that raises suspicion is they way the pendant is suspended. The ends of the suspension ring are secured by a rivet. But usually the ends of suspension rings, and indeed rings in general, are simply twisted together. The plain globular knob, through which the suspension ring passes, is also out of place.  Thor’s hammers from England usually have either a looped end, as is the case with this example from Norfolk, or a simple perforation through the ‘handle’.

This genuine Thor's hammer pendant from Norfolk has a
typical suspension loop (image © PAS)
But what clinches it for me is its metal content. Analysis at the British Museum revealed that the pendant was made of 93% silver, with the rest being copper. What was missing was gold. Gold naturally occurs in silver and is found in small amounts in most ancient ‘silver’. Crucially, however, it is refined out of modern silver.  Its absence in this Thor’s hammer therefore suggests that the pendant is modern.

So how did the maker of this pendant (whose identity is not known) come up with this design? As James Graham-Campbell points out, in the early 1980s a very popular Vikings Exhibition was held in York. The accompanying exhibition catalogue contains a drawing of a man in reconstructed Viking dress. Around his neck hangs a Thor's hammer decorated with triangular stamps containing 6 pellets. Although it can't be proved, it's likely that this artistic interpretation served as the inspiration for the modern-day forger.

Tuesday, 4 November 2014

The Dumfries and Galloway hoard: an archaeological context

In October this year a remarkable Viking hoard hit the headlines. It was discovered in Dumfries and Galloway (southwest Scotland) and contained both gold and silver items. To judge from the pictures posted by Treasure Trove Scotland, it’s a corker. Alongside familiar objects such as silver ingots and arm-rings, the hoard contained an Anglo-Saxon silver Christian cross with enamelled decoration, an Anglo-Saxon gold bird pin and, most spectacularly of all, a lidded Carolingian silver vessel filled with (as yet unexcavated) objects. Textile remains suggest the pot was originally wrapped in two layers of cloth. 

Some of the objects from the Dumfries and Galloway hoard. (© Treasure Trove Scotland)

The as yet unexcavated lidded Carolingian
silver cup, with textile remains
 (© Dumfries and Galloway Council)
My map of Viking-Age hoards (and single finds) from northwest England. Dumfries and Galloway is the territory at the top, hosting the compass. Red lines = Roman roads. Blue lines = Rivers (© Jane Kershaw)
The objects are stunning, and reveal the far-flung contacts and sources of wealth of Vikings operating in this area sometime around 900 AD. But is the hoard really so extraordinary? And what is its broader context?

Although it contains some stand-out objects, the Dumfries and Galloway hoard appears to fit into an established pattern of Viking hoarding in the lands surrounding the Irish Sea. Northwest England has produced a whole series of silver Viking hoards (17 at my last count) mainly dated to the first three decades of the tenth century. These include famous ‘historic’ hoards, such as that from Cuerdale, Lancashire, discovered in 1840, as well as recent discoveries, such as this hoard uncovered in Silverdale, Lancashire, in 2011. The distribution of the hoards (see below) shows that they are clustered around Morecombe Bay and, further south, around Chester.  The Dumfries and Galloway hoard is a more northerly findspot, but would still sit happily on this map.

Many of the artefact types contained within the D and G hoard have parallels from the same area. These include silver ingots and the punched-decorated arm-rings, known in academic circles as ‘Hiberno-Scandinavian broad-band’ arm-rings (if you say it enough times it just rolls off the tongue!). These artefacts, forming the circle in the group picture above, were produced in Scandinavian Dublin, and are also found in hoards from Huxley (Cheshire) and Silverdale (Lancashire) among others.

In fact, the Carolingian vessel has precedent. Ninth-century silver Carolingian cups formed part of the Halton Moor (deposited in c. 1025) and Vale of York hoards (deposited c. 927-8) and look very similar to the new find. These cups were originally used as liturgical containers for bread, incense or consecrated oil and were likely seized during Viking raids on the Continent. They must have been heirlooms when buried in Britain. Neither of the two known cups has a lid, however: a unique feature of the Dumfries and Galloway vessel.

So why was so much silver concealed around the Irish Sea in the early tenth century? The Sea was a hub of Viking activity, connecting Scandinavian settlements in Dublin, the Isle of Man, the western coast of mainland Britain and York. And these were turbulent decades. Scandinavian elites had been exiled from Dublin in 902 and sought new lands in Cheshire and coastal Lancashire. At the same time, Anglo-Saxon rulers from the kingdoms of Mercia and Wessex looked to extend their authority into a power vacuum left by the earlier collapse of Northumbria. The, often violent, upheavals would have necessitated the safeguarding of wealth accumulated through raids or trade at York and Dublin.

Clearly, this hoard has much still to tell us, not least regarding the contents of the Carolingian cup. It's a very exciting addition to Britain's Viking hoards.