This article is part of a series about our increasing understanding of the meanings behind the designs of Anglo-Saxon art. For other chapters click here.
Secrets in the Stones: Decoding Anglo-Saxon Art. Part 3
Early Anglo-Saxon Temple Discovered
Secrets in the Stones: Decoding Anglo-Saxon Art. Part 3
Early Anglo-Saxon Temple Discovered
A gold-and-garnet pommel from the Staffordshire Hoard, which once adorned the jewelled hilt of the sword of an early Anglo-Saxon prince or king, shows what appear to be vaults or arches beneath 'triangular pediments', hiding in the patterns of its garnets.
Once dismissed as an imaginative vision of the architecture of Rome by a culture whose architecture was limited to wooden huts and halls, new analysis we present here supports a radically different interpretation: that the pommel is a precise representation of a sophisticated and uniquely 'Anglo-Saxon' building, made decades or even centuries before such structures were previously thought to exist.
Could this sword mount provide the earliest glimpses of a lost Anglo-Saxon temple?
Once dismissed as an imaginative vision of the architecture of Rome by a culture whose architecture was limited to wooden huts and halls, new analysis we present here supports a radically different interpretation: that the pommel is a precise representation of a sophisticated and uniquely 'Anglo-Saxon' building, made decades or even centuries before such structures were previously thought to exist.
Could this sword mount provide the earliest glimpses of a lost Anglo-Saxon temple?
|
Copyright and citations notice
The synthesis between observations made in this article is the intellectual property of Aed Thompson, James D. Wenn and Thegns of Mercia. The form of words is copyright Aed Thompson and Thegns of Mercia, 2023. Pictures, diagrams and links remain property of their respective copyright holders. Licences may be sought for use of Byrga Geniht Ltd images and text by contacting Byrga Geniht Ltd. Thegns of Mercia hold a permanent licence for use of Byrga Geniht Ltd text and images within this article. Thegns of Mercia is the publisher of this article. Citations of this article must reference Aed Thompson, James D. Wenn, and Thegns of Mercia, and link to this article or reproduce its URL. |
|
Pommel cat.52 (formerly k284) is the most elaborate of 23 sword pommel-caps in the Staffordshire Hoard with gem settings, themselves a subset of the at-least 74 sword pommel-caps* in the Hoard. These pommel-caps were the most visible part of swords when worn, so became an important focus for visual display, from communication of status and/or identity to symbolism and apotropaic ‘magic’.
Most of the gold-and-garnet cloisonné items in the Staffordshire Hoard belong to Fern’s ‘Hoard Phase 3’ (Fern et. al. 2019) alongside the more mature Salin-Style-II animal interlaces, and Staffordshire pommel-cap cat.52 (henceforth, ‘pommel 52’) is a rare example in the Hoard where figural designs (rather than merely tessellating pavements) are worked into the arrangements of the garnets, requiring considerable design as well as technical skill. The most famous comparanda are the Staffordshire Hoard garnet seax fittings, and the beasts on the Sutton Hoo Mound 1 shoulder-clasps, both of which skilfully marry cloisonné with animal-art.
*The 'pommels' in the Staffordshire Hoard were delicate decorative hollow mounts which originally fitted over one, or sometimes two other nested fittings of similar shape which are also called 'pommels'. Strictly, therefore, they are "pommel-caps" or even "pommel-cap sheaths". For more on this distinction and how the various parts fitted together, see our article on early Anglo-Saxon sword anatomy here. For the purposes of this article, the sword fitting under discussion is referred to as "Pommel 52" as this was how it was named in the Staffordshire Hoard publication Fern et. al. (2019).
Another, of similar craftsmanship, is the gold pommel-cap from Dinham, Ludlow, where garnets are used to show a Calvary / crucifixion scene. At the base of the larger central cross of three, are two snake-like beast heads likely to be a split-representation allusion to the Anglo-Saxon cognate of Norse Níðhöggr - the snake or dragon which gnaws at the roots of Yggdrasil - the world tree.
Gold and garnet cloisonne pommel-cap from Dinham, Ludlow, Shropshire.
In Norse mythology Odin hung himself on Yggdrasil, speared in the side, in a ritual self-sacrifice to gain knowledge of other worlds. This syncretism (-the attempted amalgamation of different mythologies or belief systems; an important feature of Anglo-Saxon religion) between Yggdrasil and the crucifix (and by extension, of the hanged gods of wisdom; Woden and Christ) is described in clear detail in the Anglo-Saxon poem 'Dream of the Rood' and incised in runes on the Ruthwell Cross. The Dinham pommelcap - drawing together pagan and Christian imagery, is an important precedent which, as we will see, is of great relevance to Staffordshire Hoard pommel 52.
It is thought likely that most of Phase 3 of the Staffordshire Hoard (Fern et. al 2019) was manufactured in the same royal workshop(s) that produced the Sutton Hoo treasures, serving the kings of East Anglia, between around 610 and 650 CE, and that these items were looted by the Mercians during one or more of their wars with the East-Angles in the mid 7th century. The particular design of paired quadrupeds on pommel 52 has, however, also been noted to be similar to the pommel-cap from Crundale Down (Fischer & Soulat 2011; Fern 2019) so a Kentish origin is also possible.
Staffordshire Hoard pommel 52 showing Salin Style-II interlacing 'zoomorphic quadruped' design in gold & garnet cloisonne (Photo thanks to Matt Bunker)
Upper guard and pommel-cap from Crundale Down, Kent, dated to mid 7th century, and featuring Salin Style-II interlocking zoomorphic quadrupeds motif similar to the Staffordshire Hoard pommel 52.
At a glance pommel 52 may seem perfectly representative of the standard ‘cocked hat’ form typical in the Staffordshire Hoard (Menghin’s Typ Beckum-Vallstenarum) (Fischer & Soulat 2011), just as the Crundale pommel (above) is, but pommel 52 is actually modified in some important ways:
There is a hint of this development in the unique cast-gold pommel cat.57 of the same phase (here), and both arguably foreshadow the development of the bizarre ‘early insular’ pommels cat. 75 (here), pommel 76 (here), pommel 77 (here) and pommel fragment 78 (here), all of which (due to the end of sword-burial and/or regional biases in its occurrence) lack any satisfying comparanda.
Staffordshire Hoard pommel 52 therefore sits in the middle to a spectrum of evolution from the pommels of typical shape in the Hoard and elsewhere (pinned to the upper guard held together with separate rivets) to the larger ‘early insular’ pommels which entirely filled the space on the upper guard, with their hidden rivets also fixing the guard together, and onwards to mid-to-late Anglo-Saxon swords where the pommel and upper guard effectively merged to become one unit.
With its extra “wings”, as well as the tricky multi-faceted garnet settings on its shoulders bridging the faces (a very rare feature) this is not just the most elaborate gold-and-garnet pommel in the Staffordshire Hoard, but greatly exceeds the sophistication of the Sutton Hoo pommel and almost all comparanda from continental Europe and Scandinavia. It therefore seems likely that the sword it came from belonged to a king or prince for whom the typical design simply “wasn’t fancy enough”.
- The tunnels for the four pins - one on each corner - which would ordinarily serve to fix the pommel onto the upper guard of the sword, though maintained in the design, are redundant and instead capped off with tiny discs of plain gold.
- Buttresses, further decorated with cloisonné, extend from the vertical side-walls of the conventionally shaped cocked-hat element, together giving the pommel a more oval rather than rectangular footprint, effectively fully colonising the top of the sword hilt.
- Holes in the base of these elements show where hidden rivets once fixed the pommel to the guard, and also served to fix the upper guard together.
There is a hint of this development in the unique cast-gold pommel cat.57 of the same phase (here), and both arguably foreshadow the development of the bizarre ‘early insular’ pommels cat. 75 (here), pommel 76 (here), pommel 77 (here) and pommel fragment 78 (here), all of which (due to the end of sword-burial and/or regional biases in its occurrence) lack any satisfying comparanda.
Staffordshire Hoard pommel 52 therefore sits in the middle to a spectrum of evolution from the pommels of typical shape in the Hoard and elsewhere (pinned to the upper guard held together with separate rivets) to the larger ‘early insular’ pommels which entirely filled the space on the upper guard, with their hidden rivets also fixing the guard together, and onwards to mid-to-late Anglo-Saxon swords where the pommel and upper guard effectively merged to become one unit.
With its extra “wings”, as well as the tricky multi-faceted garnet settings on its shoulders bridging the faces (a very rare feature) this is not just the most elaborate gold-and-garnet pommel in the Staffordshire Hoard, but greatly exceeds the sophistication of the Sutton Hoo pommel and almost all comparanda from continental Europe and Scandinavia. It therefore seems likely that the sword it came from belonged to a king or prince for whom the typical design simply “wasn’t fancy enough”.
Some Technical Notes
While these gold-and-garnet pommels may appear quite solid, this is an illusion. They were actually fabricated from very thin gold sheet, soldered together, with a carefully shaped lattice of cell walls made of gold strip, soldered into place. It’s into these built-up cells that the garnets (once carefully and laboriously ground into shape) would then be added, with the tips of the soft gold walls pressed /rubbed, clenching over the corners of the stones to hold them in place. This means that wherever a design features an area of plain gold on the surface, rather than a garnet setting, a void has been capped off by soldering a ‘lid’ of gold sheet over the cell in place of inserting a garnet (Fern et. al. 2019). This is important, because it means that areas of plain gold, on cloisonné items, themselves required careful planning and precise work, and therefore do not occur by accident or laziness; they are generally avoided in the cloisonné designs unless the artist was trying to draw attention to particular garnet shapes, such as creating spaces of plain gold between the garnet beasts on one side of pommel 52 so they can be seen more clearly.
On pommel 52 the beasts were further augmented by use of contrasting garnets of different origins and colours. Spectroscopy revealed them to be a mixture of Rajasthan almandines, southern Indian almandines, and Czech pyropes (Fern et. al 2019). These were selected and arranged to enhance the contrast between the beasts on one side. The garnets of the other side (discussed below) included all three origins, but were selected to closely match in colour.
On pommel 52 the beasts were further augmented by use of contrasting garnets of different origins and colours. Spectroscopy revealed them to be a mixture of Rajasthan almandines, southern Indian almandines, and Czech pyropes (Fern et. al 2019). These were selected and arranged to enhance the contrast between the beasts on one side. The garnets of the other side (discussed below) included all three origins, but were selected to closely match in colour.
A Vision of Rome?
While one side of Pommel 52 shows a wonderful pair of style-2 beasts in the cloisonné argued to have ‘pagan’ significance (Fern et. al., 2019) the other side features what initially appears to be a purely pattern of garnets, of the kind which wallpapers many of the other Staffordshire Hoard items.
Closer examination, however, reveals that the pattern contains three equally sized rounded arches, and above them, what Fern (2019) describes as ‘triangular pediments’; the schematic provided in the publication highlights two of the ‘triangular’ elements but not the third which completes the design and reaches for the pommel’s apex.
Closer examination, however, reveals that the pattern contains three equally sized rounded arches, and above them, what Fern (2019) describes as ‘triangular pediments’; the schematic provided in the publication highlights two of the ‘triangular’ elements but not the third which completes the design and reaches for the pommel’s apex.
Staffordshire Hoard pommel 52, reverse, showing a representation of architecture. (Image from Staffordshire Hoard catalogue; "The Staffordshire Hoard: an Anglo-Saxon Treasure" Barbican Research Associates, 2017. Updated 2019. Funded by Historic England and the Archaeology Data Service. Used here under a modified CC BY-NC-S(PD) licence as per ADS Terms of Use
Byzantine Loom Weight (CC British Museum) We are not the first to notice that this scheme is architectural; Fern describes it as a “Christian Romanitas image - probably based on a basilica”, and compares it to a depiction on a Byzantine loom weight from the British Museum (BM. 1985,1015.2) but this comparison is not particularly convincing.
For clarity ‘basilica’ is used here to refer broadly to large, secular Roman civic halls which provided the blueprint for (or in many cases, became) churches in late antiquity (and thus, in modern parlance is effectively synonymous with ‘church’). The fashion of emulating all things ‘Roman’ (how ever clumsily misremembered or misunderstood) among the early Anglo-Saxons and related cultures, especially from the 7th century onwards, 'Romanitas' is well documented and studied, and this is a theme which runs strongly through much of the Staffordshire Hoard material. I have argued elsewhere, for example, that the serpentine filigree interlaces in Hoard Phase 2 should not be viewed purely through the lens of “Germanic” Style-II animal art, but also in the context of surviving late Roman mosaic patterns in Britain which were undoubtedly part of the early Anglo-Saxons’ ‘artistic landscape’ and which they were highly motivated to emulate. Looking past the material, to the patterns themselves, designs of Anglo-Saxon filigree and Roman mosaic interlaces show a high level of correspondence.
For clarity ‘basilica’ is used here to refer broadly to large, secular Roman civic halls which provided the blueprint for (or in many cases, became) churches in late antiquity (and thus, in modern parlance is effectively synonymous with ‘church’). The fashion of emulating all things ‘Roman’ (how ever clumsily misremembered or misunderstood) among the early Anglo-Saxons and related cultures, especially from the 7th century onwards, 'Romanitas' is well documented and studied, and this is a theme which runs strongly through much of the Staffordshire Hoard material. I have argued elsewhere, for example, that the serpentine filigree interlaces in Hoard Phase 2 should not be viewed purely through the lens of “Germanic” Style-II animal art, but also in the context of surviving late Roman mosaic patterns in Britain which were undoubtedly part of the early Anglo-Saxons’ ‘artistic landscape’ and which they were highly motivated to emulate. Looking past the material, to the patterns themselves, designs of Anglo-Saxon filigree and Roman mosaic interlaces show a high level of correspondence.
Thus, when he describes Pommel 52 as 'Romanitas' Fern (2019) is suggesting that the image on the pommel is a representation or emulation of Roman stately architecture - perhaps based on seeing ruins or preserved classical buildings in Britain, experienced via travel, or more likely, half-remembered via art, or imagined. The use of an obscure Byzantine loom-weight to illustrate this point rather than comparison to architecture directly suggests he favours the latter route for the seeding of such an image in the early Anglo-Saxon artist’s mind.
It is understandable when seeing a representation of stately architecture in a “Dark Age” British context that one’s mind might immediately jump to the crumbling legacy of the Roman Empire, or contemporary Constantinople.
But should we necessarily dismiss out of hand the idea that the Staffordshire Hoard jeweller was representing a real building in Britain - an Anglo-Saxon monumental structure serving as a symbol of power?
It is understandable when seeing a representation of stately architecture in a “Dark Age” British context that one’s mind might immediately jump to the crumbling legacy of the Roman Empire, or contemporary Constantinople.
But should we necessarily dismiss out of hand the idea that the Staffordshire Hoard jeweller was representing a real building in Britain - an Anglo-Saxon monumental structure serving as a symbol of power?
Anglo-Saxon Architecture?
The image of the early Anglo-Saxon built environment we are familiar with is one of humble villages of dark, dusty, thatched, sunken-featured buildings (so called grubenhauser / “grub huts”) as represented at sites such as West Stow, or, occasionally, slightly grander communal halls.
Typical floorplan 'Romano-Celtic' Temple - excavations at 56 Gresham Street, London, 3rd century - and artists impression (unknown source). Sites such as Yeavering / Ad Gedrin in Northumberland, Rendlesham in East Anglia, Sutton Courtney in Oxfordshire and Atcham in Shropshire (among others) have provided us with insights into grander “great hall complexes” - aisled halls of timber, up to 30x10m in size, which likely served as bases for royal / elite warrior courts (McBride, 2020).
Unfortunately as these are only represented by post-holes and footprints, and no reconstruction has ever been attempted, it is hard to picture the above-ground architecture or grandeur of these buildings. Their development from the late 6th and 7th centuries is, like princely burial, considered to mark a shift toward ‘monumental display as an expression of kingship and elite authority’ (Scull & Thomas 2022). Such sites have varied histories but many were relatively short-lived; the many layers of rebuilding at Yeavering indicate the short lifespan of such halls, and Blair (2018) suggests early Anglo-Saxon courts preferred to remain mobile, moving between sites, so investment in grand and truly permanent architecture only occurs with the development of church / monastic sites.
Contrary to popular belief, Anglo-Saxons did build monumental structures in stone.
Far too few such buildings survive today- victims of a thousand years of deliberate desecrations, ruinations and redevelopments, and none remain entirely in their original form. Survivors mostly date to the 9-11th centuries, but there are some which are much older.
Christian religious buildings are understood to have existed in Anglo-Saxon kingdoms prior to the arrival of St Augustine’s mission in 597 CE. St Martin’s Church, Canterbury - the personal chapel of Saint Bertha - the Frankish queen to the 'first' convert-king Æthelberht of Kent, is believed to have been rebuilt from Roman spolia in mid 6th century though relatively little of this original phase can now be seen. On the other hand, what little survives of structures associated with early Celtic Christianity suggests they favoured smaller, intimate spaces for private prayer rather than cavernous basilicas for collective worship, reminiscent of Roman shrines and smaller temples.
Looking further back, identified Romano-Celtic temples on mainland Britain mostly took a centrally-planned square form with a nested internal sanctuary surrounded by an ambulatory, perhaps providing an alternative blueprint for later religious buildings in Britain than the more typical 'basilica'.
Bede describes conversion-period kings keeping pagan temples and altars- but so far no building which can be described as a dedicated Anglo-Saxon ‘pagan temple’ has been convincingly identified. Building D2 at the Northumbrian royal site at Yeavering was interpreted as a pagan temple or ‘heathen hof’ based on a lack of any signs of other use, save for a pit containing ox skulls under one wall (Hope-Taylor, 1977). This identification may be wishful thinking, for it was otherwise identical to other small buildings on the site - approx 5x10m, post-built, and rectangular with single side entrances.
A totally different form - an open-air temple - may be represented by the baffling archaeology of Blacklow Hill 'ritual site', Warwickshire, where two Anglo-Saxon burials were found among countless precisely-cut shallow pits in the bedrock, enclosed with a double ring of massive square-section wooden posts creating what some have called a “sacred grove” (Foster, 2008); the squareness of the postholes, and their sub-rectangular layout, is suggestive that these were indeed early medieval rather than prehistoric. This site is entirely unique, however, and remains poorly understood. Burial archaeology suggests some level of early Anglo-Saxon engagement with prehistoric monuments including earlier barrows and henges, and both the early Mercian settlement at Catholm, Staffordshire, and the Bernician / Northumbrian royal settlement at Yeavering were, curiously, situated very close to prehistoric ritual complexes / woodhenge-style monuments. Truly convincing ‘heathen’ ritual or temple structures of early Anglo-Saxon date, if they existed at all, have so far eluded archaeologists.
However, if ‘pagan’ Anglo-Saxons did have a tradition of building, or using temples, perhaps on the Romano-Celtic model, it’s far from inconceivable that some might have survived the conversion to Christianity, becoming chapels, baptisteries, church-towers or mausoleums. The writings of Pope Gregory the Great - architect of Augustine’s mission to convert the Anglo-Saxons, gave clear instructions that pagan temples should be repurposed rather than pulled down.
Unfortunately as these are only represented by post-holes and footprints, and no reconstruction has ever been attempted, it is hard to picture the above-ground architecture or grandeur of these buildings. Their development from the late 6th and 7th centuries is, like princely burial, considered to mark a shift toward ‘monumental display as an expression of kingship and elite authority’ (Scull & Thomas 2022). Such sites have varied histories but many were relatively short-lived; the many layers of rebuilding at Yeavering indicate the short lifespan of such halls, and Blair (2018) suggests early Anglo-Saxon courts preferred to remain mobile, moving between sites, so investment in grand and truly permanent architecture only occurs with the development of church / monastic sites.
Contrary to popular belief, Anglo-Saxons did build monumental structures in stone.
Far too few such buildings survive today- victims of a thousand years of deliberate desecrations, ruinations and redevelopments, and none remain entirely in their original form. Survivors mostly date to the 9-11th centuries, but there are some which are much older.
Christian religious buildings are understood to have existed in Anglo-Saxon kingdoms prior to the arrival of St Augustine’s mission in 597 CE. St Martin’s Church, Canterbury - the personal chapel of Saint Bertha - the Frankish queen to the 'first' convert-king Æthelberht of Kent, is believed to have been rebuilt from Roman spolia in mid 6th century though relatively little of this original phase can now be seen. On the other hand, what little survives of structures associated with early Celtic Christianity suggests they favoured smaller, intimate spaces for private prayer rather than cavernous basilicas for collective worship, reminiscent of Roman shrines and smaller temples.
Looking further back, identified Romano-Celtic temples on mainland Britain mostly took a centrally-planned square form with a nested internal sanctuary surrounded by an ambulatory, perhaps providing an alternative blueprint for later religious buildings in Britain than the more typical 'basilica'.
Bede describes conversion-period kings keeping pagan temples and altars- but so far no building which can be described as a dedicated Anglo-Saxon ‘pagan temple’ has been convincingly identified. Building D2 at the Northumbrian royal site at Yeavering was interpreted as a pagan temple or ‘heathen hof’ based on a lack of any signs of other use, save for a pit containing ox skulls under one wall (Hope-Taylor, 1977). This identification may be wishful thinking, for it was otherwise identical to other small buildings on the site - approx 5x10m, post-built, and rectangular with single side entrances.
A totally different form - an open-air temple - may be represented by the baffling archaeology of Blacklow Hill 'ritual site', Warwickshire, where two Anglo-Saxon burials were found among countless precisely-cut shallow pits in the bedrock, enclosed with a double ring of massive square-section wooden posts creating what some have called a “sacred grove” (Foster, 2008); the squareness of the postholes, and their sub-rectangular layout, is suggestive that these were indeed early medieval rather than prehistoric. This site is entirely unique, however, and remains poorly understood. Burial archaeology suggests some level of early Anglo-Saxon engagement with prehistoric monuments including earlier barrows and henges, and both the early Mercian settlement at Catholm, Staffordshire, and the Bernician / Northumbrian royal settlement at Yeavering were, curiously, situated very close to prehistoric ritual complexes / woodhenge-style monuments. Truly convincing ‘heathen’ ritual or temple structures of early Anglo-Saxon date, if they existed at all, have so far eluded archaeologists.
However, if ‘pagan’ Anglo-Saxons did have a tradition of building, or using temples, perhaps on the Romano-Celtic model, it’s far from inconceivable that some might have survived the conversion to Christianity, becoming chapels, baptisteries, church-towers or mausoleums. The writings of Pope Gregory the Great - architect of Augustine’s mission to convert the Anglo-Saxons, gave clear instructions that pagan temples should be repurposed rather than pulled down.
“Tell Augustine that he should by no means destroy the temples of the gods but rather the idols within those temples. Let him, after he has purified them with holy water, place altars and relics of the saints in them. For, if those temples are well built, they should be converted from the worship of demons to the service of the true God. Thus, seeing that their places of worship are not destroyed, the people will banish error from their hearts and come to places familiar and dear to them in acknowledgement and worship of the true God.”
Gregory the Great, Letter to Mellitus, c597 CE.
[Epistola 76, PL 77: 1215-1216. Halsall, 2023]
St Peter-on-the-Wall, Bradwell-on-Sea, Essex, begun in the 660s today appears rather like a modest Anglo-Saxon hall in stone, for only the nave survives. That it had an apse accessed, reportedly, via three archways (as at the now ruined royal Kentish church at Reculver, from 669) places it firmly in the basilica tradition, as is the innovative late 7th century Mercian church at Brixworth.
At Brixworth, however the first phase focuses on the tall, narrow square tower (with later-added round stair-tower to provide easier access to the upper stories) - a monumental structure, visible in the landscape but itself providing only a small intimate sanctuary. Many surviving Anglo-Saxon churches were originally “turriform” - the tower, and space beneath it ('tower-nave' or 'sanctuary') being the central feature around which other extensions were added. It is tempting then, to imagine the earliest phase of Anglo-Saxon sacred spaces taking the form of small, usually square sanctuaries below a tower.
That such towers were important symbols of royal legitimacy and power can be seen in their representation on coinage- specifically coins of Edward the Elder and Æthelstan, where elaborate and decorative pilastering is clearly shown, reminiscent of the archetypal 10th century Anglo-Saxon church tower of Earls Barton, Northamptonshire.
That such towers were important symbols of royal legitimacy and power can be seen in their representation on coinage- specifically coins of Edward the Elder and Æthelstan, where elaborate and decorative pilastering is clearly shown, reminiscent of the archetypal 10th century Anglo-Saxon church tower of Earls Barton, Northamptonshire.
Other than heavily altered church towers, remnants of the 7th century include a small number of ‘crypts’ or vaults. At both Hexham and Ripon, small rectangular underground rooms constructed from Roman spolia are accessed by ambulatories in somewhat haphazard layouts. It might be productive to consider the parallels between these small rectangular crypts and the wooden burial chambers beneath mounds at Sutton Hoo, Taplow and Prittlewell, in which Anglo-Saxon kings and princes had been buried only a generation or two earlier. The wooden burial chamber at Prittlewell appears to have been kept open for some time, for the deceased occupant to be visited, with a folding chair beside the coffin for use by anyone sitting vigil. Environmental analysis suggests the mound continued to be tended generations after the burial (Blackmore et. al, 2019).
Were the crypts at Ripon and Hexham originally intended for this use? They have been compared to the tomb of Christ, but equally could reference the underground Roman cell / cistern - the Tulianum, in which St Peter and Paul were imprisoned. This has been commonly referenced in sacred architecture for smaller spaces of spiritual contemplation or transformation.
Were the crypts at Ripon and Hexham originally intended for this use? They have been compared to the tomb of Christ, but equally could reference the underground Roman cell / cistern - the Tulianum, in which St Peter and Paul were imprisoned. This has been commonly referenced in sacred architecture for smaller spaces of spiritual contemplation or transformation.
The Temple at Repton
At the Mercian royal site, double-monastery and ‘spiritual capital’ of Repton, a famous ‘crypt’ beneath the presbytery of medieval St Wystan’s Church, and approx three-quarters below ground level, is thought to have been constructed in the 8th century as a ‘baptistry’ (for more on the debated phases of construction of the Repton 'crypt' see footnote). Shortly after, it began to be used as a mausoleum for Mercian kings. Unlike the more haphazard layouts at Ripon and Hexham, this 'crypt' is a perfectly architecturally planned space. Precisely square except for niches in each wall (which once held sarcophagi / reliquaries), the room’s four famous spiralling columns, the square pilasters which align with them in the corners of the niches, and the arches which run between them, divide the room precisely into a ‘nine grid’ of squares. A raised lip at the base of the walls and columns suggests the uneven floor we stand on today was originally only the foundations, and that the original floor level - perhaps decorated - has been stripped away.
The "Repton Crypt" - a semi-subterranean baptistry / temple, and royal mausoleum, begun in the early 8th century. (Photo: Æd Thompson)
Plan of 5th century Baptistery of St John Lateran Small light wells (which have since been enlarged, losing their original shape) high up in the three niches allow shafts of light to enter the space and concentrate in the centre, between the four columns, such that anyone who stands there in the otherwise dark space is ‘bathed with light’. The scheme thus reflects square Romano-Celtic temples with a central square sanctuary - representing one ninth of the building - surrounded by an ambulatory. The “crypt” aligns closely, vertically, with the presbytery of the church above, which, featuring degraded Anglo-Saxon pilastering as well as characteristic, now redundant access doors for the upper stories, was clearly once an Anglo-Saxon tower which has been rebuilt at least once in the late Anglo-Saxon period (Biddle, 2018). The room originally had only one entrance, through what is now the niche in the west wall, so the original approach to the ‘crypt’ has been obliterated by development of the later church. This entrance, together with the three niches, thus formed the shape of a cross. Roughly hewn passages leading down to the northwest and southwest corners of the room are thought to have been added in the 9th century, probably following the interment of the remains of St Wystan (Biddle 2018), to manage the flow of pilgrims – effectively creating a ring-crypt layout – with the original entrance blocked and converted into another burial niche. (The door in the north niche was inserted in the 14th century.) (For more on phases of construction, see footnote).
The footprint of at least two 7th century great-halls surround the location of the 'crypt', evidencing a likely Mercian royal vicus preceding development of the monastery, and another enigmatic semi-subterranean stone building nearby - reused as a Viking mass grave - may be even older, perhaps built in the late 7th century (Biddle, 2018). Nevertheless it remains a mystery why Repton in particular was chosen (as opposed to other royal vici) for this special religious monument. The possibility of an earlier Christian, or even pagan temple on the site would provide a satisfying explanation. We cannot archaeologically trace the evolution of the ‘crypt’ on this site further back than the early 8th century but its plan, and how skilfully it was executed, suggests it is part of a well-established tradition of such buildings, possibly with more ancient antecedents. The design of the Repton 'crypt' appears to have been influential, for example being very similar to the ‘freestanding’ crypt with chapel above, which formed part of the 9-10th century phases of the development of St Oswald’s Abbey church at Gloucester (Blockley, 2000); likely commissioned by Æthelflæd of Mercia to house the newly recovered bones of St Oswald. Just like at Repton, when this square free-standing tower/chapel with crypt below, was integrated, a larger nave was attached to its west, converting the square chapel into the new presbytery. A great many later structures emulating Repton have already been identified.
Through the 8-9th century, if not afterward too, the Repton baptistery/mausoleum, and the altar directly above it, would have been the 'holiest' places within the monastic complex and royal settlement, and possibly, the entire kingdom. It therefore seems likely that the most charged ceremonies - such as any ritual approximating a later coronation, took place on this small square of land. It is, however, extremely unlikely that the Repton room was the first of its kind to be built in Anglo-Saxon Britain; similar buildings must have existed elsewhere, especially in the initially wealthier eastern kingdoms which Christianised earlier.
We see broadly the same design scheme (a centrally planned building with an ambulatory around, and an inner illuminated space defined by columns) in the earliest Roman baptisteries, including at St John Lateran in Rome, part of the palace of Constantine, which went on to serve as the blueprint for future baptisteries across Europe (and which, therefore, we argue descend from Romano-Celtic temples). The 9th century polygonal apse-above-crypt at Wing, Buckinghamshire, quite starkly appears to reference the Lateran Baptistery.
The architectural case that the Repton building was originally a baptistery before it was a mausoleum is thus very strong, though no evidence of a font or pool remains. A channel which surrounds the building (dating to “no later than 740”, Biddle & Kjølbye-Biddle 2001) is suggested to have been for accessing groundwater for baptism, but could easily have been for management of rainwater/groundwater which would flood the partly subterranean structure (see footnote).
Instead of directing water in, the room’s design controls the way in which light enters the space, like chapels which reference the Tulianum. It can be argued that Repton is, therefore, designed not for baptism with water, but baptism in light/fire, like Pentecost.
The footprint of at least two 7th century great-halls surround the location of the 'crypt', evidencing a likely Mercian royal vicus preceding development of the monastery, and another enigmatic semi-subterranean stone building nearby - reused as a Viking mass grave - may be even older, perhaps built in the late 7th century (Biddle, 2018). Nevertheless it remains a mystery why Repton in particular was chosen (as opposed to other royal vici) for this special religious monument. The possibility of an earlier Christian, or even pagan temple on the site would provide a satisfying explanation. We cannot archaeologically trace the evolution of the ‘crypt’ on this site further back than the early 8th century but its plan, and how skilfully it was executed, suggests it is part of a well-established tradition of such buildings, possibly with more ancient antecedents. The design of the Repton 'crypt' appears to have been influential, for example being very similar to the ‘freestanding’ crypt with chapel above, which formed part of the 9-10th century phases of the development of St Oswald’s Abbey church at Gloucester (Blockley, 2000); likely commissioned by Æthelflæd of Mercia to house the newly recovered bones of St Oswald. Just like at Repton, when this square free-standing tower/chapel with crypt below, was integrated, a larger nave was attached to its west, converting the square chapel into the new presbytery. A great many later structures emulating Repton have already been identified.
Through the 8-9th century, if not afterward too, the Repton baptistery/mausoleum, and the altar directly above it, would have been the 'holiest' places within the monastic complex and royal settlement, and possibly, the entire kingdom. It therefore seems likely that the most charged ceremonies - such as any ritual approximating a later coronation, took place on this small square of land. It is, however, extremely unlikely that the Repton room was the first of its kind to be built in Anglo-Saxon Britain; similar buildings must have existed elsewhere, especially in the initially wealthier eastern kingdoms which Christianised earlier.
We see broadly the same design scheme (a centrally planned building with an ambulatory around, and an inner illuminated space defined by columns) in the earliest Roman baptisteries, including at St John Lateran in Rome, part of the palace of Constantine, which went on to serve as the blueprint for future baptisteries across Europe (and which, therefore, we argue descend from Romano-Celtic temples). The 9th century polygonal apse-above-crypt at Wing, Buckinghamshire, quite starkly appears to reference the Lateran Baptistery.
The architectural case that the Repton building was originally a baptistery before it was a mausoleum is thus very strong, though no evidence of a font or pool remains. A channel which surrounds the building (dating to “no later than 740”, Biddle & Kjølbye-Biddle 2001) is suggested to have been for accessing groundwater for baptism, but could easily have been for management of rainwater/groundwater which would flood the partly subterranean structure (see footnote).
Instead of directing water in, the room’s design controls the way in which light enters the space, like chapels which reference the Tulianum. It can be argued that Repton is, therefore, designed not for baptism with water, but baptism in light/fire, like Pentecost.
'Baptism in fire' among columns and arches; the Pentecost scene in the Benedictional of St Æthelwold, 10th century.
Our main historic source for the events in the century of the making of the Staffordshire Hoard pommel, and preceding the first extant phase of the 'temple' at Repton - Bede’s Ecclesiastical History - is utterly preoccupied with the conversion / baptism of Anglo-Saxon kings; it makes sense that following the mission of St Augustine, baptisteries would be the first purpose-built monumental Christian buildings, sponsored by royal patronage, and that these would become symbols of royal legitimacy and power.
The Temple on the Sword
Looking at the Repton room from any side, we would see two of the four columns, and two pillasters, with three arches spanning them. We compare an antiquarian schematic of the Repton crypt’s architecture, unaltered, with an unaltered image of the Staffordshire Hoard pommel....
We see that the arches correspond perfectly. The pommel - made around 625-40 CE - shows us a stylised view of a vaulted Anglo-Saxon room like the Repton crypt.
What of the ‘triangular pediments’? Careful examination shows these not to be triangles, but lozenges. (We examined the early to middle Anglo-Saxon obsession with the lozenges in the previous chapter.) Another lozenge is added above, locking into the other two, and reaching the apex of the pommel.
Is this merely a decorative infill? No. Just as plain gold areas were built up, to provide space between the beasts on the other side of the pommel so they could be seen more clearly, so too, here, decorative infill of garnets on the shoulders of the pommel are interrupted with a narrow gold triangular insert on each side, to clearly define the outline of the three lozenges, so they can be seen as a precise design and not an infill pattern.
There is a slight change of angle between the silhouette of the lozenges; the upper one is very slightly flattened, as if distorted by perspective. It represents a roof.
Very few Anglo-Saxon tower roofs survive to this day, built as they were predominantly from wood, but rare survivors take a very particular shape called the “Rhenish Helm” - named for the greater number of surviving examples from (mostly post-11th century) churches in the Rhineland. The best surviving example in England is the 11th century St Mary’s Church, Sompting, where, like the ‘Ship of Thesus’ or ‘Trigger’s Broom’ the timbers have been continually replaced to preserve the roof’s original, iconic form. Internally the Rhenish Helm requires a fearsomely complex arrangement of trusses and was designed to produce both an image of geometric perfection and a powerful visual statement in the landscape through its uncanny height, and although Sompting is no older than the early 11th century, we know that older towers including the 10-11th century St Bene’t’s church, Cambridge, the 10th century church at Earls Barton Northants, and 9th century Barton-on-Humber, originally had Rhenish Helm roofs.
What of the ‘triangular pediments’? Careful examination shows these not to be triangles, but lozenges. (We examined the early to middle Anglo-Saxon obsession with the lozenges in the previous chapter.) Another lozenge is added above, locking into the other two, and reaching the apex of the pommel.
Is this merely a decorative infill? No. Just as plain gold areas were built up, to provide space between the beasts on the other side of the pommel so they could be seen more clearly, so too, here, decorative infill of garnets on the shoulders of the pommel are interrupted with a narrow gold triangular insert on each side, to clearly define the outline of the three lozenges, so they can be seen as a precise design and not an infill pattern.
There is a slight change of angle between the silhouette of the lozenges; the upper one is very slightly flattened, as if distorted by perspective. It represents a roof.
Very few Anglo-Saxon tower roofs survive to this day, built as they were predominantly from wood, but rare survivors take a very particular shape called the “Rhenish Helm” - named for the greater number of surviving examples from (mostly post-11th century) churches in the Rhineland. The best surviving example in England is the 11th century St Mary’s Church, Sompting, where, like the ‘Ship of Thesus’ or ‘Trigger’s Broom’ the timbers have been continually replaced to preserve the roof’s original, iconic form. Internally the Rhenish Helm requires a fearsomely complex arrangement of trusses and was designed to produce both an image of geometric perfection and a powerful visual statement in the landscape through its uncanny height, and although Sompting is no older than the early 11th century, we know that older towers including the 10-11th century St Bene’t’s church, Cambridge, the 10th century church at Earls Barton Northants, and 9th century Barton-on-Humber, originally had Rhenish Helm roofs.
Tower of St Mary's Church, Sompting, Sussex, with Anglo-Saxon Rhenish-helm roof surviving in its original form (left) and 9-10th century cast-bronze censor cover from Pershore, Worcestershire (right) (CC British Museum)
We further see Rhenish Helm roofs represented on 9-11th century Anglo-Saxon cast bronze censer covers, such as the one from Pershore, Worcs, designed to look like model church towers. These representations were once dismissed as ‘fantastical’ but can now be recognised as lifelike depictions of real Anglo-Saxon buildings. These predate all examples on the continent and suggest the Rhenish Helm originated in Anglo-Saxon England. It is not part of the architectural repertoire of ‘Romanitas’, and based on other clues we will explore in a future chapter, is likely to have only arrived in the Rhineland from Britain during Carolingianisation.
The pommel was made no later than 650, but most likely around 625-40; it logically follows that whatever ‘temple’ it represents - probably in East Anglia or Kent - must have been completed before this date.
Further, the Rhenish helm can only project from a square tower, telling us that the structure the jeweller is showing us, with its three arches, was not a generic three-arched portico, but had a precisely square footprint like the Repton baptistery built a century later.
The Staffordshire Hoard pommelcap shows us a vision of an Anglo-Saxon tower temple or baptistery, with a distinctively Anglo-Saxon roof, decades or even centuries before such things were previously thought to exist. Was this the first purpose-built Christian sacred space in the kingdom of East Anglia - or could it be older? Might this be an image of the temple of King Rædwald in which, famously, according to Bede, dual altars were kept; to both Christ and the pagan gods?
Either way this unique image of an early Anglo-Saxon temple tells us that early Anglo-Saxons were more intellectually sophisticated, knowing more about complex geometry and architecture, than they have ever, hitherto, been given credit for.
This provides us with one more spiralling column, upon which our next Rhenish-Helmed tower of analysis, decoding Anglo-Saxon art, soon to be published, has been built.
Further, the Rhenish helm can only project from a square tower, telling us that the structure the jeweller is showing us, with its three arches, was not a generic three-arched portico, but had a precisely square footprint like the Repton baptistery built a century later.
The Staffordshire Hoard pommelcap shows us a vision of an Anglo-Saxon tower temple or baptistery, with a distinctively Anglo-Saxon roof, decades or even centuries before such things were previously thought to exist. Was this the first purpose-built Christian sacred space in the kingdom of East Anglia - or could it be older? Might this be an image of the temple of King Rædwald in which, famously, according to Bede, dual altars were kept; to both Christ and the pagan gods?
Either way this unique image of an early Anglo-Saxon temple tells us that early Anglo-Saxons were more intellectually sophisticated, knowing more about complex geometry and architecture, than they have ever, hitherto, been given credit for.
This provides us with one more spiralling column, upon which our next Rhenish-Helmed tower of analysis, decoding Anglo-Saxon art, soon to be published, has been built.
Footnote: Issues with dating the phases of the Repton 'Crypt'
The phases of construction and/or modification of the Repton 'Crypt' are the subject of academic debate and have changed various times, such that information in the public domain about its construction is often contradictory. Dating the phases of modification of a relatively plain stone structure, with no satisfying or directly relevant architectural comparanda, from a time of quite wide architectural diversity, is fraught with difficulty. There is no 'carbon dating' for worked stone and we must accept that a final answer to these mysteries may never be forthcoming, with commentary instead based on scarce clues, logical conjecture, and informed opinion.
There is, in particular, scholarly disagreement as to when the iconic spiral / 'Solominic' columns and vaulting were added. They necessarily must pre-date the extant presbytery above, but Biddle (2018) suggests they cannot have been original, as one column overlies a now-hidden water channel beneath the floor which he believes to be for channeling baptismal water, and by comparison to the spiral columns of the tomb of St Peter in Rome, believes they can only denote sainthood and therefore cannot predate Saint Wytstan (c840 CE). Due to their similarity to columns either side of the ring-crypt-style ambulatory entrances, he suggests they likely accompanied the 9th century rebuilding of the adjoining church. It is well documented that Anglo-Saxon bishops and kings personally took pilgrimages to Rome from the mid 7th century onwards, with an Anglo-Saxon exclave (the 'Borgo' district) and 'Schola Saxonum' established beside the Vatican as early as the late 7th century. The seeding of this architectural reference to St Peters' could, thus, have occurred surprisingly early, and indeed, we see spiralling columns with remarkably similar capitals to those of the Repton 'crypt' on the canon-table pages of the late 7th - early 8th century Anglo-Saxon bible - the Codex Amiatinus.
Fernie (2018) suggests we cannot rule out that the columns and vaulting were added as renovations to the site in the 10th century, following its occupation by the Great Heathen Army and subsequent recapture (as part of Æthelflæd's recapture of nearby Derby), but Biddle (2018) points out there is no documentary evidence to support this later revival of the site, which must have been 11th century, leading to Cnut's removal of the relics of St Wystan to a new shrine in Evesham.
Although Fernie (2018) may strictly be correct that a later date for these features can't be totally ruled out, the case for it is not convincing. The vaults and columns appear to have been referenced, carved into the natural caves of nearby Anchor Church - recently found to be an 8-9th century hermitage (Simons, 2021). The carving of the columns (and pilasters - with matching capitals, likely to be of the same phase) do not match any known late Anglo-Saxon stonework from Derbyshire - an area unusually rich in surviving Anglo-Saxon stone, nor is there any strong resemblance to the huge collection of surviving mid-late Anglo-Saxon stonework at the priory church of Breedon-on-the-Hill, only 8 miles away, which was founded in the 7th century, sacked by Vikings in the 9th, and only re-founded in the 13th. Among Breedon-on-the-Hill's stonework, the best match are some 8-9th century friezes (in turn comparable to Peterborough's 'Hedda Stone' c870CE) showing saints, in threes, under three arches divided by columns. These columns have slightly offset lateral ribbing interrupted by a midline, as if we are seeing spiralling Repton-like columns in cross-section, but they terminate in devolved Corinthian capitals more associated with late Anglo-Saxon art. If anything, then, these pieces of frieze might reference the already-present vaults and columns of the Repton crypt, with embellishment, rather than being contemporary with it. Intriguingly the best analogue we can find locally, for the precisely-worked yet thick and austere pilasters (with their proud but elegantly rounded corner ribs) is actually the famous 'Repton Stone' which Biddle (2018) dates to the early 8th century and considers to have been part of a monument to King Æthelbald.
The pilasters - which are often overlooked- appear seamlessly integrated into the corners of the bays such that its hard to imagine they were not original features. Square in section but with their midribs reaching out of the middle of each side, with their corner ribs joining the capital with a gentle arch, these pilasters are in a sense an architectural microcosm of the structure as a whole. Roughly similar arrangements of proud, rounded corner-ribs terminating in an arch, defining a cartouche for decoration (of highly dateable c8th 'Mercian style') can be seen among remains of Anglo-Saxon masonry in Derbyshire including the collection of All Saints Church, Bakewell, some of which curiously feature corner-ribbing in the form of spiral columns. The robust, precise carving of the columns and the pilasters in Repton, eschewing any delicate decorative detailing we might expect to see on mid-late Anglo-Saxon carving (cf. Breedon-on-the-Hill) tempt us to consider an earlier date.
The case that the 'saintly' spiralling columns cannot pre-date Saint Wystan is twice undermined - Firstly, spiralling columns were not exclusive to saintly shrines in Anglo-Saxon architecture (cf. the use of spiral-column balusters in the tower of All Saints Church, Earls Barton). Second, we cannot say for certain that the mausoleum did not house remains considered saintly, prior to Wystan in the mid 9th century, given the paucity of the historic record for 8th century Mercia, the prior use of the room as a mausoleum by the Mercian royal dynasty, and the many saints associated both with them, and with Repton, not all of whom are accounted for.
The main remaining detail which suggests the columns were a later addition, is that one overlies the water-channel or gutter which runs around, and under, the room. Biddle (2018) considers this channel to have provided access to ground-water within the building, due to an observed tendency for the water-table on the site to periodically rise. However it seems odd that a feature so 'central' to the purpose of the original room should be haphazardly sited diagonally off-centre so as to only intersect with only one of the columns, and there is no sign in the architecture of the room that there was any feature for accessing this water in a more dignified way than scooping it off the floor - a curious ritual for baptism at a site right beside a river. Biddle (2018) notes that the other, possibly earlier stone 'crypt' on the site (which was later repurposed as a Viking mass-grave) also had a water-channel running away from the building, which he interprets as being a gutter for carrying away rainwater. Perhaps the channels around and under the main 'Repton Crypt' served the same mundane purpose, for dealing with the 'rising groundwater' already noted, which would otherwise regularly flood these semi-underground rooms. The architecture, nevertheless, arguably does place the original building firmly in the baptistry tradition, and as we have argued in the main article, it is quite possible for a baptistry to serve its spiritual purpose without a structural water-feature.
Finally, for the purposes of analysing the architecture's design and function here the precise date of installation of the columns and vault may be moot. The geometry of the room - with its niches in the form of a cross, and implicitly dividing the square into a nine-grid - belongs to the original phase and is not dependent on the columns or pilasters, although is made more obvious by them. We, further, cannot know what wooden structures may have formed part of the earlier phases, having later been replaced by stone but preserving continuity of meaning. It is plausible that the first phase of the Repton room therefore contained much the same pilasters and columns, of similar design and the same arrangement, but made of oak, and supporting a suspended wooden chapel floor above, with its own wooden roof, as was typical for all Anglo-Saxon churches.
For simplicity for the purposes of this article we have therefore taken the Repton room as-we-find-it, as being representative of a form of monumental, sacred architecture of the 8th century, to compare with the Staffordshire Hoard pommel 52.
References
Biddle, M., 2018. The Repton Sculptures. Corpus of Anglo-Saxon Stone Sculpture, 13, pp.51-68.
Biddle, M. and Kjølbye-Biddle, B., 2001. Repton and the “great heathen army”, 873–4. Vikings and the Danelaw, 45, p.96.
Blackmore, L., Blair, I., Hirst, S. and Scull, C., 2019. The Prittlewell Princely Burial: Excavations at Priory Crescent, Southend-on-Sea, Essex, 2003.
Blair, J., 2018. Building Anglo-Saxon England. In Building Anglo-Saxon England. Princeton University Press.
Blockley, K., 2000. The Anglo-Saxon churches of Canterbury archaeologically reconsidered (Doctoral dissertation, Durham University).
Fern, C., Dickinson, T. and Webster, L., 2019. The Staffordshire Hoard. An Anglo-Saxon Treasure (p. 640). Society of Antiquaries of London.
Fernie, E., 2018. The eastern parts of the Anglo-Saxon church of St Wystan at Repton: function and chronology. The Antiquaries Journal, 98, pp.95-114.
Fisher, S. and Soulat, J., 2011. The Typochronology of Sword Pommels from Staffordshire Hoard. Papers from the Staffordshire Hoard Symposium, British Museum
Foster, S., 2008. Religion and the Landscape–How the conversion affected the Anglo-Saxon landscape and its role in Anglo-Saxon ideology. In Univeristy of Newcastle-The School of Historical Studies Postgraduate Forum E-Journal (Vol. 6, pp. 1-19).
Halsall, P., 2023. Pope Gregory I letter to Abbot Mellitus c. 601 AD. Internet Medieval Sourcebook. [Online] [Url= https://sourcebooks.fordham.edu/source/601gregory1-lettertomellitus.asp] [Accessed 23/03/2023]
Høilund Nielsen, K., 2011. Style II and all that: the potential of the hoard for statistical study of chronology and geographical distributions. Papers from the Staffordshire Hoard Symposium, British Museum
Hope-Taylor, B., 1977. Yeavering: an Anglo-British Centre of Early Northumbria, Dept. of the Environment Archaeol. Reports, 7, pp.73-8.
Scull, C. and Thomas, G., 2020. Early Medieval great hall complexes in England: temporality and site biographies. Anglo-Saxon Studies in Archaeology and History, 22, pp.50-67.
McBride, A., 2020. The Role of Anglo-Saxon Great Hall Complexes in Kingdom Formation, in Comparison and in Context AD 500-750.
Simons, E., 2021. Anchor Church Derbyshire: Cave Hermitage or Summer House?: A Case Study in Understanding a Rock-Cut Building. In Proceedings of the University of Bristol Speleological Society (Vol. 28, pp. 347-60).
RSS Feed