08 January, 2012

Post-Roman York: the headquarters building

York was an important military, ecclesiastical and political centre in Late Roman Britain. In the early seventh century it was under royal control of the English (‘Anglo-Saxon’) kings of Deira, and later in the seventh century it developed into a major ecclesiastical centre and the seat of an archbishopric, a status it holds to this day. In between, the historical record is a blank; definite references to York between the fourth century and the seventh century number precisely zero, although there are one or two snippets whose meaning is less than clear (see earlier post on Post-Roman York: the documentary evidence for a summary of the documentary records). Evidence from archaeology provides some clues that may help to fill in the gap.

Minster excavations

Roman York was a legionary base from the date of its founding, and remained so throughout Roman rule in Britain. The legionary headquarters building (principia) was the most important building in a legionary fortress, always placed in the centre facing the main gate. It housed administrative offices, a great aisled cross-hall where the commander could address his assembled troops, a strong room for the legion’s pay chest and the soldiers’ savings, and the legionary shrine where the standards were kept. The principia at York was the centre of Roman military power in what is now north and north-east England.

The present York Minster is sited partly on top of the Roman principia (see my earlier post on the possible location of the seventh-century church in York for a sketch of the relative position of the two). When the Minster required urgent structural underpinning in the late 1960s to save the central tower from collapse, the engineering work provided a rare – possibly unique – opportunity for archaeological excavation on the site. One can only admire the archaeologists who carried out the excavation, which must have been the archaeological equivalent of keyhole surgery, conducted under difficult conditions in the middle of a major building project frantically trying to shore up a collapsing cathedral.

Clear, if sparse, evidence emerged of activity in the cross-hall in the late and/or post-Roman period. The Roman flagstone floor had been removed (at an unknown date), and replaced by multiple layers of trodden sand and charcoal. In some places these layers also contained fragments of ‘York ware’ pottery, which was in use in the eighth or ninth century (Rahtz; Carver 1994). In other places the layers contained large quantities of animal bone. Unusually, the animal bone contained a high proportion of pig (over one-third) and sheep (about one-third), instead of being dominated by cattle, and still more unusually, many of the bones were from juvenile animals less than a year old (Carver 1994). A sample of the bone was radiocarbon-dated to 343-416 AD (Rahtz). These layers were overlain by the collapsed roof of the cross-hall.


So, it seems that at some time after the removal of the Roman flagstone floor and before the collapse of its roof, the cross-hall of the principia had seen some activity that resulted in the accumulation of multiple layers of sand, a lot of juvenile animal bones and a few sherds of pottery. As so often, the dating of this activity is the subject of much debate. The earliest possible date of the roof collapse is constrained by the latest date of objects sealed beneath the collapse layer (provided the objects have not found their way under the collapse layer at a later date, see below). There are two groups of datable artefacts in the layers under the collapsed roof, the animal bone radiocarbon-dated to the late fourth-early fifth century, and the York ware pottery fragments dated to the eighth-ninth century.

  • If you take the deposition of the York ware pottery fragments as the last event before the roof collapsed, this suggests the cross-hall was standing until at least the late eighth or ninth century.

  • If you take the deposition of the radiocarbon-dated animal bone as the last event, this suggests the cross-hall was standing until at least the early fifth century.

In either case, the cross-hall could have been standing much later – it must have collapsed after the latest object underneath was deposited, but how long after is a different question.

The original excavator, Derek Phillips, interpreted the findings as showing accumulation of material, and therefore some sort of activity in the principia, up to the ninth century, whereas Martin Carver interpreted them as fifth-century activity (Rahtz). When two respected practitioners disagree by 400 years, it probably tells you that there is not a definitive answer. It seems to me to depend on whether you think the York ware fragments were lying on the floor underneath the roof when it collapsed, or whether you think they were intruded underneath the collapsed roof at some later date, perhaps when the remains of the Roman principia was being used as a giant stone quarry for Anglo-Scandinavian and medieval York. I prefer the idea that the the cross-hall stayed standing until around the ninth century, but you can take your choice.

Animal bone
What might the animal bone represent? There is a lot of it, so someone slaughtered a lot of juvenile sheep and pigs in the vicinity. If the bones accumulated over a period of years or decades, which is quite possible even if the same radiocarbon date range applies to them all as radiocarbon dating is approximate, this might help to explain the quantity. I don’t know if it is possible to estimate the number of animals represented by the bones, which could help to assess how long they would have taken to accumulate. If, for example, there are more year-old animals than could be produced by the surrounding agricultural area in a year, that might indicate that the bones accumulated over several years (or were brought in from a distance, although transporting large numbers of juvenile animals over long distances would require considerable transport resources). The exact relationship of the sand layers to the bone might also be able to distinguish between a single event or a succession of events.

The proportion of juvenile animals is unusually high, quite different from the findings from the Roman town, the later Anglian settlement at Fishergate or the Anglo-Scandinavian town (Carver 1994). So it seems to me unlikely that the animal bone represents the routine debris of a subsistence farming community. Subsistence farming needing to maximise the amount of food from each animal would be more likely to let animals grow to full size, and probably get some useful work out of them (cattle) or some wool clips (sheep) into the bargain before eating them. The high proportion of juvenile animals suggests to me the consumption of expensive luxury food. Sucking pig in particular was a favourite Roman delicacy, if the recipes for cooking sucking pig in Apicius’ late Roman cookbook are anything to go by. Taken with the radiocarbon date, this would be consistent with the animal bones representing high-status feasting in or shortly after the Roman period, by people who had Roman tastes in food and/or wished to proclaim some sort of Roman identity and status.

As the bones are still there on the floor of the cross-hall, evidently no-one swept up the debris in antiquity and threw it out on a rubbish heap. This sits rather uneasily with the feast being held in the cross-hall itself. It seems unlikely (though not impossible) that an expensive Roman-style luxury feast would be served and consumed amongst a pile of smelly discarded bones. Possibly the cross-hall was being used as a butchery and/or kitchen, with the feast being consumed somewhere else in the vicinity, either in a different part of the headquarters complex or in another nearby Roman building. The commander’s house, baths complex and granaries would all have been nearby, and it is possible that one of these had become the focus of activity (as happened at Birdoswald, where the granaries seem to have been successively adapted for use as living halls, see earlier article on Birdoswald), with the headquarters building relegated to more workaday uses. This would be consistent with the evidence of metalworking hearths found in a room behind the cross-hall (Ottaway 2004, p. 146), which may suggest that the building was being used as a sort of industrial unit. This may seem a bit of a come-down for such a grand building, but something similar happened at the baths basilica in Wroxeter, which was used as a builders’ yard and bakery for a while in the late fifth century before demolition (see earlier article on Wroxeter). Another possibility may be that the cross-hall had gone out of use so completely that it was being used as a rubbish tip, although if this was the case one might have expected a thicker layer of debris, more like the ‘dark earth’ deposits elsewhere in York that are thought to result from dumping of domestic rubbish.

If the cross-hall was being used as a kitchen or butchery, it may still raise the question of why the bone debris was not cleared away. Possibly the cross-hall was used only intermittently, and the debris from the previous feast had ceased to be noisome by the time the next one came round and could be ignored. Perhaps the layers of trodden sand associated with the bones were scattered over the bones specifically to form a new floor surface from time to time, perhaps each time the cross-hall was used. One might imagine a ruling group progressing round various strongholds and consuming the local resources at each in turn, like a medieval king. Or if all the bone was deposited in a single event (as mentioned above, I don’t know if the quantity or layer structure is consistent with this), perhaps the debris was simply abandoned afterwards. One might imagine a scenario of one of the various usurper Emperors of the late fourth and early fifth century throwing a grand feast of Imperial style and scale for his troops and followers before marching off to try grabbing the top job, leaving the clearing up to look after itself.

If the animal bone is all late fourth-early fifth century (I am not sure how much of the bone was sampled for radiocarbon dating), and the cross-hall stayed standing until after the York ware fragments were dropped on its floor in the late eighth to ninth century, this raises the question of what happened in between. (Needless to say, if the roof collapsed in the fifth century and the pottery is a later intrusion, this question does not arise).

If there was activity in the cross-hall in the sixth-seventh-eighth centuries, it left no trace that was detectable by the Minster excavations (unless some of the bone is later than the radiocarbon date range). This may indicate that the cross-hall was disused or used only infrequently or intermittently, even if it was still standing and more or less intact. The York ware fragments are consistent with activity in the eighth or ninth century (or later), although the cross-hall may have been abandoned previously and then re-used for some reason. The Life of Wilfred says that the seventh-century stone church in York (location unknown, although it may have been in or near the principia, as discussed earlier), built in the 630s, was so neglected by the time Wilfred took it over in about 670 that the roof leaked and birds were nesting in it (Tweddle 1999, p. 126). If the church in York fell into temporary disuse for a time perhaps the principia cross-hall did too, either because its function was no longer needed at all or because it had been replaced by a different building.

Alternatively, the lack of finds could merely indicate that later users were tidier than their early fifth-century predecessors and either did not drop much debris or cleared away on a regular basis. If the animal bone was covered with trodden sand, it would have ceased being unpleasant after a while and could have consolidated with the layers of sand to form a relatively innocuous surface. If the cross-hall was later brought back into use, perhaps as a statement or as a replacement for another building nearby that had gone out of use, this surface could have formed a floor that was regularly swept, or possibly covered by a later floor that has left no detectable trace. The Minster excavations were confined to the areas where structural work was being carried out and did not cover anything like the full floor area of the cross-hall, so it is possible that isolated or slight traces could have been missed. I need hardly say that this is speculative.


For what it is worth, my interpretation of the post-Roman fate of the principia at York is broadly as follows. First, in the late fourth and early fifth centuries, a period of use as a political centre for a ruler or series of rulers who were either the direct successors of Roman authority or liked to claim they were, who had sufficient power and resources to consume Roman luxury foods like sucking pig in or near the cross-hall. Whether the cross-hall was itself a site of feasting or whether it was used as a preparation area for feasts that were consumed nearby is open to question. I would tend to favour the latter. The presence of large quantities of young animals also implies the presence of an agricultural economy and associated population, in or at no great distance from the city. This phase would be represented by the animal bone and trodden sand layers. How long it lasted is uncertain; it might have been only a few years, or it might have been decades or more if the bone accumulated over time.

Following this phase, a period of tidier or infrequent use, perhaps punctuated by periods of disuse if the focus of activity shifted between different locations, until the cross-hall roof came down some time in or after the ninth century (either by itself, or as a result of the Norse invasions of the later ninth century, or by deliberate demolition to clear the site for the Anglo-Norman cathedral).*


Apicius, De re coquinaria, translation available online
Carver MOH. Environment and commodity in Anglo-Saxon England. In: Rackham J (ed). Environment and economy in Anglo-Saxon England. Council for British Archaeology Research Report 89, 1994. ISBN 1-872414-33-8.
Ottaway P. Roman York. Tempus, 2004. ISBN 0-7524-2916-7.
Rahtz P. Review of Phillips D, Heywood B. Excavations at York Minster I. Medieval Archaeology, available online
Tweddle D, Moulden J, Logan E. Anglian York: a survey of the evidence. Council for British Archaeology/York Archaeological Trust, 1999. ISBN 1-902771-06-0.

Map links


*In Paths of Exile, I have imagined the cross-hall of the principia in use as a royal hall during the later sixth and early seventh centuries, by the Brittonic kings of Eboracum until Peredur’s death in battle in 580 and then by the Anglian king Aelle of Deira. Then a period of disuse or occasional use during the rule of the Bernician king Aethelferth, who controlled Deira and Eboracum but whose heartlands were further north.


Rick said...

It is all very odd. Presumably the cross-hall was visually impressive, since it was a stage for the display of command. Not what you would use as a dumpster for the aftermath of feasts. Unless, as you mentioned, someone had one huge blowout before marching away and leaving the mess behind.

Perhaps another possibility is that while still intact and functional, the cross-hall was worn and showing its age. It might then have been replaced for direct elite use by a better looking wooden structure, the foundations of which are either still buried under the cathedral, outside the excavated 'keyhole,' or were obliterated when the cathedral was built.

Sheerest speculation! But that is about all we have when the evidence is both limited and puzzling.

Rick said...

I hate it when I forget to check the follow-up email box!

Carla said...

Yes, it was certainly an impressive building originally, probably on a scale similar to the current Minster. One of the fallen columns was found in the excavations and re-erected in the square outside.

I quite like the mental image of Magnus Maximus or Constantine III throwing a grand luxury feast of suckling pig for their troops, then marching off to (short-lived) Imperial glory and leaving the clearing up to be somebody else's problem. A sort of large-scale version of Frodo and his friends eating their last meal at Bag End and leaving the washing-up for Lobelia :-) Magnus Maximus (383) is slap in the middle of the radiocarbon date range, and Constantine III (407) is towards its later end. This is pure speculation. But, hey, speculation is fun :-) As long as it doesn't pretend to be fact.

It's likely that the cross-hall would have become tatty over time if it was neglected, even if it remained structurally intact. It may also have been too big for use by a unit smaller than a legion (and it's unlikely that there was anything the size of a full-strength legion in York in late- or post-Roman Britain, especially after Magnus Maximus and Constantine). If it was seen as an oversized and embarrassing symbol of a decaying past, downgrading it to industrial use may have seemed a sensible thing to do with it. There may also have been an element of deliberate rejection of Roman Imperial identity by someone who wanted to proclaim 'this is a new start and we do things differently now'. Britain had a succession of short-lived usurpers before Constantine III and possibly afterwards, and it's not hard to imagine one or more of these embarking on a grandiose rebuilding project (which may or may not have outlasted the government that started it). Refurbishing one of the other nearby buildings, or building a smart new timber hall somewhere nearby, and downgrading the big old cross-hall to industrial use would be consistent with this. Large structures get converted and re-used for different purposes nowadays, e.g. country houses and old factories get redeveloped as flats. No doubt people also adadpted existing resources to changing circumstances in the past. One of the striking things about the Wroxeter baths basilica, where there is a long sequence of phases that can be related to each other, is how the site changed in use over time.

At Birdoswald and Wroxeter large timber buildings were built some time in the fifth-sixth century on the foundations of Roman buildings, so it is entirely possible that something similar happened at York. The traces left by timber buildings are very slight; at both Birdoswald and Wroxeter relatively large areas were excavated and that allowed the outlines of the timber structures to be recognised. Whether similar traces would have been recognisable in the very difficult circumstances of the Minster excavation must be doubtful. If any such building was on a slightly different site, as at Birdoswald where the chieftain's hall was on top of the granary not the principia, then it may not be under the cathedral at all.

Doug said...

Fascinating. Sand being spread over the bones so that the atmosphere could be tolerated seems likely, and would allow the process to have been prolonged. The principle you use in "Paths of Exile" could be continued - Eadwine would have used York frequently, Oswald and Oswiu more rarely. I suspect the Northumbrian kings eventually started using York more regularly despite their Bernician ancestry, but I don't know when.

Carla said...

Doug - yes, the Bernician kings might be expected to make less use of York than a king from the Deiran dynasty would. That would be consistent with the church in York falling into disrepair by 670, towards the end of Oswy's reign. Oswine (from the Deiran dynasty) ruled in Deira as a sort of joint-king or sub-king with Oswy until 651, and maybe Oswy ignored Deira during Oswine's tenure and then carried on ignoring it after Oswine's death. (Except for collecting taxes and tributes, no doubt!). York seems to have flourished as an ecclesiastical centre after it became an archbishopric; maybe the kings found it convenient to keep it at arm's length. It retained its strategic importance, though; the Vikings were keen to get control of it, and made it the centre of an independent Norse kingdom for a while.

Gabriele C. said...

Using grand Roamn buildings for pedestrian purpose isn't so unusual. The Frankish count who made his seat in Trier in the 5th century lived in one of the side wings of the Aula Palatina, while the main hall was used as storage room for the entire village (and the windows were walled in).

Carla said...

Gabriele - Yes, something like that may have happened in York. If there was a local king/ ruler still based in the city in the immediately post-Roman period, there's no reason why they couldn't have used another part of the principia, or a nearby building, or a new building as the 'royal hall' and downgraded the old cross-hall to industrial and support services use. That would be consistent with the animal bone, if the feast(s) was prepared in the cross-hall and consumed elsewhere. Without more evidence from elsewhere on the site or nearby it's impossible to say.
I came across another example recently; I'll post something about it later.
How is it known that the windows in the hall at Trier were walled up in the fifth century, by the way?

Rick said...

Nice reference to Bag End!