31 October, 2012

Post-Roman York: Fishergate

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 Anglian (‘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.  There are no definite references to York between the fourth century and the seventh century, although there are one or two snippets whose meaning is less than clear (see earlier post on Post-RomanYork: the documentary evidence for a summary of the documentary records).  Evidence from archaeology provides some clues that may help to fill in the gap. In earlier posts I have discussed the headquarters building, the Anglian cremation cemeteries at The Mount and Heworth, the inhumation cemetery at Lamel Hill, and the possible Anglian cemetery at Castle Yard.
The cemetery evidence reviewed in the previous posts indicates that people were dying and being buried in the region around York in the centuries after the end of Roman administration.  What of the living?  It is a curious feature of early medieval England that the dead are much more visible in archaeology than the living.  The early English (‘Anglo-Saxon’) funerary customs of cremation and accompanied inhumation left cemeteries of distinctive pottery cremation urns, brooches, beads and weapons for modern archaeology to find and recognise.  By contrast, early ‘Anglo-Saxon’ settlement sites tend to leave a few post-holes and/or foundation trenches and a scatter of unglamorous domestic  debris such as loom weights, spindle whorls and bits of bone.  Such slight traces are prone to damage by later ploughing or other disturbance, easily missed without skilled excavation, difficult to date and difficult to interpret, particularly in small excavations as the significance of a group of post-holes may only be recognised when they are seen in relation to one another over a wide area.  Even when post-holes and foundation trenches have survived intact and have been excavated over a large enough area to reveal  sufficient of a pattern to be recognised as a building, they preserve at best only the ground plan, which may or may not give much of an idea of the original superstructure (as discussed in an earlier post on the possibilities of timber architecture). Material culture using perishable organic materials such as bone, wood, textiles and leather often does not survive at all – bone disappears in acid soil, wood, textiles and leather decay to nothing except in exceptional circumstances such as a waterlogged site.  Readily dateable artefacts such as coins and pottery are rare or absent until around the beginning of the eighth century, making dating difficult unless sufficient organic material has survived to allow radiocarbon analysis.  So identifying early medieval life is something of a challenge.

Nevertheless, there is some evidence of people living (as well as dying) in the York area in the early medieval period.  The best-known example is the site at 46-54 Fishergate, excavated by York Archaeological Trust in the 1980s (Kemp 1996).

Evidence – 46-54 Fishergate
Fishergate is on the east bank of the River Foss, outside the southern walls of the Roman fortress

The arrow shows the approximate location of 46-54 Fishergate.  Zoom out to see the site in relation to the rest of York.

In 1985-1986, redevelopment of the site of a former glass factory at 46-54 Fishergate provided an opportunity for archaeological investigation, carried out by York Archaeological Trust (Kemp 1996 p5).  Archaeological deposits had survived on about half the site, in the south-eastern corner, underneath a Victorian factory building with shallow foundations (on the rest of the site, modern factory foundations had removed the archaeological deposits).  Excavation revealed traces of ditches, pits and possible structures dating to around the seventh to ninth centuries.
Possible boundary markers
A curving ditch ran from north to south across the site, narrow and shallow (approximately 0.4 m wide and 0.45 m deep) at the north end and becoming broader and deeper towards the south (approximately 2.1 m wide and 0.7 m deep).  The bottom of the ditch contained a layer of silt deposited in standing or slow-moving water, and a notable absence of remains of insect species that normally live alongside human habitation, although conditions would have been expected to preserve them.  This is interpreted as indicating that the ditch was dug, perhaps to mark a boundary, and then the site left uninhabited for a year or more, long enough for a diverse community of invertebrates to become established naturally.

On top of this silty layer, the ditch had filled up with loamy soil interspersed with charcoal, containing animal bones, debris from antler-working, slag, fragments of glass and a 24-cm length of gold wire.  The ditch fill also contained traces of whipworms (human intestinal parasites), indicating the presence of latrine waste.  So the ditch appears to have become a sort of giant linear rubbish pit for domestic and craft-working waste. Part of a comb made from antler, dated to the seventh or eighth century, and a coin dated to around 700-735 were also found in the ditch fill; the coin was much worn, suggesting that it had been in circulation for a long time before it ended up in the ditch (Kemp 1996, p 18-23).
A line of six large pits ran westward from the south end of the ditch.  Three of the pits contained animal bones, antler-working debris and human parasite eggs, and one also contained a bone sword guard dated to the first half of the 8th century. These may be a line of rubbish pits marking a boundary (Kemp 1996 p. 23-24).

Possible structures
To the west of the curving ditch and north of the line of pits, groups of post-bases (shallow post-holes, possibly the foundations for padstones) and foundation slots indicated traces of several possible structures.

Structure 1 was rectangular, 5.5m wide and between 14m and 19m long (the exact length is uncertain because one end of the structure was underneath an unexcavated baulk), oriented with the long axis roughly north-south.  A shallow slot running part way across the structure east-west may be the foundation slot for a beam supporting an internal partition that would have separated off a smaller chamber at the northern end.  No traces of any timbers remained (Kemp 1996 p 27-31).

A second group of post-bases and slots was interpreted as another rectangular structure 5.5m wide and at least 13m long  (only 13m was excavated, so the actual length is unknown), oriented with its long axis east-west (Structure 2). Like Structure 1, there was a crossways slot part-way across the structure, consistent with an internal partition separating off a smaller chamber at one end (in this case the east end).  Four coins dated to approximately 700-735 were found in the fill of the slot, indicating that the timber was removed and the slot filled some time after this period.
A third possible structure was represented by a single line of post-bases with a blank area adjacent; if the post-bases represent one wall and the blank area the building interior , the structure would also have been 5.5 m wide and at least 11m long, with the long axis east-west.  The area where the western wall would have been had been extensively disturbed, which would have destroyed any traces.  However, the area where the east wall would have been had not been so disturbed and no traces of a wall were seen, so this may not be a structure (Kemp 1996 p 34).

Other groups of slots and post-holes may represent the remains of more structures, but the traces were too fragmentary to interpret (Kemp 1996 p 36-37).
All of these possible structures, the curving ditch and the rubbish pits were sealed underneath an extensive charcoal-rich layer, indicating that they all belong to the same period.
Dating evidence is limited.  Some artefacts could date to the mid to late seventh century, and the earliest coins found on the site date to the early eighth century (approximately 700-735). Coins of an earlier type from the period between 670 and 700 were absent. The findings are consistent with a foundation date for the Fishergate settlement in the late seventh or very early eighth century (Kemp 1996 p 66). Pottery types characteristic of the late ninth and tenth century, which are abundant on the Coppergate site elsewhere in York, were almost absent from the Fishergate site, suggesting that the site at Fishergate was not occupied during this period (Kemp 1996 p 83). The settlement may have shifted to Coppergate in the mid ninth century, abandoning the Fishergate site.  It has also been suggested that the site at Fishergate replaced an earlier (mid sixth to mid seventh century) settlement further west on the same gravel moraine at Heslington Hill (Spall and Toop 2008).

Finds from the Fishergate site included ironsmithing slag, debris from copper- and lead-working, woodworking and leatherworking tools, and bones from beaver and pine marten consistent with preparation for furs (Kemp 1996 p 71).Food remains included animal bone (mainly cattle), fish bones, barley, wheat, rye, apples, sloes, hazelnuts, eggs and possibly peas (Kemp 1996 p 71). Sherds of Ipswich-ware pottery indicate contact with East Anglia, and fragments of stone were identified from Cumbria, Wensleydale, Swaledale and the Yorkshire Wolds (Kemp 1996 p 72-73). Lava querns for grinding grain into flour and pottery sherds imported from Germany and northern France/the Low Countries indicate contact across the North Sea (Kemp 1996 p 73).
The Fishergate excavation represents part of a larger settlement, but the size of the settlement is unknown (Kemp 1996 p 75).  It was hypothesised that the settlement could extend along the east bank of the River Foss, perhaps covering 10-25 hectares, maybe as much as 65 hectares, which would be comparable with known eighth-century manufacturing and trading sites at Ipswich, London and Hamwic (near Southampton) (Kemp 1996 p 75-77; Tweddle et al 1999 p 193).  However, recent excavations at Fishergate House and Blue Bridge Lane, just south of the 46-54 Fishergate site, suggest much lower density of occupation in this area, and the Fishergate settlement is now suggested to be much smaller than originally thought, perhaps 4 hectares or so (Spall and Toop 2005).

The debris in the ditch and rubbish pit fills indicates that the 46-54 Fishergate site was concerned with various crafts, including textiles, fur production and the working of leather, wood, bone, antler and various metals including iron, lead and copper.  The length of gold wire (someone must have cursed when they realised they had lost that!) may indicate that precious metals were also worked on the site, perhaps making jewellery.  The pottery and stone from elsewhere in Britain and overseas suggests that Fishergate had regional and international contacts.  The most obvious interpretation is of Fishergate as a centre of manufacturing and trade in the eighth century, a smaller version of the known manufacturing and trading sites (wics) known at Ipswich, London and Hamwic near Southampton (Kemp 1996 p 64). 
The curving boundary ditch, which was dug and the site then apparently left uninhabited for a year or more, may indicate that the Fishergate settlement was deliberately planned and marked out as a site for development before the actual settlement was built.  This would be consistent with some sort of ‘official’ planned development, perhaps by a landowner who marked out the site and then permitted/ persuaded people to move into it and establish craft workshops and dwellings.  The animal bones were less diverse than those typically found on self-sufficient rural village sites such as West Stow, which may suggest that the food supply at Fishergate was restricted, perhaps provided or controlled by a central authority (Kemp 1996 p 74). This is consistent with the possibility that Fishergate was a specialist manufacturing/trading centre controlled by a lord, who provided its inhabitants with access to a restricted range of food, perhaps obtained as food rents from other sites (Kemp 1996 p 74).

York was called Eoforwic or Eoforwiccastre in Old English (Eoforwic was later turned into Jorvik by Norse speakers, and then further shortened over time to eventually become the modern name of York).  The –wic element in place names is commonly associated with sites engaged in trading and/or specialist production, which would fit the evidence from the Fishergate excavation very well.
An account of the Life of St Liudger refers to a colony of Frisian traders based in York in the early eighth century (Kemp 1996 p 65). This would fit well with the suggested foundation date for Fishergate, and the imported material at Fishergate from across the North Sea in Germany, northern France and the Low Countries.

If Fishergate was only about 4 hectares in size as recently suggested, this is much smaller than known –wic sites such as Ipswich or Hamwic (about one-tenth the size).  This may suggest that it was a different type of site – perhaps a foreign enclave, established especially for the early eighth-century Frisian traders mentioned in the Life of St Liudger?  Or perhaps it was intended to be bigger and for some reason did not develop to the same size as Ipswich or Hamwic.  Or the Fishergate site could be one of several sites scattered in and around Roman York.  Various traces of possible structures and pits have been identified in and around the Roman city (Tweddle et al 1999 p 191-199). The traces are generally insubstantial and the date range wide (often no closer than some time between the Roman and Anglo-Scandinavian periods), so it is impossible to say whether these represent other settlements contemporary with the Fishergate site or traces of habitation from different periods, perhaps shifting from place to place.  If there were several small contemporary settlements, could they have added up between them to something resembling a London-sized –wic, but that for some reason was dispersed across multiple sites?

The Fishergate site provides clear evidence for domestic occupation, craft working and regional and international trade on a substantial scale from approximately the late seventh century to approximately the mid ninth century.  By this time York has reappeared in the historical records as a royal and ecclesiastical centre (see earlier post on the documentary evidence) and as the location for a group of Frisian traders mentioned in the Life of St Liudger.  It would make sense for an important royal and ecclesiastical centre to have a trading and manufacturing population nearby to do the work and provide necessary goods.  The levying of tolls and/or taxation on trade and manufacture may have also made a substantial contribution to the economy, especially if Fishergate was part of a larger settlement or a component of a network of related sites dispersed around the environs of the Roman city of York.
Since York evidently had royal and ecclesiastical significance by 627 when King Edwin (Eadwine) of Northumbria chose the city for his baptism and built a church there (Bede Book II Ch. 14), it seems likely that there was also a working population in the area. If the suggested foundation date for Fishergate is correct, such a population was not based there in the early seventh century. If Fishergate was indeed a newly established planned settlement, it may have replaced one or more earlier sites in the vicinity that carried out some of the same sorts of activity. Developing a new site to do something that already happens locally, perhaps on a larger or more organised scale, is an easier proposition than starting from scratch.  The suggested mid-sixth to mid-seventh century settlement at Heslington Hill may represent such an earlier site.  Some of the sites represented by the other fragmentary remains in York that have not been closely dated may also belong to the seventh century, and other sites may have existed that have not (yet) been identified, but (obviously, unless further evidence turns up) this cannot be substantiated*. 

Bede, Ecclesiastical history of the English people.  Translated by Leo Sherley-Price.  Penguin Classics, 1968, ISBN 0-14-044565-X.
Kemp RL. Anglian settlement at 46-54 Fishergate. Council for British Archaeology/York Archaeological Trust, 1996. ISBN 1-872414-70-2.
Spall C, Toop N. Blue Bridge Lane and Fishergate House. Excavation Period 3: Anglian settlement. 2005. Available online
Spall C, Toop N. Before Eoforwic: new light on York in the 6th and 7th centuries. Medieval Archaeology 2008;52. Abstract 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.

*Paths of Exile is set in 605-606, much earlier than the date suggested for Fishergate.  My speculation is that ad hoc seasonal trading was already established near the Roman fortress at York long before the Fishergate settlement was founded.  The river and anything that remained of the Roman harbour infrastructure would have offered a convenient site for traders from across the North Sea to arrive in the summer with goods to buy and sell, especially if (as I also speculate) the old Roman fortress was still a royal power centre and thus a likely market for luxury imports.  My speculation  is that the York area also formed a convenient site for local seasonal trading fairs where agricultural and craft produce could be exchanged, the sort of place where farmers and part-time craft-workers might trade a sack of grain for a new cauldron, or a couple of piglets for a dagger, or a length of woven cloth for a colourful new brooch or string of beads, perhaps under the protection of a local lord who could provide some sort of security so that one could be reasonably confident of not being mugged at market (although what happened on the way home may have been another matter).  I have placed Eoforwic in 605 at the site of a (fictional) nobleman’s hall on the opposite side of the River Foss to Fishergate, between the Rivers Foss and Ouse and south of the Roman fortress.


Rick said...

I can easily imagine that, amid a general upward trend in prosperity, a previous informal trading area would be expanded and rebuilt as a 'steelyard'.

The interval of a year or more between digging the ditch and occupation of the area invites any amount of speculation. Perhaps the expansion came with added regulation. So the Frisians initially decamp - then decide that the market is too profitable to ignore, and move back.

Or the expansion could follow a successful war, with the Frisians staying away until the outcome was clear. Or whatever along those general lines.

Carla said...

Yes, that apparent year's delay does lend itself to possibilities, doesn't it? A variant of your first possibility is that the (hypothetical) informal market was a local affair, and the Frisians did not so much decamp and then return as arrive de novo when a formal site was established. If Frisians were seen as 'foreign' (which may depend on the degree of connection across the North Sea), a site where they could settle with some sort of formal authorisation from a local ruler who could take their side in any disputes could be an attractive proposition. One might speculate that an enterprising nobleman laid out the site, showed possible traders round and gave them the sales spiel, and then they came back next trading season and moved in, hence the year's gap.

Although since the full extent of the site isn't known, a prosaic explanation might be that a large site was marked out by the ditch and then filled up gradually, starting from one end, and the excavated area happened to be in one of the areas that filled up later. Modern industrial estates and retail parks sometimes develop like this, with the units being built and let sequentially.

Rick said...

Yes, there is the prosaic explanation, too! Something that strikes us as notably significant might not have been significant at all.

Carla said...

Quite so. When you have only a fragment of the whole picture there's a tendency to assume that it must be the most significant piece, but of course it may not have been.
I suppose another variant of the prosaic explanation is that the ditch was originally dug for some other purpose, e.g. as a livestock boundary marking out a pasture, fulfilled that function for a while, and then people started building houses in the pasture.

tenthmedieval said...

If Fishergate was indeed a newly established planned settlement, it may have replaced one or more earlier sites in the vicinity that carried out some of the same sorts of activity.

It seems to me that you must be right that there was production and exchange going on in York before this place was opened up. The church would have needed such things and would have had to start a market of some kind of there wasn't one already. So, if a new site to do that is opened up, that seems to suggest several possibilities:

(i) firstly, as you say above, this is to replace the old sites, either because they were becoming too small (though in that case the delay in filling the new one up is hard to explain) or otherwise inaccessible. I wonder about repairs to defences or similar here.

(ii) the old sites kept going, and this was opened either as a supplement, or as competition. Even the former might have worked out as having the effect of the latter, like a new mall taking trade away from the high street even though all its shops were different, now.

The thing that makes these two options different is that in the latter situation, whoever's opened the new site is presumably getting something new out of it, a kind of control or profit that wasn't available, or sufficiently available, from the first one. (That might indeed be setting up a small colony of Frisians whose range of goods would be different, more expensive, but who would appreciate the market access enough to pay a bit over.) If it's the former, though, then there's no reason to suppose anyone was getting anything new from the new site that they weren't from the old one. I suppose both are possible together: Edwin or Oswy or whoever repair the city walls and clear a load of buildings and stalls away from the base of them, having previously opened a new market site to house the traders but on newer and less advantageous terms... But it would still have considerable impact on the way we imagine this all happening to know if there was production or import of luxuries in other areas of the city beforehand and when they stopped functioning...

All very interesting, thankyou for summarising a tricky set of sites so clearly!

Jonathan Jarrett

Carla said...

Tenthmedieval - repairs to defences is a possibility. If the Anglian Tower dates from this period it might be indicative of a larger programme of defensive works, which might have caused people to move from somewhere else to Fishergate. It's also possible that Fishergate was an addition to or replacement of an existing site. I think both are possible together, or perhaps in close succession.

It would indeed be very helpful to know whether there was anything similar going on elsewhere in the city. Maybe archaeology will turn something sooner or later!