The missing link in York's Anglo-Saxon archaeology has at last been found by the York Archaeological Trust in excavations just completed on the site of the Redfearn National Glass Factory at Fishergate. There has always been a great contradiction between the historical and the archaeological evidence for Eoforwic, the Anglian precursor to Jorvik. Historical evidence – for instance, Alcuin's great poem on the church and saints of York – suggests a booming trading settlement, full of foreign traders and merchants. Yet, despite 15 years of city centre excavation, no substantial traces of this period had ever been found. This led some to question the historical evidence, notably Richard Hodges who stated in 1984 that 'At some stage we have to accept the power of this kind of negative [archaeological] evidence'.

Now Alcuin and other 8th-century sources are vindicated. Two kilometres (1.2 miles) downstream from the area of the Roman, medieval and modern centre of York, near the confluence of the Rivers Foss and Ouse, on a terrace of brickearth unencumbered by Roman remains, we have at last found the missing archaeological evidence, locked beneath a sprawling modern glass factory and a medieval Gilbertine priory.

Finding the priory ... and more

Work began on the site in February 1985, when the factory was abandoned but still standing. Trial transects were cut through the factory floors in order to locate the Gilbertine house, known to be somewhere within the five-acre site. Within just 11 working days, well preserved remains of the priory church were located, and within two months, an outline plan of the whole priory had been recovered. Trial work showed the remains to be very well preserved, but, more exciting, beneath the priory gardens numerous large rubbish pits of the 8th century were found, along with structural traces which, though difficult to interpret in the narrow trial transects, seemed to suggest extensive settlement.

These discoveries set the wheels of the Trust's grant- and resource-gathering machinery into action, and in due course HBMC and York City Council agreed to provide matched amounts towards further work.

The site was threatened with housing redevelopment, and once the new owners had been established, access for open area excavation was negotiated under the Areas of Archaeological Importance provision of the Ancient Monuments and Archaeological Areas Act. The developers, Costain Homes (North Eastern) Ltd, generously designed their demolition and building schedule around the excavation and also awarded the Trust the contract to remove the estimated 1,000 human burials from the site under the Disused Burial Ground Amendment Act.

Timber halls

Open area excavations began in January 1986 and continued until mid-September of the same year. The area stripped archaeologically corresponded to that of the entire Gilbertine priory and was approaching 2,500 sq m (26,910 sq. ft).

The Gilbertine deposits were excavated rapidly, so that the earlier deposits could be given greater priority. These Anglian deposits revealed some seven or eight timber halls located between two north-south linear gravel spreads interpreted as roads. The buildings – some difficult to recognise – are, on the whole, post-built. Later there was a change to sill beam construction for buildings of similar proportions but on different alignments. There was clear evidence for property boundaries, and for changes to these boundaries.

There were a large number of rubbish pits outside these structures. Although Fishergate was not waterlogged, like the famous Coppergate site, nevertheless a large environmental sampling campaign was carried out on site, and something approaching 60 tonnes of soil was processed using a variety of sieving techniques developed by the Environmental Archaeology Unit at the University of York.

This sieving campaign had a number of advantages. As the entire contents of all Anglian pits were processed, we could excavate faster as spoil was not hand sorted on site. The range of finds, particularly in the smaller artefacts, has been greatly increased and a bone assemblage without the biases brought by hand selection may yet revolutionise ideas on diet, economy and husbandry for this period.

Foreign trade

Finds from provisional sorting of residues already point to the economic base for the settlement. Thus foreign trade is revealed by fragments of Niedermendig lava, Porcupine series scettas, Frisian combs and continental pottery (Rhenish and possibly Frankish). Weights, balances and coins attest to the monetary basis of the economy. Bone comb billets reveal manufacture, crucibles and slags reveal industry, and spindle whorls and loomweights reveal crafts.

Local handmade, gritty pottery – itself dated by coins – provided the dating evidence for most features. Although hardly ubiquitous, this pottery was nevertheless present in many features, and represents by far the largest series yet recovered in York or its region. Among the exceptional finds was a bone handle decorated with 8th-century animal ornament. There were 32 coins, mainly sceattas and stycas, dating between 700-740 and 830-50.

We can now begin to see the outline of the settlement at York in the pre-Viking period. The main 8th-century discoveries hitherto had been those found in 1972 at Paragon Street, on the site of York's former cattle market, only 450 m (500 yd) to the east of the Redfearn site and on the same spur of land. In December 1985, a further 450 m beyond the cattle market site, the Trust uncovered more 8th-century material buried beneath the medieval defences near Walmgate Bar.

Extrapolating from this, it looks as if the Anglian settlement might have covered as much as 25 hectares (62 acres) and, far from having been a huddle of merchants on the side of the river, seems to have extended over an area comparable to Quentovic (in France), Ipswich, Londonwic or perhaps even Hamwic.

Preliminary study of the finds and pottery from the site suggests that occupation came to an end in the later 9th century. It is precisely at this period, of course that occupation began in earnest at Coppergate, which may therefore mark a return to the old site of York at the start of the Viking Age. At last Eoforwic seems to be taking its rightful place in Dark Age archaeology as the historical evidence has long suggested it should.

Timber and stone churches

In addition to the discovery of Eoforwic, the remains of a small timber church were found in the southern part of the site. It overlay Anglian features and is provisionally dated to the 11th century. It consisted of a clay floor covering an area some 4 m by 6 m (13 by 20 ft), limited to the west and south by shallow slots that meet at a large post hole. About 40 burials were associated with the structure, one of which contained a skull. There was also an intriguing ‘translated' burial beyond the eastern end of the church and aligned on it. This was presumably the parish church of St Andrew mentioned in Domesday, and as the later Gilbertine Priory of 1202 was also called St Andrew's, a direct link seems likely.

The timber church was superseded by a much larger stone church placed immediately to the north, and dated to the early Norman period by the architecture as well as the finds. This stone rectangle may have been the parish church of St Andrew given to Newburgh Priory in 1143 and subsequently redonated to the Gilbertine order as a basis for their priory in 1202. By examining the different foundation cores, we could see how a chancel, cloisters and other priory buildings were added to this earlier building, and a detailed picture of development from the small timber church into the complex institution of a Priory could be built up. This is the first time that any of York's eight medieval monasteries has been fully examined in this way.

The last vestige of activity on the site was a great limekiln build in the middle of the cloister garth by a demolition team in the 16th century. The site then lay beneath an orchard until the glass industry arrived in the 1800s.

Back to York and Archaeology