Excavating the Oxford Jewry
Excavations in the historic heart of Oxford have shed light on the city’s origins and development – including uncovering some of its earliest-known Anglo-Saxon structures, remarkable evidence for the medieval city’s Jewish inhabitants, and aspects of city life away from the colleges, as Edward Biddulph explains.
It all began with a road. The origins of Oxford can be traced back as far as the late 7th century AD, when a river crossing presumed to have carried a north–south routeway (later known as St Aldates) was built across the channels of the River Thames. St Frideswide’s minster is believed to have been founded close by in the early 8th century, and by the early 10th century Oxnaforda had flourished into a burh, one of several defended, planned towns established in England in the late Anglo-Saxon period. This settlement was laid out around two principal roads, one (Cornmarket/ St Aldates) orientated north–south, the other (Queen Street/High Street) east–west. Their axis, a crossroads later known as Carfax, would become the heart of the burgeoning medieval urban centre – and in 2016 Oxford Archaeology (OA) had the opportunity to bring this area’s past to light once more.
Planned redevelopment on the corner of Queen Street and St Aldates, on the south-west side of Carfax, offered a rare chance to explore this part of Oxford’s city centre. It was an exciting prospect, as this was an area with a rich and interesting history. During the 12th and 13th centuries, the north end of St Aldates was home to Oxford’s Jewish quarter, following William the Conqueror’s invitation to Jews from France to settle in England and establish a network of credit and trading posts. This hospitality was not to last, though, and the Jewry came to an end in 1290, if not before, when the Jews were expelled from England by Edward I. By the 14th century, all Jewish properties were in Christian ownership, and the area’s use had transformed dramatically: a plethora of medieval drinking establishments sprang up along Queen Street and St Aldates, and some of these establishments survived in one form or another until the modern day.
How far would our finds reflect this area’s diverse past? Promisingly, an evaluation by OA in 2015 had identified extensive evidence of undisturbed late Anglo-Saxon to early medieval layers preserved underneath existing basement levels on the site. From historical evidence, we also knew that the area we were to excavate encompassed part of two properties known to have been in Jewish ownership: Jacob’s Hall, reputedly one of the most substantial private houses in Oxford, and a property owned by an ‘Elias’ or ‘Elekin’. Would the investigation yield any archaeological evidence for these properties or for the Jewry more generally? This project also represented a chance to reach levels contemporary with Oxford’s middle or late Anglo-Saxon settlement, and to identify the remains of two drinking establishments known to have existed on the site: Battes Inn and the Red Lion. The search was on.
As our work began, we quickly established that – as is often the case on urban sites – the construction of deep basements and the foundations of more modern buildings had taken large bites out of the earliest archaeological horizons. In two areas, however, the remains of middle or late Anglo-Saxon structures had survived. One of these (labelled 1.1 on the plan shown on p.20) comprised a series of floor layers and an associated post-hole, while another (labelled 1.2) was represented by a sunken floor. A stake-hole found against the side of this surface and filled with organic matter is likely to be a remnant of a wattle lining. Radiocarbon analysis suggested that these buildings were broadly contemporary: a date obtained from charred grain from Structure 1.1 spanned the late 8th to late 10th century AD, while Structure 1.2 ranged from the late 7th to late 9th century cal AD. Further Anglo-Saxon traces came from a sequence of pits uncovered in another area of the site, which yielded pottery indicating that they had continued to be used until the late 10th or early 11th century.
While archaeological remains from this phase of the site’s life may seem limited, they offer exciting insights into a period that is otherwise not well represented in the material record, providing some of the earliest known evidence for settlement pre-dating Oxford’s late Saxon burh. We can reconstruct the buildings as simple wooden or wattle-and-daub houses, with the sunken floor of Structure 1.2 forming a cellar. Pits like those excavated would have been used for the disposal of household waste such as pottery and animal bone, and were dug in backyards away from the streets onto which the buildings probably fronted.
We can also see how the site developed after the burh was established: Structure 1.2 was replaced by another dwelling (Structure 2.1), also with a sunken floor or cellar, and post-holes along the edge of the house suggest that it had a timber revetment around its sides. Pottery collected from its gravel floor suggests that the building was in use during the 10th or early 11th century. Yet while buildings came and went, environmental evidence from both phases of later Anglo-Saxon settlement suggests that life within the burh was not much different from what had gone before. Certainly the diet of its inhabitants remained much the same: charred plant remains testify to free-threshing wheat and hulled barley, which would have been made into pottages, porridges, and bread.
Smaller quantities of oats and rye were also recorded, the former providing fodder for livestock, such as sheep, that would have been kept within the settlement. Beef, mutton, and pork were also consumed, with butchery marks found on cattle vertebrae suggesting that the emphasis was on good-sized portions, rather than the quality or attractiveness of the cut. Chickens were kept for eggs, and other dietary refuse hints that oysters and freshwater fish were also occasionally eaten. As for what this well-fed community was doing, we have found clues to spinning and bone-working – the latter utilising deer bones as well as waste from domesticated animals – being among the craft or domestic activities carried out within the properties represented here or nearby.
For all this continuity, though, the pottery record does preserve some indication of changes afoot. The locally produced late Saxon Oxford shelly ware – which comprised all the pottery from the site’s pre-burh phase – was completely replaced after the burh came into being: instead we find St Neots shelly ware from eastern England. Exotica in the form of a jar or pitcher in Beauvais ware from France, possibly used for transporting wine, was also identified. What does this abrupt change mean? For now, the answer remains unclear – late Saxon Oxford shelly ware is found in burh-related deposits elsewhere in the city – but it is possible that the area of the burh around this part of Queen Street and St Aldates was settled by a different social group with high-status connections or wider trade links. That said, evidence of burning on the pottery indicates that meals in both phases were prepared in the same way: in ceramic jars over a fired-clay hearth or in an oven.
IDENTIFYING THE JEWISH QUARTER
What of the medieval Jewish settlement known to have existed in this area? We were able to trace the site’s late 11th and 12th century phase, during which it appears to have become an open area or yard scattered with pits that had received household waste, presumably from nearby properties. There are other indications of occupation during this period: a stone-built structure (labelled 3.1) has been interpreted as a latrine that would have stood outside the building it served, while post-holes to the south-east relate to a fence or other timber structure. Another construction (3.2) formed part of a wall that divided two buildings – one extending to the north, the other to the south – and rectangular holes and stone projections seen on its north face mark the position of a wooden floor. This phase corresponds with the earlier part of the period ascribed to Oxford’s Jewish quarter, which is conventionally dated to the 12th to late 13th century, but what do these ephemeral traces mean?
Judging by mapping of medieval properties in Jewish ownership, produced by Pam Manix in 2004, the latrine appears to lie within the boundary of Jacob’s Hall, the substantial house mentioned above. Significantly, it was also this structure that yielded archaeological evidence consistent with a Jewish population. How do you identify the archaeological ‘signature’ of such inhabitants? One useful method is comparison of the animal-bone assemblage with Jewish dietary law or kashrut. A Jewish assemblage would not be expected to contain any wild animals, pig bones, horse bones, bones showing disease, or bones from the hindquarters of cattle. Shellfish and fish without fins or scales (for instance, eels) would also be absent, as their consumption is similarly proscribed.
What is fascinating about the assemblage from the latrine is that pig specimens were indeed completely absent, while domestic fowl (mainly goose) dominated the group, being even more common than cattle bones. Fish bones from the feature included herring, but no eel, and of the 89 fragments of marine shell recovered from late 11th-/12th-century deposits on the site, just one came from the latrine. This body of evidence stands in sharp contrast to that of the preceding late Saxon phase, in which pig bones were plentiful and bird bones barely present at all, and to the 13th- and 14th-century contexts post-dating known Jewish activity on the site, in which pig bones are seen once more.
Between these two assemblages, we have a distinctive window in which the site’s inhabitants do seem to have been following Jewish dietary rules. Material from the late 11th-/12th-century pitting is strikingly different, however: these features contained pig bones, marine shell, and eel bones, and it is probable that these deposits pre-date the latrine and relate to a period in the earlier part of the phase before the property was transferred to Jewish ownership.
FOOD FOR THOUGHT
The results of this faunal analysis are of profound importance. Until now, no Jewish signature had been identified in British zooarchaeology, and just two animal-bone assemblages from medieval Jewries on the Continent have been published. Could the findings be corroborated by other means? One possibility was through organic residue analysis (ORA), which is used to identify types of animal fats or plant oils and waxes that have been processed or stored in pottery vessels. Four ceramic jars from late 11th-/12th-century deposits and six jars from 13th-/14th-century deposits were examined by Julie Dunne and Richard Evershed of the University of Bristol, and, for purposes of comparison, ORA was also carried out on a contemporaneous assemblage (ten samples) from Queen’s College, Oxford. This latter material derived from domestic and industrial deposits some distance from the Jewry and, as it also pre-dated the college, should not correlate to restricted or high-status diets. Furthermore, we also compared eight pottery samples from the middle/ late Saxon phases of the St Aldates site, and two further samples from the medieval phases that had previously been analysed by Lucy Cramp and Helen Whelton.
The results were clear: non-ruminant (that is, porcine) fats were absent from all the samples of medieval date from St Aldates. All but one of those samples, however, had traces of ruminant (cattle and sheep) fats. In stark contrast, samples from the middle/late Saxon phases were mixed, with both ruminant and non-ruminant fats represented, and, while seven of the Queen’s College samples showed ruminant fats only, the remainder fell within the ruminant/ non-ruminant part of the spectrum, also demonstrating mixed use. In summary, the medieval pots from St Aldates show that they were not used to process or cook pork, suggesting that – in keeping with the animal-bone assemblage from the latrine – pork was absent from the pottery users’ diet. Pottery was used for the processing of pork in the middle/late Saxon phases of St Aldates, though, and at the early medieval location of Queen’s College at least some of the time.
One other interesting observation is that dairy fats, with no other animal fats evident, were only occasionally recorded in jars from both the mid-/ late Saxon and medieval phases of St Aldates. This suggests that the production of cheese or butter was rarely undertaken at a domestic level in this part of Oxford whatever the period – possibly such products were acquired ready-made from sellers – but, when it was undertaken, it was done so in jars specially dedicated to the task. This aspect is also tantalisingly consistent with Jewish dietary laws, as, according to the principles of kashrut, it was important to keep meat and dairy separate (although this evidence is admittedly slight, and a similar separation is apparent in pre-Jewish phases too).
Taken together, our finds and environmental evidence reveal interesting aspects of daily life in and around the site during the late 11th and 12th centuries. Apart from abstinence from pork, shellfish, and eel during the Jewish phase of the site’s use, diet consisted of bread and grain-based pottages, the latter including beef or mutton at least occasionally. Goose was consumed by Jewish residents. Fish was presumably purchased from stalls at the north end of St Aldates (formerly known as Fish Street) and included freshwater species such as pike, burbot, and (before the Jewish phase) eels, as well as marine species like herring, haddock, and plaice, which arrived in barrels preserved in salt or brine. Ceramic building material suggests that buildings of this period had tiled roofs with chimneys and (at least partially) tiled floors, something consistent with the stone walls of the latrine, which denote buildings of high status. The pottery was generally of a domestic character but also included ceramic crucibles that point to small-scale metalworking – further craft or industrial activity is also suggested by a whetstone, while a possible hoe among the metal objects hints at horticultural work. Finally, a tangible possible link to one of the site’s Jewish residents comes from a leather shoe recovered from a pit that was broadly contemporary with Structure 3.1.
It should be noted that none of the ORA pottery samples came from the latrine, but instead were taken from vessels recovered from the upper fills of the pitting which, in contrast to the lower fills, contained no pig bones, eel bones, or marine shell, and may belong to episodes of levelling related to the construction of Jacob’s Hall. Given the stratigraphy and the dating evidence from the latrine and pitting, a case can be made for dating this Jewish phase from the later part of the 12th to the early 13th century, certainly after the period of pit-digging, and perhaps spanning only a single generation of residents. This is some decades before the expulsion of the Jews from England in 1290, but the brevity of activity suggests that the property might have been taken relatively abruptly out of Jewish ownership, possibly having been confiscated, rather than being allowed to pass to heirs.
During the 14th and 15th centuries, as historical records attest, all the properties on our site reverted to Christian ownership – and this change also seems to be reflected in the archaeology, indicated not least by the reappearance of oyster shell, pig bones, and eel bones within refuse deposits associated with this phase. It was at this time that a plethora of drinking establishments were set up in the area – excavated walls belonging to structures from this phase correspond to property boundaries (mapped by H E Salter) belonging to Battes Inn and the Red Lion, while part of a cellared structure that we recorded may relate to the Swindlestock tavern, whose cellar extended below the south-west corner of Carfax.
What can we deduce about the diet, activities, and status of the people who lived here at the time? Once again, refuse pits are very much our friends: we can tell that grain continued to be used mainly for bread and pottages, while small quantities of barley might have been used for malting and brewing ale, though no definitive evidence for such activities was found. Beef, mutton, and to a lesser extent pork were consumed, as well as the occasional hare and freshwater and cured marine fish. A relatively broad range of fruits, among them sloe, cherry, damson-type fruit, grape, fig, wild strawberries, and pears or apples, were also on the menu, recovered from a layer of domestic and faecal waste, with coriander and fennel also recorded. Meals were cooked and served in a standard range of domestic pottery, but a fragment of early Malagan or Valencian lustreware suggests that rare and exotic household items were also available to some of the site’s inhabitants.
As for what their homes looked like, buildings here or nearby had tiled roofs but their floors were generally plainer – timber boards or trampled-earth floors were more usual than tiles – and part of the properties also provided accommodation for animals, as indicated by the recovery of stabling waste. At least some of the windows were glazed, and additional light was provided by oil lamps (one of which was recovered). The overall impression is of high-status households, perhaps those of merchants with connections to coastal trade.
The creation of cellars and further division of properties during the late medieval period (15th to mid-16th century) hints at the construction of several new structures, while the composition of the pottery assemblage from this time is noticeably different to that of the previous phase – it comprises almost exclusively drinking-related vessels (cups, mugs, and jugs), rather than the more typically domestic jars and bowls that preceded them. The mercantile households, it seems, had been replaced by the commercial activity of an inn, probably Battes Inn, within whose boundary the structures lie.
As we move into the mid-16th and mid-17th centuries, parts of the site were becoming more open as one area was levelled, while new rubbish pits speckled other areas. It is uncertain whether these changes relate to activity associated with the inns that were by now thriving in the vicinity, as the recovered pottery is of a decidedly domestic character. There was some activity that could be linked more readily to drinking establishments, though. A stone-lined latrine, dating to the mid-17th to early 18th century, had been inserted against what has been mapped as the southern wall of Battes Inn, its function as a latrine revealed by the presence among its contents of faecal matter, traces of sedge collected from the riverbank or the floor of stables to dampen aromas, and a chamber pot.
While the latrine’s use by the patrons of the inn cannot be confirmed, the latrine was certainly filled with some evidence of drinking-related activity. Glass recovered from the feature included fragments of ‘globe and shaft’ wine bottles, which carried the seals of two other drinking establishments: the Mermaid (formerly the Swindlestock), which was demolished c.1708, and the Crown, the latter being located at no.3 Cornmarket. A large collection of clay tobacco pipes was also recovered from the feature; 20 or so complete or near-complete examples could be reconstructed from the fragments. Among the environmental evidence that we recovered were the remains of fig, blackberry, raspberry, grape, strawberry, elder, and apple. Such fruit was also recorded at Corpus Christi College and Merton College, Oxford, and points to a high-status diet. The animal-bone assemblage, comprising domestic mammals and fowl, was more typical of an urban pattern of consumption.
Given their quantity and the nature of their deposition, the pipes are likely to relate to a level of use beyond that of a normal household. The glass bottle seals point to the supply of wine from a variety of local drinking establishments, while the fruits represent the consumption of relatively exotic foodstuffs. Taken together, the contents of the latrine are likely to represent the dumping of waste deriving from a source of some affluence, perhaps accommodation or an establishment favoured by wealthy individuals from a mercantile or collegiate background. The results of the excavation at St Aldates and Queen Street have been astonishing not only in terms of their having revealed rare archaeological evidence of a medieval Jewry in Britain, but also for demonstrating the enormous value of carefully focused analysis that combines traditional finds and stratigraphic analysis with scientific techniques.
The investigation at 114-119 St Aldates and 4-5 Queen Street, Oxford, was commissioned by Gilbert-Ash on behalf of Reef Estates and the British Airways Pension Trust, with consultation by Will Bedford of Orion Heritage. The fieldwork was managed by Carl Champness and supervised by Vix Hughes and Ben Attfield. The post-excavation project was managed by Edward Biddulph.
The site report will be published later this year in an Oxford architectural and Historical Society occasional paper devoted to the archaeology of Oxford.