Fishbourne - New Thoughts on Society

Ernest Black argues that if the building recently excavated to the east of Fishbourne Roman Palace was a shrine in the native tradition and that if the north range indicates accommodation for a native aristocrat’s household, then, indeed, the palace’s owner was a Briton.

In 1971, Professor Barry Cunliffe published his report on the large Roman building with its numerous early mosaics at Fishbourne in West Sussex. Since then he has maintained that this was the home of Togidubnus (then believed to be called Cogidubnus), a client-king ruling his people by arrangement with the Romans. Other candidates for ownership have been favoured by other scholars – perhaps the Roman governor (so making the palace imperial property?) or perhaps a successful negotiator (businessman). In discussing these possibilities little reference has been made to the palace itself, beyond the obvious point that its architecture and decoration are fully Roman and that it was clearly the product of great wealth. Perhaps it is time to give it a closer look at the palace, to see what the accommodation, and the arrangement of its various components, suggests about the sort of household that may have lived there.

First, we need to ask what we know about the social structure of a native chief’s or king’s household at the time of the Roman invasion in the mid-first century AD. There is no direct evidence, either written or archaeological, only hints. The historian Tacitus tells us that Queen Cartimandua of the Brigantes replaced her husband, Venutius, with his armour-bearer, Vellocatus. Vellocatus is assumed to have been of high social standing himself so perhaps there was a class of office-holders – druids, bards, doctors and the like, as well as armour-bearers – at the top level of a royal household whose status reflected that of their lord. This can be supported by the evidence from the cemetery at Stanway near Colchester (see plan above). The Colchester Archaeological Trust excavated four large ditched burial enclosures spanning the decades either side of AD43. In each enclosure was a sunken “mortuary chamber”, where a body may have been laid out for a period for rituals and mourning before eventual cremation. Burial was not within the mortuary chamber. Instead, the ashes may have been laid to rest within a barrow that has since been totally removed or were disposed of in some unknown way. However, there were subsidiary burials which have survived. In enclosure 3, a burial was found with articles including an inkwell, and another with a spear and possibly a shield. Could they have been the burials of a scribe and a bodyguard/champion respectively, belonging to the household of the aristocrat whose body had been displayed in the mortuary chamber at the centre of the enclosure? In enclosure 5, there was a burial which deservedly attracted much publicity in the national media when it was excavated in the summer of 1996. The burial contained a set of gaming counters still in the positions they had occupied on the board when the grave was sealed. It also contained a set of surgical instruments, indicating a doctor’s grave. So these rich satellite burials seem to provide evidence for the importance and high status of functionaries attached to a native aristocratic household in the first century AD, but what sort of accommodation would reflect the structure of such a household?

This cannot be successfully measured for the Iron Age where, apart from size, distinctions in housing have left no surviving evidence, but it may be visible in Roman villas. J.T. Smith has identified “units” of accommodation and postulated a form of “joint proprietorship” at many villas. Could what appears as “joint proprietorship” sometimes represent the structure of an aristocratic Celtic household?

At Fishbourne little is known about the south range of the palace. However, major units of accommodation, comprising suites of rooms accessed from a hallway (medianum), can be identified in the northern part of the west range and in the north range (far left). Two further units, similar in plan to simple villas, each lay to the side of its own small courtyard in the east range. There is no difficulty in identifying the quarters of the owner of the palace. The domestic hypocaust in the west range, an early type designed to heat an adjoining room indirectly, is the only one known in the original building. It is also here that some rooms with mortar floors, rather than of mosaic, attest to the permanent presence of servants. The rest of the range may have been assigned to members of the owner’s immediate family. A second suite can be identified in the northern part of the range and another two could have existed in the southern part. Since the west range was raised on a podium above the level of the other ranges there can be no doubt about the higher status of its occupants. In the north range the rooms were floored with mosaics. Were these intended for guests? This seems unlikely because the scale of the individual suites is comparable to that of those in the west range and there seem to be no rooms suitable for accommodating the personal servants who would have accompanied their masters. There is little information about the two suites in the east range, but the presence of rooms with mortar floors as well as one probable mosaic suggests a mixture of servants and high-ranking occupants. This and the provision of private courtyards makes it likely that these two suites were designed to house guests and some of their retinues. Is the answer that the north range was occupied by high-status functionaries of the household?

It is notable that when the easternmost suite in the north range was demolished in the second century the remaining two suites were subdivided so that the number of rooms remained the same. It looks as if the requirement for accommodation remained constant over time, as would be the case with the office-holders of a great household.

Excavations by David Rudkin and John Manley for Sussex Archaeological Society from 1995 onwards have revealed the complete plan of a masonry building to the east of the palace. The plan has a superficial resemblance to a military headquarters building (principia) but there are no other military features to go with it since it is dated to Period 1B or 1C, after the military occupation of Period 1A. When the palace was built (Period 2) this building remained standing and continued to do so through the second century AD.

Indeed, it is clear that the planning of the east range of the palace took careful account of this pre-existing building. It was situateded between, and in advance of, the two most important structures in the palace’s eastern facade. Although floor levels and doorways do not survive it is a reasonable inference that this building faced east also. But what was its function?

Small rectangular structures from Iron Age settlements have sometimes been identified as shrines. Similar structures, built of masonry, can be found in the plans of Roman villas such as Hambleden (Buckinghamshire), Cosgrove (Northamptonshire) and Beddington (East Sussex). They are sometimes prominently situated in front of the main entrance, as at Darenth, Kent, where a monumental pool lay between house and shrine. Such buildings may have provided the focus for the worship of a protective deity or even the settlement’s founders or ancestors. Admittedly, this is a guess and no religious artifact or bone assemblage from a sacrifice is generally found in association with them. Could the building east of the palace at Fishbourne have been such a shrine, more elaborate than those elsewhere but embodying the same spiritual protection for the settlement? On this interpretation the room on the east side of its courtyard would have been an entrance hall and the open-fronted room on the west side would have been the shrine itself.

It is notable that there was access from the north and south galleries to the area in the courtyard just in front of this room, where an altar would be expected. The shrine, if it existed, was not a substantial or a permanent feature since no foundations were found. The Triangular Temple at Verulamium, St. Albans, echoes some of the features of the Fishbourne building, and individuality of design was an option in the third quarter of the first century AD when what was later to be the standard form of the Romano-Celtic temple was just beginning to be introduced. The re-channelling of the stream in Period 1C so that it flowed between the “shrine” to the east and the other known masonry buildings to the west could even have been intended to enhance its liminal status.

If the building east of the palace was indeed a shrine in a native tradition and if the accommodation and the palace’s north range does reflect the high status of office-holders in a native aristocrat’s household – in contrast to the slave or freedman status of such men in a Roman context – we can conclude that the palace was built for a man who was probably a Briton.

It would be going too far to say that this was Togidubnus, since we do not know whether he was still alive when the palace was built. However, it may have been the home of Togidubnus’ heir, Tiberius Claudius Catuarus, whose inscribed gold signet-ring was found close by in 1995.

ERNEST W. BLACK