Version classiqueVersion mobile

Urbanisme civique en temps de crise

Laurent Brassous
Alejandro Quevedo

I. — Histoires provinciales

Public buildings in the cities of Roman Britain: successes or failures?

Simon Esmonde Cleary

Texte intégral

  • 1 Millett, 1990.
  • 2 Mattingly, 2006, chap. viii et xi.
  • 3 Frere, 1987, chap. xi.
  • 4 Reece, 1980.
  • 5 Esmonde Cleary, 1989, pp. 66-85; Faulkner, 2000; Rogers, 2011, chap. iii.

1In contrast to its continental neighbours the third century does not seem to have marked a «crisis» or turning-point in the history of Roman Britain, either in general terms or as regards its cities and their public buildings and monuments in particular. Therefore the historiography of this period for Britain is very different to that for say Gaul or Spain, and this needs to be taken into account in our consideration of civic monuments. In fact, the third century is a period which in Britain remains little studied and with no clear characterisation of how that century differs from the second or the fourth centuries. Historians and archaeologists working on the textual and the material evidence for the island have always been aware of the «third-century crisis» afflicting the Continent, but have been hard put to it to detect that crisis in the evidence for Britain. The historical sources that bear on the «barbarian invasions» do not include Britain amongst the provinces that suffered in this way, such invasions and their possible consequences have never formed part of the master narrative for this century in Roman Britain. Indeed, militarily the century appears to have seen an extended period of tranquillity on the northern frontier with Hadrian’s Wall falling into disrepair in places by the end of the century. The gradual construction of the «forts of the Saxon Shore» around the coasts of the south-east of the island seems to betoken a growing menace from across the North Sea, but there is no good evidence that menace ever resulted in destructive incursions. Neither does the archaeology afford a narrative with which to characterise the century. The first and second centuries have most commonly been explained in terms of «Romanisation»1; the fourth century has sometimes been interpreted in terms of the «decline and fall» of Roman Britain or increasingly within the framework of «late Antiquity». The third century has fallen uncomfortably between these two stools, being rather too late for the «Romanisation» debate and rather too early for the debates about the fourth century; many general treatments of Roman Britain have very little to say about it2. To turn to our specific theme here, urban public buildings and monuments, there has been no consideration of them for the third century specifically. If they are considered at all it is often as an addendum to the second century, the final working-through of the trends established under the High Empire. There has, on the other hand, been a vigorous debate over the nature of late Roman (fourth-century) cities and urbanism in Britain which can be summarised in chronological order of the development of the arguments as: were they a late version of «Classical» Roman-style urbanism3; were they, like cities across the Channel, shrunken «administrative villages» in the fourth century4; or were they an expression of a distinctive fourth-century urbanism not to be judged against a yardstick devised for the High Empire5? In this debate the public buildings and monuments have played a central role since they were regarded as a barometer of the development of the cities in which they stood, with their growth, flourishing and decline mirroring the growth, flourishing and decline of Romano-British cities more generally. So the fate of these structures has significance both for our understanding of urbanism in Roman Britain more generally and also for how we see the passage from the urbanism of the High Empire to that of Late Antiquity more widely than just in this island.

  • 6 Wacher, 1995, chap. ii for a summary of knowledge up to 1990; for more recent information see Roge (...)

2In this contribution the evidence from those sites which ranked as civitas-capitals will be considered (fig. 1). Thanks to the labours of many antiquarians and archaeologists a large number of public buildings and monuments of the first and second centuries A. D is now known from the cities of Roman Britain6. Building on these foundations we may construct a picture not only of the origins and early development of individual structures but also, for some of them at least, of patterns and rhythms of use or disuse in the later Roman period. It has to be admitted, though, that the later history of these buildings was not always of major interest to early excavators. By contrast, more recent excavations have often been more careful in recording and analysing the later deposits, but because the areas excavated have often been limited it has been more difficult to relate this sample to the entirety of the monument concerned. The cities and monuments for which we have the most complete plans are to be found at those cities which having been abandoned after the end of the Roman period were not reoccupied in the middle ages or the modern era: Caerwent, Caistor St Edmund (also referred to as Caistor-by-Norwich), Silchester, Verulamium (St Albans), Wroxeter. At these «abandoned» cities some or all of the major public monuments have been recovered in their (near-)entirety over the last century and more. In addition, excavation in advance of development since the ‘sixties of the twentieth century has brought to light parts of the public buildings at a number of Romano-British cities whose sites are still occupied, such as Canterbury, Colchester, Lincoln or London. These cities furnish us with examples of a fairly limited range of public buildings and monuments: fora/basilicae; baths; macella; theatres; amphitheatres; a circus. Though these structures tend to be fairly simple compared with their continental counterparts (for instance the fora do not have temples, amphitheatres are small and constructed mainly of earth rather than stone), they nevertheless are of great value in the context of this publication since we do know a considerable amount about the later histories of at least some of them; enough to propose some general trends and possible explanations. In what follows, the monuments will not be grouped according to the cities at which they were found but they will be considered by type of building or monument as listed above: fora/basilicae; baths; macella; theatres; amphitheatres; a circus. Their later histories will then be synthesised to see if there are any discernible patterns and if so what their significance may be. «Success or failure» will be considered not in a simple linear sense but rather in conjunction with other developments at these cities in the third and fourth centuries to assess what the changes at the building-types under consideration here may mean in terms of the overall development of the cities of Roman Britain.

FIG. 1. — Map of Britain with the towns named


  • 7 Blagg, 1980.

3The plans of Romano-British fora and basilicae which are known in their entirety are remarkably similar, the chief difference being in overall size. The majority have a simple plan consisting of a central courtyard surrounded on three sides by offices, usually fronted by colonnades, with on the fourth side a basilica. The similarity of this plan to that of military principia has been remarked upon, but is now felt not to show direct military parentage but rather a similar range of needs and uses7. The feature that distinguishes Romano-British fora from so many of their continental counterparts is the lack of a temple integrated into the complex, as in Gaul, Spain or Italy. On the other hand one may point to Gallic examples such as the fora at Glanum or Ruscino or sites in the Rhineland such as Ladenburg, to show that this plan was not specific to Britain. According to Tacitus his father-in-law Agricola while governor of Britain ca. 77-84 encouraged the construction of fora,… hortari privatim adiuvare publice ut templa, fora, domos extruerent (Agricola, 21), and therefore those fora in Britain excavated in the later nineteenth or earlier twentieth century were dated to the Flavian period. More recent, scientific excavations have shown that at Caerwent, Leicester, London and Silchester they date to the Hadrianic/Antonine period, and this probably holds good for most Romano-British fora. Caerwent and Silchester are two of the «abandoned» cities of Roman Britain where the complete, or near-complete, plan of the forum/basilica complex is known from earlier excavations but where there have been major recent excavations that have brought important new evidence to light. These two sites will be considered first, before moving on to more recent excavations on parts of the complexes at Cirencester, Exeter, Leicester, Lincoln and London. After these the more debatable evidence from other early excavations at Caistor St Edmund and Wroxeter where the plan of the entire forum is known but the detailed sequence remains unclear

  • 8 Brewer, 1993.
  • 9 Fulford, Timby, 2000, pp. 573-581.

4The forum at Caerwent (Venta, civitas of the Silures) was originally uncovered in 1905, 1907 and 1909 and then reinvestigated from 1987-93, examining the eastern third of the basilica along with the rooms to its rear8. This showed that in the late third century the basilica underwent a major programme of reconstruction which apparently involved the construction of a scaffold to enable the removal of the roofs and the dismantling of the columns separating the nave and the aisles. Some of the walls and the southern stylobate were strengthened, suggesting that the basilica was then rebuilt and put back into commission. The basilica remained standing into the later fourth century when it may have started to collapse; it seems that it was demolished and its site cleared in the 380s. Even then the range of rooms to the rear of the basilica remained in use as probably did the tribunal at the eastern end of what had been the basilica (I am grateful to Richard Brewer for information on this sequence in advance of final publication). By contrast, the re-excavation of the forum basilica at Silchester (Calleva, civitas of the Atrebates) showed a very different history9. The stone forum and basilica were constructed in the Hadrianic to Antonine period (fig. 2), but there is considerable debate as to whether the basilica was ever completed. The basilica had an eccentric plan with only one stylobate for an internal colonnade where symmetry would demand two: was this purposeful or the result of the second stylobate not being constructed? The excavators note the lack of any solid flooring to the basilica or of much in the way of stonework from a superstructure of plaster from the walls. The basilica may have been left unfinished, possibly even unroofed. This is a salutary reminder to archaeologists and other workers that an excavated two-dimensional plan should not automatically be assumed to represent a finished, three-dimensional reality; from other periods we know of major building enterprises that were never finished or that took decades, even centuries, to complete. From the later third century on the area of the Silchester basilica was used for metal-working, principally iron but also bronze and pewter, on a considerable scale and associated with some temporary timber structures. The basilica would seem no longer to have been used for civic administration, but it is possible that this was still exercised from other parts of the forum complex, which have not been subject to modern re-excavation.

FIG. 2. — Plan of the basilica at Silchester (courtesy of Roman Society)

5The area of the basilica seems to have continued in use, possibly with some of the walls still standing, into at least the fifth century, with the stonework finally being robbed in the seventh century or later. The two very contrasting histories of the basilicas of Caerwent and Silchester suggest that there will be no simple, uniform pattern across Britain; Caerwent points to the maintenance of the civic vocation of the basilica and its annexes till at least the late fourth century, whereas at Silchester this vocation had been lost at least a century earlier and the basilica turned over to industrial uses.

  • 10 Brigham, 1990; Milne (éd.), 1992.
  • 11 Holbrook, 1998, pp. 99-121.
  • 12 Bidwell, 1979, chap. v et vi.
  • 13 Jones, 2002, pp. 62-64; Ibid., pp. 127-130.
  • 14 Hebditch, Mellor, 1973, pp. 41-42.

6Turning to other forum/basilica complexes where there have been modern excavations of a part of the structure, we may start with London (Londinium, later Augusta), where the eastern of the two hills on which the Roman city stood was crowned with the largest forum complex in Britain, including a basilica whose dimensions were comparable with those of the Basilica Ulpia at Rome10. Excavations in 1984-86 on rooms north of the eastern part of the basilica when taken in conjunction with other excavations show that the basilica and its annexes were constructed in the Trajanic/Hadrianic period. In the middle of the second century they were damaged by fire but restored, with a lesser fire later in the century. The complex was maintained through the third century, but at the end of that century the basilica and its associated rooms were comprehensively demolished and the building material cleared away, though probably leaving the apse that had closed the eastern end of the nave intact as a free-standing feature. Again by contrast, at Cirencester (Corinium, civitas of the Dobunni) limited excavations seem to show the apse of the basilica and the rooms around it remained in use from their construction around the turn of the first and second centuries through to the end of the fourth century. Part of the range at the far end of the forum from the basilica also remained in use with modifications including the laying of mosaics as well as re-paving of part of the forum courtyard in the second half of the fourth century11. At Exeter (Isca, civitas of the Dumnonii), where the southern end of the basilica of the forum was excavated, there was evidence for a refurbishment towards the end of the third century. This was followed by a major re-planning and reconstruction of the southern end of the basilica in the middle of the fourth century or later12. At Lincoln (Colonia Lindum) part of the eastern range of the forum seems to have remained in commission through to the late fourth century, with a timber building identified as a church, built in the forum courtyard at this date. In addition, the north wall of the forum complex remains standing to a height of 7m to this day13. The limited excavations at Leicester (Ratae, civitas of the Corieltauvi) likewise seem to show that parts of the forum complex remained in use through to the later fourth century14. The very limited evidence from the forum and basilica at Verulamium (municipium Verulamium, civitas of the Catuvellauni) again suggests continued activity into the later Roman period. Nevertheless, one must be careful with small-scale interventions as at Leicester or Verulamium and to an extent Exeter and Lincoln; though continuity of use of forum structures can be demonstrated, that does not of itself prove that the structures continued to constitute a forum. The example of the Silchester basilica shows that existing structures might be reused for purposes very unlike those of a civic amenity.

  • 15 Frere, 1971, pp. 11-14.
  • 16 Atkinson, 1942, chap. vi; White, Barker, 1998, p. 112.
  • 17 Atkinson, 1942, pp. 105-07.
  • 18 Ibid, chap. vii, pl. 73.

7The complete plans of the fora at two of the «abandoned» cities are known from excavations before World War II, at Caistor St Edmund and Wroxeter. The excavations at Caistor (Venta, civitas of the Iceni) showed that there were two successive forum/basilica complexes. The earlier probably dated to the second half of the second century and the excavator’s notes state that it was burnt down, but the evidence for this is debatable. In the later third century the forum was rebuilt, slightly smaller and somewhat less elaborate than its predecessor. Whether there was still a basilica on the fourth (western) side is uncertain: there are no foundations of the period of the second forum, though it is possible the remains of the basilica of the first forum could have been used in some way15. Nevertheless, it was clearly felt a forum of some sort was still necessary in the third- to fourth-century city. The situation at Wroxeter (Viroconium, civitas of the Cornovii) is much less clear due to the quality of the 1920s excavation and publication16. Here the forum is dated to the reign of Hadrian by an elegantly lettered dedicatory inscription. It is clear that the basilica of the forum burnt down somewhere at the end of the third century and was not rebuilt17. The excavator did come across a number of features high up in the stratigraphy of the eastern half of the forum complex, unfortunately these could not be accurately dated nor is it possible from the records to establish whether they relate to a late use of the forum or, as may be more likely, to reuse of the site of the forum at a later date, analogous to the late occupation on the site of the baths basilica (see below)18.

8Overall, no single clear pattern emerges for the fate of Romano-British forum and basilica complexes through the later Roman period. Instead there appear to be two major trends. On the one hand at Caerwent, Caistor St Edmund, Cirencester, Exeter, Lincoln and possibly Leicester there is evidence that the complex was maintained. This maintenance was not just the «passive» preservation of the legacy of the second century but often «active» preservation involving major works of restoration or rebuilding. It is worth, nonetheless, iterating the caution that because there continued to be occupation within the structure of a forum, where the excavated sample is limited this is not proof that the structure remained in use as a forum (e.g. Leicester). On the other hand, there are cities such as London, Silchester and Wroxeter where all or a substantial part of the forum and basilica did not outlive the third century. Continuation of use seem to outnumber cessation so there seems to have been no generalised abandonment of the buildings that had materialised civic autonomy, indeed a significant number of the cities of Roman Britain apparently wished to curate this expression of their civic status. Equally, some other cities took other decisions. London seems purposefully to have dismantled and cleared away its massive forum and basilica; Silchester turned the shell of its basilica (which may never have been finished) over to other uses, principally metal-working; Wroxeter abandoned at least the basilica of the forum after it was engulfed by fire. At these cities status and autonomy no longer had to be expressed in the ways inherited from the High Empire. This does not, of course, mean that such status and autonomy no longer existed; it was just that the powers and the roles that they entailed were exercised in other ways and in other locations.


9Alongside the forum/basilica, public baths were the most numerous class of public building at the cities of Roman Britain, demonstrating the importance of the practices and rituals they housed, in particular the creation of the «Roman body» through the processes of bathing, exercise and bodily grooming. All the cities where there has been sufficient survey or excavation have yielded a major set of baths, occasionally two. Roman Britain cannot as yet demonstrate any examples of the so-called «imperial baths» such as the «Barbarathermen» at Trier/Trèves, or even of more modest but still substantial establishments such as the «Thermes de Cluny» at Paris. Romano-British public baths are essentially of the Reihentyp albeit on a monumental scale and with some elaboration of plan. The question which needs to be addressed here is whether these baths show significant change in the later Roman period, and if they do whether the reasons are simply pragmatic and to do with the structure of the baths, and/or whether they are more ideological and to do with changing social and cultural norms.

  • 19 White, 1999.
  • 20 Barker et alii, 1997.

10Bath-buildings are a class where though we have a number of good examples, including some excavated with modern techniques, the later history of the buildings is often not easy to determine. Four bath-houses have been excavated in their entirety, but generally the results are disappointing for our purposes. The complex for which we have the fullest information is the one at Wroxeter, where the main baths structure was uncovered from 1859 and subsequently allowed to decay for a century before being consolidated and put on display. Despite the loss of information this entailed, it has been possible to establish an outline sequence of developments19. Construction commenced in the Hadrianic period, and the baths essentially consisted of a north-south row of frigidarium, tepidarium, caldarium with to the north a large basilica, the «baths basilica». Through the second and third centuries additions were made on east and west sides, resulting in two smaller subsidiary bath-suites. By the end of the third century the original main caldarium may have been abandoned. Thereafter the original main tepidarium was also put out of use and the structures of both these large rooms robbed. This left the baths consisting of the original main frigidarium serving the two lateral, smaller suites of baths. The date of this cannot now be established, but the main baths of the city remained in use, albeit on a much-reduced scale. Excavations on the «baths basilica» from 1968 to 1990 has produced one of the most famous and most debated sequences for the late and post-Roman use of a public building in Britain20. What concerns us most here is the sequence of disuse of the «baths basilica» and its dating. The structure was maintained in use through the fourth and into the fifth century, though the poor quality of the late re-floorings shows it was probably in an increasingly dilapidated state. By the end of the fifth century the interior of the «baths basilica» was being colonised by ephemeral timber structures, before in the mid sixth century the building had its roof removed and was systematically demolished.

  • 21 Kenyon, 1948, p. 36.
  • 22 St John Hope, Fox, 1905, pp. 341-366.
  • 23 Frere, 1971, pp. 20-22.
  • 24 Nash-Williams, 1930.
  • 25 Blockley et alii, 1995, pp. 121-207.
  • 26 Down, 1978, pp. 145-152.

11Wroxeter is the most detailed sequence we have for the later uses of a public baths in Britain (fig. 3). One other such complex which has been excavated to a comparable scale is the main bath-house at Leicester, the so-called Jewry Wall baths21. These were probably of second-century date, but no floor-levels survived within the bath-building proper, only some patchy layers in the courtyard containing third-and fourth-century coins, but these can tell us nothing about the baths themselves at this time. The baths at Silchester were completely uncovered in 1903-4 and the account of them allows some sense of their chronological development to be gained22. Probably built in the later first century, the baths were added to over time and seem still to have been in use into the late Roman period. The excavation of the simple public bath-suite at Caistor St Edmund did not yield enough dating evidence to show if they were in use in the late Roman period23. The baths at Caerwent, probably constructed in the first half of the second century, seem to have continued in use into the later Roman period, though part of the colonnade had been suppressed and a timber structure built over it24. Of the incompletely excavated baths, the baths at St Margaret’s Street, Canterbury (Durovernum, civitas of the Cantiaci) and the portico fronting them continued in use till the latter part of the fourth century when parts of the baths seem to have gone out of use and the portico was affected by flooding25. What little is known of the baths in the northwest quadrant of Chichester (Noviomagus, civitas of the Regnenses) also suggests continued use into the later fourth century, though probably on a declining scale26.

FIG. 3. — Plan of the late development of the baths at Wroxeter (WHITE, 1999, with permission)

  • 27 Fagan, 1999.

12Though the evidence leaves much to be desired, there does seem to be a consistent pattern for the public bath-houses of Roman Britain. After their construction in the later first or second centuries, the bulk of them were enlarged and reconstructed and remained in use through into the fourth century, sometimes on a reduced scale, and sometimes showing decline or abandonment towards the end of that century. It is noteworthy that despite the large size of these complexes, the complexity of their use, and the fact that they were up to two hundred years old, the effort and expense of maintaining them was still thought worthwhile, as was the expense of firing them up, maintaining the water supply and drainage, and presumably staffing them. Given the very particular and important role of bathing and baths in creating a «Roman» identity, it is surely significant that the public baths of the cities of Roman Britain continued to be maintained and used so late. This also points to a continuing cultural and social value for communal and public bathing27. This argument is strengthened by the contrast between the provision of private bath-suites at aristocratic rural as opposed to urban residences in late Roman Britain. Bath-suites were a regular feature of the rural residences, the villas, in the late Roman period and were sometimes the most architecturally elaborate parts of these structures. By contrast, bath-suites at urban residences at the same period were rare, and generally unpretentious. So it would seem that whereas luxurious bathing in the private context was as important in the villas of late Roman Britain as in the villas of late Roman Gaul or Spain, in the cities the Romano-British elite did not withdraw into the private sphere for this activity but still cultivated the public arena.


  • 28 Richardson, 1944.
  • 29 Fabre, Paillet, 2010.
  • 30 Wacher, 1995, p. 352.
  • 31 Ibid., p. 362.
  • 32 Ellis, 2000, pp. 25-90.
  • 33 Holbrook, 1998, pp. 187-188.

13There are two possible macella in Britain whose plans are fully known and two other structures claimed as macella. A macellum of basilican plan has been fully excavated, at Verulamium28. Lying across the main street of the city from the theatre, it was probably built at the turn of the first and second centuries, and consisted of a rectangular structure containing two rows of rooms/shops along its long sides facing onto a central area. The building was substantially modified in the later second century, being reduced to half its original length with the central area surrounded by a drain (for rainwater?) and closed at its eastern end by an apse. The identification by the excavator of the complex as a macellum was and remains tentative, though in its rebuilt form it has some resemblance to the macellum at Saint-Bertrand-de-Comminges29. Somewhere around the turn of the third and fourth centuries it was again rebuilt, resulting in a rectangular building, within which the rooms/shops had been suppressed and replaced with two rows of piers at 3m. intervals yielding a basilican building. The apse was demolished and in the central vessel two substantial, load-bearing, rectangular foundations installed. Whether at this time the building still acted (if it ever had) as a macellum cannot be established because of serious mediaeval robbing. But it still constituted a substantial and quite possibly public structure. Part of what may well be another basilican macellum has been excavated at Leicester30, across a street from the north side of the forum. The excavated part consists of the western end of a basilican hall with a range running north from its western end. Again the identification as a macellum is far from certain. The complex seems to have remained standing into the later fourth century, when it was destroyed by fire31. The other complete plan of a building interpreted as a macellum is from Wroxeter32. Lying in the south-western quadrant of the insula containing the public baths it consists of three ranges of rooms/shops surrounding a colonnaded courtyard with the fourth side consisting of a wall with two doorways into the colonnade fronting onto the main north-south street. The identification as a macellum is of long standing, but again there is no direct evidence: the identification is based on the similarity of the plan to sites such as Baelo and Paestum. The building seems to have remained standing into the fourth century, with possibly some demolition at that date. Part of a structure at Cirencester consisting of a range of rooms with a corridor or colonnade to either side in the insula immediately to the west of the forum/basilica has been identified as a macellum, partly because of general similarity to the Wroxeter plan, but again there is no certainty33. This building would seem to have been abandoned in the second half of the fourth century.

  • 34 White, Barker, 1998, pp. 114-116.

14There is clearly a problem with these structures in that at none of them is it demonstrable with certainly that they were indeed macella. Nevertheless their scale and their plans along with their siting in the heart of the cities in which they stand strongly argue for them being public structures and macella is an attractive possibility. All of them seem to have remained in use into the fourth century, though as noted for the Verulamium example and as has been argued for Wroxeter34, it is possible that they may no longer have served as macella but may have been turned over to some other use.


  • 35 Frere, 1970.
  • 36 Crummy, 1982.
  • 37 Dunnett, 1971, esp. p. 41.
  • 38 Holbrook, 1998, pp. 142-145.
  • 39 Kenyon, 1935, pp. 232-241.

15Theatres seem never to have been very common in Roman Britain and comprised structures both in the «classical» D-shape and one of the «Gallic» theatre/amphitheatre hybrid form. The best-known example of the «classical type is that at Canterbury35 dating in its original form perhaps to the late first century. In the early third century this was completely rebuilt to a strictly semi-circular plan. Its history in the later Roman period is unknown, though parts of it may well have remained standing into the early middle ages. Another D-shaped theatre is known at Colchester (Colonia Claudia Victricensis Camulodunum) immediately to the west of the precinct of the Temple of Claudius (see below) and dating probably to the second century. No evidence for the later Roman period was recovered36. Also at Colchester, some 5 km. to the south-west of the walled city, was a theatre forming part of the large extra-mural religious complex at Gosbecks37. A stone theatre, replacing an earlier timber one, and of a somewhat unusual design was constructed in the mid to late second century, abandoned and systematically demolished probably within a century of its construction. A probable D-shaped theatre lay at the north-western edge of Cirencester, but nothing is known of its date or later history38. The best-known urban theatre in Britain is also somewhat atypical, being the hybrid theatre/amphitheatre at Verulamium (fig. 4), associated with a substantial temple precinct39. The seating banks consisted of earth mounds retained by flint-and-mortar walls and defined about three quarters of a circle, with a small stage-building in the remaining space. Constructed apparently around the middle of the second century, the theatre was completely rebuilt and extended around A.D. 300. But by the later fourth century the theatre seems to have passed out of use and to have been used for dumping large quantities of domestic waste, containing coins down to the House of Theodosius. The small number of theatres known from Britain and the little that is known about them precludes any general statement, but it is perhaps worth noting that the major reconstructions at Canterbury and Verulamium are amongst the latest major refurbishment projects on earlier public buildings.

FIG. 4. — Plans of the theatre at Verulamium (Left plan: second century; right plan: fourth century)


  • 40 Fulford, 1989, esp. pp. 57-58.
  • 41 Holbrook, 1998, pp. 145-175.
  • 42 Bradley, 1975, esp. pp. 78-79.
  • 43 White, 1936.
  • 44 Little, 1971.
  • 45 Bateman, Cowan, Wroe-Brown, 2008.

16Amphitheatres were considerably more numerous at the cities of Roman Britain than theatres, perhaps because of the strong military presence in the province. Even so, they are of modest dimensions and constructed with earthen seating-banks retained by timber or stone walls. None of them approaches the scale of major stone-built amphitheatres in Gaul from Arles to Amiens. The most extensively excavated example is that just outside the north-eastern walls of Silchester40. The amphitheatre was originally constructed in earth and timber towards the end of the first century A.D. and subsequently reconstructed probably in the mid second century, with a major reconstruction involving rebuilding all the retaining walls in flint-and-mortar towards the middle of the third century. There is evidence of activity in the arena to at least the middle of the fourth century, but whether the amphitheatre was still being used for spectacles at this date is impossible to say. Excavations on the amphitheatre at Cirencester41 have shown that it was constructed in earth and stone early in the second century and went through various rebuilds thereafter. By the early fourth century it seems to have been becoming derelict, and after the middle of that century there was a major reconstruction involving the removal of all the stonework and some replacement in timber. It seems most unlikely that it was still functioning as an amphitheatre from this date, though there was evidence of activity including the narrowing of one of the entrances and metalling in the arena in the very late or post-Roman period. The evidence from the amphitheatre at Dorchester (Durnovaria, civitas of the Durotriges), which seems to have been constructed in the first century A.D. is that it may have been out of use and decaying by the end of the second century with burials being made in one of the entrances42. The limited excavation on the amphitheatre at Chichester suggests that it was abandoned and robbed within the Roman period, possibly as early as the late second century, though this is based largely on the absence of later material43. Nothing is known of the later history of the amphitheatre at Carmarthen (Moridunum, civitas of the Demetae)44. London has an amphitheatre, but it may well be military at least in origin and associated with the nearby Cripplegate fort45. Constructed in the later first century in timber, it was reconstructed in stone with timber seating in the earlier second century, but was abandoned and partially robbed in the fourth century, when it may have been used for rubbish dumping. Overall, amphitheatres seem to have been more an expression of urban development under the High Empire. Some clearly did not survive into the later Roman period, with others being abandoned or put to other uses by the fourth century.


  • 46 Crummy, 2008.

17Only one circus is known from Britain, and that discovered only recently at Colchester46. The survival of late deposits is not good, but the outer wall of the structure appears to have been removed in the late Roman period. The coin profile from the circus suggests that it may have ceased to be used in the later third century, but this is not certain.


18Before going on to consider what the fate of individual buildings or classes of buildings in the later Roman period in Britain was and what this may signify, it is prudent to make two methodological points. Both of these have arisen in the discussions above, but should be made explicit as part of a general treatment of the evidence. The first is to do with the use of structures: merely because a structure continued to be used into the later Roman period does not of itself mean that it was being used for the same purpose as the one for which it was constructed. An obvious example of this is the Silchester forum-basilica. This point is particularly important when considering more recent excavations which may only concern themselves with part of a much larger structure; modifications in the later Roman period may mean continued use of the building for its original purpose, equally they may signal conversion to another purpose. The second is to do with the disuse of a building; sometimes it is difficult to tell when a building or other structure was disused in default of evidence for collapse or demolition or robbing. Good examples of this are structures such as theatres and amphitheatres; does the lack of evidence for collapse or demolition mean they were still in use, or had been abandoned and left to rot? In the case of structures such as the theatres at Canterbury or Colchester this problem is accentuated by the severity of post-Roman disturbance and robbing which has removed all evidence for the use-surfaces leaving only the wall foundations. On both these counts we need to be careful about how much we can assume about continuity or discontinuity both of use and of function.

  • 47 Laurence, Esmonde Cleary, Sears, 2011, chap. vii-viii.
  • 48 Esmonde Cleary, 2005.

19Bearing this in mind, what can we say about the fate of the buildings and monuments of the cities of Britain inherited from the first and second centuries? And what may their continued use of their disuse say about the nature of urban landscapes and urban ideologies in this later period? The first thing to say is that there was not any comprehensive suppression of the structures to make way for other expressions of the urban phenomenon. There was thus not a widespread remodelling of urban form and function as took place, for example, across the Channel in Gaul north of the Loire at this period. But equally, not everything remained the same; it is clear from the presentation of evidence above that some monuments at some cities were suppressed and others were substantially modified. It is this very variability that makes it difficult to offer generalising conclusions since there is an inherent tension in the levels at which explanation must operate. On the one hand, there is the level of the individual monument in the individual city: monuments were abandoned, demolished and dismantled for particular reasons to do with the monument itself or the needs of the city concerned. But to conduct debate at this level would be to present a catalogue without an argument. On the other hand there is the level of classes of monuments and the monumental urban heritage of the entire diocese of the Britains. It is this level that is the one to be employed here, and of course general trends may well have influenced decisions at individual cities. In the presentation of evidence above, it became clear that a significant proportion of the major public building complexes of the High Empire, the fora/basilicae and the baths, survived into the fourth century. It was noted when discussing the fora/basilicae that this involved not just the continuation in use of a building but its active maintenance and rebuilding, as for instance with the forum-basilica at Caerwent or the baths at Wroxeter. So what we are seeing here, it was argued, was not just the passive reception of an earlier building but its active curation so that it continued to be of use for the urban community. It is noticeable that the two building classes under consideration here, fora/basilicae and baths, were not only the most numerous and widespread of the public buildings of Roman Britain but also those most involved in the creation of a sense of communal and individual citizenship. Communally, the forum and basilica expressed the civic autonomy of the civitates concerned and functions such as administration, justice and the preservation of the memory of the city through archives, statues and inscriptions. They were also places that were «public» in the sense that they were accessible (as far as we can tell) to all sorts and conditions of persons, so all classes of citizen and non-citizen might be encountered in these spaces. Likewise the baths were important centres of public resort where again access was largely not conditional on questions of status, though of course it was conditioned by gender. Thus the cities of later Roman Britain for the most part continued to express a sense of civic community and identity through their major public buildings. Even at cities, such as Silchester or Wroxeter, where the forum and basilica became disused, it would appear that the baths continued to be maintained and used, so this argument for a continuity of civic identity holds at least in part. But fora/basilicae and baths were not just expressions of the communal, they were also experienced by the individual. They thus helped shape the citizen both ideologically through participation in public and communal activities and physically, especially in the baths through the regimens of bathing, grooming and exercise47. If the fora/basilicae and the baths were for the most part classes of buildings that continued in use and might actively be curated, then what of the other classes. The evidence for the macella of Romano-British cities is numerically small, though both of the best-known examples, at Verulamium and Wroxeter, seem to have continued standing into the fourth century, even if that at Verulamium had been totally rebuilt and could well have changed its function. Theatres seem to have continued in use, so far as can be told on the basis again of a small numerical sample, and it was noted that at both Canterbury and Verulamium major rebuildings there were amongst the latest such works of monuments of the High Empire. By contrast, the amphitheatres lying at the edges of the cities seem on the whole not to have fared so well, with the majority of examples apparently no longer in use by the fourth century, and those that were still in use such as Cirencester and Silchester seem to have ceased operating in the fourth century. So in the terms of the title of this article, «successes and failures», it would be difficult to claim that these monuments and the ideologies they materialised were failures, since more of them continued in use, often it appears serving their original functions, that had been abandoned or demolished. A traveller through the cities of late Roman Britain would often have found them to retain much of their monumental equipment inherited from the first and second centuries. It is worth noting that there was another aspect of this heritage that was still being curated, the street-grid. There is clear evidence from many of the cities for re-metallings of the streets into and through the fourth century, with only a small number of instances of encroachment by buildings onto the streets towards the end of the Roman period. Thus again the urban form of the cities retained their earlier aspect, and the streets also acted as spaces accessible to the generality of the citizens, and perhaps acted as the settings for civic rituals and other expressions of communal identity48.

  • 49 Faulkner, 2000; Perring, 2002, esp. chap. iii.
  • 50 Esmonde Cleary, 2013, chap. v.

20The approach adopted so far, by privileging certain classes of building and monument, risks decontextualising them from wider phenomena of urban development and change, and indeed as has happened in the past making them stand proxy for that process of urban change. So these developments in the public buildings need now to be replaced into that wider context and a broader picture gained of what happened to the cities of later Roman Britain. A major change in the urban fabric of later Roman Britain that has long been recognised was the increasing colonisation of the urban space by the residences of the richer classes, the so-called «town-houses», which became more widespread and consumed more urban land through the later third century and on into the first half of the fourth century49. This obviously has important implications for the social composition of the urban populations and also probably for economic activities. But it also had important implications for the situations in which power was exercised. Many of these houses had substantial principal reception rooms, often floored in mosaic, sometimes taking the form of an apsidal room. In this they reflect developments at rural residences in Britain, villas, and also further south in southern Gaul, Spain and elsewhere in the West where the domestic context became more important as a seat of power50. In this case it may not be surprising that some public buildings such as the forum and basilica may have fallen out of use as power passed from the public into the private setting. Nevertheless, as already noted, more fora and their basilicae remained in use than not. One other feature of these late «town-houses» that was noted above when discussing the fate of the public baths is that few of them had private bath-houses, and those that did exist tended to be relatively small and simple. It would appear that in late Roman Britain the tradition of using the public baths and thus behaving as a public citizen retained its force.

21Fora/basilicae, baths, macella, and so on were not the only public buildings and monuments ever possessed by Roman cities in the West. By the late Roman period there was another hugely important public undertaking that defined the city physically and ideologically and needs to be considered here, again to set the development of public buildings and monuments in its proper context. This undertaking was of course urban defences.


  • 51 Esmonde Cleary, 2003; Id. 2007.

22Britain is unique in the Western Provinces in its urban defences, with a large number of cities and «small towns» defended before the end of the second century. The sequence of development of defences at the principal cities of Roman Britain is well enough known51. A small number of cities, chiefly coloniae, were walled by the end of the first century A.D. In the late second century nearly all the cities along with a number of «small towns» received earthwork defences, sometimes with stone gates (the defences of London constructed at this time were entirely of stone). At the cities these defensive circuits encompassed the entire area then inhabited, sometimes more. In the course of the third century stone walls were inserted into the front of these earthen ramparts. A small number of cities not already defended, such as Canterbury and Chichester, received stone walls for the first time in the later third century. Unlike contemporary urban defences the other side of the Channel at the same period, these late circuits did not defend a small area at the core of the city but like other British city defences encompassed the then occupied area (in the case of the two cities just mentioned, 53 and 40 ha, respectively). The closest parallel for this is the set of city walls in north-western Spain such as Astorga and Lugo. At a number of cites and «small towns» the walls were further strengthened in the fourth century by the addition of projecting towers, but this was by no means consistent.

  • 52 Ibid. 2007.
  • 53 Faulkner, 2000, esp. figs. 2.14 and 2.15.
  • 54 Esmonde Cleary, 2013, p. 124.

23The traditional explanation for the construction of defences, particularly the major phase of earthwork defences in the later second century and wall-building in the third, was to assimilate them to supposed historical events. Dissatisfaction with this explanation, particularly with problems of dating, has led to an explanation in terms of civic monuments being currently more favoured52. This is not to deny the defensive nature of these undertakings; it is, though, to acknowledge that there may have been other motives behind their construction and also to recognise that these defences may have taken some time to construct. Examination of the fabric of the wall-circuits in particular show no sign of haste in their construction, and where the facing survives it is carefully finished in petit appareil with regular cordons of brick or stone at intervals. The gates, particularly the ones on the principal access-roads, were often monumental in appearance, with vehicular and pedestrian portals and flanking rectangular or semi-circular towers, rather than smaller and more easily defensible. There is therefore good evidence that monumentality was an important element in the provision of these circuits and that they were designed to impress. In this sense they can be seen as much as a public building or monument as a forum or a set of baths; indeed in terms of sheer volume of materiel used to build them they were clearly a much larger undertaking. Analysis of the chronology of different type of construction work at the cities of Roman Britain53 shows that whereas the construction of public buildings and monuments (Faulkner’s «official buildings») peaked in the second century, the labour expended on defences was essentially a phenomenon of the third century. It is possible to interpret this as the diversion of expenditure on public works from the expression typical of the High Empire (fora, baths theatres…) to the new civic monumentality of the Late Empire (walls). The role of walls as the defining monument of the late antique city is well known, and can be seen artistically in a number of mediums from sculpture to manuscript illumination as well as the use of muri and moenia as literary metonyms for the city54. In such a reading the walls at the cities of Britain in the third and fourth century express a changing view of urban monumentality and of the self-representation of cities and their communities.

  • 55 Laurence, Esmonde Cleary, Sears, 2011, chap. v-x.
  • 56 Blagg, 1990.

24So if we replace the public buildings and monuments of the first and second centuries in the context of the development of Roman-style urbanism in Britain we find that «successes or failures» perhaps imposes too limiting an agenda on the subject, one too much implicated in notions of «third-century crisis». In the first and second centuries it would seem that a great deal of resource was invested into public buildings and monuments of the types widespread in the Latin-speaking provinces of the Roman empire55 and which materialised elite competition both within cities and between cities, or rather within civic elites and between civic elites, since ultimately these are the people who financed such ventures. In fact the evidence from Britain is that individual or family patronage of the type prevalent in Italy and the provinces bordering the Mediterranean was rare in Britain, such epigraphic evidence as we have suggests the patronage of groups of patrons56, so part of the reason for the unambitious nature of most Romano-British public buildings of this epoch may be that there was not the same level of competitive stimulus. So far as we can see, many of these buildings and monuments in the first and second centuries adorned cities that were in many ways under-urbanised since areas of the street-grids remained un-built-up, and what building-stock there was seems largely to have been artisan rather than elite. In the later second century, from the Antonine period, this began to change with the increasing appearance of the larger elite residences, «town houses», a process that continued for the next hundred years and more. Interestingly, this development coincides both with a slackening of investment into the «traditional» monuments (the example of the Silchester forum basilica warns us that some structures may even have been left incomplete) and with the move to investment in defences, first in earthwork later in stone. So in one perspective we may be looking at a change in preferred urban status-marker, though as we have seen fora, baths etc. were by and large not abandoned but now lay within the new, monumental sets of defences. In another perspective, it would seem that as the elites increasingly came to reside in the cities they spent their money on their private residences and on the defences, suggesting that what we are seeing is not «successes or failures» but changing priorities in the expenditure of elite surplus and changing views on elite auto-representation. Unfortunately we have no inscriptions to show who ordered or paid for these wall-circuits. The public buildings and monuments of the Romano-British cities of the High Empire were expressions of their time and place, as were their cousins across the Western Empire. As time moved on and with it political, social and cultural expectations, then new expressions developed. The monumental legacy of the High Empire was adapted to meet these changing expectations, but for many cities in Britain some of their inherited parure monumentale continued to have meaning and to perform its original functions; equally some of it no longer served its original functions and was either modified to serve new needs or suppressed.


1 Millett, 1990.

2 Mattingly, 2006, chap. viii et xi.

3 Frere, 1987, chap. xi.

4 Reece, 1980.

5 Esmonde Cleary, 1989, pp. 66-85; Faulkner, 2000; Rogers, 2011, chap. iii.

6 Wacher, 1995, chap. ii for a summary of knowledge up to 1990; for more recent information see Rogers, 2011, chap. v.

7 Blagg, 1980.

8 Brewer, 1993.

9 Fulford, Timby, 2000, pp. 573-581.

10 Brigham, 1990; Milne (éd.), 1992.

11 Holbrook, 1998, pp. 99-121.

12 Bidwell, 1979, chap. v et vi.

13 Jones, 2002, pp. 62-64; Ibid., pp. 127-130.

14 Hebditch, Mellor, 1973, pp. 41-42.

15 Frere, 1971, pp. 11-14.

16 Atkinson, 1942, chap. vi; White, Barker, 1998, p. 112.

17 Atkinson, 1942, pp. 105-07.

18 Ibid, chap. vii, pl. 73.

19 White, 1999.

20 Barker et alii, 1997.

21 Kenyon, 1948, p. 36.

22 St John Hope, Fox, 1905, pp. 341-366.

23 Frere, 1971, pp. 20-22.

24 Nash-Williams, 1930.

25 Blockley et alii, 1995, pp. 121-207.

26 Down, 1978, pp. 145-152.

27 Fagan, 1999.

28 Richardson, 1944.

29 Fabre, Paillet, 2010.

30 Wacher, 1995, p. 352.

31 Ibid., p. 362.

32 Ellis, 2000, pp. 25-90.

33 Holbrook, 1998, pp. 187-188.

34 White, Barker, 1998, pp. 114-116.

35 Frere, 1970.

36 Crummy, 1982.

37 Dunnett, 1971, esp. p. 41.

38 Holbrook, 1998, pp. 142-145.

39 Kenyon, 1935, pp. 232-241.

40 Fulford, 1989, esp. pp. 57-58.

41 Holbrook, 1998, pp. 145-175.

42 Bradley, 1975, esp. pp. 78-79.

43 White, 1936.

44 Little, 1971.

45 Bateman, Cowan, Wroe-Brown, 2008.

46 Crummy, 2008.

47 Laurence, Esmonde Cleary, Sears, 2011, chap. vii-viii.

48 Esmonde Cleary, 2005.

49 Faulkner, 2000; Perring, 2002, esp. chap. iii.

50 Esmonde Cleary, 2013, chap. v.

51 Esmonde Cleary, 2003; Id. 2007.

52 Ibid. 2007.

53 Faulkner, 2000, esp. figs. 2.14 and 2.15.

54 Esmonde Cleary, 2013, p. 124.

55 Laurence, Esmonde Cleary, Sears, 2011, chap. v-x.

56 Blagg, 1990.

Table des illustrations

Légende FIG. 1. — Map of Britain with the towns named
Fichier image/jpeg, 148k
Légende FIG. 2. — Plan of the basilica at Silchester (courtesy of Roman Society)
Fichier image/jpeg, 169k
Légende FIG. 3. — Plan of the late development of the baths at Wroxeter (WHITE, 1999, with permission)
Fichier image/jpeg, 112k
Légende FIG. 4. — Plans of the theatre at Verulamium (Left plan: second century; right plan: fourth century)
Fichier image/jpeg, 56k

© Casa de Velázquez, 2015

Conditions d’utilisation :

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search