------- - - - - - - - - - - - - - - --- -

TOWNS IN LATE ANTIQUITY:

IOL CAESAREA AND ITS
CONTEXT

by

T.W. Potter

Deputy Keeper, Department ofPrehistoric

and Romano-British Antiquities
The British Museum
 CONTENTS

IAN SANDERS MEMORIAL FUND

Occasional Publication 2

LIST OF ILLUSTRATIONS VI
Published by the Ian Sanders Memorial Committee

I
ABBREVIATIONS Vfl
Department of Archaeology and Prehistory
University of Sheffield, Sheffield SlO 2TN PREFACE 1
1 INTRODUCTION 3
Distributed by Oxbow Books,
Park End Place, Oxford OXI IHN 2 THE CHERCHEL EXCAVATIONS 7
The history of the town 7
The Forum excavations, 1977—81 22
© T. W. Potter 1995 i) The early phases: Punic and Juban lol 23
(Phases 1 and 2)
“ii) lol Caesarea in Imperial and Vandal times 32
A cataloguing in publication record/or this book (Phases 3 and 4)
is available from the British Library iii) The Byzantine period and its aftermath (Phase 5) 48
iv) The Islamic period 61
ISBN 0 9521073 1 7 100Q 7 9 9 /3 CIVIC CENTRES IN LATE ANTIQUITY: THE WIDER CONTEXT 63
Some north African evidence 64
Late-antique cities in the eastern Mediterranean 80
Cover design by Karen Hughes
Italy 90
Some conclusions 99
A 4 CHANGE IN LATE ANTIQUITY: AN ITALIAN EXAMPLE 103
BIBLIOGRAPHY 109

Printed in Great Britain

at the Short Run Press. Exeter

V
f N
LIST OF ILLUSTRATIONS 26. Plan of the site in the sixth-seventh centuries. 53
27. Interpretation of the early medieval phase. 54
1. Principal sites mentioned in the text. 4 28. Corner of an early medieval house on the basilica 55

I
2. lol Caesarea, plan. mosaic.
10
3. Cherchel, the theatre. 29. Postholes in the church floor. 57
11
4. Roman sites in the vicinity of Cherchel. 30. The early medieval kilns in a pit in the church floor. 58
12
5. The Labours of the Field mosaic. 31. Briquetage from the early medieval kilns. 59
14
6. Tipasa, the cathedral. 32. Stratigraphy over the basilica. 60
18
7. Trois-lots, plan. 33. Belalis Major, plan. 65
20
8. Plan showing the relationship between the forum site and 34. Oil press on a street at Sbeitla. 68
24
the theatre. 35. The entrance to the forum at Sbeitla. 70
9. Statue of Eirene. 25 36. The forum and a large church at Diana Veteranorum. 72
10. General view of the excavation. 27 37. The Christian quarter at Djemila. 74
11. Plan, Punic phase. 28 38. The agora at Philippi, Thrace. 76
12. Section showing the Punic town defences. 30 39. Features cut into the edge of the agora at Philippi. 77
13. Capital of Augustan date. 31 40. Ephesus, plan. 81
14. General plan of the site. 33 41. Ephesus, the Embolos. 82
15. Plan of the site, c. AD 430. 35 42. Small church built across the main street at Perge. 86
16. One of the late-Roman wooden stalls. 37 43. Plans of successive main streets at Antioch-on-the 87
17. Detailed plan of the forum. Orontes.
38
18. Section through late-Roman deposits in trench T. 44. Late-Roman blocked colonnade at Olympia. 89
40
19. The basilica mosaic. 45. Aerial photograph of Como. 91
41
20. A general view of the probable church. 46. Colle San Pietro, Tuscania: the early medieval Street. 97
42
21. Plan of the church. 47. Late-antique Rome, plan. 100
43
22. Graffito on the church threshold. 48. Monte Gelato, phase plan. 104
45
23. Some of the late-Roman amphorae. 49. Monte Gelato, late-Roman wooden bins or pens. 106
46
24. Collapsed masonry over the southern end of the basilica. 50. Monte Gelato, chronology of deposits overlying a late- 107
50
25. A fallen column and capital, later incorporated into an Roman floor.
51
early medieval house.

vi
vii
I
ABBREVIATIONS
I.
Afr. Roin. L ‘Africa romana. Atti dei convegni di studio, Sassari.
PREFACE
Ant. Afr. Antiquités africaines.
ANRW Aufstieg und Niedergang der römisclzen Welt.
i offer here an enlarged and revised version of the Fourth Ian Sanders
Arch. Med. Archeologia Medievale. Memorial Lecture that I gave at the University of Sheffield on 14 March
1990. It was a particular privilege to be invited to do so, and it was an
BAC Bulletin archéologique du Comité des travaux occasion that I found moving and memorable. I thank all who made it so,
historiques. and especially Professor Keith Branigan, Professor Johii Collis, Dr John
CIL Corpus Inscriptionum Latinarum. Drinkwater, Dr John Lloyd and Dr John Moreland (who has additionally
been more than helpful at publication stage, as has Mr. Russell Adams).
RAI Académie des inscriptions et belles-lettres. Comptes- Ian Sanders worked in the east Mediterranean while my fieldwork has
rendu. always been in the west. It is a divide that is all too common in
archaeological studies, and one that I personally have tried to bridge in
JRA Journal of Roman Archaeology, recent years by travelling as much as possible in the east (not least thanks to
JRS Journal of Roman Studies. Swan I-{ellenic Cruises). Thus, while the main focus of this modest work lies
in Algeria, and particularly upon modern Cherchel’s illustrious Roman
MAAR Memoirs of the American Academy at Rome. predecessor, lol Caesarea, I have sought to range widely in search of a
MEFRA Mélanges de l’Ecolefrancaise de Rome. Antiquité. context. My work at Cherchel made me think hard about what happened in
cities in late antiquity’, and gave a focus to my visits to other sites.
MEFRMA Mélanges de l’Ecolefrançaise de Rome. Moyen Age. Something of this will emerge in the pages that follow.
Here I particularly thank my Algerian co-director at Cherchel, Dr Nacera
PBSR Papers of the British School at Rome. Benseddik, for a memorable five seasons of excavations between 1977 and
1981; Professor Philippe Leveau, the acknowledged authority on the ancient
city; and the many other scholars who have contributed indirectly to this
volume, especially Dr D M Bailey, Dr Gillian Clark, Dr Ian Freestone, Dr
John Riley, Dr Susan Walker and Dr Roger Wilson. I owe a great debt of
gratitude to Mrs Kate Down of my department, for so cheerfully and
efficiently converting my handwritten manuscript into an elegant
computerised script; and especially to my wife, Sandra, who did everything
she could to ensure that I finished to time, and remained patient, as did our
two children, with my desk—bound hours.

T W Potter
Summerfield
Svdenham
April 1993

viii
1
I

INTRODUCTION

The real genesis of this volume lies many years back with youthful
speculation about the fate of the vast number of late-Roman villages and
farms of the Fenland of eastern England. How could it be that the population
seemingly disappeared without trace in the fifth century, leaving an empty
landscape? Later on, working on the South Etruria field survey in central
Italy, much the same problem arose. The countryside is dominated down to
this day by an impressive series of medieval villages, many deserted but
some still occupied, set on hilltops, spurs and promontories. The
documentary evidence is such that it is clear that most were political
creations of the tenth and eleventh centuries, a process of concentrating the
population in new settlements known as incastellamento. Yet the villas and
farms which formed the principal sites of the Roman landscape rarely
seemed, on artifactual grounds, to have outlived the fifth century, leaving an
uncomfortable archaeological gap of many centuries. Nearer to Rome, some
early medieval farms are known; further north they appeared conspicuously
rare, leading to the suspicion that the medieval villages might have come
into being at a date much earlier than the documents implied.’
Several excavations, and a great many years, later, that problem (like that
of the Fenland) still remains largely unresolved; but that is not our central
concern here. This enquiry stems in particular from a chance opportunity to
excavate an area in the heart of one of the more important cities of north
Africa, the capital of the Roman province of Caesarea Mauretanensis, Jo!
Caesarea, modern Cherchel in Algeria (Figure 1). The circumstances that
brought about this opportunity will be described in the next chapter, as will
something of the results. Here, we must say that, as luck would have it, the
site at Cherchel proved to have a remarkable sequence of buildings and
deposits, up to five metres deep, which registered archaeologically a
significant part of the towns history over the past 2,500 years. Particularly

T. W. Potter, Arch. Med. 2 (1975), 215-36; id., The changing landscape of south
Etruria (London 1979).

3
 Towns in Late Antiquity: Jol Caesarea and Its Context Introduction

I— —

especially from layers overlying the fora of many towns. This is one reason
why the data from Cherchel are set out as fully as possible, so that readers
can make up their own minds about the chronological conclusions that are
drawn. Nevertheless, a study of some of the comparative material from north
Africa and elsewhere does make it possible to say something about changes
in the urban landscape in late antiquity. Much, inevitably, is not novel. But
Cherchel does have its surprises and they cannot be properly appreciated
without examining something of the broader picture, however schematically.
However, it is first to Cherchel that we must turn.

Figure 1 Principal sites mentioned in the text.

important was the fact that part of what seems to have been the forum and
basilica lay within the excavated area and, moreover, that it was possible to
say a good deal about what happened to them in the late-Roman and early
medieval periods. This we shall document in the next chapter.
These discoveries brought into focus the whole question of the fate of
public areas like the forum in late-antique times. ‘Broad-sweep’ syntheses
take wildly opposing views. For some scholars, fora disappeared early,
symbolising the decline of the ancietit city in the fifth and sixth centuries,2
while others, especially historians of early medieval Italy, suggest that they
often survived to become the market places of the Middle Ages.3 In the east
Mediterranean, on the other hand, scholars propose that they gradually gave
way to urban foci based on shop-lined streets: linear markets which were in
time to develop into Islamic souks.4
To scrutinize fully all these matters would take a very much longer work
than can be attempted here. But there is also a dearth of detailed evidence,

2 Hodges and Whitehouse 1983, 52; Christie 1989, 266. See also K. Randsborg, The
first millennium in Europe and the Mediterranean (Cambridge 1991), 82f.; and now
the essays in Rich 1992.
Wickharn 1981, 83; rather more circumspectly Ward-Perkins 1984, 183f.; Ward-
Perkins 1988.
Kennedy 1985. Liebeschuetz 1972, Liebeschuetz and Kennedy 1988; Claude 1969,
58f.; and recently, on Sardis and its context, Crawford 1990, 107f.

4 5

- -i-il
 r
2

THE CHER CHEL EXCAVATIONS

THE HISTORY OF THE TOWN

The site of Cherchei has received a good deal of attention in recent years,
most notably from Philippe Leveau. His study of the ancient city of Jo!
Caesarea, and its hinterland, has in many senses become something of a
classic. Based mainly upon field survey, combined with a detailed
examination of the existing documentation and some excavation, it provided
a remarkable picture of the evolution of one of the most important cities of
Roman north Africa.5
Cherchel’s known history is a rich one. It seems to have originated as a
Punic trading centre, occupied from at least as early as the sixth century BC
and, as we shall see, quite probably before. It is referred to as such by
Pseudo-Scylax,6 and the quality of its harbour is emphasised by Strabo
(17.3.12). Later it was to become a royal capital, although exactly when is
much disputed. The Numidian king, Micipsa (148—118 BC), is a strong
possibility, as there is an inscription which alludes to a funerary cult to him,
found in the western necropolis.7 It must also have been a royal capital of
Bocchus II, for he bequeathed it and his kingdom to Octavian, on his death
in 33 BC. Then in 25 BC Augustus gave lol, the kingdom of Mauretania and
parts of Gaetulia to Juba II. Juba was the son of the Numidian king. Juba I,
who had died in a suicide pact after siding with Scipio, and sharing in his
defeat at Thapsus in 46 BC. The young Juba was taken to Italy where he was
brought up and given Roman citizenship, probably by Octavian, whose close
friend he became. Jo! and the kingdom was his reward.

Leveau 1984. Amongst many interesting reviews, cf. Fentress 1984, and D. J.
Mattingly. Libyan Studies 18 (1987), 151f.
6 c• MOller (ed), Geographici Graeci Minores (Paris 1882), section III, 90.
Cf. P-A. Février, Linscription funeraire de Micipsa, Revue d’assyriologie et
d’archeologie orientale 45 (1951), 139—50; H. P. Roschinski, Die Mikiwsari-Inschrift
aus Clierchel, in Horn and ROger 1979. 111—16; Leveau 1984, 11—13.

1I
 Towns in Late Antiquity: Jol Caesarea and Its Context The Cherchel Excavations

Juba was to rename his capital as Caesarea, in honour of the emperor, and can be specifically attributed to him. After his death, the kingdom was
he then set about converting it into a splendid classical city a ‘yule annexed and Caesarea as Colonia Claudia Caesarea was made the
vitrine’ of Augustan Rome, but set in Mauretania, as Leveau has put it. capital of the new province of Mauretania Caesariensjs, administered by a
There were many parallels for this, although entirely in the east procurator.
Mediterranean. The most striking is the foundation by Herod the Great of Its ports, one military and one commercial, were soon developed, so that
another city, also called Caesarea, on the shores of his kingdom. It is it became a base for the African fleet, second in importance only to
described for us in some detail by Josephus (Jewish Wars 1.2 1.6—8), who Carthage. Analysis of the inscriptions indicate that most of the names are
leaves no doubt that it was intended to glorify the emperor himself in the not African — about a quarter are Greek suggesting a city that was
magnificence and luxucy of the monuments.8 becoming ever more cosmopolitan. A large proportion of the first-century
lol Caesarea, to judge from the archaeological remains, was certainly also inscriptions are of people with servile or freed status; this is a feature that is
splendidly endowed. The theatre (Figure 3) was definitely built by Juba,9 unusual in north Africa, and may reflect lol Caesarea’s exceptional status
and a number of other public works have been attributed to him, including and importance.b
the extension of the street grid, the curious elliptical amphitheatre,’° temples 1 Public monuments built in early-mid imperial times include three sets of
to Augustus and to isis and the aqueduct.11 Traces of what may be the royal baths, a circus and very possibly a stadium. There was also a substantial city
palace have also been recorded.12 Although, in reality, there is little concrete wall, 4.46 km in extent. It enclosed some 370 hectares, running prominently
dating evidence for these structures, Pensabene’s study of the capitals leaves along the crest of the ridge to east, south and west of the modern town
no doubt that there is an abundance of Augustan work, much of it done by (Figure 2). Although only the flatter ground at the foot of the ridge was built
craftsmen from Rome (cf. Figure 13).13 Moreover, there is a fine assemblage up, amounting to about one third of the defended area, the scale of these
of sculpture from the town (Juba was a noted collector), and his reputation fortifications is impressive. Many have sought to date them to the period of
as a man of scholarship was considerable: Pliny the Elder and Plutarch were Juba’s reign, although they could just as easily be placed in the period when
amongst those who used his works. the city became a colonia. 16
mba was married first to Cleopatra Selene, the daughter of Antony and Whilst precise chronologies for these monuments are lacking, it is clear
Cleopatra there is Egyptian sculpture in his collection 14 and then to
—
that, as elsewhere in north Africa, tlç Severan epoch wasa time of especial
Glaphyra, a daughter of a king of Cappadocia. His standing was therefore prosperity at Caesarea. A porta triumphalis was constructed for the circus; a
considerable. When he died after a long reign, in AD 23, his son Ptolemy temple of Aesculapius was adorned with a pronaos, a pool with trees and
became king, and remained so until murdered at the behest of Caligula in marble statues and a colonnaded portico; and the city’s decurions returned a
AD 40. Whatever the motives for this (they are much discussed: Cassius Dio road leading up to a major gate ‘to a condition both worthy of, and
(59.25) thought that Caligula was jealous of Ptolemy’s affluence), little is appropriate to, their splendid native city’. There is a real sense of pride in
known of what was achieved at Caesarea during his reign, and no buildings these evocative words.17
An important figure behind this work was the eminent governor of
8 Cf. D. Braund, Rome and the friendly king (London 1984), 107f Mauretania Caesariensis in AD 201, Publius Aelius Peregrinus Rogatus. He
9 G. Ch. Picard, La date du théâtre de Cherchel. RAI (1976). 386—97; Golvin and is associated epigraphically with two of these building projects, and was
Leveau 1979.
10 honoured by the erection of at least two statues of him at Cherchel. He is
Golvin and Leveau 1979. J. C. Golvin. L’amphithéhtre romaine (Paris 1988),
catalogue 81 (1121’.).
11 Temples: Leveau 1984, 63; aqueduct: Leveau and Paillet 1976; Leveau 1984, 61 (on 15 Leveau 1984, 199—200; Fentress 1984, 488.
further consideration of the chronology). 16 Duval 1946; Leveau 1984, 26f.
12 Leveau 1984, 42f. 17 The circus: Leveau 1984, 39f.; J. H. Humphrey, Roman circuses (London 1986),
13 Pensabene 1982. 308—10. The whole monument may date to the Severan period, as may the West
14 Horn and ROger 1979, taf. 75. Cf. K. Fittschen, Juba II und seine Residenz Baths. Temple of Aesculapius: CIL VIII, 9320. Road: CIL VIII, 20982. See also
Jol/Caesarea (Cherchel), in Horn and Ruger 1979, 227—42. note 29 below.

8 9
 Towns in Late Antiquity: Jol Caesarea and Its Context The Cherchel Excavations

I ‘S

Figure 2 Plan of lol Caesarea. a

known to have promoted work elsewhere in the province, particularly the

C.)
construction of frontier works, including a road and forts. His activities at C)

Caesarea, of which we shall have more to say later, may well have been
directed at further embellishing a city which was widely recognised as
having a long and distinguished past.18
Leveau’s survey around Cherchel (Figure 4) disclosed a relatively densely
settled landscape of villas and other forms of site (‘agglomerations’). For the
most part, the villas lack many of the characteristics of those in other parts
of the western empire notably a pars urbana. The aristocrats who owied
—

them preferred to reside in the city leaving bailiffs to run the estate.

18 H.-G. Pflaum, Les carrières procuratoriennes équestres II (1960), no. 233; N.

Benseddik, Les troupes auxiliaires de l’armée romaine en Maurétanie Césarienne
sous Ic Haut-Empire (Alger 1982), 108; P. Salama, Libyca 1 (1953), 243f., and now
N. Benseddik in Afr. Rom. 9 (1992), 425—38. I owe much to Dr Susan Walker for
this discussion, which is based on her report in the excavation volume.

10 11
 Towns in Late Antiquiti’: Jot Caesarea and Its Context The Cherchel Excavations

1 Particularly conspicuous on the sites are the remains of olive presses,

suggests a peak in rural settlement in the second and third centuries. It is
suggesting that this was the main focus of the economy. This was certainly based on finds from only 39 of his 237 sites, but he shows that 80% of the
the case at the one ‘villa’ to have been excavated, that at Nador, although African Red Slip ware (out of a sample of 1063 sherds) belong to this
wine production is a possibility and there is a building identified as a period. Whether this is a true reflection of activity, as opposed to
horreurn. 19 The most striking feature of the complex at Nador, however, is fluctuations of supply, is more debatable, as recent studies are beginning to
its monumental façade, facing the Cherchel-Tipasa road. Built in the early make clear.20 But it is interesting that Severan prosperity in the town is
fourth century, an inscription proudly proclaims that this is the praediurn apparently matched in the countryside, and recalls the themes of two of
(farm, estate) of M Cincius Hiliarianus, Jiamen Augusti perpetuus. The Cherchel’s most famous mosaics, the Labours of the Field (Figure 5) and the
contrast between this imposing frontage, and the modest farming buildings Vendange. Dunbabin dates these to the period AD 200—2 10, and they
behind, could hardly be more striking. certainly evoke in a most compelling manner the theme of rural
The Nador ‘villa’ seems to have been founded in Ptolemy’s reign, c. AD abundance.2’ It recalls the words of Tertullian, a north African writing in
40, and was still in occupation in the sixth century (although with a possible this period, who describes ho ‘marshes have been blotted out by fair
gap in the third century). Leveau’s surface collection, on the other hand, estates, forests have been conquered by ploughed fields, wild beasts have
been put to flight by flocks of sheep, the sands are sown, the rocks planted
there are now more towns than there used to be huts everywhere there
...

are houses, people, organised government.22

It is fascinating that these literary and artistic images of African
prosperity and productivity in the early third century should find a reflection
in the artefact assemblages from sites elsewhere in the Mediterranean,
especially parts of Italy. This is the message to be gleaned from the
extraordinarily widespread distribution of African Red Slip ware, produced
in Tunisia, which is even found on shepherd encampments high in the
central Italian Apennines.23 But it is also implicit in statistical analyses of
table ware, amphorae and even some classes of cooking vessels from sites
like Ostia and Luni, where African products dominate the assemblages from
the late first century onwards.24 Even at a seemingly remote rural site like
the Mola di Monte Gelato, 25 km to the north of Rome, there are large

20 E. Fentress and P. Perkins. Afr. Ram. 5 (1988), 205—14; Fentress 1984.

21 Dunbabin 1978, 114—5 and 254.
22 Tertullian, De Anima 30.3. See Cl. Lepelley, in L’Afrique dons I’Occident Romain
(Collections de l’Ecole francaise de Rome 134 (1990), 403f.).
23 In Prof. G. Barker’s survey of Mouse. There are however areas like northern Italy
where there seems to have been much more limited availability.
24 The literature is now vast, and ever more complex. S. J. Keay, reviewing Amphores
0 5 20km roinaines et histoire. Dix ans de recherches (Collection de 1’ Ecole francaise de Rome
— 114, Rome 1989) in IRA 5(1992), 353—60, is a very useful overview of an important
Figure 4 Roman sites in the vicinity of Cherchel. based on Leveau’s field survey. book. A. Tchernia, Le yin de I’Italie romaine (Rome 1986) remains fundamental, as
do many’ essay’s in A. Giardina (ed.), Le inerci, gli insediamenti (Società romana e
impero tardoanrico III, Rome and Ban 1986): cf. particularly Carandini (3—19),
19 Anselmino et al. 1989. Anselmino on Ostia (45—81), and Panella (431—59). See also the following note.

12 13

TI
 Towns in Late Antiquity: lol Caesarea and Its Context The Cherchel Excavations

quantities of African pottery, despite the existence of a flourishing local

industry. Probably significantly, a comparison of two refuse dumps, one of
early second century date, the other of the Severan period, shows that the
African component rose by a factor of ten or more in the later deposit.
Whilst pottery can only be a very rough index of trading patterns, figures of
this sort must surely reflect a general trend. Whether in grain, olive oil or
other commodities, business was good for the African-based merchant of the
second to third centuries, and the archaeological record of cities like
Mauretanian Caesarea shows it.25
We should at this point pose the interesting, if problematic, question of
the size of Caesareas population in its heyday. We have already noted that
only the flatter ground, towards and beside the sea, was heavily built up,
amounting to about 150 ha. This renders figures like the 100,000 advanced
by Gsell rather high, and Leveau follows Lézines suggestion of c. 140—150
inhabitants per hectare, giving a total of about 20_22,000.26 Duncan-Jones
argues on epigraphic evidence that urban densities could vary between
towns by a factor of five; at Caesarea, given the space available, high-rise
buildings like those at Ostia can hardly have been prolific, if they ever
existed, and a relative low density is to be expected. Yet calculations for a
modest civitas and its territory, namely Siagu (Tunisia), give a figure of 5—
9,000, while Oea (Tripoli) may have had as many as 28,000 or more
inhabitants. This makes the total for Caesarea sound on the low side, given

25 Grain: G. E. Rickman, The corn supply of ancient Rome (Oxford 1980). Oil: H.
Camps-Fabrer, L’olivier et l’huile dons lAfrique romaine (Algiers 1953). D. J.
Mattingly has recently produced many important studies, including: Oil for export? A
comparison of Libyan, Spanish and Tunisian olive oil production in the Roman
Empire. JRA 1 (1988), 33—56; The olive boom. Oil surpluses, wealth and power in
Roman Tripolitania. Libyan Studies 19 (1988), 21—42; Olive cultivation and the
Albertini Tablets. Afr. Rom. 6 (1989), 403—15. I am grateful to Dr Roger Wilson for
reminding me of the tombstone of an oil merchant, P. Livius Pilerus, from Cherchel
(Camps-Fabrer, op. cit., 68). Note that for M. Gelato ; is the fine wares that show
an increase in the later second century, not amphorae, where local forms
predominated (in contrast to Rome and Ostia). They were clearly largely self-
sufficient in wine and oil (P. Arthur, forthcoming in the M. Gelato final report). See
also Carandini, in Amnphores romaines et histoire (1989; see previous note), 505—21,
arguing for greater wine production in central Italy, directed towards the Rome
Figure 5 Cherchei: the Labours of the Fields mosaic, c. AD 200-2 10. In Cherchel market, in the later second century.
Museum. 26 S. Gsell, Cherchel, antique Iol-Caesarea (Alger 1952), 222; Leveau 1984, 79; A.
Lézine, Ant. Afr. 3 (1969), 69—82.

14 15
 Towns in Late Antiquity: lol Caesarea and Its Context The Cherchel Excavations

its size and status; even so, it seems unlikely that its population exceeded houses, some thirty of which are known.29 Their mosaic floors are
30,000.27 particularly splendid, such as famous pavements like the Vintage from the
lol Caesarea produced one emperor from amongst its citizens, namely ‘Tennis Club’ mansion; datable to the late fourth or early fifth centuries AD,
Marcus Opellius Macrinus. A praetorian prefect under Caracalla, in AD 217 they attest a prosperous and wealthy community, where urban life was
he arranged for the ruler’s assassination and was proclaimed Augustus by his sustained with considerable vigour.3°
troops. He was however to survive in office for little more than a year, Christianity was also well established. St Marciana was martyred here in
before himself being put to death. There was hardly time, therefore, to 303, quite probably in the amphitheatre; Christian cemeteries are known;
embellish his native city as Severus had done for Lepcis Magna if, — and St Augustine himself came to preach in 418, because of the conflict that
indeed, he ever contemplated it. No inscriptions that mention him have been had broken out between two bishops in 411, the celebrated Donatist
found at Cherchel, and Dio (Hist 79, 11) is at pains to stress the modesty of Emeritus and the Catholic Deuterius.31 What have apparently not so far been
his origins and, indeed, that he had a pierced ear, in accordance with the found are the great churches that characterise so many north African sites,
custom followed by most of the Moors. Macrinus was to attract no praise not least at nearby Tipasa (Figure 6). The scale of the cathedral there is
from the ancient writers, and his reign was short and brutal.28 enormous: it has seven naves and measures 58 x 42 m; beside it is a large
Little is known of the history of lol Caesarea in the later Roman period baptistery and bishop’s palace. Three cemetery churches, again of
before the events of the early 370s. These are recounted for us by, inter considerable size, are also known, surrounded by innumerable sarcophagi.32
alios, Ammianus Marcellinus, who describes at length (29.5) the revolt led Similar buildings certainly existed at Caesarea, for Augustine is said to have
by Firmus. A son of an African prince, Nubel, he is said to have been spoken in the principal church, and its political importance and abundant
‘unable to endure the greed and arrogance of the military officials and had wealth will have ensured their existence. Indeed, there were rural churches
aroused the Moorish tribes’ (Amm. Marc. 30.7.10). Caesarea was quickly in its territoriuni, most notably at the coastal site of Trois-Ilots (Figure 7),
occupied, and was only retaken three years later, in AD 373, by Count some six km to the north-east of Cherchel. Here the church measured 26 x
Theodosius, Valentinian’s general. Ammianus describes it as follows 16 m, a substantial building in such a context. Three other probably
(29.5.17): examples have been identified by Leveau, pointing to a not inconsiderable
penetration of the countryside by the Christian faith in later Roman times.33
On entering the city, and finding it almost wholly burned down from
widespread fires, and the paving-stones white with mould, he decided to station
In 429, Cherchel fell to the Vandals. Interpreting the consequences of this
the first and second legions there for a time, with orders to clear away the invasion has become one of the minefields of scholarly contention. No less
heaps of ashes and keep guard there, to prevent the place from being devastated
by a renewed attack of the barbarians. 29 Rich houses: Leveau 1982. On the revolt of Firmus as it concerns Cherchel,
It is difficult to know how literally to take this picture of such total Lepelley 1981, 513f. See also Salama 1988, who relates the burial of three coin
hoards of this period from Cherchel to the sack of the city. Lepelley (1992, 59)
devastation (although confirmed by Orosius: 7.33.5). But a letter of reminds us of inscriptions on statues from the West Baths in late-style lettering,
Symmachus (1, 64) supporting the request for a moratorium on tax demands recording that they were ‘transferred from sordid places’, i.e. from temples,
by the Bishop of Caesarea, Clement, and dated to AD 380, is usually seen as presumably those closed by decree at the end of the fourth century. It is a further
implying that Ammianus’ account was exaggerated: the aristocracy was reminder of the respect for ancient tradition at Caesarea.
30 Dunbabin 1978, 255.
temporarily in a state of financial embarrassment, but not ruined. This is
31 Leveau 1984, 215; Gui et al. 1992, 16f., Février (1986) provides a sensitive analysis
certainly the impression gained from study of their magnificent town
of the impact of Christianity upon Mauretania. For the cemeteries, Ph. Leveau in
Ant. Afr. 11(1977), 254—6; id., Ant. Afr. 12 (1978), 94—5; id., Ant. Afr. 19 (1983),
27 R. P. Duncan-Jones, The economy of the Roman emnpire (Cambridge 1974), 259f. 85—173.
28 Dio. Hist. 78.11. Cf. A. R. Birley, The African emperor. Septimius Severus (2nd ed. 32 Lancel 1971, 40f.; Gui et al. 1992, 21f. The basilica may also have been converted
London 1988), 191f., and 148f. for the changes at Severus home town of Lepcis into a church, although the evidence is not full, and this is far from certain.
Magna during his reign. Leveau 1984, 407-8. Trois-Ilots: see now Gui et al. 1992, 18-19.

16 17
Towns in Late Antiquity: lol Caesarea and Its Context The Cherchel Excavations

an authority than the Cambridge Histoty of Africa (1978, 483) can proclaim
/ that in the Vandal period ‘urban life as it had been understood even in the
I fourth century had come to an end’, a view somewhat supported by a recent
I survey of the later history, as perceived through the archaeological evidence,
of Carthage: the abandonment of the theatre, odeon and Antonine baths;
neglect of the streets and city defences; and the tipping of rubbish into once
grand city houses. Yet the evidence of the pottery from Carthage is for
C.)
C
considerable trade, especially with the east Mediterranean,34 while
documents like the Albertini Tablets, found south of Tebessa, point to a
literate and prosperous rural population, still operating under the terms of
Roman law.3
‘ We shall return to this considerable enigma in later pages, for it is a
subject on which the new work at Cherchel can be said to shed a certain
amount of light. That said, there is a dearth of literacy or epigraphic
evidence relating to this period of the city’s history. Yet the early sixth-
century grammarian Priscian, who was born and presumably educated at
Caesarea, points to an urban environment where traditional values still held
sway. He took a chair at Constantinople, and his surviving works show that
he was an accomplished and widely read scholar, in both Latin and Greek.
His writings were much consulted in the Middle Ages, and he would seem
to provide a priori evidence that Caesarea was still a place of
consequence.36
Under the Vandals large parts of Mauretania fell into Moorish hands.
Ceuta (Septem) remained the last outpost in Tingitana; otherwise, all the
land to the west of Caesarea was out of Vandal control. Thus, when
Justinian ordered the reconquest of Africa, the four duces were instructed to
create bases at Lepcis Magna, Capsa or Thelepte, Cirta and Caesarea.
Procopius (4.5.5; 20.32) tells us that a company of infantry was sent to

Hurst and Roskams 1984, 44; Humphrey 1980; Fulford and Peacock 1984; C.
Panella, Le anfore di Cartagine, Opus 2.2 (1983), 53—74. See also note 57.
C. Courtois, L. Leschi, C. Perrat, C. Saumagne, Tablettes Albertini (Paris 1952).
See also D. J. Mattingly in Afr. Rom. 6 (1989), 33f. The literature that attempts to
assess the impact of the Vandals is enormous, much stemming from C. Courtois, Les
Vandales et l’Afrique (Paris 1955); a minimalist view is currently in vogue. See
Lepelley 1979, 1981; C. Bourgeois, Les Vandales, le vandalisme et lAfrique, Ant.
Air. 16 (1980), 213—28; and now Thhbert and Biget 1990, who deny the impact both
of the Vandals and even of the Arab invasions; and Lepelley 1992.
36 Jones 1973, 991: J. R. Martindale, Prosopographv of the later Roman Empire 2
(Cambridge 1980), 905; R. A. Kaster, Guardians of language: the grammarian amid
society in late antiquity (Berkeley 1988), 346—8.

18 19
 Towns in Late Antiquity: Jol Caesarea and Its Context The Cherchel Excavations

this was far beyond their resources.37 It would seem most likely that they
Trois-Ilots constructed a much more modest circuit as happened in so many other north
African towns.38 No trace, however, is known of it, nor of any other
Byzantine monuments in the town.
irch Exactly when Byzantine rule of Caesarea came to an end is similarly
unclear, at any rate from the literary sources. The battle of Sufetula in 647
between the Arab forces and those of Byzantine Africa effectively marked
b the cessation of eastern control, although a Roman military presence
remained until the early eighth century. But, given that the primary focus of
Byzantine rule lay much further east, it is hard to imagine that there was
much interest in the fortunes of Caesarea after the mid sixth century.
Certainly the field survey of Leveau, and the Italian excavations at the villa
0 500 1000 of Nador, would seem to point to a total breakdown of the rural economic
rn
—-

system by that time, and a not dissimilar picture has emerged from the
recent excavations within the town. It was the sixth century, rather than the
fifth, which saw the demise of the classical city of Caesarea.
We shall have more to say of this below, but must first offer a few words
church on Cherchel’s subsequent history. Interestingly, linguistic studies suggest
that the Arabic of the Cherchel rcgion includes elements thought to be
I __) characteristic of the first Arabic conquerors.39 However, El Idrisi, writing at
the beginning of the twelfth century, describes the town as a thinly
populated place, a conclusion that the archaeological evidence largely
corroborates.40 Pensabene has identified capitals with 10th and 11th century
parallels, and the mihrab of a mosque, probably of this date, has been
located; but there is little more.41 Not until the arrival of the Andalusians, in
the late 15th century, did the town’s fortunes begin to rise. It soon fell into
the hands of the Turks of Algiers, who in 1 5 1 8 constructed a fort. A
considerable number of elegant courtyard houses, built mainly in pisé,
survive from this period, creating a pleasant ambience that is the old town of
today.
Figure 7 Plan of the site at Trois-Ilots, to the north-east of Cherchel. The French occupied Cherchel in May 1840, and there began a somewhat
dismal saga of investigation, and destruction, of Cherchel’s past. Little was
Caesarea in the winter of 533—534 by Belisarius, under the command of
John. By the time of Solomon’s campaigns early in the 540s, however, Duval 1946, 98 and 168.
38 Pringle 1981, 119-20..
Caesarea had become totally isolated by the Moors, and could only be
reached by sea. Although it has been claimed that Caesarea’s enormous Benseddick, Ferdi and Leveau 1983, 14.
40 El Idrisi, Description de I’Afrique et de I’Espagne (Leiden 1968), 103. Note,
circuit of walls was restored under the Byzantines, the reality must be that
however, that Thébert and Giget (1990) maintain that the Arab invasion of the
seventh century caused relatively little disruption, especially in the coastal cities.
41 Peusabene 1982, 67—8.

20 21
 Towns in Late Antiquity: Jol Caesarea and Its Context The Cherchel Excavations

published, and even less in any sort of detail. Not until Leveau (who laboratory analysis. On the whole these arrangements were satisfactory,
describes this sorry story) began his own work was the situation to change, although deficient when compared with those of other north African projects
with his monograph on the town and its territory as by far and away the most like the UNESCO excavations at Carthage. They were, however, the terms
significant achievement in the archaeological study of Cherchel.42 under which we had to operate43
Meanwhile, modern Cherchel continues to encroach upon the ancient The site that was chosen for investigation was a block of land, some 30 m
monuments in an alarming way. This indeed was the context of the Anglo- square, in the heart of Cherchel, 11 5 m to the north of the theatre (Figure 8).
Algerian excavations of 1977—81, and it is to these that we must now turn, Work on building a cinema, initiated early in 1977, brought to light
in an attempt to assess their significance both for our understaiiding of the considerable evidence for older structures, as well as fine statues: a copy of
development of Cherchel, and for the history of urbanisation in the Cephisodotus’ Eirene (Figure 9), and a torso with a short tunic possibly
Mediterranean as a whole. representing Ganymede. Subsequent rescue work then revealed that the
corner of a colonnaded piazza lay within the area, as well as a large building
THE FORUM EXCAVATIONS, 1977-8 1 with a late-Roman mosaic; they seemed prime candidates for the forum and
basilica.
This programme of excavation was the outcome of an agreement between Here, then, was a key site, with public buildings and a deep stratigraphy,
Algeria and the United Kingdom, following a visit by Professor S S Frere in and it was an obvious target for the project. In the event, it more than
1975. It envisaged annual visits by a small British team, the participants in fulfilled its potential, yielding a sequence from at least as early as 500 BC
an excavation where new techniques of investigation could be introduced. down into the present century. As such, this opening into Cherchel’s past,
The provision of training was thus at the heart of the accord which, so far, however small, is an important index of the archaeological development of a
has resulted in two projects; the one at Cherchel, directed by Dr Nacera town with, as we have seen, a distinguished and fascinating history. Few
Benseddik and myself; and another at Lambaesis, directed also by Dr excavations in north Africa have produced quite so lengthy a chronological
Benseddik, together with S P Roskams of the University of York. sequence upon a single site, especially one with public buildings sealed
Publication is through the supplementary volumes of the Bulletin within the stratigraphy. It encourages doubtless overbold generalisations,
d’Archeologie Algérienne; however, these are not easy to get hold of and, in especially concerning the later Roman and early medieval history of
the case of Cherchel, the manuscript was prepared in the early l980s and Cherchel; but that, in a sense, is the challenge of this small book.
submitted in 1985. There are many respects, therefore. in which the broader
discussion in very out of date. There is thus advantage in providing a précis
of the results, and a more sustained attempt to set them in context than was (i The early phases: Punic and Juban Jol (Phases 1 and 2)
possible before.
It should be emphasised from the outset that the work at Cherchel was
carried out with a very small team of UK archaeologists four in 1977, The oldest deposits on the site comprise a distinctive brown, clay-textured,
three in 1978—80 and just two of us in 1981. All processing of the objects colluvial hillwash, formed as a result of soil-creep down the slopes to the
was done during the season of excavation, and the finds drawn at the same south. A scatter of sherds, distributed through the upper part of the deposit,
time. The schedule was thus extremely demanding, but was helped by visits indicate activity in the vicinity, but there was no clear division between
from a few specialists, to examine some classes of pottery, and the marble sterile and sherd-rich layers. Gravelly bands suggest some form of water
and architectural remains. Other specialists were able to produce reports on action, and could mark wetter climatic episodes, while fragments of burnt
the basis of drawings and photographs, and we were given a permit to bring branches are likely to attest scrub clearance.
back to England the animal bones, environmental samples and material for

42 Cf. Leveau 1984, 1-6. I particularly thank my two principal assistants Denise Dresner and Sarah Philpot.

23
 Towns in Late Antiquity. lot Caesarea and Its Context The Cherchel Excavations

05 50 125
— m
Figure 9 Cherchel: statue of Eirene, found in rescue work at the forum site in April 1977,
Figure 8 Cherchel: plan showing the relationship between the forum site and the theatre. and now in Cherchel museum.

24 25
 Towns in Late Antiquity: Jol Caesarea and Its Context The Cherchel Excavations

A nearly complete Punic lamp of a type that is at least as early as the

sixth century BC, came from the top part of this deposit and, together with
other finds, amongst them a similar lamp, point to utilisation of the area far
back into Punic times. Given that the site lies some 400 m from the port —

the natural focus for the original settlement nucleus it is reasonable to

suppose that it lay on the periphery of what was probably already a
considerable town.44 However, no structures were identified and a thin
scatter of material, largely occurring residually in later contexts, points to
sporadic activity in the fifth and fourth centuries BC. Agricultural activity
might easily account for the presence of these sherds on the site.
until towards 200 BC is there any evidencefor building (Figures 11,
12). The principal feature of this period manifested itself as a pronounced B
depression, some five metres in width, which ran in a slight curve across the
southern part of the site. Investigation in one small trench (which, for safety
reasons, could not be taken deeper than 2 m) showed it to be filled with dark ci

silts, typical of those formed in a large ditch. There was an abundance of

refuse in this fill, including a little material of the third century BC, and ci
C)

large quantities of finds of the second to first centuries BC. The northern lip en
ci
-D
of this ditch-like feature was also examined at one point. Despite the C)

presence of a later footing, this work disclosed a large block of dressed

masonry, set in red clay, in a position that indicated that it belonged to a C)

wall bordering the ditch. We might therefore suggest that, together, they
formed part of a major defensive system around the southern side of the
town of lol. . 0

It is unfortunate that circumstances (particularly the overlay of later QfD

buildings) did not permit a ‘classic defences section for, if we have

interpreted our results correctly, the identification of loIs southern perimeter . .

of c. 200 BC is a matter of some importance. Not only does it show that lol, cc
e— C)
like many other north African urban sites,45 was erecting town walls in the
mid Hellenistic period, but it also demonstrates the quite substantial size e
o-C’,
0

some 8—10 ha of the settlement at this time. It may well already have
.2 t
>5)

I-’
C)
44 G. Vuillemot had previously identified Punic material in excavations at the port; he 00
4)0
believed that the oldest finds dated back at least as early as the fifth century BC
(Leveau 1984, 12). 4)
.0
45 Cf. for a useful survey, C. M. Daniels, Town defences in Roman Africa, in (eds) J. I)
0
Maloney and B. Hobley, Roman urban defences in the west (London 1983), 5—19. I
am grateful to Ph. Leveau for pointing out a reference to what may be another part of
the defences, observed by J. C. Glénat in 1921; cf. Leveau 1984, 13, referring to A. C?

Ballu, Rapport sur les travaux defouilles ...(Alger 1922), 11.

26 27
 Towns in Late .4ntiquity: Jol Caesarea and Its Context The Cherchel Ewavations

become a Mauretanian capital, and adds weight to the suggestion, supra,

that it was the royal centre of the Numidian king, Micipsa (148—118 BC).
Behind the putative town defences lay a building, limited elements of
which were exposed (Figure 11). Constructed in two phases, some of the
foundations were very solid, comprising large undressed stones set in red
clay (like the town vall), and with a width of over a metre. Too little was
CHERCHEL: PUNIC uncovered to gain much idea of the plan, beyond an impression that it may
have consisted of rooms placed around a court. But there seems no reason to
1
doubt that they belong to one or more houses, which the pottery suggests
were occupied in the second to first centuries BC.

1 /
-

UA
The deposits associated with this period yielded a not inconsiderable
number of imports. These were Italian lamps and amphorae (and also a
variety of Punic forms); black glaze vessels; and Nurnidian lead coins and
issues of Ebusus (Ibiza). Vine seeds from the fill of the town ditch may
indicate local wine production (and Bocchus II favoured the vine leaf as an
emblem upon his coins), while the stock economy appears to have been
based particularly upon the exploitation of cattle and pig, and to a much
-I lesser extent sheep/goat. Even though Leveaus field survey brought to light
-
virtually nothing of this period, lol must surely have possessed an extensive
/Th
agricultural base in the surrounding countryside.
/

Towards the end of the first century BC, the top of the town ditch was
V -3 filled in. It had largely silted up by this time, and can no longer have served
any real purpose. To the north, the Punic’ building was demolished, and
some footings completely robbed out. En its place was constructed a
rectangular structure, with at least three rooms, one of them sunken. They
were built of well dressed masonry, with opus signinuni floors.
Significantly, the orientation was changed from that of the older building, to
take on that of the street grid of the eastern part of lol. Since the new
building was very probably constructed during Jubas reign, it is perhaps a
In’ pointer to new regulations in town planning in the ‘yule vitrine’.
d.,cfl

town
to° 2n,
The finds suggest that the building was the focus of considerable
industrial and commercial activity. Bronze-working is attested by crucibles
,eo
and slag; iron-forging by the bottom of a smithing hearth; and there were a
number of pottery wasters, both of arnphorae and of cups and bowls. One
type of cup finds very close parallels in central Italy, and an immigrant
0 15 ro potter is a not unlikely possibility.46 Moreover, the animal bones support the

46 A cup resembling the Form IV of M. T. Marabini Moevs, The Ronan thin-walled

Figure 11 Plan of the Punic phase. 2nd-Ist centuries BC.
pottervfronCosa. MAAR xxxii (1973).

28 29
 Towns in Late Antiquitv: Jo! Caesarea and Jts Context The Cherchel Excavations

<
tiS’
-
a

0
0)

0—
0

r -7

Figure 13 Cherchel. forum site: capital of Augustan date, probably carved by a craftsman
from Rome.

idea of on-site butchery of cattle and pigs (sheep and goat remained of little
/ importance), while the discovery of four pastry-cook moulds might point to
the existence of a nearby bakery.
The cumulative effect of this evidence is to suggest that the site had now
been incorporated into a commercial quarter of the town. Given that, as we

H’ shall see, it is likely to have lain close to the Juban forum, this seems
entirely appropriate. Coupled with evidence of imports
Arretine and thin-walled vessels, a little glass
amphorae,
—

it is a modest but
—

30 31

.aoAL
 Towns in Late Antiquity: lol Caesarea and Its Context The Cherchel Excavations

satisfactory indication of the way that Jo! Caesarea under Juba became
strongly integrated into the broader classical world, as was so overtly his
intention.

CHERCHEL 1977-1981
(iii) Jol Caesarea in Imperial and Vandal times (Phases 3 and 4)

The annexation of Mauretania in AD 40 finds no reflection in the

archaeology of the present site. Indeed, in both structural and stratigraphical
terms, it is extremely difficult to demonstrate what was happening there in
the later first and second centuries AD. There are finds, especially coins and
pottery, of the period; but few were in context, and it is not impossible that
much was swept away by the grand building work that took place around
AD 200.
The remains of that period divide into two parts (Figure 14). One is the
north-west corner of a very large building, occupying the eastern part of the
site, and resembling a basilica. The other is a paved piazza, surrounded by a
colonnaded portico and ambulatory. It was, and remains, our feeling that this
was an extension of the Juban forum, which has for long been regarded as
occupying an area extending northwards from the theatre.47 Indeed, at first
sight, it was taken to be the Juban forum itself, since there were fine capitals
and sculpture of that period (Figures 9, 13). However, when the makeup
levels of the surrounding ambulatory were excavated, it soon became clear
from the pottery and coins that the forum can hardly have been constructed
before the second half of the second century, and that a late second or early
third century date would best suit the group of finds.
About 130 sq m of the paving was exposed, showing it to be rather untidy
and haphazard work. There were the bases of two podia, presumably for
statues, and steps up to the ambulatory. The building that we term a basilica
lay on the eastern side of the ambulatory, and it was fortunate that the north
west corner of the building just lay within the excavated area, so that it is at
least possible to determine the limits in this area. Its foundations were at
least two m deep, and were offset in such a manner that at the base they a recent pit
cement surface
were 2.30 in in width. The masonry was in limestone and the interior was 0 S
metres early medieval
wall
provided with a marble veneer. In the north wall the beginnings of an apse 0 10
Islamic wall

were encountered in the far north-eastern corner of the excavation which, sondage

Cf. Leveaus careful discussion of the evidence: 1984, 40f.; also E. Albertini, BAC Figure 14 General plan of the forum site.
1924, xxxiv.

32 33
 Towns in Late Antiquity: lol Caesarea and Its Context The Chereliel Excavations

were it a single apse, would give an approximate total width for the building
in the order of 20—25 m; it may be, however, that some more complex
apsidal arrangement was chosen, giving a still wider building. In the west
wall, there was an elaborate entrance onto the forum ambulatory. Its full
extent was not determined, but five portals were uncovered, separated by CHERCHEL c.AD430
double columns. The columns (some of which were certainly of granite) PHASE 4 basilica I

rested on separate blocks, which were morticed together, and between each
pair was a further block with a central groove perhaps for a screen. In the church
area of the entrance were a number of limestone voussoirs, probably from
the arches over the entrance, and various other architectural elements, ?wooden screen

particularly capitals, all in limestone. The styles imitate those of Severan Inarthex
door
nave
orders in marble, which is satisfactory support for the proposed date of the
forum 48 d———I 7wooden
Screen
The floor of the basilica had been replaced in late antiquity, and no 4
1 I
internal columns were found; but the grandeur of the structure is not in doors

doubt, and it is intriguing to wonder how far it stretched to the south; there
is space for it to have been over 100 in in length. More puzzling is the north ambulatory
question of what may have occupied the space to the north of the forum. We
found no traces of floors or walls of this period in that area, and thus have to
assume that, if not destroyed by later works, this was open ground at this east
time; this seems surprising, but it is difficult to read the archaeological
evidence in any other way.
That these buildings probably belong to the Severan period is of some
HLJJ stalls
ambulatory

9
interest. We have already referred to the epigraphic evidence for the

j
forum
activities of the governor of Mauretania. Aelius Peregrinus, and the city’s
decurions. in restoring Caesarea ‘to a condition both worthy of, and
appropriate to, their splendid native city’ at this time. Although the cheaper
alternative of limestone, rather than marble, was chosen for the basilica, it . mosaIc
would not be inappropriate were Aelius Peregrinus to have encouraged the ?podia
5 work. This might also provide a context for the conversion of the theatre
into a modest amphitheatre, especially as there is at least one mason’s mark .
on colunm bases common to both sites. The reuse of the finely crafted
capitals (and perhaps columns) would be entirely suitable to a project aimed
at underlining lol Caesarea’s ancient and distinguished past.
No structure or stratigraphical deposits of the third or fourth centuries _fift J
were found, but a wealth of evidence points to a major phase of
0 5 lb
refurbishment and new building in the early part of the fifth century. m

48 Figure 15 Plan of the forum site c, AD 430.

As suggested by Dr Susan Walker in the final excavation report.

34 35
 r

Towns in Late Antiquity: Jo! Caesarea and Its Context The Che,-chel Excavations

probably c. AD 420—30. It is thus tempting to link this with the seizure of

\ the city by the Vandals in 429, and to suppose that it was Caesarea’s new —.

masters that ordered the work. Alternatively, it could have been the outcome
of the havoc wrought by Firmus and his rebels in 371, which on the
evidence of Ammianus (29.5.17), resulted in catastrophic fires. As we
commented earlier, Ammianus is commonly supposed to have been
exaggerating; but it was instructive to find that, prior to the new work, there
::
was robbing of earlier walls, the trenches beiiig backfilled with very large
quantities of heavily burnt material. It remains at least possible, therefore,
that this is tangible evidence of Firmus’ sack, especially as the forum area
may well have been a prime target.
The work of rebuilding extended to almost every part of the site (Figure
1 5). In the forum, the ambulatory was resurfaced, and at least one of the
columns was provided with a new foundation, packed with marble
chippings. In addition, there was some patching of the flagging, including
the insertion of a few new paving stones: some comprised reused pieces of
white marble (Figure 17), which on the uiiderside had tooled lines of a type
that is identical to those on a marble threshold in the church (to be described
below). It is one of the many strands of evidence which point to a
contemporaneous and planned building operation.
A further innovation within the forum was the construction of a series of
wooden stalls around its perimeter (Figures 16, 17, 25). Three were
identified on the north side and there were two conjoined examples on the
east. Their position was marked by post-holes and grooves, and they seem to
have been tied into the ambulatory colonnade. The positions of the entrances
were well marked, aiid near one was a graffito, chiselled into a paving stone,
which could identify one of the stall—owners. It is not unrealistic to
reconstruct these stalls as structures with wooden frames, covered over with Figure 16 Cherchel, forum site. Grooves and postholes of one of the late-Roman wooden
suitably decorative awnings, and selling a wide variety of market goods; one stalls (Fig. 17, no. 2), looking south-east, together with the east wall line of stall
might envisage a somewhat souk-like appearance. no. 1.
The principal dating evidence for the stalls consisted of a number of coins
that had fallen down the cracks between the paving stones in the area of the that is strikingly borne out by the Cherchel finds.49 Of the total site
stalls themselves. None of these coins post-dated the fourth century and the assemblage only three Vandal coins were found, and there were a mere 14
initial conclusion was that these represent their main period of use. issues of Sth—6th century type and just six diagnostic coins of the period AD
However, it is iiow well recognised that the bulk of the small change of the 430—500. By contrast fourth-century coinage was extraordinarily prolific
fifth century comprised bronze coinage of the previous century, something with 117 issues, as well as 63 of general 4th—Sth century type.

R. Reece, OAford Journal of Archaeology 3, no. 2 (1984), 205, and in Hurst and
Roskams 1984, 171f. The pattern is well exemplified at Sabratha: cf. Kenrick 1986,
Fig. 110.

36 37
 f
Towns in Late Antiquity: lol Caesarea and Its Context The Cherchel Escavations

CHERCHEL: FORUM Layer Date from pottery 4th century coins Other coins
(in stratigraphical
order)

188, 189 early-late 6th C. AD 425—55 (1)

193 6th C. 2 5th—6th (1)
205 early 6th C. 7 5th—6th (2)
211.219 5th C. 45 AD 410—23 (1)
210 early 5th C. 5

This pattern of a dominance of fourth-century coinage in fifth-century

layers was well demonstrated stratigraphically. One trench immediately
adjacent to the north wall of the church (T: Figure 18) proved to be very rich
in finds, especially of datable pottery and coins. The figures are shown in
the above table.
The prevalence of fourth-century coinage in fifth-century contexts does
not, of course, rule out a fourth-century origin for the stalls. The coins could
be taken literally as evidence for use in that century. However, the
relationship of the stalls with the church is such as to suggest that both were
in use at the same time, and a later date of use is probably to be preferred.
As with the forum, the basilica also seems to have been refurbished at this
1 time. Some of the marble veneer was replaced, including the use of
Proconnesian from the Sea of Marmora in north-west Asia Minor, and
cipollino, from Carystos. Euboea; both were quarries that remained in use
into Byzantine times. In addition, a new mosaic was laid (Figure 19). Made
with black and white tesserae, arranged in a geometric design, the overall
effect is crude and sloppy. The origins of the motifs lie with Italian black
and white mosaics of the second century, and were ones that were
commonly copied by African mosaicists, although usually in polychrome;
here is evidently a late and debased version. singularly lacking in
imagination.50 Unfortunately, there was no good dating evidence, other than
the style of the mosaic itself; but it is reasonable to assume that it was laid at
the same time as part of the general programme of renovation. If so, it is a
far cry from the mosaics of the late fourth or early fifth centuries that so
embellished Caesarea’s mansions, and must imply either a lack of craftsmen
lumps (454) — or funds.
o i 5 10
T.
The one completely nevv building on the site was a long rectangular
m structure, tucked into the corner between the north end of the forum and the
Figure 17 Cherchel. detailed plan of the forum.
50 The mosaic is discussed in detail by Dr Roger Wilson in the final excavation report.

38 39
 Towns in Late Antiquity: Jo! Caesarea and Its Context The Cherchel Excavations

B Trench T: section S2
basilica

B ground level
c
plan
post- medieval
wall

DI .ro -:

p1 . — basilica
wall

church
wall

interpretation

Figure 19 Cherchel, forum site. The black-and-white mosaic, of fifth-century date, in the
north-west corner of the basilica.

basilica (Figures 20, 21). It was almost entirely built out of reused masonry,
including pieces with mouldings, and incorporated a good deal of tile to fill
in the gaps. There were three rooms, the most westerly of which extended
0 3 beyond the edge of the excavation. The eastern room was raised some
m
twenty cm above the floor level of the other rooms, and was apsidal in plan,
the apse being made of concrete. Virtually all of the floor had disappeared.
Figure 18 Cherchel, forum site. Section through the late-Roman deposits to the north of the The central room was floored with opus signinurn, and access was by means
church, together with a chronological interpretation.
of three contiguous doors, set in the south wall and thus providing access

40 41
 Towns in Late Antiquity: Iol Caesarea and Its Context Tue Cherchel Excavations

E
fl .

I E

Figure 20 Cherchel, forum site. A general view of the probable church, looking east
towards the basilica.

from the forum. The interior was heavily disturbed by later pits, including a
large intrusion in the centre of the room; however, post-holes and grooves in
the opus signinuin left no doubt that there had been a wooden screen inside
the entrance, and a further screen within the room (Figure 29). Furthermore, 0)
‘—4
there was a well-defined path, marked by heavy wear of the opus signinum,
which led through the doors and towards the centre of the room.
As was suggested earlier, the building is most plausibly identified as a
small church. Although it lacks side aisles (quite possibly because of

42 43
 Towns in Late Antiquity: lot Caesarea and Its Context The Cherchel Excavations

limitations upon space), the apsidal form of the building, together with its L)

tripartite division into a narthex, nave and raised chance!, all support this
explanation. Furthermore, there is a graffito on one of the threshold blocks,
which may include a chi-rho (Figure 22), and the wooden internal fittings 0
recall Eusebius’ description of a church where the altar was enclosed within
a wooden lattice-work fence (Ecc. Hist. x, 4, 37—45). If, as seems likely, the
U
altar was more-or-less centrally placed within the nave, then the groove
could well mark the position of just such a fence, just as the passage of feet
to and from the altar would account for the worn path between it and the
doors.
Other explanations are, of course, possible a schola has been
suggested, for example but, on balance, an ecclesiastical function
remains the most likely interpretation, especially given the extent of the
building work. Certainly, the finds show that this part of the city continued
to be relatively vibrant for another century or so, as we will see.51
First, however, it is necessary to say a little more about the overall dating
evidence, if only because the building work falls within the period of a
critical historical interface. The data are shown in the table on p.47 (with
residual material excluded).
The coins provide, a firm terminus post quem of AD 410 for the building
work, whilst the pottery would seem to imply a still later date. When the
shoddy nature of the basilica mosaic and of the church walls are taken into
account, then a Vandalic context becomes still more plausible. Certainly, if
the Vandals were not responsible for the work, they nevertheless ensured
that the old civic centre continued to function as such. Although refuse was
piled up behind the church, the buildings and piazza remained in use.
Moreover, the finds indicate that the city maintained its long-distance
trading networks. There are amphorae (Figure 23) from the east
Mediterranean (of Carthage Late-Roman types 1, 3, 4 and 5) and, as at
Carthage, handmade vessels were imported.52 African Red Slip ware occurs
in surprising abundance, given the built-up nature of the site, with a wide
range of forms, spanning the fifth and early sixth centuries. There are also
some fine late-Roman glass vessels, some with good parallels from the

51 Cf. the reservations concerning tile identification as a church ill Gui et al. 1992, 16f.
I have not seen tile more detailed critique by N. Duval, in Revue des Etudes 00
Augustiniennes 34 (1988), 247—66. The probable Christianisation’ of tile forum is
discussed further in chapter three.
52 Tile amphorae are reported upon by Dr John Riley in tile final excavation report. For
handmade vessels at Carthage, cf. Fuiford and Peacock 1984, 156f.

44 45

----
Towns in Late Antiquity: lot Caesarea and Its Context The Cherchel Excavations

Context African Red Slip ware Coins

(a) Robber trench, before dish, late 4th—early 5th C. 4thC.: 1

building of church Hayes 26, late 4th—mid 5th C.
(b) Makeup deposits for church 2 Hayes 91, late 4th—late 5th C. 4th C.: 7
2 Hayes 76, c 425—75
2late 4th C. +
(c) Construction level in Trench stamped sherd, 5th C. 4th C.: 21
T, outside church (layers 1 Hayes 67, 360—470 4th—Sth C.: 11
210, 211) 1 Hayes 88/81, late 4th—mid 5th 388—402: 1
C. 395—402: 1
4late4thC. + 410—21: 1
(d) Late ambulatory makeup dish, probably 5th C. 4thC.: 1
4th—Sth C.: 1

Note: The pottery dates are those suggested by Dr John Riley, after study of the material in
198

I eastern end of the Mediterranean. All the archaeological pointers are,

therefore, to a flourishing economy, matching the evidence from the farm at
Nador, between Cherchel and Tipasa, where much fifth-century African Red
Slip ware (although not amphorae) was found.53
On the other hand, the study of the animal bones does suggest some
interesting changes in the stock economy at this time. During the fifth
century, cattle became markedly less common than in early Imperial times,
as did the number of pigs. Sheep and particularly goat show, however, a
significant increase, coupled with some alterations in overall strategy: some
animals were killed as yearlings for meat, while other sheep were kept for
wool and many of the goats for dairy produce.55 Whatever the reasons for
this, it is at least a hint that there was some reorientation of the local
economy, although the broader implications must remain obscure without
further work.

o 53 Ct. D. Manacorda in Anselmino et al. 1989, 209t.

- -

For the African Red Slip ware forms, cf. I. W. Hayes, Late Ro,nan pottery (London
1972); Supplement (1980). Also Atlante delle forme cera,niche I (Supplement,
Enciclopedia dellArte Antica, Rome 1981).
Reported on by Dr Gill ian Clark in the final excavation monograph.

46 47
 Towns in Late Antiquity: lol Caesarea and Its Context The Cherchel Excavations

Context African Red Slip ware Coins and over the north part all the masonry was removed. To the south, however,
(a) Over north ambulatory Vandal (490—540)
there was a jumble of masonry (Figure 24), which had not been cleared
(b) Destruction deposit 1 Hayes 87, second half 5th C. 4th C.: 6 away. Some columns of the ambulatory also remained, and at least one had
within church 2 Hayes 91, late 4th—late 5th C. 4th—Sth C.: 17 fallen when some 20 cm of hillwash had accumulated over the forum
3 late 5th—early 6th C. 5th—6th C.: 4 paving; its capital was still in place when it fell (Figure 25) and it vividly
?6thC.: I resembles a photograph of 1932, showing a column and capital lying on
Vandal: 1 (490—540)
(c) Deposits over the forum 1 Hayes 87, late 5th—early 6th C.
about 50 cm of fill over the paving of the Basilica Ulpia in Rome.56 Two
4th C.: 7
I Hayes 103, early—late 6th C. 4th—Sth C.: 1 statues, found in rescue work in April 1977, also came from this general
1 Hayes 104, late 5th—early 6th 5th—6th C.: 3 context (Figure 9).
C. Vandal: 1 (490—540) The dating evidence is set out in the table on p. 48. It is comparatively
(d) Trench T, surface with 1 Hayes 64, early—mid 5th C. 4th C.: 1 full, and would seem to suggest that the complex went into desuetude
much burnt pottery 1 Hayes 12/102, mid 5th—early
(201) 6th C.
around the beginning of the sixth century, perhaps c AD 500—20. The
1 Hayes 84—86, mid—late 5th C. surface in trench T with burnt material (probably from the destruction of the
1 Hayes 99, early 6th C. church) is unlikely to have formed before about 530, and might well have
1 Hayes 100—102, 6th C. come into being when the buildings were being destroyed. The silts that
1 Hayes 104, 530—80.
overlay the surface yielded a sherd of African Red Slip of c 5 70—600 from
(e) Trench T, layers on top I Hayes 62—63, 5th C. 388—410: 1
of surface 201 (188, 1 Hayes 86, 5th C.
towards the top of the deposit and is the latest datable ‘Roman’ artefact from
4th—Sth C.: 2
193) 2 Hayes 12/102, mid 5th—early 5th C.: 1 the site.
6th C. Sth—6th C.: 1 It may be, therefore, that we have a record of the decay of the civic centre
1 Hayes 91B, late 5th—early 6th in late-Vandal times, perhaps accelerated by the burning down of the
C. church. This would be consistent with the picture for, as an example,
1 Hayes 104, 530—80
I Hayes 104B, 570—600
Carthage, where many public and private buildings were apparently in sharp
decline by this time.57 Much the same is true of Tipasa, if Baradez’s
evocative picture is to be believed.58 Equally, the data could be interpreted
Note: The pottery dates are those suggested by Dr John Riley, after study of the material i to suggest that the demise of the forum is to be associated with the arrival of
1981. For Trench T, see Figure 18. the Byzantine army in the winter of 533—4. That the Byzantine occupation
brought about significant changes within many of the north African cities is
(iii) The Byzantine period and its aftermath (Phase 5) not in doubt. Fortresses were constructed over some civic centres such as
that at Dougga in Tunisia, for instance, and spoliation of older monuments
was authorised and widespread. The pulling down of the buildings at
In the first half of the sixth century, there was in archaeological terms a Caesarea might well reflect just such a situation.59
dramatic event on the site. The church was accidentally or deliberately set
on fire, and burnt beams, together with large structural nails, fell onto the
56 R. Meneghini, in Arch. Med. 16 (1989), 545; I am grateful to him for pointing this
floor. These were in turn buried by a thick layer of tiles, leaving no doubt
out to me.
that the roof had collapsed inwards. This was evidently followed (exactly
Hurst and Roskams 1984, 42—7. The monumental centre may have remained in
when is not clear) by the systematic demolition of the walls down to floor existence into Byzantine times: Deneauve 1990. See also Gros 1985, 147 and, for
level. more optimistic general appraisals, Clover 1982 and Christol 1992.
Similarly, the basilica was pulled down, mostly to floor level or one 58 j• Baradez, Libyca 9., i (1961), 198-9.
course above it. In the process, some of the mosaic floor was stripped away, Poinssot 1958, 40; see generally Pringle 1981, 119; for the legalisation of spoliation:
Cod. Theod. 15.1.36 (as early as AD 397) and 10.3. See further chapter three.

48 49
Towns in Late Antiquity: Jo! Caesarea and Its Context The Cherchel Excavations

U
a

I-.
U

0
U

C’s
U

[
C>

C3
C

It

50 51
 ___________

Towns in Late Antiquity: lol Caesarea and Its Context The Cherchel Etcavations

Whatever the view on this matter may be, perhaps the real surprise was
that the forum and basilica (if that is what they were in this late period) did
survive in use quite so long. Some functions of the old Roman city were
preserved into late antiquity and a forum bedecked with stalls recalls the
paintings from the block owned by Julia Felix at Pompeii, apparently
showing scenes of daily life in the forum there. Here the stalls look to be no
more than trestle tables; but there is a real bustle of activity, splendidly
evoked by Roger Lings description: ‘The background is formed by a portico,
with garlands slung from column to column. In the foreground we see
equestrian statues (one of which is being sketched by an artist), bystanders
reading a public notice, street-stalls with tradesmen selling fabrics, shoes,
conversations, an old beggar with a dog receiving alms from a lady with a
maid, a horse-drawn cart, and a pack animal being loaded. There is even a
mobile soup kitchen, not to mention an open-air school, where a miscreant
pupil is being painfully chastised.’60 Although archaeological evidence for
stalls in fora is very hard to find, especially ones so permanently marked as
at Cherchel, there is nevertheless some hint of continuity of tradition. It is a
subject that will be taken up again in later pages; meanwhile. we must
present the evidence for the remarkable archaeological sequel to Caesareas
once grand civic centre.
Soon after their demolition, brown clayey hillwash began to encroach
over what remained of the buildings. To the south, apart from the blocks in
the basilica, this was largely free of any inclusions save for some abraded
pottery (Figure 32), while to the north there were many tiny fragments of
stone and pottery, but little else. The formation of this layer of wash (which
in places was to reach a depth of over 1.5 m before further building-work, of
16th century date, took place) was however briefly interrupted over much of
the site by the construction of a series of interesting and unusual structures.
unexcavatad later pit
In all. four buildings could be identified in plan (Figures 26, 27) as well as
recent pit
the pisé wall of a fifth, visible only in section. Strangely, there was virtually excacated pit
no stratigraphical build-up that could be associated with these structures Pc xt Pd yOu

merely a very thin layer within the hillwash in their immediate vicinity
and hardly any contemporary refuse.
The buildings were of rather different construction, which could well
imply some difference in date. Two (Figure 26, nos. 1, 2) were made largely 15n,

of reused masonry, including vertical orthostats, with an infilling of cobble

and pisé (Figure 28). Other sections of walling were the more conventional

60 Figure 26 Plan of the forum site in the later sixth and seventh centuries.
R. Ling. Roman painting (Cambridge 1991). 163—4.

52 53
 Towns in Late Antiquity: lol Caesarea and Its Context The Cherehel Excavations

CHERCHEL
EARLY MEDIEVAL P1-IASE5

:rus:3c -

house

L -l 7-
L.
/

L - .-

0 5 15
Figure 28 Cherchel, forum site. South-east corner of early medieval house 1, showing
cobble and pisé filling between re-used blocks. In the foreground the basilica
Figure 27 Interpretation of Figure 26. mosaic.

block-upon-block, the variation perhaps reflecting a scarcity of materials; few traces of a trodden surface; the existing Roman floors may well have
even a fallen column and capital was pressed into service as a footing served.
(Figure 25). The plans of both buildings were highly irregular, with a The two other buildings (Figure 26, nos. 3, 4) were both made of wood.
marked tendency towards the trapezoidal, and one (no. 2) possessed at least Lying over the north ambulatory was the north-west corner of a structure
two rooms (the other, no. 1, being too badly cut away by later intrusions to with a shallow foundation trench, probably for a sill-beam. The seating of at
determine the internal arrangements). No floors were preserved, save for a least one buttress post was apparent, but the full extent of the structure never
became manifest. Nevertheless, it was clearly rectangular in plan, and much
more regularly laid out than the buildings in stone. The other building,

54 55

_%—.-
--
 Towns in Late Antiquity: lol Caesarea and Its Context The Cherchel Excavations

which was put within the ruins of the church, was of quite different
construction. Rather than using a continuous construction trench, its walls
were made up of free-standing posts, set in post-holes and anchored with a
pad of mortar placed beneath the post (Figure 29). This building was also
rectangular in plan, with what might have been a verandah or lean-to along
the east side. Immediately beside it was a large pit, entered by means of
steps. \Vithin the pit were the remains of baked clay kilns, representing two
main periods of construction (Figure 30). The older was rectangular in plan,
with a central longitudinal spine, made up of two vertical conjoined bars.
Originally, it may have been open at one end, but it was subsequently
reduced in size and completely enclosed. This was eventually replaced by a
smaller kiln, realigned by some 90 degrees from its predecessor, but
otherwise with many similarities of design.
Although these kilns most closely resemble domestic cooking ovens,
there was an absence of associated refuse. There were, however, four
substantially complete items of briquetage, as well as a further three
fragments (Figure 3 1). Laboratory examination showed that one piece had a
potassium glaze, derived from wood ash. The briquetage was not, therefore,
used for salt manufacture, and the presence in one corner of the pit of what
seemed to be part of an unfired clay vessel suggests that they may have been
in fact simple pottery kilns. It has to be said that no wasters were found, and
that the kilns seem more adapted in form for domestic purposes than for the
manufacture of pottery; but they are generally hard to parallel.
The kilns were subsequently backfilled with a large quantity of debris,
including much pottery of the first centuries BC and AD. There was also
some building material, amongst which was wall-plaster of Augustan type,
and burnt wood. This produced radiocarbon dates of 40 BC ± 170 (BM
1909R), and 110 AD ± 120 (BM-1910 R), and it may be that old Roman Figure 29 Cherchel, forum site. Postholes, grooves and pits (one containing early medieval
kilns), cut into the floor of the probable church, looking south-east. The division
houses were being plundered for firewood. into light and dark areas of the floor is caused by differential drying between two
On archaeological grounds, it is tempting to infer two main periods of seasons of excavation. Compare with Figures 21, 26 and 27.
construction, with the two wooden buildings, nos. 3 and 4, together with the
kilns, as the earlier. In support of this is the fact that they are well
constructed, while the use of mortar to bed in the posts is an indication that Luni, in north-west Italy. These buildings are dated to about AD 600, when
the art of its manufacture had not been forgotten. Moreover, there is a Luni was still in Byzantine control, and seem to reflect a widespread
consistent use of right angles, and the kilns were built to a high standard. It reversion to the use of wood for lower-class housing in late-antique Italy.6’
is iHdeed probable that the later sixth-century sherds of African Red Slip The stone-built houses at Cherchel are rather different. There is the
ware found in the top deposits in Trench T (Figure 18: see table above) scantest attention to right angles, no use of mortar only pisé and cobbles
relate to the use of these buildings. They particularly recall the two post- as filling material and the general effect is, to say the least, haphazard.
built structures which rested upon some 30 cm of fill covering the forum at 61 Ward-Perkins 1981; Brogiolo 1987. See further chapter three.

56 57
 Towns in Late Antiquity. lol Caesarea and Its Context The Cherchel Excavations

0 15

Figure 30 Cherchel, forum site. The early medieval kilns (sixth-seventh centuries) in a pit
cut through the church floor Figure 31 Cherchel, forum site. Briquetage from the early medieval kilns.

58 59
 Towns in Late Antiquity: Jol Caesarea and Its Context The Cherchel Excavations

outside house 2. which appeared to belong to the early medieval period,

gave a result of 1260 AD ± 110 (BM-2130R), which made no stratigraphical
sense: but the other, using burnt material from the floor of house 2,
produced a figure of 640 AD + 160 (BM-2129R), which seems entirely
appropriate. The seventh century, then, would seem to mark the period when
all long distance trade had broken down, and a medieval community had
come into being. We might term it ‘sub—Roman, whilst recognising that the
scant archaeological evidence is incapable of indicating the ethnicity of the
inhabitants.
These early medieval structures do not seem to stand in isolation at
Cherchel. Excavation on the west side of the town in 1981 brought to light
not dissimilar stone walls, resting upon Roman street surfaces. Subsequent
work, likewise not yet published, has apparently revealed many more traces
of such buildings. Although undated, they may also reflect the emergence of
a sub-Roman village in this quarter of the town. Indeed, if a relatively
abundant quantity of cereal seeds, together with some barley, from the
forum excavations is any guide, there may well have been cultivated fields
within the walls of Lol Caesarea.

(iv) The Islamic period

Given the dearth of occupation deposits associated with the early medieval
buildings, it is hard to see them as being long in use. Once abandoned, they
were again gradually enveloped in clayey brown colluvium. washed down
the hill slopes to the south. The site then lay empty until at least the
thirteenth century. the date of the oldest identifiable medieval sherds.
Figure 32 Cherchel. forum site. Stratigraphv over the fifth-century surface of the basilica.
Islamic building levels of the fourteenth century onwards, overlie sterile hillwash. Buildings eventually began to encroach over the site, especially houses with
internal courts of typical Islamic type. Interestingly, there was a further
period of pottery manufacture, producing distinctive brightly painted table
vessels, mainly during the 18th and 19th centuries. However, these are
Whereas the wooden buildings do in some senses hark back to the Roman
aspects of the excavation which, although of importance, are of less
world, the stone structures emphatically do not. In combination, however,
relevance to the theme of this monograph. We shall therefore now turn to
they are an extraordinarily evocative symbol of the demise of the city, and
the broader context of the results from Cherchel.
the emergence of a new, perhaps village-like, community, with a very
impoverished material culture. Indeed, it was because there was an absence
of diagnostic artefacts that radiocarbon dating was employed to try to gain
an idea of the chronology. One determination, of charcoal from a surface

60 61
T

Civic CEIvTRES IN LATE ANTIQUITY: THE WIDER

CONTEXT

The Cherchel excavations provide considerable insight into the construction,

restoration and demise of part of the civic centre of a major north African
town. Although oniy a very small window into the city’s ancient landscape,
it encourages speculation on the nature of urban change in late antiquity,
especially concerning public spaces like the town’s forum. Here, if
anywhere, it would seem, one ought to have a barometer which measures the
waxing and waning of urban fortunes. In fact, the mailer is somewhat more
complicated than that, as we shall try to demonstrate below.
A forum served a number of purposes. Its original role was an outdoor
place of assembly, where market transactions might be undertaken; but it
soon developed into an enclosed piazza, with some of the city’s most
monumental and prestigious buildings. The urban market focused on the
inercatus or macchum (strictly, a meat market) which Février characterises
as follows: ‘the Roman market at least in the examples we know well in
—

African towns — is an enclosed space attached to a courtyard, or simply a

basilica not necessarily linked to a square’.62 As Shaw has stressed, these
urban markets should be clearly distinguished from periodic rural fairs
(nundinae), for which there is also good epigraphic evidence from Africa.
Although in Italy nundinae were in some instances absorbed into urban
markets, in Africa there were in effect two market systems, one based on

62 Février 1977, 330. See also Février 1982 and his: Permanence et heritages de
lantiquité dans Ia topographie des villes de occident pendant Ic haut moyen age.
Settimane del Centro di Studio sull’alto inedioevo 21(1974), 41—138; and C. de Ruyt,
Macchum. Marché alimentaire des romains (Louvain 1983). On fora, cf. the most
useful essay by R. Martin, Agora et forum. MEFRA 84 (1972), 903—33; and now for
a splendid series of plans. and exhaustive bibliographies, for sites empire-wide, Balty
1991; also Los foros romanos de lasprovincias occidentales (Madrid 1987).

63
 Toiins in Late Antiquity: Jol Caesarea and Its Context Civic Centres In Late Anriquiti’: The Wider Context

towns and vici and the other wholly rural; between them was what Shaw
terms a demonstrated disj unction’ 63
Mitch has been written on the fate of the fora, especially in terms of Italy, HENCHIR EL FAOUAR
where so many Roman towns survived in some form or another into BELALIS MA/OR
medieval and, very often, modern times. It is frequently suggested that
Roman fora may in a number of instances have evolved into medieval
market squares.64 The evidence is mainly literary. In Milan, for example,
there is a reference of 879 which refers to a public forum, not far from the
mint, which seems to have been close to the site of the Roman forum. By Islamic
952 there were permanent market booths. Similarly, open market squares
overlie the forum at Florence, Brescia, Verona and Spoleto. and Verona is
described in about 800 as having a ‘broad spacious forum paved with stone,
where in each of the four corners stands a large arch’.65 Likewise, an
inscription of about the same period from Terracina records how ‘the forum
was cleaned up in the time of George, consul and dux; there seems every
reason to suppose that it was the Roman forum that was being referred to.66

SOME NORTH AFRICAN EV1DLNCE

We shall return to the subject of Italy below, but we must first scrutinise the
north African evidence more closely. The situation is of course different, church
since relatively few Roman cities developed into major Islamic centres.67

late
63 Shaw 1981, 72. See also R. Macmullen, Market days in the Roman empire. Phoentv
24 (1970), 333—41; E. Gabba, Mercati e fiere nelI’Italia romana. Studi classici e
orientali 24 (1975), 141—63.
64 Wickham 1981, 83; Ward-Perkins 1984, 184—5.
65 Ward-Perkins 1984. 185.
66 A. Guillou. MEFRMA 83 (1971), 149—58.
67 For continuities and discontinuities in Libya, see G. R. D. King, in Libyan Studies 20 baths
(1989), 193—208. At Carthage, there is little evidence for Islamic occupation before
the tenth century (G. Vitelli, Islamic Carthage: an assessment of the evidence
(Dossier Cedac 2, 1981)); but a number of Algerian sites demonstrate ‘continuity’:
cf. eg. R. Lequément, Fouilles a I’atnphithédtre de Tébessa (2e Supplement au
Bulletin d’ArchCologie AlgCrienne, nd.); Pringle 1981, 424 (n. 101). At SCOt the
baths were in use until the beginning of the 7th century and probably demolished
_•5p
towards the end of that century; the first Islamic building dates to the tenth century, Q_ 1°m
so there is here a clear gap (A. Mohainedi, A. Benmansour, A Amamra, E..

Fentress, Fouilles de Sétif 1977—1984 (Se Supplement au Bulletin d’Archeologie

Alghrienne, Alger 1991), 29, 93. Note that Thébert and Biget (1990) strongly favour Figure 33 Plan ofBelalis Major (Tunisia). After Mahjoubi.
‘continuity’, especially in coastal towns.

64 65
 Toivns in Late Antiquity. lot Caesai’ea and Jts Context
T Civic Centres In Late Antiquiti’: The Wider Context

Moreover, there are few places where the question of the late history of the were maintained well into the next century, or later, and many nearby
forum has been addressed in detail. Even so, there are useful pointers, as private houses remained in occupation. The late-Roman Christian centre
some excellent recent discussions make clear.68 seems, as at Belalis Maior, to have lain well away from the old civic centre,
The evidence is fullest for sites in Tunisia. There has been important some 230 in to the west. The town was apparently still the seat of a
work on the Byrsa Hill in Carthage (briefly considered in the final section of bishopric in 646, and the remains of what has been claimed as a Byzantine
this chapter), where an impressive monumental centre has been investigated. fort have been identified at the western limits of the area settled in Roman
But particularly illuminating are the results from excavations at the small times. The impression gained is once more of a significant shift in the focus
town of Belalis Major (Henchir el-Faouar), which lies in hilly country of civic life, away from the traditional centre, towards once peripheral parts
overlooking a valley to the north of Beja, some 90 km to the west of Tunis of the town.
(Figure 33)69 The forum and surrounding buildings have been extensively The same seems also to be true of Thuburbo Maius and Sbeitla
investigated. Epigraphic evidence shows that the curia and a portico were (Sufetula).7’ It is clear that at both sites public areas, including the forum,
restored in Constantinian times, as there is another inscription, of 3 83—392, were given over to the production of olive oil. Even streets were titilised in
recording further work. What this may have been is unclear, although there this way (Figure 34). Sbeitla nevertheless possessed some splendid
are repairs to the forum paving, which might belong to this period. Around churches, and there is considerable evidence, epigraphic, archaeological and
the end of the fourth century, there were however substantial modifications literary, to show its importance in the Byzantine period.72 It may be to this
to the rooms around the forum. Partition walls made of rubble and ‘mortier period that the walling-up of the forum precinct belongs (Figure 35).
de terre’ were constructed, often over fourth-century mosaics, and the forum Although unquestionably late-antique in style, it is however a relatively
paving itself began to be covered with a layer of earth some 15—20 cm thick. puny construction, and cannot really be regarded as a Byzantine fort; on the
Through this deposit were cut a number of inhumation burials. While not other hand, it is yet another indication that the old Roman forum no longer
closely dated, they are a decisive illustration that the forum no longer had a functioned in the same way as before.73
part to play in urban life. Although the forum baths were operating until Other i2th African fora were converted into Byzantine forts, notably at
about AD 500. and there are late-Roman structures in the vicinity of the Dougga, Madauros (M’daourouch. in eastern Algeria) and at Tubernuc (AIn
forum, a new Christian focus, some 200 m to the north, came into being. Tebournok. in north-eastern Tunisia). That at Dougga had already been
There is a clear implication that the forum, with all its traditional pagan deprived of its statues, which were taken to the theatre, and had been
associations, was deliberately put out-of-use. covered over with a layer of rubble. It has been suggested that this was due
This pattern seems to be widely repeated in north Africa, and has been to subsideiice of the north portico, and this may indeed be the case; but,
especially vell studied at Bulla Regia.7° An important Numidian settlement, given the history of other fora at Tunisian sites, it is by’ no means impossible
also in northern Tunisia. it is situated in rich farmland, and prospered first as that it had already lapsed from use. For Madauros and Tubernuc there is less
a iflunicipiuhil and then as a colonia. The forum was elaborately laid out, evidence, but in neither instance can it be demonstrated that the forum was
with a monumental entrance, and around it a Capitolium, a temple of Apollo functioning as such when the Byzaiitine forts were built.74
and the basilica. It seems that here too the forum went out-of-use towards
the end of the fourth century, even though public monuments like the baths
71 Thuburbo Maius: Maurin 1967. Sbeitla: Duval 1964; for the rich late-Roman
68 Thébert (1983) has produced a particularly good attempt at synthesis of the north churches, Duval 1971, 1982; and Duval and Baratte 1973. See now also N. Duva in
African evidence, and I gladly ackHowledge my debt to it. There is also much L’Afrique dans I’Occident Roinain (Collection de l’Ecole francaise de Rome 134,
valuable matrial in Pringle (1981) and Lepelley (1979, 1981 and now 1992). Baltys 1990), 495—535.
main interest is in the curia, and lie touches relatively little on their late history; but, 72 Pringle 198t, 284-5 and references; Duval 1977; Gui et at. 1992, 156-8,
like Pringle’s and Lepelley’s, his catalogue is invaluable (Balty 1991). Duval 1982, 622.
69 Mahjoubi 1978; Id. Afr. Rain. 1 (1984), 63—72. For Carthage, see note 140, infra. Dougga: Poinssot 1958, 40; Madauros: Balty 1991, 79f; Gui et at. 1992, 327f.;
70 Beschaouch et at. 1983. Pringle 1981, 214f. Tubernuc: Pringle (1981, 249f.) is an accessible source,

66 67
Towns in Late Antiquity: lol Caesarea and Its Context Civic Centres In Late .lntiquTh:: The Wider Context

Much the same seems to be true of the forum at Rougga (Bararus), to the
south—east of El Djem. An extremely meticulous excavation showed that the
drains were becoming clogged in the third century, and that there was little
fourth or fifth-century material. Later, probably after c. AD 525, structures
were built against the south entrance and there is nearby what may have
been a small Byzantine fortlet. Although this was destroyed in or after 647,
marked by the burial of a hoard of 268 solidi, it became a cemetery,
associated with poor housing. It is a remarkably well documented
sequence
Further east, in what is modern-day Libya, the Italian excavations at
Cyrene have shown with great clarity how buildings began to encroach over
the agora from the later third century.76 This is also the case for the adjacent
forum. A rather different story emerges. though, from a sensitive study of
the old British excavations at Sabratha, in Tripolitania. Sabratha, as with
many towns in this part of north Africa, was apparently severely affected by
an earthquake of AD 365. Subsequently, the forum paving was repaired.
largely with marble and inscriptions from other damaged buildings,
especially old pagan temples, which were evidently not restored. Both the
: basilica and curia were rebuilt, also largely with reused materials and,
probably in the late fourth or early fifth century, the basilica was converted
into a church (which was later refurbished b Justinian).77 It is uncertain
whether the forum colonnades were reconstructed and, indeed, how long the
forum itself remained in use. What is clear is that it was eventually turned
into a cemetery, although when is disputed. Ward-Perkins favoured a fifth
century date but, on the basis of sepulchral inscriptions, Duval regards it as
a necropolis of the Byzantine period (as, indeed, the pottery implies).78
There is a case, therefore, for supposing that at Sabratha the forum may have

Guéry 1981.
76 Bacchielli 1981. Fig. 129.
Kenrick 1986. The basilica of Lepcis Magna was also converted into a church in the
6th century, and barracks were built over the forum. However, the site was largely
deserted in the Vandal period, and had been buried beneath silts and sand dunes. Cf.
R. G. Goodchild and J. B. Ward-Perkins, PBSR 21 (1953), 42—73; R. Bianchi
Bandinelli, E. Vergara Caffarelli and G. Caputo, Leptis Magna (Rome 1964); M. F.
Squarciapino, Leptis Magna (BasIc 1966): J. B. Ward-Perkins, Princeton
En cyclopaedia of Classical Studies (1976), 499f.
78 Kenrick 1986, 84f. For the Byzantine date of the inscriptions associated with the
a
burials in the forum, N. Duval. Recherches archéologiques HaIdra I. Inscriptions
Girétiennes (Rome 1975), 484.

68 69
Towns in Late Antiquity: lot Caesarea and Its Context Civic Centres In Late Antiquity: The Wider Context

> continued in use into the fifth century, if only as a piazza adjoining a major
ecclesiastical building.
>.‘
a In Algeria comparable evidence is harder to seek. A detailed study has
been published of the splendid forum, 76 x 43 iii in size, at Hippo Regius.
> but it is sadly lacking in chronological information. The forum was flanked
On its long axis by shops, and’there was a temple at the southern end. Statue
bases were discovered on the paving, as well as the famous inscription of C.
0
Paccius Africanus. proconsul and municipal patron of the city, dating to AD
77—78. At the north end, resting on the paving, is a roughly constructed
masonry structure; it has one area of herringbone paving, surrounded on
three sides by a groove, apparently for a wooden screen. The excavator
I-.
I) merely describes it as of ‘basse époque’, but makes it clear that it preceded a
period (again not dated) when, as at Belalis Major and Sabratha, the forum
a
0
CO
was turned into a cemetery. No nearby church is known the main and
—

strikingly sumptuous Christian quarter (as befits the city of St Augustine)

lay over 200 m away; but there is sufficient unexcavated ground by the
forum for one to have existed, and a cemetery church may one day come to
light.79
One would dearly like to know more about the history of other fora in
be Algeria. One such is the now remote and desolate site of the colonia of
a
DiaHa Veteranorurn (modern Zana), on the southern edge of the great upland
0
CM plain which stretches down from Sétif. Much of the forum has been laid
bare, and is strewn with masonry, including many inscriptions, which at first
CO
sight look like crude buildings of the sort excavated on the forum at
F—0 Cherchel. However, detailed study by Duval suggests that they form part of
a substantial church (Figure 36). There is also a qsur, or fortified structure,
. 0
I-.
perhaps a farm, which incorporates as its north wall an arch of Macrinus
(who, as we have seen, was born at Caesarea), which once afforded entrance
into the forum. Bordering its east side is an extremely well preserved
Byzantine fort, with walls still standing to a height of four m. There is, alas,
no dating evidence for this transformation of the colonia’s civic centre
(although there is fifth and sixth-century African Red Slip ware upon the
ground today); but it clearly belongs to the world of late antiquity, and
Duval regards the church as a Byzantine construction.80

In

be
Marec 1954. For the Christian quarter: Gui et al. 1992, 346f.
!.. f 80 N. Duval in MEFRA 89 (1977) 846f.; Gui et al. 1992, 156-8; Pringle 1981, 256
(who did not recognise the church).

70 71

zzi
Towns in Late Antiquity: Jol Caesarea and Its Context Civic Centres In Late .1ntiquity: The Wider Context

Still less is known of the history of civic centres of towns like Tipasa and
Djemila. Of that at Tipasa. Lancel observes that ‘cest tin forum rnuet, rendu
a Ia nature, ourlé de pins et d’oliviers’:81 evocative words, as befits a site of
extraordinary beauty, but reflecting a period of research when the primary
aim was to expose what remained of the monument, rather than work out the
nuances of its chronological development. The forum and basilica were
situated on a spur which juts out into the sea in the central part of the city.
V
CL) However, as pointed out earlier, the main Christian focus lay at the west
end, where a huge cathedral (Figure 6) was built, probably in the late fourth
century. Although what may be a Christian chapel was constructed amidst
houses to the north-east of the forum, (and there are those who believe that
the judicial basilica was made into a church) it seems likely that this was no
longer the primary centre of urban life.82
U Djemila offers a similar picture. The original town, founded in the late
>
first century AD, was built towards the end of a long ridge, with broad
valleys on either side. It soon expanded from its original nucleus and the
1’ I
principal Christian quarter (again datable to the late fourth or early fifth
century) lay far distant, at the top (southern) end of the ridge. Its Christian
basilicas, and imposing baptistery, surrounded by a medley of private
houses, is one of the most compelling images depicting the impact of
Christianity upon late-Roman north Africa (Figure 37)83
Enough has been said to suggest that in many, and perhaps most, north
African cities and towns, the forum as a public square went out of use
around the end of the fourth century. At the same time, new and specifically
Christian foci came into being, often in quite different parts of the town. It
seems inconceivable that these two phenomena are not linked, and reflect

I
both the rejection of the pagan associations of so many fora, and the potent
force, religious and political, of Christianity.
Seen in this light, the evidence from Cherchel appears somewhat unusual.
It would seem to be the only known north African site where new building-
work was being undertaken in the civic centre in the first part of the fifth
century, while equally unusual is its continued role as a commercial centre,
albeit with market stalls of wood. Apart from the Pompeian wall paintings
described above, these stalls are extremely hard to parallel, both in the

81 Lancel 1971, 31. See also Lancel 1982.

82 Gui et al. 1992, 21f. They are sceptical about the Christian chapel (29—3 1), probably
rightly, but more convinced that the basilica was converted into a church, comparing
it with Sabratha and Lepcis Magna (note 77, supra).
83 P-A. Février, Dje,nila (Alger 1978), 74f.; Gui et al. 1992. 92f.

72 73
Towns in Late Antiquity: Jol Caesarea and Its Context Civic Centi-es In Late Antiquity: The Wider Context

literature and on the ground. It should at once be said that it took some time
during the excavation to recognise them for what they were: oniy in the
raking light of the early evening did they show as a distinct pattern, and it
required a certain amount of luck to realise their nature. Subsequently, a
good deal of time has been expended on scrutinising other fora for similar
traces of grooves and postholes, especially on two extensive journeys
around Algeria and Tunisia, but also in other countries. The enquiry was not
on the whole successful. Possible postholes and grooves were seen on the
fora at Tipasa and Timgad, and in one corner of the forum at Hippo Regius;
but in no instance did they form any very recognisable pattern. Slightly more
convincing were marks (Figures 38, 39), bordering the east side of the forum
at Philippi in Thrace (a civic centre flanked by huge Christian basilicas); and
there is an elusive reference to postholes for a market stall on the forum at
Paestum.84 But this is the sum outcome of an investigation which, while far
from systematic, has been quite extensive. That said, it should be pointed
I, out the condition of the paving of many fora is such as to render it extremely
difficult to observe such exiguous traces. Nevertheless, the impression from
fifteen years of intermittent search is that the wooden stalls at Cherchel are a
rare, and so far probably unique, phenomenon.
How, then, is one to explain their existence, and the building work of this
period? The answer must surely lie in the fact that at Caesarea the forum (if
that is what we have) in this case did become a focus for Christianity. It
helps to lend credence to the identification of our apsidal structure as a
church, and it raises the intriguing question whether the refloored basilica
may also have been converted into a cathedral. Sabratha. where the forum
likewise appears to have continued in use, and the basilica was rebuilt for
ecclesiastical purposes, may’ offer a telling parallel (as, in terms of urban
topography. does the city of Philippi).
The presence of wooden stalls around the forum at Cherchel shows that
its role as a commercial centre was enhanced in the fifth century. At first
sight this is surprising since, as we have seen, city markets in north Africa
tended to be housed in special buildings. However, we know very little of
the late history of these macella, or of how goods were bought and sold

84 Paestum: Greco and Theodorescu 1980, 37; Philippi: M. Sève and P. Weber, Le cOte
nord du forum de Phi11:pes. Bulletin de Correspondence Hellénique 110 (1986),
531—81 (without, however, a suitably detailed plan of the paving); P. Lemerle,
Philippes ct/a Macédoine orientale hl’epoque Chrétienne et byzantine (1945). Gui et
a!. 1992. 263f. provide a useful discussion of the Christianised urban landscape of
Timgad.

74 75

A
Towns in Late Antiquity; lol Caesarea and its Context Civic Centres in Late Antiquity: The Wider Context

C)

C)

C)
C)

C)

C)

C)

I
C

0
‘C

.4
U

76 77
 Towns in Late Antiquity: Jo! Caesarea and Its Context Civic Centres Jn Late Antiquity: The Wider Context

within the late-antique town. Yet, as Shaw has shown. there is a individual owner,89 and there are some signs that production of oil and wine
considerable body of evidence which indicates that north African bishops, may have doubled. Moreover, there was an abundance of African Red Slip
priests and deacons were heavily involved in commercial matters, especially ware, coins, amphorae (the parallels for which lie mainly in the west
rural markets.85 This much is clear from repeated warnings against any Mediterranean, especially Spain) and other finds, which indicatestistaiied
entanglement with the world of business and profit-making’, a role that rural activity throughout the fifth century and into the early part of the
Christian ecciesiastics seem to have taken over from their predecessors, the sixth.90
priests of Saturn. However, clerics were allowed to participate in commerce, Nador was clearly never the residence of an aristocratic owner (even
if for example, the profits of their shops or workshops went into the coffers though it was owned by such), and archaeology is silent about the identity of
of the church (Codex Justinianus I 3.2 (357)). They could also act as those who farmed there in the fifth century, whether they were Roman or
negotiatores if it was for the benefit of the poor.86 Given the proximity of Vandal. But it is striking that the early fifth century remodelling of the farm
two possible churches, one is tempted to speculate whether the Cherchel coincides chronologically with the rebuilding of the forum site at Cherchel,
stalls might indicate some sort of periodic market, under church control. The and that the two sites entered a demise at about the same time. This may be
3talls might well have been put tip on a temporary basis, the grooves being coincidence, but it would be unrealistic not to allow for the possibility that
used to delineate the location of each stand; it is hard, otherwise, to attach a these sites were affected by the Vandal annexation of the region and, a
structural purpose to such ephemeral features. century later, by the Byzantine reconquest.
A juxtaposition of an important ecclesiastical centre and commercial To sum tip, the evidence from north Africa clearly points to the demise of
activity is also to be found in the territoriuin of Caesarea, notably at Trois public fora in the late fourth or early fifth century at many sites. This
hots, on the coast, a few km to the north-east.87 We have already referred broadly coincides with the emergence of usually qtiite separate Christian
earlier to the relatively large church, which was evidently the centre of a foci, which were often located at a considerable distance from the old town
substantial vicus (Figure 7). Although only the barest details have been centre. Cherchel and Sabratha (and probably Carthage)9’ stand out as
published, the sites industrial role is not in dotibt. Still conspicuous today exceptions to this pattern, in that the forum appears to have been maintained
are a series of tanks for the production of garum, and there is also evidence in use well into the fifth centtiry. In both instances, it is likely that this was
for the manufacture of pottery and traces of oil presses. In addition, quite due to the construction of adjacent churches. Moreover, at Cherchel it can
elaborate residential rooms have been excavated. While earlier occupation is he demonstrated that urban vitality is also reflected at some rural sites in its
probable, surface finds (especially African Red Slip, fabric D) indicate that territoriuni; here the Church may also have played a role in promoting
it flourished particularly in late antiquity, and is certainly contemporary with production. Whatever the political and social consequences, the Vandal
the late-Roman buildings and stalls at the forum site at Cherchel. intervention as measured by the archaeological record had a limited impact:
The same is true of the rural complex at Nador, on the Cherchel-Tipasa it was certainly not catastrophic. Whether this vitality persisted throughout
road, briefly discussed in the previous chapter.88 Founded in the period of the fifth century is another matter, and the Cherchel evidence may be read to
Ptolemy, in the fourth century it was a farm owned by M Cincius suggest that this was not the case. But, whatever the truth of that, it is quite
Hiliarianus, flameti Augusti, and his wife, Vetidia Impetrata. In the early clear that there is not one north African Roman town, possibly other than
fifth century, however, it was extensively modified, so that the 23 rooms or Carthage, where it can be demonstrated that the old civic centre survived in
areas of the fourth-century building were sub-divided up into no fewer than use into Byzantine times. Instead, a very different townscape was emerging.
57. It may have passed into the hands of a number of families, rather than an
89 Cf. the important review by Mattingly and Hayes 1992.
90 Cf. J. J. Rossiter s interesting discussion, Villas Vandales: Ic suburbiu,n de Carthage
85
86
87
88
Shaw 1981, 69.
Whittaker 1983, 167—9.
Leveau 1984. 248f.; J. Lassus, BAC 1955-56. 119-21; Gui eta!. 1992, 18-19.
Anselrnino et al. 1989.
I 91
au debut du VIe siècle de nOtre ère. Carthage et son territoire dans l’antiquité (Paris
1990), 221—7; also Fentress (1984) important discussion on the interpretation of late-
Roman rural settlement around Cherchel in the light of field-survey data.
Deneauve 1990; cf. the final section of this chapter.

78 79
 Towns in Late Antiquity: lol Caesarea and Its Context Civic Centres In Late Antiquity: The Wider Context

LATE-ANTIQUE CITIES IN THE EASTERN MEDITERRANEAN

The evidence from oriental Mediterranean cities may seem of peripheral

relevance to this enquiry. But the eastern amphorae found at Cherchel EPHESUS
(Figure 23) remind us of the commercial ties with that part of the world,
especially in the fifth century, and a good deal is known of the late-Roman
in late antiquity
archaeology (and history) of many eastern cities. What is said here must
necessarily be very superficial; but it will be useful and, it might be
—

suggested, instructive — to seek comparisons.

I Ephesus is a key site, not least because of Foss’ admirably balanced and
sensitive synthesis of the historical, epigraphic and archaeological data for 1hof St Mary
the late-antique, Byzantine and Turkish city.92 Throughout most of the
fourth, fifth and sixth centuries, it remained a thriving place which received
considerable patronage (Figure 40). Some of this was imperial, such as the
Baths of Constantius II, the great colonnaded harbour road of Arcadius and Baths of
Constantius II
the vast basilica of St John, constructed at the behest of Justinian to replace
a church of c AD 390—420. Governors also promoted municipal welfare.
While the free bread dole enjoyed by the needy (as in Rome. harbour
Constantinople, Carthage, Antioch and Alexandria) may have been an
Street of
imperial benefaction, governors certainly took steps to ensure ‘the blessing Arcadius
of grain of fruitful Demeter for the citizens’.93 They also maintained the
water supply the Library of Celsus was, for example converted into a
fountain - and restored streets, baths, the great theatre and much more.
There were also a few private patrons, most notably illustrated by the baths
paid for by a wealthy Christian lady, Scholasticia, in the late fourth century;
but personal benefactions were largely a thing of the past, and the state,
whether through the governor or through the municipal officers and Council
(Boule). effectively took over the burden.
Celsus
As far as the city council was concerned, this was not to last. Councillors,
— —

whose duties and commitments were heavy, became steadily poorer, and, as
in all parts of the empire, the council became of little consequence or ceased
to exist.94 Increasingly, the Church became a major benefactor, not only ,
—
\
building churches, but providing other public services such as hospitals, —

Temple of
poorhouses and inns (all of which are attested at Ephesus). Two Church \ Domitian
councils, those of 431 and 449, were held at Ephesus, underlining its \ N
N -

92 0 100 500m.
Foss 1979. N
N P
. — 93
Foss 1979, 27.
Jones 1973, 737f., especially 760—1; Liebeschuetz 1972, 167—86. But see iiote 145
Figure 40 Plan of late-antique Ephesus (Turkey). Mainly after Foss.
below.

80 81
 Towns in Late Antiquity: Jo! Caesarea and Its Context Civic Centres In Late Antiquity: The Wider Context

Seven Sleepers. Refugees from the persecution of Christians by Decius, they

slumbered in a cave near Ephesus for two centuries, before waking up
towards the middle of the fifth century. One of their number was sent into
the city, and was so astounded by the changes that he had to enquire of the
name of the place. Much indeed had altered, not least in the upper agora,
once the political heartland of the city. It is possible, on the evidence of an
inscription, alas not closely provenanced, that this was known as the Forum
of Theodosius, although whether 1 (379—95) or II (402—50) is not
specified.96 But it is more likely to refer to the lower market agora for, in the
state agora, much of what went on around this time was a process of
destruction. Nothing, unfortunately, can be very closely dated, but it is clear
that the Prytaneion (a type of town hail, associated with the goddess of the
hearth, Hestia) was destroyed, as was the Temple of Domitian and the
Temple of Rome and Caesar, over which houses were built. Similarly, the
Temple of isis, which lay within the agora, was pulled down, and houses
began to encroach over the paving. The basilica was, however, given a
Christian connotation by the erasure of the name of Ephesus’ patron deity,
Artemis. and by the incising of crosses on the foreheads of statues of
Augustus and Livia (an interesting mark of respect). The bouleuterion also
remained in use for a time, although the city council of Ephesus is not 11
attested after the Theodosian period, and such councils became effectively
extinct after the financial reforms of Anastasius in the early sixth century.97
While the chronological indicators are extremely imprecise, it does seem
evident that the main civic centre at Ephesus was transformed in the late
fourth or fifth centuries. No longer was there the elegant piazza of early
Imperial times, but a landscape that must have appeared a curious medley of
ruinous and standing buildings, many composed of spolia from torn-down
pagan edifices. The main colonnaded street (one of the architectural glories
Figure 41 The Embolos at Ephesus. looking towards the Library of Celsus. In late antiquity
of the eastern Roman cities), was known in late antiquity as the Embolos; ii
this became the main centre of civic life.
m. wide, it ran down into the lower town (Figure 41), and was likewise
drastically altered. Shops and markets filled the space where once were
political importance, and were held in the citys cathedral, the Church of St shady pavements behind the colonnades, and in time many of the columns
Mary. This enormous structure had been built in the mid to late fourth were taken down and the space filled in with rubble and spolia. Christian’
century, overlying, perhaps significantly, a former market building; this symbols were ubiquitous and, although there were elegant houses set into
cathedral is thought to have been the chief centre for the distribution of the hillslopes which flank the Embolos, the grandeur of before had gone.
grain, a role iii which the Church was to come to play a significant part.95 While the lower market agora may have continued to function (perhaps until
The Christianisation of Ephesus is, as in north Africa, the most striking
feature of the late-antique city. It is epitomised by the famous legend of the
96 See note 139.
Whittaker 1983, 167—9. Foss 1979, 13.

82 83
 I Towns in Late Antiquity: Jo! Caesarea and Jts Context

excluded by the Byzantine city wall, itself possibly of seventh-century date),

Civic Centres In Late Antiquity: The Wider Context

(Figure 42). Here are eloquent echoes of what we have been describing at
the Embolos now formed the main commercial centre of the city; but it must Ephesus. 100
have begun to acquire an appearance with medieval echoes. Indeed, in some The Ephesus model, if it may be so described, appears to be widely
neighbouring districts, there was encroachment upon some streets, as upon L I repeated in the east Mediterranean. In the more far-flung sites, like the
the upper agora, both being actions that were severely prohibited, apparently desert city of Dura-Europos in Syria, the Hellenistic agora was by the mid
ineffectually, by late—fourth-century legislation.98 third century covered with buildings, many of them shops.’°’ It has been
It must be stressed that, as in so many north African cities, the main compared to a medieval bazaar, and recalls the contemporary spread of
Christian basilicas lay some distance away. The Church of St Mary was 650 structures over the agora at Cyrene (although without necessarily implying a
m from the bottom of the Embolos, while the Church of St John was built on similarity of function).’°2
a hill, Ayasoluk, over two km away; this is explained by the fact that St Elsewhere, as has been often demonstrated, colonnaded streets frequently
John. according to ancient tradition, had lived and died upon this hill. became commercial foci, to the detriment of public squares. Antioch-on-the
However, although Avasoluk became a citadel in the later Byzantine period, Orontes (now Antakya. in south-eastern Turkey) is a much discussed
neither church became centres for major urban foci in the manner of, for example, partly because there is both literary and archaeological
example, Djemila, in late-antique times. While the upper state agora was evidence.’03 An important source is Libanius. Born in the city in 314, from
effectively lost to view, the Embolos was the heart of the city, reflecting 354 to his death at the end of the fourth century lie held a chair of rhetoric
perhaps the commercial vigour of this great metropolis of Asia Minor. As there. Amongst his writings was an encomium on Antioch, in which he
far as possible, the old pagan associations were swept away not least the lavishes praise upon the colonnaded streets. In so doing, he makes it clear
Temple of Artemis, which lay’ at the foot of the Ayasoluk hill. and was one that these were the places to shop. rather than the agora,’°4 and it is easy to
of the Seven Wonders of the Ancient World; but the principal bones of imagine how stalls in front of permanent shops were increasingly colonising
Ephesus’ urban topography. as established in classical times, remained. the shady walkway beside the streets: that. in effect, a linear market was
We have dwelt at some length upon Ephesus, because it is possible to coming into being. Archaeology has not documented these stalls, but it has
model in detail the changes to the urban landscape that took place in late shown how in a later period the principal street was narrowed, the
antiquity. But there is also a wealth of data from other cities in the east colonnades blocked up and structures were built over the sidewalks (Figure
Mediterranean. both literary and archaeological, which has encouraged 43). Dating is problematic, but these changes may belong to the Justinianic
broadly based syntheses.99 The evidence for change is almost everywhere period. At any rate the old alignment was essentially maintained and,
conspicuous in the many excavated urban centres which did survive iiito
late-Roman times. Once has, for example only to walk down the main street
at Perge in Pamphylia to notice the blocked-up colonnades, the inserted
shops and the buildings that straddle the road, one at least a small church

100 I can find no very satisfactory publication of Perges late-Roman archaeology,

perhaps because, in reference to the excavation of two colonnaded streets, it is
remarked that some unimportant Byzantine building remains, which were not
accurately planned, were removed’ Anatolian Studies 24 (1974), 48). See also Arch.
Anzeiger 90 (1975), 57—96, and Turk Tarih Kurumu 16 (1983), 178f.
101 Excavations at Dura-Europos, Ninth season. pt. 1 (1944); J. B. Ward-Perkins 1974,
98 Prohibition of building upon streets: Cod. Theod. cv. 1.39; upon an agora: Cod. 21.
Theod. x.1.22. 102 See note 76.
9 Kennedy (1985) is fundamental. C. Foss’ essays are now most usefully collected 103 Downey 1961; Liebeschuetz 1972.

I together in History and archaeology of Byzantine Asia Minor (Aldershot 1990).

84
104 Or. xi, 251; Liebeschuetz 1972, 55—6, reminding us that the agora did still exist.

85

V.
Towns in Late Amiquity; lol Caesarea and Its Co,iiext
Civic Centres In Late Antiquity: The Wider Context

C’s
Antioch Justinianic

5)

5)
0
z
0

Roman

.1
street
0

1.
0
1

0 10

(Turkey), showing how

Fgure 43 Plan of successive main streets at Anlioch-on-the-Orontes
it was narrowed. The dating is tentative. After Lassus.

e parts
remarkably, is still largely preserved today, as are not inconsiderabl
of the street grid.105

.. U

1992. There
105 Lassus 1972; Liebeschuetz and Kennedy 1988, 65—6; and now Kennedy
before the end of the sixth
may have been further encroachment over the street
early, and significant, discussion of these
century. Claude (1969) is a relatively
lly somewhat adrift. since he regarded many of these
questions, albeit chronologica
changes as of Islamic date.

86
87
 Towns in Late Antiquity: lol Caesarea and Its Context Civic Centres In Late Antiquity: The Wider Context

Other examples of street porticoes which housed shops and stalls in late
antiquity include Gerasa (Jerash, in Jordan), and Palmyra and Apamea, in
Syria, to name but a few.’°6 At Apamea, the main colonnaded street was
eventually closed off by the construction of a church, apparently in the late
sixth century; commerce rather than communications was now what
mattered.107 Naturally, developments of this sort attracted opposition from
the Imperial authorities. Justinian is supposed to have embellished Antioch
on-the Orontes with stoas and agoras; streets and waterworks; and theatres
and baths; but our source, Procopius, is notorious for his exaggeration, and
there is little archaeological confirmation.108 By Justinian’s day, Antioch
must have looked a very different place from early Imperial times. Likewise,
I,
we hear from Joshua the Stylite, writing between 494 and 506, of a new
governor at the important city of Edessa, in north-east Mesopotamia,
I
ordering the destruction of stalls between the pillars flanking the main
street.’09 We must surely doubt, however, the effectiveness of such edicts.
The burden of these reinarks, then, is to suggest that, in the eastern cities,
there was a fundamental change of emphasis in the urban landscape in late
antiquity. Public squares lost their role as centres of assembly, display and
commerce, and churches became the principal form for expressing
/ architectural grandeur. The colonnaded streets, on the other hand, became
I the hub of the city, especially as markets. The colonnades were walled in
(Figure 44) and, as time went on, these came increasingly to resemble the
souks of the Islamic town, with narrow streets, lined with shops and stalls.
Thus, when Muqiiddasi visited Damascus in the tenth century, he was to
find that all but the main street had become covered markets,0 probably
not too dissimilar from, say, the Kapali çarsI market in Istanbul or the north
African souks of Tunis or Fez today. It is a startling transformation, but one
which is firmly rooted in the fourth-fifth centuries, as people increasingly

106 Crawford (1990, 107f.) has now provided a very useful discussion of porticoed
streets that were converted into shops in Asia Minor, Greece and the Balkans in his
report on the late-Roman shops at Sardis.
107 j• Balty and J. Ch. Balty, Actes du Colloque Apamée de Sync (Brussels 1969—80, 3
vols); Liebeschuetz and Kennedy 1988, 66; Lassos 1972, 133—4, 145—6. See also J.
Ch. Balty in JRS 78 (1988), 91f. for the city’s earlier Roman history.
108 A. Cameron, Procopius and the sixth century (London 1985); cf. B. Croke and J.
Crow in JRS 73 (1983), 143f. for an archaeological testing of Procopius’ veracity.
109 Chronicle of Joshua the Stylite (ed. W. Wright, Cambridge 1882), 29, cited by
Liebeschuetz 1972, 260. For Edessa, J. B. Segal, Edessa, ‘The Blessed City’
(Oxford 1970).
110 Cf. Kennedy 1985, 12, and generally for the possible reasons behind such changes.

88
89
 Towns in Late Antiquity: lol Caesarea and Its Context Civic Centres In Late Antiquit,’: The Wider Context

“turned their back upon the monuments of the pagan past. And, more than
that, have we here some sort of echo of this ‘commercialisation’ with the
market stalls of the forum at Cherchel?

ITALY

But we must go back to the west Mediterranean, and especially to Italy, to

assess developments there. As we observed at the beginning of this chapter,
it is often argued that, especially in central and northern Italy, the Roman
forum was to develop into the market square of the medieval city. Equally,
the survival of so many Roman street grids, even down into the modern age,
has seemed good reason to postulate a degree of urban continuity, with a
semblance of civic discipline, through the ‘Dark Ages’. It is indeed
astonishing to look at aerial photographs of towns like Aosta, Como, Fano,
Verona, Lucca, Pavia or Piacenza, to name but a few, and see how
completely the Roman street layouts are fossilised in the modern landscape
(Figure 45).hhl It points to a significant measure of control over street
frontages, and careful maintenance of the streets themselves. In this context,
the survival of fora as market places would seem to make good sense, even
though this was far from being the case in the regions that we have hitherto
glanced at.
In reality, the picture is rather more complicated, and two clear opposing
schools of thought have emerged over the past decade or so. One (proposed
mainly by historians) favours a strong degree of urban continuity, while the
other (primarily advanced by archaeologists) envisages a sharp ‘ruralisation’
of the city in late antiquity and early medieval times. What is not in dispute
is that a significant number of towns many of them in the south of Italy
did disappear completely during this period; one calculation suggests 11 6
out of 372 towns.H2 It is what happened in those that did survive that is
controversial.
We will continue to focus upon what may have happened to the fora.
Wickham, whose evidence is primarily documentary, has no doubts: ‘The
forum remained in most cities
... In the early Middle Ages, it had two
...

rivals, the royal palace and the cathedral, reflecting the two major powers in
every city, the state and the bishop. The forum lost a direct political role
after the city council disappeared in the sixth century, although it remained
Figure 45 Aerial photograph of Como (Italy), showing how the Roman Street pattern is
fossilised down to this day as a chessboard grid. (Photo by courtesy of the British
Ward-Perkins 1974 is an accessible source; also Bryan Ward-Perkins 1984, 179. School at Rome).
112 Schmiedt 1973; Ward-Perkins 1988 and also mArch. Med. 10(1983), 111—24.

90 91
1 Towns in Late Antiquity: lol Caesarea and Its Context Civic Centres In Late Antiquity: The Wider Context

an economic centre, and the market was still held there.’113 He goes on, some 30 cm of silt. The finds included imported amphorae and ARS, and the
however, to point out that the cathedral was usually built on the edge of the stratigraphically latest building a radiocarbon date of AD 640 ± 80.
city (‘as the last major late-Roman civic building’), and was to become the Coincidentally, but nevertheless strikingly, this is precisely the same result
prime political centre, as time went on and the church achieved ever more (the standard deviation apart) as the radiocarbon date for house 2, over the
power. Ward-Perkins is somewhat more circumspect, but he does conclude forum at Cherchel.
that the ‘evidence does suggest continuity in some towns from the Roman
... Most have interpreted this degeneration of the town centre as a symbol of
forum to the early and later medieval market. Certainly, too, there is no decline, brought about by the collapse of the marble trade, and a primarily
evidence from an early date of markets having moved to new medieval mountainous hinterland. vvhere it was hard to sustain agricultural
settings’. And he adds: ‘with the decline of traditional [private] munificence, production. Yet Luna was to remain a place of consequence for some
Italian towns kept their squares and streets, but lost their splendid paving.” 14 centuries. It had a bishop from at least the middle of the fifth century, and
Those who oppose this view do so upon a more generalised was a major Byzantine centre for Italia Maritima. Aldio, the inagister
archaeological appraisal of the nature of the urban landscape in Italian inilitum, had his base there in the later sixth century and the city was to
towns of late antiquity. They paint a broad picture of urban space invaded participate in important councils during the next two centuries. Although
by burials, long prohibited in Roman law; of impoverished houses, built eventually abandoned in favour of the town of Sarzana, somewhat further
often in wood; and of layers of ‘dark earth’, brought in to facilitate inland, Luna cannot be dismissed as a place wholly gripped by decay in late
cultivation within the city walls. Brogiolo can offer evidence of a ‘notable antiquity.’ 18
crisis’ within the western quarter of Brescia, with poorly surfaced or buried It did in fact have a sumptuous cathedral, S. Maria, built in the late fourth
streets, and collapsing buildings.’ 5 There can rarely have been so radical a century on slightly higher ground close to, but within, the city walls, about
divergence of historical and archaeological interpretation. 130 iii from the site of the forum. It was embellished with fine floral
As far as the late-antique history of fora is concerned, hard archaeological mosaics, donated by Gerontius (probably a bishop) in the sixth century, and
facts are in short supply. The best documented example is the much saw significant building work in the eighth and ninth centuries.1 19 Other
discussed case of the colonia of Luna, modern Lunij16 This was a coastal churches are also attested at Luni.’20 and it seems clear that, whatever the
town which prospered particularly through supplying white marble from the fate of the forum and associated monuments like the Capitolium, the town
nearby Carrara quarries, although by the end of the fourth century these remained of significance. Surely it is a parallel for what we have been
were no longer being exploited.’’7 By this time the forum may also have tracing in many north African towns, namely the creation of a separate,
been out of use; it would seem its paving had been largely stripped, and silts specifically Christian, focus, set well away from the old civic centre.
were forming over it (although this is not as closely dated as once thought). A not dissimilar situation is to be found at the old Greek city and Roman
Then, about 550, several wooden houses were constructed, resting upon colonia at Paestum, in Carnpania.’2’ There was building work on the south
side of the forum in the second and third centuries, and fourth-century
113 Wickham 1981, 83. 1’inscriptions attest meetings in the curia. But by late antiquity the forum was
114 Ward-Perkins 1984, 184—5. La Rocca (1986) would broadly support this view; see /1 no longer the place it was. There are archaeological hints, as we mentioned
also now La Rocca 1992. earlier, of possible wooden stalls, and refuse pits were cut into the paving; it
115 Brogiolo 1987, 44; 1989. P. Hudson (Arch. Med. 12 (1985), 289) describes Verona
in this period as having an image of desolation and ruralisation’. Hodges and
Whitehouse (1983, 32—3) offer a similar picture. Private buildings in early Lombard 118 Christie (1990) sets the broader scene. The view of Lunis late-Roman decay is
Milan were certainly impoverished (D. Caporussi, Scavi MM3 (Milan 1991), various almost universally depicted: eg. Liebeschuetz 1992, 16.
excavation reports). Contrast this with the carefully balanced appraisal of La Rocca 119 S. Lusuardi Siena, Arch. Med. 12 (1985), 303—12; id., Luni palaeocristiana e
1992. altomedievale nelle vicende della sua cattedrale. Atti del Convegno. Quaderni del
116 Ward-Perkins 1978, 1981, 1986. centro di Studi Lunensi 10—12 (1985—7). Sarzana.
117 E. Dolci, Carrara: cave antic/ic (Carrara 1980); J. B. Ward-Perkins, Marble in 120 P. M. Conti, Luni nd/a/to medievo (Padua 1967), 180f.
antiquity (London 1992), 21—2, with note 30. 121 Pedley (1990, 163f.) is a useful summary.

92 93
I

Towns in Late Antiquity: lol Caesarea and Its Context Civic Centres In Late Antiquity: The Wider Context

was clearly entering a period of decay. However, as at Luni, a new Christian Arce (5th century), by the Augustan temple to the genius of the colony, on
centre came into being: the old Temple of Athena, 70 rn to the north, was the Arce hill; S Pietro (late 5th—early 6th century) and S Giovanni (AD 625),
converted into a church, and another church was built near by. Crudely- just inside the western city walls; and the extra-mural church S Faustino
constructed houses are attested, as well as an extensive necropolis of fifth to (second half of the 4th century). The ecclesiastical organisation therefore
seventh-century date, and there are sixth and seventh-century references to a remained strong in late antiquity. However, excavations on two sites showed
bishopric (later transferred to a new centre in the hills at Capaccia).’22 The building in ruins in late antiquity, and streets covered with rubble and ‘dark
parallel with events at Luni is noteworthy. earth’, an unmistakable symptom of decline. Yet these excavations were of
Another site of interest in this respect is luvanum, tucked away in the places on the east side of the ancient town, not particularly far from the
Abruzzi mountains to the south of Chieti. Once an important sanctuary of forum, but well away from the palaeochristian centres. The area was not
the Frentani tribe it became a prosperous Roman municipiuin. The forum has totally abandoned, for an early Lombard (i.e. later sixth century) wooden
been completely cleared, and there have been recent, sensitive excavations building was identified, overlying a Roman domus that had been destroyed
upon the surrounding shops and houses. These have shown that, although by fire, at S Giulia Ortaglia. Moreover, the line of the Roman street
there is epigraphic evidence to indicate repairs in the fourth century to the (although not the surface) was perpetuated by a wooden fence along the
city walls and streets,123 the buildings around the forum manifest every edge. In the same period, the nearby site at Via A Mario was turned over to
indication of decline from the third century. One was turned, for example cultivation, and with these demonstrations before us, it is hard to imagine
into a store-room, just one indication that a once graceful urban centre was that the old civic centre continued to be maintained.126
no longer maintained as such. Thereafter, there are few coins or fragments The matter has in fact been investigated at Verona. Here, the forum was
of fine table ware, in sharp contrast to early imperial times: it is hard to covered with a deposit some 70 cm thick, which has been interpreted as a
resist the conclusion that the forum, for whatever reason, was no longer the sign of agricultural activity. This was in turn cut by a robber trench
town’s main focus. Indeed, it is suggested that, although there was later to be containing Lombard-type pottery, which gives a rough terminus ante quein
a medieval monastery on the site. Tuvanum was more-or-less completely of the second half of the sixth century for the demise of the forum. However,
abandoned by the mid fifth century.’24 Could, we must ask, a new nucleus a nearby excavation showed that a wall, probably of fifth-century date, had
have arisen in late-antique times, as at Paestum and Luni? been constructed over the stairway leading up to the Capitolium; it is
Something of the same pattern is to be observed at the Apulian town of therefore to be doubted that the forum was still in use as such. Moreover,
Ordona, ancient Herdonia. In late-Roman times, the population was although quite elaborate stone-built houses like that at the Via Dante, built
increasingly concentrated around a Christian basilica in the northern part of in the late fifth century, did survive through the early Middle Ages, others
the settlement, a process of change which is thought to have begun in the did not. Thus a house near the Cortile del Tribunale had earthen floors over
third century. In the forum, the temple there was eventually converted into a its mosaics by the end of the fifth century; was burnt down in the early
cemetery church, and other funerary chapels built out of buildings around seventh century and covered with dark earth; and was not rebuilt until the
the forum. No longer did the old prohibition upon burial within the city ninth century.127
walls hold sway.125 Scholars have chosen to interpret the results from Verona in rather
In the north of Italy, Luni apart, the available evidence is much more different ways; but however town-life is characterised in the ‘Dark Ages’, it
fragmentary. There has however been much work at the colonia of Brixia, is difficult to reconcile the archaeological evidence for the forum’s demise
modern Brescia. Four palaeochristian churches are known: S Stefano in with the description of it in about 800 as a broad and spacious paved piazza,
with arches at each corner.128 Is it a new forum that is being referred to,
122 Greco and Theodorescu 1980, 37.
123 CIL IX, 2956, 2957. 126 For the churches, see C. Stella, Brescia roinana I (Brescia 1979), end map. For the
124 Fabbricotti 1992.
125 excavations, see Brogiolo 1987, 1989.
Although not yet published in detail, cf. J. Mertens, Ordona VIII (Brussels 1988), 127 Forum: Cavalieri Manasse 1990; La Rocca 1986.
21 f. 128 Versus, lines 10 and 12 (from Ward-Perkins 1984, 185).

94
95
 Towns in Late Antiquity. Jo? Caesarea and Its Context Civic Centres In Late Antiquity: The Wider Context

matching the well-attested spread of churches across the urban landscape in

the eighth to tenth centuries?129 The same may also be true of Pavia, where
tenth-century references to a Foro Cluso with adjacent stationes (booths) are
thought unlikely to refer to a direct successor of the Roman forum.130
Equally, this may provide a context for the Terracina inscription, mentioned
earlier, describing the ‘cleaning up’ of the forum.131
This may also help to explain the survival of so many Roman street grids.
As Bullough has argued for Pavia, the streets were periodically restored
during the Middle Ages, not least because the Roman drainage system was a
valued asset, and main streets were regarded as viae publicae.’32 The old
paving may often have been buried; but this was not always the case. At the
Roman town 011 Colle S Pietro, Tuscania, part of a stretch of basalt-paved
road was certainly retained in use in the late first millennium, and was also
extended, although on a slightly different alignment and with whatever
materials came to hand (Figure 46); a post-built structure fronted onto this
section of road.133 Similarly, a major highway like the Via Flaminia could
be resurfaced about the same time with basalt blocks, as recent excavations
at Malborghetto, near Rome, have demonstrated, while elsewhere in south
Etruria a minor country road was likewise refurbished in the early ninth
century. Many, probably most, Roman paved roads became buried, and lost;
but medieval resurfacing USillg Roman materials may have been more
common than one might suppose.134

129 La Rocca 1986, tav. 4. Figure 46 Excavations in 1974 at Colic San Pietro, Tuscania (central Italy). The Roman
130 Bullough 1966, 110—11. street has been partly realigned in early medieval times, and is flanked with
131 A. Guiilou, MEFRMA 83 (1971), 149—58. Note how the forum temple at Pola was postholes (? of eighth-century date), and later walls.
made into a church and that the forum temples at nearby Nesazio were destroyed in
the fourth century, further hints of significant changes to urban centres in late
antiquity (R. Matijatic, in La citth nell’Italia settentrionale in eth romana (Collection
de lEcole francaise de Rome, 130, 1990), 635f.). But this is to digress. We must now conclude this section with a few
132 Bullough 1966, 98.
133
words about Rome (Figure 47). Unlike Athens, where the agora finally
T. W. Potter and P. A. Gianfrotta, Arch. Med. 7 (1980), 446—7.
134 relinquished its traditional role as a place of popular concourse after the
I am grateful to the excavators, especially Dott. S. Coccia, for showing me the
excavations at Malborghetto. The demonstration of a minor road in S. Etruria being Herulian sack of AD 267,135 Rome’s fora, and the associated monuments,
resurfaced in the ninth century comes from work at Monte Gelato, described briefly lasted longer. The Basilica Aemilia was burnt down in the early fifth
in Chapter Four. For the use of major Roman trunk roads in the Middle Ages, see
now S. Quilici Gigli (ed.), La Via Appia (Archeologia Laziale 10.1, 1990), especially 135 On which now A. Frantz, The Athenian agora. 24. Late antiquity AD 267—700
the essays of Marazzi (117—26) and Coste (127—38). (Princeton 1988).

96 97
 Towns in Late Antiquity: lol Caesarea and Its Context Civic Centres In Late Antiquity: The Wider Context

century, and not restored, although shops along its front did remain in use. SOME CONCLUSIONS
Likewise, the secretarium of the Senate was destroyed, but immediately
rebuilt. A little restoration work is recorded during the fifth century, but the We should doubtless extend this enquiry into the sites of Greece, Gaul,
picture derived from such sources as we have is of slowly escalating decay. Spain and even Britain.138 But enough has been said to discern something of
Then, in the sixth and seventh centuries, there began a gradual process of a pattern in the late-Roman archaeology of the forum. On the evidence here
Christianising the area. Thus, in 526—30, the audience hail of the city presented, it would seem to be very difficult to support the idea that the
prefect, by the Forum of Peace, was converted into a church, dedicated to forum remained at the heart of civic life much after the end of the fourth
Saints Cosmas and Darnian, and later in the century (the precise date is century in many towns. There were exceptions, of which Cherchel would
unknown) S Maria Antiqua was built from part of the Imperial palace at the seem to be one. Another is Constantinople where both Theodosius I (3 79—
foot of the Palatine. Eventually, at least ten churches were to be constructed 95) and Arcadius (395—408) built grand fora, as befitted so splendid and
within the forum, mainly by converting older structures, during late important a city. Interestingly, the Forum Bovis, where cattle were bought
antiquity and the early Middle Ages. In Krautheimer’s words was the idea and sold, remained a market square (Aksaray) until very recent times; this
to consecrate to Christ the centre of Old Rome, still filled with memories of may be a genuine example of a continuity of commercial traditions, given
pagan times, just as the Pantheon was turned from a temple of the gods into the way that other eastern cities seemingly anticipated the Islamic souk from
a temple of God7136 late antique times.139
/ Krautheirner’s interpretation of all this is fascinating. He suggests that the Likewise, at Carthage it has been suggested that the monumental centre
/ creation by Constantine of a Christian administrative and political quarter at on the Byrsa hill persisted in use into Byzantine times, despite the picture
the Lateran, on the south-eastern edge of Rome, failed to attract a painted earlier of a city in deep decline. The evidence has yet to be
significantly large residential population (Figure 47). As the number of presented in full, but the excavators maintain that the pagan temple set
people living in the city shrank in late antiquity, they preferred to live in the within a piazza adjoining the forum was pulled down in the early fifth
bend of the Tiber, to the north of the forum, and also across the river, in century, and that the basilica was turned into a church. As with Rome, this
Trastevere. St Peters was, and has remained, the popular’ church. Thus, the
eventual Christianisation of the forum was an attempt to endow it with some
spiritual meaning, especially as after the Byzantine reconquest, the Palatine 138 The literature is of course vast, especially for Gaul, where the notion of the
and the area to the north of the Imperial fora became a sort of ‘government emergence of ‘islands’ within cities, or ‘double cities’, one an ecclesiastical
compound, both militarily and administratively. The Lateran remained the foundation, is well established. Cf. P.-A. Février et al., La yule antique, des
centre of ecclestiastical power; but secular authority was vested in the origines au IXe siècle. Histoire de Ia Frcnce urbaine, I (Paris 1980); E. James. The
traditional heartland of the city. Indeed, as church welfare-centres orIgIns of France (London 1982); H. Galinié (in Hodges and Hobley 1988, 61f.) is a
useful case study of Tours; J. F. Drinkwater and H. W. Elton (eds.) Fifth-century
(diaconiae) developed in the early Middle Ages, these were concentrated Gaul; a crisis of identity (Cambridge 1992); and S. T. Loseby, JRS 82 (1992), 165,
around the ancient markets of cattle (Forum Boarium), vegetables (Forum for a fascinating essay on the very prosperous late-Roman Marseilles. For the decline
Holitorium) and close to the Tiber-side granaries. However powerful the of fora in Spain in the fourth century, S. J. Keay, Roman Spain (London 1988),
authority of the Popes, the ordinary populace of Rome was stubbornly 184f.; for Britain, S. Esmonde Cleary, The ending of Roman Britain (London 1989),
conservative in the face of urban change. A deep and ancient sense of 71f. For an equally fascinating glimpse of urban life in late-Roman Greece, see T.
Gregory, Cities and social evolution in Roman and Byzantine south-east Europe, in J.
tradition may have had much to do with it)37
Bintliff (ed.), European Social Evolution (Bradford 1984), 267f., esp. 273. Barnish
1989 is an extraordinarily wide-ranging and interesting overview of the broad
development of late-antique cities, to which can now be added the essays in Rich
1992 (including J. Harries on Christianity and the city in late-Roman Gaul, pp. 77—
136 Krautheirner 1980, 75. See also Ward-Perkins 1984, 220f. 98).
137 We might however note Barnish’s emphasis upon the frequency of private fora in 139 C. Mango, Byzantium. The empire of New Rome (London 1980); id. Le
fifth-century Rome: Barnish 1989, 385; id., PBSR 55 (1987), 166. développeinent urbain de Constantinople, IVe—Vile siècles (Paris 1985).

98 99
 Towns in Late Antiquity: lot Caesarea and Its Context Civic Centres In Late Antiquity: The Wider Context

certainly crucial. The forum was not only the political and (often) the
commercial centre of the city; it was also usually its religious heartland as
Rome: intramural Christian well, especially of official cults like that of the Capitoline Triad. When
centres, c. AD 500 Constantinian church Theodosius became emperor, he soon initiated what was to become a
++ church fanatical assault upon paganism. There were attempts to contain sacrifices
. titulus (meeting place)
by legislation enacted in 381 and 385 and then, on February 24th 391, an
order closing all temples and wholly prohibiting sacrifices. The penalties for
non-compliance were savage, and careful provision was made for the
enforcement of the law.141 It was a decisive moment in the history of
Christianity.
Paganism did not of course wither and die; the hold of the old gods was
much too strong for that. There is also evidence that, under Eugenius, the
law was less firmly pressed in the West, where he was Augustus.’42
However, it is inconceivable that such punitive measures did not have a
significant effect, especially as far as temples were concerned, and not least
those prominently situated in town centres. It seems not at all surprising that
major churches were often located far away from their pagan counterparts,
and that many of the populace should turn their backs upon that aspect of
the past. In north Africa, when the phenomenon is so pronounced,
Christianity was deeply entrenched well before the Edict of Toleration. In
AD 256, there were at least 130 bishops in north Africa, and a distribution
map of churches founded before Diocletian’s persecutions of AD 304 shows
a heavy clustering in eastern Algeria and Tunisia. This is precisely where
most of the north African sites that we have discussed are located, a matter?
0 1km
which may not be coincidental.143
Christianity appears, therefore, to have exerted a powerful force in the,
shaping of the late-antique urban landscape, especially in north Africa, but
also elsewhere, such as Italy (and Gaul). Churches became the major form
Figure 47 Plan of late-antique Rome (after Krautheirner). of public architecture in the later fourth and fifth centuries, and often acted
as magnets, drawing the populace around them (a process surely encouraged
may be an instance where there was pressure to maintain a long-settled, by the clergy, as power fell increasingly into their hands). When coupled
nodal point of the city in existence.140 with the decline of town councils, and the virtual disappearance of private
But these are exceptions: elsewhere the pattern seems to be different. Yet patronage for the construction or maintenance of public works, a clear
the reasons behind the abandonment and destruction of fora at so many
other towns were undoubtedly complex. The victory of Christianity was 141 Jones 1973, 167f.
142 Jones 1973, 168. In 399, Honorius forbad the closing of temples (Cod. Theod.
16. 10.19): Lepelley 1992, 59, fn. 59.
140 Gros 1985, esp. 147; Deneauve 1990. Also Clover 1982, 8—9; Christol 1992. Note 143 R. Lane-Fox, Pagans and Christians (Harmondsworth 1986), 272 and map, 273—4.
the evidence for the survival of the Imperial Cult in Vandalic times (Lepelley 1979, Note, however, Lepelley’s insistence that this was the most Romanised, and
362—9). urbanised, part of the Maghreb (1992, 65).

100 101
 Towns in Late Antiquity: Jol Caesarea and Its Context

context emerges for the archaeological evidence for the decay of civic
centres.
Changing patterns of commercial activity may also have played their part. 4
We have sketched out the way in which the colonnades which lined the
principal streets of so many of the eastern cities were often converted into
shops and stalls in the fourth and fifth centuries. Open markets turned into CHANGE IN LATE AIVTIQ UITY: AN ITALIAN EXAMPLE
linear ones, eventually evolving into the Islamic souk. Explaining this is not
easy, but it may be related to a declining appreciation of the value of public
squares and lofty colonnades, especially if there was an increasing It would be perverse not to add a postscript about an Italian project which
reluctance on the part of private sponsors to maintain these monuments. drew much inspiration from the work at Cherchel. The site in question, the
Also of relevance are the changing patterns of trade in antiquity. This is a Mola di Monte Gelato, lies about 32 km to the north of Rome, in the
matter that remains full of controversy, especially in terms of the scale and volcanic terrain of South Etruria. It was excavated between 1986 and 1990,
nature of commerce in the period.144 We can make no contribution to that with the specific intention of examining the late history of what seemed to
debate here, but it is worth recalling again the heavy involvement of the be a Roman villa with traces of early medieval occupation. Given that there
Church in business activities. As Whittaker puts it: ‘the Church’s role as ai was apparently a well-preserved stratigraphy, there seemed to be every
agent of redistribution was frequently massive and always increasing’.’4 chance of recovering a rural sequence to compare with those from urban
This must surely have been another factor in drawing away from the forum centres like Cherchel (Figure 48).
and the agora the commercial activities that had for so long focused upon In the event these expectations were more than fully met, and provide a
the civic centre. fascinatingly complementary picture. It is therefore well worth highlighting
Cumulatively, therefore, a combination of the closing of temples, the the main points, albeit in very summary form.146 The site lies on a paved
creation of new Christian foci, the decline of private patronage and the country road which joined with the Via Cassia further south, and thus
reorientation of market processes all helped to bring about the demise of provided good communications with the city of Rome. Founded in the
many fora, especially in the late fourth and early fifth centuries. Where they Augustan period, when an elegant courtyard building was constructed, it
did survive, the Church may have played a crucial role in ‘Christianising’ the was graced by fine statuary and other marble items, such as an elegant
surrounding buildings. On the scanty evidence available this did not happen labrum (ornamental basin). There was also an elaborate tomb monument,
commonly. In some areas, the sites of fora appear to have been revived as found in a late-Roman lirnekiln, of four freed people; one, C Valerius
market places in the later centuries of the first millennium, although this has Faustus, describes himself as a cattle merchant and head of the Imperial Cult
not been tested archaeologicallv. It must remain a prime target for future at the town of Veii, 18 km to the south)47 Whatever the status of the site at
//
research. But, whether described as urban change or urban decline, there Monte Gelato, villa or vicus, it was evidently patron ised by people of some
seems no doubt that the centres of Roman cities looked very different in late / social and political consequence.
antiquity from their counterparts of two or three centuries before. In the second century a small bath-house was added, and also a
substantial mausoleum, situated on a low hill on the other side of the paved
road. The community was by now becoming more cosmopolitan, since two
144 Wickham 1989 provides an excellent overview. See also his review of A. Giardina Greek potters are attested on a waster of a drinking cup, and there are
(ed.), Società romana e impero tardoantico. III. Le merci. Gil insediamenti (Rome
Ban 1986), in JRS 78 (1988), 183f. 146 Interim reports include Arch. Med. 15 (1988), 253-312; and Arch. Med. 16 (1989),
145 Whittaker 1983, 178. Although we have stressed the rise of the Church (and 103—20. See also T. W. Potter, Power, politics and territory in southern Etruria. In
government) as the principal patrons in late antiquity, current research is tending to B. Herring, R. Whitehouse and I. Wilkins (eds.), Papers of the Fourth Conference on
stress the survival of private euergetism: e.g. JRS 66 (1986), 144; C. Roueché, Italian Archaeology 2 (1991), 173—84.
Aphrodisias in late antiquity (London 1989). 147 C. M. Gilliver, PBSR 58 (1990), 93—6.

102 103
 Towns in Late Antiquity: lot Caesarea and Its Context Change In Late Antiquity: An Italian Example

MOLA DI MONTE GELATO 1986-90: phases several other Greek graffiti as well and an unusually high proportion of
Greek wine amphorae.’48 Towards the end of the second century, however,
there were significant changes Cisterns and a latrine were filled in with
1. Augustan 2. c.AD100
cisterns refuse, including the marble labrwn and the top of a marble vase, and the
indications are that the place no longer had the elegance of before. As time
mausoleum
-; went on it became still more run down.Wooden bins were built into one
section of corridor (Figure 49); rooms ere modified and turned over to
metal-working and, in one case, perhaps a byre; and a large lime-kiln was
constructed. Then, in the later fourth and fifth centuries, hearths and rough
road road floors began to accumulate over some of the Roman cocciopesto surfaces;
baths these might once have been classified as ‘squatter occupation, but the
L construction of a small church in one corner of the site suggests otherwise.
road
We may here be glimpsing a rather typical, fairly impoverished late-Roman
community, coinciding chronologically with the demise of most urban civic
3 c.AD300-550 4. late 6th-8th C. gentres.
Between about 550 and 575, on good coin evidence, there was a

Ft
postholes
substantial collapse of roofs and walls. The church may have survived, for
there are burials of this period; but the only other traces of continued
habitation are a few postholes, some of which appear to have been propping
up a Roman wall. Virtually no material culture can be associated with this
phase, which irresistably recalls both the similarly impoverished,

L
Dcrch

..JILme-klln
D contemporary buildings over the forum at Cherchel, and the wooden
structures of this date in north Italian cities like Brescia and Milan.149
A case could be made for the subsequent complete abandonment of the
site at Monte Gelato. There are, however, no stratigraphical signs for this
(Figure 50), and desperate impoverishment, combined with the use of
5. cAD 800 6. 10th C. organic materials like wood and leather for artefacts, may be a preferable
explanation for what was going on)50 What is not in doubt is that around
AD 800, the church was rebuilt and lavishly decorated with carved marble;
habtaton
DJ there was an adjoining baptistery; and the populace housed itself in rock-cut
ttery kiln caves and old Roman buildings like the mausoleum. Even pottery
production was started up, using a single kiln tucked into the corner of a
cave-, o baptistery DsL Roman room.
church
track
graces
(all phases)
0 10 50 m 148 o Murray, et al., PBSR 59 (1991), 177—96. The amphorae will be published in the
final report by Dr Paul Arthur.
149 Brogiolo 1987. 1989.
Figure 48 Successive phases at the rural site of the Mola di Monte Gelato, near Rome. 150 T. Mannoni, in Randsborg 1989, 152f.

104 105
 Towns in Late Antiquity: Jol Caesarea and Its Context Change In Late Antiquity: An Italian Example

M. east chronology

M, part of east section

tIoor

0 2n

Figure 49 Monte Gelato: grooves and sockets for wooden bins or pens set into the floor and Figure 50 Monte Gelato: hearths and floors overlying a Roman floor surface, and their
wall of a corridor in late-Roman times. (Photo: Kate Warren). chronology. The baptistery entrance walls probably belong to the tenth century.

place in the Carolingian age, not least in Rome, but also much more
The interpretation of this phase has been discussed elsewhere and the widely.152 There was an enormous investment in churches, city walls,
arguments need not be rehearsed in detail. It is quite clear on documentary aqueducts and tne like, and we might wonder whether this provides a
evidence that the site had become a papal farm, belonging in this case to an context for the restoration of fora in places like Terracina and Verona,
estate (domusculta) known as Capracorum, founded by Pope Hadrian I referred to at the beginning of chapter 3. It would seem to make more sensç
(772_95).151 It is just one symbol of the extraordinary renaissance that took

152 P. Delogu, The rebirth of Rome in the eighth and ninth centuries. In Hodges and
151 N. Christie (ed), Three South Etrurian Churches (London 1991), 6f. Hobley 1988, 32—42.

106 107
 Towns in Late Antiquity: lol Caesarea and Its Context

than supposing some shadowy form of continuity in these civic centres from
late-Roman times, for which there is no evidence at all. BIBLIOGRAPHY
L Of Monte Gelato’s later history we shall here say nothing, for it takes us
into a quite different set of questions and problems. t will, however, have
Note: Many other references are given in abbr
become clear that this excavation of a rural site affords a fascinating glimpse eviated form in the notes.
into processes of change which do in many ways seem to mirror those of Alselmino, L., Bouchenaki, M., Carandini, A., Leveau
many towns and cities, not least Rome itself. Whilst one cannot build too , Ph., Manacorda, D, Pavolini,
C, Pucci, G, Salama, P. 1989. Ii castellum del Nado
much upon the results of one modest excavation, as with the findings at r. Storia di unafattoria
Ira Tipasa e Caesarea (I— VI sec. d.c.). Monogra
fie di archeologia Libica 23
Cherchel, it does allow the proposal of hypotheses. Testing these is the (Rome).
challenge of the future. Bacchielli, L. 1981. L ‘agora di Cirene, II i. Mon
ografie di archeologia Libica 15
(Rome).
Balty, J. Ch. 1991. Curia ordinis. Recherches d’arc
hitecture et d’urbanis,ne antiques
sur les curies provincial du monde romain (Brussels
).
Barnish, S. J. B. 1989. The transformation of class
ical cities and the Pirenne debate.
JRA 2, 385—400.
Benseddik, N., Ferdi, S., Leveau, Ph. 1983. Cherchel
(Algiers).
Benseddik, N. and Potter, T. W. 1986. Fouilles
du forum de Cherchel. Rapport
prdliminaire. 4e Supplement au Bulletin d’Ar
cheologie Algerienne (Algiers).
Benseddik, N. and Potter, T. W. 1994. Fouilles
du Forum de Cherchel. Rapport
finale. 6e Supplement au Bulletin d’Archeologie
Algerienne (Algiers).
Beschaouch, A., Hanoune, R., Khanoussi, M.,
Olivier, A., Thébert, Y. 1983.
Recherches archdoiogiques Franc o- Tunisienn
es a Bulla Regia. Collection de
lEcole francaise de Rome 28.1 (Rome).
Brogiolo, G. P. 1987. A proposito dellorgan
izzazione urbana nellalto medievo.
Arch. Med. 14, 27—46.
Brogiolo, G. P. 1989. Brescia: building transform
ations in a Lombard city. In
Randsborg 1989, 156—65.
Bullough, D. A. 1966. Urban change in early med
ieval Italy: the example of Pavia.
PBSR 34, 82—130.
Cavalieri Manasse, 0. 1990. 11 foro di Verona: recen
ti indagini. In La cilIa nell’Italia
settentrionale in eta romana. Collection de lEcole
française de Rome 130,
579—6 16.
Christie, N. 1989. The archaeology of Byzantine Italy:
a synthesis of recent research.
Journal of Mediterranean Arch. 2.2, 249—93.
Christie, N. 1990. Byzantine Liguria: an imperial prov
ince against the Longobards,
A.D. 568—643. PBSR 58, 229—72.
Christol, M. 1992. 1. Bardo. App. 36. Grands trava
ux a Carthage a Ia fin du IVe
siècle et au debut du Ve siècle. Afr. Rom. 9, 337—43.
Claude, D. 1969. Die byzaiu½ische Stadt im 6 Jahrh
undert. Byzantinisches Archiv
13 (Munich).
Clover, F. M. 1982. Carthage and the Vandals. In J.
H. Humphrey (ed.), Excavations
at Carthage 1978 conducted by the University of Mich
igan 7, 1—22.

108
109
 Towns in Late Antiquity: Jol Caesarea and Its Context Bibliography

Crawford. I. S. 1990. The Byzantine shops at Sardis. Archaeological exploration of Gui, 1., Duval, N., and Caillet, J. P. 1992. Inventaire et typologie des basiliques
Sardis 9 (Harvard). chrétiennes d’Afrique du Nord. Collection des Etudes Augustiniennes, Série
e Deneauve, J. 1990. Le centre monumental de Carthage. In Carthage et son territoire.

I
Antiquité (Paris).
Histoire et archeologie de lAfrique du Nord. IVe Colloque (Paris), 143—55. Hodges, R. and Hobley, B. (eds.) 1988. The rebirth of towns in the West, AD 700—
Downey, G. 1961. A histoiy ofAntioch in Syria (Princeton). 1050. (Council of British Archaeology Research Report 68), 16—27.
Dunbabin, K. M. D. 1978. The mosaics ofRoman north Africa (Oxford). Hodges, R. and Whitehouse. D. B. 1983. Mohammed, Charlemagne and the origins
‘ Duval, N. 1964. Observations sur lurbanism tardifde Sufetula (Tunisie). Cahiers de of Europe. Archaeology and the Pirenne Thesis (London).
Tunisie 12, 87—103. Horn, H. G. and Roger, C. R. (eds.) 1979. Die Numider (Bonn).
Duval, N. 1971. Les eglises africaines a deux absides. Les basiliques de Sbeitla. Humphrey, J. H. 1980. Vandal and Byzantine Carthage. In J. G. Pedley (ed). New
Ecoles francaises d’Athènes et de Rome, 218.1 (Paris). light on ancient Carthage (Ann Arbor), 85—120.
Duval, N. 1977. La “basilique” de Zana (Diana Veteranorum). MEFRA 89, 847—73. Hurst, H. R. and Roskams, S. P. 1984. The Avenue du Président Habib Bourguiba,
Duval, N. 1982. L’urbanisme de Sufetula— Sbeitla en Tunisie. ANRW Band II. 10.2, Salammbo: the site and finds other than pottery. Excavations at Carthage:
596—632.
Duval, N. and Baratte, F. 1973. Les ruines de Sufetula. Sbeitla (Tunis).
Duval, P.-M. 1946. Cherchel et Tipasa (Paris).
Fabbricotti, E. 1992. Cambiamenti di vita a luvanum 1. In E. Herring, R.
I the British Mission 1.1 (London).
Jones, A. H. M. 1973. The later Roman empire, 284—602 (Oxford).
Kennedy, H. 1985. From ‘Polis’ to Madina: urban change in late antique and early
Islamic Syria. Past and Present 106, 3—27.
Whitehouse and J. Wilkins (eds.), Papers of the Fourth Conference ofItalian Kennedy, H. 1992. Antioch: from Byzantium to Islam and back again. In Rich 1992,
archaeology 4 (London), 77—82. 1 81—98.
Fentress, E. 1984. Caesarian reflections. Opus 3, 487—93. Kenrick, P. M. 1986. Excavations at Sabratha 1948—1951. JRS Monograph 2
, Février, P.-A. 1977. Towns in the western Mediterranean. In M. W. Barley (ed.), (Gloucester).
A European towns. Their archaeology and early history (London), 315—42. Krautheimer, R. 1980. Rome. Profile ofa City, 312—1308 (Princeton).
Février, P.-A. 1982. Urbanisation et lurbanisme de l’Afrique romain. ANRW Band II, La Rocca, C. 1986. “Dark Ages” a Verona. Edilizia privata, aree aperte e strutture
10.2, 32 1—96. pubbliche in una citta dell’Italia settentrionale. Arch. Med. 13, 3 1—78.
Février, P-A. 1986. Aux origines du Christianisme en Maurétanie Césarienne. La Rocca, C. 1992. Public buildings and urban change in northern Italy in the early
MEFRA 98, 767—809. medieval period. In Rich 1992, 16 1—80.
Février, P.-A. 1990. Approches du Maghreb romain. Pouvoirs, dffe’rences et conflit. Lancel, S. 1971. Tiasa de A’taurétanie (Algiers).
2 vols (Aix-en-Provence). Lancel, S. 1982. Tipasa de Maurétanie I. ANRW Band II. 10.2, 739—86.
Foss, C. 1979. Ephesus after antiquity: a late antique, Byzantine and Turkish city Lassus, J. 1972. Portiques d’.4ntioch. Antioch-on-the-Orontes 4 (Princeton).
(Cambridge). Lepelley, Cl. 1979, 1981. Les cites de l’Afrique romaine an Bas-Empire, 2 vols.
Fulford, M. G. and Peacock, D. P. S. 1984. The Avenue dii Président Habib (Paris).
Bourguiba, Salammbo: the pottery and other ceramic objects from the site. Lepelley, Cl. 1992. The survival and fall of the classical city in late Roman Africa.
The British Mission 1.2 (London). In Rich 1992, 50—76.
Golvin, J.-C. and Leveau, Ph. 1979. L’arnphithébtre et le théâtre-amphitheâtre de Leveau, Ph. 1982. Les maisons nobles de Caesarea de Maurétanie. Ant. Afr. 18, 109—
Cherchel: monuments a spectacles et histoire urbaine a Caesarea de 65.
Maurétanie. MEFRA 91, 817—43. Leveau, Ph. 1984. Caesarea de Maurétanie. Une yule romaine et ses campagnes.
Greco, E. and Theodorescu, D. 1980. Poseidonia — Paestun2 I. La “Curia’ Collection de l’Ecole francaise de Rome 70.
Collection de l’Ecole francaise de Rome 42. Leveau, Ph. and Paillet, J. L. 1976. L’alimentation en eau de Caesarea de
Gros, P. 1985. Byrsa III. Rapport sur les campagnes defouilles de 1977 a 1980: Ia Maurétanie et l’aqueduc de Cherchel (Paris).
basilique orientale et ses abords. Collection de lEcole francaise de Rome 41. Liebeschuetz, J. H. W. G. 1972. Antioch. City and imperial administration in the
Guéry, R. 1981. Loccupation de Rougga (Bararus) daprès Ia stratigraphie du forum. later Roman empire (Oxford).
BAC 17B, 91—100. Liebeschuetz, J. H. W. G. 1992. The end of the ancient city. In Rich 1992, 1—49.

110 111
-r -

Towns in Late Antiquity: lol Caesarea and Its Context

Bibliography

Liebeschuetz, I. H. W. G. and Kennedy, H. 1988. Antioch and the villages of Ward-Perkins, B. Mills, N., Gadd, D.
northern Syria in the fifth and sixth centuries AD: trends and problems.
,
and Delano-Smith C. 1986. Luni
Ager Lunensis: the rise and fall of and the
Nottingham Medieval Studies 33, 64—90; = Liebeschuetz, From Diocletian to a town and its territory.PBSR 54,
Ward-Perkins, i. B. 1974. Cities of ancient 81—146.
theArab Conquest (Aldershot 1990), chp. XVI. Greece and Italy: planning in classical
antiquity (New York).
Mahjoubi, A. 1978. Recherches d’histoire et d’archéologie a J-Jenchir el-Faouar Whittaker, C. R. 1983. Late Roman trade
(Tunisie,) Tunis. and traders. In P. Garnsey, K. Hopkins
C. R. Whittaker (eds), Trade ni the and
Marec, E. 1954. Le forum d’Hippone.Libyca2, 363—418. ancient economy (London), 163—80.
Wickham, C. 1981. Early medieval
Mattingly, D. J. and Hayes, J. W. 1992. Nador and fortified farms in north Africa. Italy. Central power and local
1000 (London). society, 400—
JRA 5,408—18. Wickham, C. 1989. Italy and the early
Maurin, L. 1967. Thuburbo Maius et Ia paix vandale. Cahiers de Tunisie 15, 57—60, Middle Ages. In Randsborg 1989,
l40—151.
240—50.
Pedley, J. G. 1990. Paestuni. Greeks and Romans in southern Italy (London).
Pensabene, p. 1982. Les chapiteaux de Cherchet Etude de Ia decoration
architectonique. 3 e Supplement au Bulletin d’Archëologie Algérienne
(Algiers).
Poinssot, C. 1958. Les ruines de Dougga (Tunis).
Pringle, D. 1981. The defence of Byzantine Africa from Justinian to the Arab
Conquest. British Archaeological Reports, International Series 99 (Oxford).
Randsborg, K. (ed.) 1989. The birth of Europe: archaeology and social development
during the first millennium AD. Analecta Romana Instituti Danici,
supplement (Rome).
Rich, J. (ed.) 1992. The city in late antiquity (London).
Salama, p. 1988. Vulnérabilité dune capitale: Caesarea de Maurétanie. Aft. Roni. 5,
253—69.
Schmiedt, G. 1973. Citté scomparse e cittã di nuova formazione in Italia in relazione
al sistema di communicazione. Settimana di Studio sull’A Ito Medievo 21,
5 03—607.
Shaw, B. D. 1981. Rural markets in north Africa and the political economy of the
Roman empire. Ant. Aft. 17, 37—84.
Thébert, Y. 1983. L’évolution urbaine dans les provinces orientales de l’Afrique
romaine tardive. Opus 2, 99—132.
Thébert, Y. and Biget, J.-L. 1990. L’Afrique aprés Ia disparition de Ia cite classique.
In L’Aftique dans l’Occident Roinain. 1cr siècle av. J. C. IVe siècle ap. J. C.
-

Collection de l’Ecole française de Rome 134.

Ward-Perkins, B. 1978. Luni: the decline and abandonment of a Roman town. In H.
Blake, T. W. Potter and D. B. Whitehouse (eds.), Papers in Italian
Archaeology I British Archaeological Reports, International Series 41
(Oxford), 3 13—21.
Ward-Perkins, B. 1981. Two Byzantine houses at Luni. PBSR 49, 91—8.
Ward-Perkins, B. 1984. From Classical Antiquity to the Middle Ages. Urban
public building in northern and central Italy, AD 3 00—850 (Oxford).
Ward-Perkins, B. 1988. The towns of northern Italy: rebirth or renewal? In Hodges
and Hobley 1988, 16—27.

112
113

-r