Project Muse 772168

The Middle Euphrates and Its Transformation from the Third
to the Seventh Centuries: The Case of Dibsi Faraj
Anna Leone, Alexander Sarantis
Journal of Late Antiquity, Volume 13, Number 2, Fall 2020, pp. 308-351
(Article)
Published by Johns Hopkins University Press

DOI: https://doi.org/10.1353/jla.2020.0025
For additional information about this article

https://muse.jhu.edu/article/772168
[ Access provided at 14 Nov 2022 12:22 GMT from Durham University ]

Anna Leone and Alexander Sarantis
The Middle Euphrates and

Its Transformation from the
Third to the Seventh Centuries:
The Case of Dibsi Faraj
Dibsi Faraj is a fortified citadel situated on the middle reaches of the Euphrates
River in modern Syria which was occupied until the ninth century, when it was
abandoned and then reoccupied in the twelfth and thirteenth centuries. This
article examines the series of fortification works, churches, and bathing com-
plexes that were erected at Dibsi Faraj during the late Roman and early Byzan-
tine periods. The wider political and religious contexts for these works include
the wars fought between the eastern Roman and Sassanid Persian empires and
their Ghassanid and Lakhmid allies, as well as the growing fame and popular-
ity of the shrine of Saint Sergius at Resafa. Among the most interesting finds
are fragments of the Anastasian military inscription, known also from sites in
Palestine, which confirms the presence at Dibsi Faraj of a military garrison,
probably comprising limitanei troops, in the early sixth century. Continued
activity at the site across the seventh century reinforces the argument that the
last eastern Roman-Sassanid war and early Islamic conquests did not pro-
foundly damage the society and economy of northern Oriens.
The remains of the fortified settlement of Dibsi Faraj bear witness to an

important military and cultural settlement on the Middle Euphrates (modern
Syria) in Late Antiquity. This paper will present new insights into unpublished
material from this site, excavated by Richard Harper, then deputy Director
of the British School at Ankara, in the 1970s. Harper’s excavation took place
We would like to thank all the authors who have contributed to the study of the finds and the
stratigraphic sequence at Dibis Faraj: Agnès Vokaer, Coralie Clover, Philip Kenrick, Massimiliano
Munzi, Robert Taylor-Wilson, Mark Jackson, Nairusz Haidar Vela, Victoria Leitch, Denys Feissel.
We would like in particular to thank Denis Feissel for the reading and the integration of the inscrip-
tion of Anastasus’s law that is mentioned in this article. Finally, we would like to thank the Journal
of Late Antiquity’s three anonymous reviewers for their constructive comments. Any remaining mis-
takes are our own. This paper is dedicated to the memory of Richard Harper and Tony Wilkinson.
308 Journal of Late Antiquity 13.2 (Fall): 308–351 © 2020 Johns Hopkins University Press
LEONE &SARANTIS ^ The Middle Euphrates and Its Transformation 309
in collaboration with the Department of Antiquities of Syria, as one of the

activities conducted in preparation for the construction of the Tabqa Dam,
completed in 1973.
The following discussion of the building history of Dibsi Faraj will con-
tribute to a fuller picture of its appearance, function, and possible popula-
tion in different phases. As well as discussing what Harper’s evidence tells us
about the fortress, the paper will, for the first time, set this material in a wider
regional context, using the site as a case study for a discussion of the military,
infrastructural, and cultural history of the Middle Euphrates region in Late
Antiquity. This was in many ways a unique limes zone, the Middle Euphrates
carving its way through the northern Syrian steppe landscape, running from
west to east following the great bend at Barbalissus.1 Its strategic importance
stemmed from the fact that it both connected and, because of its fortifica-
tions, separated Roman and Persian territories. At the same time, it bisected
the Upper Mesopotamian frontier region north of the river from the limitanei
and federate Ghassanid controlled northern Syrian frontier zone to the south.
Here, the focus will, in particular, be on the stretch of the river east of Dibsi
Faraj as far as Circesium, and the Strata Diocletiana road, which ran south
of Sura, connecting the Euphrates with Resafa and then Palmyra in the south.
This overview of the area across Late Antiquity, tying archaeological evi-
dence for the built landscape to historical sources for military and political
events in the region has not been attempted to date. Instead, studies of the area
have tended to focus on individual sites, concentrate on the earlier Roman
period, or synthesize archaeological evidence from the eastern frontier regions
as a whole.2 This paper aims to fill this gap, bringing new archaeological
material to light and using it to answer broader historical questions concern-
ing the strategic importance of the Middle Euphrates frontier, the nature of
its elites, the organization of defense and frontier military resources, and reli-
gious life. It is to the identification of the site that we first turn.
The Location and Identification of Dibsi Faraj

The fortified citadel of Dibsi Faraj3 was located 17 kilometers east of Barbalis-
sus-Balis, occupying a limestone area at the east end of the Dibsi plateau from
1
Geography: Millar 1993, Chapter 12.1; Comfort 2008, 9–11 and 131. Geography of the
Roman-Persian frontier zone: Naval Intelligence Division 1942, 19–194.
2
Chapot 1907 is the main earlier synthesis. Books including substantial sections on the Middle
Euphrates in the Roman and late Roman periods: Edwell 2008; Millar 1993, Chapters 4, 5, and
12; Pollard 2000. On specific sites: Lauffray 1983. Discussion of the Strata Diocletiana: Konrad,
Baldus, and Ulbert 2001. Synthesis of the literature: Sarantis 2013a, 348–52.
3
Dibsi Faraj is the modern name of the site, but we do not know how and when the nearby mod-
ern settlement acquired this toponym.
310 Journal of Late Antiquity
which it controlled access to seasonal grazing areas and fertile agricultural

land. It was also thus on the route to the ancient site of Resafa-Sergiopolis,
named after Saint Sergius who was martyred there.4
The site has been identified with the ancient settlement of Athis which was,
allegedly, renamed Neoceasarea in Late Antiquity. Although some scholars
have accepted this identification uncritically, it remains debatable. According
to the argument, the site’s earliest name was Athis.5 Ptolemy registers Athis
as the only site between Barbalissus and Sura, in the region of Chalyboni-
tis.6 Meanwhile, the section of the Peutinger Table showing the road between
Palmyra and Resafa indicates that a site named Attas was located twelve
kilometers from Barbalissus.7 However, the Table’s depiction of this section
of the Euphrates is heavily distorted, all of its distances and locations being
incorrect.8
Other scholars have identified Dibsi with ancient Thapsacus, based on
their interpretation of a passage in Pliny’s Natural History and the narrations
of Alexander’s crossing of the Euphrates.9 Their argument is founded primar-
ily on the testimony of Arrian, which informs us that Alexander ordered a
fleet to be built in Phoenicia and Cyprus and brought overland to Thapsacus.10
It has thus been reasoned that Thapsacus could not have been far from the
Mediterranean coast, and this supports its identification with Dibsi, one of
the closest points on the Euphrates to the Levantine seaboard. An additional
argument is that the toponyms Thapsacus and Dibsi are vaguely similar pho-
netically, specifically in the sound “psi.”
This second hypothesis has been refuted on the basis that Thapsacus had
a long history of occupation stretching back to the Hellenistic period, whereas
4
Reasons for the importance of the cult of Sergius in the East: Haarer 2006, 37. Resafa-Sergiop-
olis is about 70 km south of the site of Dibsi Faraj.
5
Harper 1975, 321–24. Also in favor of the identification of Dibsi with Athis: Dussaud 1927,
453. According to Strab. Geog. 16.1.27, this was a place where the crossing of the river was easy.
Dibsi Faraj is indeed in an area where the crossing was possible in some seasons.
6
Ptol. Geog. 5.15.17.
7
Tabula Peutingeriana, Segment 11.2 (thirteenth-century maps of the Roman roads probably as
they were in the fourth century).
8
On using the medieval Peutinger Map for interpreting Late Antiquity: Salway 2005. General
discussion on the composition and date of the table: Talbert 2010.
9
Plin. HN 5.22. Alexander’s crossing of the Euphrates: Arr. Anab. 3.7, 7.19.3. Identification of
Dibsi Faraj with Thapsacus: Chapot 1907, 204, note 4; Bell 1910; and Honigmann 1934. In Der
Neue Pauly 12.1: 242, Thapsacus is identified with Qal’at Nagm. Discussion of the evidence and
proposed identification: Gawlikowski 1996.
10
Arr. Anab. 3.7.1 and 7.19.3: “Aristobulus says that he found at Babylon the fleet with Nearchus,
which had sailed from . . . and another which had been conveyed from Phoenicia, consisting of the
Phoenincian quinqueremes, three quadriremes, twelve triremes and thirty triacontors. These had
been taken to pieces and conveyed to the river Euphrates from Phoenica to the city of Thapsacus.”
Figure 1. Map of the Region (by Abby George)
archaeological evidence from Dibsi indicates that the site was in use only
from the first century ce and lacks a clear Hellenistic occupation phase.11
Bronze Age material found west of the site certainly suggests the existence
of an earlier settlement. However, even though the existence of a Hellenistic
settlement located in an area not explored at the time of the excavation cannot
11
Harper 1975, 321.
be ruled out, the lack of Hellenistic material from either this or the main site
casts doubt on its identification with Thapsacus. The Athis identification also
remains open to debate in light of the lack of evidence of an early Imperial
settlement in the excavated area sufficiently monumental to attract the atten-
tion of geographers as eminent as Pliny and Ptolemy.
As for Late Antiquity, it has been suggested that Athis was renamed Neo-
caesarea during the third century, when it received monumental fortifications
for the first time.12 The change of name is attributed to the imperial govern-
ment, perhaps due to the institutional upgrade of what had been a mere vil-
lage at Athis. While there is no evidence to support this hypothesis, we do
have a reference in the fifth-century Patrum Nicaenorum nomina to a Bishop
Paulus of Neocaesarea under the heading “VII – Provincia Syria Coele,” as
well as the testimony of George of Cyprus, who refers to its location in the ter-
ritory of Saint Sergius’s martyrdom.13 This would at least place Neocaesarea
in approximately the same region as Dibsi.
Moreover, Neocaesarea is mentioned by Procopius as one of the Mid-
dle Euphrates sites that Justinian fortified “on the furthest borders of
Euphratesia.”14 Following his descriptions of Zenobia, Sura and Resafa, Pro-
copius lists here the settlements of Barbalissus, Neocaesarea, Gaboulon, Pen-
tacomia, and Europos before digressing to describe work at Hemerium and
Hierapolis.15 He then discusses refortification work at Zeugma and Neocae-
sarea, which he refers to as fortified towns.16
This section of the text thus mentions a series of Euphrates sites between
Sura and Zeugma in no particular order, in addition to Hierapolis and Heme-
rium, the location of which is unknown but which presumably was situated,
like Hierapolis, away from the river. The Middle Euphrates sites that have been
located with some certainty are all located on the southward-flowing stretch
of the river, from Zeugma to Barbalissus (modern Balis), near the Euphrates
bend: in addition to Zeugma, which is comparatively well researched, Gabou-
lon has been identified with modern Jabbul and Europos with Jerablus.17
While neither Neocaesarea nor Pentacomia has been located, it would
make sense for these to have been situated along the river east of Barbalis-
sus, a strategically vital region which would otherwise have been neglected,
12
Harper 1975, 322.
13
On the Patrum Nicoenorum Nomina, see Paulus 63 (ed. Gelzer et al., Teubner), and for the
edition, see Gelzer, Hilgenfeld, and Cuntz 1995. Dibsi Faraj was located on the pilgrimage route to
Resafa. For discussion of the statement of George of Cyprus, see Harper 1975, 322.
14
Proc. Aed. 2.9.10. Modern discussion: Harper 1975, 321–22.
15
Proc. Aed. 2.9.10–17.
16
Proc. Aed. 2.9.18–20.
17
Identification of Europos with Jerablus: Chapot 1907, 280. Example of work on Zeugma:
Comfort and Ergeç 2001; Gawlikowski 1996. Balis: Golvin 1945.
uncharacteristically, by Procopius.18 This inference is reinforced by the texts

mentioned above, which place Neocaesarea in the vicinity of Resafa and
therefore en route to the Strata Diocletiana.
While Dibsi Faraj could theoretically have been a fortress not mentioned
in our textual sources, its imposing fortifications and hilltop location suggest
that it played an important strategic role in Late Antiquity, thereby adding
weight to its association with either Neocaesarea or Pentacomia. Indeed, while
neither site is mentioned by Procopius in his Wars narrative on the Roman-
Persian conflicts in the reign of Justinian, his inclusion of both in the Buildings
suggests that they had been the focus of some sort of imperial expenditure.
The Archive of the Excavation at Dibsi Faraj

The discovery of the archive of unpublished excavation materials from Dibsi
Faraj was fortuitous. It included 128 boxes of pottery and glass, over 500
photographs and negatives, around eighty plans, all the excavation notes, cat-
alogs of finds, and various preliminary reports. The excavation project took
place between 1970 and 1973 and was mainly funded by Dumbarton Oaks,
with some contribution from the Kelsey Museum at the University of Michi-
gan. It was part of a UNESCO-coordinated effort to salvage and document
as many sites as possible before the surrounding area was inundated by water
due to the creation of Lake Assad on the Euphrates, the so-called Tabqa Dam
project.19 Richard Harper, who was the director of the excavation, initially
shipped some of the finds and the entire excavation archive to London in
1973. This archive of research materials ended up in Harper’s garage in Dur-
ham for over thirty years before it was donated by Harper’s daughter (Eleanor
Glenton) to the Department of Archaeology at the University of Durham in
2013. Tony Wilkinson, who had worked as a young geographer at the site,
brought the archive to Durham and started to work on it with Anna Leone. 20
After five seasons of excavation, three preliminary reports were produced,
one appearing in the 1975 issue of Dumbarton Oaks Papers, and two fur-
ther papers on some of the Roman to Islamic period finds, as well as work at
another Euphrates fort flooded by the dam, Pağnik Öreni.21 However, the full
excavation has never been published. The site’s historical importance, com-
bined with its current inaccessibility, makes full analysis and publication vital.
18
Procopius’s tendency to exaggerate building work in his coverage of the East in the Buildings:
Ulbert 2000.
19
Freedman and Lundquist 1979.
20
Tony Wilkinson sadly passed away a few year later, leaving the legacy of the archive to Anna
Leone.
21
Harper 1974a; Harper 1974b; Harper 1975. Finds in Pağnik Öreni: Harper 1977; Harper
1980.
The Citadel and Its Phases

The site presents different phases of occupation, with the most important
being concentrated between the third and the seventh centuries. The earliest
period dates to the first century ce and is represented by a rock-cut funerary
chamber which had mostly been destroyed by the later setting of the city wall
and other small walls. The original function of the city is hard to identify. The
latest phase is well attested archaeologically by one large L-shaped building
and some domestic structures. Evidence of occupation suggested that the site
continued to be inhabited until 895, when it was abandoned following an
earthquake, before being occupied again in the twelfth and thirteenth centu-
ries.22 These earlier and later phases of occupation will be not discussed in the
article, the focus of which will be the late Roman to early Byzantine periods.
The excavation extended over eight distinct areas (Area 7 does not exist,
but there is an area 0). The excavation was conducted in two ways: 1) in
Wheeler boxes and elongated trenches, and 2) by extensive excavation of the
buildings through the open-area method.23 The article will discuss the differ-
ent phases of occupation rather than the different areas.
The First Fortification at the End of the Third Century

The first significant building phase at Dibsi Faraj is represented by the con-
struction of the citadel in the Diocletianic period at the end of the third cen-
tury. The chronology of the first phase of fortification is confirmed by the
stratigraphic sequence.24 This initial fortification surrounded the hilltop and
was characterized by square interval towers and four corner towers (Figure 2).
The intramural area was around five hectares in size. A lower-lying site was
22
The analysis of the later phase of occupation will be included in the forthcoming volume pub-
lished by Dumbarton Oaks: Leone forthcoming, as will a recent reconsideration of the finds, which
has shown that there was a second Islamic phase in the twelfth and thirteenth centuries. Discussion
of the Islamic phases: Harper 1975, 324.
23
For an analysis of the different methodologies, see Sanders, James, and Carter Johnson 2017.
24
A similar chronology has also been suggested, for instance, for the inner fortification of
Palmyra. On the different phases of reconstruction and restorations of the walls of Palmyra, see
Chapot 1907, 284; Zanini 1995. For further bibliography on issues of dating the fortress of Pal-
myra and the building techniques used, characterized by large blocks and a rubble core, see Juch-
niewicz, As’ad, and al-Hariri 2010. On Diocletianic frontier work: Arce 2015, 102. See also notes
36 to 38 below. In reference to Dibsi, a detailed publication of the stratigraphic sequence will be
provided in the full publication of the excavation. Traces of the earlier phases of the citadel have
been identified in several areas of excavation. Across the site, it was discovered that the primary
wall was fairly deeply founded, wherever possible set into the natural limestone bedrock and built
probably exclusively with massive blocks of the same soft limestone. The Byzantine renovation of
the wall probably early in the reign of Anastasius was essentially a re-facing exercise: in general, the
remains of the original structure were found to either underlie or be used effectively almost as an
inner wythe (a vertical unit of walling) of the defensive circuit when it was renewed.
Figure 2. Plan of the Citadel with excavated areas (Harper 1975), Areas 8 and 9. Extra-mural basilica is outside of this map; there
are no plans indicating its exact position
located outside of the citadel and covered an area of around twenty hectares.
This extramural area was later occupied by a modern settlement and excava-
tion there was only limited. It did nevertheless include the discovery of an
extramural church and a building with a mosaic, of which only one room was
excavated and the function of which is hard to determine.25
Seventeen interval towers of the wall were identified out of the thirty-four
that were probably part of the full fortification. A pair of towers also pro-
tected each of the four identified gates. The external face of one of the small
interval towers was exposed on the west wall. The main gate was positioned
on the southern citadel wall.
From its hilltop position the citadel dominated the surrounding landscape
as well as the Euphrates, the right bank of which it overlooked. This adapta-
tion of a fortress to elevated terrain resulting in its irregular form is typical
of later Roman fortifications.26 The peculiarity of Dibsi Faraj in compari-
son with the architecture of other late antique fortifications is the presence
of rectangular towers.27 These were not normally considered to be as defen-
sively effective as square, circular, or multi-angular towers because their long
straight edges were more vulnerable to flanking attacks by battering rams and
other siege engines.28 This may be one of the reasons for the renovation of
these structures in the next phase of building work, discussed below.
Inside the wall, a building dating to the same period as the fortification
was excavated in Area 1. It was located adjacent to the northern defenses,
to the east of the north-western gateway. The building was organized in two
parts. The northern part was probably the principia, the headquarters of the
fort, while the southern area possibly served as accommodation for a garrison.
The entrance to the south was located in the vicinity of two latrines and led
to a central room paved with mosaics.29 Three apsed rooms paved with mosa-
ics were located west of this hall. The southern part of the building was only
partially excavated but it contained a large apsed area paved with limestone.
The construction of the citadel complex should probably be viewed
in connection with the strengthening of the frontier region following the
25
The excavation of this extramural structure is not mentioned in Harper 1975.
26
Differences between early imperial and Late Roman fortifications with bibliography: Sarantis
and Christie 2013, 255–61.
27
In the Roman period, square towers appear to have been more common in the eastern frontiers
(Gregory 1995, 167). On projecting towers, see Johnson 1983, 37.
28
Johnson 1983, 38–50. Siege warfare tactics in Late Antiquity: Whitby 2013.
29
It needs to be stressed here that the nature of the excavations conducted by trenches very rarely
uncovered entire buildings. This makes it particularly difficult to fully reconstruct the plans of the
buildings.
Figure 3. Aerial Photograph of the Citadel, in the lower left, overlooking the Euphra-
tes (Harper’s archive)
Roman-Persian conflicts of 297 and 298.30 After his defeat near Callinicum in
297, the Caesar Galerius defeated the Persians in their territory and captured
King Narseh’s baggage train, treasury, and royal harem in 298.31 Galerius
led his army back to Roman territory along the Middle Euphrates. This cul-
minated in an extremely favorable treaty for the Romans, confirming their
occupation of Upper Mesopotamia as far as the Tigris basin and giving them
five client states on the opposite banks of the Tigris in the Mesopotamian-
Armenian frontier zone.
30
Leadbetter 2009, 88–96.
31
Galerius’s campaign and subsequent treaty: Millar 1993, 177–81; Dignas and Winter 2007,
84–88; Leadbetter 2009, 88–96. Contemporary accounts of Galerius’s campaign: Dodgeon and
Lieu 1991, 125–34.
Figure 4. Aerial photograph of the Principia at the end of the excavation, during the
inundation (Harper’s archive)
This agreement did not affect Roman control of the Euphrates, the east-
ern limits of which remained Circesium. However, the strategic importance
of the Euphrates region had become clear during the various third-century
Roman-Persian wars. In particular, both empires had proved to be vulner-
able to attack along this river, which connected the wealthy Roman north-
ern Syrian and Persian Lower Mesopotamian provinces.32 In 252/53, Shapur
I invaded Roman Syria along the Middle Euphrates, capturing a series of
Roman bases, including Sura, before defeating a Roman army at Barbalis-
sus and then sacking northern Syrian cities, such as Hierapolis, Chalcis, and
Antioch.33 Zeugma on the western stretch of the Middle Euphrates also was
32
Third-century Roman-Persian military campaigns: Millar 1993, 141–73; Dignas and Winter
2007, 70–88; Edwell 2008, Chapter 5; Börm 2016, 617–26; Hartmann 2017, 1047–67, especially
1048–51 and 1054–57.
33
Contemporary accounts: Dodgeon and Lieu 1991, Chapter 3.
captured. Subsequent raids reduced Dura and Circesium, the former never
recovering from its sack in about 256 or 257, as attested by the incredible
archaeological remains of the Persian siege.34 Dibsi’s location along the main
route between Sura and Barbalissus makes it very likely that it was affected by
these events, especially when we bear in mind its lack of defenses.
Diocletian (reigned 284 to 305) invested heavily in the infrastructure of
the Middle Euphrates, as he did in other frontier regions.35 Having failed to
withstand the Persian invasions of 252–256 and 260, bases like Callinicum
and Circesium now enjoyed a substantial upgrade, although there is little evi-
dence that any forts were built between Sura and Circesium.36 It is possible
that a Palmyra-era mud wall existed at Zenobia, but there is no evidence of
renovation at this site prior to the reign of Anastasius.
Diocletian also funded the construction of forts and fortress cities along
the road that became known as the Strata Diocletiana, from Sura on the
Middle Euphrates to Palmyra.37 This defensive system related closely to that
of the Middle Euphrates, designed to protect northern Syria from invasions
emanating from north-eastern Arabia and, beyond it, southern Mesopota-
mia.38 It thus was intended to compensate for the demise of the client state at
Palmyra, which had previously defended north-western Arabia from nomadic
Arab raiders. Legionary bases along this communication included Palmyra,
Resafa, Oresa, and Sura. In each case, archaeological research has revealed
the presence of small military forts integrated or built within the walls of
civilian settlements. What remained of the city of Palmyra, for example, was
now endowed with a military camp as well as a new urban circuit wall which
reduced the intramural area of the city.39 The roughly thirty-hectare settle-
ment at Sura comprised a small square fort and surrounding canabae.40 A
similar situation was true of the quadriburgium at Tetrapyrgium.41 Watch-
towers dotted the road between these bases. Fortification works at the major-
ity of sites consisted of rubble core and mud brick facings in this Diocletianic
phase. Some of these sites, including Cholle, were, like Dibsi Faraj, located on
hilltops or mounds, while others such as Resafa were situated in open terrain.
34
Leriche 1993; James 2011.
35
Summary of the works: Millar 1993, 181–90; Pollard 2000, 71–73.
36
For the Diocletianic fortifications of Circesium, see Amm. Marc. 23.5.1–2.
37
Konrad, Baldus, and Ulbert 2001. Wider bibliographical discussion of these fortifications:
Sarantis 2013a, 348–49 and 351–52.
38
Konrad 1999; Liebeschuetz 1977, 496–97.
39
Overview: Intagliata 2017 and 2018, especially Chapters 5–6. Debate concerning the date of
fortifications: Seyrig 1950; Van Berchem 1954. See also note 24 above.
40
Konrad 1999, 398–400.
41
Konrad 1999, 400–404.
The location of Dibsi presumably was determined by the topography of the

area overlooking the river, providing easy access to the water.
The fortification of Dibsi Faraj complements this picture by providing
physical evidence of Tetrarch-era Middle Euphrates fortification work west
of the immediate frontier bases of the Strata Diocletiana and Circesium. This
confirms the importance of defending the entire Middle Euphrates route to
Syria in light of the earlier Persian attacks mentioned above. Like the bases
just mentioned, Dibsi Faraj contained a garrison with a military headquar-
ters. However, it differs in its lack of a small, quadrilinear fort, structured
around a typical cruciform street grid system, contained within the walls of a
larger civilian settlement. Instead, this was a larger, irregularly fortified settle-
ment, encompassing military and non-military structures. This might suggest
a closer integration of military and civilian populations, a trend that would
intensify across the empire in Late Antiquity.42
The Fourth- and Fifth-century Interventions:

The Period of Major Expansion
The fourth century saw another phase of major expansion and monumentaliza-
tion of the site. At the beginning of the century, construction started on a large
residential complex in Area 0. The building was left unfinished, though, and in
the second half of the century a large basilica was built inside the walls which
cut across this residential area. The basilica had different phases of transfor-
mation. The original plan was characterized by three naves and two external
porticoes, running down the longer sides of the church. A second phase saw
the enlargement of the structure into a five-aisled church and the addition of a
small apse, to the side of the main apse, where relics were most likely placed.
Based on the mosaics and the stratigraphic evidence, this second phase can
be dated to the first half of the fifth century. The mosaics depicted geometric
motifs and birds. The two floor mosaics at the entrance contained motifs
which recall those found in jewelry and on silver plates.43 Similar decorative
schemes have also been noted in churches in Antioch.44 Mosaics in the middle
of the northern side of the church featured three octagons containing Greek
names: Melitis, Sergis, and Paulos. On the west side of the church, another
votive inscription advertised the name of Βάχχος θουρουρός ̣, or “Bacchus the
porter.” It was certainly inscribed there, near to the entrance, in order to be
read by catechumens leaving church after their prayers. From an architectural
42
Lee 2007, especially Chapters 5–7; Fear 2007; Whately 2013, 234–38.
43
Donceel-Voûte 1988, 85–86.
44
Figure 5. Aerial photograph of the citadel church (Harper’s archive)
perspective, it is worth mentioning the presence of a narthex, quite uncom-

mon in Late Antiquity, although, interestingly, another has been recorded
at Basilica A in nearby Resafa-Sergiopolis; the relationship between the two
settlements will be considered further below.45
The fifth century was in fact a period of major refurbishment and expan-
sion at the site. Not only was the citadel church expanded at this time, but a
cemetery martyr church was also built just outside of the citadel wall at the
beginning of the century. The latter church can be precisely dated thanks
to a mosaic inscription stating that the construction of this holy martyr-
ium was realized by the priest Jacobus and the perioudeutes 46 Paulos in the
45
46
For the definition of periodeutes: Chorepiscopus, in Khazdan 1991, 430. The term martyrium
probably refers to a larger complex of which only the church was excavated.
Figure 6. Mosaic of the citadel church with inscriptions (Harper’s archive)
month of Xanthikos of the year 740 (of the Seleucid era), which corresponds
to 429 ce.47
The important fifth-century phases of the two churches at Dibsi Faraj can
be placed in the context of region-wide church building in this period. The
construction of several fifth-century churches along the Euphrates is recorded
following the building of the church dedicated to Sergius at Resafa by Bishop
Alexander of Hierapolis in 431. This also precipitated an increase in the num-
ber of pilgrimages to the tomb of the martyr.48 The nearby location of Dibsi
Faraj, architectural similarities between its churches and those at Resafa, and
the reference to a man named Sergius in its citadel church all indicate that the
cult and its attraction to pilgrims played an important role in the development
of religious architecture at such Middle Euphrates sites.49
47
Feissel forthcoming.
48
Fowden 1999.
49
Spread of the cult of Sergius: Fowden 1999.
Figure 7. Inscription in the extra-mural basilica (Harper’s archive)
According to the tradition of the cult of Saint Sergius, which became

extremely popular in the fifth century, Sergius was a soldier in the service
of the emperor Maximian when he and a companion, Bacchus, were outed
as Christians. After refusing to sacrifice to a pagan god, they were sent to
Barbalisssus, where they were dressed in women’s clothing and led through
the streets of the fort. They were subsequently scourged until Bacchus died.
Sergius was finally made to walk with nails in his feet to Resafa, where he was
beheaded after further tortures.
The booming popularity of pilgrimages to Resafa, which would poten-
tially have stopped at Dibsi Faraj, was not the only reason for the construction
of Christian monuments at the settlement. More broadly, these monuments,
embellished with mosaics donated by laymen, can also be understood in the
context of the Christianization of settlements in this region, which gathered
pace in the fifth century.
Dibsi Faraj also benefitted from secular building work in the fourth to
fifth centuries. Around the mid-fourth century, the settlement expanded for
the first time beyond its walls with the building of a bath complex located
west of the fortification. The exposed portion of the bath-house complex had
Figure 8. Plan with mosaics of the extra-mural Basilica (Harper’s archive)
five main components: 1) an entrance corridor or vestibule to the north, lead-

ing to 2) a large courtyard or atrium to the east; 3) a cold room (frigidarium)
with a bath, situated immediately south of the entrance, though accessed
from the courtyard; 4) a large, central rectangular room, either a warm room
(tepidarium) or hot room (caldarium), again accessible only from the east
and sub-divided into northern and southern parts (possibly a combination
of warm and hot rooms), with a small bath (possibly laconicum or a sudato-
rium) in its southern part; and finally 5) an apsidal hot room (caldarium) to
the south, accessible possibly only from the room to the north. The furnace
(praefurnium) which probably lay on the south (or possibly the east) side of
the caldarium was not revealed by the excavation, though there was some
evidence to confirm that it was located there, mostly represented by a context
with ashes, located to the south (Figure 9).
This bath complex was abandoned in the middle of the fifth century when
new baths were built outside of the north-eastern gate of the citadel. These
were dated by an inscription at their entrance to 764 of the Seleucid era, that
is, sometime in 452 to 453 ce.50 These new baths were built in brick and dec-
orated with mosaics which were mostly preserved. The complex contained a
frigidarium, connected to a vestibule that conducted heat to three hot rooms.
The fact that the outstanding fourth- and fifth-century architectural
works at Dibsi Faraj included bathing facilities and churches, some of them
extramural, can be related to the generally peaceful situation on the east-
ern frontier. Even the fourth-century Roman-Persian wars, which raged from
338 to 350 and from 359 to 363, consisted mostly of attrition campaigns
and sieges in Upper Mesopotamia rather than the Middle Euphrates.51 More-
over, the emperor Julian’s ill-fated invasion of the Persian empire along the
Euphrates in 363 bypassed the stretch of the river on which Dibsi Faraj was
located.52 According to Ammianus, Julian and his army crossed the Euphra-
tes near Hierapolis and, having passed through Batnae, proceeded down the
River Balikh to Callinicum. From Callinicum, Julian led the army along the
left bank of the Middle Euphrates to Circesium and then across the Khabur
to Persian territory.53
Although no major fortification work took place during the reign of Con-
stantius II, the emperor divided what had been the province of Syria Coele
into three new provinces, one of which was Euphratensia, comprising the
50
51
Fourth-century Roman-Persian wars: Matthews 1989, especially Part 1; Dignas and Winter
2007, chapter 3.2; Isaac 1997, 437–60. Contemporary accounts: Dodgeon and Lieu 1991, chapters
7–9.
52
Smith 1999.
53
Ammianus’s account of Julian’s expedition: Amm. 23.2–25.3.
Figure 9a. Area 3 – Structures outside of the citadel – phase 1 (elaborated from Harper’s archive drawings; by M. Brizzi)
Figure 9b. Area 3 – phase 2 (elaborated from Harper’s archive drawings; by M. Brizzi)
Middle Euphrates frontier region.54 The main military commander responsi-

ble for these three provinces was the dux Syriae et Euphratensis, who had his
headquarters at Callinicum, Euphratensia, although he was also sometimes
based at Chalcis in Syria I.55 This provincial reorganization shows a recogni-
tion of the strategic importance of the Middle Euphrates region. Indeed, even
though there were no major Roman-Persian wars after 363, isolated attacks
on the Middle Euphrates region by non-Persian groups are recorded in our
sources. The Huns, who invaded Roman territory via the Caucasus in 395,
travelled down the Upper and Middle Euphrates before attacking Euphraten-
sia and Syria.56 During the brief Roman-Persian conflict of 421 to 422, the
Persians’ Lakhmid Arab allies tried to persuade the Persians to invade Syria
via the Middle Euphrates. There is no evidence, however, that this actually
took place.57 Instead, the Lakhmids were defeated by a Roman force, presum-
ably on or near the Middle Euphrates in which many of them drowned trying
to escape. When the Lakhmids attacked Euphratensia in 498, they were again
defeated, this time by the Roman dux Syriae et Euphratensis at Resafa.58
Nevertheless, Hunnic threats from the north to both Roman and Persian
empires meant that the peace they agreed upon in 363 was maintained for the
most part during the next 139 years.59 The consequent lack of major re-fortifi-
cation work at Dibsi Faraj was typical for the region. Some Middle Euphrates
forts—Pağnik Öreni, for instance—were even abandoned in the fifth century,
and others likely fell into disrepair.60
At the same time, the construction of the bathing establishment at Dibsi
Faraj, the proliferation of coin finds at the site,61 and finds relating to artisanal
and commercial activity, can be understood in the context of the economic
boom enjoyed by the eastern Mediterranean and Near Eastern provinces in
this period. Indeed, northern Oriens, although partly a borderland zone, was
one of the breadbaskets of the empire, a tax-producing rather than consum-
ing area.62 As well as being agriculturally productive, this was a commer-
cially vital region crossed by trading networks linking the Mediterranean
and the Near East.63 When we bear in mind that the Euphrates was one of
54
Comfort 2008, 270.
55
Schwarze 2017, 308.
56
Greatrex and Lieu 2002, 17–19.
57
Shahîd 1989, 26–35.
58
Shahîd 1989, 120–25.
59
Dignas and Winter 2007, 192–93; Chrysos 1993, 183; Sarris 2011, 125–26.
60
Harper 1977, 455.
61
Around 980 coins were found during the excavation (Munzi forthcoming).
62
Decker 2007; Decker 2009; Wickham 2005, 609–26; Walmsley 2007; Mango 2011.
63
Comfort 2008, 244–58; Pollard 2000, 175–77; Banaji 2016, 78–86.
these routes, and nearby Callinicum one of the official Roman-Persian trading
depots, it seems unsurprising that Dibsi Faraj shows sign of expansion across
the fourth and fifth centuries.64
Legislative decrees issued in 414 addressing the wanton or corrupt behav-
ior of the dux Syriae et Euphratensis and his men bolster an impression
that even the military could find time to enjoy themselves.65 The former was
accused of misappropriating public funds for the construction of private bath-
ing facilities and the latter of spending months on end in Hierapolis (where
they were presumably having a lot of fun, judging by the complaint received
in Constantinople from the provincial governor) when they were supposed to
be transporting wild beasts to Constantinople.66 It is not hard to imagine the
occupants of the Dibsi Faraj principia enjoying themselves in similar ways.
The Anastasian Phase

During the reign of Anastasius (491–518), the site was significantly re-forti-
fied. The wall was reconstructed with major work taking place primarily in its
south-eastern sector. While the wall was rebuilt in masonry, the Diocletianic
towers were cut down to their foundations and larger towers erected in their
place. At the same time, the circular corner towers were replaced by hexago-
nal towers and larger rectangular towers were placed at greater intervals. The
gateways were also modified, in particular the central gate, which received a
large freestanding brick redoubt that served to protect water supply shafts.
The eastern gateway, finally, was reduced to a postern.
In one of his preliminary reports, Harper indicated that this re-building
work was Justinianic.67 This followed from his identification of the site with
Neocesarea and the fact that it was named by Procopius among the Mid-
dle Euphrates sites whose defenses were upgraded by the emperor Justinian
(reigned 527 to 565). However, this dating is placed in doubt by the archaeo-
logical evidence. In particular, an inscribed military edict and the stratigra-
phy recorded during the excavation indicate that the reconstruction work was
instead Anastasian.68
The identification of several fragments of a Greek inscription is an espe-
cially important element in confirming the site’s military function during the
64
CJ 4.36.4. Discussion: Pollard 2000, 215–16.
65
CTh 7.11.2 and 15.11.2.
66
These were presumably lions, which still inhabited the Euphrates Valley in the nineteenth
century (Comfort 2008, 196).
67
Harper 1975, 325–27.
68
Munzi forthcoming. The highest number of coins recorded from the excavation date from the
reign of Anastasius: around half of the 130 Byzantine coins.
Figure 10a. Area 4 – corner tower of the citadel built above the earlier square tower – phase 2 (elaborated from Harper’s archive draw-
ings; Autocad M. Brizzi)
Figure 10b. Aerial photograph of corner tower (Harper’s archive)
reign of Anastasius. According to analysis by Denis Feissel, the inscription

is part of the law Anastasius promulgated to protect the limitanei.69 The
inscription at Dibsi preserves only some lines, but Feissel has been able to
decipher almost the full text by comparing these to the more complete ver-
sions of the same inscription found at Qasr Al-Hallabat, Bostra, and Umm
el-Jimal in Jordan.70 The heading of the inscription leaves no doubt that the
emperor Anastasius was the author of this law. Though some chapters of
this decree have been entirely lost (maybe fewer than 20%), we are now in
a position to reconstruct 135 lines, or 1,400 words, of the first part without
interruption.
The Anastasian refortification of Dibsi Faraj is understandable in light
of the outbreak of war between the Roman and Persian empires in 502. The
69
70
On the nature, content, and chronology of the text: Arce, Feissel, and Weber 2014.
ensuing four-year conflict was initiated by the Persian king Kavadh, who
invaded Roman Armenia and Mesopotamia in 502 and 503 and captured a
series of major cities by siege, including Theodosiopolis and Amida.71 Kavadh
was motivated by a need to bolster his domestic political standing following
a rebellion which had overthrown him, temporarily, in 496.72 Kavadh thus
used the Romans’ refusal to contribute to the defense of the Caspian Gates
pass, threatened by Hun raiders from the Eurasian Steppe, as a pretext for
his 502 invasion.73 Most of the war from 502 to 506 was fought to the north,
in Roman Armenia and Mesopotamia, although the Middle Euphrates did
feature on two occasions. At some point in 503 to 504, Kavadh took his army
back to Persian territory via the River Balikh-Middle Euphrates route follow-
ing his failed siege of Edessa;74 and in 502 the Lakhmids crossed the river on
their way to Osroene, from which they carried off 18,500 prisoners according
to the Chronicle of Joshua the Stylite.75
Following the war, the defenses of the Middle Euphrates were upgraded
significantly as part of a wider program of fortification work in the eastern
frontier provinces.76 Most of this work had been completed by the 530s. Anas-
tasius and his successors Justin I (reigned 518 to 527) and Justinian organized
this building program in response to the ease with which the Persians had
captured major cities in Armenia and Mesopotamia in 502 to 504.77
Although Procopius’s Buildings attributes most of this fortification work
to Justinian, archaeological investigations make clear that a major initial wave
was carried out by Anastasius.78 The addition of larger projecting towers,
of different shapes, and larger walls, incorporating a larger intramural area
as at Dibsi Faraj, were developments typical of this period. At Zenobia, for
instance, the hilltop acropolis was now encircled by the early sixth-century
71
502–6 war: Greatrex 1998, chapter 5.
72
Börm 2007, 93; Daryaee 2014, 26–29.
73
Greatrex 1998, 8–19, 76–79; Meier 2009, 174–94. See also Börm 2008, arguing that the Per-
sians demanded money from Rome mainly as a symbol of their own superiority.
74
Greatrex 1998, 105–7.
75
Josh. Styl. 52.
76
Sauer et al. 2017, 246, have recently played down the Romans’ military infrastructure and
capability in Late Antiquity, pointing to the lack of evidence for large, “organized” military bases
compared to the Early Imperial period. But, while there can be no doubting the impressive infra-
structural works of the Sasanians (another of the paper’s key points), Roman fortifications were
more sophisticated and monumental in Late Antiquity than they had been in previous periods, even
if they enclosed smaller spaces, and soldiers were based in fortress cities as well as purpose-built
military bases. For more detailed argument: Sarantis 2013b.
77
Sixth-century building work in Armenia and Upper Mesopotamia: Sarantis 2013a, 341–48.
Key works include: Whitby 1986a; Crow 2007; Comfort 2008, especially 229–43; Howard-John-
ston 2013, 872–84.
78
Croke and Crow 1983; Ulbert 2000; Meier 2009, 212–13.
Figure 11. 3-D model obtained by a mosaic of aerial photographs (Harper’s archive;
photoscan by M. Brizzi)
city fortification, which included thirty projecting rectangular towers and six
gates.79 The walls at Resafa have also been attributed to the reign of Anas-
tasius and, once again, included thicker, larger fortifications and projecting
prow-shaped, rectangular, and hexagonal towers.80 The U-Shaped towers at
Palmyra seem likely to have been constructed at some stage from 506 to the
530s.81 These early sixth-century fortifications were in most cases built with
more durable and sophisticated masonry techniques than the Tetrarch-era
defenses, with ashlar blocks instead of mudbrick on the outer faces and, in
places, opus mixtum with brick courses.
Epigraphic, legislative, and literary evidence referring to eastern fortifica-
tion works makes clear that these were organized, funded, and implemented
on a local level, even if initiated by the imperial authorities.82 While building
79
Lauffray 1983, part 2.
80
Konrad, Baldus, and Ulbert 2001, 14–15; Gussone and Sack 2017, 125–27.
81
See notes 24 and 39 above.
82
Zanini 2003, especially 218–20 on the center-periphery interactions involved in the implemen-
tation of even major works such as Dara. For a discussion of local contributions to fortification
work in the Balkans: Sarantis 2016, 219–24.
inscriptions on occasion acknowledge the distant authority of the emperor,

they also refer to the direct role of local military commanders, administra-
tive officials, and architects. The best example is the inscription from Chal-
cis, which mentions a general named Longinus, an ex-consul, Anastasius,
and Isidore, an engineer (probably the famous military architect, Isidore the
Younger, also responsible for fortification work at Zenobia, according to
Procopius).83 Literary sources, meanwhile, acknowledge the role of bishops
in initiating infrastructural works at sites such as Resafa and Dara.84 Non-
official secular elites, such as Thomas, attested in the acropolis inscription at
Androna, could also be responsible for such defensive works.85 Even though
we lack fortification inscriptions at Dibsi Faraj, it is not hard to imagine a
combination of a local military official, leading churchmen, and the lay indi-
viduals responsible for donating mosaics to the basilica contributing in vari-
ous ways to this project.
The publication of the Anastasian edict at Dibsi Faraj, as at other east-
ern borderland sites, can be understood as another way in which the impe-
rial authorities sought to assert stronger control over the supply, discipline,
and accountability of frontier military divisions and their administrative staff
following the renewal of war with Persia.86 The edict in question, published
also at sites in Jordan, as mentioned above, dates to 492 or 507.87 Some have
preferred a 490s context for Anastasian military edicts based on the fact that
most of his military supply-related legislation was published in this decade.88
While this is plausible, a date of 507 would make more sense in the context
of the post-war fortification and recovery program, especially given that the
edict sought fundamentally to improve the conditions and protect the rights
of frontier limitanei soldiers. Moreover, a number of Anastasius’s key sup-
ply- and corruption-related decrees could be dated to any time between 491
and 505.89 Whatever the case, the edict confirms the continued presence of a
military garrison, apparently comprising limitanei troops, at Dibsi Faraj in
the early sixth century. Interestingly, the only edited section from the edict
found at al-Hallabat referring to the province of Euphratensis concerns the
supply and payment of the dux and his soldiers.90
83
Chalcis inscription: Feissel 2000, 98; IGLS 348–49. Isidore the Younger: Martindale 1992,
724–25. Zenobia fortification work: Proc. Aed. 2.8.11–25.
84
Dara: Zach. 7.6. Resafa works and role of bishops: Gussone and Sack 2017, especially 124.
85
IGLS 1682. Discussion: Mango 2017, 202.
86
Feissel 2014; Shahîd 1989, 131–33.
87
Text: Marcillet-Jaubert 1982. Dating: Feissel 2014, 34. Context and discussion of the al-Hal-
labat imperial edict: Arce, Feissel, and Weber 2014. Bosra inscription: IGLS 9045–46.
88
Onur 2016, 188–89.
89
See, for example, CJ 12.37.16.
90
SEG 32, 1554B, 6.
Although its refortification work is typical, Dibsi Faraj differs from some
of the other major sites in its lack of a major church building drive in the
Anastasian period. Major new churches and episcopal complexes at Palmyra
and Resafa were erected at this time.91 However, in spite of the lack of similar
activity at Dibsi Faraj, its fifth-century churches remained in use and there is
no reason to suppose that the site did not remain an important stopping point
on the pilgrimage route to Resafa.92
The Post-Anastasian Sixth Century

Even though the remainder of the sixth century witnessed few architectural
changes at the site, the enlargement of the baths of the principia, discussed
above, represented a significant development.
These new structures were added to the west of the complex, covering an
area of 500 square meters. They included a caldarium, a tepidarium, and a
hypocaust. Their floors and walls were originally covered with marble opus
sectile, which was almost entirely robbed in the Islamic period. This is the
only major building activity identified during the excavation which seems to
have been carried out in the Justinianic era.
As well as the renovation of the principia baths, finds at the site suggest
that it continued to have a military function throughout the sixth century.
Artefacts dating to the early Byzantine period include: a copper alloy bucket,
apparently a component typical of a military officer’s equipment; several fibu-
lae, including a bronze crossbow fibula worn usually by rank and file soldiers
or officers; and a bone pin with a grooved collar and teardrop-shaped head,
common from eastern Mediterranean contexts and apparently introduced to
western Europe by Syrian soldiers.93
The renovation of the baths of the principia at Dibsi Faraj and its appar-
ently continued military function can be understood in the context of further
military insecurity. Roman-Persian warfare broke out again during the reign
of Justinian, the war of 527 to 532 being the first of four more major wars
across the sixth century (from 527 to 532, 540 to 545, 548 to 556, and 573
to 591).94 While the second of these wars mainly affected Lazica, south of the
Caucasus Mountains, the others featured more regular Persian and Lakhmid
Arab attacks on the Middle Euphrates region and northern Syria than had the
91
Palmyra church in the Temple of Bel: Jastrzębowska 2013. Resafa: Gussone and Sack 2017,
124–27.
92
Importance of the Saint Sergius cult at Resafa and other Syrian-Mesopotamian frontier bases:
Fowden 1999.
93
For finds, see Leone forthcoming.
94
Sixth-century Roman-Persian Wars: Greatrex 1998; Dignas and Winter 2007, chapter 3.4;
Whitby 1988, part 3.
fourth-century or Anastasian wars. Indeed, following in Shapur I’s footsteps,

the Persian kings, Kavadh (488 to 531) and Khusro I (531 to 579), launched
invasions of Roman Syria along the river in 531, 540, 542, and 573.
In 531, a Persian-Lakhmid army commanded by the general Azarethes
was halted at Chalcis by an army led by the Roman general, Belisarius.95
Belisarius proceeded to shadow the Persian force as it retreated along the
Middle Danube via Barbalissus. Reluctantly, Belisarius was persuaded by his
troops to fight a battle with the Persians on the right bank of the Euphrates
opposite Callinicum.96 This bloody encounter concluded in a Pyrrhic victory
for the Persians, the remnants of the Roman army attempting to flee across
the Euphrates, many drowning in the process.
In 540, Khusro I led another attack on Syria via the Middle Euphrates.97
Bypassing the heavily defended city of Circesium, he nevertheless sacked Sura,
and ended up capturing or holding to ransom the major cities of northern
Syria, west of the Euphrates bend, including Hierapolis, Beroea, and Antioch.98
After occupying southern Syrian cities such as Apamea, Khusro turned north,
crossed the Middle Euphrates, and launched a series of failed attacks on the
major Mesopotamian fortress cities of Dara and Edessa. Khusro attacked
again in 542, leading his army along the right bank of the Euphrates before
besieging but failing to capture Resafa.99 He was deterred from his plan to
invade Palestine by the presence of Belisarius and his army in the region and
so crossed the Euphrates and sacked Callinicum before retreating to Persian
territory via Mesopotamia.
A third Persian incursion into Syria via the Middle Euphrates took place in
573.100 After travelling along the river, Khusro I divided his army at Circesium,
leading one part of it up the Khabur valley towards Dara, which eventually
fell after a long siege, and sending another along the Middle Euphrates under
the general Adarmahan to invade Syria. Our main source for this attack, John
of Ephesus, only specifically mentions the sack of Apamea, although he refers
vaguely to the destruction of many other fortresses and villages.101 Finally,
Khusro II, fleeing the rebel general Bahram, travelled to Roman territory via
95
Greatrex and Lieu 2002, 92–93. The most detailed contemporary account of the Callinicum
campaign: Proc. Pers. 1.18.
96
The Battle of Callinicum: Greatrex 1998, 200–207; Rubin 1960, 287–89; Shahîd 1995, 134–42;
Whitby 2018, 1205–8. Contemporary accounts: Joh. Mal. 18.60; Proc. BP 1.17.30–56; Zach. 9.4.
97
Contemporary accounts: Greatrex and Lieu 2002, 103–8. See also Börm 2006.
98
Sack of Antioch, 540: Downey 1961, 542–44; Börm 2007, 172–75; Whitby 2017, 1208–10.
This event as part of the period between 540 and 542, a turning point in the reign of Justinian:
Meier 2003, 649–50.
99
Procop. Pers. 2. 20–21; Zach. 10.8.
100
Greatrex and Lieu 2002, 146–47.
101
Joh. Eph. 6.6.
the Middle Euphrates in 591.102 He was received by the local Roman com-
mander Probus at Circesium, before moving on to Hierapolis where he met
the general of the Roman eastern field army, Comentiolus.
The Middle Euphrates was also attacked by the Persian-backed Lakhmid
confederation of Arab tribes, ruled by the formidable Al-Mundhir III (some-
time between 503 and 505 to 554) in 518/19, 525, and 529.103 On the lat-
ter occasion, Al-Mundhir attacked the village of Amis between Chalcis and
Antioch, carrying off 400 virgins and an anchorite named Dodo according
to The Chronicle of Pseudo-Zachariah Rhetor.104 In 531, in a repeat of the
Hun incursion of 395, the Sabir Huns invaded from the north, crossing the
Euphrates and raiding Euphratensia and Cilicia.105
Dibsi Faraj would surely have been affected by a number of these attacks,
especially those in 531 and 540. It is therefore, surprising perhaps, that, in
spite of work expanding its principia, there is no material evidence for further
fortification work. Procopius’s Buildings records Justinianic building work
at the site, stating that, as at Zenobia, the walls were low and constructed
of loose stones, before enjoying a major Justinianic renovation.106 While this
is clearly a rhetorical ploy to exaggerate Justinian’s achievements, the con-
struction of a new, more monumental northern wall at Zenobia, with larger
projecting towers, has indeed been attributed to his reign.107 Therefore, while
there was no obvious second phase at the Anastasian walls of Dibsi Faraj, it is
not impossible that more minimal renovation work took place at some stage
later in the sixth century.
Moreover, in addition to Zenobia, there is good archaeological and epi-
graphic evidence for another wave of eastern frontier fortification work across
the region during the reign of Justinian (reigned 527 to 565). This includes the
newer, taller arcaded fighting platforms at Dara and the enlargement of and
addition of vaulting to the towers at Resafa.108 Sura acquired a larger, roughtly
39.4-hectare fortified annex to the west of the initial fortress-canabae complex,
dated by the excavators to the Justinianic era.109 The U-shaped towers at Pal-
102
Theoph. Sim. 4.10.4–11.
103
Sources for and historiography on raids of 518/19 and 525: Greatrex 1998, 131, note 30; and
Greatrex et al. 2011, 297, note 66; Shahîd 1995, 43–45. For 529, see Greatrex et al. 2011, 298,
note 70.
104
Zach. 8.5.
105
Contemporary accounts: Zach. 9.6; Joh. Mal. 18.70. Historiography: Greatrex et al. 2011,
328, note 92.
106
Proc. Aed. 2.8.8–25.
107
Lauffray 1983, especially Part 2, chapter 4; Ulbert 2000, 141.
108
Dara: Whitby 1986b; Keser-Kayaalp and Erdogan 2017, 155. Resafa: Gussone and Sack 2017,
127–30; Konrad, Baldus, and Ulbert 2001, 14–15; Karnapp 1976.
109
Konrad 1999, 398–400.
myra, as mentioned earlier, could date from any time in the early sixth centu-
ry.110 While many of these sites on or near to the frontier seem likely to have
been renovated in the earlier part of Justinian’s reign, from 527 to 540, a later
wave of fortification work is attested epigraphically in northern Syrian regions
to the west.111 The building inscriptions at Chalcis, Hierapolis, Androna, and
Cyrrhus are dated between the late 540s and 559.112 The latter works were
probably constructed in the aftermath of the Persian invasions of 540 and 542.
The socio-economic impact of Persian and Lakhmid invasions on north-
ern Syria is suggested by signs of socio-economic stagnation at major cities
such as Antioch and Apamea.113 The considerable booty, prisoners, and in the
case of Antioch, architectural works that the Persians extracted during their
attacks in 540 and 573, meant that these were the most traumatic military
episodes the region had experienced since Shapur I’s raid in 252. There are
no signs of destruction or major socio-economic dislocation at Dibsi Faraj,
although there are fewer coins dated to the post-Justinianic later sixth-cen-
tury phase and the main intramural church was razed to the ground at some
stage.114 The apparent continuity at the site mirrors the general trend observed
at sites along the nearby Strata Diocletiana.115 This may be explained by the
fact that the interior, wealthier Syrian Levantine provinces were the main tar-
gets of Lakhmid and Persian raiding, frontier fortresses along the Euphrates
often being bypassed by enemy forces.
In addition, this northern sector of the sub-Euphratean eastern limes
shows no signs of abandonment by the limitanei frontier forces in the sixth
century, unlike the southern limes Arabicus, which by the mid sixth century at
the latest had apparently lost an overt military function.116 While the Ghassa-
nid Arab phylarchs may also have been responsible for defending the northern
section of the limes by the later sixth century, as is apparent from the palace or
reception hall of al-Mundhir III just outside Resafa, it resulted in less obvious
changes to the topography of sites along the Strata or the Middle Euphrates.117
110
See notes 24 and 39 above.
111
For the earlier wave, see notes 77 and 78 above.
112
Cyrrhus: Alpi 2016. Androna: note 85 above. Hierapolis: Mouterde et al. 1945, 209, note 39.
Chalcis: note 83 above.
113
Foss 1997, 190–97 and 205–26; Kennedy 1985, 149–63; Liebeschuetz 1988; more recently
Walmsley 2007 and 2013. Earthquakes such as those in 526 and 528 also caused damage in north-
ern Syrian cities such as Antioch, but there is no evidence they affected areas as far away as the
Middle Euphrates: see, for example, Ambraseys 2009, 184–85.
114
Coins: Munzi forthcoming. Destruction at intramural church: Harper 1975, 333.
115
Konrad 1999, 408–10.
116
Arce 2015 sees this development as a consequence of Justinian’s Ghassanid policy. See also:
Whately 2013.
117
Ghassanids and northern Syrian defence: Shahîd 1995, chapters 4–6. For criticism of Shahîd 1995:
Whittow 1999, especially 207 and 212–15. Al-Mundhir reception hall: Konrad 2015; Brands 1998.
The Seventh Century

Stratigraphically it has been extremely difficult to single out specifically sev-
enth-century developments at the site. It has, however, been possible to iden-
tify some changes at the two basilicas. The mosaic depicting a portico in the
elevated apse of the extramural martyr church has been the subject of discus-
sion by Donceel-Voûte. She considered the mosaics from Dibsi Faraj in her
corpus of church mosaics of Syria and Lebanon and proposed that this mosaic
dated later than the fifth-century foundation of the church.118 She pointed out
that its architectural iconography has parallels in manuscript illustrations, in
particular the tables of Canons – Vat Syr 268, dating to the thirteenth century
(Figure 9). However, representations of porticos in these later manuscripts
are normally associated with human figures. The aniconic representation of
a portico found in the apse at Dibsi Faraj should instead be dated on stylistic
grounds to the seventh century according to Donceel-Voûte.119 This is con-
firmed by the stratigraphy, which shows that, at this time, the apse of the
church was elevated, a tomb placed next to it and this new mosaic laid.
By this phase, the citadel basilica, meanwhile, had already been aban-
doned. A large L-shaped building was constructed on top of it following
the Arab occupation. The function of this structure is difficult to identify,
although it is imposing in size and thus clearly significant. L-shaped buildings
appear commonly in the early Islamic period.120
The continued intramural and extramural works at Dibsi took place
against a backdrop of enormous political upheaval in the East. From 603 to
628, Khusro II became the first Persian ruler in Late Antiquity to attempt the
all-out conquest of the Roman Near East.121 To this end, he spent seven years
reducing Mesopotamia and Armenia before breaching the Middle Euphrates
frontier in 610 and conquering Syria over the next two to three years. From
that point, the Middle Euphrates featured little in a war that was fought pre-
dominantly in Asia Minor and the Transcaucsus.
The peace of 630 was short-lived, however, in light of the Early Islamic
conquests which followed from roughly 632.122 After four centuries of remark-
able geopolitical stability, the Near East now witnessed a major change to
its strategic geography with the collapse of the Sasanian Persian empire and
the rise of the Umayyad Arab Caliphate. However, once again, it is unlikely
that the Middle Euphrates region suffered during the Early Islamic wars of
118
119
120
Eger 2015, 252. Unfortunately, the function of these buildings is unclear.
121
Howard-Johnston 2010, 436–45; Kaegi 2003, chapters 3–5; Greatrex and Lieu 2002, chap-
ters 13–14.
122
Howard-Johnston 2010; Donner 1981; Hoyland 2015.
conquest. Most of the major fighting took place in the Roman Levantine prov-
inces and Persian southern Mesopotamia.
While coin hoards in northern Syria and destruction evidence and settle-
ment changes in Asia Minor demonstrate the impact of the Persian war in
those regions, Dibsi Faraj was typical of the Middle Euphrates and northern
Mesopotamian regions in showing no sign of major economic dislocation.123
Coin finds, albeit fewer in number, and work at the extra mural basilica dem-
onstrate that life went on at the site without much disruption. This can again
be understood in a regional context. Longer-term, the Christian Arab tribes
on both sides of the Middle Euphrates held out against Islamicization, and the
region north of the Middle Euphrates, at least, was ruled with a soft touch
by the Umayyads in the second half of the seventh century.124 Survey evidence
suggests settlement pattern continuity around Edessa and increase along the
Balikh and Khabur valleys and other areas of the Middle Euphrates.125 This is
bolstered by textual evidence for the continued wealth and vibrancy of secular
elites at cities like Edessa and of Christian monastic cultures on and around
the Tur Adbin holy mountain.126 Along and south of the Middle Euphrates,
some cities were abandoned, including Sura. However, others were re-pur-
posed by the early Islamic rulers, Resafa and Callinicum-Raqqa being the
most spectacular examples.127
Conclusions: Dibsi Faraj in Its Historical Context

The site of Dibsi Faraj provides us with a unique set of data for a fortified site
between the Strata Diocletiana and the Middle Euphrates bend. The evidence
from Dibsi Faraj thus augments research at Zenobia, Resafa, and the Strata
Diocletiana to enrich our understanding of the northern Syrian borderlands.
While its Tetrarchic and early sixth-century fortification works were
typical of the region, Dibsi Faraj was distinctive in other ways. Unlike com-
parable sites in the region, it did not transform markedly in size, location, or
character between the fourth and sixth centuries. Whereas legionary forts
like Resafa evolved into larger settlements dominated by churches, Dibsi
Faraj retained its largely military character across the period according to the
continued functioning of the principia. It only acquired two churches, one
of which was extramural and experienced no new wave of church building
123
See note 113 above for the impact of warfare on northern Syria. Continuity in northern
Oriens: Decker 2007; Decker 2009; Wilkinson and Tucker 1995; Morony 2004.
124
Wickham 2005, 778–80.
125
Wilkinson and Algaze 1990; Kennedy 2011.
126
Edessa: Segal 1970. Tur Adbin: Palmer 1990.
127
Callinicum-Raqqa: Mango 1991; Meinecke 1991; Heidemann 2006. Resafa: Brands 2011.
in the sixth century. The lack of a bishop’s name in the fifth-century build-
ing inscription from the extramural church could mean that this was not a
bishopric. Nevertheless, the architecture features of and epigraphic evidence
from the two churches which were built reveal possible religious connections
with Resafa and the cult of Saint Sergius. This is unsurprising, bearing in
mind that Dibsi was on the route taken by pilgrims from the West to Ser-
gius’s shrine at Resafa.
Otherwise, it is the secular building work at the site which impresses
most. The monumental fortifications, principia and sophisticated water sys-
tem bear witness to considerable investment. The water system consisted of
cisterns and water channels developed around the site to supply its numerous
baths, some of which were lavishly decorated. Indeed, the addition of a new
bathing complex to the principia, expensively decorated with marble, was the
most conspicuous sixth-century expenditure on an architectural structure.
The presence of baths outside of the city wall does not indicate that the defen-
sive citadel fell out of use but rather the existence of suburban or extra mural
quarters. Similar evidence has been identified for instance at Kifrin, where
archaeologists have recorded another albeit smaller and shorter-lasting citadel
on the Euphrates.128
This all suggests the presence of a discrete number of wealthy occupants at
the site who would have at least partly contributed to the initiation, organiza-
tion, funding, and erection of infrastructural works. The ruling elite presum-
ably consisted of military officers or military administrative staff, to whom
the majority of the provisions in the Anastasian military edict were addressed.
Indeed, the public display of the inscribed military law of Anastasius along
with the refurbishment and reinforcement of the citadel identify this as an
important base of the limitanei.129
With evidence from other sites taken into account, these provisions bolster
the impression that an important secular, sometimes military elite controlled
northern Syrian sites along and in the hinterland of the Middle Euphrates.
This includes epigraphic evidence from Chalcis, Cyrrhus, Hierapolis, and
Androna, and material and textual evidence for the estate of Magnos the
Syrian, as well as other apparent desert villas in the Syrian steppe such as
ibn al-Wardan.130 Even if the Ghassanid federates took over this section of
the limes, there is no evidence that they were organized in dramatically
128
Lippolis 2007.
129
Although field army units were also permanently stationed at bases in this region by the sixth
century, for example, at Palmyra: see Joh. Mal. 18.2. Roman strategic deployments in northern
Syria: Liebeschuetz 1977, 495–97.
130
Inscriptions: see notes 83, 85, and 112 above. Magnos and other northern Syrian aristocra-
cies: Kennedy 2010.
different ways from Roman soldiers.131 Whatever the identity of elites based
at sites such as Dibsi, they were presumably responsible for liaising with and
managing tribal groups in the desert regions. Policing the area and ensuring
the security of commercial routes would have been important priorities in
addition to defending approaches to the Roman empire from Persian and
Lakhmid raiders.132
Local elite wealth is unsurprising, bearing in mind the agricultural and
commercial prosperity of northern Oriens. Indeed, the coin series at Dibsi
Faraj peaks between the end of the fifth and the beginning of the sixth cen-
tury (fifty nine coins) followed in number by those minted in the remainder of
the sixth century (twenty four coins), indicating the economic vitality of the
site in this period.133 The settlement must have continued to be important in
the Justinianic period, considering the refurbishment of the bathing complex
just mentioned.
Finally, the fact that the site continued to function and receive renova-
tion work across the late sixth and seventh centuries demonstrates that the
Roman-Persian and early Islamic wars did not necessarily have a long-term
socio-economic impact on this region, as is assumed for other parts of north-
ern Syria. Dibsi Faraj must have been affected in the short-term by the numer-
ous Persian and Lakhmid Arab incursions. But the presence of wealthy elites,
bound economically and, presumably, culturally to the region, would explain
how such sites recovered from short-term shocks.134
University of Durham
anna.leone@durham.ac.uk
Römisch-Germanisches Zentralmuseum Mainz /

Johannes Gutenberg Universität Mainz
sarantis@rgzm.de
131
Studies on the limes Arabicus sometimes assume that the Ghassanids ceased using fixed mili-
tary bases (e.g. Arce 2015 seems to equate the employment of the Ghassanids with the cessation
of military activity at the forts of the limes Arabicus). This overstates their nomadic lifestyle and
ignores the fact that they were effectively a dimorphic chiefdom like the Lakhmids. On dimorphic
Lakhmids: Fisher and Wood 2016; Rowton 1977. Sedentary Ghassanids: Shahîd 1984, Ch.4. The
Ghassanids’ ability to link the nomadic and sedentary worlds of the Syrian steppe prefaced the rise
of the Umayyads: Whittow 1999, 224.
132
Policing functions of frontier forts: Isaac 1990. Others, such as Parker 1991, stress the forts’
broader strategic role in defending the empire against enemy raids. There is no reason why fron-
tier forts could not have performed both functions. For a summary of the debate: Sarantis 2013a,
358–60.
133
Munzi forthcoming.
134
For a more detailed discussion of resilience in northern Oriens, see Sarantis forthcoming.
References
Alpi, Frédéric. 2016. “Les inscriptions justiniennes de Cyrrhus (Euphratésie).” Syria:
Archéologie, Art et Histoire 93: 171–84.
Ambraseys, Nicholas. 2009. Earthquakes in the Mediterranean and Middle East: A
Multidisciplinary Study of Seismicity up to 1900. Cambridge: Cambridge Uni-
versity Press.
Arce, Ignacio. 2015. “Severan Castra, Tetrarchic Quadriburgia, Justinian Coeno-
bia, and Ghassanid Diyarat: Patterns of Transformation of Limes Arabicus
Forts During Late Antiquity.” In Roman Military Architecture on the Frontiers:
Armies and Their Architecture in Late Antiquity, edited by Rob Collins, Mat-
thew Symonds and Meike Weber, 98–122. Oxford: Oxbow.
Arce, Ignacio, Denis Feissel, and Thomas M. Weber. 2014. The Edict of the Emperor
Anastasius I (491–518 AD): An Interim Report. Amman: DAAD.
Banaji, Jairus. 2016. Exploring the Economy of Late Antiquity: Selected Essays.
Cambridge: Cambridge University Press.
Bell, Gertrude. 1910. “The East Bank of the Euphrates from Tel Ahmar to Hit.”
Geographical Journal 36: 513–37.
Blume, Fred, et al., eds. and trans. 2016. The Codex of Justinian: a New Annotated
Translation with Parallel Latin and Greek Text. Cambridge: Cambridge Univer-
sity Press.
Börm, Henning. 2006. “Der Perserkönig im Imperium Romanum. Chosroes I.
und der sasanidische Einfall in das Oströmische Reich 540 n. Chr.” Chiron
36:299–328.
. 2007. Prokop und die Perser: Untersuchungen zu den römisch-sasanidischen
Kontakten in der ausgehenden Spätantike. Stuttgart: F. Steiner.
. 2008. “‘Es war allerdings nicht so, dass sie es im Sinne eines Tributes
erhielten, wie viele meinten . . . ’: Anlässe und Funktion der persischen Geldfor-
derungen an die Römer (3. Bis 6. Jh.).” Historia 57.3: 327–346.
. 2016. “A Threat or a Blessing? The Sasanians and the Roman Empire.” In
Diwan: Untersuchungen zu Geschichte und Kultur des nahen Ostens und des
östlichen Mittelmeerraumes im Altertum. Festschrift für Josef Wiesehöfer zum
65. Geburtstag, edited by Carsten Binder, Henning Börm and Andreas Luther,
615–46. Duisburg: Wellem Verlag.
Brands, Gunnar. 2011. “City and Territorium of Rusafa in Late Antiquity and Early
Islam.” In Le Proche-Orient de Justinien aux Abassides, Actes du colloque
international de Paris, Inha, 18–20 Octobre 2007, edited by Antoine Borrut,
Muriel Debié, Dominique Pieri and Jean-Pierre Sodini, 59–76.
Brooks, Ernst W., ed. and trans. 1919–1924. Pseudo-Zachariah Rhetor: Chronicle,
Historia Ecclesiastica. Paris: E Typographeo Reipublicae.
Brunt, Peter A., ed. and trans. 1976. Arrian: Anabasis of Alexander. Cambridge,
MA. Harvard University Press.
Chapot, Victor. 1907. La Frontiere de l’Euphrate de Pompee à la Conquete arabe.
Paris: Albert Fontemoing.
Chrysos, Evangelos K. 1993. “Räumung und Aufgabe von Reichsterritorien. Der
Vetrag von 363.” Bonner Jahrbücher 193: 165–202.
Comfort, Anthony. 2008. “Roads on the Frontier between Rome and Persia: Euphra-
tensia, Osrhoene and Mesopotamia from AD 363 to 602.” PhD diss., University
of Exeter.
Comfort, Anthony, and Rifat Ergeç. 2001. “Following the Euphrates in Antiquity:
North-South Routes around Zeugma.” Anatolian Studies 51: 19–50.
Croke, Brian, and James Crow. 1983. “Procopius and Dara.” Journal of Roman
Studies 73: 143–59.
Crow, James. 2007. “Amida and Tropaeum Traiani: A Comparison of Late Antique
Fortress Cities on the Lower Danube and Mesopotamia.” In The Transition to
Late Antiquity: On the Danube and Beyond, edited by Andrew Poulter, 435–55.
Oxford: The British Academy.
Daryaee, Touraj. 2014. Sasanian Persia: The Rise and Fall of an Empire. 1st ed.
London: I. B.Tauris.
De Boor, Carl, ed. 1887. Theophylacti Simocattae Historia. Leipzig: Lipsiae.
Decker, Michael. 2007. “Frontier Settlement and Economy in the Byzantine East.”
Dumbarton Oaks Papers 61: 217–67.
. 2009. Tilling the Hateful Earth: Agricultural Production and Trade in the
Late Antique East. Oxford: Oxford University Press.
Dewing, Henry B., ed. and trans. 1914–1954. Procopius. 7 vols. Cambridge, MA:
Harvard University Press.
Dignas, Beate, and Engelbert Winter. 2007. Rome and Persia in Late Antiquity:
Neighbours and Rivals. Cambridge: Cambridge University Press.
Dodgeon, Michael H., and Samuel N. C. Lieu. 1991. The Roman Eastern Fron-
tier and the Persian Wars (AD 226–363): A Documentary History. London:
Routledge.
Donner, Fred McGraw. 1981. The Early Islamic Conquests. Princeton, NJ: Princ-
eton University Press.
Donceel-Voûte, Pauline. 1988. Les pavements des églises byzantines de Syrie et du
Liban: décor, archéologie et liturgie. Louvain: Oleffe
Downey, Glanville. 1961. A History of Antioch in Syria: From Seleucus to the Arab
Conquest. Princeton, NJ: Princeton University Press.
Dussaud, René. 1927. Topographie historique de la Syrie antique et médiévale. Paris:
Librarie Orientaliste Paul Geuthner.
Edwell, Peter M. 2008. Between Rome and Persia: The Middle Euphrates, Mesopo-
tamia and Palmyra under Roman Control. London: Routledge.
Eger, A. Asa. 2015. The Islamic-Byzantine Frontier: Interaction and Exchange
among Muslim and Christian Communities. London: I. B. Tauris.
Fear, Andrew. 2007. “War and Society.” In The Cambridge History of Greek and
Roman Warfare, Volume 2, edited by Philip Sabin, Hans Van Wees, and Michael
Whitby, 424–58. Cambridge: Cambridge University Press.
Feissel, Denis. 2000. “Les édifices de Justinien au témoignage de Procope et de
l’épigraphie.” Antiquité Tardive 8: 81–104.
. 2014. “The Reconstructed Text: A Unique Document on Military Admin-
istration in the Late Roman East.” In The Edict of the Emperor Anastasius I.
(491–518 AD), edited by Ignacio Arce, Denis Feissel, and Thomas M. Weber,
33–36. Amman: DAAD.
. Forthcoming. “The Greek Inscriptions.” In The Excavation of Dibsi Faraj:
Life in the Middle Euphrates, edited by Anna Leone. Washington, DC: Dumbar-
ton Oaks.
Fisher, Greg, and Philip Wood. 2016. “Writing the History of the ‘Persian Arabs’:
The Pre-Islamic Perspective on the ‘Nas ̣rids’ of Al-Ḥīrah.” Iranian Studies 49.2:
247–90.
Fowden, Elizabeth Key. 1999. The Barbarian Plain: Saint Sergius between Rome and
Iran. Berkeley: University of California Press.
Freedman, David Noel, and John M. Lundquist. 1979. Archeological Reports from
the Tabqa Dam Project—Euphrates Valley, Syria. Cambridg,e MA: American
Schools of Oriental Research.
Gawlikowski, Michal. 1996. “Thapsacus and Zeugma: The Crossing of the Euphra-
tes in Antiquity.” Iraq 58: 123–33.
Gelzer, H., et al., eds. 1898. Patrum Nicaenorum nomina, Latine, Graece, Coptice,
Syriace, Arabice, Armeniace. Teubner: Leipzig.
Golvin, Lucien. 1945. “À la recherche de la cité médiévale de Bâlis (Moyen-Euphrate).”
In Le Moyen Euphrate: zone de contacts et d’exchanges, Syrie romaine, 154–56.
Paris: P. Geuthner.
Greatrex, Geoffrey. 1998. Rome and Persia at War, 502–532. Leeds: Francis Cairns.
Greatrex, Geoffrey, and Samuel N. C. Lieu. 2002. The Roman Eastern Frontier and
the Persian Wars: A Narrative Sourcebook. London: Routledge.
Greatrex, Geoffrey, Robert R. Phenix, Cornelia B. Horn, Sebastian P. Brock, and
Witold Witakowski. 2011. The Chronicle of Pseudo-Zachariah Rhetor: Church
and War in Late Antiquity. Liverpool: Liverpool University Press.
Gregory, Shelagh. 1995. Roman Military Architecture on the Eastern Frontier.
Amsterdam: Adolf M. Hakkert.
Gussone, Martin , and Dorothée Sack. 2017. “Resafa / Syrien. Städtebauliche
Entwicklung Zwischen Kultort Und Herrschaftssitz.” In New Cities in Late
Antiquity: Documents and Archaeology, edited by Efthymios Rizos, 117–36.
Turnhout: Brepols.
Haarer, Fiona K. 2006. Anastasius I: Politics and Empire in the Late Roman World.
Cambridge: Francis Cairns.
Harper, Richard P. 1974a. “Excavations at Dibsi Faraj, Northern Syria, 1972.” Les
Annales archéologiques arabes syriennes 24: 25–29.
. 1974b. “Second Preliminary Report on Excavations at Dibsi Faraj.” Les
Annales archéologiques arabes syriennes 34: 30–37.
. 1975. “Excavations at Dibsi Faraj, Northern Syria, 1972–1974: A Prelimi-
nary Note on the Site and Its Monuments with an Appendix.” Dumbarton Oaks
Papers 29: 319–38.
. 1977. “Two Excavations on the Euphrates Frontier 1968–1974: Pağnik Ŏreni
(Eastern Turkey) 1968–1971 and Dibsi Faraj (Northern Syria) 1972–1974.” In Zu
den Militärgrenzen Roms II: Vorträge des 10. internationalen Limeskongresses
in der Germania Inferior, edited by Dorothea Haupt and Heinz Günter Horn,
453–60. Köln: Rheinland-Verlag.
. 1980. “Athis-Neocaesareia-Qasrin-Dibsi Faraj.” In Le Moyen Euphrate:
zones de contacts et d’échanges (Actes du colloque de Strasbourg, 10–12 Mars
1977), edited by Jean Margueron. Leiden: Brill.
Hartmann, Udo. 2017. “The Third-Century ‘Crisis.’” In The Encyclopedia of
Ancient Battles, edited by Michael Whitby and Harry Sidebottom, 1047–67.
Oxford: Wiley-Blackwell.
Heidemann, Stefan. 2006. “The Citadel of Al-Raqqa and Fortifications in the Mid-
dle Euphrates Area.” In Muslim Military Architecture in Greater Syria: From
the Coming of Islam to the Ottoman Period, edited by Hugh Kennedy, 122–50.
Leiden: Brill.
Honigmann, E. 1934. “Thapsacos.” Revue des études byzantines 5A.1: 1272–79.
Howard-Johnston, James. 2010. Witnesses to a World Crisis: Historians and Histo-
ries of the Middle East in the Seventh Century. Oxford: Oxford University Press.
. 2013. “Military Infrastructure in the Roman Provinces North and South of
the Armenian Taurus in Late Antiquity.” In War and Warfare in Late Antiquity:
Current Perspectives, edited by Alexander Sarantis and Neil Christie, 853–92.
Leiden: Brill.
Hoyland, Robert G. 2015. In God’s Path: The Arab Conquests and the Creation of
an Islamic Empire. Oxford: Oxford University Press.
Isaac, Benjamin. 1990. The Limits of Empire: The Roman Army in the East. Oxford:
Clarendon Press.
. 1997. “The Eastern Frontier.” In The Cambridge Ancient History, 13: The
Late Empire, AD 337–425, edited by Averil Cameron and Peter Garnsey, 437–
60. Cambridge: Cambridge University Press.
James, Simon. 2011. “Strategems, Combat, and ‘Chemical Warfare’ in the Siege
Mines of Dura-Europos.” American Journal of Archaeology 115.1: 69–101.
Jastrzębowska, Elżbieta. 2013. “Christianisation of Palmyra: Early Byzantine Church
in the Temple of Bel.” Studia Palmyreńskie 13: 177–91.
Jeffreys, Elizabeth, Michael Jeffreys, and Roger Scott, trans. 1986. The Chronicle
of John Malalas. Byzantina Australiensia 4. Melbourne and Sydney: Australian
Association for Byzantine Studies.
Johnson, Stephen. 1983. Late Roman Fortifications. London: Batsford.
Jones, Horace L., ed. and trans. 1930. Strabo, Geography, Volume VII: Books
15–16. Cambridge, MA: Harvard University Press.
Juchniewicz, Karol, Khaled As’ad, and Khalil al-Hariri. 2010. “The Defense Wall in
Palmyra after Recent Excavations.” In Studia Palmyreńskie XI, edited by Michal
Gawlikowski, 43–48. Warsaw: Wydawnictwa Uniwersytetu Warszawskiego.
Kaegi, Walter Emil. 2003. Heraclius, Emperor of Byzantium. Cambridge: Cam-
bridge University Press.
Kaldellis, Anthony, trans. 2014. The Wars of Justinian. Indianapolis: Hackett.
Karnapp, Walter. 1976. Die Stadtmauer von Resafa in Syrien, Denkmäler antiker
Architektur, Bd. 11. Berlin: De Gruyter.
Kazhdan, Alexander P. ed. 1991. Oxford Dictionary of Byzantium. Oxford: Oxford

University Press.
Kennedy, Hugh. 2010. “Syrian Elites from Byzantium to Islam: Survival or Extinc-
tion?” In Power and Politics in Early Islamic Syria, edited by John Haldon,
181–200. Abdingdon: Routledge.
. 2011. “The Feeding of the Five Hundred Thousand: Cities and Agriculture
in Early Islamic Mesopotamia.” Iraq 73: 177–99.
Keser-Kayaalp, Elif, and Nihat Erdogan. 2017. “Recent Research on Dara/Anasta-
siopolis.” In New Cities in Late Antiquity: Documents and Archaeology, edited
by Efthymios Rizos, 153–76. Turnhout: Brepols.
Konrad, Michaela. 1999. “Research on the Roman and Early Byzantine Frontier in
North Syria.” Journal of Roman Archaeology 2: 392–410.
Konrad, Michaela, Hans Roland Baldus, and Thilo Ulbert. 2001. Der Spätrömische
limes in Syrien: Archäologische Untersuchungen an den Grenzkastellen von
Sura, Tetrapyrgium, Cholle und in Resafa. Mainz: von Zabern.
Lauffray, Jean. 1983. Halabiyya-Zenobia: place forte du Limes Oriental et la Haute-
Mesopotamie au VIe Siècle. Paris: Geuthner.
Leadbetter, Bill. 2009. Galerius and the Will of Diocletian. London: Routledge.
Lee, A. Doug. 2007. War in Late Antiquity: A Social History. Oxford: Blackwell.
Leone, Anna, ed. forthcoming. The Excavation of Dibsi Faraj (Syria): Life on the
Middle Euphrates (1st-13th Century AD). Washington, DC: Dumbarton Oaks.
Leriche, Pierre. 1993. “Techniques de guerre sasanianes et romaines à Doura-Euro-
pos.” In L’armée romaine et les barbares du IIe au VIIe siècle, edited by Françoise
Vallet and Michel Kazanski, 83–100. Rouen: Association francaise d’archéologie
mérovingienne et Musée des Antiquités nationales.
Liebeschuetz, John Hugo Wolfgang Gideon. 1977. “The Defences of Syria in the
Sixth Century.” In Studien zu den Militärgrenzen Roms, 2: Vorträge des 10.
internationalen Limeskongresses in der Germania Inferior, edited by Dorothea
Haupt and Heinz Günter Horn, 487–99. Köln: Rheinland-Verlag.
Lippolis, Carlo. 2007. “La fortezza romana di Kifrin.” In Giornata Lincea in onore
di Giorgio Gullini (Roma, 10 maggio 2006), 151–64. Rome: Accademia nazio-
nale dei Lincei.
Mango, Marlia. 1991. “Kallinikos.” In The Oxford Dictionary of Byzantium, edited
by Alexander Kazhdan, 1094. Oxford: Oxford University Press.
. 2011. “Byzantine Settlement Expansion in North Central Syria: The Case of
Androna/Andarin.” In Le Proche-Orient de Justinien aux Abbassides: peuplement
et dynamiques spatiales, edited by Antoine Borrut, 93–122. Turnhout: Brepols.
. 2017. “Androna and the Late Antique Cities of Oriens.” In New Cities in
Late Antiquity: Documents and Archaeology, edited by Efthymios Rizos, 187–
204. Turnhout: Brepols.
Marcillet-Jaubert, Jean. 1982. “Les inscriptions greques de Hallabat, II.” Annual of
the Department of Antiquities of Jordan 26: 145–58.
Martindale, John Robert. 1992. The Prosopography of the Later Roman Empire 3,
A.D. 527–641. Cambridge: Cambridge University Press.
Matthews, John. 1989. The Roman Empire of Ammianus. Baltimore, MD: Johns
Hopkins University Press.
Meier, Mischa. 2003. Das andere Zeitalter Justinians: Kontingenzerfahrung und
Kontingenzbewältigung im 6. Jahrhundert n. Chr. Göttingen: Vandenhoeck und
Ruprecht.
. 2009. Anastasios I.: die Entstehung des byzantinischen Reiches. Stuttgart:
Klett-Cotta.
Meinecke, Michael. 1991. “Raqqa on the Euphrates: Recent Excavations at the Resi-
dence of Harun Er-Rashid.” In The Near East in Antiquity: German Contribu-
tions to the Archaeology of Jordan, Palestine, Syria, Lebanon and Egypt II,
edited by Susanne Kerner, 17–32. Amman: Goethe-Institut Amman.
Millar, Fergus. 1993. The Roman Near East, 31 B.C.–A.D. 337. Cambridge, MA:
Harvard University Press.
Mommsen, Theodor, et al., eds. 1905. Theodosiani libri xvi cum constitutionibus
Sirmondianis. Berlin: Weidmann.
Morony, Michael. 2004. “Population Transfers between Sasanian Iran and the Byz-
antine Empire.” In Convegno internazionale, la Persia e Bisanzio: Roma, 14–18
Ottobre 2002, 161–79. Rome: Academia nazionale dei Lincei.
Mouterde, René, A. Poidebard, J. Lauffray, and Soubhi Mazloum. 1945. Le limes de
Chalcis: organisation de la Steppe en Haute Syrie Romaine, documents aériens
et épigraphiques. Paris: Librairie Orientaliste Paul Geuthner.
Müller, Karl O., ed. 1878–1885. Fragmenta historicorum graecorum. Paris: Ambro-
sio Firmin-Didot.
Munzi, Massimiliano. Forthcoming. “The Coins from Dibsi Faraj.” In The Excava-
tion of Dibsi Faraj: Life in the Middle Euphrates, edited by Anna Leone. Wash-
ington, DC: Dumbarton Oaks.
Naval Intelligence Division, Great Britain. 1942. Turkey, Geographical Handbook
Series 507. Oxford: Oxford University Press.
Onur, Fatih. 2016. “The Anastasian Military Decree from Perge in Pamphylia,
Revised 2nd Edition.” Gephyra 14: 133-212.
Palmer, Andrew. 1990. Monk and Mason on the Tigris Frontier: The Early History
of Tur ‘Abdin. Cambridge: Cambridge University Press.
Parker, S. Thomas. 1991. “The Nature of Roma’s Arabian Frontier.” In Roman
Frontier Studies 1989: Proceedings of the XVth International Congress of
Roman Frontier Studies, edited by Valerie A. Maxfield and Michael J. Dobson,
498–504. Exeter: Exeter University Press.
Pharr, Clive, trans. 1952. The Theodosian Code and Novels and the Sirmondian
Constitutions. Princeton, NJ: Princeton University Press.
Pollard, Nigel. 2000. Soldiers, Cities, and Civilians in Roman Syria. Ann Arbor:
University of Michigan Press.
Rackham, Harris, ed. and trans. 1942. Pliny, Natural History, Volume II: Books
3–7. Cambridge, MA: Harvard University Press.
Rolfe, John C., ed. and trans. 1963. Ammianus Marcellinus, Res Gestae. Cambridge,
MA. Harvard University Press.
Rowton, M. B. 1977. “Dimorphic Structure and the Parasocial Element.” Journal of

Near Eastern Studies 36.3: 181–98.
Rubin, Berthold. 1960. Das Zeitalter Iustinians. Berlin: Walter de Gruyter.
Salway, Benet. 2005. “The Nature and Genesis of the Peutinger Map.” Imago Mundi
57.2: 119–35.
Sanders, Guy D. R., Sarah A. James, and Alicia Carter Johnson. 2017. Corinth Exca-
vation Archaeological Manual. Grand Forks: The Digital Press of the University
of North Dakota.
Sarantis, Alexander. 2013a. “Fortifications in the East: A Bibliographic Essay.” In
War and Warfare in Late Antiquity: Current Perspectives, edited by Alexander
Sarantis and Neil Christie, 317–70. Leiden: Brill.
. 2013b. “Waging War in Late Antiquity.” In War and Warfare in Late Antiq-
uity: Current Perspectives, edited by Alexander Sarantis and Neil Christie, 1–85.
Leiden: Brill.
. 2016. Justinian’s Balkan Wars: Campaigning, Diplomacy and Development
in Illyricum, Thrace and the Northern World AD 527–65. Prenton: Francis
Cairns Publications.
. Forthcoming. “The Socio-Economic Impact of Raiding on the Eastern and
Balkan Borderlands of the Eastern Roman Empire, 502–602.” Millennium.
Sarantis, Alexander, and Neil Christie. 2013. “Fortifications in the West: A Biblio-
graphic Essay.” In War and Warfare in Late Antiquity: Current Perspectives,
edited by Alexander Sarantis and Neil Christie, 255–98. Leiden: Brill.
Sarris, Peter. 2011. Empires of Faith: The Fall of Rome to the Rise of Islam, 500–
700. Oxford: Oxford University Press.
Sauer, Eberhard W., Jebrael Nokandeh, Konstantin Pitskhelauri, Ivane Javakhishvili,
and Hamid O. Rekavandi. 2017. “Innovation and Stagnation: Military Infra-
structure and the Shifting Balance of Power between Rome and Persia.” In Sasa-
nian Persia: Between Rome and the Steppes of Eurasia, edited by Eberhard W.
Sauer, 241–68. Edinburgh: Edinburgh University Press.
Schwarze, Marcel F. 2017. Römische Militärgeschichte 2: Studie zur römischen
Armee und ihrer Organisation im Sechsten Jahrhundert n. Chr. Pfungstadt:
Books on Demand.
Segal, Judah B. 1970. Edessa: “The Blessed City.” Oxford: Clarendon Press.
Seyrig, Henri. 1950. “Antiquités syriennes.” Syria 27.3–4: 229–52.
Shahîd, Irfan. 1984. Byzantium and the Arabs in the Fourth Century. Washington,
DC: Dumbarton Oaks Research Library and Collection.
. 1989. Byzantium and the Arabs in the Fifth Century. Washington, DC:
Dumbarton Oaks Research Library and Collection.
. 1995. Byzantium and the Arabs in the Sixth Century, Vol. 2.2. Washington,
DC: Dumbarton Oaks Research Library and Collection.
Smith, Roland R. R. 1999. “Telling Tales: Ammianus’ Narrative of the Persian Expe-
dition of Julian.” In The Late Roman World and Its Historian: Interpreting
Ammianus Marcellinus, edited by Jan W. Drijvers and David Hunt, 89–104.
London: Routledge.
Stückelberger, Alfred, and Florian Mittenhuber, eds. and trans. 2009. Klaudios
Ptolemaios Handbuch der Geographie: Ergänzungsband mit einer Edition des
Kanons bedeutender Städte. Basel: Schwabe.
Talbert, Richard J. A. 2010. Rome’s World: The Peutinger Map Reconsidered. Cam-
bridge: Cambridge University Press.
Thurn, Ioannes, ed. and trans. 2000. Ioannis Malalae Chronographia. Berlin: W. de
Gruyter.
Trombley, Frank R., and John W. Watt, trans. The Chronicle of Pseudo-Joshua the
Stylite. Translated Texts for Historians 32. Liverpool: Liverpool University Press.
Ulbert, Thilo. 2000. “Procopius, De Aedificiis: Einige Überlegungen zu Buch II,
Syrien.” Antiquité Tardive 8: 137–47.
Van Berchem, Denis. 1954. “Recherches sur la chronologie des enceintes de Syrie et
Mésopotamie.” Syria 31.3–4: 254–70.
Walmsley, Alan. 2007. Early Islamic Syria: An Archaeological Assessment. London:
Duckworth.
Whately, Conor. 2013. “Organisation and Life in the Late Roman Military: A Bib-
liographic Essay.” In War and Warfare in Late Antiquity: Current Perspectives,
edited by Alexander Sarantis and Neil Christie, 209–38. Leiden: Brill.
Whitby, Michael. 1986a. “Procopius and the Development of Roman Defences in
Upper Mesopotamia.” In The Defence of the Roman and Byzantine East: Pro-
ceedings of a Colloquium Held at the University of Sheffield, April 1986, edited
by David L. Kennedy and Philip Freeman, 717–35. Oxford: BAR.
. 1986b. “Procopius’ Description of Dara.” In The Defence of the Roman and
Byzantine East: Proceedings of a Colloquium Held at the University of Sheffield,
April 1986, edited by Philip Freeman and David L. Kennedy, 737–87. Oxford:
BAR.
. 1988. The Emperor Maurice and His Historian: Theophylact Simocatta on
Persian and Balkan Warfare. Oxford: Clarendon Press.
. 2013. “Siege Warfare and Counter-Siege Tactics in Late Antiquity (Ca. 250–
640).” In War and Warfare in Late Antiquity: Current Perspectives, edited by
Alexander Sarantis and Neil Christie, 433–59. Leiden: Brill.
. 2017. “Justinian and Persia, 527–562.” In The Encyclopedia of Ancient
Battles, edited by Michael Whitby and Harry Sidebottom, 1199–1213. Oxford:
Oxford University Press.
Whitby, Michael, and Mary Whitby, trans. 1986. The History of Theophylact Simo-
catta: An English Translation with Introduction and Notes. Oxford: Oxford
University Press
Whittow, Mark. 1999. “Rome and the Jafnids: Writing the History of a 6th-c. Tribal
Dynasty.” In The Roman and Byzantine Near East: Some Recent Archaeologi-
cal Work, vol.2. JRA Supplementary Series 31, edited by John H. Humphrey,
207–24. Portsmouth, RI: Journal of Roman Archaeology.
Wickham, Chris. 2005. Framing the Early Middle Ages: Europe and the Mediter-
ranean 400–800. Oxford: Oxford University Press.
Wilkinson, T. J., and Guillermo Algaze. 1990. Town and Country in Southeastern
Anatolia. Chicago: Oriental Institute of the University of Chicago.
Wilkinson, T. J., and D. J. Tucker. 1995. Settlement Development in the North Jazira,
Iraq: A Study of the Archaeological Landscape. Warminster: Aris and Phillips.
Zanini, Enrico. 1995. “Il restauro giustinianeo delle mura di Palmyra.” In Arte
profana, arte sacra a Bisanzio, edited by Antonio Iacobini and Enrico Zanini,
65–104. Rome: Argos.
. 2003. “Christian Topography in the Late Antique Town: Recent Results and
Open Questions.” In Theory and Practice in Late Antique Archaeology, edited
by Luke Lavan and William Bowden, 196–224. Leiden: Brill.

Project Muse 772168

Uploaded by

Copyright:

Available Formats

Project Muse 772168

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Project Muse 772168

Uploaded by

Copyright:

Available Formats

The Middle Euphrates and Its Transformation from the Third

to the Seventh Centuries: The Case of Dibsi Faraj

Anna Leone, Alexander Sarantis

Published by Johns Hopkins University Press

For additional information about this article

[ Access provided at 14 Nov 2022 12:22 GMT from Durham University ]

The Middle Euphrates and

The remains of the fortified settlement of Dibsi Faraj bear witness to an

in collaboration with the Department of Antiquities of Syria, as one of the

The Location and Identification of Dibsi Faraj

which it controlled access to seasonal grazing areas and fertile agricultural

Figure 1. Map of the Region (by Abby George)

uncharacteristically, by Procopius.18 This inference is reinforced by the texts

The Archive of the Excavation at Dibsi Faraj

The Citadel and Its Phases

The First Fortification at the End of the Third Century

The location of Dibsi presumably was determined by the topography of the

The Fourth- and Fifth-century Interventions:

Figure 5. Aerial photograph of the citadel church (Harper’s archive)

perspective, it is worth mentioning the presence of a narthex, quite uncom-

Figure 6. Mosaic of the citadel church with inscriptions (Harper’s archive)

Figure 7. Inscription in the extra-mural basilica (Harper’s archive)

According to the tradition of the cult of Saint Sergius, which became

five main components: 1) an entrance corridor or vestibule to the north, lead-

Middle Euphrates frontier region.54 The main military commander responsi-

The Anastasian Phase

Figure 10b. Aerial photograph of corner tower (Harper’s archive)

reign of Anastasius. According to analysis by Denis Feissel, the inscription

inscriptions on occasion acknowledge the distant authority of the emperor,

The Post-Anastasian Sixth Century

fourth-century or Anastasian wars. Indeed, following in Shapur I’s footsteps,

The Seventh Century

Conclusions: Dibsi Faraj in Its Historical Context

Römisch-Germanisches Zentralmuseum Mainz /

Kazhdan, Alexander P. ed. 1991. Oxford Dictionary of Byzantium. Oxford: Oxford

Rowton, M. B. 1977. “Dimorphic Structure and the Parasocial Element.” Journal of

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.