The mosaic of Theodoulos from Sousse (Tunisia) - article ; n°1 ; vol.16, pg 229-239
12 Pages
Downloading requires you to have access to the YouScribe library
Learn all about the services we offer

The mosaic of Theodoulos from Sousse (Tunisia) - article ; n°1 ; vol.16, pg 229-239


Downloading requires you to have access to the YouScribe library
Learn all about the services we offer
12 Pages


Antiquités africaines - Année 1980 - Volume 16 - Numéro 1 - Pages 229-239
11 pages
Source : Persée ; Ministère de la jeunesse, de l’éducation nationale et de la recherche, Direction de l’enseignement supérieur, Sous-direction des bibliothèques et de la documentation.


Published by
Published 01 January 1980
Reads 98
Language English
Document size 1 MB


David Parrish
The mosaic of Theodoulos from Sousse (Tunisia)
In: Antiquités africaines, 16,1980. pp. 229-239.
Citer ce document / Cite this document :
Parrish David. The mosaic of Theodoulos from Sousse (Tunisia). In: Antiquités africaines, 16,1980. pp. 229-239.
doi : 10.3406/antaf.1980.1066
Antiquités africaines
t. 16, 1980, p. 229-239
The Mosaic of Theodoulos ' holds special interest as one of the best known, yet least discussed,
examples of later North African pavement art (fig. 1). This mosaic, which comes from Sousse, Tunisia,
and is now located in the Archaeological Museum in that city 2, forms an unusual variation of a type of
paradisial imagery common to both pagan and Christian art, whose precise meaning in the Theodoulos
pavement has never been satisfactorily explained. There is also uncertainty about this mosaic's date, for
although the majority of scholars have assigned it to the Byzantine era, or sixth century A.D. 3, there
is evidence for extending the range of possible dates to the period of the fifth to mid-sixth centuries. In
what follows, I shall consider each of these issues and provide a close description of the Theodoulos Mosaic.
It is hoped that the results of this analysis will shed new light on this work of art and on related pavements
of the North African corpus, which pose similar problems of dating and interpretation.
The nearly-square Mosaic of Theodoulos, which measures approximately 2.30 m. on a side, shows
a centrally-placed gold crater with grapevines and birds, immediately above which appears a palm tree.
At the top of the pavement is the Greek name ΘΕΟΔΟΥΛΟΥ, written in the genitive case. A large gap
occurs in the mosaic's lower right corner, and other, smaller gaps are visible in the field and border. A row
of alternately-reversed lotuses with white tops and striped bases, colored reddish-pink, ochre, and dark
1 Excavated in 1902. See esp. : Gouvet (M.), Séance de la Commission de l'Afrique du Nord, 13 janvier 1903. B.C.T.H.,
1903, p. cxxxix-xl. — Gauckler (P.), Note sur les mosaïstes antiques. M. S.A. F., t. 63, 1904, p. 196, 3°, n° 2; Id, Inventaire
des mosaïques de la Gaule et de l'Afrique, t. 2, Afrique proconsulaire (Tunisie). Paris, 1910, p. 63 and pi. (hereafter cited as Inv.
Tun.) — C.I.L. VIII, 32011 a. — Truillot (Α.), Séance de la Commission de l'Afrique du Nord, 13 novembre 1944. B.C.T.H.,
1944, p. 304-305 — Picard (G.), Une schola de collège à Carthage. Karthago, t. 3, 1951-1952, p. 178, n. 15, 2° — Alexander
(M.), Early Christian Tomb Mosaics of North Africa (diss. New York University, 1958), t. 1, p. 84, 123 and fig. 70 (Hereafter
cited as Alexander) — Cintas (J.) and Duval (N.), L'Eglise du prêtre Félix (région de Kélibia) Karthago, t. 9, 1958, p. 238-
239 and pi. XXXVII, a (Hereafter sections of the article by Duval are cited as Duval, — Foucher (L.), Inventaire des
mosaïques. Feuille n° 57 de l'Atlas archéologique, Sousse. Tunis, 1960, p. 75, n° 57. 164 and pi. XXXVIII ; Id., Hadrumetum.
Tunis, 1960, p. 362 and n. 1486 — Dunbabin (K.), The Mosaics of Roman North Africa. Oxford, 1978, p. 169, n. 174, 193,
n. 29 et col. pi. F.
2 Mus. Inv. no. 57. 164 ; Sousse is ancient Hadrumetum.
3 Gauckler {supra n. 1) p. 196, n. 5 — Picard (supra n. 1), Loc. cit. — Alexander, t. 1, p. 84 — Duval, Kélibia, p. 239,
— Foucher (supra n. 1, 1960), loc. cit. D. PARRISH 230
Fig. 1 . — Mosaic of Theodoulos, overall view.
green-olive and separated by a wavy white ribbon, frames the field. The border curves inward at the lower
left to skirt the well-head of a cistern, over which the pavement was originally laid 1. Unfortunately, the
function of the building which this mosaic decorated is unknown.
Several details of this pavement image are noteworthy (fig. 2). The tall crater has a distinctive form,
characterized by a sharply-receding belly decorated with a narrow band, a flaring mouth, S-shaped handles,
a low bowl with gadroons, and a sloping base that rises to small points at the sides and is separated from
the bowl by a small knob. Inside the vessel is a minute black and white checkerboard, perhaps representing
liquid. The two symmetrical grapevines rise in pairs of alternately large and small volutes, laden with
bunches of pink, yellowish, and light and dark green grapes, on which some of the birds feed. Starting at
the panel's base and proceeding upward, the volutes contain four pairs of naturalistically colored birds
of the following species : partridges, peacocks, ducks, and pheasants. The partridge on the viewer's left
and the three upper pairs of birds all face inward. The left pheasant is inverted to draw attention to the
adjacent inscription. The right partridge faces outward, perhaps to emphasize visually the majestic peacocks
above. The image of abundance of the grapevines is echoed by the palm tree's clusters of ripe dates.
1 Gouvet (supra n. 1 p. 229), p. cxxxix. The ancient building in which the mosaic was found is now covered by a modern
apartment house, located at no. 8, rue de l'Église. THE MOSAIC OF THEODOULOS 231
Fig. 2. — Mosaic of Theodoulos, detail of left side.
Various motifs in the Theodoulos Mosaic have paradisial associations common to both pagan and
Christian art. One such motif, symbolizing divine abundance and heavenly bliss, is the crater with spread
ing grapevines inhabited by birds. This image occurs, for example, in each of the four corners of a third 232 D. PARRISH
Dionysos' triumph ί . A Christian work of art with century pagan pavement from El Jem, representing
this motif is a late fourth to early fifth century tomb mosaic from Kélibia, which shows a crater with
inhabited vine scrolls supporting a funerary inscription 2.
The heraldic pair of peacocks and crater in the Sousse panel form another distinct paradisial image,
which elsewhere may appear alone 3, but which is subtly woven into the overall floral scheme of the
Theodoulos Mosaic. The peacocks are visually accented in various ways 4, yet occupy plant volutes
like the other birds. In a fourth century pavement from Cherchel 5, the two heraldic peacocks remain
separated from the surrounding grapevines and other animals. By incorporating the of his
mosaic into the grapevines, the Sousse artist created a unified set of four different types of birds. This
particular selection of birds may have no special significance, since each of the species shown forms part
of a common repertory of birds which appear in other North African pavements of paradisial content 6.
Yet the precise combination of birds in the Theodoulos panel is unique, and suggests the possibility
that these animals symbolize the four Seasons. AU of the species represented have associations with the
four Seasons in earlier African mosaics. The various birds are linked with the following Seasons elsewhere :
ducks-Winter ; peacocks-Spring ; partridges-Summer ; and pheasants- Autumn 7. A seasonal interpretation
would perfectly accord with the paradisial theme of the rest of the Theodoulos Mosaic, since the Seasons
themselves have celestial connotations 8. The chief objection to such an interpretation is the lack of other,
corroborative references to the Seasons, such as human or plant symbols, in the same mosaic. The fact
that all the birds in the Sousse example are surrounded by grapes rather than plants representative of the
different times of year makes the situation more doubtful. Nevertheless, because of the precise number
and types of birds shown a seasonal reference remains possible.
The one other motif of paradisial significance in the Theodoulos Mosaic is the palm tree. The palm,
as the crater, can have a meaning in both pagan and Christian art ; the former motif symbolizes
the Tree of Life 9. A pagan example with this association is a Roman tombstone which shows a man and
his wife at a funerary banquet, flanked by a palm tree with ripe fruit 10. Among Early Christian works of
art with paradisial palms are the two apse mosaics in the fourth century church of Santa Costanza in Rome,
1 Yacoub (M.), Le Musée du Bardo. Tunis, 1970, p. 62, Α. 287 and fig. 63. Besides birds, the vines in this pavement
enclose putti, quadrupeds, a satyr and a silene.
2 Duval, Kélibia, p. 187, no. 11 and pi. XVII, b.
3 Cf a fifth century baptistery mosaic form Stobi, Wiseman (J.), and Mano-Zissi (D.), Excavations at Stobi, 1971.
A.J.A., t. 76, 1972, p. 423 and pi. 90, fig. 47.
4 In addition to their heraldic placement alongside the crater, the peacocks are emphasized by their large size, by the
fact that both of these birds feed on bunches of grapes, and by the outward-turned partridge below.
5 Gsell (S.), Antique lol-Caesarea. (Alger, n.d.) p. 102 and fig. on 103.
6 Cf Duval, Kélibia, p. 187, no. 1 1 and pi. XVII, b ; 189, no 16 and pi. XV, b ; 199-200, no 44 and pi. XXIV, a ;
200-201, no. 46 and pi. XVIII, a.
7 All of these birds except Spring's peacock appear as seasonal symbols, identified by plants, in a third century mosaic
from El Jem representing Diana the Huntress and the Seasons, Yacoub (supra n. 1) p. 130, Inv. 2751 and fig. 139; there,
a dove or pigeon signifies Spring. Another Season panel of late third to early fourth century date from Haïdra associates pea
cocks with Spring, Baratte (F.), and Duval (N.), Les mines d'Ammaedara-Haïdra. Tunis, 1974, p. 71 and fig. 29 ; the peacocks
frame a seasonal personification, who is identified by his basket of roses.
8 When shown, e.g., on Roman sarcophagi ; see Hanfmann (G.), The Season Sarcophagus in Dumbarton Oaks. Camb
ridge, Mass., 1951, t. I, p. 238 ; also Cumont (F.), Recherches sur le symbolisme funéraire des romains. Paris, 1942, p. 489 ff.
9 For a discussion of the palm motif as a paradisial symbol in pagan Roman art, see Goodenough (E.), Jewish Symbols
in the Greco- Roman Period, t. VII, 1. New York, 1958, p. 107-116; for its use in Christian art, see Leclercq (H.), Palme,
Palmier, Dictionnaire d'archéologie chrétienne et de liturgie, éd. F. Cabrol and H. Leclercq, t. 13, Paris, 1937, p. 947-961.
10 Amelung (W.), Die Skulpturen des Vaticanischen Museums, t. 2. Berlin, 1908, p. 624-626, no. 411 and pi. 52, fig. 411. THE MOSAIC OF THEODOULOS 233
which represent Christ and some of his apostles in heavenly settings that include palm trees 1. The fact
that the Theodoulos Mosaic includes both the palm and crater (plus its associated birds) makes its parad
isial meaning emphatic.
This combination of motifs may also indicate the pavement's specific religious content, and in an
original way. Many other works of art besides the Sousse panel join the palm and crater in a common
format, and all of these examples, for some unexplained reason, are either Christian or Jewish. None are
pagan. One example is a late fourth to early fifth century Christian tomb mosaic from Kélibia 2, which
shows three superimposed registers, with a crater and emerging plants placed below, a pair of palm trees
at the top, and two peacocks in between. In a different arrangement, the now-divided central panel of the
mosaic which decorated the main sanctuary of the fifth century synagogue at Hammam Lif 3 represented
a central crater with heraldic peacocks, flanked by two palm trees. Because the Theodoulos pavement
juxtaposes the same motifs of crater and palm (as well as peacocks), it therefore most likely had either a
Christian or Jewish usage. Pagan mosaics, by contrast, may associate other types of plants, such as millet,
with a crater and grapevines 4, but not the palm.
Of the two possible religious identifications for the Theodoulos panel, the Christian one seems the
more likely. Some of the reasons are artistic. The only other mosaic besides the Sousse pavement which
shows a palm (or palms) placed above a crater, i.e., the aforementioned panel from Kélibia 5, is Christian.
Jn that case, the palms are separated from the vase by an intervening register of birds. The Theodoulos
Mosaic is unique in showing its palm placed immediately above the crater. Apparently, no Jewish pave
ments employ this vertical alignment of motifs. Another reason for labeling the Sousse panel Christian
is the close resemblance between its overall design of a central crater and numerous bird-filled grapevines
and a mosaic generally attributed to the sixth century A.D., which comes from the choir of the small
church at Sidi-Abich 6, a site near Sousse (fig. 3). Both of these mosaics also have a similar squarish
shape. The pavement, which is dedicated to a deceased Christian, thus has all the same artistic
elements as the Theodoulos panel, except for a palm tree. The two comparisons from Kélibia and Sidi-
Abich argue strongly in favor of identifying the Sousse mosaic as Christian.
The one other piece of evidence in the Theodoulos pavement which may support a Christian interpre
tation is the inscription. The name Theodoulos (« Theodolus » or « Theodulus » in its typical Latin trans
literation) had a primarily Christian usage in Late Roman times 7. Indeed, the meaning of the name,
« servant of god », represents a concept foreign to Greek and Roman pagan tradition 8, but popular
among early Christians, who frequently used the name Theodoulos 9. Nevertheless, a few pagan instances
of this name have been identified, and among these one epigraphist includes the inscription in the mosaic
from Sousse 10. No reason is given, however, for calling that example pagan, so we must consider the iden-
1 Matthiae (G.), Mosaici medioevali delle chiese di Roma. Rome, 1967, t. I, p. 35-41 ; t. II, fig. 27-28.
2 Duval, Kélibia, p. 195-96, no. 35 and pi. XX, a.
3 Biebel (F.), The Mosaics of Hammam Lif. The Art Bulletin t. 18, 1936, p. 549 and fig. 23 ; dating, 550.
4 Yacoub {supra n. 1 p. 232), p. 121, Inv. 3341 and fig. 130.
5 Supra n. 2.
6 Inv. Tun., no. 248, 2°. This mosaic is now in the museum at Enfidaville.
7 See Pape (W.), Benseler (G.), Wörterbuch der griechischen Eigennamen. Repr. Graz, 1959, p. 490 Θεόδουλος, where
the majority of examples cited seem to be Christian ; also Kajanto (I.), Onomastic Studies in the Early Christian Inscriptions
of Rome and Carthage. Acta Instituti Romani Finlandiae, t. II, 1, 1963, p. 117.
8 The concept of hierodulia underlying the name Theodoulos apparently has an Oriental source ; see Sittig (E.), De
Graecorum nominibus theophoris. Dissertationes Philologicae Halenses, t. 20, 1912, p. 164.
9 Kajanto {supra n. 7), p. 103.
10 Ibid. 234 D. PARRISH
Fig. 3. — Mosaic from Sidi-Abich, view of lower half.
tification questionable. In view of the popularity of the name Theodoulos among Christians generally,
it seems preferable to allow the possibility that the Sousse inscription is Christian, and in fact to consider
it more likely Christian than pagan. Thus, a variety of Christian references, both artistic and epigraphic,
seems to be present in the Theodoulos Mosaic, suggesting we interpret this pavement as a Christian work
of art, despite the absence of any explicit religious symbol, such as a cross.
We must now try to determine how the Theodoulos pavement's inscription relates to its imagery.
Who was Theodoulos ? The use of the genitive case in the inscription allows three possible interpretations. \
One alternative, favored by several scholars, is that the name Theodoulos is the mosaic artist's signature i.
The fact that the single-name, genitive inscriptions of certain other African pavements have been identified
as artists' signatures supports this suggestion 2. Theodoulos, therefore, may have been a Christian artist.
A second possibility is that the use of the genitive at Sousse implies ownership in the sense of personal
property, thus indicating that Theodoulos was the mosaic owner 3. Unfortunately, we have no archaeolog
ical evidence from the destroyed building which the pavement decorated to determine what type of
establishment it was, and whether it might have had a private owner.
I now propose a third alternative, namely that Theodoulos was a deceased person, honored by the
mosaic, but buried elsewhere. The pavement did not cover a grave, but a cistern. The purpose of the mosaic
would thus be commemorative, but not funerary. In this interpretation, the closest analogy we can make
with the Theodoulos panel's inscription is the Christian, single-name Greek inscriptions in the genitive
case which have been found on painted ceramic tiles in the catacombs of Rome 4. Other, comparable tiles,
bearing Latin, rather than Greek, single names in the genitive, combined with a date or the phrase in pace,
have been discovered in the catacombs at Sousse 5. This epigraphic analogy depends on our being able
to identify the name Theodoulos as Christian, which is a likelihood, but not a certainty. The pavement
would thus be related to the tiles in its function of honoring a dead person, but would differ from them in
the fact that it did not mark a place of burial. Perhaps the site where the mosaic occurred had other asso
ciations with Theodoulos, for example as the house where he lived or the church of which he was a member.
This commemorative interpretation of the mosaic inscription would also perfectly accord with the parad
isial theme of the pavement's imagery. The type of design used in this mosaic, i.e., a crater with floral
forms and birds, plus a dedication at the top, is a common one in Christian tomb pavements from Africa 6,
which would be related to the Theodoulos pavement in their commemorative purpose. Moreover, one of
these tomb mosaics, from Kélibia 7, includes a palm tree (actually, two palms), like the Sousse panel.
Finally, the pavement which the Theodoulos panel most closely resembles in the details of its imagery
(except for the palm), i.e., the mosaic from Sidi-Abich (fig. 3), would also be the most closely related in
function, for it honors a deceased Christian 8, but apparently did not cover a grave.
Among the three possible explanations of the Theodoulos pavement's inscription, I am inclined
toward the latter, commemorative interpretation, because of the close parallels between the Theodoulos
Mosaic and the above-mentioned pavements from Kélibia and Sidi-Abich, which themselves honor
deceased Christians, and like the Theodoulos panel, show their inscriptions prominently placed above
1 G AUCKLER (supra n. 1 p. 229), loc. cit.— Dessau (H.), C.I.L., VIII, 23011a — TRUiLLOT(i«pran. 1 p. 229, 1944), p. 304 —
Duval, Kélibia, p. 239. — Foucher (supra n. 1, p. 229, 1960), loc. cit.
2 For examples, see (L.), Note sur des signatures de mosaïstes. Karthago, t. 9, 1958, p. 131, η. 2 (Macari)
— PoiNSSOT (L.), Séance de la Commission de Γ Afrique du Nord, 10 novembre 1941. B.C.T.H., 1941, p. 150 and n. 4, 6° (The-
bani p. 229). See also infra p. 236 (Ασπασιον).
3 Foucher (supra n. 1 p. 229, 1960), loc. cit.
4 Diehl II. Berlin, 1925, 3961 (Ογρβανου), 3973, 3977B.
5 Leynaud (Α.), Les catacombes africaines : Sousse-Hadrumète. Alger, 1922, 161-62, no. 64; p. 300, n° 51.
6 Cf Alexander I, figs. 29, 43, 45, 71 ; Π, 6-7, no. 8, 7-8, no. 10, 174-75, no. 282, 185-86, no. 295 ; Duval, Kélibia, p.
187, no. 11 and pi. XVII, b ; p. 189, no. 16 and pi. XV, b ; p. 189, no. 17 and pi. XV, a ; p. 190-91, no. 21 and pi. XVI.
7 Supra no. 2 p. 233.
8 The inscription, which is now damaged, is apparently the same as one recorded by P. Monceaux in its complete
form, Séance de la Commission de Γ Afrique du Nord, 30 mai 1905. B.C.T.H., 1905, p. α xxxvrii, 2o, i. The latter reads as follows :
a central crater, or crater plus palms 1. At the same time, we must stress the distinctiveness of the Theo-
doulos Mosaic, both epigraphically and artistically, within the repertory of African pavements of paradisial
content, which makes our present conclusions tentative.
This discussion now turns to the other main issue concerning the Theodoulos pavement, its date.
As before, the lack of archaeological evidence poses a problem, with the result that scholars have relied
on epigraphy and style for dating this monument. Both types of evidence are considered below.
Some scholars have thought the presence of a Greek inscription in the Theodoulos Mosaic sufficient
reason for dating the pavement in Byzantine times, i.e., the sixth century A.D. 2 This conclusion, however,
is unwarranted, for although the majority of Greek inscriptions from North Africa originated during the
Byzantine occupation, beginning in 533 A.D., there exist many other African inscriptions in the Greek
language, which are clearly pre-Byzantine in date. Some of the examples occur in mosaics. One of these
is a late second century pavement from Lámbese, signed with the name Ασπασιον 3. Another mosaic,
from Carthage and of late fourth century date, shows four circus charioteers with their names inscribed
in Greek 4. In addition, there exist several African defixionum tabellae with lengthy Greek inscriptions,
some of which come from Sousse, the site where the Theodoulos pavement itself was found, and which
have been dated in the first to third centuries A.D. 5 The Greek-inscribed, African monuments of pre-
Byzantine date also comprise a lead vase from Carthage of fourth to fifth century date 6, decorated with
a variety of pagan and Christian motifs, including crosses, which almost certainly indicate this is a Christian
work of art ; the Greek inscription is a quotation from the Book of Isaiah 7. It has been argued that this
vessel was made outside Africa 8 ; however, the presence on the vase of a figure who seemingly person
ifies the city of Carthage favors an African origin for the object 9. In view of the above evidence, there
fore, the Greek inscription of the Theodoulos Mosaic can not by itself be used to assign the pavement to
Byzantine times. Nor do any paléographie features of the inscription support such a date.
1 For a drawing of the Sidi-Abich mosaic, showing the location of its inscription within the pavement as a whole,
see Lasteyrie (R. de), U Architecture religieuse en France à V époque romane 2nd ed. Paris, 1929, p. 112, fig. 91. We should
add that the several similarities between the Sousse and Sidi-Abich panels, regarding shape, imagery, and possible commemora
tive references in both, suggest that the Theodoulos Mosaic, like the Sidi-Abich floor, may have once decorated the nave
of a church. In this context, it is interesting to note that a cistern overlaid by a mosaic with an inscription — the same condi
tions which apply to the Theodoulos panel — was found in the nave of Church 3 at Sabratha, although the design of the Tripo-
litanian pavements is geometric rather than floral, and its inscription refers to a mosaic donor, not a deceased person ; see Ward-
Perkins (J.) and Goodchild (R.), The Christian Antiquities of Tripo litania. Archaeologia, t. 95, 1953, p. 17 and pi. VIII, a, d.
2 This seems to be the case with Gauckler {supra n. 1 p. 229), p. 196, 3°, no. 2 and n. 5 ; and Foucher {supra n. 1 p. 229,
1960), loc. cit.
3 Inv. Alg. no. 190. I thank K. Dunbabin for this reference.
4 Dunbabin (K.), The Mosaics and Pavements. Excavations at Carthage Conducted by the University of Michigan,
Tunis, 1976, p. 21, 30-31 and pi. 12.
5 Audollent (Α.), Defixionum tabellae. Paris, 1904, p. 365-66, 378, no. 267 ; p. 373-78, no. 271. For the latter, see also
Maspero (G.), Sur deux tabellae devotionis de la nécropole romaine d'Hadrumète. Études de mythologie et d'archéologie, Paris,
1893, t. 2, p. 303-11 — Deissmann (G.), Bibelstudien. Beitrage, zumeist aus den Papyri und Inschriften, zur Geschichte der
Sprache, des Schrifttums und der Religion des hellenistischen Judentums und des Urchristentums. Marburg, 1895, p. 25-54.
6 Rossi (G. de), Secchia di piombo trovata nella Reggenza di Tunisi. B.A.Crist., t. 5, 1867, p. 77-87 and pi. — Garrucci
(P.), Storia dell'arte cristiana nei primi otto secoli della Chiesa, t. VI. Prato, 1880, p. 33-34 and pi. CDXXVIII, 1-2 — Le
Blant (E.), Les ateliers de sculpture chez ¡es premiers chrétiens. M.E.F.R., t. 3, 1883, p. 439-46 — Goodenough {supra n.
9 p. 232) p. 117 and fig. 129.
7 Isaiah XII, 3. For a discussion of the inscription, see Goodenough {supra η. 6), loc. cit.
8 De Rossi {supra η. 6), p. 85-86.
9 Garrucci {supra η. 6) p. 34. THE MOSAIC OF THEODOULOS 237
The other criterion used to date the Theodoulos pavement is style. In only rare instances have scholars
made their stylistic judgements explicit *, and opinions favor a sixth century date 2. However, one scholar,
apparently also for reasons of style, proposed a fifth century date 3. A major problem in using stylistic
analysis for chronological purposes is the scarcity of securely-dated mosaic comparisons. The best we
can do at present is to indicate the range of possible dates for the Theodoulos panel, restricting our compa-
randa primarily to pavements from the eastern coastal region of Africa Proconsularis.
Various stylistic features of the Theodoulos Mosaic indicate a date within the period of the fifth to
mid-sixth centuries A.D. As previously noted, the Theodoulos pavement's overall design, showing a
central crater and many symmetrically-arranged, bird-filled grapevines, has its closest artistic parallel
in the mosaic from Sidi-Abich (fig. 3), generally dated in the sixth century A.D. 4 Certain details of the
two pavements, such as the tall, gadroon-decorated form of their craters and the rounded shape of their
plant voluves are also similar. The primary evidence for dating the Sidi-Abich panel is architectural ; the
form of baptismal font found in the building which the mosaic adorned is characteristic of Byzantine
churches in North Africa 5. However it is unclear just when the Sidi-Abich mosaic was made in the By
zantine period. This panel is one of a series of pavements from the same church, which were laid in two
successive stages 6. The mosaic of interest to us belonged to the earlier flooring, into which a series of
tomb pavements were later inserted. The sharp contrast in style between the crudely made sepulchral
panels and the more finely executed choir pavement probably reflects a general decline in quality among
later sixth century African mosaics and suggests that the choir pavement at Sidi-Abich was made signif
icantly earlier than the tomb mosaics. The former pavement probably has an upper chronological limit in
the middle of the sixth century A.D. Because of the resemblance between this mosaic and the Sousse panel,
we can assume a similar upper limit for the latter pavement.
At the same time, the Theodoulos Mosaic's style suggests a lower chronological terminus in the fifth
century A.D. Some of the features which the Theodoulos and Sidi-Abich pavements share, such as a tall
crater with receding belly, S-shaped handles, and gadroons, also occur in other North African mosaics
(e.g., fig. 4, a pavement from Sousse), which have been dated on archaeological and stylistic grounds to
the late fourth to fifth centuries A.D. 7 The rounded plant volutes of the Theodoulos and Sidi-Abich
floors also have late fourth to early fifth century counterparts 8. In fact, by comparison with the vine
scrolls from Sidi-Abich, the volutes of the Theodoulos panel appear somewhat more open and lightly
drawn, in this respect resembling the plant forms of a late fourth to early fifth century mosaic from Kéli-
bia 9. Sixth century vine forms tend to have a denser, more crowded appearance 10. We also note the parti-
1 E.g., Duval, Kélibia, p. 238-39.
2 E.g., Ibid. — Picard {supra n. 1 p. 229), loc. cit.
3 Hinks (R.), Catalogue of the Greek, Etruscan and Roman Paintings and Mosaics in the British Museum. London,
1933, p. 125.
4 E.g., N. Duval dates the building where the Sidi-Abich mosaic was found as « incontestably Byzantine» (i.e., sixth
century), Les mosaïques funéraires de VEnfida et la chronologie des mosaïques funéraires de Tunisie. R.A.C., t. 50, 1974 (Mis
cellanea in onore di Luciano de Bruyne e Antonio Ferrua S.I. Ili), p. 173.
5 Freshfield (E.), Cellae Trichorae and other Christian Antiquities in the Byzantine Provinces of Sicily with Calabria
and North Africa including Sardinia, t. II. London, 1918, p. 146.
6 Alexander, t. 1, p. 16-17.
7 Cf pavements from these sites : Sousse (PI. 4 of the present article), Foucher {supra n. 1 p. 229, 1960), p. 103-104,
no. 57.227 and pi. LII ; Kélibia, Duval, Kélibia, p. 190-191, no. 21 and pi. XVI ; p. 199-200, no. 44 and pi. XXI V ; Carthage,
Hinks {supra n. 3, p. 123, 125, no. 49 and fig. 140 ; Thuburbo Majus, Yacoub {supra n. 1 p. 232), p. 85, Inv. 1394 and fig. 97.
8 Cf a mosaic from Kélibia, Duval, Kélibia, p. 199-200, no. 44 and pi. XXIV ; dating, p. 243.
9 Ibid.
10 Cf the grapevines in pavements of sixth century date from Iunca, Picard (G.), Séance de la Commission de Γ Afrique
du Nord, 19 juin 1950. B.C.T.H., 1950, p. 125 and pi. VII ; and El Mouassat, Duval (N.), Le dossier de V Église d'El Mouassat
(au sud-ouest de Sfax, Tunisie). Antiquités africaines, t. 8, 1974, p. 159-160 and fig. 4. In contrast to Duval, who finds the
El Mouassat and Theodoulos mosaics' vines to be similar in style, I see them as quite different.