Middle Byzantine Pottery in Athens
Middle Byzantine Pottery in Athens
Middle Byzantine Pottery in Athens
The desolation of mediaeval Athens and the deplorable state of industry so lamented
by the cosmopolitan Michael Akomiinatos' are well reflected in its pottery. Contemporary
wares from Corinth anidThebes, of aii eleoance rarelv afforded by mediaeval Atlhenians,
attest the fact that even in the humbler arts these cities had triumphed over their
former rival. But although the Byzantine pottery of the Ao,ora is, in general, frag-
mentary and poor, the circumstances of finding have sometimes prov-ided valuable
chronological evidence, while an occasional fine piece rejoices the excavator and relieves
the dark picture of Bvzantine Athens.2
The chronological evidence is provided by the finding of closed deposits or of refuse
dumps of pottery, the contents of cisterns, pits or pithoi, filled up either all at one
timne, or (as evidence of stratification shows) by dcegrees over a period of years. Each
such dump may be considered as a unit, or, rather, as a group, the elements of which
may be expected to show some logical relation to each other. Five such groups, rangino
in date fromnthe tenth or early eleventh century to the thirteenth, have been selected
for discussion here. Except in one case (Group B) the dating is based on coins found
with the pottery. These groups represent almost all the types of Byzantine pottery
found in the Agora, and the conclusions are consistent with those offered, with less
decisive evidence, by other Agora material. The present discussion does not pretend to
offer a complete chronology of Byzantine pottery but may perhaps establish a few
convenient landmarks for future studv.
The most comprehensive system previously devised for the classification of Byzanitine
potterv is that of Rice.3 But since the Agora material is limited in range it seems
preferable for our purposes to adopt a simpler method based on the pottery of the
res;tricted area. Only a summlary description will be given of wares corresponding to
those of Rice's classification. The reader is referred to his publication where they are
discussed in full.4 Our classificationifollows.
1 4 IX R0nVMtZ.X sxva y?sXpa zo~t )j, MiXctj) !lxou,ttrov iov XceXdTovrV aw6uEdC, ed. X:tup. Ii. Axu.ap,
Atlhens, 1879-80, II, p. 69.
2
The great quiantity of material found since the first puiblication of Byzantine pottery fromnthe Agora
(F. 0. Waage, IHesperia, II, 1933, pp. 308-328) has niecessitated a modification of the dates and classifications
there suggested.
D. Talbot Rice, Byza7tine Glazed Pottery, Oxford, 1930. This book forms the basis of all recent
3
Byzantine ceramnicstudies and contains a ftull bibliography of earlier works on the suibject.
4 Rice, op. cit., pp. 5 ff.
L Ptlain-glazedWares.
Three varieties are comionly found in the Agora:
a) White Ware (Rice A 3).
Over a white to light pink or orey clay of a satndy texture the glaze is applied
directly. Although the fabric varies, the tvpe is called for convenience white ware.
b) Brown Glaze. B 1, 2.
Here also the glaze is laid over the clay without a slip, but the body is red
andI the glaze, which is colorless or light yellow, appears browni. The term
brown glaze has the sanction of several years' usage.'
c) Plain Glaze on Slip. A 1, 13, 14, 63, 64, 82-84, 86, 87; C 1.
The glaze is applied as il the soraffito wares, the only difTerence beinig the
absence of decoration. The clay varies from buff to all shades of red and is
covered with a white slip before the addition of the glaze.
GROUPA
The pottery in this group (Figs. 1-18) was found in an anicient cisterni to the north
of the iephaisteion re-used as a refuse pit in connection with the complex of Byzantine
buildings in the area.4 The stratification noted in excavation madle it possible to
distinguish four periods of use, and coins found in the three lower layers helped to
establish a tentative chronologry. The catalogued pieces represent the proportion of
different wares in each period.
1 Waag6, op. cit., p. 323. Thedecidedly liniear quality of the slip-painted wares from the Agora leads
me to reject Rice's classification of this technique with incised ware. Cf. also H. MIegaw, "Byzantine
Architecture in MauIi,"B.S. A., XXXIII, 1932-33, p. 148.
2 op. cit., pp. 32 if.
Rice,
I have avoided the terimi"fine sgraffito" sinice this miglht imply a uiniformexcellence of executioni.
3
To call tlhe techlnique "'early sgraffito" is to ignore its continiued existence over several centuiries.
For the location of this cisterni (Sectioi AA, 87AIAE) cf. D. B, Thompsoni, Hesperia, VI, 1937, p. 399,
fig. 2, and p. 401.
432 M. ALISON FRANTZ
Section KK,Pithos K.
I am indebted to Dr. Ernst Ktilhiel of the Islamic Department of the Vorderasiatischeii Museum in
2
Berlin for the identification of this sherd as an Islamic product of the tenth or eleventh cenltury, probably
from Alexandria.
3 The same coneluision is reached elsewhere in the Agora, where a gradual reversal of the proportion
of painted to sgiaffito wares is noted ill areas where the stratification has been preserved.
MIDDLE BYZANTINE POTTERY IN ATHENS 433
are all of red clay with a white slip and thin colorless glaze on the outside. The
shape of A 68 is particularly notewortlhy. The menlding holes in A 69 suo,gest that
it may have been a decorative piece preserved from a slightly earlier period. Thle
palmette medallion of the soraffito bowl A 31 (Fig. 7), from period II, is repeated
in sirnplified form in the incised medallion A 78 (Fig. 14).
GROIJPB
No coins were found in. this cistern.' The contents (Figs. 3, 19-21) however are con-
sistent and illustrate the type of pottery found regfularly in deposits underneath those
containing black and green painted ware. The polyclhrome cup B 3 (Fig. 19) belongs to
Rice's class A 1 which he dates in the tenth or eleventh century. The small cup B 4
(Fig. 19) is simpler but of a somewhat similar fabric. The brown-glazed ware represented
by the double cooking, pot B 1 and the spouted jugf B 2 (Fig. 19) has been found in
abundance in Athens and Corinth, almost invariably in early contexts. The imost
common shape among the unglazed pots is the rounld-bodied. two-handled cooking pot
of which B 6 is an example (Fig. 20).
An interesting exception to the secular character of the subjects of the designs on
Byzantine pottery is provided by an Adoration of the Magi stamped on the lid of a
coarse cooking pot (B 5; Fig. 21). Undoubtedly too much stress must not be laid, on
this very simple representation, but it may be noted that iconographically it is related
to the Ravenna type of Adoration rather than to the Byzantine, in which the Angel
guiides the Magfi to the Christ Childl.2 The composition finds its closest parallel in a
fouLrthcentury relief on the sarcophagfus of the Exarch Isaac in Ravenna,3 although the
IThlle cisterni (Section 00 at 27/1(A) lies to the east of the Valeriant Wall, niear the Church of the
ITypapanti.
2 Cf. Baldwin Smith, Early ChristiantIconzoygaphyawcla School of Irory Carcersint Provence, Princetoni,
1917, pp. 36 ff.
3 . Diitschke, Ravennatisehe Stucdien,Leipzig, 1909, p. 10, fig. 3 ancdpp. 228 ff.
434 M. ALISON FRANTZ
differences of date and iiedium naturally forbid close comparison. Such stamped
decoration of unglazed wares is most unusual. The stamp had no doubt some other
purpose than the decoration of cooking pots, but either for use as a bread stamp or
for the decoration of impressed wares the lowness of the relief seems unsuitable.
The only type of pottery from this cistern not represented among the catalogued
objects is white ware. Several such sherds were found, but all in a very fragmentary
state. Among them was a small piece of the rim and handle of a licht green-glazed
double cooking-pot.1 Not to pass over this important ware withlout illustration, an
example of the same fabric from a mixed context is included (Fig. 22). This small table
brazier, or chafing dish,2 is an unusually sumptuous exaimipleof a shape frequently
found in brown glaze. The same principle of construction applies to all: a bowl on a
stand, with a rectangular openingfon one side of the stand and air-holes on the other.
The theory that coals were put in the stand to keep the food in the bowl warm is
substantiated by a slight blackening,of the clay in almost all examples. The white-ware
piece differs from most in that a second bowl lholdsthe coals, rather than the stand itself.
Fragrmentsof these chafing dishes or of their characteristic high domed lids are found
in most brown glaze deposits, and bear witness to a refinement not usually associated
with ordinary Byzantine life. Sometimes they are of the simple type of B 1; often they are
decorated with grotesque plastic figures from rim to base. The general character of the
design of the fragmentary lid in Figure 23 may be compared with the plastic decoration
of the chafing dish in Figure 24. The adoption of a more subdued method of decoration
is not perhaps to be regretted; there is in the plastic ware a singular premonition of
the art of eight or nine centuries later.
GROUPC
The most satisfactory group (Figs. 25, 26) from the chronological point of view was
found in a pit3 in the east end of the South Stoa, over the mouth of which was an
apparently undisturbed Byzantine fill. The pit was small, measuring 95 cm. in diameter
and only 80 cm. in depth. In addition to the catalogued pieces it contained some coarse
ware and a very few glazed sherds, either black and green painted or with simple
sgraffito designs. Conclusive evidence for the datingfof this group is provided by fifteen
coins all belonging to the period between 1057 and 1118 A.D.4 We are therefore fairly
I This fragment was foutnidin the disturbed fill at the top of the cistern, but it is of the samc fabric
as tlle other sherds, and probably belongs with the lower fill.
2
Inv. P 3075; D., 0.21; P. H., 0.115. The lower part of tlhe stand is not preserved, but the breaks
slhow that there was at least one more row of openwork squiares. Greyish white clay; yellow-green glaze,
fired darker in places. 'T'heglaze covers the inside of the stanid and the underside of the lower bowl.
Traces of buirning inside the lower bowl.
3 Section H, Pit at 23/KIT.
4 The distribution of the coins is as follows: oine Anonymous Byzantine, Class VI (1057-1059); three
Nikephoros III (1078-1081); one Anonymous Byzanitine, Class IX (1078-1081); two Anonymous Byzantine,
Class X (1081-1118); five Anonymous Byzantine, Class XI (1081-1118); two Alexios I (1.081-1118); and
MIDDLE BYZANTINE POTTERY IN ATHENS 435
safe in concluding, that the breakage of the pots took place not later than the second
decade of the twelfth century and that they were made at some time about the year 1100.
The lion on C 4 (Fig. 26), althoughl the glaze is dull and discolored, is a fine piece of
drawing, ancl it is gratifying to be able to date it with some precision.
GROUP D
The contents of the group (Figs. 27, 28) were found in a well used in connection
with the Byzantine house built over the north part of the Odeion.' The presence of a
number of unbroken water jars is evidence of its use as a source of water and not as
a refuse pit, and in this it differs from the preceding deposits. The single coin of
Johin 11 (1118-1143) suggests that the well was in use around the middle of the twelfth
century. The decorated wares were scanty and included black and green painted sherds,
and sgraffito of the same general character as the two catalogued pieces. The plate
D 2 (Fig. 27) is a remarkably fine and complete example of the red fabric with external
slip and glaze.
GROUPE
The pithos immediately to the west of the Stoa of Attalos 2 in whichl the pieces
from this group (Figs. 29-31) were found appears to represent two periods of use as a
rubbish pit. The lower fill, in which there were no coins, contained the lamp E 1 (Fig. 29)
as well as a few sherds of rather careful sgraffito and some black and green painted
ware. A coin of Alexios I and four of Manuel I point to the dating of the upper fill
in the second half of the twelfth century. Although most of the pottery from this
pithos was fragmentary and comparatively ordinary, one piece is full of interest. The
warrior (E 2; Fig. 30) is of a type found on a few fragments in Corinth, and a sherd
with a small piece of skirt, leg and scabbard from Constantinople, now in Berlin,3 probably
comes from a similar figure. There is a certain resemblance to the warrior on A 91
(Fig. 17), although the one is full face and the other in profile. Both wear conical caps
with knobs at the corners and high collars, and carry somewhat similar shields. The
artist of the later fragment (A 91) was less accomplished and made no attempt to
represent the long hair which hangs down over the shoulder of the other figure and,
from the little that is left, we may judge, that his clothingf was drawn much less carefully.
There seems to be no reason to look for the identification of either of these figures
with any saint or historical personage, for the almost exclusively ecclesiastical character
one identified as tenth to elevenith cenittiry. The only other coin in the pit was Athenian Irnperial. For
the classification and dating of these coins, see J. P. Shear, Hesperia, V, 1936, pp. 148-9.
1 Section Z, Well at 54/MT.
2 Sectioll I Pithos at 15-17/FE-HllT.
W F. Volbach, Bildwerke des Kaiser-F'riedrich-Museums:Mittelulterliche Bildwerke aus Italien und
XV.
Byzanz, pl. 19, no. 6448.
436 M. ALISON FRANTZ
of Byzantine art finds a marked exception in the pottery. Genre scenes, on the other
hand, are common, and warriors may find a place beside ball-players, charioteers and
scenes of domestic felicity.-
The recent finds in the Agora have made it possible to make a few generalizations
on the characteristics of the pottery of several different periods. Most of the tenth or
early eleventh century pottery of the Agora, in so far as it has been identified, is
represented in Group B. In contrast to Corinth, where it has been found in abundance,
the Agora has produced only a very few examples of polychrome ware (B 3, 4; Fig. 19),
but in every deposit wlich, according to excavation evidence, precedes the period of
black and green painted or sgraffito wares there is either brown glaze, or white ware,
or both.
At exactly what point these fabrics were supplanted by the painted and sgraffito
wares we are not yet in a position to say. Indeed, brown glaze seemnsto have continued
in use long, after the introduction of the new wares, but in diminished quantity, and
primarilv as a kitchen ware. The evidence from GroupA indicates that the more elaborate
wares had been established by the third quarter of the eleventh century. From the great
predominance of black and green painted ware over sgraffito in the two early periods
of Group A, as well as in other deposits where a similar stratification has been observed,
we shall probably be safe in assuming that the painted wares were introduced at some
time early in the eleventh century and reached the peak of their popularity at the end
of the same century. From the time when the sgraffito technique first makes its appearance,
toward the end of the eleventh century, its greater decorative possibilities lead to the
gradual displacement of the painted wares, although these continued to exist for some
time. By the end of the twelfth century it is not uncommon to find deposits with no
black and green ware at all, as in Periods III and IV of Group A, and in Group E.
Although there is no evidence for the existence of sgraffito in Athens before the
middle or the third quarter of the eleventh century, we note with. interest that the incised
technique followed not long after, since a plate dating from about 1100 A.D. has incised
details (C 5; Fig. 25). This new method is established by the middle of the twelfth century,
both for principal and accessory design. Both styles draw from the same repertory of
decorative motifs, but special adaptability leads to the preference of some patterns over
others. Rinceaux and running spirals, for example A 31-33, 36-45 (Figs. 7-9), are
common in the sgraffito technique, while ornamental Cufic and other patterns of a
rectilinear character (A 53, 54; Figs. 7, 8) are more frequently found in incised wares.
Birds, and animals, especially rabbits, are common to both styles in all periods.
Although the static quality of Byzantine art led to the preservation in unchanged
form of many designs, the period from the beginning of the twelfth to the middle of
the thirteenth century was not without some changes in the method of drawing. The
tendency was toward an increasing broadness of style, with consequent elimination of
inner drawing. An extreme example of th-is may be seen in the comparison of two pieces
taken from the beginning and end of our period. The lion from a plate dating abotLt
1100 (C 4; Fig. 26) is drawn with such minute detail that it is at first difficult to
distingcuish his manie fromntlle rinceau background. On the other hand the rabbit from
the latest period. of Group A (A 89; Fig. 15), probably from the middle of the thirteenth
century, represents the simplification achieved by constant repetition of the motif. That
the difference is not priimarily one of skill is eloquently attested by the goat from a bowl
of the first half of the twelfth century (A 52; Fig. 8), where the artist faithfully, but
unsuccessfully, drew as mucll detail as lie could. The virtue of the late style was that
an inlcompetent draughtsman coul(d achieve a presentable effect by following a simple
formula.'
Another interesting tendency of the sgraffito technique is the development of incised
sgraffito. It is wortlh noting that the only incised sg,raffito in Period II of Group A
occurs in neat bands and consists entirely of short straight lines (A 53, 54; Figs. 7, 8).
With Period III, however the character is completely changed by the use of the technique
of curvilinear instead of rectilinear patterns. The restraining bands disappear and the
dlesign spreads in unconfined exuberanice over the whole field (A 66; Fig. 13). The new
mode also becomes popular in conjunction witlh sgraffito designs (A 91; Fig. 17). The
absence of this curvilinear form in the first two periods of Group A and its appearance
in the last two, and in other deposits of the same period or later, leads to the belief
that it was introduced late in the reign of Manuel I.
Byzantine pottery does not exhibit subtle development or refinement of shape. Some
slhapes, however, associate themselves with certain styles of decoration and with certain
periods. The widely flaring bowls with sharply defined rims, usually flat around the top,
are found almost exclusively in the black and green ware, with occasionally a little poor
accessory sgraffito (A 18, 26, 27; Figs. 5, 32). On the other hand, plates with an almost
vertical rim and low ring foot are almost always decorated with rather fine sgfraffito
(C 5, D 2, E 2; Figs. 25, 27, 30, 33). In the former case, the rim is usually quite or
nearly straight, in the latter it has a slight curve. Simple bowls with a fairly low ring
foot and flaring sides of the type of A 31 (Fig. 32) are found in all periods and all wares,
but they are most commonlv found in the earlier periods decorated with a little simple
soraffito (A 9, 39, 40; Figs. 2, 9). Later, perhaps toward the end of the reign of Manuel I
(Period III of Group A), the sides beoin to curve in and during the thirteenth century
one of the most comnmon shapes is a bowl with sligfhtly incurving sides and a widely
flaring foot frequently decorated with rabbits and other animals of the type of A 88-89
(Fig. 15). The beginning of this tendency may be seen in A 79 (Fig. 32).
The Agora material sheds no light on the difficult problem of provenience. It is difficult
to believe that no pottery was made in Athens, but criteria for distinguishing the imported
wares from the local imitations lhave in only a few cases been established. We have no
I This trend corresponids to a similar one noted in manuscript illumination and prefaces the final
break-up of the Byzantine style; ef. Art Btull., XVI, 1934, p. 72.
438 M. ALISON FRANTZ
evidence for the existence of potters' workshops in the Agora before Turkish times, and
the inferiority of the pottery of Athens confirms the belief that we must look elsewhere
for the main centres of production. The general similarity between the pottery of Athens
and that of Corintlhand Sparta indicates a common source for much of it, and excavations
in all these cities have produced wares similar to those found in Constantinople and
some of the cities of Asia Minor.' The lack of resemblance between the pottery of Athens
and that of northern Greece is striking.2 It has been noted, for instance, that the
fabric of the pieces with the best and most careful decoration differs from that of the
poorer examples. The clay is red and fine, and, when attended by sgraffito decoration,
it is usually covered on the outside with a white slip and thin colorless glaze. Thlis
clay in no way resembles any fabric known to be of Attic manufacture in any period,
nor, so far as I am aware, any fabric of Greek manufacture. It is tempting to suppose
that these wares were exported in relatively small quantities from some centre outside
of Greece to various cities, including Athens, Corinth and Thebes. In the pottery with
similar designs but of the pinkish buff to buff clay familiar to all excavators in Greece
we may perhaps recognize the copies achieved with greater or less success by the local
artisans. Further systematic excavation on Byzantine sites will no doubt provide evidence
as to centres of production.
CATALOGUE
Unless otherwise specified, all vases are unslipped and unglazed on the outside, except
for an irregular band around the rim, and have a ring foot and plain rim. The following
abbreviations are used: H. Height; W. = Width; Diam. = Diameter; P. H. = Preserved
Height; P. W. = Preserved Width; Max. Dim. = MaximumDimension. A number preceded
by P is the Agora Inventory Number.
GROUPA
1 Cf. Volbach, op. cit., pls. 13-31, passim. Clear evidence of the manuifacture of certain wares in
Corinth has been found in the recent excavations of thc American School of Classical Studies there. This
material will be discussed by C. H. Morgan iin his publication, now in preparation, of the Byzantine pottery
from Corinth: CGorinil,Vol. XV. I am indebted to Mr. Morgan not only for the opportunity of examining
this material in detail, but also for much helpful discussion.
2 Examples of the Saloniki wares may be examined in the Byzantine Museuimin Athens. Cf. also
D. M. Robinson, Olynthus,V, pls. 204-208; the Byzantine pottery is discussed by A. Xyngopoulos.
MIDDLE BYZANTINE POTTERY IN ATHENS 439
A43
_q_~~~A 44 4A42
Fig. 11. Sgraffito and Plain-glazed Wares from Group A, Period III
Fig. 14. Sgraffito and Plain-glazed Wares from Group A, Period III
Fig. 17. Sgraffito Fragment from Group A, Period IV. (Slightly reduced)
GROUP B
FI 2i
GROUP C
GROUP D
the rim a wider band with a running pattern of heart-shaped leaves. Greenish yellow glaze;
white slip anid colorless glaze on the outside. Red clay.
GROUP E
/' '~~-'
32:s E~h *
E2 -0 :
SqLuat pot tapering slightly toward bottom; low vertical rim. Wheel-run grooves arounid
middle of body. Dark red micaceotis clay, unglazed; blackened by fire.
M. ALISON FRANTZ