DEUTSCHES ARCHAOLOGISCHES INSTITUT ABTEILUNG ISTANBUL

BYZAS 7 Veroffentlichungen des Deutschen Archaologischen Instituts Istanbul

YAYINLARI C:ANAK

LATE ANTIQUE AND MEDIEVAL POTTERY AND TILES IN MEDITERRANEAN ARCHAEOLOGICAL CONTEXTS . .. . AKDENIZ (:EVRESINDEKI ARKEOLO]IK KAZILARDA ELE GE(:EN GE(: ANTiK VE ORTA(:AG SERAMiGi VE MiMARi SERAMiGi

Proceedings of the First International Symposium on Late Antique, Byzantine, Seljuk, and Ottoman Pottery and Tiles in Archaeological Context ((.:anakkale, 1-3 June 2005)

Birinci Uluslararasl Seramik Sempozyumu Arkeolojik KazIiarda Ele Gee;:en Gee;: Antik, Bizans, Sele;:uklu ve Osmanh Seramigi ve Mimari Seramigi ((.:anakkale, 1-3 Haziran 2005)

Editors Beate Bohlendorf-Arslan Ali Osman Uysal Johanna Witte-Orr

printed with funds from CERAMICA - STIFTUNG BASEL

destegiyle basIlml~tIr B. l3oltlend01fArslan, A. O. Vysal, j. Wille-OIT (ed.) , (;anall, BYZAS i (2007) 5 15-522

The early Turkish stratum on the Belgrade Fortress

Vesna Bikic

Abstract Subject of the communicati on is the character o f pottery which was discovered in units dated in the period after the Ottoman conquest of Belgrade in 1521 or slightly later, in the mid 16'h century. Viewed as a whole, the pottery seems homogeneous since vessels are similar from a technical point of view. This pottery is clearly Medi­ eval in character and represents the sublimation of potters' experience coll ected in the Balkan region in the preceding century. Conte nt of this settlement level makes, first of all , po ttery fo r everyday use. It is heterogeneous, because, according to origin of the shapes of the vessels and deco­ rati on styles, we could recognize characteri stics of the Serbian pottery inheritance, then Middle European and Otto man pottery. It can be assumed that this potte ry originated from o ne of the potters' centres in the Serbian region in which produc­ tion was renewed under Turkish rule, perhaps Novo Brdo, Krusevac or Smederevo. Fragments of Turkish f aience were discovered as well.

Ozet Belgrad Kal esinde Erken Turk Tabakalan O smanhl11n 1521 yllll1da ya da bundan kl sa bir siire sonra 16. yUzyllll1 ortalannda Belgrad'l fethetrnesinden sonraki doneme tarihlenen seramiklel- karakteristik ozel­ likl ere sahiptir. Genel olarak, seramik mallar ho moj en bir goruni.ime sahiptir. Bu seramikler apk<;:a Orta<;:agll1 karakteristik ozelliklerini gosterir ve bir onceki yi.i zYll­ da Balkan bolgesinde t o plal11TI1 ~ <;:o mlek<;: ilerin kazandlklan tecri.ibeleri gozler onii­ ne sererl er.

Ye rl e~ imin konteksleri seramiklerin gunluk ihtiya<;: lar i<;: in kullal11ldl gll1l gosterir. Biz ozellikle Slrp seramiginin karakteristik e tkil e~ m e l e rini goruyo ruz. Bu Orta Av­ rupa ve O smanh seramiginden d aha e tkilidir. Orijinal formlar ve dekorasyon stil­ lerine gore farkhhklar bulunmaktadlr. Bu farkh seramik tiirleri i<;: inde bi <;: im ve de­ korasyon ozellikleri yonunden karakteristik olan Slrp se ramigi, Orta Avrupa veya O smanh se ramiklerinden d aha etkilidir. Bu seramik Sirbistan bolgesindeki , Ti.irk gucu altll1da ye ni iiretimler yapan mesela Novo Brdo, Krusevac ya da Smederevo gibi seramik uretim merkezlerinin bir tanesinde ortaya <;: Ikml ~ oldugu onerilebilir. 516 Vesna Bikic r------__ «( ....- .:- ----.:: .. : r--1 r------, ~------~"L'Lr ~,.~ i @

D huts _ pits

_ kiosk

Fig. I Base plan of the Metropoli tan palace with units of earlie r Turkish stratum.

The Belgrade fortress was changed in many ways by the Turkish conquest at the end of August 1521. These changes not only influenced in many ways the character of the settle­ ment within the fortress walls, but also determined the direction in which the whole town would develop further. For the consideration of our topic, there are two important histori­ cally documented facts that clearly illustrate the essence of these changes. These concern the events that marked the first year of Turkish rule in the town. The first fact relates to the emigration of the entire Christian population from the fortified part of the town, which was one of the first steps taken after the occupation of Belgradel . There fo llowed the demolition of the remnants of old structures and the conversion of any remaining into public buildings. These acts marked a virtually complete break with the pre-existing cultural heritage, to which the contents of the Early Turkish cultural layers clearly offer proof. They reveal exclusively Ottoman material. This is the premise that simplifies re­ search into the character of the material culture of this period, as does the separation of pottery that reached the fortress area directly after the establi shment of Turkish rule.

From the existing fortifications, which had an exclusively military function, only the Lower Town continued to be a fortified part of the civi lian settlement, with an exclusively Mus­ lim population. The structure of this settlement also changed its appearance, apart from some public buildings, such as the Mosque of Sultan Suleiman and the caravanserai of Pasha Piri Mehmed. Most buildings were constructed of wood, earth and thatch. Among the buildings in the Lower Town there is a partly demolished Metropolitan palace, dat­ ing from the 15lh century, of which individual parts were used to construct new buildings (fig. 1)2. Here, the first residential buildings were lightweight, of ancient construction. In the western part of the former palace the remains of two such dwellings were discovered, which we labell ed huts. In the eastern part, within th e palace yard, there was a small er

I Samardzic 1961, 164. 193.

2 Popovic - Bikic 2004, 109-11 7: in addilion LO consideration of the stratificati on , findings from individual sites are mentioned. Early Turkish stratulll Oil Ihe Bc lg rClclc Fortress 5 77 square building, probably a ki osk. Nearby there are pits for th e disposal of domesti c waste - damaged ceramic vessels and other obj ects. These buildings were all built be tween the d emoli shed wa ll s o f the palace, in a space \,'here rubble and d ebris had previously been levell ed. One of th e huts had been joined to an underground passage and the cell ar of the fo rmer palace. It was th atched, covered with mud, with a floor of wooden boards. The second, better-defi ned h Lit was made of wooden beams, " 'i th a fl oor of packed mortar rub­ bl e. Similar to this second hut is the kiosk, whose corners are defin ed with stone bases fo r wooden posts. These buildings were built in the first decades of Turkish rule or sli ghtly later, in the mid-16th century. They we re destroyed by a fi re at the end of that century, or at the beginning of the 17th century at the latest, to which clear soot marks, carbonised wood and beams testi fy.

However, the situatio n during investigation was not as easy or clear as it seemed from his­ torical data and theoretical assumptions. During exca\'atio n it proved especiall y difficult to separate the layer aboye the hut floors, sin ce in that space several de molished layers replaced each other in a very short period. T he discOl-e red layer shO\\'s two phases of the destruction of the palace in close succession. The first phase documents the last period of palace life and its destruction by fire; it consists of soot and ash , pieces of carbonised wood, and demoli shed parts of the wa ll s.

CJ7

Fig. 2 POlle ry fro m h illS level - T mkish fa ie n ce. 518 Vesna Bik ic

••

3

Fig. 3 POltery from h UlS level - th e m ost com mon jugs.

The second phase is defin ed by the partial demolition of th e remaining wa ll s, th e removal of sto ne suitable fo r building, and levelling of the space. T his is evide nced by the thick lay­ ers of mo rtar rubble and fin e sto nes, as we ll as brick parts. Above these layers of destruc­ ti o n, huts we re quickl y built whose fl oors are very rough . T he short d uratio n of the huts ended in the fire, and sho rtly after that the same area was fl attened o nce again.

There are numerous and vari o us findings that are reli ably linked to this peri od of life in the buildings3. In this presentatio n we will focus on po ttery which was the most in terest­ ing to interpre t. In this coll ecti on , the most luxurious goods comprise T urkish faience, of mostl y blue-white style with fl o ral decoration (fi g. 2) . Among the faience there is a vessel of j ade colour with painted decoratio n, red within and blue o utside. Other than th e fact that examples of this basic colour are rare and the decoration is not of classical f aience style, we already find it in maj o li ca ware fro m Faence, from the fi rst half of the 16th century (fi g. 2, 3). Judging by the exceptio nal quali ty of the coll ecti o n of luxurious vessels, there also belongs to it a jug of specifi c dark red colo ur, or rather a nuance of Armenian clay, which was often used as a pigment in pain ted vessels (fi g. 2, 5).

The remaining findings are from a coll ectio n of vessels for everyday use. T hese were goods which , afte r the arrival of the Turks, suppressed all previo usly known kinds by quantity. They are vessels of solid quali ty and unifo rm characteristi cs, seriall y produced , according to a standard of volumes. In the coll ecti o ns that are the foc us of our atte ntio n , most of the vessels are fo r wate r - ewers and jugs. Among these, of special importance are two gro ups of vessels, o r rathe r, two types (fi g. 3). T he first ty pe comprises six spo uted jugs, which all have a similar profil e and capacity (fi g. 3,4-6) . This is basicall y an orien tal shape which was first seen in earl y Otto man po ttery, and it is recognizabl e in the Balkans fro m the end of the 14th century. During the period of Ottoman rule a great number of several

3 Pottery considered briefl y in the texl, bUl il is described in delai l in the ca la logue or rindings: PopoviC - Bi kic 2004, 140-145. Ea rl y T urkish sU-a lU m o n th e Belgrade Fortress 519

• ------. - .- ~ OY-- 6 Fig. 4 Potlel) ' from huts level - less freque nt types.

similar variants also appear in the same territo ry4. In Belgrade, all these jugs have identical technological pro pe rties: purified cl ay, which after firing acquires approximately the same shade of red colo ur; and a green (exclusively dark brown) glaze with a leaded shine. The second type comprises fOUl-jUgS, the smallest of which could be possibly labelled as drink­ ing vessel-goblet rather than jug (fi g. 3, 1-3) . Their technological properties are similar to the jugs mentio ned above, except that the glaze is transparent, colo urless or ochre-ye llow in colo ur. In contrast to the spo uted jug such vessels we re no t unknown befo re the Turk­ ish conquest. In the producti o n of the later Mi ddle Ages in the Balkans they were viewed as products that represented a fu sio n of Byzantine and Oriental traditio ns. Such jugs and goblets we re very popular in the territory of Bulgari a, where they we re produced in all im­ po rtant regional ceramic centres during the 1 4 th century5 . This type of vessel is also o ne of the basic serial products within O ttoman producti o n of the 16th to 17th centuries.6

The remaining types of jugs and goblets, o nly re presented by o ne example, are important, above all, for a wide r view of the to tal character of producti o n of this period , altho ugh thei r appearance in Belgrade is less frequen t (fi g. 4, 4-7 )7. There is also a relative ly small number of bowls and dishes (fi g. 4, 1-3). However, the excavated types are considered to fall basically within 16th to 17th cen tury Otto man producti o n , and appear in great num­ bers througho ut this period. Generally, ceramic vessels have unified pro perties, especially in the ir structure, basic colo ur and glaze. Also, these vessels have a large specifi c weight, while their glaze in individual cases has fl aked, which indicates that during first firing the necessary quantity of water had no t been freed from the clay.

The appearance of a few specifi c po ttery ty pes makes this coll ecti o n significant. The first of these is a large pot, used to sto re small quantities of supplies (fig. 5,3) . It is made of

4 Bikic 2003, Types 111 /2. 111 /20. 111 / 24 p. 132-1 34; Pl etnj ov 2004, 15 1-152 tab!. 11 3- 11 4 wi th literature . 5 Georgieva 1974, 12/ 2-4.53.56; Bikic 1994, 85 fi g. 32, 9-10. 6 Bikic 2003, 126-1 27 Iig. 20; Pletnjov 2004, 63-64 tab!. 10-17 wi th literature. 7 Bikic 2003, 10 1-102. 520 Vesna Bikic

/El-- ~ 1 " 1 ....' i' f-· ~ ,

-- - ._. 6

I \ \ 12

Fi g. 5 Pottery from huts level - specifi c pottery types. clay mixed with sand, brown in colour, while along its body thin strips have been applied at approximately equal intervals8. Then , there are large dishes and covers used for baking bread, pies or meat (fig. 5, 4-5). AJtho ugh characteristic for the Medieval period, these ves­ sels were practicall y no n-existent in Belgrade during the peri od of Hungari an rule, th at is, fo r the whole century which preceded the Turkish conquest9. T hey not only appear in Ot­ toman pottery in large numbers, but also have certain features, such as vessel profil e and type of decoration, which connect them wi th the po ttery heritage of Serbian territori es.

In that sense, more eloquent examples are perhaps po ts with decorati on painted in whi te (fig. 5, 6-10) . The appearance of those vessels as a whole cl early indicates a model ado pted for pottery of the later Medieval period . On Serbian te rri tory, there was painted pottery in a comple tely d efin ed style in the 14Lh and early 15Lh centuries, similar to that found

8 Another, practicall y idenLi cal, example was discovered on the area of the Smederevo fo rtress. Il has nail inde nLa­ tions alo ng the applied band 9 Bikic 2003, 98-101. larl ), T lIrkish stralUm on the Ik lgrade Fo rtress 521 in ne ighbouring areas, which also belo nged to the earlier Byzantine cultural circle lo. In late Medieval productio n , painting in whi te was widely applied on vessels fo r everyday use (pots, lids, jugs and ewers) . Products fro m the Novo Brdo (tod ay in Kosovo) wo rksho p we re especiall y we ll known . However, o n these vessels fro m the earl y Turkish layer, simpli­ fi ed o lder painted mo ti fs are no ti ceable, altho ugh spirall y intertwined lines re main the basic decoratio n, in combina ti on with wavy and ho rizontal lines.

At the level of the huts, several more fragmen ts were discovered. These have characteristics of Cen tral European - Hungari an po ttery. One of these is a drinking vessel which, judging by its profil e and grain sandy structure, is completely parallel with vessels from the 151h cen­ tury (fi g. 5, 11 ). Since this is the o nly example found to date in the context of the Turkish stratum of the Belgrade fo rtress, the possibili ty remains that it originates fro m the late 15th cen tury layer (that is, fro m the time befo re the destructi on of the palace), but was mixed with material fro m the huts during the later levelling wo rks. In contrast to the drinking ves­ sel, the fragme n t of a jug has fea tures similar to o ther Otto man pottery in this collection (fi g. 5, 12). The appearance of Orie n talized shapes, especially among j ugs/ ewers, which clearly leant o n the pro fil e of earthenware water jugs, illustrates a new te nde ncy within Hungarian pottery (or we could say separate producti on line) II , purposefully made for new custo mers. The general features of these products are their pale red colo ur, thick walls, and a greater specifi c weight of vessel, as we ll as the wide use of ochre-yellow and green glaze. Vessels made in this new style appeared in the Belgrade area in the last decades of the 1 5 th cen tu ry, while in the following century this new pottery was comple tely formed.

The presen ted exa mple of the ceramic con ten t o f th e Early Turkish stra tum is o ne of the most illustra ti ve in the Belgrade fo rtress. This ve ry good sample comprises mate ri al relevant fo r the de tail ed analysis of the character of po ttery fro m the who le area, as well as the chro no logical cl assificati o n of each kind. In that sense, we have data abo ut the earli­ est kinds of Faie nce, which provides an excell ent chrono logical be nchmark in the event of future research and analys is of luxurio us Otto man po ttery. On the o ther hand, po ttery for everyday use is very specifi c in comparison to materi al fro m o ther units, above all that fro m the was te pits from a somewh at later peri od. In these units there are no Mi ddle Eu­ ro pean-Hungarian types of vessels, and absolute ly no po tte ry which has features of that craft circle (such as po ts with handles or bowl s wi th handles o n three legs, made of cream or ye ll owish-white sandy cl ay) . Vi ewed as a wh ole , the po ttery seems homogeneous since vessels are similar fro m a technical point of view. Having in mind that these are vessels that the conquerors bro ught with the m (together with o ther necessary items fo r ho use­ hold use) , these findings offer indicators of the basic character of 16th century Otto man po ttery in the Balkans.

Above all , this po ttery is clearly Medieval in characte r and re presents the sublimati o n of potters' experi ence collected in the Balkan regio n in the preceding century. The ceramic h eritage of regio nal Balkan wo rkshops, especially Bulgari an and Serbian , was built very

10 Alllo nova 1995, fi g. 89, 4. 6.8. 10.28. 30; Biki6 200 '1 , 102- 104 with citcd lite ratu re; Pl etnjo\" 2004,4.';-49 T. 1-4 with cite d literalUre. II Biki6 2003 , 111 - 11 2 with cit.c d lite rature. 522 Vesna Bikic easily and skilfully into Ottoman pottery, since inspiration and models fl owed from practi­ call y the same sources, that are Asia Minor (mostly Persian ) and Chinese pottery. Thus the ecl ectic Byzantine heritage was built into the eclecti c Ottoman pottery of the 16lh century, but in a much more natural way than was the case, for example, with Itali an or Western Mediterranean pottery.

Among the guardians and carriers of old models and ideas, alongside emigrants and the assimilated populati on in the Balkans, there were certainly conve rts. These were educated people, artists, potters, who with the change to a new fa ith, Islam, entered into a new busi­ ness circle. Having in mind, in addition, that the Ottoman practice was to use the services of domestic craftsmanship in occupied areas, interpre tati on of the origin of the sample of heterogeneous pottery has a stronger confirmati on. For this reason it can be assumed that this pottery originated from one of the potters' cen tres in the Serbian region in which production was renewed under Turkish rule, perhaps Novo Brdo, Krusevac or Smederevo. Likewise, the presence of jugs belonging to the Ottoman seri es on the territory of south­ ern Hungary in the case of Belgrade shows a ti gh t mutual connecti on between these two markets, established long before the Ottoman T urks arrive d. As early as the first decade of Turkish rule, and especially after the conquest of Buda in 1526, the pe rmeation of Oriental-Balkan and Middle European pottery set the course for the whole region. In that sense, stylistic heterogeneity is the essential characteri sti c of the Early T urkish po ttery in Belgrade, which demonstrates the true value of this material.

Bibliography

Antonova J 995 V. An tonova, Sumen i sumenskata krepost (= Shoumen and the Shoulll en Fortress) (1 995) . Bikic 1994 V. Bikic, Srednj ovekovna keramika Beograda (= Medieval pottery from Belgrade) (1 994) . Bikic 2003 V. Bikic, Gradska keramika Beograda (16-1 7. vek) (Belgrade cerami cs in the 16th_ 17th cel1lury) (2003). Georgieva 1974 S. Georgieva, Kerami kata ot dvoreca na Carevec, Carevgrad Trnov 2, 1974, 7-1 S6.

Pl etnjov 2004 V. Pl etnjov, Bi tovata keramika \IV Va rna XV- XVl II ve k (Ho useho ld ceramics fro m Va rna 15th_I S th century) (2004). Popovic - Bikic 2004 M. Popovic - V. Bikic, Kompleks srednjovekovne mitropolij e u Beograd u - istrai iva nj a u Donj em grad u Beogradske tvrd ave (The Complex of the Medieval Mitropolitan Church in Belgrade - Excavati on of the Lowe r Town of the Belgrade Fortress) (2004). Sa mardi ic 1961 R. Samardiic, Beograd i Srbija u spisim a francuskih savremen ika XVl-XVlI ve k (La vill e de Belgrade er la Serbie du XVl e et d u XVlIe siecles dans les ecrits des contem­ pOl'ains fran.;:a is) (1961).