«MIDDLE AND LATE BRONZE AGE ARCHITECTURE AND POTTERY IN GÖZLÜ KULE, TARSUS: A NEW ANALYSIS Dr. Dorothy SLANE dslane TARSUS ...»
ÇÜ Sosyal Bilimler Enstitüsü Dergisi, Cilt 15, Sayı 3 (Arkeoloji Özel Sayısı), 2006, s.1-14
MIDDLE AND LATE BRONZE AGE
ARCHITECTURE AND POTTERY IN
GÖZLÜ KULE, TARSUS: A NEW ANALYSIS
Dr. Dorothy SLANE
TARSUS GÖZLÜKULE’NİN ORTA VE GEÇ TUNÇ ÇAĞI’NDA SERAMİK VE
MİMARİSİ: YENİ BİR ANALİZÖZET Bryn Mawr College’a 1987’de sunduğum doktora tezinde, Hetty Goldman’ın aynı kurum için 1930’lu yıllarda yapılan kazılarda ortaya çıkardığı seramik ve mimari kalıntılar yeniden incelenmiştir.
Tüm bu araştırma sonuçları burada, kısaltılmış şekliyle sunulmuştur. Bu yazıda oluşturulan çizelgede yeni analiz sonuçları da dikkate alınmıştır. Ancak 1987 yılından sonraki gelişmeler bu yazıda kullanılamamış, yeni yayınlar değerlendirilmemiştır.
Çizelgedeki benzer yerleşim yerlerinin tamamı için Bryn Mawr College’daki orijinal çalışmadan yararlanılmıştır. Bu çalışmada en önemli ve tek değişiklik I. Geç Tunç (A.VI) ve IIa. Geç Tunç (A. IXA) Dönemi arasındaki geçiştedir, bu konuda kanıtlarım sunulmuştur. Her ne kadar, bu geçiş Goldman’ın çalışmalarında ve özel notlarında açıkça belirtilmişse de, üzerinde durulmamış ve göz ardı edilmiştir. Ayrıca bu çalışmayla Hetty Goldman’ın yayınlanan kronolojisinden daha iyi bir seramik kronolojisi de oluşturulmuştur. 1987’de ufak parçalar da dahil, mimari kalıntılar da sınıflandırılmıştır. Goldman kazı sonucunda 4 genel tabaka (MBA, LBI, LBIIA ve LBIIB) kaydetmiştir. Oysa çalışmalarım sonucunda mimari, alt evreleri de dahil 10 farklı bölüme ayrılmıştır. Seramikler bir evreden diğer evreye farklılık göstermemesine karşın, bu mimari ayrımı yapmaya yeterli olmuştur. Bu hala doğru ve geçerli bir sınıflamadır ve tarihleme açısından yararlıdır. Çalışmada genel hedef de budur.
Tarsus’da mimari tabakaların aşamalarının tekrar değerlendirilmesinde, sadece Goldman’ın (1956) kazı raporlarına bağlı kalınmamış, Robert W.Ehrich, Ann H. Ehrich ve Dorothy H. Cox’un bilimsel notları da kullanılmıştır. Eldeki tüm veri ve materyallerin yanı sıra yapıların bir dizi içinde inşa edildiğini saptamak için gerekli araştırmalar da yapılmıştır.
Anahtar Kelimeler: Tarsus, Gözlükule, Orta Tunç Çağı, Geç Tunç Çağı, Seramik, Mimari.
1 ÇÜ Sosyal Bilimler Enstitüsü Dergisi, Cilt 15, Sayı 3 (Arkeoloji Özel Sayısı), 2006, s.1-14
ABSTRACTIn 1987, I completed my doctorate at Bryn Mawr College with the submission of a dissertation which analyzed anew the ceramic and architectural remains uncovered in the 1930’s excavations conducted at Tarsus, Turkey by Miss Hetty Goldman for Bryn Mawr College. The results of those researches are presented here in abbreviated form;
the chart at the end of this article summarizes, without notes, the conclusions about dates reached by this new analysis. For full citations of parallels, consult the original work from Bryn Mawr College.
And so the following information and analysis is offered. The chronology of Gurney which I followed for the Hittite kings may no longer be accepted by all scholars, but it is the more conservative choice for the non-expert. Anyone who does not agree with Gurney’s dates may instead read the concluding time chart by king. I also realize that my bibliography reflects work done only through 1987; again, I do not see that this is a general indictment of my work. More excavation has certainly been done since then, but no major changes of chronology have resulted.
Key Words: Tarsus, Gözlükule, Middle Bronze Age, Late Bronze Age, Pottery, Architectural.
Introduction The single biggest change for which I have evidence is an unemphasized gap between the Late Bronze I (my Level A.VI) and the Late Bronze IIa (my Level A.IXA) periods. While this gap was noted in Goldman’s publication and is certainly clear from the field notes, it was not emphasized and has been consistently overlooked. The stratigraphy at Gözlü Kule, Tarsus is not continuous.
Because we now possess a more refined ceramic chronology than was available in the 1950’s when the Goldman excavation was published, in 1987 I divided the architectural levels into smaller segments. Thus, while Goldman wrote of four general levels (MBA, LBI, LBIIA, and LBIIB), I have divided the architectural levels into 10 distinct levels with sub-phases. The architecture lends itself to this division though the ceramic evidence is not so distinct from one phase to another. Still, it is a valid division and could be useful for dating purposes, our ultimate goal.
In re-evaluating the phases of the architectural levels at Tarsus, I relied not only on the published plans and Goldman’s commentary in her excavation report (1956) but also on the field notes of Robert W. Ehrich, Ann H.Ehrich and Dorothy H.Cox.
With all this material in hand, I have studied the evidence of the buildings to determine construction sequences within units, major re-modellings and destruction levels. This has produced a detailed sequence of levels at Tarsus which structured the presentation of the ceramic inventory.
Levels are differentiated either by the construction of new foundations for all the exposed rooms and / or by a change in orientation of the habitations. The Middle 2 ÇÜ Sosyal Bilimler Enstitüsü Dergisi, Cilt 15, Sayı 3 (Arkeoloji Özel Sayısı), 2006, s.1-14 and Late Bronze Ages are here represented by ten levels in Section A and by three in section B. It has become probable through re-evaluation of the field notes that some of the terrace rooms which were thought to be Late Bronze Age I by Goldman must be reassigned and rooms from the terrace which were not published should be included in order to determine the constructional history of that part of the town. The following chart correlates Goldman’s levels with my architectural sequence. From this point on, all architectural levels will be referred to by the new nomenclature.
Having established through detailed analysis that the architecture of Tarsus could thus be charted, I applied these levels to the pottery extant in both the museums of Adana, Turkey; and the Tarsus Archive at Bryn Mawr College, both sherds and photographs of the pieces that could not be found. Only those pieces for which a firm find spot could be derived from the various surviving records were included in my ceramic analysis. And it was for these that parallels were sought at contemporaneous sites. This information can be summarized to add a bit to Miss Goldman’s analysis (Goldman,1956).
Summary of the Middle and Late Bronze Age Sequences The Tarsus Middle and Late Bronze Age architecture has been re-studied, producing a detailed progression of Levels from the beginning of the Middle Bronze Age at Tarsus to the resettlement after the destruction of the side by the “Sea Peoples”.
The accompanying ceramic sequence indicated clearly that there was continuity from one Level to the next at all times. With this basis, we have also seen the sequence of outside influences at work on the ceramic repertoire, and it is through these influences that our relative dating has been derived. The Middle and Late Bronze Age architectural and ceramic evidence helps us to construct a coherent sequence of contacts between Tarsus and the central plateau.
In the analysis of the pottery by level (Slane, 1987: 85-465), correlations between archaeological levels have been determined for Tarsus with other Anatolian sites. These have yielded ceramic evidence which suggests that Levels A.I-A.III correspond to the Karum levels IV or III-Ib at Kültepe (as well as to the Middle Bronze Age in Syria). It is not possible to determine the exact correlation of the Tarsus and Kültepe levels with in the architectural sequences of the two sites, but they span the same general time period.
The break noted in the architectural layout of Tarsus between Levels A.III and A.IV has been seen to correspond to a change in the ceramic repertoire. Level A.IV is so ill-represented in the cataloque presented in 1987 that it is impossible to say whether this change came about at the end of Assyrian Trading Colony Period or at the beginning of the Old Hittite period. By Level A.V, however, Tarsus must be into the Old Hittite period.
It was probably sometime during these first two Old Hittite Levels, A.IV and A.V, that the land deed with the Tabarna seal found at Tarsus, in a secondary context (a pit in Section B), was executed (Gelb in Goldman, 1956: 246, 253-254, no.64). The occurrence of Tabarna seals on land deeds has been discussed by Balkan in reference to just such a document discovered at İnandık. He has dated his tablet to the reign of Hattusilis I (Balkan, 1973:) ; Riemschneider had argued in 1958 for the dating of such seals to the pre-Telipinus era (Riemschneider, 1958: 321 vd) and, most recently, Easton has determined that such seals date between reigns of Hattusilis and Arnuwandas I, with most dating to the reign of Mursilis I (Easton, 1981: 33).
In Level A.VI, for the first time in these Levels at Tarsus, there was a strong correspondence between the ceramic repertoires of Tarsus and the Old Hittite sites.
Parallels were found for the Tarsus material at İnandık, Ferzant and Alacahöyük, as well as at Boğazköy. The length of time spanned by Level A.VI is, of course, unknown but the lack of major remodelings or a series of new floors in the rooms associated with the Pottery Storage Room and in the Pottery Storage Room itself tends to argue for a shorter existence. That we are in the full Old Hittite period is certain and it may be that it is to this Level that we can assign the completion of a treaty between the Kizzuwtnean king, Isputahsu, and the Hittite king, Telipinus (1525-1500 B.C.; dates for the Hittite kings are taken from Gurney, 1980: 218).
The treaty concluded between these two kings has not survived in full and is only known from fragments found at Boğazköy and by reference made to it by Suppiluliumas I in the 14 th century (ibid. 25). The treaty was drawn up in both 5 ÇÜ Sosyal Bilimler Enstitüsü Dergisi, Cilt 15, Sayı 3 (Arkeoloji Özel Sayısı), 2006, s.1-14 Akkadian and Hittite; fragments in both languages have been identified in the archives at Boğazköy (Götze, 1940 : 4, n.10a). We also have a second piece of evidence about Isputahsu, and the position he held in Kizzuwtna: the seal impression from Tarsus, found in a secondary context, on which Isputahsu is identified as a “Great King” (Gelb,I. in Goldman,op.cit. 246-247, no.1). As Götze pointed out, this identifies the area in which Tarsus is located as Kizzuwatna (Götze,op.cit. 74).
It is certainly tempting to assign Level A.VI at Tarsus to the era of Isputahsu solely on the basis of the good ceramic assemblage of Old Hittite date from that Level.
Support comes from the preceding Levels A.IV and A.V, in which a central Anatolian ceramic influence was becoming more and more evident, culminating in the Old Hittite corpus of Level A.VI. The central Anatolian influence seen in the Level A.VI pottery is so pervasive that it must indicate some change in the relationship between Tarsus and the central plateau. Historically, the most significant change is that of the growing political relationship between the Hittites and Kizzuwatneans, documented in the Isputahsu-Telipinus treaty.
The evidence at Tarsus for the period between Telipinus and Suppiluliumas I is very meager. I have hypothesized that Level A.VI. should be dated to the early part of the 16th century, and Level A.IX to the early part of the 14th century. This means that the ill-preserved buildings of the intermediate Levels A.VII-VIII span at least the 15th century. It was during this period that the series of treaties between the Hittite and Kizzuwatnean kings were concluded (Kümmel, 1980: 627-631).
In Level A.IX/B.IX, the so-called Hittite “Temple” was built in Section A. The construction of this building with its associated Empire-type pottery necessitated the leveling of the site and the sinking of deep foundation trenches. It is because of this that uncertainty remains concerning the correlation of finds of Levels A.VII-VIII with the pottery of Boğazköy and other Hittite sites. Little remains in these intermediate Levels.
The foundation date of the “temple” cannot be determined.
The pottery associated with this “temple” does include pieces with religious function (Libation Arms and votive Miniatures). These belong to the final destruction phase of that building.
During the reign of Mursilis II (1339-1306 B.C.) or already under Suppiluliuma II, Kizzuwatna was made a province of the Hittite Empire. In the period following his reign, the relationship between the central plateau and Kizzuwatna strengthened, culminating in the marriage of Hattusilis III to a Kizzuwatnean woman, Pudu-Hepa (Gurney,op.cit. 37). A seal impression with her name and the title “Great Queen” was found in a secondary context at Tarsus (Gelb, in Goldman,op.cit. 248-249, no.15).
Kizzuwatna was destroyed by the Sea Peoples, according to the report of the Egyptian records. The final Level at Tarsus, A.X/B.X, is the resettlement of the site after the destruction by the Sea Peoples. Following this, buildings were constructed on the debris by people making the same plates, miniature vessels and other pottery as had appeared in Level A.IX/B.IX. On the first floor level of the rooms in Section B, though some new Monochrome Ware shapes occurred (not included in the catalogue), there was no Mycenaean pottery. It was not until at least the second floor level of these rooms that Mycenaean IIIC1 pottery started to appear and not until the third floor level that it became common.