ETHNOARCHAEOLOGICAL INVESTIGATIONS IN RURAL ANATOLIA Volume 4
ETHNOARCHAEOLOGICAL INVESTIGATIONS IN RURAL ANATOLIA Volume 4 Edited by
Turan Takaoğlu
2007
Ethnoarchaeological Investigations in Rural Anatolia Volume 4 Advisory Board /Danışma Kurulu
Curtis Runnels Jak Yakar Ronald Marchese Liza Hopkins
ISBN 978-975-807-221-7 Copyright © 2007 Ege Yay›nlar›, ‹stanbul
Cover Design
Savaş Çekiç
Printed in Turkey
Production & Distribution
Zero Produksiyon Ltd. Şti. Aslan Yatağ› Sokak, No.19/2 Cihangir 34433 Istanbul - Türkiye Tel: +90 (212) 249 05 20 - 244 75 21-22-23 Fax: +90 (212) 244 32 09 www.zerobooksonline.com
[email protected] C on t ents BRADLEY J. PARKER and M. BARIŞ UZEL The Tradition of Tandır Cooking in Southeastern Anatolia: An Ethnoarchaeological Perspective ............................................................................................... 7 ZAFER DERİN Neolithic Shellfish Gathering at Yeşilova: An Ethnoarchaeological View ........................................................................................................... 45 DAVUT KAPLAN Traditional Modes of Viticulture in Coastal Northwestern Anatolia ............. 59 ABDÜLKADİR ÖZDEMİR An Experimental Study of Mat Impressions on Pot Bases from Chalcolithic Gülpınar (Smintheion) .......................................................................................... 73 TURAN TAKAOĞLU Modern Site Surveys in Ethnoarchaeology: The Case of North-Western Anatolia .......................................................................................... 87
T he Tr ad it ion of Tan dır C ook i n g i n S out hea s ter n A n atol i a: A n Et h noa r ch aeolog ica l Per spec t ive B r a dl e y J. Par k e r an d M. B ar ı ş Uze l
The tradition of tandır cooking is deeply ingrained in the fabric of rural society in many parts of the Middle East and Central Asia. Any traveler passing through these regions cannot help but notice the ubiquitous domed clay ovens that punctuate compounds and neighborhoods in many villages and towns. Particularly in traditional communities, tandır ovens can be found in nearly every alleyway, in yards and on street corners. Often alive with fire and surrounded by women and children, tandır ovens are more than just traditional cooking facilities. They are, in many cases, communal spaces utilized and shared by a number of families that serve as anchors for local female social networks. Although tandır ovens can vary somewhat in size and shape, their basic layout and function are surprisingly uniform. Modern tandır ovens are large hollow clay structures that usually measure around one meter in diameter at their base and about one meter in height. They are beehive shaped, sometimes tipped, with one large ‘loading hole’ at the top and one smaller “ash removal hole” at the base (Fig. 1). These ovens are often the centerpiece of outdoor work areas that consist of one 7
or more elevated clay work surfaces and, in many cases, a protective awning (Fig. 2). Tandır ovens are designed and almost exclusively used to bake unleavened flat bread. For baking, the sticky bread dough is adhered to the interior of the oven’s large hollow firebox (Fig. 3). The heated interior walls of the oven along with the cooking fire bake the unleavened dough until it is removed from the oven by use of a stick or metal bar. The result is a flat round loaf of what is, at least in Anatolia, usually referred to as tandır bread (Fig. 4). Tandır cooking is today considered an integral part of tradition and heritage in many parts of rural Anatolia. The tandır’s centrality to rural life is evidenced by the fact that in Turkish folklore the sharing of fresh tandır bread between a young man and woman is occasionally used as a metaphor for love (Başgöz 1952: 332). This cooking technique not only has deep roots in tradition, but it also has deep roots in time. The antiquity of tandır cooking is attested by the fact that the word tandır itself has a long and prestigious genealogy. Tandır derives from the Akkadian term tinu-ru which passed into Arabic and Hebrew as tanu-r (Greppin 1991) and later into Turkish as tandır. Bottero (1985) has noted that there are as many as three-hundred varieties of bread attested in the cuneiform sources. As is the case today, the tandır variety was baked by adhering dough to the walls of a heated clay cylinder (Bottero 1985: 39). This flat unleavened tinu-ru bread was probably prepared in communal tandır style ovens the remains of which have been recovered at numerous archaeological sites across the Middle East. In Anatolia, tandır cooking was perhaps the most common method of bread making during the Seljuk and Ottoman periods when the remains of tandır ovens become ubiquitous (Yavuz 1997). Despite the fact that data pertaining to the construction, use and decomposition of tandır or tandır-like ovens are relatively abundant in the archaeological and ethnographic records, studies combining these two datasets have not yet reached the ethnoarchaeological literature. This chapter addresses
8
this lacuna by combining ethnographic data from four locations in the Upper Tigris River region of southeastern Turkey with archaeological data from the nearby site of Kenan Tepe (Fig. 5) where excavations have revealed a number of oven or oven-like features that bear a striking resemblance to tandır bread ovens still in use in the region today. The main purpose of the research here described was, therefore, to gather data that might help guide the excavation of these remains. However, once this research began, it became clear that the emerging data had much to contribute, not just to excavation and sampling strategies, but also to the analysis and interpretation of the ancient ovens recovered at Kenan Tepe. In the following section we briefly describe the ethnographic research carried out as part of this study.1 In the second section we discuss the ovens excavated at Kenan Tepe. And finally, the third section combines our ethnographic and archaeological data to offer a number of interpretations of the archaeological record based largely on ethnographic observations of modern tandır ovens. We conclude that ethnographic research on tandır ovens in the Upper Tigris River region revealed data that are integral to creating hypotheses that go well beyond those possible using archaeological data alone. Ethnographic Observations The Manufacture of Tandır Oven Cores At first glance modern tandır ovens appear simply to be large hollow clay domes. However, closer examination shows that tandır ovens consist of two parts. At the heart of the tandır is the oven core. Tandır cores typically measure about 80 centimeters at their base and about 80 centimeters in height (Fig. 6). They are beehive shaped and have a large (ca. 40 cm) hole at the top. Tandır ovens are constructed by positioning a core and then surrounding that core with mud bricks or clay to create the larger domed structures that are so common in many villages and towns. The core is the most important part of a tandır oven. To begin with, the core forms the hollow chamber within which
9
the fire is made. Interviews with oven users confirm that higher quality cores are less susceptible to breakage and fragmentation and, since this is the hottest part of the oven, higher quality cores last longer. As mentioned above, modern tandır ovens are almost exclusively used to make unleavened flat bread. Since the bread dough is stuck to the inside of the core of the oven for baking (Fig. 3), the quality of the core is also important. Users claim that higher quality cores have a lower rate of bread loss because bread dough sticks better to higher quality cores than it does to cheaper, lower quality cores. Quality construction is clearly very important and for this reason we found no users who claimed they themselves had made the cores of the ovens they use. Instead, core manufacturing is, at least in our study area, in the hands of a few seasonal specialists. An ethnographic survey of the Tigris basin between Diyarbakır and Bismil revealed only two locations where tandır cores are manufactured. The first is the village of Çarıklı, which is located on the Diyarbakır to Mardin road approximately 5 kilometers east of Diyarbakır and the second is a single core manufacturing site located in the town of Bismil (Fig. 7). Çarıklı is the larger of the two operations. In this location members of at least six families are regularly engaged in core manufacturing and cores, in various states of manufacture are prominently displayed along the main road (Fig. 8). In Bismil we found another location where core manufacture was taking place. This is a much smaller operation consisting of a segment of only one family. Tandır cores are manufactured from clay and thus core production is restricted to the summer months when the warm and dry conditions are conducive to sun drying and because the demand for cores is said to be higher during the summer. Numerous interviews in both Çarıklı and Bismil have shown that producers are very particular about the clay they use to manufacture cores, particular enough that none of the women interviewed felt comfortable disclosing the location of their clay source.
10
Our informants in Çarıklı did say that senior male members of the core manufacturing families gather the clay from the clay source and, using a rented tractor, intermittently transport large quantities of clay to Çarıklı for processing. Interestingly, our interviews suggest that this is the only activity related to tandır core manufacture that is carried out by males. Before making tandır cores the clay must be properly prepared. To do so, children spray down the clay with water one day before it is to be used. The following day, under the supervision of the senior female and elder female family members, children prepare the clay by mixing in straw, goat hair and salt with their feet (Fig. 9). Clay preparation appears in both locations to be delegated to children. Tandır cores are given their circular shape by molding clay around a tire or a large metal cylinder. An access hole is made by inserting a pipe or can into the base (Fig. 10). The body of the oven core is then constructed in six stages. First the tire or cylinder is removed. Then the core is built up by hand in stages of approximately 10 to 15 centimeters each (Fig. 10). Each stage is allowed to dry in the sun for 5 to 6 hours before the next stage is added. This process takes a minimum of two days. The completed core is then sun-dried for two more days (Figs. 6 and 8). In both Çarıklı and Bismil core manufacturing is exclusively carried out by women. Intensive ethnographic research focused on three production groups in Çarıklı, and one in Bismil. In all four of these cases production was conducted and supervised by the eldest female of the family who was usually aided by her daughter or daughter-in-law. In some cases daughters or daughters-in-law may relieve each other but work units were always made up of two individuals. During our observations, the ranking female always participated fully in core production. Children were often present in the work areas and occasionally aided their mothers or grandmothers on an intermittent base.2
11
Interviews suggest that the skills involved in core manufacture are passed from mother to daughter through this kind of working child care.3 Seyran Öncel, for example, the seventyfive year old matriarch of the core production family in Bismil, says that she learned how to make tandır ovens and their cores from her mother when she was a child. She made her first complete oven when she was sixteen. At the time, her family made ovens and oven cores for their own use. Seyran began making and selling tandır cores in 1980 and since then this seasonal production has generated a significant amount of her family’s income. Another informant, one Hatice Zinginoğlu from Çarıklı, also learned to make tandır ovens when she was very young. Hatice’s grandmother is said to have been a skilled core producer and it is from her that Hatice learned. Hatice married when she was seventeen years old and recalls making an oven at her husband’s residence soon after her arrival. Now seventy-nine, Hatice claims to have started the first core production business in Çarıklı some thirty years prior to our interview and is credited with having taught the finer points of core production to all of the other core producers in the village. Pairs of laborers devote approximately six hours per day to core production; two hours in the morning, two hours in the afternoon and two hours in the evening. The hours between are spent on housework, cooking and child rearing. The male members of each of the families interviewed do not participate in production. In Çarıklı at least one male family member works outside the home and thus proceeds generated by the sale of tandır cores are considered supplemental income.4 Informants claim that this income is used for domestic needs and that these funds remain under the control of the female laborers. The story is slightly different in Bismil where the patriarch of the family is deceased and the only male offspring works in a resort town on the Turkish coast. The only remaining family members capable of working are the matriarch and her daughter-in-law. These women divide their time in a similar way to their counterparts in Çarıklı. Although they do not consider 12
the funds generated from tandır core sales to be in any way supplemental, they do receive a significant amount of the family income from the itinerant male family member. In Çarıklı, the only task pertaining to this otherwise female dominated household industry performed by males is selling. Taboos about relations between village women and strangers usually prevent the producers of the cores from interacting with potential buyers.5 Only at times when there are no male family members available do women act as salespeople and even then, only the matriarch does so. Tandır cores are usually sold one at a time directly to consumers who often travel long distances to buy them at Çarıklı. Occasionally however, dealers buy cores in bulk, and at a discounted rate, and transport them as far as 200 kilometers for sale. In one case we interviewed an itinerant dealer who returns to Çarıklı as often as once a week during the summer months to fill his small truck with tandır cores. He then travels as far as Mardin, Nusaybin and Cizre, passing through small villages on the way where he solicits customers.6 Tandır Oven Construction If used daily, tandır ovens will generally last two to three years at which time the core becomes brittle and begins to flake and chip. Once an oven reaches this point it is no longer economical to use since bread dough is less able to stick to an aged core; thus, the older the core, the higher the bread attrition rate (Yakar 2000: 153). In the course of this study we observed the construction of a tandır oven at the village of Çöltepe which is located on the Diyarbakır to Batman road approximately 15 kilometers east of Bismil (figure 7). The construction began by removing most of the debris from a dilapidated tandır oven. The patriarch of the family purchased a new core from the same family of core producers we interviewed in Bismil and brought it to Çöltepe by means of a horse cart. Since tandır cores are extremely heavy (they can weigh between 70 and 80 kilograms), it took six men (three male family members and three 13
neighbors) to lift the core and carry it to the desired location. The matriarch of the family then instructed the male family members on the positioning and angle of the core. All of our informants agreed that the positioning and angle are important since access to the hollow chamber created by the oven core is through the top (the loading hole) and through a smaller hole in the bottom (the stoking or ash removal hole). The core must first be positioned in such a way that the ash removal hole is facing the work area. The core must then be tipped approximately 25 degrees toward the ash removal hole to allow for easy access and cleaning through the loading hole (Fig. 11). Once the position was decided, the men used mud bricks and broken oven core pieces to balance the new core at the proper angle. The men then retired to the garden for tea while the women completed the rest of the construction by surrounding the core with successive layers of mud mixed with straw (Fig. 12). No mud bricks were used. Once the structure was complete the matriarch inserted a young girl into the oven through the loading hole to seal the inside of oven chamber with mud. The construction took a total of 3.5 hours Since tandır cores are sun baked the first firing of a newly made oven is very important. In fact, our informants told us that the manner in which the first firing is performed influences both the oven’s performance and its longevity. After the construction of the oven in Çöltepe, the matriarch used copious amounts of dung, straw and small sticks to keep an intense fire going for more than four hours. The fire was then kept alive at a lower intensity for another ten hours. The next day the women cleaned the oven by removing the resulting ash through the ash removal hole and the following evening they used the oven to bake bread. Tandır ovens are usually protected by an oven shelter. These shelters, which allow users to work in bad weather and protect the fire from wind, can be as simple as a makeshift lean-to or as elaborate as well-built three-sided mud brick structure with 14
a tin roof (Fig. 1 and 2. Also see Yakar 2000: 153 and fig. 77). In most cases oven shelters consist of three walls forming a three-sided square or rectangular building that is roofed in some way. Built onto most ovens and as part of the oven shelter are several tables made of mud or mud bricks and plastered with mud. These tables, which usually appear on both sides of a tandır oven, function as work spaces (Figs. 1 and 2). Tandır Oven Users Tandır ovens are used almost exclusively to make a type of unleavened flat bread that is referred to as “tandır bread” (tandır ekmeği) by the inhabitants of our study area. This type of bread is said to be the most widely consumed bread in many parts of rural Anatolia (Yakar 2000: 176). When asked why locals prefer to make tandır bread rather than purchase loaves of bread from a shop or bakery (known as “factory bread” [fabrica ekmeği]), users usually respond in one or more of three ways. First, most people agree that tandır bread tastes better. Second, although the process of making tandır bread can be quite time consuming, in the long run making tandır bread is considerably cheaper than buying bread in stores. It should also be noted that many villages do not have stores so in some instances consumers would have to travel, sometimes great distances, to purchase factory bread (this is the case for example at the village of Bozçalı [Fig. 7]). And finally, several informants responded that making and eating tandır bread is part of local custom and heritage. One informant commented that “we have always done it this way, my mother did it this way and my grandmother did it this way.” The owner of the single small store in Çöltepe informed us that he does not sell factory bread because people do not want to but it saying, “they prefer to eat bread made in a tandır.” In spite of this, several older women interviewed as part of this study commented that more and more younger women, especially those who live in towns, tend to buy factory bread. They insinuate that young women of today tend to be lazier than a generation ago. Accusations aside, these comments underline 15
the fact that this, and many forms of traditional life in rural Turkey, are rapidly disappearing (Marchese 1995; 2005; Runnels 2005; Yakar 2004). As part of our study of tandır oven users we focused our research on eleven families: three from Çarıklı, three from Bismil, four from Çöltepe and one from Bozçalı. Generally speaking, the results of our interviews and observations were very similar. To begin with, in all cases it is clear that tandır ovens are almost exclusively used to make unleavened flat bread. This is reflected in the structure of tandır ovens which are clearly not designed for other types of cooking. Only on very rare occasions did users cook foods other than tandır bread (on two occasions we observed users roasting eggplants and on one occasion we observed users roasting peppers) and in those cases this cooking was done as a supplement to making bread. All of the tasks involved in making tandır bread were performed by women and girls. The process begins with preparing the dough that consists of flour, salt and water. The dough is then made into small balls. Baking is usually conducted in the late afternoon to avoid the heat of the day and so the bread is fresh for the evening meal. Before each baking, the interior surface of the oven core must be wiped clean. An hour or so before baking the oven is heated. Fuel consists of dung, which is lit with dry brush, sticks or dried cotton plants. Before baking, the balls of dough are flattened, brushed with egg and stuck to the inside of the oven core by hand. Approximately ten to fifteen tandır breads can be cooked simultaneously (Fig. 3). In the town of Bismil (Fig. 7) most houses do not have a tandır oven. Instead, ovens are located in small shelters on the street outside of house compounds (figure 2). On an average street there might be only one oven for every 7 or 8 families. Bismil’s Yıldırum Street, for example, has only 3 ovens. These ovens are shared by as many as twenty families (Meyers 2002: 25). Families coordinate which oven or ovens to light on a given 16
day and take turns baking their bread in the designated oven while the fire is hot. All of the women interviewed claim that this system is meant to conserve fuel. Nevertheless, it cannot be denied that this system also creates social opportunities (Yakar 2000: 139). Late afternoon baking is a time of socialization on Yıldırum Street. Women and children of all ages gather near lit ovens. The children play and the women socialize. In fact, this is the only time that young females are regularly seen outside of their house compounds. Not only do families share in the cooking and use this time to socialize, they also share in the construction and maintenance of tandır ovens. Socialization around ovens also works in reverse. In the town of Çöltepe for example, two brothers who share a courtyard had to build two ovens right next to each other because their wives did not get along and refused to share an oven. Archaeological Data The archaeological data utilized in this study comes from the 4.5 hectare multi-period site of Kenan Tepe (Figs. 4 and 10). Kenan Tepe is located on the north bank of the Tigris River approximately 15 kilometers east of the modern town of Bismil close to both Bozçalı and Çöltepe (Fig. 7). The site is composed of a 32 meter tall high mound and a lower town to the north east of the main mound. Archaeological research conducted by members of the Upper Tigris Archaeological Research Project (UTARP) over the past seven years has shown that Kenan Tepe was occupied during five periods.7 The earliest remains unearthed thus far belong to the so-called Late Northern Ubaid (carbon dated to ca. 4650 BCE). Ubaid or Ubaid-related material culture has been identified in three areas of the site. Ubaid ceramics were first discovered in a sounding in Area E on the southeastern slopes of the high mound during the 2000 field season (Fig. 13). In 2001, UTARP team members discovered the remains of an Ubaid period structure in trench D5 on the eastern slopes of the high mound. In 2002, Ubaid period remains were encountered at the bottom of our step trench in A9. This research sug17
gests that Ubaid period occupation is restricted to the eastern and southern slopes of the high mound (Parker et al. 2006). Remains dating to the Late Chalcolithic period have been discovered in abundance in the easternmost area of Kenan Tepe’s lower town, in several soundings near the high mound and on parts of the high mound itself (Parker et al. 2003; 2006). Carbon-14 analysis from Late Chalcolithic contexts has yielded dates in the late LC 3 or early LC 4 period (between ca. 3600 and 3500 BCE) and the LC 5 period (ca. 3100 BCE [Rothman 2001; Schwartz 2001]). Four more carbon dates from fortification/ retaining walls on the high mound show that occupation continued through the Late Chalcolithic to Early Bronze Age transition (ca. 3000 BCE). An analysis of the ceramics from various areas at Kenan Tepe combined with two carbon dates confirms that occupation at the site probably continued at least through the first half of the Early Bronze Age (Creekmore 2007). Middle Bronze Age remains have been recovered on the eastern, western and northern slopes of the high mound. Carbon-14 analysis places these remains around 1800 BCE (Parker et al. 2003; Parker and Dodd 2003; 2005). Kenan Tepe was again occupied in the Early Iron Age as evidenced by the presence of Early Iron Age Corrugated Wares dated to between ca. 1050 and 900 B.C. (Parker et al. 2003). Eight seasons of field research at Kenan Tepe have unearthed the remains of twelve ovens or oven-like structures, in various states of preservation, all of which resemble the tandır ovens still in use today. Five of these ovens date to the Late Chalcolithic period, two date to the first half of the Early Bronze Age, four date to the Early Iron Age and one is of uncertain date. Ovens and Their Contexts To date, no tandır or tandır like features have been excavated from Ubaid contexts at Kenan Tepe. We have, however, excavated a number of domestic contexts including a number of hearths (Parker and Dodd 2005; Parker et al. 2006). One of 18
these domestic contexts contained ample evidence for grain processing (Parker and Dodd 2005:72, figs 3-5). The lack of ovens in this period may be due to the luck of excavation – it is certainly possible that tandır style ovens existed in this period and have simply not been excavated. It is also possible that tandır style ovens did not exist in this period. Although it is likely that the inhabitants of Kenan Tepe’s Ubaid period village did use wheat (or barley?) flour to make some sort of bread or bread-like food, this may have taken any number of forms that did not necessitate this specific type of cooking facility. Two large dome-shaped ovens dating to the Late Chalcolithic period (D.5.5126 and D.9.22) have been excavated in Area D on the eastern slopes of Kenan Tepe’s main mound (figure 13). Although both of these ovens clearly belong to the Late Chalcolithic period, neither of them can be assigned to a specific phase within that period. In both cases Kenan Tepe’s Late Chalcolithic inhabitants cut niches into the steep slopes of the existing mound to form protected level surfaces for construction. In at least one case (D.5.5126), mud bricks were used to brace the resulting section before oven construction began. Visible remains of oven D.5.5126 included a circle of mud bricks that lined the vertical section of the protective niche (figure 14a and Parker et al. 2006; figs. 3 and 4) and an arch of mud bricks that made up the oven structure. The resulting oven measured approximately 1.2 meters in diameter at its base. No evidence of an oven core, such as those discovered in Early Iron Age contexts or recorded ethnographically, was discovered (contra Parker et al. 2006: 78). We are uncertain whether or not mud bricks were packed around the exterior of oven D.5.5126. However, it does appear that a coating of pisé was applied to the oven’s exterior. Oven D.9.22 was much better preserved than is counterpart in trench D5 (figure 14b). The unusual preservation of this oven allows a description of the method and order of its construction (Parker et al. 2006: 78). A level surface was created by digging 19
a niche into the slope of the mound. This niche probably provided some protection from the prevailing winds which, at least in modern times, tend to blow from the southwest. In contrast with the oven in trench D5 (D.5.5126), we do not have evidence that mud bricks were used to brace the vertical section created by digging the niche. Debris was thrown downhill, packed and covered with a layer of mud bricks, thus creating a relatively level surface in the steep slope of the mound. This surface was probably about 3 meters in diameter. It was identified in excavation as packed earth and brick with numerous flat-lying pot sherds and other cultural debris. No oven core was discovered. Instead, the beehive-shaped mud brick dome was achieved by narrowing the diameter of each successive row of bricks. Four mud bricks were set laterally within the oven, probably to elevate fuel or cooking pots and allow air circulation. Mud bricks measuring ca. 25 x 25 x 20 cm. were then packed around the dome. Most of the oven was covered with one row of bricks, although the bottom two courses appear to have been two rows thick. Finally, this brick lining was covered with a thick coat of pisé. The result was a large dome-shaped structure measuring 1.4 meters in diameter at its base. It was preserved to a height of 80 centimeters. No clearly definable openings were discovered in oven D.9.22 although part of this structure was contained in the south baulk and thus it is possible that either an opening may exist on the south side of this feature, or preservation did not allow us to identify such an opening. In any case, there must have been some sort of access to the lower portion of the oven to facilitate inserting fuel and extracting ash. The fact that no side openings for baking, cooking or other activities were discovered during excavation suggests that this was, like its modern parallels, top-loading. A few small domestic artifacts were discovered in association with D.9.22 including a stopper (D.9.20.11), and a loom weight (D.9.20.4). Burnt clay pieces with smoothed concave surfaces were discovered in association with the same feature (D.9.20.12). We assume that these are the remains of 20
earthen pot stands that were subjected to repeated contact with heated ceramic vessels. Another group of Late Chalcolithic dome-shaped ovens has been excavated in Area F in Kenan Tepe’s lower town (Fig. 13). The best preserved of these features is F.4.4027 (Fig. 14c). This is also the best dated and presumably the oldest oven excavated so far at the site. Although approximately half of this feature was contained within the western baulk of trench F4, we determined that its diameter was approximately 2 meters. Oven F.4.4027 was extraordinarily well preserved. It was composed of mud bricks laid lengthwise in two rows. Twelve courses measuring 1.3 meters in height were preserved. The walls of oven F.4.4027 curved slightly inward, suggesting that this structure was originally dome-shaped (Parker et al. 2003). The entire feature was sunk into virgin clay suggesting that this was one of the earliest features constructed on this part of the site. No remains of mud or mud brick debris other than the bricks that formed the oven dome were observed during excavation, no stoking hole was detected and no evidence of an oven core was discovered. Oven F.4.4027 was filled with debris that we interpret as secondary trash deposits that accumulated after the oven fell out of use. The debris consisted of numerous layers of black, gray and white ash with occasional lenses of clay, numerous animal bones and ceramics. About 25 percent of the animal bones from within oven F.4.4027 were burnt (this is almost triple the normal percentage of burned bone from the site as a whole). The nature of the burning does not suggest cooking or roasting, but secondary burning as garbage. Species represented include pig and sheep. Oven F.4.4027 also contained numerous artifacts including spindle whorls or loom weights made from bored pot sherds, a 12.4 by 2.8cm chert blade, and an andiron or kiln stand (Parker et al. 2003:118-119 and Figure 17). A carbon sample (KT4061) excavated from near the top of this debris yielded a 2 sigma calibrated carbon-14 date of 3350 to 2910 21
B.C. Three more carbon samples were extracted from the earliest layer of debris (L4023) at the base of the inside of this feature. These samples yielded 2 sigma calibrated dates of 33603030 BCE; 3630-3570 BCE/ 3540-3360 BCE, and 3660-3620 BCE/3600-3520 BCE (Parker et al. 2003:115). A pit (F.4.4022) located directly adjacent to the south side of oven F.4.4027 contained a series of ash and clay layers deposited against the outer wall of the oven. These deposits appear to derive from the same source as the debris inside the oven. The base of this pit terminated in virgin clay. Archaeobotanical remains from in and around oven F.4.4027 included charcoal from various tree types suggesting that wood or wood charcoal was used to fire oven F.4.4027. The remains of an LC 5 (late fourth to early third millennium) oven were discovered in the courtyard of a collapsed domestic structure in trench F1 (oven F.1.1045 [Fig. 14d]). Oven F.1.1045 did not have an oven core. Instead, it appears to have been constructed entirely of mud bricks. Unfortunately, only one 10cm tall course of its bricks was preserved, presumably because the oven was destroyed when the surrounding building burned and collapsed. Oven F.1.1045 was set into the floor of the courtyard. Its presence was revealed by a large ring of ash and a single row of mud bricks measuring ca. 1.5 meters in diameter (Fig. 14d). A rectangular clay stand was discovered inside the oven. This artifact was composed of a large mud brick with carefully smoothed sides measuring 64 x 30 x 10 cm. A fifth Late Chalcolithic oven was excavated in trench F7 (F.7.7124). Stratigraphically, F.7.7124 lies between the LC 5 oven discovered in trench F1 (F.1.1045 [discussed above]) and the Early Bronze Age ovens discovered in trenches F2 and F8 (see below). F.7.7124 consisted of a disarticulated mass of tandır oven core sherds excavated in association with a mud brick structure that protruded into the trench from the northwest baulk (Fig. 15a). Although these oven core sherds were resting atop a significant ash layer, the sherds themselves were 22
completely disarticulated and thus the size and structure of the feature to which they originally belonged cannot be ascertained. A terminus post quem is provided by a carbon sample taken from strata above this feature. This sample yielded a two sigma calibrated date of 3360 - 3020 BCE (Creekmore 2007). The stratigraphic position of this oven therefore suggests that it is the latest of the Late Chalcolithic ovens thus far excavated at Kenan Tepe (dating to the end of the LC 5). F.7.7124 is significant because it is the earliest oven excavated thus far that utilized a prefabricated oven core. These remains therefore mark a transition in oven technology at Kenan Tepe from the mud brick dome shaped ovens of the Late Chalcolthic period to the smaller ovens that include a separately produced core surrounded by a pisé structure in the later periods. Ovens F.2.2002 and F.8.8002 (figure 15b and 15c respectively) belong to the Early Bronze 1 (the first centuries of the third millennium B.C.). These ovens were discovered in association with an outdoor domestic area spanning a number of excavation units in Area F of Kenan Tepe’s lower town (Creekmore 2007). The total area covered by this surface was approximately 35 by 35 meters. This outdoor work area was composed of layers of river cobbles and contained a variety of cultural debris including ceramics and animal bone. Little or no architectural remains were discovered in association with this surface. Both of the Early Bronze 1 ovens (F.2.2002 and F.8.8002) were recognized by the presence of a clay core. As far as we could tell neither of these ovens was surrounded by mud bricks. Instead, the structure surrounding the core was manifested as eroded mud collapse. Oven core F.2.2002 measured 70 centimeters in diameter. It was preserved to a height of approximately 15 centimeters. Oven core F.8.8002 measured 55 centimeters in diameter and was preserved to a height of approximately 40 centimeters. In neither case did we find obvious indications of a stoking hole. The remains of both oven cores tapered slightly inward suggesting that both were dome-shaped. Both were 23
sunk between 15 and 20 centimeters into the surrounding surface. Near oven F.8.8002 we discovered three stone installations consisting of rings of cobblestones with a cobblestone base, or rings of odd-sized rocks and broken potsherds with no base. Although we did not detect burning associated with these installations, we interpret these remains as pot stands and/or food preparation facilities. To date no tandır style ovens have been excavated from Middle Bronze Age contexts at Kenan Tepe. There are several possible explanations for this. First, in spite of the fact that we believe we have excavated a variety of Middle Bronze Age domestic contexts at Kenan Tepe (Parker and Dodd 2003; 2005), one very likely explanation for the lack oven remains is that we simply have not been lucky enough to find them. A second possibility is that there are significant differences in the foodways of this period that made tandır style cooking facilities unnecessary, unpopular or redundant. It should also be noted that no Middle Bronze Age tandır style ovens have yet been reported from other sites in the Upper Tigris River region. Although it is obviously impossible to argue from negative evidence, such an absence might reflect, for example, a shift away from a bread-centered diet or a change in local or regional bread preferences. Three Early Iron Age tandır-like ovens were excavated from a single trench. Trench C3 is a 10 x10 meter unit located on the gently sloping western side of Kenan Tepe’s main mound (Fig. 13). Whether or not this apparent clustering is the result of luck, preservation, or of meaningful patterning, is uncertain. Nevertheless, this patterning does underscore the domestic nature of many of the contexts excavated in this area. C.3.3003 consisted of a well preserved dome-shaped oven core bonded to an extensive cobbled surface and associated with a wall (Fig. 15d and Parker et. al. 2003: 110-11). Very few remains of the structure of the oven (other than the core) were detected during excavation. The core measured approximately 24
60 centimeters in diameter and was preserved to a height of approximately 40 centimeters. Sherds extracted from C.3.3003 varied between 2 and 3 centimeters in thickness. C.3.3003 appears to have been manufactured in a similar way to modern cores. It was not made using a coil technique. Breaking points suggest instead that the core was built up by hand in stages using what appears to be a similar technique to those we observed at Çarıklı and in Bismil. Thumb prints were clearly visible along several of the joins. C.3.3003 was bonded to a cobbled surface (C.3.3027) made up of small stones and worn flatlying pot sherds. A small ash pit was discovered near by. The surrounding contexts contained copious amounts of animal bone, chert lithics, ceramics and a few pieces of slag. Among the ceramics were a number of blackened cook pot sherds and a fragment of a strainer. Surface C.3.3027 also bonded to another, this time rectangular, fire installation (C.3.3030) measuring approximately 2.5 x 1 meter. Although this feature was not well preserved, the presence of ash as well as animal bones in the surrounding loci lend evidence to the theory that this represents the remains of a cooking installation. The work area described above as the contexts associated with surface C.3.3027 appears to have been reused in a slightly later phase when another tandır style oven (C. 3.3008 [Fig. 16a]) was constructed a few meters away from, and at a slightly higher elevation than, C.3.3003. Although not as well preserved, oven core C.3.3008 is very similar to C.3.3003. It consisted of a dome-shaped hand-made oven core approximately 50 centimeters in diameter. Sherds from C.3.3008 vary between 2 and 3 centimeters in thickness. In this case, however there is evidence of at least part of the mud brick structure that presumably surrounded the core. These remains (C.3.3007) consisted of a disarticulated mound of mud and stone discovered in direct association with oven core C3.3008. C.3.3007 had presumably been heated on numerous occasions and thus appeared much harder and bonded together than the surrounding contexts. A few Early Iron Age cook pot fragments were discovered in the 25
debris inside C.3.3008 and two metal fragments were found in association with C.3.3007. Unfortunately we have very little information about the third and fourth Early Iron Age tandır style ovens (C.3.3006 and C.3.3023) recorded in trench C3 for the simple reason that, except for a very small portion, these features were almost entirely contained within the baulk dividing trenches C3 and C2 (Parker et al. 2003). Thus the stratigraphic relationship between these oven cores (C.3.3006, C.3.3023) and contexts in both trench C3 and trench C2 is uncertain. Nevertheless a few general remarks about the Early Iron Age remains, especially from trench C2, may be in order here. The Early Iron Age remains in trench C2 were characterized by a large round building which protruded into the trench from the north and a series of ashy areas that probably represent the remains of hearths. These data suggest that oven cores C.3.3006 and C.3.3023 rest on an outside surface, and that they may have been associated with one or more open hearths. It should also be noted that the ephemeral remains of a thin mud brick wall that may have been associated with these features was discovered in trench C3. One possible interpretation of this debris is that they are the remains of a structure meant to protect these ovens from the elements. Finally, a well preserved tandır style oven core was excavated in Area G which is located in the western portion of Kenan Tepe’s lower town (figure 13). These remains (G.9.7) were very close to ground surface in trench G9 and, although they may have been associated with an ephemeral presumably outdoor work surface, the dating of this feature and its associated loci is uncertain (figure 16b). However, it should be noted that, excluding a number of burials at the site, the latest remains excavated thus far in Kenan Tepe’s lower town date to the Early Bronze I. Oven G.9.7 consisted of a well preserved oven core surrounded by copious pisé debris. The oven core (G.9.7) measured 60 centimeters in diameter and was preserved to a height 26
of approximately 35 centimeters. Core G.9.7 did have a stoking hole measuring approximately 10 centimeters in diameter. Numerous tandır core sherds were found in and around the ashy matrix surrounding the oven core. Ethnoarchaeological Interpretations Combining the ethnographic and archaeological data outlined above allows a number of observations that might aid us in the interpretation of archaeological remains and in doing so guide excavation and sampling strategies. The most basic observation pertains to ceramics. During the first season of excavation at Kenan Tepe (in 2000), excavators unearthed an assemblage of thick low-fired handmade ceramic sherds from a number of different archaeological contexts. Subsequent ethnographic research showed that these ceramics were not the remains of large storage vessels or water containers but rather belonged to tandır-style oven cores. This observation not only lead to the reexamination of various categories of ceramics, but also helped guide the excavation and interpretation of various contexts at the site. For example, the appearance of oven core sherds often foreshadowed the existence of an oven and aided researchers in excavation and sampling (particularly botanical sampling) strategies. Ethnographic observation also aided us in interpreting the remains around oven cores. Since oven cores are fired ceramic material, their remains preserve relatively well in the archaeological record. The opposite is true of the pisé or mud brick material that makes up the oven structures surrounding the cores. Ethnographic observations alerted us to the fact that the hardened mud debris that we initially excavated as ‘hardened fill’ was likely the remains of the structures that surrounded the oven cores (Fig. 17). Observation of the construction and use of tandır ovens allowed us to compile a list of things to look for in and around tandır cores. For example, if and when an ash removal hole could be located in an ancient core, we deduced 27
from the ethnographic data that this hole would face the work area and would thus indicate where the ‘front’ of the oven was. This very simple observation was instrumental in guiding our sampling strategies and in aiding our context interpretation. Another feature that we soon began to look for during excavation was the protective structure that often surrounds modern ovens. An examination of the oven remains from Kenan Tepe shows that at least seven tandır style ovens had some sort of protection: two were built in niches cut into Kenan Tepe’s main mound (D.5.5126 and D.9.22), two were placed against the wall of a domestic structure (F.4.4027 and C.3.3003), one was located within the courtyard of a house (F.1.1045), and two may have been associated with a separate mud brick protective structure (C.3.3006 and C.3.3023). Ethnographic and archaeological work carried out as part of this project support the following observations about oven shelters. First, in many cases ovens are placed to take advantage of shelter provided by existing architecture or topography. In other cases ovens are protected by separately constructed oven shelters. Second, free standing oven shelters may be relatively ephemeral in the archaeological record since, like their modern counterparts, they are likely to have been rather flimsy constructions (Yakar 2000: 153. Also see Figs. 2 and 7). In cases where an ash removal hole can be distinguished in an oven core, this hole should face the open end of the three-sided shelter. Third, researchers should also keep in mind that freestanding oven shelters are more likely to be found in areas with warm climates. Ethnographic work in northeastern Anatolia, for example, shows that there tandır ovens are often to be located indoors (to conserve heat) and in some cases a family might have a summer tandır located in a courtyard or on a street and a winter tandır located within the house. Ethnographic work in the Upper Tigris has shown that, if regularly used, tandır ovens have to be rebuilt every few years. This research has also shown that, in many cases, multiple families 28
shared ovens. This has important archaeological implications. Several ethnoarchaeologists have observed that separate households often have their own tandır oven. They have gone on to say that households might be identified by the presence of a tandır and thus the number of tandırs in a given area may be a good indicator of the number of nuclear families (Cribb 1991; Kramer 1979). Based on our ethnoarchaeological work in the Upper Tigris and elsewhere (Yakar 2000: 153), we would argue that this correspondence cannot be assumed. Since the life expectancy of a tandır oven is, in fact, very short, any given household would have to rebuild an oven numerous times in a generation. Unless such ovens are rebuilt in exactly the some spot, a single household could potentially produce numerous tandırs over a given period.8 Furthermore, in most situations that we observed, ovens were shared by multiple households. The tandır ovens in use today in the Upper Tigris River region have a very specific function. They are, almost exclusively, used to bake flat bread. The limited function of these ovens is reflected in their construction: there are no openings that would allow pots or trays to be inserted, the heat cannot be directed to a burner or smoker, there are no racks or access points that might allow skewered foods to be cooked, etc. As noted above, the only access to these ovens is through either the ash removal hole, or the loading hole. The oven core is the most important component of a tandır oven. The function of the core is both to aid in heat retention and, most importantly, to provide a surface for the adherence of tandır bread dough. It is conceivable that tandır ovens could be used for firing ceramics, but the total absence of ceramic wasters in the archaeological record at Kenan Tepe and the lack of evidence for such a practice in the ethnographic record, makes this interpretation doubtful. Given the direct correspondence in construction and form between the modern and ancient tandır ovens discussed here, it is difficult to imagine that they served completely divergent functions. Instead, the ethnographic and archaeological data suggest that the ancient tandır ovens, like their modern 29
counterparts, were at least primarily used to bake flat (probably unleavened) bread. This conclusion is supported by the ethnohistorical record where tandır ovens appear to fill a very similar role to their modern counterparts.9 This comes as no surprise since bread has arguably been one of the most important nutritional sources in many Middle Eastern societies for millennia (Meyers 2002). One of the advantages to excavating a multi-period mound is the fact that this type of site gives a researcher the opportunity to view a large body of data in diachronic perspective. Stepping back to view larger trends in the archaeological data presented here, one diachronic observation is immediately obvious. There is clearly a change in the structure of the tandır style ovens between the Late Chalcolithic Period and the Early Bronze Age. To begin with, the Late Chalcolithic ovens are much larger than their Early Bronze Age counterparts (the LC ovens from Kenan Tepe are as much as ca. 1.5 meters in diameter while the Early Bronze Age and Iron Age ovens are about half that [see above]). Furthermore, the larger Late Chalcolithic ovens do not have prefabricated cores whereas the Early Bronze Age and Iron Age ovens do. Unless the Late Chalcolithic ovens were plastered on the inside (for this possibility see Yakar 2000: 139), they must not have been used for baking tandır bread since tandır dough would not adhere to an uneven surface. Thus the data suggest that the larger Late Chalcolithic ovens may have had a different function than the Early Bronze Age and Iron Age ovens. If we assume that tandır style ovens with separately manufactured oven cores were used for baking some form of bread, then this observation has important implications. Although the specific use of the Late Chalcolithic ovens is unknown, these data suggest that tandır bread became part of the foodways of the inhabitants of Kenan Tepe at or around the end of the LC 5 (as evidenced by F.7.7124) or the Early Bronze I (as evidenced by F.2.2002 and F.8.8002). Although it is obviously dangerous to argue from negative evidence, it is at least worth mentioning that the lack of tandır ovens in the Middle Bronze Age may be indicative of a shift in foodways during that period. 30
Conclusion This chapter has shown that ethnographic research can be a crucial component of archaeological field work. In addition to providing the archaeologist with a more intimate knowledge of the culture within which he or she conducts field work, ethnographic research can supply information critical to the interpretation of excavated remains. Furthermore, when archaeological and ethnographic research is conducted in tandem, researchers have the opportunity to continually readjust their research strategies as the data come in. In the case study described in this chapter, ethnographic research directly affected the classification of a particular class of ceramic material and, as a result, helped guide and refine archaeological excavation and sampling strategies. In addition, ethnographic data aided us in the interpretation, not only of particular archaeological contexts, but also of larger trends in the archaeological data. Since the behaviors that created the archaeological record are no longer observable, it is our position that combining ethnographic and archaeological data to produce ethnoarchaeological interpretations can allow researchers to construct theories that go beyond what can be said using archaeological data alone. Notes 1. The field research for this project was conducted by M. Barış Uzel including all of the interviews and any interaction with local informants. Interview planning and project guidance were shared by Parker and Uzel. The data processing and composition of this chapter is the work of B. J. Parker 2. Note the parallels described in Hopkins 2003: 55. 3. For parallels see Bakır 2004 and Ertuğ 2004. 4. This system is paralleled by potters in the Elazığ region (Ertuğ 2004). 5. Note the parallels observed by Ertuğ in the village of Uslu (Ertuğ 2004). 6. Hopkins has noted that although tandır cores are still produced in the village of Yiğittaşı, some villagers buy cores from itinerant merchants who transport them from as far away as Muş (Hopkins 2003: 55). A similar system was recorded by Angle and Dorarelli (1989). 7. For information about the Upper Tigris Archaeological Research Project please visit www.utarp.org. 8. This observation has been previously made by Hopkins (2003: 131). 9. Note however that Hopkins has documented tandır ovens being used to boil milk in order to make cheese (Hopkins 2003:43).
31
Bi bl io g r aph y Angle, M., and R. Dottarelli 1989 “Ethnoarchaeolog y at Uslu (Elazığ): A Preliminar y Repor t on Contemporary Pottery Manufacture in Eastern Anatolia.” VII. Araştırma Sonuçları Toplantısı II: 467-479. Bakır, T. 2004 “Domestic Pot Making in Yiğittaşı Village in Northeast Anatolia.” In Ethnoarchaeological Investigations in Rural Anatolia, vol. 1, edited by T. Takaoğlu. Istanbul: Ege Yayınları, pp. 67-76. Başgöz, I. 1952 “Turkish Folk Stories about the Lives of Minstrels.” The Journal of American Folklore 65: 331-339. Bottéro, J. 1985 “The Cuisine of Ancient Mesopotamia.” The Biblical Archaeologist 48:36-47. Creekmore, A. 2007 “The Upper Tigris Archaeological Research Project (UTARP): A Summary and Synthesis of the Late Chalcolithic and Early Bronze Age Remains from the First Three Seasons at Kenan Tepe.” Anatolica 33: 75-128. Cribb, R. 1991 Nomads in Archaeology. Cambridge: Cambridge University Press. Ertuğ, F. 2004 “Pottery Production at Uslu in the Elazığ Region.” In Ethnoarchaeological Investigations in Rural Anatolia, vol. 1, edited by T. Takaoğlu, Istanbul: Ege Yayınları, pp. 77-96. Greppen, J. A. C. 1991 “The Survival of Ancient Anatolian and Mesopotamian Vocabulary until the Present.” Journal of Near Eastern Studies 50: 203-207. Hopkins, L. 2003 Archaeology at the North-East Anatolian Frontier, VI: An Ethnoarcheological Study of Sos Höyük and Yiğittaşı Village. London: Peeters. Kramer, C. 1979 “An Archaeological View of a Contemporary Kurdish Village: Domestic A rchitecture, Household Size and Wealth.” In Ethnoarchaeology: Implications of Ethnography for Archaeology, edited by C. Kramer, New York: Columbia University Press, pp. 139-163. Marchese, R. 1995 “Assimilation and Culture Change: A Recent Journey Among the Turkish Nomads of Antalya.” Anatolica 21: 213-232. 2005
“Archaeological Perspectives in an Ethnographic Context: Interconnected Material Culture.” Ethnoarchaeological Investigations in Rural Anatolia, edited by T. Takaoğlu, Istanbul: Ege Yayınları, pp. 27-74.
32
Meyers, C. 2002 “Having their Space and Eating There Too: Bread Production and Female Power in Ancient Israelite Households.” Nashim: A Journal of Jewish Women’s Studies and Gender Issues:14-44. Parker, B. J., A. Creekmore, L. S. Dodd, R. Paine, C. Meegan, E. Moseman, M. Abraham, and P. Cobb. 2003 “The Upper Tigris Archaeological Research Project (UTARP): A Preliminary Report from the 2001 Field Season.” Anatolica 29:103-174. Parker, B. J., L. Dodd, A. Creekmore, E. Healey, and C. Painter 2006 “The Upper Tigris Archaeological Research Project (UTARP): A Preliminary Report from the 2003 and 2004 Field Seasons.” Anatolica 32:71-151. Parker, B. J. and L. S. Dodd 2003 “The Early Second Millennium Ceramic Assemblage from Kenan Tepe, Southeastern Turkey: A Preliminary Assessment.” Anatolian Studies 53: 33-69. 2005
“The Upper Tigris Archaeological Research Project (UTARP): A Preliminary Report from the 2002 Field Season.” Anatolica 30: 69-110.
Rothman, M. S. 2001 “The Local and the Regional: An Introduction.” In Uruk Mesopotamia & its Neighbors: Cross-Cultural Interactions in the Era of State Formation, edited by M. S. Rothman, Sante Fe: School of American Research Press, pp. 3-26. Runnels, C. 2005 “Ethnoarchaeolog y as a Sub-discipline of Archaeology.” In Ethno archaoelogical Investigations in Rural Anatolia, vol. 2, edited by T. Takaoğlu, Istanbul: Ege Yayınları, pp. 7-14. Schwartz, G. M. 2001 “Syria and the Uruk Expansion.” In Uruk Mesopotamia and its Neighbors: Cross-Cultural Interactions in the Era of State Formation, edited by M. S. Rothman, Santa Fe: School of American Research Press, pp. 233-264. Yakar, J. 2000 2004
Ethnoarchaeology of Anatolia. Tel Av iv: Emer y and Claire Yass Publications in Archaeology. “Ethnoarchaeolog y in Rura l A natolia.” In Ethnoarchaeolog ical Investigations in Rural Anatolia, vol. 1, edited by T. Takaoğlu, Istanbul: Ege Yayınları, pp. 7-14.
33
Fig. 1. A working tandır oven in the town of Bismil.
34
Fig. 2. Working tandır oven with protective shelter in the town if Bismil.
35
Fig. 3. Tandır bread adhering to the inside of a tandır oven during baking in the town is Bismil.
Fig. 4. Woman displaying fresh tandır bread in the town of Bismil.
36
Fig. 5. View of Kenan Tepe looking north with the Tigris River in the foreground.
Fig. 6. Tandır oven cores drying in the sun after production in the village of Çarıklı.
37
Fig. 7. Tandır cores drying in the sun.
Fig. 8. Map of modern Turkey with an enlargement of the Upper Tigris River region showing the location of the study areas. Note that the distance between Diyarbakır and Bismil is 70 kilometers.
38
Fig. 9. Boy mixing goat hair and salt into mud for the production of tandır oven cores in the in the village of Çarıklı.
Fig. 10. Stages of tandır core production as observed in the village of Çarıklı. The first stage of core production can be seen in the foreground. Cores in various further states of production can be seen in the center and background. A complete core is visible on the far left.
39
Fig. 11. Tandır oven during construction in the village of Çöltepe.
Fig. 12. Women constructing the outer structure of a tandır oven during construction in the village of Çöltepe.
40
Fig. 13. Topographic map of Kenan Tepe showing the various excavation units.
41
Fig. 14. The remains of ovens from Kenan Tepe. 11a is D.5.5126; 11b is D.9.22; 11c is F.4.4027; 11d is F.1.1045.
Fig. 15. The remains of ovens from Kenan Tepe. 12a is F.7.7124; 12b is F.2.2002; 12c is F.8.8002; 12d is C.3.3003.
42
Fig. 16. The remains of ovens from Kenan Tepe. 13a is C. 3.3008; 13b is G.9.7
Fig. 17. A dilapidated tandır oven in the village of Çöltepe. Note the mud and mud brick debris accumulating around the weathered loading hole.
43
Neol it h ic Shel l f ish G at her i n g at Ye şi lov a : A n Et h noa r ch aeolog ica l Vie w Z afe r D e r in
The significance of shellfish gathering has long been underestimated among archaeologists studying the role of marine resources in the subsistence base of prehistoric societies. As a result, marine shells identified among the cultural assemblages of prehistoric sites did not receive much attention in the literature until several decades ago when the importance of shellfish gathering as part of the prehistoric diet was accepted. Shellfish represent an easily-gathered year round food source found in shallow bays and no specialized equipment is required to procure them. By analyzing the shellfish remains, archaeologists can consider how prehistoric people interacted with the coastal environment. There are various studies proposing that shellfish exploitation was an important economic activity for the settlers of coastal western Anatolia and the Marmara region from Neolithic to Early Bronze Age. For example, archaeological investigations undertaken at the settlements of Ege Gübre, Yeşilova, Ulucak, and Yassıtepe in the İzmir Region (Derin 2006: 1; 2007, 381, fig. 17; Sağlamtimur 2007: 376; Çilingiroğlu 2007: 368), Fikirtepe, Pendik, and Ilıpınar in the eastern Marmara region (Buitenhuis 1995: 151), Coşkuntepe, 45
Beşik-Sivritepe, Kumtepe, and Gülpınar on the coastal Troad (Takaoğlu 2005: 424; 2006a: 412; 2006b: 49; 2006c: 311; Boessneck 1986: 332; Sperling 1976: 324), and Hocaçeşme at the mouth of River Meriç and Toptepe in Turkish Thrace (Özdoğan 1999: 15, fig. 25) have already provided us with great evidence about the exploitation of marine resources. Such studies are gradually changing the way archaeologists look at shellfish. It seems that past coastal communities supplemented their subsistence base by exploiting marine resources in shallow bays located close to their settlements on the coastal zones of western Anatolia (Takaoğlu 2006b: 49; Çakırlar 2006: 42). Shellfish gathering in this sense might have been an important supplementary economic activity along with agricultural pursuits in the Aegean and the Marmara region (Çakırlar 2006: 45; Karali 2002: 201). In addition to giving us information about what people ate and how they procured it, studying shellfish is also of archaeological importance for a number of reasons. In prehistoric times, marine shells were also actively used for various sother purposes. For example, they were often used to produce beads, adzes, and fishhooks. Crushed marine shells were used as a pottery temper. Marine shells, furthermore, often served as valuable items of trade during the Neolithic and Chalcolithic periods. There is also evidence to show that marine shells as valued artifacts were transported to the inland regions of Anatolia through exchange (Duru and Umurtak 2005: 138; Özdoğan et al. 1994: 112, fig. 3a, c). The inland societies evidently acquired marine shells through various forms of exchange for raw materials, agricultural products and artifacts from the coastal societies. Thus, it may be stated that marine sources played an important role in the development of prehistoric coastal communities. Examinations of modern patterns of shellfish gatherers in certain coastal zones of İzmir may help us to obtain a better picture of this economic strategy in relation to social and economic processes.
46
The Site and its Archaeological Context Recent archaeological excavations undertaken in the İzmir region in central-western Anatolia has begun to enrich our archaeological data of the Neolithic period in this geographically important region. Excavations of such sites as Ulucak, Ege Gübre, and Yeşilova illuminated various aspects of Neolithic life in the İzmir region, including subsistence strategies. Among these three Neolithic sites, the mound of Yeşilova provides us with valuable information about the ways coastal Neolithic villages supplemented their diet with marine resources, since a large assemblage of marine shells were identified in the course of archaeological excavations undertaken at the site in 2005 and 2006. The Neolithic mound of Yeşilova is located in the Karacaoğlan district of the town of Bornova in the province of Izmir (Fig. 1). Today, the mound is to be found between the Manda and Gökdere streams in the center of the Bornova Plain, roughly 3.5 to 4 km. from the coast. Yeşilova is basically a mound-type settlement covered by alluvial deposits (Fig. 2). Soundings undertaken in the settlement area have demonstrated that the site, with thick cultural deposits, was located on hill-formed alluvial deposits 14 m. above sea level during Neolithic times. These soundings also indicate that the mound was subsequently covered by alluvial deposits. Three major cultural levels have been identified at the site. Level I represents the Late Roman/ Early Byzantine period, the pottery remains of which have been found scattered over the surface on the site. Level II (phases 1-2) belongs to the Chalcolithic period, while Level III (phases 1-8) represents the Neolithic occupation at the mound. The Neolithic and Chalcolithic levels were determined about 80 cm. below the surface of the site. Archaeological excavations demonstrate that the site was among the earliest Neolithic villages so far known to us in the Aegean. In particular, the phases 6-8 of level III represent the oldest stages of the Neolithic in central western Anatolia.
47
The Subsistence Strategies of the Settlers of Yeşilova Evaluation of the archaeological evidence recovered from the mound of Yeşilova has so far provided us with significant information about the subsistence strategies and other aspects of life of Neolithic settlers of the İzmir region during the Neolithic period. As in most Neolithic Anatolian settlements, the inhabitants of Yeşilova fulfilled their subsistence needs in two ways. They either attempted to exploit wild species of animals and plants found around their settlements, or obtained their food through agricultural activities. It is clear that the exploitation of ready food resources was a common strategy during the earliest phases of the Neolithic period. Thus, the availability of food resources was an important factor in selection a site prior to settlement. The mound of Yeşilova in this context is located in an environment rich in natural food resources between two streams in the center of the Bornova Plain (Derin and Batmaz 2004: 78, 88-90) (Fig. 2). In addition to natural vegetation cover, the area of Yeşilova was also rich in fresh water springs. Although they are few in number, food processing implements discovered there, such as millstones, mortars, and pestles, can be accepted as evidence for the practice of cereal production at the site. Evidence for barley, wheat, and lentils has been identified among carbonized plant remains recovered from the floors of the houses and hearths confirms this.1 This can be accepted as evidence showing that the Neolithic settlers of Yeşilova practiced crop cultivation to supplement their subsistence base, since the land surrounding the site was optimal for tilling and there were ample water sources in the area for a productive crop yield. Analysis of animal bones from the cultural deposits of the site demonstrates that they also hunted wild boar and deer for meat. Faunal analysis also shows that stock-raising was practiced at the site as well. The recovery of large animal bones in the earliest phases (7 and 8) of the Neolithic Level at Yeşilova seems to imply that large animals such as cattle were consumed for meat. However, there was apparently an increase in the consumption of sheep, goats, and pigs in the later phases (3-6) of 48
the Neolithic level. At first glance, it seems that consuming sheep and goats was the prevailing pattern in the last phases (1-2) of the Neolithic level. This apparently points to a decline in the kill-off pattern of the cattle and pigs.2 This may imply that raising of sheep and goats gained importance in the local economy of the site. In addition to cereal production, stockraising, and hunting, the Neolithic inhabitants of Yeşilova also included oysters and mollusks as part of their subsistence base. The shallow waters located very close to the site to the west were also optimal for the exploitation of marine resources such as the edible species of oysters and mollusks. It appears that the prehistoric settlers of Yeşilova also relied heavily on shellfish as a source of food to supplement their diet. Over 2000 shells were identified in the cultural deposits of the site. Among this shell assemblage, the most numerous species represented are Cerastoderma glaucum, Arca noae, Callista chione, Hexaplex trunculus, and Bolinus brandaris, Ostrea edulis and Spondylus gaederopus (Fig. 4-5).3 Cerithium vulgatum, Patella spp., and Pectenidae are also represented among the marine shell assemblage, albeit in small numbers. Examples of Unio sp. peculiar to riverine environments have been identified among the assemblage, too. Moreover, the recovery of fragments representing the shell of a sea turtle demonstrates the variety of marine resources exploited for food. In addition to marine- and riverine-based shells, it seems that the settlers of Yeşilova occasionally included terrestrial species such as land snails as part of their diet (Derin 2007a: 381, fig. 17; 2007b: 125). Nearly all of these shellfish represented by numerous species of univalve and bivalve mollusks can still be found in the shallow bays and sandy environment near the coast close to the site. This gathering strategy appears to be a prevailing pattern at Yeşilova, since the site was located very close to the shoreline (Fig. 6). The prehistoric settlers of Yeşilova apparently consumed shellfish as a food source from the Neolithic to Chalcolithic period. A shell midden has been identified in ashy remains starting from phase 8 of the Neolithic Level (III), which dates 49
somewhere around the middle of the 7th millennium B.C. or earlier. The lack of traces of fire on these shells seems to indicate that shellfish were prepared for food in the ashes of fire in areas located outside the living units. These shellfish appear to have been eaten in the area where they were cooked (Fig. 8). There seems to be an increase in the consumption of shellfish towards the last phases of the Neolithic level and throughout the following Chalcolithic period, implying that shellfish consumption was as important as other animal-based food consumption at the site. This pattern might have been based on the rise in sea level during this period, eventually bringing the shoreline closer to the settlement than before.4 This, in fact, made shellfish gathering a practical strategy for the settlers of the site. The evidence in particular indicates how coastal Neolithic societies included alternative non-agricultural food resources such as shellfish in their diet. Ethnographic Evidence for Shellfish Gathering Examination of present-day shellfish gatherers on the shores of the İnciraltı and Bostanlı districts of coastal İzmir help us to obtain a picture of past coastal exploitation strategies. Because gathering shellfish at present is primarily the work of both women and men who came to İzmir from east and southeast Anatolia to earn their living, this ethnographic evidence provides an opportunity to see how these people took advantage of the marine environment on this coastal part of İzmir. Evaluation of these modern shellfish gatherers in the region shows that shellfish gathering is a seasonal activity, lasting from the beginning of November to the end of June. Although it is not an activity geared merely towards local consumption in this region, we can still get an idea of what kind of implements might have been used in shellfish gathering in the past by examing the modern implements used by shellfish gatherers nowadays. Modern shellfish gatherers use wooden framed rectangular sieves, shovels with wide a blade, and mesh sacks. Each individual gathers a maximum of 4 kilograms of shellfish in a single day. Examination of the behavioral aspects of 50
modern shellfish gatherers also show that men undertake the gathering of shellfish from the shallow waters, while women are responsible for the cleaning of the newly gathered shellfish on the shore. The shellfish gatherers, who live in the district of Kadifekale in İzmir, also turn the shells into ornaments to sell them in regional and inter-regional markets. Although the current shellfish gathering activity appears to have been developed only over the past few several decades, this modern case can still be used for archaeological purposes. For example, ethnographic evidence for shellfish gathering in this coastal area rich in marine resources can show us one way how the Neolithic populations of the region might have adapted themselves in such an environment and benefited from natural marine resources located around their settlements located near natural bays, since shellfish had a potential importance for the prehistoric diet. More archaeological and ethnographic studies are needed for a better understanding of the role of marine resources within the diet of the prehistoric populations of coastal western Anatolia. Acknowledgement I am thankful to Turan Takaoğlu for his contributions and comments during the preparation of this essay. Notes 1. Ayla Erkal (M.A.) has been preparing her Ph.D. dissertation on the plant remains of ther sites of this region. This information supplied in the text is based on her preliminary observations. 2. The prelimnary report about the animal remains from the site of Yeşilova is prepared by Can Yümni Gündem. 3. The preliminary study of marine shells recovered from the site of Yeşilova has been undertaken by Canan Çakırlar. 4. The sea level was 40 m lower than present day one about 10.000 years before present. It gradually first rose up to 2 m above present day sea level about 5000 years ago. See Kayan 1988 and 1996. The fact that the sea level of the Aegean started to rise from the beginning of Neolithic might have effected the situation at the site.
51
R efer enc e s Boessneck, J. 1986 “Die molluskenfunde.” Archäologischer Anzeiger, 332-337. Buitenhuis, H. 1995 “Chapter 9: The faunal remains.” In The Ilıpınar Excavations I. Istanbul: Nederlands Historisch-archaeologisch Institut, edited by J.Rodenberg, pp. 151-156. Çakırlar, C. 2006 “Arkeomalakoloji: Yabancı Bir Terim, Tanıdık Buluntular.” Arkeoloji Dergisi VII, Ege Üniversitesi Edebiyat Fakültesi Yayınları, pp. 41-50. Çilingiroğlu, A. and Ç. Çilingiroğlu 2007 “Ulucak.” In Türkiye’de Neolitik Dönem, Yeni Kazılar, Yeni Bulgular, edited by M. Özdoğan and N. Başgelen. Istanbul: Arkeoloji ve Sanat Yayınları, pp. 361-372. Derin, Z. 2006 “İzmir’den İki Yeni Prehistorik Yerleşim: Yassıtepe Höyüğü, Çakallar Tepesi Höyüğü.” Arkeoloji Dergisi VII, Ege Üniversitesi Edebiyat Fakültesi Yayınları, pp. 1-16. 2007a
“Yeşilova Höyüğü.” In Türkiye’de Neolitik Dönem, Yeni Kazılar, Yeni Bulgular, edited by M.Özdoğan and N.Başgelen. Istanbul: Arkeoloji ve Sanat Yayınları, pp. 377-384.
2007b
“Türkiye’de 2006 yılında Yapılan Araştırmalar ve Kazılar- Yeşilova.” Turkish Academy of Scienes Journal of Archaeology 10: 125-127.
Derin, Z. and A. Batmaz 2004 “Bornova- Kemalpaşa (İzmir) Arkeolojik Envanteri 2003.” Tüba Kültür Envanteri Dergisi 2004.2: 75-100 Duru, R. and G. Umurtak 2005 Höyücek. 1989 -1992 yılları Arasında Yapılan Kazıların Sonuçları. Ankara: Türk Tarih Kurumu. Erlandson, J. 1988 “The role of Shellfish in Prehistoric Economies: a Protein Perspective.” American Antiquity 53: 102–109. Karali, L. 2002 “Ftelia on Mykonos: The Molluscan Material.” In The Neolithic Settlement at Ftelia, Mykonos, (A. Sampson), Rhodes: The University of the Aegean, pp. 201-220. Kayan, İ. 1988 “Late Holocene Sea-Level Changes on the Western Anatolian Coast.” Paleogeography, Paleoclimatology, Paleoecolpgy 68: 205-218. 1996
“Holocene Coastal Development and Archaeology in Turkey.” Zeitschrift für Geomorphologie 102: 37-59.
Özdoğan, M.
52
1993
“The second millennium of the Marmara Reg ion.” Istanbuler Mitteilungen 43: 151-162.
1999
“Anadolu’dan Avrupa’ya Açılan Kapı Trakya.” Arkeoloji ve Sanat 90: 2-28.
Özdoğan, M., A. Özdoğan, D. Bar Yosef, W. Van Zeist 1994
“Çayönü Kazısı ve Güneydoğu Anadolu Karma Projesi: 30 yıllık Genel bir Değerlendirme.” XV. Kazı Sonuçları Toplantısı I: 103-122.
Shackleton, J. C. and Tj. H. van Andel 1986
“Prehistoric Shore Environments, Shellfish Availability, and Shellfish gathering at Franchthi, Greece.” Geoarchaeology 1: 127-143.
Sağlamtimur, H. 2007
“Ege Gübre Neolitik Yerleşimi.” In Türkiye’de Neolitik Dönem, Yeni Kazılar, Yeni Bulgular, edited by M.Özdoğan and N.Başgelen. Istanbul: Arkeoloji ve Sanat Yayınları, pp. 373-376.
Sperling, J. 1976
“Kumtepe in the Troad. Trial excavations.” Hesperia 45, 305-64.
Takaoğlu, T. 2005
“Coşkuntepe: A Neolithic Quern Production Site in NW Turkey.” Journal of Field Archaeology 35: 419-433.
2006a
“2004 yılı Coşkuntepe Yüzey Araştırması.” 23. Araştırma Sonuçları ToplantısıI: 411-438.
2006b
“Homeros’un Gölgesinde Troia öncesi Troas Araştırmaları.” In Sevim Buluç Anı Kitabı/In Memoriam Sevim Buluç, edited by V. Tolun and T.
Takaoğlu, Çanakkale: Olay Matbaası, pp. 47-62. 2006c
“Late Neolithic in the Eastern Aegean: Excavations at Gülpınar in the Troad.” Hesperia 35: 289-315.
53
Fig. 1. Map locating the Neolithic mound of Yeşilova in the İzmir-Bornova Plain.
Fig. 2. A view of the mound and the excavated area from the south .
54
Fig. 3. Shells of landsnails in a bowl from the Neolithic level III.
Fig. 4. Some of the marine shells recovered from Yeşilova.
55
Fig. 5. A Spondylus gaederopus shell from Yeşilova.
Fig. 6. Present day shellfish gathering at shallow waters at the district of İnciraltı of İzmir Province.
Fig. 7. Necklaces made from Cardium shells at Yeşilova.
56
Fig. 8. A view of the southwestern corner of trench M15 b where marine shells were processed for food.
57
Tr ad it ion a l Modes of Vit ic u lt u r e i n C oa s t a l Nor t hwes ter n A n atol i a D av ut K apl an
Grape cultivation for wine production was an important component of the agricultural economy of the societies of northwestern Anatolia until several decades ago. Vineyards dominated certain parts of the landscape of this region for millennia until a gradual decline started to occur in the intensity of grape cultivation following the population exchange that took place between Turkey and Greece in 1923 under the Treaty of Lausanne. Joined by incoming Turks from the Balkans, the local Turkish populations of this region gradually started to replace most vineyards with olive groves because of religious prohibitions and high taxes levied on wine producers. As a result of this switch from wine to olive oil, most farmers of the region begun to discard primitive installations used in processing grapes for wine, while some Turkish farmers casually continued to employ their traditional implements to produce wine (şarap) and grape molasses (pekmez). The region was scene to grape growing as early as the Early Bronze Age, since domesticated grape seeds dating from the Early Bronze Age were recovered from Kumtepe and Troy, as well as from Yenibademli Höyük on the nearby island of Imbros/Gökçeada (Riehl 1999: 382; Oybak-Dönmez 2002: 25). 59
The quality of grapes and wines of the coastal zones of northwestern Anatolia and adjacent islands such as Tenedos/Bozcaada and Lesbos/Midilli were particularly renowned during Classical Antiquity, as well as in the succeeding Byzantine and Ottoman periods. There is unfortunately not enough archaeological evidence for installations and implements that could demonstrate the economic importance of grape growing for wine making in the region. Evaluation of extant evidence, however, in part helps us to picture the methods and techniques adopted by farmers of the region in the production of wine during the past several centuries. Available evidence shows the installations used in wine production have not changed much since Byzantine times, in fact, making it possible to reconstruct in part the agricultural history of the region in terms of grape growing or wine making. In order to identify those structural remains that could be interpreted as belonging to installations deliberately constructed for the processing of grapes for wine, I conducted ethnoarchaeological surveys in the western parts of coastal northwestern Anatolia in the summers of 2005 and 2006 (Fig. 1). Interviews with the oldest living farmers of the region helped me to identify the fields that once served as vineyards. I noted that most vineyards were once located particularly on small plains on the slopes of natural bays or on alluvial plains and surrounding slopes near the shoreline. In such abandoned vineyards, I observed the remains of dry stone-built field houses and attached features such as water-wells, together with grape processing installations, including stone grape treading tubs, stone collecting vats, and parts of screw presses. It seems that such stone-built field houses with attached features were purposely constructed in the vineyards to process grapes for wine or molasses during cycles of the agricultural calendar (Figs. 2-3). This pattern of exploitation of vineyards through such seasonal field houses might have had its roots far back in the past. Therefore, I focused on surveying abandoned vineyards located on coastal zones of northwestern Anatolia 60
to obtain as much information as possible about these rapidly deteriorating traditional wine making installations in order to speculate about the region’s extinct modes of wine making. Although installations and implements used in wine making are mainly found near the seasonal field houses located in the vineyards, there are cases in which the grapes were taken to special press-rooms located at villages or towns. However, wine producers often sought to minimize transportation costs by having the grapes processed in the vineyards rather than having to transporting them to distant installations in villages or towns. Those farmers who opted to process their grapes for wine constructed seasonal field houses in their vineyards to minimize the cost of production. A typical seasonal field house of the region is composed of a modest small one-room dry stone-built dwelling with a flat roof. Such dwellings at time were characteristic of the region’s architectural landscape. Farmers seasonally inhabited these dwellings to cultivate, harvest and process the grapes for wine. These fairly small stone-built dwellings were often located about half an hour’s walking distance from the villages. Structures located in vineyards also bore certain features such as wells, indoor or outdoor ovens, treading tubs, collecting vats, screw presses, and pottery containers. A similar pattern involving the exploitation of vineyards from a major center of population through seasonal stone-built small structures has also been well documented on the nearby island of Tenedos/Bozcaada (Takaoğlu and Bamyacı 2005). It is reasonable to argue that most parts of coastal northwestern Anatolia and the island of Tenedos/Bozcaada shared similar agricultural practices. Because most such vineyards were abandoned long ago and related wine- making installations were discarded, I also attempted to reveal the traditional methods of wine making by benefiting from the knowledge of the oldest living farmers of the region. They provided me with useful information about the stages of wine making and related implement use. The steps of 61
wine making can be reconstructed as follows: After the grapes had been sorted and selected, they were trodden in tubs carved out of local andesite (e.g. Fig. 4). The pressed juice of grapes from the treading flowed directly from the tub into a stone collecting vat placed beside the tub. The grape juice that filled this stone vat was then transferred into large pottery fermentation containers. The grape pulp remaining in the stone treading tub was subsequently placed under a screw press to extract more juice, which is of inferior quality to the juice extracted from treading. Such screw presses are predominantly installed near these stone treading tubs. The use of such mechanical presses in conjunction with treading tubs and stone reservoirs attached to them seems to have been a very common pattern in northwestern Anatolia during the last several centuries, if not earlier. The type of screw press commonly used by wine producers of the region to extract juice from crushed grapes is a rigid-frame wooden screw press (e.g., Fig. 5). It has a single rotary screw press made of oak. Such wooden screw presses exerted direct pressure on the pressing box filled with the crushed grapes. Pressure is applied through a plate placed on top of the wooden square box, which is forced down onto the grapes through direct pressure from the screw. The juice flows through an opening in the lower end of the wooden box. A rectangular channel was cut into the top of the press-beds so that the lower ends of the wooden box could be embedded into it. A second rectangular channel surrounding the former one was also cut into the top of the press. This channel terminates in a spout which directed the extracted grape juice into a collecting vat attached to it. Opposite these channels, two holes were bored all the way through the press-bed. Two wooden poles were set into these holes. A threaded plank was then mounted horizontally across these wooden posts. The wooden screw passed through the threaded hole in the center of this horizontal plank. A wooden board was often placed at the tip of the screw. To extract more juice out of crushed grapes placed in the wooden box, the screw was turned 62
by hand until it pressed on the wooden board. The screw was operated with the help of a wooden rod fitted into the hole in the base of the screw. Examinations of ethnographic evidence show the press-beds of screw presses are of either wood (Figs. 6-7) or stone (Fig. 8). I have identified dozens of such wooden and stone press beds during surveys undertaken in the region. Most of those stone press-beds were often made from local andesite or marble blocks belonging to ancient monuments. As mentioned above, stone press-beds were frequently found at locations near abandoned vineyards. There are also cases in which such press-beds were inserted into the modern terrace walls after the wine making installations and implements were discarded (Fig. 9). Interviews with the oldest living farmers of the region proved that wooden screw presses were actively being used in the region until several decades ago. We find evidence for such discarded screw presses in several remote villages in northwestern Anatolia. It is likely that such wooden screw presses were in use in northwestern Anatolia since medieval times. There are a number of studies on oil and wine production in the Holy Land that help us to interpret the function and dating of presses from northwestern Anatolia. For example, Frankel (1999: 71-73) demonstrated that installations employing stone press-beds bearing two mortises were not exclusively used in oil extraction plants. He showed that such press-beds for rigid-frame screw presses were also used in wineries for pressing the grape mash which remained after treading. These comparable examples documented in western Galilee in the Holy Land date from the medieval period, although such installations appear to have been in use there until the last decades of the nineteenth century. Avitsur (1999: 123) also demonstrated the mechanisms of rigid-frame screw presses used in olive oil extraction. However, there are differences between the rigid-frame screw press type common in northwestern Anatolia and that of the Holy Land. The placement of a wooden box below the end of the screw and the cutting of channel on the press-bed to secure the box do 63
not appear on the examples from the Holy Land. Wooden screw presses found in northwestern Anatolia have close parallels in both medieval and post-medieval Europe. It is difficult to estimate the efficiency of wine making in the past of northwestern Anatolia. The presence of wine making installations in the vineyards instead of major population centers points to a form of production for local consumption. Although much of the wine making was undertaken by Greek populations, Turkish populations were also involved in wine production for commercial purposes to meet the demand from internal markets. Although most farmers of the region preferred to process their grapes in their vineyards, some opted to turn the grapes into wine in specially designed rooms located at towns or installations near major population centers. We occasionally find useful information on the locus of grape processing in the accounts of travelers who visited the region in the nineteenth century. For example, during his journey to the Dardanelles in 1830s, William Knight noted an installation just outside the town of Çanakkale where a Greek farmer produces wine in a specially designed room. He states that “Our supply of fowls we got from a female higgler, closely muffled up, in all but one eye; and we obtained our vegetables from the garden of a Greek outside the town. Before entering this garden we passed through a large room in which was a high wine-press, the top of which we reached by a ladder, and found a Greek inside stamping with naked feet upon cart-loads of grapes, with which he was being supplied by men whose wicker arabahs, filled with this luscious fruit and horsed by buffaloes, were at the door (Knight 1849: 22).” During his journey alongside the coast of Anatolian mainland opposite of Tenedos/Bozcaada, the traveler Thomas MacGill noted the presence of many vineyards close to the shoreline (McGill 1808: 142). He informs us that these vineyards were cultivated by the settlers from the island of Tenedos. Most farmers from the island reportedly came over in the spring and ereceted temporary buildings in order to exploit these vineyards. MacGill also informs us that the grapes 64
were carried over and only trodden out on the island instead of the vineyards where they were collected. Similar information can also be found in the accounts of John P. Brown published in a New York-based Magazine entitled Knickerbocker (1844: 330). Brown mentions a native of Tenedos owning a vineyard in the Anatolian mainland opposite of the island. This man from Tenedos reportedely came over in the spring to cultivate his vineyard, and in the fall to collect its fruit. The ethnographic implements collected in this area seem to have mainly belonged to such activities dating to the nineteenth century. The coastal areas just opposite of the island of Tenedos/Bozcaada seem to be exploited by those farmers from the island. This pattern was also in existence in the region in the Classical and Helelnistic periods in the region, when Tenedos was one of the major centers of wine making and the lands of this small island was not enough to fulfill the needs of domestic and external markets (Takaoğlu and Bamyacı 2005: 121-122). Evaluation of available ethnographic evidence from northwestern Anatolia shows that the mode of grape growing and wine making might not have changed much in the region since medieval times. Despite religious prohibitions and high taxes levied on wine producers, local Greek and Turkish populations of the region continued to grow grapes for wine and grape molasses. It seems that the modes of grape processing for wine was very uniform in this part of northwestern Anatolia, as most households involved in wine making had adopted similar methods. In light of the fact most such primitive winemaking locations with grape prcessing equipments are rapidly disppaearing from the region’s ethnographic record, it is hoped that this brief essay aimed to record the available information on the traditional modes of wine making will be useful contribution to more detailed studies that would be undertaken in the future. This is important because there are not many studies carried on the ancient and ethnographic examples of wine making. Fortunately, the archaeological and ethnographic record of wine making activities in the ancient regions of Caria 65
and Cilicia has been the subject of studies by Adnan Diler, who has already accumulated a great deal of evidence about the technological aspects of grape processing for wine (e.g., Diler 1994 and 2005). Thus, it is hoped that this essay will further stimulate archaeologists and ethnographers to study the technological aspects of grape growing and wine making among extant populations of northwestern Anatolia as a key to understanding the modes of similar production among extinct populations. More works also need to be done on the social and economic relatiosnof wine making process in order to integrate grape growing to its larger framework. A closer examination of the Ottoman written sources will definitely be very useful to reconstruct the social and economic processes that gave action to grape growing for wine in the region’s last five hundred years.
66
R efer enc e s Avitsur, S. 1999 “Olive Oil Production in the Land of Israel: Traditional to Industrial.” In Wine and Oil Production in Antiquity in Israel and other Mediterranean Countries, edited by E. Ayalon, Sheffield: Sheffield Academic Press, pp. 91-158. Brown, John P. 1844 “A Visit to Mount Ida, in Mysia, Asia Minor.” The Knickerbocker/New York Monthly Magazine 24: 225-342. Diler, A. 1994 2005
“A kdeniz Bölgesi Antik Çağ Zeytin ve Üzüm Presleri, 1993.” XII. Araştırma Sonuçları Toplantısı I: 441-457. “Karya’da Zeytinyağı ve Şarap Üretimi.” Ramazan Özgan’a Armağan, edited by M. Şahin and İ. H. Mert, Istanbul: Ege Yayınevi, pp. 79-86.
Frankel, R. 1997 “Presses for Oil and Wine in the Southern Levant in the Byzantine Period.” Dumbarton Oak Papers 51: 73-84. 1999
“Ancient Oil Mills and Presses in the Land of Israel.” In Wine and Oil Production in Antiquity in Israel and other Mediterranean Countries, edited by E. Ayalon, Sheffield: Sheffield Academic Press, pp. 19-90.
Knight, W. 1849 A Dairy in the Dardanelles, Written on Board the Schooner “Corsair” while Beating through the Straits from Tenedos to Marmora. London: Hunt. MacGill, T. 1808 Travels in Turkey, Italy and Russia during the Years 1803, 1804, 1805 & 1806. London: Hohn Murray. Oybak-Dönmez, E. 2003 “Arkeobotanik Çalışmalar Işığında Tarih Öncesi Anadolu’da Asma.” In Türkiye V. Bağcılık ve Şarapçılık Sempozyumu, edited by S. Ağaoğlu. Ankara: Ankara Üniversitesi, pp. 22-30. Riehl, S. 1999 “Archäobotanik in der Troas.” Studia Troica 9: 367-409. Takaoğlu, T. and A.O. Bamyacı 2005 “Continuity and Change in Rural Land Use on Tenedos/Bozcaada.” In Ethnoarchaeological Investigations in Rural Anatolia, vol. II, edited by T. Takaoğlu. Istanbul: Ege Yayınları, pp. 113-136.
67
Fig. 1. Map locating major modern and ancient sites mentioned in the text.
Fig. 2. An example of a seasonal stone-built field house found in an abandoned vineyard at Tuzla.
68
Fig. 3. An example of abandoned dry stone-built seasonal field house located on a small coastal plain near the ancient city of Polymedeion.
Fig. 4. An example of abandoned modern grape treading installation near a seasonal field house illustrated in Figure 3. Note the stone treading tub and collecting vats, as well as pools to store collected grape before being trodden.
69
Fig. 5. A discarded wooden rigid-frame screw press mounted on wooden posts. Such presses were typical of the region until several decades ago.
Fig. 6. Wooden basin for a rigid-frame screw press.
70
Fig. 7. A tentative drawing of a wooden rigid-frame screw press mounted on wooden posts. Such presses were commonly used to extract more juice from the pulp left in the treading tub until several decades ago (drawn by A. Özdemir).
71
Fig. 8. Andesite press-bed for a rigid-frame screw press inserted into a stone-built fence of a field that was once a wineyard near Gülpınar.
a
b
Fig. 9a-b. Examples of discarded stone beds for a screw driven wine press have been used as spolia within the walls of abandoned seasonal field houses dating to 1930s.
72
A n E x per i ment a l St udy of Mat Impr es sion s on Pot B a ses f r om C h a lcol it h ic G ü lpı n a r (Sm i nt heion) Ab dülk a dir Ö z d e mir
Recent archaeolog ical excavations under taken at the Chalcolithic site of Gülpınar (Smintheion) in northwestern Anatolia have revealed nearly 100 pot bases bearing the negative impressions of mats (hasır) and clothes. Although negative impressions of clothes or textiles (wool and cotton) have also been identified on the base of pots from Gülpınar, here I aim to interpret only mat impressions left on pot bases by carrying out experimental studies and benefiting from ethnographic parallels. Mats are basically two-dimensional or flat items used mainly for covering the floors of houses in prehistoric times. The discovery of mat impressions on the bases of hand-made pots at Chalcolithic Gülpınar is important because very little archaeological evidence exists regarding the use of mats in the daily life of prehistoric societies. Besides Gülpınar, indirect archaeological evidence for mat-making is known to us only from casual finds identified at other 5th millennium B.C. Troadic sites such as Beşik-Sivritepe and Alacalıgöl (Gabriel: 2006: 358, fig. 2; Aslan 1997: fig. 30). Although mat making was common in Neolithic Anatolia, we do not have much evidence about them in the Chalcolithic period. The discovery 73
of numerous pot sherds with impressions of mats at Gülpınar now provides us with an opportunity to better interpret the use of mats in prehistoric times. Those archaeologists aiming to explain the occurrence of negative impressions of mats on pot bases mainly adopt two different views (Carrington Smith 1977). Some maintain that they are derived from the fact that potters place freshly-shaped pots on mats to dry before they are fired. This method results in light impressions of mats on pot bases. Others argue that the impressions result from the use of mats as primitive turntables, which leave deep impressions on pot bases. By taking these two different views, I aim to interpret the occurrence of negative impressions of mats on pot bases from Chalcolithic Gülpınar by undertaking several experimental studies. The results of my preliminary replication studies, indeed, show that rectangular and circular mats were occasionally used as a form of primitive turntable by the potters of Gülpınar. At Chalcolithic Gülpınar, plaiting and coiling appear to have been the two different matting methods adopted at the site. Analysis of negative impressions left on pot bases points to the use of two different methods of plaiting at Gülpınar: simple plaiting (plain-weave matting) and twill plaiting. Simple plaiting is the most elementary of all mat-making techniques. Single warps and wefts pass over and under each other at a 90 ° angle in a 1/1 interval (Fig. 1a). This technique is so far represented by only 5 examples at Gülpınar (e.g., Fig. 3a). In twill plaiting, single elements pass over each other in a 2/2 interval. Twill plaiting is the most common technique at Gülpınar as it is represented by nearly 30 examples (e.g., Fig. 3b). Although plaiting was used in the production of threedimensional items like baskets, containers and bags, it was also used for two-dimensional items like floor mats, wall hangings, and screens. Available specimens appear to be mainly made of reeds that look like Juncus sp., Scirpus sp. and Typha sp., which are still to be attested in marshy areas around the site.
74
Coiled mats are made by taking a long bundle of grass and coiling it around and around upon itself in a spiral, each new turn of the coil is attached to the preceding one by being sewn on with stitches. Although the coiling method was often used in the production of three-dimensional items such as baskets, containers, bags, and hats, archaeological evidence shows that coiled matting was seldom used at Gülpınar. Negative impressions of coiled matting have so far been found on the base of only one large jar at Chalcolithic Gülpınar (Fig. 5). Its diameter at base is about 15 cm. No traces of mats have been observed on the preserved side walls of this base of a large jar. It seems that that the potters of Gülpınar also made their pots on circular coiled mat bases. The clay was probably placed on a circular coiled mat so that it could easily be rotated by hand on the ground. This type of base allowed the potter to move the pot around as the work progressed. A slightly thicker knot occurring on the center of the coiled mat prevents continuous contact with the ground surface, making it act like a primitive form of turntable for the manufacture of large bowls and jars. In the case of the Gülpınar example, the circular mat base on which the foundations were laid is of the same diameter as the intended base of the jar. This is because the center of the negative impression of coiled matting matches the center of the pot base on which impressions were found. One may argue in this context that the diameter of the circular mat on which the clay was shaped determined the diameter of the intended pot. If the potter had formed the clay on a large circular mat, then the center of the mat would have been quite further from the center of the pot base. One may also expect that this allowed the potter to carry the newly formed pot to the drying area together with this circular mat base in order to prevent the pot from sticking to the ground while drying before it was fired. However, since each time a given pot was formed the circular mat had to be removed from the base in order to be used for the production of another one, then the freshly formed pots were probably not left on a circular mat to dry firing.
75
We know of presence of coiled mats in prehistoric Anatolia only from negative impressions on pots at Çatal Höyük and Domuztepe (Mellaart 1967; 198, pl. 119; Wendrich 2006 and 2007: 231; Carter et al. 2003). The closest parallels for the use of mat coiling are known to us from several sites such as Saliagos, Kephala, and Amorgos in the Aegean (Evans and Renfrew 1967: 71, pl. 55; Myres 1898: fig. 1). The negative impression of coiled mats observed on the bases of vessels from Saliagos was thought to have resulted from their placement on circular mats to dry before they were fired. Myres was the first to consider the possibility that pots were formed on circular coiled mats of the same size as the intended base (Myres 1898: 179). The use of circular mats made out of coiled matting as primitive potters’ turntables was also suggested for pre-Dynastic Egypt (Lucas 1962; Johnston 1974: 93). Grace Mary Crowfoot (1938), who examined mat-making and basketry traditions in Palestinian villages in the 1930s, first pointed out that a circular coiled mat was ideal for use as a primitive turntable. She also stated that this was a practice which continued into the nineteenth century. It is clear that the settlers of Chalcolithic Gülpınar knew how to make both rectangular plaited mats and circular coiled mats. With the help of pottery specialist Ergun Arda from the Ceramics Department of the Fine Arts Faculty at Çanakkale Onsekiz Mart University, I undertook a series of replication studies in his workshop. After obtaining several specimens from modern mat-makers in the Biga Town of Çanakkale Province, I formed four clay vessels of varying sizes in order to place them on mats to dry before they were fired (Figs. 5-6). I placed two of them on a rectangular plaited mat, while the other two were left on a circular coiled mat. Two days later, I observed the light impression of mats on both small and large sized pots. I noticed that these light impressions on the bases of experimentally-produced pots are not similar to those deep impressions found on the archaeological examples from Gülpınar. Therefore, my experimental studies do not favour the 76
argument that the Chalcolithic potters of Gülpınar placed their freshly-shaped pots on mats covering the floor to dry before they were fired. Instead, this replications study implies that the certain pots were actually formed on rectangular plaited mats or circular coiled mats. I carried out a further replication study to show that mats might have been used as a form primitive turntable. I formed two clay vessels in two different sizes on a rectangular plaited mat by hand (Fig. 7). When the experimentally-produced clay vessels were removed from the rectangular plaited mat after being shaped on it, each time I observed protrusions left on the edges of the base. I smoothed these protrusions by pressing the clay inwards with my fingers (Fig. 8). This was apparently the process adopted by the prehistoric potters of Gülpınar. The finished appearance of the bases of the experimentallyproduced pots is almost identical to those observed on numerous archaeological examples from Chalcolithic Gülpınar (Fig. 9-10). Most of the archaeological examples bear deep impressions of plaited mats, which could only have resulted from the use of mats. This means that freshly formed pots were not left on the mats to dry before firing. Thus, the results of my replication studies show that pots were occasionally built upon rectangular or circular mats. The fact that nearly 100 examples of pot bases with negative impressions of mats constitute only 5 % of the total base assemblage among the pottery repertoire clearly proves that this was not the prevailing method at the site. The disappearance of mat impression on pot bases in the middle stages of the Early Bronze Age might have been a result of the introduction of the potter’s wheel in this period. Füsun Ertuğ has presented a detailed ethnobotanical account of the social significance of plaited crafts in Turkey (Ertuğ 2006a). She has also provided us with information about the processes that are related to plaited crafts, taking into consideration such issues as weaving techniques, tools used, plants preferred, centers of production, craftspeople and their work 77
organization, learning patterns, and use of plaited crafts. In her recent article, Ertuğ also examined the contribution of ethnobotanical investigations to archaeology (Ertuğ 2006b: 13). It is unfortunate that no evidence to prove the potters’ use of rectangular plaited and circular coiled mats as primitive turntables have so far been attested in the ethnographic record of Anatolia. Examination of present day mat makers and their products clearly helps us to better interpret the mat impression on the pot bases found in archaeological excavations. However, there is some comparable evidence about the use of woven mats in pot-making in the ethnograhic record of Africa, where modern potters form their vessels by hand by putting clay onto woven mats. There are also cases in which the African potters make their pots in woven–mat covered shallow depressions opened in the ground (e.g., Nicholson 1929; Culwick 1935). The African potters appear to have used woven-mat bases in order to keep the clay from contact with the ground during pot-making. My experimental study also shows that such was also the case at Chalcolithic Gülpınar in northwestern Anatolia. This is to say that the potters of Gülpınar occasionally formed their vessels either on rectangular plaited mats or on circular coiled mats to prevent clay sticking to the ground during the first steps of vessel forming. In the absence of potters’ wheel in this period, this method might have enabled potters to shape their vessels in a more efficient way than the one that did not employ mat bases. Acknowledgements I thank to Coşkun Özgünel and Turan Takaoğlu for allowing me to study the the Chalcolithic pot bases with mat imptrssions from the Gülpınar excavations. I am also thankful to Ergun Arda and Meral Başaran from Çanakkale Onsekiz Mart University for their helps and suggestions during my experimental studies. Osman Çapalov also deserves special thanks for introducing me to a family of traditional mat and basket makers in the Biga town of Çanakkale Province.
78
Bi bl io g r aph y Aslan, R. 1997 Troia und Seine Siedlungskammer: Vom Neolithikum bis zu den Anfangen des Historischen Nationalparks. Tübingen University, Unpublished Masters Thesis. Barber, E.J.W. 1991 Prehistoric Textiles: The Development of Cloth in the Neolithic and Bronze Ages with Special Reference to the Aegean, Princeton: Princeton University Press Carrington Smith J. 1977 “Cloth and Mat Impressions.” In Keos I. Kephala: A Late Neolithic Settlement and Cemetry, Volume I, edited by J. Coleman. Princeton: Princeton University Press, pp. 114-125. Crowfoot, G.M. 1934 “The Mat Looms of Huleh, Palestine.” Palestine Exploration Quarterly, pp. 195-198. 1938
“Mat Impressions on Pot Bases.” Annals of Archaeology and Anthropology, University of Liverpool 25: 3-11.
Culwick, G.M. 1935 “Pottery among the Wabena of Ulanga.” Man 35: 165-169. Edgar, C.C. 1904 “The Pottery.” In Excavations at Phylakopi in Melos, edited by T. D. Atkinson. London: Society for the Promotion of Hellenic Studies, Supplementary Paper 4, pp. 80-176. Ertuğ, F. 2006a “An Overview of the plaited Crafts of Turkey.” In Proceedings of the fourth International Congress of Ethnobotany (ICEB 2005), Ethnobotany: At the Junction of the Continents and the Disciplines, (Anatolia and Thrace) pp. 297-306. 2006b
“Ethnobotanica l Investigations in Rura l A natolia.” In Ethno archaeological Investigatiıons in Rural Anatolia, Volume 3, edited by T. Takaoğlu, Istanbul: Ege Yayınevi, pp. 7-22.
Evans, J.D. and C. Renfrew 1967 Excavations at Saliagos near Antiparos. London: The British School of Archaeology at Athens. Gabriel, U. 2006 “Ein Blick zurück-Das fünfte Jahrtausend vor Christus in der Troas.” Troia: Archäologie eines Siedlungshügels und seiner Landschaft, edited by M. Korfmann, Mainz: R. Habelt, pp. 355-360. Johnston, R. H. 1974 “The Biblical Potter.” The Biblical Archaeologist 37: 86-106. Lucas, A. 1962 Ancient Egyptian Materials and Industries. London. E. Arnold.
79
Myres, J.L. 1898 “Negative Impressions of an Early Clay Vessel from A morgos.” The Journal of the Anthropological Institute of Great Britain and Ireland 27. 178-180. Nicholson, W.E. 1929 “The Potters of Sokoto.” Man 31: 187-190. Takaoğlu, T. 2006 “The Late Neolithic in the Eastern Aegean: Excavations at Gülpınar in the Troad.” Hesperia 75: 289-315. Wendrich, W. 1999 The World According to Basketry: An Ethnoarchaeological Interpretation of Basketry Production in Egypt. Leiden: CNWS. 2006
“Çatalhöyük Basketry.” Changing Materialities at Çatalhöyük: Reports from the 1995-1999 Seasons. Çatalhöyük Research Project, Vol. 5. edited by I. Hodder, pp. 333-338.
2007
“Neolithische Korbflechterei.” In Vor 12.000 Jahren in Anatolien. Die altesten Monumente der Menschheit, Karlsruhe, pp. 230-235.
Wendrich W.Z. and G. van der Kooij, eds. 2002 Moving Matters. Proceedings of the Seminar on Ethno-archaeology in the Near East. Leiden: CNWS.
80
Fig. 1. Ethnographic examples of mat from the Çanakkale region It represents a twill plaiting methods (“two over, two under”).
Fig. 2. Ethnographic example of straw coiled matting, from the Çanakkale region. It is likely that pots were occasionally shaped on such circular mats in prehistoric times (Diamater: 15 cm).
81
a
b
Fig. 3. Bases of a Chalcolithic bowls from Gülpınar bearing negative impression of mats. a) plain-weave matting (Diamater at Base: 12 cm) (b) twill-plaiting matting (Diameter at base: 11 cm)
Fig. 4. Negative impressions of straw matting in the base of a large cooking pot from Chalcolithic Gülpınar (Diameter at Base: 15 cm). The matting would have been the circular mat on which the pot was formed.
82
Fig. 5. One of the few remaining mat makers in the Çanakkale region. Here she uses twill-plaiting method to make a rectangular mat.
Fig. 6. Mat maker flattens the newly-woven mat with a wooden hammer.
83
Fig. 7. Clay vessel is being shaped on a rectangular plaited mat by hand.
Fig. 8. Another experimentally produced clay vessel removed from the rectangular plaited mat after being shaped on it. The sharp edges are pressed by fingers to smooth the edges of the base.
84
Fig. 9. Base of a hand made Chalcolithic bowl from Gülpınar. Note that the protrusions left on the edges of the base were subsequently pressed inwards with fingers.
Fig. 10. Base of a hand-made Chalcolithic jar from Gülpınar. Note that the protrusions left on the edges of the base were subsequently pressed inwards with fingers.
85
Moder n Site Su r ve y s i n Et h noa r ch aeolog y: T he C a se of Nor t h-Wes ter n A n atol i a Tur an Ta k a o ğ lu
Ethnographic survey of modern sites has recently become an established technique of archaeological research. The main purpose of modern site surveys is to collect a body of data that could form the basis of hypotheses to explain the functions of archaeological sites (Murray and Kardulias 1986; 2000). The beginning of modern site surveys in ethnoarchaeology particularly owe much to the Stanford University Archaeological and Environmental Survey conducted in 1982 in the southern Argolid, Greece. This project highlighted the importance of studying modern structures and their associated features as if they were archaeological sites. In these modern site surveys conducted with archaeological questions in mind, the material remains of both those presently in use and those structures falling into disuse such as field houses, farmsteads, storehouses, animal folds, garbage dumps, threshing floors, mills, wells, and kilns are recorded in relation to location, size, and material content. This research design helped to reconstruct the recent culture history of the study area and to hypothesize the ancient site functions that existed in the same district. Prior to the initiation of this project by the Stanford University, there 87
were other important ethnoarchaeological studies undertaken in the Argolid which underlined the potential significance of field studies relating the interests of ethnographic data to those of archaeology, as the Argolid region of Greece has already been the focus of archaeological field studies since the 1950s by teams from the University of Pennsylvania and Indiana University (Jacobsen 1985). As parts of these field studies, for example, modern pastoral sites were examined in relation to their architecture, size, locations, and discard and storage behaviors. Such studies of traditional pastoralist activities helped to formulate hypotheses about the formation of past pastoral sites and the recognition of seasonal schedule of human activities in the archaeological record (e.g., Chang 1981 and 2000; Chang and Tourtellotte 1993; Murray and Chang 1981). These pioneering studies of recent sites undertaken as part of the Argolid Exploration Project clearly have played important roles in the theoretical development of ethnoarchaeological research. Modern site surveys continued to be undertaken in various parts of Greece following various ethnoarchaeological projects undertaken in the Argolid. There are several ethnoarchaeological studies dealing with the study of modern rural sites and patterns of land-use on the Aegean islands such as Kea and Mykonos (Whitelaw 1991; Sampson 2002). In eastern Crete, modern field houses and other rural agricultural installations were also examined as part of the Vrokastro Survey Project (Hayden et al. 1992; Brümfield 2004). These ethnoarchaeological surveys examined the physical remains of modern sites and their associated features in order to reconstruct the patterns of agricultural exploitation and rural sites. One of the most important features of these ethnoarchaeological researches is that those specifically trained ethnographers and cultural anthropologists conducting modern site surveys often opted to benefit from the historical sources of the Ottoman Empire, as well as travelers’ accounts and oral history. The modern site survey conducted as part of the Vrokastro Survey Project placed special attention upon the historical sources and 88
material remains of the Ottoman past (Brumfiel 2000). The Ottoman archival sources in certain cases served as the basis of arguments in establishing the culture history of a given region in relation to agricultural or economic activities. Such studies show that Ottoman written sources could be a great use for ethnoarchaeologists working in Greece and elsewhere. As a result, those studies concentrated on the Ottoman activities therefore become very important sources for a proper understanding of the rural life in the Aegean and Mediterranean (e.g., Davis 1991; Davis et al. 2005). In Turkey, the importance and necessity of ethnographic survey of modern sites with archaeological questions in mind has recently been recognized. Based on the methodology used in the projects undertaken in the Argolid, a modern modern site survey has been undertaken on the island of Bozcaada (Tenedos). Modern agrarian sites such as field houses and animal folds belonging to the recent past have been examined for the purpose of exploring some relationships between the landuse, agricultural potential, and settlement size and distribution (Takaoğlu and Bamyacı 2005). This methodology helped to a great extent to visualize the ways how the countryside of the island might have exploited in ancient times. In the case of Bozcaada, the Ottoman sources confirmed most what we have learnt from ethnoarchaeological survey. The ethnographic survey of both recently abandoned and still presently used modern sites, combined with the information gathered from the Ottoman written documents related to the human activities of the past five hundred years, helped to the reconstruction of the agricultural history of Bozcaada particularly in relation to activities such as grape growing and stock-raising. This is because the Ottoman archival sources are rich in information about the demographic patterns and the number of rural sites and farms used in grape growing and stock-raising. The pattern of agricultural exploitation and land-use obtained from the combination of ethnographic survey of modern sites and Ottoman archives proved to be very useful in reconstructing 89
the patterns of land-use dating to the ancient times. This ethnoarchaeological research undertaken on the island of Bozcaada demonstrates that it might be useful for those archaeologists working in Turkey to acquaint themselves with material remains and historical records from the recent Ottoman past and the early decades of the modern Turkish Republic. In the case of Bozcaada, the Ottoman archives served as important ethnographic data in reconstructing the recent economic and cultural history of the island. Similar to the island of Bozcaada, modern site survey can be an important research technique in northwestern Anatolia, where the rough environment facing the Aegean Sea provided constraints for the populations of the region. Olive and grape growing have been the two main farming activities in coastal zones of this region, while animal herding and secondary products obtained from them chiefly formed the basis of the economies of populations living in higher elevations. Although rough coastal zones of northwestern Anatolia do not provide optimal conditions for societies to develop because of harsh environment and dense vegetation cover, this region saw intensive habitation since the appearance of first village settlement around 6000 B.C. Because the soil of the rough environments is not very suitable for arable farming and the land is mainly covered with maquis vegetation, the questions of how the past societies exploited the rural lands for agrarian purposes and how they adapted themselves to the coastal rugged environments can be elucidated through modern site surveys. For example, modern patterns of shepherding and other animal-related activity areas have recently been examined in northwestern Anatolia to formulate hypothesis about the role of animal related activities among the Chalcolithic and Early Bronze Age coastal village societies of the region (Takaoğlu 2006). In rural parts of northwestern Anatolia, modern site surveys documenting entire ethnographic evidence related to rural land-use and agrarian economy can help to form an explanatory model. Ethnographic evidence related to traditional
90
economic activities includes structures such as grape and olive processing workshops (Figs. 1-2), field houses (e.g., Fig. 3), animal dams (e.g. Figs. 4-5), bread ovens (e.g. Fig. 6), threshing floors, windmills, and water mills. This information source characterized by such traditional structures reflecting the past life ways of the village societies of the region is gradually disappearing from the ethnographic record of the region. Thus, any ethnographic evidence and architectural evidence related to agrarian land use and rural economic activities must systematically be documented as early as possible. This could be done by documenting all available ethnographic evidence and place them accurately in a database of Geographic Information System, which in turn can be used to establish a feedback relationship between ethnographic evidence (e.g., architectural structures, workshops, activity areas) and geographic and geologic factors (e.g., soil quality, vegetation cover, land-use capability). A model derived from the analysis of ethnographic evidence on the traditional patterns of rural land-use and agrarian economy in relation to these geologic and geographic factors may help us to reconstruct the ways how past populations of the region made their livings in similar environments and how they exploited the lands surrounding their settlements for economic purposes. The modern site survey technique employed in the Argolid of Greece can well be applied to northwestern Anatolia as the natural environments are very similar in both study areas. The archival sources from the Ottoman past and early decades of the Turkish Republic are very useful in better interpreting the function of modern sites. Because the decades following the First World War witnessed dramatic change in the patterns of agriculture in northwestern Anatolia, any attempt involved modern site survey must benefit from these Ottoman archival sources. This must also be supported through oral history, as well as the accounts of European travelers. Yearly increasing archaeological studies show one more time how important it could be to conduct ethnoarchaeological research in northwestern Anatolia.
91
R efer enc e s Brumfiel, A. 2000 “Agriculture and Rural Settlement in Ottoman Crete, 1669-1898. A Modern Site Survey.” In A Historical Archaeology of the Ottoman Empire: Breaking New Ground (Contributions to Global Historical Archaeology), edited by U. Baram and L. Carroll. New York: Plenum Publishers, pp. 37-78. Chang, C. 1981 The Archaeology of Contemporary Herding Sites in Greece. State University of New York, Unpublished Dissertation. 2000
“The material culture and settlement history of agro-pastoralism in the Koinotis of Dhidhima: an ethnoarchaeological perspective.” In Contingent Countryside: Settlement, Economy, and Land Use in the Southern Argolid since 1700, edited by S. B.Sutton. Stanford, Calif.: Stanford University Press, pp. 125-140.
Chang, C. and P.A. Tourtellotte 1993 “Ethnoarchaeological survey of pastoral transhumance sites in the Grevena Region, Greece.” Journal of Field Archaeology 20: 249-264. Davis, J. 1991
“Contributions to a Mediterranean Rural Arcaheology: Historical Case Studies from the Ottoman Cyclades.” Journal of Mediterranean Archaeo logy 4: 131-216.
Davis, J., F. Zarinebaff, J. Bennett 2005 An Historical and Economic Geography of Ottoman Greece: The Southwestern Morea in the Early 18th Century. American School of Classical Studies at Athens, Hesperia Supplement 34. Hayden, B.J., J.A. Moody, O. Rackham 1992 “The Vrokastro Sur vey Project 1896-89; Research Design and Preliminary Results.” Hesperia 61: 293-353. Jameson, M.H. 1985 “A nother modest proposa l: eth noarchaeolog y in Greece.” In Contributions to Aegean Archaeology: Studies in Honor of William A. McDonald, edited by N.C. Wilkie and W.D.E. Coulson. University of Minnesota Publications in Ancient Studies I. Dubuque, Iowa: Kendal/ Hunt Pub. Co., pp. 91-107. Jameson, M.H., C.N. Runnels, and Tj. H. van Andel 1994 A Greek Countryside: the Southern Argolid from Prehistory to the Present Day. Stanford, Calif.: Stanford University Press. Murray, P. and C. Chang 1981 “An ethnoarchaeological study of contemporary herder’s site.” Journal of Field Archaeology 8: 372-381. Murray, P. and P.N. Kardulias 1986 “A modern-site survey in the Southern Argolid, Greece.” Journal of Field Archaeology 13: 21-41.
92
2000
“The present as past: an ethnoarchaeological study of modern sites in the Pikrodhafni Valley.” In Contingent Countryside: Settlement, Economy, and Land Use in the Southern Argolid since 1700, edited by S. B. Sutton. Stanford, Calif.: Stanford University Press, pp. 141-168.
Renfrew, C. and J. M. Wagstaff, (eds.) 1982 An Island Polity: The Archaeology of Exploitation in Melos. Cambridge: Cambridge University Press. Sampson, A. 2002 The Neolithic Settlement in Ftelia, Mykonos. Rhodes: University of Aegean. Takaoğlu, T. and A. O. Bamyacı 2005 “Continuity and Change in Rural Land Use on Tenedos/Bozcaada.” In Ethnoarchaeological Investigations in Rural Anatolia, vol. 2, edited by T. Takaoğlu, Istanbul: Ege Yayınevi, pp. 113-136. Takaoğlu, T. 2006 “Patterns of Dairying in Coastal Northwestern Anatolia.” In Ethno archaeological Investigations in Rural Anatolia, vol. 3, edited by T. Takaoğlu, Istanbul: Ege Yayınevi, pp. 23-44. Whitelaw, T.M. 1991 “The ethnoarchaeology of recent rural settlements and land use in northwest Kea.” In Landscape Archaeology as Long-Term history: Northern Keos in the Cycladic Islands from Earliest Settlement to Modern Times, edited by J.F. Cherry, J.L. Davis, and E. Mantzourani. Los Angeles: UCLA Institute of Archaeology, pp. 403-454.
93
Fig. 1. An abandoned stone-built olive mill workshop at Küçükkuyu. Such structures have been a characteristic feature of the rural landscape of coastal northwestern Anatolia.
Fig. 2. An abandoned olive crusher once operated by animal power. Such primitive installations are commonly to be found throughout rural northwestern Anatolia.
94
Fig. 3. A small farming complex surrounded with a stone-built fence topped with bushes. Such field houses with attached features such as ovens and wells are very common in the barren landscapes of northwestern Anatolia.
Fig. 4. A typical stone-built animal dam used to keep sheep or goats in rural northwestern Anatolia.
95
Fig. 5. A view of an animal fold built of tree branches and bushes. The study of such agriculture related features built of durable materials can provide new looks for ethnoarchaeological research on the patterns of site location and duration.
Fig. 6. An abandoned stone-built bread oven characteristic of most villages in western Anatolia.
96
Fig. 7. A nineteenth century drawing of illustrating a water-mill in the Çanakkale region. Such images found in the accounts of travelers represent important source for ethnoarchaeologits.
Fig. 8. A photograph of bread makers in the Çanakkale region, dating to nearly a century ago. Such illustrations are useful to interpret abandoned architectural features belonging to the recent past.
97
Contributors BRADLEY J. PARKER University of Utah, Department of History Carlson Hall 380 S 1400 East Rm 00211 Salt Lake City - Utah 84112 e-mail:
[email protected] ZEFER DERİN Ege Üniversitesi Edebiyat Fakültesi, Arkeoloji Bölümü Bornova-İzmir e-mail:
[email protected] DAVUT KAPLAN Ankara Üniversitesi Dil ve Tarih-Coğrafya Fakültesi, Arkeoloji Bölümü 06100 – Sıhhıye/Ankara e-mail:
[email protected] TURAN TAKAOĞLU Çanakkale Onsekizmart Üniversitesi Fen/Edebiyat Fakültesi, Arkeoloji Bölümü 17100 - Çanakkale e-mail:
[email protected] ABDÜLKADİR ÖZDEMİR Çanakkale Onsekizmart Üniversitesi Sosyal Bilimler Enstitüsü, Arkeoloji Bölümü 17100 - Çanakkale e-mail:
[email protected] M. BARIŞ UZEL Ege Üniversitesi Edebiyat Fakültesi, Arkeoloji Bölümü Bornova-İzmir e-mail:
[email protected] 99