Along the northern edge of he Mesopotamian lowlands ies a piedmont zone consisting of rolling plains with grassy steppe vegetation and occasional stands of oak. This area, lying between the mountains of the Anatolian Plateau and the arid steppeto the south, is sometimes called the “hilly flanks of the Fertile Crescent.” It provides excellent farmland without the need for irrigation, as well as good pasture for domestic animals such as sheep, goats, and cattle. Near the eastern end of this zone, the Euphrates River cuts through chalky hills forming a valley 8 to 12 kilometers wide as it flows from its source in the Taurus Mountains in the north toward the Mesopotamian plain (Fig. 1).
During the past twenty years, a series of darns have been built across the Euphrates in Turkey, Syria, and Iraq. At present, the Turkish government is constructing two darns in Adiyaman and Malatya provinces that will provide hydroelectric power, as well as water for a major irrigation scheme. The area within the projected reservoir includes rich alluvial plains that were important population centers during the past, and contain hundreds of archaeological sites. Because so little was known about the archaeology of the region, an emergency salvage program was begun in 1977, involving scholars and institutions from the United States as well as Europe and Turkey. As part of this program, excavations at the site of Gritille were carried out for four seasons (1981-1984) under the direction of Richard S. Ellis of Bryn Mawr College.
Gritille is a small mound located on the west bank of the Euphrates (Fig. 2). It rises 13 meters above virgin soil, and is composed of the remains of settlements dating from ca. 7000 B.C. to the 13th century A. D. Excavations have focused on the two periods best preserved at the site: a Medieval fortress occupied by Christians that covers the top of the mound, and an early fanning settlement at its base. This article will report on the early farming or Neolithic settlement which has been tentatively dated from 7000 to 5500 B.C. (pending the analysis of a large series of time had moved into the region from the south. Only two other Neolithic sites were known within the salvage area (Hayaz Höyük and Cafer Höyük). Gritille provided an unusual opportunity for the investigation of the initial occupation of this segment of the Euphrates by farmers and herders, and the evolution of their way of life as they adapted to a new environment.
Site Location and Resources
The site lies in close proximity to resources that were important to the first settlers. At the base of the mound, emerging from a conglomerate bed (rock composed of rounded pebbles and cobbles cemented together), are springs that yield a steady stream of cool fresh drinking water even in the hottest and driest months of a drought year. The surrounding land is nearly flat and easily cultivated. In the hills 3 to 4 kilometers away are bands of good quality nodular flint, a necessary raw material for the manufacture of chipped stone tools. These nodules occur in chalk and limestone beds which provided a source of easily carved stone. Cobbles of harder rock were to be found in nearby wadi beds.
Both the hills and the plain were used by herds of large game animals, and provided good grazing for domesticated animals during the wet season in winter and spring. In the summer, when the temperature reaches 50° C in the shade and the natural vegetation of the plain is seared and dry, large flat islands in the middle of the river provide lush pastures. Although the Euphrates here is a broad swift river not easily crossed, modern villagers and their animals reach the islands by swimming, and float to the eastern bank of the river on rafts made up of inner tubes and/or scraps of wood. This crossing is of minor importance today, but must have been significant in Early Bronze Age and Medieval times when it was controlled by Gritille on the west and the large settlement at Lidar on the east. It may also have been used in Neolithic times as a means of traveling to related settlements located to the east and north (see below). Finally, growing along the river banks are trees, a valuable commodity for construction and fuel.
Excavation of the Neolithic Settlement
In order to limit the volume of earth that had to be removed by pick, shovel, and bucket to reach the Neolithic settlement, investigation of this period was carried out on the eastern edge of the mound where at least part of the later settlements (as well as some of the Neolithic village) had slid down into the encroaching river (Figs. 4, 5). This location had two other major advantages. First, excavated deposits could easily be disposed of down the slope at the edge of the excavation units. Second, buckets of soil could be carried down to the river hank and washed in screens, radiocarbon samples).
The Neolithic deposit at Gritille was discovered in 1981 in a 3-meterwide trench (Operation I) cut into the eroding river edge in order to document the archaeological sequence (Voigt and Ellis 1981). It lies under approximately 9 meters of later occupation debris. Given labor costs and constraints on archaeological budgets, this deposit would not be chosen for investigation under ordinary conditions of research. We were working, however, in an area that will he flooded by the new darn within this decade. From the small sample of artifacts recovered in Operation I, we knew that there were at least 4 meters of deposit dating to the Neolithic period, and that the people who occupied the site at this ensuring the recovery of small or fragile pieces of stone and bone. Twenty percent of the deposit was treated in this way. The pump that supplied water to the screens also powered our flotation device, a modified oil drum fitted with a shower head which washed carbonized plant remains from their matrix and deposited them in a piece of muslin where they could be dried and preserved for study (Fig. 3).
As a result of two seasons’ work, the entire Neolithic sequence to virgin soil has been documented, although only a small sample was obtained for the earliest/deepest part of the deposit. The maximum contiguous area cleared is about 210 square meters, located at the southern end of the excavated strip. During the latest major phase of the occupation, this portion of the village contained three relatively well preserved buildings constructed of mud brick. Building 2 was rectangular in plan, with the interior divided into a series of small rooms Figs. 6, 7). Immediately adjacent to the south was a building with less substantial exterior walls, larger rooms, and a less regular plan (Building 1). To the north was a second well-built rectangular structure (Building 3) of which only a small section lay within the excavation area and which was disturbed by later pits. A small fragment of wall within the interior of Building 3 suggests that it may have been similar in plan to Building 2. A crumbling brick wall leading south from Building 3 enclosed an unroofed area which probably served as a courtyard for Building 2. The deposit inside Buildings 1 and 3 consisted of decayed mud brick, with virtually no occupation debris on the floors which were simply hard-packed surfaces. On the floor of Building 2, however, was a large collection of ground and chipped stone artifacts (see Figs. 16a, 20). Although there were no hearths or other features associated with Buildings 2 and 3, it seems most likely that they were used as houses. The difference in plan and construction of Building 1 suggests that it served some other function.
To the east of this group of buildings was an area with no architectural remains (see Fig. 5). The de posit here consisted of thin layers of silt (eroded building material) and trash which sloped down to the east. Based on the available information we are not able to state whether this represents an open space within the village or an area on the village edge. This open zone provided most of the artifacts, animal bones, and seeds recovered at Gritille, as well as direct evidence for a variety of activities. In some cases, bonfires seem to have been built on the surface, leaving patches of dark ash and occasionally chunks of charcoal suitable for radiocarbon dating. More complex features include carefully shaped oval or round pits filled with rounded cobbles and ash that were probably used for cooking (Fig. Sac). Since pottery was absent or very rare at this time (the few sherds recovered came from the latest part of the deposit), containers of skin or basketry must have been placed on rocks that had been previously heated, a method documented ethnographically for North American Indian groups.
A group of shallow, irregularly shaped pits filled with ash, angular cobbles, and large quantities of chipped stone tools and chipping debris were also found. Although these may have been simple trash disposal features, there is some evidence to suggest that they were used as part of a more significant activity, the heat treatment of flint. Approximately one-third of the chipped stone artifacts (both tools and waste or chipping debris) have a slight luster on the surface. Samples of raw materials with a similar color and texture obtained from flint sources in the nearby hills did not have this luster, but acquired it when experimentally heated. Thus a comparison of artifacts and raw materials suggests that the Neolithic people were heating flint to improve its flaking qualities. If the irregularly shaped pits were used for heat treating raw materials, the presence of angular (heat split?) non-flint cobbles may be related to the regulation of temperatures, since nodules of flint placed directly in a hot fire simply explode, destroying their utility for controlled flaking.
Although an analysis of the excavated material and a division of the stratigraphic column into chronological units has only just begun, a few preliminary statements about change within the Neolithic architectural sequence can be made. Throughout most of the occupation, techniques of building construction were similar to those of the latest phase. Mud brick walls one or (more frequently) two bricks thick rested directly on the surface or on a foundation made up of scattered rocks, usually chunks of the conglomerate that runs beneath the site. Buildings were oriented with their axes along the cardinal directions. There was a strong tendency for walls to be built one above the other in succeeding phases, so that in section they appear as slightly offset stacks of bricks. The earliest buildings differ primarily in the use of white lime plaster on walls and for floors; in one case, traces of red coloring were found on a plaster floor. An open space containing pits and evidence of activities using fire was also present in all phases, although it was located farther to the north in the earliest part of the occupation. One kind of feature found only in the middle of the sequence was a deep cylindrical pit dug into the clean clay debris of abandoned structures. These pits were of an appropriate size and shape for food storage; the presence of numerous rodent burrows in the surrounding deposit tends to support such an inference, while testifying to the inadequate protection provided by such storage facilities!
Plants, Animals, and People
The study of plant and animal remains from Gritille has begun at The University Museum’s Applied Science Center for Archaeology (MASCA). Dr. Pamela Crabtree reports that within the sample of seeds recovered in 1983 from the latter part of the sequence, the most common cultivated plants were emmer wheat, lentils, and bitter vetch; barley and wild plants were present only in very small quantities. Analysis of the animal bones has been undertaken by Gil Stein, a graduate student in Anthropology at the University of Pennsylvania. Based on his analysis of a portion of the sample (over 2600 bones recovered in 1983), the Neolithic settlers at Gritille relied primarily on domestic animals for food, with sheep and goats comprising over 40 percent of the sample; pigs and cattle were less common (9 and 2% respectively). The age at which sheep and goats were slaughtered, with a high percentage killed before the age of two, is consistent with the use of these animals primarily for their meat rather than for dairy products. One other domestic animal was present, the dog.
Bones of wild animals were rare, but included large herd animals such as aurochs (wild cattle), wild goats, red deer, roe deer, fallow deer, and gazelle, as well as hare. Although the small number of game animals in the sample suggests that hunting did not contribute greatly to the diet, the presence of many projectile points (arrow and perhaps spear heads) at the site suggests that this activity was an important one (see Fig. 17). Wall paintings at the contemporary village of Catal Hüyük on the Anatolian plateau (Mellaart 1967) show men with bows and arrows pursuing deer and aurochs, providing direct evidence of the value placed on hunting by early farmers and herders.
There is little information on the health or other characteristics of the human population since only two burials were found. One contained the skeleton of a small child placed in an oval pit cut into virgin soil; this burial was exposed while clearing a path to the river, and was poorly preserved due to weathering as well as to conditions of discovery. The second individual was represented only by the top of his or her skull which had been placed in a pit within the settlement.
Use of Clay
In many of the stone-filled cooking pits were found lightly fired fragments of clay figurines. These have well-preserved surfaces but were usually broken in antiquity, perhaps as a result of throwing raw clay models into the fire. The animals can be easily recognized, and show quadrupeds with sagging bellies (Figs. 10e, 11). Most seem to be cattle with long upward curving horns, hut at least one head has the curled horns of a ram.
The human figures are less conventional. The majority seem to be seated individuals with legs extended or crossed. Some are highly schematic triangles, and can be recognized as humans only within the context of the entire group of clay figures. One of the most intriguing seated figures has plump legs, a snakelike torso, and a flat face with two eyes impressed with a sharp tool (Fig. 12a,b). Quickly nicknamed “E.T.,” this small person has markings on its chest and legs left where some attachment (perhaps a child?) has broken away. A rare standing figure has stublike arms projecting slightly behind the torso which has a bulging stomach made of a separate lump of clay (Fig. 13a,b). At the time of recovery the stomach had broken away, revealing an impressed naval beneath
matching that on the outer surface of the attached piece. Tiny tab-like feet project from beneath a skirt which was once decorated with pinprick impressions, and broken fragments of clay on the shoulders may indicate the presence of a necklace or perhaps hair. The latter hypothesis is supported by another fragment, a pinched, lizardlike head which had a coil of hair wrapped around its pointed skull. Two human representations were carved of chalk or soft limestone (Fig. 101; g). A flat-faced head was pierced at its finished neck edge, perhaps for attachment to a body made of another material such as wood. A seated female with massive legs and globular feet was broken at the waist and had chipped areas on its feet (Figs. 10g, 14).
Studies of the use of small human and animal figures among historic and modern groups in the Near East suggest that the clay figurines found in burnt cobble pits were used in rituals intended to protect people from unwanted situations or events (for example, poor health, hot winds), or to obtain a desirable result (pregnancy, good harvests). The chalk figures may represent deities or figures used in teaching young people the values of the community (Ucko 1968; Voigt 1983). Rare clay figures of animals found in domestic contexts at Gritille have poorly preserved surfaces, with heavily worn feet and heads. These were probably children’s toys.
Small geometric pieces made of lightly fired clay were found in the burnt cobble pits, as well as in the trash and silt layers of the open area (Fig. 15). lei one case, at least 14 sun-dried day spheres had been deposited in a cluster. Such artifacts have been interpreted as “tokens,” evidence for a recording system that preceded the invention of writing (Schmandt-Besserat 1978). Several clay cylinders were found with designs on one flat end, presumably for stamping designs on a surface (clay, leather, fabric, human skin?). One stamp has a spiral design which seems to represent a snake with a diamond-shaped head. Fired clay containers seem to have come into use only at the very end of the sequence, and were very rare. These pottery vessels were burnished on the outside surface and fired to a dark gray.
The Chipped Stone Industry
The most common artifacts recovered were pieces of chipped stone, including retouched or utilized flakes and blades (tools), cores and core fragments, unused flakes and blades, and small chips (waste). A study of this material is being conducted by Dr. Richard Davis of Bryn Mawr College, with the assistance of Linda Kurtz (U. of Michigan) and Tineke Van Zandt (U. of Texas at Austin). At present approximately 18,000 of the estimated 80,000 pieces recovered have been described. The predominant raw material is good quality brown or cream colored flint from local sources; however, obsidian, a natural volcanic glass that had to be imported from sources on the Anatolian Plateau, is also relatively common. The Gritille obsidian is transparent and greenish black in color, attributes characteristic of obsidian sources around Lake Van in eastern Turkey.
The presence of distinctive types of flint debris which result from the process of blade core preparation, as well as numerous flakes with areas of cortex or the outer surface of intact flint nodules, indicates that flint was brought to the village in the form of nuclides, and that all phases of the tool production process took place at the site. Obsidian on the other hand occurs in the form of tools made from large, thick blades, used or retouched microblades (blades less than 1 centimeter in width), rare waste flakes, and several fragments of microblade cores. Such evidence strongly suggests that core preparation and the production of blanks or even tools took place at or near the obsidian source, and that only the final states of tool manufacture (retouching, micro-blade production and retouch) took place at ancient Gritille. The toolkit produced included projectile points, sickle blades, knives, scrapers, piercing tools (borers, reamers), pieces with notched or denticulate edges, choppers, and tools with strong bevelled edges (burins) (Figs. 16-18). Such tools would have been used in hunting, harvesting cereals, the butchering of wild and domestic animals, the processing of animal skins to make rawhide or leather, woodworking, and the manufacture of tools and ornaments from hone and stone.
In addition to the information that the chipped stone industry provides on the technology and economy of the settlement, it also documents contacts with other regions. Gritille (as well as other nearby villages such as Hayaz Hüyük) lies at the northern end of a chain of sites (including Abu Hureyra and Mureybit) that represents a progressive movement northward along the Euphrates by groups sharing a tradition of agriculture and material cul ture that apparently had its origins in the Levant (Israel, Jordan, and Syria) in the 9th millennium B.C. (Fig. 19). By 7000 B.C. such groups (known in archaeological jargon as Pre-pottery Neolithic B and PPNB-related sites) extended in an arc from south of the Dead Sea (Beidha) up into the Taurus mountains (Çayönü) and east to the Tigris (Maghzaliyah). The presence of a PPNB-related settlement at Maghzaliyah is particularly significant. Until this site was excavated in the late 1970s, many (though not all) scholars thought that the initial settlement of the northern Mesopotamian plain was accomplished by people moving clown out of the Zagros mountains to the east. In the Zagros a second tradition of farming and herding, utilizing highland plants and animals, was developed in the centuries after 9000 B.C. at sites such as Ganj Dareh and Tepe Guran (Smith 1971). Maghzaliyah with its ties to the PPNB tradition of the western lowlands shows that the process of agricultural colonization was initially restricted to zones that provided conditions similar to those of the areas in which each of these two major traditions originated.
The Search For New Materials
A final and surprising aspect of the village at Gritille is revealed by the ground stone industry. In addition to utilitarian pieces such as mortars, flat grinding stones, polished bowls, and numerous stone spheres ranging in size from marbles to cannon balls and celts (Fig. 20), this group of artifacts included pendants and beads made of a variety of brightly colored stones (Fig. 9). Some of these materials may have been imported from great distances: at least one bead appears to be made of turquoise which occurs in eastern Iran and Sinai (this identification has not been confirmed by a mineralogist); an orange pendant is of carnelian which occurs in the mountains of Iran.
An interest in blue and red minerals apparently led to experimentation with metal ores at Gritille as well as at other early Neolithic sites such as Çayönü where copper artifacts were found. In several of the burnt cobble pits from the latest levels at Gritille were found yellow or red lumps that have been identified as pieces of iron ore by Professor Robert Maddin of Harvard University. According to Maddin, not only were some of these pieces heated, but one shows evidence of deliberate working by people. The experiments with iron at Gritille were apparently a dead end; however, they provide direct evidence of an attitude toward the environment which is characteristic of these Neolithic people, whose experiments with plants and animals had resulted in an agricultural way of life within a few favorable zones, and allowed them to colonize new areas, including the Euphrates valley.
This evidence for continuing technological innovation challenges the old idea that the early villages were stable and even stagnant communities which did not change significantly over millennia. New archaeological evidence from sites dated between 7000 and 5000 B. C. indicates that during this period there were major developments within the economic and social organization of Neolithic settlements as well as in technology and architecture. We now know that the early villages must have been lively places, built and maintained by people who were innovators and explorers. Unfortunately, we will never be able to learn more about early Gritille. Within ten years the top of the site will be under an estimated 120 meters of water.
Messages from Bits of Stone
Within each of the major categories of chipped stone tools found at Gritille (e.g., “projectile point,” “borer”), there is a considerable variety of form. Some of this variation is undoubtedly related to the specific function for which a tool is used; for example, a scraper used for working wood will be heavier and will have sharper edges than one used to remove flesh and/or hair from animal hides being made into leather. Such differences will eventually be defined through a more detailed study of the characteristics of artifacts within each major category and through a study of wear traces on the tools which can be linked to specific activities.
In other cases, tools apparently used for the same purpose differ in minor ways; for example, a group of six complete tanged projectile points found inside Building 2 vary in tang size and shape, and in the proportions of the body of the point. This kind of “stylistic” variation may reflect individual or group preferences in shape or technology. When there is a consistent pattern in the distribution of specific stylistic variables in space within a single period of time they can be used to define groups that share a common tradition or technology (for example, households or multi-household units within a village, groups of villages within geographical regions). A consistent pattern in the variation of stylistic attributes through time is one of the building blocks used to establish the contemporaneity of wit as well asbetween regions.
Because of the limited size of the excavated area at Gritille and the depth of the sequence, the major emphasis in any study of artifact variation will necessarily fall on change through time. At this preliminary stage of our analysis only a few tentative comments can be made on the nature and meaning of observed changes, but these do suggest the direction of our research. For example, there are significant changes through time in the relative frequency of at least one major class of artifact: burins, and flakes removed in resharpening burins (spills), are very common in the uppermost levels of the deposit, but rare in earlier parts of the sequence (Fig. 17x,y,z,aa). Burins have been linked through ethnographic observation and experiment with the manufacture of bone artifacts such as needles or awls, and such tools together with bone beads were recovered from the site (Fig. 21). If the number of burins is indeed a direct reflection of’ changes in the activities conducted at the site (rather than a reflection of the limited area excavated and changes in the locales within the settlement in which burins were used), it should be possible to confirm this relationship through a study of the frequency of other types of debris associated with bone working, for example, pieces of cut bone and bone splinters.
If we compare the chipped stone industry with that from other sites, the potential of the Gritille se quence as a source of information on relationships between regions becomes apparent. There are strong similarities between the chipped stone industry at Gritille and that at sites lying along the Euphrates River to the south (Mureybit, Abu Hureyra), in the Syrian desert (El Kom), and in the Levant. Sites in the latter region belong to a well-known tradition originally defined at Jericho, and known as Pre-pottery Neolithic B or PPNB (Mellaart 1975; Redman 1978). To the north and west of Gritille, sites such as Ceder, Çayönü, and Maghzaliyah show some similarities to the PPNB chipped stone tradition, but also have distinctive characteristics including a heavy reliance on obsidian, and a highly diagnostic tool type made from a thick obsidian blade that has been steeply retouched along both edges and has striations along the long axis away from the edges (Fig. 17s). These tools occur at Gritille in the lower half of the deposit. Thus during the initial phases of the settlement its primary ties were with groups living in a similar steppe environment to the south, but it was also in contact with groups in the mountains who controlled access to resources such as obsidian. During the later phases of the settlement, when sites such as Çayönü had apparently been abandoned, similarities in architecture as well as chipped stone industries suggest that Gritille was most closely related to villages within northern Syria and Mesopotamia, including Ras Shamra, Abu Hureyra, and Bouqras.