368 Documenta Praehistorica XLIV (2017) An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) Jarrad W. Paul 1 , Burçin Erdog ˘u 2 1University of Melbourne, Melbourne, AU jwpaul17@gmail.com 2Trakya Üniversitesi, Edirne, TR burcinerdogu@trakya.edu.tr< berdogu@gmail.com ABSTRACT – Worked bone and antler tools were regularly used by prehistoric societies in northwest and western Anatolia to create and maintain everyday items. Ugurlu, one of the most important pre- historic sites in the north east Aegean, shows extensive evidence of bone and antler tool manufac- ture. This article examines the Ugurlu osseous assemblage from its inception during the Early Neo- lithic (6800 cal BC) to the middle Chalcolithic (4300 cal BC). A typology is established which labels the 534 items uncovered thus far, supported by contextual information. A comparison with other bone tool assemblages in the region is also presented, highlighting group similarities and type dif- ferences. The results show that few bone tools were found in the Early Neolithic Phase VI (6800– 6600 cal BC), while pointed tools dominate Phase V (6600–5900 cal BC). The established Neolithic Phase IV (5900–5600 cal BC) witnesses a dramatic growth in the number of bone tools produced, which is steadily adapted with the advent of the Chalcolithic Phase III (5500–4900 cal BC). During this transition between the Neolithic and Chalcolithic, certain tool types decline in number and man- ufacturing style changes. In the middle Chalcolithic Phase II (4500–4300 cal BC), bone objects seem to be crudely made, possibly reflecting the emergence of local traditions. IZVLE∞EK – Prazgodovinske skupnosti v severozahodni in zahodni Anatoliji so za izdelavo in popra- vilo vsakdanjih predmetov redno uporabljale obdelana ko∏≠ena in ro∫ena orodja. Na najdi∏≠u Ugurlu, enem najbolj pomembnih prazgodovinskih lokacij na severovzhodu Egejskega prostora, so odkrili ∏tevilne dokaze o izdelavi ko∏≠enih in ro∫enih orodij. V ≠lanku preiskujemo ko∏≠eni zbir iz Ugurla od nastanka najdi∏≠a v ≠asu zgodnjega neolitika (6800 pr. n. ∏t.) do srednjega halkolitika (4300 pr. n. ∏t.). Na podlagi zbira 534 tovrstnih predmetov, ki so bili do sedaj odkriti na najdi∏≠u, smo izdelali tipologijo, ki je podprta s podatki o kontekstih. Podatke smo nato primerjali z drugimi kostnimi zbiri v regiji, ki ka∫ejo tako na podobnosti kot na razlike v tipologiji. Rezultati ka∫ejo, da so v ≠asu zgodnje neolitske faze VI (6800–6600 pr. n. ∏t.) uporabljali malo ko∏≠enih orodij, medtem ko v fazi V (6600–5900 pr.n.∏t.) prevladujejo konice. V obdobju uveljavljenega neolitika v fazi IV (5900– 5600 pr.n.∏t.) se je ∏tevilo obdelanih ko∏≠enih orodij dramati≠no pove≠alo, ki se postopoma prilago- di v za≠etku halkolitika v fazi III (5500–4900 pr. n. ∏t.). Prisotnost dolo≠enih tipov orodij se zmanj- ∏a v ≠asu prehoda med neolitikom in halkolitikom, spremeni pa se tudi stil izdelave. V ≠asu srednje- ga halkolitika v fazi II (4500–4300 pr. n. ∏t.) so ko∏≠ena orodja grobo izdelana, kar morda ka∫e na pojav lokalnih tradicij pri izdelavi. KEY WORDS – Neolithic-Chalcolithic; bone tools; northwest Anatolia; typology KLJU∞NE BESEDE – neolitik-halkolitik; ko∏≠ena orodja; severozahodna Anatolija; tipologija Pregled zbirov obdelanih kosti in rogov na najdi[;u Ug ˘ urlu (Gökçeada, Tur;ija) DOI> 10.4312\dp.44.23 An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) 369 sites in Greece corresponds with the Anatolian Chal- colithic period). Peter Höglinger’s (1997.157–196) detailed report on the 272 stratified bone objects di- scovered at Karanovo is a notable addition to Bulga- rian worked bone tool studies. Site reports from Ya- balkovo and Kova≠evo also contain further informa- tion for this region (Leshtakov et al. 2007.185–234; Sidéra 2013.173–178; Lichardus Itten et al. 2006. 83–94). Positioned in the north Aegean off the northwest coast of Anatolia, Ugurlu, on the island of Gökçeada, is one of the earliest sites with evidence of farming and animal husbandry in the Aegean. Stratigraphic excavations have clarified the spatial extent of the settlement from the pre-pottery Neolithic occupation (c. 6800 cal BC) and brought to light evidence of the earliest examples of pottery production (c. 6600/ 6500 cal BC). A clear, yet gradual transition from the Neolithic to the Chalcolithic period is also obser- vable (c. 5500 cal BC). Bone and antler tools were frequently created and used by this community to support their burgeoning lifestyle. This paper explo- res the bone tool assemblage with a focus on a typo- logical and contextual analysis of the material. The subsequent discussion will provide an initial asses- sment of these tools and their position within the wider region. The Ugurlu excavation Ugurlu is located on the western part of Gökçeada (Fig. 1). The site is a low mound covering an area of approx. 250 x 200m on a gentle slope at the eastern foot of Mount Isa (Doganlı). The main Ugurlu-Dere- köy road cuts through the site. So far, during the eight years of excavation, six main cultural phases, designated as I–VI (counting from top to bottom), and at least 15 layers of occupation have been re- vealed. The earliest Phase VI dates to the pre-pottery Neolithic period, c. 6800–6600 cal BC. Phase V is marked by the first pottery Neolithic, c. 6600/6500– 5900 cal BC. During Phase IV, around 5900/5800 cal BC, the settlement expanded, and the islanders may have formed their own culture and identity. Around 5500 cal BC, major changes occurred at Ugurlu. These variations occurred in all aspects of cultural life, as indicated by modifications in settle- ment pattern, spatial organisation, the plan of buil- dings, art, and pottery production. The spatial orga- nisation of settlements and building plans were fur- ther modified around 4500–4300 cal BC (Erdogu 2014.157–178; 2016.89–94). Introduction Animal bone was an important raw material source at Ugurlu, as its durability and manageability was utilised to create strong and versatile items. The sy- stematic study of worked bone tools had its incep- tion in the mid-20 th century. Focus on typologies quickly expanded into other related areas, including use-wear analysis and experimental replication stu- dies (for early concepts of bone tool typologies, see Camps-Fabrer 1966.82–135; Schibler 1980; for use- wear analysis, see Semenov 1964; LeMoine 1997; for replication studies, see D’Errico et al. 1984; Goodar- zi-Tabrizi 1999). In recent years, significant scientif- ic advances have been made regarding the study of bone tools, expanding the knowledge base for this discipline (Bradfield 2015.3–14; Spangenberg et al. 2014.11–25; Griffitts 2011.52–66; Choyke 2007. 641–665; Legrand, Sidéra 2007.67–80). In Anatolia and the Balkans, prehistoric bone tool studies now largely integrate typological analysis with an empha- sis on scientific and theoretical procedure (Russell 2005.339–369; Dekker 2014; Vitezovi≤ 2011.117– 136; Zidarov 2008.57–62). Specialised publications of bone tools from Neolithic northwest Anatolia are lacking, with many site re- ports incorporating this material group with other small finds. A number of sites (Asagı Pınar and Bar- cın Höyük in particular) (Erdalkıran 2017.235 vd; Azeri 2015; Dekker 2014) have specialists currently working on their worked bone material, while in- depth analysis from non-worked bone experts is in- cluded in the reports from Yesilova, Yenikapı, Hoca Cesme, Fikirtepe, and Pendik (Derin 2012.182; Kızıl- tan, Polat 2013.123; Özdogan 2001.33–34; 2013. 174–182). Furthermore, the prehistoric sites of Ya- rimburgaz Cave, Toptepe, Aktopraklık, Ege Gübre, and Ulucak contain limited amounts of worked bone information in their published reports (Özdogan 2001.8–11, 64–66; Karul, Avcı 2011.1–15; Saglam- timur 2012.197–225; Çevik 2012.143–158; Çilingi- roglu et al. 2004.50). However, there are some exam- ples of detailed worked bone investigation, such as Marinelli’s report on the Ilıpınar bone and antler as- semblage (Marinelli 1995.122–143). Excavators at Neolithic sites in modern-day Greece and Bulgaria pay close attention to worked osseous material. Rosalie Christidou and Michel Séfériadés examined the Neolithic assemblages at Limenaria and Dikili Tash to emphasise aspects of bone exploi- tation in the region (Christidou 2005.91–104; Séfé- riadés 1992.99–112. The term Neolithic used for Jarrad W. Paul, Burçin Erdog ˘u 370 Worked bone and antler assemblage Methodology A ten-stage procedure was designed to collect infor- mation. After recording contextual information, the raw material (bone, antler, or tooth) of the object was recorded. The tool was then placed within a group, type, and if needed, sub-type. The object was measured (including maximum length, width, and thickness) and described. In this regard, each object was divided into a base, shaft, and tip and describ- ed individually. Surface condition, colour, and modi- fication (in the form of striation direction) were then recorded. Striations were identified with the naked eye, a x3 magnification eye-piece, and in some cases, a USB microscope with x200 zoom. This mi- croscope has a dual axis x50 and x200 lens with a manual focal range of 10mm onwards. The taxono- my and element were also recorded, limited to broad categories of small, medium, and large animal bones that were either long, short, flat, or irregular. All items were photographed and illustrated, with data recorded in a searchable database. Typology In total, 534 worked antler and bone items have been recorded thus far (Tab. 1). From these, 455 have been typed, while the remaining 79 have been classified as undefined. Objects identified as unde- fined are either too fragmented to classify or have no formal shape. Some are waste from manufacture. In order to create a more distinct picture of the assem- blage, any data using percentages relates to the 455 typed objects only. The typology is first separated into four groups: ‘pointed tools’, ‘cutting tools’, ‘polishing tools’, and ‘other’. The most prominent type in the pointed tool group are points (Fig. 2). This type is used to de- scribe any worked object with a pointed tip, excluding needles or pins. They were employed for a variety of tasks relating to perforation. The creation of ev- eryday items such as clothes and shoes would have required the use of points to create stitch- ing holes. Points were also re- quired in ceramic decoration and textile manufacture to ma- nipulate the visual form of an item. The decision to label this type ‘point’ rather than ‘awl’ is based on the unique functional characteristics of awls. Awls have a specific morphological function, and thus la- belling all points as awls is misleading. Due to the large quantity of points (n = 191, 42% of the over- all assemblage) eight sub-types have been defined, divided by morphology, raw material selection, and additional decoration (Tab. 2). Type A points are most common, with 115 exam- ples, and are characterised by a break on the shaft with a pointed proximal and no epiphysis or base. Type B consists of a longitudinally split bone with an intact distal end, while type C are split length- wise, with the base or distal end split in half. Type D points have a rounded and smoothed distal epi- physis, and type E are identified by an angled proxi- mal caused by abrasion on one lateral. Type F points are defined by decorations or additions made to the object’s surface. Two motifs are apparent: one is a zigzag pattern seen on two examples, while another has a double-‘v’ incision (Fig. 3). Type G has a flatten- ed and wide shaft that narrows into a small rounded tip, and what I describe as an awl, while type H com- prises pointed objects made on other elements be- side metapodial or tibia bones. Due to the variety of type H bone elements, mean length, width, and thick- ness varies greatly (the largest being a 88.42mm tool crafted from a femur and the smallest being a 39.78mm tool constructed from an incisor). Types A through G are mainly made on medium sized animal long bones, mostly metapodial bones of sheep and goats. Parallels exist between regional and inter-re- gional point types (Tab. 3). Fig. 1. Map of Ug gurlu on the island of Gökçeada (adapted from ©2016 Google Map data). An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) 371 There are five other types in the pointed tool group (Fig. 4). Rounded points are distinguished from points by having rounded and blunt tips. They are di- vided according to their raw material, with type A made from bone, while type B are made from antler. Only two bi-points are recorded, and are described as having both a pointed tip and base. This tool is associated with textile production or fishing related activities. Regarding fishing, three fish hooks are present. They vary in size and form, with breakage normally occurring on the bend, where the greatest amount of force is exerted. Two decorated headed bone hooks are also noteworthy. They were possibly used for securing clothing. Needles are well represented and defined by a point- ed tip either incised on the distal lateral (type A) or perforated at the base (type B). Of particular importance was the discovery of a large number of type A needles in trench P5, unit B91, discovered inside a pit and dated to 5600 cal BC. Consi- dering their apparent deliberate placement, and no obvious pa- rallels in the wider region, these objects may have had significant cultural value for the inhabitants of Ugurlu (Fig. 5). The final type in the pointed tools group is pins. They are the sec- ond most frequent pointed tool and are circular, with a slender shaft and pointed tip. The majo- rity fit into this form (type A); however, some (type B) have a non-circular shaft, but are slen- der and thin. Based on striation direction, these tools are related to perforation activities. The second tool group are cutting tools, which comprises four types (Fig. 6). The tools in this group are associated with activities that require cutting, scraping or strik- ing. There are 14 cutting tools in total, making this the smallest group (3% of the overall collec- tion). Chisels are defined by a bevelled tip and compact shaft, with no set base profile. Use-wear analysis from Çatalhöyük suggests these types of tools were used for woodworking activities (Giffitts 2011.52–66). Gouges have a hollowed shaft, varied degrees of base modification and a modified tip. Tips were most commonly strengthened by burning until blackened. This reinforces the bone cortex, making it easier to use on hard materials. A pick and a punch, both made from antler, were also found. The pick may have been used for digging, while the punch was used by striking the base with another tool, transferring the energy to the tip to exert force. Af- ter continual use, the base and tip are left damaged, exposing the spongy bone. There are only two types in the polishing tool group; however, it is the second most frequent group (n = 123, 27% of the overall assemblage). These tools are associated with the removal of excess material dur- Group Type Sub-type Number Total point Type A 115 Type B 11 Type C 12 Type D 27 Type E 13 Type F 3 Type G 5 Pointed Type H 5 191 tools rounded point Type A (bone) 17 Type B (antler) 14 31 bi-point 2 2 hook 6 6 needle Type A (notched) 20 Type B (perforated) 3 23 pin Type A (pin) 43 Type B (pin-point) 4 47 chisel 4 4 Cutting gouge 8 8 tools pick 1 1 puncher 1 1 smoother Type A 83 Polishing Type B 8 91 tools spatula Type A 23 Type B 7 30 component\ composite elements Type A – handle 4 Type B – shaft 6 10 Other perforated object 2 2 preform 4 4 utensil Type A – spoon 2 Type B – spatula-spoon 2 4 Total 455 455 Undefined 79 79 Total 534 534 Tab. 1. Ug gurlu worked bone and antler typology. Jarrad W. Paul, Burçin Erdog ˘u 372 ing hide preparation and pottery moulding (Fig. 7). The first of these are smoothers. Smoothers are also referred to in the Turkish literature as mablak. Some authors label morphologically similar items as chisels or gouges. They are characterised by a long curved shaft, either open (type A) or flat (type B), with a be- velled tip and base worked to varying degrees. Type A smoothers outnumber type B and are made from long bones, including metapodial, radius, and tibia bones, of medium- and large-sized animals. Striations usually appear on the second aspect of these objects, running horizontally and diagonally. The second type in this group is spatulas. A spatula has a flat, broad shaft with elongated ends and is made exclusively on rib bones split lengthwise. Type A and B are differentiated by tip morphology: type A is rounded, while type B is pointed. Their primary purpose was pottery moulding. Closely linked with pottery production, experimental studies show that they are extremely useful in regularising the shape of ceramics and removing excess material (Ma ˘r g a ˘rit 2017.206). Items in the ‘other’ tool group are not functionally associated with pointing, cut- ting, or polishing activities. This includes artefacts that appear not to be tools but, rather, to be worked bone and antler objects that may fulfil a variety of other functions, including use in ritual settings or as items with purely aesthetic value. There are 20 items in this group, divided in- to five types and four sub-types (Fig. 8). Component/composite elements are separated into handles (type A) and shafts (type B). Handles in this context refers to handles for spoons, where the bowl has been broken off, while shafts are made exclusively from deer antler and hollowed out at one end to hold a sharp stone object. Shafts acted as handles to protect the user’s hands, resulting in the ability to exert greater force. Two perforated objects are included in this group. The first was made of deer antler and was perforat- ed in the middle of the shaft, while the other is recti- linear, with a rounded and smoothed base. I suggest the latter perforated object may have functioned as a toggle used to hold a garment together, as it is si- milar in appearance to modern examples. Four preforms are also recorded. A preform is any object displaying early signs of manufacture, but ne- ver made into a complete type. It is important to re- cord them, as they can highlight procedural steps taken in constructing certain types. For example, one object is a needle preform which has the beginning of a perforation on one aspect and a dimpled impres- sion on the other. Much of this object is rough, sugge- sting drilling was an initial step in the manufactur- ing process. Fig. 2. Ug gurlu points. Left to right: type A (CC21B26x1); type B (P5B148x4); type C (DD20B4x1); type D (P6B7x9); type E (P5B106xz2); type F (BB20-21B9x1); type G (QP6B6x4); type H (BB20-21B59x7) (de- signed by J. W. Paul). Fig. 3. Point type F: P5B87x5, P6B41x2, BB20-21B9x1 (designed by B. Erdog gu and J. W. Paul). An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) 373 Utensils complete this group. Type A are described as spoons, with a rounded and bowled-shaped tip and handle, or broken shaft where the handle would have been. These items are associated with dairying activities, scraping, or ritual/symbolic settings. When slanting occurs on the bowl, as with sample DD20- B8x1, repeated scraping actions are implied. Type B are labelled spatula-spoons. Like type A, they have a concave head and handle, but the shaft is elongated and the head is wider. The remaining 79 objects are labelled ‘undefined’ (Fig. 9). Some items in this group have clearly been modified, but their function remains un- known. Modifications involve notches, cut marks, and stria- tion patterns. In addition, some objects are tools, but their function remains un- known. For example, three objects all share the same morphology and are manu- factured on large animal long bones. It has been suggested that they may have acted as fishing rods. Another object from Phase II, c. 4300 cal BC is shaped much the same as a pipe, but there is no hollow connection between the stem and bowl. It may have been used as a digging instrument. There is also a collec- tion of smoothed objects that are too small to be formally identified. Context Important contextual levels relating to the worked bone and antler collection will be outlined in the fol- lowing section. Not limited to the worked osseous Sub-type Phase Total VI V V–IV IV IV–III III II surface unknown A 1 4 5 37 5 42 8 5 8 115 B– 11 612 –––1 1 C– 21 413 ––11 2 D– 4– 7–1 0 1232 7 E– –– 5 –4 3–11 3 F– –1 2 –– –––3 G– –1 21– 1––5 H– –2 –12 –––5 Total 1 11 11 63 9 63 13 7 13 191 A 45.8 15.3 8.0 3.1 2.5 0.9 B 59.4 18.9 12.0 3.2 2.7 1.6 C 60.7 14.4 12.3 3.5 2.9 1.7 D 60.9 19.2 8.6 3.2 2.9 0.8 E 50.2 18.2 8.5 2.5 2.8 1.1 F 44.5 0.6 6.8 0.4 2.1 0.6 G 35.0 10.0 8.9 0.2 2.0 0.9 H–––––– Tab. 2. Ug gurlu points sub-type contextual distribution, measurements, and parallel types. Fig. 4. Ug gurlu pointed tools. Left to right: rounded point type A (BB22B10x2); rounded point type B (P6ppx2); bi-point (P6B33x5); fish hook (BB20-21B59x6); needle type A (O5B18x1); needle type B (O5B36x3); pin type A (P6B25x3); pin type B (P6B30x6) (designed by J. W. Paul). Sub-type Mean max. length (mm) Standard Deviation Mean max. width (mm) Standard Deviation Mean max. thickness (mm) Standard Deviation Jarrad W. Paul, Burçin Erdog ˘u 374 material, this investigation will take into account re- lationality with other items in the overall material record. This section is divided into chronological phases, beginning with the earliest, to highlight de- velopmental change in the assemblage. Phase VI (Pre-Pottery Neolithic 6800–6600 cal BC) The earliest occupation at Ugurlu is represented in two deep trenches in the eastern part of the settle- ment, namely BB20-21 and CC21. Five bone tools have been recovered so far, comprising three spatu- las, one spatula-spoon, and one type A point. Phase V (Early Neolithic 6600–5900 cal BC) The worked bone objects in Phase V can be describ- ed as skilfully constructed, extensively used, and de- posited largely intact. Twenty-three objects have been recorded so far and are confined to the east- ern side of the site in trenches BB20-21 and BB22. The majority of these items are pointed tools, with points (n = 11) common. Sub-typed points range from A–D, with type A (n = 4) and type D (n = 4) predominating. Overall, points were modified to a great extent, as shown by finer measurements, sur- face condition, and overall preservation. Only one object is made from antler, while the rest are made from bone. The early inhabitants of the site had li- mited access to this material and, considering the di- stance to the mainland, access to deer, and therefore antler, would have been restricted, with sheep and goat favoured on the island for size and adaptability. Unit B58 in trench BB20-21 contained the most bone tools (three points, a smoother, and an undefined Sub-type Regional Parallels Inter-Regional Parallels A Ulucak Höyük point type A Tóth’s type 1\7–1\9< Camps-Fabrer point type IV< Makri type 1.1b B Ulucak Höyük point type B Camps-Fabrer type II< Schibler’s type 1\1< Choyke no. 1< Ilıpınar awl type A Payne type 2, type 3< Dikili Tash type IIIB, VIIA< Cave of Cyclops type B< Karanovo point\awl type A C Ulucak Höyük point type C Camps-Fabrer type III< Schibler type 1\2< Chomko type II< Ilıpınar awl type B Dikili Tash type IVB, VIIA< Karanovo point\awl type B D Ulucak Höyük point type D Camps-Fabrer Type V Ilıpınar awl type B E Ulucak Höyük point type E Dikili Tash type VIIB< Karanovo type C G Ulucak Höyük point type G Camps-Fabrer type IVa H Ilıpınar awl type C Karanovo awl type C< Camps-Fabrer type I Tab. 3. Regional and inter-regional parallels for point sub-types. Fig. 5. Worked bone assemblage in ‘box-like pit’ from trench P5, unit B91: P5B91x2, x3, x6, x9, x10, x13, x14 (designed by B. Erdog gu, N. Yücel and J. W. Paul). An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) 375 item) with worked sea shell and stone also recorded. The proximity of these items with stone and sea shell objects suggests an early connection between mate- rial groups. Relatively low numbers of finds in this period may be a result of two major factors. The first is access to raw material during an initial settlement process, and the second is the constricted excavation area. The shift from the Early Neolithic Phase V into the established Neolithic Phase IV (5900 cal BC) again sees pointed objects dominate, with points the most frequently recorded (n = 11), while polishing tools are also well represented (smoothers n = 3; spatulas n = 4). Only one object was not undefined, reflecting a general trend in the assemblage towards formality of types. Phase IV (Late Neolithic 5900–5600 cal BC) Phase IV at Ugurlu saw an expansion of material cul- ture and architectural remains spread towards the western part of the site. Trenches O5, O6, P5, and P6 all contained vast amounts of worked bone ma- terial, more than was found on the eastern part of the site. 161 tools in this phase make it the second largest. Again pointed tools dominate, with points the item recorded most often (n = 63). A definite trend is now emerging, with type A points compri- sing the majority (n = 37). The zigzag motif carries into this period, as seen on one object. Another mo- tif also appears at this time, a double ‘v’-shaped in- cision on the middle of the shaft of one item, disco- vered in a fill level with four other worked bone ob- jects. This complete object dates to Phase IV level 1, 5900–5800 cal. BC. These incisions seem to have served no functional purpose; rather, they have been added for purely artistic effect. Ad- ditional time is required when inci- sing an object for purely decorative reasons, accounting for the relati- vely low number of decorated points in the overall assemblage (n = 3). After this period, no further decorat- ed points are found in the assem- blage. Thus, Phase IV represents a peak in the production of decorated pointed objects. Pins (n = 18) are numerous in the Neolithic, with their slender shafts displaying high levels of polish. Ne- edles (n = 12) are also well repre- sented and their numbers peak dur- ing this phase. Cutting tools also reach their peak during Phase IV, with chisels (n = 3) and gouges (n = 3) represented, while polishing tools also becom- ing more frequent, with smoothers (n = 29) the se- cond most frequent type. Spoons and spatula-spoons reach their highest numbers during this period. Flat- ter and wider spatula-spoons are larger in size, while object DD20B8x1 (Fig. 10) shares characteristics with similar regional examples. Large amounts of worked bone and antler material from this period are spread across a variety of con- texts. Only a few worked bone objects from the Late Neolithic originate from known contexts. On the east- ern side, sixteen artefacts were discovered on floor unit B24 in trench BB20-21, 11 of which were wor- ked bone. The remaining objects included two poli- shed stone axes and one bead. Besides floor levels, worked bone was also found in pits during the Neolithic. Unit B91 in trench P5 is a pit which contained 14 bone tool objects enclosed with yellow-coloured plaster. Many of the tools in- side were type A needles. Interpreted as idols (Erdo- gu 2014.159), these symbolically charged objects can be seen to represent an anthropomorphic form, with their notched bases representing a head, possibly in- serted into clay figurines. Fill deposits in trench P5 (units B14, B108, B104) also contained numerous bone tools alongside other objects, including worked sea shell, worked stone, and figurines. In all instances, breakage is an appro- priate reason for the discard of these items. The transition from Neolithic to Chalcolithic (5500 cal BC) was moderate when observing the distribution Fig. 6. Ug gurlu cutting tools. Left to right: gouge (P5B98x7); chisel (CC21 B19x4); punch (P6ppx3) (designed by J. W. Paul). Jarrad W. Paul, Burçin Erdog ˘u 376 of bone tool types, although in some cases, initial shifts in type frequencies become fully recognisable during the subsequent phase. Phase III (Early Chalcolithic 5500–4900 cal BC) The Chalcolithic period at Ugurlu is defined by ma- ximum settlement growth and a variation in materi- al culture. The worked bone reflects broad patterns of change, with new tool types favoured. It would be wrong to suggest that the quality of these items de- clined. Instead, a clear move towards homogeneity in some tool types of tools is noticeable, indicating shifts in manufacturing practice. A total of 176 wor- ked bone objects from this period have been found, with pointed tools predominating (n = 63). The de- mise of types F and G, coupled with the inflated pre- valence of type A points (n = 42), shows a move to- wards regularity in the point type. Additionally, pin and needle distribution all favour one sub-type of the other (type A pin: n = 20, type B pin: n = 0; type A needle: n = 6, type B needle: n = 0). Type A nee- dles decline in frequency (n = 6), an interesting ob- servation given their potential symbolic meaning. Perhaps their production became so regularised they were manufactured on materials easier to procure (but degradable), such as wood. Cutting tools are re- latively uncommon, consistent with previous peri- ods, while the number of smoothers remains the same (n = 23); however, they also be- come more homogenised, with only one type B recorded. The three com- ponent/composite elements are all type B shafts made from deer antler, while type B rounded points (made from deer antler) also increase dur- ing this period (n = 5). The more fre- quent use of this material might be linked to enhanced access to deer. Spoons and spoon handles also de- cline in number, with only one exam- ple recorded. This object, a small cir- cular spoon head, with a missing long thin handle, is reminiscent of the spo- ons at Pendik and Kova≠evo. Only one perforated object was found from this period. This item has one complete hole and one fragmented hole drilled from both sides, and pos- sibly functioned as a toggle. Floor levels offer the best contextual evidence. Trench V18 yielded four floor levels containing worked bone and antler material. Unit B12 contained two work- ed bone items and a piece of worked stone, while unit B11 yielded one worked bone pointed tool and one worked stone example. Interestingly, only one worked stone object was found in each of the two units, a trend with floor levels found in this trench. Fill contexts contain the most worked bone mater- ial in trenches on both sides of the site. Trench DD- 19-20, unit B2, is located inside Building 3 and in- cluded two type A pins, one stone chisel, and one worked sea shell. The pins are parallel in size, form, and colouration. They are well worn, with notches and scrapes covering their surfaces, and were unco- vered next to one another. Unit B4 in trench BB22 produced some interesting results, with four spatulas found together: two type A and two type B. These objects, used for shaping the interior and exterior of pottery, were found within a dense concentration of animal bones. With only one type of bone object – worked bone spatulas – uncovered in this unit, it is safe to assume the area was associated with ceramic production. In the west, trenches P5 and P6 also con- tain numerous worked bone examples from fill con- texts, especially unit B60 in trench P5. This has been labelled a pit, with a vast amount (n = 18) of bone tools damaged by burning. For example, one expert- ly crafted type A needle with two layers of notching at the base appears blackened because of burning. Fig. 7. Ug gurlu polishing tools. Left to right: smoother type A (P6B- 40x1); smoother type B (P5B137x6); spatula type A (P5B22x1); spatula type B (P5B148x2) (photo by J. W. Paul). An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) 377 Due to the symbolic nature of type A needles, the destruction of such a finely crafted object suggests it was unintentional. However, its placement within a pit may suggest that this object was in fact intentio- nally burnt. Perhaps this may have been a part of the ritual surrounding them, which explains their ab- sence in the material record past the Neolithic period. Overall, the worked bone assemblage of Phase III marks a period of adaptation. There is maximisation of homogenous forms, increase in the use items made from antler (Phase V: n = 1; Phase IV: n = 3; Phase III: n = 14), and a reduction and sometimes exclusion of certain tool types (type F and G points, type B needles and spoon handles are all absent dur- ing the Chalcolithic). During this phase, raw material was readily available and construction techniques followed known patterns. Most of the developments beginning in Phase III continue into Phase II. Phase II (Middle Chalcolithic 4500–4300 cal BC) Fifty-three worked bone and antler objects were re- covered from this phase. Although pointed tools re- main the most recorded group (n = 21), polishing tools are a close second (n = 15). Points dominate (n = 13) and are defined by their homogeneity, with the majority type A (n = 8). Needles occur in rela- tively smaller numbers during this period (n = 3), with the inclusion of one type B needle, with a per- forated base, and two type A needles. The two type A needles appear rudimentary when compared to their earlier counterparts, although they retain their slender shaft and appearance. No cutting tools were found during Phase II. This may reflect the use of other tool groups to replace these larger and more robust bone tools. 14 smoothers are disproportiona- tely divided between type A (n = 13) and type B (n = 1) and are seen as crude variations on previous forms, with some displaying unworked distal epiphy- ses, while others have had their bases removed en- tirely. Undefined objects dramatically increase in this phase, 28% of the overall Phase II collection, with higher occurrences revealing a noticeable amend- ment to tool construction methodology. This is not to say skill levels dropped without exception. O11B- 7x1 (Fig. 11) shows evidence of perforation on one aspect, made by a drill-like instrument. The second aspect has been indented for use as a guide. When considered together with the type B needle, it be- comes clear expert craftspeople were still operating on bone and antler, just not to the same degree as in previous phases. Almost all Phase II material comes from the western side of the site (in trenches O11, OP11, P5, and P6), with only two objects from the eastern side (trenches BB14 and BB15). Unit B4 in trench OP11 is associated with Building 1, a rec- tangular building 7 x 9m, and in- cluded two worked bone objects, with a pestle, spindle whorl, and worked sea-shell on its floor. Spin- dle whorls, an object used for tex- tile manufacture that has not oc- curred frequently up to this point, are also found with worked bone items in trenches P6 (unit B2) and P5 (unit B2). Their increased pre- sence could indicate an on-site shift in textile manufacture. Regu- larity in tool form also seems ap- parent. For instance, unit B2 in trench P6 contained three large, open-shafted smoothers, all with a slanted tip and similar use-wear characteristics from striation pat- terns. Phase II worked bone was influ- enced particularly by the preced- Fig. 8. Ug gurlu other tools. Left to right: component/Composite ele- ments type A- handle (BB20-21B29x2); component/composite ele- ments type B- shaft (P6B5x6); perforated object (P5B68x1); preform (O11B7x1); utensil type A- spoon (DD20B8x1); utensil type B- spatula- spoon (CC21B26x2) (designed by J. W. Paul). Jarrad W. Paul, Burçin Erdog ˘u 378 ing Chalcolithic Phase III period. Most of the type tendencies from the Neolithic had by now changed, with the exception of pointed tools. There was also a reduction in quality, but construction skill for cer- tain objects remained consistent. By the conclusion of Phase II, worked bone tools at Ugurlu were being associated with a wider variety of objects than ever before. Discussion Ug gurlu worked bone and antler: a local per- spective The worked bone and antler at Ugurlu can be de- scribed as an adaptable assemblage, adjusted to suit the needs of a society which sourced, modified, and used their tools to create essential items for daily use (Tab. 4). These tools were not created and used in isolation, but were rather connected to the larger material collection. Much of the raw material used to create these tools came from medium-sized ani- mal bones (n = 116), which included sheep and goats. Access to animals was the first step in the ma- nufacturing sequence at Ugurlu. A select number of animals would have been chosen for the initial jour- ney, with small- and medium-sized ones (sheep, goat, and pigs) favoured due to limited space on sea- rafts (Camps 1986.23; Broodbank, Strasser 1991. 240–241; Simmons 2014.77). Animal selection had a direct influence on bone tool manufacturing, as the choice of animal would have been initially re- stricted. This is reflected in the assemblage, with tools made from large animal bones barely notice- able in the earliest periods (Phase V: n = 1) and stea- dily growing as time passed (Phase IV: n = 12). Ani- mals were not slaughtered specifically for the pro- duction of tools; rather, they were butchered only when every available resource was depleted, includ- ing breeding with other animals to increase the size of the herd and extracting any meat or marrow. Due to size regulation, 70% of tools produced during the Neolithic from a known origin came from medium- sized animals such as sheep and goat. Fish hooks are also recorded in this early assemblage. Fish hooks and other fishing-related items such as nets would have been of great use on an island site, as fish and other molluscs could have supplemented a rather restricted diet (Powell 1996.104–105). As a result, the emergence of bone fish hooks around 6500 cal BC reflects an early adaptation of material culture not entirely land-based. Moving into the Late Neolithic period (5900–5600 cal BC) at the site, Phase IV sees an increase in poli- shing tools. Moreover, decorated points (type F) are manufactured for the first time. Decorated points make a relatively short appearance in the cultural record, and then completely disappear by the end of Phase IV. These incisions were probably purely de- corative. The introduction of artistic embellishments on functional tools thus shows a shift towards in- creased time allotment and greater manufacturing skill. Two clothing bone hooks also contain mark- ings. One decorated with anthropomorphised fea- tures might be interpreted as a figurine, while the other contains a bird motif, which suggests some sort of symbolic importance. Phase IV also witnessed a dramatic growth in the number of bone tools produced. With each object created, the operational chain is altered to provide optimal effectiveness for the user. For instance, the awl-shaped points (type G) give the user greater room to manoeuvre when boring holes through dried animal hide. Consequently, the array of point types reflects an expansion of tool preparation, link- ed with greater accessibility of raw material or in- creased skill. Needles are not found in Phase V, but increase in frequency during Phase IV (Phase IV: n = 12). They then decline drastically in number during the Chalcolithic (Phase III: n = 6; Phase II: n = 3). Given their shape, these needles may have been in- serted into figurines found in contextually compara- ble locations, to represent a human form. Perhaps Fig. 9. Undefined tools. Left to right: Q6B10x2; OP11B8x10; P5B50xz3 (designed by B. Erdog gu and J. W. Paul). An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) 379 these needles also played a symbolic role, contribu- ting to the notion that Gökçeada may have been a ‘sacred landscape.’ It is conceivable that the island was prominent during prehistory due to its visible position from the mainland and anthropomorphised natural features (Erdogu 2003.7–22). These needles may have substituted the religious/social identity of the island (Erdogu 2003.19). Supporting this claim is the fact that these needles were found inside yel- low plastered storage pits. Objects found inside, or associated with, plaster pits are common across Ana- tolia during the Neolithic, protecting the items hou- sed inside (Perlès 2005.275). However, plaster pits also seem to serve a ritual or religious function, with plaster used at prehistoric Anatolian sites in conjunc- tion with ritual activity, particularly during moments of transition (Balter 2001.2278–2281; Hodder 2005. 10–11; Cessford, Near 2005.179). The apparent short-lived popularity of needle production may have been the result of raw material choice, as other ma- terials, such as wood, may have been favoured dur- ing later phases. Although spatula-spoons are found in Phase VI, rel- atively complete examples of spoon heads or han- dles are first seen in Phase V and IV. Spoons are common on mainland Anatolia and Thrace, with fine- ly worked objects frequently found (Lichardus-Itten et al. 2006.88–89; Özdogan 2013.182–187). Spoons may be linked with an original toolkit transported onto the island by mainland farming groups. These objects do not appear in the assemblage during the Chalcolithic, perhaps replaced again by raw material alternatives such as wood. New procedures in the manufacturing sequence, seen in Phase IV, meant bone tools could be used with more force. Techniques such as the controlled burn- ing of medium-sized animal bone tips during the fi- nal stage of manufacture solidified the tool’s work- ing end. This type of manufacturing sequence can be found on a number of items labelled gouges and chisels. Significant amounts of large animal long bones were also used to create gouges and chisels during the Neolithic (Phase V: n = 0; Phase IV: n = 6), resulting in two outcomes. First, bone tools made from large animal bones indicate an increased avai- lability of large animals, or at least improved access to them. Second, the apparent willingness to try dif- ferent manufacture methods suggests stability. If re- sources are scarce, or tools are needed for immedi- ate use, time and experimentation give way to expe- diency and functionality. For example, tools delibe- rately burnt in a controlled setting emphasise a com- munity relatively abundant in resources, with little time constraints to maintain production on an im- mediate basis. From Phase IV onwards, greater numbers of unclas- sified objects began to emerge at Ugurlu (Phase IV = 9.9%; Phase III = 15.3%; Phase II = 28.3%). This may have been a direct result of experimentation, as dif- ferent manufacturing methods were used to enhance tool capability. Underscoring this was the develop- ment of raw material not previously used at Ugurlu to make bone tools. As well as the increased use of large animal bones, deer antler was used in greater quantities (Phase V: n = 1; Phase IV: n = 7; Phase III: n = 11). Bone tool production steadily changed and adapted from the Neolithic to the Chalcolithic. There was no abrupt change in types or groups. Rather, Fig. 10. DD20B8x1 (photo by J. W. Paul). Fig. 11. O11B7x1 (photo by J. W. Paul). Jarrad W. Paul, Burçin Erdog ˘u 380 certain production characteristics that began around 5500 cal BC continued during the Chalcolithic Phase III period. Many of the pointed tools recorded in Phase IV carry over to Phase III, including points, rounded points, needles, and pins. Despite this, regularity and pro- duction contrasts with their Neolithic counterparts, as sub-types seem to be favoured. Most of the Chal- colithic Phase III points (67%) are broken at the shaft (type A), with round-based versions (type D) also well represented (16%). There is no evidence of de- corated (type F) or awl-shaped (type G) points dur- ing the Chalcolithic Phase III or II. An apparent ma- ximisation of broken points may reflect an overall expanded manufacturing output, as many of these tools are found in pits and pit levels, rather than floor levels. Increased numbers of disposed items might correlate with amplified production and use, as tools were discarded in greater numbers. Round- ed points made from bone and antler expand expo- nentially when compared to previous phases (Phase IV: 2.7%; Phase III: 8.1%). During Phase IV, spatulas are scarce in the overall record and evenly distrib- uted across two sub-types (type A: n = 3; type B: n = 3). In contrast, during Phase III, they double in num- ber, and almost completely conform to type A (type A: n = 10; type B: n = 2). Greater frequency of spat- ulas during Phase III (8.1%) when compared to Phase IV (4.1%) may also be linked to expanded pot- tery production, as these tools were primarily used to smooth ceramics during the moulding stage. Choice of raw material also develops during the Chalcolithic. For instance, antler production, which seems to begin at the end of the Neolithic Phase IV (6000/5900 cal BC), grows in subsequent phases. This material was used to create rounded points (type B), shafts, and a pick. Not every tool type grew in frequency, however. Most notably, needles (Phase IV: 8.3%; Phase III: 4%) and utensils (Phase IV: 2.1%; Phase III: 0.7%) reduce in volume when compared to Phase IV. Type A needles, possibly associated with ritual activity during the Neolithic, were still used, but halved in number (Phase IV: n = 12; Phase III: n = 6). When compared to previous phases, Phase II bone objects seem to be crudely made. A foremost sense of functionality over any visual appeal pervades the collection. Unclassified material is much larger when compared to earlier phases, with a ratio of almost 2:1 in terms of classified to unclassified material. In com- parison, Phase III is roughly 5:1, while Phase IV is even higher. The number of smoothers also dramati- cally rises in Phase II (36.8% of the typed assem- blage). In fact, they outnumber points (34.2%), a tool type consistently favoured throughout the site’s hi- story. They are smaller (mean length of 68.6mm; SD 20.4) when compared to Phase III smoothers (mean length of 79.1mm; SD 32.1). The adaption of textile manufacture during this phase, coupled with the in- crease of spindle whorls, might explain the lower fre- quency of points. Additionally, other material types, such as copper, may have been used, although limit- ed evidence has been discovered to support this claim. Group Type Phase VI V V–IV IV IV–III III II surface unknown Total Pointed point 1 11 11 63 9 63 13 7 13 191 tools rounded point – 3 – 4 4 12 3 2 3 31 bi-point – 1 – 1 – – – – – 2 fish hook – – 1 2 – 2 – – 1 6 needle – – – 12 – 6 3 1 1 23 pin – 1 – 18 6 20 2 – – 47 Cutting chisel – – – 3 – 1 – – – 4 tools gouge – – 1 3 – 3 – 1 – 8 pick – – – – – 1 – – – 1 puncher – – – – – – – – 1 1 Polishing smoother – 1 3 29 6 23 14 6 9 91 Tools spatula 4 1 4 6 1 12 1 1 – 30 Other component\composite –1111311 11 0 elements perforated object – – – – – 1 – – 1 2 preform – – 1 – 1 1 1 – – 4 utensil – – – 3 – 1 – – – 4 – undefined – 4 1 16 6 27 15 5 5 79 Total 5 23 23 161 34 176 53 24 35 534 Tab. 4. Ug gurlu worked bone and antler contextual distribution. An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) 381 The modification of tool types, and the context in which they were discarded, provides insight into shi- fting site processes. For example, the constant reli- ance of pointed tools highlights confidence in the sourcing of certain raw material, including sheep and goat metapodial bones. Likewise, the disappea- rance of spoons after the Neolithic indicates a shift perhaps to a more localised toolkit. Needles, with their distinctive bases, coupled with a different cera- mic technique, also emphasises a shift towards a lo- calised material culture. Thus, an overall initial im- pression of the Ugurlu worked bone material is stabi- lity at a tool group level, while variation and adap- tions occur at a tool type level. Ug gurlu worked bone and antler: a regional perspective The spread of Neolithisation to the corners of Ana- tolia is seen as a complex issue not easily solved by a simplistic model of migration (Colledge et al. 2004. 35; Tomkins 2004.56; Özdogan 2011.415–430; Bu- dja 2013.47; Düring 2013.75–100; Lichter 2002. 161–169; Pinhasi et al. 2005.2220–2228; Thissen 1999.29–39). In this regard, a web of interconnect- ed strategies should be determined on an individual basis, with an emphasis on certain ‘push’ and ‘pull’ factors (Anthony 1990.895; 1997.22). Despite Ugurlu’s distance from the mainland (ap- prox. 12km during prehistory), common regional material was evident at the site. Certain ceramic types at Ugurlu, such as Impresso ware, were com- mon on the Anatolian west coast (Özdogan 2010. 886). Additionally, white-on-red painted ware, found in Phase IV at Ugurlu, was widespread north of the island, originating from Karanovo (Erdogu 2014. 160; Marinova 2007.93). Obsidian from Melos is also evident at Ugurlu, and was widespread in the region, with examples seen to the west and east (Er- dogu 2014.161; Sampson 1998.20; Takaoglu 2013. 36; Miliç 2014). The extensive use of worked bone is also evident in this region, with certain common tool groups com- mon to both Ugurlu and contemporary sites. Point- ed tools were the predominant tool group at most sites in the region, with points (or awls, as they are also labelled) the most common type. Points were the most common type at Ulucak, Ilıpınar, Barcın Höyük, Dikili Tash, and Makri all have a predomi- nance of points in their assemblages (Marinelli 1995. 124; Dekker 2014.64–65; Gerritsen et al. 2013.93– 112; Séfériadés 1992.102–104; Stratouli 1998.36– 40). The most common sub-type seems to be the Ugurlu type B point, with morphologically similar sub-types evident at Ulucak (type B), Illıpınar (type A), Dikili Tash (type IIIB), the Cave of Cyclops (type B), and Karanovo (type A). The large amount of pointed tools found across the region is the result of several factors, including the relatively simple set of manufacturing steps needed to produce points, a wide selection of raw material to draw upon during construction, and the ease of breakage. Smoothers at Ugurlu also find regional parallels, for example at Ilıpınar, Dikili Tash, and Karanovo, although they are often labelled as chisels (Marinelli 1995.126; Sé- fériadés 1992.100–101; Höglinger 1997.161). However, when investigating regionally compara- tive worked bone assemblages, certain characteristic tool types become evident. For Ugurlu, this is the type A needle, not seen in collections across the re- gion, except for a wooden parallel at Yenikapı (Kızıl- tan, Polat 2013.123). For sites in Anatolia, Ulucak Höyük is characterised by its increased production and use of larger animal bones, while Ilıpınar has a well-represented collection of spoons, toys, and figu- rines (Marinelli 1995.121–143). Barcin Höyük, Pen- dik, Fikirtepe, Hoca Çesme, and Asagı Pınar all have high-quality bone spoons in their assemblages. These are well made, richly decorated items. Northwest of the island, Dikili Tash is reliant on deer antler as a raw material source, more so than any other in the region. The assemblage at Makri is defined by the ab- sence of items such as spoons, fish hooks, pendants, and spatulas (Stratouli 1998.36–40). Karanovo, emerging from the Anatolian tradition, contains a large section of well-crafted spoons, also common to the Thrace area (Höglinger 1997.157–159). Sites in the region, therefore, have certain unique worked bone tool types that are not always widespread or shared between sites. As a result, not all types seen at Ugurlu are shared with other sites. Overall, the Ugurlu assemblage can therefore be seen as one that shares common regional tool groups, while also displaying localised tool types (for exam- ple, type A needles). Regarding the Neolithisation of the island, a land-based agricultural society relied on domestic animals, grains, and a toolkit largely made up of types consistent with other mainland cultures. As time passed, however, certain localised traditions took effect, most evident in the change of ceramics. This is also reflected in the worked bone assemblage, for instance with the decrease in bone spoons, high- lighting shifting manufacturing practices. Although Ugurlu borrows various elements from the Anatolian tradition, bone spoons (and the associated handles) Jarrad W. Paul, Burçin Erdog ˘u 382 are rare and were produced only during the Neoli- thic (Phase V: n = 2; Phase IV: n = 4). The decrease in the number of utensils in Chalcolithic Phase III at Ugurlu does reflect larger trends north and west of the island. The lack of spoons is synonymous with sites located in modern Greece during the same pe- riod (Perlès 2005.278). Sites in Anatolia (particular- ly Thrace) had a continual use and manufacture of bone spoons during this time, which makes the lack of widespread transfer to the Aegean prominent in the material record. Thus, it is clear that the worked bone and antler col- lection at Ugurlu was entirely consistent with nearby sites, its dynamic assemblage working within a glo- bal tradition that, like other sites investigated, had its own distinctive local traditions. Conclusion When Ugurlu was first settled in the 7 th millennium cal BC, bone tool production was already underway, as it was an essential component of the prehistoric toolkit. Throughout the Neolithic and Chalcolithic, as the site grew, so too did bone tool manufacture, incorporating a wider variety of animal bone and antler during subsequent phases. The worked bone assemblage did not exist in isolation, but was rather one element assisting and enhancing the Neolithic way of life. As a result, these bone objects were used to sew animal hide into shoes and bags, mould clay into pots for cooking and storage, secure fish to a line, and conceivably assist in ritual performance. Future studies regarding the Ugurlu worked bone collection will focus on further use-wear analysis to support functional assignment. Further investigation of analogous tool types with those found at Ugurlu, such as type A needles and decorated points, would also be worthwhile as part of a larger comparative analysis including sites to the north and east. In sum, stability defines the Ugurlu worked bone as- semblage, with their production, use, and discard re- maining constant over roughly 2000 years. This col- lection played an extremely important supportive role in the establishment and maintenance of a com- munity located at a geographic and cultural cross- roads. Jarrad W. Paul wishes to acknowledge the following support: Australia Postgraduate Award, Jessie Webb Scholarship, Lizette Bentwitch Scholarship, and the Graduate Research in Arts Travel Scheme (University of Melbourne, Australia). The Ugurlu Excavation pro- ject is supported by The Turkish Ministry of Culture and Tourism and University of Thrace. ACKNOWLEDGEMENTS References ∴ ∴ Anthony D. W. 1990. Migration in Archaeology: The baby and the bathwater. American Anthropologist 92(4): 895– 914. 1997. Prehistoric migration as social process. In J. Chap- man, H. Hamerow (eds.), Migrations and Invasions in Archaeological Explanation. Archaeopress. Oxford: 21–32. Azeri H. 2015. The Evaluation of the Bone Assemblage of Asagı Pınar Mound’s Chalcolithic Period. Unpublished MA thesis. Istanbul University. Istanbul. Balter M. 2001. Did Plaster Hold Neolithic Society Toge- ther? Science 14(294): 2278– 2281. Bradfield J. 2015. Investigating the potential of micro-fo- cus computed tomography in the study of ancient bone tool function: results from actualistic experiments. Jour- nal of Archaeological Science 40: 2606–2613. Broodbank C., Strasser T. F. 1991. Migrant farmers and the Neolithic colonisation of Crete. Antiquity 65(247): 233–245. Budja M. 2013. Neolithic pots and potters in Europe: the end of ‘demic diffusion’ migratory model. Documenta Praehistorica 40: 39–55. Camps G. 1986. The Young Sheep and the Sea: Early Na- vigation in the Mediterranean. Diogenes 34: 19–45. Camps-Fabrer H. 1966. Matière et art mobilier dans la préhistoire Nord-Africaine et Saharienne. Arts et Métiers Graphiques. Paris. An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) 383 Cessford C., Near J. 2005. Fire, Burning and Pyrotechno- logy. In I. Hodder (ed.), Çatalhöyük perspectives: themes from the 1995–1999 seasons. McDonald Institute for Ar- chaeological Research/British Institute of Archaeology at Ankara. Cambridge: 171–182. Çevik Ö. 2012. Ulucak Höyük 2009–2011 Yılı Kazı Çalıs- maları. Kazı Sonuçları Toplantısı 34: 143–158. Choyke A. M. 2007. Objects for a Lifetime- Tools for a Sea- son: the bone tools from Ecsegfalva 23. In A. Whittle (ed.), The Early Neolithic on the Great Hungarian Plain: In- vestigations of the Körös culture site of Ecsegfalva 23, County Békés. Archaeological Institute of the Hungarian Academy of Sciences. Budapest: 641–665. Christidou R. 2005. Aspects of bone exploitation in the Neolithic sites of Eastern Macedonia, Greece. In H. Luik, A. M. Choyke, C. E. Batey and L. Lõugas (eds.), From Ho- oves to Horns, from Mollusc to Mammoth – Manufac- ture and Use of Bone Artefacts from Prehistoric Times to the Present. Teaduste Akadeemia Kirjastus. Tallinn: 91–104. Çilingiroglu A., Derin Z., Abay E., Saglamtimur H. and Ka- yan I. 2004. Ulucak Höyük Excavations Conducted Be- tween 1995 and 2002. Peeters. Louvain. Colledge, S., Conolly J. and Shennan S. 2004. Archaeobo- tanical Evidence for the Spread of Farming in the Eastern Mediterranean. Current Anthropology 45: 35–58. Dekker K. 2014. What tools can tell. The Bone Tools of Barcın Höyük. Unpublished MA thesis. Free University of Amsterdam. Netherlands. Derin Z. 2012. Yesilova Höyük. In M. Özdogan, N. Basge- len and P. Kuniholm (eds.), The Neolithic In Turkey. Vo- lume 4. Western Turkey. Archaeology and Art Publica- tions. Istanbul: 177–195. D’Errico F., Giacobini G. and Puech P.-F. 1984. Varnish Re- plicas: A new method for the study of worked bone sur- faces. Ossa International Journal of Skeletal Research 9(11): 29–51. Düring B. 2013. Breaking the Bond: Investigating the Neolithic Expansion in Asia Minor in the Seventh Millen- nium BC. Journal of World Prehistory 26: 75–100. Erdalkıran B. 2017. Barçın Höyük 2015 Yılı Kemik Alet- lerinin Ön Raporu. Arkeometri Sonuçları Toplantısı 32: 235–249. Erdogu B. 2003. Visualising Neolithic Landscape: the early settled communities in Western Anatolia and Eastern Ae- gean Islands. European Journal of Archaeology 6(1): 7–23. 2014. Gökçeada Ugurlu Archaeological Project: A Pre- liminary Report from the 2011–2013 Field Seasons. Anatolica 40: 157–178. 2016. The Neolithic landscape and settlement of the Is- land of Gökçeada (Imbros, Turkey). In M. Ghilardi (ed.), Géoarchéologie des îles de Méditerranée. Centre Natio- nal de la Recherche Scientifique. Paris: 89–94. Gerritsen F., Özbel R. and Thissen L. 2013. Barcın Höyük: The beginnings of Farming in the Marmara Region. In M. Özdogan, N. Basgelen and P. Kuniholm (eds.), The Neoli- thic in Turkey. Volume 5. Northwestern Turkey and Istanbul. Archaeology and Art Publications. Istanbul: 93– 112. Goodarzi-Tabrizi S. 1999. The Worked Bone Remains of Hallan Çemi Tepesi, an Early Neolithic Site in South- eastern Turkey. Unpublished PhD thesis. University of California. Berkley. Griffitts J. L. 2011. Use wear analysis on Catalhoyuk bone tools. In S. Farid (ed.), Çatalhöyük 2011 Archive Report: Çatalhöyük Research Project: 52–66. http://www.catalho yuk.com/archive_reports/2011. Hodder I. 2005. Changing entanglements and tempora- lities. In I. Hodder (ed.), Changing materialities at Çatal- höyük: reports from the 1995–99 seasons. McDonald In- stitute for Archaeological Research/British Institute of Ar- chaeology at Ankara. Cambridge: 1–22. Höglinger P. 1997. Neolithisches Bein- und Geweigherät. In S. Hiller, V. Nikolov (eds.), Karanovo I.1, Die Ausgra- bungen im Südsektor 1984–1992. Verlag Ferdinand Ber- ger & Sohne. Salzburg and Sofia: 157–196. Karul N., Avcı M. B. 2011. Neolithic Communities in the Eastern Marmara Region: Aktopralık. Anatolica 37: 1–15. Kızıltan Z., Polat M. A. 2013. The Neolithic at Yenikapı. In M. Özdogan, N. Basgelen and P. Kuniholm (eds.), The Neo- lithic in Turkey. Volume 5. Northwestern Turkey and Is- tanbul. Archaeology and Art Publications. Istanbul: 113– 165. Legrand A., Sidéra I. 2007. Methods, Means, and Results when Studying European Bone Industries. In C. Gates St- Pierre, R. Walker (eds.), Bones as Tools: Current Methods and Intepretations in Worked Bone Studies. Archaeo- press. Oxford: 67–80. LeMoine G. M. 1997. Use Wear Analysis on Bone and Antler Tools of the Mackenzie Inuit. Archaeopress. Ox- ford. Leshtakov K., Todorava N., Petrova V., Zlateva-Uzunova R., Özbek O., Popova T., Spassov N. and Iliev N. 2007. Preli- minary report on the salvage archaeological excavations at the early Neolithic site Yabalkovo in the Maritsa Valley, 2000–2005 field seasons. Anatolica 33: 185–234. Lichardus-Itten M., Demoule J-P., Pernicheva L., Grebska- Kulova M. and Kulov I. 2006. Kova≠evo, an Early Neoli- thic site in South-West Bulgaria and its importance for Eu- ropean Neolithisation. In I. Gatsov, H. Schwarzberg (eds.), Aegean – Marmara – Black Sea: the Present State of re- search on the Early Neolithic. Beier and Beran. Langen- weissbach: 83–94. Lichter C. 2002. Central Western Anatolia – a key region in the Neolithisation of Europe. In F. Gérard, L. Thissen (eds.), The Neolithic of Central Anatolia: internal deve- lopments and external relations during the 9 th –6 th mil- lennia cal. BC. Ege Yayınları. Istanbul: 161–169. Ma ˘rga ˘rit M. 2017. Spatulas and braded astragalus: Two types of tools used to process ceramics? Examples from the Romanian prehistory. Quaternary International 438: 201–211. Marinelli M. 1995. The bone artifacts of Ilipinar. In J. Ro- odenberg (ed.), The Ilipinar Excavation I. Five Seasons of Fieldwork in Northwestern Anatolia, 1987–91. Neder- lands Historisch-Archaeologisch Instituut te Istanbul. Lei- den: 121–142. Marinova E. 2007. Archaeobotanical data from the early Neolithic of Bulgaria. In S. Colledge, J. Conolly (eds.), The Origins and Spread of Domestic Plants in southwest Asia and Europe Left Coast Press. Walnut Creek: 93–109. Mili≤ M. 2014. PXRF characterisation of obsidian from cen- tral Anatolia, the Aegean and central Europe. Journal of Archaeological Science 41: 285–296. Özdogan M. 2001. Yarımburgaz Cave Excavations. In O. Belli (ed.), Istanbul University’s Contributions to Archa- eology in Turkey (1932–2000). Istanbul University Rec- torate Research Fund. Istanbul: 8–11. 2010. Westward expansion of the Neolithic way of life: sorting the Neolithic package into distinct packages”. In P. Matthiae, F. Pinnock, L. Nigro and N. Marchetti (eds.), ICAANE 6, Proceedings of the 6 th International Con- gress of the Archaeology of the Ancient Near East. Vo- lume 1. Near Eastern Archaeology in the Past, Present and Future. Harrassowitz Verlag. Wiesbaden: 883–893. 2011. Archaeological Evidence on the Westward Expan- sion of Farming Communities from Eastern Anatolia to the Aegean and the Balkans. Current Anthropology 52 (4): 415–430. 2013. Neolithic Sites in the Marmara Region: Fikirtepe, Pendik, Yarımburgaz, Toptepe, Hoca Çesme, and Asagı Pınar. In M. Özdogan, N. Basgelen and P. Kuniholm (eds.), The Neolithic in Turkey. Volume 5. Northwest- ern Turkey and Istanbul. Archaeology and Art Publica- tions. Istanbul: 167–269. Perlès C. 2005. From the Near East to Greece: Let’s re- verse the focus Cultural elements that didn’t transfer. In C. Lichter (ed.), How did farming reach Europe? Anato- lian-European relations from the Second half of the 7 th through the first half of the 6 th millennium cal. BC. Ege Yayınları. Istanbul: 275–290. Pinhasi R., Fort J. and Ammerman A. J. 2005. Tracing the Origin and Spread of Agriculture in Europe. PLOS Biology 3(12): 2220–2228. Powell J. 1996. Fishing in the Prehistoric Aegean. Paul Åströms Förlag. Jonsered. Russell N. 2005. Çatalhöyük Worked Bone. In I. Hodder (ed.), Changing materialities at Çatalhöyük: reports from the 1995–99 seasons. British Institute at Ankara. London: 339–369. Saglamtimur H. 2012. The Neolithic Settlement of Ege Gü- bre. In M. Özdogan, N. Basgelen and P. Kuniholm (eds.), The Neolithic In Turkey. Volume 4. Western Turkey. Ar- chaeology and Art Publications. Istanbul: 197–225. Sampson A. 1998. The Neolithic and Mesolithic Occupa- tion of the Cave of Cyclops, Youra, Alonnessos, Greece. The Annual of the British School at Athens 93: 1–22. Schibler J. 1980. Osteologische Untersuchung der cor- taillodzeitlichen Knochenartefakte. Die neolithischen Ufersiedlungen von Twann. Volume 8. Staatlicher Lehr- mittelverlag. Bern. Séfériadés M. 1992. Bone tool report. In R. Treuil (ed.), Dikili Tash, village préhistorique de Macédoine orien- tale, I. Fouilles de Jean Deshayes (1961–1975). Volume 1. Bulletin de Correspondance Hellénique Supplément XXIV. Athens: 99–112. Semenov S. A. 1964. Prehistoric Technology: an Experi- mental Study of the oldest Tools and Artefacts from tra- ces of Manufacture and Wear. Barnes and Noble. New York. Sidéra I. 2013. Manufacturing Bone Tools: The Example of Kova≠evo. In N. Miladinovi≤-Radmilovi≤, S. Vitezovi≤ (eds.), Bioarheologija Na Balkanu: Bilans i perspektive. Srpsko arheolo∏ko dru∏tvo. Sremska Mitrovica. Beograd: 173– 178. Simmons A. H. 2014. Stone Age Sailors: Paleolithic Sea- faring in the Mediterranean. Left Coast Press. Walnut Creek. Jarrad W. Paul, Burçin Erdog ˘u 384 An examination of the worked bone and antler assemblage at Ug ˘urlu (Gökçeada, Turkey) 385 Spangenberg J. E., Ferrer M., Jacomet S., Bleicher N. and Schibler J. 2014. Molecular and isotopic characterisation of lipids staining bone and antler tools in the Late Neoli- thic settlement, Zurich Opera Parking, Switzerland. Orga- nic Geochemistry 69: 11–25. Stratouli G. 1998. Bone artefacts from Makri. In N. Efstra- tiou, M. P. Fumanal, C. Ferrer, D. Urem Kotsos, A. Curci, A. Tagliacozzo, G. Stratouli, S. M. Valmomti, M. Ntinou, E. Ba- dal, M. Madella and K. Skourtopoulou (eds.), Excavations at the Neolithic settlement of Makri, Thrace, Greece (1988–1996): A preliminary report. Papeles del Labora- torio de Arqueología de Valencia. Saguntum: 36–40. Takaoglu T. 2013. Coskuntepe: A Neolithic Village in the Coastal Troad. In M. Özdogan, N. Basgelen and P. Kuni- holm (eds.), The Neolithic in Turkey. Volume 5. North- western Turkey and Istanbul. Archaeology and Art Pub- lications. Istanbul: 35–43. Thissen L. 1999. Trajectories towards the Neolithisation of NW Turkey. Documenta Praehistorica 26: 29–39. Tomkins P. 2004. Filling in the ‘Neolithic Background’: Social Life and Social Transformation in the Aegean be- fore the Bronze Age. In J. C. Barrett, P. Halstead (eds.), The Emergence of Civilisation Revisited. Oxbow Books. Oxford: 38–63. Vitezovi≤ S. 2011. The Neolithic Bone Industry from Dre- novac, Serbia. In J. Baron, B. Kufel-Diakowska (eds.), Written in Bones: Studies on technological and social contexts of past faunal skeletal remains. Institute of Ar- chaeology University of Wrocław. Wrocław: 117–136. Zidarov P. 2008. Pointed bone tools from Ma ˘gura Gor- gana. In S. Hansen, M. Toderas, A. Reingruber, I. Gatsov, F. Klimscha, P. Nedelcheva, R. Neef, M. Prange, T. D. Price, J. Wahl, B. Weninger, H. Wrobel, J. Wunderlich and P. Zi- darov (eds.), Der kupferzeitliche Siedlungshügel Ma ˘gu- ra Gorgana bei Pietrele in der Walachei. Ergebnisse der Ausgrabungen im Sommer 2007. Eurasia Antiqua 14. V. Zabern. Mainz: 57–62.