<<

and Protohistory

New excavations in Northwestern : The settlement of Avgi,

G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

with contributions by

N.H. Andreasen, E. Kalogiropoulou, D. Kotsachristou, Y. Maniatis, E. Margaritis, O. Metaxas, K. Papayianni, T. Theodoropoulou and V. Tzevelekidi

Corresponding Author: Georgia Stratouli Ephorate of Antiquities [email protected]

1. Introduction

Neolithic Avgi (Figure 1) is one of several Neolithic sites that have been recently investigated in . During the last decades, there has been an outburst of excavations in Western and Central , Greece that are changing our knowledge of the Neolithic period, to the point that it has no precedent in the history of research for any other region of Greece.1 These are intensive, large-scale excavations, with meticulous documentation of the stratigraphic information, and systematic collection of findings of all sorts.2 Many follow internationally- standardised sampling protocols and have a highly interdisciplinary rationale, with an emphasis on the collection of bioarchaeological, geoarchaeological, and micro-environmental proxies. The first visible result is high-quality excavation research, only a fragment of which is unfolded here.

The Neolithic settlement of Avgi is located at c. 740m asl, in a hilly terrain of clayey deposits (Figure 2), c. 10km southwest of and the city of Kastoria, close to the modern town of . The occupation dates to the Middle and , c. 5,700/5,500-4,500/4,300 cal BC, and exhibits clear patterns of a dispersed or ‘flat extended’ settlement3 across a total estimated area of c. 5.5-6.0 ha.

The site has been investigated for the past eighteen years.4 The interdisciplinary nature of the research programme,5 the extensive area excavated, the state of preservation of architectural features and the rich datasets of bioarchaeological finds allow for a thorough study of material culture, subsistence practices and landscape management, built environment and use of space, as as of multiple social dynamics involved. They offer insights into different aspects of the Neolithic community, and the opportunity to examine these within the wider cultural context of northern Greece and the southern . In particular, the information provided enhances our understanding of the Neolithic environment and lifeways in and beyond.

1 Kotsakis 2013; 2014a; Grammenos 2010. 2 Kotsakis 2014b: 710. 3 cf. Andreou and Kotsakis 1987, 1994; Kotsakis 1999. 4 cf. Stratouli 2004a, 2004b, 2005, 2007, 2012a, 2012b, 2012c, 2013; Stratouli and Bekiaris 2008; Stratouli et al. 2009, 2010, 2011, 2013, 2014a. 5 For an overview regarding the archaeological programme at Neolithic Avgi as well as for the preliminary results of individual studies visit the official website (www.neolithicavgi.gr).

Journal of Greek 5 (2020): 63–134 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 1. Map of Northern Greece, showing the location of Avgi.

Figure 2. The site and its immediate landscape (Photo: George Vlachou).

64 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

In this paper we aim to summarise and synthesise the results of the archaeological work conducted at Neolithic Avgi to date. We discuss the chronology of the site, in the light of new radiocarbon evidence, and review the excavation data within the scope of the chronostratigraphic scheme that was developed by the contextual correlations and radiocarbon dating of the occupational remains. In our synthesis, we present primary data deriving from the contextual analysis of several categories of the archaeological material, most of which are unpublished. Contributions from several team members briefly discuss subsistence practices based on the study of plant and faunal remains, as well as different aspects of various material culture categories.

2. History of research

The archaeological site was discovered in the mid-1990s as a result of non-systematic surveys conducted by the local Ephorate of Antiquities and the excavation team of the nearby Neolithic site of .6 The abundance and distribution of surface finds, particularly fragments, burnt daub fragments and grinding , indicated a significant Neolithic occupation c. 500m north of the modern village of Avgi. The archaeological research in the wider region of Kastoria has brought to light a number of Neolithic sites (Figure 3), notably, the lakeside settlement of Dispilio,7 the riverside site of Kolokynthou,8 the site of Kastro Nestoriou9 at c. 900m asl, the riverside site of Koromilia,10 as well as a in the nearby gorge.11

The excavations onsite commenced in 2002 and continued for six consecutive field seasons until 2008. They were conducted by the former ΙΖ΄ Ephorate of Prehistoric and Classical Antiquities of the Hellenic Ministry of Culture, under the directorship of Dr Georgia Stratouli. A total area of 2000 m2 was uncovered (Figure 4). Due to its large extent, the site (Figure 5)

Figure 3. Google Earth map of the Kastoria region, showing the major neolithic sites.

6 Stratouli 2013; Touloumis 2002: 93. 7 Hourmouziadis 2002. 8 Tsouggaris et al. 2004. 9 Tsouggaris 1999: 25-26. 10 Stratouli et al. 2014b. 11 Trantalidou et al. 2005, 2010.

65 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 4. The excavations at Neolithic Avgi (Photo: George Vlachou)

Figure 5. The site, the ditches and the excavated area (Plan: Tasos Bekiaris).

was divided into three Excavation Sectors (Western, Central and Eastern Sector), and different Excavation Areas (Areas 1-8, Areas N5 and E5) within each sector (Figure 6). The excavation grid was a system of 4x4m arbitrary trenches, with 4x1m baulks in between.

The first soundings were carried out in 2002 and were followed by rescue excavations during the subsequent two years.12 During these early stages, the focus was on the protection of the site from further destruction due to mechanical ploughing, and on the documentation of its chronological range and major stratigraphy. To obtain a more accurate estimate of its size, trial trenches were excavated at different locations across the site. In 2004 and 2005 Dr Grigoris Tsokas, Professor of Exploration of Geophysics at the University of , and his team conducted conducted a geomagnetic survey, in an area of c. 3.5 ha.13 Τhe geophysical survey provided a better- grounded view of the extent of the site and the density of structures within it (Figure 7). The

12 The results of the first seasons were published in Stratouli 2004a, 2004b, 2005, 2012a, 2012b, 2012c. 13 Tsokas et al. 2005.

66 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figure 6. The excavated part of the site, the different sectors and areas (Plan: Tasos Bekiaris).

Figure 7. The results of the geophysical survey.

67 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 8. Public actions of the Archaeological Project Neolithic Avgi. Top left: Georgia Stratouli, the Project’s director, presents the excavation to the locals. Top right: Experimental grinding by Elementary School students in the context of the Educational Programme ‘Learning about food preparation at the Neolithic Settlement of Avgi’. Bottom left: Presentation of the Educational Programme ‘At the Dawn of the First Farmers’ at the yard of the excavation’s laboratory at Avgi, Bottom right: Simulation of a prehistoric excavation for Pre-school kids (Photos: Tasos Bekiaris).

magnetic anomalies, probable burnt remains of earthen structures and imprints of cuttings, were quite dense at the centre of the settlement and rather dispersed at its periphery. In 2005, the excavation methodology was enhanced, by the application of high-resolution micro-analytical techniques14 to the study of the stratigraphy, and the excavation of building remains in 1x1m micro-grids in 2006-2008.

From 2005 to 2008, the project was funded by the European Programme INTERREG IIIA/Cards Greece-, whilst from 2006 to 2008 by the Programme ‘PINDOS’. This funding critically enabled the more systematic and interdisciplinary nature of the research and the integration of new techniques.15

Three extended field seasons in May-October 2006-2008 were followed by more or less consecutive study seasons in 2009-2019 that focused on the analysis of selective archaeological assemblages.16 Research on several categories of archaeological, bioarchaeological and geoarchaeological material,17 has offered valuable contextual insights, into the material culture and , onsite activities and spatiotemporal patterns in the use of space, dietary practices, social networks, management of natural resources, and burial practices, all of which considerably enrich the archaeological record of northern Greece and the Greek Neolithic.

Throughout the course of the programme, public and community archaeology initiatives have constituted an integral part of its objectives (Figure 8). Priority was given to the dissemination of the results of the archaeological research to a wider public and to the active engagement of the local community into the archaeological process. Two small exhibitions of representative finds and photographic material were organised at the Documentation Centre of the excavations at Avgi, while more detailed information regarding the project and its public actions became available through social media and the programme’s official website.18 In addition, more than 5,000 preschool and elementary school students from the Prefecture of Kastoria and other regions of Greece, as

14 The pioneering application is outlined in Kyrillidou 2016. 15 Stratouli 2007, 2013. 16 Selected material studies were funded by The Institute for Aegean Prehistory (INSTAP) in 2011-2020, The Psycha Foundation in 2011- 2013, and The I.F. Costopoulos Foundation in 2011 and 2013. 17 e.g. Andreasen 2011; Bekiaris 2018; Kalogiropoulou 2013; Kloukinas 2014; Katsikaridis forthcoming; Koromila 2015; Kyrillidou 2016; Margaritis 2007; Ntinou 2008; Papayianni 2008; Tzevelekidi 2013; Papadias and Saridaki 2016; Stergiou and Theodoridou 2017. 18 http://www.neolithicavgi.gr.

68 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

well as undergraduate and postgraduate archaeology students from the Aristotle University of Thessaloniki, have participated in a series of educational programmes between 2009 and 2015.19

3. The environment

The site of Avgi is situated in an inter-mountainous terrain at the basin of the Mesohellenic Trough between high formations of the Pindos Range. The surrounding landscape comprises both mountainous and semi-mountainous areas, as well as river valleys of the Aliakmon River and its tributaries.

The geoarchaeological study of the landscape reconstruction conducted by Dr Athanasia Krachtopoulou,20 has identified that the settlement developed on the gentle slopes of a wide terrace, bounded by a -Early stream to the North, and by a second stream- course, which was active during the Classical period, to the South. Both streams have deposited alluvial sediments, forming three discrete river terraces during the Late Quaternary. The Neolithic settlement was founded upon the older terrace, which is preserved in the form of a wide, gently undulating surface and was seriously affected by erosion prior to the Neolithic occupation.21 The second river terrace was most likely formed during the late Pleistocene-early Holocene. During the Neolithic period, this was possibly still an extended active floodplain, suitable for cultivation and other activities. The younger river terrace is much later and was probably formed later than AD 670- 830. Geomorphic processes, such as stream migration and hilltop erosion, as well as anthropogenic activity, including arable cultivation, have seriously influenced the preservation and morphology of the site, now resembling a hill.

Contrary to the modern landscape, which is characterised by extensive cultivation, the Neolithic landscape was more densely covered by vegetation. The preliminary results of the charcoal analysis conducted by Dr Maria Ntinou22 point to a wide spectrum of plant species that are dominated by oak and black pine. The available data suggest that oak woodlands including various deciduous species were dominant in the immediate surroundings of the settlement, while hydrophilic vegetation, comprising willow/poplar, elm and hazel were growing in nearby humid woodland areas and streams. Forests of black pine and (sparse) fir were present in higher altitudes, whereas the zones of altitudinal transition were dominated by mixed deciduous and coniferous woodlands. The evidence is, by and large, consistent with those provided from the neighbouring Neolithic site of Dispilio,23 as well as with the palynological evidence from Lake Orestis.24

The study of the bioarchaeological remains25 has indicated that the local environment was suitable for the cultivation of cereals and pulses, such as wheat and lentils. Moreover, the study of the faunal assemblage suggests the grazing of sheep, goats, pigs and cattle.26 The natural environment also favoured hunting, fishing, and the gathering of fruits, while a wide range of raw materials was accessible through local outcrops for and artefact manufacture, as well as for construction practices.

The site is located in Miocene molasse sediments, which incorporates mainly sandstones, marls and conglomerates.27 Other rocks suitable for tool-making belonging to the Pelagonian Zone are located at the north/northwestern parts of the region. They consist mostly of gneisses, schists and

19 Stratouli et al. 2012, 2013. 20 Krachtopoulou 2009. 21 For a comparable situation from see: Bintliff et al. 2006. 22 Ntinou 2008. 23 Ntinou and Badal 2000; Ntinou 2002, 2010. 24 Bottema 1974; Kouli 2002: 308-310. 25 Margaritis 2007. 26 Tzevelekidi 2013. 27 For a detailed description of the geology and lithic sources of the region see Bekiaris et al. 2017: 416-418; Stergiou et al. forthcoming.

69 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

amphibolites. These bedrocks are interrupted by Permo-Triassic granites and overlain by Triassic– Jurassic recrystallised limestones. Less extensive are the ophiolitic rocks from the bordering /Axios zone, overthrusted to the west on the Pelagonian zone. Suitable soils and sediments for pottery production and the construction of dwellings and earthen structures were available in the immediate environment of the site. These comprise the natural soil and the clay sandy marls exposed in the surrounding areas.

4. Phasing and Stratigraphy (by G. Stratouli, T. Bekiaris, Y. Maniatis, G. Koromila, S. Kyrillidou, D. Kloukinas)

The absolute chronology of the site has been determined by a series of 26 radiocarbon dates on carbonised seeds (14 samples) and charred wood (12 samples). Thirteen samples were analysed at the Curt-Engelhorn-Centre for Archaeometry at Mannheim, , in collaboration with the Laboratory of Archaeometry, National Centre for Scientific Research (NCSR) Demokritos at , Greece; eleven samples were analysed at the Centre for Isotope Research in Groningen, Netherlands; and three at NCSR, Demokritos, Athens. The dates were calibrated using the program OxCal v.4.3.228 with the latest available dataset of 2013.29

14 Sample Code Sample type Context C Age (BP) Calibrated date 1σ Calibrated date 2σ (68,2%) (95,4%) DEM – 3071 Seeds Building 7 6782 ± 27 5711-5660 5721-5637 (MAMS-30666) GrN-28445 Charred wood Building 2a 6660 ± 40 5625-5535 5658-5484 LIH-623 Charred wood Building 1 6509 ± 76 5600-5370 5620-5320 DEM-3294 (MAMS- Seeds Building 7 6577 ± 29 5541-5486 5610-5480 39839) GrN-29857 Charred wood Building 2a 6430 ± 90 5475-5327 5550-5220 Mouse 1713930 Mouse Building 1 6375 ± 60 5466 - 5308 5474-5288 mandible (Mus spicilegus) GrN-28446 Charred wood Building 5 6340 ± 50 5460-5260 5470-5150 GrN-29856 Charred wood Building 1 6340 ± 30 5363-5302 5461-5222 GrN-29858 Charred wood Building 2a 6320 ± 30 5325-5227 5359-5222 GrN-29855 Charred wood Building 1 6310 ± 35 5321-5226 5357-5218 GrN-30684 Charred wood Layer 2 6210 ± 80 5300-5060 5340-4950 DEM – 3072 Seeds Open area (Thermal 6255 ± 24 5296-5218 5308 – 5209 (MAMS-30667) structure ΑΕ 253101) DEM – 3073 Seeds Open area 6271 ± 25 5299-5226 5307 – 5216 (MAMS-30668) DEM-3288 (MAMS- Seeds Building 2a 6265 ± 21 5296-5221 5303-5217 39833) GrN-30682 Charred wood Building 5 6220 ± 25 5285-5077 5296-5069 DEM – 2483 Seeds Layer 2 6194 ± 25 5213-5077 5221-5054 GrN-28447 Charred wood Layer 2 6130 ± 40 5206-4961 5221-4870 DEM – 3075 Seeds Open area (Thermal 6191 ± 25 5212 – 5078 5219 – 5056 (MAMS-30670) structure ΑΕ 252301) DEM-3290 (MAMS- Seeds Pit 194101 6189 ± 22 5211-5078 5217-5058 39835) DEM-3293 (MAMS- Seeds Pit 109401 6183 ± 21 5209-5077 5214-5057 39838) DEM-3292 (MAMS- Seeds Burial pot ΑΑ 225409 6138 ± 22 5204-5011 5207-5002 39837)

28 Bronk Ramsey 2017 29 Reimer et al. 2013. 30 The mouse bones were radiocarbon dated during the post-doctoral project named MOUSETRACK, for which Dr Katerina Papayianni was the main postdoctoral researcher. The project was hosted at the Zooarchaeology Laboratory (UMR 7209 du CNRS) of the Museum of Natural History of Paris (France) under the direction of Dr Thomas Cucchi, and it was funded by the Agence Nationale de la Recherche under the LabEx ANR-10-LABX-0003-BCDiv (see Cucchi et al. 2020).

70 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

DEM-3289 (MAMS- Seeds Building 5 6121 ± 22 5193-5001 5206-4986 39834) DEM – 2489 Seeds Pit 282302 5999 ± 25 4932-4846 4959-4801 GrA-30214 Charred wood Pit 253201 5925 ± 40 4840-4728 4927-4711 DEM-3291 (MAMS- Seeds Stratum 2? 5819 ± 21 4719-4619 4765-4602 39836) DEM – 2942 Seeds Burial pot ΑΑ 225405 5801 ± 30 4708-4616 4721-4555 (MAMS-26698) GrN-30683 Charred wood Pit? 5710 ± 70 4670-4460 4720-4370

Table 1. Radiocarbon dates and contexts from Neolithic Avgi.

The results of the radiocarbon dating are summarised in Table 1. Samples DEM-3071, DEM-3294, GrN-28445 and LIH-623 have provided the earliest range and date the initial occupation of the site at around 5,700-5,600 cal BC, during the Middle Neolithic period. The majority of the dates, however, fall within the range of c. 5,500-4,600 cal BC, indicating a possible habitation range of approximately a thousand years, throughout the Late Neolithic period. This is in consistent with the evidence from pottery. Bayesian analysis modelling was performed on the dates by Dr Yannis Maniatis using a three-phase model, Avgi I, Avgi II, and Avgi III. The overall agreement of the model is very good, 89%. The numerical output of the model for the three phases is presented in Table 2. The occupation of the site seems to start at a modelled mean date of around 5680 cal BC and the first phase is the longest one, covering almost 500 years. The settlement was likely abandoned around the mid-5th millennium.

Age Range Phase Mean Median Consensus (95.4%) Start of Avgi I 5,776 – 5,633 5,693 5,686 5,680

Transition Avgi 5,199 – 5,097 5,152 5,155 5,150 I to II

Transition Avgi 5,173 – 5,061 5,115 5,114 5,100 II to III End of model 4,769 – 4,248 4,567 4,607 4,600

Table 2. Plot of the output of three-phase modelling of the calibrated dates using the program Oxcal v.4.3.2 (Bronk Ramsey 2017).

71 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

The radiocarbon dates from Avgi are in accordance with the broader chronological schemes that have been developed for the Neolithic period in northern Greece and .31 Northern Greek sites that provide similar dates or are commonly considered as more or less contemporary with Avgi include Makriyalos in ,32 Toumba Kremasti Koilada,33 Megalo Nisi Galanis,34 Kleitos35 and LN Servia36 in , Dispilio37 and Koromilia38 in Kastoria.

The excavations have led to the distinction between three key stratigraphic phases (Avgi I-III). Their vertical superimposition is best attested in the Western Sector. The occupational sequence varies in thickness from a maximum of 1.10-1.30m to a minimum of 10cm in certain parts of the site.

The earliest phase of Avgi (Avgi I) dates to the 6th millennium (5,700/5,600 – 5,200/5,100 cal BC). This occupation develops upon a very firm light brown clay loam, which constitutes the natural soil of the site. This basal layer has served as a constant, secure guide for correlations among separate stratigraphic sequences across site from the base up. It has been observed, however, that the upper boundary of the natural sequence exhibits undulations, forming gentle rises and depressions. It is, therefore, possible that active surfaces that were contemporaneous existed at different elevations, at least at the early stages of the site’s history. Due to frequent lateral discontinuities in the stratigraphy, we speculate that this configuration was to some extent the result of anthropogenic intervention, by reshaping the natural surface into terraces to accommodate spatially distinct activity.

The detailed microstratigraphic study that was conducted by Stella Kyrillidou39 has identified that there are marked truncations throughout the settlement history, as well as redeposition of truncated virgin soil and collapse deposits in other parts of the site, which were used as the new foundations of occupation. The microstratigraphic study strongly suggests that the horizontal shifting of the occupation is recorded through the existence of low relief and slightly overlapping terraces across the site. This indicates that, despite the relatively small thickness of the occupational sequence, the formation of the settlement was extremely complex.

The features and sequences that are founded directly upon or are close to the natural foundations of the site have been designated as Avgi I. These include the collapsed burnt superstructure of buildings and other earthen structures, as well as more massive accumulations in the extensive open areas in between (Figures 9 and 10). It is noted that, based on stratigraphic observations and radiocarbon dating (cf. Table 1 and Table 2) the buildings that have been assigned to Avgi I were not all contemporary.

The earliest occupational remains are followed by more homogenised layers, which seem to represent a transition to different sedimentation and preservation processes in the Western Sector. These layers are also clearly anthropogenic, with abundant evidence of intense activity and accumulation of materials, which were assigned to the phase of Avgi II - samples from these deposits date at 5,200/5,100 – 4,900 cal BC. The deposits of Avgi II consist of variations of a largely undifferentiated thick brown layer, with a diffused basal boundary and no traces of any internal layering, which stratigraphically exists at the end-life of unearthed buildings. Nonetheless, some constructed features, such as fire installations, have been identified in all three sectors of the excavations, and these may also be attributed to this phase.

31 For the periodisation of the Neolithic of Northern Greece see Andreou et al. 2001: Table 1; Tsirtsoni 2016: Table 1; Maniatis 2014; Maniatis et al. 2016; Reingruber et al. 2017. 32 Maniatis and Pappa 2020. 33 Hondroyianni-Metoki 2009. 34 Fotiadis et al. 2019. 35 Ziota 2014. 36 Ridley et al. 2000. 37 Facorellis and Maniatis 2002. 38 Stratouli et al. 2014b. 39 Kyrillidou 2016.

72 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figure 9. Stratigraphic profile at an open area of the Western Sector (Photo: Tasos Bekiaris).

Figure 10. Stratigraphic profile at Area 2 of the Western Sector (Photo: Tasos Bekiaris).

The uppermost part of Avgi II is truncated, as indicated by the sharp and unusually flat basal boundary of the overlying topsoil throughout the entire excavated part of the site. The latest activities at the site, therefore, are only partially known through the negative imprints of features, such as pits, postholes and the foundation trenches of buildings that were cut through sequences of earlier deposits. Although not necessarily contemporary, these features were assigned to the phase of Avgi III (4,900 – 4,500/4,300 cal BC). The massive deposits of Avgi II and the cuttings of Avgi III are overlain by the topsoil, which is often 20-30cm thick, and significantly disturbed by erosion and mechanical ploughing.40 Although the topsoil also contains abundant archaeological material, this is reworked and out of context.

4.1 Formation processes and taphonomy, by Georgia Koromila

It is evident that a range of intersecting formation processes were operating at Avgi, from short- term episodes of activity and accumulation to longer-term and more complex, cumulative, anthropogenic and natural processes of accretion and re-organisation of deposits.

40 Krachtopoulou 2009.

73 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Macro- and microstratigraphic analyses suggest that an important distinction in deposit formation can be made between Αvgi I and Avgi II: The sediments of Avgi I preserve more episodic remains, which are often concealed inside the burnt rubble concentrations and other thin deposits and lenses (e.g. ash lenses, discussed in more detail in Section 5); better preservation of episodic deposits in the early phases may be indicative of rapid burial due to intensive, spatially focused activity. In contrast, the heavily homogenised sediments of Avgi II were formed by accumulation of remains with more complex histories, such as dispersion, disaggregation, and weathering due to prolonged surface exposure, trampling, truncations, and natural processes such as slope, water, and wind transportation; other homogenising factors include post-burial vertical movement due to bioturbation by root action and micro- and meso-faunal burrowing, as well as fine particle translocation and calcite reprecipitation due to water action, observed in thin section.

It is, therefore, evident that the temporal and spatial resolution of the deposits of Avgi I is better suited to microscalar approaches, whereas the data of Avgi II are more cumulative in nature. A key implication of this comparison is that the information obtained is qualitatively distinct between different deposit types and, therefore, caution should be applied to comparative analysis of contexts due to taphonomy.

In contrast, post-burial taphonomic processes seem to have affected in similar ways all stratigraphic sequences and remains; namely, decay (and, rarely, mineralization) of organic matter, calcite reprecipitation, and mechanical bioturbation are the key post-depositional processes observed. Partially altered calcareous materials such as calcitic ash, plaster, and natural sediments were still recognisable in thin section, although the distinction between these and neoformed calcite is challenging macroscopically. The decay of organic materials produced altered features such as dark brown or yellowish green amorphous staining, as well as rare occurrences of vivianite at the basal levels of the anthropogenic stratigraphy; vivianite formation41 was likely enabled by prolonged water saturation just above the poorly-drained natural substrate. This observation provides supporting evidence to a hypothesis based on topographic field observations that buildings at Avgi were constructed on low rises for better drainage, while depressions were used as open areas and .

4.2 Field microstratigraphic and thin-section micromorphological study, by Stella Kyrillidou

As most of the architectural remains were preserved in the form of rubble, and the open spaces were also diffused and massive, it was difficult to determine the internal anatomy, plan and arrangement of this settlement solely on the basis of conventional excavation methods, regardless of how systematic these methods might have been. High-resolution microstratigraphic and thin- section micromorphological study was integrated into the excavation practice of Neolithic Avgi from 2005, in order to provide more robust archaeological interpretations on the nature of its contents, and to enable the study of micro-contexts.42

A wide range of contexts was examined that comprised suspected houses, open areas next to, underneath and above buildings, and constructed features. In these spaces, the nature of deposits sampled and studied comprised suspected floors, accumulated residues from use on floors, other accumulated residues, for example, in -like areas, laid surfaces other than floors (meaning surfaces that were prepared and laid, but not trampled), foundations and fills, collapse deposits,

41 See Karkanas and Goldberg 2010 on the formation of vivianite. 42 This micro-contextual investigation was inspired by the extensive research of Wendy Mathews in the e.g. Matthews 1995, 2001, 2005a, 2005b. In addition to funds and resources provided by INTERREG IIIA/Cards Albania–Greece, it was funded by the Malcolm H. Wiener Laboratory of the American School of Classical Studies at Athens (ASCSA), The Greek Archaeological Committee in the UK (GACUK), and Mrs Matti Egon in particular. Professor Kostas Kotsakis, Director of the research project Thales, EXPLORING (Exploitation of Resources in the Neolithic Period in Northern Greece. Material Culture and Environment, 2012–2015) at the Aristotle University of Thessaloniki kindly provided access to exceptional polarising facilities.

74 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

and select construction materials. A holistic approach was advanced that does not separate indoor from outdoor spaces on the basis of excavation plans, as these were thought to be archaeological constructions that may depart significantly from the nature of a space, when this is examined at higher resolution temporal scales.

Microlayers, c. 2-3cm thin were examined, rather than archaeological layers, or excavation units, to study microhistories and events. These events were examined in large resin-impregnated sections, c. 14x7cm large and 30 microns thin, which were prepared and studied through sequences of spaces to enable knowledge of the nature of activities and their magnitude and significance, repetition, intensity and frequency of use. The preservation of micromorphological variables at Neolithic Avgi was outstanding, and comprises, for example, wood ash, druses, siliceous phytoliths of monocotyledons, phytoliths of dicots, dung spherulites, and possible coprolites, which complement the suite of macroscopic finds. Regrettably, due to small size these minute plant, animal and human residues often escape excavation, and they are not retrieved by wet-sieving and flotation methods because of their non-carbonised nature.43

A range of activities was identified that were not visible macroscopically, and new contexts emerged under the microscope, by study of the micro-contextual associations and life histories of individual components, in intact micromorphological sequences of excavation contexts. Some of these comprise, for example, concealed cultivation plots and plant processing areas, food preparation, and consumption inside and outside buildings, animal penning inside buildings, fire installations and platforms at the end-life of buildings, and possible symbolic contexts.

5. Built environment and use of space

The 2002-2008 excavations have provided ample evidence for the organisation of the built environment and its transformation through time. Below we examine the spatial and contextual excavation data along with contextual observations derived from specialised analyses of the finds in an attempt to reconstruct some key patterns in the organisation and use of settlement space.

5.1 The outer ditches

Α system of enclosures, comprising two ditches, was identified during the geophysical survey of the site in 2004 and these were further investigated by trial trenches in 2005.44 Both ditches are better preserved at the western part of the site (Figure 11). The outer Ditch A is a U-shaped, curvilinear which follows a roughly South-North orientation and then turns eastwards, enclosing the northern part of the site, as indicated by the geophysical survey. Its width ranges between c. 5.5m and 11m45 and its measured preserved maximum depth reaches 2.15m from the topsoil down to the natural sequence of the site. Trial trenches that were excavated in 2013 in the southern, partially eroded part of the site revealed traces of a second-deep cutting, which on the basis of similarities in size, shape and nature of fill deposits was attributed to the extension of Ditch A towards the south.

An inner Ditch B was identified approximately 10m east of Ditch A, which was oriented almost parallel to the western part of Ditch A. The section of this ditch has a W shape, or two intersecting cuts. The westernmost part is U-shaped, c. 1.2m wide and almost 1m deep, while the easternmost part is basin-shaped and c. 1.4m wide and 1.2m deep. The total width of Ditch B is approximately

43 Matthews 2010; Shillito 2014; Shillito et al. 2011. 44 Stratouli 2005; Tsokas et al. 2005. 45 The maximum width was only estimated according to the results of the geomagnetic survey. It is yet to be confirmed by trial trenches.

75 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 11. The cross- sections of Ditch A (Top) and Ditch B (Bottom, Photos: Tasos Bekiaris).

2.8m, and its preserved total length is c. 18.45m. Its fill comprised various organic and inorganic remains, while daub fragments were particularly abundant in the upper parts of the fill.

Based on the preserved extent and configuration of these two features, we speculate that the outer Ditch A was probably enclosing the whole site, while the less sizeable Ditch B may have been used to reinforce the western boundary of Ditch A. No residues were identified in situ to enable the exact chronology of these enclosures. Therefore, their relation to the settlement remains unclear. 46

5.2 Avgi I (5,700/5,600 – 5,200/5,100 cal BC)

As already mentioned, the occupation at Avgi I is characterised by the remains of buildings that are preserved in the form of dense concentrations of fire-hardened daub. The rubble areas are separated by more or less extended open spaces with midden-like deposits that are rich in anthropogenic material. These two basic features of the built environment are inextricably linked when it comes to daily life and the socio-economic organisation of settlement space.

5.2.1 The buildings of Avgi I

Building remains that were assigned to Avgi I have been revealed in all sectors of the site (Figure 12), but they are not fully excavated. The suspected dwellings were detached and above ground structures following roughly square or rectangular ground plans. Due to the state of preservation it is not easy to define their orientation. Based mainly on the uncovered floor layouts and occasional rows of postholes, the buildings do not always seem to share a common orientation. Most of them, however, could have followed either a West-East or a North-South axis. In the section that follows we briefly present the results of the excavations and the integrated thin-section micromorphological study, beginning with the Western sector of the site, where the excavations were more advanced, and moving to the Central and the Eastern sectors.

46 Enclosures and double ditches, with or without palisades, are often reported in Greek and Balkan Neolithic settlements (Kalicz and Raczky 1987: 106-111; Whittle 1996: 174-176; Chapman and Gaydarska 2006).

76 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figure 12. Plan of Avgi I, indicating the rubbles, the open areas and the thermal structures attributed to this phase (Plan: Tasos Bekiaris).

At the Western Sector of the site, two rubble areas have been uncovered, which were denoted as Building 2a and Building 5 and are separated by extensive open spaces (Figure 13). The rubble of Building 2a (Figure 14) covers a total area of c. 65-70m2 but it has been severely disturbed by later cuttings that were assigned to the later Building 2b of Avgi III, as well as by a series of pits that were cut through its western limit.47 In the less disturbed part of Building 2a, a large grinding slab was found along with a storage or cooking vessel and small concentrations of agricultural products (triticum dicoccum and lathyrus sativus-cicera). Evi Margaritis has identified that these products were at an already processed state, with their hulls removed, while some of them have been also submitted to pounding or coarse grinding and may represent foodstuffs in the form of bulgur wheat.48

Approximately 12m southwest of Building 2a the rubble of Building 5

. The Western Sector of the excavations (Photo: George Figure 13 47 Vlachou). Stratouli and Bekiaris 2008. 48 Margaritis 2007.

77 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 14. Building 2a, Western Sector. Top: the poorly preserved rubble (Photo: George Vlachou). Bottom: A grinding slab resting on the building’s floor (Photo: Tasos Bekiaris).

Figure 15. The rubble of Building 5, view from the North (Photo: George Vlachou).

(Figure 15) is less disturbed by post-depositional modifications. Nevertheless, later cuttings, such as ditches, pits and postholes are present here as well. The rubble area alongside the partially excavated surface of the underlying floor suggest a roughly rectangular ground plan, which has an East-West orientation and measures c. 75-85m2. The postholes identified seem to define an area of similar size and orientation. Fragments of probable thermal structures were identified inside the collapse as well as on the floor of the building, but they were not found in situ.49

The eastern part of this building is truncated. Residues from the interior of Building 5 have been deposited in the open area to the East, next to the building. The thin-section micromorphological study has further revealed that Building 5 had an upper storey, on the basis of floor fragments that

49 Kloukinas 2014.

78 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figure 16. Left: scan of large-format thin-section TS 2824MB3 through the sequence of Building 5. Top right: micrograph showing a variegated loamy sand with moderately oriented burnt clays and charred plants in reflected light (RL). Bottom right: micrograph showing dung spherulites in cross-polarised light (XPL). are laid and trampled and were identified above the ground floor level, inside the collapse. Some of the activities that have been identified in the roofed excavated part of the building comprise possible animal penning on the ground floor level, in the excavated area to the East, and probable storage of grains in the loft or upper storey (Figure 16).50

At the Central and Eastern Sectors of the site (Figure 17), the concentrations of burnt and collapsed structures are denser. The rubble areas are in closer proximity to one another, and they sometimes overlap, obscuring the identification of the material boundaries of suspected individual buildings. Rather than the result of diverse attitudes towards space, the density of architectural remains seems to relate to the practices of both horizontal and partially overlapping relocation of buildings and other structures, across space through time. As has been identified by the combined field microstratigraphic and thin section micromorphological study, these practices of relocation and settlement formation are more evident in the Central-Eastern Sectors due to rapid burial.

Three complex rubble areas have been partially excavated in these Sectors, which were denoted as Building 3, Building 7 and Building 1, and are briefly discussed here.

The rubble of Building 3 (Figure 18) is situated in the Central Sector of the site and is c. 85m2. Strategic test trenches have indicated the presence of two superimposed levels of which the uppermost one

50 Kyrillidou 2016.

79 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 17. The Eastern and Central Sectors of the excavations (Photo: George Vlachou).

Figure 18. The rubble of Building 3, Central Sector (Photo: George Vlachou).

may be contemporary to Avgi II in the Western Sector, on the basis of stratigraphic correlations. The architectural remains comprise burnt clays that are collapsed with large negative imprints of wood facing down, but also other burnt clays and constructed features that are in situ. Two contexts in particular have been identified to date in this area by high-resolution microstratigraphic and micromorphological study.51 The first is a cooking that was used for meat preparation and consumption in an open space next to the collapsed western wall of the building. The second is a platform that was used for repetitive firing events, next to two large plinths, in the northeast part of the suspected dwelling rubble, at the end-life of the suspected building.

Less than 5m northeast of Building 3 is the rubble of Building 7 (Figure 19), which covers an area of c. 45-55m2. Although it is almost adjacent to the rubble of Building 1 and exists at a higher level, Building 7 predates Building 1, on the basis of the field microstratigraphic study.52 This observation was also confirmed by the recent radiocarbon dates, which indicate that Building 7 is one of the earliest excavated structures onsite (see Table 1). Interestingly, Building 7 reveals a more extensive application of the -and-daub technique compared to other buildings at Avgi I.53 This difference in building methods may also be attributed to its early date. The spread of the

51 Kyrillidou 2016. 52 For the spatiotemporal interrelation of the two adjacent buildings and the formation processes of this area see Kyrillidou 2016. 53 Kloukinas 2014.

80 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figure 19. Building 7, Central Sector. Top: the rubble and the floor of the building. Bottom: whole pots on the building’s floor (Photos: George Vlachou).

collapse layer suggests an oblong, roughly rectangular structure, which follows an approximate North–South orientation. The sequence of the building comprises at its foundations a yellowish clay loam, which resembles the texture and composition of the underlying natural soil but is richer in calcareous aggregates. Obviously, some sort of ground preparation was required, and this layer perhaps reinforces the idea that at least small-scale levelling practices were occasionally employed. A large number of sherds and intact pottery vessels of various shapes and sizes were uncovered on the floor of the building, suggesting that numerous activities were taking place indoors, or that the vessels were stored inside the building when not in use.54 Grinding tools, many polishers, abraders and edge tools comprise the inventory of Building 7.55

The rubble of Building 1 (Figure 20) is >65m2 in area, as it extends into yet unexcavated building remains that were uncovered during the last excavation season further east. As is the case with Building 3, it comprises two partially superimposed levels in the south and eastern part of the rubble. Rows of post holes and preserved burnt timber posts have been identified across the surface of this floor. Their orientation from NW to SE suggests that the building probably followed a similar orientation. Two cooking pans (plateaux) were uncovered in the NW part of the rubble, whilst remains of a burnt mouse were retrieved, suggesting storage of grains.56 Burnt grinding tools, edge tools and polishers were also found within Building 1.57

54 Katsikaridis forthcoming. 55 Bekiaris 2018: 329. 56 Papayianni 2008. 57 Bekiaris 2018: 329.

81 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 20. Building 1, Eastern Sector. Top: the rubble of the building. Bottom: a burnt cooking pan enfolded within the collapsed walls of the building (Photos: George Vlachou).

The thin-section micromorphological study has further identified a plant processing area, alongside short-term storage, in an unenclosed roofed yard, which is concealed inside the uppermost rubble in the SE part of the building, and possible cooking and animal penning in the enclosed roofed parts to the north. Significantly, although it gives the impression of a building by excavation, the lowermost rubble is redeposited in this location. This daub layer has neither burnt nor collapsed in situ. A midden-like area was identified at the foundations of this building, which suggests that the lowermost level of this sequence does not represent a building.58

5.2.1.1 Building , by Dimitris Kloukinas

The vast amount of information on building technology derives from the numerous fire-hardened daub fragments that were collected during the excavations. Postholes, remains of charred timber and other features complement the corpus of the available evidence. The reconstruction of the foundation techniques is not without problems. Unlike other sites in the region,59 identification of well-defined rows of postholes was not always possible. At least seven postholes have been located in the rubble area of Building 1 and have been associated with the load-bearing elements of the structure(s). In addition, a number of postholes defining a roughly rectilinear ground plan have been identified in the area of Building 5. As is the case with other buildings, however, it is not always straightforward to establish their stratigraphic correlation to the Avgi I rubble. Nevertheless, the evidence seems to support the use of upright posts sunk directly into the soil for the construction

58 Kyrillidou 2016. 59 e.g. MN/LN Servia (Mould and Wardle 2000a), LN Makri (Efstratiou et al. 1999).

82 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

of the building’s framework. The structures follow the basic principles of earthfast architecture. This implies that the weight of the roof and other vertical loads were taken down to the ground independently of the walls, probably by using continuous wall-plates or tie-beams linking the tops of the vertical posts. As for the roofs, these are thought to be either gabled or double-pitched, based mainly on ethnographic parallels and a series of clay house models found in Greece and elsewhere in the Balkan region.60

The analysis of the superstructural remains offers valuable insights into the building and tempering materials used, as well as into the wall construction techniques applied.61 The latter refer to the screening of ‘non-structural’ walls and have been recognised on the basis of the multiple impressions of the timber frame recorded on daub. The main walling technique seems to comprise almost parallel and closely set stakes and split timbers that were arranged vertically and were probably fixed with transverses.62 The diameters of the stakes or thin poles range primarily between 0.06-0.10m, while the split timbers used in place of the round ones present comparable dimensions. Judging from the rarity or absence of associated stake holes, these timber elements could have rested either directly on the ground or on an earthen or wooden foot (e.g. a sill or sleeper beam). The wall screen was packed with successive layers of construction earth and was later lined with calcareous, finishing plaster. It is noted here, that the first layer of construction earth is commonly poorly tempered, while the following layer(s) bear large quantities of flexible plant tempers that are preserved as voids or, more rarely, in a silicified form.

The second wall construction method comprises thin, pliant branches and/or reeds that are weaved between more sturdy uprights in a wattle-and-daub fashion. The wattles commonly range between one and 2.5cm in diameter/size, while upright staves (either round or split) measure approximately 4-6cm. The wattle frame is again packed with a layer of plant tempered earth and finished with a thin layer of calcareous plaster. The application of this technique is generally restricted and could be related to the construction of internal partitions or specific parts of the superstructure. Nevertheless, a more widespread application of this method is indicated in the case of Building 7. An alternative technique is reflected by a number of daub fragments bearing wavy impressions. This comprises split or plank-shaped timbers either set parallel or partly overlapping. As has been suggested in the case of Neolithic Servia,63 they could have been used as cladding.

5.2.1.2 Micromorphology of floors, by Stella Kyrillidou

Six main groups have been identified across the investigated part of the settlement, on the basis of the combined field microstratigraphic and thin-section micromorphological study in large resin-impregnated thin-sections: a) dark orange brown silty clays in enclosed roofed areas inside buildings, b) orange silty clays in enclosed roofed areas inside buildings, c) yellow clay loams in unenclosed roofed areas outside buildings, d) yellow clay loams in unenclosed unroofed areas outside buildings, e) orange brown and reddish brown sandy clays in surfaces of constructed features, for example, fire installations, or platforms, and f) light grey to white silty clays in suspected , or small threshing floors. Other variations also exist, but these are the most frequent groups that have been identified to date in thin section.

The floors, meaning surfaces that are trampled, and were created to enable movement, are single, rather ‘self-contained’ units, c. 2-4cm thin. They differ significantly from floor sequences of tell sites in the Near East.64 Two main earths and mineral sources have been identified in the manufacture

60 Toufexis 1996; Trenner 2010. 61 The study of the material as part of the author’s PhD thesis was funded by the Greek State Scholarship Foundation (IKY). 62 The use of transverses has been reported in the case of Neolithic Dikili Tash (Martinez 1999: 64−65). It could also be supported in the case of Avgi by a single daub fragment from the rubble of Building 5. 63 Mould and Wardle 2000b: 82. 64 cf. Matthews et al. 1996.

83 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

of surfaces at Neolithic Avgi, the local virgin soil and the underlying marine Miocene marl of the site. In enclosed roofed spaces indoors, the floors are processed and laid, and may comprise minute residues from the overlying accumulation of in situ use due to trampling. Surfaces of furniture or fixtures are easily distinguished from the floors of buildings, as they have a planar vegetal structure, which is non-trampled. They comprise abundant impressions of plants and frequent siliceous phytoliths of monocotyledons, and strongly resemble daub from wall construction in thin section. Laid surfaces of suspected heating hearths, or small threshing floors are almost consistently crafted with marl, but they are not heavily tempered with plants.

There is a significant difference between the floors of buildings and the floors of open areas. In open areas, floors are also laid and trampled, but they are unprepared.

5.2.2 The open areas of Avgi I

Open spaces in-between the building areas are commonly characterised by their richness in bioarchaeological remains, a high frequency of artefacts and the presence of features uncovered in situ. The latter comprise varied forms of thermal structures, such as suspected hearths and ovens. The picture emerging is one of highly active areas affected by complex formation processes and linked to a broad spectrum of domestic and non-domestic activities.

In some cases, the preservation and arrangement of features and the concentration and spatial distribution of finds provide insights into the activities and tasks being performed in specific contexts. For example, in the open areas that surround Building 5, different attitudes towards the use of outdoor space are suggested based on the available evidence.65

More specifically, the area North of Building 5 seems to be closely related to food preparation, cooking and other fire-related and craft activities. This is indicated by a cluster of three thermal structures, including two subterranean domed ovens and a hearth66 (Figure 21), as well as at least three nearby concentrations of ashes. Micromorphological analysis67 suggested that the latter should be attributed to open hearths. These ash layers were well preserved, likely by rapid accumulation and subsequent burial, and are partly overlapping, suggesting distinct, successive episodes of burning. The calcitic composition of the ash residues in these deposits indicates wood as the primary fuel used (as opposed to dung or grass-rich fuel). The temperatures reached by these fires exceeded 500oC, which is the minimum required for the production of calcitic ash.68 In addition, redeposited or disturbed fuel residues were also observed in thin section as thin lenses, which is consistent with episodes of fuel rake-out in this area (Figure 22). All the above evidence indicates persistent fire-related activity, plausibly for outdoor cooking. The presence of intact grinding slabs and grinders near these spaces could indicate that food grinding practices had also occurred in the open in the vicinity of the cooking locales.69 What is more, the pottery found in the area under consideration seems to be associated with everyday practices, such as food preparation and consumption, as well as storage and transportation.70 The presence of fine ware may also suggest food consumption on a supra-household level with various social connotations.71 The open areas north of Building 5 may have also hosted practices related to ground stone manufacture, curation and destruction, as suggested by the presence of flakes, raw materials, and deliberately

65 Stratouli et al. 2011. 66 Kalogiropoulou 2013. 67 Koromila 2015. 68 Canti 2003. 69 Kalogiropoulou 2013; Bekiaris 2020. 70 Katsikaridis forthcoming; Stratouli et al. 2011. 71 see Urem-Kotsou 2006.

84 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figure 21. Top: in situ grinding tools located in the Avgi I open areas at the North of Building 5, Western Sector. Middle: Hearth 1 located in the Avgi I open areas at the South of Building 1, Eastern Sector. Bottom: Cluster of thermal structures at the Avgi I open areas at the North of Building 5, Western Sector (Photos: George Vlachou).

Figure 22. Trench 252.4, west bulk profile, pre-sampling; the outline indicates the origin of the thin section (centre). Sequence with two distinct in situ burning episodes, as indicated by thick and relatively undisturbed calcitic ash deposits, separated by redeposited charred plant remains and a mixed/reworked accumulated unit. The photomicrograph (right) shows concentrated calcitic ash crystals (under Plane-Polarised Light – PPL). The abundance of fuel remains in the area north of Building 5 reinforces its interpretation as a focal place for burning- related activities.

85 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

destroyed grinding tools.72 The same areas have also yielded several by-products belonging to various stages of the chipped stone chaîne opératoires.

The area East of Building 5 seems to entail more intense rubbish disposal and secondary formation processes of disturbance. This is attested by the scarcity of in situ thermal structures or other features, as well as by the composition of the accumulated assemblages of several artefact categories. Pottery presents considerable variability in terms of shape and function, although the presence of tableware is probably more pronounced compared to other areas (e.g. north of Building 5).73 The ground stone category includes mainly tools involved in the production sequences of other objects, like abraders and polishers.74 Based on the presence of chaff, the study of the archaeobotanical remains suggests that the area east of Building 5 could have served for the disposal of food byproducts.75 Another possibility for the presence of this chaff is daub manufacture, on the basis of construction materials, the groundmass of which comprises abundant chaff that coexists with soil aggregates that are similar to the accumulation east of Building 5.76 This is also in accordance with the presence of ‘tolerated’ inclusions (pottery sherds, bone fragments, chipped stone, etc) within daub fragments,77 which indicate the preparation of daub within the settlement.

Open spaces have been also identified in the Central and Eastern Sectors of the site, but their excavation is less advanced. These have also yielded significant results. The nature of the accumulation has a grey-looking appearance due to enhanced content in ash and charred plants, and it resembles midden-like areas that are reported from sites in temperate European regions.78 Residues that are both in situ as well as in secondary use have been identified in these midden- like areas underneath and next to buildings by high-resolution microstratigraphic and thin- section micromorphological study.79 Τhe in-situ activities identified are mostly associated with the practice of agriculture and includes gardening, due to changes in microstructure and the possible processing of plants. Other activities and contexts of secondary deposition include discard, as well as more massive clearance events from the interior of buildings.

Constructed features often appear at the end-life of buildings, above the rubble of burnt buildings, or next to these. A good example is Feature 139401,80 c. 1m2 in area, <2m apart from Building 1 (Figure 21). The field microstratigraphic study suggests that this feature was renewed and maintained in this area for a life span that may have significantly exceeded the life cycle of the nearby Building 1 or be unrelated to this life-cycle. The micromorphological study of the sequence of this feature further suggests that the surfaces of the feature were prepared, laid, compacted, and fired. There is little residual evidence to suggest cooking and consumption, for example, burnt plants, or burnt bone, perhaps due to taphonomy and location outdoors.81 Other possibilities, however, are also conceivable. Some of these may suggest plant processing and food preparation, rather than cooking and consumption (see also section 5.4, Type 3).

The content of the open areas and the variability of their contexts indicate that the outdoor space played a significant role in social interaction in the everyday life of the inhabitants. Household activities, including food preparation, cooking and consumption, as well as possible tool manufacture and/or repair, were taking place in the open areas around the house. The lack or

72 Bekiaris 2018: 341-347; 2020. 73 Katsikaridis forthcoming. 74 Bekiaris 2018: 347-351; 2020. 75 Stratouli et al. 2011. 76 Kyrillidou 2016. 77 Kloukinas 2014. 78 cf. Guttmann et al. 2006. 79 Kyrillidou 2016, in preparation. 80 Referenced as Hearth 1 in Kyrillidou 2016; these are features that are generally interpreted as ‘hearths’ in Greek Neolithic sites, albeit that their role and specificity of uses are little understood. 81 Mallol et al. 2007.

86 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

low visibility of spatial features demarcating the boundaries of these open areas may imply that movement and sensory experience were, by and large, unchannelled.82 It is difficult to establish a relationship between specific dwellings and open spaces, in the sense of ‘house and garden complexes’ that were perhaps fixed, as has been proposed by Chapman and Bogaard83 for the flat- extended sites of Balkans. It is plausible to postulate a certain degree of ‘openness’ and sharing between neighbouring houses or households. This sharing may have involved both equipment (e.g. thermal structures, grinding tools) and finished products, such as cooked food and various artefacts. On the other hand, it is equally possible that intra-site boundaries existed, but these were either symbolic or are archaeologically untraceable.

5.2 Avgi II (5,200/5,100 – 4,900 cal BC)

The remains attributed to Avgi II have been primarily investigated in the Western Sector of the site, where they comprise open areas. These are deposits that have accumulated on top of the remains of Avgi I and have partially covered the destruction layers of the buildings in the Western Sector. They were subsequently cut by the foundations of traces of buildings and by other cuttings of phase III. These deposits, thereby, comprise evidence of activity in between the phases of Avgi I and Avgi III. Macroscopically, these sediments are homogenised, with diffused boundaries and no internal layering, which suggests that they are significantly modified by post-depositional alterations.

Within this homogenised stratigraphic unit, spatial features and concentrations of laid artefacts were parallel oriented and linearly distributed at certain levels (Figure 23), but they were

Figure 23. Open areas, finds and thermal structures attributed to the Avgi II open areas, Western Sector (Photos: George Vlachou).

82 Robb 2007: 90; Stratouli and Kloukinas in press; see also Kalogiropoulou 2013. 83 Chapman 2010: 75-76; Bogaard 2004.

87 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

discontinuous. Inferred active surfaces may have been disturbed by soil-forming processes, as well as disturbance by later cuttings. Intensive trampling and other anthropogenic activity could have also contributed to the homogenisation of the deposits in these particular areas due to the sandier texture and softer nature of these deposits.

The micromorphological investigation conducted by Georgia Koromila84 can shed additional light on the complex formation processes of those sediments that are associated with Avgi II in the Western Sector of the site. The examined samples in thin section indicate that diverse components of multiple origins with prolonged pre-depositional biographies were progressively accumulated and reworked into homogeneous layers. Pre-burial processes include fragmentation, disaggregation, mixing, and transportation of material, possibly by exposure to rain, wind and slope-induced movement, discard practices, and human and animal trampling.85 Homogenising post-burial processes include burrowing by animals and root intrusion, as evidenced by channel- shaped voids and disturbed areas in thin section and animal burrows identified in the field; water- induced re-organisation of fine material, indicated by some reworked translocation features and surface crusts; intentional removal, as indicated by numerous postholes and pits excavated; and possible levelling to create surfaces, or compaction caused by trampling, suggested for instance by the fragmentation pattern identified on a pottery sherd.86

In general, active surfaces and depositional episodes are difficult to trace in these deposits. In such sedimentary contexts, however, horizontally laid inclusions could indicate depositional surfaces, as large-sized and hard material is less susceptible to vertical post-depositional movement, i.e. is not easily dispersed by wind or rain, not easily transported by bioturbation, nor easily fragmented in the degree that would render it susceptible to the above. In thin section, areas underlying such finds did not exhibit any traces of preparation or increased compaction.

Rich anthropogenic assemblages were found often associated with spatial features found in situ. These comprise mostly thermal structures in the Western Sector of the site, but no buildings. The better preserved examples include domed ovens with or without vent-holes, as well as suspected open hearths with and without pebbly substrates.87 That thermal structures of various forms are occasionally clustered in groups suggests that specific areas were the focus of possible cooking and other fire-related activities. It is difficult, however, to establish whether these structures were contemporary, due to the complex formation processes of the surrounding open areas. The presence of intact grinding slabs right next to some thermal structures suggests a strong spatial relation between various food preparation practices,88 while the concentration of finds (including pottery, tools and other artefacts) and bioarchaeological remains in the periphery of the thermal installations attests to the existence of outdoor activity areas.

The presence of surfaces suggested by macroscopic clusters of finds and in situ features is further supported by microscopic observations, e.g. fragmented and very locally preserved evidence of surface crusts, i.e. thin (<1mm) bands of silt to fine sand with graded sorting, likely formed through settling in stagnant water.89

As already mentioned, the above assemblages have not been linked stratigraphically to any identified building residues in the Western Sector; a question was, thereby, formulated on whether these traces of activity relate to domestic occupation or to permanent habitation in the Western Sector of the site, in particular. However, the rich accumulations of predominantly anthropogenic

84 Koromila 2015. 85 Banerjea et al. 2015; Matthews et al. 1997; Nielsen 1991; Shillito and Ryan 2013. 86 Miller et al. 2010. 87 Kalogiropoulou 2013. 88 Bekiaris in press. 89 Courty et al. 1989: 155; Koromila 2015.

88 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

material emphasise the continuous use of the site. The continuity is also supported by the fact that certain features of the Avgi II phase were built directly on top of the Avgi I remains, suggesting that the earlier rubble was still visible.

It should be noted that the absence of evidence is not evidence of absence, and that low visibility of building remains may be partly attributed to the perishable nature of building materials used, or be explained by the possibility that buildings were not preserved as they were not subjected to conflagration during their destruction. The virtual absence of domestic buildings in the archaeological record could also raise questions on a possible rearrangement of settlement space: Are there buildings that remain to be found in other parts of the site? This could result from a major horizontal shift of the occupation and the re-organisation of the focus of the settlement’s residential area towards the southernmost, eroded part of the site. Could some of these features and concentrations that have been identified in the Western Sector be contemporary with features attested in the microstratigraphy of the Central and Eastern Sectors? Another possibility is that people may have returned to this particular area at a later phase, and that this part of the Western Sector of the excavations was selectively used in Avgi II-III. Importantly, all these are areas of archaeological inquiries that remain to be explored in the future.

5.2.1 Cooking installations, by Evita Kalogiropoulou

By the end of the final excavation season, an assemblage of twenty-six cooking installations dated to the phases of Avgi I and Avgi II were uncovered in all sectors of the site and were systematically recorded thereafter. 90 Fifteen cooking installations are securely identified in Avgi I, while eleven features are recorded in Avgi II. An ongoing micromorphological study of this assemblage conducted by the author, revealed new types of cooking related features and changed the initial classification, based on nuanced observations on their construction techniques and evidence on their functional properties.91

The structures are grouped in four main categories: ovens, hearths, food preparation benches and fire pits (Table 3). Classification follows morphological, structural and functional characteristics. Preservation varies, while many structures are considerably eroded. Although technological know-how is indicated by the variability of building techniques and the selection of raw materials used, the features from Avgi show low building investment that produced plain and rough constructions. Their size and structure suggest small-scale and domestic use. Based on empirical and ongoing microscopic observations, none of all twenty-six features exceeded 800οC, an evidence that supports their domestic use. 92

Overall, eleven ovens are identified making this the dominant category of cooking related features in the settlement. In Avgi, ovens are horse- shaped and semi-circular structures with vault compartment and front-loaded entrance. 93 The dome is curved and solid (type 1a), except for three features with a circular hole of average 0.20m diameter on their roof compartment (type 1b). Ovens, here, are medium sized above ground features built directly on soil or semi-subterranean structures, constructed in a basin-like shallow cut in the ground. Heating floors were made with basic smoothening of the surface, whereas upper structures and domes were built with the coil technique. Inside the vault compartment roasting and boiling of food in small pots, or skins is possible.94 Parching and smoking is also common in Neolithic cooking practices and it is also probable in the case of Avgi.

90 Kalogiropoulou 2013. 91 The macroscopic study was conducted as part of the author’s PhD research and was funded by Cardiff University (School of History, Archaeology and Religion), the Greek Archaeological Committee in UK and the Gilchrist Educational Trust. The micromorphology study was initiated at the McBurney Laboratory for Geoarchaeology at the University of Cambridge during a Visiting Fellowship. 92 Mallol et al. 2017; Roussos and Kyparissi-Apostolika 2018. 93 A cooking installation clay model that is comparable to the in situ structures was found in Avgi (Metaxas and Stratouli 2017: 192). The clay model adds important information to the missing decorative elements of the features. 94 Atalay and Hastorf 2006.

89 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

ID AE Phase Type Shape Size Construction techniques Function

*TS 7 283202 Avgi I oven, type 1a horse-shoe 0.95 x 0.60 m heating floor(s) bake shape upper structure from coil roast TS 8 252301 Avgi I oven, type 1b horse-shoe 0.61 x 0.41 m layers boil shape basin-like shallow cut smoking above ground or semi- TS 9 252302 Avgi I oven, type 1a horse-shoe 0.45 x 0.35 m subterranean shape TS 14 - Avgi I oven, type 1a horse-shoe 0.97 x 0.35 m. shape TS 15 254405 Avgi I oven, type 1a horse-shoe 0.81 x 0.57 m shape TS 25 253201 Avgi I oven, type 1a semi-circular 0.60 x 0.50 m

TS 22 196202 Avgi I hearth, type 2a circular 0.70 m heating clay floor(s) grill smoothed heating roast ⌀ TS 23 - Avgi I hearth, type 2a irregular 1.23 m surface(s) boil pebbled underlay smoking ⌀ above ground TS 24 - Avgi I hearth, type 2a circular 0.80 m ⌀ TS 11 253202 Avgi I hearth and food irregular 3.00 m bench, types 2 and 4 ⌀ TS 26 223301+223302 Avgi I fire-pit, type 3 circular 2.50 m shallow pit grill plastered side-walls roast ⌀ TS 29 223303 Avgi I fire-pit, type 3 circular 0.95 m smoking ⌀ TS 10 252301 Avgi I food preparation irregular 0.50 m above ground prepare food bench, type 4 flattened clay surface(s) ingredients ⌀ TS 17 139401 Avgi I food preparation irregular 1.35 m reconstruction works bench, type 4 ⌀ TS 27 - Avgi I food preparation irregular length 2.84 bench, type 4 m TS 1 - Avgi II oven, type 1a semi-circular 0.50 x 0.40 heating floor bake upper structure from coil roast TS 2 - Avgi II oven, type 1b horse-shoe 1.00 x 0.55 m layers boil shape basin-like shallow cut smoke above ground or semi- TS 3 - Avgi II oven, type 1a horse-shoe 0.55 x 0.35 m subterranean shape TS 13 - Avgi II oven, type 1a semi-circular 0.50 x 0.40 m

TS 16 - Avgi II oven, type1b semi-circular 0.60 x 0.47 m

TS 5 - Avgi II hearth, type 2a irregular 1,54 m heating clay floor(s) grill smoothed heating roast ⌀ TS 6 - Avgi II hearth, type 2a irregular 0.60 m surface(s) boil pebbled underlay smoke ⌀ above ground reconstruction works TS 12 - Avgi II food preparation irregular 1.90 m above ground prepare food bench, type 4 flattened clay surface(s) ingredients ⌀ TS 18 - Avgi II food preparation irregular 1.95 m reconstruction works bench, type 4 ⌀ TS 19 - Avgi II food preparation irregular 1.54 m bench, type 4 ⌀ TS 28 - Avgi II food preparation irregular * 0.76 m bench, type 4 ⌀ *TS stands for the general term Thermal Structures ** TS 28 is partly unearthed Table 3. List of the 26 cooking installations in the Neolithic site of Avgi, Kastoria.

90 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Hearths (type 2) constitute a category of fewer structures, counting for six features overall. Their shape is either circular or irregular and their average size is about 1.10m in diameter. Hearths are uncovered structures; therefore, heating is developed in open-air conditions. All hearths in Avgi are above ground features, with finely polished and smoothed heating floor surfaces, that were applied on top of pebbled underlays. When preserved, heating floors are finely constructed, demonstrating evidence of advanced technological style and pyrotechnological knowledge. Reconstruction works, however, are not recorded, possibly indicating low sequence use or practices related to cooking habits. Hearths can support various direct or indirect cooking practices, such as indirect boiling in pots, baskets or skins, but also for direct cooking, such as grilling, smoking and parching.95

Another category of cooking related structures found at Avgi is food preparation benches (type 3). These are auxiliary features for food processing. Overall, seven such benches are recorded in both phases of the site. Food preparation benches are above ground structures with flattened and smoothed clay surfaces that include high concentrations of organic mixture. The floors of these features were built directly on the ground with no additional preparation, such as the construction of basins or pebble settings. The main element that distinguishes them from hearths is the complete lack or low fire effects on their floor surfaces. Reconstruction works are occasionally recorded on these structures suggesting a desire to prolong their use life and to sustain their spatial arrangement.

The last category of cooking related structures at the site comprises two fire pits (type 4), both dating to Avgi I. Their shape is circular, but their size is uneven with one considerably larger than the other. Construction techniques, however, are alike with shallow cuts made directly on the ground, plastered walls and floor surfaces. Both structures contain thick ash concentrations, which in the case of the larger structure could support the hypothesis of a particular or large scale cooking practice.

All twenty-six cooking installations were built in open–air spaces with direct or indirect association to buildings. They are found as single standing features or in clusters of two or three interrelated structures. Cooking installations at Neolithic Avgi formed dynamic spaces of social interaction and domestic shared experiences around food preparation and cooking practices.96 Their spatial distribution in open-air spaces challenges the boundaries of Neolithic households by developing outdoor focal areas that daily brought people together in community-wide and visible venues.

5.3 Avgi III (4,900 – 4,500/4,300 cal BC)

The upper parts of the occupational sequence of the site comprise various cuttings that interrupt earlier deposits. These include, for instance, the foundation trenches,97 post-pits and postholes of post-framed buildings, ditches or trenches of different functions, as well as several pits (Figure 24). The fill of these cuttings comprises anthropogenic materials that have been radiocarbon dated to the Late Neolithic II period. The various features sometimes intersect, suggesting the existence of separate sub-phases of removal and depositional activity. It is difficult to determine the chronological sequence of these features. As the basal boundary of the topsoil is truncated throughout the entire site, features that are not contemporary appear at the same stratigraphic level.

Traces of features that were assigned to Avgi III have been identified throughout the site, but substantial architectural remains have been documented at the Western Sector of the excavations (Figure 25). The

95 Atalay and Hastorf 2006. 96 Kalogiropoulou 2014; Bickle and Kalogiropoulou 2017. 97 Also referred to as foundation ditches or wall slots.

91 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 24. Plan of Avgi III, indicating the foundation trenches of the suspected buildings, their intermediate ditches and the various pits (Plan: Tasos Bekiaris).

Figure 25. The foundation trenches of Avgi III buildings, ditches, pits and other cuttings at the Western and Central Sector (Photo: George Vlachos). archaeological evidence attests to certain innovations in building technology (i.e. the use of foundation trenches), as well as to the transformation of the settlement space. The general impression is one of a more neatly ordered built environment, where buildings and associated features follow a similar orientation and a roughly parallel arrangement. The abundance of pits may also suggest a change in practices of removal and deposition, as well as in waste management and disposal.

5.3.1 Traces of buildings of Avgi III

The most prominent architectural features of Avgi III comprise the underground parts or imprints of two neighbouring, post-framed buildings, Building 2b and Building 6. They were both detached, following rectangular ground plans and sharing a common NW-SE orientation. They also present significant structural affinities. The north, east and west walls of both buildings are preserved in the

92 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figure 26. Schematic representation of the superimposition of buildings in Area 2. Building 2a dates to Avgi I and Buildings 2b and 2c to Avgi III (Plan: Tasos Bekiaris).

Figure 27. The foundation trenches of Building 2b, pits, post-holes and other cuttings in Area 2 of the Western Sector (Photo: George Vlachou).

form of U-shaped foundation trenches and postholes inside these. In both cases, the position of the south walls has not been securely identified.98 It is possible that alternative foundation techniques (e.g. posts sunk directly into the soil) were applied, which may also suggest the presence of an entrance on this side. No floor sequences or intact supra-structures have been preserved from these buildings of Avgi III.

Building 2b was partly built on top of the remains of the earlier Building 2a (Figure 26) It represents, however, a building with a different orientation and size c. 98m2 (c. 14x7m). The foundation trenches uncovered (Figure 27) measure up to 0.50m in width and were preserved at a maximum depth of c. 0.30m. Their initial depth must have been much larger as implied by similar features at the site, and

98 Stratouli and Bekiaris 2008.

93 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 28. The foundation trenches of Building 6, pits, post-holes and other cuttings in Area 6 of the Central Sector (Photo: George Vlachou).

by comparative evidence from other northern Greek Neolithic sites.99 Postholes of different sizes were identified inside the foundation trenches. Their arrangements suggest the existence of single or double rows of structural timbers. The larger postholes (c. 18-20cm in diameter) may correspond to the load-bearing elements of the timber frame, while the smaller ones (c. 8-12cm in diameter) could be linked to timbers used for extra support or for wall framing. Several postholes and six post-pits for the erection of load-bearing posts or buttresses were also identified in the building’s interior. The latter are circular and measure approximately 0.50m in diameter. They were all located in the northern part of the building, close to the external walls, in a Π-shaped arrangement. This arrangement could support the hypothesis of a second storey or a loft in this part of the building. It has also been suggested that the post-pits could be linked to a technical solution (e.g. a platform- like construction) in order to deal with the inclination of the site.100 We also note that traces of two successive foundation trenches have been identified to the east of the eastern wall, following a parallel orientation; these may be indications of subsequent reconstructions of Building 2b at roughly the same place. The above characteristics of Building 2b, i.e. size, substantial speculated load- bearing capacity of timbers, persistence in reconstruction, have been interpreted by the excavator as indications of a building that could have been of a special status or a supra-household function.101

Seven meters east of Building 2b, the remains of Building 6 were identified (Figure 28). They attest to a smaller structure, measuring c. >45m2 (c. 8x6m). The foundation trenches are a bit narrower (0.45m) and were preserved at a depth of c. 0.40-0.50m. The number of postholes identified inside the trenches was also smaller. These attest to a foundation technique that comprises sizeable corner posts that are either single or double ones, as well as a few intermediate, possibly load bearing, posts. In the building’s interior, three elliptical cuttings (c. 0.20m in diameter) were located at the east-central part and were arranged almost parallel to the eastern wall. These have been interpreted as cuttings that surrounded standing timber posts in the middle, and could represent possible roof supports, based on their location.

Trenches or ditches of different sorts, either U- or V-shaped, were identified in other parts of the site. Some could be associated with the foundation of poorly preserved buildings. In other

99 Sites include, for instance, EN Nea Nikomedeia (Rodden 1962), EN Yiannitsa B (Chrysostomou 1991), MN/LN Servia (Mould and Wardle 2000a) and LN Makriyalos Ib (Pappa and Besios 1999). 100 Stratouli 2013; Stratouli and Kloukinas 2020. 101 Stratouli 2007: 12, 2013.

94 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

instances, however, the lack of postholes does not favour such an interpretation. It has, therefore, been proposed that certain features may represent intra-site boundaries between buildings and/ or activity areas. These could also have served to facilitate the run-off of rainwater.102 It should be noted that all these features share a common orientation, contributing to the more ordered settlement layout, which is a key characteristic of the Avgi III phase.

Further to the north of Building 2b, another structure was partially brought to light. This was attributed to the latest stages of the site’s occupation. Building 4 comprises an elliptical, U-shaped foundation trench pointing to an elliptical or an apsidal ground plan. The foundation trench is narrower than those of the other buildings (c. 0.30m), perhaps reflecting a lighter superstructure. An even narrower trench running perpendicular to the former one, could indicate the presence of an internal partition. Two single postholes measuring c. 14-18cm in diameter were identified inside each foundation trench. In addition, two circular pits (0.50m and 0.34m in diameter), resembling the post-pits of Building 2b, were found in the building’s interior. In short, Building 4 presents certain technological affinities with the other buildings attributed to Avgi III, mainly in terms of the foundation techniques applied. Nevertheless, its ground plan remains unique. Traces of elliptical built forms are known in the LN and FN architectural record of northern Greece (e.g. the sites of Dispilio, Makriyalos IIb and Kleitos 2).103 Additionally, apsidal dwellings have been unearthed at LN Arkadikos104 and, probably, the site of Limenaria on the island of Thassos.105

In terms of building technology, the use of trenches for the foundation of the post-framed walls seems to be an innovative feature of the Avgi III phase. This technique is, of course, well known since the earliest phases of the Neolithic in the wider northern Greek and Balkan region.106 The application of the technique may be associated with the need for extra stability, durability and waterproofing of the timber frame.107 An alternative explanation could be that the trenches were cut through the earlier deposits so that the load bearing elements would sink to the natural subsoil with no need for the extensive clearance of the underlying rubble.108

Regarding the diverse arrangement of posts placed inside the foundation trenches, it is difficult to determine whether this could also reflect diverse wall construction methods. It has been argued109 that Building 2b could follow a wattle-and-daub or a similar technique, while the walls of Building 6 could have been built in a rammed earth or composite pisé de terre technique. This is, however, very difficult to test, as no superstructural material belonging to the Avgi III phase has been securely associated with a specific building. The fire-hardened fragments that were found inside the foundation trench of Building 6, and which have been attributed to the fill of this trench, most likely represent packing using older material in secondary deposition. Overall, the daub fragments that have been retrieved from the later cuttings and the ploughed topsoil of the site suggest that the wall construction techniques had some similarities with the ones applied during the Avgi I phase.

5.4 Pits and other cuttings

The 2002-2008 excavations have uncovered more than forty pits (Figure 29). These were found throughout the site, either scattered or in clusters, and their great majority has been assigned to

102 Stratouli and Bekiaris 2008: 8-9; Stratouli 2013. 103 Kloukinas 2017: 172. 104 Peristeri 2004. 105 Malamidou and Papadopoulos 1993: 561. 106 Northern Greek sites include, for instance, EN Nea Nikomedeia (Pyke 1996), EN Mavropigi Filotsairi (Karamitrou et al. 2013), MN and LN Servia (Mould and Wardle 2000a), LN Makriyalos IIb (Pappa and Besios 1999), EN Yiannitsa B (Chrysostomou 1991) and Axos A (Chrysostomou 1996). Examples from the Balkan region include the sites of Selevac (Tringham and Krstić 1990) and Divostin (Bogdanović 1988). 107 Stratouli 2013. 108 Kloukinas 2014: 234. 109 Stratouli 2013.

95 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 29. Various pits of Avgi II & III. Top left: Pit 109401, Eastern Sector. Top right: Pit 253201, Western Sector. Bottom left: Pit 281401, Western Sector. Bottom middle: Pit 254404, Western Sector. Bottom right: Pit 312202, Western Sector (Photos: George Vlachou). the later stages of the occupation (the phases of Avgi II and Avgi III). Although they vary in terms of their morphological characteristics, most features are U-shaped with a circular or elliptical ground plan ranging between 1m and 3m in diameter. They were commonly identified immediately underneath the topsoil and their preserved depth reaches up to 0.50m. We emphasise, however, that this does not necessarily reflect their initial depth, as most appear along the basal boundary of the topsoil, which is truncated. As already noted, they sometimes intersect, indicating the existence of chronologically separate activities in the same location. The stratigraphic correlation among the cuttings, as well as their association with architectural features, is indeterminate. Nevertheless, the position and arrangement of a series of pits that were excavated west of Building 2b suggest possible use of certain cuttings by specific buildings of Avgi III.110

In terms of their function, several features could have been initially cut for the procurement of raw materials for the construction of the post-framed buildings and associated structures.111 Although pits are sometimes interpreted as used for storage or as work areas,112 at Neolithic Avgi their content indicates that most were at some point filled with refuse material. The pits at Avgi II- III indicate that the disposal of rubbish rather than taking place in the open, may have been more ordered, suggesting a different management of the outdoor space.

The infillings of pit features were rich in archaeological material. Commonly, the excavated pits contained a diverse range of deposited material. Pit 253201 (Figure 29), dated to Avgi III (cf. Table

110 Stratouli and Bekiaris 2008. 111 Stratouli et al. 2014a. 112 Chapman and Gaydarska 2007: 13.

96 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

1), contained mostly ashes, pottery sherds, mainly carinated black-burnished bowls and footed vessels, broken and heavily weathered ground stone tools, chipped stone implements, a few bone tools, clay objects, faunal and food remains, as well as >300 charred husks.113

Certain pits, however, seem to reflect different patterns of deposition that are often described as ‘structured’.114 The term ‘structured deposition’115 is used here to refer to a socially organised, intentional and/or selective depositional practice of anthropogenic material in specific contexts, that is clearly differentiated from mundane deposits in terms of the materials included and/or the manner of their deposition. During the last two decades, ‘structured deposition’ patterns have been macroscopically recognised in Neolithic northern Greek sites, including Toumba Kremastis Koiladas, Makriyalos and .116

Among pit features with a higher degree of selectiveness, Pit 312202 (Figure 29) contained exclusively construction rubble, which could be related to symbolic acts of closure or cleansing.117 Pit 109401 (Figure 29) contained grinding tools associated with food processing. The large quantity, the morphological and technological variability of the tools, and especially the quantity of the handheld grinders, suggest that these could represent the inventory of more than one household.118 It is, therefore, possible that the social practices associated with specific depositional contexts were of a supra-household character. However, the size of this feature cannot support large-scale events of commensality, as it has been proposed in the case of Pit 212 in LN Makriyalos.119 Finally, the deposition of still usable ground stone tools and fully preserved pots inside pits raises questions related to the reasons for their withdrawal. This is, for instance, the case of Pit 281401 (Figure 29), which among other finds contained several intact pots along with pebbles that may have been used as burnishers (Figure 30).

Figure 30. Left: Pit 282302 of Avgi III (Photo: George Vlachou). Charred emmer wheat (Top right) and the intact grinding tool (Bottom right) from Pit 282302 (Photos: Tasos Bekiaris).

113 Stratouli et al. 2014a: 354. 114 Stratouli et al. 2014a. 115 Chapman 2000b. 116 e.g. Hondroyianni-Metoki 2009; Tzevelekidi 2012. 117 Tringham 2005: 108; Kloukinas 2014. 118 Bekiaris 2018: 387-88. 119 Pappa et al. 2004.

97 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 31. The contents of Pit 282302 (Photo: Tasos Bekiaris and Nikos Katsikaridis).

Pit 282302 (Figures 30-31) is c. 75cm in diameter and c. 45cm deep and marginally interrupts the rubble of Building 5. It has been recently dated to Avgi III (cf. Table 1), although on the basis of the microstratigraphic study,120 the feature could have been initially cut at Avgi II and continued to be used at Avgi III. The small pit is a unique and rather intriguing example of a symbolic context. It contained more than 70,000 charred seeds of emmer wheat, as it has been identified by Evi Margaritis, alongside dominant fragments of building materials, and a small grinding slab, which was placed perpendicularly on the walls of the pit, among the burnt seeds.121 In contrast to the seeds, neither the grinding slab, nor the walls of the pit exhibited any traces of burning macroscopically. The thin section micromorphological study of the content of the pit suggests that the seeds do not represent food that was meant for consumption on this occasion, as they have been deposited inside the pit whilst they were already burnt. The seeds have not burnt in situ, as would have been expected, for example, in a suspected storage facility. In addition, they have not been accumulated massively, as a single event of deposition, but at separate moments in the life history of this pit. The intentional burning of agricultural products and their subsequent deposition into pits, funerary pots and graves, comprise a well-documented practice in the Neolithic of northern Greece and beyond.122 Τhe deliberate burning of plant remains and their deposition in those contexts has been interpreted as a ritual act.123

5.5 The ‘burial area’

One of the most exceptional features of Neolithic Avgi is the identification of a ‘burial ground’ at the approximate centre of the excavated part of the site. It covers a rather restricted area of c. 3m2 and comprises ten small pots containing tiny amounts of cremated human remains (Figure 32).124 In two cases, small quantities of carbonised seeds (lentils) were also found. Their presence could be

120 Kyrillidou 2016. 121 Stratouli et al. 2011. 122 see Valamoti 2011; Margaritis 2014. 123 Margaritis 2014: 281 124 Stratouli et al. 2009, 2010.

98 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figure 32. Burial pots and their contents (Photos: Tasos Bekiaris and Nikos Katsikaridis).

related to the agricultural cycle and to notions of death and re-birth.125 Regarding their morphology, the burial pots show remarkable variability in terms of manufacture, surface treatment and shape. Some of these are elaborate and seem to imitate shapes that are common in the settlement’s ceramic assemblage (e.g. hole-mouth and necked jars). Wear marks on the base of certain pots indicate that they have been used in other activities prior to their deposition. Contrary to that, other pots are rather crude and were probably manufactured for the purpose of the burial itself. The burial pots were commonly covered by larger pottery sherds (Figure 33). Their arrangement indicates that they were probably buried in pairs, close to each other, and in more than one episode of deposition.

The osteoarchaeological analysis that was conducted by Prof. Sevi Triantaphyllou126 has indicated that the majority of the cremated fragments are consistent with the burning of bones while the flesh was still attached. The burning temperatures should have reached at least 700oC; the firing process seems to have been time-consuming and involving the continuous addition of fuel. Regarding the demographic information provided, six out of ten burials belong to adults, while only one burial belongs to an infant. More accurate sex and age determination was not possible due to the restricted quantity and the fragmented status of the preserved material.

The available evidence allows for the reconstruction of several stages of the burial ritual. After the of the deceased, which took place in an area not yet identified, only small quantities of the cremated remnants were collected and placed inside the burial pots, with no preferential

125 Bradley 2005; Williams 2003. 126 Stratouli et al. 2010.

99 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 33. Top: Two burial pots as located at the excavation. Bottom: The different stages of unearthing a burial pot covered with pottery sherds (Photos: Tasos Bekiaris). selection between cranial and post-cranial skeletal elements. Following this, the pots were deposited, either individually or in pairs, in a specific area of the site and were covered by two or more layers of large pottery fragments. The latter could belong to pots that took part in the funerary ritual and were later intentionally broken. The consistency of the ritual puts emphasis on group identity; it is apparent that only a small group of the total settlement was treated in such an exceptional way post mortem, while the unique characteristics of each burial pot may reflect the individuality of the deceased.

Regarding the dating of the burial ground, two radiocarbon dates on seeds from the content of the burial pots (see Table 1) indicate that the area may have been used for a considerable period of time, during the first half of the 5th millennium cal BC.

6. Discussion

The excavations at Neolithic Avgi significantly enrich the regional record of northwestern Greece and beyond. The Kastoria basin, where the site is located, although not systematically investigated so far, includes several Neolithic sites identified in different environmental settings. These seem to share various features in terms of material culture and subsistence practices, but they also exhibit their own idiosyncratic characteristics. Neolithic Dispilio by Lake Orestis127 has provided significant insights into the everyday practices and the organisation of lakeside settlements in Greece. The riverside Neolithic site of Koromilia128 presents a different configuration where several pits and domestic structures may be associated with more transient forms of habitation. A cave in the nearby gorge of Koromilia129 had also yielded evidence of Neolithic occupation that could be

127 Hourmouziadis 2002. 128 Stratouli et al. 2014b. 129 Trantalidou et al. 2005, 2010.

100 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

associated with the movement of people, materials and animals in the landscape. The riverside Neolithic site of Kolokynthou130 has offered additional information on Neolithic habitation during the late MN/early LN and the FN periods, while the site of Kastro Nestoriou131 indicates that human occupation was not confined to lowland areas.

The diversity of habitation patterns indicates that the Neolithic settlement of Avgi was probably taking part in a dynamic intra-regional network of communication, which may have involved the exchange of artefacts, foodstuffs and other resources, as well as various forms of social interplay. It could further be suggested that this social network was built upon notions of convergence and conviviality132 between the Neolithic communities involved. Evidence supporting the existence of such communication has been provided by the analysis of the material assemblages, and especially ground stone edge tools. The total absence of raw materials and preforms and the intense curation patterns133 that have been observed on the edge tools from Neolithic Avgi may indicate that these implements were imported to the site; a likely place of origin could be the nearby site of Dispilio,134 which has yielded significant evidence for the intensive, large-scale production of such artifacts.135 The edge tools from both sites are primarily made of ophiolites originating from the Mavro Vouno at Maniaki,136 an ophiolitic outcrop situated in the vicinity of Dispilio.

Other goods that reached the settlement as finished products may also support the existence of regional networks that were not necessarily hierarchically organised. In addition, the presence of certain imported artefacts indicates the operation of wider, inter-regional networks of exchange, in which Neolithic Avgi was also taking part. Recent mineralogical and geochemical analyses on two marble vase fragments from Avgi have indicated the Cycladic island of Naxos as their most probable raw material source. 137 Since marble is not encountered in the geological formations of the region we may argue for the exogenous - maybe also Cycladic - origin of other marble artefacts (figurines, ornaments and maceheads) found at Neolithic Avgi, the analysis of which is still in progress. A Cycladic origin has also been suggested for the marble rings found at Neolithic Dispilio.138 This may point towards specific networks in which both communities had participated. Other artefacts of non-local origin from Neolithic Avgi include Melian and probably Carpathian obsidians,139 an pendant and several malachite beads, the sources of which are yet to be determined.

On the other hand, the analysis of different material categories indicates that a wide range of crafts and practices was also regularly carried out within the habitation area or in the settlement’s vicinity. The inventory suggest that on-site production of stone implements was rather modest. It appears to have been limited to the small-scale manufacture of grinding tools made of fine to medium grained sandstones and some granites, as documented by the presence of unmodified raw materials and large cortical flakes deriving from the initial reduction stages of the cobbles.140 On the contrary, curation practices, such as the maintenance, redesign and deliberate breakage of most ground stone types were extensively practiced on-site. The chipped stone assemblage141 suggests that on-site knapping was connected with expedient flake production and modest / bladelet making, while there is also evidence for the maintenance of retouched tools.

130 Tsouggaris et al. 2004. 131 Tsouggaris 1999: 25-26. 132 Bailey and Whittle 2008. 133 Bekiaris 2018. 134 Bekiaris et al. 2017: 425-426; Stratouli 2019: 118-119. 135 Stratouli 2002. 136 Melfos and Stratouli 2002; Bekiaris et al. 2017: 425. 137 Stergiou et al. forthcoming. 138 Ifantidis 2008: 81. 139 The presence of Carpathian obsidian in Neolithic Dispilio has been attested by portable XRF analysis (Milic 2014). 140 Bekiaris 2018: 222-223. 141 See Andreasen in Appendix 2.

101 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

In addition, the on-site manufacture of ornaments made of specific marine shells that could have been collected from the Aegean or the Ionian Sea (e.g. Spondylus gaederopus, Cerastoderma glaucum, Hexaplex trunculus, Aequipecten opercularis) and sedimentary stones is suggested by a few unworked and half-worked specimens. Moreover, as indicated by the preliminary results of pottery thin- section petrography,142 the provenance of the clays and their non-plastic inclusions may point towards local production, although no specialised pottery or craft production areas or workshops associated with manufacture processes and curation practices have been identified at Neolithic Avgi.

It is plausible that certain crafts or stages of manufacture were carried out by the inhabitants in the immediate or more distant surroundings of the settlement. The manufacture of grinding tools made of coarse grained sandstone and conglomerates possibly near the procurement loci of their raw materials would have minimised the effort for their transportation. Other objects, such as ornaments, fashioned mostly through abrading, could have been manufactured at the nearby streams so as to take advantage of the available water resource and sand. Routine mobilities at shorter or longer distances around the settlement would have also included the procurement of wood, which was a key resource at Neolithic Avgi both as fuel and as a building construction element, and other agricultural and food gathering activities.

The wealth of data obtained from the present study enables us to discuss in more detail the activities and contextual interrelationships within the settlement. The extent of the excavations and the application of different techniques, including geophysical survey and micromorphological analyses, allow for an in-depth analysis of the site’s built environment and its transformation through time. The uncovered remains of buildings and other domestic structures of different sorts, and open spaces accommodating various materials and practices, can help us examine how social space was arranged and (re)produced, and how these spatial features may have operated as agents of social interaction by channelling movement, orientating experience and generating shared, culturally informed perspectives.143

A major component of intra-site spatial organisation is the system of peripheral ditches. Although their dating remains to be determined, it seems that at some point the Neolithic inhabitants of Avgi chose to enclose the settlement space and to physically demarcate it in relation to the surrounding landscape. Without doubt these massive structures were the result of a major communal undertaking that would have demanded the collaboration between different social groups and subjects. Such earthworks, also known at several other Neolithic sites in northwestern Greece,144 are generally thought to indicate an emphasis on collectivity and communal identities.145 Their construction and maintenance must have been massive efforts that would have nourished collective memory, while their constant presence would have acted as a projection of the community.146

Intra-site spatial configurations differ distinctly between the early and latest building phases of the site (Avgi I and Avgi III); these differences, refer to the degree of spatial segregation and the visibility of intra-site boundaries, which should be seen as socially meaningful.

As already pointed out, in the case of Avgi I, it is difficult to attribute the activities attested in outdoor areas to the residents of specific buildings. Suspected dwellings seem to follow different

142 Papadias and Saridaki 2016. 143 See Barrett 2006; Robb 2007. 144 In the case of northern Greece, ditches are reported from Stavroupolis I and II (Grammenos and Kotsos 2004), EN/MN Yiannitsa B (Chrysostomou 2001), MN Apsalos (Chrysostomou et al. 2001), EN/LN Nea Nikomedeia (Rodden 1965: 84), LN Makriyalos (Pappa and Besios 1999), EN/MN Paliambela (Halstead and Kotsakis 2002: 80), MN Servia (Mould and Wardle 2000a), LN Kremasti (Hondroyianni- Metoki 2009), LN/FN Kleitos (Ziota 2014). 145 Kotsakis 1999. 146 See Aslanis 1990; Kotsakis 1999; Triantaphyllou 1999 for different approaches to the issues of Greek Neolithic boundary construction.

102 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

orientations and are quite distant from one another. Their immediate surroundings probably comprise small plots for cultivation,147 roofed and unroofed outdoor spaces where different domestic activities took place, as well as midden-like areas encompassing various practices, crafts and refuse disposal.148 The absence or low visibility of intra-site boundaries may indicate that the domestic activities of each residential unit were not entirely secluded. On the contrary, outdoor space was a dynamic locale of shared everyday experiences, which would have favoured social interaction, but also the interconnection between traditions, crafts and other practices, as well as the circulation of knowledge and skills.149

In the case of Avgi III, settlement space seems to be more neatly ordered, while intra-site demarcations are more clearly visible. Suspected dwellings are built parallel and in close proximity to one another, while linear cuttings that follow similar orientation seem to segregate space and to emphasise the autonomy of residential units. In general, a stronger sense of boundedness and a less pronounced openness or sharing of the outdoor space could be argued.150 The use of pits during this phase may also support this hypothesis. Although the accumulation belonging to Avgi III is missing, the abundance of pits may be associated with a greater need to confine refuse disposal to specific areas. In general, when comparing settlement space between the two building phases, the spatial arrangements of Avgi III seem to be less flexible.

The differences observed indicate that living space did not remain static throughout the settlement’s history. Its shaping is not irrelevant to the relationship between social units and the community that structures and is structured on the basis of daily routines and social bonds. Although the exact context of this interplay is difficult to reconstruct, the evidence from Neolithic Avgi invites discussion on the possible workings of social groupings at different social scales (individual, household, community-wide).

It is noted here, that the identification of domestic units is not an easy task. Such processes necessitate the recognition of building remains as dwellings or residential units, as well as their subsequent association with key social units. The ambiguities entailed by this twofold process have already been discussed in the .151 In the case of Avgi I, the features uncovered seem to attest to the residential character of the excavated buildings. As is the case for other Neolithic Greek sites,152 the mean size of the Avgi buildings could have sustained a social unit of some sort.153 Moreover, the portable finds (fully preserved pottery vessels, stone tools and other implements, foodstuff etc) from the destruction layers and the floor deposits of Avgi I are generally consistent with an inventory that could support several daily practices (e.g. food preparation and consumption, storage, crafts etc.). In terms of interior features, no in situ hearths or ovens have been identified macroscopically in those assemblages which were more extensively excavated. Nevertheless, both the study of rubble material154 and micromorphological analysis of buildings155 support the existence of fire installations that were related to in situ cooking and food preparation within buildings, while further indications for internal furnishings are provided by the micrographic repertoire.156

147 Kyrillidou 2016. 148 Stratouli et al. 2011. 149 Stratouli et al. 2011; Kalogiropoulou 2013; Bekiaris 2020; Stratouli and Kloukinas 2020. 150 Stratouli and Kloukinas 2020. 151 Efstratiou 1990; Souvatzi 2008. 152 Halstead 1999: 80. 153 The exact composition of the group inhabiting the Avgi I houses cannot be easily determined. It is sometimes assumed that Greek Neolithic houses were sheltering nuclear families. Nevertheless, the indirect evidence from the well-known house model from Platia Magoula Zarkou, Thessaly (Gallis 1985), probably comprising two adult couples and four children, seems to imply the existence of extended families (see Souvatzi 2008: 96-98 for a relevant discussion). 154 Kloukinas 2014. 155 Kyrillidou 2016. 156 See Metaxas in Appendix 2.

103 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Activity space is essentially extended beyond the building interiors, to the surrounding open areas that constitute an integral part of the Neolithic community. Subsistence practices, such as cooking, appear to have been performed both within and outside houses.157 Plant processing activities, such as dehusking and the sieving of glume wheats, were also practiced in the open.158 The on-site slaughter of domesticated species that is attested by the presence of all anatomical units of the animals may also have been practiced outside the houses, as well as other interactions with animals related to their routine care.159 Other outdoor activities comprise stone and shell object manufacture, bone and antler working, hide processing and the tempering of daub, as well as possibly small-scale gardening.

In terms of socioeconomic organisation, certain categories of material culture may suggest an overall homogeneity across the excavated buildings, which may indicate a relative self-sufficiency of separate units. In particular, the presence of at least one set of grinding tools for food preparation,160 as well as the evidence for modest production and maintenance of chipped stone tools, inside individual dwellings161 may support this notion. Short-term or small-scale storage of cereal grains has been recorded in the case of Building 5, while similar archaeobotanical samples derive from Buildings 1 and 2a.162 Besides, the occurrence of small storage vessels163 and evidence for rodent infestation of foodstuff further support storage within buildings.164 Food-preparation and cooking activities also seem to be organised on a small scale, as hinted by the small size and the distribution of cooking facilities associated with grinding implements and food remains.165

Other features, however, suggest domestic activities taking place on larger scales. This could be supported, for example, by a cluster of three thermal structures in the area north of Building 5, with several food preparation installations in close proximity166 and documented successive events of open-air burning.167 These features suggest that at least some cooking and other burning- related activities were recurrently taking place outdoors, in an area of unrestricted access and visibility, and thereby inclusive, in conditions that may have enabled and/or enhanced interaction, collaboration and sharing. In addition, indications for collective practices and the organisation of production on a larger, even community-wide scale, also exist. These refer, for example to the construction of outer ditches, or to the absence of raw materials for stone tool manufacture inside individual dwellings, which could suggest that the exploitation of certain resources was organised on a supra-household level.

Interestingly, the activities taking place within the habitation area were not restricted to what are sometimes considered as purely ‘domestic’ practices. Although the boundaries between the domestic and the ritual are not always clear-cut,168 Neolithic Avgi has offered ample indications for the existence of symbolic practices, including acts of destruction, structured deposition and burial rites.

Destruction practices are attested in various material categories, either through their intentional burning or breakage. The destruction of houses by fire is a widespread phenomenon in the Neolithic of southeastern , which has been associated with the destructive and regenerating

157 Stratouli et al. 2011, 2013. 158 See Kotsachristou and Margaritis in Appendix 1. 159 See Tzevelekidi in Appendix 1. 160 Bekiaris 2018; for the presence of grinding tools in houses of see Bekiaris et al., this volume. 161 See Andreasen in Appendix 2. 162 Margaritis 2007; Kotsachristou and Margaritis in Appendix 1. 163 Katsikaridis forthcoming. 164 Papayianni in Appendix 1. 165 See Kalogiropoulou 2013; Bekiaris 2018; 2020. 166 Kalogiropoulou 2013. 167 Koromila 2015. 168 Bradley 2005.

104 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

properties of fire and the symbolic death of houses and households. Intentional conflagration was, therefore, approached within a ritual context marking social transitions and renegotiations of social identity.169 It is indeed tempting to perceive the destruction of the Avgi I buildings as a symbolic marker of closure and reinstitution of households. The events of house burning must have been remarkable moments in the life of the community, marking the end of an established constituent part of the community and at the same time a new beginning, a rebuilding accompanied by renegotiation of social identities and relationships.170 The production of the rubble itself provided a permanent reminder of the dead house and its place in the settlement’s history, ensuring social memory and continuity of place.

Similar notions of death by fire and regeneration may be evoked by the presence of burned seeds within two burial pots uncovered at the site. Moreover, the placement of more than 70,000 seeds, which have been burnt elsewhere and prior to their deposition, into a small pit on separate occasions may be approached in comparable terms. The intentional transformation of agricultural products through fire and their deposition into various contexts, such as pits, funerary pots and graves, is a practice well documented in prehistoric sites in northern Greece and beyond.171 Τhe presence of carbonised plants has been interpreted as a ritual act that aimed not at the destruction of the seeds, but at their preservation through their sacrifice and transformation by fire.172

Grinding tools were regularly subjected to deliberate breakage,173 achieved through intense, multiple and, in many cases, multidirectional impacts.174 In addition to this, some edge tools have also been decommissioned through breaking and flaking, while the fragmentary state of the maceheads and stone vases may imply a similar treatment.175 The high degree of fragmentation of the Spondylus bracelets and stone rings could also be indicative of such destructive acts. In addition, the presence of clay legs deriving from figurines and the absence of conjoined parts could further support this hypothesis. Finally, the smashing of pots during funerary rites and the subsequent use of their fragments to cover burial urns and to form small tumuli has also been suggested.176

Expressions of socio-symbolic or ritual character are also evident through structured deposition practices, attested by the intentional and/or selective incorporation of a diverse array of materials in the interior of pits. As discussed elsewhere (Stratouli et al. 2014a), the contents of these cuttings could be the remnants of specific individual or collective events, that were manipulated and arranged in a way that goes beyond the mundane acts of rubbish discard, towards a ritualised and socially organised practice that seeks to convey various meanings, by incorporating aspects of the Neolithic past into the Neolithic present.

The placement of two cattle horns may have symbolically marked the beginning and the end of the small Pit 254404. 177 Similarly, the selective placement of intact pots is evident inside Pit 281401.178 Pit 223101 contained exclusively fragments of grinding slabs, all of which have been recycled as grooved abraders,179 while Pit 312202 was filled almost exclusively with burned daub pieces that could have originated from the destruction of the same structure.

169 Stevanovic 1997, 2000; Tringham 2000, 2005; see also Bailey 1999, 2000; Hodder 1998; Chapman 1997, 1999, 2000a. 170 Stevanovic 1997: 341-343, 385-388; 2000: 55, 59; Hodder 1998: 92, 100. 171 see Valamoti 2011; Margaritis 2014. 172 Margaritis 2014: 281 173 For the deliberate breakage of ground stone artifacts see Adams 2008; Stroulia and Chondrou 2010; specifically for the grinding tools see Bekiaris et al. this volume. 174 see Bekiaris 2018: 312-324. 175 Bekiaris 2018: 313-314. 176 Stratouli et al. 2010: 99-100. 177 Stratouli et al. 2014a: 354. 178 Stratouli et al. 2014a: 352-353. 179 Bekiaris 2018: 386.

105 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

The contents of other pits may have been associated with collective events. This is the case with Pit 109401 which has yielded several grinding tools, probably consumed in the context of a single small-scale event of social character (i.e. some sort of gathering or feast). The moderate use of these implements and their sufficient preservation may indicate their quick disposal after the completion of specific practices.180 This event may have included the communal preparation and consumption of food, as supported by the pottery and other finds from the same pit (e.g. the presence of cooking slabs, animal bones and seeds).181 Finally, the example of Pit 282302, where more than 70,000 charred seeds of emmer wheat were placed along with a small grinding tool and fragments of daub, has already been discussed. The co-existence of three vital and interconnected elements of the Neolithic life, the grains, the grinding slab and the burnt daub creates a discrete material metaphor with conspicuous and strongly symbolic meaning, perhaps expressing the importance of food processing practices for the economic, social and ideological reproduction of the Neolithic community of Avgi III.

Perhaps the most distinct and rare example of highly symbolic practice documented at Avgi is the funerary ritual the remains of which were uncovered in the ‘burial area’. The last stages of this ritual comprised the burying of small pots with cremated remains at the same or almost the same spot within the settlement area. This practice could be seen as part of exceptional events embedded in social memory and related to specific circumstances, or the deaths of particular members of the community. Obviously, this area of the site was at some point devoted to the manipulation of death. This change in use signifies a change in the character of social discourse and suggests a place charged with meanings and potential for negotiation of power and social (inter)relationships. The close association of a group of deceased (ancestors?) with the living social environment may indicate the desire of Neolithic people to negotiate their own past by it into their own present. Such practices of materialising and managing the past could be considered as acts of remembering and, according to archaeological and anthropological theory,182 constitute important components in the formation of social identities and the construction of social relations.

In conclusion, the evidence from Neolithic Avgi has brought to light a farming community with a long history of continuity and change, including important interactions and negotiated relationships to the world beyond the limits of the settlement, and with its own sets of beliefs and rituals. Above all, our work at Avgi has tested the limits of our own knowledge and pre- conceptions and has challenged us to pose questions and find solutions through interdisciplinarity and collaboration; some of these questions were discussed here, but a lot more remains to be challenged as our analyses progress.

APPENDIX 1: Subsistence Practices

The plant remains, by Dimitra Kotsachristou and Evi Margaritis

The archaeobotanical study undertaken at Neolithic Avgi is based on the identification and analysis of charred (and a few mineralised) plant remains. The plant remains were recovered by manual flotation of sediment samples from the excavation units of the corresponding Neolithic layers of the site, representative of Avgi I, II and III. The samples were taken at locations with visible concentrations of charred plant remains, concentration of pottery, surfaces of floors, destruction layers and pits, from inside the different buildings, but also from open areas and identified structures. The wide range of context types sampled at the site during the excavations allowed for both the investigation of the spatial distribution of archaeobotanical material within buildings, as well as for a comparison between buildings and other spaces, illuminating a wide variety of

180 Bekiaris 2018: 388-389; Bekiaris 2020. 181 Stratouli et al. 2014a: 352. 182 van Dyke and Alcock 2003.

106 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

human behaviours. It was also possible to record differences between the chronological phases of the site and therefore understand its development through time. The charred remains derive principally from the residues of fires used for heating, cooking and food processing, remains that were burnt as fuel, or as they formed rubbish removed for the purpose of cleaning, representing mainly secondary or even tertiary deposition; in cases, they also exist as primary depositional episodes as they represent in situ destruction of stored material.

The archaeobotanical material of Neolithic Avgi is approached in order not only to reconstruct agricultural practices and the economy of the site, but also to detect changes in food preparation and consumption technologies, which will provide evidence for interaction with surrounding geographical regions and sites, especially that of Dispilio.

Identification of plant remains has revealed a remarkable diversity of several crop, fruit/nut and wild species. Specifically, cereals include emmer wheat (Triticum dicoccum), einkorn wheat (Triticum monococcum), bread/macaroni wheat (Triticum aestivum/durum) and barley (Hordeum vulgare). A variety of pulses are present at the site such as lentils (Lens cf culinaris), grass pea (Lathyrus sativus) and bitter vetch (Vicia ervilia). Fruits and nuts include grape (Vitis vinifera), fig (Ficus carica), blackberry (Rubus fruticosus), terebinth nut (Pistachia terebinthus) and cornelian cherry (Cornus mas). The wild/weed species category is represented by Galium spurium, Polygonum aviculare, Rumex sp., Bugglossoides arvensis and Lolium temulentum.

The predominant cereals are emmer and einkorn in all three occupation phases; from the pulses, lentils and grass pea occur more frequently, while the latter was found in concentrations within the domestic contexts of the site. Fruits and nuts repeatedly occur throughout the site, while wild/ weed species occur sporadically and in very low numbers in all three phases.

At Avgi I, the areas of interest are Building 5 and the open areas to the east and to the north of the building. Starting with Building 5, most of the samples derive from the destruction layer and the floor of the building and were often recovered among burnt areas and pottery. The predominant species found is the emmer wheat, which occurs in most samples, while there is an almost complete absence of chaff in the assemblage from this specific building. Additional species that occur in some samples are lentils, grass pea and blackberries in relatively small quantities.

The open area North of Building 5 revealed additional plant remains, with a greater variety of species: seeds of cornelian cherry, parts of whole fig fruits and also emmer and einkorn chaff. The open area east of Building 5 revealed emmer and einkorn chaff, low quantities of barley grains, and seeds of wild/weed species. Some of the samples consist solely of emmer grains, others were of emmer and einkorn grains accompanied with chaff of both species, while some were merely einkorn chaff. These rich chaff samples suggest the byproducts of dehusking, which took place in these areas of the settlement. At Avgi, crop processing activities are one of the focal areas of research. The creation of cultural artefacts involves a sequence of events to produce the end product, whether it is a stone tool, a ceramic vessel, or bread. Each stage of this chaîne opératoire produces waste, and thus evidence for each stage of the sequence in the archaeological record.183 The nature of subsistence economies, which is agricultural systems, food processing and consumption technologies, can be thought of in terms of this chaîne opératoire. With regard to cereal agriculture, this sequence includes primary production, procurement, early processing, distribution, preparation (late processing), consumption, and disposal; each with its associated material culture and potential charred plant byproducts. Ethnographic research has suggested that dehusking of glume wheats can be done by the pounding of spikelets in deep wooden mortars

183 Hillman 1981, 1983; Jones 1983.

107 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

in order to release the grains, and this may be a more effective approach than processing with grinding slabs.184

In Avgi I, the evidence of Building 5 suggests the presence of the last stages of crop processing indicative of storage, while the evidence from the areas east and north of Building 5 suggest processing activities for the dehusking and sieving of glume wheats. It is not certain if these activities (crop processing, cooking episodes) were undertaken in these open areas, or the byproducts were charred there as fuel, or deposited in these areas charred as part of cleaning operations.

The species present in Avgi II and III are not different from the ones present in Avgi I, although the statistical analysis of the different phases is still ongoing and the quantity of the plant remains retrieved is less.

Great exceptions are two very interesting contexts that belong to Avgi III. The first one is a Pit 282302, where more than ten kilos of clean grain, mainly emmer has been located. The storage of emmer as a clean product represents the end of the chaîne opératoire of grain processing. It is interesting that emmer is the main crop in the majority of the samples; the same observation has also been made for the Neolithic site of Dispilio, located close to and being contemporary to Avgi, next to Lake Orestis.185 The second context of interest of Avgi III is the burial area, where ten small vessels containing cremated human remains were located.186 Some of these vessels, in addition to the cremated bones, also contained charred lentils, providing insights into the ritual landscape of Neolithic Greece. It is suggested that the charred plant remains were not part of the actual cremation, as they would have not survived, but rather were either added in the periphery of the cremation area and got charred, or they were deposited already charred in the small vessels.187

The study of the plant remains of Neolithic Avgi in the final publication of the site will be approached from an anthropocentric point of view offering information on changes in subsistence, agricultural strategies, long-distance trade, different consumer groups, diversity within the population, etc. But the emphasis is firmly on how people used the plants (and animals), that is, on plants as objects.

The faunal remains, by Vasiliki Tzevelekidi

The terrestrial faunal assemblage from Neolithic Avgi has yielded a total record of 8516 Minimum Number of Anatomical Units (MinAU),188 including a rather impressive range of identified species (Table 4). Domesticated taxa constitute the vast majority of the recorded material, with wild game percentages ranging from 1.6 to 2.1%. More specifically, sheep/goat are the dominant species, followed by pigs and cattle (Table 5). A closer examination of Table 5, however, reveals lower percentages of pigs in Avgi I and respectively higher percentages of sheep/goat and cattle. Statistical comparison of species composition among these three phases, suggests that the diversification noted for Avgi I is not fortuitous, implying therefore some degree of differentiation in animal husbandry, which remains to be further explored.

Avgi I Avgi II Avgi III All phases

CATTLE 202 554 125 881 PIG 329 1271 322 1922 SHEEP 950 2983 751 4684 GOAT 148 452 116 716

184 Alonso et al. 2013. 185 Margaritis 2011. 186 Stratouli et al. 2009; 2010. 187 Margaritis 2014. 188 For a detailed analysis of the methodology adopted, including explanation of quantification methods, see Tzevelekidi 2012: 24-25.

108 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Avgi I Avgi II Avgi III All phases

SHEEP and GOAT and SHEEP/GOAT 1098 3435 867 5400 DOG 28 96 38 162 RED DEER 4 17 9 30 ROE DEER 4 9 2 15 DEER 1 1 1 3 CATTLE/RED DEER 1 0 0 1 AUROCHS 2 5 1 8 FOX 4 5 0 9 BOAR 7 18 2 27 HARE 8 23 13 44 MARTEN 0 3 0 3 BADGER 0 0 0 0 BEAR 3 1 0 4 WOLF 0 1 0 1 LYNX 0 1 0 1 BEAVER 0 2 0 2 HEDGEHOG 0 2 0 2 TORTOISE 1 0 0 1 TOTAL MinAU per phase 1692 5444 1380 8516 TOTAL % per phase 19.9 63.9 16.2 DOMESTICATED % 97.9 98.4 98.0 98.2 WILD % 2.1 1.6 2.0 1.8

Table 4. Species representation in Neolithic Avgi, presented per phase and in total (MinAU; sheep/ goat proportionally assigned to sheep and goat).

Zooarchaeological analysis of the assemblage from Neolithic Avgi has a two-fold scope; on the one hand, it attempts to verify the degree of post-depositional biases (weathering, recovery, fresh breaks, etc.) that have affected the material under study and tend to distort identification and explanation of various patterns. On the other hand, it approaches questions related not only to management of animal capital but also to its role in the development of intra- and inter-settlement relations, as a means of exchange, gifts, etc. and, more generally, of creating and (re-)negotiating social bonds.

So far, evaluation of taphonomy for Avgi I, II and III, as well as comparison between different features uncovered within the same phase, indicates variation in terms of preservation, e.g. in carnivore attrition, especially obvious between open areas and buildings, implying as a result variation in discard patterns. In addition, there seems to be differential access by carnivores to the remains of different species, however, this observation needs to be further evaluated in the light of the age information of the culled livestock.

Avgi I Avgi II Avgi III ALL PHASES CATTLE 12.4 10.5 9.5 10.7 PIG 20.2 24.2 24.5 23.4 SHEEP and GOAT and SHEEP/GOAT 67.4 65.3 66.0 65.8 Total in MinAU 1629 5260 1314 8203 chi-square tests Avgi I: Avgi II χ2= 13.216; p=.001 Avgi I: Avgi III χ2= 11.782; p=.003 Avgi II: Avgi III χ2= 1.182; p=.584 Table 5. Percentages of main domesticates per phase (MinAU).

109 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Avgi I Avgi II Avgi III All phases

MinAU 1541 5249 39 6829 Open areas % 92,0 99,6 100 97,8 MinAU 134 19 0 153 Building interiors % 8,0 0,4 0 2,2 Total (MinAU) 1675 5268 39 6982 chi-square test: χ2= 346.926; p=.000 Table 6. Distribution of faunal material between deposits identified as open areas and building interiors (MinAU).

The presence of all anatomical units points towards the on-site slaughter of at least the main domesticated species. Body part representation of wild game might have been subjected to different ‘rules’ of carcass management, sharing and consumption. The vast majority of the assemblage is fragmented and the types of recorded fragments suggest a rather intensive exploitation of all bone nutrients (marrow extraction). In addition, based on fragmentation data, different species exhibit different percentages of (more/less) complete bones, an observation that could be related either to varying cooking techniques employed (e.g. roasting versus stewing) or different numbers of consumers involved. In any case, there are several factors that also need to be examined prior to reaching a conclusion, such as culling profiles of different species.

As far as butchery patterns189 are concerned, the percentages of identified cut-marks (dismembering, filleting, skinning, chopping, etc.) are very low, however, there appears to be also a differentiation between butchery techniques among the three phases, an observation that needs to be checked against evidence for weathering of bone surfaces (erosion, encrustation, etc).

Given the presence of good contextual control, the faunal assemblage from Neolithic Avgi offers the potential for identifying changes not only through the lifespan of the settlement, but also between different types of deposits within the same phase. So far, variability in different aspects of animal management and carcass handling has started to emerge, at least between Avgi I, II and III, while differential treatment of live- and dead-stock may also be visible in comparing not only open spaces and building interiors but also in juxtaposing groups of different buildings and their surrounding areas combined, with implications for potential variation in access, handling, distribution, use and ownership of animals and, as a consequence, their secondary products, within the same phase, demonstrating the presence of more complicated social structures.

The fish remains, by Tatiana Theodoropoulou

The aquatic element was also part of the subsistence strategies of the Neolithic inhabitants of Avgi, as suggested by the presence of fish bones and shell remains. More than 7,425 fish bones have been essentially retrieved from wet-sieved samples. They belong to freshwater species, primarily small- sized Cyprinids, i.e. small species or smaller individuals of larger species, such as carp, bream, rudd, roach, or tench. These fish are typical of lowland waters with gentle gradients, slow rates of flow, silty substrates and well-oxygenated water but often turbid or opaque, and were caught in the streams around the site. Most species live in shoals and some, such as roach and carp, prefer densely vegetated riverbeds, while during spawning or in the juvenile phase they can be found along riverbanks. Fishing could take place from late winter to autumn, but it would be particularly efficient in summer and early autumn for catching these nearshore shoal fish. The fish behaviour and range of size, generally between 15-25 cm, suggest that fishing took place with fine mesh nets, either fixed, cast or seine nets attached to the riverbank.190 Creels or baskets may also have been

189 Analysis of butchery practices follows Binford 1981 and Isaakidou 2004. 190 On the use of small round or rectangular cast nets at Kastoria Lake in 1860, as witnessed by Mary Walker, see Rouskas 1997: 52.

110 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

used.191 Angling, used to catch some carnivore Cyprinids, is a less efficient technique in terms of quantities of catches, and more difficult to catch smaller fish. Although there are no indications of hooks, net-weights or netting needles at the site,192 natural materials -i.e. perishable or not easily identified in the archaeological material (wood/bark, grooved/notched stones or shells)- could have been used as sinkers and floaters, while needles made of wood may have been used for net mending.193 Freshwater mussels (120 complete or fragmented valves and more than 3,000 fragments) were also collected in the soft substrates of the neighbouring streams, flooded pits or canals, either simultaneously with or independently of fishing activities.

It seems that both fish and molluscs acquired a complementary role in the diet of Neolithic Avgi.194 Fresh catches of small Cyprinid fish would arrive whole to the village and were probably prepared whole, as the presence of all anatomical areas -including fragile cranial bones and fins/spines - in the material suggest. The latter also accounts for the increased numbers of fish remains, which should be interpreted with caution in terms of actual yields, as they probably represent a significantly lower number of fish caught. The reconstructed sizes also point to fairly low fish protein input. The absence of larger fish is interesting. It could be related to specific ecological conditions in the neighbouring waters at the time of occupation of the site, to technological constraints (e.g. absence of adequate fishing tackle to catch larger fish), or to cultural choices (e.g. distaste for the flesh of larger fish, etc.). It is also possible that Avgi completed the fish protein input with imported fish fillets from the neighboring lakeside village of Dispilio, as such an activity has been suggested for this settlement.195

Preliminary spatial analysis suggests significant concentrations of fish remains in the Western Sector of the site, possibly food waste, especially inside Pit 2812001. Other assemblages that produced fair numbers of fish include open areas of Avgi II at the Western Sector. Fish bones do not preserve direct evidence of fire, so possibly cooking took place in vessels, in some liquid. Freshwater mussels are mostly distributed in the western and central parts of the excavated area. They were probably consumed after short heating to open their valves, as tool marks to force their opening are scarce on the material. Occasional presence of burnt valves suggests a direct, post- refuse contact with fire. Few shells and fish have been recovered from the destruction layers of the Avgi I buildings.

However, aquatic resources did not only serve as food. The Neolithic settlement of Avgi offers a more complex relation of Neolithic inland groups of the North Aegean to the resources offered by aquatic environments, both freshwater and more distant, marine ones. The robust and shiny shell of the freshwater mussel Unio sp. was occasionally exploited by the inhabitants of Avgi, potentially after consumption of the meat. Some valves exhibit abrasion marks that indicate a possible use as scrapers. Others are cut and modified into shapes, potentially as preliminary steps for bead- making.

A yet more complex relation, beyond subsistence, to more distant waterscapes is attested through the presence of marine shells at the site. The latter follows a well-attested trend among inland N. Aegean of this period, particularly during the Late Neolithic.196 For these inland communities, the sea was not part of their subsistence ground and they did not depend on it for survival. Their interest for the raw materials that the sea had to offer should be seen in a wider

191 On a passage from (V.16), describing the use of baskets to catch fish from boats in the middle of a lake. 192 cf. the lakeside settlement of Dispilio, Stratouli 2008; Theodoropoulou and Stratouli 2009. 193 The use of nets is both archaeologically and ethnographically attested (Cleyet-Merle 1991; Pétrequin 1984, 1997). 194 cf. with other Neolithic settlements in Western Macedonia, Theodoropoulou 2008; 2014b. 195 With respect to the complete absence of caudal parts of large catfish in the last occupation phase, Theodoropoulou 2008, and ethnographic references on the trade of fresh or cured catfish from the lake of Kastoria in the eighteenth century in Rouskas 1997: 85. 196 Theodoropoulou 2011, 2014b.

111 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

cognitive and cultural context, as argued elsewhere.197 The distance and difficulty to acquire them may have ascribed materials from this remote and seemingly meaningful waterscape with a more symbolic character.

Most marine specimens found at Avgi had been collected dead on the Aegean and/or Ionian shores and travelled over a long distance through direct harvesting expeditions to the coasts -or in combination with other activities - or via exchange networks between coastal sites and inland villages (see below details on the ornaments made of marine shell).198

The microvertebrate remains, by Katerina Papayianni

An assemblage of almost 1500 microvertebrate remains has been recovered from the heavy residue of the soil samples, which were floated after systematic recovery from all contexts and all sectors. The majority of the samples comes from the Western Sector of the site, which has also been the focus of the excavation and the area that yielded the denser habitation remains.

The array of identified genera and species includes hedgehogs (Erinaceus concolor), mice (Mus macedonicus, Apodemus sylvaticus vel flavicollis), dormice (cf. Glis glis), voles (Microtus sp.), shrews (Crocidura cf. suaveolens), lizards (Lacertidae), as well as unidentified snake and amphibian bones. Among those vertebrates, only the hedgehog could have been hunted and consumed either for its meat or for its pelt and thorns by the Neolithic Avgi inhabitants; a burnt mandible of a hedgehog was recovered from the collapse debris of Building 5 in the Western Sector.

The two mice, the vole and the shrew that were identified have been recovered from all phases of the site and from the deposits of buildings and pits: the destruction layers of Buildings 1, 3 and 5; the fill of different pits of either refusal or symbolic use; in open areas among the buildings. Some of their bones were burnt, others bore evidence of the digestive system of some bird of prey. The discovery of some hundreds of small mammal remains among the Neolithic debris signifies the synanthropic behavior of them, which was triggered by the abundant food and the safety against predation that they found in the human habitat.199

The small mammals found in Avgi are all native to the Greek and Balkan fauna200 and were attracted to the settlement by the stored grain or by the food leftovers; since mice and shrews were found inside buildings and pits, they were definitely infesting them. Nevertheless, some of them were also preyed upon by birds, since their remains have also been found in the deposits of open areas.201 This could mean either that the birds were hunting their prey in the close vicinity of the settlement and in the open areas among the buildings, where the small mammals could not hide, or that they were roosting in the Avgi ruins, after the settlement’s abandonment.

In terms of environmental reconstruction, all identified small mammal indicate an open grassland alternating with water sources. The modern natural surroundings of Avgi are cultivated fields; there is a small stream near the site. Such would be the habitat during the Neolithic, judging by the small mammal species. The existence of a water source is supported by the unidentified amphibian remains.

197 Theodoropoulou 2011, 2014a. 198 Theodoropoulou 2014a. 199 Vigne 2011. 200 Mitchell-Jones et al. 1999. 201 Andrews 1990.

112 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

APPENDIX 2: Material Culture Pottery, by Nikos Katsikaridis

The pottery sample under study since 2006 comes from selected features and assemblages. These include buildings 2a, 5, 7 and the ‘open areas’ of Avgi I, the ‘open areas’ of Avgi II and the pits of Avgi III.202

The study conducted as part of the author’s PhD thesis focused on approximately 2 tons (94989 sherds) of selected material. It is estimated that this sample represents approximately 25% of the total assemblage unearthed.

Methodological issues

The goals of the ceramic analysis are the classification of the material, observations on technology, shape and decorating techniques, function of the vessels and finally the distribution of the material in the contexts of the Neolithic settlement. Overall, the study of the pottery assemblage provides a chronological assessment for the three phases of Avgi. The pottery sample was chosen beforehand, according to the excavated spaces of activity. The basic tool is macroscopic analysis, which aids the typological classification of vessels and sherds into pottery wares, which are constituted by the color and surface treatment of the pots. Decoration is not considered as a basic part of the ware, but as an additional step implemented by the Neolithic craftsmen which adds meaning to the pot. Other data that are considered are fabric, form and technological characteristics of the pottery. Beside macroscopic analysis, petrographic analysis of selected sherds is underway, while the preliminary results are already available.203

Technology

The majority of the pottery is manufactured using the ‘coil building’ technique, which in the case of LN pottery is usually very difficult to discern, because in most cases the secondary surface treatment is of high quality and capable to erase the marks of primary forming.204 In some cases, the technique is established because vessels break at the point of joints between the coils, or, in extremely rare cases, coils are barely visible on the surface, because secondary treatment could not erase them. Smaller vessels are probably manufactured using pinching, although it is not certain if combinations of various techniques were at use, along with moulds or other tools, as is suggested by ethnographic case studies.

The pottery fabrics are characterized by mica, quartz, other types of small stones, sand, organic inclusions and limestone, which is generally in accordance with local geology. Preliminary petrographic analysis suggests that there is possibly mixing of clays or pastes during manufacture.205 The surface treatment is generally of high quality, with most sherds having a burnished or polished exterior, but also some smoothed and rough exterior surfaces are attested.

Regarding firing, there are probably two basic firing techniques in use, an oxidising and a non- oxidising one, which produce the different pottery wares and colors, both of them never exceeding 900o C and of short duration, as is attested by the non-fully oxidised cores of the pots. Up to this point, no pottery firing installations were excavated at Avgi.

Pottery wares

The pottery assemblage from the older layers of the settlement is characterised by black topped, black burnished and black polished wares, which form a widespread tradition in northern Greece

202 As the floor levels of the Avgi III buildings were not preserved, the material attributed to the latest episodes of occupation derives exclusively from pits and pit-like features. 203 Papadias and Saridaki 2016. 204 Kotsakis 2010: 66-75. 205 Papadias and Saridaki 2016.

113 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

and the Balkans for the given era.206 These three wares constitute nearly half, 43,76%, of the total sherd count. Black topped and black burnished vessels appear simultaneously at Avgi, so there is no clear chronological differentiation between these two wares. Brown polished, burnished and smoothed are also recorded in large numbers, which comprise 26,30% of the sherds. Red categories, polished, burnished and smoothed comprise 9,18% of the total. It seems that the preferred surface treatment for the Neolithic potters of Avgi is burnishing. The preferred decorating techniques are rippling, incising, impressing, but also plastic decoration and scraped technique are attested in a few sherds. In many cases, combinations of the aforementioned decorating techniques are recorded.

Painted categories are polychrome, black-on-red, red-on-cream, matt-painted, brown-on-white and brown-on-red, but all painted categories count very few sherds in number (681 in total – 0,72%) and are seldom whole painted pots. Painted categories at Avgi could not be associated with the Thessalian ‘classical Dimini’ painted wares (Dimini brown-on-cream, Otzaki black-on-red) or the Eastern Macedonian black-on-red and ‘Akropotamos style’ painted pottery.207

The final ware that is recorded in the assemblage is a painted and incised category, that is detected in the layers of Avgi III. The characteristics of the ware are deep incisions, probably filled with paste, that create geometrical patterns which are in turn filled with dark red and black colors. This ware is associated with a limited number of shapes, possibly only scoops and footed bowls.

Pottery shapes (inventory)

Pottery shapes are divided in three major categories: open vessels (Figure 34), closed vessels (Figure 35) and hole-mouth vessels (Figure 36).

Open vessels have a rim diameter that is (or equals) the largest diameter of the body. They are further separated by their rim diameter into small (up to 15cm rim diameter), medium (16-30cm diameter), and large categories (more than 31cm in diameter), thus resulting in cups, bowls and basins. Taking body shape into consideration and based on basic geometrical shapes the open vessels were classified into the following types: cups are hemispherical, conical, carinated with ‘S-’shaped profile, cylindrical and with inverted rims. Most cups have flat bases, but also some slightly convex bases are attested. These vessels have no true handles, but lugs on the outer body or near the lip are recorded. Cups are typically considered as pots for personal use.

Bowls are hemispherical, conical, carinated, carinated with inverted or everted rims, cylindrical, with ‘S’-shaped profile, and finally very shallow vessels called ‘dishes’. Bowls are formed mostly with flat bases, but some slightly convex, concave and ring bases are attested. These vessels are associated with lugs on the body or near the rim, some of them perforated horizontally or vertically. Bowls, which are the medium sized open vessels, are considered suitable for a number of uses. They are also the most abundant shape in the inventory. Basins are the largest open vessels, which are suitable for a multitude of uses in the social life of the settlement, and are fewer in number than cups and bowls. Their shapes are hemispherical, conical, carinated and with everted rim and are all formed with flat bases.

In the wider category of open vessels there are also some shapes that could not be classified in the above mentioned categories: these are plates (or pans), ‘strainers’, ‘sieves’ or ‘cheese-pots’, ‘fruitstands’, pithoi, ‘lids’ and ‘altars’.

206 Demoule et al. 1988: 35; Demoule and Perlès 1993: 391-393; Kalogirou 1994: 133-140, 149-153; Hitsiou 2003: 82-83; Urem-Kotsou 2006: 92-95; Vlachos 2009: 53-57; Kalogirou and Urem-Kotsou 2013; Bonga 2013: 133-179. 207 Koukouli-Chrysanthaki 1996: 115; Malamidou 2014.

114 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Plates or pans are elliptical–shaped, large shallow vessels, usually related to cooking activities.208 Pans (also known as ‘plateaux’ in French) are widespread in the LN of northern Greece. At Avgi, some vessels of the type have characteristic triangular lugs at the long sides of their elliptical body. ‘Strainers’, ‘sieves’ or ‘cheese-pots’ are open vessels that have numerous holes on their bodies, which are considered vessels for dairy products, lighting or even beekeeping.209 Strainers appear at the layers of Avgi II, but are few in number.

‘Fruitstands’ are recorded in three types: small footed bowls usually brown burnished or smoothed, medium sized polychrome painted ‘fruitstands’ and large red polished and slipped vessels, usually decorated with scraped technique. In rare cases, large fruitstands are black burnished or black-topped.

Pithoi are large open vessels with a rim diameter from 20-69cm and have conical or cylindrical shape or spherical lower body. The vessels of this type are formed with flat bases, ranging from 12-40cm in diameter. True handles or large lugs at the lower part of the vessels are recorded.

‘Lid’ is the conventional name given to small vessels with cylindrical body and flat edges.210 They are black or brown burnished with small holes near the flat edge, probably serving as tying points. Quadrangular vessels, very few in number, with small cylindrical feet Figure 34. Open vessels: bowls (Drawing: Odysseas Metaxas, Photo: Nikos Katsikaridis). called ‘altars’ are found in the two first phases of the settlement.

Closed vessels have restricted openings, therefore rim diameter is not a clear indication of the vessel’s size. Their diameter ranges from 9-36cm. There are collared jars, carinated collared jars, collared jars with spherical body and pear-shaped collared jars. Collared jars appear in many sizes, from small to medium

208 Tsirtsoni and Youni 2002: 103-110; Urem-Kotsou and Kotsakis 2007: 225-246; Urem-Kotsou et al. 2008: 619-629. 209 Kalogirou 1994: 154-161; Katsikaridis forthcoming and references therein. 210 Korkuti 1995.

115 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

and some large sized, so they can be paralleled with the cups and bowls. Amphorae are considered closed vessels with vertical handles, but also a conical neck, which either opens or restricts the vessel’s rim.

The wider category of closed vessels includes biconical shapes, small aryballoid vessels and ‘scoops’, these peculiar vessels are characterised by their diagonal opening, and are generally considered as vessels for lighting, carrying coal or incense burning.211 Scoops appear in the layers of Avgi II.

Hole-mouth vessels are medium and large jars with a large rim diameter, which could not be classified either as open or closed vessels, and with a diameter between 15-46cm. Their body shape generally resembles closed shapes, but their wide opening allowed easy access to their contents, and also enabled potters to apply surface treatment to the inner walls. Hole-mouth vessels have spherical or pear-shaped bodies and usually flat bases.

Pottery function/use

Pottery analysis of Avgi identified four main functional categories of Figure 35. Closed vessels: jars (Drawing: Odysseas Metaxas, Photo: Nikos Katsikaridis). vessels: cooking vessels, storage vessels, vessels for serving and consumption of food or drink (tableware) and a small group of vessels used as funerary urns in the burial area.212

Cooking vessels form a heterogeneous category, suggesting that several food preparation techniques were at use. There are plates (or pans) for baking, closed vessels and hole-mouth vessels for stewing or boiling, but also black burnished and usually decorated quadruped vessels with traces of fire at the feet or near the base, probably used in the simmering position. Traces for use over fire are recorded at some carinated vessels with small bases, probably used as chance cooking vessels or signifying vessels that became cooking utensils in a secondary use. Cooking vessels are found in many contexts, inside buildings, near hearths and thermal structures or even at open areas of activity.

211 Chapman 1988; Weinberg 1965; Mlekuz 2007; Biagi 2003; Bonga 2013: 219-235, 2014. 212 Stratouli et al. 2009, 2010.

116 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figure 36. Left: Plate/pan and black burnished lid, Right: hole-mouth jars (Drawing: Odysseas Metaxas, Photo: Nikos Katsikaridis).

Storage vessels are associated with long-term or short-term storage, of liquids or other products. Pithoi, collared jars, amphorae and hole-mouth vessels were available at Avgi for storing products: food, water or other drinks. Also, other infrequent products, such as honey, beeswax, dried fruit or nuts, or oils for cosmetics could be stored in the small ‘aryballoid’ vessels, most of which were found in the burial area, after they had fulfilled their primary function. Storage vessels were unearthed in many areas of the settlement, inside buildings, in the open areas of activity or in pits.

Small and medium sized open vessels, such as cups and bowls, are suitable for serving and consumption of food and drink, either communally or individually. Cups and especially bowls are the most abundant vessel type in the settlement. They are also the most elaborate and highly decorated in the whole assemblage. Other vessels that could be related to serving are ‘fruitstands’ or some highly decorated and painted closed vessels such as amphorae. Vessels for consumption and serving of food/drink are found in many contexts of the site, but they are strongly associated with the open areas North of Building 5 and East of Building 5.

Most of the pots were discarded when broken, but some sherds were reused even broken, recycled or with modified characteristics. Some closed pots acquired secondary use when the neck broke off and the point of breaking was modified as a new rim. Ceramic discs were created by chipping and rubbing broken sherds. Some of them were perforated, while others are broken in half or four pieces, after they were modified into discs. Some sherds were used as tools for scraping, most often at one or two of their sides. Finally, broken pots were used as building material in the base of thermal structures.

Chipped stone, by Niels H. Andreasen

Excavations at Neolithic Avgi have uncovered a large assemblage of chipped stone (Figure 37). This short review includes only material from secure Neolithic contexts in the Western Sector, where excavation has produced 1073 pieces of chert.

117 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Chert is dominant at Avgi and falls into two main groups: The larger group includes a variety of shiny, translucent, and cloudy cherts of mainly grey and light brown colors. On the microscopic level, the density of Radiolaria is low, but some large fossils do occur.213 The second group is of matt and opaque, dark reddish-brown cherts with a moderate occurrence of smaller- sized radiolarian fossils. There are also small numbers of chipped quartzite and obsidian. The color spectrum indicates that some of the settlement’s procurement took place at a secondary source that mixes chert from several geological formations. Smooth, abraded cortex shows that some raw materials have been subject to intense, mechanical weathering. This points to deposits from a high- velocity river, perhaps the nearby Aliakmonas, where pebble deposits could have been exposed in the edges of the river valley in the past. Yet, a survey of the river- and stream transects within the valley failed to produce cherts suitable for tool production within at least a 10km radius from the settlement. Only a few radiolarian dark reddish- brown pebble cherts of inferior quality were found, and they are

Figure 37. Chipped stone artefacts. Top: Bifacial projectile points from unsuitable for tool production. Western Sector, Middle: Prismatic blades of various raw materials, Bottom: Selection of cores (Photos: Niels H. Andreasen). Most of the cortical and a few cores, however, retain a chalky limestone surface, which is often smooth but with no signs of battering as would have been expected from river pebbles. It seems that such materials were extracted and imported from unidentified, non-fluvial deposits outside the Aliakmonas valley. While some procurement of pebble cherts could have taken place within or immediately around the Aliakmonas valley, it took place in combination with the import of lithics that were brought, traded, or exchanged into the valley. Given the several phases of occupation at Avgi that stretched throughout a large part of the Middle and Late Neolithic, it is highly probable that several sources of raw material were exploited, and different strategies were pursued to obtain them.

The small assemblage of cores mostly falls within the ‘exhausted’ category, and few have a maximum dimension that exceeds 40mm. The average weight is only 12 grams, and most are worked down

213 The identification of the two groups of chert was supported by a petrographical study of nine archaeological samples (Andreasen 2009).

118 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

from larger nodules. In their discarded form, many cores are small, opportunistic flake cores, none of which are evidence for the production of the regular blades and bladelets that are so common at the site. Some evidence for blade manufacture is based on the recovery of a sub-conical blade core, a couple of blade core fragments, and a few crested blades. A few cores would have produced bladelets or a combination of flakes and bladelets before they were discarded.

The presence of working debris and rejuvenation flakes suggest some knapping activity at Avgi. Yet, on-site knapping appears to be connected with expedient flake production, modest blade/ bladelet production and the maintenance of retouched tools, rather than the first phase of core reduction and organised blade production.

Perhaps not surprisingly, the most common tool type within the assemblage is laterally retouched blades, often truncated. Many unretouched blades carry traces of mostly subtle or moderate usewear. The presence of gloss and uni- or bilateral usewear suggests that these segments were used as inserts in sickles used for cutting vegetation that contained silica. Such plants include various wild grasses as well as reeds, straw, etc.214

There is modest evidence for the curation of tools. Although some tools demonstrate more prolonged use or recycling than others, the typical use-life of a tool does not include substantial reworking or reshaping for the completion of other tasks. Once tools broke, they were discarded with minimal attempts to rejuvenate them or transform them into other functional implements.

In terms of on-site production locales, accumulations of lithic debitage are primarily represented by construction fill and some pit deposits, obviously indicating secondary refuse. Local maintenance and modest production were likely occurring in or near individual households as demand required, rather than within more specific, circumscribed workshop areas.215

Ground stone (or macrolithics), by Tasos Bekiaris

Excavations at Neolithic Avgi brought to light one of the largest ground stone assemblages known from Neolithic Greece, comprising approximately 8000 artefacts.216 So far, the study of Neolithic Avgi’s ground stone assemblage has focused on the macroscopic and contextual analysis of 2907 ground stone objects217 (Figure 38) deriving from undisturbed Neolithic deposits and on the petrographic investigation of their raw materials.218

The dominant rock types are sandstones, conglomerates, limestones, granites, serpentinites and gabbros. Most of them were collected from regional secondary sources, such as the Aliakmonas River and its tributaries, while a limited number seem to derive from primary outcrops (i.e. marl sandstone slabs, ophiolites).219 Two marble vases originate from the , specifically the island of Naxos, as their geochemical analysis has indicated.220 Based on their technological attributes and use-wear patterns the ground stones of Neolithic Avgi have been ascribed to the following categories:

214 Anderson 1980; Unger-Hamilton 1985. 215 Costin 1991. 216 Similar assemblages, in terms of quantity, technology and typology, are known from other Neolithic sites in Northern Greece, such as the recently excavated site of Koromilia at Kastoria (Bekiaris 2016), the sites of Kleitos (Chondrou 2018), Toumba Kremasti Koilada (Stroulia et al. 2017) in Kozani and Makriyalos (Tsoraki 2008) in Pieria. 217 Bekiaris 2018; 2020. 218 Bekiaris et al. 2017; Stergiou and Theodoridou 2017; Stergiou et al. forthcoming. 219 Bekiaris et al. 2017; Stergiou et al. forthcoming. 220 Stergiou et al. forthcoming.

119 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Figure 38. Ground stone implements. a: Grinding slab, b: grinder, c: abrader, d: macehead, e: edge tool, f: polisher, g: (Drawing and photos: Tasos Bekiaris).

1. Grinding tools:221 The group includes 704 grinding slabs and grinders that were used for the processing of foodstuffs and other substances (e.g. pigments). They are usually made of sandstones, conglomerates, granites and gneisses. Flaking and pecking were employed for the making of most grinding implements. Some of them were produced onsite. Grinding tools were heavily curated artefacts, subjected to regular maintenance via pecking, but also to reshaping and recycling in order to acquire new functional roles when needed (e.g. grinding slabs used as grinders, grinders used as pounders, grinding tools used as abraders). As documented in many other Greek Neolithic sites,222 the majority of the grinding tools (607 out of 704 objects) of Neolithic Avgi are fragmentarily preserved. Moreover, several of them bear evidence of deliberate breakage, indicated by the presence of impact points and the recognition of specific breakage patterns.223

221 Bekiaris 2020. 222 cf. Stroulia 2010; Stroulia 2018; Stroulia et al. 2017; Bekiaris et al. in press. 223 Bekiaris 2018.

120 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

2. Abrading tools: The group comprises 82 expediently designed artefacts made almost exclusively of quartz sandstone. Their workfaces are either flat, or in the form of an open cavity or groove with a V or U transverse section. They were probably used for shaping objects made of stone, wood, antler and bone.

3. Pebble tools: This subset comprises 884 items. These are expediently designed limestone and sandstone pebbles that were used with their faces, sides and edges in an active abrasive mode. Most of them exhibit high sheen on their work faces, often accompanied by subtle scratches. Such tools, commonly characterised as polishers, have been associated with pottery burnishing or with the accentuation of the appearance of wood or bone.

4. Edge tools/celts: Τhe category comprises 176 stone implements that have acquired an acute edge through deliberate modification. These tools were originally hafted to a shaft made of wood, bone or antler. Most edge tools were made of serpentinites and gabbros from the outcrop of Mavro Vouno, at a distance of about 10km from the site. The techniques of flaking, pecking, sawing, grinding and polishing have been exploited in order to manufacture and repair the edge tools. Almost every shows traces of multiple episodes of resharpening, redesigning and/or recycling (e.g. as hammers/pounders). Some of them also bear traces of deliberate breakage. There is no evidence for the onsite production of celts. Therefore, it has been suggested,224 that the inhabitants of Neolithic Avgi could have acquired those tools from the nearby Neolithic settlement of Dispilio.

5. Percussion tools: The category includes 105 expedient rocks that share a globular shape and have been used mostly in an active percussive manner with their faces, ends and edges (e.g. as hammers and pounders). They are made of hard rocks, including gabbros, peridotites and quartzes. Such implements could have been used in activities involving the breakage or the reduction of materials.

6. Discoid tools: This group comprises 393 water-rolled pebbles and cobble of a discoid shape. They are all expediently designed implements, made of siltstones and fine-grained sandstones and have traces of abrasion and polish on one or both their broad faces.

7. Pestles: Only 10 elongated cylindrical artefacts can be classified as pestles. With the exception of two specimens made of gabbro through pecking, all of them are expedient items made of limestone. Their ends bear traces of abrasion and light pounding.

8. Multiple-use tools: The group includes 116 active and 3 passive implements that could have acted in an abrasive and percussive manner concomitantly.

9. Miscellanea: This heterogeneous category comprises 183 ground stone objects of several subtypes. The dominant groups are (a) burnt slabs: 63 marl sandstone slabs, exhibiting traces of fire. Based on ethnographic parallels and experimental work it has been suggested that they may have been used for cooking,225 (b) maceheads: 5 elaborately crafted, perforated items, of a spheroid shape made of serpentinites and marble. All of them are preserved at a fragmentary state, (c) burnt pebbles: 84 limestone pebbles, heavily burnt. They were used as the substructure for the floors of hearths,226 (d) spheroid pebbles: 22 expedient sandstone and limestone pebbles with no signs of use. They could have acted as ‘slingshot bullets’, (e) stone vases: 2 fragments of elaborately crafted marble vases. They seem to belong to open, hemispherical bowls, (f) pigments: 3 pieces of bauxite, with multiple facets bearing traces of abrasion.

224 Bekiaris et al. 2017; Stratouli 2019. 225 Bekiaris 2018; Stratouli et al. 2013. 226 cf. Kalogiropoulou 2013.

121 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

10. Raw materials/Debitage: 155 items of various raw materials.

11. Items of undetermined function: 96 extremely fragmented objects that cannot be classified to any category. They don’t preserve signs of use or manufacture.

Ornaments, by Tasos Bekiaris, Tatiana Theodoropoulou and Georgia Stratouli

Figure 39. Various ornaments from Neolithic Avgi: Top Left: Amber pendant. Bottom Left: Fragment of a Spondylus ring with mending holes. Middle: Suggested reconstruction of bracelets and necklaces with stone, bone, shell and malachite beads. Left: A synthesis of various beads, pendants, bracelets and suggested reconstructions of necklaces (Photos: Tasos Bekiaris and Nikos Katsikaridis).

The ornament assemblage from Neolithic Avgi comprises nearly 300 pieces (Figure 39). The study of the assemblage is ongoing. The ornaments are made of several materials, including stone, bone, antler, shell and clay. Several beads - most of which belong to the discoid type - are made of malachite, a copper carbonate hydroxide mineral, resulting from the weathering of copper ores. Copper carbonates are typically associated with various types of copper mineralization, which are sparse throughout the broader region of Kastoria. Therefore, the malachite beads could originate from both local and non-local sources. One pendant is made of amber. As far as we are concerned, this is one of the oldest amber artefacts known from SE Europe. The provenance of both malachite and the amber will be further investigated in the future. Finally, some shell ornaments are made of Spondylus gaederopus, a bivalve marine mollusk, harvested from the Aegean and the Ionian Sea and distributed across the mainland through exchange and communication networks. These shells were worked into elaborate items, commonly found in this period in Western Macedonia (Ifantidis 2019). Beads, ‘buttons’, pendants and ‘bracelets’ made from Spondylus gaederopus feature among the most impressive items.

The ornaments from Neolithic Avgi were classified into the following categories: bracelets/rings, pendants, beads, ear-studs, miscellanea. The beads group was further divided into the following subtypes: barrel beads, cylindrical beads, ‘button-like’ beads and discoidal beads. The preliminary examination of the artefacts indicates that most ornaments were created with the techniques of abrasion and drilling. Some were further decorated with incisions, or subjected to polishing in order to acquire a ‘glossy’ appearance. The production of some ornaments was probably local, as suggested by the presence of preforms or raw materials. A number of half-worked or unworked marine shells (Spondylus gaederopus, Cerastoderma glaucum, Hexaplex trunculus, Aequipecten opercularis) suggest that some marine raw materials may have arrived at the settlement to be transformed into objects on site. Other ornaments would have arrived at the settlement as finished objects. Several items, mostly bracelets or rings made of Spondylus¸ were further curated after their (accidental or intentional?) breakage in order to be maintained or retransformed into a new piece of adornment. Items with mending holes or traces of abrasion on their broken parts are often encountered in the assemblage.

122 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Figurines and other miniature clay objects, by Odysseas Metaxas

Figure 40. Miniature clay objects from Neolithic Avgi: a: Rectangular block with incised decoration, b: Clay stamp seals with various incised motifs, c: miniature clay vessel with incised decoration, d: miniature of an oven with checkerboard decoration that could represent a calendar, e: various anthropomorphic figurines (Drawings: Odysseas Metaxes, Photos: Tasos Bekiaris and Nikos Katsikaridis).

The Neolithic settlement of Avgi revealed a broad array of clay objects, namely figurines and various miniatures, as well as stamp-seals (Figure 40). Regarding the stratigraphic and spatial picture, a rise in the absolute number of figurines and miniatures is evident after the Avgi I phase. Specimens found in open spaces, and more specifically in association with pits, clearly outnumber those found in areas occupied by buildings.

So far 91 anthropomorphic figurines have been recognised, mostly in fragmentary condition. Only one figurine fragment can be identified as male, due to the presence of minute imprints denoting a beard. Every other sexed figurine is female, indicated by specific anatomical features. It is notable that when clothing elements, ornaments or body decoration are indicated, they are accompanied by female genitalia. This might be indicative of processes of gender construction or conceptualization relative to specific activities or modes of body treatment.227 However, most figurines appear as sexless. 228

It is not clear if the large degree of fragmentation is related to the fact that most of them were found in pits, possibly discarded after they were rendered useless due to breakage, or if circulation of previously broken parts was at play. The occasional incidence of the latter might be deduced from the presence of several quite naturalistically made legs without a pair or other conjoining piece, often with incised decoration, which seems to be at odds with the mostly flat and generic body fragments. It is likely that the spatial configuration and the degree of fragmentation was a result of different processes of use and deposition.229 Evidence of some kind of multi-layered biography is not often discernible: a figurine of a pregnant female had developed significant wear at the back, suggesting a long period of use. Furthermore, the marks of the head’s breakage were largely concealed by abrasion.

In line with the picture emerging from the neighbouring site of Dispilio,230 instances of incised decoration are limited to a few examples. This also seems to conform quite well with data from sites in the Republic of . 231 In the central part of Greek Macedonia, incised decoration is more widespread during the Late Neolithic period.232 The region of Thessaly to the south is characterised by a growing percentage of stone figurines during the Late Neolithic.233 This

227 Mina 2014. 228 It is not certain if they were indeed perceived as sexless by their users (bearing a generic ‘human’ attribute in such a case), cf. Bailey 2012 for a discussion on the issue of ‘sexless’ figurines. 229 Whether fragmentation was intentional or not is a contentious issue, see e.g. Budja 2003 for a positive evaluation of intentional fragmentation. 230 Marangou 2001. 231 Naumov 2014. 232 Nanoglou 2006. 233 Nanoglou 2009.

123 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

trend is not represented in the assemblage from Avgi, which includes only two stone figurines. In any case, the figurines from Avgi, which are mostly schematic, share some morphological affinities with all neighbouring regions. Two chronologically late figurine heads, dated to Avgi III, show a marked similarity to figurines popular in southern Greek Final Neolithic sites.

Animal figurines (n=21) are mostly generic representations of domesticates, often hastily made and without defining anatomical features, except for horns in the case of cattle. A carefully formed head of a bear stands out. More important however is the rare fact that it is removable, with a hole facilitating its attachment to a body, possibly made from some perishable material. The above difference might stem from a different conceptualization of domestic and wild animals, with the former being treated as a collective unit, but with the latter having individual (and perhaps hazardous) agency.

Four miniatures depicting elements of furniture, such as tables or chairs, have been unearthed. Four additional enigmatic quite oblong rectangular blocks can bear intricate incised decoration, but one specimen is devoid of any decoration and one has a concave large rectangle face. That might suggest that at least some of them were understood as miniatures of beds (which are otherwise elusive), but more data is needed for a better informed view.234 The miniature of an oven is remarkable due to the presence of a specific kind of checkerboard decoration, which elsewhere in the eastern Balkans has been deemed as indicative of a calendar.235 Miniatures of tools or other small portable objects are rarely recognizable with certainty. A few exceptions exist though, as is the case with a miniature perforated clay hammer. Vessel miniatures are more numerous, since 56 have been found, in a variety of different shapes. Two miniature vessels incorporate aspects of anthropomorphic anatomy, which provides a link with the figurine assemblage. The presence of various types of miniatures brings up the question whether they were entangled in a web of interconnections along with figurines, contributing to the compilation of representational syntheses. It is possible that whatever meaning was attributed to figurines was dependent on such arrangements.236

Stamp-seals (n=10) can in many respects be treated separately, since they do not partake in aspects such as miniaturization and the construction of representational syntheses. They vary in terms of size, shape and decorative motifs. A uniform use cannot be posited for all stamp-seals, as is indicated by the variety in their size and most importantly the width and depth of the decorative incisions or .237 It is interesting that there is one instance of two stamp-seals closely associated within the same spatial and stratigraphic context, even though it has to be stressed that they do not share an identical or similar decorative pattern.

Acknowledgements

Thanks go to the anonymous reviewer for her/his valuable comments, as well as to John Bintliff for accepting the present article for publishing in the Journal of Greek Archaeology. The rescue excavations at Avgi were made possible with the support of a number of people, institutions and funding bodies. We would firstly like to thank the former IZ′ and ΚΘ′ Ephorates of Prehistoric and Classical Antiquities, especially their Directors Dr Maria Lilimbaki-Akamati and Dr Konstantinos Soueref, for their invaluable help in different stages of the project. We particularly thank the mayors of the former Municipality of Aghia Triada, Ioannis Antoniadis and Christos Pantazis, for

234 A specimen from the area of in Eastern Macedonia (Renfrew 2003: 408-9), has been related to tablets with similar decoration from the wider Balkan area, but the undecorated piece from Avgi demonstrates that these were not just the medium for the application of decoration. 235 Chokhadzhiev 1984. 236 Marangou 1996. 237 Regarding stamp-seals some sort of specific use has not been generally pinpointed, and a variety of functions is likely, cf.. Prijatelj 2007.

124 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

providing us resources and accommodation, as well as the mayor of the Municipality of Argos Orestiko, Nikos Totonidis, for his significant collaboration.

The excavations at Avgi were mainly funded by two development programs, i.e. INTERREG IIIA/Cards Greece - Albania and PINDOS. We warmly thank the Institute for the Aegean Prehistory (INSTAP) for funding the study of various archaeological materials during the years 2011-2020. We would also like to thank the Psycha Foundation for the funding of geoarchaeological and other studies of the Avgi research programme, as well as the I.P. Costopoulos Foundation for the financing of various side projects during the years 2013-2014.

Many thanks go to the several dozen colleagues that took part in the Avgi excavations throughout the years, including the young archaeologists and students from the Universities of Thessaly, Athens, Thessaloniki and , our special collaborators (e.g. topographers, architects, geologists, photographers), as well as the workers and the support staff. Finally, special thanks go to a number of friends and residents, who have supported and facilitated our stay in the region in various ways. Special thanks go to our invaluable friend Pantelis Zattas, journalist and photographer, who passed away unexpectedly during the early years of the project.

References Litteraires de l’Universite de Besançon 508) : 17-25. Paris: Diffusé par Les Belles Lettres. Alonso, N., D. Antolín, D. López, F.J. Cantero and Andreou, S., M. Fotiadis and K. Kotsakis 2001. G. Prats 2013. The effect of dehusking on Review of Aegean prehistory V: The Neolithic cereals: experimentation for archaeobotanical and of northern Greece, in T. Cullen comparison, in P.C. Anderson, C. Cheval and A. (ed.), Aegean Prehistory: A review (American Journal Durand (eds) Regards Croisés sur les Outils Liés au of Archaeology Supplement I): 259-327. Boston: Travail des Végétaux (Actes des XXXIIIe Rencontres Archaeological Institute of America. internationales d’archéologie et d’histoire d’Antibes, Andrews, P. 1990. Owls, and fossils. Chicago: The 23-25 Οctobre 2012): 155–168. Antibes: APDCA. University of Chicago press. Anderson, P.C. 1980. A testimony of Prehistoric Aslanis, I. 1990. Οι oχυρώσεις στους oικισμούς του tasks: Diagnostic residues on stone tool working Βορειοελλαδικού χώρου και η περίπτωση του edges. World Archaeology 12: 181-194. Διμηνίου, in Μελετήματα του Κέντρου Έρευνας Andreasen, N. 2009. The influence of availability Ελληνικής και Ρωμαϊκής Αρχαιότητας 10: 19-64. and accessibility on direct-access exploitation Atalay, S. and C. Hastorf 2006. Food, meals and daily of siliceous rock in Late Neolithic and Bronze activities: food habitus at Neolithic Çatalhöyük. Age Greece: a petro-archaeological study American Antiquity 71: 283–319. (unpublished Fellowship Report submitted for Bailey, D.W. 2012. Figurines, corporeality and the the Wiener Laboratory, ASCSA). origins of gender, in D. Bolger (ed.) Companion to Andreasen, N. 2011. The chipped stone artefacts Gender Prehistory: 244-264. Oxford: John Wiley. from Neolithic Avgi: some observations from Bailey, D.W. and A. Whittle 2008. Living well the continued study of the West Sector in 2011. together? Questions of definition and scale in (Online publication - http://www.neolithicavgi. the Neolithic of south-east and central Europe. gr/?page_id=492). In D.W. Bailey, A. Whittle and D. Hofmann (eds) Andreou, S. and K. Kotsakis 1987. Διαστάσεις του Living well together: settlement and materiality in χώρου στην κεντρική Μακεδονία: αποτύπωση the Neolithic of south-east and central Europe: 1–7. της ενδοκοινοτικής και διακοινοτικής Oxford: Oxbow. χωροοργάνωσης, in Αμητός Ι: τιμητικός τόμος για Banerjea, R.Y., M. Bell, W. Matthews and A. Brown τον καθηγητή Μ. Ανδρόνικο: 57–86. Thessaloniki: 2015. Applications of micromorphology Aristotle University of Thessaloniki. Andreou, S. and K. Kotsakis 1994. Prehistoric rural to understanding activity areas and site communities in perspective: the Langadas formation processes in experimental floors. Survey Project, in P. Doukellis and L.G. Mendoni Archaeological and Anthropological Sciences 1: 89- (eds) Structures Rurales et Sociétés Antiques. Actes 112. du colloque de Corfou, 14-16 Mai 1992 (Annales Barrett, J.C. 2006. A perspective on the early architecture of western Europe, in J. Maran,

125 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

C. Juwig, H. Schwengel and U. Thaler (eds) Bogaard, A. 2004. The nature of early farming in Constructing Power: Architecture, Ideology and Social Central and South-east Europe. Documenta Practice: 15–30. Hamburg: Lil Verlag. Praehistorica 31: 49-58. Bailey, D.W. 2000. Balkan Prehistory: Exclusion, Bogdanović, M. 1988. Architecture and structural Incorporation and Identity. London and New York: features, in A. McPherron and D. Srejović (eds) Routledge. Divostin and the Neolithic of Central : 35–141. Bekiaris, T. 2012. Ground stone artefacts from the Pittsburgh: University of Pittsburgh Press. Neolithic Settlement of Avgi: A preliminary Bonga, L.A. 2013. Late Neolithic Pottery from report. (Online publication -http://www. Mainland Greece, ca. 5300-4300 B.C. Unpublished neolithicavgi.gr/?page_id=494&langswitch_ PhD dissertation, The Temple University. lang=en). Bottema, S. 1974. Late Quaternary Vegetation History Bekiaris, T. 2016. Τριπτά αντικείμενα από νεολιθικές of Northwestern Greece. Unpublished PhD κοινότητες της ΠΕ Καστοριάς: Παρατηρήσεις και dissertation, University of Groningen. ερωτήματα, in Το Αρχαιολογικό Έργο στην Άνω Bradley, R. 2005. Ritual and Domestic Life in Prehistoric Μακεδονία, Δ΄ Eπιστημονική Συνάντηση, 17-18 Europe. London and New York: Routledge. December 2016, Kastoria (oral presentation). Budja, M. 2003. Seals, contracts and tokens in the Bekiaris, T. 2018. Ground stone technologies in Balkans Early Neolithic: where in the puzzle. the Neolithic of Northern Greece: the case Documenta Praehistorica 30: 115-130. of Neolithic Avgi, Kastoria. Unpublished PhD Canti, M.G. 2003. Aspects of the chemical and dissertation, Aristotle University of Thessaloniki. microscopic characteristics of plant ashes found Bekiaris, T. 2020. Ground stone technology in in archaeological soils. Catena 54: 339-361. context: consumption of grinding tools and Cleyet-Merle, J.J. 1990. La Préhistoire de la Pêche. Paris: social practice at Neolithic Avgi, NW Greece. Editions Errance. Journal of Lithic Studies 7(3): 24 pages. Chapman, J.C. 1988. Ceramic production and social Bekiaris, T., Ch. Stergiou and S. Theodoridou differentiation: The Dalmatian Neolithic and the 2017. Making choices in a Neolithic landscape: Western Mediterranean. Journal of Mediterranean raw materials and ground stone technology Archaeology 1 (2): 2–25. in Neolithic Avgi, NW Greece, in A. Sarris, E. Chapman, J.C. 1997. Places as timemarks. The social Kalogiropoulou, T. Kalayci and L. Karimali construction of Prehistoric landscapes in Eastern (eds) Communities, Landscapes, and Interaction in Hungary, in G. Nash (ed.) Semiotics of Landscape: Neolithic Greece, Proceedings of the International Archaeology of Mind (British Archaeological Conference, 29-30 May, 2015: 415- Reports International Series 661): 31-45. Oxford: 433. of America: International Archaeopress. Monographs in Prehistory. Chapman, J.C. 1999. Deliberate house-burning in Bekiaris. T., D. Chondrou, I. Ninou and S.M. Valamoti the prehistory of Central and Eastern Europe, this volume. Food-processing ground stone in A. Gustafsson and H. Karlsson (eds) Glyfer och tools in the Aegean Neolithic and Bronze Age: A Arkeologiska rum eu Vänbok till Jarl Nordblach: 113– synthesis of the published data. Journal of Greek 25. Göteborg: Norum Grafiska. Archaeology 5. Chapman, J.C. 2000a. Fragmentation in Archaeology. Biagi, P. 2003. The rhyton of the Balkan Peninsula: People, Places and Broken Objects in the Prehistory chronology, origin, dispersion, and function of a of South Eastern Europe. London and New York: Neolithic ‘cult vessel’. Journal of Routledge. XVI-XVII: 16–26. Chapman, J.C. 2000b. Pit-digging and structured Bickle, P. and E. Kalogiropoulou 2017. ‘Very like the deposition in the Neolithic and Copper Age. Neolithic’: the everyday and settlement in the Proceedings of the Prehistoric Society 66: 61-87. European Neolithic, in P. Bickle, V. Cummings, Chapman, J.C. 2010. Houses, households, villages D. Hofmann and J. Pollard (eds) The Neolithic of and proto-cities in southeastern Europe, in D.W. Europe. Papers in Honour of Alasdair Whittle: 7–23. Anthony and J.Y. Chi (eds) The Lost World of old Oxford: Oxbow Books. Europe: the Danube Valley, 5000–3500 BC: 58–73. New Binford, L.R. 1981. Bones: Ancient Men and Modern York and Princeton: New York and Princeton Myths. New York: Academic Press. University Press. Bintliff, J.L., E. Farinetti, K. Sarri and R. Sebastiani Chapman, J.C. and B. Gaydarska (with K. Hardy) 2006. Landscape and early farming settlement 2006. Does enclosure make a difference? A view dynamics in central Greece. Geoarchaeology 21 from the Balkans, in A. Harding, S. Sievers and (7): 665 - 674. N. Venclová (eds) Enclosing the Past: Inside and

126 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Outside in Prehistory: 20–43. Sheffield: J.R. Collis mouse. Nature Scientific Reports (19/5/2020) Publications. 10:8276. DOI: 10.1038/s41598-020-64939-9 Chapman, J.C. and B. Gaydarska 2007. Parts and Demoule, J.P. and C. Perlès 1993. The Greek Neolithic: Wholes: Fragmentation in Prehistoric Context. a new review. Journal of World Prehistory 7: 355– Oxford: Oxbow. 416. Chokhadzhiev, S. 1984. Arheologicheski danni za Demoule, J.P., K. Gallis and L. Manolakakis 1988. kalendar v nachaloto na kamenno-mednata Transition entre les cultures néolithiques de epokha. Arkheologiya 2-3: 1-7. et de Dimini: Les catégories céramiques, Chondrou, D. 2018. Η τριπτή λιθοτεχνία από τους Bulletin de Correspondance Hellénique 112: 1-58. προϊστορικούς οικισμούς Κλείτος Ι και Κλείτος Efstratiou, N. 1990. Prehistoric habitation ΙΙ της λεκάνης της Κιτρινής λίμνης Κοζάνης. and structures in northern Greece: an Unpublished PhD dissertation, Aristotle ethnoarchaeological case-study, in P. Darque and University of Thessaloniki. R. Treuil (eds) L’ Habitat Egéen Préhistorique: Actes Chrysostomou, P. 1991. Οι νεολιθικές έρευνες στην de la Table Ronde International par le Centre National πόλη και την επαρχία Γιαννιτσών κατά το 1991. de la Recherche Scientifique, l’ Université de Paris et l’ Το Αρχαιολογικό Έργο στη Μακεδονία και Θράκη 5: École Française d’ Athènes, 23-25 Join 1987 (Bulletin de 111–26. Correspondance Hellénique Supplement. XIX): 33–4. Chrysostomou, P. 1996. Η νεολιθική κατοίκηση στη Paris: L’ École Française d’ Athènes. βόρεια παράκτια ζώνη του άλλοτε Θερμαϊκού Efstratiou, N., M.P. Fumanal, C. Ferrer, D. Urem κόλπου (επαρχία Γιαννιτσών). Το Αρχαιολογικό Kotsos, A. Curci, A. Tagliacozzo, G. Stratouli, Έργο στη Μακεδονία και Θράκη 10: 159–72. S.M. Valamoti, M. Ntinou, E. Badal, M. Madella Chrysostomou, P. 2001. Νέα στοιχεία από τη and K. Skourtopoulou 1998. Excavations at the Neolithic settlement of Makri, , Greece Νεολιθική έρευνα στην επαρχία Γιαννιτσών. (1988-1996): preliminary report. Saguntum 31: Μια άγνωστη μορφή προϊστορικής γραφής. Το 11–62. Αρχαιολογικό Έργο στη Μακεδονία και Θράκη 15: Facorellis, G. and Y. Maniatis 2002 Αποτελέσµατα 489-499. χρονολόγησης δειγµάτων µε τη µέθοδο του 14C Chrysostomou, Α., Poloukidou, C. and Α. Prokopidou από τον προϊστορικό οικισµό του Δισπηλιού 2001. Επαρχιακή οδός Αψάλου-Αριδαίας: η Καστοριάς, in G.H. Hourmouziadis (ed.) Δισπηλιό: ανασκαφή του νεολιθικού οικισμού στη θέση 75000 Χρόνια Μετά: 289-294. Thessaloniki: Γραμμή. Το Αρχαιολογικό Έργο στη Μακεδονία και University Studio Press. Θράκη 15: 513–524. Fotiadis, M., A. Hondroyianni-Metoki, A. Kalogirou, Costin, C.L. 1991. Craft specialization: Issues in Y. Maniatis, A. Stroulia and C. Ziota 2019. Megalo defining, documenting, and explaining the Nisi Galanis (6300-1800 BC): constructing a organization of production. Archaeological cultural sequence for the Neolithic of West Method and Theory 3: 1-56. Macedonia, Greece. The Annual of the British Courty, M.A., P. Godberg and R. Macphail 1989. Soils School at Athens 114: 1-40. and Micromorphology in Archaeology. Cambridge: Gallis, K. 1985 A Late Neolithic Offering from Cambridge University Press. Thessaly. Antiquity 59: 20–24. Cucchi, T., J.C. Auffray and J.D. Vigne 2012. History Grammenos, D.V. 2010. Οι νεολιθικές έρευνες στην of house mouse synanthropy and dispersal in Ελλάδα – με έμφαση στη βόρεια – κατά τις the Near East and Europe: a zooarchaeological πρόσφατες δεκαετίες. In N. Papadimitriou and insight, in M. Macholan, S.J.E. Baird, P. Z. Tsirtsoni (eds) Η Ελλάδα στο ευρύτερο πλαίσιο Munclinger and J. Piàlek (eds) Evolution in των Βαλκανίων κατά την 5η και 4η χιλιετία π.Χ.: Our Neighbourhood: The House Mouse as a Model 30–7. Athens: Museum of Cycladic Art – N.P. in Evolutionary Research: 65–93. Cambridge: Goulandris Foundation. Cambridge University Press. Guttmann, E.B., I.A. Simpson, D.A. Davidson and S.J. Cucchi, T., K. Papayianni, S. Cersoy, tL. Aznar- Dockrill 2006. The management of arable land Cormano, A. Zazzo, R. Debruyne, R. Berthon, A. from prehistory to the present: Case studies from Bălășescu, A. Simmons, F. Valla, Y. Hamilakis, F. the Northern Isles of Scotland. Geoarchaeology Mavridis, M. Mashkour, J. Darvish, R. Siahsarvi, F. 21(1): 61-92. Biglari, C.Α. Petrie, L. Weeks, A. Sardari, S. Maziar, Jones, G.E.M. 1984. Interpretation of archaeological C. Denys, D. Orton, E. Jenkins, M. Zeder, J. Searle, plant remains: The application of ethnographic G. Larson, F. Bonhomme, J.C. Auffray and J.D. models from Greece, in W. van Zeist and W.A. Vigne 2020. Tracking the Near Eastern origins Casparie (eds) Plants and Ancient Man, Studies and European dispersal of the western house in Palaeoethnobotany, Proceedings of the Sixth

127 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Symposium of the International Work Group for Crete. Unpublished PhD dissertation, University Palaeothnobotany, Groningen, 30 May–3 June 1983: of London. 43–61. Rotterdam. Kalicz, N. and P. Raczky 1987. The late Neolithic of the Halstead, P. 1995. From sharing to hoarding: the Tisza region: a survey of recent archaeological Neolithic foundations of Aegean Bronze Age research. In L. Tálas and P. Raczky (eds) The society, in R. Laffineur and W.D. Niemeier (eds) late Neolithic of the Tisza region. A survey of recent Politeia: Society and State in the Aegean Bronze Age excavations and their findings: Hódmezővásárhely- (Aegeum 12): 11–20. Liège: University of Liège Gorza, Szegvár-Tűzköves, Öcsöd-Kováshalom, Vésztő- Press. Mágor, Berettyóújfalu-Herpály: 11–30. Budapest- Halstead, P. 1999. Neighbours from hell? The Szolnok: Kossuth Press. household in Neolithic Greece, in P. Halstead Kalogiropoulou, E. 2013. Cooking, space and the (ed.) Neolithic Society in Greece: 77–95. Sheffield: formation of social identities in Neolithic Sheffield Academic Press. Northern Greece: evidence of thermal Halstead, P. and K. Kotsakis 2002. Paliambela, in D. structure assemblages from Avgi and Dispilio in Blackman (ed.) Archaeology in Kastoria. Unpublished PhD dissertation, Cardiff Greece (Archaeological Reports 48): 80 University. Hill, J.D. 1995. Ritual and Rubbish in the Iron Age of Kalogiropoulou, E. 2014. Αναζητώντας κοινωνικές Wessex: a Study on the Formation of a Particular ταυτότητες: η συμβολή των θερμικών Archaeological Record (British Archaeological κατασκευών στην οργάνωση του χώρου στη Report 242). Oxford: Archaeopress. νεολιθική Μακεδονία, in Ε. Stefani, N. Merousis Hillman, G.C. 1981. Reconstructing crop husbandry and A. Dimoula (eds) 1912-2012. Εκατό Χρόνια practices from charred remains of crops, in R. Έρευνας στην Προϊστορική Μακεδονία. Πρακτικά Διεθνούς Συνεδρίου. Αρχαιολογικό Μουσείο Mercer (ed.) Farming Practice in British Prehistory: Θεσσαλονίκης 22-24 Νοεμβρίου 2012: 359–372. 123–162. Edinburgh: Edinburgh University Press. Thessaloniki: Archaeological Museum of Hillman, G.C. 1984a. Interpretation of archaeological Thessaloniki. plant remains: the application of ethnographic Kalogirou, A. and D. Urem-Kotsou 2013. models from , in W. van Zeist and W.A. Νεολιθική κεραμική στη Μακεδονία. (Online Casparie (eds) Plants and Ancient Man, Studies publication - https://proistoria.wordpress.co in Palaeoethnobotany, Proceedings of the Sixth m/%CF%80%CE%B1%CF%81%CE%AC%CF%8 Symposium of the International Work Group for 1%CF%84%CE%B7%CE%BC%CE%B1/?fbclid=I Palaeothnobotany, Groningen, 30 May–3 June 1983: wAR3uUgBmeirsBESuIrFk6hdOjcZXntVsuo- 1–42. Rotterdam. UD1kqYhiAbf5dKF5gsASKIJA). Hitsiou, E. 2003. Production and circulation of the Kalogirou, A. 1994. Production and Consumption of Late Neolithic pottery from Makriyalos (phase Pottery in Kitrini Limni, West Macedonia, Greece, B), Macedonia, northern Greece. Unpublished 4500-3500 ΒC. Unpublished PhD dissertation, PhD dissertation, University of Sheffield. Indiana University. Hodder, I. 1998. The domus: some problems Karamitrou-Mentesidi, G., N. Efstratiou, J.K. reconsidered, in M. Edmonds and C. Richards Kozlowski, M. Kaczanowska, Y. Maniatis, A. Curci, (eds) Understanding the Neolithic of north-western S. Michalopoulou, A. Papathanasiou and S.M. Europe: 84–101. Glasgow: Gruithne Press. Valamoti 2013. New evidence on the beginning Hondroyianni-Metoki, A. 2009. Μη Oικιστικές of farming in Greece: The Early Neolithic Xρήσεις του Xώρου στους Νεολιθικούς Οικισμούς: settlement of Mavropigi in western Macedonia Το Παράδειγμα της Τούμπας Κρεμαστής (Greece). Antiquity 87. PAGES? Κοιλάδας. Unpublished PhD dissertation, Karkanas, P. and P. Goldberg. 2010. Phosphatic Aristotle University of Thessaloniki. features, in G. Stoops, V. Marcelino and F. Mees Hourmouziadis, G.H. (ed.) 2002. Δισπηλιό: 7500 Χρόνια (eds) Interpretation of Micromorphological Features Μετά. Thessaloniki: University Studio Press. of Soils and Regoliths: 521-541. Oxford: Elsevier. Ifantidis, F. 2008. Λίθινοι δακτύλιοι από το Νεολιθικό Katsikaridis, N. (forthcoming) Νεολιθική κεραμική Δισπηλιό: Ιστορίες κόσμησης και θραύσματα από τον οικισμό Αυγής Καστοριάς: τεχνολογία, ιστοριών. Ανάσκαμμα 1: 79-92. χρήση και κατανομή στο χώρο. Unpublished PhD Ifantidis, F. 2019. Practices of Personal Adornment in dissertation, Aristotle University of Thessaloniki. Neolithic Greece. Oxford: Archaeopress. Kloukinas, D. 2014. Neolithic building technology Isaakidou, V. 2004. Bones from the Labyrinth: Faunal and the social context of construction practices: Evidence for Management and Consumption of the case of northern Greece. Unpublished PhD Animals at Neolithic and Bronze Age Knossos, dissertation, Cardiff University.

128 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Kloukinas, D. 2017. Pictures of home: regional Krachtopoulou, K. 2009. Geoarchaeological perspectives into the Neolithic building investigations of the environs of the Neolithic technology of northern Greece, in A. Sarris, site of Avgi. (Οnline publication - http://www. E. Kalogiropoulou, T. Kalayci and L. Karimali neolithicavgi.gr/?page_id=118&langswitch_ (eds) Communities, Landscapes, and Interaction lang=en). in Neolithic Greece. Proceedings of International Kyrillidou, S. 2016. Settlement form and high- Conference, Rethymno, 29-30 May 2015: 167–86. resolution microhistory of houses in prehistoric United States of America: International Greek Macedonia. Unpublished PhD dissertation, Monographs in Prehistory. University of Reading. Korkuti, M. 1995. Neolithikum und Chalkolithikum in Kyrillidou, S. in preparation. On the outdoors Albanien. Mainz am Rhein: von Zabern. of Aegean prehistory: the micromorphology Koromila, G. 2015. From Archaeological of open spaces in Neolithic and Bronze Age Sediments to Human Practice: a comparative communities of Northern Aegean, in E. Hatzaki geoarchaeological study of open areas in the and P. Zafeiriadis (eds) On the Outdoors of Aegean Neolithic of northern Greece. Unpublished PhD Prehistory. Proceedings of a Colloquium held at The dissertation, University of Reading. Norwegian Institute at Athens, 17-18 May 2018. Kotsakis, K. 1999. What tells can tell: social space and Malamidou, D. 2014. Ταυτότητες ανθρώπων και settlement in the Greek Neolithic, in P. Halstead διακοσμημένα κεραμικά σκεύη κατά τη Νεότερη (ed.) Neolithic Society in Greece: 66–76. Sheffield: Νεολιθική στη Βόρεια Ελλάδα: παραδείγματα Sheffield Academic Press. από την κεραμική παράδοση ‘μαύρο σε ερυθρό’, Kotsakis, K. 2010. Η κεραμική της Νεότερης in Ε. Stefani, N. Merousis and A. Dimoula Νεολιθικής στην Βόρεια Ελλάδα, in N. (eds) 1912-2012. Εκατό Χρόνια Έρευνας στην Papadimitriou and Z. Tsirtsoni (eds) Η Ελλάδα στο Προϊστορική Μακεδονία. Πρακτικά Διεθνούς Ευρύτερο Πολιτισμικό Πλαίσιο των Βαλκανίων κατά Συνεδρίου. Αρχαιολογικό Μουσείο Θεσσαλονίκης την 5 και 6 χιλιετία π.Χ.: 66-75. Athens: Museum of 22-24 Νοεμβρίου 2012: 527-536. Thessaloniki: Cycladic Art. Archaeological Museum of Thessaloniki. Kotsakis, K. (ed.) 2013. Νέες Aνασκαφικές Έρευνες Malamidou, D. and S. Papadopoulos 1993. στη Νεολιθική Μακεδονία. Archaeology Online. Ανασκαφική έρευνα στον προϊστορικό οικισμό (Online publication - https://www.archaiologia. Λιμεναρίων. To Αρχαιολογικό Έργο στη Μακεδονία gr/blog/2013/08/26/%CE%BD%CE%AD%CE%B5 και Θράκη 7: 559–572. %CF%82-%CE%B1%CE%BD%CE%B1%CF%83%CE Mallol, C., F.W. Marlowe, B.M. Wood and C.C. Porter= %BA%CE%B1%CF%86%CE%B9%CE%BA%CE%AD 2007. Earth, wind, and fire: ethnoarchaeological %CF%82-%CE%AD%CF%81%CE%B5%CF%85%CE signals of Hadza fires. Journal of Archaeological %BD%CE%B5%CF%82-%CF%83%CF%84%CE%B7- Science 34(12): 2035-2052. %CE%BD%CE%B5%CE%BF%CE%BB%CE%B9%CE Mallol, C., S.M. Mentzer and C.E. Miller 2017. %B8%CE%B9/). Combustion features, in C. Nicosia and G. Kotsakis, K. 2014a. Domesticating the periphery: Stoops (eds) Archaeological Soil and Sediment New research into the Neolithic of Greece. Pharos Micromorphology: 299–326. Hoboken, NJ: Wiley. 20 (I): 41-73. Maniatis, Y., Z. Tsirtsoni, Ch. Oberlin, P. Darcque, Ch. Kotsakis, K. 2014b. Εκατό χρόνια νεολιθικής έρευνας Koukouli-Chryssanthaki, D. Malamidou, T. Siros, στη Μακεδονία: τάσεις και κατευθύνσεις, M. Miteletsis, S. Papadopoulos and B. Kromer in Ε. Stefani, N. Merousis and A. Dimoula 2014. New 14C evidence for the Late Neolithic- (eds) 1912-2012. Εκατό Χρόνια Έρευνας στην Early Bronze Age transition in Southeast Europe. Προϊστορική Μακεδονία. Πρακτικά Διεθνούς Open Journal of Archaeometry 2: 43-50. Συνεδρίου. Αρχαιολογικό Μουσείο Θεσσαλονίκης Maniatis, Y. and M. Pappa 2020. Radiocarbon dating 22-24 Νοεμβρίου 2012: 707-711. Thessaloniki: of the Neolithic settlement at Makriyalos, Pieria, Archaeological Museum of Thessaloniki. North Greece. Radiocarbon 62: 467-483. Koukouli-Chrysanthaki, H. 1996. Μακεδονία-Θράκη, Marangou, C. 1996. Assembling, displaying, and in G.A. Papathanasopoulos (ed.) Νεολιθικός dissembling Neolithic and Eneolithic figurines Πολιτισμός στην Ελλάδα: 112-116. Athens: and models. Journal of European Archaeology 4: Museum of Cycladic Art. 177-202. Kouli, K. 2002. Δισπηλιό και παλυνολογία: Marangou, C. 2001. Three-dimensional clay προσεγγίζοντας το παλαιοπεριβάλλον, in G.H. representations from Dispilio, Lake of Kastoria, Hourmouziadis (ed.) Δισπηλιό: 7500 Χρόνια Μετά: Northern Greece, in B. Raftery and J. Hickey 303–316. Thessaloniki: University Studio Press. (eds) Recent Developments in Wetland Research

129 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

(Monograph Series, vol. 2 and WARP Occasional Paper Matthews, W., C.A.I. French, T. Lawrence and 14): 171-181. Dublin: University College Dublin. D.F. Cutler 1996. Multiple surfaces: the Margaritis, E. 2007. Αρχαιοβοτανικές μελέτες στο micromorphology, in I. Hodder and The Νεολιθικό oικισμό Αυγής Καστοριάς 2005–2007. Çatalhöyük Research Trust (eds) On the Surface: (Online publication - http://www.neolithicavgi. Çatalhöyük 1993-95: 301-342. Cambridge and gr/?page_id=114&langswitch_lang=el). London: McDonald Institute for Archaeological Margaritis, E. 2010. Storage, lathyrism and gathering Research and British Institute of Archaeology at in Neolithic Dispilio. Aνάσκαμμα 5: 76-88. Ankara. Margaritis, E. 2014. Acts of destruction and acts of Melfos, V. and G. Stratouli 2002. Η προέλευση των preservation: plants in the ritual landscape of πρώτων υλών για τα τέχνεργα του οικισμού, in Prehistoric Greece, in G. Touchais, R. Laffineur, G.H. Hourmouziadis (ed.) Δισπηλιό: 7500 Χρόνια F. Rougemont, H. Procopiou and S. Andreou Μετά: 175–183. Thessaloniki: University Studio (eds) PHYSIS. L’Environnement Naturel et la Relation Press. Hommemilieu dans le monde Egéen Protohistorique. Metaxas, Ο. and G. Stratouli 2017. Figurines from Actes de la 14e Rencontre Égéenne Internationale, the neolithic settlement of Avgi, the neolithic Paris, Institut National d’Histoire de l’Art, 11-14 site of Trita near Koromilia and the hellenistic Décembre 2012, Aegaeum 37: 279-286. lakeside nekropolis of Krepeni, Kastoria District, Martinez, S. 1999. A new look at house construction NW Greece, in: P. Adam-Veleni, A. Koukouvou, O. techniques: current research at Dikili Tash, Palli, E. Stefani and E. Zografou (eds) FIGURINES, A Neolithic site of Eastern Macedonia. Το microcosmos of clay, Exhibition Catalogue: Catalogue Αρχαιολογικό Έργο στη Μακεδονία και Θράκη 13: Nos. 24-29 & 632. Thessaloniki: Archaeological 63–68. Museum of Thessaloniki. Milic, M. 2014. PXRF characterization of obsidian Matthews, W. 1995. Micromorphological from central , the Aegean and central characterisation and interpretation of Europe. Journal of Archaeological Science 41: 285- occupation deposits and microstratigraphic 296. sequences at Abu Salabikh, Southern Iraq, in A.J. Miller C.E., N.J. Conard, P. Goldberg and F. Berna. Barham and R.I. Macphail (eds) Archaeological 2010. Dumping, sweeping and trampling: Sediments and Soils: Analysis, Interpretation and Experimental micromorphological analysis Management. Papers from the 10th Anniversary of anthropogenically modified combustion Conference of the Association for Environmental features, in I. Théry-Parisot, L. Chabal, S. Archaeology held at the Institute of Archaeology, Costamagno (eds) The Taphonomy of Burned UCL, July 1989: 41-74. London: University College, Organic Residues and Combustion Features in Institute of Archaeology. Archaeological Contexts, Proceedings of the round Matthews, W. 2001. Methodological approaches in table, May 27-29 2008, CEPAM, Paleοethnology 2: 25- microstratigraphic analysis of uses and concepts 37. of space at Tell Brak, in M. Fortin (ed.) Recherches Mina, M. 2014. Don’t throw the baby out with the en Archéométrie: 177-197. Québec: CELAT. bathwater: integrating sex in the study of Greek Matthews, W. 2005a. Micromorphological and Neolithic anthropomorphic figurines, in C.M. microstratigraphic traces of uses and concepts Ursu and S. Țerna (eds) Anthropomorphism and of space, in I. Hodder (ed.) Inhabiting Çatalhöyük. Symbolic Behaviour in the Neolithic and Copper Age Reports from the 1995-99 Seasons: 355-398. Communities of South-Eastern Europe, Bukovina, Cambridge and London: McDonald Institute for Studies into South-East European Prehistory, Archaeological Research and British Institute of Bucovina Museum: 123-160. Suceava: Editura Karl Archaeology at Ankara. A. Romstorfer. Matthews, W. 2005b. Life-cycle and life-course Mitchell-Jones, A.J., G. Amori, W. Bogdanowicz, B. of buildings, in I. Hodder (ed.) Çatalhöyük Kryštufek, P.J.H. Reijnders, S. Spitzenberger, M. Perspectives: Reports from the 1995-99 Seasons: 125- Stubbe, J.B.M. Thissen, V. Vohralík and J. Zima 149. Cambridge; London: McDonald Institute for 1999. The Atlas of European Mammals. London: Archaeological Research and British Institute of Academic. Archaeology at Ankara. Mlekuž, D. 2007. Sheep are your mother: Rhyta and Matthews, W. 2010. Geoarchaeology and taphonomy the interspecies politics in the Neolithic of the of plant remains and microarchaeological Eastern Adriatic. Documenta Praehistorica 34: 267– residues in early urban environments in the 280. Ancient Near East. Quaternary International Mould, C.A. and K.A. Wardle 2000a. The stratigraphy 214(1–2): 98-113. and phases, in C. Ridley, K.A. Wardle and C.A.

130 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Mould (eds) Servia I: Anglo-Hellenic Rescue preliminary report on the 1993–1995 excavations. Excavations 1971–73: 21–70. London: The British Journal of Field Archaeology 26: 177–95. School at Athens. Pappa, M., P. Halstead, K. Kotsakis and D. Urem- Mould, C.A. and K.A. Wardle 2000b. The architectural Kotsou 2004. Evidence for large-scale feasting at remains, in C. Ridley, K.A. Wardle and C.A. Mould Late Neolithic Makriyalos, northern Greece, in (eds) Servia I: Anglo-Hellenic Rescue Excavations P. Halstead and J. Barrett (eds) Food, and 1971–73: 71–105. London: The British School at Society in Prehistoric Greece: 16–44. Oxford: Oxbow Athens. Books. Nanoglou, S. 2006. Regional perspectives on the Pétrequin, P. 1984. Gens de l’eau, gens de la terre. Neolithic anthropomorphic imagery of northern Ethno-archéologie des communautés lacustres. Paris: Greece. Journal of Mediterranean Archaeology 19(2): Hachette. 155-176. Pétrequin, P. 1997. Les sites littoraux néolithiques de Nanoglou, S. 2009. Representing people, constituting Clairvaux–les–Lacs et de Chalain (Jura), III, Chalain worlds: multiple ‘’ in the southern station 3, 3200-2900 av. J.-C., Vol. 2. Paris: Éditions de Balkans. Documenta Praehistorica 36: 283-297. la Maison des Sciences de l’Homme. Naumov. G. 2014. Together we stand – Divided we Peristeri, Κ. 2004. Ανασκαφική έρευνα 2003 fall: the representation and fragmentation στον νεολιθικό οικισμό Αρκαδικού Δράμας among Govrlevo and Zelenikovo figurines, στα αγροτεμάχια 542, 545 ιδιοκτησίας F. Republic of Macedonia, in C.M. Ursu and S. Țerna Χατζηγιαννίδη. To Αρχαιολογικό Έργο στη (eds) Anthropomorphism and Symbolic Behaviour in Μακεδονία και Θράκη 18: 25–32. the Neolithic and Copper Age Communities of South- Prijatelj, A. 2007. Digging the Neolithic stamp-seals Eastern Europe, Bukovina, Studies into South-East of SE Europe from archaeological deposits, texts European Prehistory, Bucovina Museum: 161-186. and mental constructs. Documenta Praehistorica Suceava: Editura Karl A. Romstorfer. 34: 231-56. Nielsen, A.E. 1991. Trampling the archaeological Pyke, G .1996. Structures and architecture, in R.J. record: An experimental study. American Rodden and A.K. Wardle (eds) Nea Nikomedeia Antiquity 56: 483-503. I. The Excavation of an Early Neolithic Village in Ntinou, M. 2002. Η παλαιοβλάστηση γύρω από τον Northern Greece 1961–1964: 39–54. London: The οικισμό και η χρήση της, in G.H. Hourmouziadis . (ed.) Δισπηλιό: 7500 Χρόνια Μετά: 175–183. Reingruber, A., G. Toufexis, N. Kyparissi-Apostolika, Thessaloniki: UniversityStudioPress. M. Anetakis, Y. Maniatis and Y. Facorellis 2017. Ntinou, M. 2008. Η εφαρμογή της ανθρακολογίας Neolithic Thessaly radiocarbon dated periods στο νεολιθικό οικισμό της Αυγής. (Online and phases Documenta Praehistorica 44: 34-53. publication - http://www.neolithicavgi. Renfrew, C. 2003. Special clay objects: cylinders, gr/?page_id=116&langswitch_lang=el). stamp seals, counters, biconoids, and spheres, Ntinou, M. 2010. Παλαιοπεριβάλλον και ανθρώπινες in E.S. Elster and C. Renfrew (eds) Prehistoric δραστηριότητες: η ανθρακολογία στο λιμναίο Sitagroi: Excavations in Northeast Greece, 1968–1970, οικισμό στο Δισπηλιό Καστοριάς. Ανάσκαμμα 3: Volume 2: The Final Report: 403-419. Los Angeles: 45–60. The Cotsen Institute of Archaeology at UCLA. Ntinou, M. and E. Badal 2000. Local vegetation and Ridley, C., K.A. Wardle and C.A. Mould (eds) Servia I: charcoal analysis: an example from two Late Anglo-Hellenic Rescue Excavations 1971–73. London: Neolithic sites in northern Greece, in P. Halstead The British School at Athens. and C. Frederick (eds) Landscape and Land Use Robb, J. 2007. The Early Mediterranean Village: Agency, in Postglacial Greece: 38–51. Sheffield: Sheffield Material Culture, and Social Change in Neolithic . Academic Press. Cambridge: Cambridge University Press. Papadias Y. and Ν. Saridaki. 2016. Neolithic Rodden, R.J. 1962. Excavations at the Early Neolithic Settlement of Avgi, Kastoria, NW Greece: site at Nea Nikomedeia, Greek Macedonia (1961 Preliminary results of the pottery petrographic season). Proceedings of the Prehistoric Society 28: analysis. (Online publication - http://www. 267–288. neolithicavgi.gr/?page_id=1956&langswitch_ Rodden, R.J. 1965. An early Neolithic village in lang=en). Greece. Scientific American 212: 83–91. Papayiannis, K. 2008. Microfauna in Neolithic Avgi. Rouskas, G. 1997. Το καστοριανό καράβι. Athens. (Online publication - http://www.neolithicavgi. Roussos, D.N. and N. Kyparissi-Apostolika 2018. gr/?page_id=501&langswitch_lang=en). An integrated geoarchaeological methodology Pappa, M. and M. Besios 1999. The Neolithic to reveal site formation processes linked to settlement at Makriyalos, northern Greece: building practices and pyrotechnology at Middle

131 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Neolithic Imvrou Pigadi, Greece. Geoarchaeology έρευνες. Αρχαιολογικόν Δελτίον 56-59 (2001-2004), 34: 200–220. Χρονικά Β΄3β (2002), ΜΑΚΕΔΟΝΙΑ – ΘΡΑΚΗ: 436-439. Shillito, L.M. 2014. Geoarchaeology at the microscale. Stratouli, G. 2012b. Νομός Καστοριάς. Αυγή. Ο European Geologist 38(2): 58-62. νεολιθικός οικισμός. Αρχαιολογικόν Δελτίον 56- Shillito, L.M., W. Matthews, M.J. Almond and I.D. 59 (2001-2004), Χρονικά Β΄3β (2003), ΜΑΚΕΔΟΝΙΑ – Bull 2011. The microstratigraphy of middens: ΘΡΑΚΗ: 472-473. capturing daily routine in rubbish at Neolithic Stratouli, G. 2012c. Νομός Καστοριάς. Αυγή. Ο Çatalhöyük, Turkey. Antiquity 85(329): 1024-1038. νεολιθικός οικισμός. Αρχαιολογικόν Δελτίον 56- Shillito, L.M. and P. Ryan. 2013. Surfaces and streets: 59 (2001-2004), Χρονικά Β΄3β (2004), ΜΑΚΕΔΟΝΙΑ – phytoliths, micromorphology and changing ΘΡΑΚΗ: 512-514. use of space at Neolithic Çatalhöyük (Turkey). Stratouli, G. 2013. Νέες ανασκαφικές έρευνες στη Antiquity 87: 684–700. Νεολιθική Μακεδονία (μέρος Δ’): Νεολιθικός Souvatzi, S.G. 2008. A Social Archaeology of Households οικισμός Αυγής Καστοριάς: μία αγροτική in Neolithic Greece. Cambridge: Cambridge κοινωνία της 6ης και 5ης χιλιετίας σε μετάβαση. University Press. (Online publication - http://www.archaiologia. Stergiou, Ch. and S. Theodoridou 2017. gr/blog/2013/10/07/νέες-ανασκαφικές- Ορυκτολογική-Πετρογραφική Μελέτη. έρευνες-στη-νεολιθι-4/). (Onlinepublication - http://www.neolithicavgi. Stratouli, G. 2019. ‘Ετούτοι… του νερού’ και gr/?page_id=1983&langswitch_lang=en). ‘εκείνοι… του βουνού’: η δυναμική τόπων και Stergiou, Ch., T. Bekiaris, V. Melfos, St. Theodoridou η διαπραγμάτευση κοινωνικών χώρων στη and G. Stratouli forthcoming. Sourcing Νεολιθική της Καστοριάς, in K. Kotsakis (ed.) H marcrolithics: minerological, geochemical and Aντίπερα Όχθη. Κοινωνικός Χώρος και Ιδεολογία στις provenance investigation on stone artifacts Προϊστορικές Κοινότητες: 107-123. Thessaloniki: from Neolithic Avgi, NW Greece. Aristotle University of Thessaloniki. Stevanović, M. 1997. The age of clay: the social Stratouli, G. and T. Bekiaris 2008. Αυγή Καστοριάς: dynamics of house destruction. Journal of Στοιχεία της βιογραφίας του Nεολιθικού Anthropological Archaeology 16: 334–95. Οικισμού. Το Αρχαιολογικό Έργο στη Μακεδονία Stevanović, M. 2002. Burned houses in the Neolithic και Θράκη 22: 1-10. of south-east Europe, in D. Gheorghiu (ed.) Fire in Stratouli, G. and D. Kloukinas 2020. Transformations Archaeology: 55–62. Oxford: Archaeopress. of settlement space at Neolithic Avgi, NW Greece, Stratouli, G. 2002. Τα εργαλεία από λειασμένο λίθο, in N.N. Tasić, D. Urem-Kotsou and M. Burić (eds) οστό και κέρατο, in G.H. Hourmouziadis (ed.) Making Spaces into Places: the North Aegean, the Δισπηλιό: 7500 Xρόνια Mετά: 155-174. Thessaloniki: Balkans and western Anatolia in the Neolithic: 43-52. University Studio Press. Oxford: BAR Publishing. Stratouli, G., 2004a. Νεολιθική Αυγή Καστοριάς. Ένα Stratouli, G., S. Triantafyllou, T. Bekiaris and N. χωριό πριν από 7.500 χρόνια. Archaeology and Arts Katsikaridis 2009. Η διαχείριση του Θανάτου: 91: 110-116. Χώρος ταφικής πρακτικής στο Νεολιθικό Stratouli, G. 2004b. Νεολιθική Αυγή Καστοριάς Οικισμό Αυγής Καστοριάς. Το Αρχαιολογικό Έργο 2003-2004: Τα πρώτα βήματα ενός ερευνητικού στη Μακεδονία και Θράκη 23: 9-19. προγράμματος. Το Αρχαιολογικό Έργο στη Stratouli, G., N.H. Andreasen, E. Kalogiropoulou, Μακεδονία και Θράκη 18: 661-668. N. Katsikaridis, D. Kloukinas, G. Koromila, E. Stratouli, G. 2005. Μεταξύ πηλών, πλίνθων Margaritis and T. Bekiaris 2011. Σπίτια και αυλές και πασσάλων, μαγνητικών σημάτων στο Νεολιθικό Οικισμό Αυγής Καστοριάς: Το και αρχαιολογικών ερωτημάτων: Τάφροι ‘Κτίριο 5’ και οι γειτονικοί ανοιχτοί χώροι. Το οριοθέτησης και θεμελίωσης στο Νεολιθικό Αρχαιολογικό Έργο στη Μακεδονία και Θράκη 25: Οικισμό Αυγής Καστοριάς. Το Αρχαιολογικό Έργο 7-18. στη Μακεδονία και Θράκη 19: 595-606. Stratouli, G., Τ. Bekiaris, Ν. Katsikaridis and Stratouli, G. 2007. Nεολιθικός Οικισμός Αυγής V. Tzevelekidi 2014a. Ενσωματώνοντας Kαστοριάς 2006-2007: Χωρο-οργανωτικές το παρελθόν, προσδιορίζοντας το παρόν, πρακτικές 6ης και 5ης χιλιετίας. Το Αρχαιολογικό νοηματοδοτώντας το μέλλον: Πρακτικές Έργο στη Μακεδονία και Θράκη 21: 7-14. δομημένης (εν-)απόθεσης στο Νεολιθικό Οικισμό Stratouli, G. 2008. Νεολιθικά αγκίστρια από το Αυγής Καστοριάς, in E. Stefani, N. Merousis and Δισπηλιό Καστοριάς. Ανάσκαμμα 2: 13–24. A. Dimoula (eds) 1912-2012: A Century of Research Stratouli, G. 2012a. Νομός Καστοριάς. Αυγή. Ο in Prehistoric Macedonia, International Conference νεολιθικός οικισμός. Πρώτες ανασκαφικές Proceedings, 22-24 November 2012, Archaeological

132 New excavations in Northwestern Greece: The Neolithic settlement of Avgi, Kastoria

Museum of Thessaloniki: 349-357. Thessaloniki: Ελλάδας: ένα πανόραμα των αρχαιοζωολογικών Archaeological Museum of Thessaloniki. δεδομένων, in E. Stefani, N. Merousis and A. Stratouli, G., T. Bekiaris, N. Katsikaridis, E. Dimoula (eds) 1912-2012: A Century of Research in Maousidou, G. Petsalnikos and E. Tsiola 2014b. Prehistoric Macedonia, International Conference Τρίτα Κορομηλιάς: μια παραποτάμια προϊστορική Proceedings, 22-24 November 2012, Archaeological θέση στην περιφερειακή ενότητα Καστοριάς. Το Museum of Thessaloniki: 453–464. Thessaloniki: Αρχαιολογικό Έργο στη Μακεδονία και Θράκη 28: Archaeological Museum of Thessaloniki. 1-8. Theodoropoulou, T. and G. Stratouli 2009. Fishbones Stratouli, G., T. Bekiaris and N. Katsikaridis 2012. vs. fishhooks: a comparative study from the Στην Αυγή των πρώτων αγροτών: Εκπαιδευτικές Neolithic lakeside settlement of Dispilio, Greece, δράσεις στο Νεολιθικό οικισμό Αυγής. Το in D. Makowiecki, S. Hamilton-Dyer, I. Riddler, N. Αρχαιολογικό Έργο στη Μακεδονία και Θράκη 26: Trzaska-Nartowski and M. Mirosław (eds) Fishes- 1-8. Culture-Environment Through Archaeoichthyology, Stratouli, G., T. Bekiaris and N. Katsikaridis 2013. Ethnography & History, 15th Fish Remains Working Μαθαίνοντας για την παραγωγή της τροφής στο Group (Stowarzyszenie Archeologii Srodowiskowej Νεολιθικό οικισμό Αυγής. Αρχαιολογικές και SAS, Environment and Culture vol. 7): 126–130. βιωματικές δράσεις 2011-2013. Το Αρχαιολογικό Poznań: Bogucki Wydawnictwo Naukowe. Έργο στη Μακεδονία και Θράκη 27: 1-9. Toufexis, G. 1996. House models, in G.A Stratouli, G., S. Triantaphyllou, T. Bekiaris and N. Papathanassopoulos (ed.) Neolithic Culture in Katsikaridis 2010. Manipulation of death: A Greece: 161–162. Athens: Museum of Cycladic Art. burial area at the Neolithic Settlement of Avgi, Touloumis, K. 2002. Η οικονομία ενός Νεολιθικού NW Greece. Documenta Praehistorica 37: 95-104. λιμναίου οικισμού, in G.H. Hourmouziadis (ed.) Stroulia, A. 2010. Flexible stones. Ground stone tools Δισπηλιό: 7500 Xρόνια Mετά: 89–114. Thessaloniki: from Franchthi Cave (Excavations at Franchthi Cave University Studio Press. Greece, Fascicle 14). Indiana: Indiana University Trantalidou, Κ., P. Κarkanas, E. Belegrinou and T. Press. Hatzitheodorou 2005. Σπήλαιο σε μια σημαντική Stroulia, A. 2018. Macrolithics: Ordinary things in δίοδο της δυτικής Μακεδονίας, στο φαράγγι an extraordinary place, in A. Papathanasiou, του Λιβαδοπόταμου. Πρώτη παρουσίαση. Το W.A. Parkinson, D.J. Pullen, M.L. Galaty and P. Αρχαιολογικό Έργο στη Μακεδονία και Θράκη 19: Karkanas (eds) Neolithic Alepotrypa Cave, in the 579-94. Mani, Greece: In Honor of George Papathanassopoulos: Trantalidou, K., E. Belegrinou and Ν. Andreasen 2010. 200‐241. Oxford: Oxbow Books. Pastoral societies in southern Balkan Peninsula: Stroulia, A., L. Dubreuil, J. Robitaille and K. Nelson the evidence from caves occupied during the 2017. Salt, sand, and saddles: exploring an intriguing work face configuration among Neolithic and era. ANODOS, Studies of grinding tools. Ethnoarchaeology 9(2): 119–145. the Ancient World 10: 321- 334. Theodoropoulou, T. 2008. Ο άνθρωπος και η Trenner, J. 2010. Untersuchungen zu den Sogenannten λίμνη: ψαράδες και ψαρέματα στο προϊστορικό Hausmodellen des Neolithikums und Chalcolithikums Δισπηλιό. Ανάσκαμμα 2: 25–45. in Südosteuropa. Bonn: Verlag Dr. Rudolf Habelt Theodoropoulou, T. 2011. Fishing (in) Aegean GmbH. seascapes: early Aegean fishermen and their Triantaphyllou, S. 1999. Prehistoric Makriyalos: a world, in G. Vavouranakis (ed.) The seascape story from the fragments, in P. Halstead (ed.) in Aegean prehistory (Monographs of the Danish Neolithic Society in Greece: 128-135. Sheffield: Institute at Athens 14): 51–69. Athens: Danish Sheffield Academic Press. Institute at Athens. Tringham, R. 2000. The continuous house: A Τheodoropoulou, T. 2014a. Dead from the sea: sea- view from the deep past, in R.A. Joyce and worn shells in Aegean prehistory, in K. Szabó, S.D. Gillespie (eds) Beyond Kinship: Social and C. Dupont, V. Dimitrijevic, L.G. Gastélum and Material Reproduction in House Societies: 115–34. N. Serrand (eds) Archaeomalacology: Shells in Philadelphia: University of Pennsylvania Press. the Archaeological Record. Proceedings of the 11th Tringham, R. 2005. Weaving house life and death into ICAZ International Conference, Archaeomalacology places: a blueprint for a hypermedia narrative, Working Group, 23-28 August 2010 (British in D.W. Bailey, A. Whittle and V. Cummings (eds) Archaeological Reports International Series 2014): 77- (Un)settling the Neolithic: 98–111. Oxford: Oxbow 90. Oxford: Archaeopress. Books. Theodoropoulou, T. 2014b. Οι αλιευτικές Tringham, R. and M. Stevanović 1990. Field research, δραστηριότητες στην προϊστορία της Βόρειας in R. Tringham and D. Krstić (eds) Selevac: A

133 G. Stratouli, T. Bekiaris, N. Katsikaridis, D. Kloukinas, G. Koromila and S. Kyrillidou

Neolithic village in : 57–148. Los Angeles: Unpublished PhD dissertation, Aristotle University of California Press. University of Thessaloniki. Tsirtsoni, Z. 2016. The chronological framework in Urem-Kotsou, D., K. Kotsakis, C. Beck and E.C. Greece and between the late 6th and Stout 2008. Organic residues from the Late the early 3rd millennium BC, and the ‘Balkans Neolithic Makriyalos cooking pots, in Υ. 4000’ project, in Z. Tsirtsoni (ed.) The Human Face Facorellis, N. Zacharias and K. Polikreti (eds) of Radiocarbon. Reassessing Chronology in prehistoric Proceedings of the 4th Symposium of the Hellenic Greece and Bulgaria, 5000-3000 cal BC. New edition Society for Archaeometry (British Archaeological [online]: 15-46. Lyon: MOM Éditions. Reports International Series 1746): 619-629. Oxford: Tsirtsoni, Z. and P. Youni 2002. Neolithic cooking Archaeopress. vessels from Dikili Tash (Eastern Macedonia, Valamoti, S.M. 2011. Σπόροι για τους νεκρούς; Grain Greece): a technological and functional approach, in V. Kilikoglou, A. Hein and Y. for the dead? Αρχαιοβοτανικά δεδομένα από Maniatis (eds) Modern Trends in Scientific Studies τη Μαυροπηγή Κοζάνης, θέση Φυλλοτσαΐρι. To on Ancient Ceramics (British Archaeological Reports Αρχαιολογικό Έργο στην Άνω Μακεδονία 1 (2009): International Series 1011): 103-110. Oxford: 245-256 Archaeopress. van Dyke, R. and S.E. 2003. Archaeologies of Tsokas, G., G. Vargemezis, A. Stampolidis and memory: an introduction, in R. van Dyke and G. Stratouli 2005. Πρώτα Συμπεράσματα της S.E. Alcock (eds) Archaeologies of Memory: 1-13. Γεωφυσικής Διασκόπησης στο Νεολιθικό Οικισμό Massachusetts and Oxford: Blackwell. της Αυγής Καστοριάς. Το Αρχαιολογικό Έργο στη Vigne, J.D. 2011. The origins of animal Μακεδονία και Θράκη 19: 323-332. and husbandry: a major change in the history Tsoraki, C. 2008. Neolithic society in Northern of humanity and the biosphere. Comptes Rendus Greece: The evidence of ground stone artefacts. Biologies 334 (3): 171-181. Unpublished PhD dissertation. Sheffield: Vlachos, D. 2009. Ceramics from Makriyalos II, University of Sheffield. Northern Greece. Unpublished PhD dissertation, Tsouggaris, C. 1997. Ανασκαφικές έρευνες στον University of Sheffield. Νομό Καστοριάς το 1997. To Αρχαιολογικό Έργο Weinberg, S.S. 1965. Ceramics and the supernatural στη Μακεδονία και Θράκη 11: 19–30. cult and burial evidence, in F. Matson (ed.) Tsouggaris, C., Τ. Salonidis, Α. Douma and Η. Sarrigiannidou 2002. Κολοκυνθού: ένας Ceramics and Man: 187–201. Chicago: Aldine παραποτάμιος νεολιθικός οικισμός του Ν. Publishing Company. Καστοριάς. Το Αρχαιολογικό Έργο στη Μακεδονία Weissbrod, L., F. Marshall, F. Valla, H. Khalaili, G. Bar- και Θράκη 16: 625–640. Oz, J.C. Auffray, J.D. Vigne and T. Cucchi 2017. Tzevelekidi, V. 2012. Dressing for Dinner. Bone Butchery The origins of house mice in ecological niches and Deposition at Late Neolithic Toumba Kremastis- created by settled hunter-gatherers in the Koiladas, Northern Greece (British Archaeological Levant 15,000 y ago. Proceedings of the National Reports International Series 2541). Oxford: Academy of Science 114 (16): 4099-4104. Archaeopress. Whittle, A. 1996. Europe in the Neolithic: the creation Tzevelekidi, V. 2013. The zooarchaeological of new worlds. Cambridge: Cambridge University assemblage (terrestrial fauna) from Neolithic Press. Avgi: preliminary results (2012–2013). (Online Williams, M. 2003. Growing metaphors: the publication - http://www.neolithicavgi. agricultural cycle as a metaphor in the later gr/?page_id=110&langswitch_lang=en). prehistoric period of Britain and North-Western Unger-Hamilton, R. 1985. Microscopic striations Europe. Journal of Social Archaeology 3: 223–255 on flint sickle blades as an indication of plant Ζiota, Ch. 2014. Ο οικισμός του Κλείτου Κοζάνης στο cultivation. World Archaeology 17: 121-126. ευρύτερο φυσικό και ανθρωπογενές περιβάλλον Urem-Kotsou, D. and K. Kotsakis 2007. Pottery, cuisine and community in the Neolithic of North της Νεότερης και Τελικής Νεολιθικής Greece, in J. Renard and C. Mee (eds) Cooking up περιόδου in E. Stefani, N. Merousis and A. the Past: Food and Culinary Practices in the Neolithic Dimoula (eds) 1912-2012: A Century of Research in and Bronze Age Aegean: 225-246. Oxford: Oxbow. Prehistoric Macedonia, International Conference Urem-Kotsou, D. 2006. Νεολιθική Κεραμική του Proceedings, 22-24 November 2012, Archaeological Μακρύγιαλου. Διατροφικές συνήθειες και Museum of Thessaloniki: 323-336. Thessaloniki: οι κοινωνικές διαστάσεις της κεραμικής. Archaeological Museum of Thessaloniki.

134