Documenta Praehistorica XLIII (2016) The Cretan Mesolithic in context> new data from Livari Skiadi (SE Crete) Tristan Carter 1, Danica D. Mihailovic´ 2, Yiannis Papadatos 3 and Chrysa Sofianou 4 1 McMaster University, Department of Anthropology, CA [email protected] 1 2 University of Belgrade, Department of Archaeology, RS 3 National and Kapodistrian University of Athens, Department of History and Archaeology, GR 4 Hellenic Ministry of Culture and Sports, Ephorate of Antiquities of Lasithi, GR ABSTRACT – Investigations at Livari (south-eastern Crete) produced a small Mesolithic chipped stone assemblage, whose techno-typological characteristics situate it within an ‘early Holocene Aegean is- land lithic tradition’ (9000–7000 cal BC). The material provides antecedent characteristics for the lithics of Crete’s founder Neolithic population at Knossos (c. 7000–6500/6400 cal BC). The idiosyn- crasies of the Knossian material can be viewed as a hybrid lithic tradition that emerged from inter- action between migrant Anatolian farmers and indigenous hunter-gatherers. Small quantities of Me- lian obsidian at Livari attest to early Holocene maritime insular networks, knowledge of which likely enabled the first farmers’ successful voyage to Crete. IZVLE∞EK – Raziskave na najdi∏≠u Livari (na jugovzhodu Krete) so pokazale navzo≠nost drobnega mezolitskega kamnitega inventarja, katerega tehnolo∏ke in tipolo∏ke zna≠ilnosti ga postavljajo v ok- vir ‘zgodnje holocenske egejske tradicije kamnitih artefaktov’ (9000–7000 cal BC). Ta material pred- stavlja predhodnike tipov kamnitih orodij, ki so zna≠ilna za kretsko ustanoviteljsko neolitsko popu- lacijo v Knossosu (ok. 7000–6500/6400 cal BC). Zna≠ilnosti materiala iz Knossosa ka∫ejo na hibrid- ne tradicije v tipologiji kamnitih artefaktov, ki so se pojavile kot posledica interakcije med migran- ti – poljedelci iz Anatolije in domorodnimi lovci in nabiralci. Manj∏a koli≠ina obsidiana iz Melosa na najdi∏≠u Livari pa ka∫e na zgodnje holocensko morsko omre∫je, ki je potekalo med otoki, kar je omogo≠alo prvim poljedelcem, da so pridobili znanje za uspe∏no plovbo do Krete. KEY WORDS – Greece; Aegean Islands; Mesolithic; Neolithisation; lithics; Knossos Introduction Until recently, the received wisdom was that Crete ago (Broodbank 2006.205–209; dates hereafter gi- (Fig. 1), the fifth largest island in the Mediterranean, ven as cal BC). Even in the Aegean, pre-Neolithic sites had remained unoccupied until the founding of the had begun to be reported by this time on a number Initial Neolithic (hereafter ‘IN’) village at Knossos by of significantly smaller islands, including Ikaria, Kyth- migrant Anatolian farmers around 7000 BC (Brood- nos, Melos, and Youra (Broodbank 2006.204–205, bank, Strasser 1991; Evans 1994; King et al. 2008). 211; Sampson et al. 2010; 2012), while a DNA study Crete was thus something of an anomaly, given that argued that some modern Cretans embodied traces later Palaeolithic populations were known on the of a local Middle Pleistocene population (Martinez islands of Corsardinia (22 000–18 500 BC), Cyprus et al. 2007). In 2008, evidence for Palaeolithic and (11th millennium BC), and Sicily some 30 000 years Mesolithic activity was finally discovered on Crete’s 1 Correspondence to: Tristan Carter, Department of Anthropology, CNH 524, McMaster University, 1280 Main Street West, Hamilton, Ontario, L8S 4L9. Email: [email protected] DOI> 10.4312\dp.43.3 87 Tristan Carter, Danica D. Mihailovic´ , Yiannis Papadatos and Chrysa Sofianou south-west coast, the former da- ted to at least 110 000–130 000 BP (Strasser et al. 2010; 2011). These new data force us to con- front a number of questions. First- ly, do they imply Crete’s contin- ual occupation from the Middle Pleistocene, or are these traces of intermittent visitations and/ or failed colonisations? Secondly, when the migrant farmers arrived at Knossos, did they land on un- occupied territory, or were they confronted by an indigenous hun- ter-gatherer population? If the lat- ter, what implications does this have for our understanding of early Knossos and the origins and nature of Neolithic society Fig. 1. Location of main sites and regions mentioned in the text. on Crete? Does the material cul- ture of IN Knossos attest to interactions between the Situated on Crete’s south-eastern littoral, Livari today two groups? Thirdly, in coming to Crete, did these comprises a small and relatively flat coastal plain Anatolian seafaring farmer-herders paddle into un- enclosed by low steep hills to the north. It is cut by known waters, or were they drawing on ancient ma- several streams, two in quite deep gorges, while a ritime knowledge, taking advantage of pre-existing spring provides a year-round water supply (Fig. 2). routes and interaction networks, with the settlement Aeolian and sea erosion has largely removed the Ho- at Knossos being the result of prior reconnaissance locene soil deposits, exposing large areas of the na- trips (cf. Strasser 1996.327–328)? tural bedrock, a Miocene conglomerate consisting of limestone, dolomite and chert (Brandl 2010). In This paper attempts to answer these questions 2008–2010 the Greek Archaeological Service exca- through detailing a new Mesolithic chipped stone vated a small Early-Late Bronze Age cemetery at Skia- assemblage from Livari Skiadi in south-eastern Crete. di, 50m from the modern seashore (Papadatos, So- The assemblage is contrasted with broadly contem- fianou 2015). While the work’s focus was the 3rd– porary data sets from the Eastern Mediterranean to 2nd millennia BC burials in a tholos, rock shelter and see if Cretan practices can be located within larger ‘house tomb’ (Fig. 3), it subsequently became ap- regional traditions. We then reflect on the materi- parent that there were also traces of Mesolithic oc- al’s relationship to that from IN Knossos. Our work cupation. This is not an insignificant claim, and as thus employs lithic technology as a means of con- such requires substantiating. tributing to debates on the nature and dynamics in- volved in the Neolithisation of the larger region (cf. The evidence for Mesolithic Livari Cauvin 2000; Kotsakis 2003; Perlès 2001; Pinhasi et al. 2005). The evidential basis of Livari’s Mesolithic occupation is comprised primarily of chipped stone artefacts Livari Skiadi (Figs. 4–6). The material derives from thin remnant Holocene soil deposits that were protected from ero- The existence of a Cretan Mesolithic was first claimed sion by the rock shelter and Bronze Age tombs (sug- via the publication of stone tools from surveys at gesting that most soil was lost after the 2nd millen- Moni Kapsa and Plakias on the southern coast (Gala- nium BC). It is important to state that the excava- nidou 2011.224; Strasser et al. 2010); one of the tors found no stratigraphic distinctions between latter sites, Damnoni, having since been excavated the material we claim to be Mesolithic and Bronze (Strasser et al. 2015). Here we report on Livari Skia- Age, i.e. these were mixed deposits due to the later di (hereafter Livari), the second Mesolithic site to activities associated with the 3rd– 2nd millennium ce- have been dug on Crete (Fig. 1). metery. The total lithic assemblage from the excava- 88 The Cretan Mesolithic in context> new data from Livari Skiadi (SE Crete) sent bulbar scars (Carter 2015.114–115; see also Tixier 1984). Their dihedral plat- forms, lack of lip and over- hang attests to the use of a copper-tipped flaking tool (Pe- legrin 2012.485–490). The other three pressure blades were made of chert, one grey- ish- blue, the other two tan in colour (Fig. 4.CS10–C11). The recovery of obsidian pres- sure blades from a Cretan Bronze Age cemetery is enti- rely in keeping with the islan- ders’ funerary traditions of Fig. 2. Map of the Livari plain and early Holocene coastline reconstruction. the 3rd–2nd millennia cal BC. tion comprised 469 pieces, of which 251 are identi- This Livari material has numerous parallels. Indeed, fied as Mesolithic. This claim is based on an integrat- Carter (1998;1999; 2010) has studied/published ed four-fold approach that considers the role and over 20 of these data sets from eastern and central nature of chipped stone tools in Cretan Bronze Age Crete. These obsidian-dominated assemblages are burial practices/contexts, the techno-typological cha- highly structured, lacking cores or production debris. racteristics of the assemblage, raw material selec- The fragmentary state of these implements is like- tion, and intra-site artefact distribution. A further po- ly due to their post-depositional breakage, rather sited 20 Mesolithic artefacts were subsequently col- than deliberate fragmentation. The blades’ freshness lected from the site’s surface within a 20m radius of and almost complete lack of use-wear further sug- the excavation area in 2014. gest that they were produced specifically for funer- ary consumption. The Livari chipped stone assemblage is made up of two techno-typologically distinct components. The While Crete is known to have small deposits of knap- first comprises obsidian pressure-flaked blades of pable cherts and other siliceous materials through- Bronze Age date (Fig. 4). The second component of out the island (cf. Blitzer 2004.511; Brandl 2010; the Livari material consists of a microlithic flake- Carter 2007.685–688; inter alia), tools of these raw based tradition, most of which were produced using materials are rarely documented in these burial as- local raw materials (Figs. 5–6); this material, we semblages. Moreover, such local resources tended to argue, is Mesolithic. have been ignored by Cretan/southern Aegean Bronze Age populations, with obsidian being the The Bronze Age assemblage Just under half of the Livari chip- ped stone (n = 218/469) com- prises fragmentary prismatic blades (Fig. 4). Most of these im- plements are made of obsidian (n = 215/218, 99%), a raw ma- terial that is exotic to Crete, the closest sources being located in the Aegean islands (Carter 2009).