Introduction

The cultivation and use of Linum usitatissimum as a crop in prehistoric Greece, unlike cereals, pulses and certain fruit species (the grape-vine and olive), has received relatively little discussion in the context of prehistoric socioeconomic organisation in the area. Yet, archaeobotanical evidence clearly testifies its presence among archaeological deposits of the region throughout the Neolithic and the Bronze Age. Based on archaeobotanical remains preserved by charring from more than 80 prehistoric sites from Greece, this paper provides an overview of the available archaeobotanical evidence for flax (L. ussitatissimum) retrieved from Neolithic and Bronze Age sites in Greece and investigates the potential uses of this ‘extremely useful plant’ as its Latin name implies by the prehistoric inhabitants of the region. To this end the specific archaeological and archaeobotanical contexts of certain rich flax finds, both published and unpublished, from five sites in northern Greece inhabited during the Neolithic and Bronze Age are examined in detail. The results of the archaeobotanical analysis of the charred linseed remains presented in this paper form the basis for approaching aspects of flax cultivation and uses in prehistoric Greece.

Materials and methods

More than 80 sites in Greece have yielded charred plant remains published either in detail or a summarized form, yet few are represented by a considerable number of samples systematically processed by flotation and fully studied. Those sites with archaeobotanical data that could be used as a basis for discussing the crops used by their prehistoric inhabitants, are located mainly on the eastern part of mainland Greece and in their majority belong to the Late Neolithic and certain periods of the Bronze Age, mainly the late Bronze Age; the most abundant evidence both in terms of number of samples and quantity of plant remains come from the regions of Macedonia and Thrace in northern Greece (Kroll 1991; Hansen 2000; Megaloudi 2006; Valamoti 2009). The remains of flax, retrieved from prehistoric sites of Greece have been preserved by charring and in most cases retrieved by flotation, as is the case with the vast majority of archaeobotanical remains in the region. L. usitatissimum is reported from Early Neolithic levels continuing through to the end of the Bronze Age (7th millennium-2nd millennium b.c.). Published remains of flax have been tabulated (Table 1) indicating at the same time its presence in concentrations of more than 30 flax seeds in a single sample, where this information is known. The geographical distribution of the sites where flax has been identified in archaeobotanical remains is provided in Fig. 1.

Table 1 Greek prehistoric sites where Linum usitatissimum has been identified
Full size table
Fig. 1
figure 1

Neolithic and Bronze Age sites with flax remains from Prehistoric Greece. White arrow indicates Neolithic sites, black arrow indicates Bronze Age sites (based on Valamoti 2009)

Full size image

Results

Overview of evidence for flax in Neolithic and Bronze Age Greece

Compared to the remains of cereals and pulses which occur at nearly every site sampled for plant remains in Greece, those of flax are not as common and certainly not as numerous, perhaps partly due to their oil-rich seeds which might burn up more easily than other seeds (e.g., Wilson 1984). This vulnerability of flax seeds to fire has led to the suggestion that flax cultivation can be inferred for a region even when a few flax seeds are found present in an archaeobotanical assemblage (Kroll 1993). All flax finds from Greece consist of charred seeds or conglomerations of charred seeds fused together, while no stems or capsules have been reported, unlike at other sites in south-western Germany for example where they are present in abundance (e.g., Maier 2001, 2004; Herbig 2009). This lack of flax finds other than seeds from Greek sites is probably due to preservation by charring as the archaeobotanical finds from the site of Hornstaad Hörnle IA in the vicinity of Lake Constance (Bodensee) in southern Germany indicate. Only those plant remains preserved by waterlogging appear to preserve remains of flax stem and capsules while they are absent from the charred plant remains from the same site (Maier 2001, 2004). Other factors might also have contributed to the absence of flax remains other than seed at the prehistoric sites of Greece. As Kroll (1993) points out, stems are highly flammable and may have been kept away from the settlements. Moreover, processing of the plant might have taken place outside the settlement, especially if stems were processed for fibre extraction, therefore only processed fibre and/or processed clean seed might have been brought to the site, stored and/or further processed.

Flax as both an oil and fibre crop has been reported for Greek sites in individual site reports and in comprehensive works by Hansen (1988, 2000), Kroll (1991), Megaloudi (2006) and Valamoti (2009). Reference to its use for its medicinal properties is much rarer (e.g., Valamoti 2004, 2009). Early remains of flax from Neolithic Greece are rather scarce for the Early and Middle Neolithic (approximately 6600–5500 b.c.). They have been reported from the Thessalian sites Sesklo and Otzaki Magoula, where it is present in more than half of the samples, i.e., in three samples (Kroll 1981). Besides a few Middle Neolithic (5800–5500 b.c.) records from Thessaly (no quantitative data is provided; Kroll 1982), flax is encountered in a few samples from Middle Neolithic Apsalos in very small numbers (only four seeds from four samples out of a total of 101 samples sorted for plant remains; Valamoti 2006). Although no exact figures are provided for most of the early Thessalian and northern Greek finds, given the limited chances of preservation of this plant, these finds of flax may be indicating that this crop was already cultivated from the 7th millennium b.c. in Greece. This is also the case with Neolithic communities of the Near East, where flax corresponds to one of the crops brought to cultivation by the first farmers (e.g., Zohary and Hopf 2000). Finds of flax are more common and rich among Late Neolithic (5300–3500 b.c.) archaeobotanical assemblages, mainly from the region of Macedonia in northern Greece, perhaps due to a more systematic recovery of plant remains from archaeological sites in relation to other parts of Greece. Flax seeds have been identified at Late Neolithic Kremasti Kozanis, Makri, and Pefkakia (Kroll 1981; Karathanou and Valamoti in press; Valamoti 2004), where flax is present but only in very small numbers. Besides these sporadic finds at a number of sites, flax seeds have also been identified as pure, relatively rich concentrations at Makriyalos, Arkadikos, Mandalo (Valamoti 2004; Valamoti and Jones 2003). A few seeds (1–2) had been reported from Dikili Tash (Valamoti 2004) recently (July 2010), however, samples ‘rich’ in flax have been identified and will be presented in detail below. These finds clearly imply that during the Neolithic flax was a cultivated crop at least at these sites and possibly at the other Neolithic sites where only a few seeds have been identified.

Flax is also encountered in archaeobotanical assemblages from Bronze Age sites of Greece. An impressive find, among the earliest to be reported from Greece, comes from the site of Lerna in the Argolid and has been dated to the Middle Bronze Age (Hopf 1961). It consists of a conglomeration of flax seeds fused together in a lump, clearly illustrated in a figure in the original publication. Another conglomeration of fused flax seeds due to high temperatures is reported from Late Neolithic layers at Kastanas (Kroll 1983). Bronze Age flax remains are reported by Kroll (1981) from the sites of Pefkakia and Argissa. Overall, flax seeds are frequently encountered during the Bronze Age in northern Greece as they are found at the majority of sites that have been systematically sampled, for example Archondiko (Valamoti et al. 2008), Kastanas (Kroll 1983), Agios Mamas (Becker and Kroll 2008) and Assiros (Jones and Valamoti 2005). At Middle Bronze Age Agios Athanasios only two seeds were identified in the samples (Moniaki and Valamoti unpublished) while flax seeds have not been found in samples from Middle Bronze Age Toumba Thessalonikis (Kotsachristou 2008). In the south, flax seeds are reported from Late Bronze Age Tiryns where very few seeds have been identified (Kroll 1982).

Among the sites presented above in relation to flax, certain of them have yielded rich concentrations of flax seeds. The remains of linseed from among some of these sites are discussed in the following section using contextual information in combination with archaeobotanical composition, in order to explore the status and uses of flax among Neolithic and Bronze Age economies of the area. The sites are Makriyalos, Mandalo, Archondiko, Arkadikos and Dikili Tash, located in the region of Macedonia in northern Greece (Fig. 2) and all five of them have been systematically sampled for plant remains, with soil samples being processed by flotation using a 300 μ mesh as the smallest mesh size. Most of these flax rich deposits belong to the Neolithic with those from Makriyalos being the earliest, dated to the second half of the 5th millennium b.c. (Pappa 2008), while those from Mandalo, Arkadikos and Dikili Tash belong to the Final Neolithic, i.e. between approximately 4500 and 4000 b.c. (Kotsakis et al. 1989; Andreou et al. 1996). The flax remains from Archondiko belong to the end of the 3rd millennium b.c. and have been dated to 2100–1900 b.c. (Papaefthymiou-Papanthimou et al. 2002). The finds from Makriyalos, Mandalo and Arkadikos have been fully published (Valamoti and Jones 2003; Valamoti 2004), those from Archondiko are published in preliminary reports (Valamoti et al. 2008; Papanthimou et al. in press) while the remains from Dikili Tash are reported here for the first time.

Fig. 2
figure 2

Location of Neolithic and Bronze Age sites discussed in detail in the paper: Makriyalos, Mandalo, Archondiko, Arkadikos, Dikili Tash

Full size image

Some neolithic concentrations of flax from Greece

Makriyalos

The richest Neolithic assemblage of flax seeds that has been encountered so far is at the site of Makriyalos. Makriyalos is a flat extended site southwest of the modern city of Thessaloniki in northern Greece, inhabited between the middle of the 6th millennium b.c. and the middle of the 5th millennium b.c., i.e., the period referred to as the Late Neolithic (Pappa and Bessios 1999; Pappa 2008). It is characterized by pits and ditches, two occupation phases during the Late Neolithic, and a wealth of finds including numerous figurines, Spondylus shell, decorated pottery and the remains of what could be described as Neolithic feasting, consisting of huge quantities of animal bone and broken pottery, deposited in perhaps a few depositional episodes (Pappa et al. 2004; Halstead 2007; Urem-Kotsou and Kotsakis 2007). Plant remains from the site are dominated by cereals and in particular glume bases of einkorn, new glume wheat type and emmer, with the first two species being the dominant ones (Valamoti 2004, 2005). The rich flax remains from Makriyalos originate from a pit situated at the southern end of a rectangular, bipartite structure defined by the negative features of its foundations (Fig. 3). The temporal relationship between the various negative features of the site, the pits in particular, has been difficult to establish and according to the excavator the pit and the rectangular structure may have been contextually and temporally unrelated (Pappa 2008).

Fig. 3
figure 3

Makriyalos, the area of the megaroid building and associated pit where flax was found (photograph courtesy of M. Pappa)

Full size image

This particular pit from Makriyalos, unlike many others at the site, stands out due to the large amount of seed and grain found in it. Although flax seeds in the remaining samples from the site are represented by a single or a few seeds in 11 samples out of a total of 53 samples, in this sample approximately 2,100 flax seeds have been retrieved (Fig. 4a). These 53 samples have been amalgamated from a total of 114 individual samples on the basis of similarities in composition while the number of flax seeds is based on absolute counts of floated material and fractions sorted from the heavy residue, subsequently multiplied thus providing an approximation of the total number of seeds in the heavy residue (Valamoti 2004).

Fig. 4
figure 4

Charred flax seeds from: a the Makriyalos pit 554 located to the south end of the megaroid building; b Dikili Tash, excavated in July 2010; c Archondiko, end of 3rd millennium B.C.; d Late Neolithic Makriyalos (top row) and from Early Bronze Age Archondiko (bottom row) indicating differences in seed size

Full size image

The pit in question, besides samples rich in glume wheat chaff (e.g., samples 962, 983), contained concentrations of lentil seeds (e.g., sample 983), nuts of Pistacia terebinthus and fig (Ficus carica) seeds (e.g., sample 962). One of the samples (965) was dominated by flax which, unlike the other samples from the pit, was also rich in seeds of wild plants, identified mainly to family or genus level: Brassica/Sinapis, Camelina sp., Urtica sp., Anagallis cf. arvensis, Hypericum sp., Trigonella sp., Medicago sp., Solanaceae, Asteraceae, Fabaceae and Poaceae. The occurrence of these wild/weed species in this sample is impressive and exceptional for Makriyalos due to a general lack of wild/weed seeds in the samples from the site. It seems very likely that these wild/weed species are associated with the flax sample as they occur only or mainly in this sample. Perhaps the most straightforward interpretation of these wild/weed species would be that they correspond to weeds of flax due to their contextual co-occurrence. Certain species of the genera present in the sample, such as Camelina, Brassica and Urtica, have been found with flax in other archaeological deposits (Pals and van Dierendonck 1988, p. 245). There are various reasons, however, that cast doubts on a ‘weed’ interpretation of the seeds of wild species found in the pit together with flax. The overall limited presence of crop weeds in general at the site, among contexts containing spent fuel rich in charred glume bases, i.e., contexts likely to preserve weed seeds had they been present, makes this flax sample stand out for its wealth of wild/weed seed. Although the possibility that the wild/weed species closely associated with flax represent weeds of flax fields cannot be ruled out entirely, an alternative interpretation emerges in light of the medicinal properties shared by some of the species or genera represented in the pit and in the flax sample in particular. Besides figs and Anagallis cf. arvensis, identified to species level, different species of Sinapis/Brassica, Urtica and Solanaceae as well as Hypericum perforatum are well known in traditional folk remedies for their medicinal properties (Bown 1995; Brockhampton 1996). For example Anagallis arvensis is used as a diuretic, antipyretic and for treating depression. Brassica and linseed, present together in the Makriyalos pit, are mentioned as used in combination to treat cough in recent times (Brockhampton 1996; Bown 1995). Hypericum perforatum is a widely known medicinal plant, used externally on wounds and as an infusion against psychological problems (Gennadios 1914; Bown 1995); it is, however, the flowers, stems and leaves that are used medicinally, not the seeds of this plant. The presence of a rich concentration of Pistacia terebinthus nuts next to the flax sample further underlines the exceptional character of the archaeobotanical composition of this pit. Terebinth is yet another plant well known for its aromatic and medicinal properties in antiquity, of which best known were its resin and the oil from its nuts. Dioskourides informs us that terebinth resin (occasionally oil) was applied externally, together with other substances to treat for example scorpion bites, skin ulcers and pain, while terebinth resin could also be burnt as incense with aromatic and therapeutic powers. Besides its aromatic and medicinal properties, the nut oil of this plant is edible (e.g., Gennadios 1914; Özcan 2004; Marcopoulos 1965) and was a precious component of royal meals in ancient Persia (Simpson 2005). Of course the lack of identification to species level of the wild/weed seeds present in the pit limits the certainty with which a medicinal use of these wild plants and flax could be argued for convincingly. Yet, the specific archaeobotanical composition of the Makriyalos flax sample stands out from the rest of the site. Thus, we have a location at Late Neolithic Makriyalos where a unique combination of plant species is encountered, standing out from the remaining samples of the site and sharing as common features their medicinal properties and in part their oil content (flax and terebinth). In the light of the above, it seems plausible that flax at Makriyalos may have been used for its medicinal properties and/or its oil (see discussion below).

Mandalo

To the northwest of Makriyalos, in the foothills of mount Paikon, lies a small tell site called Mandalo (Papaefthymiou-Papanthimou and Pilali-Papasteriou 1997), with excavated deposits dating to the Late and Final Neolithic (4775–4010 b.c.) and the Early Bronze Age (2950–2200 b.c.) (Kotsakis et al. 1989). This small site covering an area of 0.2 ha and reaching a height of 7 m of archaeological deposits, comprised two stone enclosure walls of Final Neolithic date, and yielded evidence of copper smelting. Among the plant remains from the site (Valamoti and Jones 2003), flax seeds are occasionally encountered among otherwise cereal grain or pulse seed dominated samples, originating mainly from burnt destruction layers from probable house interiors. Only in one case are flax seeds found in relatively high numbers; together with a concentration of emmer grain, 76 seeds of flax were identified. All samples from this burnt floor destruction contain a rich wild/weed flora, consisting mainly of the following species: Setaria viridis/verticillata, Eragrostis minor, Polycnemum sp., Chenopodium album and Rumex acetosella while the flax-rich samples (152 and 702) contain a large proportion of seeds of Lolium sp., some of which have been identified as L. remotum. This species has not been found in the other samples from the same house containing seeds of wild/weed species and the same is true of seeds of Silene sp. Unlike the rest of the emmer-rich samples, where the wild/weed species have been interpreted as mainly deriving from animal dung, those in the flax samples, i.e. L. remotum and Silene sp. seeds, were apparently associated with flax and represent in all likelihood flax weeds (Valamoti and Jones 2003).

Arkadikos and Dikili Tash

Moving westwards, to the large Drama plain, at the outskirts of the modern city of Drama and close to a formerly extensive peat bog known as the Tenagi of Philippoi where a deep pollen sequence has been obtained (Wijmstra 1969), lie the sites of Arkadikos and Dikili Tash. Arkadikos is a flat extended site on the outskirts of the modern city of Drama in northern Greece. A very small proportion of the site has been excavated and presented in preliminary reports published in Greek. On the basis of pottery finds the excavated layers were dated to approximately the end of the 5th millennium b.c. (Peristeri and Touloumis 1994; Vargas et al. 1995). The archaeobotanical remains have been fully published (Valamoti 2004) and reveal a wealth of stored crops charred during destruction of the post-framed houses. These crops include an impressive variety of pulse species such as lentils, peas and grass pea. Cereals are also present, especially barley. At Arkadikos, 103 flax seeds were found dominating a sample that also contained glume wheat chaff. Most of the remaining samples from the destruction level in the vicinity of the flax rich sample were rich in pulse seeds. Two Rubiaceae and one Poaceae are the only wild/weed species in the flax sample and it is difficult to associate them with flax rather than the glume wheats in the sample. Although the flax seeds are relatively few in number, they probably represent clean grain stored for use or sowing as has been suggested for the more or less pure pulse seed concentrations found in the same destruction level.

Dikili Tash is a tell site near the ancient city of Philippoi, in the Drama plain in eastern Macedonia in northern Greece, approximately 20 km to the southeast of Arkadikos (Koukouli-Chrysanthaki et al. 1997). The archaeobotanical samples from Dikili Tash originate from four rectangular Late Neolithic houses, parallel to each other, containing ovens, hearths, storage bins, pots and a wide variety of objects related to daily activities. The houses were destroyed by fire in a major conflagration episode dated to the second half of the 5th millennium b.c. The archaeobotanical remains from the site are best known for the remains of grape pressings (Valamoti et al. 2007), yet they also comprised a rich assemblage consisting of cereal grain (einkorn and barley), pulse seed (bitter vetch, lentil, grass pea), fruits and nuts (wild pear, acorns), presumably stored within the houses at the time of the destruction (Valamoti 2004). Only a couple of flax seeds (L. usitatissimum) had been identified in the studied and published samples, yet, as excavation of the destruction debris of House 1 resumed in July 2010, three samples from the same location contained a considerable number of flax seeds (Fig. 4b). The samples originate from among the destruction debris to the northeast of a clay-lined large oven within which a double-headed zoomorphic figurine was unearthed in 2008. A total of 201 flax seeds have been retrieved from the >1 mm mesh of floated material and more seeds are expected to emerge from sorting of the heavy residue in the near future as the material is still under study. Flax is a new addition to the crop repertoire of House 1, and Dikili Tash as a whole.

Bronze Age concentrations of flax: the finds from Archondiko

Besides published information on flax remains from Bronze Age sites in Greece presented earlier, Archondiko has yielded several rich concentrations of the seeds of this plant. Archondiko is a multi-period site near the modern town of Giannitsa west of the river Axios not far from Mandalo. Systematic excavations at the site have brought to light three Early Bronze Age habitation phases, of which the earliest, where flax has been identified, has been dated to 2135–2020 b.c. A major conflagration episode destroyed and preserved post-built structures containing numerous cooking and storage structures as well as a rich selection of pottery (Papaefthymiou-Papanthimou et al. 2002). A wide range of crops has been identified from this early phase dated to the end of the 3rd millennium b.c., including einkorn and emmer wheat, possibly spelt and free-threshing wheat, barley, ground barley, acorns, Lallemantia sp., Linum (Valamoti et al. 2008) and Papaver somniferum (Valamoti unpublished). The flax finds from Archondiko consist of rich, pure caches of flax seeds (Fig. 4c) and have been encountered at various locations in the excavated area, corresponding to different post-framed habitation units. Some of the concentrations contain seeds of wild plants that are currently under study. Although no measurements for individual seeds are as yet available, it is interesting to note that the flax seeds from Archondiko appear overall to be of a smaller size than those encountered at the Late Neolithic sites of Makriyalos, Arkadikos and Dikili Tash, as can be inferred from the images of flax seeds presented in this paper (Fig. 4d).

Discussion

The body of archaeobotanical data from Greece is unevenly distributed both spatially and temporally. Thus, generalizations regarding the importance of flax, or any other crop for that matter, for prehistoric communities of Greece, inhabiting different regions (e.g., the north and the south) and living in different time periods should be carried out with caution, at least for the time being. Although flax has been reported as a crop from Neolithic and Bronze Age sites in Greece, this species has usually been left out of discussions of economic organisation and daily activities. Yet, the preceding overview of archaeobotanical remains of flax from Greece has demonstrated the regular occurrence of seeds of this species at several prehistoric sites from the region, despite the admittedly limited chances of its preservation for reasons discussed earlier on. Moreover, it is clear that flax was cultivated and/or used as a crop, at least at those sites where a significant amount of seed has been identified in individual concentrations and possibly also at those where only a couple of seeds have been retrieved.

The cultivation of flax is highly demanding in terms of soil and moisture requirements as well as weed elimination (e.g., Dewilde 1999), something already well known from classical and Roman Antiquity (e.g., Kroll 1993; Tzahili 1997 referring to ancient writers such as Pliny and Columella, as opposed to Megaloudi 2006). Thus the Neolithic and Bronze Age flax fields that can be inferred from the presence of their seeds in archaeobotanical assemblages from Greece, would have required skilled agricultural knowledge, necessary for maintaining high nutrient levels in the flax fields through fallowing, manuring and/or crop rotation with pulses. The location of some at least of the prehistoric sites discussed in detail in this paper would be ideal for providing high moisture levels in flax fields. Arkadikos, Dikili Tash and Archondiko were situated close to marshland and to rich, permanent water sources, though well drained patches of land would have been selected for the cultivation of flax. Mandalo and Makriyalos are also both in close proximity to streams. By the end of the 5th millennium b.c. a specialized flax weed flora, similar to flax weeds known from recent times, had already probably developed as Lolium remotum and perhaps Silene sp. accompanying flax at Mandalo suggest. L. remotum used to be a very common weed of flax in recent times (e.g., Behrendt and Hanf 1979) and L. remotum and S. linicola are typical weeds of flax cultivation (Tutin et al. 1964–1980). Unfortunately in the other sites from the Neolithic where flax has been found in some quantity in Greece, flax seeds appear to have been cleaned of their weed seeds, had there been any, or, as is the case with the Makriyalos concentration, it is questionable whether all or most of the wild/weed species accompanying flax are field weeds of this crop. L. remotum continues to be associated with flax and therefore a likely flax weed at Early Bronze Age Kastanas (Kroll 1983). It is possible that flax was sown in winter both at Mandalo in the Neolithic and at Kastanas in the Bronze Age as L. remotum rarely occurred in spring sown crops (Behrendt and Hanf 1979).

Finds of flax seed concentrations do not occur as regularly as those of cereals and pulses at the vast majority of the sites examined in this paper. This, however, may not be an indication that it constituted a rarer crop, but rather a crop less likely to survive in the archaeobotanical record. The evidence from Archondiko stands out in this respect due to the identification of several caches of charred seeds in different locations, suggesting that at least at that site, flax was a common crop, grown perhaps by different households as it was found in separate concentrations, in different locations within the post-framed constructions. Inferences for the uses of flax are rather difficult to make as the flax seed that survived within the archaeological deposits could represent seed kept for sowing the next crop intended for oil or fibre or both, seed kept for extracting oil or seed kept for medicinal uses (see for example Chauhan et al. 2009 for a recent review of the uses and properties of L. usitatissimum products, seeds in particular).

The examination of the context of the flax finds at Makriyalos has led to the suggestion that flax at that site might have been used for medicinal purposes. Flax seeds are mentioned in various medicinal recipes in the work of Dioscourides as well as in recent publications of traditional plant based medicinal remedies (Pyrros 1838; Gennadios 1914; Bown 1995, see also Kislev et al. this volume, for medicinal uses of flax oil). Many treatments in the work of Dioskourides involve the use of flax seed, occasionally roasted and mixed with honey, vinegar, roasted figs and occasionally several other ingredients, to treat various problems, externally as cataplasms against pain, scorpion bites and vaginal infections or internally to treat stomach and bowel problems. As pointed out already, at Makriyalos many wild/weed seeds of a wide range of species and genera, many of which have medicinal properties, have been found together with flax seed, an exceptional find at that site. A possible use of flax for medicinal purposes at Makriyalos does not of course preclude the other two uses.

The use of flax for oil has been suggested for the Neolithic of northern Greece (Valamoti 2003, 2004; Jones and Valamoti 2005), especially in light of the lack of other oil bearing plants, except perhaps Terebinth. It is during the Bronze Age, however, that several crops yielding oil-rich seeds make their appearance: Lallemantia sp., Papaver somniferum, Brassica sp., Camelina sativa (Jones and Valamoti 2005; Kroll 1983; Valamoti and Jones 2010). These crops have been associated with a possible local oil/ointment production, making use of special small sized vessels intended as oil/fat containers (Fig. 5; Andreou et al. in press). In an area where the olive is at the limit of its cultivation, alternative oil plants could have been used for the preparation of luxury substances in the north of Greece as is known to be the case with olive in the Late Bronze Age palaces of southern Greece. At present there is no way of knowing whether linseed, found in prehistoric Greece, northern Greece in particular, was indeed intended for oil production. At Archondiko, pottery vessels with a spout, similar to vessels used for Argan oil extraction in Morocco (Fig. 6) have been found, yet this obvious similarity could be misleading unless substantiated in the future by other analyses, e.g., chemical analyses of the pottery residues themselves. Linseed oil is rich in omega three fatty acids and it is edible when extracted using a cold water method (Serpico and White 2000; Breslow 2006; Hu et al. 1999; Kapoor and Huang 2006). Had it been intended for human consumption, it would have probably been consumed within a limited time period as it quickly turns rancid. Linseed oil was produced in parts of Anatolia (Ertuğ 2000) and in Kappadokia in particular as olives do not grow in the region (Bozi 1997; Ertuğ 2010). There is ethnographic evidence from Anatolia to suggest that the ideal timing for extracting linseed oil was in winter, when animals needed high quality fodder, thus linseed would be stored until winter when it would be processed for oil extraction (Ertuğ 2010).

Fig. 5
figure 5

Vessels with encrusted and incised decoration, probably used as oil/ointment containers from Toumba. Thessalonikis (photograph courtesy of S. Andreou)

Full size image
Fig. 6
figure 6

a Open bowl with spout from House D at Archondiko (photograph courtesy of A. Papanthimou and E. Voulgari); b Argan oil extraction in Morocco (photograph courtesy of F. Ertuğ) (http://archaeobotany.dept.shef.ac.uk/wiki/index.php/)

Full size image

Weaving was another major activity at Archondiko as the numerous spindle whorls suggest and different types of threads and woven fabric were being produced at the site (Andersson 2008; Papadopoulou 2010). Therefore flax at Archondiko might have been used for both purposes. Size of flax seeds has been associated with the different uses of the plant, big seeded varieties used for oil while small seeded varieties were used for fibre, while varieties intended for both uses are also reported (Zohary and Hopf 2000). Whether the small size of flax seeds at Archondiko is related to specific uses of the plant, in this case for fibre, or to harvesting of plants not fully ripe, a practice adopted when good fibre quality is desired (e.g. Gennadios 1914; Willson 1979), is difficult to tell. No systematic measurement has been carried out on the flax seeds from Greece though it may well be worth undertaking as size changes may be related to the development of small seeded varieties of flax with specific, sought after properties. The development of different flax varieties in southwest Germany during the 4th and 3rd millennium b.c. has recently been inferred on the basis of measurements taken on waterlogged flax seeds (Herbig and Maier 2011). Flax varieties could have developed in northern Greece locally or been introduced from elsewhere. There is ample archaeobotanical evidence for crop introductions during the Bronze Age in northern Greece (Valamoti 2007), likewise varieties of crops with desired properties may well also have circulated in prehistoric times with the movement of people and regional contacts. The small-sized flax seeds found at Archondiko could be related to a variety intended for fibre extraction, as is the case in modern times; however the seeds, once ripened in the capsules, could have been used for the extraction of their oil.

Flax has been closely linked to prehistoric weaving in Greece as one of the main fibre sources, the other being wool. Wool has been associated mainly with Bronze Age weaving, thus flax has been considered as a main fibre crop for the Neolithic (Tzahili 1997, p. 38, referring also to Halstead 1981). Spindle whorls are eloquent indicators of weaving and constitute a common find not only at Archondiko as mentioned above but also in the vast majority of prehistoric sites excavated in the region. Weaving is evidenced through a range of artefacts related to this activity such as loom weights and spindle whorls. Although wool and flax provide different kinds of threads, experiments have shown that no association between spindle whorls of specific characteristics (e.g., shape, weight) and fibre source can be established (Mårtensson et al. 2006). The differences in shapes and weight of spindle whorls are related to different types of threads that could be produced in terms of thickness/fineness and quality (e.g., Barber 1991; Andersson 2003). It is therefore impossible, at least on present evidence, to relate the archaeobotanical finds of flax from Greece with finds of spindle whorls. At Kastanas, Kroll (1983) observed that there is a correlation between the presence of linseed finds and loom weights in different archaeological layers. Such correlations, however, are largely dependent on preservation conditions and are likely to vary both within and among sites, and therefore should be treated with caution.

Spindle whorls and loom weights suggest weaving and flax seeds indicate a potential fibre source from Neolithic and Bronze Age Greece. Moreover, the location close to water sources of most sites examined in detail here, would have been favourable for retting of harvested flax and subsequent preparation of the flax fibre. Nevertheless, the archaeobotanical remains leave little scope for further exploring the specific uses this cultivation was intended for, except for those cases where textual evidence contributes relevant information. Compared to the rich archaeobotanical evidence for Neolithic and Bronze Age cultivation of flax in northern Greece, a paucity of finds from the south can be observed. Yet, for the south, other lines of evidence provide valuable insights into the cultivation and uses of flax. During the Late Bronze Age, textual evidence from the Mycenean palaces in the Peloponnese and Crete provides clear indications of flax cultivation and its use in weaving and probably in other activities. The archives of Pylos and Knossos mention flax under the name RI NO (LI NO) or SA, focusing on cultivation, production and taxation of the product (e.g., Killen 1984; Robkin 1979; Foster 1981; Tzahili 1997). Flax production appears to have been centrally controlled by the Mycenean palaces and at the palace of Pylos a special group of women workers called rineja (flax ladies) were involved in the processing of flax which was located usually in the countryside. Although it is not clear what their precise tasks were, they were probably involved in preparing the fibre starting from the early stages of retting and beating the flax stems (Killen 1984). Moreover, flax fibre production was probably more intensive at Pylos in the Peloponnese rather than at Knossos on the island of Crete (Killen 1984). Flax in the Linear B texts seems to be associated with a wide range of professions such as hunters and ship builders, perhaps because they might have used different types of products from flax fibre (Tzahili 1997). Based on the Linear B tables, it appears that flax was not distributed in relation to ritual offerings, unlike other plant based commodities such as cereal grain, figs, olive oil etc. (e.g., Bendall 2007).

Conclusions: flax in prehistoric Greece

The preceding discussion of flax finds from prehistoric Greece has attempted an overview of the evidence for the cultivation and possible uses of flax in Neolithic and Bronze Age Greece. Flax has been identified at many sites from Greece; it seems, however, that pure, dense concentrations can only be retrieved from sites extensively sampled and processed by flotation. The available archaeobotanical evidence from prehistoric Greece as regards flax suggests that flax was cultivated both in the Neolithic and the Bronze Age and that by the 5th millennium b.c. a specialized weed flora had probably developed, at least in the north. The clean seed finds from Greece indicate a series of stages proceeding from field preparation, to harvest and processing. L. usitatissimum was cultivated and processed in prehistoric Greece for extraction of the seeds, and the plant was used either for oil or fibre or both and probably also for its medicinal properties.