Defining the problem: the Later Neolithic Stages in Southern Greece

1This brief overview of the latest work on the Aegean Later Neolithic Stages is representative of a growing awareness of the shortfall of existing views on Later Neolithic cultural phases and chronological associations connected to them4. It is no coincidence that most of this work is related to regions such as Crete and the Cyclades, for which available evidence has for decades been coming only from a few reference sites, such as Knossos, Saliagos, and Kephala. In recent years, more sites have been excavated, while existing archaeological remains have been reconsidered within a broader perspective in terms of chronology, cultural affinities, socio-economic characteristics, etc.5.

2Regardless of the preferred terminology one adopts (i.e. Late Neolithic I, II, Early/Middle/Late/Final Chalcolithic, Final Neolithic I, II, etc.) [table 1], which is a matter that deserves an independent discussion beyond the scope of this paper, it is becoming gradually accepted that the term “Final Neolithic”6 cannot be productively used without further refinement. The Final Neolithic was introduced in Aegean archaeology approximately 30 years ago based on the available information at that time and it has been used as a catchall determinative for a period that extends approximately between 4500‑3200 BC, i.e. more than a millennium7. New excavations in the Aegean islands8, new data from Knossos and Crete in general9, excavations or re-examination of sites such as Tsepi in Attica10 or Mikrothives in Thessaly11 (just to mention some of the currently available data), make apparent the need for a fresh look at both the old and the new evidence.

3The present volume is a good starting point for incorporating the Aegean Later Neolithic Stages into a wider perspective and discussion. Moreover, a closer comparison of the existing evidence with the developments in other areas, such as the Balkans or Anatolia, in which field research and theoretical/interpretative discussion of this time-frame is more advanced, is necessary.

4Terminology is always a subject of debate, while the preferred methods and approaches one uses to address it are constantly reconsidered and re-examined. Considering the title of the project behind this volume, it is a matter of debate whether chronological limits represent one phenomenon with common characteristics and can, therefore, be studied isolated, or whether this section of prehistory needs to be connected with previous and/or later developments. In this respect, the question remains open on whether only the study of ceramics, together with their context and stratigraphy, and the exploration of similarities and differences in the style of material culture12, are the most appropriate approaches to the investigation of cultural associations, social transformations, and changes in the use and meaning of material culture. It seems more and more necessary to introduce into our research a wider range of methods. For example, in addition to stylistic considerations, to identify discontinuities and gaps in the archaeological record we should examine depositional practices and uses of space to understand why some phases are better attested in the landscape than others. A characteristic example of this are the Neolithic tells; for decades archaeologists have been examining their basal levels for traces of the Mesolithic, to prove cultural continuity and local development13. However, there was never any reliable support for the assumption that the Mesolithic use of space should be identical with that of the Neolithic14.

5Surveys or other synthetic works15 make reference to a phase usually called the “Final Neolithic”. However, which subphase of the “Final Neolithic” is present in each case is usually not clear, often because of poor preservation of the archaeological material, but also because of the general use of the term “Final Neolithic” for everything dated after the mid‑5^th millennium BC. The term “Final Neolithic” includes in one term several sub-phases with different characteristics in relation to the use of space as well as the use and meaning of the material culture. It is a period during which many Aegean islands are colonized, painted pottery looses its importance, and other important changes occur16. It is the main phase during which a large amount of material is deposited in caves17, especially during the second half of the 5^th millennium BC. At the cave sites, traces of earlier as well as later material are often also attested; however they represent a different use of space and material culture and, likely, different meaning assigned to them that produced differences in depositional practices.

6This paper touches only upon some of the questions presented in the preceding discussion, mostly through the prism of the evidence from the excavations at the Agia Triada Cave in Karystos, Southern Euboea. Testing some of the ideas outlined above were some of the principal reasons for starting this project. Primarily, we wanted to explore the site of purportedly earliest occupation in this part of the Aegean, since previous surveys inside the cave produced evidence for the white-on-dark Late Neolithic pottery, which is absent from the rest of the Karystia18. Second, we hoped to address some issues connected to the transition from the Late Neolithic I to the Final Neolithic/Late Neolithic II and from there to the Bronze Age in this geographical context, as well as to explore why a cave with a lack of suitable habitation space was used in prehistory.

Excavation, stratigraphy and finds

7The Agia Triada Cave is located at the foot of Mount Ochi, close to the village of Kalyvia (fig. 1). It was formed at the point where marble and schist masses meet along an underground riverbed and it is the product of chemical and mechanical action of water. The entrance to the cave is about 50 m uphill from the modern church with the same name (Agia Triada = Holy Trinity), and in the immediate vicinity of a perennial spring. The cave entrance is well sheltered by the configuration of the terrain and vegetation and is almost impossible to see from below (fig. 2). At the same time, one can easily observe all the approaches to the valley from the cave entrance as well as see all the way to the sea.

8The excavations inside the Agia Triada Cave took place over a period of four consecutive field seasons between 2007 and 2010 (fig. 3). In the course of the research we opened twelve excavation trenches inside the cave and its immediate surroundings. All trenches except trench 10 were located inside the Agia Triada proper, in two distinct areas: (1) the long and narrow entrance corridor and (2) the double chamber we refer to as the “East Chamber”19. Trench 10 was placed in a small rock shelter, which is located about 30 m south of the cave’s entrance, exactly above the Agia Triada church. Trench 10 did not produce any evidence for the location’s prehistoric use and we omit it from further consideration in this paper. Trenches 1‑3 and 5‑7 were located in the main entrance corridor. We excavated trenches 4, 8, 9, 11, and 12 inside the main section of the East Chamber (fig. 4).

9The entrance corridor is approximately 40 m long from the entrance of the cave to the area where it makes a sharp left (east) turn that takes it deeper inside the cave. The corridor starts steeply at the entrance but gradually levels out to become almost horizontal close to the section where it turns east. From that point, it continues for another 20 m roughly due east before it reaches a ca 6 m deep vertical crevasse. We placed trench 1 about half way between the crevasse and the main section of the entrance corridor. This trench is shallow and it produced only two distinct layers of soil including the surface stratum. We observed no evidence for cultural stratification although the trench did produce a small quantity of Neolithic pottery. Trench 2 was located at the very end of the main section of the entrance corridor, at the spot where the corridor turns left. The stratigraphic sequence as well as the archaeological inventory from this trench is much better preserved and more informative than the one from trench 1. We were able to discern seven distinct stratigraphic layers based on soil color, inclusions, and consistency. The archaeological material found in trench 2 consists chiefly of pottery with some obsidian and disconnected fragments of human and animal bone. Based on its stylistic attributes, the pottery from trench 2 can overwhelmingly be dated to the Late Neolithic. In this trench we also found some ceramics that can be dated to the Early Bronze Age and small quantities of post-prehistoric pottery (the latter found in the topmost mixed strata). The Early Bronze Age material from this trench does not appear to be stratified and is commonly found mixed with more recent deposits.

10The most important element of the trench 2 stratigraphy is feature 1. This feature consists of a roughly circular patch of reddish hardened and burned soil with a substratum of large pottery sherds and flattish slabs of local rock (fig. 5: 1). From this circular burnt area a small platform made of flattish rock slabs extends towards the northeast corner and into the profile of the trench20. Roughly below the middle section of the circular feature we found an anthropomorphic protome handle that belongs to a white-on-dark painted vessel whose other parts we also found in trench 2 (fig. 5: 2; 6). It appears that such a positioning of the prosopomorphic handle was intentional. The handle was found within a leveled layer (layer 4), which contained significant amounts of charcoal and ash. The purpose of this feature is unclear. The most straightforward explanation is that it represents the remains of a fire, which must have burned with some degree of intensity21 to create a patch of burnt soil that is almost 2 cm thick. On the basis of other excavations in caves such as the Drakaina Cave on Kephalonia22 and the Leontari (= Lion’s) Cave in Attica23, it appears that the intentional placement of certain materials below floors was a common practice24. Regardless of its purpose, feature 1 and the clearly layered stratigraphic sequence with much archaeological material in trench 2 confirmed the presence of intact cultural layers in the Agia Triada Cave, which merited the continuation of research.

11Trenches 7 and 3 were opened close to trench 2 in order to examine whether the structure (feature 1) in trench 2 extends over a wider area. This, however, was not the case and the archaeological material found consists chiefly of pottery with some obsidian and disarticulated fragments of human and animal bone. At the same time, in trenches 3 and 5 the existence of possible shallow pits and burned areas (fire places?) enabled the preservation of a greater amount of Late Neolithic I material.

12The area of the cave called the East Chamber produced, so far, the richest prehistoric deposits, which are dated to the Late Neolithic I and II25 and the Early Bronze Age (primarily Early Bronze II26). The East Chamber is approximately 6 x 3 m in extent. The original surface of the chamber is uneven and mostly sandy, with some areas of relatively recent thin layers of stalagmitic crust. The ceiling of the chamber is triangular, resembling somewhat a gabled roof in appearance. We excavated the area of the East Chamber with five separate trenches (4, 8, 9, 11, and 12), leaving stratigraphic balks among them (see fig. 4). During the last excavation season (2010) we removed all the dividers and connected trenches 4, 8, and 9 into one unit. Our aim was to create extended unified profiles and to facilitate better understanding of the connections between various horizontal stratigraphic features observed in each trench separately. Only trench 9 was excavated until the bedrock (fig. 7). We only removed the topmost layers (1‑5) in trenches 11 and 12.

13The Final Neolithic/Late Neolithic II strata in the East Chamber are, for the most part, clearly separated from the Early Bronze ones by a layer of sand of variable thickness (layer 6, in most trenches). The sand in layer 6 is in some areas dark reddish-yellow in color, indicating exposure to fire. This fire is probably connected to Early Bronze layers as copious amounts of burned fruits and grains were found in them. Layer 6 contains considerably fewer finds than either the upper (Early Bronze) or lower (Final Neolithic/Late Neolithic II) layers; also, the finds are very fragmented. From the preliminary sorting of the material it seems that pottery from layer 6 belongs to different phases. Although most of the sherds are undiagnostic, in general appearance they resemble the Final Neolithic/Late Neolithic II material. Few Early Bronze Age sherds are also present as well as a large part of a pot with a “trumpet” lug that has parallels, among other sites, at Emborio on Chios27.

14Layer 7 in the East Chamber consists of two strata: (1) a leveled upper stratum paved with flattish slabs of rock and large pottery fragments (fig. 8), and (2) a lower stratum, which contained much pottery and chipped stone, and lay mixed with or below the paved stratum. We excavated both strata as one stratigraphic layer. In terms of sheer quantities of excavated material, layer 7 is one of the richest Neolithic layers at Agia Triada. The soil in this layer largely consists of loamy sand with some gravel. The paved “floor” seems to have had a substructure consisting of smaller rocks (in comparison to the ones forming the upper part of the paved surface) and very small, fragmented, and worn ceramic sherds. In some excavated sections there was a second layer of larger flat rock slabs, indicating intentional leveling of the entire surface. This mode of floor construction, where a surface of large flat rocks lies on a substructure of smaller stones, gravel, and crushed pottery, is not uncommon in Aegean prehistory, especially during the Neolithic period28. The floor and associated substructure extends inside trenches 4, 8, and 9, i.e. it covers most of the East Chamber.

15It is more difficult to interpret deeper layers that contain Late Neolithic II/Final Neolithic pottery. The underlying layer 8 consisted of sand that is more compact and contained less gravel than in preceding layer 7, although it was similar in color. This layer contained a possible fireplace (unit 5) in the north section of trench 8. It is possible that the fireplace is delimited by a row of rocks arranged in a rough circle. The finds from layer 8 are much fewer in number than in layer 7, and chiefly consist of obsidian fragments with very little pottery.

16Layer 9 contained mottled but mostly yellowish sandy soil. Sections with more gravel were observed and they appeared to contain more material than other areas in this layer. The patches with less gravel contained charcoal mixed with sand. This mixed soil also extended below the areas with more gravel. Associated with the sandy soil with charcoal was a patch of burnt sand mixed with ash – a likely fireplace (unit 6). A large amount of burnt animal bones, which was found in the sandy part of layer 9, is probably linked with the fireplace. The finds from this layer consist of pottery and other terracotta objects (i.e. spindle whorls and perforated clay objects), and obsidian implements, including two possible projectile points. Unit 6 extended across trenches 4 and 8.

17Layer 10 consisted of finer yellowish sand with a much greater concentration of ceramics than the previous layer. Among the finds from this layer we should emphasize a relatively large quern stone. A fragment of a large typical Late Neolithic II/Final Neolithic rope-decorated pithos was also found. This layer, unlike the preceding layer 9, contained almost no bones.

18The following layers (11‑24) are of varying thickness and mostly consisted of sand with different percentages of intermixed gravel. In some cases the gravel dominated. The gravel consisted of small rounded pebbles and coarse sand, and was, generally speaking, mottled in color. These layers resemble water-borne deposits the most, although it is unclear whether this was indeed their origin. All layers contained archaeological material with varying degree of frequency, and many also contained charcoal fragments. The material from these layers consists of pottery fragments, obsidian, and other types of worked stone.

19Interpreting the Late Neolithic II/Final Neolithic stratigraphy at the Agia Triada Cave is a difficult task, particularly so for deposits lying below layer 11. These strata are much more ambiguous when compared to the Early Bronze Age layers, which more or less unequivocally represent signs of use with a particular symbolic significance of at least the cave’s East Chamber. Below we provide some evidence to support this interpretation. The paved surface we call “floor” in layer 7 indicates that some care was invested in preparing the chamber for whatever purpose it served during the part of the Neolithic that this layer represents. A somewhat similar rock-built platform exists in trench 2 as well (supra), although it is of much smaller extent and possibly even of a different purpose. In the East Chamber we also found a relatively large number of “scoop” fragments. An unusual characteristic of the Agia Triada’s assemblage dated to this period is the relatively large number of spindle whorls. It is difficult to imagine regular spinning or weaving activities in an environment characterized by very damp and cramped conditions and the complete absence of natural light. Hence, the presence of these, otherwise mundane, objects inside the cave can be, in this case, very cautiously interpreted as evidence for their symbolic significance. Several bone tools were also found in the Late Neolithic I‑II layers of the Corridor area (fig. 9)29. It should also be noted that the pottery from the Neolithic layers is very fragmented. Nevertheless, on closer examination in laboratory conditions we determined that the percentage of sherds that can be joined is much larger than we originally anticipated.

20At the same time, other finds from Agia Triada’s Late Neolithic II/Final Neolithic layers do not depart much from an assemblage expected at a typical Attica-Kephala site (e.g., Kephala on Kea itself). The range of pottery seems to be fairly typical (fig. 12), including the “scoop” fragments, which are frequently found at other locations of the same period (e.g., at Kephala, the Ancient Agora in Athens, Skotini Cave on Euboea, etc.)30. According to preliminary observations all typical shapes and wares seem to be present in the assemblage, however, the detailed statistical analysis of the finds will give a more secure picture. The obsidian implements are also common, since obsidian is the primary raw material used for chipped stone tool production throughout central and Southern Greece and the Cycladic Islands during most of prehistory. Since obsidian artifacts are much more numerous in deposits associated with layer 7 than anywhere else in the cave, we conclude that activities that produced layer 7 required frequent use of cutting and scraping tools. One excellently preserved copper awl from layer 10, trench 9 (fig. 10), is the only piece of evidence for metal use in the Karystia during this phase; however, this reflects more on the general scarcity of excavation research than anything else. Finally, a large monochrome pot was discovered in the northeast part of trench 9. The pit in which this vessel originally lay (fig. 11) cut through several Late Neolithic II/Final Neolithic layers. It bears evidence of a perforation in its lower section and it was placed in this location intentionally, since we found horizontally placed flattish rock slabs and large sherds at its base serving as its support. The upper preserved part of this pot was first observed in layer 7 (stone-paved floor), while its bottom together with the rest of the supporting construction was found in layer 10, which could also possibly represent a surface with intentionally placed stones.

The Use of Space

21Agia Triada is different from most other Aegean caves, which were often used for habitation during prehistoric times. Past the first couple of meters from the narrow entrance, there is virtually no light in the cave unless provided by artificial means. At the same time, there are no easily accessible large galleries in Agia Triada that could accommodate a large group of people. Aside from the East Chamber, most of the first 100 m of the Agia Triada Cave consists of relatively narrow and sometimes steep corridors. Larger galleries exist, but they are located deep inside the cave and are difficult to reach, especially without caving equipment. Another factor that would probably not have prevented at least seasonal habitation in the cave, but that would have made it significantly more uncomfortable, is its dampness. Agia Triada is still geologically active and there is much running and dripping water everywhere in the cave. The dampness increases significantly during the fall and winter months, when precipitation levels in the Karystia rise, and especially during spring when snowmelt from Mount Ochi swells local water sources. Agia Triada contains an underground river and lake system that springs out of the ground only a few meters below the cave entrance.

22Based on preliminary geomorphological examinations conducted by Dr P. Karkanas31, it seems possible that the East Chamber was completely flooded several times during its attested cultural use. According to Dr Karkanas, within sandy deposits there are several observable lenses of reddish sand that probably represent remains of ancient fires washed away by the rising water levels32. Finally, the use of the East Chamber would not have been possible without a fire to illuminate the space and to provide some warmth33. The reduced oxygen (O₂) and accumulated carbon dioxide (CO₂) levels resulting from a burning fire in a restricted and poorly ventilated space such as the East Chamber would likely have quickly become intolerable, especially during a prolonged stay of a group of people consisting of more than two or three individuals.

23Hence, it is difficult to imagine a permanent year-round Neolithic occupation of the parts of the Agia Triada Cave we explored thus far. Alternatively, because of its relatively hidden and well-protected position and easy access to water, the cave could have been used as a place of refuge for a population from a nearby community. It seems unlikely, however, that its location was unknown to other synchronous Karystian communities. For example, if the threat that would have caused people to hide inside Agia Triada came from within the Karystia, then the cave would not have been the best hiding space; if the threat came from outside of the Karystia and if it affected more than one community, the cave would have been unlikely to accommodate everyone in need of refuge. This is the reason we are, at the moment at least, leaning toward the explanation of the Late Neolithic II/Final Neolithic evidence from Agia Triada as remains of symbolic significance. If that were the case, human presence in the cave would have been intermittent and connected to the specific rites conducted inside the cave. Ritual activity could have taken advantage of and incorporated the sensory deprivation produced by Agia Triada’s darkness and dampness. Even the physiological and psychological effects of depleted O₂ and increased CO₂ levels produced by burning fires could have played a part in the ritual. Moreover, caves, particularly the ones like Agia Triada, where separation from the outside world is nearly complete starting already at a short distance from the cave’s entrance, have been used as settings for ritual activity since Palaeolithic times. Solid archaeological and ethnographic evidence exists for ritual use of caves virtually on every continent and in most time periods, and this practice possibly predates anatomically modern humans34.

Chronology and cultural associations

24The available radiocarbon dates (table 2) provide evidence for two major phases of Neolithic occupation at the cave, one during the 5^th millennium BC and the second at the beginning of the 4^th millennium BC. We briefly discuss further in the text how these dates are related to the stratigraphic sequence presented above. Furthermore, we present some of the relevant material evidence from the cave’s Neolithic layers, with a caveat that all our observations are preliminary, since the detailed study of the material is still ongoing.

25Layers 6‑7 in trench 4 contain several pattern-burnished sherds (fig. 13), some sherds with grooves that were possibly filled with white paste, as well as several pieces of “cheesepots”35. There are also many “scoop” fragments (mainly fragments of handles), “elephant” lugs, coarse sherds with rope decoration, few sherds with other relief decorative patterns, and handles with vertical plastic protrusions (“spurs”). In addition, there are fragments of many coarse and red-burnished pots. Large and medium sized vessels predominate with a variety of shapes being present. This assemblage finds its best parallels at the settlement and cemetery of Kephala on the island of Kea36, as well as at Pefkakia in Thessaly, especially in its early and middle strata37. Geographically the closest parallels come from the briefly excavated site of Plakari in the Karystia38. Sherds with crusted decoration are few and not very well preserved.

26This situation seems to agree well with the radiocarbon dates from samples taken from layers 6 to 10 (Lyon-7637, DEM-2096 and DEM-2095), which span the years between 4300/4000 and 3980/3700 cal BC39. “Elephant” lugs and the first appearance of the pattern-burnished ware are usually placed, indeed, at the end of the 5^th millennium, around 4300‑4200 BC40. Agios Dimitrios in Trifyllia (Peloponnese), which produced pattern burnished sherds, can be considered as representing the terminus post quem for the beginning of the Late Neolithic II41, especially if one relates available radiocarbon dates and the similarities of the material found at this site with that from the site of Pangali on Varassova Mountain in Aetoloakarnania42, which represents the phase just before the appearance of the pattern-burnished ware.

27In the case of the Agia Triada cave, there is no stratigraphic discontinuity between the earlier and later parts of the Late Neolithic (Late Neolithic I and the Late Neolithic II)43; however, the typo-technological characteristics of the pottery indicate a change in ceramic traditions (see below). The available radiocarbon date from layer 13 in trench 4 (Lyon-7636, 5645 ± 50 BP = 4579‑4360 cal BC) points towards this transition. A date from layer 4 in trench 3 also falls within this interval (Lyon-7201).

28In the lowest layers of trench 4, a change can be clearly seen, indeed, in the pottery record, since dark-faced burnished sherds, eventually decorated with white-painted motifs, predominate (fig. 14). The available radiocarbon dates for the lowest cultural layers are also in accordance with the stratigraphic sequence and the appearance of the characteristic material culture, since they range from the first half to the middle of the 5^th millennium BC (DEM-2097, 5820 ± 52 BP = 4791‑4546 cal BC and DEM-2183, 5867 ± 60 BP = 4896‑4552 cal BC). However, it has to be noted that the material from these layers is very fragmented.

29The preservation of older cultural strata is better attested in trench 8. Dark-faced pottery becomes more numerous starting from layer 9. In layers 11‑13 we found characteristic dark-burnished sherds, open shapes with mostly S-shaped profiles, straight-sided bowls, etc. Unfortunately, the decoration on all painted sherds is very poorly preserved, making it difficult to examine in detail the specific characteristics of the painted wares in order to reach more secure chronological conclusions44.

30The material from trench 3 also consists mainly of a dark-faced, burnished ware. The shape repertoire includes a hemispherical bowl, two straight-sided bowls, one vessel that has vertical walls, and three vessels with everted rims. A monochrome Late Neolithic “scoop” was also found, similar to the shape known from the Skotini Cave, where it is similarly associated with white-on-dark sherds, which are dated to the LN Ib subphase by the excavator45. The painted decorative motifs found on the white-on-dark pottery from trench 3 include zigzag and wavy lines and concentric lozenges; however, multiple chevrons in many different arrangements seem to represent the most characteristic pattern. These shapes and motifs are characteristic of the Saliagos cultural stage at sites such as the Antiparos Cave46, Akrotiri on Thera47, Saliagos near Antiparos48, and many others. The material from Agia Triada seems also to be close to that from the Zas Cave on Naxos49. The Zas Cave is usually considered to be later than Saliagos or even Grotta on Naxos, but this horizon should probably be dated to the period around the middle or the second half of the 5^th millennium BC.

31According to preliminary observations, and taking into consideration (1) the dating of the white-on-dark ceramics from Skotini50, (2) the above brief discussion concerning the typological characteristics of the material from the Agia Triada Cave51, (3) the date of white-on-dark sherds from the Youra Cave in the Sporades further to the North52, (4) the new evidence from sites such as Ftelia on Mykonos53, the Antiparos Cave54, and Akrotiri on Thera55, it seems that the Late Neolithic I material from the Agia Triada cave mainly belongs to an advanced stage of the period56. However, the ongoing detailed study of the material from the deepest excavated layers will provide us with a better idea of the sub-phases of the Late Neolithic I present in the cave.

32The available radiocarbon dates for Agia Triada’s Late Neolithic II fall into the end of the 5^th and the beginning of 4^th millennium BC. It appears that the relatively late date of the white-on-dark pottery tradition found in the cave (middle-later part of the 5^th millennium BC, perhaps starting even earlier) indicate not too wide a gap between the two phases57. The Cycladic Late Neolithic material (and, now, the one from Agia Triada) is of major importance. The sites such as the Zas Cave, Grotta on Naxos, Akrotiri on Thera, the Antiparos Cave, Ftelia on Mykonos, as well as the material from the upper phase at the Saliagos site itself, provide evidence for the presence of characteristic late elements at the Late Neolithic I sites. This evidence implies continuity or at least some chronological and cultural association between the two traditions58. If we broaden our geographical perspective, we can see that the evidence from the Troad, such as the Beşik-Sivritepe horizon, can be placed before the Late Neolithic II, between ca 4800‑4370 cal BC59. This date corresponds to the advanced Late Neolithic I in Southern Greece, which is, in turn, in accordance with the evidence from the Agia Triada Cave. However, it has to be mentioned that the evidence from the cave discussed above is based on the current stage of study of the material and the stratigraphy, and the presence of even earlier phases cannot be excluded.

33The 4^th millennium BC dates in relation to Late Neolithic II are also not unusual. Besides the ones from Agia Triada, samples dated to the 4^th millennium come from the Skotini Cave, as well as from the Sarakenos and Drakaina Caves, among others60. Also, in the Leontari Cave, the Late Neolithic I early and advanced phases are present, as well as the Late Neolithic II, which seems to be associated with elements such as the rolled rim bowls and the red-burnished ware61. This evidence implies that a date in the 4^th millennium BC is not that uncommon. Also, material remains such as those from the Kitsos Cave in Attica, where pattern-burnishing is found on rolled-rim bowls, indicate that the so-called “Kephala culture” could have continued well into the 4^th millennium62.

34Moreover, some of the Cycladic sites usually dated to Early Bronze Age I (such as Agia Irini phase I) are of major importance for understanding the transition from the Neolithic to the Early Bronze Age, especially in relation to cultural continuity and the general difficulty in establishing a secure record of sites and absolute dates related to the 4^th millennium BC. Deposits from Agia Irini phase I63 are thought to belong to a phase later than Kephala, but earlier than the beginning of the Bronze Age64. Also, there are many other possibly related sites that have been recently located on the Aegean islands65. Of similar importance are also sites such as Tsepi in Marathon (Attica), with characteristic early Early Cycladic I material and crusted decoration66, or Mikrothives in Thessaly67. It seems also that the Thessalian “Rachmani phase” coincides with the later stages of the LN Ib subphase68, as it is known in Southern Greece69, and it could possibly continue well into the Late Neolithic II70. The early Late Neolithic II/Final Neolithic starts at the same time or a bit later, between 4300‑4200 BC, and seems to continue at least until the beginning of the 4^th millennium, around 3800‑3700 BC, as it is indicated by the radiocarbon dates available from Agia Triada, Tharrounia, Kephala, and other sites.

35With the presence of the Late Neolithic I white-on-dark ware well into the second half of the 5^th millennium BC, it seems that the gap with the “Kephala horizon” is minimized, indicating a well-established occupation before the appearance of the pattern-burnished ware. Moreover, the increasing evidence from radiocarbon dates, together with the existence of many candidate sites that need to be published or excavated in detail, could possibly fill this gap, especially in relation to the second half of the 4^th millennium BC. It is still unclear how long into the 4^th millennium the Kephala horizon extends; however, several currently available radiocarbon dates and other data (e.g. pattern-burnishing on rolled-rim bowls) provide indications that it is also attested during the 4^th millennium BC.

Conclusions

36Material dating to three distinct periods – Late Neolithic I, Late Neolithic II/Final Neolithic, and the Early Bronze Age – has been discovered in the Agia Triada Cave. The Late Neolithic I material is the earliest documented archaeological material in Southern Euboea found thus far, and by that fact alone is very important for understanding the cultural connections of what could be the very first Karystians. This material (pottery) seems to have been in use until an advanced phase of the Aegean Late Neolithic. The earliest pottery found in Southern Euboea is considered to be mostly dominant in the southern part of the Cycladic zone (which is something that still needs to be confirmed). Because of its presence in many insular and coastal contexts, it seems obvious that its producers/users maintained extensive relations with the rest of the Aegean area at this time. Moreover, this poses some interesting questions about the place of origin of the earliest Karystian settlers. Regardless, it is likely that there was a reciprocal, two-way process of contacts during this phase of the Late Neolithic between Euboea and the other Aegean islands.

37The character of the cultural developments in Southern Euboea from this earliest stage on is not yet very clear, especially since the next phase discernible in the archaeological evidence is clearly connected to the Late Neolithic II/Final Neolithic period as found at sites in Attica, at Kephala, Strofilas on Andros, and the Zas Cave on Naxos (see below). The existence of the white-on-dark ware during the middle/second half of the 5^th millennium BC bridges the chronological gap between the Aegean Late Neolithic sites and the Late Neolithic II Kephala tradition. In the Cycladic context, the presence of typical late features (e.g., crusted ware, cheesepots) gives some clues about the continuity between these traditions.

38The Final Neolithic/Late Neolithic II pottery from Agia Triada has features typical of the Attica-Kephala cultural horizon. It is similar to the one found at Kephala71, the Kitsos Cave in Attica72, and Plakari in the Karystia73, which is located just a few kilometers south of the cave. From the available absolute dates from several sites, it seems that this stylistic tradition continues well into the first half of the 4^th millennium BC. The lower limits of the Kephala tradition are still obscure, as well as their relationship with sites that seem to be dated to the second half of the 4^th millennium BC. However, only with new field research will the exact dates of these sites be better determined.

39The advanced Late Neolithic I and early Late Neolithic II represent subphases during which major changes took place, reflected in the use and meaning of material culture as well as in the settlement patterns74. They can be considered as new expansion episodes in previously underused or unused landscapes; new regions were inhabited (especially marginal areas, such as islands and higher altitude locations) in a more permanent way75. In the process, likely new ideas of how people relate to the newly inhabited regions without previous ancestral connections were created76. The so called “Later Neolithic Stages” in the Aegean represent a complicated cultural period, during which new ideas about the use of the landscape were developed while, at the same time, the use of the material culture took on a new content and meaning. It is a long period, since these changes started at least from an advanced phase of the Late Neolithic I, and it is a mosaic of stylistic and cultural traditions whose cultural and chronological relations have just started to be understood.