HOUSE (CENTRIC) SOCIETIES ON THE PREHISTORIC GREEK MAINLAND

Summary. The concept of ‘ House Societies ’ was introduced by Lévi-Strauss. It not only considers the social aspects of houses and households, but also the economic and political ones. By so doing, one gains a much more complete understanding of how a community is structured and organized, and how changes can slowly unfold within seemingly static communities. In this article it is argued that Early Helladic III and early Middle Helladic communities on the Greek mainland can be considered proto-house societies, while House Societies proper do emerge at some places during the later Middle Helladic and Late Helladic I periods. The arguments are based on an analysis of domestic architecture, mortuary practices and the domestic economy. These indicate that the concept of property became more important during the Middle Helladic period. The house was perpetuated through rebuilding and mortuary practices, thereby referencing the concept of kinship, in order to transmit property. Such habits define House Societies.

economy requires that households share and pool goods and services.Clearly, households of primitive communities are usually not entirely self-sufficient, but are to some extent dependent on help from other households or communities (e.g.Bintliff 1999, 526).Crops can always fail, in which case households share produce, based on reciprocity.If a household wants to become more prosperous and rise above others, both economic and social changes need to take place first: production needs to be increased, an incentive is needed for this, and the sharing of foodstuffs must be less of an obligation.Less affluence can lead to a decreasing obligation to share, and as a result an increasing sense of property.For example, the availability of land and resources affects the concept of property and the development of social differences (Sanders and Webster 1978).If land is scarce, people will try to maintain possession of the land they have (Meggitt 1965, 218).Instead of sharing, households will now hoard more of their resources.Of course, some sharing still takes place, as households are not likely to be entirely self-sufficient.Less sharing can be an incentive for the household to produce more, as the extra produce can be of direct benefit to the household.Cross-cultural analysis has showed that larger households are more likely to benefit from such a situation than smaller households (Netting 1982).Concepts of property and ownership in prehistoric Greece have been examined, especially for Neolithic Greece (Halstead and O'Shea 1982;Halstead 1989;1995;1999;2006). Halstead (1995) argues that in order to cope with a more chronic and widespread scarcity of land and resources during the later Neolithic, Greek households became independent economic units, isolating themselves from the community at large and withdrawing from community obligations to share food.However, these concepts have barely received consideration in studies of the subsequent Greek Bronze Age.
The house-society model is one of many to approach communities and their built environment and economy.This is not to say that it is the most suitable or only model to understand specific communities, but as opposed to some other models it functions in a more inclusive way: it not only considers the social role of houses, but also their economic and political role.As such, one gains a much more complete understanding of how a community is structured and organized, and how changes can slowly unfold within seemingly static communities.Indeed, the concept of house societies has become popular in archaeology and anthropology because of its applicability to the archaeological material (e.g.Carsten and Hugh-Jones 1995a;Coupland and Banning 1996;Gillespie and Joyce 2000;Beck 2007).This applicability is related to the (im)possibilities of the archaeological record.First, it is very difficult to reconstruct kinship systems of non-literate prehistoric societies.This can be resolved by taking instead a house-centric perspective (Waterson 1995, 47).Secondly, what archaeologists can study are material remains and patterns that inform us about social actions and practices in the past.Indeed, scholars using the house-society model see the house as the expression of social groupings, and focus their studies on processes and practices (Gillespie 2000a, 41-2).As such, two meanings of the house are brought together: the mundane in which household members undertake daily activities, and the ritual, embodied in property, transmission and linked to ancestors and kinship (Carsten and Hugh-Jones 1995b, 45).Thirdly, houses are multigenerational, their survival depending on their consolidation, maintenance and own social reproduction.Among these factors are various actions that can be detected by archaeologists, for example the continuity of place, the life cycle of houses, and the use of heirlooms (Driessen 2010, 41).
Lévi-Strauss proposed house societies as a type of social structural formation, but not one necessarily applicable to every pre-modern society.The concept 'should motivate an investigation of the interconnected pragmatic actions and strategic motivations that link persons over time to and through objects or places and thereby serve to define a social group, enable its relations with other persons and groups, and facilitate its social (and accompanying material) reproduction' (Gillespie 2000a, 50).Study of house societies appears especially valuable in contexts of socio-political transformations where egalitarian kin-based societies become ranked, and are in a transitional phase towards social stratification and territory-based systems (Waterson 1995;  Moore 2005, 184).Egalitarian societies where houses play a central role, but strong socio-political inequalities and power competition do not exist (beyond age and gender) may better be defined as house-centric societies (Gonzáles-Ruibal and Ruiz-Gálvez 2016, 390).Houses have also been referred to as 'proto-houses' (Schrauwers 1997, 364-5) or 'embryonic houses' (Lévi-Strauss 1987, 192;Sandstrom 2000, 68) when not entirely adhering to the definition of Lévi-Strauss.
Socio-political transformation takes place during the EH III-LH I periods on the Greek Mainland.Earlier studies highlighted the continuity of the house and the increasing importance of property and transmission thereof (Georgousopoulou 2004;Milka 2006;2010;Wiersma 2014).In the following, I set out how at some communities a transformation indeed takes place from 'proto-houses' to Houses.Following the definition of Lévi-Strauss, I specifically address the following three aspects: 1) continuity in house architecture, 2) the exploitation of the ideas or concepts of kinship or alliance, and 3) the passing down of (im)material property.

EARLY HELLADIC III
The EH III period (c.2200-2000 BC) is marked by destruction of settlements and an overall depopulation (Forsén 1992): it has for long been considered a 'Dark Age'.More recently, some have argued for considering this rather as a period dynamic and rich in innovation, bringing about the loosening of traditions and regulations and opening up possibilities for a renegotiation of social values (Weiberg and Finné 2013;Weiberg 2014;2017).The available body of data consists of a comprehensive overview of domestic architecture (Wiersma 2014, 191-201), encompassing 79 houses located at 19 settlements.Most settlements are represented by a few houses, but a few settlements dominate.Architectural patterns attested at well-published or well-represented sites are often also attested at smaller and less well-published sites.Site size could often not be established due to limited investigations, but when reconstructed ranged from generally a few hectares to (only occasionally) 5-10 ha.

Continuity
EH III domestic architecture was very uniform, homogenous, usually consisting of apsidal one-or two-roomed houses (Fig. 1).Continuity is seen in the house shape, and in rebuilding practices.Approximately one-third of the houses were rebuilt.Most of the rebuilt houses were done in such a way that they made practical use of several or all former walls to construct the new house on, while some other houses were rebuilt in such a way that they did not make use of the former foundations, but still replicated the former house with respect to its general location, orientation and shape.
The homogeneity of houses implies that conformity, accepting the norms, and hence belonging to the community were important.The house architecture solidified community relationships by its homogenous appearance.EH III communities were generally small, under 2 ha, and we must assume that community involvement and integration must therefore have been high.But the house architecture also communicated a certain sense of isolation of the household, as houses were freestanding and usually had a single entrance.That the household was an important and independent unit is also indicated by the rebuilding practices.The differences in the meticulousness of rebuilding seem to rule out the interpretation of this action as solely functional.A more symbolic interpretation should be considered: this action resulted in the physical continuation of the house and the household in space and over time.I therefore suggest that rebuilding of houses was a means to symbolize continuity and transmission of the house and household and to reinforce rules of transmission of the household vis-à-vis the wider community.

Kinship or alliance
Besides the rebuilding practices, kinship or alliance are little alluded to in the archaeological record.The intermural burial of some people (especially premature babies and neonates) may be interpreted as an act to make them symbolically part of the community.In addition to graves, a few tumuli may date to the EH III period, though whether these were for burial or religious usage remains sometimes unclear.The tumuli, because of their 'size and visibility in the landscape, make it all the more likely that they functioned as communal symbols and played an important role in the religious life of the communities' (Whittaker 2014, 93).In conclusion, the mortuary practices especially emphasize the community.

Property
The dualism earlier observed in EH III house architecture is also reflected in the household economy.On the one hand the household is an independent production unit, which safeguards its domestic welfare.On the other hand, the household is submerged in the larger community and needs to comply to broader communal or kinship relations that force members to share resources.In such an economy, it is not possible for a household to become rich or rise above others.As expected, storage took place at the EH III household level in very modest quantities.Some evidence has been found of stone and mudbrick storage structures inside houses, storage vessels in the partitioned back rooms and/or storage pits, the so-called bothroi.There are only two exceptions that indicate some communal storage may possibly have taken place as well, namely at House 4A at Argissa (Hanschmann and Bayerlein 1981) which contained an extremely large storage construction and at Tsoungiza House E which contained nine storage vessels and where communal grain processing may have taken place (Nilsson 2014).Based on the available archaeological data, I argue that household property and ownership were seemingly not important concepts for most EH III communities.Households were largely economically submerged in the larger community.Some houses may have had a larger storage capacity, but this potential economic difference did not translate into social or political differences between households.
In conclusion, the EH III communities do not adhere to the definition of House Societies.They could though be considered as proto-house societies: the household was the most important social unit, emphasized by the freestanding (and rebuilt) house and the private storage of produce.The divisive effects of private storage may have been countered through meat-sharing practices on various group levels (Halstead 2011) and small-scale drinking events (Rutter 2008), while intramural burials and tumuli may have been a means to symbolize the importance of the community.

MIDDLE HELLADIC I-II
The MH I-II period (c.2000-1750BC) sees a slow and uneven increase in population and settlement numbers (Zavadil 2010).The mainland becomes less isolated when one considers the increase of trading relations (see various contributions in Philippa-Touchais et al. 2010).The body of data now consists of 92 houses located at 20 settlements (Wiersma 2014, 202-12).As during the previous period, most settlements are represented by a few houses, but at six settlements five houses or more were uncovered.Site size could often not be established due to limited investigations, but when reconstructed ranged generally around 2-3 ha and in one case 8-10 ha.

Continuity
At first sight, the same kind of architectural continuity is seen as in EH III: the domestic architecture appears homogenous and approximately one-third of the houses are rebuilt (Wiersma 2014, 209-11).However, some minor changes do take place.Increasing variation is seen for example in house shape, size and number of rooms.With respect to house rebuilding a slight increase is seen in MH I, followed by decline in MH II.When houses are rebuilt less use is made of old walls than during EH III, though houses are rebuilt on more or less the same location.The rebuilt house seems to fix the memory of the location of the former house, without necessarily being an exact replica of the former house.In short, architectural continuity remains important during the MH I-II period.

Kinship or alliance
Kinship and alliance are more clearly communicated now through the purposeful placement of burials: first, burials represent all age groups, secondly, they cluster around houses, thirdly, they alternate with houses at a few locations and finally they appear in extramural cemeteries.I elaborate on each of these developments.
The MH I-II mortuary record consists generally speaking of intramural burials, and single contracted inhumations in simple, small tombs (pits, cists, jars) with few if any offerings.Unfurnished graves cannot be dated closely, which is problematic and makes observations difficult.However, there are a few well-studied sites in the Argolid, namely Asine, Argos and Lerna: we see here in the mortuary practices the inclusion of the ancestors by actions designed to foster their memorial and connection to the house or household.The fact that adults are also buried among the settlement remains has thus far not received much consideration and emphasis.It is unclear where the adult population was buried before, during EH III.Apparently, during MH I, there was deemed to be a certain need for deceased adults to (symbolically) dwell among or near the living.Georgousopoulou (2004, 211) argues that the adult burials among MH I settlement remains at Asine may have been used to demarcate property, as representative of the essence of a group's identity.The dead were placed within plots of land as markers of traditional property.
This leads to the second aspectburials clustering at or around houses, as can be seen at for example Asine and Lerna (Voutsaki and Milka 2017).At Lerna, families or kin groups seems to have been the basic principle by which social relations were structured, as is manifested by the clustering of graves and their association with specific houses, but also by the shared features within burial groups and their persistence through time.Milka (2006;2010), in studying graves at Asine and Lerna more closely, concludes that burials opened inside houses still in use were mostly of neonates and infants, and that they date to the early MH phases.The majority of the adult graves were dug into ruined houses or abandoned settlements.Milka also observed at a few sites a pattern in a sequence of events: an alteration of habitation followed by desertion and burial and then again habitation.This practice implies a growing emphasis on demarcation and continuity of the household and the safeguarding and transmission of resources.
The marking out of social groups also happens outside the settlement.The first extramural cemeteries appear in MH II (e.g.Voutsaki et al. 2007).What can also be observed are localized burial habits, as in the appearance of extramural tumuli in the south-west and west Peloponnese (Boyd 2002;various contributions in Müller Celka and Borgna 2012).Tumuli usually consist of pithos burials horizontally inserted into the mound; they contain an adult burial and are accompanied with modest grave goods, though there are some exceptions (Rambach 2007;2012; Korres 2012).The appearance of extramural cemeteries has thus far been particularly interpreted as evidence for the rise of an elite.For example, it is suggested that the social group using the East Cemetery of Asine (Fig. 2) claimed status by burying their group members in a distinct cemetery (Voutsaki, Ingvarsson-Sundström andDietz 2012, contra Milka [2019, 454-8] who argues that people of different origin were buried here).From the perspective of (proto) House Societies, the appearance of extramural cemeteries seems to signify the growth of social houses that encompassed more than a single household.Therefore, burial in conjunction with a single house is no longer suited to communicate messages of transmission, and so new burial locales are created.
In conclusion, mortuary practices emphasize the continuity of the house during MH I-II.This practice is not limited to a few houses, but widely shared within the community.Some exceptional architectural and mortuary monuments existedwhich due to a lack of space cannot be discussed here.These may have been expressions of individuality, indicating a relatively modest departure from the communal norm under the impetus of an impromptu personal decision (Boyd 2002, 94;Georgousopoulou 2004, 212).Extramural cemeteries became new spatial foci where people could negotiate and change their positions or relations (Milka 2006, 7) within a larger social group.This change in focus during MH II could go some way to explain the decrease in the rebuilding of houses.

Property
The clearer articulation of household or kin groups suggests that property became more important during the MH period.At some settlements, (possible) storage structures were constructed among the houses (Wiersma 2014, 234-5).The construction and use of these facilities, possibly by multiple households, suggest increasing economic cooperation.At the same time, signs of increased segregation of some households may also be seen, as a few households fenced off an area beside the house (Fig. 3).Settlement space was as a result demarcated and appropriated by the households, signalling the security of resources and property.The two economic developments of increased cooperation and yet isolation seem contrasting, but may be part of the same phenomenon: some households started to move away from obligations to share with the community at large, but their aspirations were still restricted by social obligations to (part of) the community.
The location of storage installations inside houses seems significant in this respect.Some storage facilities were moved in some houses from the inner recesses to the front room, porch or even into sheds or buildings in the yard (Fig. 3).In most cases the storage facilities remain entirely or partly shielded for plain view (by means of [partition] walls), but at the same time become more accessible.Some space remains devoted to additional storage inside the house.It therefore seems that some households divided their produce: one portion is privately stored inside (the back of) the house, and another portion is stored in a more accessible, even external, area.Perhaps this produce is owned by a group of people that is not limited to those inhabiting the house.
Too little data are available for mainland Greece to securely assess whether the outlined developments regarding the domestic economy took place at specific settlements, in specific areas, or at specific moments in time.The available evidence suggests that economic developments occurred especially in the Argolid and in central Greece and at settlements of 2 ha or more in size.Large settlement size could have been responsible for increased economic cooperation in two ways.First, the larger population size could put pressure on the availability of land and resources, as well as on interpersonal relationships and decision-making processes.One means of coping with such pressure could have been the enlargement of the social group.This enlargement could have led to increasing economic cooperation of the households forming the social group, bringing more economic stability to cope with such pressures.Sharing of resources would take place within this group, rather than with the community.Second, settlements of 2 ha or above were probably large enough for social groups to be endogamous, which could have increased the importance of fostering social relationships between households within a single settlement.Therefore, depending on settlement size, settlement density and socio-economic dependency, some communities and households could have collaborated with their neighbours, while others competed with them.
Pulling together the evidence from the MH I-II domestic and mortuary contexts, we can conclude that most mainland communities still qualify as proto-house societies, with some beginning to show characteristics of House Societies.Changes in the domestic mode of production, the increasing concern for transmission and the use of the concepts of continuity, kinship and alliance all indicate the appearance of social and economic groups that in some cases encompass more than a single nucleated household.In MH II, we are on the threshold of potentially major changes in both the social and economic spheres.During the MH III-LH I periods (c.1750-1640 BC) settlement numbers rapidly increased (Wright 2004;Zavadil 2010, 162).Interaction with other areas, especially with Crete, further intensified at certain locations on the mainland, and significant changes occurred in the mortuary sphere (Voutsaki 1999).A total of 80 buildings from 24 different sites are attributed to this period, but most belong to the earlier MH III period (Wiersma 2014, 212-21).The underrepresentation of LH I buildings is, among other reasons, due to their obliteration by later and often extensive building activities.Reconstructed site size ranges generally around 0.5-4 ha, with an occasional outlier of up to 12 ha.

Continuity
The falling away of rebuilding activities that had set in during MH II continues: indeed houses almost cease to be simply rebuilt in the same way and on/close to the same spot.There are three social and economic explanations for this: 1) overall, it seems that instead of rebuilding, now making internal modifications or additions to the house became the preferred modus operandi; 2) new types of houses (room complexes or buildings with multiple axes) started to replace the freestanding, axially designed houses (Fig. 4); and 3) some settlements were deserted towards the end of the MH period.Maran (1995) suggests this phenomenon may be related to a restructuring and reorganization of settlements, which in turn arose out of a polarizing of social disparities within society.Perhaps the new types of houses required virgin territory to build upon.But we also see settlement relocation to higher, more defensible locations.
With the introduction of new types of houses, the earlier observed homogenous architecture slowly disappears.More and more architectural variety is seen in house shape, layout and size.Houses become increasingly detached from communal obligations and homogenizing forms of behaviour.I return to this development below, when discussing the domestic economy.

Kinship or alliance
Kinship or alliance became even more important realities during the MH III-LH I period.Both continuity and change are seen in the way these concepts underlie mortuary behaviour.Long-term continuity is seen in the mortuary practices, as the dead are still incorporated among the ancestors (Boyd 2002, 94), whereby clusters of burials are also seen in relation to houses, for example at Asine and Lerna (Milka 2019).New grave types especially designed for multiple interments (shaft grave, tholos, chamber tomb), the reuse of graves, secondary burial of remains, and the clustering of burials all indicate an increasing emphasis on ancestors, kinship and descent in a period when ranking in status becomes important (Cavanagh 1978;Voutsaki 1995;2010;Voutsaki and Milka 2017).Additionally, certain groups are being elevated well above the mass of the community, marked out in the mortuary record: the Grave Circles at Mycenae are the best known examples.Age and gender differences become also more pronounced (Pomadere 2010;Milka 2019).
In conclusion, kinship and alliance become even more important concepts in the MH III-LH I mortuary record.At the same time, a stronger compartmentalization of the society is seen, in that the community as a whole becomes more subdivided into separate social groupings, in which growing distinctions are also made along age and gender lines.The increasing variety seen in the architectural record is thus reflected in the mortuary practices.However, it is in these last that the appearance of elite families is most clearly detected, groups able to command and exploit (inter) national networks, bearing in mind the valuable or imported grave goods deposited with them.The obvious conclusion is that the clearer articulation of kin groups or larger social groups ties in with those changes in property or wealth taking place during MH III-LH I. Wealth distinction was not however so clearly visible in the MH III-LH I house architecture.Nonetheless, we return now to the MH III-LH I houses to assess if and what kind of changes took place in the domestic economy.

Property
It appears that in some regions and at some settlements pressure on the availability of land and/or resources was experienced.Settlement numbers increased during the later MH, with marginal and inland areas settled and increasing interaction attested on both a local and on a regional scale, often expressed in the increasing strength and influence of Minoan Crete.We can well imagine that these circumstances may have caused increasing local competition, tensions and threat.It caused changes in the relationships between and within settlements; some sites (Mycenae, Kolonna) are more clearly ranked.Some communities start to demarcate the settlement area through the construction of defensive and circumference walls, and the positioning of extramural cemeteries.This is especially attested in areas experiencing greater interaction, such as the Saronic Gulf.
The increasing size and segmentation of houses (Fig. 4) suggest that developments took place regarding cooperation and/or activities and production.We see in some houses a process of upscaling (Knappett 2009): houses become larger and were supposedly inhabited by a larger social group of people.For example, at Asine, it was suggested that multiple households may have inhabited a single structure, and economically cooperated (Voutsaki 2010).Other possibilities to account for the large size of houses could be the enlargement of the household, the performance of an increasing number of specialized tasks, or the simultaneous performance of more tasks in the house.The domestic architecture indicates that kinship relations remain important, but undergo a transformation from the household or small kin group to an extended kin group.Some of the developments first observed in the MH I-II period continue to occur during the MH III-LH I period: the construction of courtyard houses or house complexes, the construction of separate storage-facilities, and the movement of some stores to more accessible areas.At MH III Eutresis, evidence of what appear to be distinctive food processing, storage and preparation facilities have been found, including hearths, an oven and storage bins and vessels (Fig. 5).It remains unclear by what kind of groups these facilities were used, but the benefits of economic cooperation have been outlined above.It seems clear that the decreasing economic dependence of households on the wider community resulted in less homogenizing forms of behaviour, as seen in the increasing variety in domestic and mortuary architecture.
Artefact assemblages in the mortuary and domestic sphere also suggest intensifying importance of household economics.Storage vessels as gifts in mostly late MH burials could be a means to communicate the success of economic reproduction achieved.More precisely, the inclusion of vessels intended for storage and consumption in MH III-LH I high-status burials at Asine and Mycenae is interpreted as a sign of control over specialized agricultural produce (Wright 2008, 245).The acquisition, production, consumption and transmission of (im)material goods of the social group are thus more emphasized than hereto.At the same time, fostering relationships with other social groups is also considered important.The pairing of tableware vessels in domestic contexts and especially late MH-LH I (richer) tomb contexts could signal the value attached to inter-household relationships through drinking ceremonies (Nordquist 2000, 3-4).The pairing and sets of vessels here imply the importance of hospitality, sharing and conviviality (Nordquist 2000, 15).
I argued earlier that in MH II mainland communities were on the threshold of potentially major changes on the social and economic scale and that some communities showed even then some characteristics of House Societies.During MH III-LH I this threshold is crossed by various communities, with many more to follow during the subsequent centuries.The appearance of House Societies on the Greek mainland is evidenced in the continuity of the House (a social group extending beyond the nucleated household), where the ideas or concepts of kinship or alliance and the passing down of (im)material property are mobilized.That the physical house no longer demonstrates during MH III-LH I the previous clear continuity of being meticulously rebuilt could be explained in the changes occurring in the domestic economy, which required new kind of buildings, with some houses or settlements being relocated in response to changes in the economy, such as new attitudes emerging as to land-holding.

EH III-LH I HOUSE CENTRIC SOCIETIES: CONCLUSIONS
Although the concept of House Societies was first developed in the field of anthropology, it also has value in archaeology.Especially in studying societies undergoing socio-political change and moving from being ranked to even more socially stratified, and others where social organization is 'upscaled' from the nucleated household to a social group that encompasses more than kin-related household members.Although the House is often materialized in a structure or building, it can also be a sacred place or a burial ground.
The preceding discussion has maintained that EH III and MH societies can be considered proto-house societies: the household, embedded within a structure, was the most important social unit as indicated by rebuilding practices, intramural burials, and the alternation of cycles of burial and building at a few locations.Strong socio-political inequalities and competition for power and position did not exist beyond matters of age and gender.However, changes took place in the domestic mode of production during the MH-LH I period: property, the demarcation of property and the transmission of property became more important, with reference to the concepts of kinship or alliance, and culminating in the appearance of House Societies at some mainland settlements.The social alliances made during the MH III-LH I period seem to have been unstable, constantly shifting and changing.It is possible that Houses were therefore not materialized in specific grand buildings, but rather in monumental and exclusive burial grounds.
Different concepts of houses existed later during the LH period as is suggested by textual evidence, stemming from Linear B tablets: ko-to-no, referring to an estate; don referring to the individual or nuclear house; and wo-i-ko referring to the home or house proper (Driessen 2010, 55).Further research into domestic architecture is needed to explore to what extent also later LH communities can be considered House Societies.LH mortuary practices have been intensively studied and indicate that the mortuary sphere continued to be used as an arena where kinship or alliance are emphasized.The fact that some burials inside collective funerary structures, such as tholoi or chamber tombs, contained many precious gifts, while others did not, may be related to the prevailing domestic economy and the presence of surplus (Hayden 1995).The next step here would be to investigate if and how the appearance of House Societies on Mainland Greece are related to the emergence of palaces during LH III, and if these palaces should be considered as extremely successful Houses.

Figure 1
Figure 1 Lerna.Plan of early EH III settlement (author, modified after Rutter 1995, plan III).
HOUSE-CENTRIC SOCIETIES ON THE GREEK MAINLAND OXFORD JOURNAL OF ARCHAEOLOGY 150 © 2020 The Authors Oxford Journal of Archaeology published by John Wiley & Sons Ltd MIDDLE HELLADIC IIII-LATE HELLADIC I