Lesson 6: Narrative

The Early Minoan Period: The Tombs

Cave Burials

Burial in caves was the standard form of burial during the Late Neolithic period in the north and east of the island. There is no evidence so far for LN burials in southern Crete predating the earliest use of tholos tombs in the Final Neolithic. During Early Minoan (EM) I-II, burial in caves continues in the north and east and lasts into EM III and even MM IA in the east.

Burials in caves are almost always found in a highly disordered state with the bones of numerous individuals all jumbled up together (a single exception at Ellenes Amariou was an extended inhumation lying on its back). Often a number (but not all) of the bones are burnt. The reason for the jumbling up of the bones is probably repeated burial in the same small area with no regard being paid to earlier burials. Burnt bones are probably the result of periodic fumigations rather than of sacrificial rituals or of funerary meals involving cooking. There is no significant evidence for a cult of the dead.

Rectangular Built Tombs (“House Tombs”)

Typology – There are two broad divisions of the form known as the {Minoan house tomb}

(1) Complexes consisting of a series of long and narrow parallel chambers within a single rectangular building (Palaikastro, Archanes, Gournes, Siva, Platanos)

(2) Complexes consisting of a group of square and oblong “rooms” within a single building (Mochlos, Gournes, Palaikastro, etc.).

There are no obvious chronological differences between these two classes and they are sometimes both found at the same site. The tombs of the second class, best represented at Mochlos, have walls built of stone slabs which sometimes still stand to a height of 2.5 meters. In some cases, a bedrock cliff-face is used as the back wall. These buildings were certainly roofed, presumably with flat rather than pitched roofs. Their door jambs were built of several courses of stone, while their entrances were closed with large stone slabs.

Origin – Parallels with contemporary domestic architecture have suggested to some that these tombs were intended to copy houses of the living and hence to serve as true “houses for the dead”.

Distribution and Chronology – This form of tomb is most common in the northeast and east. The earliest dated examples begin to be used in EM II and they generally go out of use during the MM II period, although the latest dated tomb at Mochlos (VI) was used as late as MM III. In the Mesara, complexes of long parallel chambers (i.e. Class 1 above) appear next to the tholoi at Siva and Platanos and seem to be late EM or even MM additions designed to contain bones and grave goods removed from the tholoi proper (i.e. ossuaries).

Late Types – During MM I, monumental versions of this tomb type appear. At Ayia Varvara B, the tomb consists of a complex ca. 13 meters square into which are packed seventeen cells of irregular shape and size.

The most impressive monumental example of this tomb type is Chrysolakkos at Mallia which measures 38.50 by 29.80 meters. In its initial use (dated to the final pre-palatial phase, MM IA), this tomb consisted of a complex of small rectangular chambers fronted on the east by a rectangular unit consisting of a long N-S corridor on the west and a series of “cult rooms” at the south and east. The west side of the corridor features a set of {orthostate}s [cut blocks set in such a way that their greatest dimension is oriented vertically] alternating with niches. To the east of this facade unit is a raised altar and a kernos set into the floor at its base. A {kernos} is a stone slab having multiple shallow circular depressions hollowed out of its upper surface, usually lined up in a partial or complete circle or quadrilateral with rounded corners; often found in the immediate vicinity of entrances, such slabs have been variously interpreted as gaming boards or as some kind of receptacle for small offerings (so-called “first fruits”). The odd corridor featuring niches and orthostates, as well as a type of cup with a pointed base that has been found associated with this phase of the tomb complex, have been compared by Watrous with contemporary Egyptian forms; neither the architectural feature nor the vase shape has a convincing Minoan pedigree.

Shortly after its construction, this whole complex of chambers was enclosed within a single wall, much of it consisting of a splendid series of hard gray limestone blocks that are characterized by clear saw-cutting marks and deep dowel holes made with a large tubular drill. These technically very sophisticated building blocks are all in secondary use in the positions in which they were found at Chrysolakkos, so must originally have been used in an earlier building, very possibly sited at some distance away from their present location and quite likely not to have been a tomb. In this second phase of construction at Chrysolakkos, the level of the area outside of the surrounding wall to the east was raised and the earlier corridor and “cult rooms” were buried. A paved area extended around the whole complex, bounded along the east side by an imposing colonnade of large piers, square in plan. Inside the main enclosure wall, larger and more substantial interior walls than heretofore divided up the grave chambers in a new way. The absence of preserved entrances suggests that access to the individual chambers was through the roof.

Method of Burial – Tombs of this form were used for multiple inhumations. They appear to have been periodically swept out and the bones and grave goods piled up in one area against a wall or jammed into a particular chamber. In the EM III and MM IA periods, burials in larnakes and pithoi began to be placed in tombs of this type. The appearance of individual larnax and pithos burials and the roughly contemporary multiplication of “cells” at Ayia Varvara and Chrysolakkos indicate a concern for individualized burial which is absent in the fully communal tombs of EM II.

Both Soles and Whitelaw have commented on the fact that the house tombs of Mochlos and other sites in northern Crete (e.g. Gournia and Sphoungaras, Mallia), on the assumption that each tomb of this type served a kinship group probably to be identified as a single nuclear family, provide evidence for a hierarchical social organization, since the tombs are differentiated by size, proximity to the settlement, architectural elaboration, and quality of grave goods. But this claim has been emphatically denied by Watrous, who provides alternative explanations for the distinctions noted by Soles at Mochlos and Gournia and who considers the tombs at Mallia cited by Soles to be almost entirely of protopalatial (MM IB-II) date.

Cist Tombs

Typology – The types are essentially those in use in the contemporary Cyclades.

Distribution and Chronology – Cist graves are relatively rare in Crete and occur chiefly in the northeast (Mochlos, Pseira, Zakro, Ayia Photia) where Cycladic influence was strongest, although there is at least one example in the south at Arvi. The type does not continue beyond the EM period, and is undoubtedly a facet of the strong Cycladic influence detectable in several different media in EM II Crete. At Ayia Photia, the ceramic and bronze finds from tombs of this type exhibit unusually close Cycladic connections dating from as early as the EM I phase.


This is by far and away the most distinctive and also most problematic of all EM tomb types. More than seventy of these tombs are known from upwards of forty-five different sites. The vast majority of these tombs, and all of the early ones with the exception of that at Krasi, are located in the south of the island, mostly in and around the Mesara plain. The earliest tholoi, in at least one case (Lebena A) in use as early as the Final Neolithic, are found south of the Mesara plain proper on the southern slopes of the Asterousia mountains. For over a millennium, this type of tomb appears to be the only one commonly employed for burial in southern Crete and in some instances (e.g. Kamilari) such tombs were still being used well into the Late Bronze Age (certainly as late as LM I, possibly as late as LM IIIA).


In plan, the {Minoan tholos tomb} is circular, with a diameter of between four and thirteen meters. The generally thick walls (0.70-2.50 m.) are constructed of unworked fieldstones (although Kamilari has roughly worked blocks), usually with larger stones on the inside and outside faces enclosing a core of smaller stones; clay is used as a binding medium. These earlier Minoan tholoi are always founded on bedrock, and are built either on a level surface or against a rocky overhang (Ayios Kyrillos, built partially underground, is a notable exception). That is, these tholoi are not sunk into the earth or into a hillside. Entrance into the tomb is almost invariably from the east. Doorways, usually small (rarely over 1.00 x 1.00 m.) and in the EM examples normally consisting of a simple trilithon arrangement, are almost invariably closed by a large rectangular slab on the outside. Lintel blocks are occasionally thicker in the center, an indicator that builders were concerned about the possibility of breakage due to the heavy loads these blocks carried, but only one tomb (Megali Skini) preserves evidence for a triangular corbel-vaulted void above this block, a so-called {relieving triangle}. This common feature of later Mycenaean subterranean tholoi is designed to redistribute the weight of the superstructure above the center of the lintel by angling it at a diagonal to the ends of the lintel and the jambs of the doorway below; as a result, the lintel itself is “relieved” of the weight and so is less likely to crack and eventually break at the midpoint of its span.

Complexes of rectangular rooms (“annexes”) are frequently built against the circular wall of the tholos around the entrance. Only in the MM I tholos at Apesokari do these annex walls bond with the tholos itself, thereby indicating unambiguously that the annex and tholos were built at one and the same time; at this site, however, the bonding in question appears to belong to a rebuilding of the circular tomb chamber at some point following its original construction. Thus certainly in most, and perhaps in all cases, the annexes were secondary accretions against the outer wall of what was originally a plain circular tomb chamber. Some annex rooms which lack normal doorways can only have been entered from the roof.

Most tholoi reveal a slight overhang of the interior wall – that is, most tholoi are corbelled to some extent. In some tombs, the interior faces of the walls have been hammer-dressed to form an even curve, an element of sophistication unattested in contemporary domestic architecture. A number of tholoi have a series of slabs projecting from the exterior wall. These slabs do not always occur in the same area of the plan nor is their number fixed nor do they appear only in tombs of a particular date or area. They are usually placed between 0.3 and 1.0 m. above ground level and between 0.2 and 1.0 m. apart. Explanations for these slabs include: (a) (Parabeni, Xanthoudides): They were used to “key” a covering mound of earth over the tomb [but there is no evidence at all that such a mound ever existed] (b) (Levi, Xanthoudides): Such slabs constitute a type of scaffolding used during the tomb’s construction [but why do they occur only at a low level and extend around only a part of the tomb’s circumference?] (c) (Branigan): They facilitated access to the top of the wall so that the wooden roof over the tomb could be removed during fumigation of the tomb [but again, why do they appear only at such a low level?]. A number of tombs have a thickened wall in one portion of the circumference (e.g. Apesokari). Platanos A and Marathokephalon II have a series of short wall stubs projecting perpendicularly from the circular wall, and some have interpreted these walls as buttresses, although they seem much too flimsy for this purpose. No doorways are so small as to suggest that burial was made through the roof instead.

The Problem of the Roof

The question which for a long period was dominant in studies of these tombs is whether or not they were covered with a complete stone vault. That is, what did such tombs look like in elevation?

Arguments in Favor of a Stone Vault

(1) The tholoi must have had a roof of some kind (although Seager did in fact suggest at one point that the tombs were simply unroofed). There is no evidence for a flat roof in the form of internal supports, which would have been absolutely necessary for the large span of the bigger tombs. Therefore, the roof was vaulted.

(2) Evidence for a collapsed stone vault is frequently found in the form of masses of collapsed stonework inside the tomb above the floor.

(3) The corbelling of the interior wall face indicates that the tombs were covered by a corbelled vault.

(4) The exterior walls are often extremely thick and tend to be thicker the larger the tomb. This extreme thickness is designed precisely to sustain the immense weight of a complete stone vault.

(5) The thickening of the exterior wall at some points, the occasional use of protruding wall stubs as buttresses, and the occasional building of a tholos against a natural stone outcrop are all additional clues which point to a stone vault.

(6) The projecting slabs were used as scaffolding in the process of laying the upper courses of the stone vault.

Arguments Against a Stone Vault

(1) Although collapsed stonework is indeed found inside the tombs, not nearly enough of it has been found in such a position to complete a stone vault. Usually, enough is preserved to raise the level of the wall only by another 0.50 m. or so. In addition, the collapsed stonework is often of a slightly different type (triangular stones) suggesting a possibly different method of construction in the upper walls. (Incidentally, proponents of a full stone vault argue at this point that much of the original stonework has been quarried away by modern villagers living nearby.)

(2) The construction of the preserved walls is nowhere near strong enough to support a stone vault. The construction technique utilizing large amounts of clay packing and relatively small stones is simply too weak. While an earthen mound heaped over the tomb would probably serve to withstand the outward pressure of the vault, there is no evidence that such earthen mounds ever existed, and tholoi were never sunk into the earth like later Mycenaean tholoi (except for Ayios Kyrillos, which is founded only some two meters underground).

Possible Solutions

(1) Pini concludes that at least the smaller tholoi up to a diameter of ca. 7.50 m. were fully vaulted in stone. The most cogent objections to such vaulting really only apply to the largest of the tholoi.

(2) Hood has suggested that the top of the tomb was covered by a light, timber-reinforced mudbrick vault [but there is no evidence whatsoever of decomposed mudbrick in the tombs].

(3) Branigan suggested in 1970 that the top of the tomb was covered by a light, flat roof built of wood which could easily be removed when the tomb had to be fumigated. The projecting slabs would have provided access to the top of the stone portion of the tomb in order to facilitate the removal of the roof. The spans of the larger tombs would not be impossibly large for such flat timber roofs since the spans would have been reduced to between some four and seven meters by the corbelling of the tomb’s inner face.

(4) The discovery at Archanes in the 1970’s of a relatively small MM IA tholos tomb whose vault was very nearly fully preserved in stone seemed to substantiate Pini’s view, at least insofar as the smaller and later tholoi were concerned. But then it was claimed by Cavanagh and Laxton in 1981 that the structural mechanics of a corbelled vault require a minimum ratio of 0.53 between the wall thickness and the radius of a fully vaulted structure. Only one of the forty-four tholoi for which measurements were supplied in 1976 by Pelon satisfied this criterion, and so for a brief period there was a swing in the pendulum of at least some archaeological opinion to the view that no virtually no Early or Middle Minoan could have been fully vaulted in stone.

(5) In the revised edition (1988) of his now classic work on the Mesara tholoi, Branigan came to the conclusion that the recent evidence from Archanes and other sites indicates, on the contrary, that most, if perhaps not quite all, tholoi were, in fact, fully vaulted in stone, and this opinion has more recently been enthusiastically endorsed by Watrous (1994).

Method of Burial

Hundreds of inhumations were deposited in most tholoi with no regard being paid to previous interments. The insides of the tombs therefore often consist of a confused mass of bones and grave goods. Traces of fire in a number of tombs (but not in all) suggest periodic fumigations. In some cases, a layer of sterile white sand separates one burial stratum from another, and this occasional renewal of the floor would presumably be an alternative method of “cleaning up” the tomb. There is good evidence, in the form of nearby built ossuaries, that the contents of the tomb were periodically removed and redeposited in subsidiary structures. Perhaps the best evidence of this kind for a long-term sequence of usage in the communal burial structures from a single site, tholoi as well as rectangular ossuaries, comes from Archanes. The long history of these structures’ construction and use from EM II through MM II has been conveniently summarized by Watrous (1994).

Two or even three tholoi were often in contemporary use at the same site, a fact which suggested to Branigan in the early 1970’s that the tholoi might have “belonged” to individual clans within a village rather than always to the village as a whole or to even larger “tribal” groupings. More recently, fieldwork by Blackman and Branigan in the Ayiofarango valley of south Crete and Whitelaw’s analyses of earlier mortuary data from the Mesara indicate that each tholos may actually have been the tomb of an even smaller EM social unit, the nuclear family. Every EM settlement identified by intensive survey in the Ayiofarango has one or two associated tholoi, while some tholoi even seem to be paired with small outlying farmsteads. The placement of tholoi, in Branigan’s view, is governed largely by considerations of proximity to the settlement which it served: tholoi almost always lies within 250 meters of a contemporary settlement.

The very long periods of use for which the finds in many tholoi are evidence reveal that the social groups which such tholoi served must have been very small if they were intended to contain all the corpses generated by that group. In cases where two or more tholoi served a single settlement, the tombs are normally of about the same size and have provided roughly comparable ranges of grave goods, thus suggesting that Mesara society in the EM period, with the probable exception of Phaistos, may have been more egalitarian than the supposedly stratified social order suggested for such north Cretan settlements as Mochlos, Mallia, and Gournia (see above).

The artifacts from the tombs (vessels of both pottery and stone, tools and weapons, jewelry, and seals) show that a dead person was buried with his or her personal belongings as well as with food and drink (for the next life?). Since the dead were supplied with food and drink, they were presumably primary burials (i.e. fully articulated bodies) and not secondary collections of bones. Most of the grave goods show signs of use in mortal life (i.e. they were not designed specifically for funerary purposes). Two groups of finds, however, may well have been designed exclusively for funerary usage: stone vases, many of which are too small for any practical use and examples of which are rarely found in settlement contexts; and anthropomorphic and zoomorphic ceramic vessels which are found only very rarely in settlements (e.g. “The Goddess of Myrtos”). During EM I-II, the tombs appear to have been used for funerary rituals only. Such rituals appear to have required only a restricted range of grave goods and to have been attended by relatively small numbers of celebrants. A significant change takes place in the subsequent EM III – MM I periods, when tholoi began to be sited regularly above and overlooking the settlement which they served (e.g. Kamilari, Ayios Kyrillos, Gypsadhes, Vorou). At the same time, the range of goods placed in tombs was dramatically expanded, while the architecture of the tombs was enhanced with paved courts and terraces (Archanes, Apesokari, Moni Odegetria) where ceremonies attended by larger groups could have been held.

A “Cult of the Dead”?

(1) Apesokari (MM I): The rectangular complex in front of the tholos includes two constructions identified as altars, one inside the annex and one just outside. Interpretations of this arrangement have been that: (a) one altar was for religious authorities (inside), one for the common people; (b) one altar was for members of the immediate family or of the clan (inside), the other for the rest of the village; (c) one altar was used during the actual burial ceremony (inside), the other for a continuing “cult of the dead”. In fact, there is not much support for any of the above interpretations.

(2) Terracotta Model from Kamilari (MM III-LM I): [Hood, APG 105 Fig.88] Are the two smaller scale figures worshipping the four larger seated figures? Are the seated figures divinities or simply heroized dead? Or does the whole model, one of several multi-figured terracottas from a relatively late stage in the use of this tomb, have no connection with either death or cult?

(3) The discovery of masses of plain, ordinary drinking cups stacked in piles in some annexes suggests that some kind of funeral ceremony or cult ritual took place at some tholoi. Branigan has suggested that this ceremony was a farewell toast to the dead.

Theories on the Origin of the Tholos Tomb Form

External Origin

(1) (Hutchinson, Branigan): EM tholoi are somehow derived from mudbrick tholoi of the Syrian Halaf culture attested at the site of Arpachiyah [but the Halaf tholoi date to the late 5th millennium B.C. (i.e. are at least 500 years too early) and are evidently domestic rather than funerary architectural forms].

(2) (Evans, Xanthoudides, Pendlebury, Alexiou): EM tholoi are derived from circular tombs in Nubia or from Old Kingdom vaulted tombs in Egypt. This theory was very popular when many of the finds in the Mesara tholoi (figurines, amulets, stone bowls, scarabs) were considered to have been imported directly from Egypt or Libya. However, these objects, whether or not they came to Crete directly from Egypt rather than through the hands of Levantine intermediaries, are now considered to be later than EM I in date and so have become irrelevant in any discussion of the tholos’ origin. As far as pure architecture is concerned, Nubian tombs are flat-roofed and usually solid cairns, similar in their circular plans to the EM tholoi but radically different in their sections. In addition, they are of late 3rd millennium B. C. date and thus postdate most EM tholoi. Old Kingdom tombs in Egypt have barrel rather than corbelled vaults and thus also are not closely comparable to the EM tholoi.

(3) (Hutchinson): EM tholoi are to be derived eventually from Early Neolithic circular houses at Khirokitia on Cyprus by way of the small circular tombs of the Final Neolithic period at Kephala on Keos. The objections to this theory are: (a) the Cycladic tombs are minuscule when compared to the large EM tholoi at sites like Platanos, Koumasa, and Ayia Triadha; (b) EM tholoi cluster in the south of Crete, whereas Cycladic influence in the island, particularly in EM I, is restricted almost entirely to northern Crete; (c) there is no good reason to connect the EN houses from Cyprus with the FN tombs from Kephala.

Indigenous Origin

(1) (Branigan): There is no evidence in earlier or contemporary Cretan architecture for a circular plan. However, tholoi may be viewed as free-standing imitations of caves, the only sites used for burials in LN Crete. Caves are relatively rare in the Mesara area, and EM expansion into the Mesara plain may have necessitated the invention of the tholos form. A difficulty with this theory is that the tholos type diffuses over a wide area geographically and becomes remarkably standardized as an architectural type all within the EM I period. Moreover, most of the earlier tholoi were built not down in the plain of the Mesara, but rather in the foothills of the Asterousia Mountains, on both their northern and southern slopes. Since EM I is a long (ca. 600 years) period, some argue, there may have been sufficient time within it to encompass full development and standardization of the tholos form. In further support of such a view, some of the earliest tholoi show peculiarities that may be symptomatic of the type’s recent development (e.g. entrance at the south, two entrances at both south and east, construction incorporating rock outcrops, etc.).

Chronologically Significant Features

Most tholoi were built either in EM I or in MM I. Although there is no consistent development of the tholos form, there are nevertheless a few features which appear to be characteristic of the later tombs:

(1) Later tombs are generally smaller (only 6 of 28 early tombs have a diameter of below 5 m., while 8 of 14 late ones do).

(2) Later tombs feature built doorways, not trilithons (7 of 8 late tombs for which the relevant details have been published).

(3) Later tombs have higher doorways (3 of 4 late tombs whose doorway heights we know have doorways over 1.5 m. in height). Larger doorways were no doubt required due to the use of larnakes and pithoi as individual burial containers within tholos tombs beginning in the EM III period

(4) Later tombs feature a rectangular annex which incorporates the vestibule/antechamber to the tomb (5 examples; similar annexes at Platanos B and C appear to be later additions to pre-existing structures).

(5) Later tombs feature improvements in masonry, such as coursing and the employment of cut blocks.


A larnax in the EM period is elliptical in plan and relatively low in elevation. It has no legs and is never painted. In origin, it is probably a copy in clay of a simple wooden trough. Burial in larnakes begins in EM III. At Pachyammos and Gournia in east Crete, larnax burials consist simply of larnakes being placed in individual pits cut into sterile soil or bedrock. At other sites in north Crete, larnakes are deposited inside of built tombs of the “house” type. By the MM period, this custom has been extended to tholoi of the EM or Mesara type, especially at Archanes (notably in Tholoi Epsilon and Gamma). Larnax burial in the Early and Middle Bronze Age Aegean world is peculiar to Crete. Within Crete, it reflects a trend at the end of the EM period toward individualized burial.


This form of burial appears at the very end of the EM period at the same sites where larnax burial is more or less contemporarily introduced. It becomes far more popular in the MM period. Pithos burial is common in western Anatolia in the Early Bronze Age and also occurs in the Final Neolithic cemetery at Kephala on Keos and in the EH II “round graves” on Lefkas. It is, however, relatively rare during the Early Bronze Age on the Greek Mainland and in the Cyclades, while in the Middle Bronze Age in these areas it is a form used mostly for the burial of children and infants.


The following peculiarities of pre-palatial (i.e. EM I through MM IA) Minoan burial customs are worthy of particular note:

(1) Burial practices during this period are highly regionalized, with house tombs prevalent in the north, tholoi dominant in the south central region (i.e. the Mesara), cave burial favored in the far east and far west, and cist graves typical of some sites in the northeast.

(2) The native forms of burial that are particularly characteristic of the pre-palatial Bronze Age on Crete – that is, excluding cave burial as a holdover from the Neolithic and cist burial as a fashion imported from the Cyclades, and perhaps even by brought to Crete by immigrant islanders – are both collective and supraterranean (i.e. above-ground) ways of disposing of the dead. They stand in marked contrast to the preference of both islanders and mainlanders for burial of the dead below ground, either individually or else in small numbers in relatively small tombs. In their emphasis on togetherness of an entire community, Minoan burial customs closely parallel their preference for living virtually on top of one another, cheek-by-jowl, in their settlements.

(3) These collective, above ground burial structures of the Minoans had extraordinarily long lifetimes, sometimes in excess of 1000 years. Like their decision to live together for upwards of 1500 years at Knossos before they began to colonize the island of Crete in the fifth millennium B.C., this unusual conservatism speak in favor of a culture in which inherited custom played a far more significant role than elsewhere in the Bronze Age Aegean.

Comments are closed.