Settlements in the Early Minoan Period


Early Minoan Settlement at Trypiti / Wikimedia Commons

By Dr. Jeremy B. Rutter
Professor Emeritus of Classical Studies
Sherman Fairchild Professor Emeritus in the Humanities
Dartmouth College


Chronological Problems

The construction of a useful and valid relative chronological framework for Early Minoan (EM) Crete has been made difficult by a number of independent problems:

(1) Most of the EM material known to us until fairly recently was excavated between 1900 and 1920 when excavation techniques in the area were in their infancy and the standards for excavation recording were generally rather poor.

(2) A number of important sites excavated between 1930 and 1965 have never been published in any detail (tholos tombs at Lebena; village at Ellenes Amariou; EM I well at Knossos; etc.).

(3) At the major palace sites of Knossos and Phaistos, the EM levels were largely terraced away when the first palaces were built in Middle Minoan (MM) IB. There is every indication that these sites were of major importance already in the EM period, but the architecture of this period, as well as what must have been a valuable series of stratified deposits, have been lost forever.

(4) The practice of re-using EM tombs for large numbers of inhumation burials (circular tholos tombs in the Mesara Plain in the south; rectangular house tombs with multiple chambers in the northeast and east) has made it impossible to isolate significant numbers of individual EM burials with their associated grave goods. As a result, we have a mass of EM material from these tombs which can be relatively dated only in stylistic terms. We are therefore unable to trace chronological development in detail for numerous classes of objects which are found primarily in tombs: seals, stone vases, bronzes, figurines, etc.

(5) Evans’ highly artificial division of the period into EM I, II, and III, largely on the basis of the stratigraphy observed at Knossos, has caused a good deal of trouble because EM pottery in general, and that of the EM III phase in particular, varies considerably from region to region within Crete. Only recently has enough pottery of the EM period been studied and published in sufficient detail (see especially the work of Day, Whitelaw, and Wilson, building on the earlier work of Warren) to allow such ceramic localism and its development to begin to be assessed properly.

Early Minoan I (ca. 3100/3000-2700/2650 B.C.)

Knossos / Wikimedia Commons

The existence of such a period as a distinct phase was for a time denied by Levi on the basis of his excavations at Phaistos, but at the end of the 1950’s Hood found a well at Knossos which was filled with pottery of exclusively EM I types. Alexiou’s subsequent discovery of a mixed FN-EM I level stratified below an EM IIB level in Tholos Tomb 2 at Lebena, together with Warren’s and Tzedhakis’ documentation of a distinct EM I phase at Debla in west Crete, established by the mid-1970’s that a distinct EM I assemblage did indeed exist throughout the island.


EM pottery in general is distinguished from that of the preceding LN-FN periods by the introduction of the jug shape and by the presence of large amounts of painted ware. At least four major ceramic classes or wares of the EM I phase are worthy of mention:

Pyrgos Ware (Betancourt 1985: 26-29)

The favorite shape in this pattern-burnished ware, normally black, gray, or brown in color, is thechalice. Shapes and decoration suggest that this ware may be derived from prototypes in wood, the pattern-burnish being a conscious attempt to imitate the grain of the wooden original. This ware is found most commonly in northern Crete, occasionally in the south, but not in the east. It disappears rather quickly after EM I.

Incised Ware (Betancourt 1985: 32-33)

The favorite shapes in this ware, also dark-surfaced, are the bottle and the low pyxis. Found only in the north and northeast of Crete and either imported from the Cyclades or closely modelled after Cycladic prototypes, this ware is important for establishing the contemporaneity of the EM I period and the Kampos phase of the Grotta-Pelos culture of the Early Cycladic I period.

Ayios Onouphrios Ware (Betancourt 1985: 29-31, 40-43)

The paint of this dark-on-light pattern-painted ware varies from red through brown to black depending on firing conditions. The favorite shapes are jugs, two-handled cups, and bowls. Painted patterns are almost exclusively rectilinear (usually groups of diagonal parallel lines). Examples of the ware may be divided into two broadly defined groups or styles which appear to have some chronological significance. Thus Style I of the EM I period is characterized by round bottoms on the vases and by simple decoration which is designed to emphasize the shape of the vessel. Style II of EM II, on the other hand, features flat-bottomed or footed vases, an extensive use of cross-hatched patterns, and generally a less intimate connection between vessel shape and decoration. Ayios Onouphrios ware is at home primarily in the north and south central portions of the island; Style I does not occur in the east, while Style II is never common there. Local EM II developments of Ayios Onouphrios I include Koumasa ware in the south center and Myrtos ware in the southeast.

Lebena Ware (Betancourt 1985: 31-32)

This ware, characterized by patterns in white added over a solidly painted dark red ground, is apparently complementary in terms of its shape range to Ayios Onouphrios ware; there are no jugs, most of the shapes being low dishes, plates, and bowls. Vases often have round bottoms and the patterns are similar in design to those appearing on Ayios Onouphrios ware. Like the latter, Lebena ware is at home primarily in north and south central Crete and presumably originated in the Mesara plain. It probably survived into the earlier part of the EM II period but clearly lacked the popularity and enduring influence of the dark-on-light-painted Ayios Onouphrios class.


Phaistos / Wikimedia Commons

At the end of the Neolithic period, there were open settlements at Knossos and Phaistos, but most Late Neolithic sites are in caves (Trapeza, Miamou, etc.). Some caves continue to be occupied in EM I, but the number of open sites now increases. However, relatively little is known about EM I architecture. There are traces of rectangular buildings at Mochlos; at Debla in west Crete there are small free-standing buildings on a hilltop far up in the foothills of the White Mountains, perhaps seasonally occupied shepherds’ huts rather than a permanent settlement and hence atypical rather than characteristic; at Ellenes Amariou there are multi-room blocks as well as one free-standing building; and at Phaistos there is a rectangular room with a red-plastered floor. A well over ten meters deep of this period has been found at Knossos. There were probably fairly large villages during this period at both Knossos and Phaistos, but little of either has survived.


The lack of stone bowls from this phase at Lebena suggests that this industry had not yet begun on Crete; two Predynastic or Early Dynastic Egyptian stone bowls said to have been found in a LN building at Knossos are claimed by some to show that such objects were already being imported. No seals definitely datable to EM I have as yet been found, but they may well already have existed. Marble figurines of the so-called Ayios Onouphrios type may already have been made in EM I (for the type, see Pendlebury, The Archaeology of Crete Pl. XII:1, middle row; XII:2, right, top and bottom). A marble animal figurine of this date is known from Lebena.


No metal has yet been found in a purely EM I context, but the copper axe from LN Knossos shows that metallurgy was undoubtedly being practised. The discovery of copper daggers and awls at Pyrgos and Kanli Kastelli where no material of EM II date has been found may indicate that objects such as this were already being produced in EM I.

External Relations

Pottery in the form of Incised Ware shows contact with the Grotta-Pelos culture of the Early Cycladic I islands of the central Aegean. There may also have been some loose contact with western Anatolia and the Levant, but no external contacts are very strong.

Early Minoan II (ca. 2700/2650-2150 B.C.)

Ruins of Pyrgos on the island of Santorini / Wikimedia Commons


Ayios Onouphrios ware and probably also Lebena ware continue from EM I. Pyrgos ware disappears as does the coarse Incised ware with its Cycladic connections. The following new wares are significant:

Fine Gray Ware (Betancourt 1985: 40)

Very fine in texture, gray in color, and normally featuring a polished surface, this ware is typical of the earlier part of the period that is therefore designated EM IIA. The favorite shapes are spherical and cylindrical pyxides. Decoration, exclusively incised, typically takes the form of short diagonals, semicircles, rings, and dots. This ware occurs throughout the island, though never in very large quantities.

Vasiliki Ware (Betancourt 1985: 43-48)

Solidly painted but intentionally mottled and generally dark-surfaced, this ware occurs in small amounts in some EM IIA deposits in east Crete but becomes overwhelmingly dominant among the fine wares in the subsequent EM IIB phase throughout the eastern and southern portions of the island. The favorite shapes are flat-bottomed jugs, teapots, dishes, spouted bowls, and goblets. Jugs and teapots often have applied pellets (“eyes”) on either side of the spout. Always relatively rare in the north central and western regions of Crete, this ware declines steeply in popularity even in the east and south after EM IIB.

Aside from strictly utilitarian pottery, terracotta was also used occasionally as a medium for anthropomorphic and zoomorphic vessels which presumably functioned as ritual vessels both within settlements (e.g. “The Goddess of Myrtos”) and in burial ceremonies (most of them having been found in tombs).


Country house ruins at Myrtos on the island of Santorini / Wikimedia Commons

The three major excavated sites are Vasiliki on the west side of the isthmus of Ierapetra, Fournou Korifi (Myrtos) on the south coast ca. 12 kms. west of Ierapetra, and Trypiti on the south coast some 40 kms, further west of Myrtos. The first two sites peak in the EM IIB phase, near or at the end of which both are destroyed by fire; the last appears to reach its peak in the subsequent EM III phase, although it was already occupied as early as EM I and had become a substantial settlement by EM II. In the earlier EM IIA phase, there are traces of substantial buildings at Vasiliki and Palaikastro, so the large EM IIB complex at Vasiliki is not entirely without a predecessor.


Preserved and excavated are two wings, at the southeast and southwest of what was originally considered to be a single very large building with four such wings arranged around a large central courtyard. The surviving wings are each about one hundred feet long and consist of numerous rooms including both storage areas and what appear to be residential apartments connected by corridors. The southwest wing is preserved at ground level, while the southeast wing exists only in the form of a basement, due to the slope of the low hill on which the site is located. To the west is a paved court, over part of which the southwest wing was built. The construction technique may be described as “half-timbered rubble masonry”: the limestone fieldstones of the walls are packed in a mortar of mud and are supported by a timber frame consisting of squared timbers running horizontally and vertically just under the plastered wall faces and connecting rounded beams running transversely through the thicknesses of the walls. The upper walls were built entirely of mudbrick, while wooden beams overlain by reeds, cane, and mud were used for ceilings. Some walls were covered on the interior with red-painted lime plaster. The original American excavator of the site, Seager, assigned virtually all the EM remains at the site to a single building, huge for its time, which he christened the “House on the Hilltop” and restored along the lines of the later palaces of the Middle and Late Minoan periods. Reinvestigation of the complex in the late 1960’s and early 1970’s by Zoïs led to the realization that the two surviving wings are in fact two separate buildings datable to slightly different periods within the EM IIB phase. There are therefore no grounds for considering the composite “House on the Hilltop” to be an EM II predecessor of the Minoan palace plan complete with large central court, paved west court, and so forth. However, the now separate structures, renamed the “Red House” (old southeast wing) and the “West House” (old southwest wing), are still evidence for complex and relatively sophisticated architecture quite different in terms of both plan and building technique from the “House of the Tiles” and other monumental buildings of the contemporary Greek Mainland. Continued excavation at Vasiliki through the early 1990’s strongly suggests that the EM II settlement here never consisted of more than about four large houses extending over an area of roughly 80 x 40 m. or 0.32 {hectare} [1 hectare = 10,000 sqaure meters]. A carbonized olive pit from a sealed EM IIB deposit is a welcome supplement to the less securely dated olive pit recovered at Myrtos, but neither the size of these olives nor, clearly, their number provide much evidence for the degree to which this fruit was exploited, much less whether it was “wild” or “domesticated”. A two-part mould from Vasiliki attests to EM II metalworking on the site.


Fully cleared in only two short seasons (1967-68) and fully published just four years later, this tiny (30 x 50 m.) but enormously important site has two main phases of occupation, both datable within the EM II period. Minor differences in the pottery between the two phases permit the distinction of EM IIA (Period I) and EM IIB (Period II), the first characterized chiefly by the presence in small quantities of Fine Gray ware and by a wide range of dark-on-light pattern-painted pottery, the second by a restricted range of dark-on-light pattern-painted pottery and by the dominance of Vasiliki and other solidly painted wares. The site was completely destroyed by a violent fire which has been dated by radiocarbon to between 2292 and 1770 b.c. (= 2960-2150 B.C.). Thermoluminescence dating of some of the pottery ranges from 2580-2170 B.C., with an average of 2373 B.C. The archaeological date originally proposed by the excavator for EM II was ca. 2600-2170 B.C., within which he placed the date of Myrtos’ destruction at ca. 2200 B.C. (For comparative purposes, the House of the Tiles at Lerna was destroyed ca. 2300 B.C.). More recently, the excavator has reported the Myrtos destruction dates to be 2850-2305 B.C.

One possible reason for the foundation of the site in EM IIA is a population rise in central Crete which caused expansion to the east and the foundation of numerous settlements there (including Vasiliki). There is no apparent gap in the occupation of Myrtos between Period I of EM IIA and Period II of EM IIB.

Relatively little architecture of Period I is preserved, the most interesting by far consisting of a small potter’s workshop (Rooms 47-51, especially 49 and 51). The complete plan of the settlement of Period II was excavated as preserved. It consists of a conglomeration of ca. 65 rooms, corridors, and open areas all packed into an area of 0.15 hectare. This complex is bounded on the exterior to the south and west by a wall of ordinary width, hardly to be considered a “fortification” but nevertheless pierced by just two openings: a main entrance at the south approached by a raised walkway from the outside, and a second entrance at the west. The east side of the site lies along the edge of what amounts to a cliff, while the northern boundary is unique in being relatively poorly defined. Major passageways lead through the settlement from north to south and from east to west, in each case beginning at one of the main entrances. Walls are built of mud-packed rubble but lack the half-timbered framework of contemporary Vasiliki. Walls and roofs were extensively plastered, walls sometimes being painted red. Roof construction was simple and flat: olive branches overlain by reeds and lime plaster. The absence of roof-spans greater than 2.5 m. implies that there were no large trees in the area which could have supplied large roofing beams. Floors consist simply of bedrock covered with a variable thickness of white clay to make the floor surface approximately level. There is no evidence anywhere in the settlement for a second storey, but the flat roofs of many rooms were probably used as outdoor work and living spaces and several roofed areas appear to have been accessible only from above (i.e. through a trap-door in the roof).

Warren initially interpreted the architectural complex of Period II at Myrtos as a unity, presumably administered by a central authority and featuring a number of functionally specialized workrooms, storage rooms, and other activity areas such as cult rooms which served an entire community estimated at ca. 100-120 individuals. In other words, Warren saw the complex as a miniature version in some important respects of the later Minoan palaces and hence possibly ancestral to them. Even before the site’s full publication, Branigan had outlined a somewhat different interpretation according to which the entire complex was the residence of a local ruler in which he, like Warren, was prepared to see a direct antecedent of the later palaces. In a fundamental re-interpretation of the evidence a decade later, Whitelaw identified the complex as an agglomeration of five or six independent units, each of essentially the same size and complexity, which he suggested were individual family dwelling units. The growth of the settlement in stages from the original dwelling unit to the eventual group of five or six can be traced by a painstaking analysis of wall joints and abutments. This reconstruction of the settlement’s history and basic organization is also supported by a careful analysis of the movable finds and their contexts: such functionally specialized spaces as occur within the settlement are multiple rather than single (e.g. kitchens; storerooms; workspaces housing looms, washtubs, etc.) and are in most cases recurring features of the individual dwelling units (two rooms fronting on public courtyards which may have been designed to accommodate public rituals would be exceptions). There is no evidence in either the architecture of, or the finds from, the individual units for the existence of a social hierarchy within the settlement. According to Whitelaw, the whole settlement in its final stage is likely to have housed a small kinship group of some fifty individuals descended in part from the builders of the first dwelling unit of Period II. Rather than being the residence of a rather humble central authority of some kind, the complex is instead good evidence for the basically egalitarian nature of EM society, the basic social unit of which may have been a group closely comparable to the contemporary nuclear family. Thus Myrtos appears quite closely comparable in terms of size, population, and length of occupation (three to four generations) to the typical Cycladic settlement of the Early Bronze Age (see preceding handout).


So far published only in brief preliminary reports, this spectacularly photogenic but small cluster of rectilinear rooms closely resembles Myrtos in its location on a fairly steep hill within a couple of hundred meters of Crete’s south shore. As at Myrtos, two principal building phases have been identified and the settlement as a whole amounts to no more than six or so individual buildings crowded very tightly together.

Architecture Material


The long-lived Minoan stone vase industry begins during the EM II phase. Numerous examples were found in the tombs at Mochlos in the northeast and also in the tholoi of the Mesara plain in the south center. Stone vase fragments from the settlement at Myrtos were probably imported from the north coast, perhaps from Mochlos. Stone vases made of chlorite schist and bearing incised decoration (e.g. Hood, APG Figs. 130-131) belong specifically to EM II, but a large number of other materials were in contemporary use and continued to be exploited in subsequent periods (e.g. marble, serpentine, tufa).

Sealstones are also now definitely being produced: four finished and three unfinished seals, as well as one sealing, were found in the EM IIB levels of Period II at Myrtos, a discovery which also indicates that seals were both being produced on that site and used to create impressions on lumps of moist clay. Another sealing has been found at Trypiti. All stamp seals with simple rectilinear designs (usually nothing more complicated than cross-hatching), the Myrtos seals are certainly not up to the standard of the EH II seals of Lerna III from a purely aesthetic point of view.

Marble figurines become more common. The Ayios Onouphrios type continues to be produced, folded-arm figurines (FAF’s) of Cycladic types are imported from the central Aegean, and eventually a specifically Cretan imitation of the Cycladic FAF known as the Koumasa type is manufactured locally in Crete.


Objects of metal now occur in quantity for the first time. Numerous daggers of copper and bronze are found in tombs, as well as a good deal of jewelry (especially at Mochlos). A typical dagger was found in the EM IIA strata of Period I at Myrtos. Metal tools (i.e. saws, chisels, axes), however, continue to be rare.

Internal and External Relations

Minoan letter seals / British Museum

Both Vasiliki and Myrtos were completely destroyed by fire near or at the end of the EM IIB period. Myrtos was never reoccupied; Vasiliki continued to be a village for centuries but never again produced complexes as impressive as the Red and West Houses. There is no question of an invasion of Crete by a foreign population, since the cultures of the EM III and MM I periods develop smoothly from that of EM II, but some sort of centralization of political power may have been taking place within Crete during the last century or two of the Early Bronze Age. Much more evidence is, however, required to prove the existence of such a process of centralization and, if it is real, to chart its path.

Extensive contact with the Cyclades is documented by the marble figurines of folded-arm type, by incised stone vases produced in chlorite schist, by the overall similarity between the two areas in dagger and tweezer types of copper, and by imported Cycladic pottery in both settlement and funerary contexts. Contacts between Crete and the Greek Mainland are established by the discovery of EH II Urfirnis sauceboat fragments in an EM IIA context at Knossos and in mixed EM-MM deposits at Platyvola in west Crete, as well as by the presence of a few probably EM seals and amulets in EH contexts and an occasional EM vase in a Cycladic context. Minoan activity at the site of Kastri on Kythera is considered by some to mark the establishment of a Minoan colony at that site as early as the EM II period, by far the earliest Minoan settlement outside of Crete itself, but the earliest colony there is probably no earlier than MM I, earlier Minoan remains there simply being debris left behind by west Cretan fishermen who made seasonal trips to the area. Contact of some sort with the Levant is attested by the ivory used to make Minoan seals. Stone bowls of Egyptian manufacture may show contact either with Egypt (direct) or with the Levant (indirect).

Early Minoan III (ca. 2150-2050/2000 B.C.)

Andreou’s Kouloures archaeology team / Wikimedia Commons

Problems of Definition

Evans defined this period ceramically by the appearance of a white-on-dark pattern-painted pottery decorated in a style altogether different from the earlier Lebena ware of EM I-II. This definition relied, however, primarily on the ceramic sequence typical of sites in eastern Crete (e.g. Gournia, Vasiliki) and not on the sequence at the site of Knossos which Evans himself was excavating. For many years, Evans’ definition of EM III caused considerable confusion for excavators working at sites outside of eastern Crete and was a particular problem for archaeologists interested in the period immediately preceding the construction of the first palaces at Knossos, Phaistos, and perhaps Mallia in MM I. Andreou’s dissertation of 1978 resolved much of this confusion by publishing in some detail a large deposit from Knossos (the Upper East Well group) which certainly postdates EM II but which equally clearly precedes Knossian (or north central Cretan) MM IA, a period itself represented by Andreou’s Kouloures group. The result of Andreou’s work is that EM III is now definable as a distinct period in north central as well as in eastern Crete. It is, however, a brief period in comparison either to the EM II phase which precedes (500 years long) or to the MM I phase which follows (250-300 years long).

Pottery (Betancourt 1985: 53-63)

In eastern Crete, EM III pottery is characterized by a white-on-dark pattern-painted style which includes both rectilinear and curvilinear (e.g. circles, spirals) ornament. Such pottery seems to develop directly from a light-on-dark pattern-painted class present in small quantities in the EM IIB period at sites such as Myrtos. In north central Crete, there is relatively little light-on-dark-painted pottery in this phase and none whatsoever decorated with spirals. In fact, dark-on-light-painted pottery is more popular in the Knossos area, though patterns are rare and most of the painted decoration is purely linear (i.e. banding). Particularly characteristic of the north central region is the footed or flat-based goblet or “egg-cup” (Hood, The Minoans 38 Fig.14), a shape which occurs in the eastern part of the island only in the form of imports. What is happening in the south at this time (i.e. in the Mesara plain) is not yet clear. Neither barbotine nor polychrome painted decoration make their appearance anywhere until MM IA.


Assignable to this period at Knossos is the so-called “hypogeum” (possibly an underground, corbel-vaulted granary) at the south end of the later palace, as well as one large wall, presumably from a monumental building of the period, located in the northern part of the later palace’s west court. Hood has suggested that the latter wall, built entirely of small stones rather than of the large squared blocks which are typical of the MM IB palace, may be part of an EM III palace at Knossos. In view of the extensive terracing and consequent demolition of most EM structures in the central part of the palace hill at Knossos, this suggestion will probably never be capable of either substantiation or refutation.