free trial issue                                             subscribe                                                  back issues

Athena Review, Vol.3, no.3: Minoan Crete

The Minoan Palace at Petras, Siteia

Metaxia Tsipopoulou

Archaeological Museum, Agios Nikolaos

History of Research at Petras: Fieldwork conducted in the region by two of the greater names in Minoan archaeology, Nikolaos Platon and Costis Davaras, raised the probability that an important Minoan settlement was to be sought in the area of the modern town of Siteia. Chief indicators were the two religious sanctuaries overlooking the Siteia plain, both active in the Protopalatial and Neopalatial periods (19th - 15th cent. BC). It is known from elsewhere in Crete that during the Bronze Age such sanctuaries were connected to major administrative centers, frequently palaces. The Piskokephalo sanctuary was considered exceptional by virtue of its low altitude, since it failed to answer to the then prevalent definition of a peak sanctuary. Yet its finds place it among the most significant of such sites: several terracotta models of shrines capped with horns of consecration, a surprising number of large figurines, both male and female, as well as the usual multitude of small figurines depicting human figures and animals.

[Fig.1: View of Siteia Bay from the peak sanctuary of Prinias (photo: M. Tsipopoulou).]

During the 1950s, a series of Neopalatial (17th-15th cent. BC) finds were made by chance, as new roads were opened up in this isolated part of the island. At Achladia, Zou, Prophitis Ilias, and Klimataria, Platon excavated examples of a type of structure that were known in the specialized literature as "villas," "country houses," "mansions," or "houses of local chieftains." They were understood as local agricultural centers, consisting of a mix of refined architecture, such as representative areas and domestic shrines, and functional features, including storage and industrial production areas. Already in the 1920s and 30s, larger examples had been excavated in central Crete at Vathypetro, Sklavokampos, Amnisos, and Nirou. Clearly, the Siteia area was organized in a manner similar to that of other economic and political centers in Neopalatial Minoan Crete, with the notable exception that no palace had as yet come to light. Despite the clues offered by the sanctuaries and the "villas," none of the archaeologists then active in eastern Crete deemed it worthwhile to search for such an architectural unit. The Siteia Bay area was too small, it was thought, to have required such a center, and the previously known Minoan palaces at Malia, Gournia, and Zakros (fig.2)  rendered its existence even more unlikely. Zakros formed, it was argued, the palatial focus of eastern Crete, particularly given the absence of a palace at Palaikastro, a site which offered far better arguments than the Siteia area. The excavator of Zakros, Platon, reconstructed the economic and political organization of eastern Crete as centered on his site, to which the "villas" were tied. These housed the local chieftains who answered to Zakros, an arrangement rendered necessary by the distance and the harsh topography, and the attendant difficulties in communication.

In this intellectual climate, the coastal area of Siteia was considered unpromising, a mere vista on the road to the great sites further east. It was forgotten that already in 1900, R.C. Bosanquet had spent three days sinking a number of trenches on a low hill opposite the modern town of Siteia known as "Petras." A short report witnesses to interesting finds, leading the excavator to conclude that here lay the gateway to the Eastern Mediterranean, the stepping-off point for Minoan ships trading with the East. Yet Bosanquet abandoned the site, believing it to be too poorly preserved to warrant further attention. He continued on his road eastwards, destined to discover and excavate Palaikastro, which was to become the crowning achievement of his career. When, in the early 60s, the palace of Zakros was discovered, and the spectacular finds argued heavily in favor of seeing this site as the gateway, the fate of Petras appeared sealed.

[Fig.2: Map of eastern Crete showing Neopalatial polities and palace sites (after M. Tsipopoulou).]

The hill, however, was covered with walls piercing the surface of the soil, and with a dense coverage of sherds and other artifacts. After a visit in 1983, the present author decided to undertake a trial excavation in 1985. The topography of the hill, having a large flat plateau on the northern edge of a long ridge, and a series of terraces on the lower hill as it slides down onto the valley floor, suggested that a major structure was to be sought on the plateau. The trial trenches were placed further down, on one of the terraces, so as to allow the archaeologists to uncover the stratigraphical sequence of the site before turning their attention to the main feature. A survey in the second season indicated that the Petras site consisted of four separate concentrations of finds: the main settlement on Hill I; a secondary, mainly later occupation on Hill II; scattered finds on the ridge south of the site, Hill III; and a series of later Minoan chambertombs cut into the side of Hill IV. After 17 years of excavation and study, it is evident that the administrative center for the area is not to be sought further east, but that it lies on Hill I, at the palace of Petras.

[Fig.3: Plan of the Palace at Petras (M.Wedde, M.Tsipopoulou, and M. Klontza).]

The Palace Plan: The palace of Petras (fig.3) is, in fact, two palatial buildings, the one superceding the other on the same emplacement. Continuity throughout is attested by the constant north-south orientation of the Central Court (fig.3), around which, as is customary with Minoan palaces, the entire structure was organized from its Middle Minoan IIA (early 19th cent. BC) inception until the final destruction in Late Minoan IB (1480-1425 BC). The building suffered two significant intermediary destructions, one at the end of the Protopalatial period (Middle Minoan IIB, early 17th cent. BC), the other possibly connected with the Santorini eruption towards the end of Late Minoan IA (ca. mid-16th cent. BC). Each reconstruction of the west wing re-employed the system of six west-east axis walls that characterized the first construction. The significant feature of the Petras palatial plan is that the Central Court is not surrounded by a separate wing on each side; the sole wing that can be followed throughout the life of the building is the west wing. On the north side, a large storage facility was constructed in Late Minoan IA (1600-1500 BC), when the bedrock was cut back to create the necessary flat surface. What existed here before remains unknown; the only indications available today concern the Prepalatial period (2900-2000/1900 BC). To the east of the Central Court there was a narrow stoa-like construction facing a large open space limited to the east by the cyclopean-like wall that delimits the plateau. The south side remains an enigma. Remains from the Protopalatial period in the southeast corner could suggest that a south wing was part of the original plan, but to date nothing has come to light.

The earlier Protopalatial court measures ca. 6.5 H 18 m and was originally enclosed by substantial walls on all four sides. The court was covered with plaster of fine quality. A substantial drainage system removed the rain water from the roof, one branch running along the western edge of the court, a second running eastward as a channel cut into the bedrock. At the western half of the north wall of the court, a monumental staircase (fig.5) with fine plaster surface functioned as the main entry into the building complex. In the Late Minoan IB phase (15th cent. BC), the court was reconstructed in reduced dimensions, about 4.5 H 10 m. Its eastern side was now open, with alternating columns and pillars, to create a stoa. The monumental staircase had been badly damaged in the LMIA destruction (16th cent. BC) and was covered, thus creating an open space. To its immediate north, a square room was added to the original plan and used as a magazine. Access to the LMIB palace was from the southeast corner, where a much more modest entrance was appended, comprising a staircase leading to the upper floor.

[Fig.4: View of the earlier (Protopalatial-Late Minoan IA) and the later (Late Minoan IB) central courts from the north-northeast (photo: M. Tsipopoulou).]

In the Protopalatial and the Late Minoan IA periods, the North Façade stretched 10 m westward. In the first period, one of the entrances to the palace existed here. Into this doorway there fell, when the Protopalatial palace was destroyed (in MMIIB, early 17th cent. BC), an archive of clay documents inscribed with signs of the Hieroglyphic script (fig.6). In the Neopalatial reconstruction, the northwest corner of the building was extended beyond this point, a decision that assured the conservation of the archive, since the floor upon which it had come to rest was not reused. A curious feature of the building in this area is a series of parallel walls with narrow spaces between them, serving no other purpose than that of support for the floor. They do not, however, fall outside the general pattern of construction of the Petras west wing, for this building was raised on a system of six west-east axes on deep foundations, between which shallower north-south walls divided off the resulting space into rooms as required.

The west wing is divided in two by a corridor that runs to the west façade. A very interesting area of the Protopalatial building is the complex of rooms along the south edge of the preserved part of the west wing. The main room is equipped with a plastered bench four meters long, a flagstone floor, and orthostates along the walls. Based on the architecture, it is assumed to be the sacred space of the first palace. It was discovered empty because, after the MMIIB destruction (early 17th cent. BC), it was walled off and sealed. Only the westernmost corner continued in use until the very final phase of the palace, and was probably used as a shrine, as suggested by the find of a plaster offering table. Access to the putative shrine was through a rectangular room with a flagstone floor. This room continued in use until the end of the palace. In the fiery Late Minoan IB destruction (15th cent. BC), the walls were rubified in the intense fire, and from the upper floor there fell numerous ashlar architectural fragments, some with mason's marks.

In the area to the southeast of the palace, massive Protopalatial walls (19th-18th cent. BC) came to light. They appear to be part of a unit comprising the large rectangular building (4 H 15 m) to the south, as well as the eastern terrace wall, and the bastion, a roughly rectangular projecting structure, preserved to a height of 2.5 m. It forms part of the extensive defensive architecture along the eastern flank, probably to be associated with the remains of a massive fortification wall, preserved to a height of three meters, lower down the side of the hill, just below the modern road leading to the site. The Protopalatial period in eastern Crete saw substantial political tension, leading to the erection of defensive structures at a number of sites, including Petras.

[Fig.5: Monumental staircase in the Central Court of the Palace at Petras (photo: M. Tsipopoulou).]

The northern sector of the plateau is covered by the North Magazines. In the Late Minoan IA period, the bedrock was cut back on line with the North Façade to seat a substantial wall forming the southern limit of the magazines. Their manner of construction was sufficient to support a substantial weight, indicating that these magazines, some three meters lower than the rest of the palace, carried at least one further story, if not two. A staircase in the westernmost magazine supports this interpretation. In Late Minoan IB, an additional magazine was added in front of the monumental staircase, no longer in use. The excavation uncovered remains of some 30 pithoi (large pottery storage vessels) on the ground floor, some with Linear A inscriptions, while there was enough space for around 60 pithoi in the magazines. Of particular importance was the discovery of a discoid label inscribed on both sides in the Hieroglyphic script. The label was found with one of the pithoi in the corridor, and was probably initially attached to a handle or to the lid. This is the first evidence for the contemporaneous use of the two Minoan scripts in Late Minoan I. In the area covered by the palace, no industrial installations were uncovered. Since workshops form an important feature of Minoan palaces, the search for them at Petras extended the excavation to the southwest, onto a small plateau overlooking the main building. Here, a series of rooms was excavated, containing half-finished stone vases and various raw materials.

Although the palace of Petras is a small structure when compared to the better-known buildings, a checklist of prerequisite features associated with the building type sees every item except one present on the site, even if only, as the case may be, in a modest manner. Fine ashlar masonry was incorporated into the building at several points on the plan, and many blocks were marked with mason's marks. The LMIB southeast entrance features a pier-and-door partition. In addition, carefully cut jamb bases, column bases, stairways, dadoes, and slab pavements, as well as dove-tail clamp cuttings, and - not always a prerequisite in Minoan architecture - right-angle corners, illustrate the care that went into the construction. Plaster floors are a recurrent phenomenon throughout the several phases of the complex, some of them being of a very high quality, comparable with the finest examples at the large palaces. As a central administrative center, the site saw a continuous traffic of subjects on official business; it was part of the Minoan way of life to offer a drink (probably an alcoholic beverage) to the visitor in a ritualized reception ceremony. The most common container was the conical cup, ubiquitous at Minoan palatial and "villa" sites. Petras has produced thousands of these cups, the majority from three major destruction deposits connected with entrances. Wall paintings have also been preserved in representative areas, although no figurative scenes have been found. The administrative function was filled by substantial storage spaces associated with the central building; the pithoi at the site are of the large size found only in palaces.

Writing at Petras: The bureaucratic role of the palace is also indicated by the presence of writing: scattered documents - including three Linear A tablets - have come to light in several trenches; inscriptions, painted and incised, appear on pithoi and small vases; and the Protopalatial northwest entrance produced the Hieroglyphic deposit. This archive, fallen from the upper floor, is probably the most well-preserved hieroglyphic deposit of the four known on Crete. It came from a closed, securely datable context, and contained inscriptions (fig.6), and a large number of sealings from 42 different seals, including gold rings.

[Fig.6: Clay bar with hieroglyphic inscriptions, from the Palace at Petras (Photo: Erik Hallager).]

Some of the material suggests that the archive was working at the time of the catastrophe. This observation, unique for a Cretan archive, is of particular significance for the reconstruction of the historical situation at the moment of the destruction. It may also contribute to understanding the nature of the catastrophe itself. The forced departure, coupled with the absence of signs of violence, and the complete clearing out of valuables, argue for a natural cause. An act of war is unlikely, given the undoubted continuity of the time period leading into Neopalatial Petras. The houses in the settlement were rebuilt, for all intents and purposes, immediately, and retained the orientation of their predecessors. The palace was repaired, including the major elements such as the Central Court, the monumental staircase, and the North Façade, which continued in use into the first Neopalatial phase.

Dating phases at Petras: Most walls, floors, and alterations in the palace at Petras are dated by the scraps of life - the shattered and scattered remains of clay vases that, in the form of perhaps only one sherd from any one vessel, form the main class of data recovered by the archaeologists. A vast amount of sherd material corresponds with the moment that the previous architecture on the plateau was torn down to make room for the palace in its earliest form. In Middle Minoan IIA the entire plateau was transformed, the old buildings (of which very little is known) torn down, and the surface leveled by the removal of huge amounts of fill. This fill was carried down the hill and deposited in a depression in an area in which the township was later to flourish. It consisted of high quality sherds from vessels today identified as mainly belonging to two ceramic styles, Light-on-Dark and Kamares.

On a number of occasions, however, the dating of a level is provided by a destruction horizon (the sudden cessation of all activity, a transformation brought about in a non-orderly fashion). Four such horizons have been identified at Petras on the basis of the pottery associated with them. The first is dated to the Middle Minoan IIB period, and brings the Protopalatial palace to its end. A hieroglyphic archive associated with a number of vases fell down from an upper floor into a doorway, where it remained undisturbed (fig.7). Although small in size, this deposit proves that Petras functioned as a palace in the Protopalatial period.

[Fig.7: Middle Minoan IIB pottery found with the hieroglyphic archive at Petras Palace (photo: M. Tsipopoulou).]

Two destructions are dated to the Late Minoan I phase, the first in the IA subphase, the second in IB. In both periods the palace was destroyed. The main hint of these destructions is a large deposit of drinking cups, in an entire pantry that crashed down from the upper floor, thus creating a thick deposit consisting of pottery, bones, and other objects. The LMIA destruction is not well-known from other deposits; on the one hand, it would appear that the inhabitants had time to flee with their valuables, while on the other, the immediate reconstruction would have removed much of the debris. The IB destruction is spectacular: the smashed pantry from the upper floor; fallen ashlar blocks, rubefied by the intense heat; an oil storage room going up in flames and leaving a thick, black deposit; pithoi in the magazines and in the Central Court destroyed. The inhabitants appear to have fled - none were found in the ruins, and the precious objects had all been removed. The fourth and final destruction horizon is that of the Late Minoan IIIB phase. After a hiatus of perhaps a generation, the plateau was inhabited again, no longer as a palace, but a modest village. Two or more houses were built on top of the ruins. When these houses were destroyed, no spectacular deposits resulted, and a silence of 26 centuries spread over Petras, broken only once, when a Byzantine cemetery was established on the plateau in the late 12th c. AD.

Conclusions: The palace of Petras occupies the center of a well-defined economic zone which comprises its site catchment. On the slopes of Hill I around the palace, a large urban settlement spread out. Three building complexes have been excavated to date, producing large two-story buildings with architectural refinements housing various economic activities, including wine production, wool dyeing, weaving, stone vase manufacture, etc. The hill was inhabited from the Early Minoan period onwards, always a prosperous settlement with external relations. The economic and political functions that led to the status of palace grew gradually out of the existing framework. In Middle Minoan IIA, the existing architecture no longer sufficed and a major transformation of the structures on the plateau of Hill I was undertaken, resulting in the construction of the palace.

As the central place, Petras administered to the entire Siteia basin (fig.2). The "villas" of Achladia, Zou, Prophitis Ilias, and Klimataria are to be understood in this framework as answering to the palace in a tripartite hierarchy, with each "villa" forming a local center for a number of farmsteads. In the Minoan period, the sea extended far into the present day plain, the coastline being at the base of the foothills. At the deepest reaches of the bay lay the mouth of the Pandelis river. Consequently, Hill I lay on the waterfront, with a protected anchorage in a small bay between Hills I and II, and while Petras was not the large harbor facing the east that Bosanquet had sought, and which Platon found at Zakros, it served as an important way station on the route along the northern coast.

[Note: This is an abridged version of the article "The Minoan Palace at Petras, Siteia" by Metaxia Tsipopoulou, whose full text and illustrations appear in the printed issue of Vol.3, no.3 of Athena Review (pp.44-51). Copyright 2003, Athena Publications, Inc.]


Bosanquet, R.C. 1901. "Excavations at Petras." BSA VIII, 282-285.

Macdonald, C. and J. Driessen. 1997. The Troubled Island, Minoan Crete before and after the Santorini Eruption. Texas, Liege, pp. 226-227.

Platon, L. 1997. "The Minoan 'Villa' in Eastern Crete: Riza-Akhladia and Prophetes Elias, Praisos: Two different Specimens of One Category," in R. Hägg (ed.). Acts of the International Symposium, The Function of the Minoan "Villa." Athens, pp.187-202.

Rupp, D. and M. Tsipopoulou. 1999. "Conical Cup Concentrations at Neopalatial Petras: A Case for a Ritualized Reception Ceremony with Token Hospitality." In MELETIMATA, Studies in Aegean Archaeology presented to Malcolm H. Wiener as he enters his 65th year. Liege, pp.729-739.

Tsipopoulou, M. 1990. "Potters' Marks from Petras, Siteia." Kadmos XXIX, 2, 92-106.

1994. "Evidence for a New Minoan Palatial Building at Petras, Siteia." Abstract in AJA 100, 1996, 387.

1995. "Late Minoan III Reoccupation in the Area of the Palatial Building at Petras, Siteia." Acts of the Symposium Organized by the Danish Institute in Athens. August 1994: Late Minoan III Pottery. E. Hallager and B.P. Hallager (eds.). Athens, pp. 209-257.

1997. "Palace-Centered Polities in Eastern Crete: Neopalatial Petras and its Neighbors." In W. Aufrecht (ed.) Aspects of Urbanism in Antiquity, from Mesopotamia to Crete. Sheffield, pp.263-277.

1998. "A New Hieroglyphic Archive at Petras, Siteia." Abstract in AJA 103, 390.

1999a. "From Local Centre to Palace: The Role of Fortifications in the Economic Transformation of the Siteia Bay Area, East Crete." In Polemos, Le contexte guerrier en Egee a l'Age du Bronze. Liege, pp.179-191.

1999b. "Before, During, After: Stratigraphical Trenches at the Palatial Building of Petras, Siteia." In MELETIMATA, op.cit., pp .847-856.

Tsipopoulou, M. and E. Hallager. 1996. "Hieroglyphic and Linear A Inscriptions from Petras, Siteia." SMEA XXXVII, pp.7-46.

1997. "A New Hieroglyphic Archive at Petras, Siteia." Kadmos, pp.165-167.

Tsipopoulou, M. and A. Papacostopoulou. 1996. "Villas and Villages in the Hinterland of Petras, Siteia." In R. Hägg (ed.) Acts of the International Symposium, The Function of the Minoan "Villa." Athens, pp. 203-224.

Tsipopoulou, M. and L. Vagnetti. 1995. Achladia. Scavi e ricerche della Missione Greco-Italiana in Creta Orientale (1991-1993). Roma.

1998. "Workshop Attributions for some Late Minoan III East Cretan Larnakes." In R. Laffineur and P.P. Betancourt (eds,) TEXNH, Craftsmen, Craftswomen and Craftmanship, Acts of the International Symposium, Philadelphia, April 1996, Aegaeum 16, pp. 473-479.

Athena Review Image Archive™   |  Paleoanthropology in the News   |  Guide to Archaeology on the Internet   |   Free issue   |   Back issues

Main index of Athena Review   |   Subject Index   |   Travel Pages   |   Galleries and Museums   |  Ad rates   |  Current issue index |

Copyright  ©  1996-2004  Athena Publications, Inc.  (All Rights Reserved).