ARCHAEOLOGICAL SURVEY IN SPHAKIA, CRETE
L. Nixon, J. Moody, S. Price, and O. Rackham, Echos du Monde Classique/Classical Views 32 n.s.8 (1989) 201-15
We are most grateful to the Editor of Echos du Monde Classique/Classical Views for permission to reproduce this article here
The second season of the Sphakia Survey was conducted in the summer of 1988. Most of our time was spent on intensive survey of the Anopolis plain, with some extensive work in other areas.1
The general objectives of the Survey are to investigate the environment and vegetation of Sphakia; to clarify the nature and chronology of human exploitation of the area; to determine when and how Sphakia was linked with the outside world; and to study the stronghold effect in Sphakia, which has resulted in marked preferences for inland rather than coastal settlement locations in some periods.
The pilot season of 1987 enabled us to work out a sampling programme suitable for a research area of such generous size (the eparchy of Sphakia is circa 470 km2), based on eight environmental zones.2 Our specific objectives in 1988 were to investigate the mountain plain of Anopolis in some detail and to continue the sampling of other environmental zones, i.e. the foothills of the Araden area (up to 800m); middle slopes (800-1200m) at Mouri; two more gorges, the Sphakiano and the Trypiti, near whose mouth lies the site of Poikilasion ; and the coastal plain at Frangokastello, where we could work with geologists studying the sedimentation history of the area. We also wanted to visit sites in the adjacent eparchy of Selino, to serve as comparison (Elyros, Hyrtakina, Lissos and Syia). We were able to achieve all these objectives (and to clarify details of the plan of Tarrha, at the foot of the Samaria Gorge); and to begin work on an instructional video about the Sphakia Survey.
FIELD METHODS AND RECORDING
Nearly all fieldwork on the Sphakia Survey this year was done intensively, by the systematic walking of transects.
During the pilot season of 1987, one team of 4 walked "line" or "contour" transects, designed to run through or cut across selected areas. In 1988, we were aiming for more detailed coverage, particularly of the Anopolis Plain, and therefore walked transects in selected 1 km squares. There were two teams (A and B) of 3 or 4 people each, led by Nixon and Moody. Members of each team walked in parallel, 10-15m apart. Each team could cover 1/4 km2 per day, walking 5 N-S transects, 500m long and 50m wide, with a 50m gap between them. Thus in 2 days, the survey could cover 1 km2.
Walking transects in this manner requires a great deal of precision, both in map-reading and in pace-counting. Our work this year was enormously facilitated by the 1:5000 Greek Army maps, whose grid corresponds closely to that of the more accessible 1:50,000 British Military maps. For that reason we have used the co-ordinates of the latter maps to locate our sites (e.g. 17/26.1 etc; see fig. 1). Compasses were necessary for accurate navigation, especially in forested areas where sighting on a distant point was impossible.
At the beginning of every transect, all members of a team recorded archaeological and environmental information on their pace record forms. Every 50m, the team would stop for an 'official' reading (10 per transect, 50 per 1/4 km2); in addition one team member would continue to record environmental data on the environmental record form. 04.044 Initially, we also recorded every artifact in between 'official' readings; later, we recorded only diagnostic material between 50m readings. Brief comments were possible on the pace record forms, but more detailed remarks, and 1/4 km2 and unit summaries went on each team leader's transect comment form. When artifacts were collected, they were placed in small self-sealing paper envelopes or plastic bags with white plant labels, marked as follows: unit, team, transect no., metres (or paces), recorder's initials, and date.
student stayed at home every day on a rotational basis, to input pace
record forms and environmental data on the computer, and to wash and label
finds. The 1988 finds have been bagged by team and transect, and arranged
in order by map unit. The 1987 finds have been labelled by area, transect
and pace, and bagged by date; thanks to the kindness of the Khania Museum,
we had last year's material with us throughout the 1988 season, which
made comparison between site collections much easier, especially for areas
that have now been visited several times.
The definition of the word "site" has often been a problem in survey archaeology. For the Sphakia Survey, any locality with significant human activity is a site. This flexible definition means that not all our sites need be settlements; they could also be a set of ancient terraces; the area round a spring in the White Mountains, and so forth. Our sites are discovered in the field, often out loud, as the members of a team call out their sherd densities. Because we survey areas where sites might not be expected, and because we take readings every 50m, we can be fairly certain that we know where sites are, and where they are not. (In most of the areas covered so far, there seems to be very little background noise.) When a site is diagnosed, a special collection of material may often be made to include representative sherds for all fabrics and periods observed plus stone tools, glass, etc. Short site summaries are included on each team leader's transect comment sheets. Sites were then selected for a revisit by both team leaders, at which point a detailed site record form was filled out, and a general collection of artifacts to supplement previous collections was made. Site recording also included photography of the area, and of the finds.
This year we also used a video camera and recorder. Video has been used in excavations, but not, so far as we know, on surveys.3 The purpose of the video is to show our method of survey; to record sites and different types of landscape; and to include ethnographic information as necessary.
i. Prehistoric Period (PH)
This is the period about which the least was known before our Survey began, in terms of both the number and the distribution of sites. Only six prehistoric sites had previously been identified in Sphakia: the caves at Asphendou and Kapsodasos, plus two sites in the Frangokastello Plain; Loutro; and a possible Mesolithic site found by Mortensen in the Samaria Gorge. There were also the three Late Minoan III vases from "ta pharangia" said to be from Sphakia.4
After two seasons' work, we have located another 33 sites with PH material, both on the coast (e.g. at Poikilasion and Khora Sphakion), and inland (Livaniana, Mouri, and especially the plain of Anopolis). Thus there are currently 39 sites with a PH phase, with material of this date found in every region except the Madhares and Askyphou (though obsidian, usually to be associated with earlier sites, has been found in the latter).
While PH pottery is not usually difficult to recognise, PH sites can be difficult to define and measure. The pottery is sometimes scattered over large areas without a focus, presumably because it has been disturbed in the Graeco-Roman and later periods.
The earliest PH material in Sphakia may come from Mortensen's site in the Samaria Gorge, to which he assigns a tentative date of Epipalaeolithic or Mesolithic; a Palaeolithic date has been given for the cave at Asphendou, though later dates have also been suggested.
But most of the PH material from Sphakia consists of pottery dating to Final Neolithic/Early Minoan to Early Minoan, and stone tools. Though we do find some ground tools, chipped stone is far more common: obsidian, including at least one core; and chert, usually black or grey but in one instance rose. Black chert occurs in abundance in Sphakia; grey chert also occurs but less frequently. Obsidian occurs at nearly every site with a PH phase; it obviously has to be imported, and is Melian in appearance. The tools found in the survey have a wide range of specialised shapes; one site alone produced at least 10 different types (Fig. 1, 18/24.1 = 4.53).
Coarse ware chronology
Our pottery dates are based on the coarse ware chronology developed in NW Crete by Moody. The problem in W Crete generally is that painted decoration on pottery is rarely preserved, with the exception of the mottle burnished surface that dates the Final Neolithic/Early Minoan material (even when the sherds are relatively small). Moody built up her coarse ware chronology for the Akrotiri Peninsula by comparing surface finds with well-stratified excavation deposits. The assemblages of fabrics found in Sphakia are sufficiently close to those used by Moody to suggest that the rough outline of her chronology can also be applied to Sphakia. One particularly good example is an Middle Minoan fabric with calcareous temper found at the Anopolis site mentioned above (photo row 3, 2nd sherd from left), which was already known on the N coast from two excavated sites and Moody's survey.5
It is fortunate for the Sphakia Survey that these coarse ware ceramic parallels exist, as without them it would be difficult to date our finds even in broad terms. But the presence in Sphakia of pottery fabrics similar to those on the N coast (or possibly imported), and the amount of obsidian found both suggest that the area was not totally cut off from the rest of Crete, nor from the outside world.
The choice of site location suggests that access to the sea (and its resources) was often important. For example, 71% of the Frangokastello sites have a PH phase (10 out of 14 sites so far), compared with only 27% of Anopolis sites (11 out of 41 sites). This is not a function of exploration, as we have now surveyed most of the upland Anopolis plain, with relatively little time spent so far in the coastal area of Frangokastello. We also found PH material, which includes a stirrup jar disc [now, 2000, thought not to be a stirrup jar disc], at Poikilasion above the mouth of the Trypiti Gorge (Profitis Ilias SW slopes 1.02 ).6
It is important to note, however, that there are PH - often early PH (Final Neolithic/Early Minoan) - sites at inland and upland locations, e.g. Anopolis and Aradena (Fig. 1, 18/23.5 = 4.44 and 14/25.1 = 3.16; and the tripod foot from Mouri (6.02) (at c.1000m) suggests that the altitudinal range for PH activity was broader than one might have expected.
ii. Archaic, Classical, Hellenistic and Roman Periods
After the Bronze Age there is little sign of human occupation in Sphakia until the Classical period. Despite considerable prosperity elsewhere on Crete in the Archaic period, there is Archaic material from Sphakia only from graves at Tarrha.7
Four city states
But by the Classical/Hellenistic period there had emerged four city states: Anopolis, Araden, Poikilasion and Tarrha; the names of the first three, plus that of Phoinix, survive in modern toponyms. There is Classical occupation known at Anopolis and Tarrha, and by the Hellenistic period Anopolis, Araden and Tarrha were communities looking outward to the wider Greek world. All three were signatories (along with 28 other Cretan states) to a treaty with Eumenes II of Pergamon in 183 B.C., and at around the same time all three sent sacred ambassadors (theoroi ) to Delphi; so too a theoros went from Apollonia, just east of our survey area, to the Ptolemaia in Alexandria, where he died in 233 B.C.8
The early Hellenistic period was, however, a time of turbulence. In the late third century B.C. Anopolis was conquered by its enemies (perhaps Araden), to receive liberation at the hands of Kharmadas, one of its own citizens.9 The earlier fortification wall at Anopolis probably dates to this period. Poikilasion, whose Hellenistic location was first discovered by the Survey, occupied a site of some physical inconvenience, some 400 metres above the sea and higher than all the cultivable land . Such defensiveness reflects the general prevalence of wars and of piracy in the period. Indeed, many Cretans themselves served as mercenaries in the armies of the Hellenistic kings. Kharmadas, the liberator of Anopolis, served in the Ptolemaic army, and died at Gaza.
Such strife and insecurity was counterbalanced by the formation of political leagues. In the fourth and third centuries B.C. the Oreioi included Tarrha and Poikilasion with three cities further west, Elyros, Hyrtakina and Lissos; between c.280 and 250 B.C. the Oreioi formed an alliance with the king of Cyrene, the oath being by the deities of Lissos and Poikilasion and other deities. In the mid-third century B.C. a federation of Cretan cities came into existence, and formed the basis for the signing of the treaty with Eumenes II.10 The coming of the Romans in the first century B.C. marked the end of overt strife and insecurity within the island.
The prosperity of the area in the Hellenistic and Roman periods is difficult to assess. The surviving stonework shows no sign of ashlar construction, which contrasts with other sites of the same period on Crete (such as neighbouring Lissos) and elsewhere in the Greek world. The main local limestone (Plattenkalk) is unsuitable for squared blocks, and only a limited amount of other stone was brought in to make good the deficit (as, for example, at the main site at Anopolis). There are also few monumental civic buildings. There was a temple at Tarrha and perhaps two at Poikilasion (one of the latter quite small),11 but none of the four towns is known to have had a theatre, odeion, stoa or even agora. Parts of columns have been found at Anopolis and Araden, but they could be either Hellenistic-Roman or Byzantine, and are too small to have been load-bearing. This contrasts with the remains of Roman theatres at Elyros and Lissos, and the lavish architecture and sculpture revealed by excavation at Lissos.12
The settlement system of the area has been much clarified by the survey, and will be further illuminated once study of pottery enables a proper distinction to be drawn between Hellenistic and Roman wares. There seems to be a change in the khora/skala relationship between the Hellenistic and Roman periods. The main site of Poikilasion, where we observed no pottery later than the Hellenistic period, was perhaps abandoned in favour of the more congenial setting of the Trypiti gorge. And the importance of Phoinix as the port for Anopolis and Araden probably increased in the Roman period. But unlike examples of change elsewhere on Crete, when the inland city declined in favour of the port, Araden and Anopolis remained vigorous throughout the Roman period.13
Within the territory of Anopolis the survey has identified 35 out of 41 sites (85%) as Hellenistic-Roman, which is the most common of all periods; intensive survey has not so far been carried out on the territories of Araden, Poikilasion or Tarrha. On the Anopolis plain, in addition to the main site of ancient Anopolis (Fig.1, 17/23.1 = 4.21; ), our finds suggest two types of habitation sites: small hamlets of a few houses; and isolated farms. Both are scattered across the whole territory of Anopolis, often on south-facing slopes, on slight rises in the plain and at heights of up to 800 m. The precise dating of these sites is not yet possible, but one may note that outside Anopolis there is very little black glazed pottery (e.g. Fig.1, 17/25.2 = 4.28), and that is found at the larger sites; while in the Roman period Samian is found very widely, even at smaller sites (e.g. Fig.1, 18/23.2 = 4.41, 19/25.1 = 4.67).
Some evidence for agricultural practice are the terraces within the Anopolis plain and elsewhere in the survey area where a general scatter of Graeco-Roman pottery might suggest cultivation in this period (e.g. Fig.1, 16/26.1 = 4.17). This evidence joins the other, very limited, evidence for the use of terrace walls in the Greek world.14
connections between Sphakia and other parts of Crete and other parts of
the Mediterranean in the Hellenistic-Roman period require investigation.
There may be evidence of Hellenistic-Roman activity at a site in the Madhares;
and some Sphakiote personal names are peculiar to West Crete, which may
suggest links to the north coast.15 Pottery
offers the most systematic evidence that the area was tied to other parts
of the Mediterranean. At some point a 4th c. Attic red figure vase was
brought to the Frangokastello area,16 while in the Roman period pottery
was imported from Italy and North Africa, and a grain ship plying from
Alexandria to Rome once gained safety in the harbour of Phoinix.16
To what extent this part of the south coast gained substantially in prosperity
in the Roman period through trade remains to be seen. But the range of
non-Cretan pottery used in Sphakia does suggest that the area was linked
to the broader Mediterranean economy.
Late Roman period
Most of the sites occupied under the earlier empire seem to have continued in use in the later Roman/First Byzantine period. The area of major expansion was the Frangokastello plain, where the earliest post-Minoan evidence for occupation (apart from the red figure vase already mentioned) is mid-third century A.D. and where settlement became more dense in the fifth and sixth centuries A.D. Why this area was apparently not settled earlier in the Roman empire is at present unclear.
The most notable architectural, and social, development was the building of Christian churches in the fifth and sixth centuries at four sites: Araden; Phoinix (at least two); Tarrha ; and on the Frangokastello plain at Ag. Efstratios and Ag. Niketas. All these churches were larger than their replacements in the second Byzantine period. The one site notably missing from this list is Anopolis, where there is no clear sign of an early basilica. And the one bishopric in Sphakia was based not on Anopolis but on Araden/Phoinix. We cannot yet explain why Anopolis should have given ground to its neighbour and rival Araden.
As a result of the Sphakia Survey, we now know the correct site of Hellenistic Poikilasion; at Tarrha, we know that the city walls were for protection against the sea (rather than fortification), which has implications for Hellenistic and Roman sea levels. In the territory of Anopolis we have found imported Samian and African pottery even at remote sites, which would be very unusual in mainland Greece, and we are in a good position to analyse the settlement hierarchy of an ancient city state.
iii. Byzantine, Venetian and Turkish Periods (A.D. 962-1900)
Settlement in 19th and 20th c. Sphakia was of two types: "nucleated" villages of the type familiar from other parts of Greece (e.g. Agios Ioannis, Kapsodasos); and "dispersed" villages made up of several neighbourhoods or hamlets (e.g. Anopolis, with 9; and Askyphou, with 5). Dispersed villages occur elsewhere on Crete, especially in the W, but are otherwise rare in modern Greece.
It is likely that this pattern of settlement goes far back into the last millennium, possibly to the Second Byzantine period (from the departure of the Arabs in 962 to the Genoese (and then Venetian) takeover of Crete in 1204). There is no evidence of any other settlement pattern, either from the work of the Sphakia Survey, or from earlier studies.17 While all settlements in Sphakia have been depopulated, there are relatively few earlier desertions before this century. We know of only one village abandoned well before 1900, but still having standing ruins: the area now known as Peradoro in the Trypiti Gorge, reported by Spratt in the 1850s, when it was already long deserted and nameless.18 And to date we have discovered only two other mediaeval habitation sites outside villages, represented by sherd scatters and building foundations, near Anopolis (Fig. 1,18/24.3 = 4.54) and Khora Sphakion.
Villages of both types (nucleated and dispersed) are distributed wherever there is cultivable land. The highest village is Mouri (6.02) (now deserted) at ca 1000m, where Byzantine/Venetian material has been found. We have seen no agricultural terraces higher than 1000-1100m, but up to this altitude almost every slope with even a little soil has been terraced. The presence of numerous threshing floors suggests that the main crop was grain. It is unfortunate that terraces are so difficult to date, so that we cannot say definitely which (if any) areas were newly terraced in the second Byzantine period.
Though there is cultivable land by the sea in Sphakia, the coast was conspicuously lacking in settlement through the 1000 years from the late Roman period/first Byzantine period up to the age of tourism. Settlements are sometimes located 1 km or so inland, but are not directly visible from the sea; examples are the deserted Trypiti village at Peradoro, and Agia Roumeli (the old village). Avoidance of coastal locations is usually attributed to insecurity; as late as 1867 eastern Sphakiotes did not live on the Frangokastello plain because they feared raids from the sea.19
But as elsewhere in Greece, churches in Sphakia were often built outside settlements as well as within them (though never, as far as we know, above 1000m), and some were built on or near the sea, where they served as conspicuous landmarks. The only structure definitely datable to the second Byzantine period is in fact the 10th c. church of Agios Pavlos on the shore below Agios Ioannis, on the route to Agia Roumeli.20
Other evidence for this period comes from documents. There is a charter of ca 1184, by which the Byzantine emperor Isaac Angelos grants or restores the governance and revenues of a territory dependent on Anopolis to the Skordylis family. This territory is defined by a perambulation, and its boundary may be roughly determined. The area began near the future Frangokastello, ran along the N side of the White Mountains, and continued down to Agia Roumeli. The Anopolis province was therefore much the same as the modern eparchy of Sphakia, but somewhat shorter at the E and W ends. The word "Sphakia" is mentioned in the charter, and seems to have been in use as a geographic term, but was not so precisely defined as to make the perambulation unnecessary. The 41 place names in the charter are not those of settlements but of natural features (e.g. Three Olive Trees). That it is possible to locate a number of these toponyms with some precision is remarkable, given that nearly 800 turbulent years have passed since they were recorded.21
In the Venetian period (1204-1669) there is some material evidence for the importance of Anopolis and a certain prosperity in Sphakia generally. Gerola first noted a house with Venetian features in Anopolis;22 it could have belonged to the Skordylides. The house is also a indication that the main settlement of Anopolis was no longer on the ridge where the Graeco-Roman city had had its centre. No church in Anopolis has yet been dated to the Venetian period, but at least three churches in the eparchy have 14th c. frescoes (one at Aradena, two at Agios Ioannis).23
The economic importance of Sphakia is more clearly revealed in documents of the Venetian period. In 1463 the Anopolis plain is listed as a major grain-producing area; 24 again, it is frustrating not to be able to date the construction of agricultural terraces. Cypress was a valuable timber of which the White Mountains were an important source in the Venetian Empire in the 15th c. and later. Some of the 11 sawmills in the Samaria Gorge observed by us may be Venetian; mills in this area are mentioned in Basilicata's report of 1630.25 The Skordylis family reappears in two 15th c. documents, in which a murderous dispute with another Sphakiote family over pasturage for sheep and goats is to be resolved by a dynastic marriage.26 Sheep were kept for their milk, to judge from a 17th c. traveller's account:27
Pastoralism seems to have been an integral part of the Sphakiote economy at least as early as the Venetian period.28 It is tempting to assume that the system of transhumance still practised in Sphakia--sheep and goats are moved in the summer to the Madhares (high pastures), where large round cheeses (graviera) are made in the corbelled huts called mitata--was already in operation in the Venetian period. But the mitata cannot themselves be dated, and at this altitude (1700-2000m) there are no churches to provide any chronological peg.29
All three products mentioned--grain, timber, and cheese--were for export to Venice. The inhabitants of Sphakiote villages, then, should not be seen as subsistence farmers and shepherds, but as the productive subjects of a highly exploitative imperial power. The Venetian yoke did not always sit lightly on Cretan shoulders, as the insurrections of 1294-99 and 1366-67 reveal; material evidence for trouble in Sphakia comes from the fort at Frangokastello built in 1371-4.30
Sphakia continued to be rebellious during the Turkish period (1669-1898). Daskalogianni, who was from Anopolis, led the revolt of 1770, as a result of which the village was burnt. The peculiar lack of older, larger olive trees in the Anopolis plain may be the result of Hüseyin Bey's order of 1824, that they be cut down for his troops' firewood.31 At least 9 new forts were built, some commanding good views of the coast (even at the cost of being difficult to maintain or reinforce; e.g. the forts at the mouths of the Trypiti and Samaria Gorges).
Sphakia, like much of Greece, has the remains of a remarkable network of carefully constructed mule tracks. They are called kaldirimia (from the Turkish kaldirim) but we have as yet no evidence for their date of construction. The main route into Sphakia was through the Imbros Gorge where there are remains of a broad, well-paved kaldirimi; Pococke travelled on this road in the 1740s.32 Lesser tracks (but even the narrow ones paved and provided with steps) connect all the villages and cultivated areas, and run even into the Madhares and the Trypiti Gorge.
The only monastery known so far in Sphakia, that of Agios Kharalambos, was built right on the coast of the Frangokastello plain (E of the fort) circa 1835.33 Other structures of the Turkish period include the Kankellaria at Loutro, built in 1821, which housed the Revolutionary Committee.34
After two seasons of field work, the Sphakia Survey has located (or relocated, in some cases) 98 sites, of which 34 have a prehistoric phase (35%); and 67 have a Graeco-Roman phase (69%). We surveyed 5 km2 in 1987, and 10 km2 in 1988.
As for our broader objectives, we have learned a great deal about Sphakiote connections with the outside world. In the prehistoric period, the presence of obsidian and a particular pottery fabric suggest links with (respectively) the Aegean (probably Melos), and the NW coast of Crete. The scarcity of Classical-Hellenistic black glazed pottery contrasts sharply with the surprisingly wide distribution of Samian ware. Similarly, the material culture and documentary evidence of the Byzantine, Venetian, and Turkish periods suggests that "remote" Sphakia was integrated into the economy of Crete and beyond.
We can also draw some tentative conclusions about preferred locations in different periods, to be tested in our next field season. The few Final Neolithic/Early Minoan sites discovered so far are all in the area of Anopolis and Aradena, i.e. at about 600-800m. Later in the PH period, however, people lived on the coast at Loutro and also in the plain from Khora Sphakion to Frangokastello. Tarrha, the only Archaic site known to date, is also a coastal settlement. But Classical and Hellenistic material tends to be found away from the sea; Poikilasion is a good example. Under the Romans, occupation continued at many inland sites, such as Anopolis and Araden, but the main settlement at Poikilasion seems to have shifted to the foot of the Trypiti gorge. By the late Roman/first Byzantine period, there were many coastal settlements, the most notable of which was probably Loutro, with a number of smaller sites in the Frangokastello plain. But for most of the second millennium A.D., people lived in inland villages, up to 1000-1100m.
Future field work in Sphakia will be carried out in environmental zones not intensively sampled so far (the Madhares; the Askyphou and Frangokastello plains), with continued revisiting of other areas; by the end of our third season we hope to have surveyed a representative sample of the eparchy. Work can then begin on the final publication of the Sphakia Survey, which will incorporate environmental, archaeological, documentary, and ethnographic evidence.
LUCIA NIXON, QUEEN'S UNIVERSITY/CANADIAN RESEARCH INSTITUTE FOR THE ADVANCEMENT OF WOMEN,
JENNIFER MOODY, UNIVERSITY OF MINNESOTA
SIMON PRICE, LADY MARGARET HALL, OXFORD
OLIVER RACKHAM, CORPUS CHRISTI COLLEGE, CAMBRIDGE
The Sphakia Survey: Internet Edition, University of Oxford 2000
Designed and Developed by ACDT at University of Oxford
This page last updated