In 2001, members of the Eastern Korinthia Archaeological Survey (EKAS) discovered a large Mycenaean architectural complex at the location Kalamianos near the village of Korphos, on the rugged Saronic coast of the southeastern Corinthia (Fig. 7.13; Rothaus et al. 2003; Tartaron et al. 2003). The importance of the site was instantly clear: walls and foundations of buildings of Mycenaean
7.13 Digital terrain model of the Korphos region.
Type, some of them monumental, are exposed on the surface of the gentle seaside slope above the cape known as Akrotirio Trelli, covering almost eight hectares on land and an unknown further extent now submerged underwater (Fig. 7.14). In 2006, the Saronic Harbors Archaeological Research Project (SHARP), which I co-direct with Daniel J. Pullen, was constituted for the purpose of initiating investigations on the site and in its surroundings.5 From 2007 to 2009, a first phase of surface investigations was carried out, comprising detailed mapping and architectural study, a surface survey on the site and in a zone of seven square kilometers around it, geomorphological and environmental research, initial underwater investigations, the recording of oral histories, and various
7.14 Aerial photograph of the Kalamianos site. Balloon photograph by Kostas Xenikakis and Symeon Gesafides.
7.15 GIS plan of architecture and other features at Kalamianos. Photo by author.
Specialist studies of the artifacts collected by the survey. (For a detailed preliminary report, see Tartaron et al. 2011.) SHARP hopes to undertake excavations at Kalamianos at a future date.
The site consists of an urban settlement preserved as stone architectural foundations and walls occupying approximately 4.5 hectares set within a town wall enclosing around eight hectares (Fig. 7.15). The “empty" 3.5 hectares seem to have been used for agricultural terraces and to quarry the settlement's building stone. Because of a unique convergence of tectonic activity, erosion, and human history, these features are exposed on the surface, giving us a rare opportunity to study a virtually complete Mycenaean settlement. The buildings employ a characteristic Mycenaean large-stone and - rubble construction, with foundations and walls preserved in situ, surrounded by massive stone collapse that indicates the considerable height of the original walls (Fig. 7.16). To date we have recorded over 1,200 walls and more than 50 buildings.
Although Kalamianos witnessed human activity at detectable levels during much of the Bronze Age, the urban settlement was a new foundation, laid out with a strong measure of central planning in a short period of time beginning around 1300 BC or a little earlier. Most buildings are oriented roughly to the cardinal directions, with long axes either north-south or east-west. Yet neither the layout nor the buildings themselves are uniform across the site. In certain areas, multiroom buildings cluster to form complexes, whereas elsewhere buildings are free-standing and often set at a distance from one another. Moreover, some of the buildings can be described as monumental while others are more modest in size and architectural elaboration. These contrasts suggest some form of differentiation that may be social, functional, chronological, or some combination thereof.
The chronology of the Kalamianos site was firmly established by a gridded intensive surface survey. Artifacts and features were recorded in regular 25 x 25 meter grids, and special collections were made from the interior spaces and rubble cores of intact buildings. The canonical masonry of the walls provides a rough chronology in the palatial period (circa 1400—1200 BC), but the retrieval of LH IIIB pottery built into the cores of the walls of many buildings provides a terminus post quem that indicates a construction date in the thirteenth century. A preliminary analysis of the pottery collected at Kalamianos shows how dominant Mycenaean material is relative to all other periods. If we remove the unidentifiable sherds, LBA makes up 86% with Late Roman coming in a distant second at 5.5%. Also significant is the fact that we have not yet recognized LH IIIC material, meaning that Kalamianos was likely abandoned by around 1200, and so may be closely tied to the palaces and their fate. Postabandonment phases from LH IIIC through Hellenistic are virtually absent.
Geomorphology of an Unlikely Harbor
We have strong evidence that Kalamianos was a harbor settlement in the Mycenaean palatial period, and we have come to believe that it served as Mycenae's principal Saronic harbor in the thirteenth century. Yet we could never have imagined making such bold statements upon first encountering the site. Kalami-anos is by no means an obvious location for an ancient harbor: a shallowly submerged peninsula off the coast makes it impossible for even small boats to avoid the shoals and approach the shore today. We approached the Korphos region as most observers would (e. g., Conlin 1999: 77), assuming that if an ancient harbor were to be found, it would be located in the sheltered, inviting Korphos Bay (Fig. 7.13), but Kalamianos provides a perfect illustration of the point, emphasized in Chapter 5, that we cannot assume that ancient Aegean coastlines possessed the same configurations as their modern counterparts.
The modern coastline in the Korphos region is rugged, dominated by a rocky shoreline that plunges to water depths of three or more meters, with the exception of Korphos Bay. Despite its rugged structure, the Saronic coast offers an abundance of small, sheltered anchorages. This was surely true in ancient times as well, but the configuration of the shoreline has changed dramatically since the Bronze Age due to tectonic displacements. In the Corinthia, tectonic movements have occurred along several major regional extensional fault systems with a complex history of differential fault motions. In low-lying, shallow water contexts like Kalamianos, these forces can bring about significant changes in coastal configuration with even small changes in relative sea level. The narrow
7.16 Example of large-rubble construction of Mycenaean buildings at Kalamianos. Photo by author.
Land shelf at Kalamianos slopes gently into the shallow offshore waters, with depths of only several meters within 125 meters of the shoreline, after which the sea floor drops abruptly to 50 meters, and within 500 meters from shore reaches more than 100 meters depth. This feature is known to local fisherman as the “chasm," and is exploited as a particularly fertile fishing ground that has sustained the fishing trade for generations.
We have followed multiple lines of geomorphological evidence to reconstruct the coastline and harbor basins of the Bronze Age. Recently, a Canadian-American team collaborating with EKAS determined that the coastline of Kor-phos Bay, about three kilometers west of Kalamianos and just southwest of Korphos village, has undergone net subsidence during the Holocene as a result of co-seismic fault motion (Nixon et al. 2009). From a series of cores taken in a salt marsh, they identified up to five phases of local coastal subsidence since the mid-Holocene, associated with seismic events resulting in rapid relative sea-level rise. The transgressive events were recognized by shifts in the abundance of microfossils (foraminifera, thecamoebians) in marsh sediments and correlated with tidal notches in the inshore area. They estimate a relative sea-level rise of about four meters in the last 5,500 years. Members of the same team recognized several beachrock platforms at depths up to 5.9 meters in the inshore areas adjacent to Kalamianos (Rothaus et al. 2003; Nixon et al. 2009). These cemented beach deposits were formed in the supratidal zone close to sea level and provide a useful indicator of former sea level (Kelletat 2006; Vousdoukas et al. 2007). In spite of the proximity of these two locations, their tectonic histories are not identical; Nixon and colleagues report that Korphos Bay and Kalamianos have distinct and independent sequences controlled by different fault blocks (Nixon et al. 2009: 51—52). This result illustrates how localized tectonic effects can be, with serious implications for coastline reconstruction, while the shared indications of multiple subsidence events support the archaeological evidence of submerged Bronze Age structures and artifacts off the coast at Kalamianos.
The next step toward identifying the configuration of the Bronze Age coastline and harbor basin was taken in 2009, when a collaborative project was initiated between the Canadian Institute in Greece and the Greek Ephorate of Underwater Antiquities (Enalion).6 More than 400 line kilometers of bathymetry, side-scan sonar, sub-bottom seismic, and magnetic survey data were acquired within a ten-square-kilometer expanse of sea in the Korphos region using a seven-meter Zodiac inflatable survey boat. The bathymetric survey generated a detailed map of the sea-bed relief around the site, and determined the location and configuration of beachrock ridges identified by previous work, which were then mapped using Differential Global Positioning System (DGPS) equipment. The sub-bottom seismic and magnetic survey data provided information on sediment thickness, bedrock structure, and location of buried ballast and pottery materials within the harbor basin. Underwater diver surveys were conducted using scuba equipment to investigate the submerged beachrock platforms and other targets identified by the geophysical survey. These were documented with underwater video and samples were obtained at several locations for ongoing laboratory analysis (grain size, micropalaeontology, pottery studies) and AMS radiocarbon dating of shell materials.
Results
Based on the results of these studies, Joseph Boyce has constructed a preliminary model of the evolving Bronze Age paleoshoreline configuration (Fig. 7.17). The bathymetry clearly identifies a submerged bedrock promontory extending east from Akrotirio Trelli and a drowned isthmus that formerly connected the small islet with the mainland coast. The submerged isthmus divides the inshore area into two separate lagoonal basins (the “western" and “eastern" basins in Fig. 7.17a). Two distinct beachrock platforms (BR-1, BR-2) appear in the bathymetry mapping and were confirmed by diver survey. BR-1 consists of two mound-like beachrock outcrops located on the submerged isthmus, about 100 meters from shore. The mounds are up to 1.2 meters in height, 30 to 40 meters in length, and about 20 meters in width. Both outcrops are elongated roughly parallel with the modern shore and have a basal water depth of 3.2 to 3.6 meters. Cemented into the calcarenite of BR-1 are thousands of Mycenaean sherds, constituting
7.17 Reconstructed coastlines and harbor basins at Kalamianos. Courtesy of Joseph I. Boyce, Despina Koutsoumba, and the Trustees of the American School of Classical Studies at Athens.
Around 30% to 50% of the beachrock volume and showing little sign of postdepositional reworking or biological alteration. This condition is consistent with rapid burial, as with a tectonic event, in a supratidal low-energy beach environment. The lowermost beachrock platform (BR-2) occurs at a depth of 5.4 to 5.8 meters on the western margin of the submerged promontory. The beachrock is about 0.4 to 0.6 meters in height and incorporates well-preserved sherds of EH pottery making up 10% to 20% of the beachrock volume. This pottery also preserves surface decorative features and lacks significant postdepositional reworking or biological alteration, consistent with rapid burial. Because beachrock forms at the interface of shore and sea, and because the Aegean is nearly tideless, we know that at one time BR-1 and BR-2 were shoreline positions. The pottery cemented into the platforms gives terminus post quem dates for the formation of the beachrock; that is, BR-1 could not have formed prior to the Mycenaean period, and BR-2 must have formed in the EBA or later. Yet because the condition of the pottery suggests rapid burial and not gradual transport or wearing away of surfaces, and because our examination of the potsherds to date indicates segregation of the pottery phases with little mixing of earlier or later material, it is highly likely that the broken sherds were incorporated into the deposits roughly during the time of their use, whether as the refuse of normal harbor activities or the result of a catastrophic tectonic event.
The provisional chronology derived from the associated pottery allows us to assign the BR-1 shoreline to LH III (circa 1400—1200 BC) and the BR-2 shoreline to an EH phase (circa 2700—2200 BC).7 As reconstructed, during the LH (Mycenaean) phase the islet was much more extensive than at present (approximately 500 square meters) but separate from the mainland. The bedrock promontory on the east side of Akrotirio Trelli would have provided a sheltered anchorage site (western basin) with a deep-water approach, the extent of which is approximate because the thickness of the post-Mycenaean sediment fill has yet to be established in seismic and core data. During the Mycenaean phase, small boats could have been pulled up onto shore, and larger ships may have anchored in the western basin or moored at the offshore island. The process of onloading and offloading may have generated much of the broken pottery preserved in BR-1. The western basin would have provided a sheltered anchorage during periods when the dominant winds were blowing from the north or west to southwest, accounting for most wind patterns throughout the year. During periods when winds were blowing from the east and southeast, the offshore island offered some protection from winds and along with the submerged promontory diminished wave energy, but ships might also anchor off the western side of Akrotirio Trelli.
During the EH phase, the local relative sea level was about 5.4 meters below present and the island was connected to the mainland via an isthmus that stood 1.0 to 1.5 meters above sea level. Together, the island and isthmus formed a natural recurved breakwater about 250 meters long and 40 to 50 meters wide,
7.18 Ballast pile identified in inshore waters at Kalamianos. Courtesy of Joseph I. Boyce, Despina Koutsoumba, and Trustees of the American School of Classical Studies at Athens.
Creating a well-protected double harbor configuration with many options for moving watercraft as required by weather conditions and a sufficiently deep approach to permit even the largest seagoing vessel of the day — the Cycladic longboat — to anchor close to shore.
Other important clues to the location of anchorage sites were obtained from the distribution of ships' ballast, which can be detected by a magnetic gradiome-ter survey even when buried at some depth (Boyce et al. 2009). Magnetic surveys in the eastern and western harbor basins at Kalamianos identified a number of magnetic “hotspots" found by subsequent examination to be associated with accumulations of volcanic ballast stones and pottery, which have a significant induced and remnant magnetization compared to the local limestone bedrock and seafloor sediments. Diver reconnaissance surveys of the western basin identified a number of small ballast stone piles and a large, partially exposed ballast mound consisting mainly of andesitic boulders and limestone cobbles (Fig. 7.18). The exposed portion of the ballast mound is four to five meters in diameter and includes scattered Mycenaean pottery fragments. Mapping the distribution of magnetic anomalies and recording their sources is helping to pinpoint the locations of anchorage sites. An intriguing and possibly telling pattern in the magnetic data shows numerous anomalies all around Kalamianos, but few in Korphos Bay. This pattern seems to confirm that Kalamianos was the area's primary anchorage, and there is some evidence that the modern Korphos Bay may have been primarily a wetland in the Bronze Age.
World History
