For all the indications that sea travel was fundamental to the Mycenaean political economy, we have little evidence for their coastal anchorages, let alone developed harbors or port towns, in the Aegean. We must ask why this is; what, if anything, we can do to recover and investigate them; and why we should want to do so in the first place. These questions form the focus of this chapter.
To tackle the last question first, the importance of identifying specific landing sites instead of merely characterizing the types that would have existed on Bronze Age Aegean coasts (e. g., Morton 2001: 6—7) must be established. Many anchorages will have been small, susceptible to alteration over time, and not necessarily accompanied by a settlement or a conspicuous artifact scatter. Is it worth the time and effort to search for them? Surely it is if we are serious about understanding the networks of interactions by land and sea that amounted to the connectivity of daily life, particularly if, following Horden and Purcell (2000: 123), our aim is to reveal “. . . the various ways in which microregions cohere, both internally and also one with another — in aggregates that may range in size from small clusters to something approaching the entire Mediterranean." Only by knowing the locations of the coastal nodes in these networks can we fully make use (and sense) of archaeological data bearing on interaction at all scales.
Dedicated harbors as well as anchorages used only intermittently or opportunistically must have been abundant in many regions during Mycenaean times, both because of the morphology of the Greek coastline and because they were needed. As communication by sea expanded, a multitude of anchorages, large and small, was required to ensure that voyaging was as safe as possible, and to facilitate economic and social relationships. Ship captains needed access to safe anchorages to shelter from winds and storms, to procure provisions, and to enter into various kinds of transactions, including trading and raiding. Running out to sea to escape a storm was a maneuver of last resort; to seek a coastal haven was much preferred. At times, ships and their crews were forced to wait for favorable winds, and this could occupy days or weeks during which their needs for sustenance would have to be met. The potential hostility of the local population was a complicating factor that must often have forced crews to seek alternative landing sites. Another function of small, scattered anchorages might have been as convenient pickup points for agropastoral products to be transported by sea in local and regional trade networks (Rothaus et al. 2003: 40).
Too often, however, the existence of safe, suitable anchorage is taken to be self-evident or based on guesswork that is quite possibly wrong. Often, it seems, the problem of long-term coastal change is simply left unexamined. A common mistake is to assume, implicitly or explicitly, that Bronze Age and modern coastal morphology are essentially the same. Even when some attempt is made to infer changes based on observations of modern coastal landscapes, in the absence of geomorphological analysis the conclusions reached can be misleading or wrong, and are at best limited in their ability to lead to a genuine understanding of ancient coastal environments. The actual changes to Aegean coastlines over time vary widely, but they tend to have localized causes that can only be charted with geoarchaeological techniques. In this chapter, I describe the process of working back to Bronze Age coastlines, and stress the importance of doing so in a systematic way.1
Conditions of Discovery
There are few places on the modern coastline to search for Bronze Age anchorages and harbors that can truly be called “obvious." Locations that boast fine harbor basins today, or that preserve harbor works from historical periods of the past, are no sure indicators of high potential for harbors of the Bronze Age. One reason for this is changes in maritime practice — mainly ship technology — over time. It is undoubtedly true that small, shallow anchorages perfectly suitable for Mycenaean shipping could not accommodate the large, heavily laden military and commercial fleets of later Greek and Roman antiquity. Thus, to limit the search to prominent historical and modern harbors would be to miss most of the Mycenaean maritime landscape.
A more important reason is pervasive change over time in the physical topography of the coastline, caused by short - and long-term geomorphological processes. The most prominent of these are sea-level change, sedimentation, marine erosion, and tectonics. Eustatic sea-level change, caused by the cycling of ocean waters into and out of the polar ice caps, has had a profound effect on the world's coastlines since the end of the last Ice Age. Following the last glacial maximum circa 18,000 years ago, at which time eustatic sea level lay
At 90 to 150 meters below present sea level, global warming induced rapid sea-level rise, interrupted only by the cool Younger Dryas event of 10,000 to 11,000 years ago, until circa 6000 BP (van Andel 1989; Wells 2001: 151). In the Mediterranean, maximum sea transgression into coastal areas was reached circa 6000 BP, followed by a stabilization of eustatic sea level. Since that time, global eustatic sea level has slowly risen by no more than five meters or so. Since the LBA, eustatic sea-level rise is only a few meters, a figure that can account neither for observed vertical changes in relative sea level caused by tectonic uplift and subsidence, nor the lateral and vertical changes effected by coastal sedimentation and erosion.
Moreover, the discovery potential for Mycenaean harbors is not the same for all parts of Greece and the Aegean, because regional-scale geomorphological histories may make detection of ancient harbors relatively easier or more difficult. As a result of regional tectonics, some coastlines abound in natural coves and inlets, while others can be characterized as virtually harborless coasts. Faulting of the Greek land mass into mountains and intervening valleys in a general northwest-southeast alignment, followed by inundation of the coasts by the sea, has produced the characteristic pattern on Greece's Aegean (eastfacing) shores of promontories separated by deep gulfs, with numerous smaller inlets on a generally rocky coast (Morton 2001: 15-16). The Aegean coast of Asia Minor developed deep inlets reaching tens of kilometers inland at places like Miletos, Ephesus, and Troy. These great estuary and delta systems hosted superb Bronze Age harbors, but have silted in completely in historical times, leaving once-great harbors stranded many kilometers inland.
There are other distinctive patterns. Some coastlines are formed by major faults that separate a mountain ridge from an adjacent collapsed and submerged block; these coasts are steep and linear, with few inlets or offshore islands, and thus little safe anchorage (Morton 2001: 137). Examples of such “harborless" coasts occur in Epirus and Thessaly (the two flanks of the northern Greek mainland), and on the long northeastern coast of Euboea. They tend to be cliffbound and very deep — difficult places to find holding ground for an anchor or to come ashore for provisions. In a few places, perennial rivers have broken through to the coast, forming broad estuaries and deltas. In Epirus north of the Ambracian Gulf, marine embayments or estuaries at the mouths of the Acheron and Kalamas (Thyamis) Rivers provided safe haven and access to fertile hinterlands. If it can be established that the separation of the two blocks forming such coastlines occurred before the Bronze Age, the candidates for harbor locations are few and the search is simplified.
By contrast, the long, west-facing coastline of Elis in the western Peloponnese has witnessed several sequences of accretional barrier and lagoon formation since the end of the Mesolithic period, with the result that most evidence of Bronze Age coastal landforms and archaeological sites is now buried under meters of sand and wetland deposits. There, investigation of Bronze Age coastal environments requires long-term programs of geological coring, and results remain tentative even after years of study (Kraft et al. 2005).
The lack of knowledge about Aegean Bronze Age coastlines is also partly a symptom of the general historical trajectory of maritime archaeology in the Mediterranean, which in the twentieth century was dominated by studies of ships and shipwrecks, at the expense of harbors and coastal landscapes (Mar-riner and Morhange 2007: 137—44). The emphasis on ships continues today in both the Old and New Worlds (e. g., Bass 2005a; Blue et al. 2006; Gould 2000). The shipwrecks at Uluburun, Gelidonya, and Point Iria have generated extraordinary information about LBA seafaring and maritime trade, but as restricted spatial and temporal events, they constitute a small part of a much larger picture (Marriner and Morhange 2007: 180). When harbors were the object of study, they tended to be the largest and best-known artificial harbors of the Greco-Roman world. Prehistoric and small, natural anchorages and harbors were rarely investigated, nor were there frequent efforts to reconstruct entire “maritime cultural landscapes" formed by networks of landing sites, coastal settlements, and their ties with other communities by land and sea (Westerdahl 1992). In this respect, Mediterranean maritime archaeology is less advanced than in northwestern Europe or the Baltic area (e. g., Chapman and Chapman 2005; Cunliffe 2001; Ilves 2009). Moreover, until recently, maritime archaeology in Greek waters lagged behind other areas in the Mediterranean, such as the Levantine and Turkish Aegean coasts. This can be attributed in part to highly restrictive controls placed on underwater archaeology by the Greek underwater archaeological authority (known as Enalion), particularly with respect to foreign teams. As a result, most of the major projects and innovations in underwater technique were developed on the coasts of countries like Israel and Turkey, where permitting was more liberal (Bass 2005c; Raban 1991). In a policy shift that occurred after 2000, Enalion has invited collaborations with foreign teams, and many are now underway alongside an expanded agenda for Enalion's own projects.
It must be pointed out that a long tradition of coastal geomorphology has existed in the Aegean, targeting coastal change over time in both local and regional settings, for example Franchthi Cave (van Andel and Sutton 1988), Asine (Zangger 1994b), Tiryns and the Argive Plain (Niemi and Finke 1988; Zangger 1991, 1994a), Pylos (Kraft et al. 1980), Dimini (Zangger 1991), Troy (Kraft, Kayan et al. 2003; Kraft, Rapp et al. 2003; Rapp and Gifford 1982), Ephesus (Kraft et al. 2000), Miletos (Brtickner 2003), and the deltas of the Alpheios (Kraft et al. 2005) and Acheron (Besonen et al. 2003) Rivers. In Greece, this work could generally be accomplished under more easily acquired permits granted by the geological service (IGME). Yet the specific areas in need of development in the Aegean are two: first, there is a lack of systematic programs of research that identify and investigate local and regional maritime cultural landscapes in a holistic way; and second, programs of coastal reconstruction have often not been closely coordinated with terrestrial surveys and excavations, in terms of research design and planning, shared fieldwork and resources, and joint publication. For example, although all the programs of coastal paleogeography cited above sought to address specific archaeological questions, most were not integrated closely with a particular excavation or survey on land. Meanwhile, excavations and regional archaeological surveys on islands and in coastal areas have rarely extended their focus beyond the shoreline, with the result that the potential of integrative concepts, such as the maritime cultural landscape, which situate coastal regions within interactive networks that unite inland, coast, and sea, has scarcely been tapped (Berg 2010). Some vestiges of the former restrictions remain as mandated by national legislation covering all archaeological activity, including a limited number of available permits for foreign researchers and a clear distinction between terrestrial and underwater permits (these come under separate authorities, and a project can rarely expect to hold both land and underwater permits in a single year). Yet there are many ways to make collaborative research work within the present regulations: much paleocoastal fieldwork can still be accomplished on land under geological (IGME) permits, and it is possible to bring together mixed Greek and foreign nationals to work in a given locality under separate projects and permits. Nevertheless, moving toward a truly holistic coastal archaeology of maritime cultural landscapes will require fully integrated programs of archaeological and geomorphological investigation of sea and land in coastal regions.
Geomorphology of Coastal Zones
Coastlines are among the Earth's most dynamic geomorphological settings. George Rapp and John Kraft (1994: 71—72) list the characteristic and interrelated processes: “local tectonism; eustatic sea level change; climatic change; ocean currents and wave regimes; the nature and frequency of catastrophic events; sources, types, and quantities of sediments available and the resultant aggradation and progradation of deltaic floodplains into erosional and tectonically derived embayments; and the nature and intensity of human activity." Over time, even the most stable coastal environments evolve as sediments accumulate or shorelines erode. Local relative sea level moves up or down in response to changes in the absolute level of the land (tectonics, isostasY2), sea (eustasy), or both. Climate and topography control the flow of energy (tides, waves, currents, sediment flux) in a coastal system. The tectonic environment (faults, tectonic events) impacts topography and relative sea level, and sometimes introduces changes to coastal topography that are catastrophic from a human perspective. Finally, human impacts on coastal environments, notably in the form of increased sedimentation, have been detected since the Neolithic period in the Mediterranean. The following discussion summarizes briefly the most important processes and materials involved in long-term coastal evolution, following the thorough treatments of the topic by Lisa Wells (2001) and Michael Summerfield (1991: 313-42).
Broadly speaking, a first-order distinction can be made between open coasts and protected coastlines (Wells 2001: 150). Many coastlines cannot be characterized as completely open or completely sheltered, but for a coastal location to have been viable as a harbor, it must have been sheltered to at least a considerable degree from winds and waves. Often, because of shifts in wind direction and intensity during the course of the year, the viability of a harbor may respond to seasonal or even day-to-day conditions.
Open coasts are exposed to the full impact of winds and wind-driven waves, creating high-energy environments and landforms dominated by beaches or rocky shorelines. Coastal erosion is common on open coasts, and sediments tend to be relatively coarse with abundant organic material including woody debris or carbonate particles. By contrast, on protected coastlines, wave energy is attenuated by refraction, creating low-energy environments exemplified by estuaries and tidal marshes in which sedimentation dominates coastal evolution and landform development. Sediments are relatively fine grained and may contain highly organic deposits such as peat. Protected coastlines are ideal locations for sheltered harbors.
Long-Term Coastal Evolution
Coastal evolution can be characterized in terms of the relative motion of the shoreline over time in response to the processes outlined by Rapp and Kraft above, and by the resulting landforms, which can appear and disappear with successive periods of coastal change. Joseph Curray (1964; see also Wells 2001: 154-55) proposed a classification of coasts based on their relative motion over a discrete period of time: (1) progradational coasts grow seaward (prograde) as a result of sedimentation; (2) transgressive coasts are submerged as a result of relative sea-level rise; (3) recessive coasts erode landward (marine erosion); (4) regressive coasts emerge as relative sea level falls; and (5) aggradational coasts grow vertically (aggrade) when the rates of sea-level rise and sedimentation are roughly equal (Fig. 5.1). The early Holocene record of the Mediterranean, outlined above, provides a good example of rapid and widespread transgression as global sea levels rose. The maximum transgression is often marked by wave-cut notches in a former sea cliff, or other signs of a paleoshoreline stranded far inland from the modern shoreline. After eustatic levels stabilized circa 6000 BP, these transgressive coasts shifted to progradational, recessive, or aggrada-tional, depending on the rate and dynamics of sedimentation and erosion in a
5.1 Classification of coasts by relative motion of the shoreline. After Wells 2001: 154, fig. 6.2. Courtesy of University of Utah Press.
Given location. The Aegean coasts of Turkey provide dramatic examples of deep embayments created by the late Pleistocene to early Holocene marine transgression, followed after circa 6000 BP by a gradual but inexorable progradation of tens of kilometers in the major river deltas. The relationship of sea and land created by the interplay of erosion and sedimentation can be altered locally by tectonic movements.
World History
