Publication: Pirates in the Bay of Pamphylia: an Archaeological Inquiry

The theoretical framework to the Rough Cilicia Archaeological Survey assumes that western Rough Cilicia, because of its abundant forestry resources and its unique location along one of the most important waterways of the ancient Mediterranean, enjoyed a strategic importance that surpassed its otherwise peripheral position within the ancient Mediterranean world system.It also assumes that the native inhabitants of western Rough Cilicia exploited this strategic importance successfully to mediate over time a favorable position for themselves vis-a-vis various competing external world empires.One of the questions we are exploring is the possibility that prior to the era of the pirates, Mediterranean core polities were either unwilling or unable to commit the necessary levels of manpower and resources successfully to exploit Rough Cilicia's valuable stands of cedar forests and maritime supplies.Both the limited availability of cultivable terrain for subsistence agriculture and the hostility of pastoral populations in the nearby highlands essentially impeded the development of coastal settlements until the era of the pirates.As a result, regional "urban" development in western Rough Cilicia lagged behind that in nearby Lycia, Pamphylia, and Flat Cilicia through the late Hellenistic era.

By offering a form of asylum to maritime labor under stress, we hypothesize that the anarchically democratic societies forged by pirate bands along this coast (and their relatively peaceful relations with the inhabitants of the interior) encouraged the immigration of skilled manpower necessary to initiate more labor-intensive activities such as forest harvesting, shipbuilding, metallurgy, and weapons manufacture.[40]In other words, that owing to the unique character of their social organizations the pirates may have succeeded where hegemonic polities failed at creating the population base and the necessary pool of skilled labor for incipient regional development to occur.Assuming that forestry resources were the region's chief attraction,[41] then we would expect to see a dramatic and progressive acceleration in the process of deforestation and site proliferation in western Rough Cilicia from this time onward.This acceleration ought to be measurable in archaeological terms through a progressive advancement of sites from the coastal areas inland along various river canyons as forestry supplies most proximate to the sea were consumed by harvesters.In other words, regional development ought to be characterized by a process of progressive deforestation further and further inland, as well as by an adaptation of now deforested areas (primarily the narrow coastal plains and coastal ridges) to alternative, agricultural forms of exploitation, including viticulture, olive production, grazing, and, where suitable, grain production.Evidence for this latter aspect of the progression ought to present itself in the form of small highly dispersed agricultural sites (farmsteads and villages) spatially removed from the urban centers, and identifiable by their locations (usually isolated on hilltops), their architectural remains (generally domestic, utilitarian, and non-monumental), and their associated assemblages of artifacts.The last might include a predominance of locally produced coarse wares, amphoras, and loom weights among the ceramic remains, alongside featural indicators such as press stones, grain-storage silos, and evidence of land terracing.Additionally, "sherd scatters" in otherwise "unworked terrain" (i.e., "off-site" locations) may likewise indicate the adaptation of deforested land units to agricultural use through practices such as "manuring."The temporal indicators of these rural assemblages (ceramic typologies and architectural construction techniques) form one means by which to date the progressive development of this region. Steepness of slope, soil depth, water courses, and similar geomorphological features are also evaluated in order to estimate the productive potential of past agricultural and other land uses.Again, what the theory calls for is a minimal phase of development prior to the pirates and a highly accelerated one from the time of the pirates onward with a deceleration at some point toward the end of antiquity.

To recover field data capable of evaluating this theoretical framework, the Rough Cilicia Archaeological Survey Team has employed a fairly consistent set of field methodologies depending on the character of the terrain under investigation.The primary role of systematic surface survey is to locate, to describe, and to analyze all sites within a selected region and to correlate this data with corroborating evidence of land utilization in the surrounding and intervening landscape (off-site evidence of human landscape disturbances).Sites of all periods within the surveyed regions are recorded and described in order to reconstruct patterns of long-term social change and to contextualize any given social system by reference to systems preceding and following it.At urban and large rural settlements with significant surface remains we have employed relatively intensive and time-consuming methods of ceramic and architectural data collection.In previously unexplored rural terrain, however, we have opted for a less intensive “course interval” methodology that enables us to find and to identify site locations rapidly over extended tracts of terrain.Also, in rural terrain previously subjected to course interval survey we have experimented with a more intensive, close interval field survey capable of extracting off-site land use data from seemingly featureless terrain.

Owing to the extreme difficulties posed by the burgeoning urban development of the Alanya vicinity, the team decided at the outset of the survey to delay work in the northern portion of the survey zone and to concentrate our efforts instead on the more open agricultural terrain in the zone's southern half, where a higher concentration of urban sites existed (see figure 2).[42]As a result, our work zone restricts itself to the watershed areas of Gazipasha, Güney, and the intervening coastal hills (from the ancient site of Iotape to Antiochia ad Cragum).[43]The only exception to this decision is the "emergency salvage" work the team conducted in 1996 at the magnificent, yet heavily looted mountain site of Laertes and at the equally vulnerable "Syedra Kiln Site," to be mentioned below

Another difficulty to confront the survey team concerns the vast scale of the urban centers--primarily architectural remains and dense concentrations in surface sherd scatters--and the need to devise a proper strategy for obtaining worthwhile data with a field team limited to ten persons or less.Using sophisticated electronic survey equipment,[44] the architectural team, Rhys Townsend, Michael Hoff, and Jason DeBlock, attempts to map and to record architectural data, ranging from large structures such as fortification walls, theaters, aqueducts, baths, and gymnasia to smaller structures such as tombs, cisterns, and "farm houses."[45]Even with the aid of equipment such as this, the three-member team inevitably labors amid dense and extremely thorny scrub brush to locate the remains of walls, joins, and returns, and to identify and to isolate these remains where possible into successive chronological phases of construction.Under normal circumstances the process of mapping a single urban site such as Antiochia ad Cragum (with architectural remains extending over 15 hectares) would require sustained work of several seasons. Given the time-restraints of the regional framework of the survey, it is virtually impossible for a team this small to record and process every architectural facet of each individual site. Consequently, the team has agreed to restrict the architectural work to those features identified as the most significant and/or representative of a given site.The uniqueness, the accessibility, and the surviving state of architectural features within a given site equally form this determination.At each urban site, for example, comparisons are initially made with plans available from previous architectural surveys, particularly those of the Rosenbaum survey,[46] to determine whether the existing plan accurately reflects the state of the surviving remains.Assuming that the existing plan is accurate, the architectural team attempts to build on this plan by recording and mapping previously undetected components to the site, such as the late Roman village and fortification system at Selinos or the domestic quarter at Iotape.In places where the previous plans are deemed inadequate, such as the gymnasium at Laertes and the colonnaded street of the "upper city" at Antioch, the architectural team has constructed new, more accurate plans.

At rural architectural sites, all of which are previously unexplored, the decision to construct a plan is usually determined by the uniqueness and importance of the structural remains, as well as by their accessibility and their relative state of survival.[47]In this manner the architectural team has worked to isolate and to identify representative typologies of architectural remains with the objective being to complete detailed plans for a broad range of architectural complexes, from large urban settlements to isolated rural “farmsteads.”The objective ultimately is to assign typologies, chronological classifications, and functional priorities to an entire range of vernacular architecture in Rough Cilicia, not only to the architectural features of the urban centers, but also to the vastly more difficult to characterize remains of the non urban sites.[48]

In a similar manner the team has repeatedly adapted its methodology to cope with the diversity and significant densities of ceramic remains in the survey zone, particularly the rich assemblages of Roman and Later Roman ceramics encountered at many of the urban sites.Making matters more difficult, a communication received from the General Directorate of Monuments and Museums during the 1997 season officially prohibited Dr. Ismail Karamut, the Director of the Alanya Archaeological Museum, from receiving and storing large quantities of our survey pottery at the museum.[49]The process of ceramics recovery and identification at urban and/or large rural sites had to be modified, accordingly, and has generally proceeded as follows.At urban and large rural sites, as soon as site limits are determined ceramics "grab collections" are conducted within designated, paced off “collection areas” (approximately 100 m. square and typically demarcated by visual landmarks).The field directors record the locations of collection areas in field notebooks as well as on 1:5000 ratio topographical maps, with a GPS location taken, where appropriate, for each collection area.[50]To accommodate the 1997 restrictions on storage, Rauh redoubled the team's efforts to restrict and to control the size of the grab collections, particularly at the urban sites where literally tens of thousands of sherds are visible amid the remains.Working in separate areas of the collection areas, team members bring their grab collections to a central location where Rauh proceeds to triage them down to a representative sample of what is typically being recovered in that area, along with any specimens of unique, unidentified, and/or previously unencountered sherds.[51]In addition, in order to maintain a certain level of accuracy to these ceramics collections, Rauh attempts personally to participate in the ceramics collections at each and every site collection area, going so far as to return to areas examined during his absence to conduct "recollections.”In this manner the team has worked to ascertain that the final collections adequately reflect the full range of the extant ceramic remains in a given area.[52]Large urban sites such as Selinos and Antioch can require as many as 16 collection areas to complete, significant rural sites generally two to four.Figure 4, for example, displays the 1997 field map of the Collection Areas for the "upper city" at Antioch, including Collection Area (CA) 4, to be discussed below.

The triaged collections are then bagged, tagged, and brought to the laboratory at the Selinus Hotel in Gazipasha, where every sherd is cleaned and coded.Pottery specialists Richard Rothaus and Kathleen Slane then proceed to "read" or process the sherd collections according to their functional, diagnostic, descriptive, and temporal characteristics and to tabulate their data in the project's Chronotype Ceramics Database.[53] Figure 5 presents, for example, the preliminary Chronotype data table for Collection Area 4 at Antioch (mentioned above), showing the chronological breakdown of the ceramics collections (grouped according to typologies) triaged and recovered in an area to be discussed below.Although the ceramics data we have produced for the urban sites are severely reduced from a quantifiable perspective, they do present a highly accurate chronological analysis of the pottery recovered from a systematically organized and narrowly delineated set of spatial locations.Since each grab collection furnishes potentially crucial information regarding the surface remains of a site-specific spatial area, each collection is processed, photographed, and treated as a separate "archaeological context."

By employing this method the walking and ceramics laboratory teams were able in 1997 to process more than 4000 sherds from six urban sites and more than 40 identified "rural" sites and "off-site" land units (transects).[54]All the ceramics data have been digitally recorded in two ceramics databases that are continually updated and revised.In addition to the chronotype database, Rothaus and Slane assembled the "Rough Cilicia Survey Pottery Study Collection" consisting of more than 300 useful diagnostic sherds of the most commonly observed forms of the region, based on the collections of the survey team.This study collection is made accessible in the field laboratory with labels displaying type, form, and use-chronology at the outset of every season, to serve as a field reference tool.A fully digitized version of this study collection is nearing completion, and will eventually provide a thorough description for each sherd, including Munsell soil colorations, dimensions, fabric and slip descriptions, profile drawings, photography, and available comparanda.[55]

These methods have enabled us to digest representative quantities of architectural and ceramics data in the urban areas thus far examined.For the rural areas the team employs slightly modified methodologies.With the exception of the terrain covered during the 1998 season, the Rough Cilicia Archaeological Survey team has generally investigated the non-urban terrain within the survey zone by employing course interval field methods.The primary goal of course interval survey is to locate, to document, and to obtain preliminary temporal information for architectural remains, both small and large, while surveying more extensive areas of terrain.High sherd densities are also designated as sites and, as with larger sites, the ceramics there are triaged and inventoried.To meet this goal, a crew of five to seven members, spaced at 15-to-25-meter intervals and in radio communication, walk down-slope transects, reporting architecture and sherd findings by radio.Archaeological remains are considered a site if architecture is encountered or when two or more individuals report one or more sherds over a 5-meter distance.Once so designated, sketch maps are made of the location, a site form is completed, GPS readings are recorded, ceramics are collected, triaged, processed (and if warranted, bagged and brought to the lab), and the site location is photographed.[56]By employing course interval survey as a base method for collecting "rural" data, the team has surveyed approximately 120 sq. km. of coastal terrain from Iotape to Antiochia ad Cragum, a distance extending more than 30 kilometers over an area of varying width, between 2-10 kilometers.[57]Through 1999 the walking team has identified more than 50 “rural” sites, as well as numerous additional “off-site” areas of human disturbances within the survey zone.

When LuAnn Wandsnider assumed responsibility as project field director during the 1998 season, she successfully implemented a more intensive survey methodology that emphasizes human landscape ecology.Under her direction the team intensively surveyed 21 transects comprising more than 17 linear kilometers of survey terrain where the walking team had previously conducted course interval survey.The 1998 systematic field survey was designed to develop and to evaluate a survey method that would incorporate evaluation of landscape human ecological questions by implementing an array of new strategies.These include collecting information on ceramic densities from differentially located landscape elements; collecting information on ceramic sherd attributes sensitive to deposition history; collecting whatever information can be gleaned about the differential distribution of ceramics with different functions; and assessing the difficulty in defining the landscape elements themselves.[58]

To address these objectives, data were collected in 1998 according to a protocol implementing a higher level of systematization, yet, by necessity one that investigates more limited areas of terrain.[59]Survey team members were responsible for walking a linear traverse and inspecting an area approximately 1.5 m wide (Figure 6).[60]Ceramic sherd totals were noted on the Survey Unit Description Form for individual transects along with the total length of the area traversed.Individual survey members picked up all temporally diagnostic sherds along their paths, with individual collections then being assembled to form a transect-unit "Grab Sample."In addition, a designated team member collected all ceramics along his traverse, referred to as the "Systematic Sample."Sites were defined as locations with pre-Ottoman architecture, however unspectacular, with at least two articulating walls. Features were defined as anthropic, non-portable phenomena that did not meet the definitional criteria for architecture.In general, wall alignments inferred to be terrace or retaining walls were identified as features.

The data collected during the 1998 season is, accordingly, highly detailed and far more capable of detecting field biases, such as the tendency of Hellenistic era ceramic sherds to be less obtrusive in the field and, as a result, to be potentially under-represented in survey results.[61]Nevertheless, our utilization of intensive systematic survey methodology has been restricted thus far to the 1998 season and to a limited portion of the terrain otherwise covered by "course interval" survey.At the current time the results from intensive interval survey are insufficiently formulated to furnish any worthwhile information to the pirate question.

For purposes of identifying piracy in this region other survey methodologies would certainly be useful, including marine archaeological survey of neighboring waters, meteorological analysis of the region's prevailing winds and currents, and geoarchaeological survey of sedimentation and pollen residues to determine chronological rates and sequences to the region's deforestation.[62]While recognizing the viability of these and other methodologies, the members of the Rough Cilicia Archaeological Survey Team have for the time being worked within the limits of available resources, permit authorizations, and assembled expertise.Given the largely unexplored character of the landscape of this region we regard our work as an essential first phase to the regional exploration of Rough Cilicia.While recognizing and acknowledging the limitations of the work we have completed thus far, we remain confident that the architectural, ceramic, and land use data we have assembled for western Rough Cilicia is sufficiently detailed to advance our awareness of the archaeological history of the region.

Archaeological evidence for piracy

Having said that, we can also assert that depite having surveyed 6 urban and more than 70 rural sites and land units, we have yet to identify a single sherd of a republican-era amphora from Italy anywhere in the survey zone.To dispel the notion that this finding results from some oversight on our part (from poor observation, for example, or from some failure in our survey technique), we should add that we have in fact recovered a minimal number of temporally coeval amphora sherds in the survey zone. These include the Classical era Phoenician and Chian sherds noted above, two Hellenistic Rhodian amphora handles (one stamped), and one apparent Hellenistic Koan handle attachment (also stamped).In addition, the survey team has recovered a rather significant quantity of "western Mediterranean" amphora sherds datable to the early Roman era (first to third centuries AD), including numerous Tripolitanian amphora rims from north Africa (extremely common), several Dressel-20 transport amphora rims and handles from Spain, and one Dressel 6 transport amphora handle from Italy.The combination of finds in these two categories--Classical/Hellenistic amphora remains from the Aegean and eastern Mediterranean and early Roman amphora remains from the West--argues strongly against the possibility that we have somehow inadvertently "missed" surviving Roman republican materials from Italy, particularly with Rauh devoting so much of his field energies to the search for these very sherds.Assuming that any of the settlements surveyed thus far (including Antiochia ad Cragum) were in fact inhabited by Cilician pirate bands, therefore, the absence of any corroborating ceramic evidence would appear to indicate that André Tchernia's "wine-for-slaves" thesis is invalid for this region.Of course, another possibility is that the trade lines of Cilician piracy were far more complicated than Tchernia’s hypothesis allows.

A related observation warrants discussion, however tentatively.The use-chronologies of temporally diagnostic Hellenistic ceramic remains recovered thus far appear generally to predate the era of the pirates.[65] Admittedly, this conclusion arises from an extremely limited sample.Were it not so preliminary, one might argue that the era of the pirates was a highly depressed one for the survey region, since it appears to have experienced little if any importation of externally produced goods (either as a result of trade or plunder).[66]

Turning to the architectural data of the survey thus far, one finds that these data provide similarly confusing results, in large part because the process of isolating and identifying pre-Roman chronological remains has proven extremely difficult.There are at least three reasons for this.One, the construction techniques employed in western Rough Cilicia generally do not conform with those observable in the neighboring regions of Lycia or, more importantly, Pamphylia.Rather, the natives of this region appear to have employed their own highly distinctive "vernacular" forms of construction that severely impede efforts to assign datable phases to building constructions based on external comparanda.Nearly all of the features that "could be" Hellenistic--such as the lower courses of roughly hewn "ashlar blocks" incorporated into the late Roman citadel atop Selinos (figure 9 and 10), the rock-cut "odeion" or "council house" at the same site (figure 11),[67] a similar, smaller rock-cut structure atop Nephelion, and the irregular “terrace” walls at Iotape (to be mentioned below)--do not resemble similar, more readily identifiable Hellenistic styles of masonry at Perge, Termessos, Sillyon, or Side, for example.

Two, the single greatest obstacle to the identification of pre-Roman construction phases is the omnipresent employment of mortar in regional building construction.A fine grained, lime-based mortar is detectable in the construction of nearly every structure inspected thus far in the survey zone, including the large "ashlar-like" block construction in the lower courses of the Selinos citadel.In view of the traditionally recognized date for the advent of rubble and mortar construction techniques (particularly in Anatolia),[68] it becomes difficult under these circumstances to assign Hellenistic or pre-Roman dates to these phases of construction. This is true even for those instances where pre-Roman ceramic evidence, such as the Classical/Hellenistic black slipped sherds recovered near the walls of the Selinos fortress, or the Rhodian and Koan amphora sherds recovered at the fortified hill site (28-C-3-d-4,to be mentioned below), are recovered in the same spatial context.[69]The members of the survey team continue to wrestle with this question.   Mortar remains a highly inappropriate device for joining large-block modes of construction.Conceivably, the mortar visible on these remains was applied to the interstices of the ashlar blocks at some later date, for example, during the Roman or late Roman eras when builders superimposed elevated phases of mortar-constructed defensive wall to the preexisting courses of "ashlar" masonry.[70]In this respect it is important to note that the 8-9 solid courses of large, ashlar block construction visible in the Selçuk era fortress at Alanya (Coracesium), even though generally recognized as Hellenistic era construction, likewise exhibit mortar along the interstices of their blocks (figure 12).[71]

But this question leads to the third, and possibly more important observation; namely, that far too few extant remains of "pre-Roman" construction exist in this region, furnishing far too inadequate a data set with which to assign dates to phases of wall construction in any confident manner.Moreover, what few examples of mortar-free or "large block" construction do present themselves at Iotape, at Selinos, or elsewhere are invariably overshadowed by a massive overlay of Roman-era construction.At Selinos, for example, the acropolis citadel and the "odeion" may contain pre-Roman elements of construction, but these are dwarfed by later, superimposed “additions.”They also pale in comparison with the large rectangular porticoed complexes on the valley floor[72], including an aqueduct and bath complex, a porticoed complex with a raised platform structure at the harbor, a seawall that appears to extend from the “harbor” to a Roman-era interior basin at the site, a mortuary precinct exhibiting a minimum of thirteen tombs,and last but not least, a heavily fortified late Roman village with an imposing system of fortification walls with a built-in network of cisterns.[73]Amid the overlay of these subsequent constructions, it becomes extremely difficult, if not impossible, to attempt a reconstruction of the featural remains of the pre-Roman settlement at Selinos.The likelihood that Roman and late Roman-era builders employed extensive landfill to create immense earthen platforms for these complexes, landfill that conceivably conceals earlier phases of site construction, obviously renders the problem more difficult.[74]With these points in mind, it is prudent to recall the description Livy offers for this region during the year 197 BC.Apart from Korakesion and Selinos, settlement in Rough Cilicia was restricted to a number of castella, or small forts.[75]The appropriateness of this description may very well explain the paucity of pre-Roman featural remains that the survey team has encountered not only at Selinos but throughout the survey zone.

When we apply these findings to Antiochia ad Cragum, a site bearing directly on the question of piracy, the difficulties posed by the absence of pre-Roman material remains grow acute.Although the featural remains of this site date predominantly to the period of resettlement by Antiochus IV of Commagene c. 52 AD, there seems little doubt that this was the location of the Cilician pirate base described by Appian.[76]In fact, Appian's topographical description of this phrouria megista appears remarkably site specific.His two impregnable citadels--the Kragos and the Antikragos--closely match the site's actual topography, combining an acropolis-like "upper city," perched high atop a precipitous seaside cliff (some 300 m. in altitude), with a lower more heavily fortified bastion on a rock promontory flanked by hidden sea coves (FIGURES 13-16). The relatively undisturbed character of the remains seemed to offer our best opportunity with which to test theoretical assumptions about piracy.However, several weeks of investigation at Antioch revealed practically no pre-Roman ceramics remains, let alone any vestiges of Roman republican amphora remains from Italy.[77]In particular, ceramics collections conducted in the vicinity of the Byzantine castle and hidden cove of the lower city produced an overwhelming mass of Byzantine ceramics dating to the 10th-12th centuries AD, with little if any deposits of preceding eras.[78]While a modicum of evidence exists to suggest pre-Roman habitation at the site’s “upper city,” therefore, practically none has survived at the so-called "pirate's cove" or at the fortress of the "Antikragos" below.To make matters more difficult still, from an architectural standpoint none of the surviving structures investigated thus far exhibits any obvious pre-Roman phase of construction.

One hesitates to draw any further conclusions about the pre-Roman settlement of the site while the data remain incomplete.At an identified cliff-top structure at Collection Areas 4 and 5 (noted above), for example, the walking team recovered a handful of Hellenistic sherds, conceivably emerging along the margin of the land terracing in this vicinity.Directly behind this structure (inland) stood Antioch’s sizable gymnasium complex, complete with enclosed porticoes and surviving two-story bath structure.Evidence of terracing in the vicinity indicates that perhaps as much as 3 meters of early Roman fill was employed to create the earthen foundation for this complex.Indeed, evidence of extensive terracing and human landscape disturbances characterizes not only the ancient remains of this site, but the modern structures of the inhabited (Güney) village as well.[79]Conceivably, pre-Roman remains lie below this fill.At Antioch, therefore, the information gathered thus far presents a highly fragmentary, uncertain picture of “pirate habitation."

That a pre-Roman settlement existed at this site seems reasonable; whether we will ever possess sufficient data to characterize it remains to be seen. Recent discoveries suggest, however, that the limitations of our initial theoretical models result in large part from our effort to evaluate internal developments in western Rough Cilicia according to a gradual assimilation of outside influences.Since the evidence thus far appears to demonstrate that western Rough Cilicia was under-populated and externally isolated during the pirate era, our attempts to measure regional development according to surviving quantities of externally produced goods may very well be inappropriate.Countervailing locally generated trends were perhaps more significant during the pirate era. For example, a greater adherence to local and regional cultural attributes and a heavier reliance on vernacular modes of production, including crude building techniques, small rugged forts, a heavy reliance on locally or regionally produced coarsewares, and perhaps most importantly, a strong connection with the more dominant cities of Pamphylia may be detectable.Rather than to look for indications that the inhabitants of this region engaged in exchanges with distant polities such as republican Italy, what we need to determine is the degree to which the inhabitants of western Rough Cilicia participated in or were influenced by an projection of Pamphylian hegemony throughout this bay.To understand how the evidence possibly supports this last interpretation, we need to evaluate three additional aspects of the archaeological record--the emerging evidence of local amphora production, our recent investigation of fortified hill sites, and the complicated record that is emerging for the site of Iotape.

One of the more interesting developments of the survey is the discovery of several possible kiln sites producing internationally identifiable forms of transport amphoras.[80]For purposes of the discussion of piracy the most significant are the Pamphylian amphora remains found at the so-called “Syedra Kiln Site” near the mouth of the Syedre River.During a preliminary visit in 1995 museum authorities at the Alanya Archaeological Museum brought to our attention the existence of these kiln remains on a beach a few kilometers south of the ancient site of Syedra.From the highway the site stands behind a farm on a small, annually plowed agricultural terrace directly abutting a campground complex to its north. The terrace (approximately 67 m. N/S by 43 m. E/W) stands more than a meter above the beach and neighboring agricultural fields to the south and is visibly retained on its south side by the remains of an ancient wall.Remains of ceramic floor tile, set in concrete, are likewise visible amid scattered debris.In July 1996 the field team conducted an intensive survey of the site, gathering a "triaged" collection of its representative ceramic forms.Elements of the team revisited the site twice in 1997 to gather additional context pottery.Richard Rothaus and Kathleen Slane analyzed these collections that season, with the extremely preliminary assessment that a kiln site probably existed here.It must be stressed that no surface remains capable of verifying the existence of a kiln structure are visible.In addition to the limited topographical indicators mentioned above, our conclusion that an ancient kiln stood here is based on the discovery of misfired and vitrified material including three pieces of kiln lining and two roof tiles.Items still usable but over fired gray and metallic and also likely to be products of this site are two shapes of basin rim, several loom weights or heat pads, a stewpot with everted rim, one or two amphora toes and about ten wall fragments of coarse vessels or amphoras.Numerous normally fired vessels in these shapes were also collected, particularly the basin and amphora fragments.

The amphora remains include a large bowing handle (oval in section, preserved length 9-10 cm.) and a short tapering spike (H. 7-8 cm.) with rounded or button end that closely resemble forms of Pamphylian amphoras published by V. Grace and K. Slane (SEE FIGURES 17-18).[81]Based on the similarities with those forms, we have tentatively identified these amphora remains as Pamphylian. Similarities in fabric between the amphora sherds and those of the site's basin and loom weights enable us to suggest that the amphoras were locally produced.Context pottery collected with the wasters and amphoras is mostly Hellenistic or early Roman, suggesting that the predominant phase of the site was late Hellenistic and early Roman.[82]

Pamphylian amphoras (use-chronology, first century BC through the third AD) are so-called because of the ethnicity of names recorded in their amphora stamps, not to mention the possibility that two of the stamps recovered at the Athenian Agora refer specifically to the city of “Aspendos.”[83]Although Pamphylian amphora remains have been identified at a number of eastern Mediterranean sites, to our knowledge no kiln sites have been identified for this amphora in Pamphylia proper.[84]Their relevance to the question of piracy is potentially great.French excavators recovered high concentrations of Pamphylian amphoras during the excavation of the “Maison des Comédiens” at Delos, prompting V. Grace and Savvatianou-Pétropoulakou[85] to suggest that the “pirate armada” of Mithradates VI of Pontus conveyed these vessels to Delos during the assaults on the island in 88/87 and/or 69 BC.Were the forms “produced” at Syedra to match those found at Delos, they would furnish a crucial indicator of a material cultural transference between Delos and Cilicia in a direction entirely contrary to that proposed by Tchernia.

Unfortunately, the published form most closely resembling our sherds is extremely difficult to determine from the remains of mere handles and toes: some of our handles appear to curve like Virginia Grace's form 3, others are straighter like form 13.While the context pottery allows for the possibility that these forms were produced at the time of the pirates, it would be hazardous to insist on such a chronology.For the time being we must content ourselves with the fact that Pamphylian-styled amphoras were apparently produced in this region.This in and of itself suggests that inhabitants along the border of western Rough Cilicia (some 20 km. east of Alanya) strongly identified with cultural attributes emanating from the urban centers of Pamphylia.[86]Although the remains at the Syedra Kiln Site probably date to the early Roman era, they serve as an important indicator of the influence Pamphylian cultural centers wielded over distant smaller communities along the periphery of the bay.This, in turn, suggests that the settlements in western Rough Cilicia functioned as satellite communities and as sources of agricultural produce and raw natural resources for urban centers more than 150 kilometers away.[87]

Our growing perception that the settlements of western Rough Cilicia functioned as satellite communities to the core populations in Pamphylia is further supported by the discovery in 1997 and 1998 of several fortified hill sites along a stretch of the coastal ridge between Kestros and Nephelion.The largest of the investigated sites is situated atop Guda Tepe, the next peak south of Kestros (Map 28-c-03-a&d, Site 4; figure 19).The upper reaches of the tepe consist of a gentle slope upon a narrow flat area, with architectural remains extending along the upper region and east and west slopes for more than 100 meters (SEE PLAN, figure 20).The character of the structures falls essentially into two types: defensive and non-defensive.Along the northwest slope are several well-constructed rectangular towers, apparently connected by walls.At the south extremity of the site is a large round tower, constructed with massive roughly ashlar blocks.Around the southwest rim of the site is a smaller circuit wall that joins the round tower.On the eastern side the high bedrock and steep slope may have made defensive fortifications unnecessary, as there is no evidence of wall here.[88]The other structures, likely houses, were built on terraces of walls along the east and west slopes.One structure is particularly well preserved to warrant mention--a two-room structure (house?) on the west slope whose walls still surpass 2 meters in height, complete with its lintel in situ.[89] Pottery collected at this site includes one Rhodian stamped amphora handle (now in the Study Collection); a possible (stamped) handle attachment of a Hellenistic Koan amphora, a few fragments of Hellenistic black slipped fineware, and a large number of Early Roman fineware sherds.[90] Although the predominant era of the pottery is Early Roman, this site furnishes the most compelling Hellenistic ceramics assemblages of any of the rural sites investigated thus far.

The architectural data of the site remains complicated: two phases of construction appear to exist, one requiring fortification walls, the other conceivably domestic in character. Based on the surviving ceramic evidence, it would be tempting to associate the fortified phase of construction with the Hellenistic ceramic remains and the later, "domestic" phase with the Roman and/or late Roman remains.If so, this site and one or two neighboring ones along the ridge offer themselves as potential confirmation of the “castles and fortresses in the Tauros mountains” that Plutarch (Pomp. 28.1) asserts were constructed by the Cilician pirates to house their families, their disabled warriors, and their treasure.

Other, equally valid interpretations need to be considered, however. The potentially early second century BC dates of the Hellenistic ceramic remains may indicate, for example, that the fortifications were constructed by settlers prior to the advent of the pirates in order to defend themselves against indigenous populations as they pioneered their way out of the Selinos valley. Then again, by weighing one’s interpretation on the presence of mortar in the fortification walls, one can conceivably argue that the defenses were constructed by inhabitants during the early first century AD, when the sources report repeated waves of assaults by marauding Isaurian tribal elements of the Cilician interior.[91] Under this scenario the Hellenistic ceramic remains could be attributed to some pre-architectural, or nomadic settlement at this location.Given the present state of our knowledge, based as it is on limited remains recoverable on the surface, any of these interpretations is conceivable. The site remains useful as a typology for rural settlement in western Rough Cilicia at the time of the pirates, and a scenario that bears watching, since it is supported by the discovery of several additional fortified hill sites in the survey zone.[92]

The remains of maritime site of Iotape, located approximately 9 km. northwest of Gazipasha and conspicuous from the modern roadway that cuts directly through the site, provide a second potential useful direction for the investigation of pirate remains in western Rough Cilicia.According to the available epigraphical evidence, the site was founded by King Antiochus IV of Commagene in the early first century AD and named after his daughter Iotape.[93]Although there is nothing in the textual record to demonstrate the existence of a preexisting site at this locale, as Table I indicates above, grab collections conducted in 1996 point to the existence of a Hellenistic settlement at this locale.

The architectural data is equally complex.The Rosenbaum survey plan of the public structures between the modern road and the sea shows two buildings, identified as baths, and two temple platforms (figure 21), all of which are located to the east of a small stream that divides the site before spilling into the bay.To the west of the stream the plan includes the circuit wall of the citadel and the line of an ancient roadway flanked by several large honorific statue bases (figure 22).On the higher ground across the modern road, the position of numerous tombs as well as scant wall traces of no discernible pattern were noted.[94]

The architectural team of the survey concentrated its efforts in 1996 on the citadel and the area of higher ground inland.[95]In each case, we were able to draw extensive surface remains of domestic architecture that provide a more representative view of the site.

The citadel wall encloses a warren of small structures (Figures 23-24).Today this area of approximately 2000 square meters is densely overgrown and obstructed by fallen rubble.Nevertheless, in two days of survey we were able to identify at least 25 rooms.Traces of stairs and vaulting demonstrate that several of these had more than one story.This evidence, together with the repeated use of shared walls and narrow passageways, shows that the citadel should be reconstructed as a honeycomb of densely packed inhabitation.

Across the modern road, our survey recorded 26 previously unrecognized buildings, mostly houses in the area north and west of the small medieval church that earlier investigators believed to lie at the northernmost extent of the site (figure 25).Here we were able to discern not only individual houses, in many cases with walls standing a meter or more in height, a few even preserving doorjambs, but lines of streets as well (figure 26).As in the case of the citadel, these structures are to be considered as a representative sample rather than an exhaustive recording of the preserved remains.

In this instance, two periods of construction are clearly identifiable in the remains: in the first period, local limestone quarried from the hill itself is roughly cut into square or nearly squared blocks that are joined in a rough ashlar fashion without mortar. In Period II, rubble masonry of small stones set in a cement mortar is employed.Most often this technique is found as repair or addition to structures of Period I.On the citadel construction is represented primarily by Period II, with the addition of a third technique, probably a separate period, in which fragments of brick are added as part of the rubble construction.

Together with our study of the architectural remains, preliminary analysis of the Hellenistic and Late Roman sherds recovered during our grab collections suggest that the site was occupied during the Hellenistic era (that is, prior to its supposed founding by Antiochus IV of Commagene, c. 52 AD).It appears to have ceased during the Late Roman era and to have been reoccupied during the early Byzantine era.[96]

Sites such as Iotape leaving no textual or epigraphical record of habitation prior to the Roman era but clearly demonstrating pre-Roman material remains present themselves as areas of potentially great significance to the piracy question.At the very least the ceramic evidence indicates the existence of a Hellenistic settlement at this site. The ceramics evidence is supported by the survival of what appears to be examples of mortar-free construction in the remains of the “domestic quarter” as well as by the possible reuse of large rectangular stone blocks in phases of construction in the citadel. Other features at the site warrant revisitation, for example, the remains of walls employing very large irregularly cut blocks in their construction (figure 27). Also, high on the ridge behind the site remains of several “satellite” villages await systematic investigation.

Bearing in mind the lessons learned at Selinos and Antioch, it is possible to suggest that a small, fortified site existed at this location during the Hellenistic era, the remains of which became heavily superimposed by the monumental constructions of Antiochus IV’s early Roman foundation.Although the extensive reuse and remodeling of this site during the Roman and later eras greatly obscure the extent of the Hellenistic era settlement, certain minimal conclusions can be drawn.If there were pirates along this shore, their settlements, based on the surviving surface remains, appear have been extremely small (with the possible exception of Korakesion), and their locations (high on the coastal ridge, or nestled on promontories that sheltered hidden coves) highly furtive.Sociological descriptions for both ancient Mediterranean and modern transatlantic piracy suggest that small-scale settlements of this sort were appropriate to pirate societies.In the final analysis, the reason for our inability to locate and to identify the remains of Cilician pirate bases may ultimately arise from the fact that the pirates themselves did their utmost to evade detection by potential foes.

Conclusion

However, the evidence for piracy generally is limited and greatly obscured by Roman and later constructions at these sites. In addition, what minimal Hellenistic era remains do survive at sites such as Selinos and Nephelion indicate that these settlements were extremely small at the time of the pirates and that they existed perhaps as little more than fortified waystations.This may have exposed their inhabitants to potential seizure by pirate bands, a phenomenon which is known to have occurred elsewhere (Plut. Cim. 13.3; Cic. Verr. 2.4.21).This, however, is something unlikely to be detected from the material remains.

What the recovered survey data for western Rough Cilicia does suggest is that the settlements along this coast were extremely small and frequently hidden from view.Some are situated as fortified hill sites high atop the coastal ridges, others like Selinos possibly settled on the less visible land ward flanks of coastal promontories, and still others such as “the Kragos” and Iotape nestled amid rock precipices and hidden sea coves.Topographically, Hellenistic era settlement in western Rough Cilicia presents itself as a series of small, fortified, and extremely well camouflaged sites along the coast.Rather than assimilate external cultural attributes or externally produced wares seemingly available from passing mercantile traffic, their inhabitants appear to have relied more on vernacular modes of construction and locally produced coarsewares.[99]In addition, the evidence of a possible kiln site producing Pamphylian-styled amphoras near Syedra suggests that the settlements along this coast looked to the urban centers of Pamphylia for cultural, economic, and political leadership and that they serviced these centers as satellite communities generating natural resources and agricultural produce.

On the face of things, Pamphylia, rather than western Rough Cilicia, would appear to have existed as the core to Cilician piracy.The cities of this coast not only acquired a reputation but were repeatedly punished for colluding with the pirates.It is important to note, however, that the extent and degree to which individual Pamphylian communities participated in the pirate menace was never monolithic.   It probably varied over time as well as by location.[100]

Overall, the behavior of these various actors in the pirate saga seems peculiarly "cat-and-mouse.”Despite the many difficulties the pirates are credited with creating, they appear never to have succeeded at mounting a genuine military threat against the dominant polities of the Mediterranean.They roved a pirate round from Phaselis to Dianium in Spain.They raided, looted, and plundered Aegean maritime cities.They surreptitiously attacked the Italian harbors of Ostia and Caeata, setting fire to sitting cargo and war fleets.They even ventured inland to engage in scortched earth tactics and high profile kidnappings.However, as far as we can tell, they never once defeated an armada fully deployed by a legitimate state such as Rhodes or Rome.Their tactics were more furtive, more "hit-and-run" in character.Appian (Mith. 92) comments that the pirates selected, "forts and peaks and desert islands and retreats everywhere, but they chose for their principal rendezvous the coastal strip of Cilicia that was rough and harbor-less and rose in high mountain peaks."He adds, moreover, that that the furtiveness and unpredictability of the pirate menace made it extremely difficult to address.[101]In the best traditions of pirate strategy, they kept their bands deliberately small, their bases hidden, and their movements frequent in order to evade detection.[102]The impact of such a behavioral pattern on the archaeological landscape need only be imagined.

The Romans and their allies appear to have recognized that cities such as Side and Aspendos secretly collaborated with pirates.There are indications that the Romans attempted to "quarantine" these cities, and when all else failed, to single them out for punishment as “rogue states," in much the same manner that the modern Western industrial nations confront the behavior of states such as Libya, Syria, Iraq, or Iran.Ultimately, this treatment took its toll, but the fact that the Romans had repeatedly to return to Pamphylia militarily suggests that as soon as their forces had departed, pirates hiding along the recesses of the Bay of Pamphylia reemerged.[103]

In fact, it seems particularly odd that the commanders of the first two Roman campaigns against the pirates (M. Antonius in 102 and P. Servilius Vatia Isauricus in 77-76) never directly confronted the obvious pirate menace at Korakesion, despite venturing as far as Side in Pamphylia.   Following his suppression of Zenicetes' piracy in Lycia and Pamphylia, for example, Isauricus suddenly adapted what had been a naval operation into a land assault on the Isaurian tribesmen deep within the interior of the Tauros Mts.One can legitimately question why Isauricus chose to ignore the infamous pirate settlement at Korakesion, with its "well stocked harbor, shipsheds, and skilled artisans and laborers chained to their stations." Some scholars reasoned that Isauricus' thrust into the Isaurian hinterland was intended to "soften up the neighborhood," clearing the way for some future assault against Korakesion.[104]Others legitimately argue that the Roman generals assigned to Cilicia were equally required to monitor the threat posed by King Mithradates to neighboring Cappadocia.[105] Another possibility, however, is that at the time of Isauricus’ campaign, the western Pamphylian empire of Zenicetes posed the greater and more visible threat to Mediterranean shipping.Conceivably, in 77-6 BC Roman authorities lacked adequate intelligence about the extent of the pirate menace in western Rough Cilicia; the settlements at Korakesion and the Kragos possibly had avoided detection as yet as the "common anchorage and military camp" of these marauders (App. Mith. 92).

It remains difficult to determine at what point before 67 BC the Romans were adequately informed of the threat posed by the pirate settlements in western Rough Cilicia.[106] Sites such as Korakesion and the Kragos were small, well hidden, and perhaps frequently relocated. With the entire Bay of Pamphylia at their disposal, pirate bands conceivably relocated several times from one end of the bay to the other during the course of the pirate menace's sixty-year life span.In time the members of the Rough Cilicia Archaeological Survey team hope that to produce effective means to answer these questions.The evidence collected thus far reminds us of the importance of furnishing an accurate regional context to a phenomenon such as the Cilician pirates and of the need to inform oneself of the sociological mainsprings to piracy before proceeding to identify pirate remains on the ground.

Bibliography

Atvur, Orhan

1999        Olympos Antik Kentı (1991-1992 Çalişmaları). Arkeoloji ve Sanat 88 (1999), 13-31.

Bagnall, Roger S.

1976        The Administration of the Ptolemaic Possessions Outside Egypt. E. J. Brill, Leiden.

Balsdon, J. P. V. D.

1939        Consular Provinces under the Late Republic. Journal of Roman Studies 29: 57-73, 167-183.

Bean, George E.

1979        Turkey's Southern Shore. London and New York, Norton.

Bean, G. E. and T. B. Mitford

1962        Sites Old and New in Rough Cilicia. Anatolian Studies 12:185-217.

Bean, G. E. and T. B. Mitford

1965        Journeys in Rough Cilicia in 1962 and 1963. Austrian Academy of Science, Vienna.

Bean, G. E. and T. B. Mitford

1970        Journeys in Rough Cilicia 1964-1968. Austrian Academy of Science, Vienna.

Bernhardt, Rainer

1971        Imperium und Eleutheria:die römische Politik gegenüber den freien Städten des griechischen Ostens.      U. Hamburg Diss., 1971.

BernhardtR.

1985 Polis und römische Herrschaft in der späten Republik (149-31 v.Chr.)Berlin, 1985.

Blackman, J.

1973        The Harbours of Phaselis. International Journal of Nautical Archaeology 2: 355-364.

Blake, M. E.

1947        Ancient Roman Construction in Italy from the Prehistoric Period to Augustus. Washington DC.

Blümel, W.

1992        Die Inschriften von Knidos, Teil I. Inschriften griechischer Städte aus Kleinasien, Band 41. Österreichische     Akademie der Wissenschaften, Rheinisch-Westfälische Akademie der Wissenschaften. Bonn:Habelt.

Blumenthal, E.

1963        Die altgriechische Siedlungskolonisation im Mittelmeerraum unter besonderer Berücksichtigung der     Südküste Kleinasiens. Tubingen.

Brandt, Hartwin

1992        Gesellschaft und Wirtschaft Pamphyliens und Pisidiens im Altertum. Bonn. Habelt.

Brixhe, C.

1976        Le dialecte grec de Pamphylie. Paris

Bryce, Trevor R.

1986        The Lycians Vol. 1:The Lycians in Literary and Epigraphic Sources. Copenhagen, Museum Tusculanum Press.

Coarelli, F.

1977        “Public Building in Rome between the Second Punic War and Sulla,” Papers of the British Sshool at Rome 45 : 1-19.

Couilloud, M.-Th.

1973        Les monuments funéraires de Rhénée, Exploration archéologique de Délos 30. Paris, E. de Boccard.

Crawford, Michael H.

1996        Roman Statues. London.

Desideri, Paolo and Anna Margherita Jasink

1990        Cilicia Dall’età di Kizzuwatna alla conquista macedone.Turin, Casa Editrice Le Lettere.

De Souza, Philip

1997        Romans and Pirates in a Late Hellenistic Oracle from Pamphylia. Classical Quarterly 47:477-81.

De Souza, P.

1999        Piracy in the Graeco-Roman World. Cambridge University Press.

Diler, Adnan

1991        Lykia Olympos Dağında Bir Ön Araştırma. Turk Arkeologiji Dergisi29: 161-176

Eilers, Cl. And N. P. Milner

1995        Q. Mucius Scaevola and Oenoanda:a new inscription.Anatolian Studies 45: 73-95.

Erdemgil, S. and Özoral F.

1975Antiochia ad Cragum. Turk Arkeologiji Dergisi22.2: 55-71.

Ferrary, J. L.

1977        Recherches sur la législation de Saturninus et de Glaucia. Mélanges d'archéologie et d'histoire de l'Ecole française de Rome 89: 619-660

Franke, P. R., W. Leschhorn, B. Müller, J. Nollé,

1988        Side, Münzprägung, Inschriften und Geschichte einer antiken Stadt in der Türkei. Saarbrücken.

Fraser, P. M.

1972        Ptolemaic Alexandria. Oxford University Press.

French, David

1991        Sites and Inscriptions from Phrygia, Pisidia, and Pamphylia, Epigraphica Anatolica 17: 53-54.

Goldman, Heddy, D.H. Cox, V. Grace, F.F. Jones, A.E. Baubitschek,

1950 Excavations at Gözlü Kule, Tarsus I:The Hellenistic and Roman Periods. Princeton University Press.

Grace, Virginia

1973 Imports from Pamphylia. Études Déliennes. BCH Suppl. I, Paris, Ecole française d'Athènes: 183-208.

Grace, V. and M. Savvatianou-Petropoulakou

1970        Les timbres amphoriques grecs, in Ph. Bruneau, L'Îlot de la maison des comédiens.EAD 27.Paris, E. de Boccard: pp. 277-374.

Grayson, A. K.

1975        Assyrian and Babylonian Chronicles.Locust Valley, New York.

Hassall, M.,M. H. Crawford, and J. Reynolds,

1974        Rome and the Eastern Provinces at the End of the Second Century B.C.The So-Called 'Piracy Law' and a New Inscription from Cnidos.Journal of Roman Studies 64: 195-220.

Hani, Jean

1964        Plutarque en face du dualisme iranien.Revue des Etudes Grecques 77: 489-525.

Hild, F., and Hellenkemper, H.

1990        Kilikien und Isaurien. Tabula Imperii Byzantini 5.Vienna: Österreichischen Akademie der Wissenschaften.

Houwink Ten Cate, P. H. J.

1961        The Luwian Population Groups of Lycia and Cilicia Aspera during the Hellenistic Period. E. J. Brill, Leiden.

Huber, G.

1964        Vorläufige Beobachtungen über die Städteplanung in den Küstenorten des westlichen Kilikien. Turk Arkeologiji Dergisi 13.2: 140-148.

Jones, A. H. M.

1971        Cities of the Eastern Roman Provinces (2nd Ed.). Oxford University Press.

Karamut, I., and J. Russell

In press  Nephelion, a Rough Cilician Settlement. Journal of Roman Archaeology.

Knoblauch, P.

1977        Die Hafenanlagen und die anschliessenden Seemauern von Side.Ankara:Türk Tarih Kurumu.

Lamprecht, H. O.

1968        Opus caementicium. Dusseldorf.

Launey, M.

1987        Recherches sur les armeees hellénistiques. second ed. revised by Y. Garlan, P. Gauthier, and C. Orrieux, Paris, E. de Boccard.

Lehmann-Hartleben, K.

1923        Die Antiken Hafenanlagen des Mittelmeeres. Klio Beiheft 14, Leipzig.

Leski, Noel

1999        Assimilation and Revolt in the Territory of Isauria, from the 1st Century BC to the 6th Century AD. Jounal of the Economic and Social History of the Orient. 42: 413-465 (418).

Lloyd, Seton, and Storm Rice,

1958        Alanya ('Ala'iyya) London.

Magie, David

1950        Roman Rule in Asia Minor to the End of the Third Century after Christ. Princeton: Princeton U. Press.

Mansel, A. M.

1957        Studien zur Geschichte Pamphyliens.Turk Tarih Kurumu, Ankara

Mansel, A. M.

1963        Die Ruinen von Side. Berlin.

Marasco, Gabriele

1987            Roma e la pirateria cilicia. Rivista Storica Italiana 99: 122-146.

Marquardt, Joachim

1881        Römische Staatsverwaltung.Leipzig, S. Hirzel, reprint Arno Press, New York, 1975.

MRR = Broughton, T. R. S.

1986        The Magistrates of the Roman Republic.Originally published, NY, 1952-53; revised edition, 1986, Atlanta, American Philological Association.

Mitford, Terence B.

1980        Roman Rough Cilicia. Aufstieg und Niedergang der römischen Welt. 7.2 Berlin and New York, Walter de Gruyter. 7.2: 1230-1258.

Mosca, P. G., and J. Russell

1987        A Phoenician Inscription from Çebel Ires Dağı in Rough Cilicia. Etudes Anatoliennes 9: 1-21.

Pereira, M.

1966        Mountains and a Shore: A Journey through Southern Turkey.London, Goeffrey Bles.

PECS = The Princeton Encyclopedia of Classical Sites, Richard Stillwell, editor. Princeton University Press., 1976 “Sillyon,” p. 840.

Potter, David

1998        Strabo and early Roman Asia Minor (Review of Syme 1995).Journal of Roman Archaeology 11: 659-665.

Ormerod, H. A.

1922The Campaigns of Servilius Isauricus against the Pirates. Journal of Roman Studies 12: 35-56.

Ormerod, H. A.

1987Piracy in the Ancient World. Liverpool, 1928; Reprint, NY:Dorsett Press.

Rauh, N. K.

1993        The Sacred Bonds of Commerce.Religion, Economy, and Trade Society at Hellenistic Roman Delos, 166-87 B.C.Amsterdam:J. C. Gieben.

Rauh, N. K.

1997 Who Were the Cilician Pirates? Res Maritimae:Cyprus and the Eastern Mediterranean from Prehistory to Late Antiquity.Edited by Stuart Swiny, R.L. Hohlfelder, and H.W. Swiny. American Schools of Oriental ResearchArchaeological Reports No. 4.Atlanta, Ga.:Scholars Press.pp. 263-283.

Rauh, N. K.

1999a Dağlık Kilikiya Yüzey Araştırma Projesi:1997 Sezonu Raporu. Araştırma Sonuçları Toplantısı I. Cilt (25-29 Mayis 1998, Tarsus). T.C. Kültür Bakanlığı, Anıtlar ve Müzeler Genel Müdürlüğü, Ankara.pp. 339-348.

Rauh, N. K.

1999b      Rhodes, Rome, and the Eastern Mediterranean Wine Trade, 166-88 BC. Hellenistic Rhodes:Politics, Culture, and Society.Edited by Vincent Gabrielsen.Aarhus:Aarhus University Press. pp. 162-186.

Rauh, N.K., R. M. Rothaus, K. W. Slane, and L. Wandsnider

In press Amphora Production in Western Rough Cilicia. Submitted to the Journal of Roman Archaeology.

Redford, Scott and Jennifer Tobin

1998 "The Sekerhane Kiosk at Selinus, Turkey: Trajan's Cenotaph," AIA Annual Meetings, Washington D.C. December 1998.

Rediker, Marcus

1987Between the Devil and the Deep Blue Sea: Merchants, Seamen, Pirates, and the Anglo-American Maritime World 1700-1750.Cambridge:Cambridge U. Press.

Rosenbaum, E., Huber, G., Onurkan, S.

1967        A Survey of Coastal Cities in Western Cilicia.Preliminary Report. TTKY VI/8. Ankara.

Ruge, W.

1922        Korakesion. Real-Encyclopädie der classischen Altertumswissenschaft (Pauly-Wissowa), Second ed., 1894--, Berlin. 11: 1371.

Russell, James

1991Cilicia--Nutrix Vivorum:Cilicians Abroad in Peace and War during Hellenistic and Roman Times. De Anatolia Antiqua (Eski Anadolu)1: 283-297.

Shaw, B. D.

1990Bandit Highlands and Lowland Peace:The Mountains of Isauria-Cilicia. Journal of the Economic and Social History of the Orient 33: 199-233.

Sherry, F.

1986Raiders and Rebels:The Golden Age of Piracy.NY:Hearst Marine Books.

Sherwin-White, A. N.

1976        Rome, Pamphylia and Cilicia 133-70 B.C. Journal of Roman Studies 66: 1-14.

Sherwin-White, A. N.

1977        Roman Involvement in Anatolia, 167-88 BC. Journal of Roman Studies 67: 62-75.

Sherwin-White, A. N.

1984        Roman Foreign Policy in the East 168 BC to AD 1. London.

Slane Wright, K.

1980        A Tiberian Pottery Deposit from Corinth.Hesperia 49: 135-177.

Slane. K. W.

1986        Two Deposits fromt he Early Roman Cellar Building, Corinth. Hesperia 55:271-318.

Syme, R.

1995        Anatolica.Studies in Strabo. (Posthumously edited by A. Birley). Clarendon, Oxford.

Tchernia, André

1986        Le vin de l'Italie romaine.Essai d'histoire économique d'après les amphores. Rome, Ecole française de Rome.

Tigrel, Gülay

1975        Alanya Yöresinde antik bir Liman. Belleten Turk Tarih Kurumu 39: 613-628.

Vann, Robert L.

1976        A Study of Roman Construction in Asia Minor:The Lingering Role of a Hellenistic Tradition of Ashlar Masonry.PhD Dissertation, Cornell University

Wandsnider, LuAnn

1998a      Landscape Element Configuration, Lifespace, and Occupation History: Ethnoarchaeological Observations and Archaeological Applications. In The Interpretative Potential of Surface Archaeological Phenomena, edited by A. P. Sullivan, III. University of New Mexico Press, Albuquerque.

Wandsnider, LuAnn

1998b      Past Regional Processes, Modern Archaeological Landscapes. Measuring Time, Space, and Material: Unit Issues in Archaeology, edited by A. Ramenofsky and A. Steffen, pp. 87-102. University of Utah Press, Salt Lake City.

Wiseman, D. J.

1956         Chronicles of Chaldaean Kings (626-556 B.C.) In the British Museum. London.

Table 1: Classical/Hellenistic Ceramic Data [107]

A. URBAN SITES

B. RURAL SITES

Table II: Transect subunits and sites (1998-1999) with unconfirmed classical/Hellenistic sherds[110]