To the Dregs: Drawing Meaning from the Rhodian Handles of Hellenistic Ashkelon

Citation with persistent identifier:

Birney, Kate. “To the Dregs: Drawing Meaning from the Rhodian Handles of Hellenistic Ashkelon.” CHS Research Bulletin 3, no. 2 (2015).

Introduction [1]

1§1 In a second-century letter to Philokrates, Aristeas describes Ashkelon as one of the four major port cities along the Levantine coast, together with Gaza, Jaffa, and (’Akko-) Ptolemais. Because of these ports, he claims, the land lacks nothing in the way of imported luxuries.[2] Indeed both his characterization and Ashkelon’s status as an international nexus during the period are confirmed by the wide range of imported goods that arrived at the site from all over the Mediterranean, including Attic finewares, Cypriot and Pergamene Sigilattas, and wines imported from Chios, Rhodes, and Kos in the Eastern Mediterranean to North Africa and Brindisi in the west. The site also has some of the earliest Italian cooking wares to appear in Israel in the form of orlo bifido and frying pans, appearing in the early first century BC. Nor was Ashkelon’s commercial development a passive one: from as early as the third century BC dedicatory inscriptions attest to the presence of Ashkelon’s traders establishing connections abroad in cities throughout the Mediterranean. By the beginning of the second century an Ashkelonian was listed as a proxenos at Delos, and in the late second century – after Delos had been transferred to Athenian control – Delian inscriptions list fifteen Ashkelonian merchants. One of these, a banker by the name of Philostratos, was the only banker from the Near East attested on Delos. Ashkelonian merchants are likewise attested in inscriptions from Rhodes, Thessaly, and Italy. By the first century BC, a time when the leading citizens of Athens were neither aristocrats nor political elites but commercial agents recognized for their success, a citizen of Ashkelon had risen to become an ephebe in Athens itself. [3]

1§2 This blend of the textual and archaeological supplies a general picture of Ashkelon as a prosperous and networked city in a Hellenistic commercial age. In the brief and preliminary study that follows, we attempt to refine that picture by addressing a single category of evidence – Rhodian amphora handles – in an effort to add nuance to our understanding of the city’s economic development and commercial connections throughout these centuries.

Rhodian Handles: Approaches to Data

2§1 The last decade or so has seen a tremendous growth in scholarship regarding Rhodian amphorae, affecting virtually every aspect of interpretation.[4] Most notable has been Finkielsztejn’s revision of Grace’s original chronology, which together with refinements in morphological seriation allow for more precise dating of these amphorae and their stamps.[5] Other advances have been possible due to greater precision in collection techniques and consideration of the Rhodian imports in relation to those from other Eastern Mediterranean suppliers. Petrographic analysis, for example, has now allowed the recognition of regional koines of both Rhodian and other East Mediterranean amphora shapes, replacing earlier assumptions about restricted or single centers of production, thus broadening our understanding both of commercial patterns and the significance of ‘branding’ (Monachov 1999, Lawall 2002, Hein et al. 2008). Likewise the rise of volumetric studies – though often controversial – are also representative of the increasingly refined questions that can be asked about the apparatus and oversight of Rhodian wide trade during the Hellenistic period (Monachov 2005).

2§2 Yet Rhodian data sets are notoriously problematic. As with any ceramic class, studies of Rhodian amphorae are plagued by the collection practices – particularly in older excavations – that often favored representative sampling over absolute quantities. It is an almost uniquely Rhodian problem, however, that since the nineteenth century collectors’ fascination with the decorative stamps has sometimes resulted in the preferential collection of the stamps themselves. As a result, Rhodian amphorae tend to be overrepresented at the expense of unstamped and less recognizable jars imported from other areas of the Mediterranean. A second problem is one of context. Setting aside the problem of handles in museum collections that can only be attributed to a given site, the majority of Rhodian stamped handles from actual excavations come from non-primary contexts. That is to say, they are most commonly found not on floors – where they might be more clearly associated with particular behaviors, practices, or social classes – but either in secondary (discard) or tertiary (redeposition of discards in building fills) contexts.

2§3 One of the most significant shifts in scholarship has been in the development of methods which accommodate these flaws, and show ways to work even with incomplete data sets and decontextualized artifacts. One such trajectory has been to examine just the quantities of the Rhodian stamps themselves, setting aside questions of context or other statistical factors.[6] Much like coins, stamped handles are not context-dependent for their dating, but can be assigned a date range usually within 25–30 years and sometimes less. As refined chronologies allow for increasingly precise mapping of these patterns, the fluctuations in imports can therefore be charted over time. Peaks and valleys can then be correlated with commercial prosperity, and used (some suggest) as barometers for political, commercial, or cultural change.[7] These approaches share a core assumption that stamped handles are generally – if imperfectly – representative of an overall pattern of Rhodian imports, and thus when mapped can be indicators – if imprecise – of broader trends.

2§4 A second subset of approaches has specifically explored the value of non-primary contexts. Lawall (2004), for example, has shown that stratification within secondary and tertiary deposits such as wells and construction fills can meaningfully reflect morphological changes in a vessel type over time. Most relevant for our study, however, is the increasing attention paid to the “object biography” of the Rhodian amphora itself. Throughout its lifespan, each amphora passes through a number of stages – from production, stamping, transport, treatment dockside, to its eventual discard and/or reuse.[8] The usefulness of a given handle for dating is variable depending upon where in the arc of its lifespan each vessel was found; likewise Rhodian assemblages from different contexts (primary, secondary, tertiary) can offer different kinds of insights into changing practices of consumption and the mechanisms of wine trade within a site (Finkielsztejn 2002, 2004).

2§5 It is these approaches we hope to harness in this preliminary study of the Rhodian stamped handles from Ashkelon. This rich site has produced a staggering amount of material over nearly thirty years of excavation. However, the sheer quantity of material, coupled with great variation in collection practices employed over the years, renders certain types of quantitative studies impossible (e.g. analysis of fabrics and vessel types sorted by weight, as modeled at Tell Anafa, cf. Berlin and Slane 1997). In the absence of statistical and detailed quantitative data, Ashkelon’s Rhodian handle assemblage is therefore well suited to analyses of the sort described above. And yet it offers something more: Ashkelon offers the opportunity for spatial study rarely possible at other archaeological sites. The Leon Levy Expedition to Ashkelon, operating at the site since 1985,[9] has uncovered more than 1300 square meters of the Hellenistic city, including fortifications, a fragmentary city center, and both elite and lower-class residential neighborhoods. The scale of excavation and the variety of spaces – public and private – thus provide a rare opportunity to assess the value of spatial analysis as a tool for interpreting Rhodian handle assemblages. We therefore present below first a spatial, and then a chronological analysis of the Rhodian stamped handles from Ashkelon, and concurrently assess the impact that inclusion of spatial data might have upon the interpretative methodologies applied to chronological patterns alone.

Spatial Patterns in the Rhodian Data Set [10]

3§1 Ancient Ashkelon is set into a natural basin created by a semicircular kurkar ridge that cups the city on its landward side and faces open towards the sea. The city’s fortifications were built and rebuilt over successive centuries atop this ridge, making the city highly visible from the sea and offering a convenient heading for passing ships. Although the ridge encloses roughly 40,000 square meters of habitable space, throughout the Persian and most of the Hellenistic period only a small portion of this area was occupied. From the early fifth to the late second century BC the area was settled in a strip parallel to the coast, with greatest density of occupation concentrated primarily around the southernmost of two raised “tells” enclosed within the natural ridge. Signs of occupation on the north tell and in the lower-lying areas to the north and east were negligible. It was only at the end of the second century BC that the city expanded, settling areas to the east closer to the edge of the basin, a pattern confirmed both by excavation and extensive coring (see Figures 1 and 2).

3§2 Seventy-five percent of the total Hellenistic stamped amphora assemblage was concentrated in two residential areas on this southern tell, one on the north side and the other on the south (see Figure 3). The remainder was scattered in smaller quantities throughout the site. None come from undisturbed primary contexts; rather the Rhodian corpus was recovered almost entirely from secondary or tertiary contexts. This is unsurprising for a city so densely occupied; most rooms and buildings were reused so consistently that such undisturbed contexts are exceedingly rare except in destruction layers.

3§3 The largest quantity of Hellenistic stamped amphorae at the site – over sixty percent – was recovered from Grids 50, 51, and 57. These were neighborhoods of adjacent insulae (two near-complete and two partial) situated on the southern slope, and built on a roughly orthogonal plan. Originally founded in the Persian period, the insulae were reused continuously throughout the Hellenistic period. The residents were not wealthy; these were lower- to middle-class neighborhoods of mixed residential, commercial, and industrial use. Here boneworking workshops and quicklime production occurred in close proximity to cooking and living spaces. Their diet too was relatively poor: faunal evidence showed a surprising preponderance of birds butchered for consumption, which indicates that some of the occupants were reduced to supplementing their diet with even the tiniest waterfowl captured in the nearby marshes.[11] The character of the finds recovered both from floors and fills within these “downtown” insulae was quite modest, and lacked the large quantities of coins, jewelry, and imported tablewares recovered from the elite neighborhood of Grid 38.

3§4 Grid 38 produced 13% of the total stamped handle collection throughout the Hellenistic period. This was a neighborhood of one near-complete and one partial insula situated on the north slope of the south tell. Although only half the size of the combined occupied areas of buildings in Grids 50, 51, and 57, these two insulae alone produced 60% of the coins from across the site (see Figure 4). In addition to boasting a broader range of luxury items (in both primary and secondary contexts) in all Hellenistic phases, at least by the second century some rooms in these buildings were decorated with Greek Masonry style wall painting, paved floors, and stucco adornments typical of elite residences in Delos and in wealthy neighborhoods up the Levantine coast at Dor and Beirut (Aubert and Eristov 1998; Aubert 2001/2). At least one of the insulae also housed a private bath. While the quantities of stamped handles found in Grid 38 constitute a larger concentration than seen in other areas of the site (such as on the fortifications to the north, or in the areas proximate to the later city center), it pales in comparison with the unusually high concentrations in the neighborhoods on the south slope. The resulting picture is that of an inverse distribution: the majority of the assemblage of stamped amphora handles appeared in lower-class neighborhoods, where coins and other imported goods were relatively rare. By contrast, stamped amphorae were relatively rare in the wealthier neighborhoods. How is this to be understood?

3§5 At first glance, the larger quantities in the southern part of town might be dismissed as due simply to the area’s larger size. Even when broken down by square meter, however, these southern neighborhoods are 2.5 times the size of the northern one, yet have 5 times the number of stamped handles. The divergence between neighborhoods thus cannot be explained on the basis of square footage alone.

3§6 A second consideration might be one of wealth and product preference. Large quantities of Rhodian amphorae are quite often correlated with “prosperity”, though rarely with an eye to whose.[12] An increase in amphorae might well be a good indicator of a city’s commercial prosperity and overall economic health, yet it does not follow that the contents of every amphora should be treated as precious. Nor in a cosmopolitan age does an imported good necessarily mean a luxury good. Could it be, then, that Rhodian wine was being preferentially consumed by the non-elite residents of Grids 50, 51, and 57? This is a difficult question to tackle in the absence of primary contexts, but one worth asking nonetheless. Such an argument would first have to be archaeologically grounded in the assumption that the Rhodian jars in the southern neighborhood be seen as items discarded very nearby their area of initial use. These assumptions are appropriate to a secondary deposit (a street or rubbish dump), but more difficult to prove for a tertiary deposit (construction fill), for which an additional degree of separation from use to endpoint must be acknowledged. This alone does not preclude such an interpretation: one might envision a neighborhood trash heap from which local leveling fills were drawn – and indeed as we shall see below, this may be the case. The greater problem is rather one of scale and degree of devaluation, as the relative quantities appearing in these neighborhoods suggest something closer to regular and repeated use, rather than a rare purchase.[13] The economic and cultural value of Rhodian wine would thus have to have been diminished to the point where it had become not merely a middle-class, but an altogether lower-class commodity. While the devaluation of luxury products in a global market is to some extent expected – and I would argue that it is more appropriate to view Rhodian wine by the late second century as a middle- and upper middle-class product, rather than an exclusively elite one – it is difficult to imagine that this import had fallen so far as to become the cheap staple wine with which the bird-scavenging residents of Grid 51 washed down their suppers.

3§7 The key to the distribution disparity, I would suggest, lies instead in the city’s urban syntax. Grids 50, 51, and 57 are neighborhoods located immediately adjacent to areas which had served as the docks, warehouses, and portside marketplaces for the city during the preceding Iron Age and Persian period (Stager et al. 2011:31–52) and likely continued this role in the Hellenistic period as well. From at least the seventh century onwards, the construction and leveling fills used in the insulae were rich in amphora sherds, and included substantial quantities of larger transport jars such as basket-handled jars and Phoenician carinated-shoulder jars, often in large pieces.[14] In both the Persian and Hellenistic periods, the leveling fills used in construction and rebuilding of these insulae seem to have been tertiary deposits of commercial refuse dumped from the area of the port. The large proportion of the stamped handles in Grids 50, 51, and 57 should therefore be seen in relation to the large proportions of amphora refuse in these areas, and is likely the result of the proximity of these neighborhoods to the commercial entry point to the city.

3§8 This distribution pattern described above helps to illuminate the “object biography” of the Rhodian amphora in port cities such as Ashkelon. Once unloaded from the ships’ holds, most of the jars did not travel far for storage in elite houses (hence their relative rarity in Grid 38) but were rather decanted into other containers at the docks or nearby, and then brought to market in other areas of town. The amphorae themselves could be reused as ballast or refilled with other commodities (Lawall 2000:65). This is not to suggest that none of the amphorae traveled beyond the docks: for example, the backfill of a private bath in Building 65 in Grid 38 included near-complete examples of one Rhodian, one Brindisian, and one Koan amphorae, suggesting that someone nearby may have kept a private cellar. It is likely too that private businesses, such as kapeleia – that dispensed larger quantities of wine – might have kept quantities of complete amphorae in their storerooms, as Kelly-Blezeby (2008) has demonstrated on the basis of well deposits nearby such establishments in Athens and Corinth.[15] Such instances appear to have been rare at Ashkelon, however, at least in the areas exposed.

Chronological Patterns in the Rhodian data set [16]

4§1 Eighty-four Hellenistic stamped amphora handles were recovered from excavations throughout the city, fifty of which were Rhodian handles. Because none come from primary contexts the dating of individual handles carries little weight; meaning is instead derived by examination of the fluctuating quantities of imports over time as marked by the stamped handles. These patterns are presented in Figure 5. The chronological ranges in which stamps appear are represented in 25-year increments, and should be considered in relation to Grace’s Periodic divisions (Figure 4).[17] Despite its limited size, when mapped chronologically the Ashkelon corpus reveals a number of potentially interesting patterns. A full analysis is beyond the scope of this study, and as such we address here only two points of interest, with an eye towards the viability of the quantitative and spatial patterns as interpretive tools.

4§2 First, even when applying the broadest range of possible dates, there is an apparent lacuna in the stamped handles between ca. 270 and shortly after 250 BC. Studies of production patterns on Rhodes itself have indicated that during the span from 268-235 BC there may have been a slight decrease in overall production of Rhodian wine, and further that the amphorae may not have been as systematically stamped at this time (Finkielsztejn 2001b:185). Viewed in isolation, this might well explain the gap, as indeed numbers of imports at ‘Akko and Maresha reflect similarly low levels of Rhodian imports during this range.[18] At Ashkelon, however, the lacuna is concurrent with the decades immediately after the destruction of the early Hellenistic city, which is dated to ca. 290-280 on the basis of coins from several different areas of the site (Stager et al. 2008:287). In a number of neighborhoods excavations revealed decaying buildings, water-washed surfaces and at least one intramural burial, suggesting that the city must have experienced a brief period of disruption or even partial abandonment prior to the rebuilding and reconfiguration of the city. A corresponding lacuna is present in the sitewide coin data, which likewise appear to show a break between 265 BC and 245 BC, during the last two decades of the reign of Ptolemy II. The absence of stamped handles also in this span thus accurately reflects the internal struggles of the city – and therein a diminished market – rather than any external shifts in Rhodian wine production or patterns of Eastern Mediterranean trade. This demonstrates the viability of using Rhodian amphora patterns as indicators of change, when confirmed by outside evidence.

4§3 A second phenomenon of note is the marked tapering in Rhodian amphorae that occurs after 150 BC. While this change does correspond roughly to the transition between the Middle and Late Hellenistic periods at Ashkelon, all areas of the city were continuously occupied throughout this transition and there are no signs of disruption, violent or otherwise. The lacuna must therefore be explained in another way, particularly given that this pattern runs contrary to trade in Rhodian amphorae witnessed at other coastal sites, where the late second century instead witnessed a spike in consumption of Rhodian wines (Finkielsztejn 1999:27). At Ashkelon, in fact, only five amphora stamps date to after 130 BC, and none after 100 – not even in residual contexts. Such paltry numbers are unexpected given Ashkelon’s commercial prowess and long history of Aegean connections. How is this to be understood?

4§4 Changing economic circumstances may have played a role in the decline of Rhodian amphorae. A similar, though not identical drop-off in imports after the mid-second century can be seen at the site of Maresha, some 30 km to the east of Ashkelon. Here, however, after the decline the quantities of Rhodian amphorae seem to recover to near-peak levels in the 130s – 120s BC. In isolation these numbers are misleading, though: despite the recovery, during these decades for the first time the Rhodian jars actually represent a smaller proportion of the total imported jars at the site than they had previously. More jars were arriving overall, and increasing proportions of Brindisi amphorae, Dressel 1A, “ovoid republicaine,” and tubular jars in particular demonstrate the broadening Mediterranean network in the face of Roman expansion (Finkielsztejn 2002 Figure 2). Similar shifts are visible at Ashkelon, although we cannot quantify them. General studies of Late Hellenistic – Early Roman amphorae carried out at the site by Barako and Johnson demonstrate that, as at Maresha, a more diverse range of amphorae made its way to the city’s ports during the last quarter of the second century BC. These included types seen at Maresha, including North African amphorae and Peacock and Williams Amphora Class 32 (Johnson 2008:141ff).

4§5 Yet a greater diversity of imports does not by itself explain a vanishing Rhodian presence in the wine trade. Only a portion of the new imports at Maresha (and Ashkelon) seem to be wine jars;[19] in fact while more jars were arriving at the city, most of these were not bringing wine. As we have no reason to expect a decrease in demand for wine at this time, then in the face of Rhodian decline we must look for alternative sources for wine. As it turns out, we may not need to look far.

4§6 Textual sources from the fourth century AD and later attest to Ashkelonian wine as a favored product in the west. It was used as currency in the church, as the basis for medical recipes, and was of sufficient prestige that it was even served at the coronation banquet of the emperor Justin II.[20] The product was well known enough that in Late Roman sources the wine jars of the region of Ashkelon and nearby Gaza (also a wine producer) were distinctive and referred to as the aksaloniton and the gazition, respectively, names that appear in a variety of papyrological inventories.[21] These late Roman forms (broadly called “Gaza jars”) have been identified and recovered in vast quantities from the areas around both Ashkelon and Gaza, including from kilns. Certainly by the late Roman period local wine production had evolved into a flourishing international industry. Is it possible that these Roman vintages could have had Hellenistic roots?

4§7 Two second-century BC Hellenistic store jar forms – the Hellenistic ribbed pear-shaped amphora and particularly the Phoenician Semi Fine Baggy Jar – may be seen as morphological precursors to both forms of the Late Roman “Gaza jars” with which they share a number of common characteristics (see Figure 6).[22] These common traits include the thickened holemouth rim, ridging along the body – either along the length of the jar or simply under the shoulder and around the base – and twisted vertical handle. Stager (1991), Barako (2008), and B. Johnson (2011) noted the appearance of an amphora with these same features appearing already in late second-century and predominantly first-century BC contexts at Ashkelon which they termed “Proto-Gazan” (Barako 2008:458 Amphora 28; Figure 6.2). These vessels are made in sand-tempered local fabric and become increasingly common in first-century BC to first-century AD contexts at the site. They were traded abroad as well, with first-century AD examples noted as far west as Rome. Johnson also noted the appearance of what she termed a “Near Gazan” jar, a transitional jar form dating to the first and second centuries AD. The body shape exhibits some variability, with the earlier forms being shorter and sack-like – presaging the askaloniton (Figure 6.3a), while gradually elongating into the more familiar tubular shape characteristic of the standard Late Roman gazition (Figure 6.3b).[23] It seems therefore that it may be possible to trace both a morphological and quantitative trajectory in local production from Late Hellenistic baggy jars (produced on a smaller scale in the late second century) through Proto- and Near-Gazan forms of the early Roman to Roman periods, produced in larger quantities, to the Late Roman baggy jars which were produced in vast quantities and became emblematic of local Byzantine wine trade.[24] If this genealogy is accurate, the decline of Rhodian amphora at Ashkelon in the late second century could be seen in light of the seeds of local wine production and the emergence of wine trade in distinctive jars for which Ashkelon was best known in the later Roman and Byzantine periods.[25]

4§8 The spatial patterns should also be considered, as they may help to identify another important shift at the site during this time – one which need not preclude the changing mechanisms of trade consistent with the increasing prevalence of local production. The late Hellenistic (late second through first centuries BC) levels in the southern neighborhoods – those which had produced the largest quantities of stamped handles – were heavily damaged by Byzantine construction activities. The limited nature of late second- to early first-century contexts in these usually amphora-rich areas might admittedly explain the apparent tapering of Rhodian imports during this period. This explanation is still unsatisfying, however. Even in the absence of primary Late Hellenistic contexts, if Rhodian amphorae were still arriving at Ashkelon’s docks, they should still be appearing in the portside dumps, and would be expected to eventually make their way into local construction fills in successive centuries as they had for centuries prior. Residual stamped amphora handles should still be appearing in tertiary contexts – much as Persian amphora continued to appear in Hellenistic and later contexts – as one presumes that builders would still be digging up local debris for the purposes of backfilling and leveling. The absence of Rhodian handles in the construction fills from all later periods in the area thus does indeed point to an overall absence.

4§9 We suggest that by the close of the second century BC the commercial point of entry for the city may itself have changed. Such shift would be consistent with a number of other significant changes occurring in the city at this time. In 113/112, Ashkelon minted silver coinage proclaiming for the first time that it was autonomous, holy, and inviolate, having achieved complete independence from the Seleucids. At roughly the same time, we see the first significant shift in urban syntax since the Phoenician refounding of the city 400 years before: Ashkelon was refortified – the first defensive system built since Nebuchadnezzer tore down the city’s walls at the end of the Iron Age. For the first time too, occupation expanded east of the southern tell to the very edges of the basin. Perhaps most significantly, a monumental colonnaded building was erected in the new settled area to the east of Grid 38. This building, unfortunately poorly preserved, lay immediately underneath a building which later became the Early Roman bouleuterion, suggesting that whatever its purpose it was likely public and situated in what had become the city’s new public center (Boehm and LeBlanc 2011). If the patterns suggested by the subsequent Roman city are relevant here, this Late Hellenistic public center was likely accessed by roads that ran to the north of the elite neighborhood of Grid 38 out to the city’s “Jerusalem” gate in the northeast of the city wall, in what was perhaps a forerunner of the decumanus. The construction of a new port would have better facilitated transfer of goods from ships on new roads to the new city center, while also conferring prestige appropriate to a newly independent city.[26] A shift in the location of the commercial point of entry would account for the sudden disappearance of the stamped handles even in secondary and tertiary contexts after ca. 130 BC, as any remaining jars arriving in subsequent decades would have been decanted elsewhere in the city, and thus remained invisible to our excavation.


5§1 The study above is admittedly preliminary and like fine wine, it will be matured with the addition of new data as studies of Hellenistic Ashkelon continue. It does, however, illustrate the importance of considering urban syntax and spatial distribution in interpreting patterns of trade, and demonstrates too the value of non-primary contexts. The “warts and all” methodologies noted above have offered better ways to navigate the limitations of flawed data sets, and in so doing to drain to the dregs every last bit of information that might be drawn from these silent stamps.

Figure 1. Coring data showing Persian and Hellenistic occupation (ca. 500-200 BC)
Figure 2. Coring data showing LH-ER occupation (first century BC to first century AD)
Figure 3. TEST Distribution of Hellenistic stamped by amphora handles (Left). Distribution of Hellenistic coins by percentage of total by percentage of total (Right).
Grace (1985) and Empereur (1987) Finkielsztejn (2001)
  Dates BC Period Dates BC
Ia End of 4th c. – 280 Ia c. 304-271
Ib c. 279 – 270 Ib c. 270-247
Ic c.269 – 240 Ic c.246 – 235
IIa c.239-225 IIa c.234-220
IIb c.224-206 IIb c.219-210
IIc c.209-199
IIIa c.205-202 IIIa c.198-190
IIIb c.201-194 IIIb c.189-182
IIIc c.193-188 IIIc c.181-176/4
IIId c.187-182 IIId c.175/173-169/7
IIIe c.181-175 IIIe c.168/6-161
IVa c.174-156 IVa c.160-153
IVb c.155-146 IVb c.152-146
V c.145-108 Va c.145-133
Vb c.132-121
Vc c.120-108
VI c.107-88/86 VI c.107-88/86
VII c.85-Augustus VIIa c.85-c.40
VIIb c.40 – Augustus

Figure 4 Comparative Chronologies of Rhodian Class Amphorae (adapted from Ariel and Finkielsztejn 2003 Tab 8.1)

Figure 5. Fluctuations in stamped handles from Ashkelon in dates BC, by percentage of total
Figure 6. Possible Lineage of the “Gaza Jar” 1a: Briend and Humbert 1980 Plate 7.8; 1b: Barako 2008 Figure 23.28; 2: Figure 23.28; 3a: Figure 23.30; 3b Figure 23.29

Works Cited

Ariel, D. I. Sharon, J. Gunneweg, and I. Perlman. 1985. A Group of Stamped Hellenistic Storage-Jar Handles from Dor. IEJ 35.2/3:135–152.

Ariel, D., and G. Finkielzstejn. 2003. Amphora Stamps and Imported Amphoras. Maresha Excavations Final Report I: Subterranean Complexes 21, 44, 70, ed. A. Kloner, 137–151. IAA Reports 17. Jerusalem.

Aubert, C. 2001/2. “Architecture et Décor de la Maison hellénistique à Beyrouth.” ARAM 13–14:73–85.

Aubert, C., and H. Eristov. 1998. “L’habitat hellénistique de Beyrouth et son décor: Site du Petit Sérail.” In La peinture funéraire antique: IVe siècle av. J.C. = IVs siècle ap. J.C. Actes du VIIe colloque de l’Association Internationale pour la Peinture Murale Antique, 6-10 October 1998, Saint-Romain-en-Gal, Vienne, ed. A. Barbet, 211–214. Paris

Barako, T. 2008. “Amphoras through the Ages.” In Ashkelon I: Introduction and Overview (1985-2006). Final Reports of the Leon Levy Expedition to Ashkelon, vol. I, eds. L. Stager, D. Schloen, and D. Master, 429–61. Winona Lake, IN.

Berlin, A. 1997. “The Plain Wares.” In Tel Anafa vol. II: The Hellenistic and Roman Pottery. JRA Supplementary Series, No. 10 Part II, I, ed. S.C. Herbert, 1–211. Ann Arbor.

Birney, K. 2010. Grid 51 Final Report 2010 Excavation Season. Available online at

———. 2011. Grid 51 Final Report 2011 Excavation Season. Available online at

———. 2012. Grid 51 Final Report 2012 Excavation Season. Available online at

Boehm, R. R. Le Blanc, and D. Master. Forthcoming. The Civic Center and Bouleuterion of Ashkelon.

Brett, A. 1937. “A New Cleopatra Tetradrachm of Ascalon.” AJA 41/3: 452–63.

Briend, J., and Humbert, J.B. 1980. Tell Keisan (1971-1976). Une cité phénicienne en Galilée. Vol. I. Fribourg.

Caprariis, F. de, C. Fiorini, and D. Palombi. 1988. “Contenitori da trasporto dell’area Siro-Palestinese.” Mélanges d’archéologie et d’histoire de l’école française de Rome 100:305–20.

Conovici, N. 2005. “The Dynamics of Trade in Transport Amphoras from Sinope, Thasos and Rhodos on the Western Black Sea Coast: a Comparative Approach.” In Chronologies of the Black Sea Area in the Period c. 400 − 100 BC, ed. V. Stolbe and L. Hannestad, 97–117. Aarhus.

Empereur, J.Y. 1990. “Η χρονολόγηση των ροδιακών αμφορέων στην ελληνιστική περίοδο.” In Επιστημονική Συωάστηση για την Ελληνιστική Κεραμεική. Χρονολογικά προβλήματα της ελλενιστικής κεραμεικής Πρατκιτά, Ρόδος 22-25 Μαρτίου 1989, 199–209. Athens.

Empereur, J.Y., and A. Hesnard. 1987. “Les amphores hellénistiques du monde Égéen.Céramiques hellénistiques et romaines 2. Centre de recherches d’histoire ancienne 70, ed. P. Léveque and J.-P. Morel, 10-54. Paris.

Finkielsztejn, G. 1999. “Hellenistic Jerusalem. The Evidence of the Rhodian Amphora Stamps.” In New Studies on Jerusalem. Proceedings of the 5th conference, December 23, 1999, ed. A. Faust and E. Baruch, 21–36. Jerusalem: Rennert Center for Jerusalem Studies.

———. 2000. “Amphores importées au Levant Sud à la époque hellénistique.” In Epistemonike synantese gia ten Hellenistike keramike: chronologika provlemata, kleista synola – ergasteria: praktika, 207–220. [CITY PRESS]

———. 2001a. Chronologie détaillée et revise des éponymes amphoriques rhodiens de 270 à 108 av. J.-C. environ. BAR International Series 990.

———. 2001b. “Politique et commerce à Rhodes au IIe a.C.: Le témoignage des exportations d’amphores.” In Les cites d’Asie Mineure occidentale au IIe siècle a.c., eds. A. Bresson and R. Descat, 181–196. Bordeaux: De Boccard.

———. 2002.“Du bon usage des amphores hellénistiques en contexts archéologiques.” In Céramiques hellénistiques et romaines: Productions et Diffusions et cote syro-palestinienne, eds. F. Blonde, P. Ballet, and J.F. Salles, 227–234. Lyon.

———. 2006. “Some Remarks on Amphora Productions and Trade in the Southern Levant.” Old Pottery in a New Century: Innovating Perspectives on Roman Pottery Studies, eds. D. Malfitana, J. Poblome, and J. Lund, 253–263. Catania.

Fuks, G. 2000. “Antagonistic Neighbours: Ashkelon, Judaea, and the Jews.” Journal of Jewish Studies 51:42–62

Grace, V. 1985. “The Middle Stoa dated by Amphora Stamps.” Hesperia 54/1:1–54.

Hein, A., V. Georgopoulou, E. Nodarou, and V. Kilikoglou. 2008. “Koan amphorae from Halasarna – Investigations in a Hellenistic Amphora Production Center.” Journal of Archaeological Science:1049–1061.

Kelly-Blazeby, C.F. 2008. Kapeleion: Casual and Commercial Wine Consumption in Classical Greece. PhD diss., Leicester University.

Lawall, M. 2000. “Graffiti, Wine Selling, and the Reuse of Amphoras in the Athenian Agora 430-400 BC.” Hesperia 69/1:3–90.

———. 2002. “Ilion and Alexander: Amphoras and Economic Archaeology.” Studia Troica 12:197–244.

———. 2004b. “Archaeological Context and Aegean Amphora Chronologies: A Case Study of Hellenistic Ephesos.” In Transport Amphorae and Trade in the Eastern Mediterranean. Monographs of the Danish Institute at Athens 5, ed. J. Eiring and J. Lung, 171–88. Aarhus.

———. 2005. “Amphoras and Hellenistic Economies: Addressing the (Over)Emphasis on Stamped Amphora Handles.” In Making, moving and managing: The New World of Ancient Economies, eds. Z. Archibald, J.K. Davies and V. Gabrielsen, 188–232. Oxford: Oxbow Books.

———. 2007. “Hellenistic Stamped Amphora Handles.” In Die Basilika am Staatsmarkt in Ephese 2. Teil: Funde klassischer bis römischer Zeit. Forschungen in Ephesos IX 2/3, eds. V. Mitsopoulos-Leon and C. Lang-Auinger, 28–60. Vienna: Verlagder Osterreichischen Akademie der Wissenschaften.

Lund, J. 2011. “A New Sequence of the Eponyms named on Rhodian Amphora Stamps in the First Half of the Second Century BC As Established through Seriation.” Acta Archaeologica 82:271–290.

Margaritis, E. 2014. “The Kapeleio at Hellenistic Krania.” Hesperia 83/1:103–121.

Mayerson, P. 2008. “The Wine of Ashkelon in Byzantine Texts.” In Ashkelon I: Introduction and Overview (1985-2006). Final Reports of the Leon Levy Expedition to Ashkelon, vol. I, ed. L. Stager, D. Schloen, and D. Master, 471–477. Winona Lake, IN.

Monachov, S. 1999. “Towards a typology of Knidian amphoras of the 4th – 2nd c. BC.” Bosposkii fenomenon: grecheskaia kultura na periferii antichnogo mira, 161–172. St. Petersburg.

———. 2005. “Rhodian Amphoras: developments in Form and Measurements.” In Chronologies of the Black Sea Area in the Period c. 400 − 100 BC, ed. V. Stolbe and L. Hannestad, 69–96. Aarhus.

Oked, H.S. 2001. “‘Gaza Jars’: A Chronicle, and Economic Overview.” In Ashkelon: A City on the Seashore, ed. A. Saason, Z. Safrai, and N. Sagiv, 227–250. (Hebrew with English Summary). Tel Aviv.

Regev, D. 2004. “The Phoenician Transport Amphora.” In Transport Amphorae and Trade in the Eastern Mediterranean, ed. J. Eiring and J. Lund, 337–352.

Stager, L., D. Master, and J.D. Schloen, eds. 2011. Ashkelon III: The Seventh Century BC Final Reports of the Leon Levy Expedition to Ashkelon, vol. III. Winona Lake, IN.

———, eds. 2008. Ashkelon I: Introduction and Overview (1985-2006). Final Reports of the Leon Levy Expedition to Ashkelon, vol. I. Winona Lake, IN.

[1] This preliminary study is based on research conducted from August 2014 to May 2015, and has been made possible through the support of the National Endowment for the Humanities, the W.F. Albright Institute of Archaeological Research, and the Harvard Center for Hellenic Studies, for whose support I am profoundly grateful. Thanks are also due to the Leon Levy Expedition to Ashkelon, Directors Larry Stager and Daniel Master, and the Leon Levy Foundation.

[2] Aristeas, Letter to Philokrates 4.114–115.

[3] Delos: IG XI iv 818; Rhodes: IG XII 118; Italy: CIL 10, 1746; Athens: IG II2 1029, 148.

[4] See Lawall 2004 for an excellent survey of the development of amphora studies and attendant methodologies as they have evolved.

[5] Finkielsztejn 2001. For a slightly divergent approach see Lund 2011.

[6] As Conovici notes in her 2005 study: “We shall ignore: 1. The total amount of amphoras imported from every centre; 2. The real content of the amphoras…the amphoras’ capacities…their prices, the reasons for their importation…” etc. (2005:97–98)

[7] Beyond commercial implications, fluctuations in quantities of Rhodian amphorae have been read as indices for the presence of Greek mercenaries or even (in the case of Jerusalem) the rejection of Hellenization. (Finkielsztejn 1999, 2000)

[8] Lawall 2000; Finkielsztejn 2002, Ariel have been champions of this point, inter alia. Lawall particularly draws attention to refilling jars from local wholesalers, selling wine from the amphorae, and refilling amphorae with other commodities for resale.

[9] The Leon Levy Expedition to Ashkelon began in 1985 under the direction of Larry Stager (Harvard University) and the project is presently co-directed by Larry Stager and Daniel Master (Wheaton College).

[10] A rudimentary and preliminary summary of the Hellenistic period was provided in Stager et al. 2008; more detailed analysis will appear in the forthcoming Hellenistic Period site report, authored by myself.

[11] D. Fulton, personal communication. D. Fulton and P. Hesse are contributing the faunal analysis chapter to the forthcoming Hellenistic site report.

[12] See especially Conovici 2005. Lawall offers a useful corrective: “Transport amphoras serve as a useful index of prosperity at least for some of the local population” (Lawall 2002/3:199, emphasis added).

[13] The data set is admittedly small to be statistically significant, but their relative quantities seem to support something beyond occasional use.

[14] See Birney 2010, 2011, 2012.

[15] Although conversely Margaritis (2014:107) in her study of the kapelio at Krania found the proportions of amphorae to be rather small in relation to cooking and drinking vessels.

[16] These are the preliminary results of an ongoing study of this assemblage, based on the complete assemblage as known in January 2015. It should be noted that many of the Hellenistic levels at the site were excavated in the 1980s, the earliest years of excavation at the site, when the archaeological methods were still experimental, and recording and collection practices were highly variable. As a result, some of the handles within the assemblage were described or photographed, but subsequently lost. The stamps included in this study were only those for which the actual handle or photographs of the stamp (rather than simply textual description) were available to confirm the reading. However, even in cases where a photograph of the stamp remained, we sometimes lack information about the style of the rim, the shape of the handle, or detailed information about the paste, each of which contribute to a more precise dating of the form.

[17]The chronological system is that of Finkielsztejn’s (2001) revised lower chronology, derived initially from the work of Grace (1985) and Empereur (1987). The chart in Figure 5 is created on the basis of Periodic dates, as narrower dates for most handles cannot yet be confirmed. Narrower dating will eventually be possible, but with our present data we can responsibly offer here only the general picture created by the broadest possible date range.

[18] Conversely, Conovici’s study on a triad of Black Sea ports showed a gradual increase of Rhodian imports in several sites, especially Kallatis, during this same stretch (2005:113 Figure 12).

[19] Finkielsztejn classifies the Brindisi and ovoid republican jars as oil jars, likewise the Mana C1/2 jars are classified as non-wine jars but with unknown contents (2002 Figure 2).

[20] Dulcia Bacchi/munera, quae Sarepta ferax, quae Gaza crearat,/Ascalon et laetis dederat quae grata colonis…prisca Palaestini miscentur dona Lyaei/alba colore nivis blandoque levissima gusto…. (Flavius Creconius Corippus In laudem Iustini Augusti minoris 63).

[21] For a full discussion see Mayerson 2008.

[22] Two Late Roman forms are identified as Gaza jars. Killebrew (1998, cited in Barako 2008) first suggested that the shorter Gaza jar (Type B) should be identified with the askaloniton and the elongated form (Type A) the gazition, both dating to the Byzantine period. Regev, following Oked (2001), notes that both the longer form and the shorter form of these Byzantine jars were produced in the same kilns, suggesting that the distinction may be not one of local city production, but instead reflects a chronological development. The rounded amphorae (more closely tied to their Hellenistic sack-shaped predecessors) date to the first and second centuries AD while the elongated type spans the fourth through seventh centuries AD (2004:349, see also Oked 2001:233–238). Despite their differences in size, both the shorter and elongated varieties share traits consistent with the Hellenistic precursors.

[23] For a fuller discussion see Caprariis, Fiorini, and Palombi 1988:305–308.

[24] Similar genealogical links have been suggested by Regev (2004) and Finkielsztejn (2006) connecting the morphology of the Hellenistic Phoenician bag-shaped amphora with the Late Roman Amphora 5.

[25] A rise in entrepreneurial activity might not even be limited to Ashkelon: Certainly by the third quarter of the second century, local stamping of bag-shaped jars has been identified at Dor (Ariel et al. 1985) and Anafa (Berlin 1997:152). Lawall notes a similar increase in local forms in this same time frame at Ephesos (2004), although at Ephesos this is the end result of a steady trend in which local amphorae can be seen gradually claiming more market share.

[26] It may also be significant that in 96 BC the city of Gaza was destroyed, an event which doubtless increased traffic through Ashkelon’s port and at least temporarily raised its prestige.