The Oxford Handbook of the Phoenician and Punic Mediterranean

by Carolina Lopez-Ruiz

  Map 36.1 (a) Distribution of rock-cut tombs in the Maltese islands; (b) distribution of tomb clusters in the Rabat and Mdina area drawn on a simplified soil map.

  Source: Drawing by M. Anastasi and N. C. Vella

  In such a scenario, it is not difficult to conceive of the Maltese islands as a node in a web of activities extending over a regional area almost by definition, a fact that finds resonance in the writings of ancient geographers and historians (in particular Diod. Sic. 5.12; Arnaud 2008). This regional area would have the earliest Phoenician settlements circle the strategic strait between Sicily, Sardinia, and North Africa, with Malta and Gozo marking its eastward entrance. What groups these settlements together, besides structures and groups of burials, which are a sign of permanence and would indicate a long-term commitment to a specific land from the eighth century bce, is epigraphy and the tophet sanctuaries which contained cremated infants and animals (Quinn 2012–2013), including the one discovered in the Tal-Virtù area, Malta, in 1816 (Vella 2013). To these markers of group identification can be added novel pottery shapes and forms unknown in the Phoenician homeland, including handmade cooking pots and thistle-headed beakers, which could be a reflection of novel dietary requirements and commensality which fostered a sense of collective, regional identity for the first generation of settlers away from home. This network, grounded in geographical proximity, also brings Malta close to that important seamark in the strait between Cap Bon in Tunisia and western Sicily, Pantelleria, if we go by the discovery of fragments of seventh-century bce Phoenician amphorae made from Maltese clays found on the volcanic island (Bechtold 2018). This link maps nicely onto the sailing route surmised for the ship that went down off the western approaches to Gozo and which contained querns fashioned from Pantelleria’s igneous rock, besides amphorae originating in central Italy (Sourisseau 2015; pers. comm. T. Gambin 2017).

  The Maltese tophet seems to have been short-lived. Likewise were the connections that the islands may have had with commerce that linked Sicily to Carthage and the Tyrrhenian by the end of the Archaic period. Certainly, there is nothing to suggest any particular dependence on Carthage for the sixth–fourth centuries bce, when the cultural links appear to have shifted southward in the subbasin of the Lesser Syrtis, toward the seaways that led to Libya and the Tunisian Sahel (figure 36.1d). The islanders had never been here before, but a late tradition preserved in the sixth-century ce writings of Stephen of Byzantium would have coastal Acholla, the present-day Ras Bou Tria, opposite the Kerkennah isles (Slim et al. 2004: 138), be a colony of the Phoenicians of Malta (St. Byz. Ethnika, 152). Indeed, already years ago, Antonia Ciasca pointed out that for this time span no pottery originating in Malta could be identified in Sicily and in Carthage (pointed out by Bechtold 2018). Links based on pottery finds are tenuous, but recent work has recognized amphorae fragments from the island of Djerba (Bechtold 2018). Moreover, the similarities that exist between the funerary practices along the Sahel and those on the Maltese islands after the fifth century bce may have been conditioned by more than just similar soft rock that could be cut into shaft and chamber tombs. After all, inhumations were the preferred method of body disposal here and in Malta when elsewhere in the Punic world, the norm was cremation (Ben Younès and Krandel-Ben Younès 2014; cf. Said-Zammit 1997: 5). In this context, it is worth repeating the remarkable similarities that have been noted between a wooden casket used as a coffin in coastal Bou Ghrara, opposite Djerba, and an equivalent in terracotta discovered at Għar Barka, Malta (Sagona 2002: 820). The cultural connections that might have brought communities around the Lesser Syrtis and the sea of Malta together may also have left an indirect trace in the writings of the later Greek geographer Diodorus Siculus (5.12), who inserted the islands of Malta and Kerkennah in a series of sea measurements and sea routes that linked Africa to southeast Sicily (Arnaud 2008: 25).

  Connectivity beyond the Maltese archipelago shifted again in the course of the third century bce, judging by the transport amphorae and table amphorae originating in Malta that have been recovered in Pantelleria, Lilibeo, and Ibiza—geographical areas that are believed to have come under the control of Carthage (Bechtold 2018). This period also coincided with the start of the Second Punic War, by which time Pantelleria and Lilibeo had fallen to Rome. It is difficult to gauge the extent of the new realpolitik. Pottery of Maltese origin appears to stop leaving the island soon after. Indeed, beyond the mid-second century bce, no Maltese-sourced archaeological material has been attested beyond the islands, despite the fact that the local ceramic production in Malta and Gozo continued to flourish (Anastasi 2015: 163–67). What has been noted is a marked increase in imported wine amphorae from southern Italy, suggesting that the archipelago witnessed a shift in trading contacts that are synonymous with the early Roman western expansion (Anastasi 2015: 168–71). Prior to the start of this influx of imported goods, the ceramic assemblages of the fourth and third centuries bce are markedly devoid of imports. Compared to the pottery recovered from neighboring Pantelleria, for instance, the Maltese islanders opted to use locally produced fine wares, as well as their own oil, probably produced at those sites that went on to become significant olive-oil production centers, and stored and carried in Maltese ovoid amphorae. Self-sufficiency seems to have been key at this time until the fame of one hallowed sanctuary dedicated to Astarte (Tas-Silġ; see later in chapter) reached rich seaborne merchants who made it a point to visit.

  Rituals and Place-Making

  The process of Phoenician settlement on Malta and Gozo—in particular the human–environment interaction—resulted in a cultural landscape embedded in maritime connectivity, on one hand, and in the exploitation of the islands’ agricultural resources, on the other. Four archaeological sites can be linked with efforts to ritualize the beginning and end of sea journeys that brought the islanders into contact with the world beyond and newcomers in contact with a productive environment.

  On two precipitous headlands on the west coast are the remains of a shrine. At Ras ir-Raħeb in Malta, amateur explorations carried out in 1961–1962 revealed the foundations of a building surrounding a courtyard paved with opus signinum. The pottery has not been properly studied, but Punic pieces are common. A handful of terracotta figurines are known, two of which probably represent Herakles, if not Herakles-Melqart (Vella 2002), a deity often taken to represent the polytheistic matrix of ancient Mediterranean religions (Malkin 2011). At Ras il-Wardija in Gozo, a rectangular rock-cut chamber, measuring 4.6 by 5.6 m, with wall niches had benches also cut into the rock and a number of cisterns; nearer the edge of the cliff, archaeologists found the remains of a building or an enclosure, measuring about 11 by 3.5 m. The earliest pottery recovered suggests a use in the third century bce (Anastasi 2015: 133–37, app. III). It is tempting to link this shrine with one of those four temples, one of which was dedicated to Astarte, mentioned in a second-century bce Punic inscription found in Gozo in 1855 (CIS, I, 132).

  Even if identifying the function of many sites is difficult given the limited finds, their locations are telling and reflect the islanders’ relationship with their sea space—that transitional area observed along the seaways rather than a panoptic gaze that encompasses the entire Mediterranean. In this sense, the west coast of the archipelago was a veritable sea mark, with its spectacular line of white cliffs that look toward Africa (figure 36.1a). The coastline would have aided wayfinding for those plying south-central Mediterranean waters, but the accidented topography made the approach of a vessel under sail a dangerous endeavor, as the archaic shipwreck and others (Azzopardi 2013) that went down off Xlendi Bay attest (for shipwreck archaeology, see chapter 27, this volume). The gods that inhabited the headlands on either side of the Comino channel would have been invoked in prayers and other rituals to placate these anxieties. Admittedly, the dates of occupation of both sites in Malta and Gozo suggest that the material ritualization of this sea mark came rather late, but many rituals of the sort envisaged here would have taken place
at sea with offerings ending up in the sea (Brody 1998: 81–83). Evidence from around the Mediterranean, in locations where the Phoenicians are known to have been active, suggests that certain coastal areas were held as sacred because they were nodes in traveling itineraries. To put it another way, a sense of place was constructed as the result of traveling through sea spaces, with knowledge embedded in stories and oral traditions that allowed seafarers to cope with different scales of environmental variation (López-Bertran 2011).

  On a hillock overlooking the southeast bay of Marsaxlokk in Malta, the long-lived site of Tas-Silġ also attracted the Phoenicians owing to unknown reasons (figure 36.2). Despite years of excavation, the nature and extent of the Archaic remains still elude archaeologists. Only a large ground altar (2.75 by 1.1 m) and the foundations of at least one freestanding chapel, as well as neat postholes for a fence that would have marked the religious space, have been dated to the seventh or sixth century bce (Rossignani 2012). By the fourth century bce, major changes were carried out, as the same altar was flanked with masonry screens ending in two pillars, probably crowned with double Egyptianizing capitals, and a wall was built to encircle the site (Rossignani 2012: 101 n17). The intervening space appears to have been occupied by stone pillars inserted in the beaten-earth floor (Rossignani 2012: fig. 3b)—not unlike the arrangement at the Phoenician shrine at Kommos (Shaw 1989) or those pillars set up to carry offerings, as at the urban sanctuary of Greek Metapontum (Doepner 2002: 19–80). A table altar now replaced the earlier chapel (Rossignani 2012: fig. 3b). A molding of the new altar was dedicated to the “Lady Astarte of Malta [’NN],” dated on account of the letter forms to the fourth or third century bce (Amadasi Guzzo 2004–2005: 288). The name of Astarte is also found on potsherds recovered from the dumps at the edge of the sanctuary, alongside numerous other votive formulae, many of which are still undeciphered (Frendo and Mizzi 2015).

  The Tas-Silġ sanctuary remained intact during the wars between Rome and Carthage. It underwent another major restructuring phase in the second or first century bce, when dedications to the Greek Hera and the Roman Juno appear, and a side chapel was kept aside for a cult featuring Isis (Bonzano 2012). Activity was centered on a courtyard, lined with two rows of columns crowned by Doric and Italic-Corinthian capitals and floors in opus signinum (17.45 by 13.45 m), which acted as an architectural frame to the older cult setup. In this period, the intensity of contacts with southern Italy jumps up the scale, and with it the cultural standing of the islanders, as revealed by the cargo of the ships calling at Maltese ports. It is likely that all this was owed to rich merchants plying the central Mediterranean waters, who patronized a sanctuary that was to be lauded for its riches by the first-century bce orator Cicero (Verr. II, 4, 103–104), who refers to it as fanum Iunonis, alongside the Heraion on Samos (Verr. II, 5, 184) (Bruno 2009: 138–39). The temple was important enough to receive a pair of coordinates in the famed second-century ce geographic treatise of Ptolemy (Geog. IV, 3, 13).

  From a maritime perspective, what we see at Tas-Silġ is continuity rather than transition or rupture implied by the phases of cultural history. Occupation reveals that from its earliest, the site was conditioned by the wish to capitalize on its location, a point of convergence for three maritime routes—from Marsascala, St Thomas Bay, and Marsaxlokk—and a route that is oriented to the island’s agricultural interior (Grima and Mallia 2011). Location itself, if not the memory of the sacred nature of this spot with its megalithic remains, may have lured the Phoenicians to a hub perfectly placed on the frontier of two domains, where foreigner and local could meet, exchange knowledge, and where cultural traits could overlap—a veritable emporion-type setup, certainly by the closing centuries of the first millennium bce.

  The material recovered from the Tas-Silġ temple dumps is very revealing in this sense. The votive objects—scarabs, amulets, inscriptions, and such (Bonanno 2015; Amadasi Guzzo 2012)—taken together with those found in the tombs point to broader spheres of “Phoenician-Punic” belief systems with a penchant for anything Egyptian (Hölbl 1989). But also the coins from the period that start in the mid-fourth century bce appear to fill that intermediate position between Sicily and Carthage (Frey-Kupper 2014). More telling are the faunal remains and the residues found in the pots offered and used in the sanctuary. The first, from deposits which predate the last major restructuring phase of the site, show a concern with feasting on resources of land (including sheep and goat, cattle, chickens, doves, duck) and sea (fish: sea bream, snapper, grouper, tuna; seashells, sea urchins) (De Grossi Mazzorin and Battafarano 2012; Fenech and Schembri 2015; Corrado et al. 2004). The pots that contained these remains were in part meant for everyday consumption and in part for ritual and ceremonial purposes, often marked with dedicatory inscriptions (Quercia 2004–2005; Amadasi Guzzo 2004–2005; Frendo and Mizzi 2015). Coral was also offered (Quercia 2005). Some of the pots were found in secondary deposits dated to the late third or second century bce, placed among amphorae that had probably carried fish sauce and wine to Malta (Semeraro 2012: 116). Fat-residue analyses have revealed that these cooking pots were used to boil meat and fish and that serving plates may have held honey (Notarstefano 2012). These are products that were harvested from the two landscape domains of which Tas-Silġ was a principal hub. Highly flavored sauces may have been the perfect condiment for the food with which farmers, shore dwellers, and exotic merchants and travelers celebrated the multifaceted Astarte, the goddess on the move and the goddess of the land, or—and we say this cautiously for want of clear, direct evidence—the goddess of the sea and agricultural production all in one (cf. Lipiński 1995: 153–54; Day 2002: 131–32).

  It is our contention that the Phoenician islanders sacralized the headlands and bays that looked out at the worlds beyond, with Tas-Silġ taking on the most important role. Besides the three sites mentioned here, however, others must have existed, concerned with more local matters, including inter-island connectivity. Such a shrine was built at a modified cave site and an artificial terrace sometime in the fourth century bce, possibly earlier, at Għar ix-Xiħ, overlooking the cove of Mġarr ix-Xini in Gozo (Azzopardi 2014: 215–33). The site underwent changes well into the third century CE. We cannot be sure if the small terracotta (molded and hand-made) and stone figurines recovered from the site represent deities rather than votaries, as the figures are unnamed, without unambiguous iconographic details. They reflect the popular culture of those who deposited offerings or feasted off the numerous small bowls recovered. The coins found, although few, do show that the visitors were drawn into a scheme of things that brought them into contact with a world of commerce and trade (Frey-Kupper 2014: table 7.8). The location of the site is telling, too, for it marks the entrance to a winding gorge, along the sides of which there is a system of rock-cut treading pans and collection vats. It is thought these were used for wine making; a similar principle is known from Punic Sardinia (Roppa 2013: 70, 87). One of the Gozitan pans, located on the northern side of the valley bottom, was associated with a terraced field that was properly excavated and dated to the sixth or fifth century bce (Azzopardi 2014: 234, fig. 57). Numerous fragments of drinking cups have been recovered from the soil matrix of the field, and it is thought that these were used in rituals marking the harvest and the pressing activity that ensued (Azzopardi 2014: 234–35).

  The production of wine appears to have been high on the productive agenda of the islanders elsewhere, too. At a time when the move into the agrarian countryside is particularly marked in the Punic world, starting from the sixth century bce and increasing steadily by the fourth (Van Dommelen and Gómez Bellard 2008: 202), in Malta we see some evidence of this not only in the archaeobotanical record with cereals and pulses existing alongside a pastoral element (De Grossi Mazzorin and Battafarano 2012: 358–59; Fiorentino et al. 2012: 174, 176) but also at a number of farm sites which go on to flourish during the Roman Imperial period, when they capitalize on the production of olive oil (Anastasi and Vella 2018).
Below the Roman villa site at San Pawl Milqi, located at the head of a major inlet at Salini/Burmarrad (see Marriner et al. 2012: 63), archaeologists have reconstructed a Punic farming establishment involving two wings built round a courtyard, an area of about 15 m square (Locatelli 2005–2006: 266–67, fig. 2). From the pottery collected in a major fieldwalking project that included the area around the site, it seems that agricultural activity in this part of Malta may go back at least to the fifth century bce (Docter et al. 2012: 119), making most of the inner reaches of the fertile plain if not also parts of the fields on the clay slopes which could now be tilled by iron-shod ploughs. At the other end of the island, 1.6 km from Tas-Silġ, in Żejtun, the first-century bce remains of the foundations of another Roman farmstead have been noted to cut into an earlier system of linear rock-cut trenches, about 2 m apart (Vella et al. 2017). These trenches are associated with the cultivation of vines, following a system that is known from Punic sites in southern Spain (Echevarría Sánchez and Vera Rodríguez 2015) and Ibiza (Marlasca Martín and López Garí 2006). From the same site in Żejtun come a handful of votive inscriptions in Punic that draw the site closer to Astarte’s domain at nearby Tas-Silġ.