3.2 Philological considerations

The consistencies established between our artefact samples and western Mediterranean ores support earlier corrections of ancient documents that mistakenly suggested Tarshish's location was to the far east or south of Cisjordan. That is, confused notions about Tarshish's location in Africa or the eastern Tropics began to take shape in antiquity. In his work on the texts of 1 Kings 10.22, Lipiński noticed that while the original Hebrew mentions metallic utensils q(w)pym and t(w)kyym as part of the cargo from Tarshish, and that these might be understood as 'knives and razors', they were incorrectly rendered in later Greek translations as 'carved and hewn stones', and 'apes and peacocks' in Targum Jonathan. Apparently relying on dubious sources and misunderstandings, the historian Josephus working in the 1st century AD similarly included 'Ethiopians and monkeys' among the goods from Tarshish in his description of Solomon's wealth (Lipiński 2004, 225-6 and 262, citing Antiquitates Judaicae VIII, 7, 2-4 §181-186). Along this conceptual line in more recent times, the publication in 1885 of Sir H. Rider Haggard's 'lost world' adventure novel King Solomon's Mines would continue to carry popular imagination about the sources of Solomon's mineral wealth further away from biblical passages mentioning Tarshish, and deeper into the unknown places of Africa.

That said, there is a biblical tradition involving ships sailing south from the Gulf of Aqaba to obtain luxuries, but the destination of such voyages is aligned more convincingly with another lost region or port called Ophir and cargoes of gold than with Tarshish and silver. This is not to say that gold could not have been obtained in Tarshish, only that Tarshish was emphatically associated with the silver trade. Some of the relevant passages include Isaiah 13.12, I Kings 9.26-28 and II Chronicles 8.17-18; I Kings 10.11 and II Chronicles 9-10; I Kings 22.48-49; I Chron. 29.4; Job 22.24 and 28.16; Psalms 45.9; Gen. 10.29 and, Jeremiah 10.9 - if Uphaz in the latter is understood as Ophir, following McKane (1986, 223). I Kings 22.48-49, for example, explicitly understands the Red Sea as a staging point to Ophir and its gold, and outside of biblical sources belonging to the Deuteronomistic history (Deuteronomy I to II Kings 25) that tend to reinforce this understanding (Lipiński 2004, 190), there is also the well-known ostracon found at Tell Qasîle (near the Yarkon River at Tel Aviv) dated to the 8th century and translated by B. Mazar (Maisler) as 'Ophir gold to bēt ḥōrōn, thirty shekels' (1951). He understood that the three horizontal strokes used to indicate 'thirty' were comparable to Phoenician numerical conventions, and considered that the reference to Beth Horon indicated the administrative centre and storage city reportedly extant in the period of Solomon (as in 1 Kings 9.17). Wherever Ophir might have been, this ostracon grounds its association to gold in a historical source free of narrative agendas and errors potentially introduced by variously motivated redactions.

Isaac Kalimi's correct observations about the 'ships of Tarshish' are also helpful in circumscribing the millennia of exegetic confusion about the relative locations of Ophir and Tarshish. Whereas the passages 1 Kings 10.22 and 22.49, Isaiah 2.16 and 23.1, 14, Ezekiel 27.25, and Psalms 48.8 have the ships or fleets of Tarshish ships sailing in the service of Hiram and Solomon in the Mediterranean and the Red Sea, the later, parallel texts of 2 Chronicles 9.21 and 20.36-37 misunderstand them as ships sailing to Tarshish from the Red Sea (Kalimi 2005, 393-4). In other words, ships of a type called Tarshish might have been built to sail south from the Red Sea, including towards Ophir (1 Kings 22.49), and there might have been expert Tyrian sailors on them who acquired exotic goods originating at great distances to the east (Jarus 2013), but that does not mean that the region called Tarshish was located in eastern Africa (Täckholm 1965) or even as far afield as the Tamilakkam in southern India and Ceylon (Origen In Ps. LXXI (LXXII), 9 (PG 12, col. 1524); Bochart Phaleg, 1646, iii.7).

Here, the data from the southern Phoenician hacksilber hoards bring our focus back to sources of silver and lead in the western Mediterranean, where, without the benefit of material substantiation, classical and some biblical sources have been directing our attention for millennia.

Among Classical authors, the Greek historian Diodorus Siculus is regularly cited for having said that the Phoenicians were able to obtain large quantities of silver in exchange for 'small things' from the natives of Iberia who did not value the metal (Bibliotheca Historica V.35, 4-5), but he lived in the time of Julius Caesar, and his work is largely a selective composite of earlier sources. It might be beneficial to question his ability to assess pre-colonial trade, but at least by c. 800-750 the Phoenicians (Sidonians) in Homer were already 'famed for their ships', associated with metals and treasures generally, but also emphatically with silver gifting and craftsmanship, and characterised as trading in trinkets (athurmata). That their associations to sea-trade, trinkets, metals in general and emphatically to silver had the appearance of a trope in the earliest narrative writing of the western tradition, which was itself recorded in Phoenician letters adapted to Greek purposes, could, on its own, arguably indicate significant proto-colonial sea-trade. At the same time, the apparently reductive examplars of Phoenician cargoes may not do justice to the range of goods carried (Winter 1998, 253; for Homeric examples, see Iliad 23.740-45; Odyssey 4.615-20, 15.9, 15.415).

Biblical passages that generally agree about the location of Tarshish in the far west include Psalms 72.10, which situates Sheba and Tarshish at opposite ends of the earth. Likewise, the relatively late book of Jonah understands that the voyage to Tarshish began from the eastern Mediterranean seaboard; its eponymous prophet's descent into the belly of the big fish began with his boarding a ship at Jaffa as an attempt to flee to Tarshish from the divine command that he go to Nineveh (Iraq). The prophesied destruction of Tyre in Isaiah 23.6 similarly orients those looking to flee from Yahweh's wrath in an occidental direction, by exhorting the profiteering, sea-faring merchants on the Phoenician littoral to 'cross over' to Tarshish. More generally, these passages tell us that Tarshish sometimes came to represent in the biblical imagination a geographically imprecise location so far opposite of the Near East that fugitives there hoped to escape the notice or influence of Yahweh himself.

The indications that Tarshish might have been an island in the distant west include Psalms 72, where a chain of scaled correlates often translated as 'mountains and hills', 'rain and showers', 'seas and river' and 'Tarshish and islands' appears in parallel construction. Tarshish's place in this patterned line-up might indicate that it was understood as a large island, or even as a large group or region of coastlands and islands. Nuanced reinforcement is found in Esarhaddon's 7th-century inscriptions that categorise Tarshish as an island, and appear to point towards its location in the west. Although they are not identical (both are damaged), the two relevant inscriptions are jointly referred to as Aššur Babylon E (AsBbE) and sometimes treated as duplicates (Leicthy 2011, text 60). One is K18096, a fragmentary clay tablet found at Nineveh, now in the Kuyunjik collection of the British Museum. The other was discovered at Aššur on a fragmented alabaster tablet, now kept in the Istanbul Archaeological Museum (EŞ6262). It reads:

gi-mir KUR-šú a-bel áš-pur MAN.MEš šá MURUB₄ tam-tim DÙ-šú-nu TAKUR.ia-da-na-na KUR.ia-man a-di KUR.tar-si-si a-na GÌR.II-ia ik-nu-šú
All the kings from the lands surrounded by sea- from the country Iadanana (Cyprus) and Iaman, as far as Tarshish, bowed to my feet.

Iaman is difficult to render with precision; it might refer to Ionia, Greece (Kalimi 2005), or to an island or islands in the Aegean, and Tar-si-si has been variously rendered Tarshish and Tarsus (emended from Nu-si-si once read as Knossos, Pritchard 1955, 290, with references). In any case, the tablet specifies tar-si-si as a region that is šá MURUB₄ (normalised qablu); this means it is 'in the middle of' or 'surrounded by' tam-tim 'the sea', and appears last in a set of geographical markers that seem to point west from Cyprus (Elat 1982, 58).

In these instances, šá MURUB₄ (qablu) might be construed as a term for coastlands or groups of islands and coasts in general, except that Esarhaddon's inscriptions regularly distinguish island locations like Cyprus, Sidon and Tyre as šá (ina) MURUB₄ tam-tim, from seacoasts that are a-ḫi tam-tim. This contrast is more apparent in the polysemic use of these terms for body-parts. The word a-ḫi that designates seacoasts, also means 'arms', whereas qablu that refers to islands, also means 'hips', 'loins' and 'waist'. From here, we see a-ḫi as 'arms' referring categorically to things that are lengthy or extended, while qablu signifies things that are rounded and can be girt – in fact, the word sometimes meant 'belt'. And so we find qablu used to refer to the centres of land-locked citadels like Nineveh, and also to the sea-girt island-cities of Sidon and Tyre (Esarhaddon Text 1:iii 41'; 2: I 14'; 60: o 7'), and distinguished from seacoasts (Esarhaddon Text 2: iv 55'). The reading of tar-si-si as 'Tarsus' is incorrect.

The word qablu has a semantic range comparable to Phoenician (ha)-gadir (or qadir) for 'hedge' or 'enclosure'. Both qablu and qadir have the capacity to reflect an eventual settlement pattern in the colonial period of Phoenicians who prized coastal promontories with adjacent or adjoined islets. The word qadir's application to fortresses, strongholds, enclosures and cities includes its adoption into the Berber languages as agadir, where it now commonly appears as a toponymn in North Africa (Lipiński 2002, 575). Of particular importance here is its survival as the name of the once-island site of Cádiz, adjacent to the region often identified with mythological Tartessus, between Huelva and Seville. There is no question that, long after early historical interactions had become legend, Tarshish was eventually identified with Tartessos in antiquity (Helm 1956), and perhaps because Cádiz was an enclosure, or even the enclosure as far as the Phoenicians were concerned, we may now want to equate it with Tarshish. On the other hand, the fact that we know the Phoenicians called it Gadir at least during the Punic period - when the word Tarshish was still known and in circulation, raises doubts about the strength of the equation.

That is, Cádiz is a place-name with a staying power of more than two and almost three millennia, and this might suggest that it did not have the alternate or earlier name 'Tarshish'. At least in Esarhaddons' inscriptions and Psalm 72, Tarshish seems to be a large island, closer to the scale of Cyprus than the once-islet of Cádiz. The case for Sardinia or a location on the same island as Tarshish is now somewhat strengthened. Its well-established metals-based connectivity to the eastern Mediterranean that certainly pre-dates 1200 BC (this has been most clear from the oxhide ingot trade), upon which pre-colonial traffic could have been built, has recommended its candidacy all along. And while we know what the Phoenicians called other colonial sites on Sardinia including, for example, Sulcis (Slky) and Cagliari (Krly), their name for their most important anchorage on the island's southern coast, at ancient Nora, is unknown. Such considerations recall W.F. Albright's assessments that the Tarshish mentioned on the Nora Stone '... must be the Phoenician name of Nora itself (or of a settlement in its vicinity)', and that 'biblical and Assyrian Tarshish may have been Sardinia' (Albright 1941, 21). The evidence from silver found in southern Phoenicia, in contexts datable to the pre-colonial period, lends material evidence to Albright's views and we agree with the notion that the site's name might have been applied by metonymic extension to the island itself. At the same time, our consideration of the documentary and metallic evidence presented here agrees with the astute simplicity of Frank Moore Cross's observation that Tarshish '... is most easily understood as the name of a refinery town in Sardinia, presumably Nora or an ancient site nearby' (1972, 16).


Internet Archaeology is an open access journal. Except where otherwise noted, content from this work may be used under the terms of the Creative Commons Attribution 3.0 (CC BY) Unported licence, which permits unrestricted use, distribution, and reproduction in any medium, provided that attribution to the author(s), the title of the work, the Internet Archaeology journal and the relevant URL/DOI are given.

University of York legal statements

File last updated: Wed Oct 16 2013