Dilmun

The Sumerian name of the Gulf trade-land, Dilmun, written with the sign group NI.TUK and the place-determinative, a spelling so fixed that it served every later period and language of cuneiform. It is among the very oldest toponyms in existence: the name already appears in the archaic texts of Uruk, at the birth of writing itself around 3200 BCE, before any connected sentence can be read, which makes Dilmun the earliest foreign land that writing records. In historical times it runs through the commercial and royal record of Sumer, from Ur-Nanshe of Lagash, whose inscriptions have the ships of Dilmun carry timber as tribute, through the Ur III merchants’ accounts of the Gulf trade. The word’s own etymology is unknown; it may well be the island’s local name, taken whole into Sumerian mouths.

What sets Dilmun apart on its own page is that the name works two registers at once, the ledger and the myth. The same land whose boats appear in tribute lists is, in Sumerian literature, the pure and pristine country of Enki and Ninhursag, where no raven croaks and no lion kills, and the place where the gods settled Ziusudra, the flood-hero, to live forever where the sun rises. No other entry in this atlas names a country that its neighbors could both invoice and imagine as paradise without changing the word. Every later name on this page, Akkadian Tilmun, very likely Greek Týlos too, descends from or continues the name fixed here, at the moment writing began.

The Akkadian name of Dilmun, Tilmun, the Sumerian name carried into the Semitic lingua franca of Mesopotamia, most often written with the same NI.TUK logogram the Sumerian scribes had fixed and voiced in Akkadian with an initial t. It appears in the inscription in which Sargon of Akkad has the ships of Meluhha, Magan, and Dilmun moor at the quay of Agade, and it frames the best-documented private commerce of the ancient world: the Old Babylonian tablets of Ur name the ālik Tilmun, the “Dilmun-farers,” the merchant partnerships that sailed south for copper, among them the archive of Ea-nasir, whose customers’ complaints about his copper are still legible. A millennium and a half later the name is unchanged in the annals of Sargon II of Assyria, who counts Uperi, king of Dilmun, among his tributaries, and it survives to the empire’s end in Neo-Babylonian administrative use.

Tilmun is the workhorse of this page, the form under which the name did its two thousand years of commercial duty. Sumerian coined the writing, but Akkadian was the language of the contracts, the loan documents, and the diplomatic annals, so nearly everything we actually know about historical Dilmun comes to us under this form of the name. It also shows the standard Mesopotamian pattern, seen likewise in Elamtu beside Elam, of one written sign group voiced two ways in two languages; the difference is that here the scribes never added a Semitic ending, keeping the foreign word foreign. The name the Greeks later heard around the Gulf, and rendered Týlos, was in all likelihood this Akkadian form on its last living breath.

The Greek name of Bahrain, Týlos, brought back by the surveyors Alexander sent down the Gulf in the 320s BCE, Androsthenes of Thasos foremost among them. Theophrastus, drawing on Androsthenes’ report, describes the island of Tylos and its wool-bearing trees, the most detailed early Greek account of cotton; Strabo places Tylos beside a second island called Arados and reports that their people pointed to temples like the Phoenicians’ and claimed the great cities of Phoenicia, Tyre and Arados, as their colonies, the Gulf-origin theory of the Phoenicians that Herodotus had already heard. The name stays on the classical map through Pliny and into Ptolemy’s geography, naming the island through the Seleucid and Parthian centuries in which it was a garrisoned and prosperous place.

Týlos is generally taken as the old name Tilmun itself, worn smooth in transmission and fitted with a Greek ending, and if so it is the hinge of this page: the point where a name that had lived its whole life in cuneiform crossed into the alphabet and the West. The identification cannot be proven outright, since no ancient writer connects the two words, but the continuity of place, the initial t of the Akkadian form, and the lack of any competing explanation have satisfied most scholarship. Read that way, a Greek botanical notice about cotton-trees quietly carries the last living syllables of the Bronze Age Gulf, a name already two and a half thousand years old when Androsthenes wrote it down.

The Latin name of the island, Tylos, taken directly from the Greek and used by Pliny the Elder in his survey of the Gulf coast, where he sets Tylos among the islands off Arabia and reports it famous for its pearls; in his botanical books the same island returns for its wool-bearing trees, the cotton notice inherited from Theophrastus and Androsthenes. Rome never sailed these waters in any strength, and Latin knowledge of the Gulf is secondhand Greek knowledge throughout, so the form is a bookish one, at home in encyclopedic geography rather than in any Roman’s sailing directions.

Latin Tylos matters less for what it adds than for where it carries the name: into the encyclopedia that medieval and early modern Europe read. Pliny’s notice, pearls and cotton-trees around a far island, is the shape in which the West remembered the Gulf’s holy isle until modern archaeology recovered Dilmun itself, and when nineteenth-century scholars first connected Bahrain’s mounds to the Dilmun of the tablets, it was partly the classical Tylos, sitting in Pliny at the same coordinates, that let the identification close. A name that began in archaic Uruk ends its ancient run as two syllables in a Roman natural history, waiting two millennia to be recognized.

The Classical Arabic name of the Bahrain archipelago, Awāl, the standard medieval name of the island that had been Dilmun and Tylos. The geographers of the Islamic world, Yāqūt among them, treat Awāl as the island of the pearl-fisheries lying off al-Baḥrayn, a name which in classical usage meant the whole eastern Arabian coastland rather than the island itself; medieval tradition derived Awāl from the name of a pre-Islamic idol worshipped by the tribes of the coast. The name served throughout the Islamic middle ages and long past this atlas’s boundary, well into the Ottoman and early modern centuries, before the mainland’s name al-Baḥrayn finally migrated offshore and settled on the island in its place.

Awāl is the one name on this page that owes nothing to the Bronze Age thread: not a descendant of Dilmun or Týlos but a fresh start, the island renamed from within Arabia rather than from Mesopotamia or Greece. That break makes it the exception that maps the page’s structure, marking where the world’s longest-running commercial toponym finally lapsed. Yet the island’s function under the new name was precisely its old one, pearls, palms, and the Gulf carrying-trade, and the later drift by which al-Baḥrayn, “the two seas,” displaced Awāl even echoes the island’s most ancient reputation, the sweet waters that rise up through the salt, which Enki and Ninhursag had made the gift of the god to Dilmun. The names broke; the island’s story did not.