The homeland of the Proto-Indo-Hittite language

No less important than the vindication of the Indo-Hittite theory and the fading of the Volkswanderung is the growing evidence that the Anatolian language family was indigenous to south-central and western Anatolia. This was proposed, although less clearly than it could have been, in Colin Renfrew’s Archaeology and Language,¹⁶ and it has been supported by subsequent research. It nevertheless is still resisted by many linguists and possibly even by most. In his chapter, “The Position of Anatolian,” soon to be published in Handbook of Indo-European Studies, Melchert writes “I follow here the long-standing majority view that the Indo-European languages of Anatolia are intrusive to Asia Minor, having moved there from some point further north in Europe.”¹⁷ Melchert concedes, nevertheless, that the supposed folk migration into Anatolia could not have occurred late in the third millennium bc, as many linguists once assumed:

Contrary to earlier views, there has now developed a consensus among linguists that entry of Indo-European speakers into Asia Minor was much earlier than previously assumed. . . . The gist of the argument is that the attested degree of differentiation of the IE Anatolian languages such as Luvian and Hittite already by the beginning of the second millennium requires at a minimum that their divergence from Proto-Anatolian began by the middle of the third millennium. It may easily have begun as early as the end of the fourth.¹⁸

Whether or not linguists have reached a consensus on this point, there could have been no migration of Indo-European speakers (or, let us say, of Indo-Hittite speakers) into Anatolia either ca. 2500 bc or ca. 3000 bc. Because the Anatolian languages did not share with PIE the terms for a wheeled vehicle and its constituents we must conclude that Proto-Anatolian and PIE had gone their separate ways by the time, say 3300 bc, that wheeled vehicles were invented.¹⁹ Moreover, because by the time that wheeled vehicles made their appearance PIE had already developed its distinctive lexicon, its feminine gender and its elaborate verb system, we must say that the separation of PIE and Proto-Anatolian had occurred a very long time before 3300 bc. The split between the Indo-European and the Anatolian branches of Indo-Hittite could hardly have occurred any later than 4000 bc and is more likely to have occurred by 5000 bc. That the Anatolian branch of Indo- Hittite left a homeland in the north and migrated to Anatolia either ca. 4000 bc or ca. 5000 bc is scarcely credible. Not only does the idea require the widely discredited Volkswanderung, but it also conflicts with what archaeologists have learned about Anatolian prehistory and the spread of a Neolithic economy.

Neolithic Anatolia was far from underpopulated. Quite to the contrary, when we compare it to Europe and the Eurasian steppe we would have to say that its Neolithic population was remarkably dense. Some of the earliest settlements anywhere—whether in China, the Near East, or Sudan and Egypt—have been found in eastern Anatolia. These settlements probably arose for the knapping of obsidian (before metallurgy, obsidian was the most prized material for the making of cutting tools and weapons) rather than for the production of food. Most important of the sites is a village at Hallan Cemi, half way between Diyarbakir and Lake Van and overlooking the Batman river, a tributary to the upper Tigris. The settlement at Hallan Cemi has been dated ca. 10,000 bc, and the villagers may have produced at least a small portion of their food (the excavators concluded that the villagers kept domesticated pigs).²⁰ More certain than the pigs of Hallan Cemi is the obsidian found at the site. The excavators found that obsidian was brought there “from both the Van and Bingol areas and used extensively in all parts of the site.”²¹ Down the Batman river from Hallan Cemi were two other settlements (Demirkoy and Kortik), which may represent relocations of the same community. Other very early settlements stood along the upper Euphrates and its tributaries.

In the ninth millennium bc Neolithic settlements began appearing on the plateau of central Anatolia. Possibly these were preceded by settlements in Cyprus, where several sites dating to the ninth millennium bc have been found.²² The earliest known settlement in the Konya plain of Anatolia, where the annual precipitation is rarely more than 300 mm (12 inches), was found under a small mound at Pmarba^i. Carbon dates indicate that the Pmarba^i settlement lasted from ca. 8500 to 8000 bc. The occupants of this village too were engaged in the knapping of obsidian, and for their meat they depended on game animals.²³ In the Aceramic and Early Neolithic periods the Konya plain had settlements large and small. In the seventh millennium bc one of the largest settlements, evidently because of the obsidian that could be collected on the slopes of Hasan Dag and knapped for exchange, was the proto-city at Catal Hoyuk, with a population of at least 5000. More typical would have been villages like Hacilar, far to the west, with at most a few hundred people. Mihriban Ozba^aran’s survey of the ninth to the sixth millennium bc on the plateau lists dozens of identified Neolithic sites, at eleven of which there has been some excavation.²⁴

Neolithic sites on the plateau were relatively conspicuous to archaeologists, while sites along the Aegean and Marmara coasts have been difficult to detect because they are under alluvial deposits. Recent discoveries and excavations, however, have shown that on Anatolia’s Aegean coast, which receives twice as much precipitation as does the plateau, the Neolithic way of life began ca. 7000 bc and here the villagers were most definitely food-producers. A line of settlements has been found here, from Cukurigi Hoyuk near Ephesus to the Troad. Most productive has been the site of Ulucak Hoyuk, near Izmir, where Turkish archaeologists have been working since 1995. From the 22,000 bones recovered at the site it is clear that from the very beginning ca. 7000 bc animal husbandry—of sheep, goats, pigs and cattle—was a mainstay for the Ulucak villagers. In the lowest level (Level VI) all four domesticates are represented.²⁵

Further north, inland from the Sea of Marmara and the Bosporus, were many more settlements, most of them sharing in the culture of the type-site, Fikirtepe. According to Mehmet Ozdogan, recent research has “securely solved the chronological position of the Fikirtepe culture, placing it between 6450 and 6100 cal bc.”²⁶ Like their contemporaries to the south, the Fikirtepe villagers depended for their meat on domesticated ovicaprids, pigs and cattle.

The spread of stock-raising and agriculture was gradual. Unlike the massive folk migrations visualized 100 years ago, the continuous “wave of advance” proposed by Renfrew brought a trickle of Neolithic settlers—a few at a time— further and further into arable land. From the western coast of Anatolia colonists carried their animals and their crop seeds westward and northward. All of this movement was obviously seaborne. The fertile and well-watered plain of central Crete, between Mts. Dikte and Ida, was one of the first places to be colonized: Level X at Knossos has often been dated as early as 7000 bc, but some specialists would date it ca. 6600 bc. The colonization was probably carried out by fewer than 100 people. These pioneers, as Cyprian Broodbank and Thomas Strasser showed in a careful but imaginative study, knew very well where they were going and how to get there, and they brought with them across the sea no pottery but all four domestic animals: sheep, goats, cattle and pigs.²⁷ Broodbank and Strasser imagined a flotilla of ten or fifteen paddled hide boats or log boats, each carrying a ton or two of cargo.²⁸

The same kind of colonizing brought the Neolithic to Europe (“Europe” in this book is the Eurasian continent west of the Bosporus, the eastern arc of the Carpathian mountains, and the Gulf of Finland). First to beckon was the Thessalian plain. Lying in the lee of the Pindus mountains, and to the south of Mt. Olympos, the Thessalian plain was the broadest expanse of arable land across the sea from Anatolia. Argissa on Thessaly’s Peneios river was apparently colonized somewhat later than central Crete, after the making of pottery had begun, and the settlers again brought with them on their boats the full roster of domesticates.²⁹ Other villages along the Peneios—Soufli Magoula downstream and Platia Magoula Zarkou upstream from Argissa, plus a cluster of other sites—attest to a considerable influx into Thessaly during the Early Neolithic period.³⁰ Sesklo, away from the Peneios but a short walk inland from the Gulf of Pagasai, became a town of several hundred people by the end of the seventh millennium bc.³¹ Settlements on the coast of Macedonia and Thrace were planted at Nea Nikomedeia and Hoca Ce^me, the latter just east of the mouth of the Maritsa/Evros river, which today forms the border between Turkey and Greece.

From the coast, settlers made their way—evidently in the “wave of advance” posited by Renfrew—up the rivers of Thrace: the Maritsa/Evros, the Mesta, and especially the Struma (Strymon). Along the middle of the Struma valley six Early Neolithic village sites have been located, and at one of them—Ilindentsi— excavation has begun under the direction of Dr. Malgorzata Grebska-Kulova.³² Earlier excavations, by a French-Bulgarian team, unearthed an informative Early Neolithic site at Kovachevo, in south-central Bulgaria. Judging from what they have found at these sites, archaeologists believe that the original settlers, usually a small group of twenty or thirty families, came from western Anatolia.³³ From these small Early Neolithic settlements grew the Karanovo-Gumelnitsa culture, which in the fifth millennium bc covered much of Bulgaria and southeastern Romania and may have included several hundred thousand people. Although many of these people would have been descendants of immigrants from Anatolia, just as many may have been descended from the Paleolithic hunter-gatherers of southeastern Europe. The security provided by food production must have attracted many hunter-gatherers to the Neolithic way of life. In any case, on the basis of changes in the material culture Ozdogan concluded that for centuries the settlements in southeastern Europe kept in touch with their Anatolian roots.³⁴

The wave of advance that brought agriculture and stock-raising to the Balkans seems to have continued into eastern Romania, Moldova and western Ukraine. The Cucuteni-Trypillian (or Tripolye) culture of this region commenced soon after 5000 bc, and is now known from several hundred sites. These were primarily located in the river valleys, since the steppe itself was not easily cultivated. Eventually the Cucuteni-Trypillian culture reached as far to the east as the right bank of the Dnieper.

The flow of Neolithic colonists from northwestern Anatolia into southeastern Europe and western Ukraine has been the topic of much archaeological discussion and literature in the last 15 years. It was the subject of workshops held in Istanbul in May of2004 and in April of 2009. The proceedings of both workshops have been published.³⁵ In 2011 Mehmet Ozdogan, who has directed many of the pertinent excavations, provided a useful survey: “Archaeological Evidence on the Westward Expansion of Farming Communities from Eastern Anatolia to the Aegean and the Balkans.” Ozdogan’s article appeared in Current Anthropology in 2011, in a supplemental fascicle titled, “The Origins of Agriculture: New Data, New Ideas.” Another survey published that same year in Praehistorische Zeitschrift traced the expansion of the Neolithic from the Anatolian plateau to southeastern Europe: between 6500 and 6000 bc successive periods of colonization brought three different “Neolithic packages” from western Anatolia to Greece and the Balkans.³⁶

In short, archaeologists have found plenty of evidence that there was—as Renfrew argued a generation ago—a sustained outflowing from western Anatolia into southeastern Europe in the Neolithic period. For a folk migration in the opposite direction—from southeastern Europe or the Pontic steppe into Anatolia—we have never had any evidence at all, nor do we have reason to suspect that in Neolithic or Chalcolithic times any such thing was attempted or even contemplated. The colonizing from Anatolia to central Crete, to Thessaly, to the river valleys of Thrace and ultimately to the lands along the lower Danube, made very good sense: the places toward which the colonists—100 or so at a time—sailed were at least as fertile as the places from which they came, and either were populated only by hunters and gatherers or (as was the case with Crete) were completely vacant. Why an entire linguistic community (thousands? tens of thousands?) would have considered moving from well-watered southeastern Europe to the Anatolian plateau, which already had as large a population as its arid climate could support, is difficult to imagine. Crossing the Bosporus or the Dardanelles, against the native population’s opposition, would itself have been a daunting challenge. Even more difficult to imagine is where on the steppe or in the Balkans a large and cohesive linguistic community (thousands? tens of thousands?) would have come from. Perhaps it is superfluous to add that had there been a migration from the Balkans into Anatolia in the fifth millennium bc, the language brought into Anatolia would have been descended from the very language brought out from Anatolia a millennium or two earlier.

If we pair the Anatolian colonization of southeastern Europe with the Indo- Hittite theory, which in 1987 Renfrew did not do, questions about the origins of Proto-Indo-European are put into a very different light. The languages brought to Europe would not have been (or would not have become) Indo-European, but would have been Indo-Hittite and would have become cognates to the Anatolian languages. As Margalit Finkelberg has well argued, one of the languages of southeastern Europe descended from Indo-Hittite was very likely the “Minoan” language of the Linear A tablets found on Crete. Another, as Finkelberg has tentatively suggested, may have been Etruscan.³⁷

All of that would depend, of course, on the premise that the Anatolian language family—and therefore Indo-Hittite also—was native to western Anatolia. Let us return to that premise and examine it. Not every language native to Neolithic and Bronze Age Anatolia belonged to the Anatolian language family. The ancestor of the Hurrian language attested in the Bronze Age may have been spoken in parts of eastern Anatolia during the Neolithic period. More certainly, Hattic was the language of Hatti, the land within the loop of the Kizilirmak (“Red river,” known to the Greeks as the Halys river) and south of the Pontic mountain range, and Hattic did not belong to the Anatolian family nor—like Hurrian—can it be easily assigned to any other known language family. During the four and a half centuries of the Great Kingdom centered at Hattusha the Hattic language gradually gave way to the language of Kanesh, but Hattic continued to be used on religious and ceremonial occasions. Scraps of Hattic were therefore inscribed on the Hittite tablets found at Bogazkale (ancient Hattusha), providing cuneiform scholars with what little is known about Hattic.

South and west of the Kizilirmak, however, until late in the Bronze Age people seem to have spoken one or another of the Anatolian languages. Of these Luwian was far and away the most widespread, followed by Palaic and finally—and perhaps only in the large city of Kanesh until the language was exported to Hattusha in the Late Bronze Age—by “Hittite.” The number of Luwian speakers can only be guessed, but as one can see by glancing at map B II 13 in the Tubinger Atlas des Vorderen Orients, archaeologists have located hundreds of Early Bronze Age settlements in western Turkey, and they are thickest in the triangle between Konya, Burdur and Eski^ehir.

In the Iron Age Luwian, Palaic, Hittite and perhaps other early Anatolian languages evolved into a variety of regional vernaculars. Thanks to the work of John Ray, Ignacio-Javier Adiego and other specialists, and to the 1996 discovery of a Carian-Greek bilingual inscription at Kaunos, it has now become quite clear that Carian belonged to the Anatolian language family.³⁸ We must therefore now say that all the known Iron Age languages of southwest Anatolia—Carian, Lydian, Lycian, Sidetic and Pisidian—were descended either from Luwian or from a cognate Anatolian language of the Bronze Age. In fact, until Phrygian and Greek were imported, in all of western and southern Anatolia no language other than those belonging to the Anatolian family is attested.³⁹ That blanket statement agrees with Onofrio Carruba’s study of the place-names in Anatolia. Carruba found that southwest of a diagonal drawn from the Sea of Marmara to the Gulf of Iskenderun none of the attested place-names can be assigned to any language other than those of the Anatolian language family.⁴⁰ Place-names ending in -ss, -nd or -nth were common southwest of the diagonal, as they were in Crete, Greece and the southern Balkans. Theories about a migration of Proto-Anatolian or Proto- Indo-Hittite speakers from southeastern Europe or the Pontic steppe into western and southern Anatolia would necessarily need to assume an event both catastrophic and undetectable: although leaving no archaeological trace, the migration would have completely obliterated whatever languages had earlier been spoken there and whatever place-names had earlier been in use.

We have, in short, neither linguistic nor archaeological evidence—and no sound reason to think—that the Anatolian languages of western and southern Anatolia had ever been superimposed on a non-Anatolian or pre-Anatolian substrate. For more than 100 years the prehistory of Eurasia has been distorted by illusions of Volkerwanderungen into Anatolia of Hittite and Luwian nations or of a mass of Proto-Anatolian speakers. That prehistory must finally be set straight. The roots of the Anatolian languages in situ go back at least to the beginning of the Neolithic period, when settled communities were first established. The Indo-Hittite language family originated in southern and western Anatolia.