History Files


Ancient Anatolia

Indo-European Daughter Languages: Anatolian

by Edward Dawson & Peter Kessler, 5 August 2016

The Indo-European group of languages, which laid down the roots of many of the world's most-important modern languages, developed from a once-small group of speakers in an isolated homeland.

A sister feature, A History of Indo-Europeans, Migrations and Language, covers in detail the proposed origins of proto-Indo-European, its language shifts, and the migrations of its speakers. Starting there would be a good way of approaching this sequel feature (see the sidebar's 'related links').

Once the proto-Indo-European language (PIE) and its speakers began to expand from their original homeland, proposed to be in the Caucuses Mountains, differences in dialect would have started to become apparent over time.

One such group were the Anatolians.

From proto-Indo-European to proto-Anatolian

They apparently were the first to separate from the rest. This is generally accepted because they preserved in their language what would seem to be early roots of words that later changed amongst the remaining PIE speakers to their north, showing that they were not part of this evolution and were therefore isolated from it.

In fact, the Anatolian branch is such an early split that some linguists have doubts over whether the Anatolians were even PIE-speakers to start with. They may have formed a 'cousin' of PIE rather than a direct descendant of it.

Whether that's true or not, any split from the other PIE groups seems to have occurred around 3500 BC at the latest, which is the dating largely used here, but it may have been even earlier.

Map of Indo-Europeans c.3000 BC
Map 3 from the earlier feature on Indo-European (IE) language and migration shows IE migration out of the Pontic-Caspian steppe by around 3000 BC, by which time the South IEs were certainly already detached from them and isolated from any language changes (click on image to see full sized)

This division and its language now becomes categorised by linguists as pre-Anatolian, showing that it is already divided from PIE and is developing its own distinctive variants in language, whilst also preserving many older usages that later change in mainstream PIE.

The length of time between the calculated appearance of pre-Anatolian and its transformation into the better-understood proto-Anatolian is uncertain, but it could be anything up to a millennium. That descended form then continued to evolve (as all language does), into three main sub-variants. How all this happened is largely expert theory based upon some archaeological evidence and some - rather accurate in the main - linguistic calculation, but overall the theory is sound.

The reasons are given in detail below but, briefly, given the historically-attested division of proto-Anatolian into its daughter languages of Luwian and Hittite when the Hittite empire was at its height in Anatolia in 1400 BC, then a split between Luwian and Hittite may have occurred as much as two thousand years before that (3400 BC). (A third language, Palaic, did not survive much beyond 1500 BC.)

That dating does suggest that the original PIE split was earlier rather than later.

Proto-Anatolian and pre-Anatolian would have been universally used for an indeterminate time before that, possibly a thousand years or more. This pushes back an Anatolian/PIE split to around at least 4400 BC - entirely within the realms of possibility. At this stage, proto-Indo-European is conjectured to be in an archaic state, not yet having fully evolved into the PIE that would be spoken by all IE groups subsequent to the departure of the Anatolian group.

Whether the split from proto-Anatolian into Hittite, Luwian, and Palaic took place around 4400 BC or 3500 BC, proto-Anatolian retained many archaic features which were lost among the other branches of Indo-European during the further evolution of the language, indicating a clear separation between the two branches at an early period in time (into south and north respectively).

Once the southern PIEs had passed over the Caucuses Mountains and into north-eastern Anatolia itself, they would have been isolated from their fellow IEs to the north. They probably had to bypass more distantly-related Kartvelian speakers on their way, sticking to a route that probably followed the western coast of the Caspian Sea and which avoided the difficult terrain of the inland highlands.

Once in the region of what is now Armenia, Azerbaijan, and north-eastern Turkey, they appear to have settled for quite some time into what was probably a semi-nomadic lifestyle that may have found a relative gap between other more settled groups that were heading slowly towards early civilisation. Their new territory could be described as the proto-Anatolian homeland.

From proto-Anatolian to Hittite, Luwian, and Palaic

So how did these proto-Anatolians get from here to occupying and ruling most of Anatolia?

Well, an examination of their language (or at least the daughter languages that were formed out of proto-Anatolian) suggests that they were there for a couple of millennia or so. During that time they probably expanded westwards, gradually creeping further into Anatolia, perhaps more so when early cities appeared there.

During this time, differences emerged in how the three main groups of proto-Anatolians were using their language. There could have been other groups too, but they would have been minor, and none survived to leave a trace in history. The differences that emerged were probably due to separation between the groups as they shifted further westwards. In time they would become separate languages in their own right - daughter languages of proto-Anatolian.

Working out how this process took place is relatively easy when using a more recent example. It seems reasonable to assume (based on a good deal of evidence from historical records and modern language forms) that modern Welsh and Irish Gaelic speakers would have been mutually intelligible around the time of the Roman invasion of Britain. Tribes from Britain were able to migrate to Ireland at that time, and Irish tribes had already migrated to western Britain in the two or three centuries before that.

It has taken two thousand years of division and language shift for Welsh and Irish to reach their current forms. This includes periods of quite intense social pressure and breakdown - the Welsh in the sixth to eighth centuries when their territory was under ever-present threat from growing Anglo-Saxon supremacy in the east, and the Irish during the period of Viking occupation of parts of the east and then the Norman invasion and occupation.

If Irish and Welsh can change so much in two thousand years, then it seems to be a safe assumption that proto-Anatolian would experience no worse a change in two thousand years in a region in which there seems to have been less social pressure, no empires to threaten domination of them, and no major invasions to face. In fact, it was the proto-Anatolians who were doing the 'invading', migrating as they were from the Caucuses into north-eastern Anatolia, and probably disturbing an established tribal order along the way.

How long does a language last? If a typical modern English speaker were to meet an Anglo-Saxon speaker from AD 1000, they would be unintelligible to one another. That's a fairly extreme example, though. English was fused with Norman French to become a very different version of the language. Spanish has remained far more stable, so two people from either end of this timescale would be able to understand each other, albeit perhaps haltingly.

Yet another example is modern spoken Frisian. This is so close to Anglo-Saxon that once a Frisian learns the AS alphabet variant he can mostly read Beowulf in the clear (even though it was composed in the sixth century AD). So you could say the degree of change varies, between the two extremes of English and Frisian, and depending on contact with other languages. English at one extreme was heavily affected, and Frisian, its speakers isolated on their islands, is almost a living fossil. So the not-too-clear answer is anything up two two thousand years depending upon circumstances.

How many non-Indo-European speakers the proto-Anatolian speakers hoovered up along the way can never be known. But it has been common practice throughout history for a new, dynamic, and potentially aggressive group to gain precedence amongst settled groups, eventually dominating them as a new ruling elite and imposing a new language and customs upon them (whilst borrowing some words along the way). The same shift can be seen clearly in Celtic groups when faced with Germanic expansion in Europe and Britain.

Accepting a proposed period of two thousand years for proto-Anatolian to devolve into its daughter languages of Hittite, Luwian, and Palaic means that between 3400 BC and 1900 BC, these daughter languages were developing, and their speakers were becoming isolated from one another - perhaps not entirely, but they certainly developed along three different lines of progression.

Given the extreme examples, above, of a modern English speaker not being able to understand an Anglo-Saxon of a thousand years ago, but a Frisian being perfectly fine with that, one has to wonder how cross-intelligible were Hittite, Luwian, and Palaic by the end of this period. The answer is that they probably were, with some dialectal differences.

Map of proto-Anatolian migration 3000-2000 BC
This map attempts to illustrate in basic terms the separate paths taken by the Luwians, Hittites, and Pala during their westwards migration and their progress from proto-Anatolians to kingdom-builders (click on image to see full sized)

Settling down to form kingdoms

The non-Indo-European peoples who dominated Anatolia during the third millennium BC founded cities such as Hattusa, Kanesh, Kussara, and Zalpa. These were organised and were trading with the more advanced Mesopotamian states by no later than 2700 BC (see links, right).

They began to be threatened by the descendants of the proto-Anatolians in the early eighteenth century BC, by which time the Luwian speakers were already occupying areas of the southern coastline of Anatolia. The Palaic speakers seem to have followed the northern, Black Sea coastline, and the Hittites were taking the middle route.

As mentioned, the Luwians established themselves first, along the coastline of Anatolia and Syria at the north-eastern corner of the Mediterranean. There they founded an early settled state that may have been known as Luwia. As it expanded further west along the Mediterranean coastline it divided into two states: Arzawa and Kizzuwatna.

The Pala (Palaic speakers) seem to have been a relatively small group. They followed the Black Sea coast to reach a region that lies to the north of today's Ankara and there they stayed, neighboured to the east by the non-IE Kaskans. One has to wonder what effect they had on established native cities that they passed. One of the most powerful, Zalpa, apparently suffered at the hands of the Hittites by the mid-eighteenth century BC, but were they weakened before that by the Pala? Unfortunately the historical record says nothing about this, and the Pala disappeared fairly early, possibly absorbed by new arrivals - the Phrygians - in the late thirteenth century BC.

The third major group, the Hittites, migrated through the Anatolian highlands to attack and swiftly subjugate the cities of the Hatti. By the seventeenth century they were in charge in central-eastern Anatolia, and they went from there to found one of the great empires of the ancient world.

Later groups also included Lycian and Lydian, and the descendants of the proto-Anatolians were dominant in Anatolia until the arrival of Greek settlers in the twelfth to eighth centuries BC (these were also Indo-Europeans, but they originated with the northern group). Luwian survived the end of the Hittite empire, but today all three language groups are extinct.


Main Sources

Anthony, David W - The Horse The Wheel and Language: How Bronze Age Riders from the Eurasian Steppes Shaped the Modern World, Princeton University Press, 2007

Online Sources

Proto-Bulgarian Runic Inscriptions:

Studies in the History and Language of the Sarmatians:

Linguistics Research Center, University of Texas at Austin:

Indo-European Chronology - Countries and Peoples

Indo-European Etymological Dictionary (J Pokorny):



Maps and text copyright Edward Dawson & P L Kessler. An original feature for the History Files.