]On New Year's Day, I posted [on the CONLANG list] about doubts I had against the paleolinguistic model on which my Hesperic family is based on. Since then, I have spent some thoughts on this matter, which I shall discuss in this post.
The Bell Beaker culture is an archaeological culture (i.e., a network of finds of similarly styled artifacts of material culture, such as the ceramic vessels this culture is named after) appearing in large parts of western Europe - basically everywhere west of a line running from Szczecin to Budapest and from there to Trieste - shortly after 2800 BC. The artifacts which define this culture do not bear any inscriptions, hence tell us nothing about which languages the people who made and used them spoke.
Basically, there are four possible answers:
1. The Bell Beaker people spoke a dialect of Proto-Indo-European ancestral to those IE languages spoken in their area in historical times, i.e. Italic, Celtic and Germanic.
2. The Bell Beaker people spoke an extinct branch of Indo-European which perhaps diverged early and may have shown archaic traits, but was later completely eclipsed by Italic, Celtic and Germanic, on which it may have exerted substratum influence.
3. The Bell Beaker people spoke a non-Indo-European language, perhaps the ancestor of Basque.
4. The question is wrongly posed, as there was no Bell Beaker ethnicity, only a fad for a particular style of material objects, and the people who made and used those objects spoke many different languages, some related, others not.
None of these answers can be rejected out of hand. But they may be unequally likely. First, there is the question of the existence or non-existence of an ethnicity corresponding to the Bell Beaker culture. There are some reasons to assume that such an ethnicity actually existed.
One reason comes from genetics. The area of the Bell Beaker culture closely resembles the area where a particular genetic signature, the Y-DNA haplogroup R1b, is dominant in present-day Europe, and analysis of ancient DNA has shown that R1b was almost completely absent from western Europe before the rise of the Bell Beaker culture, but most Bell Beaker
males (females of course do not have any Y-DNA haplogroup) showed this haplogroup. This most likely means that a population characterized by this haplogroup spread across most of western Europe, and such a population would probably also have spread their language.
Second, there is some linguistic evidence of some degree of linguistic unity within the Bell Beaker culture, even if it is weak and not unequivocal. The Bell Beaker area again closely matches that of the Old European Hydronymy, a network of recurring names of watercourses discovered by German linguist Hans Krahe in the middle of the 20th century. This network, however, is hard to interpret as we do not know the original meanings of those names. Krahe interpreted these names within an Indo-European framework, but other interpretations have been advanced, most notably the interpretation advanced by Theo Vennemann (another German linguist) who interpreted them by means of Basque (most linguists are very sceptical of this, especially as Vennemann has proposed Basque interpretations for names that can be aptly interpreted
from known languages such as Gaulish or Old High German; also, his model of Proto-Vasconic is doubted by most Vascologists). However, neither Krahe nor Vennemann drew a connection to the Bell Beaker culture.
Moreover, the Italic, Celtic and Germanic languages share a number of lexical and structural features which may be due to a common ancestor specific to them, or the influence of a common substratum. One such feature is the "centum" change, a merger of the PIE palatovelars and plain velars, which, however, is also observed in Greek, Hittite and Tocharian. The notion of an Italic-Celtic-Germanic node within the IE family tree is accepted by many though not all relevant specialists.
So we have a close correlation of artifacts, genes and (though in the latter case, the evidence is meagre) and languages, which makes it likely that the Bell Beaker culture corresponds to an ethnic unit, which means that the question of their language is probably legitimate. Of course, this was not a single, homogenic language, given the large size of the territory in question, rather a language family.
But what kind of language family was this? At first glance, the notion that the Bell Beaker people spoke a language ancestral to Italic, Celtic and Germanic seems most parsimonious. These, after all, are the languages spoken in most of the area in historical times (though there are exceptions, namely Basque, Iberian and Tartessian in the Iberian Peninsula, and Etruscan in Italy). In contrast, the extinct IE branch hypothesis and the non-IE hypothesis obviously require a large-scale language shift. But there is a problem with this reasoning, and this concerns the Celtic languages. The Celtic languages we find in Roman times in the British Isles, Gaul and the Iberian Peninsula cannot have developed out of the Bell Beaker language in situ - they are simply too similar to each other. Rather, most Celticists assume that Proto-Celtic was spoken in the early Iron Age within the archaeological context of the Hallstatt culture in southern central Europe, c. 800-400 BC. And the Italic languages that are now spoken in much of western Europe are of course the result of the expansion of Latin in the Roman Empire. Hence,
we can conclude that at least in some regions, the languages spoken there in historical times must have arrived much later than the spread of the Bell Beaker culture. This does not mean that the Bell Beaker language cannot have been the common ancestor of Italic, Celtic and Germanic, but it invalidates a common argument in favour of this scenario.
The question of the Bell Beaker language may be accessed via genetics. Of course, genes do not speak languages, and language shifts are common enough, but when a population spreads across an area, as the genes show, they usually also spread their language. The question is: Where did the Bell Beaker people come from, and who are their closest relatives? Before modern genetics, archaeologists were divided on this question. Many assumed an origin of the Bell Beaker culture in the Iberian Peninsula; others sought their origin in the Netherlands or in Hungary. Genetic studies have shown that the Bell Beaker people mostly have a steppe origin (though less so in the Iberian Peninsula than elsewhere). They seem to be descendants of a western extension of the Yamnaya culture, which expanded out of the Pontic-Caspian steppe around 3000 BC. The Yamnaya culture is today widely accepted as the most likely candidate for the community who spoke Proto-Indo-European, which probably means that the Bell Beaker people spoke an Indo-European language. It seems thus unlikely that their language was unrelated to IE and an ancestor of Basque.
But what kind of Indo-European? This is where things become difficult. As discussed above, even the hypothesis that the Bell Beaker people spoke "Proto-Italo-Celto-Germanic" does not work without later major language shifts (or migrations), which does not prove it wrong, however. At first glance, it seems tempting to equate the daughter cultures of the Yamnaya culture such that the Corded Ware people spoke the PIE dialect ancestral to Balto-Slavic and Indo-Iranian, the people of the Balkan-Danubian complex the dialect ancestral to Greek, Albanian, Illyrian, Thracian, Phrygian and Armenian, and the Bell Beaker people the dialect ancestral to Italic, Celtic and Germanic. The Afanasievo culture in southern Siberia would have spoken a dialect ancestral to Tocharian, which seems to stand entirely outside the IE dialect network, as if spoken by a group of Yamnaya people who migrated off early. But whence then Anatolian? There is good linguistic evidence that this branch broke away even earlier than Tocharian. There is some steppe DNA influx into the Balkan Peninsula before the Yamnaya expansion, such as the "Golden Man of Varna" (a man interred at Varna, Bulgaria, around 4200 BC, with a large number of gold items), who seems to have had one grandparent from the steppe. This influx may have been connected with the influx of a language from the steppe.
But was that language Indo-European? Probably not! The steppe grandparent of the "Golden Man of Varna" would have come from the Sredny Stog culture north of the Black Sea, which was one of the two parent cultures of the Yamnaya culture - but the *dominant* parent culture, and thus probably the one contributing the main framework of the language, was the Khvalynsk culture further east, along the lower Volga. (Also, Khvalynsk is closer to the homeland of Proto-Uralic than Sredny Stog.) Hence, the Sredny Stog language probably was not IE, and thus not the ancestor of Anatolian.
Of course, there are those who opine that Anatolian did not go through the Balkan Peninsula at all, but via the Caucasus. Alas, evidence does not look good for that, either. The Caucasus is of course full of non-IE languages, and in the Bronze Age, there was an unbroken chain of non-IE languages - Hattic, Hurrian-Urartian, Kassite, Sumerian, Elamite - stretching from the Black Sea to the Persian Gulf. Also, the most divergent Anatolian language is Lydian - the westernmost of all Anatolian languages. Hence, Anatolian probably entered Anatolia from the west.
Also, while Anatolian seems to have broken off early, this cannot have been *that* early. While the Hittite morphology appears to be more archaic than that of Late PIE, the Anatolian languages do not require any reconstruction of the Early PIE phonology different from the "classic" one. Early PIE and Late PIE seem to have had the same phonological system; while the phonetic values of some phonemes may have changed, there apparently were no splits and no mergers. This limits the time difference between Early and Late PIE to a few centuries.
It is in my opinion more likely that the Balkan-Danubian complex was instead linked to a PIE dialect ancestral to Anatolian. But what then about the Bell Beaker dialect? Shouldn't that dialect then have been even more archaic than Anatolian? The Balkan-Danubian complex probably emerged from a western extension of the Yamnaya culture in the lower Danube valley, which was set off from the main massif of the Yamnaya culture by a geographic bottleneck between the curve of the Carpathian mountains on one side and the marshes and wetlands of the Danube delta on the other. In such areas, languages tend to be more archaic, and it is this we see in Anatolian. But the Bell Beaker culture probably originated even further west. The Yamnaya culture had an even further western extension in southern Hungary and northern Serbia - set off by an even narrower bottleneck formed by the Iron Gates, a narrow gorge where the Danube passes between the Carpathians and the Balkan mountains. It seems absurd to assume that the innovations of Late PIE would spread there, skipping the Lower Danube extension!
This makes it likely that the Bell Beaker people spoken an even more archaic dialect of PIE. But how archaic? There seems to be some evidence that this dialect may not even have undergone ablaut, a change that is reflected even in Anatolian. The Old European Hydronymy shows a phonology different from that of Late PIE including Italic, Celtic and Germanic. Where Late PIE and even Early PIE show ablaut, the language of the Old European Hydronymy seems to show a uniform */a/. The same
predominance of */a/ is found in many of those words in western IE languages which are not found further east. It seems as if the Bell Beaker people spoke a language with just three vowels, */a i u/, of which */a/ was by far the most frequent. (It is conceivable, however, that this language has just levelled ablaut by sound changes merging */e/, */a/ and */o/ and eliminating vowel length distinctions.)
There are also some apparent substratum loanwords which nevertheless show affinity to PIE words. One is */hal-/, an element found in many names of ancient salt production sites in Germany and Austria. The correlation with salt production shows that it probably meant 'salt', and it may be related to PIE *sh2el- 'salt'. The 'sulfur' words in Latin and some Germanic languages may have a similar origin; they somewhat resemble PIE *swel- 'to burn slowly, to smoulder', but do not play ball phonologically. We may even possess a trace of the self-designation of the Bell Beaker people! This may be the Germanic word that in English has become elf. What? The ancient Germanic Elves were nothing like the tiny winged fairies fluttering through Victorian fairy tales, nor like Santa Claus's little helpers. They were a majestic people, beautifully restored by J. R. R. Tolkien. This may have been a memory of a sophisticated culture in the British Isles. The Irish had tales of similar beings living in Ireland before the Celts came there, and the Greek tales of Hyperborea and even Plato's Atlantis may refer to a lost civilization in Britain. And what did that name mean? It may be related to Latin albus 'white', and the Bell Beaker people named themselves the 'Whites' because of their pigmentation. Among the results of ancient DNA studies is that the Neolithic farmers who lived in western Europe before the spread of the Bell Beaker people were of darkish complexion, brown-skinned and dark-haired. They may have looked like people from northern India today. And the Mesolithic hunter-gatherers were also
dark. The Bell Beaker people, in contrast, probably were as pale as the people who now live in their former domain, and often blond.
There is another interesting genetic observation. Europe today is dominated by two Y-DNA haplogroups, R1a in the east (roughly the former Corded Ware area) and R1b in the west (roughly the former Bell Beaker area), with a mixture zone in central Europe (where both cultures intermeshed). In Scandinavia and on the Balkan Peninsula, neither of the two dominates, but they are not uncommon. Both haplogroups came from the Pontic steppe. They are related to each other, but that relationship is much deeper than PIE. So they were both represented in the Yamnaya culture.
But how did the mixture of these two haplogroups separate themselves that way? Examinations of kurgan burials almost exclusively reveal R1b, but the kurgan graves of course are burials of the upper class. The lower class, whose graves were plain and simple and now lost, may have been mostly R1a. This can be explained by the hybrid origin of the Yamnaya culture: it resulted from the take-over of the Sredny Stog culture by the Khvalynsk culture. The Sredny Stog were mostly R1a, the
Khvalynsk mostly R1b. And the Khvalynsk people supplied the upper class in this merger. What we may be dealing with are two migration movements originating in different social strata: the Corded Ware culture would have been founded by Yamnaya commoners who moved northwards to free themselves from their rulers, while the Bell Beaker culture was founded by Yamnaya nobles (perhaps younger brothers of heirs) who sought out new domains in the west. Now, in a stratified society like Yamnaya, the upper classes tend to speak a more archaic sociolect than the lower classes, as the upper class language is often also the language of ritual which always tends to be more conservative. So, the Bell Beaker people may have spoken a more archaic form of PIE than the Corded Ware people.
Alas, this is just speculation, and my own trying to make sense of the facts; I am not convinced that this is right. But I hope it is at least plausible enough to build an "Archaic PIE"-based conlang family on it! Or are there some major problems I have failed to consider?
