Tuesday, December 28, 2021

The Story Of Celtic

Image via Wikipedia.

Davidski discusses the genetic evidence relevant to the origins of the Celtic languages at his Eurogenes blog:
A new paper at Nature by Patterson et al. argues that Celtic languages spread into Britain during the Bronze Age rather than the Iron Age [LINK]. This argument is based on the observation that there was a large-scale shift in deep ancestry proportions in Britain during the Bronze Age.

In particular, the ratio of Early European Farmer (EEF) ancestry increased significantly in what is now England during the Late Bronze Age (LBA). On the other hand, the English Iron Age was a much more stable period in this context.

I don't have any strong opinions about the spread of Celtic languages into Britain, and Patterson et al. might well be correct, but their argument is potentially flawed . . . . Indeed, when I plot some of the key ancient samples from the paper in my ultra fine scale Principal Component Analyses (PCA) of Northern and Western Europe, it appears that it's only the Early Iron Age (EIA) population from England that overlaps significantly with a roughly contemporaneous group from nearby Celtic-speaking continental Europe.

Archaeology and linguistics have generally favored an early Iron Age date for Celtic language expansion into Western Europe including Britain and Ireland. 

But, we've know for some time now that there was massive population replacement in Britain and Ireland by Bell Beaker derived people with lots of steppe ancestry (and lighter skin) in a roughly 300-400 year time period (or less) in the mid-2000s BCE, which almost surely resulted in the replacement of the first farmer languages of Britain and Ireland (presumably derived from the Iberian first farmer language in the same family as the Cardial Pottery Neolithic that took the Mediterranean route, quite possibly a Vasconic language one, as they largely replaced Mesolithic hunter-gathers there whose language is no doubt almost completely lost), with whatever language the Bell Beaker people (radiating from a Dutch Bell Beaker hub) shared. 

This transition was aided because farming as a way of life had collapsed in Britain and Ireland, probably because the first Neolithic farming methods were unsustainable in some way, such as by depleting top soil nutrients, resulting in a reversion to hunting and gathering and probably to herding as well.

The new paper argues that this language was proto-Celtic of some kind. And, there is a lot to like about this hypothesis, since the geographic extent of the Bell Beaker people and the geographic extent of the Celtic languages coincide to a great extent, and because it was a time of language replacement. Also, most of the early Iron Age genetic impact was in Southern England, not in the periphery of places like Scotland, where Celtic persisted longer.

On the other hand, seeing a continental European genetic influx in the early Iron Age from Celtic areas at just the time that linguists (based upon the magnitude of linguistic diversification) and anthropologists (based mostly upon material culture) should have arrived and brought Celtic languages to the region has the virtue of multidisciplinary convergence of evidence. It is a hard signal to see genetically because Celtic areas of continental Europe and the British Isles were quite similar genetically immediately prior to the Iron Age. 

Iron Age technology is also just the kind of elite dominance generating factor that would have allowed for language shift with a fairly modest migration population genetic-wise, much like the Norman invasion that led to the transition from Old English to French influenced Middle English (on essentially the same geographic route) about two thousand years later.

One downside of this model, however, is that it leaves us in the dark about the nature of the Bell Beaker language. This is particularly challenging because while the genetically similar Corded Ware people almost surely spoke an Indo-European language, the Bell Beaker people adopted a lot culturally from pre-steppe Southern Iberians where this archaeological culture began, and could have been heavily influenced by Iberians linguistically as well (in which case the Bell Beaker people might have been Vasconic rather than Indo-European linguistically, as a result of a culturally driven language shift).

One could also imagine a slightly more sophisticated model in which the shared Bell Beaker language family substrate of Celtic regions facilitated the distinctive development of the Celtic languages, perhaps in a manner distinct from that of, for example, the Italic languages, that may have been part of a shared mid-tier language family with Celtic, perhaps in connection with the Urnfield culture of the 12th century BC (Late Bronze Age) expansion immediately preceding the Hallstatt archaeological culture expansion commonly associated with the proto-Celtic languages and people (and followed in much of its area by the La Tène culture). In this model, the divergence of the Italic and Celtic languages arises from differing linguistic substrates.

A lack of consensus of the relationships of the extant and extinct Celtic languages to each other among linguists (driven by scant source material) doesn't help the process of resolving the different possibilities:
Irish, Scottish and Manx form the Goidelic languages, while Welsh, Cornish and Breton are Brittonic. All of these are Insular Celtic languages, since Breton, the only living Celtic language spoken in continental Europe, is descended from the language of settlers from Britain. There are a number of extinct but attested continental Celtic languages, such as Celtiberian, Galatian and Gaulish. Beyond that there is no agreement on the subdivisions of the Celtic language family. They may be divided into P-Celtic and Q-Celtic.

The different hypotheses (quoted from Wikipedia) are lined up below: 

Eska (2010)

Eska (2010) evaluates the evidence as supporting the following tree, based on shared innovations, though it is not always clear that the innovations are not areal features. It seems likely that Celtiberian split off before Cisalpine Celtic, but the evidence for this is not robust. On the other hand, the unity of Gaulish, Goidelic, and Brittonic is reasonably secure. Schumacher (2004, p. 86) had already cautiously considered this grouping to be likely genetic, based, among others, on the shared reformation of the sentence-initial, fully inflecting relative pronoun *i̯os, *i̯ā, *i̯od into an uninflected enclitic particle. Eska sees Cisalpine Gaulish as more akin to Lepontic than to Transalpine Gaulish.

Eska considers a division of Transalpine–Goidelic–Brittonic into Transalpine and Insular Celtic to be most probable because of the greater number of innovations in Insular Celtic than in P-Celtic, and because the Insular Celtic languages were probably not in great enough contact for those innovations to spread as part of a sprachbund. However, if they have another explanation (such as an SOV substratum language), then it is possible that P-Celtic is a valid clade, and the top branching would be:


Within the Indo-European family, the Celtic languages have sometimes been placed with the Italic languages in a common Italo-Celtic subfamily. This hypothesis fell somewhat out of favour following reexamination by American linguist Calvert Watkins in 1966. Irrespective, some scholars such as Ringe, Warnow and Taylor have argued in favour of an Italo-Celtic grouping in 21st century theses.

No comments: