Taranis
Elite member
Broadly, it depends what scenario of the relationship between the various Celtic branches you prefer. There are traditionally two camps in the scholarship, namely the Continental Celtic versus Insular Celtic model and the Q-Celtic versus P-Celtic model.
Continental versus Insular Celtic argues for a closer relationship of Brythonic and Goidelic, based on the fact that they share that they have VSO word order and consonant mutations, which is by Indo-European standards quite exotic, while Celtiberian and Gaulish had an SVO word order, no sign of (regular) consonant mutations, and complex declension systems that were quite similar to that of ancient Greek or Sanskrit. The extreme variants of the Insular Celtic model argue that there's a common substrate in Insular Celtic, and an Afroasiatic language is often suggested for that (VSO is found, for example, in the Berber languages, Old Egyptian and the Semitic languages).
In the Q-Celtic versus P-Celtic model mainly concerns the treatment of the sound *kw from Proto-Indo-European. Both Celtiberian and Goidelic preserved *kw (well, Primitive Irish did, it became *k later in Old Irish, as well as in the modern Goidelic languages), while Gaulish and the Brythonic languages have shifted that to *p.
Which scenario is the more correct one? Of course, one sound change alone (much like the Centum/Satem change) isn't a pretty strong case, but in my opinion the Q/P model is the more "correct" one because Gaulish and Brythonic have more commonalities in their phonetic evolution. We don't know much about Goidelic and Brythonic from the same time that we know about Celtiberian and Gaulish (classical Antiquity), but many of the typical "Insular Celtic" features seem to be later innovations - Primitive Irish (known from the Ogham inscriptions) was essentially a "Continental Celtic" language.
Another aspect is that you have common British and Gaulish deities, as well as tribes (the Atrebates and the Parisi, in particular) that inhabited both sides of the Channel. Thus, we do know that the Britons and the Gauls were close, and it shouldn't be a surprise that their languages were very similar.
Conversely, however, the case for a "Q-Celtic" is much more dubious: People have suggested because Irish is Q-Celtic, just like Celtiberian, that a "Mil Espáine"-type scenario is correct and that the Goidels indeed arrived originally from Spain in Ireland. However, Primitive Irish is much more conservative than Celtiberian, and Celtiberian clearly is not the ancestor of Goidelic. In my opinion, if anything, a reverse Mil Espáine type of scenario is accurate there, and Celtic languages spread in the reverse direction (north to south, not the other way round) along the Atlantic seaboard.
), who proposed the existence of a distinct "Belgic" language, but I don't see where he takes his evidence from. In my opinion, Gysselings construct is unnecessery, Belgic Gaul was predominantly Celtic, except for the area at the Rhine delta, which was probably mixed/bilingual (Celtic and Germanic).