Zompist Bboard Again

In a discussion about "Dark Age" Britain in a German-language world building forum, the idea came up that the Belgae in northern ancient Gaul spoke a Germanic rather than a Celtic language. This was the first time I head such a suggestion, and I am sceptical about that, but the matron names in the Rhinelands look as if they came from a language that had undergone the Germanic consonant shift, which may speak for this hypothesis. What do you think of this?

WeepingElf wrote: ↑Sat Jul 08, 2023 9:18 am In a discussion about "Dark Age" Britain in a German-language world building forum, the idea came up that the Belgae in northern ancient Gaul spoke a Germanic rather than a Celtic language. This was the first time I head such a suggestion, and I am sceptical about that, but the matron names in the Rhinelands look as if they came from a language that had undergone the Germanic consonant shift, which may speak for this hypothesis. What do you think of this?

Could it have been a Celtic language or dialect whose consonant inventory might give the impression of that consonant shift? (and "matron names" makes me think of basically Roman word lists - grave markers, dedications, various forms of Roman paperwork with Belgae names on them)

I don't know enough about the matter to really have opinions about it, but I wonder how we today would be in any position to really ever figure anything out about the matter.

One way to tell is old geographical (and where known, personal) names. Are the names in Roman Era northern Gaul Celtic or Germanic? I don't know, but this probably has been investigated long ago, so one would expect that the scholars know, and the prevailing opinion seems to be that the Belgae spoke a Celtic language. Perhaps they once spoke Germanic but later shifted to Celtic, the language of their culturally more sophisticated southern neighbours.

WeepingElf wrote: ↑Sat Jul 08, 2023 9:18 am the matron names in the Rhinelands look as if they came from a language that had undergone the Germanic consonant shift, which may speak for this hypothesis.

The usual explanation for that is that these names were ported*) to Germanic when Germanic tribes settled the left bank of the Rhine, a process that was already in progress when Caesar conquered Gaul (hence his run-in with Ariovistus).
*) Germanic and Celtic tribes had been neighbours for millennia, and the speakers on both sides were aware of the correspondences; that's why we have doublets like the river names Latin Moenus < Celtic *Moinos vs. Germanic Main, or Rhēnus vs. Rhine (with Celtic /e:/, Germanic /i:/ as correspondent outcomes from PIE /*ey/). Plus nowadays the Germanic consosnant shift is dated rather late, so that the shifted forms may not be due to substitution, but even to loaning with subsequent shift.
In general, there are place names in the area known from antiquity that look non-Germanic, e.g. Coriovallum = Heerlen in the Netherlands, where the first element is the PIE word for army in an unshifted form.
OTOH, others have speculated that the Belgae weren't Celtic speakers; after all, this is the area of Kuhn's Nordwestblock, where he assumed a non-Celtic and non-Germanic IE language was spoken from which Germanic got its words with /*p/.

Yes, the "Nordwestblock" hypothesis is familiar to me. This IE language that was "neither Germanic nor Celtic" may have been a Para-Celtic language that was not reached by the loss of *p, or a Para-Germanic language that was not reached by the Germanic consonant shift. Or simply yet another branch of IE. Which of the three is hard to tell from a bunch of names. I dimly remember seeing a map of pre-Roman Gaul that showed "mixed Celtic-Germanic tribes" (keltisch-germanische Mischstämme) in Belgica, and according to both Tacitus and Caesar, the Belgae themselves claimed to have Germanic roots, as opposed to the Celts (or so says Wikipedia). Perhaps they were Celticized Germanic people.

And I concur with you that the Germanic consonant shift probably was quite late - later than 500 BC, perhaps even later than 100 BC (the Cimbri and Teutones seem to have had unshifted names, but these may have been Celtic or "Belgian" exonyms), though earlier than 1 AD as names like Cherusci or Chatti are clearly shifted. Perhaps the shift happened in two stages: first, the PIE voiceless stops acquired phonemic aspiration in those environments where we see the shift, which may have been quite early; second, the feature [+aspirated] changed into [+spirant], which was late. The devoicing of the PIE voiced unaspirated stops, Verner's Law and accent retraction may have happened any time between the two stages, and as Celtic voiceless stops appear to have been aspirated in similar environments as in Germanic, they would have been reflected as aspirated stops in pre-spirantization Germanic, too, so the Celtic loanwords in Germanic would have been affected by spirantization as we observe.

I now think that the Belgae were Celticized Germanic people. They spoke a Celtic language that was, however, heavily influenced by the Germanic they had spoken earlier, and thus noticeably different from Gaulish, though the difference was of course not nearly as great as that between Gaulish and Aquitanian, which was related to Basque. This is, according to the English and German Wikipedia articles, also what the Roman historians wrote about the matter, and apparently compatible with the actual evidence.

WeepingElf wrote: ↑Sun Jul 09, 2023 8:50 am And I concur with you that the Germanic consonant shift probably was quite late - later than 500 BC, perhaps even later than 100 BC (the Cimbri and Teutones seem to have had unshifted names, but these may have been Celtic or "Belgian" exonyms), though earlier than 1 AD as names like Cherusci or Chatti are clearly shifted. Perhaps the shift happened in two stages: first, the PIE voiceless stops acquired phonemic aspiration in those environments where we see the shift, which may have been quite early; second, the feature [+aspirated] changed into [+spirant], which was late. The devoicing of the PIE voiced unaspirated stops, Verner's Law and accent retraction may have happened any time between the two stages, and as Celtic voiceless stops appear to have been aspirated in similar environments as in Germanic, they would have been reflected as aspirated stops in pre-spirantization Germanic, too, so the Celtic loanwords in Germanic would have been affected by spirantization as we observe.

On Verner: it's generally assumed that spirantisation already had happened when Verner worked, as similar voicing of spirants depending on the accent has been observed elsewhere.
Generally, the Negau helmet (ca. 4th century BC) is taken as the earliest attestation of the Germanic consonant shift, so names like Cimbri and Teutones (1st century BC) would be Celtic exonyms preserving the unshifted consonants.

hwhatting wrote: ↑Mon Jul 10, 2023 5:06 am

WeepingElf wrote: ↑Sun Jul 09, 2023 8:50 am And I concur with you that the Germanic consonant shift probably was quite late - later than 500 BC, perhaps even later than 100 BC (the Cimbri and Teutones seem to have had unshifted names, but these may have been Celtic or "Belgian" exonyms), though earlier than 1 AD as names like Cherusci or Chatti are clearly shifted. Perhaps the shift happened in two stages: first, the PIE voiceless stops acquired phonemic aspiration in those environments where we see the shift, which may have been quite early; second, the feature [+aspirated] changed into [+spirant], which was late. The devoicing of the PIE voiced unaspirated stops, Verner's Law and accent retraction may have happened any time between the two stages, and as Celtic voiceless stops appear to have been aspirated in similar environments as in Germanic, they would have been reflected as aspirated stops in pre-spirantization Germanic, too, so the Celtic loanwords in Germanic would have been affected by spirantization as we observe.

On Verner: it's generally assumed that spirantisation already had happened when Verner worked, as similar voicing of spirants depending on the accent has been observed elsewhere.
Generally, the Negau helmet (ca. 4th century BC) is taken as the earliest attestation of the Germanic consonant shift, so names like Cimbri and Teutones (1st century BC) would be Celtic exonyms preserving the unshifted consonants.

Yep.