Talk:Proto-Celtic language/Archive 1

Proto-Celts?

If there was a Proto-Celtic language, theoretically the people who spoke it would have been labelled as Proto-Celts.

I'm curious as to what, if anything, has been pieced together about these people. Gringo300 08:53, 13 Jun 2005 (UTC)

It is notable that compared with the proto-Indo-Europeans, the proto-Celts lasted as a unified people much less. The Celts split from the Proto-Indo-Europeans around 1500 BC and by 800 BC the Celtic expansion already began, so the Celts (which were living from the Atlantic coast to the Black Sea coast) hadn't time to create a very unified culture and many of common features were probably inherited from the Proto-Indo-European culture.

However, there are some interesting articles which may containe some information on them: Urnfield, La Tène, Celtic mythology bogdan ʤjuʃkə | ^Talk 11:39, 13 Jun 2005 (UTC)

Overjoyed

I am overjoyed that this article which I dared to create is generating discussion about a subject that, being of some Welsh ancestry, I hold dear. May I extend my thanks to those who have enhanced the layout of this page. Some experts (Trask, Michelena) have mentioned lexical and grammatical similarites between Celtic and Basque. There is reason to believe that Basque related languages were the substratum of the Celtic languages spoken in Gaul, Iberia, Hibernia and Albion. Proto-Celtic already had words bearing similarities to known Basque ones: Celtic *Belego- 'raven' looks a lot like Basque bel-hego black wing, for example. The languages are not genetically realted, but there is no reason to assume that interaction between the cultures did not happen. GeoffMGleadall 5 July 2005 22:37 (UTC)

Tabular Summary

I like the article, but the phonemes could be arranged in a better way I think. We could either group them according to the place of articulation, or common feature changed, or whatever. The order used currently is a bit chaotic, isn't it? :) --Pet'usek 7 July 2005 18:49 (UTC)

Is it? It's a familiar enough pattern to me: voiceless stops arranged by place of articulation, then voiced stops, then voiced aspirated, then fricatives, nasals, liquids, and glides. Right up the sonority hierarchy. --Angr/_{tɔk tə mi} 7 July 2005 19:18 (UTC)

Oh, I'm sorry. I'm totally blind today :). I'm using a different arrangement in my table ({p,b,bH},{t,d,dH},{k,g,gH},{k',g',g'H},{kW,gW,gWH},{s},{r,l,m,n},{r.,l.,m.,n.},{y,w} etc.), which must have puzzled me :))) I think that my arrangement might be better in some respect though. :)

Transition to Welsh

The table is a nice addition, but unfortunately mixes together the transition from proto-Celtic to Common Brythonic and the later (and historically attested) transition from Common Brythonic to Old or Middle Welsh. I suggest splitting the table, and putting the second half under either Brythonic languages or under Welsh. The second half can then also usefully discuss the chronology of the various changes. --Nantonos 22:14, 26 July 2005 (UTC)

One of these days I intend to start an article on the Proto-Brythonic language, and this table would be perfect there. --Angr/_{tɔk tə mi} 22:31, 26 July 2005 (UTC)

Since there is now an article on the Brythonic languages, I've moved this table over to there. Paul S (talk) 20:01, 28 March 2010 (UTC)

temporarily dead links

Just for anyones info: the links to the protoceltic dictionaries are temporarily closed for web reorganization. If you find this my observation to be obsoleted, then please make a note below. Rursus 17:35, 29 December 2006 (UTC)

Lay-out a mess

Can someone please correct the messy lay-out at the start of the article? Homun 10:08, 26 July 2007 (UTC)homun

Layout looks fine to me. You must be having browser issues. —Angr 15:02, 26 July 2007 (UTC)

Phonology: fate of initial sr-, spr- and medial s

Could someone take a look at the phonology : ProtoCelt -> Welsh section and review the /s/ entries. Surely there needs to be an entry for (either) sr- (and/or poss spr-) > W "ffrwd", "ffer"?

What about loss of medial -s- in early Welsh? (See Jackson LHEB?)

We should also be clear about what it is that the table is mapping the transition TO. Are we talking about the transition to _Modern_ Welsh? (Safest thing, surely, as we then don't have to worry about periodization and datings.)--CecilWard (talk) 06:12, 11 March 2008 (UTC)

Yeah, the s-clusters should be added. But as far as the consonants are concerned (and that's practically all that's discussed here), there's no real difference between Middle Welsh and Modern Welsh. Old Welsh is more difficult since lenited consonants are usually spelled the same as unlenited ones. —Angr ^{If you've written a quality article...} 06:57, 11 March 2008 (UTC)

Proto Celtic > Old Irish

"These changes may be summarised as follows" -- heh, that list would scarcely be less summative if it were just a list of every etymon and its outcome. It would be much more readable if it were consolidated some, e.g. with a unified statement of all these V+stop+sonorant+(j)V rules. I could do it but I'd have to read through it all and see what's going on and check slowly to sure I didn't accidentally exclude a case etc. Anyone care to do it who's familiar with the changes? 4pq1injbok (talk) 03:32, 29 August 2010 (UTC)

Okay, so the lines are gone. But they were useful information (to say they were "not relevant" is indisputable hyperbole), and they ought to be preserved somewhere in some form. Maybe I'll just have to, one day. 4pq1injbok (talk) 15:02, 25 January 2011 (UTC)

Well, this article isn't the place for it in any form. Maybe as a section of Old Irish, or maybe a separate article on the Phonological history of Old Irish. But not here. —Angr (talk) 23:29, 26 January 2011 (UTC)

Intro paragraph

Two points had to be changed: First, Celtic being a "centum-language" is correct, however, an outdated criterion for branching IE (cf. Meier-Brügger, Indogermanische Sprachwissenschaft, different editions, also in English). Second, Gray and Atkinson are neither linguists nor archaeologists, and the time of split between "Irish" and "Welsh" (in their terms) was an input to calibrate their computations, not the result. HJJHolm (talk) 17:38, 5 December 2010 (UTC)

Seriously?! In that case, I can only say they are totally bonkers. Why exactly do they believe that the Irish-Welsh split is six millennia old? And where does the uncertainty of ±1500 years come from? And the number 6100 instead of 6000? That doesn't make sense if they used the figure as input. If they had used the French-Romanian split as input, putting it at 2000 BP (which would have actually made sense), then it would have been idiotic to say "OK, we aren't really sure, perhaps French and Romanian were already diverging in 1500 BC, or perhaps they were only starting to diverge in 1500 AD". *facepalm*

Just to give you an idea: The earliest texts in Romanian handed down to us are from the 1500s and look virtually identical to 20th century Romanian (apart from the script, and the fact that 20th century Romanian has taken over many words from Latin and Romance languages to replace borrowings from Slavic mainly), and French was already in the Middle French period by the time and differed from 20th century French only in details, mainly some points of pronunciation, lexicon, and of course spelling. On the other hand, 1500 BC is a realistic age for proto-Italic.

Actually, if you check Kenneth Jackson's chronology of British Celtic sound change in Language and History in Early Britain, there is no difference but the P/Q-Celtic isogloss that would have been present already in the 1st century; apart from that, British and Goidelic would seem to have been identical. So chance is that the split between Irish and Welsh is barely 2000 years old, and it is quite unlikely that it is older than 2500 years, with proto-Celtic most likely dating into the first half of the first millennium BC, the time of the Hallstatt culture. An assumption of 6100 BP for the split is not only utterly implausible, it is outright crazy. Agriculture was introduced to Ireland ca. 4000 BC – do they simply assume that Celtic was introduced along with agriculture (for no compelling reason at all, and in the face of more than one reason militating against the assumption), and that pre-Irish immediately diverged from pre-Brythonic?

I would really like to know what your claim is based on. I'm not familiar with the original paper, but it is exceedingly curious. --Florian Blaschke (talk) 22:15, 18 February 2011 (UTC)

It is also worth pointing out that from the little we can recover about Primitive Irish and the British language at the time of the Roman occupation, they were both remarkably different from their medieval successors and much more similar both to each other and to Continental Celtic languages, especially Gaulish, as well as generally similar to languages such as Latin, Greek and Germanic as well; it must be concluded that Irish and Brythonic acquired their peculiar structure in a very short time – in about the 5th and 6th centuries for the most part. If Primitive Irish and British were still so similar to each other (not to mention their closest and more distant contemporary relatives) at the time of the ancient Roman Empire, it stretches belief to assume that they had remained virtually unchanged for 4000 years! --Florian Blaschke (talk) 23:43, 18 February 2011 (UTC)

Is *ɸ actually reconstructable?

As far as I know, this phoneme disappeared everywhere, except in the combinations *sɸ and *ɸt. It seems to me that that means that this phoneme isn't actually reconstructible from Celtic evidence alone, because Celtic doesn't show any reflexes of this phoneme appearing anywhere. It can only be reconstructed by means of cognates elsewhere in PIE. Doesn't that make the reconstruction more artificial and uncertain? What if words have no non-Celtic cognates? How do we know they did or did not contain *ɸ? CodeCat (talk) 15:32, 16 March 2013 (UTC)

You're right, it disappeared everywhere except in *sɸ and *ɸt, and those two clusters are the main reason *ɸ is included in Proto-Celtic reconstructions—actually, just *sɸ since *ɸt could otherwise be reconstructed *xt. And no, if a reconstructed word has no known relatives outside Celtic we have no way of knowing if a *ɸ was there or not (except in *sɸ). Angr (talk) 19:02, 16 March 2013 (UTC)

So as far as we are able to know, what we call Proto-Celtic may actually have had no /ɸ/ phoneme in it, except in the combination /sɸ/ which should then presumably be considered a single phoneme. I think that is important information for this article because it currently says with relative certainty that Proto-Celtic itself still had *ɸ but gives no evidence for it. Is there anything from outside Celtic, such as loanwords or attestations of words, that are able to show the relative dating when it was lost? Supposedly Germanic *laudą "lead" was a Celtic borrowing from PIE *plew-; if that's true then *ɸ was already gone from Celtic when this word was borrowed (because Proto-Germanic had a *ɸ itself, at least after Grimm's law). CodeCat (talk) 22:30, 16 March 2013 (UTC)

Even /sɸ/ may not have existed; Kim McCone thinks it was probably phonetically [sp] and phonemically /sb/, the idea being that p→ɸ didn't happen after /s/, just like Grimm's Law didn't, leaving the environment after [s] as the only place where [p] survived, so that it was reassigned to the phoneme /b/ after /p/ had disappeared everywhere else. Peter Schrijver, on the other hand, believes in [sɸ] because both (1) the merger with /sw/ and subsequent development to /s/ in Irish and (2) the development to /f/ in Brittonic is more likely if we start with [sɸ] than if we start with [sp]. As for external evidence, I don't know of any for /ɸ/ directly, but the /h/'s in Latin place names like Hercynia and Hibernia may be evidence for a /h/ that descended from /ɸ/. Angr (talk) 18:11, 17 March 2013 (UTC)

Schumacher (2004) has an entire chapter devoted to this issue. Eska has suggested that the V (the digamma) in Lepontic UVAMOKOZIS is to be read as /ɸ/, and Schumacher accepts this, pointing out that in Lepontic inscriptions, /w/ is always rendered as U, never V. A fortiori, the phoneme should have been preserved as such in Proto-Celtic in all other positions, because the intervocalic position is the weakest one. --Florian Blaschke (talk) 17:15, 7 February 2014 (UTC)

I just found a new reference directly relevant to this issue on the Bryn Mawr Classical Review site:

Joseph F. Eska, ‘In defense of Celtic /φ/’, in Adam I. Cooper, Jeremy Rau, and Michael Weiss (edd.), Multi nominis grammaticus: Studies in Classical and Indo-European linguistics in honor of Alan J. Nussbaum on the occasion of his sixty-fifth birthday. Ann Arbor: Beachstave Press, 2013, p. 32–43. --Florian Blaschke (talk) 14:16, 9 February 2014 (UTC)

I'd like to add that Matasović also supports the reconstruction of a Proto-Celtic phoneme */ɸ/ (he notates it as *f, but concedes that it may have been a bilabial fricative phonetically rather than a labiodental one). See also Wiktionary's take.

Schumacher does explicate that */ɸ/ – although arguing that it was credibly still retained as such in Proto-Celtic – eventually changed into other sounds in most positions (where it was not lost entirely), and merged with other phonemes: with */b/, */w/, and – before */t/ and */s/ – with */k/; the phoneme */k/ was realised as *[x] in this position. Therefore, in the individual Ancient Celtic dialects, it became a marginal phoneme eventually existing only after */s/, and in Insular Celtic, phonologically speaking, it eventually became what could be conceived of as an allophone of */b/. (Proto-Celtic reconstructions in Wiktionary notate only *x rather than *ɸ before *s and *t, implying the assumption that the phoneme had already merged into */k/ in this position.) --Florian Blaschke (talk) 10:19, 4 November 2019 (UTC)

ei > ē

Don't some continental Celtic inscriptions still have ei in some endings? CodeCat (talk) 21:50, 16 September 2013 (UTC)

Yes. As Schumacher (2004) points out, Celtiberian and Lepontic both preserve ei, for example in Celtiberian loc. (!) sg. tokoitei and Lepontic dat. sg. PIUONEI and ATILONEI. --Florian Blaschke (talk)

If this is the case, then the vowel table shouldn't have ē in it, should it? I'm thinking another scenario is also possible. Latin is known to have lost "long diphthongs" but they were preserved in absolute word-final position. Thus, it preserves the distinction between word-final *ai and *āi (Latin ī and ae respectively) and *oi and *ōi (Latin ī and ō). It's possible that such a scenario also applied to Celtic, meaning that ēi was preserved as such in absolute final position. After that, ē > ī, ei > ē, and then ēi > ei. This is just my own speculation, but either way, the article currently contradicts itself by positing an ē for Proto-Celtic. CodeCat (talk) 01:17, 17 April 2015 (UTC)

Passive voice

Is the passive voice attested in any Celtic text? Or was it only reconstructed because of Proto-Celtic's apparent similarity to Latin and other Italic languages? Steinbach (talk) 07:34, 12 May 2014 (UTC)

The first-person plural and third-person plural passives of Old Irish are true passives (grammatical patient in the nominative); the third-person singular passive is already becoming an impersonal "autonomous" form (grammatical patient in the accusative). Angr (talk) 09:35, 12 May 2014 (UTC)

Proto-Celtic Reconstruction Etymology

I have been searching for the etymology of the reconstructed Proto-Celtic Wirdjos. I believe I have traced the first mention of the reconstruction to Wikipedia and I was hoping some here might know more about where it comes from. The information I am most interested in is the justification for reconstruction and the breakdown of the meaning. However, I would also be interested in proposed pronunciations and any side information anyone here might possess. Thank you for any help you can offer. Wirdjos (talk) 19:38, 6 September 2014 (UTC)

External links modified

Hello fellow Wikipedians,

I have just modified one external link on Proto-Celtic language. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20110711231504/http://www.indo-european.nl:80/cgi-bin/main.cgi?root=leiden to http://www.indo-european.nl/cgi-bin/main.cgi?root=leiden

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at {{Sourcecheck}}).

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 5 June 2024).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 05:03, 13 November 2016 (UTC)

Minor change on oldest Celtic literature

The unsourced statement that the oldest Celtic literature is in Old Irish appears to be a confusion of Celtic with Goidelic languages. Wikipedia's various literature articles show Old Irish works can be dated to the 7th century, whereas the Middle Welsh Cynfeirdd (recorded in the Tudor period) are firmly 5th or 6th century.

I have updated the relevant section accordingly, but happy to discus.

Cymrogogoch (talk) 10:40, 28 April 2017 (UTC)

Morphology

I'd be happy to be pointed to the source, which the reconstructed declensions and conjugations come from. I can't seem to find them on my own. Anyone? Michalite (talk) 10:24, 12 September 2017 (UTC)

Neolithic datings for Proto-Celtic

Considering that these datings precede the generally accepted dating of c. 3000 BC for the start of the divergence of Indo-European itself (at least if Anatolian and Tocharian are left out), that Celtic is highly similar to Italic (with especially reconstructed Proto-Celtic strongly resembling Proto-Italic) and may therefore well have diverged from Italic even later than that, the dates proposed by Gray/Atkinson and Forster/Toth are well outside the realm of plausibility. Worse even, at least the Gray/Atkinson paper did not date the split-up of Proto-Celtic to 4100 BC, but its split from its closest relatives, so it cannot be treated as a dating for Proto-Celtic!

While there is no certainty about the dating of Proto-Celtic, specifically the lexeme *īsarnom generally reconstructed for Proto-Celtic to mean 'iron' does strongly point to the Iron Age, which is supported by the low internal diversity of Ancient Celtic dialects and their evident resemblance to Proto-Celtic reconstructions. There is some debate about an alleged Celtic identity of Bronze Age archaeological cultures in Western Europe, which I'm highly sceptical about, but at least it seems to be treated as a legitimate topic of debate in the field. However, the datings suggested by Gray/Atkinson and Forster/Toth have been widely panned in the field as implausibly early, at least informally, such as on Language Log, or implicitly, through newer papers supporting low datings for Proto-Indo-European and refuting the Anatolian hypothesis and its implied or explicitly stated datings for Proto-Indo-European. --Florian Blaschke (talk) 09:49, 4 November 2019 (UTC)