When was Proto-Celtic spoken (offtopic from Beaker-Bell R1b)

Taranis

Elite member
Messages
2,381
Reaction score
433
Points
0
Location
Germany
Y-DNA haplogroup
R1b1b2a* (inferred)
Ireland is an endpoint and endpoints are good to test hypotheses.

The last of the four Megalithic building waves was the Wedge Tomb which briefly overlapped with the Bronze Age people. In this Bronze Age, the artifacts are Bell Beaker, which are found uniformly throughout the island in contrast to the Megaliths which are mostly in the north. The certifiably Celtic La Tène culture is also found entirely in the north, and the south is without any Iron Age artifacts at all.

No one has found any evidence of any "Celtic Invasions" of Ireland, yet the people are there. It has been pointed out that many of the pre-Christian royal sites (really more rex as in the Golden Bough, than king), are associated with Megalithic sites, especially Tara and the great Boyne Valley monuments. This seems to me to indicate a certain cultural continuity through the centuries, as opposed to some Celtic Invasion from Iberia who stumbled upon Newgrange.

So it has always seemed to me, that the only candidates for the presumed pre-Celts were the Bell Beakers, for if not them, then who?

These new R1b results are very intriguing.

You bring up some very interesting ideas there. I absolutely agree that the position of Ireland is intriguing, and the origin of the Celtic language in Ireland is puzzling.

As I mentioned before, there are several unanswered problems with the idea that Beaker-Bell was Proto-Celtic: first is the sheer area that it covered (since areas such southern Scandinavia, Sardinia, Sicily, Southeastern Iberia, North Africa, etc. were never Celtic), the second is the ancientness of Beaker-Bell.

Beaker-Bell existed circa 2,000 years before the earliest attestations of Celtic languages, and if we compare the Celtic languages in Antiquity (Lepontic, Celtiberian, Gaulish, Ogham Irish - especially the parallels between the latter two are particularly stunning), they are all fairly similar to each other (and paradoxially Oghamic Irish, from the 4th century AD is in many ways the most conservative of them all). If we compare language families that are of a roughly comparable age (ie, the Germanic and the Romance languages), they have diverged to a much greater degree. From that perspective, I find it highly dubious to argue that the Celtic languages began to diverge in the 3rd millennium BC if they are so similar to each other. As an analogy, this is like arguing that Vulgar Latin was spoken around 4000 years ago.

I absolutely agree however that the Celtic language presence in Ireland can be neither explained by a La-Tene invasion/immigration (this can be convincingly argued for P-Celtic Britain, but definitely not for Ireland!), nor by some kind of late immigration from Iberia (I'm talking about the Mil Espaine from the Book of Invasions, which are almost certainly a medieval fabrication - besides if anything I think this happened vice versa :LOL: ).
 
Last edited:
Beaker-Bell existed circa 2,000 years before the earliest attestations of Celtic languages, and if we compare the Celtic languages in Antiquity (Lepontic, Celtiberian, Gaulish, Ogham Irish - especially the parallels between the latter two are particularly stunning), they are all fairly similar to each other (and paradoxially Oghamic Irish, from the 4th century AD is in many ways the most conservative of them all). If we compare language families that are of a roughly comparable age (ie, the Germanic and the Romance languages), they have diverged to a much greater degree. From that perspective, I find it highly dubious to argue that the Celtic languages began to diverge in the 3rd millennium BC if they are so similar to each other. As an analogy, this is like arguing that Vulgar Latin was spoken around 4000 years ago.

Aren't there conservative languages that doesn't change too much over time like Greek, Lithuanian etc?
 
If we compare language families that are of a roughly comparable age (ie, the Germanic and the Romance languages), they have diverged to a much greater degree. .

Do you really think that Germanic and Romance languages have the same age? Just because the Grimm law is contemporary to the Roman empire doesn't mean that some sort of Germanic languages (without exactly the same sounds'laws but related anyway) weren't spoken in the bronze age. It can be exactly the same for Celtic languages.

Btw, celtic languages were not so similar. Typically, Lusitanian and Ligurian diverged from other Celtic languages very early.
 
Do you really think that Germanic and Romance languages have the same age? Just because the Grimm law is contemporary to the Roman empire doesn't mean that some sort of Germanic languages (without exactly the same sounds'laws but related anyway) weren't spoken in the bronze age. It can be exactly the same for Celtic languages.

Btw, celtic languages were not so similar. Typically, Lusitanian and Ligurian diverged from other Celtic languages very early.

(I've decided to move this into a different thread, because I felt this was straying too far off-topic from the original thread)


Well, let's recall what the "proto" means: it's the reconstructed ancestor language of all known branches. In the case of the Germanic languages, this encompasses East Germanic (principally Gothic), North Germanic (the Scandinavian languages) and West Germanic (ie. Anglo-Saxon, English, Frisian, Dutch, German, etc.). The reconstructed ancestor language of these is called "Proto-Germanic", and it's usually reconstructed to have been spoken around the 1st century BC/AD. The defining feature of Proto-Germanic, which sets it apart from all other branches of Indo-European is Grimm's Law. Theoretically, you do obviously have to consider the developments that took place between the divergence of Proto-Germanic into it's daughter languages, and the point when Proto-Germanic became a separate branch of PIE. This time frame we could refer to as a "Pre-Proto-Germanic" language stage(s). Theoretically, we could conceive languages to branch off there.


If we now look at the Romance languages, we actually are lucky to know the ancestor language, Latin, and we know also a number of extinct relatives (the old Italic languages, such as Faliscan, Oscan, Umbrian, and a few other ones which are poor-attested). Oscan and Umbrian, of course, are Italic languages (like Latin), but they are obviously not Romance languages.


What does this mean for the Celtic languages? One of the unique features that all Celtic languages have in common is the loss of the Proto-Indo-European *p sound (compare Irish "athair", Latin "pater" and English "father"). In contrast to this, Lusitanian preserves PIE *p (compare Lusitanian "porcom", Gaulish "orcos", Old Irish "orc", German "Ferkel").


The take of some scholars (notably Untermann, 1987) on Lusitanian was to suggest because of this that the Celtic languages should be redefined to include the *p. Note however that this generates a circular logic: if we say Proto-Celtic isn't defined by the loss of *p (amongst other things), then Lusitanian suddenly becomes a Celtic language. However, all other Celtic languages still have it in common (amongst other things) that the *p is lost. So from that perspective, we might rather say Lusitanian is "Para-Celtic" instead.


If we follow the Lusitanian "vibe", we might say that Oscan and Umbrian are "Para-Romance" languages, but there's no way that Oscan and Umbrian are "full" Romance languages: all known Romance languages still have commonalities with each other that they don't share with Oscan and Umbrian, and vice versa. We might also hypothetically conceive a "Para-Germanic" language for which Grimm's Law doesn't apply that coexisted simultaneously with Proto-Germanic, but this language would not be a proper Germanic language because Grimm's Law doesn't epply. Unless we suddenly redefine the Germanic languages and say that Grimm's Law is not the unifying feature of all Germanic languages.

So, what does this mean for Proto-Celtic? Obviously that Lusitanian (if we take it really as a "Para-Celtic" language, which is an opinion that is not shared by all authors) was the first branch to diverge and that all other Celtic languages are more closely related with each other than they are with Lusitanian. If we consider this, it's really questionable to push the date for when Proto-Celtic was spoken as much back as to 2500 BC.
 
(I've decided to move this into a different thread, because I felt this was straying too far off-topic from the original thread)


Well, let's recall what the "proto" means: it's the reconstructed ancestor language of all known branches. In the case of the Germanic languages, this encompasses East Germanic (principally Gothic), North Germanic (the Scandinavian languages) and West Germanic (ie. Anglo-Saxon, English, Frisian, Dutch, German, etc.). The reconstructed ancestor language of these is called "Proto-Germanic", and it's usually reconstructed to have been spoken around the 1st century BC/AD. The defining feature of Proto-Germanic, which sets it apart from all other branches of Indo-European is Grimm's Law. Theoretically, you do obviously have to consider the developments that took place between the divergence of Proto-Germanic into it's daughter languages, and the point when Proto-Germanic became a separate branch of PIE. This time frame we could refer to as a "Pre-Proto-Germanic" language stage(s). Theoretically, we could conceive languages to branch off there.

You have still failed to address the centtral and southern areas of Germany which where initially non-germanic in language. The Jutes, Angels, saxons, frisians all lived around the northsea coast from jutland to netherlands.

If celtic is too young for the bell-beaker , Únětice periods, then what was spoken in this area. ? Where they gallic people ?

http://en.wikipedia.org/wiki/Continental_Celtic
 
If we now look at the Romance languages, we actually are lucky to know the ancestor language, Latin, and we know also a number of extinct relatives (the old Italic languages, such as Faliscan, Oscan, Umbrian, and a few other ones which are poor-attested). Oscan and Umbrian, of course, are Italic languages (like Latin), but they are obviously not Romance languages..


What relations are there between Romance language family and Italic languages?
Romance languages are part of a bigger family (Italic languages family) and are thus closely related to non Romance Italic languages like Oscan etc.
Do you agree with me that the relationship between Irish (or Gaulish? Breton etc) and Lusitanian is exactly like the relationship between French and Oscan. Both French and Oscan are part of the Italic languages.
Then it is possible to make the same "bigger language family" to link Lusitanian and Celtic languages.
Even without the loss of "P", Lusitanian and Celtic languages are obviously relatives.


So my guess is that Celtic languages formed during the Iron age and the bigger language family (Lusitanian+Celtic languges+ Nordwestblock) formed during the Atlantic bronze age.
This is because it formed so long ago (bronze age) that Luitanian, Nordwestblock and Celtic languages diverged and formed themselves separates languages families.

The Atlantic bronze age and Unetic culture formed both out of a Bell Beaker substratum. There is indeed nothing between Bell Beaker culture and Atlantic bronze age in the Atlantic fringe of Europe as there is no culture between Unetice and the Bell Beaker in central Europe.

So that bigger Celtic language family (Lusitanian+ Celtic + Nordwestblock) probably formed long before the iron age, (because in the iro age Lusitanian, Celtic and Nordwestblock laguges were already separate) tha is to say at least during the bronze age. And as I said, the Bronze age cultures where later Lusitanian, Nordwesblock and Celtic languages would be spoken are all incuded in the Bell Beaker cultural complex.
 
its surprising that western Romance languages has celtic influences , while southern romance does not

http://en.wikipedia.org/wiki/File:Romance-lg-classification-en.png

http://en.wikipedia.org/wiki/File:Western_and_Eastern_Romania.PNG

The question most interesting to me is when did celtic enter the britsh isles and affect these areas.
Does picone in france represent the picts and a possibity the language transferred from there


Southern French , Occitan language has nothing in common with the southern italian Oscan in regards to dialect continuum

It still posses the question in the bell beaker times as to what was spoken in southern Germany prior to the migration of the germans from the north
 
its surprising that western Romance languages has celtic influences , while southern romance does not

http://en.wikipedia.org/wiki/File:Romance-lg-classification-en.png

http://en.wikipedia.org/wiki/File:Western_and_Eastern_Romania.PNG

The question most interesting to me is when did celtic enter the britsh isles and affect these areas.
Does picone in france represent the picts and a possibity the language transferred from there


Southern French , Occitan language has nothing in common with the southern italian Oscan in regards to dialect continuum

It still posses the question in the bell beaker times as to what was spoken in southern Germany prior to the migration of the germans from the north

I am not so surprised at all: the western romances languages are surely a latin expansion through celtic lands but classifications are often a matter of criteria choices- phonetically speaking, Corsican and Sardinian show some akiness with western romance area: strong lenition of stops between vowels, drop of doubled stops (geminees), excepted /ts/= z'z, 'zz' - as in Iberia romances and in Gallo-italian dialects - they know even a lenition at the initial of words as in present day celtic languages, phenomenon ignored by other western romances languages -
corsican 'Casanova': /kazanowa/ - 'la casa': /la gaza/ -
'in piazza à a ghjesgia (chiesa) trova à zia Catalina': /in pjatsa a a jézha drowa a tsia gadalina/
* sorry I have no IPA at hand: /j/ = english 'y' - /zh/ = french 'j'
on an other hand, northern corsican, as SW occitan and castillan and basque, pronounces a kind of spired /B/ or an occlusive /b/ in place of /v/ and castillan as a whole shows strong basque influences phonetically (vowels and consonnants) as the gascon dialect of Occitania - and sardignan and southern corsican and southern italian show an evolution /ll/ >> /tt/ or /dd/ as in gascon dialects
italian 'bello' (beautiful) >> corsican S, sardinian, sicilian 'beddu' - 'gallo' (cock) >> gascon 'gatt' - 'valle' >> gascon 'batt'
for centralization or palatization of 'o' /ö/ and 'u' /y/ = german /ü/ only french dialects and occitanian and gallo-italian ones show it -
what remains sure is that toscan and central and southern italian dialects kept very different from other romances languages as a whole - but I do not see any amazing result here according to history?
 
I add to come back at the topic that I would be very glad if someone could provide us some of the words that show the proximity of Lusitanian with celtic: because (maybe it is "******* the flies" (sorry, I am confused)) i feel lusitanian as a possible branch of a proto-celtic-italic stage encompassing ligurian too, without any proof by lack of documentation
 
Why not the bell-beakers? From the hydronimyc names to the celtic languages there is a long way, is not?
Are the Celtic languages only the romantic and endocentric point of view of the recent gaulish and insular languages?
Why gaulish and insular celtic languages loss one of the principal indo-european features, the sound *p? An interference by a non indo-european substratum how Mac Eoin thinks?
How can you explain the celtiberian [-]VAPORCONI, with the word *PORCOS?
What can you say to Patrizia de Bernardo Stempel when she comments about western indo-european languages of Iberia:

“[…] formado en época muy antigua dentro de un entorno linguístico céltico sin ser luego alcanzados por la más moderna eliminación de la p- […]. Lo que no sabemos es la procedencia exacta de este tipo de celta, ni tampoco si lo aprendieron en la Peninsula Iberica o si lo trajeron consigo desde otras regiones europeas“. (P. de Bernardo Stempel, 2009)

Why the loss of *p is late in lepontic: cf. Eska, 2010 or in the pictish oghamic inscriptions: pictish NEHTAN/NEHHTAN vs. irish oghamic NETAN from lat. nepto- (Griffin, 2008)?

What can you say about in Gallia we can see an invariable /p/ in oldest gaulish items like PALONI, PICOS, PINCIOS, PLATIODANOS, PLAVMORATI (cf. galician lávego < *(p)lavaeco-), Paedocaeus, Paemani, Paetinius, Paetinus, Paettusius, Pama, Pameta, Pamius, Pansiana, Pansius, Panto, Panturo, Parameius (f. hisp-celtic PARAMAECO, celtiberian PARAMICA), Paranus, Parasenus, Parassius, Parra, Parridius, Parrio (cf. callaecian place-name PARRICA, today Parga), Peintius, Penti, Pentilius, Pentis, Pentius, Pentodia (same word that celtiberian and hisp-celtic PENTIUS; PENTILUS, etc.), Peisius, Pel(l)ius, Pelidianus, Pellaeus, Pellic[ius, Pellius, Pelto, Penci, Pessiacus, Pessicinus (same word celtiberian and hisp-celtic PAESICUS, PESICUS), Pisulus, Pisus, Pitius, Pitulus, Piturix (cp. callaecian PITILUS, celtiberian PITANA), PLATODANIUS, PLASSARUS, PLASSUS, PLOXENVM, POEMANIUS (cf. callaecian godess POEMINA), POENINA, POETOVIO, POINAI, POLOS, POMPE (cf. lusitanian PUMPE), PORCIA, PORCIUS, PORCUS (cf. lusitanian PORCOM, celtiberian [-]VAPORCONI), PRENIO (cf. call. PRAENIA), PRICASSES, PRIGENUS, PRICUS, PRIGA, PRITI, PRITMANUS, PRITO, PRITONIUS, PRITILLIUS, PRITTIO, PRITO, PRIUNUS, PROCALLIUS, PROGENUS (lusitanian PROGENUS), PRUEINI, PIXTILVS, Pixtillus, Pistillos, Pistillu, Pistillus, PIXTVGENOS, Pixticenus, Pixtacus, PICTAVI (or PICTONES) similar to the callaecian PICTOLANCEA, celtib. PISTIROS from the ie. root *pik-to-, etc.. and in Lepontica, whether as P (plosive labial voiceless) or as PP (plosive labial voiceless geminated) and with graphematic representation PI: ]PEUESA, KOPLUTUS (cp. celtiberian, callaecian and asturian COMPLEUTICA, COMPLEUTO), PISA, ]PLIOISO, OIEPLU, KAPUTUS (cp. lusitanian CAPORUS), PLAI P, KOP (cp. call. COPORI), POLIOS, PUSIONIS (cp. celtiberian, callaecian and lusitanian PVSINA), APIOS (cp. celtiberian, callaecian and lusitanian APANA, gaulish APANADEVA), KEPIOS, KIPODIS, POIKANOS, IAPOS etc. and over 45 mentions with PALA ‘(grave)stone’.

IS IT NOT CELTIC? NOT GAULISH?
 
Callaeca, we've had this discussion before in another thread, and the same criticism that I have to your view on the matter still applies.

- Several authors (notably Martinez 2006, Curchin 2007) point to the existence of clearly non-Celtic but Indo-European (in the case of Gallaecia, Martinez also suggests an additional stratum of non-Indo-European languages) names that coincide next to Celtic names within the Celtic context on the Iberian peninsula. The only logical explanation for this heterogenity is that Celtic languages are a later, foreign-introduced element here, upon an older Indo-European stratum that developed indigenously in the west of the Iberian peninsula.

(for reference, Martinez 2006: "We do have some unequivocally Celtic place-names, as has been demonstrated here. It is striking, however, how many names in -briga-/-bris show a non-Celtic initial element, indicating that at their arrival the Celtic populations used non-Celtic onomastic elements already existing in the area to create the new names of these settlements.").

- The loss of Indo-European *p (it should be very clear to make that distinction) in the Celtic languages is attested in all branches of Celtic, including Celtiberian (cf. place names "Argaela", "Octaviolca", "Tenobriga", "Uxama").

(for reference, compare Curchin: "Octaviolca. Ptolemy (II, 6, 50) attests this Cantabrian town. The element olca is Celtic for “ploughed field” from IE *pelk- “to turn”.")

(likewise: "the likeliest etymology for Tenobrica is “hot fort” or “fire fort” (a hilltop site from which fire signals are sent, cf. Old Irish tene, Welsh tân “fire”) from Celtic *tep-no- “hot”.")

- The concept of a non-indo-european substratum does not explain how PIE *p doesn't simply vanish, but leaves specific reflexes in the presence of other obstruents (*χ, *b). The loss of *p can be best explained by a successive shift *p > *φ > *h > Ø. That development may be unique in the Indo-European context, but outside out Indo-European, there's a good analogue for such a sound shift found in the Japonic languages. In Japanese the same sound is reflected as /h/ which in the Ryukyuan languages is reflected as *p. When the Portuguese explorers came to Japan in the 16th century, they recorded the words that are today written with an "h" with an "f". Modern Japanese actually retains /φ/ before /u/ (for example the word "Fukushima"). It's entirely plausible that the development in Proto-Celtic functioned analogous. It's conceivable that *φ was retained in Lepontic at intervocalic positions (Lepontic inscription CO.48, "uvamokozis", as opposed to the expected "upamokozis" if *p was retained). Likewise, there's instances where Gaulish may have retained an initial *h- (cf. ethnic name "Helveti").

- In both the Brythonic languages and in Gaulish, a new *p sound arises from PIE *kʷ (a process that is also seen in Osco-Umbrian and in Greek). It is clear that the development of *kʷ can only have occured after the development *p > Ø occured, otherwise we would observe cases of PIE *kʷ > Ø in the Celtic context.

- Additionally, there's multiple occurences of *p in Irish (a Q-Celtic language, after all) which is the result of borrowings. There is no reason why this should not apply for older Celtic languages too, but the prequisite is that this borrowing occured only after PIE *p was lost.

- Additionally, there are instances occuring in the Celtiberian context where *p actually represents PIE *b or *bʰ (cf. Celtiberian ethnic name "Pelendones").

- Lastly, all of the above makes more sense than arguing that the occurence of *p and *p > Ø could be in free variation with each other in the Celtic languages, which would be the unavoidable consequence of your idea.
 
Callaeca, we've had this discussion before in another thread, and the same criticism that I have to your view on the matter still applies.


thank you Taranis for this clear demonstration (necessary indeed) -
are all Ogamic inscription in Ireland well 'C-' (Kw-) documented?
 
For Churchin the percentages between celtic topomastic of Callaecia and Cantabria is not diferent: Callaecia 41%, Cantabria 41%. And the Celtiberians topomastics names has over 33% for Churchin, similar percentage than Lusitania 30%. The percentages of Callaecia and Cantabria are similars to Gaul and Britannia if we count, like in the Hispanic names, his concept of indo-european undifferentiatted and uncertains names (roots like *tam-, *ab-, *nav-, *arg-, *sal-, *sar-, *al-, *albh-, bib-, etc.). For Churchin there are not traces of non indo-european languages in the half west of Iberia. Prósper, Villar, Beltrán LLoris, Moralejo, Búa, Armada, de Bernardo Stempel, Untermann and other authors are in the similar way than Churchin.

The work of Luján is obsolete. For example,
It is incredible to consider the callaecian place-name Tude (variants Touda, Toude < *touta-/*touti-), today Tui, as not Celtic, when in the same area is considered as Celtic the god-name Toudadigoe (< *toutatikos), similar to the gaulish Teutatis, Toutatis, Tutatis. Can you explain me, please, what is the difference? Luján Martínez, for example, says that the callaecian river Limia is non celtic (cp. place-name LEMICA CIVITATE, today Xinzo da Limia, ethnic name gen.pl. LEMICOM 'of the Limici', territorio LEMETO in the Parrochiale Seuvicum, today A Limia), then non celtic are the gaulish LEMANA, LEMINCUM, LEMANNONIUS SINUS, LEMAUSUM, LEMANE/LIMENE or the gaulish ethnic LEMOVICES.

About the existence of the word olca < ie. *pelk- in hisp-celt. is false. The galician and castilian languages have olga and huelga but it is a medieval galicism. It is an absurd named an hill fort as 'ploughed field'. Now the option is ie.*h2olkeh2 'protection, fortification'. Argaela is derivated for *h1erg- 'white, silver' (an hydronymyc name) and Tenobrica is the same word that you can see in the galician river name Tea > *tena and galician place-names Tiobre < *tenobrigs, Tebra < *tenebriga.

Do you think that the celtiberian name Aplonos, Aplonios, the familiar name Aploniocum, from ie. *h2eplo- represents b?, and Pisturus from the root ie. *pikto-? Why not in western hispania this phonetic representation? You can find there Laboena and Lapoena, Abana and Apana, Abilus and Apilus, Lacibea and Lacipea, etc.

But that is not the issue (
I maybe make a new thread about the NW hispanic topomastic names). I have asked you why and not how the gaul-roman has lost the Indo-European *p., and why authors such as Eska, Bernardo Stempel, Delamarre, Untermann, etc. considered that proto-Celtic had *p.

(please not with examples of distant agluttinative languages like japanese. You have nearly examples in the Aquitanian, Iberian and Rethian urnfield cultures, l
anguages probably similar to which had talk in Central Europe, Northern and British Isles, prior to the incorrectly pronounce of the Indo-European)

I asked you
if you consider that the list of Gaulish and Lepontic names, which is extensible, are or not Celtic.
 
Last edited:
For Churchin the percentages between celtic topomastic of Callaecia and Cantabria is not diferent: Callaecia 41%, Cantabria 41%. And the Celtiberians topomastics names has over 33% for Churchin, similar percentage than Lusitania 30%. The percentages of Callaecia and Cantabria are similars to Gaul and Britannia if we count, like in the Hispanic names, his concept of indo-european undifferentiatted and uncertains names (roots like *tam-, *ab-, *nav-, *arg-, *sal-, *sar-, *al-, *albh-, bib-, etc.). For Churchin there are not traces of non indo-european languages in the half west of Iberia. Prósper, Villar, Beltrán LLoris, Moralejo, Búa, Armada, de Bernardo Stempel, Untermann and other authors are in the similar way than Churchin.

The work of Luján is obsolete. For example, It is incredible to consider the callaecian place-name Tude (variants Touda, Toude < *touta-/*touti-), today Tui, as not Celtic, when in the same area is considered as Celtic the god-name Toudadigoe (< *toutatikos), similar to the gaulish Teutatis, Toutatis, Tutatis. Can you explain me, please, what is the difference? Luján Martínez, for example, says that the callaecian river Limia is non celtic (cp. place-name LEMICA CIVITATE, today Xinzo da Limia, ethnic name gen.pl. LEMICOM 'of the Limici', territorio LEMETO in the Parrochiale Seuvicum, today A Limia), then non celtic are the gaulish LEMANA, LEMINCUM, LEMANNONIUS SINUS, LEMAUSUM, LEMANE/LIMENE or the gaulish ethnic LEMOVICES.

Forgive me that I am asking, but why are you overlooking the inconsistencies that both Curchin and Martinez describe and which have not escaped me either?

About the existence of the word olca < ie. *pelk- in hisp-celt. is false. The galician and castilian languages have olga and huelga but it is a medieval galicism. It is an absurd named an hill fort as 'ploughed field'.

This word isn't a "medievalism" as you call it, but attested in Greek/Roman sources.

Now the option is ie.*h2olkeh2 'protection, fortification'.

Can you suggest me comparable reflexes for this root in other IE languages?

Argaela is derivated for *h1erg- 'white, silver' (an hydronymyc name) and Tenobrica is the same word that you can see in the galician river name Tea > *tena and galician place-names Tiobre < *tenobrigs, Tebra < *tenebriga.

What is your etymology for "tenobriga" then?

Besides, there are other examples from Celtiberian which clearly attest that *p is lost in Celtiberian:

"Aregrada" (Celtiberian mint: "Areikoratikos", A.52). Why is the attested form not *Paregrada?

"Clunia" (Celtiberian mint: "Kolounioku", A.67), compare with Old Irish "clúain" ("meadow"). Also compare from Lithuanian "šlapias" and Latvian "slapjš" (both meaning "wet", "damp"). If *p was retained in Celtiberian, the expected form would be *Klepnia. Instead, the name shows the regular Celtic development *eu > *ou.

Likewise, Martinez brings more examples from Gallaecia:

"If the correct form is Arotrebae, as Pliny NH IV 114 explicitly argues, we would have here a compound of are-53 (with loss of the initial IE *p) plus a form of the stem *treb- 'live in, inhabit."

"Lamecensis (Alonso (2003: 126) as related to OIr. lám 'hand', from IE *pl̥ma or *plāma, with Celtic loss of initial *p-. )"

In contrast, per Curchin: "Segontia Paramica (the second word is Indo-European but non- Celtic)"

As I said, the consequence of your view is that somehow *p and *p > Ø can exist in free variation in the Hispano-Celtic context, which is obviously impossible.

Do you think that the celtiberian name Aplonos, Aplonios, the familiar name Aploniocum, from ie. *h2eplo- represents b?, and Pisturus from the root ie. *pikto-? Why not in western hispania this phonetic representation? You can find there Laboena and Lapoena, Abana and Apana, Abilus and Apilus, Lacibea and Lacipea, etc.

But that is not the issue (I maybe make a new thread about the NW hispanic topomastic names). I have asked you why and not how the gaul-roman has lost the Indo-European *p., and why authors such as Eska, Bernardo Stempel, Delamarre, Untermann, etc. considered that proto-Celtic had *p.

If we were to assume for a moment that you were right (which you are not, see above!) and PIE *p was retained in Celtiberian, and which Untermann et al. asserted, you're essentially just creating a "hen-and-egg"-type of problem: you basically define Proto-Celtic as retaining *p. But, in this case the development *p > Ø still then holds true as an innovation for all other branches of Celtic besides Celtiberian. That way, you are explaining nothing. It's just an arbitrary decision.

(please not with examples of distant agluttinative languages like japanese.

Why not? It's evidently an analogue for the development in Proto-Celtic.

You have nearly examples in the Aquitanian, Iberian and Rethian urnfield cultures, languages probably similar to which had talk in Central Europe, Northern and British Isles, prior to the incorrectly pronounce of the Indo-European)

So, if I am getting this right:

- you're basically asserting that the Iberian peninsula was Indo-European in the Bronze Age while the rest of Western Europe was not.

- you want me to believe that this an "explanation" for the loss of *p in the various Celtic languages?

Well, I have to disappoint you, that is very much impossible, primarily because it requires reversing the direction by which Indo-European languages are supposed to have spread:

- for one, how do you explain the presence of the Germanic languages which are generally assumed to have originated in northern Europe (the proto-Germanic homeland is usually thought in Jutland and southern Scandinavia)?

- how does it come that the extend of Iberian place names ends in the Languedoc if you go northwards, but extends across a wide arc all the way southwards to central-eastern Andalusia, into an area which was never part of the Urnfield Culture?

- There is no evidence for Aquitanian or Iberian place names anywhere in northern France, in the British Isles or in Germany.

- Raetian, as can be demonstrated from the inscription, was a language closely related with Etruscan (ie, Tyrsenian languages). Etruscan didn't have any voiced stops, instead distinguished between unaspirated and aspirated unvoiced stops, e.g. *pʰ *tʰ *kʰ vs. *p, *t, *k. Why would a loss of *p result from this if there wasn't even a *b phoneme in Etruscan? Besides, there's no evidence for Tyrsenian languages in Central Europe. This whole "Raetian" argument is entirely inconsistent.

- You're still leaving explained why the *p doesn't just disappear but leaves reflexes in the context of other consonants. Considering the obstacles I just describe, it seems to me that your assertations "they were non-Indo-Europeans and couldn't pronounce the *p right" appears to me just like a very dubious argument.
 
1. Taranis, Can you tell us the difference between *toutai - > Toude (with lenition) > Tude and the first element of Toutatis/Tutatis? The difference between in territorio Lemeto, today A Limia region, (with Celtic suffix -et-: cp. callaecian Nem-et-obriga, gaulish Mog-et-io) and the Gaulish place-name Lemane/Limene? What is the criterion to consider these callaecian forms as non-Celtic?

Churchin does not comment about the existence of non Indo-European languages in NW Hispania. On other hand his etymological proposals labeled as undifferentiatted or uncertain Indo-European have always parallels in other extrapeninsular areas. Luján Martínez's work is not bad, but it is strongly influenced by the University of Salamanca and based on previous criteria. Other better sources exist as Moralejo, Báscuas, Búa, Vallejo, etc. or the own Churchin.

2.
The Galician-Portuguese olga ​​and ancient Castilian huelga should arrive under Cluniac influences, through the Provençal olca 'ploughed field' or the ancient french dialect of Moselle region olke 'vineyard'. The reason is the hispanic romances can not change the consonantic group -lk- to -lg, then it is a secondary form. Only if it belong to a similar form like the old french dialectal variant olica (in the Formulae Senonenses, c. VIII), we can reconstruct the form olga and huelga: cp. galician place-name A Olga (not A Olca).

See ie, *h2olkeh2- (IEW 32, LIV 236) 'protection, fortification', goth. alhs 'fortified temple ', old english ealgian 'to protect', better to designate an hill fort.

3. The cantabrian place-name Tenobriga have the same ethymology than the galician river name Tea < Tena or galician place-names as Tiobre < (med. Teobre) *teno-brigs, Tebra < *tenebriga, Teis < *tenes, the same word,
I think, than La Tène. I have not a secure ethymology about this item, but it is an hydronymyc word: cp too the callaecian god name + incomplet place-name COHVE TENA = COSSUE TENA[ECO?] or TENA[BRI?] (COHUE presents the same solution like old welsh, S > H. Prósper and Bernardo Stempel point this phonetical fact like a new variety of celtic in the Central Callaecia Lucensis).

4. No I am not defining proto-celtic as retaining *p. I ask you why
authors such as Eska, Bernardo Stempel, Delamarre, Untermann, Mac Eoin, Brun, Almagro and others consider recently that proto-Celtic had *p?

5. Well, what we see immediately after the urnfields is the emergence of the non indo-european languages. Why this process would not have been affected in the urnfield areas of Central Europe? Kortlandt, a great specialist in Germanic languages of the century XXI and not of 1820, affirms that the Germanic is strongly influenced by a finno-ugrian substrate and it can explain its consonantal and vocalic instability. Then we have Venneman and followers with his vasconica-semitica theory. Certainly there are hydronymyc names in Germania, Austria and NE France that do not seem indo-europeans and others close to alpine dialects (is it indicative of a previous pre-indo-indoeuropean substrate and the posterior arrived of the indo-european language from de Alps? I don't know). On the other hand, a great number of Celtologists (Khun, Schrijver, Mac Eoin, etc) observe the existence of a non Indo-European substrate in Ireland. Why was not the RSFO affected by similar languages as the neighboring Raetian (f.ex. the germanic runic alphabet derives from one of these Tyrsenian alphabet) or others linked or not with the aquitanian-iberian, but not necessarely indo-european (think that similar facies like Aquitania and Languedoc, with its local differences, we find in the urnfield areas of Isère, Loire, Seine, Rhone, Lot and Paris region)? And, where the Iberian culture come from with its urnfield features? From the north, is not?

6. The gaulish and lepontic personal and place names that I have pointed above, ar
e or not celtic names?
 
Last edited:
1. Taranis, Can you tell us the difference between *toutai - > Toude (with lenition) > Tude and the first element of Toutatis/Tutatis? The difference between in territorio Lemeto, today A Limia region, (with Celtic suffix -et-: cp. callaecian Nem-et-obriga, gaulish Mog-et-io) and the Gaulish place-name Lemane/Limene? What is the criterion to consider these callaecian forms as non-Celtic?


Churchin does not comment about the existence of non Indo-European languages in NW Hispania. On other hand his etymological proposals labeled as undifferentiatted or uncertain Indo-European have always parallels in other extrapeninsular areas. Luján Martínez's work is not bad, but it is strongly influenced by the University of Salamanca and based on previous criteria. Other better sources exist as Moralejo, Báscuas, Búa, Vallejo, etc. or the own Churchin.


1. I would suggest that you go and read Martinez' paper and individually address the cases that he explicitly referes to as non-Celtic and proposes as non-Indo-European, and demonstrate to why these forms should be Celtic after all according to you. Since Martinez' paper is readily available online, this shouldn't be much of an effort for you. It also makes more sense than insinuating me to a statement that I never made.


2. The Galician-Portuguese olga ​​and ancient Castilian huelga should arrive under Cluniac influences, through the Provençal olca 'ploughed field' or the ancient french dialect of Moselle region olke 'vineyard'. The reason is the hispanic romances can not change the consonantic group -lk- to -lg, then it is a secondary form. Only if it belong to a similar form like the old french dialectal variant olica (in the Formulae Senonenses, c. VIII), we can reconstruct the form olga and huelga: cp. galician place-name A Olga (not A Olca).


See ie, *h2olkeh2- (IEW 32, LIV 236) 'protection, fortification', goth. alhs 'fortified temple ', old english ealgian 'to protect', better to designate an hill fort.


3. The cantabrian place-name Tenobriga have the same ethymology than the galician river name Tea < Tena or galician place-names as Tiobre < (med. Teobre) *teno-brigs, Tebra < *tenebriga, Teis < *tenes, the same word, I think, than La Tène. I have not a secure ethymology about this item, but it is an hydronymyc word: cp too the callaecian god name + incomplet place-name COHVE TENA = COSSUE TENA[ECO?] or TENA[BRI?] (COHUE presents the same solution like old welsh, S > H. Prósper and Bernardo Stempel point this phonetical fact like a new variety of celtic in the Central Callaecia Lucensis).


To me, I'm afraid to say that it appears that you fundamentally reject what should be a self-evident fact on the Iberian peninsula: that in the same general area, we have both names where PIE *p > *p and where PIE *p > Ø. I have brought up a considerable number of example, some which were brought up by yourself a year ago:


Well, perhaps can we explain it like callaecian tribe Arotreba (< from *aretreba), lusitanian. name god Ateraeco (*pateraiko), callaecian placename Olca (< polka) or callaecian LANOBRIGA (< planobriga)?


examples of PIE *p > Ø:


- Gallaecian tribal name "Arotrebae" (< *par-)
- Celtiberian place name "Aregrada" / "Arekoratikos" ( < *pare-)
- place name "Clunia" / "Kolounioku" (< *klepn-)
- place name "Octaviolca": I agree with Curchin's and Martinez's etymology, it makes sense as o-grade derivative of *pelk-, additionally J. Porkorny lists French "ouche" ("meadow", "fallow land") as derived from Celtic *olka.
- place name "Tenobriga" (if Curchin's etymology is correct, which is backed by cognates in both Goidelic and Brythonic. Additionally, this is a cognate with Latin "tepidus", and Russian/Ukrainian "teplij")
- place name "Lama" (as per Martinez: "It would be thus a derivative from lama-, which occurs frequently in Hispanic onomastics, and has been explained by García Alonso (2003: 126) as related to OIr. lám 'hand', from IE *pl̥ma or *plāma, with Celtic loss of initial *p-. ")


examples of *p retained:


- Lusitanian "Porcom" (Cabeço das Fráguas)
- Lusitanian "Porgom" (Lamas de Moledo)
- place name "Pallantia"
- place name "Pintia"
- place name "Segontia Paramica"
- ethnic name "Capori"
- place name "Turuptiana"


we additionally have cases of PIE *kʷ > *p:


- Lusitanian "pumpi" (Ribeira da Venda)
- Lusitanian "puppid" (Arroyo de Luz)
- Lusitanian "petranoi" (Lamas de Moledo)


Unless you believe that sound laws can be in free variation with each other (something which is impossible, because sound laws have no "memory""), it should be clear that such an arrangement cannot have developed in-situ. One of the two strata must be foreign-introduced element that developed elsewhere, whereas the others must be regarded as an autochtonous, earlier development.

Kortlandt, a great specialist in Germanic languages of the century XXI and not of 1820, affirms that the Germanic is strongly influenced by a finno-ugrian substrate and it can explain its consonantal and vocalic instability.


I severely disagree with Kortlandt's assessment: the idea that a non-Indo-European substrate is responsible for the First Germanic Sound Shift (Grimm's Law) can be easily refuted. As you should be aware, there is a substantial amount of Celtic loanwords into Proto-Germanic, and virtually all of these occured either before or during the sound shift. Virtually scholars on the topic agree that the First Germanic Sound Shift cannot have happened before the early iron age in Northern Europe (ca. 500 BC, most notably, the word "iron" itself is a borrowing from Celtic!), and other authors (Euler) suggest that the sound shift didn't occur until the 1st century BC. In this context I recommend Wolfram Euler (2009) "Sprache und Herkunft der Germanen" ("language and origin of the Germanic peoples").


You're also the first person ever to label Grimm's Law a "consonantal instability".


You're still not fond of the very concept of sound laws, are you?


5. Well, what we see immediately after the urnfields is the emergence of the non indo-european languages. Why this process would not have been affected in the urnfield areas of Central Europe?


There is no such "emergence". The cultures that predated the Urnfield Culture in Central Europe (from the Corded Ware period onward, likely) were all likely Indo-European. From where should such non-Indo-European peoples suddenly show up from in Central Europe? Did they use teleporters?

Then we have Venneman and followers with his vasconica-semitica theory.


I don't think anybody really takes Vennemann seriously on that issue. Especially, you should be aware that Vennemann's position (the idea of a Vasconic substrate in the entire Atlantic region) is, something that is also in conflict with the observations by Martinez and Curchin.


Why was not the RSFO affected by similar languages as the neighboring Raetian (f.ex. the germanic runic alphabet derives from one of these Tyrsenian alphabet)


That is clearly a strawman argument: the Etruscans and the Raeti were not the only ones using these alphabet (notably, both the Lepontic and Gaulish inscriptions from northern Italy / the Alps were written in variants of the Etruscan alphabet). By your logic, the Greeks spoke a Semitic language because they adopted their alphabet from the Phoenician one and the Phoenicians spoke a Semitic language. Besides, there is no evidence for Etruscan- or Etruscan-like languages (Raetian in the proper sense) in the Rhine-/western Switzerland / eastern France region, as these are only found in the region of South Tyrol. The Tyrsenian-speaking peoples were certainly not autochthonous to Italy or the Alps (let alone Central Europe!) but arrived from the Mediterranean (cf. Herodotus: The Histories 1.94, and also compare the Lemnos stelae).


Also, as I would reiterate, the Tyrsenian languages very much possessed the phoneme *p but lacked the phoneme *b, so a Raetic substrate as an explanation for the loss of *p in Celtic can be eliminated, anyways.


And, where the Iberian culture come from with its urnfield features? From the north, is not?


I find the idea that the Iberians came originally from Central Europe absolutely unreconcilable with the evidence of Iberian place names: characteristic of them is the prefix *ili- or *iltir-, which may be a cognate with modern Basque "hiri" (meaning "town", "city"), as well as the suffix "-sken". If you look at the distribution, they can be found across a very large arc from the Languedoc to eastern Andalusia. The Urnfield Culture clearly never reached Andalusia, yet the Iberian language was there. Some examples:


- "Illiberre" (Ptolemy, Tabula Peutingeriana, modern Elne)
- "Ilerda" (Ptolemy, Itinerarium Antonini, modern Lleida, Iberian mint "Iltirta")
- "Iliberris" (Ptolemy, Itinerarium Antonini, modern Granada)
- "Ilipa" (Itinerarium Antonini, near Seville)
- "Iliturgi" (Itinerarium Antonini, near Mengíbar, Andalusia, Latin mint "Iliturgi")
- "Urci" (near Cuevas del Almanzora, Iberian mint "Urkesken")


If you were correct, we should not find any traces of Iberian language in Andalusia, and conversely we should find traces of Iberian language north of the Languedoc. We don't. What does that tell you about the association between Urnfield and the Iberian language?


In summary, the very existence of non-Indo-European substrate that you see in central and north-west Europe as the source of the Celtic loss of PIE *p is spurious at best.

Now, I'd like to pick up again on this:

Why not the bell-beakers? From the hydronimyc names to the celtic languages there is a long way, is not?

If the Beaker-Bell Culture really spoke Proto-Celtic, where is your Celtic substrate in North Africa, on Sardinia and Sicily?
 
Last edited:
I have not the knowledge of Taranis (I should be glad then)
But a point about phonetic evolution:
a shift in a living language, for I suppose, don't came suddenly without a background: every language tends to evolve, for the most in respect of the "lesser pronounciation effort" (sorry for my english) but with some constraints as structure and other collective habits very uneasy to explain at first sight - so the languages don't evolve at the same rythm and passing through the same stages exactly according to human groups -
the palatization, as an example, appeared in some populations, not in others, and trying to explain it by the structure alone is an oversimplification - in oil french (except very northern France), a very palaziting country, the phenomenon is yet at work nowaday - the 'cu-' and 'qu' digraph officaly /k/ because protected by 'w' sound, is more and more pronounced /kj/>/tch/ in dialects (today moribond) and popular speech sometimes, before 'i' and 'e' sounds - in "learned" languages where it seams being an acquired ready kit of
pronouciation, this evolution stopped apprently -
if some phenomenons took some centuries to find complete result, we can imagine the same for other phonetic phenomenons -
what we know about ancient languages is through writings, old texts, inscriptions of all kinds - we infer evolution on written modifications and too on loan words, without knowing the time they needed for intermediary forms - so the very precise datation of germanic evolution (consonantal shift) is somewhat uncertain even if the fork cannot be to open - even the datation of loanword sis somewhat uncertain, sometimes -
I believe yet in a nordic (finnic?) influence on germanic (as did H.HUBERT, but I have not his knowledge!!!), from people living on the northern plain of Europe - perhaps ma I wrong - so, excuse me.
 
zzzzzzzzzzzz

 
Last edited:
As an aside, Tartessian (SW Iberia) has recently been classified as a Celtic language. See WIKI "Tartessian Language," Yokum (2011) and others. Tartessian is now the most ancient Celtic language, attested as 500-700 years older than Lepontic (a variant of Gaulish).
 
zzzzzzzzzzzz
 
Last edited:

This thread has been viewed 43335 times.

Back
Top