2012/5/10, dgkilday57 <
dgkilday57@...>:
>
Enough time already. Perhaps this represents *bHr.h1-teh2- formed from
the set.-root observed in Grk. <pheretron>, Skt. <bharitram>, Lat.
<feretrum> 'bier, litter, stretcher'. The force of *-h1 as a
root-extension is not immediately obvious. Possibly it signifies 'to
completion, to the end'. A born child has been carried to the end of
child-bearing (i.e. birth), and a dead person is carried on the bier
to his final resting-place. Thus *bHer- 'to carry', *bHerh1- 'to carry
to the end'. Cf. *kel- 'to strike, hew', *kelh1- 'to strike
(decisively)'.
Bhrihskwobhloukstroy:
see my message of 2012/5/2 (21:36 Swiss time) : the only alternative
to appertinentive Vrddhi *bho:r-ti-s would from seṭ variant *bherH-
with zero-grade and, pace Jens, derivative -h2 (from feminine)
*bhrH-th2-i-s (> Celtic *barti-s like *bardo-s < *gwrH-dhh1-o-s with
'Schrijver's Law')
I'd already published that
<pruuia> /bruwya:/ :
> Gaulish bri:ua: 'bridge' insists on an onomasiologic difference in
> Celtic itself (bri:ua: vs. drochet).
DGK:
The Lepontic form is <pruiam> and I can see no principled way of
getting it out of *bHreh1wo- 'bridge'.
Bhrihskwobhloukstroy:
You probably don't have seen my analysis of bri:ua: < *bhre:u(H)ah2
(2012/5/9, 13:24 Swiss time):
"Gaulish *Bri:wotti: (bri:ua: 'bridge' = Romance ponte) < Celtic
*Bri:wottoi <- *Bri:wottu:s < Late IE *Bhre:wotnó:s < PIE
*Bhre:u(H)o-tnó-h1es < *Bhre:u(H)o-tnH-ó-h1es 'extensions of Bridge (=
Ponte)' with *Common* PIE 'neognós' laryngeal deleting"
DGK:
I suggest instead the acc. sg. of *gWrh2u-jeh2- 'heavy stone',
Bhrihskwobhloukstroy:
Why *laryngeal deletion in *gWrh2u-jeh2-? Shouldn't it yield †*baruja:
or †*bra:wja: according to syllabification?
DGK: referring to the inscribed heavy stone itself. Before Whatmough,
scholars read the last word of the text as <palai>, which could be
loc. sg. 'in the plot'. The legible part of the Vergiate text is then:
pelkui:pruiam:teu:karite:i[--------]ite:palai
Problems abound with taking <teu> as a postposition and <karite> as a
passive verb. Instead <karite> could be understood as an instrumental
or other adverbial word, along with <...ite>, with the subject <i...>
largely effaced. I hypothesize <teu> as an active verb, an
unreduplicated 3sg. root-perfect, Lep. /de:u/, corresponding to Skt.
<dadha:u>, and containing *-w as an archaic root-perfect formant, PIE
*dHeh1-w 'has placed'. In this view the Lat. /w/-perf. originated from
root-perfs. with laryngeal-final roots, <(g)no:vi:>, <-ple:vi:>, etc.
(cf. Skt. <jajn~a:u>, <papra:u>). The text is then to be partially
read as follows:
'... (acting with) ... (and) ... has placed (this) heavy stone in the
plot for Belgus.'
Possibly Lep. <tetu> (Prestino) can be similarly analyzed as a
reduplicated root-perfect *de-deh3-w (Skt. <dada:u> 'has given').
Narbonese Gaulish has <dede> which can hardly be the exact
morphological equivalent.
Bhrihskwobhloukstroy:
OK, the connection <pruia>* : bri:ua: has alternative
counteranalyses, but my argument was "<pruuia> /bruwya:/ : Gaulish
bri:ua: 'bridge' insists on an onomasiologic difference in Celtic
itself (bri:ua: vs. drochet)". I thought You would have used this
argument against the hypothesis of the Celticity of Lepontic. A
connection between <pruia> and bri:ua: would have been (so it was at
least since Pisani) an argument in favour of a lexical diversity
between Lepontic and Gaulish and I was trying to explain it, but if
You indeed don't believe in a connection between <pruia> and bri:ua:
then a possible argument for You doesn't even start and this problem
doesn't exist.
>> 2012/5/9, dgkilday57 <dgkilday57@...>:
>> >
>> > --- In cybalist@yahoogroups.com, Bhrihskwobhloukstroy
>> > <bhrihstlobhrouzghdhroy@> wrote:
>> Could You please be so patient to enumerate the ad-hoc assumptions
>> in my theory? Yours are two, aren't they? (Being ad-hoc isn't wrong;
>> it simply means "not independently founded"))
> DGK:
> Celts developing in place defy archaeology. Celtic place-names with
> conserved inherited /p/ (as you claim for Porcobera) defy the concept of
> Celtic.
>
Bhrihskwobhloukstroy:
Archaeology shows a succession of cultures (Lagozza,
Canegrate-Golasecca and so on); maybe each culture is a different
population (but not genetically different), maybe not. Let's admit
that each culture corresponds to a different population; the last one
is Celtic. Does this imply that the former weren't Celts? Obviously it
doesn't imply anything. Celts can supersede on other Celts. A
difference in culture cannot necessarily imply difference of language
family (otherwise the Suebians couldn't have attacked the Usipetes and
Tencteri...)
As I have already twice made clear, IE place-names with conserved
inherited /p/ don't necessarily represent a pre-Celtic layer until
they don't show any non-Celtic innovation. You can call them
"Porcoberians" or what You like; I had used Giulia Petracco Sicardi
label "Celtic", but they are in fact Late IE not fully celticized.
Have I been sufficiently clear?
Therefore these aren't assumptions. I don't *assume* anything on
these grounds. Archaeology is indifferent and therefore cannot prove
nor disprove my thesis nor Your one; Late IE names cannot prove nor
disprove my thesis nor Your one. So, I haven't made any assumption of
such kind and therefore these aren't ad hoc assumptions (they aren't
assumptions at all, there's no assumption: absence of population
replacement and absence of a pre-Celtic layer can only be
*consequences* of independent etymolgies).
>> Place-names that show only a part (even just the Late IE section)
>> of Celtic sound laws are pre-Celtic only in the sense they are
>> conservative. They don't prove the existence of any substrate; at most
>> they can suggest it, but they can as well (of course don't necessarily
>> need to) represent only marginal areas.
>> Place-names that show non-Celtic innovations but no Celtic
>> innovations are post-Celtic.
>> Place-names that show non-Celtic innovations and only a part of
>> Celtic innovations can (= cannot avoid the suspicion to be) later
>> incomers.
>> Only place-names that show non-Celtic innovations and all local
>> Celtic innovations suffice to establish a pre-Celtic substrate.
> DGK:
> But you regard all innovations as taking place in situ. Your doctrine
> cannot distinguish Celtic from non-Celtic at any time-depth.
Bhrihskwobhloukstroy:
Doch (= "no" or "yes"?), it can. I there were non-Celtic inovations,
they would be recognized as such. E.g., a place-name †Medioplo:nom (or
†Mediopla:nom, depending on it final member's etymology) could be
immediately recognized as Venetic or Latin, †Mefioplu:no resp.
†Mefiopla:no as Italic, †Midiaflo:na- as Germanic and so on.
The proof is that, of course, although a regular Celtic etymology
would be possible (for †Medioplo:nom *medio- + *plo- + *(f)ono-, where
*plo- < PIE *kwl(H)-o-, compounded form of *kwlH-ah2 > pala:, with
neognós laryngeal deletion, and *(f)ono- < *pono- 'water'), the risk
of a casual coincidence with real Celtic lexical items (*medio-, pala
and *ono-) would be nevertheless higher in such a three disyllables
analysis than in a straightforward comparison between the entire
tetrasyllable compound *Mediola:non and its potential Venetic match.
Therefore, the etymology of †Medioplo:nom as the Venetic outcome of
PIE *Medhyo-plh1no-m (> Celtic *Mediola:non) is statistically more
probable than the analysis of †Medioplo:nom as the Celtic output of
PIE *Medhyo-kwlHo-pono-.
Bormo- < PIE *gwhor-mo- and Barga < PIE *bhrg'h-ah2 don't have the
same statistical superiority on *bhor-mo- and *bho:rg'h-ah2 because 1)
the length of the comparanda isn't higher for non-Celtic (two
syllables) than for Celtic reconstructions (two syllables), and 2) the
non-Celtic diachronic phonology isn't independently established.
Anyway, even if it were demonstrated that non-Celtic innovations are
to be preferred in the etymology of some place-names, there would be
the question of the relative chronological priority between Celtic and
non-Celtic layer. The layer which shows more features of both
diachronic phonologies is the older one. If the two layers show the
same amount of transformations through both diachronic phonologies,
they're equally old (this almost never happens). If each layer doesn't
exhibit any trace of transformations according to the other one's
phonology, we can't decide whether the two languages has been spoken
separately or one of the two has completely erased the other one in a
prt of the region under examination.
>
>> > DGK:
>> > Furthermore there should be abundant parallel /o:/-grade formations NOT
>> > involving Osthoff's shortening from roots without resonants in this
>> > position, in Celtic and elsewhere.
>>
>> Bhrihskwobhloukstroy:
>> Every vrddhi formation, isn't it?
> DGK:
> Every one has /o:/-grade?
Bhrihskwobhloukstroy:
No, some have /e:/-grade (cf. my *bhre:u(a)ah2 > Gaulish bri:ua:)
and some have /o:/-grade, as every handbook of IE studies tells
(unless You adhere to Eric Pratt Hamp's idea that lengthened grade
never existed)
> DGK:
> Then I do not understand your objection to Rea:te as a counter-example. If
> all your place-names end in -a:te, that is ABSOLUTELY no big deal.
>
Bhrihskwobhloukstroy:
My English must be really unintelligible if You (this is patent)
don't have understood even the elementary fact that -ate-place-names
have accented long /a:/. As a consequence, You have completely missed
my argument, which is based on the analysis of (long!) /a:/ as the
conflation of *-a:a:- < *a:-ja:- < Late IE *-o:ya:- < PIE *-o-h1yah2-.
I redirect You to my message of 2012/5/1 (15:29 Swiss time); maybe as
You have seen -ate You have been so caught by Your discovery (You have
repeated six time "Reate") - although without any etymology - that You
have sauté du même au même.
I did propose an etymology for Rea:te (2012/5/3, 15:09: PIE
*H1/3reiH-ah2-ti '(place) where (Velino and Turano rivers) flow
(together)' > *Reya:ti > Italic Rea:tí > Latin Rea:te). As for -ate, I
copy and paste what I then wrote, because after 103 messages in this
thread I can't afford to write it again in other words:
"Your question (...) implies that an -ate-place-name is found
outside the area of the 200 -ate-names of Transpadana and of
Gallo-Romance place-names in (Prov.) -at, (French) -é.
As You can infer from my etymology, I separate Rea:te from Irish
áth. Not just that; I even separate a couple of dozens of
-ate-place-names from áth, precisaly those ending in -rate (: Ir. 2
ráith 'fort'), -biate (: bláth 'bloom'), and possibly -nate (if from a
hypothetical Celtic *na:ti 'hill, rise': Gk. no:ton, Lat. natis;
Ternate, XIIth c. Trinate, is on three hills).
I don't think You are really maintaining that the form of a suffix
has to be always of one and the same origin (so, for instance, Alaska
isn't a Ligurian -asca name, although Piero Riva 1964 seems to believe
that), especially if it occurs outside of the area of the rest of the
names: French, Provençal, Rhaeto-Cisalpine are Western Romance
languages, from Gaulish Imperial Latin, on a Continental Celtic
substratum as the inscc. testify; Reate is in Sabina, quite far away,
but above all in an Italic context that doesn't explain Ligurian /b/
/d/ /g/ < */bh/ */dh/ */gh/ in any way (pace Festi)."
>> >> Instead of this non sequitur I propose: let's see how many
>> >> etymologies are possible, beginning from those according to the three
>> >> languages that are historically attested in situ. After that, let's
>> >> try to establish a hierarchy of probability; all hypotheses have to be
>> >> correctly formulated, just one can be probably right (let alone the
>> >> case of folk-etymology or the like).
>>
>> > [DGK:]
>> > Such a procedure is obstructionistic.
>> >
>> Bhrihskwobhloukstroy:
>> Do You have a better one? (A less long but equally 'measuring' one, I
>> mean)
> DGK:
> Yes. First, dump the development-in-place (DIP) doctrine.
Bhrihskwobhloukstroy:
I was asking You to espose a better procedure for evaluating
competing etymologies. If You solution is to eliminate one of the
competitors, I wouldn't qualifiy it as 'better'; it rather looks like
a(n) Endlösung
> Since you
> formerly agreed with Kretschmer's Ligurian and linguistic stratification,
> and even taught this material, it would be good to know what epiphany led to
> your conversion.
Bhrihskwobhloukstroy:
I make reconstructions of and from every piece of evidence I meet.
I reconstruct all possible PIE etyma according to all possible
diachronic phonologies and after that I reconstruct all possible
Hamito-Semitic etyma.
Gradually I began to realize that Mediterranean reconstructions
could have IE etyma, that Old European hydronymy can be analyzed (as
Krahe thought) according to local diachronic phonologies and that
pre-Celtic words could have regelmässige Celtic etyma as well.
After many years of such work I asked myself to what extent could
the reconstruction of European linguistic prehistory be reduced to a
minimum of layers.
My provisional answer is that there's no need of IE linguistic
classes that aren't attested by direct textual evidence (with the
possible exception of proper Illyrian and Daco-Mysian). Maybe they did
exist, but apparently the evidence doesn't need them.
I was quite sad of this result because I loved substrates. Still
now I think and hope that they nevertheless existed, albeit as the not
direclty attested part of languages that are not recognized as IE
(e.g. I believe that Villar's "meridional-ibero-pirenáico" can indeed
be part of the Iberian complex); but for Liguria I have to recognize
that Ancient Ligurian is demonstrably only a Late IE language, without
the whole of Old Celtic transformations, but without any non-Celtic
innovation.
DGK:
Did Alfredo Trombetti's ghost appear to you in a dream?
Bhrihskwobhloukstroy:
Of course he did, but contrary to what You may think I told me
many times that (if I remember well) I've got to face the *hard*
evidence that neither Basque nor Iberian are demonstrably IE and that
"the only real language IE-ists' "PIE" can at best approximate is the
paleo-dialect of the Steppes, which apparently acted as a superstrate
to other paleo-varieties in the genesis of the historical IE
languages".
You know, he would never subscribe a theory where his Mediterranean
is analyzed as direct IE heritage in Celtic or Italic (or Greek &c.)
and Ligurian is an IE language and there has never been any pre-IE
layer anywhere...
DGK:
> Or did you recognize that Ligurianists vacillated on the outcome of *kW, and
> realized that you could get parallel results by dumping Ligurian and
> proposing various forms of archaic Celtic?
Bhrihskwobhloukstroy:
I've taken the theory of Celtic Ligurian from Giulia Petracco
Sicardi and Aldo Luigi Prosdocimi (who anyway has many times publicly
declared that his method is the exact opposite of mine; in his words,
I would reconstruct everything, he doesn't reconstruct anything, and
his pupil Patrizia Solinas has stated that she and I, her theory about
PIE and Gaulish and my one are "very, very, very, very, very
different" - but anyway I make use of their theories and add my ones,
they don't make use of these latter)
> DGK:
> I regard Ligurian as a P-language, and no Q-Ligurian is necessary. The
> apparent survival of *kW as -ku- is due instead to *kowV- > *koV- > *kuV- in
> my opinion.
>
Bhrihskwobhloukstroy:
This is what I suspect as well, although I can't be sure of it. I
had mentioned Q-Ligurian theory just to show that a P-/Q- isogloss
boundary within Ligurian is conceivable for many people.