Re: Ligurian

From: Bhrihskwobhloukstroy
Message: 69495
Date: 2012-05-03

2012/5/3, dgkilday57 <dgkilday57@...>:
> What I
> characterize as "baloney" is the assignment of Porcobera and the place-names
> in Borm- and Barg- to Celtic, and the resulting warping of the accepted
> notion of Celtic to include ad-hoc "archaic" dialects conserving inherited
> /p/.

Bhrihskwobhloukstroy:

as I've already once written, this isn't our point of discussion.
Celtic phonology is what it is and no one casts any doubt on it (to
the contrary, I religiously follow usual Celtic reconstructions).
Nor is the Celtic affiliation of Ligurian other than a matter of
terminology. We could be satisfied with an indepedent IE class,
labeled 'Ligurian'.
What is really under discussion here is:
did Ligurian language show diachronic phonological developments
(from PIE) *different* from (i.e. absent in) Celtic? You say "yes", I
say "no". This is the point and this has been the starting point of
our discussion as well.
If one (surely not You) agrees on "no", Ligurian phonology becomes
automatically identical with the Late IE phase preceding Celtic. This
is the meaning of "Archaic Celtic" on phonological level: centum
treatment (innovation), */bh/ */dh/ */gh/ > /b/ /d/ /g/ (innovation),
no consonant shift (retention), preservation of short /o/ ≠ /a/
(retention). No other IE class would fit in (not even Phrygian, if one
accepts bekos 'bread' : bake and Manezordum beside Manegordum).


>> What's straightforward is the comparison between Barga and the
>> root *bherg'h- in its meaning 'hill, mountain'. When it comes to the
>> question of the root's ablaut grade, one should at least take into
>> consideration all the five of them (we have no means of knowing the
>> vowel quantity of /a/):
>> 1) *bhrg'h- > *barg- with vocalic */r/ > /ar/ as in Greek, Anatolian,
>> Armenian, Iranian
>> 2) *bherg'h- > *barg- with */e/ > /a/ as in Iranian and Indic
>> 3) *bhorg'h- > *barg- with */o/ > /a/ as in Germanic, Baltic,
>> Messapian, Albanian, Anatolian, Iranian and Indic
>> 4) *bhe:rg'h- > *barg- with */e:/ > /a:/ as in Germanic
>> 5) *bho:rg'h- > *barg- with */o:/ > /a:/ as in Celtic
>>
>> The spare formation on *bherg'h- that we can detect in the ancient
>> Ligurian area are - beside Barga - Bargagli (Genoa; cf. Bergalei in
>> the Tabula Clesiana, now Bregaglia = Upper Mera / Maira Valley [North
>> of Lake Como], Grisons, Switzerland), Briaglia (Cuneo), and Bregançon
>> < Brigantio.
>> Bargagli : Bergalei would suggest solution n° 2, but then one
>> should explain away Briaglia (apparently formed with the same suffixal
>> complex) and Brigantio as Celticisms, in areas that Ancient Authors
>> ascribe to the Ligurians. Solution n° 1 (Your favourite one) has the
>> same shortcoming and is compelled to consider Bargagli and Bergalei
>> different ablaut formations. Solution n° 5 is the only one that opts
>> for an outcome found in a neighbouring class, i.e. Celtic, has the
>> advantage of keeping free for Ligurian (again, as in Celtic, quite
>> coherently with previous hypothesis) the outcome of vocalic */r/ as
>> /ri/ (and therefore ascribe Brigantio and Briaglia to the Ligurians,
>> without postulating special Celtic immigrations) and, just like Your
>> favourite solutions, distinguishes Bargagli and Bergalei as different
>> ablaut grades. I can't see any advantage in solutions nn. 3 and 4, so
>> I leave them. As a result, solution n° 5 is preferable, event if
>> Ligurian shouldn't be Celtic for the rest. Q.E.D.

> DGK:
> Quid est deridendum. Some of the Ligurian names have /e/-grade in *berg-
> (see Petracco-Sicardi). I do not deny the existence of Gaulish names in
> Liguria, or Liguro-Gallic hybrids like Bodincomagus. My position is that
> the Ligurian stratum is pre-Celtic.

Bhrihskwobhloukstroy:

Ligurian names in /berg/ (like Bergeggi) and /borg/ (like Borgio;
both with regular Late Latin palatalization and therefore
pre-Germanic) very aptly exclude solutions nn. 2 and 3, not the other
ones. All the rest being unchanged, solution n° 5 stil results the
best one. Quod erat demonstrandum

> DGK
> But we already have Gaulish personal names on
> Lepontic inscriptions and coins. Whatmough explained this easily in 1933,
> and he referred to the inscc. as "Kelto-Liguric" (i.e. Ligurian with Celtic
> superstrate). We have Latin names in Etruscan inscc., and that does not
> make Etruscan an Italic language.

Bhrihskwobhloukstroy:

Etruscan inscc. have plenty of Etruscan lexical items, Lepontic
inscc. unfortunately much less of Lepontic lexical items, but
nevertheless please show me that the majority of Lepontic lexicon
contrasts with Celtic lexicon.
You'll find <pala> and <pruuia>. <pala> can have many etymologies,
so it neither proves nor disproves anything; <pruuia> /bruwya:/ :
Gaulish bri:ua: 'bridge' insists on an onomasiologic difference in
Celtic itself (bri:ua: vs. drochet).

What's strange is that anywhere a univoque Celtic trace emerges in
Liguria it's interpreted as the result of Celtic infiltration. This
would be possible, although not sure.
On the other side, all remaining Ligurian glosses and names can be
Celtic (they are indeed so analyzable; You may think this isn't the
best way of analyzing them, but in any case it's a correct way, and
better or worse is a matter of discussion, their correctness isn't)
except a couple of occurrences of PIE */p/ (Porcobera, Boplo - I
myself have given a fresh etymology for it - while Pellius isn't
diagnostic). I (with and after many others) draw the conclusion that
Ligurian was predominantly completely Celtic and that there were
isolated conservative pockets just like Piario east of Larius (or, on
Larius itself, the very Plinius Caluos = Lith plynas 'bald') and
Parra.
Now, what's more probable? That all surely Celtic traces were
casually the result of Celtic infiltrations or that the isogloss /p/ >
0 had left some areas untouched?

You have written that Ligurian had, like Celtic, /b/ < */gw/ and
(like only p-Celtic!) /p/ < */kw/. This means that:
i) all Celtic and Ligurian had /b/ < */gw/ (BTW this is the first
*specific Celtic* innovation - i.e. not common to Germanic or Italic
as well - according to both Schrijver and Isaac; of course they don't
take Ligurian into consideration)
ii) all Celtic (but not Ligurian or, as Patrizia and I would say,
not all Ligurian) had dephonologization of /p/
iii) p-Celtic (not q-Celtic) and Ligurian had /p/ < */kw/.
There are instances of <qu> in Ligurian names (e.g. Quiamelius);
if they represent IE */kw/, this would mean that there were q-Ligurian
(beside p-Ligurian) dialects. If such an incomplete diffusion of the
innovation */kw/ > /p/ internally divides both Celtic and Ligurian,
there's no reason to exclude that this can have been true for the
dephonologization of PIE */p/ (which follows */gw/ > /b/) as well.
This for /p/ and 'Archaic' Celtic.

For /eu/ (not > /ou/) You implicitly maintain that /eu/ > /ou/ was
no longer operative as the only Celts You admit in Cisalpine, after
600 BC(E) or even after the IVth c. BC(E) arrived to their historical
seats, because /eu/ is thouroughly conserved in Pre-Roman place-names:
in tonic position Leucum > Lecco, pretonic Leocade > Locate, in
contrast with tonic ou > u: in Busto < Boustom = CIb. PouśTom,
pretonic Cunardo : CIb. kouno-, Ir. 1 ard 'hill'.
Pairs like Neuidunum : Nouiodunum can't be diagnostic because they
are equally well analyzable as IE ablaut, a derivative pattern not
less than suffixation in e.g. Celtic nouio- vs. nouo-.
In my view, Celtic /eu/ > /ou/ is simply successive to the
introduction of Cisalpine *-a:tus place-names into Cisalpine Latin
(II. c. BC[E]; by that time [Provincial] Latin /ou/ as well was still
a diphthong, cf. Ouf[entina tribus] in inscc. from Mediolanum and
Comum), just as /ou/ > /o:/ took place after Caesar's conquest of the
Gaul (cf. instead 16 BC[E] Vindobo:na < *Windo-bouna:, 43 AD
Lo:ndinium < *Loundenion : *Loundonion > Llundein). Also this is
perfectly possible.
As anyone can see, the Pan-Celtic status of /eu/ > /ou/ depends
i.a. on the acceptance of Celtic etymologies for -ate-place-names in
Cisalpine. Nobody denies Celtic /eu/ > /ou/; it's just a matter of
absolute and relative chronology. In order to be Pan-Celtic, the
change should embrace *all* Celtic languages from the very beginning
of the attestations or at least before the end of the attestations of
the first disappeared Ancient Celtic languages.
In any case, the absence of a Celtic innovation /eu/ > /ou/ is
much less significant than an anti-Celtic development such as /ar/ <
syllabic /r/ before stops. The Ligurian retention of /eu/ is therefore
at most another archaic feature as against /ou/ < /eu/, but in no way
sufficient to classify Ligurian as an independent IE class with its
own innovations.

You can't reject any of these consideration.
Ligurian retention of PIE /p/ and /eu/ are conservative facts and
therefore weakly significant.
Full labialization of */gw/ and, later on and partially, */kw/ are
innovations and are common to both Celtic and Ligurian.
Before stops, /ar/ from syllabic */r/ is based on an example that
can be equally well (until You prove the contrary; I've made my task)
explained as from long */o:r/ and, moreover, at the cost (which I
don't have to pay) of postulating Celtic infiltrations for cases of
/ri/, /re/ < syllabic */r/ in Liguria: possible, but ad hoc and
therefore a shortcoming if compared with my explanation.
Full labialization of */gwh/ is based on two examples (*gwhormo-
and *dhegwh-e-lis) that can be explained from */bh/. In this case the
only invoked criterion is semantics, but remember that */gwh/ > /b/ in
only paralleled by Latin -b- in internal position.

In sum, the status of Ancient Ligurian as an independent IE class
is based on:
- *three* etyma,
- of *place-names* (i.e. without full semantic verifiability; It.
debbio isn't diagnostic, because it could reflect regular Gaulish
*deuelo- < *dhegwh- as well as our ambiguous *debelo-)
- and never with parallels in the diachronic phonology of
neighbouring languages (it was Italic, not Latin, that bordered with
Ligurian, cf. Briniates : *Frinianum > Frignano).

On the contrary, the classification of Ancient Ligurian as:
- a Western Late IE dialectal complex (on this we agree: centum,
*/bh/ */dh/ */gh/ > /b/ /d/ /g/),
- taking part to early Celtic significant innovations (on this we
agree as well: */gw/ > /b/)
- and - maybe with local exceptions - even to p-Celtic (as well as
Italic) innovations (here too we agree: */kw/ > /p/) diatopically
partial,
- at least suggests caution in assuming patently non-Celtic
developments and, since it's in fact possible to explain the evidence
through Celtic phonology, shows that it's more economic to avoid the
postulation of an entire independent IE class, even at the modest
'price' of interpreting Aquae Bormiae (and the rest) as 'boiling'
(*bher-) instead of simply 'warm' (*gwher-) and Barga as 'pertaining
to mountains' (vrddhi) instead of 'mount' (which it isn't in fact;
it's *on* a mountain, not itself a mountain).

> DGK
> Strength in numbers, eh? 200 etymologies can't be wrong?

Bhrihskwobhloukstroy:
Etymologies can be a) correct or not and b) right or wrong.
Correct = in full coherence with phonology, morphology, lexicon
etc. of the implied language(s)
Right = corresponding to historical reality.
Even correct etymologies can be wrong. This applies to all of us. A
good starting point would be etymological correctness. My etymologies
are correct under (or 'from'?) an IE point of view (unter / von
indogermanistichem Gesichtspunkt) and they are hoped to be
celtologically correct as well. Kretschmer's etymologies are surely
correct from an IE point of view, but they don't refer to any
independently demonstrable IE class, so they are - rightly and
necessarily - ad hoc.
Therefore we have 200 correct etymologies vs. 2 ad hoc etymologies
(their correctness from an IE point of view being equal).
(What would You say if the respective proportions would be
inverted? That Kretschmer's 200 etymologies can be wrong?)

> DGK
> I stand corrected. I didn't realize how you differ from Alinei.

Alinei postulates a pre-Roman Latin-like (not simply "Late IE", as
the label "Ibero-Dalmatian" or "Ibero-Adriatic" would let think, nor
"Italic", as the label "Italoid" or "Italid" would suggest) language
family from Lusitania to Illyricum (does this suggest anything to You)
and that this family is the direct ancestor - not a mere substratum -
of historical Romance languages. According to Alinei, such a family
had been the result of the first arrival of Modern Humans in this
area, who spoke a modified version of PIE (with centum treatment,
*/bh/ */dh/ */gh/ > /ɸ/ /θ/ /x/ and so on).
I think - for this area (and without taking into consideration
Basque and Iberian for this matter) - that PIE as such (stricto sensu
and in its reconstructed prehistoric form) evolved directly in situ
into Ancient Dalmatian, Venetic-istrian-Liburnian, Latin, Italic, and
Continental Celtic, with marginal pockets of not completely
'developed' (in the sense of 'become completely Celtic) dialects, and
- just like the overwhelmingly majority of historically interested
people - that the Romans for the first time introduced in these areas
(outside Rome itself) the Roman variant of ancient Latin dialects,
this variant being the direct ancestor of historical Romance
languages.
As You see, there's nothing in common between Alinei and me except
for the dating of the arrival of the first IE genealogical variety
(itself different: an almost Italic IE for him, PIE for me)

> DGK
> I think we should discuss Petracco-Sicardi's etymologies involving
> *bHerg^H-.

Bhrihskwobhloukstroy:
OK, but we'd completely agree on that point. (A very small detail:
her surname is Petracco Sicardi without hyphen, she was born Sicardi
and had married a Petracco)
>
>> Facts:
>> 1) In Ireland there are place names of the structure à th 'Ford' + X
>> (Paradebeispiel: Ã th Cliath); O'Rahilly in 1943 has detected an older
>> layer with posponed -ad (where /d/ < */t/ regularly in post-posttonic
>> position)
>> 2) OIr. áth < PIE *h1iah2-tu-s 'passage' (√*h1iah2- 'go' < √*h1ei-
>> 'go')
>> 3) In Western Romance Countries there's a toponymic layer with final
>> -at
>> 4) These place-names are often named after the rivers along which
>> such places are situated (e.g. Agognate on Agogna river, Terdobbiate
>> on Terdoppio, Arnate on Arno [Lombardy], Lonate on Olona, Lambrate on
>> Lambro, Beverate on Bevera, Brembate on Brembo, Seriate on Serio);
>> 5) Their locations are not just like any other one; they are
>> precisely where the principal ancient roads between prehistoric
>> centres (Como, Bergamo, and so on) crossed those rivers (particularly,
>> the two Brembate are where Adda river is crossed by the road between
>> Bergamo and Como and respectively between Bergamo and Milan, whose
>> site was topographically relevant also before 600 BC because is 2m
>> higher than the surrounding plain and therefore could emerge as an
>> island from the ancient moor)
>> 4) There are regular correspondences between Irish and Romance
>> names, e.g. à th Bó = Bobbiate, à th Carr = Carate, à th Caoin = Cenate,
>> à th Cliath = Cedate, à th Cúile = Cugliate, à th Fearna = Vernate, Ã
>> th
>> Garbháin = Garbagnate, à th Lóich = Locate (old Leocade), à th Malain =
>> Malnate, à th Nó = Novate [still transparent], and many more
>> 5) These correspondences fall together with hundreds of Western
>> Romance (and, in this case, specifically Lombard) place-names
>> completely identical with Celtic names (e.g. Duno [Varese] = dún)

> DGK:
> That could equally well be Ligurian (in my sense), given Nevidunus (saltus)
> against Gallo-Latin Noviodunum.

Bhrihskwobhloukstroy:
"Equally well"; You're right.
What about following equations?
Barzio : OIr. bairte
Introbio : OIr. itropa
Taceno : Welsh tag, OIr. en
Crandola : Breton kran, OIr. 7 ol
Margno : Middle Ir. mairne
Casargo : Breton kasarc'h
Indovero : Ir. indber
Parlasco (XIIIth c. Perlasco) : Welsh (g[w]o-)br, OIr. lasc
Vendrogno : Welsh gwyn, rhawn
Tremenico : Welsh tremyn or Ir. tríamnach
Introzzo : Oir. 1 étrochta
Sueglio : Welsh huddigl
Vestreno : Middle Welsh gwest, Breton rûn
Corenno : OIr. cor, inn
Dervio : Welsh derwenn
Murcote : OIr. mor(rígain), cúa
Perledo : Welsh (g[w]o-)br, llwyd
Varenna : OIr. fáir, 1 rind
Mandello : OIr. menn, Breton, Welsh gwell
Borbino : OIr. borb, Welsh 1 -in
Vassena : Welsh gwas
Laorca : Or. ló, rech
Vercurago : Welsh (g(w)orchordd, -og
(Monte)marenzo : Oir. mór, étaim
Brianza : OIr. Brigit
(Val)madrera : Gaulish madera*, OIr. 3 aire
Trebbia : Welsh tref
Bartesago : Or. barr, tais, -ach
Barro : OIr. barr
Garlate : OIr. gair, Gaulish Lato-
Consonno : OIr. conn, Breton sonn
(Val)greghentino (885 Carcantinum) : Welsh carreg, Ir. étan, Welsh 1 -in
Brivio : Gaulish -briua
Airuno : OIr. 1 aile, 1 rún
Campsirago : OIr. camm, sírach
Calco (960 Calgo) : OIr. calg
Arlate : Gaulish are, Lato-
Imbersago (985 Amberciago) : Ir. amairsech
Osnago : OIr. oissinech
Lomagna : OIr. loman
Missaglia (XIIIt c. Massalia) : OIr. máss, 2 aile
Casirago : Ir. casrach
Cremella : Breton krem, OIr. 3 ell
Cassago : Ir. casach, Welsh 1 casog
Bevera : Gaul. Bibra-
Bernaga : Ir. bernach
Oggiono : OIr. 2 og, Welsh llawen
Imberido (XIIIth c. Amberado) : OIr. 2 imm, 1 rath
Ello (XIIth c. Aello) : OIr. 1 adall
Marconaga : Welsh marchonnen, -og
Sirone : OIr. sír, on(-chú)
Molteno : OIr. molt, en
Brongio : OIr. bruinde
Annone : OIr. an, on(-chú)
Asso : OIr. ass, Welsh ach
Rezzo, Rezzago : Middle Ir. reithe, -ach
Caglio : Ir. 2 caile
Sormano : Welsh Sôr, mawn
Lasnigo (1206 Asenigo) : Welsh asenig
Barni : Welsh barn
Magreglio : OIr. mérl(-ach)
Onno : OIr. onn
(Val)brona : Welsh bro, OIr. on(-chú)
Canzo, Canzaga : Ir. céite, -ach
Caslino : Gaelic Caislín (Ir. 1 cais, 1 lín)
Erba : OIr. erb
Incino : Ir. éicne
Buccinigo : Ir. Buaignech
Longone : Irish 1 long, on(-chú)
Segrino : Ir. seg, cf. rían
Corneno : Welsh, OBret., MCorn., OIr. corn, OIr. en
Suello : Bret. huzel
Bosisio : Gaelic buidse
(Costa) Masnaga (XIIIth c. Masenaga) : Welsh, Breton maen , OIr. magen, -ach
Brenno : Gaulish Brennus
Nibionno : Welsh 1 nyf, OIr. ond
Tabiago : Welsh taflog
Merone : Ir. 1 mer, on(-chú)
Alserio : Ir. ail, ire
Cremnago : OIr. crem, on(-chú), -ach
Mariano (XIIth c. Marliano): Gaulish *margila, Welsh iawn
Carimate (915 Canimallo) : Welsh 1 cain, mall
Cantù (987 Canturi) : Ir. cét, ú(i)r
Galliano : OIr. Galian
Verzago (1060 Vertezago) : Ir. feirtche, -ach
Rovello (Lat. Rodellum) : Ir. rodial
Lomazzo (988 Lemacio) : Ir. 1 lemach
Cirimido (859 Celemani) : Ir. ceilemain
Veniano (1346 Vetegnano) : Welsh gw(y)den, iawn
Appiano (XIIIth c. Aplano) : Gaulish -apa, -la:no-
Guanzate (XIIth c. Vogonzate) : Gaulish Vocontii, OIr. áth
Bulgaro(grasso) : Ir. bolg, 4 ar
Caccivio : Welsh caeth, ywydd
Oltrona (962 Altrona) : OIr. altr(-am), on(-chú)
Beregazzo (1346 Baragazio) : Ir. 1 barrach, 2 áite
Drezzo : Ir., Bret. dris, dres, Welsh drys
Trévano : Welsh trefan
Bizzarone (Oir. bidg, 4 ar, on(-chú))
Caversaccio : Ir. cuairsce
Cagno : Welsh 1 cain
Concagno : Galatian Konko-, Ir. án
Solbiate : Gaulish Su-, lubi, Ir. áth
Albiolo : Welsh 1 elfydd, 1 ôl
Faloppio : Welsh ffâl, Insubrian opulus
Camnago : Welsh cafnog
Luisago : Ir. lúisech
Cadorago : Ir. cathrach
Minoprio (1335 Minovrio) : Ir. mín, 1 brí
Asnago : Ir. asnach
Prestino : Welsh prys, 1 -in
Senna : Or. sinnae GS
Capiago : Ir. capp, -ach
Lipomo : Ir. lic, mag
Urago : Ir. úrach
Solzago : Ir. soilsech
Ponzate : Ir. cóicthe, áth
Blevio : Welsh blew
Cernobbio : OIr. 1 cern, (su-)bae, (du-)bae
Moltrasio (1019 Maltraxe) : OIr. mall, tráige
Bisbino : Breton biz, blin
Urio : Ir. úir, úr
Nesso : Welsh, Bret. nes
Zelbio : Ir. gelb(-dae)
Lemna : Welsh llwyfen
Laglio : Welsh llaill
Brienno : OIr. 1 brí, inn
Argegno : Ir. airgen
Dizzasco : cf. Gaulish Diuici-acus
Cerano (1053 Clarani) : Ir. clár, Welsh clawr, iawn
Erbonne : OIr. erb, onn
Casasco : Ir. cass, MLatin (< Gaulish) ascuum 'pascuum'
Blessagno : Ir. blisán
Pellio, Peglio : Ir. cíall, Welsh pwyll
Ponna : Ir. conn (if at least in part connected with cenn)
Laino : Ir. 1 láith, Welsh 1 -in
Colonno : Welsh 2 coll, OIr. onn
Sala : OIr. sal f.
Tavazzano : Welsh tafod, OIr. 2 án 'swift'
Lenno : OIr. linn
Bellagio (884 Beslacio) : Bret. bes, laguenn
Tremezzo : Welsh tref, Ir. mide
Griante : OIr. grían, étan
Menaggio : Ir. menaige
Grandola : Welsh gran, Ir. grend, 7 ol
Bregagno : OIr. 1 brí, -án
Sanagra : Welsh hanu, Ir. ér
Rezzonico (Ravezonico) : Ir. Roe (Ptol. Rhaouíos), coí, on(-chú), -ech
Cremia : Bret. krem
Calozzo : Welsh call, CIb. (Pou)śTom
Mianico : Ir. meile, ainech
Garzeno : Ir. 2 gairge, en
Germasino : OIr. gerr, cf. maigne
Liro : Welsh llyr
Livo : Ir. lí, Welsh lliw
Domaso (851 Domacio) : Mod. Ir. dumhach
Vercana : Gaulish uer-, Welsh cawn
Trezzone : Ir. 4 tríath
Colico (locally Colach) : Ir. colach
Porlezza : Welsh pôr, Ir. 1 leithe
Bene : Ir. benn
Carlazzo : Welsh carleg
Gottro (XIIIth c. Gotoro) : OIr. 1 got, 1 or
Corrido : Welsh cyfred
Soldo : Gaulish sold(-urii*)
Albogasio : Gaulish albos, Ir. 1 géis
Drano : OIr. drenn, drénd
Loggio : OIr. lugae, luige
Cressogno : Ir. Cresan, cres
Claino : Ir. clad, Welsh cla(w)dd, 1 -in
Osteno : OIr. Uisnech, Sc.Gael. oisinn

>
>> 6) Pre-Roman inscriptions in the area of -ate-names are *only*
>> Lepontic and Gaulish, therefore only Celtic (other linguistic layers
>> can theoretically have been present, but till now one has not found
>> any contemporary evidence of them)

> DGK:
> Excuse me? What about Reate (modern Rieti), home of Varro? What about the
> comment about Sacrani from Reate driving Ligurians and Sicels out of the
> Septimontium (Paulus ex Festo)?
>

Bhrihskwobhloukstroy:
Rea:te n., commonly analyzed as a -te formation like Ates-te and
Tea:-te; Ates-te with toponymic suffix -te on a basis like river name
Ates-is (Adige).
None, to my (little) knowledge, has proposed a *lexical* etymology for Rea:-.
For what it may be worth, my hypothesis is a connection with river
names R(h)e:nus (< Celtic *Re:nos) < PIE *Reinos < *h1/3reiH-no-s
(√*h1/3reiH- 'flow'), parallel to Ates-te : Ates-is and with reference
to the confluence of Velino and Turano rivers. Shortly, PIE
*H1/3reiH-ah2-ti '(place) where (rivers) flow (together)' > *Reya:ti >
Italic Rea:tí > Latin Rea:te.

But Your question obviously isn't about the etymology of Reate
(You have not given any etymology of this name); it rather 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).

>> Working hypothesis: -ate = OIr. áth 'ford'
>>
>> Let's see: à th Nó < Celtic *Ia:tus nowos 'new ford'; *Nowo-ia:tus
>> > Latin *Nouoia:tus > Romance †Novoggiate: it doesn't work
>> But let's try again: à th Nó < Celtic *Ia:tus nowos 'new ford' < PIE
>> *H1iah2tus ne/owos; *Nowo-h1iah2tus > Late PIE *Nowo:ja:tus > Celtic
>> *Nowa:ja:tus > *Nowa:a:tus (regular loss of /j/ between identical
>> vowels) > Gaulish *Noua:tus > Latin *Noua:tus > Romance Novate: it
>> works!
>>
>> Consequences
>> 1) -ate-names mean indeed 'fords'
>> 2) they really correspond to Irish áth-phrases and -ad-compounds
>> 3) they are made of Celtic lexemes
>> 4) they must have been generated *as compounds* before the Late IE
>> loss of laryngeals (otherwise they would yield †-oggiate, not -ate)
>> 5) as facts nn. 4-5 show, these place-names cannot have been
>> trasferred from elsewhere (like e.g. York > New York), they are
>> precisely place-names explicitly coined for their very places (unless
>> the whole river net has been transplanted; but Lombard river names
>> very rarely have correspondents North of the Alps, they don't belong
>> to the Old European layer, they are rather 'Mediterranean')
>> 6) therefore, pre-Late PIE was spoken in these places when these
>> place-names have been coined
>> 7) other compounded place-names in the same region with same
>> compounding members (e.g. *bri:ua: 'bridge') show phonological
>> treatments (e.g. neognós-rule) that imply the action of a Common PIE
>> (not simply pre-PIE) phonological rule (it would have been no more
>> operating as early as the Late IE phase)
>> 8) these place-names have passed through the whole diachronic
>> phonology from PIE to Gaulish
>> 9) no trace of any other diachronic phonology can be found (Iike
>> epigraphical evidence, see fact n° 6)
>> 10) therefore, Cisalpine Celtic has directly developed from local
>> PIE (the whole lexicon is Celtic; the whole diachronic phonology is
>> Celtic; there's no trace of other phonologies)
>> 11) since those 'Mediterranean' *river*-names can be etymologized
>> through IE lexicon (and Celtic diachronic phonology) and alternative
>> etymologies (e.g. Basque ones) are either much weaker on phonological
>> ground or (more often) lack at all, those river-names cannot be
>> considered as relics of pre-IE languages

> DGK:
> I think you overwhelm yourself with your own ingenuity. You still have not
> explained Reate, far from any Celtic influence.
>
Bhrihskwobhloukstroy:
Now I have, see above

>> Given these consequences, pairs like La:rius (Lake Como) (= Welsh
>> llawr, bottom ground of a valley): *Pla:rius (> Piario [Bergamo]; on
>> the bottom of its valley) are best explained as Celtic La:rius vs.
>> 'Late IE' *Pla:rius. This coincides with Your view. If You doesn't
>> like to lable such Late PIE 'Celtic', it's just a matter of
>> terminology.
>> The crucial point is that Late PIE *Pla:rius doesn't not represent
>> a preceding linguistic layer, but simply an older phase. What You
>> can't do is to infer that the whole area between Orobian Alps (where
>> Piario is) and Maritime Liguria (where Porcobera flows) had non-Celtic
>> features (like /ar/ from syllabic /r/) before the Celts, because such
>> divergences in developments would have been registered in the PIE -ate
>> names, which is not the case.
>

> DGK:
> Reate, Reate, Reate.

Bhrihskwobhloukstroy:
see above, see above, see above

> (...)
>> Now You can:
>> a) refuse this theory and all other theories (You'll be a Skeptical)
>> b) refuse this theory and accept weaker theories (You'll be
>> incoherent)
>> c) accept this theory and other theories as well (You'll be
>> open-minded)
>> d) accept this theory and reject other ones, incompatible with
>> this one (You'll have changed Your mind)

> DGK:
> e) continue developing my theory

Bhrihskwobhloukstroy:
Until You won't have given a regular explanations for all the 700
place-names I've mentioned, this will coincide with case b)

>
>> > DGK:
>> > I do not have a convincing etymology for Ingauni, but that does not
>> > make
>> > Patrizia's punk-rock 'Tattooed Ones' better than nothing.

Bhrihskwobhloukstroy:

Don't Brithwyr and - of Celtic or Latin origin - Picti, Pictones,
Pictavi tell You anything in favour of 'Tattoed Ones'? At least let
Punk Rock in peace and mention Gaul and Britain

(...)
>> So, we have competing comparisons, univoque esternal linguistic
>> data, and competing phonologies (but with different degrees of
>> probability). Therefore, we can continue in producing PIE etymologies
>> with Celtic diachronic phonology. We can, this is beyond doubt; are we
>> also afforded to do that? This is Your real question.
>> Your answer is 'no', this is clear. I'd like to know why, because
>> I am not able to find - in this concrete case - an objective line not
>> to be trespassed

> DGK:
> Probability refers to the future. In dealing with the past, what counts is
> plausibility. This cannot be objectively defined, and we are not dealing
> with concrete cases. That should not bother us.
>

Bhrihskwobhloukstroy:

having said that, You still haven't precised to what extent an
etymologist can do his work and when his etymologies start to become
"too many".
I've exposed my reasons; I admit that I cannot do better than so.
This means that You'll continue to think to be right and I'll continue
in turn to think that I'm right, since I've replied to Your objections
and anyone can see whether it suffices or not.
If we'll be there, I foresee that in 2057 we'll be still amusingly
discussing, unless new data will emerge.
Now it has taken 4 hours and a half to write this reply, added to
the four hours of the day before yesterday; life is short and I must
work to eat (this is indeed my work, but I also have to write other
papers instead of continuously defending some ideas I've already
published). I'll reply once again to the endless critiques by Tavi and
then I'll have completed what I can state.