Re: Reconstructed Substrates (was: Bart; was: Ligurian)

From: Bhrihskwobhloukstroy
Message: 69824
Date: 2012-06-17

2012/6/12, dgkilday57 <dgkilday57@...>:
>
>
> --- In cybalist@yahoogroups.com, Bhrihskwobhloukstroy
> <bhrihstlobhrouzghdhroy@...> wrote:
>>
>> 2012/6/8, dgkilday57 <dgkilday57@...>:
>> >
>> > --- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@> wrote:
>> >>
>> >> W dniu 2012-06-07 03:19, dgkilday57 pisze:
>> >>
>> >> > Thus the river Druantia in Liguria Transalpina (now Durance) can be
>> >> > equated with Skt. Dravanti: 'Running (River)' f. from *drew-n.tih2,
>> >> > with
>> >> > the same Lig. innov. absent from Celtic. Likewise the smaller rivers
>> >> > Drance (*Druantia) in Kt. Wallis, and Durance in De'p. Manche, with
>> >> > Drouance in De'p. Calvados, Normandie. That is, Greater Liguria
>> >> > stretched across Gaul until it was split by Gaulish invasion and
>> >> > expansion from the south (cf. Liv. 5:34).
>> >>
>> >> Would it include today's northeastern Poland and the River Drwe,ca <
>> >> *drUvoNtja, one of Torsten's favourites? (no trace of *dreu- in
>> >> Balto-Slavic, and absence of Grimm's Law excludes a Germanic
>> >> intermediary).
>> >>
>> >> http://en.wikipedia.org/wiki/Drw%C4%99ca
>> >
>> > It looks that way. Artemidorus did say that the Ligurians once ranged
>> > all
>> > the way to the Northern Ocean. Earlier I attributed this remark to
>> > misunderstanding on A.'s part of how far north Worms-am-Rhein is, but
>> > perhaps he was spot on.
>> >
>> > DGK
>> >
>> Bhrihskwobhloukstroy:
>>
>> So Your Ligurian is, like d'Arbois', Dottin's, Philipon's and
>> Kretschmer's, an IE Old European stratum like Pokorny's
>> Veneto-Illyrian.

> DGK:
> Yes. However, I do not believe in lumping Venetic (with its [h] and [f])
> together with Illyrian. I think Macro-Illyrian (Illyrian proper,
> Macedonian, Paeonian, Messapic, Japygian, probably Rhaetic and Belgic)
> belongs with Lusitanian in an "Illyro-Lusitanian" node.

> Bhrihskwobhloukstroy:
>> As I've many times pointed out, I'm find such hypotheses very
>> attractive. I constantly try to fit them all in a maximal frame; I'm a
>> kind of a collectionist of these theories.
>> I've noticed that I can found new hypothesis in this perspective
>> simply out of one and the same language: 1) In the territory and
>> linguistic documentation of language X (of IE descent), I look for an
>> ethnonym Y without any linguistic affiliation; 2) I choose a
>> place-name that can be geographically associated with this ethnonym;
>> 3) I take into consideration the PIE etymology (if any) of this
>> place-name; 4) I try to find if it can be modified (especially with
>> regard to ablaut) in order to have a similar etymology, maybe from the
>> same root and in any case a sufficiently straightforward one, but with
>> a different diachronic phonology (Y) as opposed to the one of the
>> locally attested language X; 5) I apply diachronic phonology Y to
>> every name of X I can; 6) the area of the names with which I've
>> succeedingly applied diachronic phonology Y is identified with the
>> ancient territory of the linguistic stratum named after the ethnonym
>> Y. In X You can read 'Celtic', 'Germanic', 'Baltic', 'Slavic',
>> 'Greek', 'Armenian' and so on; in Y You can read 'Sorothaptic',
>> 'Latial-Ausonian'/'Palaeo-Umbrian', 'Pelasgian', 'psi-Greek',
>> 'Themematic', 'Greltic' and so on.

> DGK:
> That sounds like a straw-man procedure.
>


Bhrihskwobhloukstroy:
I am my own straw-man; I was really describing my procedure and I
cannot see any other one in order to obtain new substrates. It's based
on what Eichner has labeled "etymolgical procedure": Equation with PIE
proto-forms directly continued by attested forms in IE languages and,
on the basis of such proto-forms, construction of a diachronic
phonology; the alternative method (of the "tentative reconstruction
postulates") is the blind application - like in the "minimalist"
section of my approach - of formerly (through etymological procedure)
established sound-laws.
If Your method differs from anything I've written, I'll be very
interested in learning it!


>> What I have to add in order to get a complete picture is to
>> examine the opposite possibility, a minimal amount of language
>> substitutions. (Alinei's Continuity is in no way minimalist on this
>> point, because it takes Pre-Latin languages as superstrata, therefore
>> implying a double language superposition - their arrival and their
>> death). There's a limit beyond which one's Reductionism is patently
>> falsified (e.g. in the case one should try to deny Illyrian or
>> Thracian or Continental Celtic); before that very boundary, I think
>> it's our duty to find out the minimalist approach. It doesn't mean
>> it's right; it suffices that it be both possible and extreme. I can
>> myself construct the opposite extreme, applying to each name of X the
>> method I've just exposed.

> DGK:
> Your opposite extreme requires indecent liberties with ablaut and other
> matters.

Bhrihskwobhloukstroy:
A Romance-speaking scholar could have signed a statement like that.
Ablaut has precise rules; anything outside them would be out of the
language system, but You can't seriously maintain that I'm violating
ablaut rules.
The opposite extreme position would be the typical Neo-Latin one:
only attested words can exhibit ablaut patterns, but nothing outside
them can be generated. You seem to tend to such a position, i.e.
ablaut as a no more productive strategy. (I can't imagine "other
matters" in which I can license myself "indecent liberties", so I
reply only to Your argument about ablaut.)
Note, however, that when we discuss about ablaut we aren't referring
to Ligurian, but to PIE: e.g., *bho:rg'h-ah2 or *g'enh1/2-o:wah2 are
PIE transposits, at morpho-phonological level they have to be analysed
as PIE - not Ligurian - formations; the specific Ligurian and Celtic
contribution is just on phonological level: *bho:rg'h-ah2 > Barga:,
*g'enh1/2-o:wah2 > Gena:wa:. I'm sure You wouldn't deny ablaut was
productive in PIE. Maybe You'd contend I don't know the details of the
*norm* (as a concrete realisation of the system) of PIE ablaut: If You
know them, I'm ready to learn, as above!

> DGK: One could equally well construct English etymologies for every
> place-name in the United States. For example, Hoboken (N.J.) could be
> analyzed as *hobo-ken 'tramp-knowledge', i.e. a good place for a tramp to
> get a handout, since tramps share knowledge of such places. This would
> disregard the evidence pointing to a dialectal Dutch etymology as 'high
> beeches'. English dictionaries are very large and can easily accommodate
> such analyses for virtually any place-name, just like your Celtic
> procedure.
>

Bhrihskwobhloukstroy:
You've chosen an amusing example of formally perfect ambiguity
between a Dutch and an English interpretation (I thank You for the
explanation, since to my ears Ho[og]-boken is of course much more
close to anything familiar - like Hochbuchen - than American English
hobo). It demonstrates that:
i) competing interpretations are possible;
ii) a Western European one is always possible with any non-Amerind,
non-Na-Dene, non-Eskaleut US name;
iii) on the North-East Coast, English and Dutch interpretations are
historically justified (evidently not only because of colonization,
but also as a consequence of later immigration);
iv) in this particular case, the Dutch interpretation is more
intuitive than the English one.

This maybe reveals the model You love. European colonization of North
America is a powerful model for Eurasian prehistory, but I hope You'll
concede it isn't the only possible one. In fact, the application of
Your example to Cisalpine Gaul leads to following conclusions:
i) competing interpretations are possible (therefore please take my
interpretations into consideration as well; as already said, I've
always taken Yours and Kretschmer's ones into consideration);
ii) a IE etymology is always possible and there isn't anything
comparable to Amerind, Na-Dene or Eskaleut in Central and Western, for
the simple reason that no pre-IE people and language is explicitly
documented in the vast area between Rhaetia, Etruria, Basque Countries
and Pictland (of whatever affiliation such languages may be, <mi
nemetie> and similar instances don't permit any Etruscan etymology of
Ligurian names);
iii) in Cisalpine Gaul, a Celtic etymology is always justified, a
Venetic one (the closest parallel to a Dutch etymology for a US
place-name I can imagine in this concrete situation) may at least be
searched, but a non-Celtic Ligurian is by no means paralleled by Dutch
in America: Dutch colonists and immigrants from Holland and
neighbouring areas are well attested and moreover Dutch and Flemish
are widely documented languages in Europe, while the very existence of
a Ligurian language with non-Celtic innovations is just the result of
a possible interpretations - among other ones - of the onomastic
material of (pre-)Roman Liguria;
iv) therefore, no matter how more intuitive a non-Celtic etymology of
a Ligurian place-name can be (with respect to a Celtic one), it will
never reach the status of a Dutch etymology in the US, but at least
the status of an etymology through a non directly attested and never
uncontroversially documented Germanic - e.g. German - dialect (just to
give an example: the hypothetical Rhineland pre-stage of Yiddish).
Sure, if such etymologies through a controversial Germanic dialect
(like Rhineland pre-Yiddish) are systematically more straightforward
than English (and Dutch, rest-German, Scandinavian and so on) ones, it
becomes more probable that such a dialect really existed and was at
the origin of these names (like the hypothetical IE substrate in
Northern Africa), but I permit myself to affirm that I still have to
find such a systematically higher straightforwardness of Kretschmer's
and Your etymologies of Ligurian place-names (not to speak of
Transalpine place-names in uncontroversially Celtic territories!) as
opposed to my reconstructions of PIE transposits through independently
assured and areally always possible Celtic diachronic phonology.
(Sorry for German length of sentences...)