mardi 15 novembre 2016

Original Lautstand (IF Really a Branching Out from a Proto-Language) - Alternative Hypothesis, Mainly a Monologue

Original Lautstand (IF Really a Branching Out from a Proto-Language) - Alternative Hypothesis, Mainly a Monologue · Kephalé / Galva ? · Discussion of Celtic Inis (Welsh Ynys) with Latin Insula, perhaps Greek Nesos

Hans-Georg Lundahl
Have you considered the possibility that PIE might have had a Lautstand close to pre-Verner Germanic?

I tested non-merging sound laws (obviously also mergers b+bh = b for Celtic and Balto-Slavic) this morning.

billy-goat *khaferos, knee *keneu, Phrygian/Usual/Useful *bruk.

drag chain:

khaferos, keneu, bruk
khaferos, keneu, vruk (b, d, g > v/β, ð, γ)
khaferos, geneu, vrug (p, t, k > b, d g)
kaperos, geneu, vrug (f, þ, χ > p, t, k)

And from v/β, ð, γ on to SIE bh, dh, gh/ph, th, kh/f, b; f, d, b; f, h, g.

[Here SIE = South Indo European is correctly used as opposite of NIE as per kephalé article]

How about non-aspiration of first of two aspirates, both Greek and Sanskrit?

dhrighs = dhrichs
dhrichs, dhrighos
dhriks, dhrighos
dhriks, drighos
thriks, drikhos > known Greek thriks, trikhos, by evening out some difference between d and th.

How about transition to diverse SIE aspirates?

vrug > frug > phrug first devoicing in order to emphasise, then taking over by a language lacking non-sibilant fricatives (confer Pilipino for Filipino).

Parallel process for Sanskrit, but without devoicing first.

erudhros > ruvros > ruber

Back to NIE:
PIE to Germanic = Verner. Merger of b, d, g and some positions of f, th, kh.
PIE to Celtic and Balto-Slavic. Merger of b, d, g with earlier p, t, k (same process as in SIE).
PIE to Celtic, also: f > h > zero.

Not worked out implications for Satemisation or sp, st, sk.

The latter COULD have been sf, sth, sch, and English/German sch > sh could go via partial satemisation of ach-laut to ich-laut.

It could also have been simply, sp, st, sk.

Tested what would happen to a word like "babbel".

Thot bablom / Tha babla ...

Tod vavlom / ta vavla.

SIE keeps nt pl or turns it into feminine:

Greek :
paphla as in Paphl-agon son of Phineus and his descendants Paphlagonians, probably meaning "speaking an uncomprehensible really foreign language".

Latin :

Germanic keeps singular.

Satemisation ...
if original for satem/sto etc involved an ach-laut which satemised to ich-laut, so much easier to understand why no language has an affricate. Ç and SH usually go less well as affricates than TH, KH and F as plosives.

For Celtic,
having "sp" as originally "sf" is of course a boon, since "sf" tends to either "sp" or - Celtic solution - "sw".

For Farsi,
having sp, st, sk as original sf, sth, skh is not bad.

For Old Irish,
only initials are doing the SIE mutation of (f), th, kh to (p), t, k.

[SIE - South Indo European, here used for non-Germanic which is other division than in previous, so change SIE to non-Germanic, since Celtic and Old Irish are NIE.]

Later still
and not separating
While meditating over "babla akonthi"* > "paphla agousin", some phonetic things struck me:

If the older pronunciation was the Germanic one (or as Germanic as one can be pr e-Verner and pre-vowel shifts), then we can also relate to:

  • Greek losing final -th and making final -nth > -n;
  • -onthi > -onsi > -ousi outside Doric (BUT there would be a chronological problem if Mycenean Greek as per Linear B had -oti and distinguished th from t)
  • Greek also losing final -t (which became -d in Latin), since final dentals had so much less rendering after loss of final -th.
  • -nth- (not as in the other -nth- of non-IE origin, which might earlier have been -ndh-) becoming -nt- in forms of participle, which might also make the chain above a partial push chain rather than drag chain.

Note * : if earliest Paphlagonian lang was Kaskian, one can relate to IE speakers saying sth like "babla akonthi". And since Venetians are Paphlagonians, one can still, in another sense, say "fabulas agunt".

In this vein, "sagum" could go back to a PIE (=Germanic pre-Verner or so) *sakom - which could be an earlier loan of the same word which was later borrowed from Semitic (I think, confer "Sackcloth (Hebrew שַׂק saḳ) is a term originally denoting a coarsely woven fabric, usually made of goat's hair.") as (as I recall) "saccus".

Nelson Goering
Can I ask what the point of this would be? Even if we can notate the necessary sound changes, it doesn't seem very likely that we'd have shifts of (for instance) *t > *d and then *þ > *t in basically every branch _but_ Germanic.

Hans-Georg Lundahl

Germanic goes, first, fadheer > fadher, OI goes faþeer > aþeer > aþiir (and no þ > t between vowels there!) ...

So some branches do keep some of the þ rather than make t. (Do you have a t-pot ready, if ever I visit you? The pun made me t-thirsty!).

Hittite seems to scramble t and d into t.

Phrygian seems to be basically as Germanic.

In other words, euphony.

As for statistic likelihood, what if Germanic is the remaining non-branched version of PIE?

If you count on a full Swadesh list by computer, extant Germanic langs are a sister branch to Gothic, and these together a sister branch to PIE (as per usual reconstruction, no doubt) before all of this is attached to other branches.

However the 30/40 word extract from Swadesh list (can't recall the Russian's name) gave another result.

I meant the Swadesh Yakhontov list. It gave another result. It was the complete Swadesh which made extant Germanic and Gothic sisters and close cognates to PIE which together were a further off cognate with other IE langs - the comparisons had been done by computer.

Swadesh, not Swadesh-Yakhontov
Germanic Gothic PIE other IE
|_ _| | |
  |_ _| |

Aucun commentaire:

Enregistrer un commentaire