From: Patrick Ryan
Message: 51967
Date: 2008-01-27
----- Original Message -----
From: "Richard Wordingham" <richard@...>
To: <cybalist@yahoogroups.com>
Sent: Sunday, January 27, 2008 1:32 PM
Subject: [tied] Re: Existence of PIE (was: Nostratic language family)
--- In cybalist@yahoogroups.com, "mkelkar2003" <swatimkelkar@...> wrote:
> You may want to check:
>
> Trubetzkoy, N. S. (2001), Studies in General Linguistics and Language
> Structure," Anatoly Liberman (Ed.), translated by Marvin Taylor and
> Anatoly Liberman, Durham and London: Duke University Press.
>
> Six specific structural features have to be present for a language to
> be classified as "Indo-European." I do not know what they are.
1. No vowel harmony. (This does not exclude assimilation in height)
2. The consonantism of the beginning of the word is no poorer than
that of the middle and end. (Geminates don't count! Trubetzkoy
claims they can't occur word initially - surely an odd claim for a
Russian to make!)
***
Wrong. I have found through intra-proto-language comparison that final *bh,
*dh, *gh, of derived from pre-PIE *bha, *dha, and *gha, are frequently
simplified to *b, *d, and *g. I attribute this to idea that, at a time
between pre-PIE and PIE, vowel glides (-*y- for *e/i; -*w- for *o/u). The
presence of these additional elements evidently prevented loss of
aspiration. The presence of these glides is supported by the palatalization
of dorsals and the sporadic velarization of apicals.
Secondly, I have established the existence of PIE dorsal nasals by the same
method. Finally, PIE has *ng(^) and *nk(^) which frequently resolves to
*g(^) and *k(^). Initially, these nasals are always resolved to the
appropriate stops.
***
3. The word does not necessarily begin with the root. (I.e. all IE
languages have prefixes.)
***
I would make the distinction between prefixes such as the late and
unlamented &d-. and adverbial/preposition elements prefixed to the root.
***
4. Words are formed not only be means of affixes but also be means of
vowel alternation within root morphemes.
***
Wrong again. Words are not formed by vowel changes in the root; these are
strictly grammatical.
***
5. Not only vocalic but also consonantal alternations play a
morphological role. (He alleges they are impossible in Semitic
languages - I can only presume Ethiopic languages and Modern Hebrew
are not Semitic.)
***
There is no regular process involving consonantal alternation (I presume he
means exchange *gh for *g or *k). Different consonantal extensions, of
course, modify the meaning to make new words but the consonants are not
morphologically relatable only semantically.
***
6. The subject of a transitive verb receives the same treatment as the
subject of an intransitive verb. (This may be exemplified by being an
accusative rather then an ergative language, or by word order rules
where there is no significant inflectional difference.)
***
Lehmann, for one that I know, and I think Beekes as well, have speculated on
an early PIE that was not accusative in form. The lack of an accusative
ending for neuters hints strongly in that direction.
***
He claims that these six features only co-occur in Indo-European and
co-occur in all Indo-European languages.
(Samples of 'Studies in General Linguistics and Language Structure'
can be got by googling 'Trubetzkoy "Indo-European" structural'.)
It hadn't occurred to me before that Thai was Indo-European! The
evidence:
1. No vowel harmony.
2. Phonation types are only contrasted syllable initially. All
syllable-final consonants can occur syllable-initially (which is not
true of English!)
3. Thai is very rich in prefixes.
4. This was the tricky one. I suggest that the vowel alternation seen
in alliterative 'compounds' satisfies this requirement (e.g.
pliang-pleng 'change'), though it is less predictable than in English
strong verbs.
5. Li has actually assembled a collection of related words differing
in voicing, though in some cases this has now become a tone difference.
6. Thai passes the test just as French does.
Mind you, it requires some dishonesty to assemble the Indo-European
vocabulary for Thai, though Indic words are by no means missing from
the 100 word list if you construct it by the book. (E.g., look up
'father' in Se-ed's Modern English-Thai dictionary, and you get
_bida:_, which comes from Sanskrit/Pali _pita:_, as the first
translation.)
I thought Cappadocian Greek would fail on the vowel harmony test, but
its 'vowel harmony' does not appear to be of the Turanian type.
Of course, it is remarkable to see that Hindi is not Indo-European -
split ergativity!
In short, the six features are not as persuasive as one might hope.
Richard.
***
What a horrendous shame that a respected linguist could formulate such
criteria!
Patrick
***