Re: [tied] Re: PIE as a pidgin?

25-12-03 18:18, elmeras2000 wrote:

> --- In cybalist@yahoogroups.com, "Andy Howey"
> <andyandmae_howey@...> wrote:
> [...] pidgin-like origin for a proto-language [...] Trubetzkoy
> (1939). [...]
>>
>> To what characteristics and from which languages would Trubetzkoy
>> have been referring?
>
> He said IE lgs. shared 6 features *with* which any language would be
> classified as IE and *without all* of which no language could be IE.
> By that definition a lg could become IE or cease to be so. And by
> piecing together some traits of Semitic, Caucasian and Finno-Ugric
> Trub. could set up a scenario whereby all IE lgs. could have
> crystallized out of fragments of the neighbors.

To be precise, Trubetzkoy defined "IE-ness" as the simultaneous
occurrence of the following _typological_ traits:

(1) There's no vowel harmony (the quality of the vowel of the root
doesn't constrain the quality of vowels in suffixes).

(2) The same classes of consonants occur in word-initial, medial and
final positions (crucially, there are no constraints on features of
initial consonants).

(3) The first morpheme of a word doesn't have to be a lexical root.

(4) Ablaut is regularly employed as a grammatical device.

(5) Consonant alternations can also be so employed.

(6) There's no ergative construction (the subject of a transitive verb
has the same form as the subject of an intransitive verb).

Well, it's an odd collection (not to say mishmash). Some of these
statements are uncharacteristic enough to be true of hundreds of non-IE
languages; (3) is questionable, since an IE word _normally_ begins with
a root and there are very few morphemes that could be called prefixes
(it is as if somebody included non-inevitability of initial stress among
the defining features of Germanic); and lumping (4) and (5) together (as
though PIE had had, say, consonant mutations on a par with ablaut) makes
little sense to me.

>
>> How valid is this suggestion?
>
> Zero percent. IE is defined by its past, and the past does not
> change.

I completely agree. The standard definition of PIE is not typological in
the first place. It refers to the reconstructed linguistic substance: a
common vocabulary plus a common morphological system (both, and the
comparative evidence for both, simply ignored by Trubetzkoy's
typological definition). IE morphology, incidentally, is the logical
opposite of what we find in creole languages.

>
>> Was this an attempt at a counter-argument against including PIE in
>> a macro-family such as Nostratic.
>
> No, it was rather an attempt to compromise the belief in "Aryan"
> supremacy.
>
> It is silly, though, for IF languages are not to be classified by
> their substance, why did Trubetzkoy go looking for a set of
> typlological features that would unite the *same* languages under
> the heading Indo-European as were already classified as such? If the
> classification was so wrong, why should its results not be changed?

Yeah. He wanted to eat his cake and have it too.

Piotr