Jens:
> But is that more than a statistic thing? Does it really matter that
> much?
Me:
> In my theory, that seemingly trivial matter matters very much. []
Jens:
> Fine. Would you be so kind as to give, say, five clear examples of it?
> Examples, that is, of /i/ replacing the thematic vowel in *pretonic*
> position, as you say the rule is. Speak here.
As long as you can listen patiently for once, we can get to the bottom
of this.
All *i-reduplicated stems are examples for one thing (eg: *gi-gnehW-,
*bHi-bHer-, etc) but no doubt you'll say that they show initial accent,
ignoring the big tip-off that the accent isn't original here because
it's fixed on the first syllable in stark contrast to athematic
stems which show accent alternation. It should be obvious that this is
caused by Acrostatic Regularization, a rule that affects all thematic
stems of that period while athematic stems like *kwon- being unaffected
show the original wonky accent pattern. The whole reason for accent
regularization here and in the other thematic stems is in order to
avoid multiple accent positions on a single stem. If a stem must have
an accented syllable, Acrostatic Regularization guaranteed that it fall
on one select syllable only, so the initial syllable was chosen for
all case forms.
The original accent in the *i-reduplicated stems is on the second
syllable. Not only does the full-grade *e in the second syllable &
the zerograde in the first show that it was _that_ syllable that was
accented, the fuller form of the verb stem is even reflected in that
syllable (ie: we don't see **bHer-bHi-, do we). Two reasons why the
accent was logically on the second syllable, and thus two reasons why
the *i here is pretonic. We then don't need an imaginary *i-infix here
because we can unceremoniously explain *i in these stems as the reflex
of phonetic schwa in an open syllable before accent. So *bHi-bHer- is
from *bHa-bHér- with pretonic unaccented schwa reflected in my notation
as the phoneme *a. The form is just the partial reduplication of the
default stem *bHer- and that's it.
Not only are *i-reduplicated stems an example of pretonic schwa, we also
have compounds with the first element in *i, like say, *xegri-kWolex
"farmer", derived from *xegro- "field" and *kWolex "tiller". Things
like this only make sense if an underlying stage had *xegra-kWálax
with the first *a being pretonic and predictably raised to *i just as
I've been saying. Common sense.
The only thing that may seem hard to follow on your part is why there
is the alternation of *o/*i in stems without there being any direct
evidence of "pretonic schwa". Such cases would be when o-stems are
converted directly to i-stems, whether it be accented on final syllable
in adjectives or initially in nouns. How to account for these
counterexamples?
I've already stated the reason for this: analogy. With the pattern of
compounds being formed with thematic stems using the i-variant as I've
shown above, and being that the first element is normally descriptive
of the second element in such compounds (example: "dog-house" is a kind
of house (the object in general)... for a dog (further specification)),
analogy took place by viewing the first element almost like an adjective,
an entity that helped to in some way modify the following noun stem.
So...
My understanding is that when adjectives become nouns in IE, they shift
the accent off the thematic *-ó- to the first syllable (a la nominal
acrostatic pattern). The adjectival accent has already been explained
away as genitival in origin, so such a noun now follows the pattern
*CVC-o-s. The question is what happens if a noun like that should be
converted back into an adjective. Do we just change the accent back?
Perhaps, but another possibility is that there is a different pattern
for denominal adjectives, the caland variation. So such a noun would
go from *CVC-o-s to an adjective *CVC-í-s with accent on the *i to
conform with adjectives. And then, because crazy things do happen in
linguistics, it would be possible to convert it back into a noun
again by shifting the accent as per the deadjectival noun process
described in the preceding paragraph: *CVC-i-s.
All from a simple analogy. I don't see how you can really argue against
this. Your explanations seem far more convoluted in design while mine
try to explain things in the most mundane way possible. No fanfare, no
pizzazz. No tonal *z or dually articulated Ferengi phonemes. Just plain
ol' boring morphology and analogy at work here, as is always the case in
language.
= gLeN