Peter recently states on the Tychicus thread:
>Latin verbs with -a- vocalism have a number of different origins, usually
>developing the -a- within Italic. There's work to be done there, someone!
And to this, I'd add that there is work to be done still on instances of
uncoloured *a in PIE.
I just realized that I'm probably wrong about *bHo:g-. Actually, it probably
_is_ derived from the o-grade of *bHeh- and thus we must write *bHoh-g-.
But now, that is an interesting conclusion because we naturally wonder:
What is *-g- supposed to convey, why do we find the o-grade in the stem,
and how does this relate to my statement that some instances of *a are
caused by voiced labial phonemes preventing a normal pre-IE change of
earlier *a to *o?
Now I think I get it. I'm now claiming that *bHoh-g- is derived from *bHeh-
and that it is the product of a stative nominal stem being attached with
an intensive marker *-g- which derives from the emphatic enclitic *ge. We
then can observe a series of these "intensive"-derived stems. Another
example of an intensive stem would be *bHux-g- "to flee, fly away" from
the secondary meaning "to rise" of the verb *bHeux- "to grow up from the
ground". (Or have I overanalyzed these verbs?)
If this is so, it shows that an IE verb could be produced simply with a
noun phrase without any verbal stem by using the intervening enclitic *ge.
This conclusion in turn validates the origin I claim for *?eg(-o:/-om) "I",
which is to be properly understood as literally "my being here; as for me"
from the synthesis of *?e "here", *ge [emphatic], and a 1ps ending.
Now that we have that straightened out, why did I mention the *a-rule that
I'm constantly revising? Well, the new rule goes like this:
Instances of uncoloured *a in PIE derive from the failure of the
general Late IE vowel shift (*a > *o) to operate on any short *a's
that neighboured voiced labial phonemes (*m, *b, *bH, *w) within the
default stem of a paradigm.
What does that mean by "within the default stem of a paradigm"? I mean
to say that if a verb was inheirantly aorist like *dehW- "give", then the
default stem is the aorist stem *dehW- (not the derived durative stem
*didehW-). Obviously. So this rule can only apply to *dehW-, not to
*didehW- or to *dedohW-.
So now, let's imagine a stem which is naturally in the o-grade in its
simple default paradigm. Now let's imagine that we are speaking early
Late IE (eLIE) where *a has not yet become *o. Here then, we must think
of a verb which is naturally in the "a-grade" instead. Now let's further
imagine that this eLIE stem of the natural shape *CaC-, which normally
develops into PIE as *CoC-, is something like **mag-. Since *a neighbours
a voiced labial in this example, *a will not become *o, and the result
is PIE **mag- instead of **mog-. However, if we imagine a pretend eLIE
durative e-grade stem **meg-, it will develop the normal perfect
**memog- (with *o, not *a). This is because we are now speaking of a
derived form which depends on the simple stem **meg-, rather than the
simple stem itself on which the above rule on *a applies.
Got it? So, back to *bHoh-g-, we would think that we should have **bHah-g-
until we realize that *bHoh- would be a derived stative form of the simple
stem *bHeh-. Therefore, the *a rule does not apply.
This rule would give insight into the origins of all a-grade stems in PIE
that do not appear to be the result of uvular colouring of *e next to *x
or *q.
Additional thoughts anyone?
- gLeN
_________________________________________________________________
Add photos to your messages with MSN 8. Get 2 months FREE*.
http://join.msn.com/?page=features/featuredemail