Jens and I, at each other's throats again:
>> Certainly a theory is not hard evidence. I say that
>> *pelh- did not derive from *plh-. Prove me wrong.
>
> Schwebeablaut quite generally works that way: [...]
> In the case of *pleH1-, Piotr has also insisted that
> the old form has "full-grade two", using the
> comparative *pleH1-yos- of Gk. pleío:n and Skt.
> prá:yas- as particularly impressive points de repère.
I didn't say any conclusive proof against my view.
> The "o-infix" as I choose to call it [...] was first seen
> in the causative. The whole theory is about the vocalism
> of the causative. The option *Vstex- is new to the debate,
The option of *Vstex- had always existed but I should have
explicitly mentioned it since it would not be obvious, I
now realize, in the way that you think about Pre-IE.
As for the O-fix, I'm more ready to accept the process in
nominal stems than in causatives, regardless of its
foundation. I checked out your Feb 10 posting of "Infixal
/o/" and found that ultimately there was no deep reasons
for positing the O-fix in the causative from what I
could see mentioned. For example, while it might be
functionally correct to say that the causitive verb root
is a little longer than it is expected to be under the
normal processes of quantitative ablaut, there's nothing
to suggest that your analysis is the only one, nor that it
is the most optimal idea around.
You use *mon-éye-ti, which is expected to be **mn-éye-ti,
and also *swó:p-ye-ti, expected to be **swép-ye-ti (this
2nd analysis I don't agree with). You deduce from this
that the stem is "an *o longer" than it should be. This
is a quantum leap since it can only be legitimately
asserted that the stems are longer than expected at
best, not "an *o longer" (and yes I'm aware of your
analyses of qualitative ablaut which can also be
interpretted differently).
On the other hand however, these examples can be
analysed a different way: They _resist_ ablaut and
are in the *o-grade. The question that needs to be
answered is whether the resistance is caused by the
O-fix as you say, or is it because they had developped
to the state they are now found in Reconstructed IE
after the automatic process of quantitative ablaut had
passed on. Nothing stated in the post shows that the
O-fix must be accepted. In my view, after giving a
valid linguistic motivation for your O-fix that was
lacking, there is nothing that would place the
causative at risk of the "infix".
All we can immediately see the causative being is a
full *o-graded root, sometimes with alternating accent
that evidently resists ablaut. It can just as easily,
if not more easily without wonky mysterious phonemes,
resist ablaut because the accent and perhaps the
causative paradigm itself was formed after the process
of quantitative ablaut had ended. Clearly at some point
it did end. The stems *wlkWo- and *suxnu- with accent
on their initial syllables prove this event
conclusively.
What we gain from this more straight-forward
explanation is a solution that doesn't need to be
further explained. In contrast, we must explain your
"consonantal" *O phoneme. Why must we? Without further
evidence, I can't be bothered scratching my head on
that one.
Ironically, as we read further down that post, you
mention *luk-eye- and the like. You state "The
conclusion is that the -o-, which is ABSENT in some
root structures (generally, but not entirely, the
more complex ones), must have been once a consonant
that was either deleted because it caused too much
complexity or (in some special types) was assimilated
to the environment to such a degree that it escapes
detection." Sounds like phonotactics to me: hence
a-Epenthesis.
Yet, as for the causitive of forms like *luk-eye-,
I have to admit I'm confused. Is the causative
of a like stem *bHeudH- not *bHoudH-eye- with *o?
What's the deal? Is it *louk-eye- or *luk-eye-?
>> Clearly your sense of logic is flawed then. Most people,
>> aside from Jens, understand that "green" means "go".
>> I personally can't be worried about my eventual
>> demise for otherwise I wouldn't be able to function.
>
> Okay, but in my case it would be *my* demise, so I do
> care very much.
Well, you can't drive the car, remember? You're too
afraid of crossing on a green! Your hesitation will
only cause more accidents so you're safer with me
behind the wheel :)
>> I don't dismiss them. I prioritize them. There are many
>> theories I've dismissed only to have them revisited and
>> assimilated, including your ideas.
>
> Yes, 100% vs. 0%, and that's the silliness. Besides, I do
> not accept your priorities.
You may not accept my priorities but we can't make choices
if we don't decide one way or the other. Choosing requires
absolutive thinking. We can't "half-choose", we can't
"choose 30.5%". You're being silly now.
> I don't see how a suffix can begin ablauting after ablaut
> is over
That's simple. Syncope causes the quantitative ablaut, or
more specifically the "zero-grade", by dropping vowels
in unaccented position. We end up with oscillations that
are baked into the paradigms. So we have *es-t versus
*?s-ent, for example. The process also becomes the normal
rule for derivations where there are accent shifts because
that was what was going on even BEFORE Syncope, albeit
without vowel _reductions_ rather than zeroing. An example
in preSyncope eLIE:
*bHer- "carry" + -as' [genitive]
=> *bH'r-as'
This even occured in MIE when *e was unaccented by stress
shift, becoming reduced to *a [&]. There is a long history
of this quantitative ablaut. It would carry on into Late
IE until a certain point where, as I say, this automatic
process ended and stems like *wlkWo- and *suxnu- were
finally allowed.
It's not very mysterious at all.
> If what we were talking about at this point was the
> underlying degree of complexity of suffixes, then I
> actually do maintain that IE suffixes are simpler
> than many roots.
I don't think this point has much to do with what
we're discussing but I agree here. I just don't agree
that *-men- comes from *-ment- because I insist on
a simpler change of *-n>*-r in MIE. Some propose
*-nt > *-n or *-r but this is illusion since *-t in the
3ps does not exist at the time of the Rhotacisation
rule and *-n in *-mn would be the product of nasal
assimilation as Miguel stated many times.
>> [...] *wertmn would now be reconcilable according to
>> Mid IE syllabics: *wérta-man. It yields eLIE preSyncope
>> *wert'm'n and thus *wertmn. This is assuming that the
>> ending existed at that time, of course.
>
> No quarrel with the last part, but I cannot accept the
> root-final "-a".
I figured as much but luckily I thought about it because
I found myself panicking, considering whether it was
really **wert-man that I should be reconstructing in
MIE. However, this cannot be so and CVCC cannot exist
at this stage. Why?
The original point of the Penultimate Accentuation Rule
was to explain away the unpredictable accent of IE. It
worked in many cases brilliantly but, as with *woid-,
it wasn't quite right. So the bran' spankin' new idea
of Quasipenultimate Accentuation solves even more than
Penultimate alone could do. We _NEED_ *-a in order
to regularize the accent that is unpredictable in
Reconstructed IE. Otherwise, the root is not always
penultimately accented and the word as a whole is
not confined to penultimate or antepenultimate
accentuation. Dropping the *-a obliterates the entire
pattern that serves to explain IE accent. Why on
earth would you want to do that?
The motivation that you fail to see is quite simply
the underlying pattern that we gain to explain away
the source of mobile accent. We can answer questions
like why the accent alternates between strong and
weak cases with the least amount of fuss.
> I was speaking of the root *dhewgh- of Sanskrit
> dógdhi, duhánti 'milk'. There is no morphological
> seam in *dhewgh-.
As I said, I get an MIE root *deuga from it, so
what is *-wgh- supposed to be?
> No. The middle endings are not derived from active
> endings, but the pre-existing middle endings are
> partly, and increasingly, adjusted to the corresponding
> active endings throughout the history of the older
> languages.
Based on what? It's clear that middle endings have
a relationship to the active endings. The active
endings are more basic in function than middle.
Hence, it's more likely that the middle endings
derive from the active ones. Where is your assertion
coming from?
> Well, what the inflections sit on in IE morphology is
> in fact a stem. Do you honestly believe there were no
> derived stems by the time of the ablaut?
No but that's not the point. We don't know which ones
come from this event of Syncope (as opposed to "ablaut"
which should be kept seperate for reason I state above).
So rather than second-guess, we restrict ourselves to
indivisible roots and their inflections just to be safe.
> How did the optative suffix come to ablaut? The stative
> suffix?
Again, there was a long process of quantitative ablaut
partly existant before Syncope by the simple reduction
of unstressed vowels and certainly long after with
zero-grading. The ablaut process ended sometime in
Stage III Late IE. It "self-avowedly disqualifies"
nothing. You only misunderstand the timeframes involved
in my theory.
> The IE word consists of root + suffix + desinence,
> sometimes with multiple suffixes, and the ablaut worked
> on the lot.
But it wasn't always this way so it's absurd to impose
these same rules on Mid IE or your version of Pre-IE.
If you have the same morphological rules in Pre-IE as
you do in later IE, then there's something fishy.
> There is no way suffixation can be a post-ablaut
> innovation.
That statement makes no sense. Suffixation is a
PostSyncope event whereas it occurs through the
period where Quantitative Ablaut was a normal
morphological process long after Syncope. ????
On the 3ps:
> This is not serious talk. Of course I did not
> think of it first, Bopp did. But whatever the
> origin of the *-t, I would assume it was present in the
> 3sg active in the IE on which the ablaut worked.
Yes and no. It didn't exist in MIE but did exist
by eLIE. All this time is the time when "the ablaut
worked". You're confusing Syncope and Ablaut and
it's getting confusing.
As for license to derive *-t from *-to, we do have
rules but you'd rather not listen to them. It
turns out that it's simply the affixing of the
demonstrative to the bare verb stem and the
process of Syncope that we can be certain of. If
you ignore Syncope, I don't know how to get through
to you. If you don't ignore it, then we do indeed
have rules and therefore your assertion "for we
have no rules to allow that" is very incorrect.
> Is the verb 'to be' an old aorist now?
Yes, and it's also a Semitic loan. It was just
a thought. It could have been dragged
into the durative at some point because the
durative came to be the default aspect of the
verbal system and because *es- became heavily
depended upon. As for the other verbs, I'm
still not sure but a possibility exists that
they derive from the aorist. Otherwise we
have to state that the thematic *& expected
in eLIE was reduced by erosion. This is
equally another possibility since these
athematic roots are used most often. Perhaps
the latter will have to do.
> It is all we need. Once we get the subjunctive
> from somewhere we do not need the simple thematic
> type at all.
The thematic is a default, the subjunctive is a
derivative.
> I am of course saying just that. I think it's
> a fair statement that there are no root aorists
> with original zero-grade in the strong parts of the
> paradigm (active singular). I know only *bhuH-t as
> a serious candidate of a root aorist that generalized
> its zero-grade alternant already in PIE.
Well when you can disqualify all zeroed root-aorists,
let me know.
>> > Syllable structure does not create IE /o/ of any kind.
>>
>> I agree. It wouldn't be natural to have *o as an epenthetic
>> vowel. It's *a, that LATER becomes *o because of Vowel
>> Shift.
>
> No, syllabic structure did not create any *prestage* of
> IE /o/ of any kind either.
The alternation of *e and *o with *a cut out of the loop
(merely an allophone of *e) is not natural, just as
unnatural as infixing *o for whatever reason that you
evidently can't come up with. The overall lack of
uncoloured *a, putting aside ablaut, is even more suspect.
So in order to naturalize the system, one of the vowels
must derive from *a. Turns out that identifying *o as
original *a works best and we end up with a alternation
of *e and *a. I'm not the first to propose this alternation.
Your O-fix is an A-fix and is misnamed.
> Because the structure containing -o- is opposed to zero.
> The two are opposed to each other and so cannot be equated.
They don't oppose each other. I reject causatives in this
theory. So sticking to nominal roots, I don't see this
opposition. I see that there is every reason to deduce
that we're dealing with zero-grading, right down to the
omissions created by Saussure which would not exist as
naturally in a fullgrade environment. Your theory as it
is left is wonderfully mathematical but remains
unnatural.
> It cannot be an insert-vowel. The structures without
> the -o- do not insert vowels, so the structures with
> the -o- must have a different background.
This sounds circular. In my reinterpretation, which
results in the same as you have it, *o is inserted
into a zerograde syllable in order to remedy "bad"
syllables. There doesn't need to be a seperate *o.
Unnecessary complexity. And, as above, makes for an
unnatural language in every way (putting aside even
more multiplication of hypotheses such as your
triple-length... yikes!).
>> In your mind, you can't see anything but *CeC >
>> *OCeC > *COeC > *CoC when in actuality, *C'C >
>> (*aC'C >) *CaC > *CoC makes far more intuitive
>> sense.
>
> No, that becomes CC.
Yes, perhaps in your theory. In mine, a resultant
*CaC simply becomes *CoC by Vowel Shift while
*CeC remains *CeC. The reason why *mon-eye-ti
would insert a vowel has nothing to do with
O-fix/a-Epenthesis. It is something else unless
you can tell me why we need to include this. So
*mntos is the result of normal ablaut but the
causitive simply doesn't operate under normal
ablaut and uses full *o-grade, probably because
it is a late inflection.
On CCVC < VCCVC:
> Okay, that's a new formulation, but fine. But why would
> it be a CVC-pattern?
As I said, Quasipenultimate Accentuation. Otherwise the
mobile accent is unexplained. Since the rule is simple
and explains an unpredictable accent, we really really
need CVC because that's what QAR depends on. As a
result, we regularize the accentuation AND simplify
the syllabics. What a deal! Nope, I'm keeping this
rule.
> And why would the language refrain from using initial
> clusters if it has some elsewhere?
I don't know. Why does Turkish do that? Why,
those crazy Turks up to their phonetic shinanigans
again. I'm not "assuming" as you say. As I said QAR, the
Quasipenultimate Accentuation Rule, is needed to
explain away mobile accent.
> The infixal o shows there were no vowels in the slots
> between consonants preceding the full-grade vowel.
> You appear to accept this, for you have not reacted
> to the points where I brought it in.
Erh, more like the *a (not *o) was infixed into the
first, MOST OPTIMAL slot in a root in order to repair
the phonotactics as best as possible. If *stex-
derives from *astex-, we have an unstressed preSyncope
eLIE *'st'x- that starts some hypothetical derivative
with an accented suffix. The first available slot,
from left to right, that is within the root is where
*e is in the guna form. The infixing should result
in a pattern of (C)CoCC and not (C)CCoC.
= gLeN