[tied] Re: Monovocalism: sequel

From: elmeras2000
Message: 33388
Date: 2004-07-05

--- In cybalist@yahoogroups.com, enlil@... wrote:
> On the subjunctive, Jens:
> > But it should also have a motivation in the subjunctive; and the
> > positing of two separate structures, one thematic "*bhér-e-ti",
the
> > other subjunctive "*H1éy-H1e-ti" to go by your instructions,
should
> > also have a motivation.
>
> There is no such motivation because it cannot be shown clearly that
> the thematic vowel is a MORPHEME. It has no succinct usage that
explains
> it's wide and varied use in all word categories. Rather, the
WYSIWYG
> solution is the simplest here: The thematic is a redundant vowel.
Duh!

I fail to grasp the full argumentative depth of the last word. You
seem to be disregarding that languages can change. Empty morphs can
easily have arisen out of something that earlier had a function.
Now, the functional impact of the subjunctive is not enormous. It
has not survived in Anatolian, Balto-Slavic and Germanic. It has
become a future in Latin, a fate shared by the present in Slavic.
The Tocharian subjunctive does not seem to continue the IE
subjunctive in any of its varieties, it rather looks like something
as odd as the perfect injunctive. All in all, the functional
distance between the subjunctive and the present indicative does not
seem to have been very great. Therefore, if we find that one type of
present marking has an empty morpheme which looks exactly the same
as the marking of the subjunctive, the only natural inference would
be, I guess, that it *is* the form of the subjunctive which has lost
its independent function.

>
> The 'subjunctive' has been given a name for the very fact that it
has a
> clear function. More duh!

Again this special argument is above me. As for the clarity of the
subjunctive, it may be faor to say that even in its clearest usages
the subjunctive has a strong affinity with the present indicative.
At the very least both are regularly used about the imminent future.


> Any analysis of the subjunctive forms at the
> very least shows a morpheme *-e- added to both athematic and
thematic
> stems, as you admit.

I sure do admit this, oh yes.

> However a single vowel void of consonant does not
> conform to known IE phonotactic restraints on morphemes.

It can only have this form if it is allowed. A suffixal morpheme
consisting of a single vowel would also be stem-*final* and would
therefore be the thematic vowel. There are desinences consisting
only of vocalic material, why no suffixes?

> All other morphemes in IE have at the very least one _consonant_
as a bare
> minimum (such as the nominative *-s, aorist *-s-, 3ps *-t and
indicative
> *-i).

Yes, for otherwise they would not be "other". It may be that simple.

> While the thematic vowel cannot be called a morpheme, the
> subjunctive clearly is and must conform to these syllable rules.
Therefore
> there is something wrong with a plain ol' subjunctive **-e- by that
> criterion.
>
> If we are to obey the rules of IE, either we say that the
subjunctive is
> very very very very different from everything else in IE or we
cave in to
> the bombardment of common sense. We conclude that the subjunctive
is
> *-he- (aka *-H1e-) because only this can satisfy the resultant *-e-
in
> daughter languages while satisfying the phonotactics at the same
time.
> So while *h is necessary for the subjunctive morpheme, it is
pointless
> by Occam's Razor for the thematic vowel because it's not by any
definition
> a morpheme. The mere possibility of something doesn't make it
right, as
> we all know.


Yeah, right, it *is* "very very very very different from everything
else" because there can even theoretically only be *one* item of
this shape. And, surprise, that single item is then unique. Is this
the correct place to say duh?

> Only *a is affected by Schwa Diffusion afterall and
> only this shows later *e/*o alternation... but this would force
one to
> admit to TWO vowels, *e and *a, in any prestages of IE.

That sounds like a serious weakness, especially when heaped on top
of the unfounded idea of Schwa Diffusion.

> That is then sufficient amongst the other things I've mentioned to
show
> that my solution is most optimal.

It does not show *me* that. Does anyone else feel they have been
shown this, and can they explain to me what this is based on?

> You say of *bHibHerti that "the ending consists of a consonant only
> and so does not move the accent" but this rule still doesn't
explain
> anything. Why should the accent alternate at all, why should it
matter
> what the syllabic shape of the ending is, and why should there be
so
> many unintuitive paradigmatic accent patterns to choose from? This
just
> dusts off your responsibility to explain what we find in IE.

> QAR addresses all that you ignore. Accent shifts in this language
would
> be based on syllable number from the end, that is, only if the
> proterodynamic and hysterodynamic could somehow be unified. QAR
and the
> acknowledgement of the already proven Syncope in even _final_
positions
> unifies both of these accent patterns and gives a simple reason
for the
> learned accentuation: an underlying _automatic_ quasi-penultimate
accent
> based only on syllable-number.

I would like there to be some such rule, but I have yet to see any
glimpse of independent motivation for the preforms it would need to
work on. I have seen only self-serving postulation of vowels which
are subsequently used to explain things. And perhaps the solution
just is not as one would like it to be.

If we discount paradigms we just cannot analyse safely
('brother', 'hound' and 'know' are examples of this), and accept
that the paradigm types interchange according to animacy, as Ved.
ápas- 'work' : apás- 'active', it will seem that the paradigms we
really *can* analyze are all accented on the last vowel they have.
The verbs do this very plainly: *gWhén-t, *gWhn-ént, *gWHn-yéH1-t,
*gWhn-iH1-ént. Nouns perhaps do the same if they have no good reason
not to: *H2nér-m, *H2nr-ós; *pH2-tér-m, *pH2-tr-ós. If you add case
endings or personal endings without any (even underlying) vowels in
them you get a weak case or a weak verbal form, and if you add a
desinence that has a vowel of its own you get a strong inflectional
form. This clear correlation must count for something.

There are deviations, but they would not be hard to explain: *pér-tu-
s, *pr-téw-s does not accent /-tew-/ underlyingly, but that could be
because the vowel of this segment is a later insertion (as per
Szemerényi). Perhaps the suffix *-men(t)- was also originally
vowelless; that would explain the radical accent of *H2ák^-mo:n,
*H1nóH3-mn, while poimé:n 'shepherd' would have contrastive accent
on the suffix because it denotes a person. The same contrastive
accent appears to have worked on compounds. The suffix of s-stems is
still recalcitrant, but not hopeless I think.

The acrostatic paradigms are not contrary to this given the
introduction of initial accent in a period after the first ablautt
reductions. Then these paradigms come out with initial accent in all
forms, no matter where they originally had their accents.

In this account there are no *original* proterodynamic paradigms at
all, for all original stems are taken to have accented their last
vowel. When prop-vowels arise in heavy suffix structures the
proterodynamic type is created automatically.

This would, in its further perspective, answer the burning question
of why the accent is where it is before the ablaut changes take
flight. I know that is what you are trying to do with your QAR whose
foundation I just fail to see. Our objective is the same.

> My QAR simply could not accomplish such a straightforward union of
the
> hystero- and proterodynamic paradigms unless this were based in
reality.
> The rule you give above is still ad hoc, unexplained and thus
undesirable.

I'm ready to give any rule of yours due consideration, especially if
you will be kind enough to present it in an easily understandable
form. But I cannot promise to do any different to suggestions of my
own.

> > The reduplicated present type is regularly *opposed* to an aorist
> > of the underived type (root aorist), so an aorist is just about
the
> > last thing it would be.
>
> No. For MIE, I'm thinking more like a reduplicated aorist *bebér-a
> (> *bHibHérti), a simple aorist *béres-a (> *bHe:rst) and a
durative
> *bér-e-ta (> *bHéreti). I wasn't saying that the first 2 forms
would've
> meant the same thing. I was just saying that they probably
inhabited
> the same aspect at one time.

The type *dhi-dhéH1-ti is of the present aspect everywhere. It is
opposed to the root aorist.

> The former would emphasize doing something
> repeatedly or habitually while the latter would simply be an action
> regardless of how many times, if at all, one did it.

This may again be just a case of misguided terminology. There is no
point in using such a traditional label as "aorist" if one does it
in a way contrary to tradition. What you describe is the function of
a traditional present stem, and since it has forms with primary
endings there is little point in calling it an aorist.

>For example, I
> could see *beresa being used in simple prohibitive statements like
> */meh tam béresa/ "Don't carry that" whereas */meh tam
bebéra/ "Don't
> keep carrying that all the time".

Yes, that's what Hoffmann demonstrated for Vedic, following the
precedence of Ammann. It is still descriptively adequate for Vedic
and Gathic Avestan. It shows the opposition of the aspects in full
clarity: the aorist introduces a new situation, the present stays
within the situation at hand.

> > The word at hand would then mean "keep on bringing", "bringing
ever
> > more ...".
>
> Yes, then we are in alignment on the use of the aorist.

No, that is a perfect description of the present aspect. It is a
continuation of the situation already introduced by an aorist, now
being elaborated by a present.


> > The athematic reduplicated structure is specifically NOT an
aorist.
>
> Speaking of multiple events of an action without any reference to a
> clear point in time is indeed specifically aorist... Does "aorist"
> not mean "without time" in the end?

Multiple events amount to a durative in this system, which is what
the socalled present aspect is. One could imagine this differently,
and I'm sure it is different in many other languages, but IE is this
way. That of course does not exclude that it was earlier different
in this respect (as in most other respects), but we have not been
presented with any evidence to make us believe so. That would be the
minimum requirement. The woord "aorist" means "unlimited", but that
is of little help if it is a silly term.

> > The root *deH3- just formed *déH3-m 'I gave'. Why the unmotivated
> > fuss?
>
> Everything about IE forces us to go through that fuss, Jens.

Root *deH3- and 1sg.act. secondary ending *-m must belong
among "everything about IE", but they do not force me to perceive of
*déH3-m as anything but a simple concatenation of these two
morphemes. It must be some very special things about IE that make
some believe they have to derive this form from something different
from what it looks like.

> IE simply
> didn't always have zero-grading.

Relevance? On or off topic, this paradigm did: Gk. 2pl.aor. é-do-te,
2pl aor.ipv dóte.

> This logically must derive from a
> fuller preform.

No, an accented root *déH3- would remain *déH3-, and an unaccented
desinence *-m would not change, nor would it change anything else.
There is no problem here, so the "solution" is undesirable.

> We have *t:ehWa-m by all the rules I've discovered and
> it would be truely the only possible MIE form to explain later
*dehWm.

That's your own personal universe. It is something you have messed
up yourself. The form is fully regular and in need of no further
justification. If you have made up stories about other things with
which this is now in conflict, my guess will be that the stories are
wrong.

> The loss of *a is trivial but serves to explain _all_ instances of
> syllabic nasals, not just here.

What? Syllabic nasals are just positional variants of consonantal
nasals. Are you staking all this on the single attestation of the
1sg aorist injunctive /daam/ (RV 10.49.1)? There is also ádada:m and
even sthá:m with no syllabicity in the underlying nasal. Surely
these are just original sandhi variants which are used for metrical
purposes: /deH3-m/ would syllabify the nasal in pausa and before a
consonant, but leave it asyllabic before a vowel.

> Syncope shows that the accusative *-m
> was once *-am and indeed we find the fuller forms outside of IE in
Uralic
> et al.

No, syncope shows nothing where it has not occurred. You just wrote
that act into the play. What I have seen of Uralic et al. does not
force me to derive the IE acc. ending *-m from a full syllable.
Indeed IE morphophonemics tell the opposite story. That is not to
say that it may not earlier have had a vowel which in that case just
is no longer of any consequence in the analysis of IE.

> We also know that alternations like *genh-/*gnh- show that indeed
> syllabic nasals must be the product of Syncope.

No, the logical modality is only "may". There may also be other
sources, as nasals just standing in the position where they get
syllabic.

Jens