Re: [tied] IE genitive

On Sat, 26 Apr 2003, Glen Gordon wrote:

>
> I will have to forego commenting directly to your quotes and to try to
> condense my rebuttal into something to-the-point. I found myself just
> dismissing half of the post because it contained no valuable rebuttals,
> just rhetoric and opposition for opposition's sake. I am categorizing
> my rebuttal into clear issues.

Then I'll make an effort and see if I can follow.

> METHODOLOGY
> -----------
> First of all, we seem to have a problem with my good buddy "Ockey".

You're in for some serious lacerations the way you keep provoking the
razorslinger by misrepresenting his wisdom.

> When you protest, "I do not accept that it can be known that there was
> only one source of the thematic vowel," you're obviously missing the
> whole point about comparative linguistics because we must take the
> _simplest_ theory (ie: that there is ONE source) before we dare consider
> graduating to something more complex.

Agreed on the point concerning simplicity; not agreed on the point
concerning single origin. An even simpler solution would we: It never
changed, it has always been this way. That principle would indeed simplify
scholarly literature by doing away with all necessity to write and read
any.


We cannot "know" anything in comparative linguistics, so we must

> accept the most economical theory, even if our imagination shows us fun
> alternative hypotheses.

Agreed.

> And I can see that you have much imagination.
>
> This is how I always operate: If I reject something, it is because I fail
> to see how it is efficient enough to bother considering. I must refuse to be
> dragged into endless what-if scenarios with you that multiply hypotheses
> simply for the sake of finding an alternative hypothesis no matter how
> absurd.
> That's not constructive reasoning and a waste of time. I don't deny there
> are
> more absurd hypotheses than mine, but I question whether there are more
> efficient ones.
>
> You say that you "believe" that thematic vowels "survived ablaut
> unimpaired",
> despite the fact that doing so automatically forces us to unnecessarily
> assume
> how it might of survived, necessitating further assumptions.

It can't be helped. I am forced to that conclusion if I want to understand
anything, since nothing else works. There was certainly no help on your
websites.

> Yet while "they must do that for a reason", as you say, "that has not
> been found". So added to your blatantly inefficient hypothesis you
> strangely undermine it yourself anyways. I don't see ordered thought in
> this respect and I think Ockham would spank you.

For turning a critical attitude towards myself? I'm sorry, but I'm sure
that's going to happen again, probably many times.

>
>
> THE NOMINAL THEMATIC VOWEL (AS OPPOSED TO THE VERBAL ONE)
> ---------------------------------------------------------
> You probably encounter "very grave difficulties" with my theory that
> nominal thematic vowels originate from genitival constructs more because
> you have failed to fully understand it. I notice a long list of
> misunderstandings on your part and these misunderstandings are
> frustrating for me and this debate. This is the immediate source of your
> grave difficulties.
>
> The thematic vowel *& is not the same vowel as in the genitive singular
> *-as since the accented genitive singular regularly becomes *-os WITHOUT
> *e/*o ablaut alternations.

But the (nominal) thematic vowel arose out of genitives reinterpreted as
nominatives, eh? Then what IS the relation between the thematic vowel (TV)
and the vowel of the genitive ending if any? In my understanding of your
theory they are identical; in my understanding of the matter, they have
nothing to do with each other.

> Secondly, there is no "ablaut" per se of genitive *-os to *-s. Rather
> *-os and *-s represent two paradigmatic variants that arose from the
> moment of syncope in Mid IE.

I'd say that is exactly what ablaut is all about: loss of an unaccented
short vowel changes unaccented /-os/ to /-s/.

> Stems originally in accented *-as first retracted their accent to the
> initial in the nominative (BEFORE Acrostatic Regularization took place)
> resulting in *-&s, not *-s, because the latter would result in a twisted
> paradigm (eg: nom **CVC-s but gen **CVCe-s !!?).

Supposing the prerequisite to be true, what would have been unacceptable
about such an inflection? It looks a bit like Gothic dags, dagis, or Latin
urbs, urbis. Or are you trying to say something entirely different which I
cannot make out?

> So instead we see *CVC&-s with gen. *CVCe-s before Acrostatic
> Regularization served to repair these stems containing two possible
> accentuations by imposing a single non-alternating accent on the stem.
> _That_ is my position.

What is all this based on? The position taken is not argued at all. You
are just issuing a decree. I need to see some legislation before I feel
forced to obey it.

>
> The genitive is SUPPOSED to show thematic *e since *s is unvoiced and
> could not have lengthened *&. In fact, it's the nominative that's
> unexpected -

Agreed.

> - but it's clear that the strange lengthening of schwa (> *o)
> here mimicks the nominative lengthening seen in athematic stems -

No, that's not at all clear. It could be true by sheer chance, but there
is nothing that compells us to view it exactly that way.

> - that was in turn caused by the compensatory lengthening after Clipping
> (*-s& > *-s + comp.length of previous vowel).

That is not a thing we can know. It's far from the only possibility. I
don't even see by what standard it could be called the simplest one.

> This same lengthening is also seen in the sigmatic aorist.

I have actually suggested the same. But is the decisive factor common to
the nominative and the s-aorist really loss of a following vowel and not
presence of the sibilant morpheme?

>
> Please understand that the nominal thematic vowel has nothing to do with
> the verbal thematic vowel other than the same use of *&, so verbal
> thematic vowels are not caused by genitival constructs!

That's a point where I believe Ockham would be on my side. It would seem
more economical to account for thematic nouns and verbs by the same set of
rules instead of keeping their stories completely separate.

> As if the above confusions aren't bad enough, you are also confusing
> schwas of different stages, confusing MIE and eLIE *& together as if
> they were interchangeable entities -- MIE *p&t:as& becomes eLIE *pet:as
> with the unstressed stem now containing *e, **never** schwa. MIE *& is
> mostly eradicated except in special circumstances that I've already
> specified and often shifted to *e.

I suppose we basically agree on this point.

>
>
> THE "ALTERNATE POSSIBILITY" (BEAM ME UP, SCOTTY)
> ------------------------------------------------
> The alternate possibility of the thematic vowel being derived from a
> third person possessive with "Gruppenflexion-like" collocation of an
> uninflected agent noun followed by a possessive is just ludicrous and by
> far inefficient for the obvious fact of having to assume truth to
> three(!) baseless ideas instead of one.

Your theory is not one idea, but a whole sequence of completely unfounded
events of wishful thinking. My suggestion actually accepts the language as
it is: if such a structure is meant, nothing needs to be changed, for the
form is already there. What is ludicrous about zero complexity? Still it
does not have to be right, the truth is often more complicated than it
looks.

> I must reiterate that thematic vowels do not show any particular function
> or meaning because its usage is too varied and widespread. Their morphemic
> status, in other words, is left to be proven by you.

Even if the TV never meant anything - and it frequently doesn't - we would
have to ascribe morphemic status to it unless it could be revealed to be
completely automatic, i.e. the effect of overriding phonotactic
constraints. That is certainly not the case. At least there is plus/minus
TV in the subjunctive vs. the injunctive/indicative, and in many
adjectival derivatives vis-a-vis their base-words.

> Your idea of all derivatives using thematic vowels as being "modifiers"
> of sorts is rejected because it doesn't make immediate sense.

Take your time, and the wisdom of it will dawn on you. Adjectives are
modifiers, so are subjunctives.

> It is more logical to just _not_ attribute meaning to something that
> doesn't have morphemic function anyways! Tell me why I must assume that
> the thematic vowel is a **seperate** entity and **does** have a
> function!
>
> Also, I don't see how *-ro- and *-to- have a thematic vowel and why this
> has been mentioned here.

It alternates this way, and it occurs in stem-final, just like all other
TVs. Both form adjectives. What IS the status of the vowel of *-ro- and
*-to-?

>
>
> REGARDING *-ske- AND ITS NEWNESS
> --------------------------------
> As I said, what disqualifies *-ske- as ancient is the unusual accentuation
> and the unusual non-CVC syllabic shape for an accented morpheme. Further,
> it seems safe to say that the suffix is comprised of the aorist *-s- and
> intensive *-g-, so it probably _was_ originally an aorist.

There are not very many mutually unrelated suffixes in IE, so there is no
sufficient basis on which to disqualify one that does not look pretty. I
agree that the morpheme of the s-aorist is contained in the *-sk^e/o-.
This is patently the present type that originally belonged to the
s-aorist. This is shown by pairs such as Ved. prcchati/apra:ksam,
yacchati/aya:msam, Lat. pascor/Hitt. pahs-, Lat. cogno:sco:/Hitt. ganeszi,
Ved. icchati/Lith. iïes^kau (with acute reflecting s-aorist lengthened
grade), and some more. It is also seen in the two forms of the inchoative
stative (ingressive) *-eH1-sk^e/o-/*-eH1-s- (Lat. sene:sco:/Hitt.
marses-zi 'get forged'). Other pairings of prs./aor. exhibit the
n-prs./root-aor. or red.pr./root-aor. types, or the prs. is formed from
the aor. by the addition of *-ye/o- (e/o mostly accented, if based on a
derived stem apparently always). That invites the analysis of the morpheme
*-sk^e/o- as consisting of the morpheme of the s-aorist followed by the
present (i.e. durative) morpheme otherwise known as *-ye/o-. I do have a
parallel case, but I do not insist that it can be proved, it just is what
functional analysis tells us. I do insist, however, that the formation in
*-sk^e/o- is a present-aspect, not an aorist formation. The function of
the sk-verbs varies from branch to branch, being either iterative or
inchoative. In the ingressive it is patently inchoative, so it would be
nice if the iterative value could be somehow differently explained. I
think it can, for the type is so frequently reduplicated that the
iterative meaning may have been properly signalled not by *-sk^e/o-, but
by the reduplication. Then it was the beginning of an action that was
expressed by the *-s-, and the *-sk^e/o- type is then in turn the
corresponding durative inchoative meaning "am beginning to ..". The type
has retained its inchoative value in Italic and Baltic (the sta-verbs).
Since the aorist expresses the action that causes a new situation, an
inchoative note was not inappropriate for the default aorist that became
so productive all over the place that its original locus can only be
determined by means of remains that do not fit synchronic patterns. I do
not know of an IE intensive type marked by suffixal *-g-.

> The aorist was _not_ the past tense -- It merely could have a past
> nuance if anything.

Am I being told this because I let you in on the general explanation of
the tudaïti type as a displaced aorist? Well, the verb forms a nasal
present, so we expect a root aorist, just as we do to go with Ved.
vindaïti, Av. vinasti, where we find aïvidat. Both tuda- and vida- are
rather plain thematicizations of root aorists alternating *teud-/tud-,
*weid-/wid-. Now, how could one of these end up in the present
system? Obviously past is the path between them. As you say, tense has
nothing to do with the stem marking, and that's precisely the reason one
could talk about the past in different ways. When past was meant, that
could be reported from different points of view (aspectu:s), so that the
very same act could be regarded as ongoing in the past (imperfect) or as
simply a fact to be reported (aorist) - and WHEN THAT DID NOT MATTER
(which must have been the case very often) THE TWO WERE SYNONYMOUS. That
must be the reason we see so many verbal stems slipping from one aspect to
the other.

> If I understand right, the aorist marked a momentaneous non-punctual
> action (an action with a beginning and end that was not abruptly
> executed).

That is not the issue here, but the aorist is typically a statement of a
past event; it is the typical form to be used when a momentaneous event
interrupts a longer process. It is very often referred to as punctual.

> Repetitive actions, which were marked with *-ske-, would have
> been momentaneous and non-punctual. This then explains the aorist-like
> accentuation for *-ske-.

I would ascribe that function to the reduplication of forms like Gk.
gigno:ïsko:, Lat. disco: (*di-dk^-sk^-o:).

>
> Now, is the accent truely on *-ye- though? Do we find *?no:mn-ye- 'to
> name' with accented *-ye-?

Denominatives accent the suffix -yaï- in Vedic. We do not find the verb
you mention with any accent. Slavic denominatives of the normal type
zvoniti 'sound a bell' have acute -i- pointing to *-e-yeï- with accent on
the second part. These thematic vowels were accented. If accented thematic
vowels in present formations are incompatible with preconceived ideas
largely based on Etruscan I am not shocked.

> THE SOURCE OF THEMATIC QUALITATIVE ABLAUT
> -----------------------------------------
> While I see clear examples of lengthening before voiced segments in
> English, you prefer to use a smoke-and-mirrors theory by Hirt using the
> ol' magical F2 to cover up the lack of real-world examples for this
> phenomenon. Logically, I must side with lengthening as the preferred
> hypothesis because this is a trivial phenomenon.

I am not disputing this phenomenon for English, but languages can be very
different. And in fact the English lengthening is not restricted to
stem-final position, but works everywhere. That is VERY different from the
IE facts. Hey, don't give all the blame to Hirt, I combined some of the
things myself. We do not differ all that much on this point which only
concerns a detail. You are actually advocating things I wrote many years
ago when I had just read about "natural phonology" and realized that /o/
is inherently more sonorous than /e/, so that the change of e to o induced
by following voicing could be explained as an event of assimilation in
which a quantum of sonority rubbed off on the vowel. But since this is the
same language which reduces e to o when unaccented (on its way to zero),
then it would be preferable to comply with Ockey's command and have
related rules cause the two very similar changes.

>
> ON MY ORDERED NAMING OF LANGUAGE CHANGES FOR PRE-IE
> ---------------------------------------------------
> Your protest against my "crude" naming of stages goes unheard. I need to
> give names to these language changes so that I don't need to constantly
> explain the changes over and over every time I invoke them. Imagine
> having to explain Grimm's Law each and every time we needed to use it to
> explain something in Germanic! Forget it, Jens. It suffices to say "Grimm's
> Law" and if you don't know what that means, you must look it up yourself.
> I already have supplied my list of changes, their names, and descriptions
> on my site.

Well, Grimm's Law does mean something meaningful. I have indeed looked up
your designations, but they are even cruder and less motivated than I
feared. A visit to the website leaves one with the impression that the
siteowner is of opinion that Indo-European is a distant dialect of
Etruscan. At any rate there are hardly any changes from
Proto-Indo-Tyrrhenian to Etruscan worth talking about, while IE has been
thoroughly recast. Thus, for all intents and purposes, IE can be derived
from Etruscan. Now, basing one's impression of the prehistory of
Indo-European on Etruscan is like forgetting about all the other IE
languages and taking only Gaulish in an attempt to link IE with Uralic. I
am sure there are things that clinch, but that would be more due to some
solid knowledge coming from Uralic studies than to evidence from Gaulish
which would only be decisive on very rare occasions. By this I'm saying
that the priorities are wrong if evidence from IE - which is a vast field
of study - is being downplayed simply in order to accord a leading role to
the meagre and insecure facts of a language as imperfectly known as
Etruscan.

>
> http://glen_gordon.tripod.com/LANGUAGE/NOSTRATIC/STEPPE/indoeuropean_rules.html
>
> We've also discussed the alternate url using...
>
> http://members.tripod.com/~glen_gordon/LANGUAGE/
>
> I must rename Paradigmatic Strengthening as Paradigmatic Resistance and
> also explain Suffix Resistance. Maybe I'll put them right under the
> Penultimate Accent Rule.

Instead of naming these suggestions you should try and motivate them.
There is not the slightest attempt at doing that in any proper manner on
any of the webpages referred to in the above.

Jens