----- Original Message -----
Sent: Sunday, May 22, 2005 8:49 AM
Subject: Re: [tied] IE *de:(y)-
'bind'.
> ***
> Patrick wrote:
>
> I have explained this before but I will explain it
again.
>
> I do not believe that laryngeals had any
coloring abilities.
Jens:
But we very definitely observe that they did.
***
Patrick writes:
Some of my views on this subject can be found at:
***
Patrick wrote:
> The different long vowels (at least, originally
long) that one
sees in conjunction with laryngeals in PIE is, in fact, the
Nostratic original vowel, preserved in its quality by being long.
***
Jens:
But this has ceased to be a serious possibility long ago. We
know
that very many of the long vowels of the attested languages are
laryngeal-based and contain a segmental reflex of the laryngeal. We
know that an underlying short /e/ was coloured by adjacent
laryngeals
in PIE, whereas an originally long /e:/ was not.
***
Patrick writes:
What they contain of the laryngeal is length.
Coloring laryngeals are an imaginative fabrication for PIE. They were
made out of whole cloth to explain PIE *a:, *e:, and *o:.
I can predict the quality of the vowel in any PIE *CV: root by the
semantics of the *CV syllable. Thus, *da:-, 'liquid, disintegrate', can only
be *da:-, not *de:- ('whirling') nor *do:- ('lump, pile').
Why not explain why laryngeals can color short vowels but not long
vowels?
If you see *e: in the root, you reconstruct *H1; if *a:, then *H2; if o:,
then *H3. The determination of the laryngeal is totally circular.
***
> Only short vowels suffered the Ablaut
reduction in PIE.
No, also underlyingly long vowels are reduced (to short full vowels)
when preceding the accent (there don't seem to be any that
originally
followed the accent). The laryngeal sequences VH of
course just lose the
vowel and have zero-grade H.
***
Patrick writes:
Sorry about my imprecision above.
I meant to indicate that only short vowels became subject to the *e/*o
oscillation.
***
Patrick wrote:
> Before you write how uninteresting
that is, Professor Lehmann
was told of the idea, and expressed his
interest in it, suggesting I
write a paper on it; and he would assist me
in placing it. Shortly
thereafter, his wife died, and I have not had the
heart to burden
him with more than he already has as a burden
***
You have not succeeded in showing me any good points about the idea.
It is not enough for me that somebody else has been
interested.
***
Patrick writes:
Try reading the short essay linked above.
But the proof of the pudding is that you cannot predict which color a
laryngeal will have in any given root whether you know the semantics of the
root or not.
I can --- in most cases.
***
> I thought you were in the picture on this when you
saw me
indicating laryngeals by *H and showing the vowel I believe was
preserved from pre-PIE. The only thing of which I am not certain is
if
there were two laryngeals: /?/ and /h/.
> ***.
I figured you might be operating on that assumption, but knowing how
wrong it is I could not really believe it. Now I know worse.
***
Patrick writes:
Into every life some rain must fall.
***
<snip>
> > Not really, but there are *no* formal
restrictions
concerning
> root
>
> structure that would make a root predictably durative or
> punctual.
> >
>
> ***
> > Patrick
writes:
> >
> > We both
know that present theory allows *-eye to form
durative
> stems of *CVC roots.
>
> No,
that forms causatives, and from all kinds of roots.
>
> ***
> Patrick writes:
>
> According to Beekes, that is not correct. On page 229,
item c.
of Stem Formation, Suffix *ei/i. For example *mn-ei-, and
thematized *spék^-ie-.
Jens:
"*mn-ei-" is a wrong analysis. There is no allomorph "-ei-". The
analysis is based on the Slavic stative se^d-e^-ti, prs. se^z^doN,
se^d-i-tU; Lith. se:d-é:-ti, sé:dz^iu, s:éd-i 'be seated'. That is
the
IE stative in *-eH1-, prs. *-H1-yé/ó-. The Slavic full -i- is
not from an
*-ei- of this type, but from the other i-verb types
where it arose put of
either *-éy-e- of causative-iteratives or *-e-
yé- of denominatives. There
are simply fewer types in Slavic than in
its prestages, so there has been
some amount of reduction and
analogical standardization.
***
Patrick writes:
You are contradicting yourself on the same screen shot.
Above, I posit a durative in *eye; and you say no, that forms
causatives.
Then, in the paragraph imemdiately above, you show stative in *eH1-,
present (that is durative in my book) as *H1-yé/ó. What is your *yé/ó if not
my durative *eye?
And now I notice you have slipped in iterative. Iterative, does not mean
'doing several times', almost like durative?
***
***
Patrick wrote:
> Next Beekes mentions your causative in
*-ei-e- but, in almost
the same breath, "-ei-e is also used to form
non-causatives with
zero-grade in the root: PIE *luk-éie, "to shine"". On
page 230, he
lists several more uses for *-ie- that are not
causative.
> ***
Jens:
You don't say? Well, I vaguely remember when I wrote about that.
These two types have been demonstrated to be but one original type.
There is a regular phonotactic alternation between -o- and zero in
the
root vocalism of the non-causative *-éye- formation.
The non-
causatives do not only have zero vocalism as Beekes may
be read to
say, but also -o-, but only with certain specific root
structures.
One example is Vedic rucáyate 'shines' from *luk-éye-tor;
another is
dha:r-áya-ti 'holds, maintains' from *dhor-éye-ti.
The thing is that
the root structure *lewk- did not contain (in my
view: did not keep)
the -o- which appears in the root structure *dher-.
Both are in
origin causative, and the stems *l(o)wk-éye- and *dhor-éye-
did
mean 'make shine' and 'make hold' resp.,
but the use of the words in
the middle voice neutralized the
causative meaning, so 'is made
shine' became just a way to say 'shines'.
So it was originally the
middle dha:ráyate that meant 'is made hold, i.e.
holds', but since
the middle voice meaning was also lost in this process
the natural
form became the active dha:ráyati.
Besides the alternating o/zero-
type without the causative meaning
there was a renewed form with
constant -o- regardless of root structure
which was used when the
causative was really meant, thus Vedic rocáyati
'causes to shine'
from *lowk-éye-ti. That must reflect a pre-PIE
analogical renewal
of -o- to be used for real, i.e. pragmatic, causatives,
while the
old form with -o- in some root structures, zero in others, lived
on
in the relic forms that had lost the causative meaning. The
complementary distribution of -a(:)- and zero (reflecting IE *-o-
and
zero) was established by Stephanie Jamison for the Vedic aya-
verbs in
1983. The explanation via the middle voice was made by
Delbrück as early
as 1888. I combined these observations and
expanded the -o-/zero analysis
to the entirety of IE in my doctoral
thesis of 1989. So, in a way I have
heard about this already.
***
Patrick writes:
You can certainly assert that *leuk- should have had an *o in *luk-éye-,
but asserting it is not proving it.
You have not proved that *luk-éye- was originally causative, only
asserted it.
In my interperation of the middle voice, the translation would _not_ be
'is made hold' (which, by the way, is really not acceptable English for 'is
made to hold') but rather 'causes one's self to hold', active rather than your
shift to passive. Though it is not something one says every day, there is no
reason to believe that the meaning of the middle voice inflection here was
lost.
Then you write: "when the causative was really meant". What precisely do
you think was "meant" when they used the, according to you, causative
inflection in the earlier example??? Gosh, I thought it might have been the
causative.
Let's see, now you write that when the causative was really, truly,
cross-my-heart meant, the normally expected form with *o in the root is used.
Where is your proof that it was not _always_ used, and so the form *luk-éye-
could _not_ have been a causative.
***
<snip>
> ***
> Patrick writes:
>
> Do you mean "obtuse"?
>
> If so,
it is because I have explained my reasoning behind this
before, and at
length.
> ***
No, O obtuse one, I meant the impression that "adding *y [to a root
of the structure *CV] would produce a durative stem of the form
*CVy-
". What is that funny statement based on? You seem very sure about
it, so it ought to be easy for you to write a list of impeccable
examples that are above any suspicion of being based on mere wishful
thinking. I'd like to see that, there must exist such a list if you
have any basis for what you are saying.
***
Patrick writes:
Who is the obtusor and who the obtusee? I have lost track.
If making a list of impeccable examples was so easy, it would already
have been done by someone if not me.
The oldest and best preserved language we have that is related to PIE is
Sumerian. Its durative is called the mâru-conjugation, in which the inflection
for the third person singular of transitive verbs is -e, which represents
Nostratic *-y. It is added to *CV and *CVC roots alike.
Nostratic *y has a limited number of uses: differentiation (motion) and
durative.
If it can be showed to be used with PIE *CVC verbal roots as a durative,
which it can, then presumably, it should also have been used in pre-Nostratic
for *CV roots since we know from Sumerian that any given verbal root, without
-e or reduplication, is punctual.
When PIE adopted the *CVC root as the standard format, previously formed
verbs with durative stems like *CVy- from *CV were neutralized as to
aspect.
In every PIE *CVC verbal root, the final *C has some meaning. It was not
just _arbitrarily_ added to create a new morpheme. In most cases, we can
identify the meaning of the second consonant, and understand the composition
as a semantic whole.
***
>
> > In this, the linguistic
sign is just arbitrary. What *is*
your
>
> point? If you are just finding occasion to say - for reasons
I
> don't
> >
understand - that short roots ending in /y/ are not
necessarily
> > durative, you are of course right, but
why in heaven's name
> should
>
> they be?
>
>
***
> Patrick writes:
>
> Sorry we
live in two different worlds. I do not believe in
anything "arbitrary" AT
ALL. Every effect has a cause.
> ***
We simply have to
accept "l'arbitraire du signe linguistique" for
most purposes. We cannot
reconstruct ad infinitum.
***
Patrick writes:
The definition of insanity is activity without rational purpose.
I think the idea of arbitrary signs is a confession of
incompetence.
***
<snip>
> And then what happens? How far above? Are you referring
to your
> statement, "some roots were simply *CV so that
adding *y would
> produce a durative stem of the form
*CVy-", meaning that, by
> substitution, "some roots were
simply *CV so that adding *H
would
> produce a stative
stem of the form *CVH-"?
>
>
>
***
> Patrick writes:
>
> I thing
he's got it!
> ***
So I have? But that's nonsense too! Therefore a last chance: Where
*do* you observe that there were well-motivated short roots of the
structure *CV which were expanded by *-y and thereby became
demonstrably durative, and when expanded by *-H became demonstrable
stative? Again, you have been insisting so much on this that it is
obviously a thing you feel you just know to be true, so I think it
is
only fair to ask you to let us in on a list of clear and obvious
examples
that can show us that what you say is indeed so and must be
accepted.
***
Patrick writes:
Short answer: Sumerian. See above. For stative, the -a suffix in
Sumerian which I discussed some time ago.
***
> What roots do you have in
> mind?
Surely the three classical roots *dheH1- 'put', *deH3-
'give',
> *steH2- 'take a stand' are not stative, for they all form
root
> aorists. The root *yeH2- 'go (by horse or vehicle)'
is durative,
but
> not stative in any common use of
this term, for its laryngeal is
> also present in
non-durative forms of it, as the s-aorist *yé:H2-
s-
>
/*yéH2-s-. Where do you really have material that allows you to
> segment the root-final laryngeal off and ascribe a
stative force
to
> it? Are you doing an empirical study
at all?
>
> ***
> Patrick
writes:
>
> Jens, we have altready been through this
once.
>
> We already looked at *CVy- and determined
it was as liable to
form a root aorist as a root present.
And is a
root *aorist* a "durative verb" all of a sudden? Sure, if
you permit
yourself to use any technical term also about the
opposite of what it
was made to mean you can say anything and win
any debate, but then without
gaining any genuine insight.
***
Patrick writes:
You know very well (or you should) that I have never characterized an
aorist as a durative.
I agree with your second statement. I do not do it, however.
***
> I extrapolate that the same is true of
*CVH-.
>
> *CVy- and *CVH- were lexicalized, and the
original formants of
durative and stative were no longer felt as
such.
I'd say they weren't, for what we observe is sometimes the
opposite.
You may be right 50% of the time, but that will be so with
anything
which you give only two possibilities. This just goes to show
that
the root structure has nothing to do with the aspect properties of
the roots.
***
Patrick writes:
Thank God a dead horse does not feel pain.
***
> It is in Egyptian and Sumerian that we
have actual CV roots; and
this means that, at an earlier stage
(pre-Nostratic), PIE is the
recipient of *CV roots + formants like *y, *H,
*t, and *w.
Only if the elucidation of the connection between Egyptian
and
Sumerian on one hand and IE on the other is so clear and safe that
such inferences are possible. I do not think that is anywhere near
the
actual state of affairs.
***
Patrick writes:
The connections between PIE and Sumeran and Egyptian/Arabic/Hebrew are so
clear that the only good reason for not acknowledging them would be to
practice anti-Semitism.
***
> We know these formants are used after the later *CVC
roots.
> ***
That does not make the phonemes they contain old markers of the same
when they occur in other places. None of the s's of English <boss>
is a genitive marker, none is a plural marker, none is the
third-
person present marker. What *is* the kind of logic you are operating
by?
***
Patrick writes:
Ockham's Razor.
***
<snip>
> > ***
>
> Patrick wrote:
> >
>
> Whether *CVy- and *CVH- were still felt as duratives and
> statives in PIE (I guess we decided they probably were
not), I
> claimed that in Nostratic they were: the product
of a *CV root +
> stem formant. This was their
origin.
> > ***
>
>
Why make such a claim? And what is its relevance for the
analysis of
> forms of IE individual languages?
>
> ***
> Patrick wrote:
>
> If, after being lexcalized as regular *CVC- roots, none
of the
former force of the formants was felt (and I cannot, at present,
prove that it was), then these originally CV+f. roots, now *CVC- are
neutral.
> ***
What is your basis for calling them "CV+f. roots" in the first
place? If you can't see any functional justification for such an
analysis, why do you make it at all? And why do you insist upon it?
And why do you give it priority over analyses operating within the
empirical scope of direct observations we *can* handle? I ought
perhaps to ask what the +f. means but I don't bother, I just take it
that you put in a morpheme boundary, and I'm asking you to motivate
it, indeed prove it.
***
Patrick writes:
The general principle that simple precedes comples. And the existence of
CV roots in Egyptian and Sumerian.
All descended from Nostratic; ergo, at some point, the ancestor of PIE
had CV verbal roots.
***
<snip>
> > ***
>
> Patrick writes:
> >
>
> I do not know what material you collected but this pattern
is
> common enough --- if you can see it.
>
>
> > For example: IE *ge:y- (*geHy-)
+ -*to yield Old Indian
> gi:tá, 'sung'; _exactly
parallel.
>
> No, give me one with following
-n-.
>
> ***
> Patrick
writes:
>
> Middle Persian bri:n-, 'determined', from
*bhreHi-.
> ***
That's correct (in principle, if not assuredly so for this particular
root), but that's after a sonant. The zero-grade sequence *CRHy- was
regularly metathesized to *CriH- in IE (just as *CRHw- became *CRuH-
).
That is chapter 2 on the "ERu/Ru:" problem of my 1989 book. So we
do have
Ved. pri:tá- 'loved' from *priH2-tó- by metathesis from
earlier
*prH2y-tó-, zero-grade of *preH2y- 'love'.
***
Patrick writes:
I mistrust introducing metathesis unless there is absolutely no other
alternative. Here, it is unnecessary. *prH(2)y-tó- should also yield pri:tá
(*6(2) -> i; y -> i; i + i = i:).
In di:ná- there
is no sonant preceding, so I do not see how
that can regularly
reflect an alternant of a root of the structure *deHy-.
It will
however be quite regular from a root of the shape *deyH-. The
etymological assignment of di:ná- is completely unsettled, so why
insist on a derivation that only makes a problem of it?
***
Patrick writes:
di:ná from *daHy- makes perfect semantic sense.
***
> The picture is different for -t-
> and -n-. Non-Aryan evidence shows *-&-to- where Indic
has
either -i-
> ta- or -i:-ta-; with following sonant
the form is consistently
with -
> i-. The Indic variant
-i:-ta (not found in Iranian) can be
explained
> from
*-&-to- also, i.e. as identical with -i-ta-, by simple
> levelling, in that -i-ta- has been adjusted to full-grade
alternants
> with -a:- taking over its length. There
are four forms in -i:ta-
:
> dhi:tá- 'sucked', gi:tá-
'sung', pi:tá- 'drunk' (but Gk. potós),
> sphi:tá-
'thrived'; but ditá- 'bound' (Gk. detós), diná-
'divided'
> (dití- 'division', Gk. datéomai 'divide'), ON gin
'abyss', Gk.
> khatéo: 'want, be without', s'itá- (Lat.
catus), Lat. po-
> situs 'placed' (and sinus 'bay'), Lat.
satus 'sowed', OS
> sad 'satiated', sitá- 'bound', chitá-
'cut off'. These reflect
*dh&1-
> tó-, *g&2-tó-,
*p&3-tó-, *sph&1-tó-, *d&1-tó-, *dH2i-nó- (*d&2-
tí-),
> *g^hH1i-nó-, *gh&1-tó-, *k^&3-tó-, *s&1-tó-
(*sH1-nó-) *s&1-tó-,
> *s&2-tó-, *s&2-tó-,
*sk^&2-tó-. The roots are *dheH1y- 'suck',
> *geH2y-
'sing', *peH3y- 'drink', *speH1y- 'thrive', *deH1y-
'bind',
> *deH2y- 'divide', *g^heH1y- 'yawn', *gheH1y- 'lack',
*k^eH3y-
> 'hone', *seH1y- 'lag behind', *seH1y-
'sow', *seH2y- 'bind',
*seH2y-
> 'satiate',
*sk^eH2y- 'cut off'. There are no examples with -i:-
no-.
> Other derivatives also showed *-&-t-, but *-(H)i-
before
sonants.
>
>
>
> Surely you noticed this word in your
survey.
> > ***
What word? di:ná-? Not really, for it is not assignable to any known
root. It is not usefull evidence.
***
Patrick writes:
I disagree.
***
<snip>
> ***
> Patrick writes:
>
> Since I wrote above that the root-vowel becomes Ø, you
can be
assured I understand it.
> ***
But then
you obviously don't understand yourself. If the root vowel
is lost, and
laryngeals are shown to be present only by their
lengthening effect on
preceding vowels, how can you insist that you
can see whether the
laryngeal is present or not in a form that
begins with either *di- or
*dHi-? I have gone to great lengths to
try and make some sense of your
many categorical statements, but I
really fail here.
***
Patrick writes:
You must be a wonderful cook to be able to mix so many things
together.
In full-grade, laryngeals are detectable by lengthening of the
root-vowel; in zero-grade, by whatever the appropriate reflexes of *6 are for
the particular language in question; for Old Idnian, -i-.
Patrick
***