Re: [tied] IE *de:(y)- 'bind'.

--- In cybalist@yahoogroups.com, "Patrick Ryan" <proto-
language@...> wrote:

>
> ----- Original Message -----
> From: Jens Elmegaard Rasmussen<mailto:jer@...>
> To: cybalist@yahoogroups.com<mailto:cybalist@yahoogroups.com>
> Sent: Monday, May 16, 2005 9:59 AM
> Subject: Re: [tied] IE *de:(y)- 'bind'.
>
>
> --- In

cybalist@yahoogroups.com<mailto:cybalist@yahoogroups.com>, "Patrick
Ryan" <proto-language@...<mailto:proto-language@...>...> wrote:

> >
> >
> > I guess that means you do not accept Lehmann's view of the

subjunctive

> as obligative.
>
> I don't believe I know what he says. Could you summarize? May I

guess that

> he considers it an innovation since it is not found in Hittite?

And that

> he also believes it has never existed in Germanic? Just a hunch

based on

> my general prejudice.
>
> ***
> Patrick writes:
>
> Glad to fill you in.
>
> The information I mentioned is in _Proto-Indo_european Syntax_,

Winfred P. Lehmann, 1974.

>
> On pages 131-35, he discusses his views of the subjunctive,

using examples from Vedic Sanskrit and Homeric Greek.

>
> e.g. índraM náro nemádhitá havante
> yát pá:ryá yunájate dhíyas tá:H
>
> 'Men call on Indra in battle, saying:
> "You must accept our prayers as decisive"'
>
>
> He also writes: "In time the obligative meaning of the

subjunctive came to be subsidiary to its function of indicating
subordination."

>
> Lehmann considers it important that the same negative (*me:) is

used for the imperative and subjunctive in Homeric Greek.

>
> He adds: The basic obligative meaning is accordingly apparent in

the first person-subjunctives. From this use a future meaning could
readily develop.

>
> Lehmann mentions Hoffmann and Geldner sharing his view.
>
> Re: Hittite, Lehmann writes: "The subjunctive survived with

obligative, necessitative meaning notably in the first person of the
Sanskrit imperative and also of the Hittite. . .The Hittite first
singular, -l(u), may therefore be equated with Skt. -n(i).

>
> Of course, he says much more there, and in other parts of the

book.

>
> Really not so far from what you have written.
>
> He mentions the possibility of the subjunctive in Germanic, and

refers to another publication of his: "The Germanic weak preterite
endings. Language 19.313-319 (1943).

> ***

Thank you for this valuable input. It seems clear that the first
person subjunctive can express a promise: I shall do that, you can
count on it. The thirds person is rather something like 'let him -'
in main clauses and 'so that he -' in subordinate clauses. I think
all the examples given for the "obligative" use in Lehmann 1974
(which I have now seen) have a pragmatic modality that could be
expressed by the adverb "consequently".

>> Patrick writes:
>
> You mention the "inchoative" with suffix -*s-. How would you go

about "proving" inchoative. In verb tense and aspect inflection, I
know only of -*s-, which, parallel to its use in nouns, indicates
a "condition".

>
> ***

If the s-aorist is the paradigmatic companion of the sk^-present, it
will be inchoative for that reason alone. There are some cases of
this kind: Ved. pr.ccháti ápra:ks.am; yácchati áya:m.sam; Lat.
cogno:sco:, Hitt. ganeszi; Latin stative inchoative in -e:sco:,
Hitt. ingressive -es-mi; Gk. ge:rásko: 'grow old', aor. ege:ra: from
*g^e:r&2-s-m. I do not think it can be strictly proved in and of
itself, but there is an intrisic likelihood in it. If the language
came in need of an aorist, what was better suited to fill the gap
than the aorist of an old derivative verb meaning 'begin to -, take
to -'? So if the old present of *weg^h- 'drive' was replaced by the
subjunctive of the root aorist *weg^h-e-ti, the root aorist itself
needed renewal, and in there was a form *wé:g^h-s-t 'began to drive,
set out driving' would be the perfect candidate for a verbal form
meaning 'drove off'.

>
> >So I've noticed. I have also noticed that of the many languages

that go

> into creating a more and more detailed impression of the

Nostratic family

> and its protolanguage, you deal with only two, Egyptian and

Sumerian. I do

> not doubt that Egyptian, being Afroasiatic, is a distant

relative of

> Indo-European, and I have also seen quite serious-looking

attempts to

> connect Sumerian to the family, perhaps on a separate line. But

I cannot

> accept an argument based on a selection of the Nostratic

material unless

> the Nostratic situation is clear on the point concerned. And the

way to

> clear up Nostratic is to utilize the secured material of the

various

> branches, not the mysteries and mistakes, in a combined grand-

scale

> comparative analysis. My impression is not that you do that.
>
> ***
> Patrick writes:
>
> Unfortunately, your impression is correct.
>
> First, with Sumerian. No one that I know of has attempted to

reconstruct Proto-Sumerian per se. I have taken that up by default.
The "experts" all believe Sumerian to be unconnected to any other
language, ancient or modern, so a higher grouping has not been
reconstructed. Although I do not see many differences between the
Proto-Sumerian I have worked out and recorded Sumerian, there are a
few.

It would be nice to see a short list of the "obvious connections",
i.e. etymologies that are so evident that it would be silly to
dismiss them. If there are none of them, the matter is precarious. I
have had that experience with Eskimo which I studied rather
intensely some years. I found the basic grammatical marker top be
highly reminiscent of IE and Uralic, but I haven't really seen a
persuasive lexematic etymology. Are there any obvious personal
markers in Sumerian that ring a bell to me? And are there some
obvious lexematic links?

> With Egyptian, I have been forced to use it be default, also.

The Afro-Asiatic reconstructions of Ehret, and then Orel and
Stolbova have not played to sold-out houses among members of their
own specialities; I am trying to be kind.

It is my impression you are being fair. But how many tickets can
your analysis of Egyptian sell? Don't get me wrong here, I do not
mean such things are decided by majority vote. Sometimes only the
lonely cowboy is right, but alas more often he is wrong too.

> I am simply not in a position to even attempt a reconstruction

of Proto-Afrasian or Hamitio-Semitic; it requires a certain
expertise in too many exotic languages for which even reference
materials are often hardly available. I have also worked on Arabic
correspondences along with Egyptian ones but when you get into
Semitic per se, you have *CVCVC roots to analyze before you can even
get into *CVC comparisons with Egyptian, Sumerian, and PIE.

I think it would be wrong not to include a language like Arabic
which looks very archaic - and of course has vowels.

> I wish it were not true but, unfortunately it is. And I do not

know of any way to improve the situation.

>
> Dravidian roots when proeprly reconstructed can be a sometimes

help.

> ***

If the closest relatives of IE are Uralic, Altaic and Eskimo-Aleut,
shouldn't they then be included in such analyses?

> > > The formant involved with *k^el- is *to, the participial
> > morpheme.
> > >
> > > I think the PIE *-to participle is clearly future-

prospective.

> >
> > And *mr-tó-s 'dead'? How did it assume the function of *past
> > participle*?
> >
> > ***
> > Patrick wrote:
> >
> > I am pretty certain you have heard of *mór-to-s, 'mortal'.
> >
> > How did it assume the function of a future-prospective?
> ***
>
> The type does not have that on a general basis. It generally has

the

> function of adjectival abstracts, which would here be 'death' as

a

> derivative from 'dead'. The process of forming nouns from

adjectives by

> moving the accent to the initial is a living one throughout the

known

> linguistic history of the older languages, as Gk.

thne:tós 'dead' =>

> thánatos 'death'. In the prehistory fo the IE protolanguage, the

process

> of course operated on the form of the adjective of the time:

when the

> participle **mer-tó-s 'dead' (the prestage of PIE *mr.-tó-

s 'dead') had

> reached the stage **mor-tó-s the derivation of a substantival

counterpart

> created *mórto-s. The compound *n.´-mr.to-s 'immortal' must

contain the

> noun in its second part (then with further weakening of the

vowel because

> of the placing of the accent in the compound), cf. Gk. athánatos
> 'immortal' from 'not having death'. The simplex noun *mórtos

apparently

> changed its meaning from 'death' to 'a mortal, a human being',

not by any

> demonstrable function of the suffix *-to-, but by a simple whim

of

> linguistic usage. It is up to you to demonstrate that *-to- has

the

> alleged "future-prospective" function on anything resembling a

general

> scale. An isolated and non-representative case does not do that.
>
> ***
> Patrick writes:
>
> Here, I believe you are being a bit dogmatic.
>
> If the languages I have used actually are related to PIE, that

should be taken into account for your analysis.

No, not for analyses applying to the immediate prehistory, unless
the testimony of the external material is of an unusual degree of
clarity and makes a very strong statement. I do not see that.

>
> I have not attempted to prove the Nostratic theory to you; I

have a website that attempts to do that. Frankly, I am much more
interested in learning from you about PIE rather than in proving the
Nostratic hypothesis to you.

I am already convinced about the correctness of the Nostratic
relationship.

>
> But, although I disagree with Bomhard on several details, I

think he has made a good case for Nostratic as a family (Eskimo,
maybe, maybe not). What do you think?

I think yes for the family as a whole, and for Eskimo I also say
yes, though I can't see it in the lexicon.

> I have a small disagreement strictly within PIE terms with what

you have written.

>
> I do not believe *mR-tó-s could have become *mór-to-s at the

time the *e/*o Ablaut was still fully functional. In my opinion, *o-
vocalism was a resukt of a stress-accented *é becoming stress-
unaccented.

There was no "fully functional" e/o-ablaut. There were a number of
sound changes that changed some e-vowels into o, and a rather
sweeping one that deleted unaccented short vowel. This is not
basically a functional matter.

>
> *mór-to-s, in my opinion, needs the pre-stages *mér-te-s then

*mor-té-s then *mór-to-s, , the second *o being retained because the
Ablaut no longer (fully) functioned.

The second -o- is -o- because it is the thematic vowel. Thematic
vowel is a cover term of stem-final vowels; in that position vowels
are not deleted, but are subject to a different rule of alternation,
being regularly -o- before anything voiced, -e- elsewhere. The
nom.sg. *-o-s is explained by the assumption of a nom. marker *-z
which was originally voiced. The to-and-fro of *mér- > *mor- > *mór-
is in my opinion unavoidable; the thematic vowel seems to have
stayed unaffected all through it.

> *mR-tó-s needs the pre-stages *mér-te-s then *mor-té-s then *mór-

to-s then *mR-tó-s, with the inrtoduction of zero-grade.

You do not have to accent the root in the input form, it just had
that form before the vowel deletion happened. So it's *mer-tó-s >
*mor-tó-s > *mr-tó-s (making reservations for the exact shape of the
thematic vowel which is independent of it all).

> I will do some more looking into -*to-.
>
> ***
>
>
>
> > "Operate" is a little sarcastic, also. Sarcasm to what

purpose?

>
> Nothing of the sort was intended.
>
> ***
> Patrick wrote:
>
> To be called an "operator" in English implies dishonesty or, at

least, lack of ethics.

> ***

I certainly didn't mean that.

>
> > Now if you cannot see that *streu-, 'strew', e.g. is an s-

mobile form

> of *reu-, 'roil up', then your eyes may be poorer than mine (and

I am

> working on cataracts). Not all related PIE languages handled

initial

> *sr- the same.
>
> You're right I can't see that. For, as you quote yourself just a

few

> seconds later, there is also an IE root *srew-. PIE *srew- and

*strew- are

> two different roots.
>
> >
> > Under *srebh-, 'drink up', Pokorny lists Latvian

strebju, 'drink up

> noislly'. Poor Pokorny, wrong again - according to you.
>
> No, Pokorny is correctly using a rule applying to Latvian here.
>
> > Under *sr-edh-, 'whirl around', we have OHG stredan.
> >
> > Under *sreu-, 'flow', OHG stroum.
> >
> > Now what is that -t- if not euphonic, a desert mirage?
> >
> > It is obvious to any objective observer that initial *sr

never occurs

> in Germanic (and, some other IE languages like Latin), and that

PIE

> roots in *sr- are provided with a euphonic -*t-.
> >
> > Back to *streng- again. Pokorny lists Middle Irish srengim.

You will,

> of course say, that -*t- in this combination disappears in

Celtic; but

> I will say that it was never there because Celtic permits

initial sr-.

> >

> Now, I guess you can still fall back on my connecting that

phenomenon

> with PIE. Well, my answer to that is that I consider Pokorny's
> entries, with the exception of the 'laryngeals', PIE. I guess I

could

> have explained at length that the proper reconstruction of

*streng-

> was **sreng-, but German and Latin and others added a euphonic -

*t-,

> and that these facts were not obviated by the circumstance that
> examples in *sr-permitting language families (outside of Celtic,

for

> which the wrong explanation would have been advanced) are
> unfortuitously not attested.

No, Latin does not add a euphonic -t- in *sr-

> ***
> Patrick writes:
>
> I guess I still have to stick we my explanation in the paragraph

above.

> ***

I don't see how you can.

>
>
>
> > As for your *H3reg^- or *H3ryeg^-, I would laugh if I did

not have to

> cry. You are conflating every possible variation into a monster-

root.

> > ***
>
> Sure, what's wrong with that?
>
> ***
> Patrick writes:
>
> It is the method whereby Starostin, for example, has earned the

extreme scepticism of his colleagues in Caucasian studies.

>
> In the case of this root, Greek o- is probably just the prefix

*o-, which is a known quantity in PIE studies.

If the remarkable Indic forms with iraj- are related it must be a
laryngeal.

> As far as the *y is concerned, if it is validated, it is a

single-language development from **reig^- or *reg^y-.

No it seems to be a pre-PIE development of ry > r, but only after
the loss of unaccented short vowels. That creates *H3ryeg^- >
*H3rig^- in the zero-grade as opposed to *H3ryeg^- > *H3reg^- in the
full-grade. There are some others of this kind: *lwep- peel' > *lep-
/lup-, *ghryebH2- 'grab' > *ghrebH2-/*ghribH2-. Hirt saw this. This
is not confined to any special branch of IE, it's just part of the
inherited material.

> There is no need to conflate them if one simply admits that

various forms of this root were present at one stage or another.

> ***

This begs the question of how the variety came about.

>
> > > I am now prepared to admit, on the basis of what you have
> > demonstrated, that *-y as a final element of *CVy is as

likely to be

> > simply semantic differentiation as to be conveying durative;

the

> > same for, at least, some *CVCy-.
> >
> > I do not see any differentiation, nor have I seen you

proving that

> > the *-y is an extension at all.
> >
> > ***
> > Patrick wrote:
> >
> > Would you prefer to regard *CVC- as a truncated root?
>
> Yes, I would. That's what I've been saying all the time.
>
> ***
> Patrick writes:
>
> That is the first time truncation has enterd this discussion to

my knowledge.

>
> But fine, it is also a possibility.
>
> Prove, if you can, why truncation is the preferred explanation.
> ***

I understand your use of truncation to mean that the underlying form
of the root alternats is the long form and the shorter forms are
derived from it by loss of some of its material. I say this is the
only sensible analysis, for there is no functional connection
between the y-variants; the alleged "y-extension" just seems to turn
up everytime the phonotactic frame allows. That makes it a matter of
pure phonology, and simple sound laws can be formulated to predict
the forms that actually turn up. They are laws of loss and
metathesis, not of y-addition. Roots that have no -y do not add one
even where those that had a -y have retained it. That looks
elementary enough.

> > No objective person can say they are not related in some

way!

> > ***
>
> They are in my analysis underlyingly *identical*.
>
>
> ***
> Patrick writes:
>
> I believe that economy of effort is a fact of life and language,

too.

>
> I will never believe hat *-y is added to a root for _no_ reason.
> ***

Very good, we're in business.

> Jens:
>
> Come on, now, if *newaH2- is 'make new (*newo-s)', isn't Lat.

canta:-

> 'sing' not 'make sung' (based on the participle cantus 'sung').
>
> ***
> Patrick writes:
>
> I do not understand. You seem to be saying that I have asserted

that *H2 (my *aH) somehow conveys a causative whereas I believe it
conveys stative.

> ***

Have you said that? You shouldn't have. The stative is formd by *-(e)
H1-. Factitive verbs are derived from thematic adjectives by means
of *-H2-. The stem *newaH2- is retained in Hittite newahh- 'renew'.

> > Also, the -*ye/o- causative is not "durativizing".
> ***
>
> Jens:
>
> The causative morpheme is *-ey-e/o-, possibly from earlier *-ey-

ye/o-. The

> causative stem is indeed a durative, for its aorist is

different: the

> reduplicated aorist. And of course because it forms a present

indicative.

>
> ***
> Patrick writes:
>
> Judging only from inside PIE, one could toy with that idea.
>
> But what proof could you advance for the gemination?

None. IE does not have *-yy- in any words I know of. We know that *-
s-s- yielded *-s-, so is it so terrible?

> I believe that the causative *y comes from pre-Nostratic *¿o as

opposed to differentiative *y from *?e.

>
> Lehmann writes: "It is well-known that causatives in -*é-yo-,

like other derived forms, 'belonged only to the present stem'
(Brugamnn 1904a:535)". We suppose that is the earlies state of
affairs though causatives were later treated as roots, and inflected
for tense.

>
> Before the development of a future, the present indicative seems

a logical stem from which to form the causative since what is caused
to happen is always future to the actual time of the causation.

> ***

Well, PIE also has a causative aorist. Only that has a different
stem-formaiton.

End of part one of the reply.

Jens