--- In
cybalist@yahoogroups.com, "Patrick Ryan" <proto-
language@...> wrote:
> You have the audacity to imply a swindle on my part while all
you have done is display a lamentable lack of knowledge about
cuneiform and how it is to be read.
Oh yeah? Let's see how it progresses.
> Now, I am going to quote Sturtevant (1951) on several issues you
bring up. Your answer to all these, I am sure, if you follow
established patterns, will be to dismiss anything Sturtevant held
(that clashes with your own opinions) as a charming relic of the
past; and label your own unsupported opinions as examples of the
wonderful progress since those sad and benighted times (dialectic
linguisticism?), but I owe it to the other readers of this list to
review some of Sturtevant's opinions on statements you have made -
without any proffered justification but your own hallowed opinion -
and less honorably, for the sake of your argument. I will not, and
they should not blithely accept your dismissal of past knowledge
unless you are willing and able to present new evidence or arguments
to support your progressive opinions. An admonition to 'read up on
it', with the implication that if only one read what you have, he
will agree with you totally, will be regarded as the _cop out_ it so
richly deserves to be labeled. Delphi has been closed some several
years now.
Little man on a mission, eh? Poor Sturtevant who did so many good
things, do you really have to drag his nose through his errors and
shortcomings? If all the answers are in Sturtevant's production by
1951, what is the world to do with the following half-century of
additional scholarship? Have you any idea why it was produced?
> 1) you write: "The Hittite form is not "tikan" but /te:kan/"
Guilty as charged. Sturtevant wrote the same.
> Sturtevant (p. 19): "At the time our texts were composed (ca.
1300 B.C.) e and i represented a single phoneme, but when Hittite
was first reduced to writing they were distinct, and in many common
words the distinction was preserved in the traditional spelling."
That is very wrong.
> Now it is possible that te-kan is spelled with 'e' (and as te-e-
kan) because the vowel was /e/ at the time the word was "first
reduced to writing" but for IH, Sturtevant does _not_ reconstruct
an /i/ so what significance can "original" 'e' have? It could not
have contrasted with IH **i since such did not exist.
This is incomprehensible. Could you translate into understandable
and unambiguous terms? The protolanguage from which Hittite and the
rest of IE is sprung, i.e. Sturtevant's IH, does have /i/ and /e/ as
separate phonemes, no matter what Sturtevant might have said on the
subject. The facts are there for anybody to check. They have now
been sorted out in two detailed handbooks of admirable quality,
Melchert's Anatolian Historical Phonology of 1994, and Kimball's
Hittite Historical Phonology of 1999.
> As for whether the word was /digan/ or /degan/, if Sturtevant is
right, such a phonemic contrast did not exist either in IH or
Hittite.
He is not.
> For Hittite, Sturtevant postulates three vowels: i, a, and u (p.
17). Therefore, if he is right, <te-kan> or <ti-kan> would _both_
have been pronounced /digan/. But, there is the question of
his "Suggested phonetic interpretation" of te(e)kán on p. 187 of his
dictionary: "degan". On the first page of his dictionary, Note 2,
appended to the heading "Suggested phonetic interpretation", he
includes the sentence: "Consistency in writing e or i is not
attempted." Though I deplore this lack of consistency, it seems
obvious that Sturtevant intends that, written <e> or <i>, the same
phoneme is intended, and by p. 17, that phoneme is /i/.
And what is his argument for this particular choice that has
convinced you? The fact that he had read more tables than me, and I
say something else? I don't follow the logic here (well, what else
is new?).
> Therefore, Jens' attempt to differentiate between tikan and te(:)
kan is meaningless and uninformed.
It's not a thing *I* am doing, it's what the material shows and the
field has discovered long ago.
> When Jens wrote above, "Since both <te> and <e> are unambiguous
signs in the Hittite syllabary being
> both opposed to i-signs <ti> and <i>, there is no possibility of
reading the form with an i-vowel".
Where's the main clause to go with this? Plain and wise words, what
am I to say?
> We have seen that, according to Sturtevant, these signs are
_not_ differentiatable.
They are. I recommend Heiner Eichner's article in VI. Fachtagung,
Wien 1978 (publ. Wiesbaden 1980): Phonetik und Lautgesetze des
Hethitischen - ein Weg zu ihrer Entschlüsselung (pp. 120-165).
There's a wonderful table p. 133 of what replacement signs are
written when the intended one does not occur. The differentiation
of /e/ and /i/ are the big problem, but the solution to it does
certainly not lie in denial of the opposition which is very real.
The rule is actually simple: If the e-phoneme is meant, one uses an
unambiguous e-sign if that exists: me-en /men/. If it exists only
for one of the two adjacent signs, that's alright: me-iz /mez/
(there is no <ez>), li-em /lem/ (there is no <le>). If none of the
signs exists with an unambiguous e-value, a vowel sign <e> may be
put in: li-e-iz, but also li-iz for /lez/. In cases like the last
where the -e- is necessary to mark the e-value there is no reason to
interpret the vowel as being long (though it may still be), but if
there is a "superfluous" -e- inserted, the vowel is surely long,
such as me-e-en for /me:n/. I find it somewhat surprising if
Sturtevant had not really understood this; as I have pointed out he
really quotes "te:kan" with a long vowel, obviously based on the
spelling te-e-kán.
> Jens writes that the vowel of this word should be reconstructed
as /e:/.
No, it should be posited as /te:kan/ with a *synchronically* long
vowel. And yes, that is based on the disambiguating spellings.
> He apparently bases that on occasional spellings like <te-e-kán>;
again, dead wrong, according to Sturtevant, who writes on p. 37:
>
> "IN Akkadian, vowels are frequently written double (U-2-
UL 'not', BE-E-EL 'lord'). A similar usage is characteristic of
Hurrian writing, including the Mitanni letter. In Akkadian writing
an extra vowel sign may indicate a long vowel, but in Akkadian texts
at Boghazköy the double writing of short vowel is unusually
frequent. There is therefore no reason to infer vowel length from
the insertion of an extra vowel sign in Hittie words. Hittite e-es-
zi is the constant spelling of the word for 'is', which must be
identified with Skt. asti, (Greek) ésti, Lat. est."
These are *etymologically* short vowels, but they are lengthened in
Hittite when accented. It's like German ge-bären which has a long
vowel today reflecting the short vowel of IE *bhér-e/o-. The rules
according to which specific vowels are lengthened under which
specific conditions are set out to the extent the material permits
in Melchert's and Kimball's handbooks; as I understand it, the
groundbreaking work was made by Eichner.
> Now, for the sake of fairness, let us pretend that the vowel of
tekán is /e/. Jens insisted on /e(:)/ to eliminate my suggestion
that tekán might be derived from PIE *dheig^h-, 'clay?', namely.
*dheig^hón. On p. 21, Sturtevant writes:
>
> "IH y after a tautosyllabic short vowel yielded Hittite e; after
a tautosyllabic long vowel it yielded Hittite ai."
Sure, i-diphthongs yield /e/ (or at least something often spelled
with the e-signs).
> Thus, PIE *dheig^hón would be expected (by Sturtevant, at least)
to yield Hittite tekán.
I'm not sure for a suffix-stressed form, by why put the accent
there? It is plainly on the /-e:-/. And yes, *regarded in
isolation*, that form could be derived from a preform with an i-
diphthong. But we have other forms of the word, so let's read on.
> Now, Jens continues to assiduously shovel out the grave of his
own argument with:
>
> "The Luwian counterpart of Hitt. /e:/ is <i>, as indeed it is in
tiyammi-, so these two fit exactly."
As they do.
> 1) There is no Hittite /e:/ that is unambiguously indicated in
the cuneiform even if it existed, which Sturtevant says it did not.
On the contrary: <te-e-kán> *unambiguously* spells a form with a
long [e:].
> 2) Even if Hittite /e:/ was assumed, it could represent PIE /ey/
so the example is interesting but the conclusion is totally
unjustified, i.e. tiyammi would not exclude the possibility of PIE *-
ei-.
Viewed in isolation, yes, but now watch out (this is getting nicer
all the time).
> We should also ask ourselves if we have ever seen a
reconstructed PIE form like *dhe:g^hóm, which is what Jens seems to
be proposing for the prototype for his Hittite te:kán. I have not.
Perhaps some of you readers have. If so, I would like to know about
it.
Mate straightened you out on this false quote, I'll pass it over in
silence, kind as I am.
> In my antiquated pre-classical thinking, a stress-accent on the
second syllable, which is supposed to produce the tongue-twister
*dhg^hóm, necessitating the non-Hittite metathesis to *g^hdhóm
(oops! *g^hðóm), would not allow *dhe:g^hóm but at the very most
*dheg^hóm (reduced grade). And why would this word have /e:/ in the
first place: Hittie vrddhi? Now these are questions of PIE rather
than cuneiform, and I wonder where Jens got these ideas.
"Why would it have /e:/ in the first place?" Because that's what
Hittite made out of accented IE *é. And the Hittite form is not
identical with that of the other branches, but represents the
*preserved* strong-case stem of the old paradigm which was levelled
in all the other branches.
>
> Then Jens writes: "I do not think there are plene writings of
unetymological vowels . . ."
It would presuppose vowel insertion with subsequent accent shift to
the new vowel. But why would a putative *dhg^hó:m move the accent to
an inserted -e- if the inserted -a- of gen. ta-ag-na-a-as keeps its
final accent?
> No one who has the slightest knowledge of cuneiform, Hittite or
otherwise, could make such a bald-faced misstatement. This is simply
ignorant of the facts we have. Furthermore, anyone who knows
cuneiform realizes that it was written inconsistently throughout its
long history. As for Hittite specifically, I think Sturtevant
possibly looked at a few more tablets than Jens has or will; you
read what he wrote above. The Academie Accadienne simply was not
omnipotent in Boghazköy.
Eichner says a few words about Sturtevant's procedure which is
itself inconsistent and illogical.
> And finally, Jens continues: "and now that we have the Luwian in-
zgan showing what did come out of the cluster produced by zero-
grade, any opposition becomes even more pitiful." This is vintage
Jens. Anyone who disagrees is not just benighted, unprogressive, and
unread, he is also "pitiful". Jens, you omitted "wretch", out of
kindness, no doubt.
Yes, I left out a lot, out of pure kindness.
> I do not doubt that tekán might, under some circumstances, have
had a zero-grade (/dgán/) though I thought the original pitch by
Jens was that tekán showed a pre-Ablaut vowel which Ablaut would
have deleted, demonstrating its relatively earlier origin than
the "metathesized" *g^hdhóm from earlier *dhg^hóm (A.A., after
Ablaut).
Pre-ablaut vowels that are lost in the ablaut cannot be *retained*,
no matter how old a form is, so of course I said no such thing.
> What Jens does not know is that Hittite phonological developments
and spelling conventions would mean that *dhig^hóm could also be
written *tekán in Hittite.
That's right, I don't know that. I know it's not so. A phoneme /i/,
or even an unaccented old *e, which became /i/ in Hittite, could not
be written with an unambiguous marking of the e-quality in three
places in te-e-kán.
> And very last, Jens writes: "The weak-case forms of the Hittite
paradigm, such as the genitive /tagna:s/ <ták-na-a-as>, completely
rule out any possibility that the root had an i-diphthong (and an i-
> vowel in the zero-grade)." If the nominative had been /deg-án/
(from *dheg^h- or *dheig^h-), we could expect a zero-grade for
genitive to yield /d(6)gnás/.
No. Read any introductory book on IE. The zero-grade of an i-
diphthong is -i-, not schwa or zero. That's what I was referring to.
> If, in fact, it resulted in just a reduced vowel in the root
syllable, then Hittite would have written tak-na-as to be sure since
Hittite <a> represented /6/ (schwa). This would have been written
the same way if the result had been /dgnás/. Hittie had no real way
to indicate it otherwise. Only if the long had been long would
Hittite have written <ai> for the nominative.
Sure, te-e- could conveivably reflect *dheyg^h- which you are
fighting so bravely for, but its zero-grade would be *dhig^h- with a
vowel [i] which could not be treated as in ták-na-a-as, allative ta-
ak-na-a. It would have needed *ti-ik-.
> Perhaps I am older than you, Jens, but one of the few things I
think I have actually learned through life is that nothing
completely rules out anything. There are only probabilities, and the
probabilities of your explanation are, because of your mistaken
understanding of cuneiform, unimpressively low.
That's just more nonsense that goes to your embarrassment. The
probability that the allomorphs of the Hittite paradigm of the word
for 'earth' are etymologically related is certainly greater than any
idea that they aren't as you are suggesting (apparently without even
knowing it).
> Have I proved my case? No. I have only showed, I hope, an
incrementally greater probability for its consideration.
>
> Frankly, Jens, this tripe could not be sold on a Mexican street-
corner carnicería.
Is that where I am? I should have realized.
Jens