[b-hebrew] Reflection on Randall's Statement
jc.bhebrew at googlemail.com
Sat Jan 23 04:28:06 EST 2010
For the benefit of the layman who may read this and think Randall actually
knows what he's talking about I would present this question to Randall so
that the layman can see how and if he can even answer it:
Akkadian being a dead language on what transliteration evidence are you
basing your 'knowledge' of Akkadian pronunciation?
2010/1/23 Randall Buth <randallbuth at gmail.com>
> vayyixtov Uri Hurwitz
> >>Randall Buth has written, inter alia:
> "... inside a language
> family where older, genetically-tight sister languags
> have established three-vowel systems.">>
> > Yes, of course the three vowel system is known
> from the earliest Akkadian evidence. It is however
> interesting to ask whether these early Semites had
> indeed only three vowels, or simply chose to compress
> their vowel markings to the three.>
> First, it depends how one counts.
> Semitists agree that the early western Semitic languages had
> six vowel phonemes /ii/ /i/, /aa/ /a/, /uu/ /u/, which can also
> be described as 'four', /i/, /a/, /u/ + 'length'. And since length
> also applies to consonants, one can simplify the description
> by talking about three vowels, since all three vowels are
> susceptible to lengthening. (Many languages with phonemic
> length in their vowel system do not show symetric patterns
> but will typically have more longer vowels than the set of
> shorter vowels.)
> In most languages one discovers forms that preserve older
> forms of a language. These are forms that were protected,
> for one reason or another, from phonological or morphological
> changes that were generally applied throughout the language.
> I can illustrate with three brief examples from Hebrew and then
> will briefly go on to the general point.
> Qal suffixtense verbs are known in three patterns: qatel, qatol,
> and qatal. However, there was a general trend in the language
> to 'regularize' the morphology toward 'a', so that 1st and 2nd
> person singular forms take 'a':
> Hafets 'he wanted' Hafatsti 'I wanted'
> consider also 'ask', that has an 'i' vowel in Arabic history, both
> regular sa'il 'he asked' and si'il 'he asked' (some dialects today)
> In Hebrew"
> sha'al 'he asked' (already regularized to 'a') sha'alti 'I asked'.
> However, forms with added suffixes could preserve an older
> form of the verb and we get a form that might surprise a beginning
> she'iltihu 'I asked him'
> The first 'a' vowel of sha'alti dropped out as a regular feature. But the
> second 'a' vowel surprisingly shifted to 'i'. "What?!"
> The 'i' is simply the ancient, original vowel of this word. It is a fossil.
> 'a' did not perversely change to 'i', but the original 'i' never got erased
> or changed from this form. It was 'buried' three syllables inside the
> form in a closed, unaccented syllable that resisted vowel reduction
> and change.
> sha'al was a stative verb.
> There are 3 kinds of these verbs, corresponding
> to the three vowels i, a, u, and corresponding to the three vowels of
> Arabic. qatel and qatol are explained as lowerings of *qatil and *qatul.
> There are three kinds of common noun pattern *qitl, *qatl, *qutl, which
> have regualrized over the ages into *qet-l/qat-l, *qat-l, *qot-l, the
> segolate forms.
> there are three super short 'shva'-vowels: Hataf-segol, Hataf-pataH,
> Hataf-qamats, corresponding to the short sounds 'e', 'a', 'o', which in
> turn correspond to the earlier *i, *a, *u. However, while these do
> often reflect the etymoloigcal *i, *a, *o vowels, they are not really good
> evidence on their own, because such 'doubly-short' vowels could have
> naturally have collapsed a larger inventory of short vowels.
> > The Masoretes, much later, had two additional ones:
> Segol and Tseireh, not to mention the differences
> bweteen Hatafim and regular vowels. >
> Yes, but these 'newer vowels', or 'younger vowels' entered the language
> and fit into many places without generating new morphological structures
> in the ancient core patterns of the language that are attested as
> cognate patterns in the genetically related languages.
> Hence, there are not five stative verbs qatil, qatel, qatal, qatol, qatul,
> only three. There are not five cognate segolate nouns
> qitl, qetl, qatl, qotl, qutl.
> > And, interestingly, while classical Arabic has only
> the three traditional vowels familiar from the Akkadian,
> the spoken Arabic has a clear Segol/Tseireh vowel: Khirbeh,
> sheikh and so on.>
> this shows a natural develop within Arabic along the same trajectory
> as Hebrew. Many, of course, show the history of 'ay' becoming 'e' and
> 'aw' becoming 'o'.
> This email, expanded into a book, would be an introduction into why
> three vowels are the satisfactory, sufficient, phonemic vowel inventory
> of early western Semitic language systems. Said another way, we
> don't find 'minimal pairs' at earlier strata of the language necessitating
> more than three vowels, though this is tempered, of course, with some
> reasonable allowance for homonyms. See Bauer-Leander
> and Jouon-Muraoka.
> Randall Buth, PhD
> randallbuth at gmail.com
> Biblical Language Center
> Learn Easily - Progress Further - Remember for Life
> b-hebrew mailing list
> b-hebrew at lists.ibiblio.org
More information about the b-hebrew