Biblical Hebrew Forum

["Henry Churchyard" <churchh AT crossmyt.com>]

> From: "Jeremy" <jnorthct AT hotmail.com>
> Date: Sun, 7 Oct 2001 04:50:29 -0400

>> From:  "Henry Churchyard" <churchh AT crossmyt.com>
>> Date: Wed, 3 Oct 2001 09:38:04 -0500 (CDT) 
 
>> The whole absolute vs. construct state distinction is not as
>> relevant for numerals as it is for ordinary nouns, since the main
>> uses of the construct state are to signal noun-noun compounds, or
>> possession, but when a numeral precedes a noun, then such a phrase
>> is not actually either a "compound" or a possessor-possessed
>> situation in any true sense.  So in the numerals there are forms
>> which cannot be classified as being either contruct or absolute
>> (e.g. "sh@neym" and "shteym" in "sh@neym-`asar" and
>> "shteym-`esre"), forms which act quite differently from nouns with
>> parallel morphology (e.g. the numerals from 20-90 end with
>> masculine plural "-im" ending, but do not form constructs ending in
>> "-ey" as nouns do), etc.
 
> But cardinals from three to ten place `tau' at the end to indicate
> the masculine construct, which, while not following the normal
> morphology of adding `yod,' does provide a consistent classificatory
> scheme for distinguishing absolute numerals from constructs (at
> least for the masculine forms for numerals from three to ten).

Not sure what you're referring to by "adding yod" (only a very few
highly irregular nouns like _'abh_, _'ah._, and _peh_ can be really
said to add yod in construct state); the masculine and feminine
numerals 2-9 basically show normal patterns of morphological and
phonological alternation in their construct and absolute forms
(i.e. orthographic final _qames._ + _he_ in absolute becomes short
vowel + _taw_ in construct; vowels in pre-main-stress open syllables
which are lengthened in absolute are instead reduced to _sh@wa_ or a
_h.at.eph_ vowel in construct; etc. etc.).  The only particularly
noticeable peculiarity that I see is that five and six have absolute
forms _h.amishshah_, _shishshah_, but construct forms that look like
they should go back to historically earlier forms *_h.amisht_ and
*_shisht_; however, true historical forms ending in _-iCt_ generally
went to _-aCt_, so that the long _s.ere_ vowel seen in construct
_h.ameshet_ and _sheshet_ is probably due to a fairly late analogical
development (however, the same long _s.ere_ vowel is seen in _'eshet_
as construct of _'ishshah_...).

> What I don't get is why, in certain instances, a numeral is employed
> in its absolute form and, for an identical expression, the construct
> form is employed.  There must be some basis for the occasional
> variations that take place other than those of grammar, such as
> regional or period differences.  Or am I to understand from what you
> are saying that the construct rules for numbers were somewhat
> ill-defined and, hence, it was left to each individual scribe to
> decide which way to go?

Individual scribes or Masoretes certainly didn't feel free to make
arbitrary changes (that's the "Kahlean fallacy").  I actually don't
have any specialized knowledge about the detailed distribution of
construct vs. absolute numeral forms in the text -- however, if this
distribution should happen to show irregularities or apparent lax
usages, then this would not be particularly surprising to me, since I
know from general principles that the construct vs. absolute
distinction is not as relevant for numbers as it is for ordinary
nouns, as I explained above.  (This means that the irregularities and
apparent lax usages might have been present in a single state of the
language, and don't necessarily need to be explained by regional,
diachronic, or scribal variation.)  A detailed study of the
distribution of these forms might reveal more subtle patterns and
tendencies in their occurrences.


--%!PS
10 10 scale/M{rmoveto}def/R{rlineto}def 12 45 moveto 0 5 R 4 -1 M 5.5 0 R
currentpoint 3 sub 3 90 0 arcn 0 -6 R 7.54 10.28 M 2.7067 -9.28 R -5.6333
2 setlinewidth 0 R 9.8867 8 M 7 0 R 0 -9 R -6 4 M 0 -4 R stroke showpage
  % Henry Churchyard    churchh AT usa.net    http://www.crossmyt.com/hc/