Biblical Hebrew Forum

["Yitzhak Sapir" <yitzhaksapir AT gmail.com>]

On 10/30/07, K Randolph wrote:

> I looked back over the homonym tables that Peter Kirk and I discussed
> a while back. In the list that Peter devised based on scholarly
> consensus, even though the percentage of homonyms for roots containing
> X is higher than average, it is within the standard deviation for
> homonyms for all roots. Because its percentage is within the standard
> deviation, it would seem that evidence from within Hebrew for X to
> have come from two original letters is weak.
>
> The expectation is that if X originally came from two original
> letters, then the percentage of homonyms containing a X should be
> significantly higher than average.

There is no such expectation.

If X1 and X2 are independent phonemes, and X1 is present in W1 words,
with H1 homonyms, and X2 is present in W2 words with H2 homonyms,
then if they coalesce into X3, we will have W3 = W1 + W2 words, with
H3 = H1 + H2 + (new homonyms)  - (multiple homonyms).
If the ratios H1/W1 and H2/W2 are somehow standard, then the ratio
H3 / W3 should also be fairly standard.  The main problem is in calculating
the new and multiple homonyms.  However, it is very likely they pretty
much cancel each other out.

We discussed this before, and I specifically dealt with this point.
https://lists.ibiblio.org/sympa/arc/b-hebrew/2007-January/031343.html

Yitzhak Sapir