[b-hebrew] Initial Consonant Clusters in Biblical Hebrew

Yitzhak Sapir yitzhaksapir at gmail.com
Fri Jun 11 07:50:56 EDT 2010

On Fri, Jun 11, 2010 at 7:24 AM, George Athas wrote:
> I think that by ‘initial consonant cluster’ Jim is referring to two consonants (not separated by a vowel) occupying the first place within the same syllable. In other words, the syllable would be of CCvC type.
> To answer your question, Jim, yes the number two, שְׁתַּיִם, does preserve an initial consonant cluster. The dagesh lene in the taw demonstrates that this is the case. It is an unusual occurrence in Masoretic Hebrew for this to happen, but there it is. Usually, Masoretic Hebrew seeks to separate such consonant clusters with a vowel or vocal shewa. This is one instance where it does not happen.
> A little more common, but still not prevalent, is ultimate consonant clustering of the CvCC type. Witness the final syllable in Qal Qatal 2fs forms of regular verbs: קָטַלְתְּ.

In Tiberian (Masoretic) Hebrew, the number two is pronounced [eshtaim].
See here -
The above provides us with:
Tiberian Hebrew - e$taim
Phoenician - ?$NM
Arabic - ithnatani

Now, in Samaritan Hebrew we have the form: $ittem (e = schwa sound).
Both Samaritan Hebrew and Tiberian Hebrew were under heavy Arabic
influence that did not allow initial consonant clusters.

But on the basis of both forms, we might suggest that the common form
in the common language from which they diverged was [$ttaim].  Now
Samaritan and Judean Hebrew could have developed alongside each
other for a long time, maintaining such a difference.

Before the Arabic influence, there was a tendency to remove the consonant
cluster already beforehand as we see in Phoenician.  But Biblical Hebrew
never has )$TM or )$NM, and so Hebrew apparently did not break
consonant cluster in this case.  This is indicative that the pronunciation
'$ttaim' was probably current in Biblical times.

This is not the only case of a consnant cluster.  Samaritan Hebrew has
in the imperative certain abnormal forms with an initial prostethic vowel:
Tiberian kato:v 'write!', Samaritan ikt at b (@ = qamats).   Tiberian
paso:l 'sculpt!' Samaritan afsel (e = schewa sound).  This is the
exception in Samaritan but it is important because it is indicative of
the original form.  (Note that neither consonant cluster 'kt' or 'ps' has
a sonorant, although the latter has a sibilant).

The initial prosthetic vowel is found as a regular feature in Arabic.  This
suggests that the language that Judean Hebrew, Samaritan Hebrew, and
Arabic diverged from had an initial consonant cluster in the imperative.

That there are such forms in Samaritan that are maintained suggests that
in Biblical times the pronunciation with an initial consonant cluster was
maintained in Hebrew.  After Samaritan and Judean Hebrew diverged,
Samaritan turned the consonant cluster into the above cases and Judean
Hebrew (and Samaritan in most cases) broke the above by placing a
vowel in between.

Modern Hebrew is based in large part on European Hebrew vocalization.
We don't have a good documentation of this tradition but we can understand
the pronunciation in general cases.  The European vocalization of Biblical
Hebrew apparently preserved the consonant clusters and sonorants.  Unlike
Tiberian it was maintained in regions that had consonant clusters and
sonorants in the primary language.  Thus, Modern Hebrew has $nayim, as
well as 'ktov!', 'psol!'.

Another interesting thing is that both Samaritan Hebrew and Judean Hebrew
seem to have used an -i- to break up a triple consnant cluster:
Samaritan: $ttaim 'two (fs.)' < $ittem
Hebrew: ktvi ('write! (fs)') < kitvi

However, I wouldn't worry about it too much.  Whenever we gave solid
linguistic evidence that Jim Stinehart's theories were untenable, he simply
ignored it.

Yitzhak Sapir

More information about the b-hebrew mailing list