[b-hebrew] Matres Lectionis in the Siloam Inscription

Yitzhak Sapir yitzhaksapir at gmail.com
Thu Aug 30 20:26:49 EDT 2007

It has been claimed that the Siloam inscription has matres lectionis.  I think
there is only one clear case of a mater lectionis in the inscription.  General
opinion of scholarship is that there are various types of mater lectionis in the
inscription, though.

The following is my take, going from what I consider more likely cases of
matres lectionis to least likely.

Final matres lectionis:

-w in wylkw: Probably a mater lectionis.

-y in ky: I do not want to discuss this at length.  According to HALOT, this
spelling is found even in Ugaritic, where mater lectionis are extremely rare to
nonexistent.  I think this is reason in itself to suggest that this
spelling does
not have to signify a mater lectionis.

-h in wzh ("and this"):  An Akkadian transliteration of the Ugaritic
is du-u, which may just indicate a long -u:, especially in light of the
consonantal spelling d in the alphabetic texts.  In other languages (such as
Phoenician, Aramaic or OSArb) we have a two consonant pronoun for the
masculine [dh]n (and we have the example of zn that Yigal mentioned, which
in the LXX is interpreted as this pronoun).  The most suggestive evidence for a
consonantal spelling, however, is that Phoenician uses z) with a consonantal
aleph alongside the spellings z, zn, and )z for this pronoun.  In other words,
just like Phoenician had a pronoun with a consonantal ), the pronoun in Hebrew
could have had a consonantal he.  (It would, in my opinion, also go some
way to explain the Massoretic vocalization of this pronoun because if it was
just one long vowel originally, how did it become a seghol?).

-h in hyh ("was"): The original form of the verb hwy, and the 3ms suffix Qal
was hawaya.  Just like the medial glide -w- became a -y-, there is no
reason not to suppose that the final -y- did not become an -h-: hayaha (later,
the h dropped, leaving a long a:, after which, this long a: became a qamats).

Final -h for feminine: It has been observed in other languages that
final feminine
-t becomes a consonantal h.  There is even a rhyme in the Quran that depends
on this.  Also, it is quite possible that t was aspirated which might further
suggest why -t became an -h.  In any case, we cannot know that -h for the
feminine singular was not consonantal.

-w in hkw: The w here is also a root of the word.  Thus the vocalization
might very well be, "hikkawu:".

-) in qr): This is a root letter of the word, and so it is reasonable to suppose
that it is consonantal.  One can compare in this regard the absence of the
aleph later in this inscription in the word lqrt.

-w in r(w: This word could be plural, but even if not, since the -w- also takes
place of a pronoun suffix we cannot be sure that the -w- isn't consonantal.
In fact, on this particular point, scholars generally agree that it is

-h in gbh: Here we have a letter of the root, that is vocalized even in
the Massoretic tradition.

Medial matres lectionis:

-w- in hmwc): To add to the evidence adduced before for -aw- being
sometimes vocalized as a diphthong, the reading in an Amarna letter from
Byblos is mu-u-ca, which could indicate that in this word the diphthong
remained.  One can compare an Akkadian transliteration of the Ugaritic
vocalization for "dwelling", mu-$a-bu.

-w- in b(wd: Scholars have suggested the -w- is consonantal if for no other
reason than that it is supposed to be if medial matres lectionis are not to
be attested in the Siloam inscription.   As it stands, Sarfatti refers
to Rainey's
evidence that -aw- was sometimes vocalized as a diphthong (thus the name
Hoshe (was sometimes vocalized as a-u-si-)a, and other times as u-si-a in
Akkadian transliteration), but he also notes that there is no compelling
reason to vocalize (awd or mawca in this inscription.  In discussing its
lexical meaning, Sarfatti notes that while the Bible uses this word as an
adverb, the use in inscriptional Hebrew is as a noun - remainder.  This
word is poorly understood and does not have many cognates to help us
out.  Also, if the word servers as a different part of speech in later Hebrew
(as it does - an adverb in place of a noun), we cannot use the later
pronunciation to work back the earlier pronunciation, because it might
have a completely different word form as well. Even though the word is
poorly understood, what we do have is a verb (wd which in the Piel
means "to surround", while cognates suggest "return", "to circle around."
(Note the meaning of "return" from "re" + "turn").  A use of a verbal noun
for the Piel would have a close meaning to "remainder" ("returning", that
which returns to you after you divide something up).  The vocalization of
the Piel infinitive would be (iwwud in the Massoretic, but the much closer
Ugaritic evidence gives (uwwadu.  In any case, the noun is poorly
understood, has a different part of speech role in the inscriptional
evidence than in the Bible, and it could very well have had a different
vocalization with a consonantal -w-.

-y- in m)ty[m]: Here the -y- is likely consonantal as we have a dual.

We note there are only three examples of possible matres lectionis in
medial position, and the first two should be considered against several
words that likely had -o- as their vowel (such as yom and qol).  The
case for medial matres lectionis remains weaker.  In the end, the only
place I think we have a clear mater lectionis is the -w in wylkw.  Indeed,
I think -w for -u: had been used already in the second millenium BCE.
The second strongest case is for -y as -i:, which may be related to the
development of -iy > -i: (that is, final original -iy developed into a long
final -i:).

Yitzhak Sapir

More information about the b-hebrew mailing list