[b-hebrew] Dating the variant readings in the Samaritan Pentateuch

Yitzhak Sapir yitzhaksapir at gmail.com
Mon Jun 20 16:12:39 EDT 2005

Dear Professor Hjelm,

> Thanks for taking interest in my work although on a negative note, 
> some of which seem to be based on misunderstandings

Again, I think these are important issues.  My own personal opinion
right now, is that J & E are possibly Samarian sources, while P & D
are probably Judaean.  This is based, among other things, on the
issues I point out below.  I think dating the Surdi debate prior to 
the 7th century CE requires proof, not just finding an appropriate 
background for the debate to have been composed.  And that this
story is significant in what it tells us of Samaritan beliefs in the
first millenium CE.  In any case, some of these arguments might
interest you in your current research and perhaps you have a
different explanation for them than the one I propose.

> > 1) Mutilations of Biblical verses.  The most glaring are
> > on page 179.  Some seem to mix Samaritan and Judaic
> > spellings.  I think this is very severe.

> mutilation, I don't think is a proper word here. I do not
> always quote biblical verses verbatim and sometimes I have 
> rearranged some verses for rethorical purposes, however,
> never in a way that pretends to give out a verbatim quotation. 
> You might disagree to some of my paraphrases of biblical 
> passages, but that's a different matter. Mixing Jewish and
> Samaritan. I have never done that. If it seems so, and that
> would only count for Pentateuch quotations, I have simply
> misquoted the Masoretic text. please give me the exact 
> examples.

Simple examples include inaccuracies in the verse numbering, 
such as quoting 1 Sam 13:9 for the mention of Sisera (p. 193), 
whereas it should have been 1 Sam 12:9 (as far as the 
Massoretic numbering is concerned), but also direct 
spelling errors: p. 208 quotes Jer 7.15 as "zr( )ptym" [t
instead of r].  The particular error that suggested to me
that you were using the SP in some cases is p. 179 which
quotes "htwrh hzwt" as opposed to "hz)t" (Deut 31:9).  But
I have now checked that against the SP and found that the
SP also has an alef in its spelling of this verse.  Slightly later, 
we have the mutilated Deut 31:13 "wbnyhm )$r l)yd(w $m(w 
wlmdw lyr)h)w-yhwh", $m(w instead of y$m(w, lyr)h)w-yhwh 
instead of lyr)h )t-yhwh, and also a missing maqqaf in l)yd(w.  
This verse is quoted as Deut 31:9-13 even though it's just 
Deut 31:13.

> > 3) Assumptions of very late datings for the Deuteronomistic
> > History (p. 301), while at the same time taking the AF (a
> > Samaritan Chronicle by Abu Fath) as valid evidence in light
> > of very meagre arguments (p. 188). 

> The Deuteronomistic history as a scholarly concept dates to 
> Martin Noth 1943!! The Former Profets in some forms date to the 
> centuries before the common era as we know from Dead Sea
> Scrolls, the Greek and Latin evidence, Josephus and rabbinic 
> sources. The form in which we have these tests in vokalised 
> Hebrew are at best medieval (the Aleppo and the Leningrad 
> codices) and the editions most use are from the 20th century. 
> Samaritan manuscripts of AF date from the 13th to the 19th
> century, but as biblical material is known to have existed at a
> much earlier time than the actual manuscripts, so there are 
> strong indications in 1st millenia Jewish and Christian 
> discussions that Samaritans claimed the sanctity of Gerizim 
> at least as early as the 2nd century BCE.

I do not doubt that Mt. Gerizim was a holy Samaritan place and 
their primary site in the 2nd century BCE and long before that.  This 
does not mean we can assume that any particular sentence of the AF 
dates from the 2nd century BCE.  The same goes for the text of the 
Samaritan Pentateuch (SP).  We cannot assume that any sectarian 
difference between the SP and the MT is that ancient, without proof.

Scholars who use Deuteronomy or the Deuteronomistic History
as historical sources, generally base their conclusions on dating
the main revision of this document to the 7th century BCE, for
example.  Once they do they are bound by this date, and they
refrain from using this information to date events more than 500
years back.  Scholars who use Herodotus to analyze history of
the Persian period, are basing themselves on dating the text to
Herodotus, even though it only survives in late documents.  Just a
simple example regarding Herodotus, Herodotus claims that Darius 
received 360 horses "for each day of the year", suggesting that
Darius's year was 360 days long.  Evidently, this suggests a solar
calendar, and we do have evidence the Persians later used a
calendar of 365 days of which 5 were considered "outside the
normal year".  Boyce has even used this to claim that at this time
in Persian history, the calendar was 30 days a month for each of
12 months.  But the Behistun inscription, clearly marks Persian
months as lunisolar, hence 354 days in 12 months.  While we may
"harmonize" this problem by suggesting that perhaps there is an 
unwritten "religious" Persian calendar side by side with the civil 
Persian calendar marked in the Behistun inscription, this becomes
unlikely in view of the lateness of the manuscripts of Herodotus.
Thus, in any document, a scholar needs to take extreme care
when using claims of past events, and one cannot just cite the
claims of a document regarding the past as an alternative to
different documents (such as the Massoretic Bible/Septuagint)
whose existence many centuries prior to the composition of said 
document (such as Abu Fath) is widely accepted.

In the case of Abu Fath, it is wrong to use the text to analyze 
history much further back than say 500 years prior to Abu Fath
given that it is clear that the text in its main revision does not
predate Abu Fath.  Any ancient sources whose remnants might
have found their way into Abu Fath cannot be so easily recovered, 
and, personally, I saw no methodological attempt to filter ancient 
from late elements in Abu Fath in ch 4 of your book.  It also cannot 
be used as an alternative to the Biblical "historical" or pseudo-
historical books, and to those pseudo-historical books that survive 
outside the Bible from Biblical times.

Anyhow, no theory is problematic simply because it dates "from
1943."  This is especially so for the Deuteronomistic History, that
is explained today in light of the most modern archaeology by 
foremost archaeologists.  As such, the Deuteronomistic History
can be classified not as an "old and ancient scholarly theory" but
as one that although suggested at an early date, has weathered
much criticism and has convinced generations of scholars.  (I
understand that you deal with revisions of the Deuteronomistic
theory such as those by Finkelstein and Halpern elsewhere, and 
my intent here is only to point out that it is faulty to discount a theory
simply because it is aged).

> > the story brought on p. 211-213, about Samaritans proving
> > to Darius that Gerizim was the proper Temple site based on
> > exegesis of the Torah, suggests that it comes from a time
> > when Samaritans did not have an "addition" to the
> > Decalogue.  As it is quoted in an Arabic source and refers
> > to the Qibla within the story, I think this would suggest that
> > as late as the 7th century CE, the Samaritans did not have
> > such a commandment

> that is much too late for that and the discussion implies that 
> the SP includes the commandment regarding Gerizim. I don't
> think this is an 'addition' to a Jewish decalogue, but one of the 
> variants between SP, MT and LXX you find in the SP. AF 
> cannot be dated on the basis of such variants. I do not imply a 
> validity of the Samaritan Darius story as reflecting 6th century 
> discussions. I don't see any evidence that such discussions 
> took place at that time. They did take place in the 2nd-1st
> century BCE and whether the Samaritan versions belong to 
> these discussions or are, in fact, later, we do not know. 

How do you suppose to argue that the 7th century CE (the time
of the Muslim conquest and hence necessary for dependence on
Muslim and Arabic terms such as "Qibla") is too late for the Surdi
debate?  Or, furthermore, that it is too late for dating the SP in the 
text as we have it today, text that is for the most part unattested 
prior to Medieval times?  Finally, on what basis do you argue that 
prior the Medieval times, the SP was not fluid as far as the text 
was concerned and that the Samaritans were as rigid about the 
text of the SP as were the Jews for the text of the MT?  All of
these are assumptions that in my opinion are not to be taken for

To check the claims of the Samaritan "Gerizim" commandment,
I am going to bring in several articles which I have come upon in
looking up your conclusions:
1) APGAPIZIN: A Criterion for Samaritan Provenance, by
R. Pummer, Journal for the Study of Judaism, Vol XVIII, no. 1.
This refers to various manuscripts which preserved a perfect tense
for the cases of ybxr/bxr in Deuteronomy.  The earliest manuscript
provided is Latin Codex 100 (Lyon; 7th cent) for Deut 16:2, and 16:7.
(I used P above in the title, because ARGARIZIN is written in Greek
letters in the article name).
2) Israel Museum News 11, p. 75, which provides a version of the
Samaritan Decalogue dated therein to the 3rd century.  The 
Decalogue is abridged in the sense that in some cases only the 
first verse of the commandment is provided, and that the verse "I
am Yahweh, your god" is missing.  For the "Gerizim" 
commandment, only the words "wbnyt $m mzbx lyhwh )lhyk" is

One can claim that perhaps there were both changes (perfect tense
and mention of Gerizim) prior to these dates.  But they are 
unsupported by the evidence.  It is possible that perfect tense bxr
appeared first in Deut 16:2-7 and later spread to other sections of
Deuteronomy.  It is possible that first the verse "wbnyt $m mzbx
lyhwh )lhyk" (="and you shall build there an altar to God.") was first
transplanted in the Decalogue and later other verses (including the 
explicit mention of Gerizim) were too.  

It is possible that because of the abridged nature of the Decalogue
and the nature of the preserved manuscripts and the choice of 
translation, that it is simply not representative of the full version
that the Samaritans had at these respective times.  So, although 
this evidence is not conclusive in the negative either, the evidence 
is not conclusive in the positive.  It is inconclusive either way.  Of 
course, there may be other pieces of evidence of which I am just
not aware and which may prove me wrong, but I thought that for the 
purposes of the discussion at hand the evidence I brought is sufficient.

This is where the Surdi debate episode in Abu Fath is significant.
I have delayed my response in an attempt to review the evidence
of Abu Fath myself.  Firstly, the discussion/"proof" in the Surdi
episode (Ch. XIX, p. 85 - 93 in Stenhouse) does not make use
specifically of such a commandment.  Stenhouse references the
verses as coming from elsewhere in Deuteronomy but not from
the Decalogue.  In relation to Ex 20:24, one point of sectarian
difference, Stenhouse correctly places the Samaritan version.
In light of this, I wonder why the verses which are so similar to
the commandment in the Samaritan Decalogue are referenced
instead of the Samaritan Decalogue itself.   This includes a 
reference to Deuteronomy 27:4 where Mt Gerizim is mentioned 
in the Samaritan version.  The difference of location is signicant.
When read in the context of the decalogue, it reads as a 
commandment to be observed continuously.  When read in
the context of Joshua's altar, it reads as a ceremony to be
performed by Joshua, once, but not necessarily for generations
afterwards.  In all, though, it seems odd that the Samaritans will
have to prove that they need to sacrifice on Mt. Gerizim, if there 
is a commandment in the Decalogue that says just that.  Instead, 
Abdal (the Samaritan) goes into an argument that appears to be 
very convoluted.  Furthermore, Abdal uses the "altar on Mt. 
Gerizim" in the sense of a singular commandment after the
entrance into Israel [which in turn specifies the choice of God].
That is, he reads it in the context of Deuteronomy and not in
the context of the Decalogue.

Briefly, the argument goes like this: God forbade bringing offerings
anywhere but the Chosen Place (Dt. 12:13).   God ordered 
Joshua to build an altar on Mt. Gerizim (Dt. 27:4-7), where 
(olot and $lamim were to be offered.  The manner of offering
such sacrifices is detailed in Lev. and this requires offering them
at the place of the Tabernacle.  Furthermore, the name of God
must be fully pronounced in the blessings given in Dt. 28, as
mentioned in Dt. 11:29 ("wntt )t hbrkh" -> "you shall pronounce
the blessing").  Furthermore, from the fact that Dt 28:10 says
"ky $m yhwh nkr) (lyk" -> "that the name of God is pronounced
upon you" we see that the name must be pronounced.  By 
comparison with Ex 3:18 "yhwh elhy h(bryym nkrh (lynw" ->
"God of the Hebrews is pronounced upon us" [Stenhouse
translates as "met", but it appears to me that the Samaritans
did not make this distinction and read it as a spelling variant
for "nkr)"], and this evidently follows Ex 3:14 where God
reveals his divine name to Moses.  So too we find that the
blessings of God to Israel must require the pronounciation of
God's name, in Nu 6:27 "w&mw )t $my (l bny y&r)l" -> "and
they [the Priests] shall pronounce My Name upon the
Israelites."  Elsewhere pronunciation of God's name is
forbidden, in Lev 24:16 "wnqb $m yhwh mwt ywmt" -> "one
who pronounces the name of God shall be put to death."
This choice to have the Blessings be read with the
pronounciation of God's name is the meaning of "the Chosen
Place" whereby Dt 12:5, 12:11 "hmqwm )$r ybxr yhwh ... l&wm/
l$kn $mw $m" -> "the Place where God will choose to place
his name there."  This argument is further supported by Ex. 20:24,
where "[bmqwm] )$r )zkyr )t $my )bw) )lyk wbrktyk" -> "In
that place that I will have my name pronounced, I will bless
you" (this is one verse that is listed as a sectarian difference from
the MT "in all places", but I don't know the original hebrew text of 
the SP in this case, so I put bmqwm in brackets.  The verse also 
mentions an altar, but here, at least in the MT, it is an earthen altar 
whereas in Dt 27 it is a stone altar).  Hence, the argument goes, it is
clear that God intended the blessings, altar, pronounciation of the 
Divine Name, and Temple all in one place, and that is Mt. Gerizim.

As can be seen, Abdal's argument above makes several
1) The SP mentions the Chosen Place in the imperfect, "the
place that God will choose."  This occurs several times.  
2) The SP apparently has no direct commandment for
observing God on Mt. Gerizim, or else the entire "convoluted"
argument would be unnecessary.

Furthermore, two statements, one by "Abdal" (the Samaritan)
and one by "Zerubabel" are significant:
1) Abdal says that the verse "the place which God will choose
as a dwelling for His Name" is "as all shades of opinion and
belief will agree" of necessity the place where the Tabernacle
must be erected, so that sacrifices (and hences the calling of
the Name of God) will be continously performed.
2) Zerubabel, upon taking the scroll of the Torah, says "I 
cannot throw it [into the fire] for .. this book is mine and his,
because the one who wrote it is God, the Messenger Moses."

The argument in front of Surdi therefore makes clear two points,
one a negative mention of a very clear SP passage (the 
decalogue commandment) and another, the use of the
imperfect whereas today's SP has the perfect ("the place which
God has chosen.")  It further makes the claims on these verses
by both sides, Jews and Samaritans, that they are working with
identical manuscripts.

This leads to the question, how can the manuscripts be identical
if it is clear that the altar is to be erected on Mt. Ebal in the 
(Jewish) MT and Mt Gerizim in the (Samaritan) SP.  Perhaps this 
is to be answered in a text some manuscripts of Abu Fath have 
at the end of the Surdi debate story: "after that time, the Jews 
began to alter the Torah."  (Notes 355 and 330 in Stenhouse are 
signficant: Abu Fath later mentions [AF81, passim] that Ezra 
changed the reading to Ebal in Dt 27:4)  Hence the story is 
operating in an environment where the MT and SP differ on the 
question of Joshua's altar, but not on the absence of Gerizim 
from the decalogue or the tense/aspect of the Chosen Place.  
The Ebal difference is explained as something Ezra did later on, 
and therefore had yet to happen in Surdi's time.  So even if during 
the story's composition the texts differed on this one point, the 
story assumes that during the time of the Surdi debate both texts 

It appears then, that during the original authorship of the Surdi 
story, the MT and SP differed only on the question of the place
of Joshua's altar.  If based on various Arabic terms ("Qibla",
for example), we date the story to 7th century or later, we must 
conclude that as late as the 7th century the SP and MT matched
on the issues of tense/aspect of the Chosen Place ("will choose")
in various places of Deuteronomy (Deut 12:5, 11, specifically) and 
the absence of Gerizim from the Decalogue.

This is evidently supported by archaeology which is inconclusive
on this issue prior to Medieval times.  Against this background, 
the story of the debate of Surdi might be explained.  Having
no commandment specifically pointing to Mt. Gerizim, but having 
a commandment to build an altar "$m = There," Abdal still needed 
to prove that "There" is Mt. Gerizim.  It further suggests the 
increasing importance of these specific verses (Deut 27:4-7) which 
appear to me to have found their way into the Samaritan Decalogue 
during the late first millenium CE.

The recent finds at Mt. Gerizim are also signficant.  I have tried to
suggest on the Aramaic list:


that the Israelite dialect was more "Aramaic-like" based on these finds.
So far I have received no response to this query, so maybe I am wildly
off track.  Also, upon reflection, it seems to me that perhaps the Mesha
stele which also has the $t spelling for year, and another "Aramaic-
like" attribute (-n endings, for example), might be closer to the Israelite
dialect.  In the end though, it seems that from what little epigraphic
evidence we do have, we know of an ")$R" being used only in the south 
and "zy" being used only in the north.  This would suggest that the 
phrase "BMKWM )$R [Y]BXR L&WM $MW $M" (= the place where He 
[will] choose/chose to place his name") -- with its relative pronoun ")$R" -- 
is not a product of the Mt. Gerizim priesthood and probably not a product 
of the Samarian community.  

In relation to this, I also found the following message by Dr. Lehmann:


This may also have bearing on the "zy" inscriptions that read in part:
1) zy hkrb )l(zr / (l )mh )ntth wb[nwhy]
2) zy hkrbt mrym (l np$h w(l / bnyh
3) zy hkrb dlyh br $m(wn (lwhy w(l bnwhy )bn[hdh ld]krn +b qdm )lh) b)tr) dnh
4) ...] wpryn kl [.. / db]x bbyt dbx) [... / ...)]nh mhr d)[...
(note the shift *dh -> d in the last case)

I hope all this helps you in your research.  Again, I am no linguist and
have no formal training.

Be well,
Yitzhak Sapir

More information about the b-hebrew mailing list