After a couple weeks in the Near East, I am back in Euorpe, and have caught
up on the discussion of the article on Shem-Tob.
Below are some comments...
First, re Jean Valentin's post:
>(1) LANGUAGE OF THE VORLAGE
>From which language was it translated? Was it directly from Latin, or was
it from some vernacular (like, >for example, the pepysian harmony which is
supposed to come from a lost French original)? The >variants give us
evidence as to the textual family it belongs to, we need more in order to
decide from >which language exactly this Hebrew version was translated. Dr
Petersen gives us an interesting >example of mistranslation or wrong choice
in the translation of a word (at § 89).
>I would like to add another example: at Mt 7.11 there's the use of the
preposition 'im in the phrase 'im >heyothkhem ra'im. This should be
translated, of course, "while you are wicked". But in Hebrew, 'im is a
>preposition meaning "with", and nothing else. So in Hebrew, this verse
simply doesn't make sense.
>What happens? In latin, "Cum" can be understood in two ways: (1)
preposition meaning "with" - this is >how it is translated, wrongly, in our
Hebrew text. (2) "Cum" can also be a subordinative conjunction, >meaning
"while", "as" and introducing temporal clauses. This is how it should have
been translated. >So our tranlator, having to choose between the
preposition and conjunction, took the wrong choice. >This is, I think,
another evidence for the fact that our Hebrew text was translated from
latin. At this point, >the sentence is nonsensical: this is a sure trace of
mistranslation, and we can demonstrate that the >mistake comes from not
understanding a latin word. Our text was not written originally in Hebrew,
but is >translated from Latin.
It would seem that Jean has answered his own question.
Latin seems the obvious choice becuase of the reasons stated in the
article: (1) we know the *Vorlage* of the Middle Dutch tradition: it was
Latin. (2) We know the *Vorlage* of Isaac Velasquez: it was, apparently,
also Latin. Beyond this, (3) the agreements of Shem-Tob's Hebrew Matthew
with the Vulgate and the Vetus Latina, as well as (4) the standard practice
of the time and place (the medieval West) point to Latin. This is all
supported by (5) readings in the text (cf. para. 89 in the article), of
which Valentin has apparently found another example (I do not have my copy
of Shem-Tob at hand, so I cannot check his example, but off the top of my
head, his analysis sounds reasonable).
>(2) NATURE OF THE VORLAGE
>It is nearly certain that the language of the Vorlage was Latin. But what
kind of text was it exactly? It >could have been a Diatessaron from which
the translator chose the pericopes that come from >Matthew (of course,
keeping the harmonizations as he didn't take the time to eliminate them).
It could >have been, as Dr Petersen seems to suggest, separate Gospels
heavily influenced by the >harmonized tradition. Or, third solution, it
could have been a lectionary - some show traces of a heavy >influence of
the harmonized tradition.
There are good reasons to exclude the first and third of Jean's
suggestions. As to the possibility that it was a Diatessaron from which
the translator excised the Matthean fragments: to "unbundle" a gospel
harmony, in an age and place where the separate gospels were circulating,
seems not only unlikely but a very obtuse manner in which to go about
things. It simply seems unlikely. Furthermore, if such a procedure were
followed, one would expect even more "confusion" than there is with the
other gopels, for the task of deciding where the text of Matthew would
begin and end, begin again, and then end again, in a pericope would be a
very difficult and imprecise task. As opposed to this, Shem-Tob's Hebrew
Matthew seems--by and large--a copy of Matthew, but one heavily influenced
by the harmonized gospel tradition.
On the third possibility, that this is a lectionary text: yes, some
lectionaries display harmonization, but no lectionary rubrics are evident
in this Hebrew text (at least to the best of my knowledge); also recall
that lectionaries don't contain the entire text of a gospel, while--albeit
with some omissions here and there--Shem-Tob's Matthew seems to be giving a
more or less complete text of Matthew.
Because of these and other reasons, the second possibility--a text of
Matthew, heavily influenced by the harmonized gospel tradition--seems most
likely to me. Remember that we are not speculating about the existence of
such a text in Latin, for Isaac Velasquez *knew and used* such a text *in
Spain* precisely *in the time-frame* we are targeting.
>I don't have a very precise opinion as to how we should study the matter
in order to decide, but let me >try a suggestion. The Hebrew text of Ben
Shaphrut is divided into peraqim - paragraphs. Do these >correspond to
divisions in Latin manuscripts - it could be the sections of the latin
diatessaron (the >Dutch harmonies are also divided in paragraphs roughly
corresponding to pericopes or narrative >units) or the pericopes of a
lectionary. If a specialist in the Western liturgy could tell us if the
peraqim of >our Hebrew text correspond to one of the Western rites of the
Middle Ages (roman, mozarabic rites...), >this could help us to determine
the origin of this text and the process by which it was translated.
I'd be happy to have someone check this out. But recall that the liturgies
will not give all the text of a gospel, and Shem-Tob seems to give
more-or-less most of Matthew. The paragraphs in the Western harmonies do
not--again, to the best of my knowledge--mirror any liturgical pattern.
Codex Fuldensis--in Latin, and our oldest extant harmony in the West--has
182 capitula, if I recall correctly (Ulrich Schmid worked on this at NIAS
last year: Ulrich, any comment?). The Liege Harmony has well over 200
"chapters," if I recall correctly. The only western harmony with obvious
litrugical links is the Middle English Pepysian Harmony, which, at least in
my researches, has very little in common with the Hebrew Matthew in
Shem-Tob's *The Touchstone.* (The Pepysian Harmony has some rubrics here
and there: "The reading for the third Thursday," etc.)
OTHER REMARKS and informations:
>- At § 38, variant 5, Dr Petersen points to "being built" instead of "set
on". I have met this variant in an >Arabic version from the Middle East
based mainly on the Byzantine text (but with many other >influences) - the
one I'm studying for my thesis: ms Sinai Arabic 69, from the XIth century.
Again, I am without my books at the moment, but my memory tells me this
reading is found in several other sourcesas well: check the specifics in
Appendix III in Quispel's *Tatian and the Gospel of Thomas* (1975).
>- At §80, variant 2. As Dr Petersen cites this variant following the
apparatus of Legg, I just checked >directly in the Georgian editions, and
can confirm that the citation is exact (I always prefer to have a >look
directly at the versions when I can, you never know...).
Good for you, Jean: always go to the original sources, if possible.
>- At § 95, the variant of Mt 25.6. My Arabic version from the XIth century
has this too: fa-Hiina kaana niSf >al-layl idha biSawt yunaadii... "and
when it was half of the night, behold, a voice announced..." The >variant
is no more a unique agreement between Liege and Ben-Shaphrut.
>The variant is also present in another Arabic version about which I posted
to this list some time ago: >that of ms Sinai Arabic 71, from the Xth
century: wa-fii niSf al-layl abadan Sawt yaquul "and at half of the >night,
suddenly (litt: always = ?) a voice was saying..."
>These Arabic versions, though translated from Greek, have many common
variants with >diatessaronic witnesses and eastern versions. They being
eastern witnesses, could we count the >variant as diatessaronic as it is
supported now both in East and West?
This is interesting. One reading, however, can't tell us much. One would
have to do a much more thorough analysis. When I stumbled across the
reading, interpolating "voice," I immediately thought of my Hebrew lessons
years ago, where "voice(s)" of god/the prophet/whomever was/were "crying
out" all the time. The reading seemed very "Semitic" to my nose. So I am
not surprised to learn that a parallel has turned up in the East. The
question, however, is the origin of the variant. On that point, I don't
know enough about this Arabic MS (Sinai Arab. 71) to give an opinion. As I
mentioned at the beginning of this paragraph, a much more thorough study of
this MS would be necessary before one could offer any intelligent comment.
A few general observations may be useful, however.
First, recall that there is a close link between the Arabic translations of
the gospels and the Syriac (cp. Anton Baumstark, Curt Peeters and B.
Levin). You mention that this MS (or its Vorlage) is translated from the
Greek; is that certain? And how profound is the Syriac influence on this
MS? This might be the genesis of the "Diatessaronic" readings in Sinai
Arab. 71. Finally, before claiming bilateral support (East and West, and
hence possibly a Diatessaronic reading), one must be certain that Sinai
Arab. 71 has been influenced by the Diatessaron; that is not yet clear--at
least to me. The influence, if it does exist, would be very "faint" and
"distant" if, as you suggest, the MS itself (or its Vorlage) were
translated from Greek. I would be cautious about claiming Diatessaronic
origins for this reading, for it is entirely absent from all the major
witnesses, both East and West, *except* for the Liege Harmony (late) and
now, MS Sinai Arab. 71 (also late, and with a Greek Vorlage, according to
you)... This is *not* horribly impressive to me. Now, if the reading were
also in *Ephrem* (4th cent.), then the cards would be stacked
Nevertheless, an interesting parallel.
>- At Mt 16.13, the name Philippus is written FYLYBWS in Hebrew. This is
interesting, as this reflects the >Arabic orthography of that name (after
all, we're in Spain!). In Arabic, there is no double pronunciation >of the
letter phe (as is the case in Hebrew). It is always pronounced "F", and the
sound "P" of the >Western languages is transcribed by the letter "B". In
older Hebrew, there would have been a "P" in >both positions (look at
Jastrow's dictionary, where several transcriptions of that name is given at
the >article PLYPA page 1182 left column). This is at least one trace of
the influence of Arabic civilization >(and grammatical tradition!) on our
Interesting; as you note, however, we are in Spain, in Moorish times.
>- About the quality of Howard edition and of its revision. In the first
edition I had found not a few typos. >Some have been corrected in the
second edition, others not. Here are those that weren't.
>* Mt 6.22 third word: read 'eyNeykha, with a nun instead of the second
>* Mt 15.7 first word: read hwy instead of hyw (correct in the apparatus,
but not in the text).
>* Mt 17.1 first word should read "aHar" (with heth, meaning after) instead
of the obscure "atar"
>* Mt 20.14 last word, erase the taw and read "kamokha" (like you) instead
of ke-mothkha (like your >dead?).
>* Mt 20.15 first word after the question mark: replace the daleth by a
resh and read "ha-yera'" (is it bad) >instead of "ha-yeda'" (does he know).
>And I didn't note them systematically! So I'm lee enthusiastic than Dr
Petersen when he praises the >quality of the edition of the Hebrew text.
Your observations are absolutely correct. There are numerous typographical
errors. While one cannot always be certain that they are errors in the
text of the edition itself (although one can suspect, and be rather certain
that they are...), they are clearly present when one compares Howard's
lists of readings in his "Introduction" (e.g., his lists of parallels with
Thomas, the Vetus Syra, etc.) with his own Hebrew text in the edition.
They sometimes don't agree--meaning that one or the other must be a typo.
Thanks, Jean, for your very interesting and careful comments.
As for other posts: Several early posts completely misunderstood the point
of the article, thinking that I was arguing that the Hebrew Matthew was
dependent upon a Middle Dutch Vorlage. My thanks to those of you who
quickly intervened and pointed out that this was incorrect, and cited me at
length, showing that it was a common *Latin* Vorlage that was the source.
I must confess that I am puzzled how such a misunderstanding could have
arisen, for in the first and last paragraphs of the body of the article,
the following is written:
(paras. 1-28 are a description of the edition, the translation, and a
summary of Howard's analysis of the *waw-consecutivum*; only in para. 29
do I offer my first comment on the possible Vorlage)
Para. 29: "...these similarities in structure [may] stem from dependence
upon a common archetype. And since we know the Middle Dutch manuscript's
Vorlage, it suggests that if there is dependence upon a common Vorlage,
then that Vorlage is Latin."
This is the very first comment on the matter from your author, at the very
start of the body of the article, which is devoted to the matter of the
At the end of this analysis section (paras. 111-141 are the evidence of
Isaac Velasquez, "What can be learned," and the Appendix on Thomas and the
Italian Harmonies), in para. 104, the first paragraph in a section titled
"The Genesis of Shem-Tob's Hebrew Matthew," one finds:
"the tradition behind the Liege Harmony--which we know to be a Latin gospel
harmony--must also be the principal element responsible for the textual
complexion of Shem-Tob's Hebrew Matthew."
This sentence is even italicized. Clearer--at the beginning and at the
end--I do not think I could be.
Finally, someone asked if such Latin manuscripts actually existed. The
answer is already in the article. See paras. 48, 51, 54. See also paras.
--Petersen, Penn State University.