Louis Ginzberg.

Geonica (Volume 1) online

. (page 14 of 20)
Online LibraryLouis GinzbergGeonica (Volume 1) → online text (page 14 of 20)
Font size
QR-code for this ebook

'Amidah prescribed by Rab Amram for the Ten Peni-


tential Days (^''^, I, 45) proves the sentences Q^'^n^— OVn,
45 a, to be an addition from the Spanish Ma/izor.

The Benediction over the Shofar, in the copy of the
Seder Rah Amram used by Ibn Gajat, read J?lpni? {^''^\ I,
261), while our text offers the formula prescribed by
Rabbenu Hai. On the other hand, the ^)ipn i'n''i5i> in his
copy of the Seder had the words N^n myD yipni? Tin DNl.
One must despair of establishing the wording of this
Benediction original to the Seder.

The prayer n^^niN is properly missing in the printed text,
27, while the MSS. Oxford and Sulzberger contain it as
an addition from the Sephardic Mahzor. It is, doubtless,
of Palestinian origin, as can be seen from the Mahzor
Romania, where it has a place in the Daily Prayer.
Besides, the closing Benediction Tpn ^Jioi? ^"^2 is known
to us to have been used as such through a passage in the
Yerushalmi Berakot, I \ and accordingly belongs to the rem-
nants of the Palestinian liturgy, which have been preserved
among the Sephardim, Ashkenazim, and Italiani.

The words nm^pn Nnn^no n^t^ ^, on 47 a, make it seem

1 Not in our text of the Yerushalmi, but in the text used by the old
authors. Comp. Ratner, D^'^^riTi ;v:? n^n^?, 33-4. Maimonides also has
this formula, as well as Rabbi Saadia, in his 'Abodah given by Dr. Elbogen
in his Studien, &c., 122. Curiously enough, Dr. Elbogen overlooked this,
on p. 70, n. I.

2 Weiss (IV, 49) reproaches the Geonim for calling the Academy
ncjiipn m^uj^ However, it is not the Geonim who use the expression,
but the scholars outside of Babylonia (R. Ibn Abitur and Moses ben Enoch,
in y"ir, 4 d, 29 ; 30 a, 9) or the correspondents of the Geonim (b"j, 9). In the
latter passage, the question contains the words : n^^nprr ni'u:'2 yn:i2n ya,
while the Gaon's reply was the simple "i^arr^a. Likewise in 'jn'^ic, 172,
n\rTipn xni^n^'O is a remark made by the compiler of the Responsa.
In general, the Geonim either cite decisions by other Geonim or the
custom of the Academy, but never a decision of the Academy, which,
indeed, would have been odd coming from a Gaon, as all decisions were
supposed to be issued by him and not by the Academy. In nV'oj, 44,
n^npn rrnTn?2n in Rabbi Sherira's reply is only a verbatim repetition of the
expression employed by the questioner. It is interesting that in the Re-
sponsum by Rabbi Sherira and Rabbi Hai jointly, found in the Responsa
Collection of Rabbi Solomon Ibn Adret, V, 25, a-b,n. 121, the question
contains the expression nvimpn nr^^^n, while the answer has instead of it


very likely that m: i>D was missing in the original Seder,
for these words were never used by the Geonim. If,
besides, we take into consideration that mj ^3 was un-
known in Babylonia, as we are told by the Geonim of
Sura and of Pumbedita without a dissenting voice ^, the
probability of its not having formed a constituent part
of Rab Amram's Seder rises almost to certainty. There
would be no explanation to offer for Rab Amram's pro-
cedure in first putting it into his Seder, and then character-
ising it as a " foolish custom." We probably are troubled
by two additions derived from different sources. The
first addition, the mj b^ itself^ came, in all likelihood, from
the Seder of Rabbi Saadia, and to this was joined, as a
second addition, the disparaging criticism upon it made
by Rabbi Natronai, introduced by the words NniTiDD n^c

To the Spanish Ma/izor, again, the prayers N'in''1 rhv'' and
life, on 48 a, are attributable. As we learn from explicit
statements in Ibn Gajat, t^>'':^, I, 61, and ManJiig, 60, it was
Rab Amram's opinion that these prayers were not to be
said on nniMn ni\ The author of the Mcmhig, and Abu-
draham as well (133), add that none but the Spanish rite
differs from Rab Amram. This point affords a striking
illustration of the heedless way in which the copyist to
whom we owe our text set aside the real Seder of Rab
Amram. On 47 a, where a list of the initial words of the
prayers for a'':3nv is given, he followed his model implicitly.
There he included neither nW nor 'jii'D. But two pages
further on he could not refrain from setting down what
he was accustomed to connect with the services of the day.

Our text contains no alphabetical ^itD^ bv, yet Abudraham,
153, cites one from the Seder Rah Ainrain.

The prayer for a mother on the day when the child to

^ Comp. ©"u:, I, 60-1. Eabbi Saadia is the only one who knows
Kol-Nidre, whence it follows that it was of Palestinian origin, as the
Seder of Rabbi Saadia follows the Palestinian customs closely ; comp.
below, pp. 166-7. Concerning the opposition of the Geonim to nm: msn,
comp. above, p. 96, n. i.


which she has given birth is circumcised, 52 b, is a later
addition, as was proved by the present writer in the
Z. H. B., IX, 106. The Geonic sources mention a prayer
for the child, but none for the mother.

The Benediction to be said at the circumcision of a
proselyte, and of a slave, as set down in the Manhig, 98 b,
from the Sede7\ is not in agreement with our text, which
should probably be modified according to the Manhig.

In view of all the passages instanced, it would be a
wilful perversion of judgment to make an inference regard-
ing the nature of the Geonic liturgy from the recensions
of the Seder at present available. Our printed text cannot
be looked upon as anything more or less than a Spanish
Order of Prayer with some additions from the real Seder
Rah Amram. The same characterisation applies also to
the MSS. Sulzberger and Oxford ^ though they deviate
here and there from the printed text. Of the two MSS.
the Oxford apparently is a more recent version, the in-
sertions in which may have been taken from the Seder
of Eabbi Saadia. This supposition is strengthened by the
long passages, given by Marx, Untersuchiingen, &c.,
Hebrew part, 4, 6, 18, which are said expressly to have
been derived from Rabbi Saadia, and p. 11, relative to
n^i, which is quoted by various authors with the name
of Rabbi Saadia attached to it ^. The grace after meals
in the MSS., having the same wording in the two, is also
more recent than the printed text of the prayer, as is
shown by b^i6 pr»n n:3"in given at the end of the Seder.
The prayer after ]'^r] pn^* in the Oxford MS. is doubtless
a later addition 2. Rab Amram would scarcely have sent
the Spanish congregations more than the main prayers.
Hence the difference between the forms of the nijnn in

^ I have given the prayers in them only a cursory examination, but
I am convinced they agree with the printed text in all essentials.

' Comp. Miiller in (Euvres Complets de E. Saadia, IX, 156.

2 The sentence (28) V7 innc . . . . 'Sf^bo occurs almost literally in an
epitaph at Brindisi, of the year 833, published by Ascoli, Inscrizione, 66.
Comp. also n"j^ , II, 635.


the printed text and the MSS., as none of them were
contained in the original Seder Rah Amram. There
is, of course, no need to lose time in adducing proofs that
the addition to Mslimat in MS. Oxford (24) is a late
insertion, nor that the extracts from the Hekalot, to be
found only in the printed text, most probably were not
of the original constituent parts of the Sede7\ It is sig-
nificant that while the Oxford MS. has no nn^'^ in the
Week-day Service (p. 3), it has it in the Sabbath Service
(13), exactly the reverse of what we find in the printed
text. As has been demonstrated, Rab Amram did not
have the n'^^l:^ in his Seder.

The Halakic Part of the Seder Rab Amram.

It now behoves us to explain how it happened that of
all old works the Seder Bab Amram was subjected to such
peculiar treatment. Like the others it suffered additions
to its essential, original form. But that is not all— the
essential original form itself was not left intact, it was
so modified, abridged, and extended, that we now have
very little of what it was in the first place, when it left
the hands of Rab Amram. Prayer-book making among
Jews is a wholly modern trade. Rab Amram did not, by
any manner of means, write a prayer-book. He merely
sent the Spanish congregations the prayers prayed in
Babylonia, well knowing that, to use a Talmudic phrase,
" every stream has its own current." He had no intention
of forcing Babylonian rites upon Spanish congregations.
Incorrect readings, which had crept into some of the
prayers in the course of the centuries, were rectified in
the Halakic notes accompanying them, and at the same
time the notes served to state the principles which had
guided the Tannaim and Amoraim in settling the form
of the prayers, and which still were to be applied as norms.
These explanations of the Gaon subjoined to the prayers
were the important part of Rab Amram's Responsum for



the Spanish Jews. There was no disposition on the part
of the latter to abolish their local rites, but when the
congregations had differing customs, or in doubtful cases,
the directions of the Gaon were resorted to, consulted,
and applied. The main task of the copyists, employed
by those interested in spreading the Seder, consisted not
in reproducing the prayers, but in recording the Halakic
directions and the important variations from their prayers
given by Rab Amram. In this way we have come into
possession of Spanish prayer-books embellished with ad-
ditions from Rab Amram' s Seder as well as his Halakic
instructions. Similarly, the Germans had their nil DmrriD
niDV ^, prayer-books embodyiug their liturgy together
with the Halakic portions of Rab Amram's Seder '^.
Of the same class is the Mahzor Vitry, which contains
the major part of the Halakic element of the Seder, but
in the prayers themselves it follows the French ritual.
In view of the close relation subsisting between some
of the prayers and the Halakot accompanying them, it
may be assumed, without further evidence, that the Spanish
congregations, and here and there others as well, yielded
to the great authority of Rab Amram, and made changes
in their liturgy in consonance with his directions, such
as the excision of the reference to the Messianic redemption
from the Geullah, which, as was demonstrated above, existed
in the old Spanish forms of the prayer. Occasionally,
compromises must have been made between the local
custom and the version recommended by the Gaon. When
we find the Sephardim using nnD for the Musaf Kedushah,
and "js^npj for the Kedushah of nnnii', it is fair to conclude

^ i"is, I, 26 b.

2 Naturally, many a Halakah was given a place in the Mahzorim that
had the sanction neither of Rab Amram's name nor any other Gaon's.
Hence, quotations from the j?""© in the works of the German authors
that cannot be traced. For instance, a contemporary of Rashi's grandsons
C?") ij'n 'ttjn, 3) cites the rhu niDbn of Rab Amram, of which not a trace
can be found in the 3?"nD, and probably it never existed there.



that we have an instance of an attempt at amalgamating
different rituals *.

The influence of the Babylonian ritual must, therefore,
have been strongest in Spain, whither the Seder was first
taken, which, however, did not hinder it from asserting
itself among the Franco- German Jews. In pursuing this
line, it must be borne in mind that frequently what is
denominated the custom according to the Seder Rah
Amram is nothing but the old Spanish rite, which agrees
with the old Ashkenazic rite, both derived from Palestine 2.

While the liturgical part of the Seder was badly abused
by the copyists, the Halakic part has reached us in com-
paratively good condition. After what has been said, the
reason is patent. The prayers the copyists knew by heart,
and they paid little attention to their model. They wrote
as their memory dictated. Besides, they knew that the value
of their work was concentrated mainly in the copying of the
Halakot. To these they therefore devoted conscientious
care. It was inevitable, of course, that in spite of all
attention, even this portion of his Seder should receive
additions from other hands than Rab Amram's, and, also,
several Responsa by him, which he seems to have addressed
to Spanish scholars independently of the Seder, have been
inserted at suitable places. For instance, the Halakot on
pp. 2>6 a-b, bearing the name of Rab Amram, are abstracts

^ Comp. G. S., p. 49.

^ The great respect enjoyed by the 2?"-ic among the Franco-German
Jews is apparent from the words of Rabbenu Tam, in -iu."n '-, ed. Rosenthal,
^, in which he maintains that the Seder was the chief source for the
prayers. Rapoport, p3 't 'n, note 29, goes too far, however, when he says
that the Germans were the only ones to accept the Seder Rab Amram,
excluding the Spaniards as he does. Yet his instinct was correct in
laying stress upon the influence exercised by the r^iD upon the German
liturgy. In his polemic against Rapoport, Weiss, Dor, IV, 121-2, is less
close to the truth when he deduces the dependence of the Sephardic
ritual upon the 5?"nD, from the agreement between the former and our
text of the Seder. We have seen that the relation is exactly the reverse.
Furthermore, Weiss is mistaken in calling Maimonides' Seder Sephardic
it is Egypto-Palestinian.


of Responsa of his addressed to the congregation of Barce-
lona, y'a, 56-7. Ibn Gajat, ^"^^ 1, 10, and Rabbi Abraham of
Lunel, Manhig, 26 a, quote these passages, but it is doubtful
whether they knew them from the Responsa as such, or
from the Seder '^. The sentence DiT^a — 131, 37 b, did not
occur in the Seder used by the author of the Manhig (43 a) ;
it is obviously a gloss calling attention to a Responsum of
Rab Amram's, which gives support to this peculiar custom
by means of the authority of the Yeshibot and the Geonim ^
— authentication which was all the more pertinent, as not
only did the European Jews know nothing of the recital of
riirT'i'D on Purini, but also the Geonim themselves were far
from unanimity upon the point, as appears from Tur, Orah
Hayymi, 693. It would seem that the custom prevailed
only in Sura ; in Pumbedita no nirfi'D were prayed on
Furim, So far as Sura is concerned, the testimony of
Rab Amram is reinforced by the fact that Rabbi Saadia
himself composed nin^i'D for Purim, published by Professor
Schechter, Saadyana, 49-50. There is the possibility, of
course, that these nin^i'D may have been intended for n^:v^
nriDN. The sentence nu^sj^'' — s'^VN*, on p. 32 a, is taken
from a Responsum by Rab Amram, quoted in full by the
author of the br]''2^, loa^.

Additions from the Seder of Rabbi Saadia occur in three
places in the printed text of the Seder Rab Amrarii, 4 b
(bis), and 52 a. But, as was observed above, the MSS.
do not contain the first two insertions, and as for the
third, we know that it did not appear in the copy used
by Ibn Gajat, as can be inferred from his words at the end
of ^''^, I. It seems to have been taken from the D'-^n ninnx,
26 c, which cites the opinion of Rabbi Saadia in opposition
to Rab Amram's.

^ In bT}"ya^, 54, it was doubtless taken from a Responsum, and not from
the Seder.

^ By a slip the author of bn'^a^", 157, writes ;\s'ira«i J•'^<:n! For the
meaning to be attached to yazn in this sentence of Rab Amram's, see above,
p. 24, n, I. On p. 29 of the hn'^ixo it has, properly, j\v:n without J'niton.

^ Comp. also Hazan, c^n ••>« , 45 a.

L 2


If the superscription (14 b) KnnnDiQ — nD^f 31 is correct,
then, naturally, we are dealing with an addition, as it is very
improbable that Rabbi Zemah could have been quoted by
Rab Amram. But one cannot help being assailed by
doubts as to the correctness of the superscription. It
is not impossible that the abbreviation 'v '"i, standing
for pXi'H m, was improperly interpreted as n»v 'n, and
then, to complete the verisimilitude, Nnnn^iQ N^i was added
after nyi:^\ as Rabbi Zemah was Gaon at Pumbedita.
Originally, it must have read npy^ JINJ nrc^^ CJ'Nn V''-|\
without specifying the Academy. As was demonstrated
at length above, only the heads of the Sura Academy bore
the title Gaon. At first, and even later, when the heads of
the Pumbedita Academy were already called Geonim, a
distinction was still made between the 2\>v^ pN3 ny^'' c^xn,
the head of Sura, and the Gaon of Pumbedita, who were
only ni?^: b^ na^CJ^M C'i^l. As early as Talmudic times
(Rosh ha-Shanah, 23 b), n^ia was synonymous with Pumbe-
dita ^. Later copyists, especially those in countries remote

^ Dncn "^ipS 9 a, has the reading no!? pni"% plainly traceable to the
abbreviation :j"-\, for which the copyists had two explanations, pni*'' 'i
and r\i2-j 'i . That ^^nnims ^^'n nn^c^'i is a later addition is confirmed by
'7"iD->r'X, I, 33, where it does not appear. The names prrr'' and p"n2 are often
confounded. Comp., for instance, Mekilta, Jethro, I, and Si/re, Deut., 38.
In both places pnr is to be read instead of pni?% as appears from Kiddushin,
32 a. The name of the Gaon Zadok is misread for \irT^' in D^rr, 56, n"K, II,
414, '5n"rr, 211, and in many other places. Comp. also Zunz, Gesammelte
Schriften, TV, 274. MS. O reads htu?''! pt<j nQ!J ai.

2 What Maimonides (Commentary on Bekorot, IV, 4) has to say on the
use of these two titles at his own time is particularly interesting. He
informs us that while ipi""" ]Miy nrtr' rxi was used in Palestine, the
Babylonians bore the title nbi3 Stt? nn>\r>'' VDi^-i. The reason for the differing
practices is obvious. In Palestine they tried to perpetuate the original
title of the Gaon, while in Babylonia the title of the head of Pumbedita
was continued, as this Academy survived that of Sura by two generations.
The Hebrew text of Maimonides is corrupt. It reads nirorr y-is instead of
bii. The Arabic text published by Lowenstein, Berlin, 1897, p. 22, has the
correct reading psny'^j^, and the same is to be found in the MS, of the
Arabic text of the Maimonides commentary in the Jewish Theological
Seminary of America.


from Babylonia, did not distinguish the Geonim from each
other by their exact titles. The mistake having been made
of reading v''"i as r\J2)i ^1 , the expression :3pr |1n: r\2^^'> K^NT
was retained, while the words Nn^nnniS N^^ were added :
they bear plainly the earmarks of an explanatory gloss.

The same error of interpreting an abbreviation incor-
rectly may have changed ^J<jntDJ 'n, 4 b, into \)^r]: 'n^.

It is a vexed question, the identity of the Rabbi Nathan
mentioned three times in the Seder, 35 b, and 37 a (bis).
In the last two places he is called nn''t^^ ^a,^, both in the
printed text and in the MSS., which gives no encourage-
ment to his identification with the uncle of Rabbi Sherira,
Rabbi Nathan ben Rabbi Judah. The latter was no tJ'N"!
nn^C^"', only an ^ji^js, and if the copjasts had desired to
confer a more distinguished title upon him, they would
have called him Gaon, the usual epithet bestowed later upon
a very prominent scholar. But there was no Rabbi Nathan
who was a ny^'' ^i^l in Babylonia, and we have the choice
of again resorting to a falsely interpreted abbreviation,
and putting ]r\: for ''XJnD:^, or identifying him with the
contemporary of Rab Amram, the Rabbi Nathan of Kair-
wan, who was a nn'^K^'' t^sn in Kairwan^. The difficulty
of identifying this Rabbi Nathan is increased by the fact
that Abudraham, even in his first edition (Lisbon, 1489),
twice has pm 'n in citing the Seder. In the first passage,
p. 79, pni is probably a mistake for Amram, while in the
second, p. no, corresponding to 37 a of our text of the Seder,
the dictum ascribed in the latter to Rabbi Nathan, is quoted
in the name of Rabbi Nahman. But pn: would seem to go

^ Comp., R. E. J., LIV, 204, where this passage of the S'""id is quoted,
but without the name of Rabbi Nahshon. There is no reason for doubting
that it is taken from the Seder.

^ An interesting example of mistaking :"i = •'SDir^: S and :"t = jn: 'i
for each other is afforded in Tur, Orah Hayyim, 190. It occurs in the
first Soncino edition, and in all following editions, while ed. Mantua,
1475, has 'SOITC: '"\ as is proper, and as is confirmed by Q"n, 187 ; for
indirect testimony by Rashi see above, p. 43, line 6 from below.

^ Comp. above, pp. 31-2.


back to pcj'm, rather than to in:. The name of the Gaon pc'nj
is elsewhere found corrupted into pn: ^ Thus the reading
in: becomes very doubtful. Besides, the decision given on
37 a in the name of Rabbi Nathan offers a difficulty in
the subject-matter. It contradicts a usage prevailing in
the Yeshihot, if we can put implicit confidence in the words
of Rabbi Natronai, s''n, 187. The last point may be
adduced in support of the assumption that the authority
referred to is Rabbi Nathan of Kairwan, who recorded
his opinion here at variance with that of the Babylonian

Apart from these additions, which can be attributed to
definite authors, there probably are a number of anonymous
passages in the Halakic part of the Seder that did not
belong to it originally, but were inserted in the course
of time. For instance, it is not at all likely that the
references to the Spanish ritual, i a and 2 a, were made by
the hand of Rab Amram himself 2. The expression ^i^D ^D
nmt^ni mi'NSi'n in the latter place is not a Gaon's way of

^ Comp. Eapoport's Introduction to p":, 9 b, and also m^nj, 47, where^
likewise, pirn: is to be read instead of pn;. The first edition of Abu-
draham reads p: instead of pn3 in n-rn 'rr, 135, in agreement with y"-\c,
35 b, while all the subsequent editions have pn: \ Schorr, He-Haluz, VII,
144-5, insists that there was a Gaon by the name of pn^, though none is
mentioned by Eabbi Sherira in his Letter. By way of corroboration, he
adduces the fact that Rabbi Sherira has no reference to the Gaon Rabbi
Menahem, of whose existence there can be no doubt. Schorr evidently
was carried away by his opposition to Rapoport. In point of fact, the
Gaon Rabbi Menahem is mentioned by Sherira. h^^dto is out of the ques-
tion, the only Gaon by that name, the son of Kohen-Zedek, not having
written any Responsa. In Abudraham, 139, the end of r"ic, 35 b, is
also given in the name of }n: 'i, but this can scarcely be correct, as in
•D-i"n:i, 125, and c^^'n 'm«, 90a, the same passage is ascribed to Rabbi
.Tehudai, whom Rab Amram followed here as in many other places. "I'V,
21 1 , has p: sm which seems to corroborate our assumption that R. Nathan
was not a Gaon, wm is never used in connexion with a Gaon.

^ Also lines 14-17, on p. 5 b, seem very suspicious to me. On the use of
T:Dt-«, Germany, comp. the Responsum of Rabbi Paltoi in '2i"'2j, 149. where
C"i::;"Ci< are mentioned.

the halakic literature 151

Relation of the Manuscripts to the Printed Text.

The fact that an old work has been subjected to additions
does not preclude the possibility of its having suffered
abridgment as well. However, it lies in the nature of
these Halakic expositions to give suitable opportunities
for additions, especially extracts from Geonic Responsa.
It may, therefore, be formulated as a rule, that only the
material common to the printed text and the MSS. can
with certainty be considered as originally part of Rab
Amram's work. Accord between the MSS. and the old
authors is not in itself conclusive as to the genuineness of
the passages found in them. At most, it proves that such
additions, if additions they be, were made in a remote
time. And in point of fact there are but few additions
in the Sulzberger and the Oxford MSS. that cannot be
followed up in one or another old author. Some of these
parallel sources to the MSS. of the *S'ec/er Rab Amram
follow : —

The resume of the ni3n2 r\i{i2 in S and O, i, is met again
literally in Mahzov Vitry, '^-^, and an abstract of it, in
D>K-i^ 'D, ed. Schiff, II, 2'^S' Besides, the conclusion |ND1
innm is cited in the Manhig, 7 b, from the Seder. Never-
theless, it does not seem at all probable that Rab Amram
would give a summing-up of the niDnn nxD sent by his
predecessor to the Spanish congregations not very long
before his own Responsum.

The regulations regarding the benedictions over the
Tefillin, the Mahzor Vitry had in the copy of the Seder
used for it, in agreement with 0, 2, as appears from the
remark of the author on p. 642, while the Manhig, 7 b, i&
in accord with our text ^

Mahzor Vitry, ^-^^ ^^^ ^^^ section n^p^mh — Nn::05r in
MSS. S and O, and also 1>T — ND^:, found only in MS. S.

1 2 3 4 5 6 7 8 9 10 11 12 14 16 17 18 19 20

Online LibraryLouis GinzbergGeonica (Volume 1) → online text (page 14 of 20)