TWO NEW MAIMONIDEAN AUTOGRAPHS IN THE JOHN …

TWO NEW MAIMONIDEAN AUTOGRAPHS IN THE JOHN RYLANDS UNIVERSITY LIBRARY

By SIMON HOPKINS THE HEBREW UNIVERSITY, J E R U S A L E M

Among the treasures of the Rylands is the collection known as the Gaster Genizah Fragments. These once formed part of the extensive library of Dr. Moses Gaster (1856-1939) and were

. acquired by the Library from the Gaster family in 1954. l The size of this collection is larger than is often realized. It comprises 10,573 fragments classified as Hebrew and a further 840 classified as Arabic. Together this gives the substantial total number of 11,413 Genizah fragments in the Gaster c~l lect ion.~ Like all collections of Cairo Genizah manuscripts, the Gaster fragments represent an enormous variety of material of both literary and documentary character. A considerable number of the parchments may confidently be dated to the second half of the first millennium A.D., while some of the later paper items are scarcely, or less than, a century old. Many parts of the Mediterranean world are represented in the collection and there are also quite a number of fragments written in Ashkenazi hands. Most of the collection is manuscript, but printed material occurs as well. There is even a rare fragment of an Arabic block print. In addition to the preponderant Genizah languages, Hebrew, Aramaic and (Judaeo-)Arabic, fragments were noted in the collection in various other languages, e.g. Coptic, Judaeo-Spanish, Turkish and Yiddish.

The great majority of the items is made up of very small pieces. Thousands of those inventoried consist of mere fragments of single leaves. This state of affairs, of course, both increases the problems of identification and, at the same time, reduces the rewards of the scholar who comes to work on the collection. A good many of the fragments give the reader so little to go on that they are likely to remain unidentified for a very long time, and this is almost certainly the main reason why the collection

See BULLETIN, xxxvii (1954-55), 4-5. See F. Taylor, "The Oriental Manuscript Collections in the John Rylands

Library", BULLETIN, liv (1971-72), 456 and 463, respectively.

TWO NEW MAlMONlDEAN AUTOGRAPHS 71 1

has not been as heavily exploited as holdings of Genizah material in some other libraries. Nevertheless, even small fragments can sometimes be of significance and produce valuable surprises.

Thus, in December 1972 M. Beit-Arik was fortunate enough to light amongst the Gaster fragments upon a new autograph of Maimonides ( l 135-1204). A preliminary account of the important fragment of Mishne Tora discovered by him is given the BULLETIN.^ During a visit to the Rylands in March-April 1982 the present writer discovered a number of additional autograph fragments of Maimonides among the fragment^.^ It is hoped that all these autographs will be published in the near future in the Library BULLETIN. The present article deals with two of the new finds. Both are literary autographs and both are written in Judaeo- Arabic. The first'is a new piece from the Guide of the Perplexed, the second is from the Commentary on the Mishna.

I. Guide of the Perplexed (Book 11, Chapter 30)

Three portions of the (an?) autograph draft of Maimonides' Guide of the Perplexed are known. The first was discovered by H. Hirschfeld, who, at the beginning of this century, identified two leaves among the fragments of the Taylor-Schechter Collection in Cambridge. These leaves, covering parts of Book I and Book 11, were published, with photographs, in "The Arabic Portion of the Cairo Genizah at Cambridge. IV. Two autograph fragments of Maimonides' Dalilat al Hairin [sic]", Jewish Quarterly Review, xv (1902-03), 677-681. Two further leaves, this time giving a continuous portion of Book I, were found among the Mosseri fragments. One of the four pages was reproduced by J. Mosseri, "A New Hoard of Jewish MSS. in Cairo" in The Jewish Review, iv/21 (September, 1913), opp. p. 210, and all fo& were sub- sequently published by D. Yellin, "a"3n~;l >W 17- 3nDn 0-57 -IID",

Tarbiz, i (1929-30), 93-106. Finally, a third fragment was discovered by M. Lutzki in the E. N. Adler Collection in New York. This fragment, covering a section from the beginning of Book I, is reproduced by S. D. Sassoon in the Introduction to Maimonidis

lvii (1974-75). 2-6. The text itself remains, I believe, unpublished. It should be noted that the Gaster collection also contains several auto-

graphs in the hand of Maimonides' son Abraham (1 186-1237). As far as I am aware, this fragment remains unpublished.

712 THE JOHN RYLANDS UNIVERSITY LIBRARY

Commentarius in Misclznam . . . Vol. I, Corpus Codicum Hebrai- corum Medii Aevi, ed. R. Edelmann, Pars I (Hafniae, 1956), pl. xlvi, and again in Encyclopaedia Judaica, Yearbook 1977 8 (Jerusalem, 1979), 107.6

During a visit to the Rylands in the Spring of 1982' I was fortunate enough to discover a further fragment, or rather two further fragments, of the autograph draft of Maimonides' Guide. The new portion consists of two small pieces, Rylands Gaster B2597 and B4094, which fit together exactly and restore to us part of the author's draft of the thirtieth chapter of Book 11. The recognizable portions of the text correspond to ed. Munk (= below M) 11, 70a,9-71a,l. Two words on the margin of the recto correspond to M 69b,18. Since, however, the two fragments together produce less than half a single leaf (the bottom half), the middle part of this section is still missing (M 70a, 17-70b,6). Also missing is the portion corresponding to M 70b,17-21, though part of this passage may yet be detected in the faded first line of the margin of the verso. In Maimonides' full manuscript the new Gaster leaf (Book 11, Chapter 30) stood only four or five folios before the second of the two Cambridge leaves discovered by Hirschfeld (Book IT, Chapter 32). Comparison with the completely preserved Hirschfeld leaf and restoration of the missing portion of the text (i.e. M 70a,17-70b,6) make it clear that appioximately a dozen lines have been lost from the full folio of which the new Rylands Gaster fragment(s) once formed a part.

Maimonides' Guide of the Perplexed is unusual, if not unique, among medieval Judaeo-Arabic texts in that it has been trans- mitted in three different graphic modes. The author himself composed the work in Hebrew letters, and this is how it appears in the editions M, J and Q referred to below in the notes to the text. But the Guide was early on transferred to Arabic script and read in

The first two of these three autographs are well-known and have been reproduced on various occasions. E.g., a leaf from the Cambridge fragment, T-S 10Ka4.1 may be seen in D. Yellin and I. Abrahams, Maimonides (London, 1903), opp. p. 128, and another in The Jewish Encyclopedia, i x (New York- London, 1905), 75. Both the Cambridge and the Mosseri autographs are republished by B. Joel in the edition of the Guide referred to below as J. ' I would like to record my best thanks to Miss Glenise Matheson and the

staff of the Rylands Library for their help and efticiency in enabling me to look through the Gaster Genizah Fragments.

TWO NEW MAIMONIDEAN AUTOGRAPHS

Muslim circles too.s An edition of the work from an Istanbul manuscript written in 1477-9 in which both the Arabic text and the Hebrew quotations are given in Arabic scriptg was recently published by H. Atay (= below A).'O An intermediate stage in this process of graphic transfer is provided by a number of Genizah fragments in which the Arabic text is written in Arabic script, but the Hebrew quotations remain in Hebrew letters."

The text of the two new fragments is given below and both are shown adjacently aligned in the accompanying plates. The manuscripts are not well preserved, being faded, torn, blotched, crumpled and generally difficult to read. It is very likely that after appropriate treatment, particularly flattening out of the creases, more words and letters could be extracted from the fragments than are visible on the photographs published here. The new portion of the autograph draft of the Guide of the Perplexed given below is principally of interest as a further example of Maimonides' handwriting and method of work; textually, it is almost identical with the printed versions. The minor variations which occur are recorded in the notes.

See e.g. M. Steinschneider, Die arabische Lireratur der Juden (Frankfurt a.M., 1902), 204, 2 18 9 n. 12; D. Yellin and 1. Abrahams, op. cit., 156.

The use of Arabic script for writing Hebrew is a rare phenomenon attested principally, and perhaps exclusively, among the Karaites; on it see J. Blau, The Emergence arid Linguistic Backgro~trtd of Judaeo-Arabic2 (Jerusalem, 1981). 4 2 8 ; H. Ben-Sharnrnai in Studies in Judaica, Karaitica and Islamica, Presented to Leon Nemoy on his 80th Birthday, ed. S. R. Brunswick (Rarnat Can, 1982), 115ff. Among the Cairo Genizah fragments (including the Rylands Caster collection) there are a number of unpublished manuscripts containing Biblical verses in Hebrew and Aramaic written in Arabic characters.

'O Samuel b. Tibbon apparently used a transcript in Arabic letters Tor (part of) his Hebrew translation of the Guide, completed in November, 1204; see M. Steinschneider, Die Hebraeischen Uebersetzungen des Mitrelalrers ... (Berlin, 1893), 416. Whether this was completely or only partly in Arabic script is not known.

l ' E.g. three leaves of such a manuscript are extant in T-S NS. 306.252.

7 14 THE JOHN RYLANDS UNIVGRSITY LIBRARY

RYLANDS GASTER COLLECTION B2597 + B4094 Recto

a. [ l

KT. 3.n. [ I

a i l 2 in^ i y l n - D > [ I

nnpaosa K T W K K ~ - I > - I T V x ~ a y n i a ipnoo

K - W K [ - l 3 9 ~ 5 - U 7aa3 [ ~ i ] K;IIDKD[K] I n ;1upn>n 75

aipnon5u ;15Kn5u [vp=n ' 1 2 5 ~ D T Y X ? K laon] Margin

NOTES

Abbreviations

M = Le Guide des ~ ~ a r k s , TraitP de ThPologie er de Philosophie par Moise ben Maimoun dit Mairnonide, publit pour la premiere fois dans I'original arabe ... par S. Munk, 11, Paris, 1861

J = 13 ;lwn nm', (D->~xI ;mn 130) r i w n ' , ~ 3'7x3'1 i,,jLLI JYJ 3"i- n-31015-~ DO ,... p n ... nxna -D', -3iu;l irpn;l ,~in-n p.3111- ;l-rr;l- l"1;l +"', Jerusalem, 193 1

' 1 [ l 2. 31 4' 51 6 ' 71 8 91 110 JORK RYLAhTS LIBRARY CMS.

PLATE I : Ryl. Gaster Coll. B2597 and B4094, recto

l I

I i 1 2 3j 4 51 6 7! 8 91 1'0 JOHN RYI,A>iS LIBR4RY CMS.

PLATE 11: Ryl. Caster Coll. B2597 and B4094, verso

TWO NEW MAIMONIDEAN AUTOGRAPHS 715

Q = ,anin1 iiFn ,riqxn'lx 'i15x3-r ,a*z1131;l ;mn ,Iin9n p nwn 13-3-1 nDxp 71'1 i1';133 71~1- ... n-73~3 on-n, Vol. 11, Jerusalem, 1972

A = Delbler 'U I-Hairin, Filozof Musa ibn Meymun el-Kurtubi 1135-1205 ... Prof. Dr. Huseyin Atay, Ankara ~niversitesi Illihiyat Fakultesi Yay~nlari 93, Ankara, 1974 This book contains an edition of Millet Library, Istanbul, Jarullah (Jir Allsh) Collection, MS. no. 1279, of which a description is given by F. Rosenthal in JAOS, lxxv (1955), 14-23.

Recto: Lines 3-1 1 correspond to M 70a,9-17; J 249,22-250,l; Q 386.1 5- 25; A 386,12-387,l. The two words of the margin appear in M 69b,18; J 249,ll; Q 385.32; A 385.16. 2. One would like to make the traces here match ;l in 1x tivnin of

M 70a,8 etc., but this does not seem to Work. Nor can the first word be convincingly taken as zunj'rx [DJ-], which occurs just afterwards. It looks as if the text of the autograph may be different here from that of the editions.

8. ax303x o;l-3u: Thus A; the other editions have 3"~. 11. The dot of x d is actually written over the daler.

X;luo3x looks as if it is miswritten X;IDP~X. N W : Thus A; the other editions have X'? 1x3.

RYLANDS GASTER COLLECTION B2597 + B4094 Verso

no 3 ] 3 ~ i > u ~ K I a i n u i [ l ~ ] - i > u i a a ; ~ > ~ i i 5aa 3


Margin

NOTES

Verso: Lines 2-1 1 correspond to M 70b,6-17; J 250,lO-19; Q387,ll-25; A 387,12-388,5. The identifiable portion of what is written in the margin corresponds to M 70b,22-71 a, l ; J 250,23-25; Q 387,32-35; A 388,lO-12.

1. No certain traces can be recovered here (at least from the photograph). 2. p: A omits. 3. ;133X11: The editions all have ; r x x l 1x1.

XIIX: All the editions spell with final yod yd' according to classical Arabic orthography.

4. x;1np5o*: All editions have xmp5u' 03. ~DP'?x: Here and elsewhere M, J and Q write Sin for historical S; Maimonides himself writes historical S with samekh.

5. I first read p ~ 9 x l here, but all the editions have [ ~ u w ~ x ] lx and a close inspection of the fragment reveals that this is indeed what is written. What looks like a waw is in fact the upper part of final nun; the bottom of the nun is still visible in the form of a dot above the qoph of pnr* in the following line. aamx : All editions have rlq>x aamx.

8. 121: M, J and Q 'ur; A 5. 10. 9755 : A prefers imperfect L&-. 11. 1-39: A has & = ;I->U (fern.), a reading which may be possible here

also. M, J anh Q have r99. prim: M, J and Q with Sin; A with sM.

TWO NEW MAlMONlDEAN AUTOGRAPHS 717

Margin I : This line has resisted decipherment and identification. The clearest letters seem to read n i x and a little later one might read mn15x, but the line as a whole remains elusive.

11. Commentary on the Mishna (Sanhedrin)

Of all the works of Judaeo-Arabic literature, the textual history of Maimonides' Commenrary on the Mishna is possibly the most intricate. The relationships between the various witnesses to the Arabic text(s), and the relationships, in turn, of these to the various medieval Hebrew translations are extremely complicated, and the discovery of new sources, such as the fragment presented here, serve only to increase the complexity.

In the Bodleian manuscripts Huntington 117 and Pococke 295' and the former Sassoon manuscripts 72 and 732 is preserved the greater part of what is generally, and I think quite rightly, regarded as Maimonides' personal copy of his Commentary on the Mishna, written in his own characteristic hand.3 These four splendid volumes, together representing five of the original six- volume set, may be studied in facsimile in the three tomes of Maimonidis Commenrarius in Mischnam ..., ed. R. Edelmann (Hafniae, 1956-66).

In addition to this authorized version, which incorporates many of Maimonides' revisions to his work, there exist a number of autograph Genizah fragments which clearly belong to an earlier stage of compilation and attest to a lengthy period of textual development lying behind the form of the Commentary as now represented by the Bodleian and Sassoon codices. These Genizah fragments are as follows:

(I) JTS, ENA 4045 f. 8. A single leaf of tractate Tohorot discovered by M. Lutzki4 and published in facsimile by

l A. Neubauer, Catulogue of rite Hebrew Manuscripts in rhe Bodleian Library, . . . i (Oxrord, 1886), nos. 393 and 404.

D.S . Sassoon, Ohel Dawid. Descriptive Catalogue of the Hebrew and Samaritan Manuscripts in the Sassoon Library, London (Oxrord-London, 1932), 92 3. These manuscripts are now in the Jewish National and University Library, Jerusalem. ' On this matter see in particular S. M. Stern, Tarbiz, xxiii (1951-2). 72-88 and

idem, Bodleian Library Record, v (1 954-6), 18 1-8. Stern's demonstration or the autograph status of these volumes is now generally accepted, cf. J . Blau, Tarbig, xxxiii ( 1 963-4), 31 6.

Hatekufah. xxx-xxxi (1946), 683.

7 18 THE JOHN RYLANDS UNIVERSITY Ll BRARY

S. D. Sassoon in his Introduction to Maimonidis Commen- tarius in Mischnam . . . Vol. I (Hafniae, 1956), pls. xxxii xxxiii, and again by A. Scheiber, Acta Orientalia, xxxiii 2 (1979), 193 4. Both reproduction and text were recently published by J. Blau and A. Scheiber in the work mentioned below (iii).

(ii) JTS, ENA Collection. A fragment of a draft of the commentary on Baba Batra identified by S. Abramson. A facsimile is given in Sassoon, ibid., pls. xxvi xxvii, and both reproduction and text in S. Abramson, IS"'n73 ;llwn;l w n m Y U ~ " "i-r*-3n3>, Hadoar, Jubilee Volume, xxxvii 28 (New York, 1957), 67-70.

(iii) Leningrad, Antonin 1095. Two leaves of the Introduction to the commentary on Tohorot recently discovered by A. Scheiber. Reproductions of all four pages are given by Scheiber in "Autograph manuscripts of Maimonides from the Leningrad Geniza", Acta Orientalia, xxxiii 2 (1979), 187- 95, and republished with full text and notes by J. Blau and A. Scheiber, An Autograph of Maimonides from the Adler Collection and the Leningrad Library, Draft of the Introduction to Seder Tohorot (Jerusalem, 1981 [in Hebrew]).

(iv) Rylands Gaster B3667. A fragment of a leaf from the commentary on Sanhedrin, published, with plates, below.

(v) Recently many more fragments of Maimonides' draft Com- mentary on the Mishna have come to light in various Genizah collections. These will be published and discussed in full elsewhere.

The very considerable and often striking differences between the texts of the Genizah fragments and the Bodleian and Sassoon manuscripts make it probable that the "final" version of Maimo- nides' Commentary was preceded not by one draft, but by two, or several drafts. At least, if we may not be entitled to speak in terms of a succession of formal draft copies, we may certainly speak in terms of various stages of more or less thorough revision. Besides the internal textual indications which point very obviously in this direction, there are various pieces of external evidence which encourage such a conclusion.

Firstly, whereas most of the Genizah fragments are surviving folios of what seems to have been one codex, the Lutzki and Abramson pieces are single loose leaves which could never have been bound into a book. Lutzki himself had already observed5

Ibid., 683, n. 16.

TWO NEW MAIMONIDEAN AUTOGRAPHS 7 19

that the text on the verso of the piece he had found was written inversely in relation to that on the recto and could therefore not have been part of a book but can only have been a single sheet.6 Stern, too, correctly concluded7 that this fragment was not a leaf from the lost commentary on Tohorot which once belonged to the Bodleian and Sassoon set, but was the "sole surviving fragment of the draft". The Genizah fragments, then, attest to earlier stages of the text in both codex form and in the form of loose-leaf additions. Such a state of affairs suggests several different stages of revision by the author.

Secondly, we have the testimony of Maimonides himself on the manner of revision of the Commentary. In one of his responsa he deals with a query in which he is called upon to explain the discrepancy between a ruling given by him in the Commentary and a ruling in his later work Mishne Tora. In his replys Maimonides informs his correspondent that the decision of the Mishne Tora is to be preferred to that of the Commentary because the copy of the latter "which is in your possession is the first version [al-nusxa al-'Plc?], which left our hands before it was properly checked [qabl al-tahrir] . . . and when we examined the sources and checked them it became clear that what we had mentioned in the Mishne Tora was right, so we corrected the Commentary, and similarly in that first version [tilka al-nusxa al-'PI51 of the Commentary on the Mishna there are many such passage^".^

Since in this particular case the correction mentioned by Maimonides is plainly visible in the "final" text of the Commen- tary contained in the Bodleian and Sassoon manuscripts, it must follow that the fair copy extant in those tomes is identical with the "first version" which Maimonides states that he had corrected. Since, further, it is quite clear that the text exhibited by the recently-found Genizah fragments represents a stage anterior to that "first version", we can plainly see that the Commentary underwent several stages of revision before reaching the form in which we now have it. If only for certain parts of the work, there

This practice is quite normal in Genizah fragments consisting of only one sheet. Whereas today we turn the page horizontally in order to continue the text on the verso, the mediaeval custom was to turn the page vertically, so that the first line of the verso lies behind the last of the recto. ' Bodleian Library Record, v, 182 n .

J . Blau, R. Moses b. Maimon Responsa, i i (Jerusalem, 1960), 383, no. 217. Nor is this the only such statement. See the literature referred to in Blau's

notes to the responsum.


must have been several such stages, attested to now by the various fragments from the Genizah, which can certainly not all be leaves from one and the same codex.

It seems, in fact, that Maimonides kept the text of his Comment- ary in a constant state of revision throughout his life. Even the so-called "final" authorized version of the Bodleian and Sassoon manuscripts is replete with additions, cancellations and corrections of all kinds, revealing the author's constant endeavour to improve his text. Indeed, in all likelihood the work was in a continuous state of development from the time Maimonides began it in Spain at the age of twenty-three until, presumably, the end of his life.1° This seems to be the only realistic way of interpreting the evidence at our disposal. Such an assumption also accounts for the frequent discrepancies between the "final" version of the Arabic original and the mediaeval rendering(s) of it into Hebrew. l l There seems to me no doubt that the divergences of the Hebrew - version(s) must usually be explained as resulting from the fluid state of the Arabic original, and as reflecting in a concrete way its constantly changing form. l 2

This, indeed, can now be proved on the basis of the new Genizah material referred to above, for more than once we find that the Hebrew version agrees with the reading of the draft, whereas the later form of Maimonides' Arabic text has been revised and the passage altered. In this way we are now enabled to trace not only the growth of the text but also the development of Maimonides' halakhic thinking. I note here in passing that his understanding of certain issues occasionally underwent funda- mental change, for we sometimes find in the newly-discovered Genizah drafts rulings precisely the opposite of those he later adopted in his fair copy. These matters will be treated in detail elsewhere.

The new autograph fragment of Maimonides' Commentary on the Mishna published here, Rylands Gaster B3667, comes from the opening of tractate Sanhedrin. The fragment is the surviving

l0 Thus, e.g., Blau and Scheiber, An Autograph of Maimonides, 5 . l 1 On these translations see Steinschneider, Die Hebraeischen Uebersetzun-

gen . . ., 922ff. l 2 See on this subject S. Lieberman, Hilkhot ha-Yerurhalmi (The Laws of the

Palesrinian Talmud) of Rabbi Moses ben Maimon (New York, 1947), 6ff.; cf., too, Stern, Bodleian Library Record, v, 185.


centre portion of a once whole leaf and is probably incomplete on all four sides, for, although it looks as if line 32 of the recto may have been the last line of the page, this cannot be insisted upon. The paper of Ryl. Gaster B3667 is extremely brittle and frail, requiring the most delicate handling. Since I first saw the fragment, several small pieces have already broken off from the edges and have taken some of the text with them. The two tiny pieces which appear beside the main body of the fragment on the accompanying photographs were attached to the top of the text when I studied the manuscript itself in the Rylands in the Spring of 1982. Were they still in place it might have been possible to restore something of the first line of both recto and verso. Working from photographs, however, I have not been able to fit these pieces satisfactorily together with the rest.

The Arabic text of the opening sections of Maimonides' Com- mentary to Sanhedrin has appeared in print on three occasions. l In 1893 M. Weisz published his Inaugural-Dissertation presented to the Friedrichs-Universitat Halle-Wittenberg with the title Maimonides ' Commen tar zum Tractat Sanhedrin. Zum ersten Male itn arabischen Urtext herausgegeben, mit verbesserter hebraischer Uebersetzung und erlauternden Anmerkungen versehen, Halle, 1893. This edition, made from two Berlin manuscripts, is designated below as W. Weisz's publication, and the many like it which appeared about the turn of the century in the form of disserta- tions, was superseded with the appearance of the whole of Maimonides' Arabic Commentary on the Mishna, with a new Hebrew translation, by Y. Qafih, p awn u*m p m * ~ nu ;Imn ninni l i?n ,linqn, 7 Vols., Jerusalem, 1963-8. This edition is for the most part based upon the autograph contained in Bodleiai~ MSS. Huntington 117 and Pococke 295 and the former Sassoon MSS. 72 and 73. MS. Pococke 295, which contains Maimonides' own fair copy of his Commentary to Sanhedrin, was published in Vol. I1 of the series Maimonidis Commentarius in Mischnam, ed. R. Edelmann, Hafniae, 1961. The portion which interests us is found on pls. 256-262. For those sections missing from Maimonides' personal copy Qafih has established his text according to predominantly Yemenite sources. In the variants below I

l 3 It is perhaps worth noting the recent publication of a translation of the text into English: F. Rosner, Maimonides' Commentary on the Mishna, Tractate Sanhedrin (New York, 198 l ) .


have designated this edition as Q. Although Qafih, Vol. IV, ryrl 170, has based his edition of the Commentary t o Sanhedrin on the autograph Pococke 295, I have nevertheless compared the original manuscript as available in Edelmann's facsimile edition. Information derived from this source is identified by the sign P.

Ryl. Gaster B3667 contains the text of Maimonides' Commen- tary t o the first two mishnas of Sanhedrin and corresponds to W 1,17-12, middle; Q 146,15-155,lO; P 256,2 1-262,20. Between the end of the recto and the beginning of the verso a section of text is missing : W 7,2-8,14; Q 150,23- 1 52,8 ; P 259,16-260,19.

In dealing with the fragment I have first reproduced Maimonides' autograph text line by line, exactly as it stands in the manuscript, retaining such idiosyncrasies as his quite unsystematic use of diacritics. There then follows an attempt at a reconstruction of the full leaf. The restorations have been made according to PIQ, and , for the sake of uniformity. diacritical points have been added even where Maimonides in the autograph left them out.14 Save for a few minor variants, mainly of an orthographical kind, the text of W is for all practical purposes identical with PIQ. In some cases I have not offered a restoration because it is clear that what once stood in Ryl. Gaster B3667 was materially different from what we now find in P/Q and W. AI1 such instance are marked by .. ..

The restoration of lacunae in a text such as this is, of course, a risky undertaking, but I nevertheless venture to believe that a reasonable degree of success has been attained. One of the most valuable guides in this matter is the amount of space estimated to be available for any particular restoration, but in this respect it is very easy to err. Firstly, the number of letters to a line is not uniform in any case, and, secondly, Maimonides' whimsical use of abbreviations and his habit of erasing certain phrases and adding others between the lines or in the margin make the chance of error even greater than it might otherwise be. For this and other reasons, some of the restorations proposed below may turn out to be mistaken, but it seemed to me preferable to offer something that may be wrong than to suggest nothing at all.

l4 The text of Q has been compared with that of the autograph fair copy Pococke 295 and I have tacitly corrected where necessary. None of these corrections amounts to very much; they mostly concern the genuine Maimoni- dean orthography of e.g. I + the definite article as 3'7 (as against Q - 3 ~ 3 ) or the spelling without aleph of nh3h etc. (as against n h ~ 3 h of Q).

r J-' r C

THE JOHN RYLANDS UNIVERSITY LIBRARY

] i n i x 113 73s x n n x 5np K ~ K 3 3 7 ll-[n5

PLATE 111: Ryl. Caster Coll. B3667, recto


NOTES TO THE TEXT

Recro : 4. The mem of ]a is oddly written, but the word is certain. 7. The reading [. . ] m looks clear enough; perhaps m a x was written

here? The verb used in Q and W is '?g. 9. There are clear traces of the first word here, but as the text is quite

different from that of Q and W I have not ventured to suggest what it might be.

10. [ ]51p1 rather than [ ]>xpl. 12. One should perhaps read >v5 without yod.

The final mem of 03->Y is uncertain on the photograph; it may be that one should edit (a);r+u.

14. As both the Hebrew subject and the Arabic predicate are feminine one would like to read *a with Q and W; but the word is clearly 13.

15. The nun of is corrected from lamed. 17. The final pe of the first word seems fairly plain; the other possibility

would be to read her. 25. h p [ ] is less likely.

For this use of gad in a nominal sentence ("we already have a text ...") in Judaeo-Arabic see J. Blau, A Grammar of Mediaeval Judaeo-Arabic2 (Jerusalem, 1980), 182 (Hebrew); idem, The Emer- gence and Linguisric Background of Judaeo-Arabic2 (Jerusalem, 1981), 65; for the same usage in Spanish Arabic see F. Corriente, A Grammatical Sketch of rlte Spanish Arabic Dialect Bundle (Madrid, 1977), 129. This provincial feature has been eliminated from the revised version of the Commentary (see "Variants").

26. The dalet of xmnx is faint on the photograph but nevertheless beyond doubt.

31. The general content of the restoration here after Q and W may well be correct, but immediately after the break there are visible what seem to be traces of the upper parts of two lameds. In this case the wording would have been slightly different.

RYLANDS GASTER COLLECTION B3667 RECONSTRUCTION - Recto


n E cP

P h X r J ' h

r r r a C C

h

m r? J' (U r w

X R 4 X X r

J' r n * e r n

n r r X r O r E r r r e P

n C x E r r r r(

n . r X E C P X P c r X r

r

T c c C

n a C r J'

r x n x J ' ~ n n t C E

c r c

n E n r ~

- n D

J ' x n C

R % * C r

E n D J '

r n c r n r r n u m n n n

4 C -

r n r n r m rJ' X r~

r r r n r r r r : r n C C E R X

J ' , - , ~ J ' r E J ' D X c n r n r - h n - P - - c c n

C U n c U r r w < - C C

r r h - a r L L E ~ C - r r n r n r c c . X


VARIANTS

Recto : 415. Q 146,2112 and W 2,4-6 are different, but the intention is the same:

n101p9x TXO -D o33n n101?9x nin -3 n9:,n xn91 .n19an9x n9n1 In .... 6. The citation of Deuteronomy xxii.14 is absent from Q 146,24 and W

2,718 which read: n m v ~ n r x h D-hna nlvp puu-. 7. Q 146,257 and W 2,9-11: *B rx1 x im 103 nxn 0111 mp9n 1.1

. . . 9 x n h 52 1x5 ntp'rrn. 8. Q 147,2: n09wn. 9. The order of Q and W is different here and the text considerably

expanded. 1-xn i:, n:,h r x l corresponds to Q 147,14 and W 2,27, but ... p:, 73~1, which in the draft follows the mention of R. Meir, in Q and W comes before it at 147,4 and 2,16 respectively. The explanation for the presence of the three judges given in Q 147,4-14 and W 2,16-27 is very much longer than the short note of a few words for which there is room in our draft. The mention of Exodus xxii.8 and xxiii.2 in this connection probably belongs, therefore, to Maimonides' later revision(s) of the text (unless, of course, one assumes that something corresponding to Q 147,4-14 and W 2,16-27 appeared in the now lost margin or was added between the lines).

10. This corresponds to Q 147,25 and W 3,11. Again, the sequence is different. Whereas in Q and W our line 10 concludes the commentary on I 2, in the draft it is placed first. It is not clear how the end of the line may best be restored and I have hence left it blank.

13. W 3,8: l;r'?v. 1314.- W 3,1011 1 apparently misunderstands the text and prints : lx 3 x

3 ~ 3 -D nrvnP xn -9v nuaix - x i unk. 14. Line 10 of the draft (mentioning Simeon b. Gamliel) occurs at this

point in Q 147,25 and W 3,ll. The chapter heading bl O - I ~ n:,-no is not quoted in Q 147,26 and W 3, line 1 of the commentary on I 3. It is worth noting that in this draft Maimonides never cites the full text of the Mishna (whereas in P he always does); cf. on this S.W. Baron, A Social and Religious History ofthe Jews, vi (New York, 1958), 353. After ox:,n'?x nntpn there is a long expansion in Q 147,27-149,l and W 3, line 2 of the commentary on I 3 4 2 9 which is quite without representation in the autograph draft. Line 15 of our manuscript corresponds to Q 149,2ff. and W4,32ff., so I have restored the end of line 14 according to Q 149,112 and W 4,3013 1.

15. Q 149,3 and W 4,3213 both times give Arabic for our Hebrew O*IW and add r15x after the second occurrence.

1516. After Tnx yb-oa Q 149,5-7 (and similarly W 4, antepenult. -5,l) reads: 91p0 n-3 r x 0-10 -1i)f Tpnr (see W,10 note 6) Txn ii m x t iu r ~ - o i n followed by an? i:, n:,5m. There is not sufficient


room to admit this passage; perhaps our draft contained a shorter version of it.

1617. Q 149,8-11 and W 5,3-7 are both differently worded from and more substantial than what could have stood here in the draft after aix33x: .D ~1.3 t p i .xn5~3 xn9n *l19 ;rwnn> 33% In1 XI~YDI

;I> 7.1. . u ~ i YUI ;15ipi .D-IW -153 rKj pwn 3x ]X mm- i w y n5xn 'Y31 YDl 11.73.

20. Q 149.19 and W 5.16: 1;1>3x. 21. Q 149,21 and W 5,18 spell 70- in accordance with classical Arabic

orthography (and again in line 22). Q 149,22 and W 5.19 substitute ir for XID (as above, line 7) and write ~ 7 1 7 without the definite article.

22. Q 149,24 and W 5,22 have ylx5x for our ;11ntSx (for dimna in this general sense see I. Friedlander, Der Sprachgebrauch des Maimonides . . . I (Frankfurt a.M., 1902) s.v.), 'ion with yod instead of aleph (see note to preceding line), and the wholly Hebrew r 3 m for our hybrid 33um (= Se-'alayhi rather than Se-'aleha).

23. For 11.1 "pay" (cf. R. Dozy, Suppliment aux Dictionnaires Arabes, ii (Leyde, 1881), 800) Q 149,27 has ~91'1; cf. Q ibid., n. 30. At the end of the line there is insufficient room to accommodate Q 149,28 9 and W 5,25 6: niupip3x 1-tpn 3nn a'rx3x 1-tpn W.

24. Q 149, ult. and W 5, ult. - 6,l read: an333 nu1 33ip3 3tndx uiinn, continuing thereafter: x1n3y.3 ;1n;13;1 nu1 a3ip IN n3p 1x3 ;r3x ixnn- x3 x i n ~ rtrpn3 x3 3np3x x a n ~ y x k x an33 p. Both Q and W omit the chapter heading.

25. Q 150,314 and W 6,4 5: .. . in1 3n335x 3np 1x3 xnlu 1x3. 26. W 6,7 : xnnmx. 27. After t n * Q 150,8 9 and W 6,10 1 1 add : niwm *rt p 33yt

7Y> r3-0 ; I W ~ D9WY3. The explicit statement ... 1 1 ~ 5 ~ i> ;1>5;n is reversed in Maimonides' later revision(s), which rules (Q 150,lO and W 6,11 2): 93 ;1>3;n ;13'?Y. The chapter heading for Sanhedrin I 5 which I have supplied here does not occur in Q 150 and W 6.

28. Q l50,l l U J W ~ X but P 259,9 as here. 30. Q 150,18 and W 6, commentary to I 5, lines 8 9: k 1x5. 3 1. Q 150,2 1 and W 6, penult. : 3171;1.

RYLANDS GASTER COLLECTION B3667 Verso


/!H

if;,' d h, ?"p21P: 2

,C,'- , I E ; '

-

PLATE IV: Ryl. Gaster Coll. B3667, verso


a l p S x r n xna i19ara i n 3 - i 15nSxa S. [ 21

1 un1xS

NOTES TO THE TEXT

Verso : 4. A blob of ink gives the second a rather strange appearance. 5. The second shin of ;~W>WJ has been corrected from another letter,

probably lamed. 10. oxl>xr is badly written; it looks more like, but cannot really be, 0x191. 1 1. The form mnjn is of double interest. Firstly, Maimonides has here a

niph'al as against thepuWal of Q and W (see "Variants" and cf. Q 153 n. 6); secondly, there is the characteristically Sephardi feature of patah for Tiberian gum*.

13. linnn5- is hardly a genuine metathesis of ti5nnn-; it is simply a mis- spelling.

16. Note the spelling ~ P X ~ D with aleph. 25. The trace of the letter at the beginning of the line may possibly

indicate yod. If so, we would here have to edit oln* [xh] as in Q and W.

27. The word preceding n n one would naturally expect to be * m O , but the traces seem to favour a final nun.

nr

rr

#r

"

!,

B'

"

C

n

L

n

r r

r

E

J'

C

r

no

Rr

(n

n

rr

nr

(r

D

n

a

c

FT

:

r P

r

U

Ca

r

E

r - P

rr

PL

nC

B

DC

-r

X

nn

n

Pc

r r

r

B&

r-

X C

l-

J'

C

oc

rr

!L

n

aB

r

J'

ra

rO

C

X

r +

C;7

>

r

r

ar

an

J'

nx

mr

C

YJ

'Z

C

J'

P

r

R J'

J'

o

n

~C

TC

~

XJ

'L

~

rr

rh

a

xr

n

ex

a

r o

r

*+

I

n

r P

C

m U

r J'

n.c

a

.P

RR

cY

E

cn

X

E. 0

r C

r r

rn

r

4B

r

~r

vp

nL

!Z

o

X X

rr

r L!

J'

R

EC

R

ar

r

c

r>

nu

r:

o~

a

TC

r

J'

rn

n

E

r-

r

X

hC

C C

a

m

rr

~!

.

X r

r0

r

C J'

r-

n

.c

-

CJ

'C

D

E

C

nJ

'z

CO

ln

C

.n

r

nX

C

I X

YO

RZ

n r

-r- J'

J'

AK

n

hR

X

r

-L

!Y

~I

IJ

'P

.~

Q~

(

Cl

Xr

rr

r

.T

D.

r

-P

m

C C

r

X

r

E

X c

J'

r

C J'

a

C

C

r

X

P

r r

w

C

C

C

n

J'

C


(17) I n 0 1 - 1 - 1 1 .nbnp 7 s f i x 35 7 3 - a 5 ~ I K

VARIANTS

Verso : 3. The "twenty-three" before p5u1 mmul is omitted in W 8, ult.

After D-nm n x 5 n Q 152,13 4 and W 9,2 3 have what is plainly a later gloss for which there is no room in our draft (unless, of course, it was added between the lines or in the margin): - n I mxnui ; l ~ l p ' ? ~ 'IyX

nl.Pl3. 4. I have restored ' ~ 5 m instead of the mistaken ;1~5m of Q 152,17 and

W 9, footnote b. See on this Q 152, n. 49 and W 12, n. 44. 4 5. The passage ;lm5m1 np5nnn1 D"IWI n-JII ;lpr +m a ~ p l taken directly


from BT Sanhedrin 17b is omitted from both Q and W, which read here: . .. lnrx xbni ;rp-rr;l a5n5 nw5wi.

6. Q 152,19 20 and W 9,9 10: an5hr nriwu -it? yn~pn5x 5 , ~ .. . riwm.

7. The separate headings of both the pereq and the mishna are absent in Q 152,22 and W 9.

8. Q 153,l: p 3 1 (but see ibid., n. 2 for the reading ~OIDI of P). W 9 prints p 3 1 in the text, but informs us in footnote i that both his MSS. in fact read r0131. The repetition of the phrase containing this word is also found in P 261,1, but the second occurrence is there marked for deletion. Q 153,l and W 9, I1 1, 3 4: .. . ;rxlun ;13u xlm p 3 1 13 ;151~1.

10. P 261,4 has ~ ~ U X I (but Q 153,6 as here in oblique case). 10 1. Q 153, 8 9 and W 10,l 2 formulate differently: 1;r 517 - i j x

mnnn (P 261,8 ~;IxTY) n;IxIxY x ix DXI??. 11. The latter part of the line presumably contained a mention of the

expression 7 n i ~ 3 11x (unlike Q and W). 12. Q 153,ll and W 10, I1 3,l: ... u p 9 x 1 ~ -3% *D xi;^. 13. For pnnny Q 153,13 and W 10, I1 3,3 naturally have the expected

p5nnn-. 5xn35x - 3 ~ : Q and W omit.

14. pn5x 71u: Q 153,15 and W 10, I1 3,6 omit. 14 5. The phrasing in Q 153,17 8 and W 10, I1 3,8 9 is slightly different:

... 1x3 DH>O>N ; r jy '11'11. Hence one could restore ... 7 7 1x31 if necessary. However, since both P 261,22 (not indicated by Q) and one of the MSS. used by W have a nominal sentence and omit 1x3 altogether, I have assumed the same construction here.

15. It may be that the restoration of the latter part of this line from Q 153,20 and W 10, I1 3,11 2 is unjustified. Q 153,20-154,4 and W, commentary to I1 3, p. 10,ll-11,2 contain a short section about the king's widow for which there is insufficient space in our autograph draft. I have simply restored here the beginning of this passage, but even this may be too long for the line.

16. The chapter heading is not present in Q 154,4 or W 11,2. Q 154,4 and W 1 1,2: mri, 15w r l l u 5 ~ mm. P has ri1u5b in the mishna (26 1,15), but miswrites riu15b in the commentary (26 1,26).

17. Q 154,617 and W 1 1,415: 509o In i x x 171 ;nox>x1 -0xi35x YK~IN p . I have here restored the chapter heading, which is absent from Q 154,9 and W l l , in accordance with Maimonides' practice.

17 8. Q 154,lO-12 and W 11, I1 4,2-5 have an expanded version: ]X n1w~;r nnn3n unm In1 .;rulm p 1nxnSx -9 xra xn3 79nu nnn9nl ... . . . I n n X?.

18. Q 154,1213 and W 1 1, I1 4 3 : rnnn r x mm. 19.15n'rx 139 : Q 154'15 1xu505x 15i . W 1 l , 11 4,8 omits 7% and,

footnote f, wrongly removes the definite article from p u 3 0 5 ~ .


19/20. Q 154,1516 and W 1 1, I1 4, ult. are more concise: 5xn IN? pt3?D?'r IXD?D?N.

20. Again the chapter heading is absent from Q 154,17 and W 11. 2011. The opening of the commentary to Sanhedrin I1 5 in the draft is

substantially different from that now extant in Q and W. Here Maimonides speaks about the multiplication of wives by a king, but in his later version(s) of the commentary he has nothing to say on this subject and begins his remarks with the matter of the king's horses (= our lines 22 ff.).

23. Q 154,20 and W l l , penult. : I n n . 25. Q 155,2 and W 12,5: D i n * xh .

W 12,5 prints with one MS.: x k x ??U. 27. Again Q 155,6-9 and W 12,9-13 have something quite different from

the draft. Our passage about the king's copy of the Book of the Law "hanging from his arm" (< BT Sanhedrin 21b) does not appear there at all, and for what does appear there there is insufficient space in our draft. The chapter heading has been restored here as usual.

28. Q I55,lO and W 12, I1 6,l : 3 ~ x 1 p?n- xn my.

Documents

TWO NEW MAIMONIDEAN AUTOGRAPHS IN THE JOHN …