BIRNBAUM ELEAZAR VT 196710 The Michigan Codex An Important Hebrew Bible Manuscript Discovered in The University of Michigan Library

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 48

The Michigan Codex: An Important Hebrew Bible Manuscript Discovered in the University of

Michigan Library
Author(s): Eleazar Birnbaum
Source: Vetus Testamentum, Vol. 17, Fasc. 4 (Oct., 1967), pp. 373-415
Published by: BRILL
Stable URL: http://www.jstor.org/stable/1516691
Accessed: 25-05-2015 07:18 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Vetus Testamentum.

http://www.jstor.org

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX
An ImportantHebrew Bible ManuscriptDiscovered in the
Universityof Michigan Library*.
BY

FrIAZAR BIRNBAUM
Toronto

The numberof Biblical codices in Hebrew which survivefromthe


finalformativeperiod of theMasorah is rathersmall.When,therefore,
a new one of importancecomes to light,its finderhas the pleasant
task of drawing the attentionof his colleagues to its existenceand
reportingon its characteristics.
The "final formativeperiod of the Masorah" referredto above
may be variouslydefined.Here it is used for the period preceding
the eleventhcentury.Not all the survivingHebrew Biblical codices
of this time have formalMasoretic notes, and not all of those that
have a Masora Parva (MP) have also a Masora Magna (MM). A manu-
scriptthathas both may deserveparticularattention.
Afterthe discoveryof the famous Cairo Genizah at the end of the
nineteenthcentury,its contentswere dispersedamong a numberof
librariesin Europe and North America. Amongst the huge number
of documents and fragmentsit yielded, there were a considerable
number of quite old Biblical MSS, or ratherfragmentsof them1),
withMasoreticnotes. No Genizah materialis known to have come to
the Universityof Michigan Library in Ann Arbor. Its Rare Book
Department is justly famous, but not for Hebrew material.When
I was Near Eastern Bibliographerat the Universityof Michigan
(1960-64), I was asked, late in 1962, to supply statisticsof that Uni-
versity'scollections of Near Eastern MSS I found that only five
Hebrew MSS were recorded in the Rare Book Department'sfiles:
four scrolls and one codex. Two were average Torah scrolls (the
older 17th or 18th century,the younger 19th century),two were
scrolls of Esther. The lone codex (numberedMS 88) was entered
thus on p. 19 of the Rare Book Department'sInventory for 1922:
* Abbreviations and References, see p. 414 f.
1) Not covered, as a matter of policy, in S. SHAKED, A TentativeBibliography
of Geniza Documents,Paris-La Haye, 1964 (= ttudes Juives, 5).
Vetus TestamentumXVII 23

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
374 E. BIRNBAUM

Pentateuch. i missing).
Hebrew.On Vellum.(defective-Gen. 17 (?) cent.
23 cm.Bought
ofGejouofParis1),in December,
1922.
This entry was probably constructedfrom the following brief
reportmade on MS 88 (hereafterM) by Leroy WATERMAN, formany
yearsProfessor of Hebrew at the of
University Michigan. (It is to be
foundin the Library'sfileon MS 88).
Thevolume of Hebrewtexthereexamined consists
of thePentateuch. It
ispractically
complete ofGen.I. It is written
withtheexception inthesquare
Hebrewcharacter withvowel pointsandis dividedintosections
withrunning
commentsandcatchlinesat topandbottom ofthepages.It is evidentlylater
thantheinvention ofprinting andbelongsat theearliest
to the17thcentury
Its valueis chiefly
probably. illustrative.
I should pounds
regardfive as a liberal
price.
Nov. 1922. [signed] L. WATERMAN
On opening the box containingthe MS, I was very surprisedto
see Hebrew calligraphytypical,not of the 17th century,but of the
9thto 12thcenturies-a good codex fullypointed and accented,with
Masora Parva and Masora Magna in the margins.When I found the
readingnDW.',whichis characteristicof the Ben Naftali(BN) school,
as compared with the Ben Asher (BA) and Textus Receptus (TR)
reading ?=t, it became clear that the MS deserved furtherstudy.
I gave a preliminary reporton theMS in a paper read beforea joint
meeting of the Middle West Branchof the AmericanOrientalSociety
and the Midwest Section of the Society of Biblical Literature,held
at the Universityof Michigan in Ann Arbor, April 26, 1963. The
presentarticlecontainsthe resultsof furtherexamination.
While the MS is undated, palaeographicalstudy showed it to be
10th century2). Rather few full texts of the Hebrew Pentateuchof
this age and calligraphictype survive. As will be shown in detail
later, M does not correspondclosely in its textualtraditionto any
single known MS. Although it has affinities in many respectswith
various MSS, it is unique in some of its readings,vocalization and
accentuation.Attentionwill be drawn to some of these,and statistics
of its readingswill indicateits relationshipto the Ben Asher versus
Ben Naftaliclassificationof MSS 3).
1) I. E. GEJouof 77bis Avenue de Breteuil,Paris,was a dealerfromwhom
the Universityof MichiganLibraryboughtmanyGreek papyrifromEgypt at
thisperiod.
2) See below, p. 383.
8) Considerable doubtshave beenexpressedas to whetherthereareany"pure"
or ever have been. The descriptionof the Aleppo Codex as
BA manuscripts,

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 375

PhysicalDescription
151 folios of vellum, each 21 x 18 cm, writtenin black ink on
both sides of the leaf. Two columns to the page 1), 21 lines to the
page. At various points in the text leaves are missing. These total
56 apparently,so thattheoriginalMS probablycontained207 leaves 2).
Binding.Black leather Islamic binding, but without flap. In the
centreof both frontand back covers is a large medallion (shamsah)
gilded inside, except for a raised relief pattern of stylized leaves
withinit, and a small lozenge above and below it. Around the edges
of the covers thereis double-linegold border but it is now mostly
covered up by "repairs": coarse black leatherwas used to re-edge
the covers. The same material forms the present backstrip, con-
sisting of two overlapping pieces of leather.In the centre of each
cover fore-edge,and abovethe leatherrepair strips,is a small metal
plate, the remains of a European type of closure. The clasp which
once linked them is gone, but it must have been put on afterthe
repairs.It seemsto me ratherlikelythatthecovers originallybelonged
to some quite different MS. Gold end-papersdecorated with large
red stylizedflowersare pasted on to the inner sides of the covers.
On the frontend-paperthereis a red ink inscriptionin a European
hand of approximatelythe 18th century:No. 365, and beneath this
a 20th centurypencil note: 15567 152 Bl. From all these indications
it seems clear thatM is notone of those MSS which came out of the
Cairo Genizah at the end of the 19th century.Where the Paris book
dealer GEJou obtained it is an unanswered question, but it went
throughthehands of a German-speakingpersonnot verylong before
its sale to the Universityof MichiganLibraryin 1922. The MS shows
evidence of having been repairedand rebound more than once. In
the process some pages were misplaced3). The folios were not num-

"pure BA" is under strong attack. Cf. Harry M. ORLINSKY, "The Masoretic
Text: a CriticalEvaluation" printedas a Prolegomenonto the photographic
re-issueof the London, 1897 edition of C. D. GINSBURG'S to the
Introduction
EditionoftheHebrewBible.New York, Ktav, [1966].
Massoretico-Critical
1) Many of the older Biblicalcodices are writtenin threecolumns,and some
later ones in two columns; but the reverse is also not quite uncommon. Thus
the numberof columnsper page in whicha BiblicalMS is writtenis no safe
criterion for dating. Cf. M. GOSHEN-GOTTSTEIN,'Biblical Manuscripts in the
UnitedStates',Textus2 (1962) p. 45 n. 7.
and
2) See AppendixA fora tabulationof the missingleaves and thechapters
verses they must have contained.
3) See below, p. 399 for details of misbinding and the correct sequence of the
folios.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
376 E. BIRNBAUM

bered, but at the end of the verso of each leaf the catchwordlinking
it to the next leaf has been writtenin a hand which looks at least a
centuryyounger than that of the MS. itself. Since some of these
catchwordsreferto missingleaves, the MS was completewhen the
catchwordswere writtensome ten centuriesago.
The consonantalBiblical textis writtenin an excellentand beautiful
dear professionalscribal hand. The Masoretic notes are in a good,
rathermashait1) hand. The ink has occasionallybrowned or faded,
especially the vocalization. Some pages are rubbed or smudged.
Sometimes words and even several whole pages were re-inkedin
antiquity2). A numberof pages are partiallybroken offand lost, and
several have had portions of theirmargin deliberatelycut away 3).
In spite of the beauty of his writing,the original scribe seems to
have been distractedat times,to judge by the numberof words or
phrases omitted by him. Some are typical cases of homoeoteleuton,
othersare plain carelessness;only a small minoritymay possibly re-
presentsome genuine textualvariant.The omissionsare markedby
a small circle at the top of the line: the missingwords have usually
been insertedin the margin,sometimes,apparently,by the original
scribein the same square script.Most of the insertionsare however
in a less formalhand, similar to that of the Masorah. Occasional
correctionshave been made in other,later,hands. There are several
cases of dittography.The superfluouswords which are to be deleted
have been markedby the scribeor laterreaders4).
The scribeof the consonantaltextwrote the charactersdependent
fromlines, which were impressedhorizontallyacross the parchment
with a blunt point (shir.tut), as required for Torah scrolls; vertical
rulings to mark the leftand rightedges of the columns are usually
pressed rather shallow and are sometimesinvisible.
A singleMasoreticmanuscriptof thistypecould have been written
by up to fourdifferent people: (1) the main scribe,fortheconsonantal
text; (2) thenaqdanor punctuator(vocalizer); (3) the accenter;(4) the
masorete, who wrote the Masora Parva and the Masora Magna.
In M it would appear that the consonantaltextwas writtenby one
scribe, and the MP and MM by another.Further,both the vowels

1) Non-squarebook hand, in the formalizedcursivewhichwas inaccurately


termed"Rabbinic" by Europeanscholarsof formergenerations.
2) E.g. ff.134b,135b,136a.
8) See below, p. 398.
4) On thesemarks,see below p. 358.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 377

and theaccentsseem to have been writtenby one and the same person
(this is subject to furtherstudy)who also appears to have corrected
some of the mistakesin the consonantaltext1).

Is M a Ben Asher or a Ben Naftali manuscript?


For a long time now Biblical manuscriptswrittenin the period
preceding the rise of printinghave been categorized as belonging
eitherto the Ben Asher or the Ben Naftalitradition;the majorityare
regarded as being pure representativesof neither,showing conta-
mination to various degrees. The exact criteriafor distinguishing
betweenBA and BN have been the subject of much argument.Both
schools are, in fact, quite close to one another, having identical
readingsforthevast majorityof words. As to theverylimitednumber
of disagreements(hillufim),even 70 yearsago GINSBURG pointed out 2)
that"it is the presenceor absence of the Methegor Gaya which cons-
titutes fully nine-tenths of the differences". GOSHEN-GOTTSTEIN
describes the hillufimas "tiny differences"which really constitute
"two subsystemsinside the Ben Asher family"3) withinthe Tiberian
Masoretic system,and blames "the medieval compilersof Masoretic
lists" for reducing"variantsto the stereotypedcontrastBen Asher-
Ben Naftali"4). Today the usual method of classifyingMSS is to
compare them with the oldest surviving source on the hillufim of
BA and BN, namely the lists given in the Judaeo-Arabictreatise
Kitdbal-Khilaf,by Mishael ben 'Uziel (MbU) 5).
M containsmanyreadingswhichare different fromthoseattributed

1) The ink of the consonantaltexthas generallyremaineddark,whereasthe


Masorahsare sometimesmorefaded.The inkof thevowels,accentsand some of
the corrections, is now oftenfadedalso, and it is frequently brownratherthen
black(e.g. ff.5b and 6a). It is notyetclearto mewhetherthescribeoftheMasorah
was thesamepersonas thevocalizerand accenter.Ink colourdifferences between
some of the vocalizationand accentingof the dittography on f. 13a seemedto
indicatethatthevocalizerand theaccenterof M weretwo different people. Exa-
minationof the originalMS. however,revealed that this dittographywas a
specialand unrepresentative some laterperson,who was offended
case: apparently
by the lack of completevowels and accentsforthesesuperfluouswords,added
themhimselfin a slightlydifferent ink.
2) p. 278.
GINSBURG,Introduction,
3) RTBT, p. 107 ff.
4) RTBT, p. 109.
5) Edited with criticalapparatusby Lazar LIPSCHOTZ in Textus2 (1962), at
summaryof the textin English and
back. See also the valuable introduction,
discussionby LIPSCHUTZin Textus4 (1964) pp. 1-29.Mishaelprobably "lived in
thefirsthalfof the 11thcenturyin Jerusalem"(Textus4, p. 2, n. 8).

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
378 E. BIRNBAUM

to BA and BN, as well as otherswhich correspondwith eitherBA


or BN. Since, however, the BA/BN division is so firmlyentrenched
in Masoretic studiesI examinedthe whole of M on the basis of the
informationgiven in Mishael ben 'Uziel's treatise.Before giving
in each Biblical book, the Kitdb al-Khilafsets
details of the hillufim
out eight general rules for distinguishingbetween BA and BN 1).
Rule 1: BA '-:t; BN '?t. M reads like BN in everycase 2).
Rule 2: Vocalization of certainformsof the verb 7"a. M reads like
BA in 12 of the 15 cases concerned.M has a lacuna in 2 of
the other instances. In only one case does M apparently
read as BN 3).
Rule 3: Punctuationof certainformsof the verb t'l. Of the three
occurrencesin the Pentateuch,M reads like BN in one case;
the othertwo are lacking,therebeing a lacuna in the MS 4).
Rule 4: Refersto a difference in pronunciationwhich is not repre-
sentedin writing.
Rule 5: Use of ga!yahwith the phrase -n; trsi-lrvt. BA without
ga!yah;BN withgayah. M reads like BA in each case 5).
Rule 6: Use of ga!yahin certaincontextswith 'i;n. BA withga!yah,
BN without.M reads like BA 6).
Rule 7: Vocalization of prefixes: and , when followed by ,. BA
--,', --- ; BN--n, --,. M reads invariably likeBA in a total
of 24 instances.There are lacunae in severalothercases.
Rule 8: Use of dageshand rafein n"5rnm in specifiedcircumstances.
BA rafe;BN dagesh.M reads like BA 7).

differenti-
Mishaelben 'Uziel gives a supplementarycharacteristic,
Textus4, p. 16-19.
1) Hebrew text,Texius2, p. iii-v; Englishsummary,
2) See plateA col. 2 line 12, and plateB col. 1 line 7.
3) Readingsas BA: Gn. iii 17; Lv. vi 11; vi 19; vii 6; Nu. xviii 10; xviii 13;
Dt. xii 15; xii 18; xii 22 (twice); xv 20; xxviii39. Lacuna at Dt. xviii 24 and
xviii25. Readingapparently as BN: Dt. xv 22. This lastone seemsto be somewhat
uncertainbecausethepositionof theshevaat an angleto theextremerightbeneath
the kaf(thus ?) would suggestthatthe vocalizermighthave intendedto add a
patah(whichwouldhaveresultedin theexpectedBA reading)butforgotto do so.
4) Readingas BN: Nu. xxii6. Lacuna at Ex. xxiii29 and xxiii30.
5) Lv. xxiii19; Nu. vii 16 (see plateB col. 1, lines1-2); Nu. xxix11. The other
eleveninstancesin whichthisphraseoccursare not pointed,sincethescribehas
leftall the repetitionsin this chapterunvocalized and unaccented.See below
p. 386.
6) Nu. xxi 35 (smudged,but withga'yah);Dt. iii 3; xxviii55.
7) Gn. xix 17; xxxix15. In anothercase (Dt. ii 16) M has a lacuna.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 379

atingBA fromBN: the pointingof the words rpp (p. 7n). In


this,M coincidesinvariably withBA. In theapplicationof theeight
generalrules which involve differences in writing M
1),therefore,
agrees with BA in six and with BN in two instances.
The situationis, however,verydifferentwhenwe compareMbU's
listsof (a) individualhillufim
betweenBA and BN, and (b) thecon-
gruences of these two great Masoretes against other Masoretes, on
theonehand,andthereadingsofM, on theother.Belowis a summary
table showingthe resultsof thesecomparisons.(Since fadingand
smudgingin the MS makes it difficult to be sure about some of the
vocalizations, such cases may have to be disregardedfor statistical
purposes.I havemarkedthembyputtingthenumber, withan aster-
isk,intoparentheses.Thus,6 (or5 *) meansthatthereappearto be six
instances,but thatone of themmay have to be discounted.Full
detailsare givenin AppendixB.

MichiganMS. 88: Total Number Mishael ben


Mishaelben of CUziel's
'Uziel's total
Biblical totalsof Agreementswith Disagreements number of
Book individual withBA and BA and BN
hillufim BN congru- congruences,
betweenBA BA BN ences (whe- as against
and BN therlistedby other
MbU or not) Masoretes

Genesis 39 19 10 8 (or 7*) 31


Exodus 20 7 9 10 28
(or 8*)
or 7*)
Leviticus 15 6 6 3 13
(or 5*)
Numbers 24 10 9 13 16
Deutero-
nomy 19 6 5 8 24

Totals 117 48 39 42 (or 41*) 112


(or 47*) (or 38*
or 37*)

From the table it is clear that,if MbU's informationis used as a


M can hardlybe classified
criterion, as a BA manuscript,
stillless as
a BN, though it is somewhat closer to BA. M's total BA and BN
readings,according to the above table (i.e., leaving out of account
1) Rules 1-3,5-8 and thesupplementary
rule.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
380 E. BIRNBAUM

the generalrules cited earlier),compriseonly about 87, whereas


MbU's total117. The lacunaein M accountforthedifference. The
large number of cases (42 or 41) whereM with
disagrees bothBA
and BN, putsM in a category ofitsown.

Ga!yahin M.
We referred above to the factthatthe vast majorityof the dis-
agreements betweenBA and BN concerned theplacingor absenceof
ga!yah.A studyoftheuse ofga!yahin M showsthatthesub-tradition
it represents was notablydifferent in stressfrombothBA and BN.
M oftenplacesga!yahs whereBA and BN do not1). Conversely, in
many cases where BA and BN are without ga!yah,2) M clearlyuses
them.Not infrequently M uses a ga!yahon a different vowel than
manyotherMSS 3). To sumup, we areleftwiththedistinct impres-
sion thatM is morelavishin its use ofga!yahsthanBH but less so
thanMG. It shouldbe pointedout thatM's use ofga!yahoftencoin-
cideswithMG's againstBH 4).
A notablescribalfeature in M is itshabitof nearlyalwaysplacing
thegayahsbefore thevowelsign(i.e. to therightofit)in contrast with
mostotherMSS, whereit is to be foundafterthevowel sign(i.e.
to theleft).Nevertheless therearesomeinstances whereM too places
it to theleftof thevowel5). A detailof someinterest is thescribe's
of
practice joining two signs graphically:when a ga!yahprecedesa
patah thecombination often appears thus 6).
Furtherexamination of M showeda verylargenumberof points
of interestof many kinds-differences in 'plene' and 'defectiva'
spelling,unexpected vocalizations,manner of divisionintosedarim,
paragraphing of and
(positions petuhot setumot) and so on. Some of

1) E.g. Gn. xii 10, 11,14 "';", (3 times);Gn. xii 16 :'::1751 (f.7a); Gn. xxiv67
rg;nRn(f. 18b).
2) At least, MbU makes no mentionof themand theyare absentfromTR
(MG) and BH.
3) E.g. Gn. xix 26 (f. 12b) l"nNR; TR RI; BH R? (no gacyah).
4) Since I do not have access to the LeningradCodex, B19a, (whichwas the
Vorlage for BH), I cannotcompareM with thatcodex. It is, of course,now
commonknowledgethatBH is unreliablein its representation of some details
in the LeningradCodex. It is quoted herewiththe reservation that,whileindi-
vidual readingsmay not therefore be correct,thegeneral picturemaybe approxi-
matelyvalid.
5) E.g. Gn. xix 11 Inl. (f. 12a), as in BH; cf. NORZI, Minhat Shai,ad loc.; Gn.
xxi 6 pn;" (f. 13b).
6) E.g. Gn. xii 16 (f. 7a).

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 381

these will be discussed in this article,but plenty of furtherwork


remainsto be done.

Examinationand Dating ofM


Palaeographical
It will by now have become clear thatM has a special position in
relation to the BA/BN classification.But what is the date of M?
Unfortunately, the finalleaf, which possibly containeda dated colo-
is
phon, missing.
Two methods of dating presentthemselves-by internalevidence
on theone hand and by palaeographyon the other.Masoretic studies
bristlewithdifficultiesand stillunresolvedproblems.Various features,
such as the dates of development of details of vocalization and
accentuation,(which vary a great deal in earlyMSS), are subject to
a varietyof divergentdatingsby different scholars.We are stilla very
long way fromhaving anythinglike a reliableabsolute chronological
account. The statisticspresentedabove may also be variouslyinter-
preted to indicate differingdates. It was thereforefelt desirable to
rely on dating by a method which is, at this stage, fairlyobjective.
Since the shapes of lettersand the relativeproportionsof the various
strokesof which theyare formeddifferfrom age to age, and these
differences are measurable,it seemed wiser to rely on palaeography
to establishan approximatedate than to build on internalevidence.
While the dated palaeographical material for this purpose is not
abundant,it is certainlyadequate.
M exhibitstwo main stylesof the Hebrew script,and the writing
is clearlyprofessionalpenmanship.The consonantal Biblical text is
in an excellent,indeed a beautiful,neat square script,while the Maso-
retic notes which are writtenin all the margins and between the
columnsare in a formof mashait, thesemi-formal bookhand developed
from cursive1). Separate palaeographical examinations of both
scriptswere made, using only dated material,mostly Bible manu-
scripts2). A preliminary examinationshowed thatM belonged to the
9th-12thcenturies,but this period was narroweddown upon further
study.In orderto date the square script,a letterby letterexamination

1) More preciselythe Masora Magna in the upperand lower marginsare in-


variablyin mashait, whereassome of the Masora Parva is, at times,ratherclose
to square.
2) It is myverypleasantdutyto expressmywarmest thanksto myfatherand
teacher,Dr. S. A. BIRNBAUM,forhis advice on methodologyand forthe use of
his unrivalledcollectionof photographsof Hebrewmanuscripts.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
382 E. BIRNBAUM

of thefollowingdateddocuments was made.Withtheexceptionof


number11,all ofthemareBiblicalcodices(or fragments)
ofthesame
generalcharacteras M 1).
1. "Cairo Ben AsherCodex of the Prophets",preservedin the
KaraiteSynagoguein Cairo.MS written 8952).
2. Leningrad,MS 2 Firkovich17. Written9303).
2
3. Leningrad.MS Firkovich159.Written 9434).
4. LeningradMS 2 Firkovich10. Dedicatedto publicuse ("de-
dared heqdesh")946; probablywritten
in the secondhalfof the9th
century5).
5. LeningradMS 2 Firkovich39. Written
9896).
6. LeningradMS 2 Firkovich115.Written994 7).

1) It will be noticedthatI have not includedthefamousAleppo Codex in my


palaeographicalstudy,since it is not actuallydated.Its calligraphy is, however,
similarto thatof MS 2 Firkovich17 (dated930) whichwas writtenby Shelomoh
ben Buya'a. A latercolophonto theAleppo Codex attributes its penmanshipto
a scribeofthesamename.(See I. BEN-ZVI,'The Codexof Ben Asher' in TextusI
(1960), p. 3-5). M. GOSHEN-GOTTSTEINhas pointed out (in 'Keter Halab ve-Ben
Buya'a ha-sofer'in Tarbiz xxxiii (1963), p. 149-156,with photographs)the
similaritybetweenthehandwriting ofthesetwo MSS. A. DOTANhas sinceargued
(in 'Ha-omnamninqad KeterHalab biydeiAharonben Asher' in Tarbiz,xxxiv
(1965) p. 136-155)thatShelomohben Buyca' cannotbe thescribeof the Aleppo
Codex, as claimedin the latercolophon,becauseof differences betweenthetwo
MSS in calligraphyand in the numberof lines in the paragraphfollowingthe
Song of Moses. For thesereasonshe does not believe thatthe Aleppo Codex
was furnishedwith its Masorah by Aaron ben Asher himself.-It would seem
quitecredibleto me thata singlescribecould use one Vorlageat one periodof his
lifeand anotherlateron. As to theminorcalligraphicdifferences,a man'swriting,
in Hebrewas in otherlanguages,maywell changea littlein a periodof 20 years
or more.Every difficulty is not resolvedby suggestingtherehas been forgery.
Whether DOTAN has disproved GOSHEN-GOTTSTEIN'Sthesis or not, BEN-ZVI's
"inferencethatthe Aleppo codex was writtenat the end of the ninthcentury,
at anyratenotlaterthantheyear910" (TextusI, p. 5) seemsto me to be a reason-
able date palaeographically.
2) Photographs: BIRNBAUM, The Hebrew Scripts (HS) II plate 92; colophon
in KAHLE,Geniza,plate 7. These photographs,as well as those noted in the
footnotesto the otherMSS below,are citedbecausetheyare easilyaccessibleto
colleagueswho may wish to checkmydating.In myown palaeographicalexa-
minationI oftenused photographsof additionalpages of these MSS in my
father'scollection.
3) Photographs:KAHLE, MdW, I, Tafel 17/1;Tarbiz,vol. xxxiv(1965) facing
p. 152; TarbiZ,xxxiii (1963) facingp. 156, reproducedfromD. CHWOLSON,
CorpusInscriptionum Hebraicum,1882,no. 104.
4) Photograph: KAHLE, MdW, I Tafel 18/2.
5) Photograph:KAHLE,MdW,I Tafel19/3.
6) Photograph: KAHLE, MdW, I Tafel 21/5.
7) Photograph: KAHLE, MdW, I, Tafel 22/6.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 383

1009in Egypt1).
7. LeningradMS B19a. Written
8. LeningradMS 2 Firkovich225. Dedicated to public use
1017 2).
("heqdesh")
9. LeningradMS 2 Firkovich58. Written1021in Kufah(Baby-
lonian) 3).
10. LeningradMS 2 Firkovich26. Dedicatedto public use circa
1036 4).
11. Cambridge MS. T.-S. 20.42recto.Marriage
Library.
University
of Yehezqeyahuben Shelomoh.Writtenin Fustat
deed (ketubah)
(Old Cairo), 1062 5).
12. LeningradMS 2 Firkovich144. Date of finaldedicationin
colophon(fol 223 a) 11216).
Comparing eachletterofthealphabetin M withthecorresponding
letterin all the datedMSS beingutilised,the resultsshowedthat
M had theclosestand mostfrequent withno. 2 (written
affinities in
930) of the above list.Next came no. 1 (written895), followedby
no. 4 (dedicatedforpublicuse in 946 but probablyup to a century
older),and,ratherless close,no. 5 (written 989). The Squarescript
of M shouldtherefore date fromabout the firsthalfof the tenth
century.
A separatepalaeographicexamination was made of the mashait
scriptin whichthe Masoreticnotesof are written.It is worth
M
pointingout that,whileitsformsare clearand thehandis obviously
thatof a prefessional, the mashaithere employedis considerably
morecursivein stylethanthe Masorahof the majorityof Biblical
MSS of thistype(suchas thoselistedabove) whosescriptis almost
Square.The shapeofthealefin M is a good example.Unfortunately,
the earliestmaterialsin a rathersimilarmashaitare undatedand
variouslyassignedbetweenthe ninthand eleventhcenturies, so I
was obligedto compareM's mashait withtheverylimitednumberof
dated documentswhich (to my knowledge)survive.The oldest

1) Photographs: BIRNBAUM,HS, II plate 186. Another page in KAHLE, MdW,


I, Tafel23/7.This MS is of coursethebasisofBibliaHebraica,3rd and lateredd.
ed. by KITTEL and KAHLE.
2) Photograph: KAHLE MdW, I, Tafel 24/8.
3) Photographs: BIRNBAUM,HS, II plate 93* (last plate in volume). Another
page in KAHLE MdW, I Tafel 25/9.
4) Photograph: KAHLE, MdW, I. Tafel 27/11.
5) Photograph in Dr. S. A. BIRNBAUM'S collection.
6) Photographs: KAHLE, MdW, I Tafel 30/14 (cf. also p. 77); BIRNBAUM,
HS II, plate93 (finaldedicationcolophon).

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
384 E. BIRNBAUM

of thesewhichis accessibleto me byway of a photograph,is T.-S. 12.9


in CambridgeUniversityLibrary,and dates fromabout 979,but there
with certaindocumentsreproducedin S. A. BIRN-
are also affinities
BAUM'SHebrewScripts,namelyplates 185, datingfrom956 (an Egypt-
ian MS) and 195, datingfrom963 (a BabylonianMS) 1). Early dated
materialin mashaitof a similartypewill no doubt become generally
available as facsimilesof more MSS from the Cairo Genizah are
published2). While the paucityof our available dated mashaitmaterial
makes a more precisedatingdifficult, M's Masorah is clearlywritten
in a hand fromabout the tenthcentury,so thatthe datingarrivedat
on the basis of the Square scriptis confirmed.
Where was M written?The MS, as we now have it, containsno
mentionof anyplace. If its colophon had survived,thismightpossibly
have yieldeda clue. If the MS had been some centuriesyoungerwe
could have ascertainedthe countryof origin fromthe writing,but
in about the tenthcenturythe penmanshipof Egypt, Palestineand,
frequently,Babylonia, too, was extremelysimilar. The constant
movement of scholars between the various centres of learning in
these countries,which was characteristicof those times, doubtless
helped greatlyto ensure a generaluniformity in the developmentof
the script,and more especiallyof theformalSquare style.This applies
also, though not in equal measure,to mashaitand even to cursive.
The mashaitof M would seem to me to bear a greaterresemblanceto
documentsoriginatingin Egypt than to Palestinianones, but I put
this impressionforwardwith some reserve.Before the question can
be decided with confidence,more mashaitmaterial,furnishedwith
place of origin,mustbe found and studied.

ScribalTechnique
(i) Line Fillers.At this point it mightbe apt to draw attentionto
the scribe's method of fillingup lines where there is not enough
space to writea completeword, or wherehe wishes to fillup the end
of a line for some Masoretic or other reason. Our scribe used a
diamond-shapeddot at the line ceilingwith a tail going down to the
leftat an angle of 45. Sometimeshe would curve this down at the

1) Later dated documentswithsome featuresin commonwith M, includea


marriagecontract,dated1002 (BodleianLibrary,MS Heb. a.2. 28051)and letters
of 1027 (CambridgeUniversityLibraryT.-S. 24.29) and 1029 (Bodleian, MS
Heb. a.3.-BP 5146.
2) Cf.now S. SHAKED,A Tentative ofGenizaDocuments,
Bibliography 1964.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 385

end, as low as the half way point between the line-ceilingand line
base, thus ,,. Sometimestwo or threeof these fillersfollow one an-
other,and occasionally two are linked together.The most striking
example of M's use of fillersmay be seen in the column preceding
the Song of Moses (Dt. xxxi 25 onwards,fol. 149 b, col. 2) in which
half of each line is taken up by them1). M never employsletter-like
dummysigns, such as are characteristic of some old MSS, including
the Aleppo Codex 2), nor the dilated letters(notably "nT;rnK,but
sometimesothers,too) which are to be met with in later centuries.
(ii) Deletions.Most words or lettersmarkedfor deletion in M are
cases of dittography.They are markedin severalways.
(a) Overdotting-adot is placed above each superfluousletter3).
Some letterseven have two dots above them4).
(b) Indotting-adot is placed inside each superfluousletter5).
(c) Overlining of the superfluousletter6).
Many of these mistakes(they were usually that) must have been
noticed soon afterthey occurred, since the naqdan often did not
vocalize such marked words. Still, it now and then happened that
such dotted words were vowelled and accented7), in spite of the
factthattheywere clearlydittographies.
(d) Several "superfluous" letters, (including genuine variant
readings) appear to have been scratchedout, although one cannot
tell when and by whom 8).
(e) In one case a letterhas been circledto indicatedeletion9)

1) See below p. 389 and plate C.


2) See GOSHEN-GOTTSTEINin Textus, I pp. 27, 56, 43 note 81; also p. 50 note
93. Some comparable Biblical codices appear to use dots only e.g. Leningrad
B19a, while others use both dots (sometimes extended with tails like in M) and
dummy letters,e.g. BM Or. 4445 (undated but probably early 10th century). See
reproductions: f. 98a in R. F. HOSKING and G. MEREDITH-OWENS,ed., A Hand-
book of Asian Scripts. London , British Museum, 1966, plate; and of f. 89, pu-
blished as British Museum postcard OPB 1.
3) E.g. f. 3a, Gn. viii 1; f. 5b bottom and 6a top, Gn. xi 14.
4) E.g. f. 5b last line.
5) E.g. f. 13a, top 3 lines, Gn. xix 39; f. 63a, first3 words, Ex. xxviii 12; f. 101b
(the Tetragrammaton!).
6) E.g. f. 23b, firstword. The overlining seems to be in a later hand. Similarly
f. 103b (Nu. x 18) where, apart from a heavy overlining, there are thinnervertical
lines to the left and right, enclosing the superfluous word on three sides.
7) E.g. f. 5b and 6a, Gn. xi 14 and f. 23b.
8) E.g. f. 146b, erasure of the letter1 of the rnKI(6th word) of Dt. xxviii 20,
thus bringing it into conformitywith TR.
9) On f. 26a. Cf. Appendix D, ad fol. 26a, Gn. xxxi 51. Regarding circling as a

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
386 E. BIRNBAUM

(iii) Omission The whole of M is vocalized and


of vocalitation.
accented,withtheexceptionof one sequenceof passages,wherethe
vocalization(but not the accentuation!)has been omitted.This
was clearlydone on purpose,sincethe passages,(f. 99a, [plateB]
Nu. vii 24 onwards,concerning theofferings of theprincesat the
dedicationof the Sanctuary)are almostidentical.The firstand last
verseof eachof theofferings,containing thenameof theprinceand
tribein question,is, however,fullyvocalized.

Vocalization
The vocalizationis basicallymoreor less standardTiberianbut
thereare a considerablenumberof differences fromthe formsto
whichwe are accustomedin TR (MG etc.) and BH. A thorough
examinationand interpretation is required.Listedbelow are a few
of
examples interesting vocalizations thatwerenoticed:

Gn. xviii23 (f. lla) l,. (BH, MG r5) 1).


Gn. xxi22 (f. 14a) [sic]mv (BH, MG ns2).
Ex. xv 17 (f. 51a) Wtn'rj?
(BH, MG twnr;pt).
Dt. xxi3 (f. 142b)-;:. (BM, MG -ran)2).
Dt. xxvi29 (f. 146b)nlzi (BH, MG 'nlu).

Shapeof Vowels, etc.:(1) Qamesis written in two ways,as and ;


bothformsare oftenusedin a singleword3).
(2) Shewais usuallyverticalbut sometimes thelowerdot projects
to therightthus.
(3) Rafeis used extensively, not only on the lettersn"Sa mtr
butalso on alef,(e.g. ian1,), he(e.g. TrhK)
and sometimes on nun(e.g.
Gn.
nlf,., xxii11, f. 15a).
We havealreadynotedthatthevocalizationwas doneas a separate
stage,veryprobablyby a different, specializedscribe(naqdan)since
thevowelsare oftenwritten in a differentink thantheconsonantal
text;theinkof thevowelsis frequently morefaded.

methodof deletion,see AlexanderSPERBER,A HistoricalGrammarof Biblical


Hebrew,p. 523.
1) NoRzI, MinbatShai, ad loc.,remarksthatcarefulcodices read withpatah,
presumably againsta readingsuchas M's.
2) M possibly re-inked?
3) E.g. ,t-l, f. 13b, col. 21 line 3.
.r

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 387

Accentuation

Shapes of the Accents.These are generallyin the archaic forms


typicalof the Aleppo Codex and otherearlycodices. One remarkable
feature,however,is the use of a shortdownstrokeat an angle beneath
a letter,thus ,. It can indicate not only merkhabut sometimesalso
ga5yahand silluq(although the two latterappear more frequentlyin
theirregularverticalposition). Converselya shortverticalbar beneath
a lettercan sometimesrepresentnot onlyga!yahand silluq,but also
merekha or tippeha1).
Use. While the accentuationis basically the usual Tiberian one,
thereare a considerablenumberof divergencies.In the few examples
below, M's readingsare compared with the accentuationin BH and
TR (as representedin MG).

8X;:
Gn. xvii 6 (f. 9b) M crIn+jrlMn7a lnX n^sm
Ipm
TR, BH Txipnn nxb tnxani n?Nn
Gn. xviii 8 (f. 10b) M BH -, ;
TR ; Rim
Gn. xix 21 (f. 12a) M U's
BH, TR 2). r,3 '
Gn. xxi 23 (f. 14a) M-nn
BH, TR i nsn7
Gn. xxii 1 (f.
14b) M n
Gn. xxii 3 (f. 14b) M
BH, TR n$j,

Gn. xxiv 31 (f. 17a) BH,


M TR npw,
BH, TR 'iR~
Gn. xxv 11 (f. 18b) M

Gn. xxv 13 (f.


BH, TR 3). Tp31 n~1 bEy^r ~:

1) Examplesare verynumerous,e.g. (f. 15b-17b):Gn. xxii21; xxii23; xxiii7;


xxiii 17; xxiv 2; xxiv 31; xxiv 38; xxiv 40; xxiv 41. These accentsat 'unusual'
anglesare interspersed withwordsbearingthesameaccentsat their'usual' angles.
This suggeststhe possibilitythatthe angle may indicatesome accentualdetail
such as lengthor continuity.
2) NORZI, ad loc., supportsthe [BH, TR] readingwithZaqefqatan on the basis
of MSS, presumably againstthereadingwithrevicarepresented by M.
3) Accordingto NORZI, ad loc.,the "best" MSS read Inl2 withzaqefqatanand
'11j with ma'arikh.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
388 E. BIRNBAUM

andMinuscules
Majuscules
According to Masoretic traditions,there are certain exceptional
cases when a letteris writtenlarger(majuscule)or smaller(minuscule)
thanits neighbours.
Majuscules:A comparison of the practice of M with a list of the
majusculeswhich are standardfor the Pentateuchat the presenttime
showedcomplete fortheextant
identity ofM (totalling
portions fourout
of the eleven or twelve "standard" majuscules)1).
The contraryis thecase forminuscules.Six (or according
Minuscules:
to some authorities,seven) minusculesare standardMasoreticpractice
at the presentday. Three of the passages are on pages now missing
fromM 2), but, of the remainingthree or four3), nota singleoneis
writtenin minuscule,or is noted as such in the margin.

The"Dots" 4)
The Masorah prescribesthata dot (nequdah)be placed above certain
letters of the Hebrew text in specific passages. Early MSS differ
somewhatas to how manydots thereshould be in each case and on
which letters.The number of passages cited where dotting occurs
is 15, of which 10 are in thePentateuch.The MP of M also gives these
figures5). In M dots appear in all the passages concerned,but some-
times on different lettersthan in TR and BH. Details are given in
Appendix C.

1) Ma1usculesinM
I in ntimtl (Lv. xiii33)
' in 7X'W(Nu. xiv 17)
3 in 1213(Ex. xxxiv7)
1 in 1]3t) (Nu. xxvii5).
The remainingpassages containingmajusculesaccordingto the standardlist
wereon pagesno longerextantin M. We cannotof coursebe surethattheywere
majusculesin M. These passages' are: 2 Gn. i 1; I Dt. vi 4; H Dt. xxxii6; 1 Lv.
xi 42; ? Dt. xxix27; v Dt vi 4; Ex xxxiv14; (and accordingto someauthori-
tiesS Dt. xxxii4).
2) K Lv. i 1; H Gn. ii 4;
' Dt. xxxii18.
3) Z Gn. xxiii 2; 3 Lv. vi 2; p Gn. xxvii46 (and accordingto some also 1
Nu. xxv 12).
4) GINSBURGuses the term "Extraordinary Points".
5) Theyare referred to in the Sifrito Nu. ix 10, whereMidrashicexplanations
of the use of the dots are given.For fullreferences to the sourcessee Talmudic
Encyclopedia,art. Otiyot,vol. 1 p. 190-192(4th ed., Jerusalem,1955). Cf. also
p. 318 ff.
GINSBURG, Introduction

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
VETUS TESTAMENTUM XVII, 4

i
A, * .*w
'X, .,1
'*4",a
*V"'",
*I, ,
Iflo~~~.10
IV
j
W,,rt,.
W

/.

Plate A
MichiganCodex.f. 92b. Lv. xxxii 30-Nu. i 16
Note(1) col. 2 line 11 "'It "'1T (article,pp. 394-395)
(2) col. 2 line 12 '1:Dt' (article,pp. 374, 378)
(3) col. 2 line 16 '1-1'I3 (article,pp. 394-395)

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
XVII, 4
VETUS TESTAMENTUM

~'*'tnn^ 1^St 2;! t


AOT,
I w^^,,,w^^ *. t
=tt$
"3 7wa ^ . 2

* ! *e . j' <- .^ ^ .
' *'*
_ ":' *. ' _ -'I '

in conflict
(4) MM bottom p. 394note1).
withMM on f.100b(article,

PlateB
Codex.f. 99a. Nu. vii 15-vii32
Michigan
Note (1) col. 1 lines 1-2: 'fliX D"-'I (article, p. 378, and note 5)
(2) col. 1 line 4: DrT7fB(article, p. 410 and p. 395 note 6)
(3) col. 2. Unvowelledbut accentedtext(article,p. 386 and p. 378 note5)

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
VETJUSTESTAMENTUM XVII, 4

.* I
Is11
. .

,H^_I
AM

: f.
#ii I .1
I

Plate C
AlichiganCodex. f. 149b. Dt. xxxi 20-xxxi 38
Note(1) col. 1 line 12. Pelugta(article,p. 393)
(2) Use of line-fillers(article,p. 384)
(3) Ornamentalarrangement of Masorahin top margin(article,
pp.396-397).

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
VETUS TESTAMENTUM XVII, 4

Eu
!T.p
4
-
a
m 4

1
1 ta%
V"w1m b4 aNS
_
R

I
I1-
4
1.

'f
..

Plate D
MichiganCodex. f. 150a. Dt. xxxii 22-xxxii 36
Note (1) 70-line (not 67-line) arrangementof Song (article, p. 389)
(2) col. 2 line 4. Plene spelling of 1K;H (article, p. 411)
(3) Ornamental arrangementof Masorah between columns (article, p. 398).

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 389

TheSongs
A point of some interestin Masoretic studieshas been the manner
in which the Song of the Sea (Ex. xv) and the Song of Moses ('Ha'a-
jinu',Dt. xxxii) are written.
1. The Song of Moses. MAIMONIDESprescribeda 67-line arrange-
ment1) for this song as being the Ben Asher systemand thatis still
the arrangementof many Spanish and all YemeniteMSS. According
to the ancient traditionembodied in Tractate Soferim2), however,
the Song of Moses is to be writtenin 70 lines, which is indeed the
practiceof TR, and is the presenthalakhahaccording to the Shulhan
'Arukh3).
In M, the leaf which contained the first21 verses is missing4),
but the arrangementof the extant portion is identical with the
70-linesystem5).
The scribe of M was at pains to make sure that, preceding the
Song of Moses, the word npIVN (Dt. xxxi 28) should stand at the
head of a page 6) in accordance with scribal custom (l"tzn n"n3).He
achieved this by making the column before that (f. 149b col. 2,
plate C) containveryfew words per line, and fillingup the remaining
part of each line with dots extendedby long tails7).
The Song is followed by a blank line (petuhah)and the next para-
graph consists of 6 lines beginning(Dt. xxxii 44) (i) Klc (ii) stim
(iii) a'Iw (iv) n:= (v) rn; (vi) arz. This differsfrom the usual
TR arrangement:(i) n:m (ii) asn (iii) 3ni:nnm(iv) a:=:= (v) uttK
(vi) rnnn and fromthe Aleppo Codex and Maimonides arrangement

1) MishnehTorah,vol. 2: Ahavah,Hilkhot SeferTorah, ch. 8, halakhah 10.


M. GOSHEN-GOTTSTEINhas proved conclusively from MSS that the original
textof Maimonidesread "67 lines" not "70 lines", as in the printededitions.
See "The Authenticity
of theAleppo Codex" in Textus,1 (1960) p. 33 ff.
2) Ch. 12, halakhah 9.
3) YorehDecah, 275, halakhah 5.
4) f. 150a beginswiththefirstextantline,starting'pi9n.
5) I.e., in Dt. xxxii25 the hemistichsare arrangedin the usual TR manner:
fromrn1? to 'in, from1"n'lD1 to ;HDIR,fromll to n1'ln, and fromp3r1to
MlI/t(see plateD), in contrastwiththeMAIMONIDESand Aleppo Codexarrange-
ment,which incorporatestwo verylong hemistichs, fromr1n1Dto 'Tnl, and
from1t to 7132.
6) This page is now unfortunately missing,but is indicatedby the catchword
at the bottom of f. 149b col. 2 in a later hand.
7) The Aleppo Codex, facedwiththe same problem,filledup alternate lines
completelywithdummyshapes thatlook like partsof letters,ratherthanonly
half lines like M. See plate containingAleppo Codex, Dt. xxxi 12-xxxi28 in
Textus1 (1960). On line fillerssee above p. 384.
Vetus TestamentumXVII 24

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
390 E. BIRNBAUM

of five lines: (i) KRt'(ii) a1t7 (iii) 'IR (iv) nrtn(v) RK1). Since M
is not a Torah scroll, we cannot be certainthat the arrangementof
these lines is intended to be on the "official"pattern,but the care
which the scribetook to place nrT'K1at the head of a column,and to
arrangethe Song according to the tradition,makes it reasonable to
suppose thathe was followinga specifictraditionhere also.
2. TheSongoftheSea (Ex. xv). This is anothersubject of difference
in earlyMSS. Once again a missingleaf in M deprivesus of the first
half of the song. The extantpart in M begins (f. 51a) at ar:,7v (Ex.
xv 10). The lines are set out exactlyas prescribedby MAIMONIDES,
with the exceptionof the finaltwo lines. Both are arrangedin three
groups of words each, while in MAIMONIDES the penultimateline
has only two. The final two words in MAIMONIDES' penultimate
line are here placed at the beginningof the last line, so that (a) the
groups consist of different words, and (b) each line has nine words
(against MAIMONIDES' 11 + 7).
WnAY'n
M:Qf Ua
fnlmIn'V n=l2 n7nIV lcD NK
.^ s:En lann
"pm "^nl ID n *m t"n '~-nX
TextualDivisions
The Biblical textis dividedin M in fiveprincipalways: into books,
verses,parashiyot,sedarimand paragraphs('open' or 'closed').
1. Books are separatedby severalblank lines 2).
2. Verse divisions are indicated by two small verticallines, one
above the other,which togetheroccupy half the height of a letter
fromthe top of the line ceilingdown, thus :a 3).
These two formsof division are uniformin most MSS, including
M. The use of the threeothertypesof division varies in the MSS.
3. Parashiyotare the divisions into weekly pericopes of Torah
readingsaccording to the annual cycle. In M theycorrespondwith
the standardMasoretic practice representedin TR. The beginning
of each one is markedby the letters 'nt in the margin.In some cases
the scribeof the Masorah has written,in the blank lines betweenthe
end of one parashahand thebeginningof thenext,the numericaltotal

1) On this problemcf. M. GOSHEN-GOTTSTEIN'The Authenticity' in Textus


1 (1960) p. 42 ff.
2) f. 92b betweenLv. and Nu., plateA; f. 127b betweenNu. and Dt.
3) The small verticalstroke,called paseq, used to preventphoneticliaison,
extendsfromthemiddleof thelineto linebase,thusl: but in lessformalwriting
it is oftenas tallas a letter.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 391

of the verses containedin the precedingparashah,and generallyalso


the standardname with the same numericalvalue (gematria)used as a
mnemonic.But not infrequentlythese have been put in by a later
hand.
4. Sedarimare the divisionsinto weeklysectionsof Torah readings
according to the triennialcycle. In M the beginningsof these are
indicatedthroughoutthe MS by a large ornamentallettero in double
outline, in the margin opposite the beginning of the firstverse of
the seder(sidra). Early Masoretic MSS differconsiderablyin their
divisionsinto sedarim1).
A comparisonof M's divisionofsedarim in Genesiswiththedivision
given in Mishael ben 'Uziel's Kitdbal-Khilaf(embodyingthe BA and
BN traditions)yieldedthe followingdifferences 2): of the 46 sedarim
recordedforthatbook by MbU, M shows no sederin eleven instances
where MbU calls for one 3). In three of these cases the sederin M
begins one verse earlierthanit does in MbU 4). We cannot conclude
that M was merelymaking the same kind of mistake three times
over, since,in one of these cases at least, the different sedergiven by
M is attested elsewhere5). Furthermore,a completely extra seder
appears in Ex. 6). I leave it to othersto compareM's sedarimn arrange-
mentin the restof the Pentateuchwith thatof otherMSS.

(Petuhotand Setumot)
Paragraphing
The paragraphingof the Biblical text is a point of considerable
importancein Jewishlaw. This is particularlythe case forthe Torah,
from the officiallyaccepted paragraphingmake a
where differences

1) Cf.,e.g., thelistsgivenby JacobMANN in TheBibleas Readin theOld Syna-


gogue,Vol. 1: The PalestineTriennialCycle: Genesis and Exodus. Cincinnati,
1940; and GINSBURG,Introduction,
p. 472.
2) WherelacunaeexistI have assumedno difference fromMbU. Furthermore,
as thebeginningof a parashahmaybe assumedto be also thebeginningof a seder,
I have construedthe absenceof the letter0 in themarginin thatpositionto be
merelya scribalconvention,and not countedsuch absencesamong the "differ-
ences". Nevertheless, in one instanceat least (Ex. xxvii20, f. 64a) the scribehas
actuallywrittenbothttS'1and 0 in themargin.
3) Gn. xvii 1; xix 1; xxi 1; xxii 1; xxii 20; xxiv 42; xxvii 1; xxx 22; xxxi3;
xl 1; xli 38.
4) M MbU
xviii33 xix 1
xxvi34 xxvii1
xxx21 xxx22
5) MANN, TheBible..., Seder24a beginningat Gn. xxvi34.
6) Ex. iii 16 not attestedin MANN, TheBible...

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
392 E. BIRNBAUM

scrollunfit(pasul)forpublic(i.e. synagogue)use. Medievalmanu-


scriptsand scrollsare known to have variedconsiderably. MAI-
MONIDEShimself complains ofthisand setsoutto remedythe situation
by listingall the 'open' and 'closed' paragraphsaccordingto the
"Ben Ashercodexwellknownin Egypt"1).
M is nota scroll,butitis,nonetheless,generally to observe
careful
thedistinctionbetweenpetuhot and setumot.
A comparison of M with
yieldsthe followingdifferences
MAIMONIDES 2):
No. ofinstances
A. M haspetuhah whereMAIMONIDEShas setumah 6 3)
B. M has setumahwhereMAIMONIDES haspetuhah 84)
C. M haspetuhah whereMAIMONIDEShas no paragraph2 5)
D. M. hassetumah whereMAIMANIDEShas no paragraph7 6)
E. M has no paragraphwhereMAIMONIDEShaspetuhah2 7)
whereMAIMONIDEShas setumah
F. M. has no paragraph 4 8)
Total paragraphing betweenM and
differences
29 1)
MAIMONIDES

1) See MAIMONIDES, Torah,vol. 2 Ahavah.HilkhotSeferTorah,chap.8,


Mishneh
halakhot3-4:
: .nDlD
IN 9"17n ,nlr n,z p0nnal nmnn0 3n tnb
nn
nn i ,r:r
mIr
1rimn n
rn p nrm nm mn 1n
5Imo-I I,nI'v1 nin:
alpa1 apwt IKN
nn;1on
of
cannot
":I) courseknow howKKtt
We n"r mn w i

) xvii;
Gn.x1; xxxiii18; xxxix1; Ex. iii 1; vi. mn
Gn.xi 1;
) xxvii E.i i; 8; i n;xv 15; x1n ; xiiL.
trjn"s? pn1P '"i: ni-rvnnnisimrin&min mmmnonr
)Lv.
Ex.
viOivii
n22.1
N tti 1n1; n no
VtVVY1 nsY1 -1 ^InQttl o^tV 1YIT1 $&on1 HNI
pagesnv
scattered which are now
oon i?z8? n
missingfromM, iandwhichconstituted
;rn* tatw n;1=7 tr"Q1v ! niv tv-10
27% of

81) 1n;i i
xxvEx.
14.1oo 1 Lv. 1nX lYo
2) We cannotof courseknow how manyfurther therewere in the
differences
27% of
scatteredpages whichare now missingfromM, and whichconstituted
originaltext.
itsoriginal
3) Gn. xvi 1; xx 1; xxxiii 18; xxxix 1; Ex. iii 1; vi 1.
4) Gn. xxiii 1; xxxvii 1; Ex. i 8; ii 1; xiv 15; xvii 1; xxii 13; Lv. xiii 29.
5) Ex. viii 1; Lv. vii 22.
6) Ex. ii 11; Lv. xxv 14.
7) Gn. xlix 5; (after xlix 4 a later hand has inserted the Arabic word 7X5 fasl
["division" or "section"] to indicate the missing petuhah);xlix 14.
8) Gn. xxxiv 1; Lv. xxv 29; xxv 47; Nu. iv 29.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 393

These maybe otherwise


summarised
as follows: No. of
instances
1. M's paragraphingcontradictsMAIMONIDES'"BA"
prescription 14
2. M's paragraphing or addi-
petuhah setumah)
(whether
tionalto MAIMONIDES' "BA" prescription. 9
3. M's omissionof paragraphing
(whether or
petuhah
setumah)prescribedin MAIMONIDES' "BA". 6
Total paragraphing differencesbetween M and
MAIMONIDES 29

MAIMONIDESsets the total number of paragraphsin the BA text of


thePentateuch
at 669. Thus the29 differences
notedin M constitute
nearly4.35% 1).
PleneandDefectiva
Spelling
M sometimes usesplenespellingswhereTR uses defectiva
and vice
versa.OtherMSS sometimes recordthesamereadingsas M, contrary
to TR. NORZI's MinhatShai and otherlate Masoreticworksalso
recorddifferences fromTR, sometimescoincidingwith M, not
to
infrequentlyrejectthemas inferior. laterhandshave
Occasionally,
corrected M, usuallyin thedirection
of with
conformity TR, by the
insertionof"missing"lettersabovethelineor byrewritingtheword
in the margin,or, sometimes, by scratchingout a "superfluous"
letter.A selectionof examplesofplene/defectiva is given
differences
in AppendixD.

Pelugtot
We haveseen,in manyof theprecedingsectionsthatfrequently
M does notconform to whatwe knowofeitherBA or BN. Nor does
it conformconsistently
withthe littlewe know of the practiceof
severalotherearlyMasoretes.At times,however,its readingsdo
supportone Masoreteagainstanother.In mostcases,thedifferences
betweenM and othersare, of course,tacit,i.e. theyare noticeable
onlyin a wordby word examination. However,fromtimeto time
the naqdan(or occasionally,one of severalother,probablylater,

1) This percentageassumesthatthe 27% of M thatis no longerextantwas


identicalin paragraphingwith Maimonides' BA prescription.If we assume,
more reasonably,thatthe proportionof differences
therewas similarto thatin
therestof the MS, we get about 6.2% paragraphing
difference.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
394 E. BIRNBAUM

hands) has repeatedthe word in the margin,with a different vocali-


zation or accentuation,followed by the word if5t (i.e. "disagree-
ment"). Most of these are notBA/BN variantsrecordedby MbU in
his Kitdbal-Khilaf.In severalinstancesno explicitindicationis given
in the marginas to which vowel or accent is the subject of the dis-
agreement.The full list of these Pelugtotis given in Appendix E,
(p. 411).
I have not noticedany referencesin M to specificMasoreticauthor-
ities by name. Apart from the abbreviationilft (and once ibr),
attentionis drawnto a disagreement by thewords:in rnrnr ntin
the margin(f. 133b re rpi Dt. vii 26) and (once in a laterhand, f. 5a,
Gn. x 9) ki, "anotherversion".
In one note in the MM there is mention of the reading of the
Ma'arbae (and thus,by implication,of the Madinhae)1). In the original
naqdan'shand thereis one marginalreferenceto the Mugah nimn, the
famouscodex whichis mentionedin otherearlyMSS, but is no longer
extant 2).

"Two-partnames"
It is well known thatcertainnames,which consistof two elements
juxtaposed,are variouslywrittenin Biblical MSS as one word or in
two words 3) (and, in the lattercase, with or withouta maqqef).An
examinationof M shows the following:
T'p :n' In twowords (Gn. iv 22, f. 2b)
n
mnlt3 In oneword (Gn. x 26, f. 5a)

1) On f. 100b:NrlnRK nrninxR R?5frIUn?nK7 11 R ':] r iO i:


n'5
*. ;lT?3
"11Y Di
`-715V iOD Kt;l [etc.] ... It is worthnoting a in the
difference
formof the Masoretictraditionhere.Whilethe same figure,of therebeing 22
verseswithoutwaworyod,is mentionedfourtimesin theMP on ff.99 and 100,
the earliestoccasion when thismatteris treated,on f. 99a (MM to Nu. vii 20,
plateB) has theclear reading"21 verses": RKil n Hrp p
lp2 r105'nlmrn
lm
[etc.]...1i.l K1 nln RK1rn")'O1 1I
1K1KR51
Both the 21-verseand the 22-versereadingsare knownfromotherMSS, e.g.,
the 22-verseformin the MM to Nu. vii in BritishMuseum (BM) MS Or. 9401
(Ashkenazic,dated 1286) and the 21 in BM MS Or. 15250 (Sefardic,of 1493)
f. 77 and 78a.
2) f. 116a on Nu. xxiii15. The consonantaltextappearsto be accented;lpjK
(withreviac), thoughthisis not quite clear.The naqdanhas notedin the margin:
KR;1l13t(i.e., with merekha).The latteraccentuationcorrespondsto TR and
BH.
3) The differencehas beenclaimedto represent a difference between
respectively
the recensionsof the Madinhae("Easterners" i.e. BabylonianJews) and the
Macarbae ("Westerners" i.e. Palestinian Jews). Cf. GINSBURG, Introduction,
p.
199 ff.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 395

7 In oneword (Gn. xii 8, f. 6b and passim)


lYsT53~ In oneword (Gn. xiv 1, f. 7b and passim)
TY' In oneword (Gn. xxxi 37, f. 25b)
'W01 In twowords (Nu. i 6, f. 92b and passim)1)
1rIS-;17 In twowords (with maqqef).(Nu i 10, f. 92b and passim2).

The Masorahin M
A small circleabove a word of the consonantalHebrew textindi-
cates (in accordance with the standardMasoretic convention) that
there is a comment on it in the Masorah Parva, usually (but not
invariably)in a parallel position in the margin. Only some of these
items are the subject of a fullercommentin the Masorah Magna.
In M the MM sometimescommentson words not even mentioned
in its MP, and occasionallythe circleabove a word in the consonantal
textrefersstraightto the MM and not firstto the MP 3). Sometimes
a word commentedon in the MM is unmarkedby a circle in the
consonantal text,and not indicated in the MP 4). Occasionally, for
lack of space or by error,theMM is not on the same page as theverse
to which the commentrefers.There are conflictsbetween MP and
MM; some are doubtless scribal errors, others presumablyresult
from the conflationof differenttraditions5). Conflicts occur for
similarreasons between the textof M and its Masorah 6). Sometimes
the MP seems to be wrong, possiblythrougha scribalerror7).
Many of the Masoretic notes contain informationsimilarto that
in otherMSS and in MG, BH, etc., althoughthe actual wordingmay

1) See plateA, col. 2, line 11. AlthoughmanyMSS writethisin two words,


BH and MG have it in one word. This coincideswiththe recommendation of
NORZI,ad loc.,quotingtheTalmudYerushalmi,TractateMegillah.
2) See plate A, col. 2 line 16. In addition,GINSBURG records (Introduction,
p. 200) thatthe expression"fromus" is alwayswrittenas I1.3?withsegolbelow
memand dageshin nunby the Macarbae,whereasthe Madinhaewrite1312?(with
serebelow memand rafeon nun).M alwayspoints1371M.
3) E.g. Gn. viii 2 (f. 3a) X'1; Gn. viii 6 (f. 3b) 1k1Il.
4) E.g. Gn. xxi 26 (f. 14b) Inl nZ11.
5) E.g. on Gn. xiii 3 (f. 7a) MP reads '7S '71 1;i 12n1k iD 7 as againstMM
1o5 A.The latteris in accordancewithTR (MG).
6) E.g. Nu. vii 17 (plateB, f. 99a, line 4). Text `7'T1Y,correctedto `Tl7nlby
a laterhand, whichalso added in the marginan extra MP in3 } i. That the
plenereadingwas regardedas authenticby theoriginalMasoreteof M is evident
fromthe MM appearingon f. 99b:
E.g. Gn. xxii24 (. m15b)
nx?) r pp"mr e n11 int ;mInn
ai nsa
7) E.g. Gn. xxii 24 (f. 15b) 1t71En marked 7 instead of i.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
396 E. BIRNBAUM

be slightly Yet somenotesseemto be peculiarto M. Much


different.
further workremainsto be done on thissubject.Sometimes M will
itselfgivethesameinformation placesin varying
in different forms1).

Decoration UseoftheMasorah
andDecorative
Quiteapartfromthefinecalligraphy ofthemainscribe,thosewho
produced our MS worked on it with considerableartistry.The
textoftheMM itselfis mostlyin plainstraight lines.At times,how-
ever,itis writtenin simplegeometricalpatterns,butitalwaysremains
entirelylegible2). Ornamentation occursin M:
1. At thebeginning usuallymarkedbya largeornamental
ofa seder,
outlinelettersamekhin the margin,with ornamentalflourishes,
chieflyaboveit.
2. At the beginningof a parashah, indicatedby the lettersvt5,
withflourishes usually above and below.
3. To markthe "middlepoints"of the Torah and of individual
books,indicatedby a largeornamental designabove and below the
words"middleof..."
a) "%DOMln (f. 20b; 55a 3); 110b) 4).
b) io: ;rinn nM (f.73b at Lv. viii 8) 5).
c) inhPnn rn (f.75a at Lv. x 16)
4. The wordsoftheMM arenotinfrequently arrangedin a number
of simplegeometricalpatternsin the upperand/orlower margins.
Among them are thefollowingtypes:

1) E.g. the followingmidrashicMasorah.In the MM to Gn. xxii 11 (f. 15a)


it appearsthus:
mt Xrrn 5Knv nl2
at =p5a.2p twK rtln Plnwnm"t
pn n pa MW 'T
o-o miP nsan n;* nrn In1n2Pon PlnSI ;ntn
The MM to Ex. iii 4 (f. 41a) is wordedthus:
IDX5 n n7np
prX,np In Rn~mnmi 1pn*asnQin
p"O,nnM mn nsn2I
o-o pD 528n Tma1?
1Tn
Inrm r van P5wn PmniN
II p. 635 para.
worded,longerformin GINSBURG,TheMassorah,
Cf.a differently
637.
2) The writingof the Masorah in patternsbecame quite common in later
increasedthe actualtextof the Masorahtendedto
centuries,and as the artistry
becomemoreand morecareless.HoweverM's use of patterns, albeitsimpleones,
seemsto me to be unusuallyearlyforthistypeof MS.
3) At Nu. xxii17, and againstMG and manyMSS, whichhave it at xxii27.
4) The latterapparently to Nu. xvii 20 (= MG and GINSBURG'S
referring
edition,and againstBH whichhas it at Nu. xvii 19).
5) As mostMSS, butagainstTalmud Babli, Kiddushin7a, which places it at
Lv. xiii 33.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 397

(f.2a bottom)
/ii

(f 2b bottom.)

(f. 14b top)


i\
(f.29b top)

/\ /\ /\
(f. 53b top)

(f. 110bbottom)

(f. 150btop)

(f. 151atop)

/^^vy^ f ^

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
398 E. BIRNBAUM

The most elaborate patternoccurs in the broad space between the


parallelcolumnsin the Song of Moses (f. 150a) where the MM takes
thisform1):

5. The MP too is sometimesgrouped ornamentallyinto the upper


and/orlower margin(normallythe location for the MM) when the
MP notationsconsist of one word, followed by the abbreviation'i
or i: ' ("unique"), in thisfashion:

6. The sign markingtheend of a MM note, a horizontalline with


a small circleat each end, is sometimesreplacedby more complicated
developments,in the form of triangles,diamonds, etc. with small
circles at their junctures; at times small additional flourishesare
attachedto them2).

Damage
Some leaves are ratherworn; the surfaceof theparchmenthas been
rubbed and, occasionally,damaged by water. In several cases whole
pages have been somewhatunskilfullyre-inked3). Not infrequently
the edges of leaves have been worn off.A number of leaves have
ragged holes in the middle,or big pieces broken offat the top or the
bottom4). Part or all of the bottommarginof severalleaves has been
cut offwith a knifeor scissors. The rectangularshape of
deliberately

1) See plate D.
2) A laterreader,manycenturiesago, scrawleda six pointedstar(magen
David)
betweenthecolumnsof f.27a.
3) Especiallyff.134b,135b,136a.
4) Badlytornand defectiveleavesincludeff.135 and 143,whichlack thelower
thirdand halfrespectively.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 399

most of these cuts suggests that the pieces were used-perhaps for
making meurtotor parashiyotof tefillin 1) at some later date, after
damage to the codex itselfhad made it unfitforregularuse,

Misbinding
As mentionedabove (p. 375) the MS was misbound at some time.
f. 2 is upside down and back to front(i.e. the present"2a" is really
2b, and "2b" should be 2a). Apart fromthis,the correct order of the
folios should be:
ff.1-53; 56-57; 54-55; 64; 63; 58-62; 65-117; 119b; 119a; 118b;
118a; 120-end.
Omissions
The margins of M show a ratherlarge number of additions to
the consonantal text. These are in most cases plainly omissions by
the original scribe, as remarkedabove. There are typicalexamples
of homoeoteleuton,and other forms of scribal carelessness; it is
possible however that some have textualsignificance.In appendix F
are listedthe cases I have noticedwhere additionshave been made to
the originaltextof themain scribe,usuallyin mashaitscript,vocalized
and accented.The additionsare in severalhands; manyare in a hand
verylike thatof the naqdan.

1) I owe thissuggestionto my father,Dr. S. A. BIRNBAUM. The cuts are on


the followingleaves: ff.47, 62, 65, 68, 70, 98, 102, 113, 119, 121, 124, 126, 131.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
400 E. BIRNBAUM

Supplement
A quick comparisonof M with MG, BH and occasionnallyother
texts,yielded the following notes on additional points of interest.
I leave the evaluation to Masoretic scholars. (BH=means "BH
readingidenticalwith M").
I
Biblical Readingor Commentof
MG (or BH or others
Folio Biblical words Readingor Comment
no. reference commented of M. where indicated) if
on. fromM. Also
different
additionalremarks.

3a Gn. vii 14 nits MP 5nt a


MP: MG 1n1 .,t
MP: BH-
4b Gn. ix 29 (BH, MG =) v1i (SomeMSS read1'1 Cf.
NORZI, MinhatShai,ad
loc.)
5b Gn. x 29 MP (no MM) 1Di1 1 MP (MG, BH)
5b-6a Gn. xi 4-5 M: the scribeomitted
end of xi 4 and first
halfof xi 5, realizedhis
mistake and rewrote
thecompleteverseim-
mediately.He cancel-
led the superfluous
wordsby placinga dot
above each letter,and
two dots above broad
letterslike n.
6b Gn. xii 10 MP i '1~, MG: MP 32"N1'
(col. 2) Withgacyah. No MM. No ga'yah.No MM.
7a Cursive scrawl in left
margin hardlylegible
butseemsto be ?

7b Gn. xiv 7 (BH =) ]Pn3 MG 2Sn3t


9b Gn. xvii6 ,IR,t'I?2Z MM pm11Z1o i MG -n

,N-
TJnnflm TN
nsunamue tanK
,r~',D~'
bXenvr
13; 1D ']~ 'iR
"3.1
n"!
91)n1bXenvr
nt'mnn
lla Gn. xviii28 MG

llb Gn. xix2 nBn,H


MG
BH anwbn
llb Gn. xix8 1VI MG 1VS1.
BH 1W

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 401
I
Biblical Reading or Comment of
Biblical words MG (or BH or others
Folio Readingor Comment
no. reference commented of M. where indicated) if
on. fromM. Also
diflerent
additionalremarks.

12a Gn. xix 11 (BH = MinhatShai=) MG fl!

12a Gn. xix 16 ij~rryw rinnnr11


litti
~aranl
4 z1=1V1r
BH 1'174114
12a Gn. xix 17 BH, MG N--1
12a Gn. xix 17 (BH, MG =) t3x1s-1:5 According to Minbat
Shai,BA readingis 5
12a Gn. xix21 BH, MG: YI X
MinhatShai: "With
.aqefqajan,accordingto
manuscripts."]
13a Gn. xx 2 (BH =) MG, MinbatShai I1
13a Gn. xx 11 BH 1
MG I
13b Gn. xx 14 MP on 7 MP: MG 'Ofl 'I
MP: BH -
13b Gn. xxi 10 BH 1i?Xi
MG VZf
13b Gn. xxi 11 MP (no MM) dn i MP: MG'Ofl 1' (with
4 examplesin MM)
14a Gn. xxi 18 MM ion trujl BH-, MG-
tron-101ma
inj%l (But similar MM in
GINSBURG, Massorab,
II,
"I
[illegible-]1'31Z1j7 p. 532.)
14a Gn. xxi20 (twice) BH, MG (twice)NV11
14b Gn. xxi26 (MG -) r-w56 BH ZV11rrx-
14b Gn. xxi27 MG '
BH 1 nY'x'7
14b Gn. xxii2 MG, BH : 1;1
14b Gn. xxii2 (BH =) MGIls1701
14b Gn. xxii3 (BH -) MG V=1141
14b Gn. xxii4 (MG -) BH ItZ)
14b Gn. xxii4 BH ~1VZf1ln
VIa~~ MG '11VZfl1
15a Gn. xxii13 BH tIM3
MG MtMr
15a Gn. xxii18 (MinhatSbai=) BH Inn
1' 1
MG ID1lfl1
15b Gn. xxiii2 (BH = normalsize n) MG, MinhatShai:
smallD
16b Gn. xxiv8 (sic!) aidfl MG, BH 11j71l
(The rafelooks faded
or scratched)
17a Gn. xxiv30 rlM
t3467?32 1) rMl. MG, BH flN1

') So too in Yemenitetradition.Cf. R. Haiyim Qarah's note in R. Yahy-


ha-Diqduq,ad loc. (ed. JosephQdfih,in the PentateuchKeterTorah,
Sdlih,HFeleq
Jerusalem,5719 AM): "VI11'il.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
402 4. BIRNBAUM

Biblical Readingor Commentof


Folio Biblical words Readingor Comment MG (or BH or others
no. reference commented of M. where indicated) if
on. fromM. Also
diflerent
additionalremarks.

18b Gn. xxiv67 MG '1XIX"I


BH MI'MVI
18b Gn. xxv 6 (BH =) MG 13-101~`Wl
37a Gn. xliii26 (BH, MG -) W',1"
MP
65a Ex. xxxvii8 (MG =) BH, GINSBURG, Penta-
teucb,(quoting many
MSS) 1?M-21-I (with
maqqef and without
accent).
65a Ex. xxxvii8 (BH, MG =) GINSBURG, Pentateuch
66a Ex. xxxviii12 BHI,MG =l?Z=
(A laterhand has in- (But manyMSS read as
serted -t before the M: *I?3 Cf. apparatus
word and above the to BH and GINSBURG,
line.) Pentateucb).
66a Ex. xxxviii21 MM IIfl% ion n MM to MG IU t (with
rn
~P71;r r'3z
V101 t'n-iz
'7111
; details of sevencases).

#**V*)i...rq*,p

87b Lv. xxiii17 (BH, MG =) IN'1


103b Nu. x 18 flXN1'~1~C(1 is dele- MG, BH I=Wl (om.
ted by overliningin a "=). (But some MSS
laterhand,to bringthe readas M. Cf.apparatus
text into conformity to BH and GINSBURG,
withTR)' Pentateucb).
104a Nu. x 35 The invertednuns"do
not look like nunsbut
ratherlikereshinverted
witha blob on top. Cf.
NORZI, ad loc.)
117b Nu. xxv 12 (Wawnormal)13*v MG: with"cutwaw" See
No MP MinbatShahlad loc.
and, further,Talmudic
vol. 9 cols.
Engcclopedia,
133b Dt. vii 26 397-398).
MlText: 1331"PVI
-yiItP
MM: 'I= 7 iz fill

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 403

APPENDIX A

Leaveslackingin M

How many Chapterand versenumbersof


leaves missingportions.(Anyverseof
Positionin M. whichanyportionis lackingis
missing.
notedas missing).

Precedingf. 1 2 Gn. i 1-ii16


Betweenf. 2 and 3 2 Gn. v 3-vii14
Betweenf. 18 and 19 1 Gn. xxv 17-xxvi14
Betweenf. 33 and 34 1 Gn. xxxix18-xli1
Betweenf. 35 and 36 1 Gn. xlii4-xlii30
Betweenf. 38 and 39 4 Gn. xlv 9-xlix2
Betweenf. 39 and 40 1 Gn. 1 4-Ex. i 7
Betweenf. 46 and 47 2 Ex. viii 15-x5
Betweenf. 48 and 49 1 Ex. xii 13-xii36
Betweenf. 50 and 51 1 Ex. xiv 18-xv10
After53 1 Ex. xviii9-xix8
After57 6 Ex. xxiii16-xxvii6
After63 3 Ex. xxviii40-xxx36
Betweenf. 61 and 62 2 Ex. xxxiv29-xxxvii7
Afterf. 68 and 69 2 Ex. xl 29-Lv. iii 11
Betweenf. 73 and 74 1 Lv. viii 14-ix3
Betweenf. 75 and 76 1 Lv. xi 24-xii4
Betweenf. 85 and 86 1 Lv. xx 20-xxi24
Betweenf. 90 and 91 1 Lv. xxv 55-xxvi30
Betweenf. 92 and 93 1 Nu. i 16-i43
Betweenf. 93 and 94 1 Nu. ii 25-iii25
Betweenf. 108 and 109 1 Nu. xv 23-xvi10
Betweenf. 115 and 116 1 Nu. xxii25-xxiii13
Betweenf. 125 and 126 2 Nu. xxxii21-xxxiv2
Betweenf. 126 and 127 1 Nu. xxxv7-xxxvi2
Betweenf. 128 and 129 1 Dt. i 37-ii19
Betweenf. 129 and 130 1 Dt. iii 6-iii29
Betweenf. 131 and 132 2 Dt. iv 43-vi15
Betweenf. 133 and 134 1 Dt. viii 12-ix7
Betweenf. 134 and 135 1 Dt. x 9-xi12
Betweenf. 136 and 137 1 Dt. xii 23-xiii14
Betweenf. 137 and 138 1 Dt. xiv 25-xv19
Betweenf. 142 and 143 3 Dt. xxi 15-xxiv13
Betweenf. 148 and 149 2 Dt. xxix13-xxxi11
Betweenf. 149 and 150 1 Dt. xxxi28-xxxii22
Afterf. 151 1 Dt. xxxiii25-xxxiv12 (end)

Total 56

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
404 E. BIRNBAUM

APPENDIX B
ofM withBenAsherandBenNaftaliReadings
Comparison recorded
in
Mishaelben'Uziel'sKITAB al-KHILAF
Notes.1. Passagesare citedin theordergivenby MbU (Kitdbal-
Khilaf,ed. L. LIPSCHUTZ), who cites the BA/BN disagreements
pericopeby pericope(of the annualcycle),and followedin each
pericopeby thecongruences. This meansthatthebiblicalorderof
citationsis sometimes slightly
interrupted.
2. Wherea hilluf is recordedby MbU but not givenin thislist,
thereis a lacunain M at therelevant passage.
3. Wherea BA/BNcongruence is recordedbyMbU butnotgiven
in thislist,(a) M agreeswithBA/BN,or (b) thereis a lacunain M.
(See tabulation oflacunaein M, in AppendixA, p. 403).
4. Thelistbelowalso notescases(marked*)otherthanthosehillu-
fimor congruences citedbyMbU, whereM's readingwas noticedto
differfromBA/BN.In thisrespectit is farfromexhaustive.

BiblicalBook M's readingin different


Chapterand M readsas fromBA and BN Comments
Verse

Gn. iii 3 BA
Gn. ix 10 BN
Gn. ix 2 * 'Dl1 (no ga'yah,no ma- BA/BN havega'yahand maqqef
qqef)
Gn. xiii15 BA
Gn. xii 20 * No gayah BA/BN havega'yah
Gn. xiii12 * No ga'yah BA/BN havega'yah
Gn. xv 6 * No ga'yah BA/BN havega'yah
Gn. xviii26 BA
Gn. xxi30 BA
Gn. xxii8 BA
Gn. xxii12 BA
Gn. xviii5 BA/BN "Some scholarsread ' withda-
butthisis incorrect."
gesh, (MbU).
Gn. xxiii4 BA
Gn. xxvi22 BA
Gn. xxvi27 BA
Gn. xxvii13 BA
Gn. xxvii27 BN
Gn. xxvii40 BN
Gn. xxx16 BN
Gn. xxx42 BA
Gn. xxxi39 BA
Gn. xxxii9 BN

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 405

BiblicalBook M's readingin different


Chapterand M readsas fromBA and BN Comments
Verse

Gn. xxxii9 BN
Gn. xxxii27 BA
Gn. xxxiv3 BN
Gn. xxxv12 BA
Gn. xxxvi16 BN
Gn. xxxii18 No ga'yah BA/BN havega'yah
Gn. xxxiv24 * 'Y1t withgacyah BA/BN have no gacyah
Gn. xxxvii10 BA
Gn. xxxix6 BN
Gn. xxxviii9 ? It is notclearfromMbU whether
thisdagesh is the point of agree-
mentbetweenBA and BN. The
accents of M are the same as
BA/BN
Gn. xli 45 BA
Gn. xli 50 BN Texthas '1" buta marginalnote
in anotherhand reads T1? (=
BA). GINSBURG,Introduction,
p.
696 statesthatBM Ms Or. 2363
f.41b notes that BN and R.
MoshehMohah read withqames,
and BA and R. PinhasRosh ha-
Yeshivahwithpatah.
Gn. xliii26 BA
Gn. xli 56 No ga'yah BA/BN withga'yah
Gn. xliv19 B
BA
Gn. xlix8 BA
Ex. iii 5
Ex. ii 17 *
BA
13t'I1 withqacyah BA/BN have no ga'yah
Ex. vi 27 BN
Ex. vii 13 BN
Ex. vii 22 BA M r (sic)
Ex. vii 28 BN
Ex. xii 3 BN
Ex. x 11 [BN?] xKli5 Against MbU's BA/BN; but
accordingto marginalnote to
MS 2 FirkovichArab.-Heb.148
(quoted by LIPSCHUTZ on p. 11
footnote 2), this is the BN
reading.
Ex. xv 13 BN
Ex. xvi 12 BN
Ex. xv 26 * BA/BNhavenoga'yahon ntnOI
Ex. xxii29 * -1 with maqqef and BA/BNhaveno maqqefand have
without gacyah ga'yah.
Ex. xxiii11 * -7 No ga'yah BA/BN withga'yah.
Ex. xxi37 * I:V' No gacyah BA/BN withga'yah.
Ex. xxviii5 BA
VetusTestamentum
XVII 25

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
406 E. BIRNBAUM

I I
BiblicalBook M's readingif different
Chapterand M readsas Comments
fromBA and BN
Verse
1
Ex. xxviii33 BA
Ex. xxxiv10 BA No ga'yah BA/BN withga'yah.
Ex. xxxiv14 * mllnnll No ga'yah BA/BN withga'yah.
Ex. xxxviii3 -ln No ga'yah BA/BN havega'yah
Ex. xxxviii1 [BN] Accordingto marginalnote of
MS 2 FirkovichArab.-Heb.148
(quoted by LIPSCHUTZ, p. 13,
footnote5), thisis a BN reading.
Ex. xxxix1 BA
Ex. xxxix28 BA
Ex. xxxix29 * ntl No ga'yah BA and BN havega'yah
Ex. xl 15 * nnlnlwithga'yah
Nu. xix 10 BN
Lv v 24 BA? '1ttR No ga!yah, but also BA -t1V; BN W'I
nomaqqef
Lv. xiii31 BA
Lv. xiii56 BA
Lv. xv 4 BA
Lv. xviii3 BN
Lv. xx 17 * -nt-l'I with ga'yah and
two maqqefs.
Lv. xxii3 BN
Lv. xxiii2 BA
Lv. xxiii4 BA
Lv. xxiii32 BN
Lv. xxiii44 BN
Lv. xxiv16 BN
Lv. xxv37 BN
Lv. xxv9 *
'pl= withga'yah BA/BN have no ga'yah
Lv. xxvii25 * .tl with ga'yah BA/BN have no ga'yab
Nu. iv 15 BA
Nu. iii 26 * No ga'yah
Nu iv 14 * No ga'yah
Nu. iv 16 * Withga'yah All againstBA/BNcongruences.
Nu. v 19 * No ga'yah
Nu. v 24 B
No gacyah
Nu. viii20
BA
BA
Nu. ix 19
Nu. x 30 BN
Nu. xi 16 BN MRv
Nu. xii 6 BN
Nu. xi 16 * No ga'yah BA/BN havega'yab
Nu. xi 23 * No ga'yah BA/BN havega'yah
Nu. xiii19 BN M
Nu. xiv 23 BA
Nu. xiv 27 BA
Nu. xiv 43 BN
Nu. xv 14 BA

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 407

BiblicalBook M's readingin different


Chapterand M readsas Comments
fromBA and BN
Verse

Nu. xvi 28 BN
Nu. xviii9 BN
Nu. xviii19 *
BA/BN have no ga'yah. LIP-
SCHUTZ (p. 19 note 2) quotes M's
readingfromthe ManuelduLec-
teur[= Tijan], ed. J. DEREN-
BOURG in JournalAsiatique, Ser.
VI, tome16, Paris,1870.
Nu. xvii12 * No ga'yah BA/BN havega'yah
Nu. xxi23 BA
Nu. xix22 M has no ga'yahand a second
maqqef.BA reads 1= W"-l~
whileBN reads 71"'S1
Nu. xxi 1 BN M;
Nu. xxi4 M's readingis also attestedin the
Mahtora Rabba (Cf. LIPSCHiTZ
in Kitabal-Khilafp. 19 footnote
a, citing NORZI, Minhat Shai, ad
loc.,q.v.) and in currentedd. of
TextusReceptus(againstBA and
BA BN).
Nu. xxii5 BA
Nu. xxiv22 BA BA
later
Nu. xxvii11 BN
BN No ga'yah
Nu. xxxi12 BA
BA
Nu. xxxi22 BN
* n11tIn-nH
No ga'yah BA and BN havega'yah
Nu. xxxiv13 *
*
i,n~nn M = TextusReceptus,(curr.edd.)
BA has no ga'yah;BN lI,nn.
Dt. vii 1 Q7ll No maqqef BA '-:"1; BN '-'1
Dt. vii 1 BN
BN M;
Dt. vii 26 BA
BA
Dt. viii2 BA
Dt. xi 19 BN
BN
Dt. x 4 * "'75 No gayah BA/BN havega'yah
Dt. x 7 BA
BN; and M reads as laterreadingof BA
laterBA (i.e. withoutmaqqef.See MbU,
p. 21, line 5 frombottom).
Dt. xvi 2 BN
Dt. xvii11 BA
Dt. xvii20 BA
Dt. xviii15 *
gaipya
TNo BA and BN haveno ga'yahs.Ma-
nueldu Lecteurreads withga'yah
on lnnp
Dt. xx 1 * No ga'yah BA/BN havega'yah
Dt. xx 15 * No gawyah BA/BN havega'yah
Dt. xix 5 * Q?S7n withga!yah BA/BN have no ga'yah
Dt. xxvi19 BA
Dt. xxviii52 * This word was omittedby the

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
408 E. BIRNBAUM

___
BiblicalBook M's readingif different
Chapterand M readsas fromBA and BN Comments
Verse

originalscribeand added in an-


other hand. BA/BN have no
ga'yah.M agreeswithreadingof
ManuelduLecteur
Dt. xxxi21 BN In the margin,the same reading
is given in contemporary square
script:ilt 'n1Dt1 (i.e. "There
is a disagreementon the vowel
qameshere.").
Dt. xxxi20 * No ga'yah In agreementwithMS 2 Firko-
vich 150 and Manueldu Lecteur,
and againstBA/BN.

APPENDIX C
The "Dotted Passages" accordingto M.
Note: Those marked* divergefromthe dottingin TR (MG)

Biblical Dotting Note in


Reference in M Masora Parva Comments

Gn. xvi 5
Gn. xviii9 * jt j The second dot is on the t not
the ", and the thirddot is between
the 1 and the ". A hole obliterates
theremainder of the MP, which
was presumably like thatto Gn.
xix 33 or Gn. xxxvii12.
Gn. xix 33 * 1 1 i1 in 71p i
Gn. xxxiii4 inlti 11n [np
I171h MP probably thus, but very
unclear.
Gn. xxxvii12 * fp ip
.'l117. No dot on alef.
Nu. iii 39 * p'lik1 (sic) inn:1 1it No dot on vav.The lastword of
the MP is a mistake.It should
readas Gn. xix 33 or Gn. xxxvii
12
'l
in !p
Nu. ix 10 0npr
Nu. xxi30 IR iM 1p3 MP is in laterhand,but thedot
of textis original.
Nu. xxix15 1t1vYPVttIl None Therearedotsabove bothwords
though that on the firstresh
seems clearer.The second dot
maybe an error
Dt. xxix28 Lacuna Lacuna

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 409

APPENDIX D
Pleneand Defective
Spellings.
Noted below are words which are spelt plene in M and defectiva
in TR, or vice versa; and also words which have been alteredin M
eitherin the body of the textor by a marginalannotation,to make a
or vice versa. The list is not exhaustive.
pleneinto defectiva
! I/

Biblical Form Formin TR


Folio no. in M Comments
Reference (MG) and BH

4a line 5 Gn. viii22 A laterhand has converted MG, BH text: Wi


thisto '1S7.This contradicts MP as MP of M.
theMP 6n .T
5a Gn. x9 In other hand in margin MG, BH '111
(second time) U121Ri No MP with MP
12 Gn. xix22 No MP. A wawhas been in- MG, BH VS%
serted(possiblybytheorigi- MP to BH 71Zi
nal hand?) to bring it into
conformity withTR.
13a Gn. xx 3
MP bri . MM refersto v. 3 BH, MG 1i'nl
and v. 6. Accordingto Min- (Yet MP to MG
hatShai, it is plenein Codex reads as that of
Hilleli and all good MSS. M).
NoRZI explains that the
position of the MP 6rn:
(which also occurs in MG)
is a printer'serrorhere,and
mustreferto Gn. xx 6 and
Gn. xxxi24. M provesthat
this MP is not a printer's
error.From NoRzI's note is
is quiteclearthatotherMSS
also read as M.
The orderof MM on this
page is not consistent.Top
marginrefersto '"1 'I" (v.
6); bottom margin, first
entry,refersto lt3n' (v. 6)
but the second entryrefers
to :5nl (v. 3). At firstsight
this would supportan ex-
planation of scribal error
similarto NoRZI's interpre-
tation,but this is made un-
tenableby the factthatthe
fullcitationsin the MM for
v. 3 referto v. 3 and v. 6
only,not to xxxi24.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
410 E. BIRNBAUM

Biblical Form Formin TR


Folio no. Reference Comments
in M (MG) and BH

18b Gn. xxv13 NORZInotesthatHilleliand BH, MG


good MSS readplene
19a Gn. xxvi18 1ram= Alteredto read l1OttV.Pre- BH, MG naltt
vious word l1lVtinsertedin
a laterhandsuggesthomoe-
oteleuton
24a Gn. xxx33 In a laterhand in the mar- BH, MG KR1n
gin: j3
26a Gn. xxxi51 The secondyodhas changed BH, MG J]=
beeninto a circle,to cancel.
(See p. 385,par. ii e). MP 'D
jM RK: i inl4 . No MM.
51a Ex. xv 16 Above theline,betweenthe BH, MG '1]:1T
reshand cayin a wawhas been MP of MG '7 1
inserted in another hand,
which also added an MP
5} i. The gacyahunder the
reshalso seemsto be an addi-
tion. Accordingto NoRzI,
Hilleli readspleneand Yeru-
shalmireadsdefectiva.
66a Ex. xxxviii14 Amendedbya laterhandby BH, MG t;1l"T
insertionof a wawabove the
line, betweenthe memand
dalet,whichalso insertedMP

82b Lv. xvii5 Amended in anotherhand BH, MG nl1K


by insertionof waw above
the line, between alef and
taw
99a Nu. vii 17 Amendedbya laterhand,by BH, MG 3Tln27
insertionof wawabove line (bothwithMP =
betweentawand daletand of M's)
a MP ij 't j (Cf. also
NoRZI,ad loc.)
103b Nu. x 24 Amended by insertionof BH, MG "m'lm"t
wawabove theline,between
cayinand daletand repetition
inanotherhandofthewhole
word "l'71T in margin.
106a-106b Nu. xiii24 In each case a laterhandin- BH, MG
sertedwaw above line be-
tweenkafand lamed.NORZI
cites Hilleli, the Masorah
and the Italian codices as
pleneand the Yerushalmias
defectiva.
117b Nu. xxv 17 Amendedinanotherhandby BH, MG IniK
theinsertion ofwawbetween

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 411

III
Biblical Form Form in TR
Folio no. Reference in M Comments
(MG) and BH

alefand taw,above the line,


and MP ;iO 3 %j
117b Nu. xxvi5 in According to NORZI, this BH, MG l[ln
shouldbe plene.
120b Nu. xxviii2 Amended in anotherhand BH, MG 1753r1m
by insertionof wawbetween
menand Cayin,above theline.
121a Nu. xxviii28 (first) The firstwawscratchedout BH, MG V'V107
to forma holem(dot).
123b Nu. xxxi6 ni8rn A neatwawhas beeninserted BH,MG nllIrn
above line betweenthe resh
and taw
126a Nu. xxxiv9 (first) Waw scratched out and BH, MG ,1rn
(51en vowel inserted to agree
with TR. A later hand has
added MP bn j
126a Nu. xxxiv11 (second) The same. BH, MG ?n1;1
51.in
150a Dt. xxxii34 Ki*n MP (if any) was in part of BH, MG K~N
(see plateD) marginwhich is worn or
broken off. Aleppo codex
and most of MSS cited in
GINSBURG'S Pentateuch read
but some read as
defectiva,
M; Norzi notes a division
of opinion on this in the
MSS and theMasorah.

APPENDIX E: PELUGTOT
The following passages in M are marked irl or similarly.(See
above p. 399)

Subjectof the M's Margi-


Biblical
Folio no. pelugtaas given nal annota- Comments
Reference
M. tion

f. 83a Lv. xvii14 (= BH, MG) "? iyfrt-I


(bottom)
f. 84b Lv. xix22 (= BH, MG) *'5
(withoutaccent)
f. 92a Lv. xxvii28 (= BH) -'lt sitb -J No marginalindicationof
nature of disagreement.
MG reads Jt (with mu-
nah and withoutmaqqef.)

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
412 E. BIRNBAUM

- I
Biblical Subjectof the M's Margi-
Folio no. Reference pelugtaas givenin nal annota- Comments
M. tion

f.97b Nu. v 30 (=BH, MG) I. an0-rn i Marginalsegoland maqqef.


115a Nu. xxi34 (= BH, MG) '? No rafeon marginalkaf,
but no indicationof na-
tureofdisagreement.
Mar-
ginal note in different
hand than otherpelugtot.
126a Nu. xxxiv12 (= BH, MG) q; flt -ID Margin has no accent,
n,?n but a maqqefinstead.Cf.
long discussion of this
reading in NoRZI, ad loc.
131a Dt. iv 23 The disagreement is pre-
sumablyover eitherthe
or thepe/fe
gacyah or both.
BH, MG 13 (withpe).
131b Dt. iv 38 (= BH, MG) *T
-nn1
132b Dt. vii 9 (= BH, MG) `-" 1s* -t' No indicationof nature
of disagreement.
133a Dt. vii 19 ift -1n Both text and margin
have no dageshin kaf.
Margin has maqqef.MG,
BH -|'
133b Dt. vii 26 [etc.]r'1iV See Supplement, lastitem,
[etc.]1t (p. 402), ad Dt. vii 26 f.
133b.Differences in accen-
tuationand ga'yahs.
135a Dt. xi 16 gif 'In
146b Dt. xxviii26 ifts Jn^al The disagreement is pre-
sumably whether there
should be a gacyahon the
lamed(as in BH, MG,
GINSBURG'SPentateuch).
1
149b Dt. xxxi26 =
i? This form with qames,
(see plate C) (whichis also TR and BH)
is a BN readingand also
thatof the Mah.ora Rab-
ba,in MS BritishMuseum
Add. 15251. See GINS-
BURG, Introduction
p. 436
The BA readinghas pa-
tab(MbU, ad loc.)

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 413

APPENDIX F
Omissions
Inserted
See p. 399. This list does not mentionthose discussed elsewherein
this article.The insertionsare vocalized and accented.

t
I
Folio Biblical Word(s) originally
no. Reference omittedand later Comments
inserted

7b Gn. xiv 7
8a Gn. xiv 16
8a Gn. xiv 19 rnl
12b Gn. xix33
13b Gn. xxi2 Apparentlybythe
scribeoftheconsonantal
text.
14b Gn. xxi31
15a Gn. xxii13
15a Gn. xxii17
17b Gn. xxiv48
18a Gn. xxiv65
18b Gn. xxv 14 Completeverse
19a Gn. xxvi18 nlwe Homoeoteleuton? See
Appendix D on this
verse.
20b Gn. xxvii36 nns
22b Gn. xxix27
23a Gn. xxx4
33a Gn. xxxviii24
36b Gn. xliii15 Homoeoteleutonat
and 16
36b Gn. xliii17
45a Ex. vi 25
55a Ex. xxi37
55a Ex. xxii9
66a Ex. xxxviii11 Added by naqdan in
Completeverse marginbut withoutini-
tial \. Presumably the
omission of verse is a
case ofhomoeoteleuton,
since Ex. xxxviii12 be-
ginswithnrix
69a Lv. iii 10 Presumablyhomoeote-
leuton,sincenextword
is also 'VS
70a Lv. iv 25
70b Lv. v 6
74b Lv. ix 19
91a Lv. xxvi64
91b Lv. xxvii3 non

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
414 E. BIRNBAUM

Folio Biblical Word(s) originally


no. Reference omittedand later Comments
inserted

92a Lv. xxvii17-18 Homoeoteleuton,cau-


sed by IHflI
96b Nu. v 10 ttr
98b Nu. vii 6 WnN BH, MG aDlM
99b Nu. vii 28
117a Nu. xxv5 Last tetragrammaton
117b Nu. xxvi3
117b Nu. xxvi5
121a Nu. xxviii13
123b Nu. xxxi6
128a Dt. i 19
131b Dt. iv 34
133a Dt. vii 20
134b Dt. ix 26 "X' n%n13
134b Dt. ix 26 - S1
7i,>n
135b Dt. xi 27
140b Dt. xviii19
142b Dt. xxi3 TS7H
(second) i>n',1
142b Dt. xxi4 (second)
(second) rl,lyrl
,n^S^n
146b Dt. xxviii29 n^s3n
147b Dt. xxviii52 12nH ha

ABBREVIATIONS AND REFERENCES


BA = Ben Asher.
BH = Biblia Hebraica ... edidit R. KITTEL. Textum masoreticum curavit P.
KAHLE... [12thed. correctedfrom3rd and 7thedd.]. Stuttgart,
Wiirttem-
[1961].(See p. note4).
bergischeBibelanstalt,
BIRNBAUM,HS = Solomon A. BIRNBAUM,The Hebrew Scripts, II: The Plates.
London, Palaeographia,1954-57.
BN = Ben Naftali.
= C. D. GINSBURG, Introduction
GINSBURG, Introduction to the Massoretico-Critical
of theHebrewBible.London, 1897. Re-issuedwitha Prolegomenon,
edition
"The Masoretic Text: a Critical Evaluation", by H. M. ORLINSKY. New
York, Ktav, 1966.
GINSBURG, Pentateuch= ThePentateuch. revised
Diligently to theMassorab
according
and theearlyeditions, fromMSS... by C. D. GINSBURG
withvariousreadings
[New edition].London, Britishand Foreign Bible Society,1926 (= The
Old Testament. Diligently revised... by C. D. GINSBURG.[New edition],
vol. 1).
GINSBURG,Massorah= C. D. GINSBURG, The Massorah. Compiled from manu-
scripts, alphabetically and lexically arranged by C. D. GINSBURG. 4 vols.
London, 1880-1905.
KAHLE, Geniza= P. KAHLE, TheCairo GeniZa.2nd ed. Oxford, 1959.
KAHLE, MdW = P. KAHLE, Masoretendes Westens.
I. Stuttgart,1927.
Leningrad= LeningradPublicLibrary.

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions
THE MICHIGAN CODEX 415

M = MS no. 88, Universityof MichiganLibrary.


MbU = Mishaelben cUziel,authorof Kitabal-Khilaf.
MG = Miqraot Gedolot[edition of the Hebrew Bible with Masoreticnotes].
New York, Pardes, 1951. [Reproduced from the Warsaw (Lebensohn)
editions,1860-68,whichsupposedlyfollowJacobben Haiyim'stext,(Venice,
1524/25).The latternot being available to me, I used MG as a roughap-
proximation to it].
Minhat Shai, see NORZI.
MM = Masora Magna.
MP = Masora Parva.
MS(S) = Manuscript(s).
NORZI, Minhat Shai = Yedidiah Shelomoh NORZI, Minhat Shai, as reprinted
in MG.
RTBT = M. H. GOSHEN-GOTTSTEIN,'The Rise of the Tiberian Bible Text' in:
BiblicalandotherStudies,editedby A. ALTMANN. Cambridge,Mass., Harvard
University Press,1963,pp. 79-122(= PhilipW. Lown InstituteofAdvanced
JudaicStudies,BrandeisUniversity. Studiesand Texts,v. 1).
SPERBER = Alexander SPERBER, A Historical GrammarofBiblical Hebrew. Leiden,
1966. [This newlypublishedwork reachedme afterI had completedmy
study.It could not therefore be used (exceptfortheadditionof a footnote).
It containsa considerableamount of informationof relevance to this
study,particularlyin its sectionIV (pp. 414-562).
TalmudicEncyclopedia,Jerusalem. vol. 1 - . 1947 - . (In progress).
Textus= Textus.Annual of the Hebrew UniversityBible Project.Jerusalem.
TR = TextusReceptusof the Pentateuch,particularly
as representedin MG.

Note:
wasmadeandthephotographs
Thisstudy are published of the University
by courtesy
ofMichigan of Toronto.
Library,andwiththeaid ofa grantfromtheUniversity

This content downloaded from 138.251.14.35 on Mon, 25 May 2015 07:18:44 UTC
All use subject to JSTOR Terms and Conditions

You might also like