Skip to main content

Full text of "Notes on Post-Talmudic-Aramaic Lexicography"

See other formats


STOP 



Early Journal Content on JSTOR, Free to Anyone in the World 

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in 
the world by JSTOR. 

Known as the Early Journal Content, this set of works include research articles, news, letters, and other 
writings published in more than 200 of the oldest leading academic journals. The works date from the 
mid-seventeenth to the early twentieth centuries. 

We encourage people to read and share the Early Journal Content openly and to tell others that this 
resource exists. People may post this content online or redistribute in any way for non-commercial 
purposes. 

Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- 
journal-content . 



JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people 
discover, use, and build upon a wide range of content through a powerful research and teaching 
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit 
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please 
contact support@jstor.org. 



NOTES ON POST-TALMUDIC-ARAMAIC 
LEXICOGRAPHY 

By J. N. Epstein, Charlottenburg. 

II 

Sheeltot. 

The language of the Sheeltot (= She.) is on the whole 
the language of the Babylonian Talmud. But, as in the 
case of the language of the Geonim generally, it shows 
a particular resemblance to the language of Nedarim, 
Nazir, Temurah, Keritot, and Me'ilah, these tractates being 
distinguished from the others dialectally 1 (?V instead of 
N, sorb, jvk t6, fU'N xb, bjidi, tom^>, and so on), and 
characterized by 'older' 2 and fuller forms in the pro- 
nominal and verbal suffixes (pr - , p ,_ > P"> P^, pn^H ; further 
r^n instead of Mn, xnn, pn, pTn, Djn' , », &c). 

There is no doubt, however, that the She. were composed 
in Babylonia, and thus furnish us linguistic matter of the 
genuine Babylonian-Aramaic type. The proofs are as 
follows : 

i. The language which is Babylonian throughout, and is 
marked as such in orthography and vocabulary (NSD1N, ndn, 
m, trm, kpdh, mn *ira, smin Tin, pnn, wrc, niTD, xr6 me>» 

1 Concerning these tractates see the article by W. Rabbinowitz in 
D^BTV, X, 233 ff., esp. 244-5. 

2 But to conclude from this that those tractates belong to an 'older 
redactional stratum ', as Margolis does in his Lehrbuch, § 1 d, is an error 
(in accordance with this conclusion the She. and the Gaonic literature 
generally must be older than the Talmud !) ; comp. on the contrary note 1. 

299 



300 THE JEWISH QUARTERLY REVIEW 

'03; Persian: NprriN, N01N, K3TU, Npnn, on on, snpns and 
Minna). 

2. Its Babylonian terminology (yoc? Nn, 'NO, V3, li»»K, 
Noi>H IN, »03 'N, nDI^ 'SOD DN, "]!> Wfl «i>, ~p 'JOT '3, 

W313 n$>i, jrjo, m»N 'n, moN *n aba, "^o -on, n?n, nd^>n, 
N"im:i>, pan iiDK, &c.) and the almost exclusive use of the 
Bab. Talmud (a few citations from the Palest. Talmud and 
Midrash prove nothing to the contrary, comp. Poznaiiski, 

7<2^.,N.S.,III,405). 

3. Passages, though citations from the Babli, which 
speak of the Babylonians and Babylonia as JJN, p, and 
Nan in contrast to ir6 and mn : No. 5 : N.ll p an >ybs> t6) 
ir6; 35: 'ia annb p-ho o; 46 (=109): pnay owa £03 ii>w 
•>ov nn pnaj? u"d xncm 'ia nbv nn; 58: up in»»mrr^B> ]3S 
pnay ; #££ : Nani> nrai DJvi nj>M«n ; and 68 : 13 un: Nam 3 
"WN (comp. IJullin 49 b-5oa and Pesahim 51 a). [Further- 
more, She. No. 100 paraphrases a difference of opinion in 
b. Ber. 35 a which is not called expressly a controversy, 
in a manner not unusual in the Babylonian Talmud 'J^B 
N3->j>D3 wiidn nn na: N"n m na m^b ■ob'N ikbh h'b 
ND'Tian "iont JN01 'iai ndhibt n'a hen nn una pjw mi 
in b. :) lex nn Nanyos wion nn na vbs rrovD 'nd nin^ 
nm man nwam £n!> notui WNian [Na -idn: (tdis 3ai >an tnm ! 

3 Cited in HG. 3^>n, ed. Venice ( = V.), 13a b = Berlin ( = B.) 549 ; comp. 
ni3 ,_ lD, fol. 129 a = B. 529. Nevertheless also R. Han. cites Pes. 51a 
from Hul.: IID'N 13 3WJ N3iTI. Besides, it was already remarked by 
J. M. Freimann to "IVlini, II, p. 4, n. 90: na'n D'iOl VOCD JN3D 

niD'K ia un: scm ana yv ^333 n<n iijn mn^xBfi. Nan for 

Babylon and D]"in for Palestine occur also in Anan : Hark. 4 'DJ Dfin f]Nl ; 
Nan, N3Nn, 14-16, 21. Likewise in Sherira's Epistle, Neubauer, Anecdota, 
I, 22, 23, 27, 28. 

3a This Baraita from the School of Rabbi, '31 UT, was considered by 
She. as post-tannaitic, Amoraic. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 301 

rrw yf?v onciMt? im D-W>n trip -iox. Since this is no 
citation from the Babli the author must have composed 
it in Babylon.] 

4. Furthermore, the single Sheeltot are to be considered 
from the very outset as of Babylonian provenance, i. e. as 
a collection of decisions and lectures 3b by Babylonian scholars 
(Geonim) — as far as they did not conflict with the personal 
opinion of R. Ahai — in part literally transmitted, as already 
suggested, but without proof, by Reifmann. 4 Thus the 
She., No. 58 : ruiewo imi c"n 'ia ^wos>a xroSi -jnx ma 
!>WDt5>a twiaivn Nncoi? 'w (quoted in the name of Ahai, 
Responsa, ed. Cassel, No. 133), is mentioned in a re- 
sponsum of Sherira as a decision delivered on a Nnae> 
N^jm before the exilarch Solomon b. Hisdai in the 
days of R. Samuel b. Mari, Gaon of Pumbeditha (it 
met, however, with opposition), Responsa, ed. Harkavy, 
No. 183, p. 8a: noib> pjj& ton nom pan jo tow 
toi now • niomn p jto jwtn t<n^n [era an "pro xpn wn 

DJ JltU ^WOE* 31 nD 'OVQ KTvhi W\ WDn na ilO^C 

xnta t?n »op t6am xnaca ntdk (r. ea = ma ,twvnaoiB) 
pi jrnn me> dw 'otn i>woe'a nwD^ ni>ie>m fonj ana 5 pddi 
niwxo i^>wn pro am rvi> now mmnw • spvn P rva psnpM> t^N 
xntano prnn «p mat* 'na kp^dbw urb worn xpn 130. 
Again, Hai Gaon in Responsa, ed. Harkavy, No. 328 : 
maw irrvoN 130 i^ra rniDo fnc nnan irao nets' irnan nnw 
<tur jKoa iw n^i w iTua 'nrn jiu6a nann nnne Umax 

3b n^DT Nni>KB> ! see below, sub i (Nn^tttP). 

4 moin ma, in. 

6 Comp. Nathan ha-Babli : ^B* ntSHBH pJJia tTIHl nta K»tO nniBl 

neu wncai 'ia twvnaoi a na^ty tptcni? men jni is Din mix 
^m Dan ins jpr noipi f|W> ns inv ma noiw wjna ^nno 

atnn p3V3 a"Wl. Now was not R. Ahai this ^J11 Dan jpl ? 



302 THE JEWISH QUARTERLY REVIEW 

mDN ini ""pine "tain nb toyv tub mow ru^tran ^>ds in -ipd 
in rfona "ibo!' iwi 1 n^i ^yaw mnn tasNi -itso "taw -iboq 
mw i»3 nnai njwa^ mm nw Tii>jnj ixtnb kw m»Ni ^ids^> 
ruwab, resembles almost verbatim She., No. 95: mus taNl 
»tur ni» j-pdnt panp jd 'n (so MS. E 6 ) jpt n^i mtaNi w 
"tai n^ mm 7 na^ |wn N-iwm -pn jd 'n -itdd "tan mb mm 
iofl taN y-n D-13 . i"h N^npn ^yo^ tidni vint? .t^> mn nsra 
-ibooi nw m»xi biDSD in -ie>a jd ^n jtv n^i S'jn'N-i nvrm 
njcwd ne>a: 'a^ noi^ nvdti dn 'la n^> in njdwd 'd 'is Tibyai 
'13 n5> IN njdtid pa^ naiWNi' nna '13. Likewise in the 
nDWB> 'n (by R. Samuel Gam'a), fol. 143 (Steinschneider 
in jfiid. Zeitschrift, II. 78): an 'th naiDJD^N maiwi'N 'Si 
xnx 2-to mnW 'si NnN'p p.j)De> 'no ma^mx 'si ^"r ww 
wanting) . . . nnN ntaa ni'jn n»na Knarie^t] Nawi njn NnNae>D 
blfbx .mn ;n . . 'Nn . . wnx tap! (in our copy of the She. 
8 rr f ? pay: -vmbn pyai noriN 'nthd pam nde>d. 

For a Palestinian origin, however, there is no evidence 
at all; for KnW, ma, inv, N"iw, and ns^n— which are 
quoted by Ginzberg— are good Babylonian-Aramaic words, 
since they occur in both the Babylonian Talmud and the 
Gaonic literature, as I shall prove further below. 

The date of the composition of the Sheeltot is therefore 
to be placed before the Gaonate of his pupil Natroi 
(Natronai) Kahana — who was the cause of his emigration 
to Palestine— which, according to Ibn Daud (Neubauer, 
Anecdota, I, 63), lasted thirteen years, until 4521 (= 107a 
Sel.), and hence began at 1059/60 Sel. Sherira Gaon, 
however, places in the year 1059 Sel. the beginning of the 
Gaonate of Rab Samuel b. Rab Mari, which is followed by 

6 E = MS. Epstein, F = Cod. Merzbacher, Frankfurt. 

7 E: '•NJl D'pirriD iN, F: "]Jn» 'N JT"T N^l Trfo\ tWH iTUB 
'13 P'DNT NIH D'anp; ed. V. njJT vb). 

8 More in another place. 



POST-TALMUDIC- ARAMAIC LEXICOGRAPHY — EPSTEIN 303 

that of Natroi Kahana (the duration is not given), then 
Abraham Kahana (again the duration is not indicated), 
and then in the year 1072 Sel. begins the Gaonate of 
Rab Dodai (Neubauer, ibid., pp. 35-6). According to this 
statement Natroi's term of office would have to be 
shortened slightly, and 1061-1070 Sel. would be the date. 
At all events, the composition took place during the 
Gaonate of R. Samuel b. Mari, under the reign of the 
exilarch Solomon b. Hisdai (see Sherira), which is confirmed 
also from another source (see above under 4). 

The place of composition is probably Pumbeditha, 9 
where subsequently Natroi became Gaon. 

Whether his above-mentioned pupil, Natroi of Bagdad, 
took part in the redaction of the Sheeltot, cannot of course 
be maintained with any degree of positiveness. Redac- 
tional glosses, however, which are exceedingly ancient and 
are found in all our MSS. (UM ^N, Wl^NBH K"1E»D "D3 'M. &c.) 
make it evident that the work has come down to us in 
the form of a later redaction considerably modified and 
shortened by his pupils. Further proof is furnished by 
many She. which are abbreviated and translated into 
Hebrew (as, for instance, No. 167 — see Hal. Ged. and 
Geonica — No. 146, see Hal. Ged., &c), and also by a 
passage in No. 138 : '13 wn p»jn 01 Nan -ion, MS. E and 
RNZB 9a '13 "2 .toTtd -idn. 10 

' J in now 'n, foi. 214b : ntsnan xra&v t<m aio ronW »b tirn 
ioio pmi» '-n maj -in jkd -i^nD^x xnn ;x nbym • . . irn^ra 
Dia 'ijtf rutOD bnx ;di \>"i pw »tnirr an n»^> (!) -rattyoi tam 

NI1H3. Steinschneider, in Jiid. Zeitschr., II, 79-80. 

9a MS. of R. N. Z. Berlin, quoted in his commentary n!w pDJ)n 
and in his Additamenta. 

10 If \HS\ ID, the pupil of Jehudai Gaon (see Aptowitzer, R£j., LVII, 



304 THE JEWISH QUARTERLY REVIEW 

Assuming that Natroi Kahana participated in this 
redaction, we should have to expect in our She. also 
something of the Aramaic dialect of Bagdad, to which no 
doubt belongs the nwon iSD of Anan (see above), who, 
as a member of the family of the exilarch, probably had 
his residence in Bagdad. 11 

I begin with the technical terms and other constant 
expressions. It should be noted at the outset that we can 
hardly speak about the She. ' citing ' the Babli ; where the 
latter is quoted it is often done in paraphrase, and even in 
citations the exact wording is not retained (see above sub 3 ; 
furthermore, e.g., sub 13 on NTrt, )fa, pj-K, 14, 15, 17, 18, 
19, ai, 35, 29, 31, 43, 48, 71, &c), but rather accommodated 
to the language of the She. 

1. Nr6w. Every She. begins with the word N^W 
(pi. hebraized ninbw, as the name of the book). The 
word is good Babylonian ; it signifies there not only 
'question' = K"jn, as Meg. 28 b : Nmo Kni>W ^KCl wp Tin 
and Shab. 30 a-b : Nr6w pajfo 'l3 '"ID «W> tw5>W *mw 

246 ff.)? Now comp. also the DWlSn T!D ^5? DlS^il, ed. Schechter (in 
print in the Festschrift su Ehren D. Hoffmann [now Hebr. division, p. 263 ff.]) 
A recto, 1. 1 f. : ']>DB> "13 fa N31 'K (= She. No. 47, "Vntni, I, 240-41) ; 

ibid., 1. 1. ff. : mne» *6 nt nn new rabn snw enn S5n 'a 
veb toi no-ns' nsbn anst? "inx anna ney» (137 ik~i / n=) 

DinC" &6 p '•S^'VBK 'SI (comp. Rabba, Hul. 12 b) K3K W3"l ; further, 

v. 23: 'bn 13 kdn irai 'N '13 'in -ihion pnv 'n (=310 be, 

II, 261, in the name of JlKJn = R. Jehudai; 1ST 'H, 6 on the contrary like 

min' n) ; further, ibid., l. 25: 'i3 be-son mbx Dme6 5>BOD p3 no Sri 'K 

(against 1N"I '7\,ibid., and against R. Dimi Hul. 15b). As to the relation 
of Mar Rabba to ' R. Aha ' we must call attention to Kalla rabbati (comp. 
Aptowitzer, /. c, note), ch. 4 : '13 NflX 3T b"ti '13 N31 n^> fppnD. 

u That Anan lived in Bagdad is proved by Ibn Alhiti (Poznanski, JQR., 
VIII, 432; Tr\>WV, X, 84). Comp. in addition the Irakiyah (i^i^.;) 
in Targ. Onk., Berliner, Massora, 23, and below to 1VK N7, sub 14. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 



'13 javnp -tabmn (so Rashi in accordance with MSS.), but 
also, as is frequently the case with the Geonim, ' responsum ' 
(comp. my Der Gaonaische Kommentar, &c, pp. 83 and 160) 
and even 'lecture' of the Resh Metiba generally, as HG. 
nni33, ed. Venice, 138 d below : m id vmn "PIT'S 1^K 
rfan xniw3 KrWiD en Win (nbl naturally is the correct 
reading ; concerning fi73 see below, the passage quoted 
sub a and sub 85). 

2. Every She. opens then (after the word Nl"6w) as 
follows: '13 5»tnt2» rrai p3»riD — % $>n-ib* nm pa«no — n 
'« i^kt", '13 bvrw rmi> pn5> -vdk — i, '13 i^n — V 2 
'13 "won (No. 19), '13 n"ap vunrmn (167 and 10a). The "i 
is the Bibl. Aram. H, Papyri n (comp. my ' Weitere Glossen, 
&c.' in ZA W., 1913, p. 235), Syr. "T = 10*6, as introducing 
the direct speech. It is preserved in the Talmud as far as 
I know only in the following places : Suk. 46 b; &V* Nab tb 
»td i? tuaWi Kpwb ; Ket. 5$ a : n«pjn !>ijn ■£ tunatp ns idk ; 
Ker. 3b: '13 b"y\ pnv 'n xn^s t<rn ^"k; #«/., 18 a: 'dm 
ton -i-kd '-1 yo xrn n^; Me'ilah 14 b: 'is "jn ND^nl no»; 
Sanh. 46 b 'Aruk : mt« n3n n^ "»« "in ('Aruk : !>"« nm ''2 
rro nnapi jnn nsjn) ; RCH. : ("|1-ik3 ironn nota Jhin T3jn. 
We might perhaps add the phrase SOK 'nosH (llD's) id^k, 
jm "iiotn ncN, comp. Rosh Hashanah 6a: nin jot nDinon niD'N. 
But it is quite frequent in Gaonic literature as well as in 
Anan's writings: HG. pou, ed. Ven., 79 a (B. 325) : |»"l 1jn 
ihd '13 n^tr "in s6 nan "run n!>33 pso wjh an no -op 
'13 iT3TJ'Di> ; Responsa, ed. Hark., No. 555 : '13 pn K3rm 
'ID liaron D'anpn * iTSlin m ; in the ' Formularies of 

12 No. 60 E has : |ND I^KT Nfl^NC, F : btTW 1 1V3 pa'TIDT NJlW 
?ND t>»1; likewise No. 115 E, F, and Oxf. c. 18 (fragment): l^Kt ; 
No. 44 F : '13 1133 ibsn "p* D" 13 ! and J 3° F = "13 ?«» l!>tn THX D13, 
otherwise here without T . 

VOL. XII. Y 



306 THE JEWISH QUARTERLY REVIEW 

Decrees, &c.', by Aptowitzer in JQR., 1913, pp. 23 ff., 

No. IX (p. 29) : •om 'is 'bi> na bsi 'ia k[j]djk3 mm snnnD 
'13 3 , 3C ^j?3 '5?sn mox ; Responsa ntlM men, No. 13 : 
'13 ^031 J31N1 |0 Wyp 8 ! tUB*^ '•NHS HiTD »3TD <K1. In She. 
not rare even in other connexions: No. 10 : 3n ^ "iDsn 
'13 *a wn *3T WH3 ; nn'' nntw xmiKon wiow *ioki (E nvntrfi 
rrrniso) 21 F: W5>3 *D33 -|1B>y WNl moNl; 60 E and F: 
W»y3 nH mcN sn (E Kp nh) ; 72 E and F : upon trb 3ns »dj 'n 
'13 »:i5>b T3 "b men ; 95 : ton mow 'is vfom -icaoi ton moxi 
nrojJN neooi (E F woyo) ; 113 : tow »an w t&i rrb ioni 
■^ srato •oru; 119: yrbn ttbl rrb yvwm (likewise E, but 
F wrongly rrbti !) ; and 130 E nwi pnt6 rr6 notn Knpem 
i? ann sin rrnjn JO. In Anan, Hark., 5 : nH ndn Kpi 
'la B>abn ; 6 : (men =) nntMl pr6 into xp nee>n viipn ndk 
nnmt6 pa^s^; 13: jvx meaoi> 'oni n^ [son] xp '■an 
'13 ^-1^ ncsi; and 34: winwa -pa^ pysnm i? ndn up. 
HG., ed. B., 63 : W3*iao sp t6 totn >ymi6 rrb ■'yanoi ; ibid., 
326: '13 nnt? (by=) "btxs unn mow nn »ai> rwitn naujn 
(ed. V., 79 a ^y lin, hence the correction by Hild. is 
unjustified) ; ed. V., nt»br\ 109 c : swyan npii> iri> -iok w 
'la "p 'in* *30 ratn npii> n^ nox '13 an3>oi> ; #£/., d : 
'13 pj'K K^an »nr yrri |ir6 'o^i (= ed. B., 448, 449) ; in 
ivrbn 03, HG., ed. V., 66 c (B., 315, ltn 'n, ed. Schl., 121) : 
'13 ^ a^D3 mn rata n^a mos pi ; Responsa, ed. Cassel 
No. 41 (comp. my ' Rechtsgutachten, &c.' in Jahrbuch der 
jiid. lit. Gesellschaft, IX, 261) : rosn vpbb i&Kb *?W1 b^y 
'13 'NyiN nna; Responsa, ed. Hark., No. 546: ntn *?"tt 
'13 jcno ; No. 554 : '13 ytotWU b y«jn 'OKI ; R. Zemah 
in Hamakrid, No. 43 : piscn (r. Tvao) lrvao xrr\m b nm 
'13 b 3n3 pyoen pisi ^n p nos w -iroi> 'ia nbv [maom ; 
Geonica, ed. Ginzberg, p. 10 1, last line: Wl »3 *0p N061 
'ia wtn ; #&/., 212, 1. 16 : '13 nti Kruno mvan pisn -ia m 'k 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 307 

'is rvjn Koyta 'nd; ibid., 213, 1. 21 : 'w Nina mat mtn rtfy mow ; 
#£/., 314 : '13 snrvN xrn -icci" bvJ? n^ rvx »o ; ibid. -. t6 w 
'13 ^n ni=n (.reV!) [Knn]»« men («^tn=); and finally 267, 
1. 3 : 'is nmjn (irn =) im -ionii ('and as to his saying : come 
ye, let us remove him, &c.\ w = friK, imp. pi., as, for 
instance, Baba batra 22 a : tne*3 tax in). Comp. Land., 
Anecdota Syriaca, III, 3, 1. 20 : Kmp £ ami irn -now. 12 ' 
bN~i^'' n-3T is an abbreviation for btrw rV3T xoy, as in 
No. 1 : ^snt^ IW1 KDy itsbib (E and F IW1 (Toy), as like- 
wise in the Bab. Talmud rootM m, KJvbl B*n Ul, in Targ. 
Yer. II to Deut. 25. 17-18 p^oy roi, in Pal. Talmud 
frequently 'ta rV31 p^N (similarly Syr.). In Targ. Yer. I 
in direct speech: *?mw "33 "oy (often), in Targ. Yer. II: 
i>*w m »djj (n:y »ey), Exod. 20. 1 ; 23. 2, 19 ; 34. 20 
(i>tne» 7V3 *»y vsy); Lev. 19. 16 (?x-\w< JV3 'oy), 26 (same); 
Deut. 14. 22 (?X-\W "33 n:y) ; 25. 4 (same) ; Pal. Talm. Ber., 

V, 9c: 'is 5>kib» "33 "oy ponnm p^Ni. 12b [Mand. nun kbk 
^'NIDy, Sidra r. I, 45, 5 fr. below, 120, 9, 10.] 

l^Ni. l^N is frequent in the Bab. Talmud, note par- 
ticularly Pes. 5b: i3"rviKn3 Tamo ibw yaina i!vsn jm 

'13 'Kp. 

3. sotii. After the introduction follow details, often 
prefaced by tovil (e. g. Nos. 6, 17, 20, &c.) in the sense ' if, 
si, of the legal codes, Syr. JK (px). In this sense N3TI 
appears here for the first time (then frequently in HG., &c). 
Comp. Will in Anan, ed. Schechter, ia, 21 ; 14, 14. 

4. 'Nil. Then follows very often (usually before ma 
inv) 'Kll as an introduction of a case about which there 
cannot be any doubt : ' this is certain ' (that if . . .), e. g 

12a See Addenda, No. 16. 

llh See also Landauer, Festschrift Harkavy, p. 26. 

Y 2 



308 THE JEWISH QUARTERLY REVIEW 

No. 17 : '13 W Jib 1CH Wl ; 20 E : '13 Wl rvb ION 'tni ; 
38 to, 43: '13 "UTO W, 50, &c. In the Bab. Talm. we 
usually find KtWS instead (also in She., No. 4a— citation 
from Sanh. 74 b — '13 '1 w '13 XW1B NtWB ; No. 154: 
'13 "jr)N NW NtD'CS; and No. 170: '13 DT3 '13 njfE> NCB>S), 
which is always rendered by 'HI ('Nil) in 1ST ntt^n, e.g. 

P-8. 

5. "pv D"i3. Then comes the 'question' which begins 
(often) with ym D"D, or D"i3 alone, as the She. from the 
Petersburg MS. in the ed. Wilna, III, 226 ; Geonica, 386 = 
No. 44 ; ibid., 37a r., 1. 35 and v., 7 ; Hark, No. 376, p. 191, 
a citation fr. No. 134; HG., ed. V., 57 b— ed. B., 375: 
T"1V EH3— ; ibid., 57 d— ed. B., 377 : 'iv DH3— , wanting in 
She.; "Vn?m I, nn», fol. 33a, She. No. 115 F: bit 012 nM'tc 
'13 K3\i, No. 134 F: ' 13 N3,,n 1 ^ 012, &c. The complete 
formula was ^tb (or UK pans) riK "p* D"tt, according to 
the testimony of Nathan ha-Babli (Neubauer, Anecdota, II, 
84, cited above, p. 301, n. 5). So it is still found in the 
ed. (ed. = ed. Venice 1546, which I use exclusively through- 
out] Nos. 1, 3, 65 ; abbreviated to 'vb T"iV 013 in Nos. 133, 
J 35 11. 1 and 3, 143 1. 3, 146, 168 ; 'tb 71* 012 in No. 139. 
Beginning with No. 97 the abbreviation 'lb, '*zb has been 
resolved erroneously into 10b, l^rbf, which is always 
missing in MSS. E and F. The complete formula men- 
tioned above is still preserved in MS. F, Nos. 35 and 44 
(margin) ; Geonica, 371 {'vb '« 'v Q13), 380 (TIN "p* D13) ; 
in my Vienna fragment, JQR-, 19 14, 420 («W> "N P 3 ^ 012, 
likewise in the pi. pny DT3, in Tanhuma, npn = No. 137) ; 
MS. Oxford (fragm.) hebr. d. 63 (28363^ fol. 55 w., 
No. 33 to, fol. 56 r., No. 34 ; (2836 60 ) fol. 99 v., No. 4, 
fol. 100 r., No. 10 to, 100 z/„ ibid., 101 r., No.. 96; HG., 
ed. Berlin, 319 (wanting in She., No. 103); abbreviated to 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 309 

tfyvb '*W D13 MS. Oxf. (fragm.) d. 48 (276o 10 ) fol. 23 r. 
(No. 6), 23 v. (No. 7), 24 z>. (No. 8) ; Oxf. (fragm.) e. 16 
(266o 8 ) fol. 12 r„ No. 4 ('d$> 'IV '13) and fol. 12 w., No. 5 
(the same). 

D13 is not confined to Pal. Aramaic, as Ginzberg main- 
tains, but was also used in Babylon in public inscriptions, as 
e. g. Baba batra 58 b : note WWH '•3KD D13, and in lectures. 
For this assertion we have not only the above-mentioned 
testimony of Nathan ha-Babli, but also such Babylonian- 
Hebrew passages in the Babyl. Talmud as the frequent 
note p DT3 "T3 ;n niyo D3"oe£ *met<B> nnan, Erub. 16 b 
(R. Nahman in Nehardea), Shab. 63 b (R. Dimi, ibid.), 
Hullin 56 a (Zei'ri, ibid.), Erub. 104 a, B. B. 127 a and Nidd. 
68 a (Rabba in Mahoza) ; comp. Yeb. 108 b : mvr an "idn 
p DT3 '13 U3"OTI t6 \b 'DV '"13 5>t«NDB» 'l ~\OX '13 31 1Dt5 
'13 iW. 

However, this word was preserved later in Babylonia 
(nor is it very frequent in Syriac), like many another word, 
only in the official language. 

"]nx signifies in She. not as in Yerushalmi ' it is 
doubtful ', but, as already mentioned, it was shortened by 
the copyists from tp'D? nx "jnv D13, and this certainly is 
good Babylonian, B. m. 101 a : rnTnsnv NT |U3 ; ibid. 16 a : 
B<xb mnv pt< it; Naz. 19 a: in"3nx t6, the frequent vb 
tonv, vbx ro-iw ttb; Bek. 31a: tontavt* twi 13 tun 3-1 
tww nb, &c. 

6. nw T~a . xsbx. A blessing precedes the answer to 
the 'question': No. i: im (F tuem T13) n"3ipn nw T13 
(F t«W>i fi^ob) Nabxi? tun nco n» by twiivoi xmix toi> 
i>tne» irai rroy. So also Nos. 64 s , 125, and No. 7 MS. 
Oxf. d. 48, fol. 24 r., margin ; abbreviated to rvotr "p 3 
n"3ipn Nos. 66, 92, 121, 145, 154, 165, 166, and 170; ed. 



3IO THE JEWISH QUARTERLY REVIEW 

Wilna, III, 226; Geonica, 381, and HG., ed. Berlin, 235 
(wanting in our copy) ; shortened to "pa, Geonica, 364, 
368, 371 (jtw Tia); "vi-itm, I, 152 (=157 n»ee> T"ia), 277 
(.TDK' T"D), II, 69, line before last (TW r. T">a), 1 01, 119, 
141, 160, 174, 188, 222, 255, and 275. In MS. F, beginning 
with No. 6 up to 58 throughout: n"3ipn nw fia, or 
'131 iTDtr ina, or even T"ia. Otherwise it is missing alto- 
gether. 

'« ann n"aipn rrac 713, comp. Ber. 54 b: tnErn T"ia 
'unann; 58a: '13 twittta am worn ina; Shab. 88a: T">a 
'13 *wv5>n INniN arm worn ; but n"aip is frequent in Babli 
also otherwise (comp. also the t?np prayer : nw '13 ■pam 
rfaipi), Hebr. HG., ed. B., 15 : '13 inatr n'apn bv let? Tian\ 
This introduction is quite frequent in the later Midrashim. 

S'S?N likewise is not confined to Pal. Aram. ; as a matter 
of fact it is found in b. Naz. 4a: vnpob i3ts> ")3tr '■tfbxb 
and in Anan, ed. Hark, 22: 'S&K, |UBli>t«, pa!>o, tyi>D; in 
a question from Basra to R. Moses, Gaon of Sura, Geonica, 
2i4 6 : »nna»i> wno s^r and Yeb. 63 a: is^D xoye 7J0 p'arr! 
s]5>K was therefore current in Babylonia alongside with t\b\ 

7. Then follows: t»!>»*i wvjnoa wan un 73 -ho'n 
(whereupon . . . pan un, or . . . pn), No. 64 2 (E and F 
have p moN), 66, 92 (which in E and F is found in NSn), 
E (wi>H x^D>3n pn p -wk) and F (N«o»an i:n p Ties 
win wwnea), 121, 154 E ('ia ia nie^F), 165 ('is is 'cn, 
E -I1DN), 166, 170 (E -non), and thereupon the Derasha. 
Now Reifmann (nioiri rva, III, 28) made the most absurd 
suggestion that it is a mistake for Yc'ir = wai nsw 'T h?, 13 
and this was incorporated in the 1S1N, j.z;. Nrw, without the 
least scruple ! But, indeed, it is the imper. "WD't? or "riD'K 

13 But 64 has indeed W3n HE'D »T by anteceding! 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 3II 

'say'! in Babli NCK, iD'N (Aboda zarah 10 a: tb'yb lO'N); 
in Pal. Aram. 110N (tid) and "iCX (Dalman, Grammatik, 
p. 246) ; in int 'n, 105 : nyn |"jj3 wn = ncn of HG., 
ed. V., 96 d, u (see also below sub $5}, comp. 11B' I N and 
Mand. "indn and "WON (Noldeke, Mand. Gram., p. 242). 
It is an exhortation to the interpreter, tniDN, NJDmno, 
standing near the learned lecturer, to begin the Derasha 
(comp. Sanh. 7 b et al.). 

ttfrl NmnD. No. 64 follows a Baraita ; 66 Bar. ; 
92 F Bar. ; 166 Mekilta de R. Ism. (but cited here from 
Zeb. 116 a); 121 Mishnah, 125 Mish., 154 Mish., 165 
Mish., and 170 Mish. ; HG. Gittin, ed. Berlin, 317 (the 
whole She. is wanting in our copy) : x&l xmnoi pm 
Mish. ; ibid., Kiddushin 340 (missing partly in our copy) 
pn Krunoa pm Mish. ; ibid., Shebuot 499 (missing partly) 
N3^n Knynca pma Mish. 

Ginzberg concludes from this fact that here to^H Nnono 
is employed for Mishnah and Baraita, that the author 
lived at that time in Palestine, where everything was called 
Nmn». But also in Babli wan (= K$n 1 pi., written 
Njwari, She., No. 48*, HG., ed. B., 200 = pn, in Yer. p'an) 
is used for Baraitot, e. g. Ber. 8 b, Taan. 21 b, B. m. 59 b, 
73 b, Hul. 29 b, where Rashi, it is true, resorts sometimes 
to the explanation NJK W (Taan., /. c. ; Ket. 53 a) ; but on 
the other hand it is evident that X3on refers also to Mishnah 
(e. g. Meg. 27 b, B. m. 4 b, and frequently p3"i um xxb won), 
likewise the frequent Win *D) pS *|N , which is used even 

" Otherwise 1KT 'fl translates NCK by TON, 28, 64, &c. The 
INI 'H construe "WK, "1D*K ("1 11D*N) as imp. also elsewhere; p. 44: 

npTi npii> nieni oidw=) ""»n; 6i: u ne*u nvi (-ncK) tick 
15ns wain n^nno; 130: p»a 'dv n pdkV -now-) idnc iick 

1DN Qli>3 nSBO ; but it occurs also in the first pers., p. 93 : fTMnK DN 
'13 1^5 E»E> "I1DW 'IS. See also below, skA 35. 



312 THE JEWISH QUARTERLY REVIEW 

for Baraita, e.g. Ker. 22 a '13 twm Win "v: pN g|K! 
Furthermore Ber. 22 b: N"n3 '13 np ^y3 (Tos. II =) "i"n 
'« omoK onm n»3 toyn '13 hkd dws nnns^ i>3N itwy!>. 
Entirely against Ginzberg's view speaks the word nin'pn, 
which is likewise used for Baraita (e. g. B. m. 23 b). Also 
pirono and pn for Baraita has survived, despite all the 
corrections of the levelling revisers : prwnD, e.g. Ned. 55 a 
(She., No. i2i, has towi instead !), Tern. 14b: '13 'DDTi vtn 
jwanoo, ibid., 16 a: '13 «|i>K son jwinDl ; on the other 
hand wnriD for Mishnah Tamid 28 a ! comp. B. m. 104 a : 
NTWriB Tiin 'in. Further, innumerable times pn for M, 
where we could not possibly invoke corrections, e.g. 
Ber. 27 a : pm Nr6 wn jkd (= Mekilta rb&2), Yoma 83 b : 
'13 DWtn tTO pm Wi, for which Hul. 106 a: 'onia N3N 'in 
(see R. Isaiah Berlin, <*</ /^.), B. m. 119 a : nhy 'M »D3 pHi; 
Hul. 6 a, 14 a, where the editions have fcOJm, but the 
Tosafists had pm — pm also Erub. 36 b, Yoma 55 b, and 
Suk. 23 b according to Tos. and R. Han., ad loc, likewise 
Git. 28 a (25 a erroneously twm), B. k. 69 a, and Me'ilah 
22 a — ; further Hul. 14 b, 15 a, 31b, 58 a bis, 69 a, 77 a, 
82a, nib, 125b, 138a (see the marginal glosses), and 
even Dm pn 87 b (see Tos., s.v. pn !) and 120 a (see the 
marginal gloss) — all this in one tractate ! — and many more. 15 
Attention should also be called to Besah 16 b: '13 Y'n 



'13 <an ^333 pit? Dnxeo nh rr?ui> n""i rbun awo (pw onx = 
rmv:n for Baraita). Similarly vice versa, N^n for pn is 
very frequent (comp. Ginzberg, Geonica, p. 190, and 
Aptowitzer, MGWJ., 1911, 637, and n"»3tn, 89, n. 8. 16a 

16 Comp. in addition Tos. Suk. 4 a, s.v. JV3, ibid., 18 a, s.v. n3V"IN3 pN, 
and Nid. 26 b, s. v. in?nrl3 , likewise Sepher Keritot. [See no w also further 
proofs by R. Baruk b. Samuel in R. Bezalel's !1D?nn V?3, Hoffmann, 
Festschrift, p. 216-17, No. 421.] 

u * [More detailed data on this theme, in my book Der Mischnatext und 
seine tlberlieferung (in MS.)- See also Jahrbuch, XII, 121 below.] 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 313 

For twan un, Hebr. D'aan up (125 = 87), UTiian up 
(167), comp. the beginning of the sixth chapter of Abot : 
'uj rupan pe^a anaan up, in iflim frequently rupaa caan upi , 
xnBDina onaan UP, &c. ; similarly in the later Midrashim. 

twan, in She., No. 66 e F : anaai> mot? n'DP iwarTnDJo 
'ia D'sava payp nnijn s^K; likewise HG., ed. B. 38: 
'la war soim iwan poxi (in ed. V. here : pds, no doubt, 
as is the case frequently, from the She.). The She. have 
usually pan lias. 

After the Darasha: 'ia p"n jwnanp nj^nkh xni>W ps&i 
(Nos. 1, 64, 92, 154, 165, 166, 170, and MS. F, Nos. 6 and 7) 
or merely sn^NP pj£i (No. 66 ; MS. F from Nos. 9-58 
throughout). On Krp'KP pajfo, see the parallel above jw# i. 
yap Nn as in Babli. 

8. tnco^. As an introduction to the closing sentence 
of an argument the She. employ Kna^ in the sense of 
•fr "rc/6 of the Baraitot (e. g. B. k. 23 a) or HKbx of the Babli: 
No. 10=37 ana sna^m vncnb, 21 bis (for the first \no^b 
Ket. 68 a b"ap), 27 end, 51 2 end, 58 bis, 80, 88, 96 bis, 103, 
126, and 152 bis. 16 Further, in the sense of iota (PIT'S) 
Nos. 14, 29 ('ia spa Nairn Ttvb), and 95 (spaooT itytb, 
F NP2DO 'p»D1 tnavb). It is used frequently in Babli to 
introduce a retort, a question: 'ia nasal 'ia toco^ ('UTVB, 
pnm), or in the sense of tm^, sou ('ia 'una NBU frequently) 
Ber. 13a bis; but in the same sense as She. Ned. 35a, 
58 a, Naz. 39 b, and Me'ilah 25 b ! 

9. na nax, na nb^, and the like, "ia nas in the 
Talmudic sense only, No. 35 : 'ia rwa na 'otn npinn nyiaP 
npinn nyiap snpna na^sa (this is said at the beginning of the 
No.); likewise HG., ed.V., 39 a above: 'ia ia natn an^P jar ; 



16 No. 32 : ia payn upan ?i>aa, f has 'ia upan k-ib»b!>. 



314 THE JEWISH QUARTERLY REVIEW 



ibid., 41 d : '13 nyryac D'"m id not* (pDS), which was stated 
previously in HG. For this purpose the She. employ 
|J'-idk Np "ID'D (see below) ; in its stead we find frequently 
in the She. the expression '13 "id veab '31 '13 no wsb vb 
and "i» with the third person of a verb inst. of the second : 
No. 31, '13 p'3C 'D2 tOfl no KD'i> "31 ; 48*, nTDVm ID KD'5> si> 
'13 ; 68, '13 J'DID 'i>J>3 no ND'i» '31 (= -i'nrm II, 3 : ncNn DN1 

'13 pdid '$>jn norm B"tr); 86, K'i»eH -id nd^ xi> (F t6 

N^BHD ND'i>!); 118, '13 Dm "IDS3 'K "ID KD'i> '31; corrupt 

No. 3a, '13 'i> n'xn no td 'd3'$> &6 (E n*b n'st no kd'^> s6, 

F fl'$> D'SO JKD ND'i> vb) ; 33, '13 n3'K»D TD i1'!> TnVD 'NT ; 

48*, 'is td rrb nsra '3T ; 127, td n'i> spy '3 ; 70 (ed. Wil., 
71) end, NDp iT^jntd n'i> 3"no Kp 'KDK, r. with E and RNZB. 
ni"Ms no (b. B. m., 43 a : ninn tuiwr ktidk rrb n3"no up 'kdk). 
Likewise HG., ed. B., 178 (where the whole is probably 
derived from the She., but is missing in our copy) : nd'j vb 
-id (ed. V., ND'n N^>), ed. V., strip (which no doubt belongs 
wholly to the She.) 12 c: '!"D 'Jm ID nd'J t6i (ed. B., 44, 
"id 'dnhi i6l !) ; comp. further Seder R. Amram, ed. Frum- 
kin, I, 84: '13 bin3 ND^C3 ID ND'i> '31; corrupt in Pardes, 
ed. Const., 57 b : '13 !>in3 rti'n ID KD'n '31 ; Siddur Rashi, 3 : 
'13 ND^Sja KDTl '31 'JD't HD3 ITVDD ("ID ND'i? '3l) 131 ; probably 
from the She. also HG., ed. V., 96 c: 'Nl '13 "id "idn w 
'13 -id -idn (=Hal. Pes., ed. Schlossberg, 39: '» nDND DKl 
'13 -iDxn dxi) 17 and HG, ed. V., 130 b (= B., 537): 
'13 -id nb «i'-ixd nsi (not in b. Hull. 55 b). 18 This expres- 

17 Sherira piX njffi>, 94 a-b quotes this as 'SCI13D f J3T1 SBOTB and 
adds: 'N113D fJ3-|T NDSJ-D VTDW1K Wl [1^31 fltf! "nD3 '!>'D ]bt< 
(1. niD^ri3) 11Diri3 JJ31 v6wn |in'D1B3 pnDI, comp. my remark, 
JQR., N. S., IV, 435, n. 11, and above, p. 302. 

18 The halakah of iTTI?J is quoted in the name of R. Ahai. 



POST-TALMUDIC- ARAMAIC LEXICOGRAPHY — EPSTEIN 315 

sion, which in the She. is addressed to the public, is 
preserved in Babli, as far as I know, only Ned. 71a: 
'13 ID -IDK *N1 '13 no "ION »N '13 ir6 WN. 18 " 

10. pnex "ico, pnos xp nons, pan -msx *id»d. In 
connexion with an already mentioned halakah, to which 
then a further query ("pv D~ia) is appended, the She. make 
use of pnoxp no'D: No. 15 end: '13 innap DN pn5S TDV3; 
36 : '13 j3'3«non p"-)DKp "id*d T~»^ D">3 nM <K (so E and RNZB., 
F has 'inosp nD'D nM "N) ; ibid. : '13 ns 'n»N xp nco ; 54 2 : 
'is pncx -i»ns ins o~a 'dj 'X. 48 F follows the second 
She. : 'is xs'm '13 mpssn jxd i^n pnnsn io'd t~i* ma <m 'x, 
whereupon comes a whole She. — somewhat abbreviated, 
it is true — which is wanting in our copy, but is still 
preserved in HG. in its greater part ; 100 : ico *px d-)3 
'13 tn»B pan mox (stated previously in the She.), and 129 E 
and F : '13 ''tap^ -|1DK pan mox -ico "p-re ma (in the She. 
above). 19 Further Geonica, 371 (corresponds to No. 96) : 
'*IOX Xp -ID'D 'D^> 'X 'IV D13 ; 380 : Xp ID'D flX T~N D"U 
'13 TDXn pnDN ; 38 1: n&'Drl 'JHX WX 31'ITDT p»TDK Xp ID'D 
'13 = HG., ed. V., 31 b (= B., 156), in accordance with 
which the She. Geonica is to be corrected and completed 20 
(which Ginzberg failed to see) ; Hebr. in She., No. 146 : 
'13 D"JB> WiOX nn 21 (the whole number is extremely abbre- 
viated, but is preserved to a great extent in HG. and Hal. 



18a On the other hand there is found in Babli mOX Suk. 38 a and Naz. 
60 a, inst. of nD TDX. 

19 No. 137 E has: '13 ^3 p3"l "MOX X"3 '03 *X, ■'TDK 
F: 13 ^3 p3T "I10X "p* D13, "HiDX (stated in the She. above). 

» p. 380, after DniSSH D1H 113'3 I^BX pniDK >D, a lacuna; 
whereupon supply from HG., /. c, as follows : fX»n pnCK Xp nD'D] 
'13 rrO 3*nD [Omean DVa ^3X1, ed. B. 156 has here: nD'D »D3 "X 

\vnax xp. 

81 IJIDXC *in, also in "Wltm, I, 35 5 «~IDX nn, II, 113 = She. No. 8 S . 



316 THE JEWISH QUARTERLY REVIEW 

Pes. fragment 21b ). Again, in the She. extant in the HG. : 
HG., ed. V., 37 b (= ed. B., 187) : '13 pnoSp" no'D nspiD pjjfo 
(almost the whole of 31B di» 'n certainly from She.) ; V., 57 b 
(= B., 375) : 'i3 harr s^n pnox xp nco d-d (B. : ins ma 

'13 nc») ; ?'&#., 57 d : '13 JVJI^sn pnCNP ICD 0*13 (B., 277 : 
'13 nD'D 'nv D13), all from a missing She. ; 1133 'n, V., 123 d 
(B., 479) : '13 npcDi i»3ND (pnox Np) pnosp ID'D (= She., 
No. 56). In Babli iat* no>o is quite frequent. 

11. 'jnp. As an allusion to something already men- 
tioned the She. employ often the term 'onp (inst. of no not* 
of the Babli) : No. 58, '13 niJIBD 'JH »:n Sp ; 74, 5>ioi> 'jnp 
'13 ns (likewise E, but F erroneously WD ynp). However, 
this is the case already in Babli, comp. Me'ilah 2 a, 7 b, 8 a, 
12 b, 13 a, to which Rabbinowicz (/. c, 235) has called 
attention. 

12. W8. In Babli CT is used both in the sense of 
' being asked ', properly ' being sought ' (comp. "I3?n ne>p3M), 
' being needed ' (comp. "pnv D13 above), and in that of ' be 
in need of, have to '. However, B. b., 58 b : p»D yono [on 
(are asked, are needed); Ned. 50b: wb >j?3nD NDD "WO (is 
necessary to him) ; Suk. 53 a (in the mouth of R. Johanan, 
it is true) : 'jonen ~\mb (whither he is to go) ; Sanh. 98 a, 
MS. Karlsruhe (in Margolis, Grammatik) : Wjnno ND^H 
(perhaps I am asked for) ; Moed k. 28 a : 'jane N^l (and 
it is not asked for), antonym ronCD N^l ; in Yerushalmi it 
occurs frequently in the sense of 'being searched', e.g. 
Erub. I, 19 a: 'rows xi>i ''MID nyHivN - , likewise in the sense 
of 'be in need of, have to', e.g. A. z., 1. 39b : np'D *jnn'D n:k . 
The She. employ predominantly in the sense of ' have to ' 
(with -b) : 'WO ('jniVD), so ed. V., Nos. 1, 3 (here E and F 

21 b [See now J. N. Epstein in the Jahrbuch d. j. I. Gesellschaft, Ffm., XII 
(1918), p. 123.] 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 317 

'WD, »ino), 4 (F W«), 8 thrice (likewise E and MS. Oxford 
d. 48, fol. 24 r., F on the other hand 'WD in the first and 
last), 15, 16 (E 'WD), 18 (F "V30), ibid, end, 19 beg. (F 'sm), 
z'&tf. (E 'WDl), 31 to (F 'WD, the first time also E 'WD), 
32 thrice (F W», 'W3), 31 once 'W/i'D (twice 'WD), 3a bis 
(F 'jod), 34 (F, E 'WJ, WD), 36 'W/i'd (F and E '»td) 
beside 'JQ 22 (F and E TO) and 'WD, 41 (F 'WD), 54 1 to 
(the second time F TO), 58, 63 thrice, beside 'WD, 89 (F 
'jd), 10 1 (F 'Wdi), in itrsj nio^ i? 'larvS (E 'Wk, F 'jok)> 
130 to, 147 to, 161, and Geonica, 373n ; in the proper 
sense No. 79 : '3'5f WD. Ed. V. and the older MSS. 
(Genizah, Oxford Fragment) have therefore mostly 'WVD, 
by the side of the rarer 'WD (Nos. 31, 36, 63). n'!> 'WD 
in ed. V. still in Nos. 2, 28, 45, 51 1 , 73, 132, T36, and 167, 
many of which, as most of the cases in E and F, are to be 
ascribed to the haste of the copyists who put down the 
form more familiar to them. Similarly 'Wi'D is frequent 
in HG., e.g. ed. V., 5b bis 23 ; likewise Jehudai Gaon 
Responsa, ed. Lyck, No. 45, frequently 'Wl'D N^. In 
Anan, 'W/id, Hark. 113, No. 47 four times, Schechter 3 7 = 
3°u> 3u5> 527> beside 'WJ, Hark. 7 bis, and Njn, Hark. 
81 bis, Sch. 5 1. 1., i5 13 _ 16> 28 10> 22 1. 1. 

13. KVn 'which' (fern.), is frequent in She. : No. 26 s , 

82 'W, also No. so ; }3'JJ3, No. 35. 

» Instead of Ned. 8 b (Ar. and RN.) : &m'J '"lt!»Db '^>K1 tib we find in 
HG., ed. V., 75 c: 'WID JO, no doubt only a paraphrase of the Syr. 
Jlo ' be proper, have to ', and not a variant, comp. Rabbinowitcz, 

D'isemv.c. 

24 For this in the parallel 161 : W'3'D VI, but alongside with it NTH . 
Also No. 26 1 : t|HJJ lTO'D VI, likewise 64, but for this in the parallel 12a 
Wvn. No. 142: '3'3niD liT'J'D VI; Geonica, 353: [1H"]3'D JUVI, 
r. XTn. 



318 THE JEWISH QUARTERLY REVIEW 



28 : nt> nvpSD tnvr? p DX (E, F and ed. 'Arak. 15 b lrwo vi), 
34, 53, 68 3 , 77, 83 to, 135 end, and 161 (on the side of <n) ; 
pvf' which' (masc), She., No. 145: p*wn 'XIJ ♦B'D }W X^>1 
'13 XD'J (r. pvn); HG., ed. V., 47 b: XD'J via. The Babli 
has »n (x^n - ) instead, the Yerushalmi, however, has XTn 
fem., pTfi masc. throughout. Basing himself on this fact 
Ginsberg again concluded that Palestine was the birthplace 
of the She. But XTn (hti) ,pvi is good Babylonian as 
proved by the following : Geonica, 390, from HG. : xn pnexi 
ni> »:n xp xvnx pjw 'n (HG., ed. V., 69 c <Tnx) ; HG. 
Berakot, ed. V., 6 c : niaxa '13 xnSD 3T nDX nvu ; *'&V/., 70 b 5 : 
xa^DX xo'^x Tnx, b. Ket. 28 b. : x^nx ; ibid., 1 1 1 b : ItW'Xl 
irwo XT <n any ; ed. B., 454: XTn, r. XTn ; ibid., 178 : NTiTl 
Hants nx>3 <rw ; R. Saadya, Responsa, ed. Hark., No. 555, 
p. 277: xvi xnm xth icna an m^o »xi; and the masc. 
and fem. is found already in b, Ker. 3 b, fHVi ; 4 a, |H Ti ; 
Ned. 26 b, ]n 71; Men. 109 a, pPVi (r. pT.1, ed. Ven. 1529 
and 1548 : p ti), comp. pTX, Tamid 32 a ; Zeb. 69 a MSS. : 
nvT xwt x!>i '13 dw -nvDE' (MS. R i : xtw \n, 1. xn vi). 
Syr. p'x, XTX; Neo-Syr. XTn. 

Here belong also the forms \hn (inst. of *J.n) and jiFx 
(inst. of Wx, lruy), pn and "inn, which are very numerous 
in the She. : pfci, No. 22 : "b'D pno (E and F like Fd. Mno) ; 
73 F: pj»en neon pf>m; 76 F: ncrcn pS>no xmijjD nkj; 
85 F : ?bn ba on trim ; 95 F : pbno xin cap >xi ; 147 : 
X'D jbib; and alongside with it *Jn, e.g. 103: n$OK> ^n!>i 
can?; 137: nnyp yn, M and constantly »5»d ym. Likewise 
Anan, Hark., 28 2 : '^o pM>; 45: ^xo pbnto; 47, 48, 51, 
56 bis, 94, 95, 96, 100 : '13 ^n 13 '13 pi>n ■£ pvotn ; 103 
58 But No. 168 mrj '•am , F and E correct lilTl \TI. No. 88 beginning : 

'13 pma pine wwd <ni pxoe irwo <m, E only wre '•m 

Jims D'KOe, F only Jjma pXOe WJ Vll, as No. 95 and others. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 319 

below, 104; Sch. 3 25 : rrnpn xosn -fin '3 xw '«!> »w pirn; 
35 9 , alongside with the constantly used »/'D "'Jm, and TW, 
lxw (Hark., 5, 89, &c.) by the side of "jlw (frequent in 
Anan, see above, pointed *!?? , Hark., 58 ; "jiw also used by 
the Geonim, see my Rechtsgutachten, &c, p. 248). "jw in 
She., No. 34 : ynnx Tw by& '« 'dyi "on ; 39 : *dj "pr6 l^sx 
'13 x!> yn kdI»h ix 'i3 (F: '13 x^ iwi> '« n;xn$> 26 i^bx) ; 
51 : <YB inn; 68 3 : KD1JPD TW1; 86: W 'r "pm; 88: "pro 
nnsv; 103: nnera "pm nenm wi; 135: ys 73m 27 (Fym). 
Mand. p^Nn /«, by the side of yw>n z'//z, Noldeke, Mand. 
Gram., p. 54. In the Babyl. tractates Nedarim, Nazir, 
Temurah, Me'ilah, and Keritot, 28 as it is well known, jbn 
occurs likewise by the side of v'D yn and "pn, all of which 
are undoubtedly Babylonian, f&tt is even more frequent 
in She. : No. 14 (on the side of WW), 15 (p'X), 16, 17, 3a — 
six times! — 34, 35, 43 end, 47 to, 64 s (citation), 74 to, 
77, 79, 86, 89 bis, 95 thrice, 103 bis, 147 to, and even more 
numerous in the manuscripts. Alongside with it is iru»K, 
e. g. a, 51 2 , 104, 114, and vwy, e.g. 14, 88, and 108. pin, 
No. 11, 36 beg., 66 4 , 74, and 126, alongside with 'Kit, 
ibid., &c. nth, No. 30: xn^D Kin (49 and 131 : xn{"D xro), 
79, 68 1 , and 121 to, alongside with xn, e.g. 47 1 (see also 
above). But all this is known to occur in the above-named 
tractates 29 and likewise in Anan throughout: p^X, with 

26 -pxn 'that' in She. No. 50: iV^ priTO Xpi TWI JTP X^H J"I?X 
'13, 76: 1WDT 'XHO 13 ItMD x!n YXDD, 137: DITTO yttn, Sec— 
Anan, ed. Hark., 13: '13 "Xn P|X 13 T'XH ?1D, 31-2 the same, 
83 pointed "p'Xn DO, «'6<rf., No. 11 the same, 84-5 the same, 48 Y'XITI 
XJXO; ed. Schechter, 8 M , I7 10 : "pXil. 

37 Gaonic also *pX (Bab. "P'X), see Aptowitzer in JQR. for 1913, p. 40. 

28 Besides also B. mez.goa: '13 Hin \b<\ ^XlDCTt iTDx5> fT 1 !? W^EH. 

29 prK also elsewhere in b. Ber. 31 b : pj'K JXOl ( = Shab. 118 b), 50 a : 

JVB3 prx. 



320 THE JEWISH QUARTERLY REVIEW 

inrx once, Anan ed. Sch., a6 22 . JWK also frequent in HG., 
e.g. 30a, and with the Geonim, e.g. Geonica, ai4 9 . 

14. m t6, rfX vb = my>b rv!>. The She. employ 
xa'x ab for xa^ and JTX xi> for n^: No. 10, ed. V. : 
nb~\v X3'x ixh :"yx, xri33D m xh (on the contrary, ibid., 
X3^ TDBTl, but Oxf. d. 63 has also here m xb, see below) ; 
16: naina xs'x xi> xm (E naina nanax xb, F xana xi> 
naina) ; 96 : rvx xh \n (F rvxn jva ! E rww nivon |M 
'13 mtr). Further, No. 17, E and F ; n*W> nwx xh K3n ; 
24 F: x-iiytr X3<x xb wi; 68 3 F: X3>x «h '13 '■x yn' 1 t6i 
xan (inst. of Nan xa^ "Wil of the ed., E anp xa'w »x) ; 170 E : 
x*in: nay fine X3"X xi> bx ; MS. Oxford (Fragm.) d. 48, 
fol. 23 v. (No. 6 end) : DJJTtD xa'X xb (the respective passage 
above 23 r. is not legible). Similarly the old fragm. Oxf. 
d. 63 has W'K ab throughout : fol. 55 r. (No. 32) : 'tMX prm 
'13 twx xs^x xh; 56 r. (No. 34): cnaip rvb irx xbi, 
'ia twx xa^x xt> w y (rpDj»)n xa*x xb 1x1 (yet txo!> $>ax 
'la n^ n^ won ; other evident cases of xaw xb in this 



No. are illegible) ; 100 r. (No. 10): r6ny xa^x xbl 33 bv *|X, 
xa'X x^ Tina nx ; 100 z». : xriMD xa^x xb~\ (as in ed. V., 
see above); 10 1 v. (No. 97); ax n xvx xbn xsxn i>ax. 
Also in Gaonic literature : HG., ed. V., 4 d : itrpbvb fwvx pa 
rvrvx xb pai x-na'n ; ed. B., 360 : xb xn rwya rnvxn 'eye 
'13 wa nwx (ed. V., 90 d like B. k. 105 a : rwb xn) ; 
Responsa nroj mon, No. 62: lrwjn minx xi> (Ket. 55 a 
injn^) ; ibid. : 'ia powx xi>l '13 btpdJ) ji? jtx ttb ; Responsa, 
ed. Lyck, No. 62 : IVX xi> ; and Geonica, 33 : JVX xin 
[R. Samuel b. Hofni cited in niawn nsD (ed. Friedlander 
in Hoffmann, Festschrift, Hebr. division, p. 90) from Git. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 321 



14a: 'an tip rpoj Mff «in nw WD, edition: 'KM xa^i 
lONp. — The inf. N3»d is a splendid parallel to K^> nWD 
nv nsaiN Git. 6a a, comp. Fleischer in Levy's Worterbuck, 
II, 454]. In Anan — it should be added here — Na^N vb, 
rVN ab throughout: ed. Sch. q 20 , 143 to, 23 28 , ed. Hark., 
j 4, ai to, 6a, 115 to, 171 (m *6); Sch. 6 31 , 9 21 _ 23> 25> 
^a.u, ^s, 3 3i, 3, Hark., 5, 18, ao, 41, 67, 107, 108, 116, 
119, and 171 (fPN N^). This is probably the older form, 
BA. wk «b (Dan. 2. 10, 11) "rriTVK «b (#«/. 11), Pal. 
Mak., I, 31 a : WK mb Of Q'vhv "3t\ (Shebi. 39 c : "TVN x!>), 
but it also occurs in Babli (alongside with the more frequent 
abbreviated form n^ and vab) : Nazir 38 b : fWN n!h 
(on the side of nh, often); Ker. 5a: nWN tib (but 10 a 
■Tn^) ; Me'ilah a a, b, 5 b, 7 a, 30 ia a : JVK k!>, by the side 
of rcb ibid, and 9b; 19 a: tO'K *6 ; J/zV/r. ha-Gadol, 52 
(from Sanh. 97 a) : XDClp WN tfi> to (^i/. rwh) ; Targ. 
Onk. Num. 24, 17 VTIJVX ah Cod. B., R., W., for TWV^I, 
variant in Cod. 7 (Berliner, Targ. Onk., p. 47, see also op. cit., 
143) ; Onk. Gen. 30, 3a Cod. Angel, has variant VTiJVN tibl 
inst. of Ttirvh ; Gen. 7. 3, 8 we have for vnrvh the nvJoy 
(isJijl^. a reading from Irak) : wn nb~\ (Berliner, Massora, 
p. 33), for which Berliner rightly suggests \"i[WK] x?T. 
This would then be a form from the dialect of Irak ! see 
also above, introduction, p. 304. 

15. ^5? instead of N. The K constituting an abbreviation 
from ?y, ba, which is very frequent in Babli, does not occur 
in Anan at all ; in She. 'we note a great tendency towards 
by -. No. 6 MS. Oxf. d. 48 : ntJOB nora by mm (? Dy)T>» ; 
No. 10 Oxf. d. 63 : hno ^ 5j? mm njn'O = No. 37, ed. V.: 

s »5b: pjJVN «h f^lDB '13 iTJVtO KDB, 7a: inen TDK KITI 
D^jn priT'N &6. fOVPK in Anan, ed. Hark., 48, |WK, 106. 
VOL. XII. Z 



322 THE JEWISH QUARTERLY REVIEW 

'13 nj by mm •no ; No. 35 : '13 m by mm "I'D ; 97 E and 
F: (F V3K DS byi) V3N OKI ION ON 5>J> mm '•TO; 157 F: 

pciy by mm nio ; No. nE: njvit sin by mb ntsn (ed. sim, 
F Kinb); 33 F: mosj wbs by Tpao xpn (ed. inbisx, d. 63 
illegible) ; ibid. d. 63 : moay |inbl3 by Tpa xbl ; No. 24 : 
na ab'Bp by jraoo ; 43 : n<a by w mb aw xpn (below : d"pid 
n'BM mb) ; 68 •. xniv ^yn (F Knirtn) ; 96 : xin by xntipcn 
ttttn (E, F tnriK) ; ibid. : smo n by mb sruc (E, F *3K) ; 
ibid. F: ibDB N3HD n» by p^apcK w (ed. and Geonica n«K , 
see below) ; 101 : nnp< by b»non nan (F mnpa), by b^no xp 
wain sip 11 (F rmp-N) ; 107 : tains by worn ; 99 F : H piD&<b 
sr» by WD; 137 F: 'is nmp an 'dk no by (ed. and E 
xobx, 'oba) ; 141 F : nou by mnnw (ed. newt*, E nam) ; 
and 15a ed. and F : nw by rwijn 3W (E nie»KK). 

This by is frequent also in the above-named tractates 
of the Babli: Ned. 14 b: NJTIN by (preceding it Nyntttt, but 
MS. M., R. N., Nachmanides and Asher : NJHK by) ; ibid. : 
by mnvn beside -s mnyn; 17 b: D«n by D<3:y by; 23 a: by 
mnyn beside mnyix ; 25 a : snyT by beside snjriN : 70 a : by 
rre>£H ; Naz. 24 b : yyo by (Kid. 56 b and else TJTOK) 
beside often "K ('WK, N3^N, Nt^nN, &c.) ; Ker. 9 a: tow 
xnn by nb ynon ; likewise 22 a : «nn >y '« tow ; 14 a: by 
noma tnn (alongside with ana, 15b); iab: by nDsnjCK 
(SB'>DN=) ttt>"D, beside fccbx, *'fo'<£ ; 26 b: NDn pn by is<b 
TiNp NDn pn byi '13 'nxp ; 27 a : nnas Kin byi ; and 27 b : 
swp wi dpk by "am dew toe>p '« (= bk>nn). 

16. Dymo, iTD. In Babli the shorter form 'TO is the 
most generally employed, while Byre is rare (see Levy, 
III, 31 b) ; especially numerous is DyTD in the tractates 
Ned., Naz., &c. (see Rabbinowicz, B^BTV, /. c). In She., 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 323 

as in Gaonic literature generally, DJTPD is predominant : 
She., Nos. 6, 11, 17, 31 bis (F has 'TO in the second passage), 
55 1 D5TPD (= 55 2 n'D), 47 beg., and abundantly elsewhere; 
however mm H'd, Nos. a, 6 (also »TD to^, but thereupon 
DVT'O N3^), 10 mm HD = 37, but Oxf. d. 63 here : DJJT'D 
mm (see above), 35 end, 50, 58, in, 113, 163; ''TD also 
e. g. 30 and 70 bis. In the Gaonic literature mention must 
be made here of D5?T9, so pointed Geonica, 183 ; , J)T'0 z'fo'^., 
173; and the pi. TO, 'OKTD, from a sing. KTO, Responsa, 
ed. Hark., 190 (see my 'Nachlese' in Jahrbuch d.j'ud. lit. 
Gesellschaft, X, 373). In Anan nUD is frequent, Hark. 3, 
J 3. 3 1 . 34. 35. 3 6 > 49 Passim, 55, 83 ; Sch. 4 10( 12 , 5 21) 23i 27 , 
8 U , 11 1. 1., I3 5> 6 ; more rarely Dymo, Hark. 18, Sch. 14 1. 1. 
17. jru, *n , D. In Babli we meet in the perf., part., and 
imper. forms of 3m ; in the imperf. and inf. forms of 3ru* : 
3JVD, iaiVD, and 3VW ; inf. also of am : amo ; impf. of jru in 
isolated forms only : Ned. 55 a: [Wivi, Gittin 56b and B. m. 
51 a: |riN ; very rare is an inf. of jn: : mi> jivd^ Sanh. 113 a 
[R. Samuel b. Hofhi in manievi nsD (in Hoffmann, Fest- 
schrift, Hebr. div., p. 89) cites from B. m. 32 a : 'plat* Dnn 
ItJ? mi> ;rvDi lab widd, edition however : imrh 3Til»i r. with 
MS. M. arvoi ; Pes. 12 a and b MS. M.: \r\il inst. of 2wn 
in ed.]. In Yerushalmi, however, jrVD is frequent, e. g. Pes. 
Ill, 30 a, &c, She., like Anan and the Geonim, employ ex- 
clusively forms of JDJ in the impf. and inf. She., No. 4 : 
mi> [mob ; 11 : mi> jn»^> ; 31 : ni> irr-oi', n!> fn^oi ; 39 : rtb jrvoi ; 
ibid. : rrb pum mm '13 npiv mb ?n^ (F jivj mm) ; 40 : irfiob 

mi»; 41 (citation from Kid. 70a) : "\b 3m5l '13 K*1B>3 "b 13n, E 
rightly ^ fnrT (F wanting, edition of Babli 3TP:i); 44 2 : 
mb fumh, rv6 rwntt; 45 s : ml? injwoi' ; 56: m$> frivol; 63: 
mnsn ?n»^» jrvDi ; ibid, end : pom '13 frvob »mn 'i3 nr6in jrin^ 
innsb- NniKn n<i> ; 113 F and RNZB.: xnuw m: ]r\*th ; 117 : 

z 2 



324 THE JEWISH QUARTERLY REVIEW 

rwpniB irpoi' ; 132 : 'la ^bb iivdi '13 vnsb jivdi ; #£/. below : 
]T\J? |nw '13 |n , 'Di', further: n^ purvh; 139 end: n^> pjrv!'; 
147 : '13 113^ }rw pDK iKl '13 SOD |ir6 JIVD^ ; 151 : irTOI 
twro$> ; 157 : nnjN jivd^ ; 168 : xpiri' snan jn s D ; Geonica, 
360 (in the Derasha to No. 43) : nnar6 [itd^ '13 nnan^ \rtxh 
(paraphrase of B. m. 69 b) ; and in my Vienna fragment 
{JQR., 1914, 420) r. t 1. 19 : |rVDi> nb nbi, »., 1. 4: nir |nxi. 

Exactly so in Anan : Sch. 3 7 iTOrVDi', 8 _ 9 ]n>nb, 4 10 ]T\^b, 
n 12 rarws, 1.1. in"D, i2 li4(9 , i3 T , h 6 , i5io,ia,ie,M. l6 is 
flTD^, I5 U [rVDI, i6 n fr«l. Also frequent in HG., e.g. ed. 
V., 5b below: 'is ^ab btr^w na? rtb tdki. 

18. Dn!?, nab instead of NDnj. In the She. only Di"6, 
xnnb: No. 73: in"»nb, orb; 77: xvrb, nr6 passim, 
Kjntn K»r6 81 ; 90: K»r6 b^ 'jm p, #£/. below: ym 
'13 xmb bv, but E: XDni'i', similarly quoted in HG., 
ed. V., 9a: tvorb bynb ■'viapi iqa ^d >:n ; 51 2 fe Nor6 
( = Ber., 42b-43 a!), F and E have here nddj in the first 
place, in the second also E HDrb ; 107 F (= HG., ed. V., 
30 b, ed. B., 146!): '13 xmb nb ^3N 'o ttiv abt; in HG., 
ed. V., 9c (from She., No. 90!) ttvrb bis; orb, warb very 
frequent in HG., also otherwise : ed. V., 6 c below, 7 a 
above, 7 b above thrice (= She. 77), 10 b bis (PDS), 10 c 
nxsn aarb, Erubin 21c, 29 d (from She., No. 74 F, see 
below). 32 In Anan orb, mrb throughout (Hark. 18, 47; 

31 =HG., 30 c. For this NJTKn tVDlb, in Ber. 38 a ("Aruk, HG., 
7 b above, and R. Hai Gaon in RSBA., adloc.) : NjntH K313 , in 1ST 'n , 16 : 
Vp~V nS ; N313 is Syr. K313 ' cake ', as rightly in Levy, II, 303 b. Anan, 
ed. Sch. n 15 : 7\b p'SK OT3 prUl, ibid., 1. 15: W, 21 K313 irO 
(annulled). To Ber. 42 a 'Ar. : »3J>3 ,| S , "1 IDnn the "Aruk has s. v. "|jn : 

Tn 'n ivd-ik pwbi rv^ao rmn ■a •oya Kim >55 -ra idd. 

Low, in Krauss, LehnwSrler, s. v. N313 , is therefore to be corrected. 
sa Against Albeck, 5>13E>Kn 'D, p. 55. ". 19- 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 325 

Sch. I4 2 7> i5 3 . io s , 1- 1-. i'26-28) # Thus the b was not, at 
that time, transformed into i in that Babylonian dialect, 
as appears also in Mandaic K»Ki"6, mzrb, likewise in Huzv. 
written, lahma (£*)), and was pronounced n&n (Quatremere, 
Journal asiat., Mars 1835, p. 256) against Htm of the 
Babyl. Talmud. The n, however, despite its being 
written so, was probably pronounced like n, as many 
another n in Babli (see Noldeke, Matid. Gram., p. 59). But 
acrh occurs also in Babli : Ber. 42 b-43 a (see above), 
Taan. 25 a bis, and Ned. 49 b passim?* Xfirb, as it were, 
was completely forced out from the Babli by the more 
frequent N»n3 ; however, not all its traces could be effaced. 
On vacillations between b and i in the Babylonian see 
above, sad 13. 

19. N»N, KDIVN, NDWN. Already in the Babli forms of 
the abbreviated KCN (for ~i»N) are employed usually in the 
impf., imper., and part, combined with to (tON), as ND»K, 
KDT1, Nrfo, tU'DN, &c. (on the Ithp. see below). But in 
Anan the short form N»N is used also in the perf. throughout 
(see above on Anan, sub i7). 34 This is the case also in 
She. in the old fragment, Cod. Oxf. d. 63 (see above), 
throughout : fol. 55 r. (No. 33) : 21 K»K tnn [ui 'otn] 36 ; 
55^.-. WvQV SdNI, N3K1 NDtH VOW Nn ; ?'fe/. (No. 34): 

n»n an Sdn ; 56 r. : 'is son 3-1 ndki, Sdk 13TW b No[ai] 
i>Ki»B> 3-1; 56 z>. (No. 35): pj)DE> un kdxi, »a^nn 3-1 kdki 
!>NW kdk 1C3K 13 ; 99 r. (No. 2) : K3tn Sotn j»p «n ; 
(No. 3) : n!> ndn per, -irybx [m] t*eta ; 99 z>. (No. 4) : 
bn -itj&n m kdki, rren^c Ninas •a-i 555, '« b prw , 3-i ndni, 

33 Sanh. 100 b, Alfasi has XDH3, comp. Rabb. ; in A. z. 35 b, Geonica, 
26, &c. in Rabb., has ND!"0, see Albeck, I.e. 

34 Once TON, Hark., 27; in the part, sing., Hark., 13: fNDl 

36 I supply in accordance with ed. Venice. 



326 THE JEWISH QUARTERLY REVIEW 

prm on SE5 ; ioo r. (No. 10) : kcn ann ndb'd nsns na VTT 
kdn pnv on 'u, #&/. jKob, kdx i&a pm !'tt "3n rr!> ndn 
on itJ» kdk 'is nb kdk, 'is kcn p nna 'ia "3K kdk 
pp5» kdn; ioo w. : ami [m]ry [p nr]yi>N una, pasa srix 

pni" 1 On KDK '13 NDN; (No. I a) 3e : $>KW KDK mm" 1 31 KDN ; 
IOI r. (No. 96) : nvin it!» KO8, lnjPD no kdk- ; (No. 97) : 

win an ndk, ndk nww on, 'tw an ndk, 'is kdk mirp an 
'la ndk jpN \xd; 101 w. : xdn K[j,na an], yan« an £55, kdk 
'la pro an, '13 jua n^ kdk 'ia «as n^ kdk. The Geonim 
likewise, Geonica, 80 below : Kaxn NDN 37 and even ndn 
r\b = rtbuti (see on Anan, sub 1 7). [Now in addition in 
the nianiOT naD of R. Samuel b. Hofni, Hoffmann 
Festschrift, Hebr. div., p. 86 7 : Nnvoc tab rbnx Up, 1. 
12, 14: rrboK, p. 87 34 : ini»»k = mi? nos, 38 : ,tS>dk, also 

P* "990, 91, 92, 95, 114, 116, 117 (l!2 : '°&V> P- 9°J25, 132, 143, 

i« 146-] Eshkol ha-Kofer, No. 139 from niplDS nobn: 
nni *6 Wl KTiyTN nBTh wn; 6Vakr i?. Amram, ed. 
Frumkin, I, 376: a^l riDN kdn 'K. Also ND'N inst. of 
kdk (= npN part.), therefore something like kok, ndk 37 »? 

She., No. 48, ed. V. : 'dw W '13 n^ nd'k 'id t3 ndv *n 
'13 1^; No. 72: 'id •'DJ 5>"n ^ KDW ; HG., ed. B., 159: 
KTI '' XDty n Q3 '3D "b NDiN (ed. V., 31 d : ^D "]i> nDNn) ; and 
even "jbo'K, Geonica, io3 12 : n^D'N K3Nn (see on Anan, /. <;.) ! 
Then ND" (= no") and ND'D (= no'D) : She. fragm., Geonica, 
357 6 , 363^ ND"; No. 1 E : Nin xnatn kd^d. Likewise in 
the name of the Amora nD" : No. 92, ed. V., nDN nd" an 
(likewise E, but F has nD") ; No. 170 E (F wanting) : 
\&xb& na 55" an bis ; in HG., ed. B., 350 : TTilia 55" ;kd ; 

36 Also in E and F No. 12 is found between 10 and 11, before 11. 

37 HG., ed. B., 32: mb WDN1, r. NDK1 ; but JMDK here = WDK, 
p. 92 &/s and 159. 

37 a See below, sub 87 (a). 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 327 

378: nwis ma rnxn nd" ^ ; no^o HG., ed. B., 131 : 
»an mi> NtFoT. 



In the Ithp. NOIVN and NOVVN — the latter, as is well 
known, already in the Babli,'but (except Dalman, Wbrter- 
buch) misunderstood — She., No. 20: nox NOWNn B>"n 
win an (=norVNn); 138: '12 mm ids win an nowki 
(=nD)VNi); 141: 'w na nan nox 'M noivni (F 'orvxi, 
E norvxi) ; No. 24 : mim an 'orvNi "ax nox (= No. 68 
NOTTW) ; ibid. : mim an 'orPKl tuna an 'on (68 and 92 
soivto) ; 55 : ass an noivni wan not< (= noww) = 68 4 
nss an 'onw wan 'ota ; 58 : pro an 'orvNi mew an am , 
and 170: no" an 'orvxi bxjw 'na xan 'ox; HG., ed. B., 
191 -. snx an xorvto "as ni?j? b«$> ; HG., B., 463, Konw ; 
comp. also MS. Oxf. d. 48 (2760) fol. 26 r. [now in Jahrbuch 
XII., p. in, 1. 1]: nanjjoa (= norvw) noTiKi. From all 
this it is evident beyond any doubt that the Babyl. 
kown = "Vy* (nos'nx) = nomx, and is not composed of 
1VK and KOK, as Levy wants it (I, 72 a), or of KD'n •>« 'and 
if you say ', as Margolis {Manual, p. 88*) explains it. This 
form was misunderstood already by the Palestinian (as 
rightly suggested by Poznanski, R&J. 63, 235 [see also 
now Jahrb., ibid., p. 99], translator of the niplDS niai>n, the 
author of the ltd )YD?n ed. Schlossberg, who likewise renders 
NOWXI by noin pto (p. 9) and not*n Dm (pp. 106, 107). As to 
<J&B K0TPN1 ^i»a nos comp. the frequent ^bsn nwo norvx 
■oi^an nwo norvto. 

20. The language of the ' redactor ' of the Sheeltot was 
likewise Aramaic (see above). His formulae are : 'OJ in 
in connecting the single ' questions ' (Dna) with one another 
(very often), see esp. No. 10 : Kineo ni?l3a xna^n poao K: >tt 
ana ^no, i.e. that the words 'iai bvm NWa nhn wni at 
the beginning of this number existed also in the form of a 



328 THE JEWISH QUARTERLY REVIEW 

Sheelta with D"ia and yoc xn, which is borne out also by 
SOD"0^ at the end of that decision, this term being usually 
placed after Jjttt? Nn (see above, sub 8). No. 142 : H D3 ''N 
■<-)po Kama p^doi norvNi 'ia wpb B>n -ion »an nd^neh k-ib»» 
fWO fjnjn nawtn "pim (F and E) ; No. 82 E, F, and 
RNZB. connects with 81 through the words: ;va 'Di ^ 
bifirb) twm natro nw aire i^bx rvya-iD nnv pnoxi, i.e. to 
place after the words (pnotn jva) T3E>D '13 i^sn in No. 81 : 
'13 ^nar^l tfom; No. 58 (E, F, and RNZB.): VM •>« 
rvihvi mS>am ro^n niDvnn ne^aa nixi^m nunin p^Bnaona 
'ia s<a-na 1TID10 'j p»jm ntam mbna w »a»n pnoio, i.e. 
after the words (p'tmaona) : ppiDio n^taa '13 nuain at the 
beginning of the number, to the place : 'ia mWj W 'aTi ; 
No. 91 after pj>DB> '"13 KnaS>rn E, F, and RNZB (see ed. 
Wilna) have the following: la soti (Wil. 'ti, F /- i) E>Vt 
x"a nM 'n jopi now enna x'jni xnWn KDB>aa (nax) xax 
'ia trnn 1^X, see ed. Wilna, 38 No. 88 : Kwbx mDpIN <DJ ^ 
i.e. as was already remarked by Reifmann: 'DJ 'X, which 
follows here, is to be placed after XW"b (XB^a, ed. V., 
35 c) ; No. 103 before the second T"IX Dia E and F have : 
no<D oyopn $>y d^vu -cmrb nna (F marg. ntm nw *x) nw , x 
no ivncn (E erroneously Km) xa s m (F in addition xnox xp) 
(F piDDi) pd»oi D'anp 't Mrfo idid mm xpi>Dn nyi xho wn 
WXa, an order of succession which HG. HNOID, ed. V., 
44 b ff. had before it ; No. 50 : 'la ne> 'a 'an x^jn'o ^o: 'x 
(comp. No. 107 !). As shown by the expressions |Va 
pnetn (81), *lo*D (103), these redactional glosses still belong 
to a time when the She. were ' spoken ', preached. 

Mention must be made here also of two other glosses 
by copyists which are likewise old: Geonica, 361 (in the 

38 What is still to be said on this does not belong to the sphere of 
lexicography. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 329 

Derasha to No. 43): an i»n nb >a*fl wnra na nral> nun 
nisi? nun nrsDn by) nunn S>y pna»B> n^vs nt^sna htisk pru, 
i.e. 'let 'ia DWa ?»na an n»N be heard (nix) previously', 
before nix5> nun, which is accordingly stricken out. Like- 
wise No. 45 2 : 'ia wjd na ejov an n»N Nnp'ya, i.e. (see 
r6st2> povn, note 'n) the words of R. Joseph are to be placed 
anterior to the words of R. Johanan, as we actually find it 
in MSS. E and F. 



And now I proceed to the lexical remarks, which are 
ranged according to the order of succession of the She. : 
21. No. 1 : r\h naji r\h rvaxo nai wro <:an twto 



•■na i>iSn wk notna kov nn nWvi Tay n<nTa<yi>; noxc, 
E: Tay n^ nOJl novo nai, F has n»J1 supplied above the 
line, nax = 'pn (i) ' fit out, prepare ' as here and frequently 
in the Babli (Levy, IV, 168 a) ; in She. also No. 44 s : nax 
KXOlK iT^ (Temurah 8b another reading 38 *), and No. 133 : JIM 
nonx n^ pnaxn paxan xuann (E pnuxn) ; in Gaonic literature : 
HG., ed. V., md (B., 457): xnaxo 'ya mm 'no r\h 'h» 
r\h, and Geonica, 325, XLV: iwnb tftw ;naxon = payn 
z'foi/., XLI. (a) ' mend, repair ', in this sense only in 
the Geonim: She., No. 170 E: n'DN nanw tcvbpx W 
Kiyion xb)ra rrniaxb (ed. ruipn$>) ; HG., ed. V., 23 a : sotii 
mniaxi' <ui> no'D^ ntr 'ia twaw nat? anyo isora wsn, ed. 
B., 102 with a marginal gloss in Arabic : ni?»y\ sjLm 'to 
mend it'; HG., ed. V., 29 d: 5riaxn5i xnan'on aa by «]Kl 
'ia tnn »ba won t& (= Nnajpw), ed. B., 143 : nraxoi snanon 
(the Ed.'s work Die Vat. Handschr., p. 21 has Knaxei) with 
the gloss : asysB' ''3 39 , Z£L , reparavit (fracta vasa) . 

38 a KXD1K n'b naS cites from Tem. 8 b, also Kafthor Wa-Pherach XVI, 
ed. Berlin, fol. 67 b. 

39 The Arab, glosses in HG., ed. B., are probably derived from the 



33° THE JEWISH QUARTERLY REVIEW 

Syr. rav, ornavit, adornavit (ndtd toy, SnrnxD asb^, 
xnrmo *6 xnropno 16 xyix, Payne-Smith 3360), also so'nia 
xnrmo (PSm. : ' Tunica acupicta'?), Arab. &£>, reparavit 
(vestem), see in addition sub 42. 

Instead of 713 W>Vi we find in Tanhuma and var. in F : 
h^a, W>3 ; but since N^Ti Tay precedes it, ^Ti would be 
more correct. nWti ' feast ', in Talmud ' feasting, wedding ' 
(Levy, I, 47a b), see esp. Sanh. 105 a : TOV N"Wl Kncwa 
kij B»an «3"ino (Ar. sao sov) tbbn, and Git. 57a: 'Wti 
"ami; Hebr. B*W>n Jud. 9, 37; Mand. nWvi, Syr. xW>n 
' feast, wedding ', PSm 1370 (comp. Noldeke, Mand. Gram., 
p. 118, and Ges.-Buhl, s.v. thbr\). The WN niDN is not 
known otherwise, hence it is taken from life. 

22. Ibid. : '13 p'HBN NnjJDtn KJpDDI ; 12: NDyCKH NJpDOl 
'la -|"N; NJpD», n3pD» also in No. 13 F (topDOl rri> mo'Di? 
«1DV ana), 18, 33 Oxf. d. 63 (fol. 55 v. ■. runjo snaSn ropDoi 
. . . JVD i^bni fJ'p xanv [nvpD3 inn 3W, abbreviated in our 
She.), 113 end: 'ia an -idn njpdidi; 95: n:i'' V'x jrpoEi ; 
97: 'an vtaa |rpi>D»i; 138: sna^n jrpDo; Geonica, 103: 
WD j>DK> x^N f^p^Di "W «03 jwetn. 40 In Babli the noun 
NJPjap only in the sense of 'result', Meg. 14b: 'on mpDDl 
wn (Levy, III, 176 b); the verb, however, in the Kal : 
B. kamma 92 a: xnairn xa^N Nnyot5> n^ p^D nin n^>, in the 
Aph. : Suk. 39 a: *WlSo 'P.IEK; Anan, ed. Hark., 39: p^KDl 

thru Y^H of R. Hefez b. Yazliah (see Halper, JQR., 19 14, p. 560). 
From this— and not from the JWIB>i>N 3X713 as Halper, I.e., thinks— is also 
derived the mDpn DWDD DKS^K TDBn, for that Baraita mBpfl DIB'3 
is found incorporated in the ed. B., 636 ! 

i0 No. 91, RNZB. (see ed. Wilna), F and E : Itf B»"li> rfb fp3^n 

nnun p<ddi, g«o«*v«, 154 : (nTijn xpi>D=) rvnjn J>» rrb apDC "b man 

'13 'D'Sm ; in our redactor (see above sub 20), No. 103 : '13 XpPDT TO) 
nt?N3 (piDDI) pDiDI, No. 142 : WTO p'DBl 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 33I 

'and concludes, ends'; Responsa ntUJ men, No. 161 : np1 
fP^DDl finals x-iU'S. In HG., and likewise in the later 
literature, as is known very frequently, KJpDC, JJ'pDD. 

23. wm. No. 5 : 'ia ktuvk xyd? Sjwis mb rrh jkoi 
frona NpjniR najrei n-iin isd wat ; *%£ : rvi> n^n paa W "on 
ktwk xynb in twnwa pDjre^ in aba SprtiS (E : kpwik rvh 
aDj^i Nnnixa pDyo!?, Oxf. c. 16, fol. ia z>. : «p:m rrb rvh 
3d:'dS NJvnixa pdvo^); No. 14: Npaw najwi w jo ^pt^o 
KTBDn!? (E and F m»Dtri>) = HG., nsDn , ed. V., 142 c below : 
nfyn D^ivo xpmx |J'$>PB> (r. xp:"tis) : further HG., ed. V., 
67 b, below : rvpym iypen Kin wn ; HG., ed. B., 209 in 
'cn^': pi?31N lDSn DlEtai KpJIlK D1B>»1; Geonica, 213 (from 
Basra to R. Moses Gaon) : NprriN nby (r. topajfioi) nspwnoi ; 
ibid. : jaism nna j» kpjtik toya «pi tnaycn mot* pjjdb> nai ; 
in the ' Gaonic Decrees, &c.' {JQR., IV (1913), 29), IX, 6: 
wb pni? ran wk nna wiBtn i»n^ sanoias pni? •oo^i 
Nam SprtiN |ini> paarrai prra-iva nwpi ; ibid. 14 : Sp?w ipiasn 
\\nb paDT no twwa MTV wb (a monthly salary) ; in a 
responsum by R. Zemah, JPTaon, No. 43: |ariN rr<b mew 
snaiii? jvddi rppen "ina$> 'ia waf> sprtis jinn 11 am "tit nso 
li> }rpoi> rri>jj jynaoon pwn (r. wao) wao sman ^ nm 
tnaiy mn "ani '13 mi>a p an' x^> k-ids? rr$> nanai Splw (r. p) 
'ia sprns ini? jnn^ "by (r. rnva) 'wa p^Dtn (means of living, 
expenses).— Kp:TiK therefore means here not only ' costs, 
expenses ', but also ' wealth, money ' (as in No. 5 : rv6 rcbn 
aDJ'cS wvniN3 puynb KpWN), and ' salary '. In Babli only 
in the sense of 'costs' (Levy, I, 52 b) ; in Targ. Est. 7, 4: 
xa^on NprriNa = Npjrwa 'in the treasury', Pehlewi pwix, 
sk.jy*, Lagarde, Hagiographa, p. 366, and Abkandlungen, 
P- 2 5> n - 3- A similar change of meaning is noticeable also 
in the Old Persian ya^a, kh 'treasure', which later on 
assumed the meaning pecunia, divitiae, census, xprnxara, rt/xia, 



332 THE JEWISH QUARTERLY REVIEW 

and <f>6poi, comp. Lagarde, Abhandlungen, p. 195. In our 
case we must also assume the change to Kprm ' means of 
living, daily needs '. KpJW Taan. 23 b and Men. 69 b, 
according to the Aruk, Syr. XW\ (Levy, IV, 431 b), Pers. 
Ujjj, Old Pers. ekjj ; , whence Arab. J^ and its derivatives 
(Lagarde, op. cit., p. 81 ; Fleischer in Levy's Wbrterbuch, 
P- 486). 

24. No. 9 : ibm wvro bp ynnwtn, similarly below. 
In Babli only the Pael T.1? 'to pipe, chirp' (Levy, IV, 
185 b); here noun, 'the piping, crying', Syr. NJS1X pipitiis, 
PSm. 3382. 

25. No. 10: .won "panBtn km bix, F: to»n bx 
rrtMon wtvki, Oxf. d. 6% fol. 100 f.: — n wen Tii-rvFn 
— but Nri33D wx ttbi inof» — n wen Tiani^ wn (agrees 
with the reading of the HG., see below) ; No. 37 : "ira ^3N 
'i3 iTjdod -maw wen nwo Swictn, E and RNZB. (Ill, 
fol. 100 a): wen(!) "jiane' , 'Kn nntj»e (RNZB.: w iibtin 
won TantrN) ; both MSS. are corrupt here. HG., ed. V., 
23 d has the correct reading : ;U3 bvie'W "iron pTBoe bin 
'xs nna'tn in wen vrwvin (B. 105: ■pan^xn tovi). iwn 
i.pl- 11WK) = TBnB"N, from ItW, NOT WK, Shab. 156 a, 
"IBWIK, TBWlK, Onk. on Num. 35. 33; comp. also HG., 
ed. V, 128 c: twi NT133 N-int5»en prvpa rustM (ed. B., 526: 
Ninwr), as if formed from K1BM Different Geonica, 320 u : 
xn^e ni> twines = mntw, from tat? ' to throw, cast away ', 
somehow in the sense of yitan NH k^ik , in Yerush. often. 

26. No. 13: nvtaaob nb avr n^ ton n^n to3 ah |vai 
D«T3l 3^>n3, E and F: t>ieDei>; No. 154: 0^33 WDCDee 
3^m , F : tHdEdd . This is the correct reading, fr. DDD = DDDD, 
Syr. 'to cure', 'nurse', DD01D, NODD1D, 'healing, nursing', 
PSm. 2653-4. Also used by the Geonim, Hark., No. 338, 
p. 169: (Yeb. 42 b KDeDee) xDDDon it?ss D'ean next? nei 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 333 

t6t n& didod n^vni 3^>n n^aso wiDBOt? '« abm ura rvb 
'13 It DlDDD l^BM '13 "D5PO (see Low, note, *Wrf., p. 369) ; 
pal. Shek., V a, 48 d : Do5 -ion p «m K*5to!> 3D inn p "n 
K*BD^, var. D.ODD 'venomous* (comp., however, Low in 
Krauss, 398). 

27. No. 15 : «nt "U ktim i^bn (ntrci>) xo^jj aina wtkh 
(nttvi> is gloss, wanting in E), F : Wo ntS KD^U JVU tOTKn 
Silt 12 tO"iX3 ntS; Syr. KIPS, PSm. no ; in this manner 
the Syrians explain also K"px by j\y>\, Sindon, PSm. 3367, 
comp. Schemel, Die Kleidung, &c, p. 33, note. N"i"i¥ as 
Aruk, s. v. is the correct reading. R. Solomon ben 
ha-Yatom on M. k. 27 b, p. 127 : "133, «m 13, KTfiO 
(N"nsa=), cn nH hs yidvti, ttx wn p^a tnp« ay 

DH33 n310 (Wim=) WTIX 1T3n (51b) KJWO K331; She., 
No. 113: NlVTiM; Oxf.c. 18: wvmxa 41 ; HG.,ed.V., 94 b: 
snKVnxa, ed. B., 378 and Aruk: tMimva ; Assyr. suduru, 
syn. of el-lu-ku (= p^>n), Delitzsch, Handworterbuch, 564 a, 
' a cloth '. 

28. No. 20 F : N'^bi kodu mm arroi* »wi Ojrn is 
cited also by Asher, Decisions, B. k., VI, 5), to K'Woi the 
margin has: KJK3 pD» '•3. KajapDD might be 'torturer', 
' one who inflicts wounds ', which does not fit here ; nor is 
it in place to emend it to K3N"lpDO ('driver of cattle', as 
Pes. 153 a: '13 TpDD Kim NDXDO {HiTJ N03 ^ TpDD, see 
Levy III, 163 a and 580 a, and Low in Krauss, s. v.), for 
the simile does not explain anything. In accordance with 
b. B. m. 93 b: mo >xro Nmw aniTai ib noa, I suggest 
WK3PM3* 'guardian of a tower', Pers. eli^, Arab, jlj^, 
Talm. Kptrp (Shab. n a Aruk), Targ. Nptrip, Syr. xptyia, 



41 Over the line in old Greek script iiiTairparas = peTcmpaTip, ' one who 
sells again ' (so also 'Aruk) ; there are more such Greek glosses here. 



334 THE JEWISH QUARTERLY REVIEW 

K3K3 = J6, ('master'), ' guardian ' (comp. KJ3T7, 'guardian 
of a door '). 

Ibid. r. with E : Kin -isn 'i3 ytrb sm 'iwin jw isra bax 
rr!>enta (ed. n^ nsnKn), Af. from W, Sb> as Git. 77 b. 

29. sr6 met). No. 21 : 'K03 i!»bk nn '•sni' two ^ ^"k 
^ KITO &6 MfDl, E and RNZB. like ed. Ket. 53 a: wo w 
nn. But the reading of the ed. V. is certainly correct. 
In Gaonic literature also the following: HG., ed. V., KTPp, 
13 b: dib>o p"bnoi 3ie>n dik K3W to»n noun m S>e> tiki 
K-nn:S> ins vbi 21 bv *)k (ed. B., 47, .tti^oon pp^noi) 42 rwvb 
JJOiao K^ ; HG., ed. V., 31D DV, 36 b : 'K '13 K"lDtt 101 lono 
jrnio n^ nnb nie>o xi> w «ni> $> man 'una ii> kito, ed. B., 
182: "pH NrtaooK (!)b' l 3e>3 m6 dki (pleonasm), in She., 107 
after li>W3 n3T KOB> F has : xb w ... K-ioir toi "rono 
'13 Tin Kn^lOOK (in b. Bez. 21 a another reading). Thus 
in both passages of the HG., ed. V., and in that of the 
She., No. 21 xnb me>0 corresponds to unroot* 'because of. 
It is a translation — let us say a vulgar translation, a popular 
etymology — of KrtaoOK divided into KninDOK, in which 
1D0K = D1C0 ! And so it is to be explained also in the 
Babli : Yeb. 29 a : .Tin mb D1ET0 nninx rrvn, Rashi explains 
.Tin i>ro Die>o, but is compelled to add tnn pe6 w xnb wn ! 
B. m. 20 b : "i»i ypa nn »Kr6 dik>o tok no 3n m khk ypa 
J?PS nn ^ni? Dlt^o TDK, Rashi also here ^to ^3B»3, while 
it is self-evident that it means 'for my sake', 'in behalf 
of my honour (because of injury to my reputation) ' ; 

42 The passage must have experienced the following changes among later 
writers : Giyat, I, 17 cites : mbn 01570, MV., No. 151, p. 117 : ND11X Ol&B, 
Siddur Rashi, 263, No. 126: ."W DWD ! [For that in my Fragment, 
Jakrbuch, XII, ir8, 23 : linaW pp^CS?,] likewise for HG., ed. V., 36 b, 

we find im 'n, 8: pmiB UK ["K 1^3B>3 1k!> DK1. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 335 



similarly ibid., 106 a: ITT -\rb two. ~[xrb two, i*rh> two 
nn 'for thy sake', 'for my sake' (Rashi i?K>, 1T\b). 
Thus alongside with hoDK, on blDDN, Nni>1t2DN, we find in 
the Babli also a vulgar ttrb DltPD ! Levy's attempt to explain 
it (s.v. aro, II, 531 a) was reversed entirely by Fleischer, 

P- 536. 

boon in She., No. 69 : "on Sddk ; 7a: jne>yi htDON; and 
13 E : "on Sddn "on pnn id (V. and F : wax). 

30. 7fe£ E : kwd '^asi WBii> h'd *w vb ion i-pti 



NTU3. The latter is wanting in the edit. Ket. 53 a; how- 
ever Erub. 53 a: Nicab KTM NWD o J3S1 ; Rashi : psyw "WV3 
prim d:321 is apw !>n«3 WIN, an explanation which does 
not fit the context in this passage. KTU3 NrWD ('a plug 
in the hedge '), in accordance with R. Han. in Erub. : ''B 
TliD nS nbvo n^ law iriiaa nnvi (quoted in Aruk, j. z>. 
JBVK), probably means ' something of little value, something 
insignificant '. 

31. Ibid.-. '13 aim Nin b »kdk mpinm nntini, E wants 
^kdn niphni nnnm = edit. Ket. 68 b, F has i"d nipSni nnnn 
'13 i>3. But 'KDN is not a scribal error, it is the usual 
introduction of a question in this dialect, frequent in Anan 
(see I, sub 15) and in the above-named tractates (see /. c.) ; 
in She. also No. 23 : '13 WEnon JV3 "NDN1 (Ned. 8 a another 
reading). 

32. No. 23 : KTlJ '"iW bwsrm Nim p*3, read with E 
and F i>WD, from W. 

33. No. 24: xhca Ninn nnna^ n^ f:nyt2>D, and below: 
hdb> nnrnb ; further 48 F : Npsa xi> r6i3T «rp3 in n^D p'aa 
'13 p"s:n terra nm!> n^dt jkos 'is (ed. nnnn) nnrni»; 
talmudic nna (properly Hna, N*in3 'in one', Bab. rnra, 
Papyri mn, see my ' Weitere Glossen' in ZAW. for 



336 THE JEWISH QUARTERLY REVIEW 

1913, p. 139), 'together with', 'while, near to', Hn7 
' opposite ', but also HH37 Zeb. 39 a : uniri 'nrab »Tit3B arhn 
uvn 'together with'. As to the form comp. Pal. Tin73 
and *pin737 (Targ. Yer. to Exod. 18. 14), <mn737 (Targ. on 
Job 1. 19), similarly biblical rWK*i3D7, Talm. rb^nrob, 
y"ia»7, 11D37, and like forms in Phoenician, as ^nao?, 
iDnoi07, n-i'a»7, &c. She., No. 77 F: nbt n«s«n unam 
"•Bin paj N7i win!? njivao ri7 (r. 'jrcana) 'jiana m-\tt »a 7»ea 
WD ' immediately ' (without effort) ; Bab. K1D7 ' directly ' 
(opposite), anton. KD^K ; ' straightway, in a direct line, at 
once' (see Levy, s. v.); Mand. tftnrb, KWrr?, rarely 
NWW7 = Syr. and Talm. vrnrb, solus (Noldeke, Mand. 
Gram., p. 141). 

Ibid. •. wwop w (n)tii utr nupup S7N mm moa «7, 
E KrWDp VTOTi but the D is scratched out, F liTO s ni 
Nn^jpup, Aruk, s. z>. ni7pip3jf cites: Snipijp inyo vii. 

34. No. 36 : on 'oi naum ri3B>, E pan on nw (F pan to), 
RNZB. : pan En <oa mum n"m na*^ pan on 'dj roum na^ 
'•a t6; No. 29: 'ia n<7 >jjd na 'an En naew ?tsi, E En 
'WD ■on, F 'jrao pan "jn ; No. 30: 'an "En wn u; 50: on 
'ia niara pan ; 75 E : ton na 'an En rtaaa TDtn oti *ai 
nturn ; 80 E : nwna "iidk »e: 'an En n^asa nb -posa w 01 . 
Dn etiam, see my remark on Dn in Jahrb. d.jiid. lit. Ges., 

IX, 236 and 25a (and above on Anan, sub 32), on jn 

X, 380. 

35. No. 35: v&tt wan ny "idini, E 'cni, F how ; 
nevertheless "iDlifl in the ed. V. is not a scribal error. It is 
imper. sing. = "nCK, IIDN (see above j«# 7), and this form 
occurs frequently : [inst. of imper. KD"K : Gaon. Kamm. to 
Kel. XXVI, 1. s.v. D'31 MS., cites from b. Sab. 91a: 
(ed. ND'N) 1T7 JVtn (-ID1K =) 'ciS ''S ; b. Ker. fragment, ed. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 337 

Schechter, 27 a: iffl^ N^^EiS, 37 b: ntaro 'din; Rabiah 
ed. Aptowitzer I, No. 193 (p. 221, see n. 21); DUjp 1D1S1 
(comp. Zeb. 18 b: NOW);] inst. of "WK, TD'n: Geonica, 
3i8 21 : nty^K 'i *i»sn (ic'N =) idin tdnpi ; #&/., 319 : 
'dkt noiN p'lDNpT wn [and also 1ST 'n, which on p. 97 
has 13 norac n»lN for p3"i nc*n "WK (Sheb. 42 a M. no^), 
had no doubt in its text of the mplDS ni3$>n the words 
JJnom TDlx (Responsa ed. Hark. 33: pnotn 11DS), for 
otherwise it renders iota "ilD'N by "iokb> "I1DK, p. 130 (see 
above sub 7 and note 14); in addition also R. Nissim in 
his nnso, Shab. 132 b (ed. Wien 50 a): ne>j; Vik 'nDKT "iion 
(edit. mow), on the side of p'TDtn "llD'K, from Yeb. 3 b. (edit, 
now); HG. ed. B. 430: n(j)DanK mv "iidk (edit. B. b. 
174 a ID^s). The form "1»1K as imperative corresponds to 
the imp. with pronominal suffix in Mandaic (Noldeke 
271 ff.): "taip, likewise HG. B. 564: rP3mn. However, 
that KO'K, lO'N (110'K) in those expressions (. . . Ty now 
and the like) are imperatives is evident from the corre- 
sponding phrases in the Yerushalmi, which exhibit Tidk, 
as p. Sota 22 b: J3 ronton o-njn s|t* (nowst) "iioki, Ket. 
32a: '13 N3n epN 110K1, and B. k. 4d: '131 nnan ^>3 »ai 
nb'ssn nrw: nono "ilDKl. Indeed, already in Sifra xnp'i, 
Nedaba, Par. 2, 3 we find : (!n»w =) "iiok ,p loi^ rVN") noi 
DnDlcn T\\l~b Dis]. Otherwise Anan, ed. Hark., 119, No. 
57 : ptnra fiJ'M ntw bn D'pn "3 nran cpn idin on. The 
end is preserved in another MS., on p. 170 : TDtn nvi 
TON N^> pe^TM. Hark, translates on p. 119: -mm 103 '3 
(nhtt^), 43 in the Index, p. 121 : low (?"il»K ?-id*s =) idin, 
on p. 170: '13 newt? Kin ptntwa . . . k> "idnje> no »a. But 

48 This is followed by Ginzberg, Geonica, p. 318, and in the Index, 
p. 411. 

VOL. XII. A a 



338 THE JEWISH QUARTERLY REVIEW 

it should be translated as follows : p«ltM3 '13 totes? no *a 
-id1nb> sin, ion, idxt on p. 170=10^1, icn, part.; tdk, 
however, is in Anan SON, see above, ,y«# 19. [iDlK, 
p. 119, is either a Hebraism ("iDiN) for "ipx, as p. 170 (as 
to "IDIN *a, comp. the Babylonian "iDNp »a) or else it is 
Aramaic written in accordance with the vulgar pronun- 
ciation of the long a like u (see J. N. Epstein in the 
Schwarz Festschrift, p. 321), to which may perhaps be 
referred also Responsa ed. Cassel, No. 53 (fol. 12a below) : 
Nnxnaw srora imtn <nq ;d pany 'an *aa pam Nna^m (we 
might have expected here ton, comp. No. 91 fol. 32 a in 
the middle: KTCK). On the other hand, Ned. 51 a: 
i? -\n\tn ib Tn'j; (previously, *'&rf., lb -id'so ^ T,pn dip, but 
MS. M. has also here "iDitn !) is probably a Hebraism (nok) 
for "TOW, impf.] 

/£*#. end: n'iD rrb tna"VNl, denom. of can 'become 
filthy, dirty'. E however B"BBW, F D'llBNl, which is 
probably D'Yinxi, see below sub 61. 

36. No. 36 : waits Nil, r. with E and F SO^BIB, Ka^BlB in 
Babli; Anan, ed. Hark., 10 xas'BlD.— F here M«BU3 pni 
inst. of xn, see above sub 13. 

37. No. 38: n»W r6 poiBi, E pi«^b6, F N^civ 

38. No. 40: NriD <D1tn (= No. 134 2 ). Likewise 'Dltn in 
the Commentary MS. Berlin oct. 333, fol. 311 a (to No. 134). 
This form, as also Geonica, 372 : T"W nym (=nojn, Nid. 
65 b: non), supports my suggestion (Jahrbuch, IX, 246) 
that Bab. DTD (NDH mn DTD) = DSTD, in the Aram, formu- 
laries for documents \ibv±>\ DlTVD^ (Jakrb., I. c, and X, 379), 
from Djn in the sense of ' be in evil repute ', ' declare dis- 
reputable '. — An explanation cited by the above-mentioned 
Comm. MS. d eserves mention : UTN DD1K ptvta mW w d""1 
aa-iin tmiM yot? «in VnaDi (Isa. 33. 15) dot wdvb 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 339 

pmo (Mak. 24 a). This is Arab. ^S (fr. p), contemptibilis, 
vilis, from which, however, our 'on is to be kept apart, 

see /. c. 

39. No. 41 : ewdt Knpno iwi 3WD^ p^o!? n^ »jjano 

'13 on^on Knpno, E : Kpnne 'is "SpnvTo '13 (!) Kruno 3D^oi> 
(where, however, the left foot of the n and p is erased) ; 
F: KP3110 3*D3 «h '13 (!)Nri3n» (from apy\-K> resulted 
Nrono = Kruno). Correctly in the parallel passage No. 153 : 
spnno Toy 1 ? np*i> xb) Drvon Kpnno awnsi p^o^ n^ »j»n*o 
sni^Ds nn rvK-i, and No. 135: rrntn spnn k^k ?yi nn* k^> 
r6 nn' (E and F Kpn !). Bab. Kpnm ' family', B. m. 59 b, 
Zend dutan, Huzv. inn (Perles, Levy, I, 395 a), corre- 
sponding to the Gaonic form. Here no doubt belongs also 
the puzzling iprai of Seder Olam Zuta : "pri won jne> cilD3i 
'13 p «int? 11m tuirntwi nnm urn »prm, Neubauer, II, 73= 
1DV vntf djm in the Hebrew translation, ibid., 76, hence 
something like 'relatives, kith and kin' 43a . 

Ibid. : i? ami, see above sub 17. 

Ibid. (= Kid. 70 a) : is 'is kjtqj? Kpn Kin u«nai Kmia 
|33T netna mrno, E and the Comm. Cod. Berlin, fol. 76 a: 



Kon-mn, F ttpTuun (Aruk twrmau), edit, twvruu. twruu, 
Krmju is the right reading, a dimin. of tmJU = tmu (Syr., 
Arab. jjL)=KYi:i u (B. Talm., B. b. a a : tnu— rwno !), ' fence', 
as Syr. KDnCK, fr. "lD'tt 'sheep', Talm. KrfiDB, fr. tnBB 
' fatted cattle ', Mand. trwsn N'Tisn ' fatted oxen ' (Noldeke, 
Mand. Gram., 315),* 5 Hebr. ma 'suet', Bab. NITS, Nid. 

43a [So already Bruit, Jahrbiicher, II, 106 n.] 

44 Noldeke, Neusyr. Gramm., 28. SOW in She., Nos. 49 and 145. 

45 Comp. in addition the reading of R. Hai (Responsa, ed. Hark., No. 46, 
p. 22) in 'Az. 11b: TTKB nO'JTI '•JUS KrWipK nnlD N^K3T iKOIK 
(R. Hananel orf/oc. : T1K3 nOW '3133 13 'pinO, in MS. quoted by Kohut, 
5. v. WVapK ITIXS) and his explanation to TIKS nD'J?. 

a a 2 



340 THE JEWISH QUARTERLY REVIEW 

27 a, Aruk, * dry filth of small cattle ' (Aruk in the name 
of the Geonim ; sma JVDTtf pB>b), comp. pno ibiy and npan 
(n. pr. and Mishnic). The J is frequently inserted in 
Aramaic, as e.g. in the word before us TIM = Tu, Taan. 
23 b: "by ntimd Np, Hebr. "rain, Ned. 81 a: by Trav xta 
"imn 'to pride oneself, to boast*. In Anan — I may be 
permitted to add it here — in the same sense: "WO, [mJlJ'tf, 
Anan, ed. Sch., 27^ f. : 'vn ntwa )b n»nn «i» <|>«]B»i> inn kdki 

iTa npDo [na kbit]' ton n>S> [m]i:nt6 "VDtn ntwa ii> nnn sb 
nyo rb?v tc[p s]nnn n^j? rmro <tn -piD:it6 'ia ; #«/. 28 2 : 
. . . aviD 1T3 p»DD na n^j> *won . . . ws tow 'to i^ "into Np 
' to boast '. 46 

/££/.: "10 w '^ neN, E and F fan, see, however, on 
Anan, sub 17. 

40. No. 43 : toanoi "an <ai 'ia nnan^ Snan inion jnd 
xair 1M 'an tniD'ta D^p. Instead of snan read with F 
Snar, as below in F : TwiaF »m by wm dhjti (E twiair, 
ed. tOD'ty) and at the end of the No. frequently Wiar\ nnsf 
(F here likewise WISP) ; Anan, ed. Hark., 10, 12: Knar, 
M. k. 28 b: sna" 1 !', Syr. Wiar. The Nna , T at the beginning 
of the No. in ed. V. and E is therefore a scribal error, 
although the Aramaic papyri have ttfiat, papyri Sachau, 56, 
11. 5-6 (see my ' Glossen ' in ZA W. for 191 2). 

Kair 'the debtor' and Kjajio 'the creditor' also below 
in this No. and above No. 31 frequently, likewise HG., 
ed. V., 98 a (ed. B., 380), Syr. vm>, Targ. Nap. 

" For the Hebr. 13 TWinb DlpO lb "WOn in Sherira's Epistle, ed. 
Lewin,— in print- 67, Span. nnjPt6 (French nvUt6), 68, nn3r»6 
' to force oneself to the front' (make oneself deserving), as the Arab, ~±L 
dignus fuit.—jSZs. from .^jXc (see Dozy, II, 229 a) has nothing in common 
with it. Hence Ben Jehuda's new formation (Millon, s. v.) : "Hi? with its 
derivatives n'rjlM, &c, in the sense of 'coquet', 'coquette', is not to be 
recommended. Is not the HB^riD good enough ? 



POST-TALMUDIOARAMAIC LEXICOGRAPHY — EPSTEIN 34 1 

41. No. 44: D2r6 it!> wen ny, Geonica, 362: nnon, 
F nnrvm ; further: A nnt6 mnt& in* xfo, F "in ns!*, 
Geonica, 363: n«i?nt6; below, also in the Ed.: rrnnt6 mo 
(F M lTnt6, Geonica, 363 rvitnt^), likewise in the citation 
from Bez., a6b: ivi> rF»rnt6 tnaia ti«ni (Geonica, 364: 
tnaia "ltnt6). Therefore Aph. of stn, instead of Kin. In 
HG., however, Bekorot, V., 139 e : nb ^nriio" tri>N 'ia 'iron ny 
onnt6, ed. B., 597: 'ia n«int6 Kin t>iv t&N 'ia nnon ny; 
in 3io DV ed. V., 36 c: "linxi', B., 183: main ""insF, in 
b. Bez., 26 b : D3r6 n^rnx for, #«£ : D3r6 nnns, Hul. 59 b : 
nwDl, MS. R 3 and Rashi rvrnDl, and in two MSS. 
of Rashi: mriDI. In She., No. 88 beginning: "p* 
inai> mint^, F ^\neb, below: rnintcS in* Natron too wt? 
inab, then: pab *pnx nyjjo nntai tyM ,-ra n^rptn i^n j"k, 
F jnai> n^iint6, 125: iwao nnx t6n snao 'ten, E 'ins. 

42. No. 44 2 : nvdin rrb nax *e>n an ni> yV« toai? no, 
F n55 ; HG., ed. V., 139 d (B., 599) : rm, in b. Temurah 
8 b kvdik n^ nav is wanting. As to nas, see above, sub 21. 
nsD is merely another form of writing of nas 48a , which, 
according to Bickell, is to be derived from Pehl. JVBD, Pers. 
j£~- (see PSm., 3360, s.v. at the beginning), and signified 
as that : (1) Ab. zarah 50 b : ntov rb nsD = nax ' prepare ' 
('offer'), Maas. II, 6 blto naiD, b. B. m. 89b: niSD< t6l 
i>atn rbt-2 'prepare'; (2) 'mend, repair', Shab. 124b: 
»bn nx na niDai> nap n:»o niaD> t6i, Rashi: jprv nisD^ 
vvpiyi vnio^a n'oni>! 

43. No. 45 : mi> nTiD Np torii 'ia ^sn rota to\n, F nnm, 
but E torn like ed. V. — ton ' repeat ', also No. 50 (E and 
F) : '■a aio ova n*jnoi> 'jni xncp ^nj wm = 107 : i^n: 'tni 
wwd? nt? l»Dn'«D t*nop, and *'&#.: ty twni 'ia tF3n \s 

«» Mand. : n'StW: riBS, riNDNXO, Nold., Gram., 48. 



342 THE JEWISH QUARTERLY REVIEW 



mws, likewise HG., ed. V., 35 a (B., 176): n"jrvo Ttno 
NliriBN N'jn '13 pnDNl 'o tinoyb (Bez., 39 a, a different 
reading) ; further, No. 95 : ina S5n 'OJ 'tt top iTU tori Tin 
Nip ; No. 170 : 47 iVn^rui h«3JvJi &V, and No. 66* (from Meg. 
4a!): ttnenD Tap wn noNT |i» '13 snsxa rrynob a^n'D 
iTontn; in Anan, ed. Sch., 35 22 : pion onS, ed. Hark., 34: 
■T'in 'snin; 90: N'jm "xni; 93: N'on "anh ; 94: wmi; 
95 : junjd 'aniii ; 83 end : tttwra = 84 : -iNin»i>, &c. (see on 
Anan, sub 31) ; HG., ed. V., 23 c below (B., 104) : IT nw 
inj^no yn ytw!> n"apn n^ idn n^o nyna. In Babli 
besides the passage in Meg.— from which it also appears 
that the more customary meaning of WD was ' to learn 
a Mishnah or Baraita ' (comp. Suk. 28 b), and that for the 
other meaning ' to repeat ' reference was made to the 
proverb rrinST xi Knene "iujjn — also Meg. 25 a : Nni"D "idk 
ni? wT; Pes. 74 b: <35n n^ in m Kin n<a o^m N$>n wi 
•T3 ts^n, Yoma 87 a: ':ro «i^n; Kid. 81 b : (*wm) n«jni ns^n 
n'op, Git. 34 a : 'iDi n^ptsw wn Tin 'iai n^psfN S3H ; Men. 
19b and elsewhere: x~\p n*a ^5n, and probably also Naz. 
38 a : N^m nnnta n-i»jj »am wnpoo j,w xn ' and it is 
repeated ' (the twnpD), i. e. when lambs are dyed, the wool 
at the root remains 'loose' (because it is not dyed), and 
hence it has to be dyed over and over again (when the wool 
grows). 47 a Elsewhere only in the Targ. (see Levy, Targ. 
Worlerbuch) ; in Babli Tin is used instead and so pro- 
bably also Bez., 29 a : pm ttrbm noa (Var. pp-i) = tip 
(see above). 

"' TvbriD, b. Men. 35 b. 

47 a This is no doubt also the explanation of Asher when he says : 3TIN1 
lip*]? DIpD b"SH "Wil PIBinD. But he has not tPjni, while in MS. M. 
the following ini is wanting (M. merely : 13 n 3T(\). Comp. nn ttbl *\3W, 
B. k. 93 b. The explanation of Pseudo-Rashi, ad loc., is forced. [Sanh. 

59 b: wn»oi yvpoi] 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 343 

44. No. 48* F and E: aiwffrmn Tin (= HG., V., 
25 d), and No. 103 : niDK jmx rrmin Tin (=■ HG., 44 a) : in 
HG. in addition V., 27 a (ruKTTin Tin), 40 a, 128 d end: 
xnjjim WTin Tin (B., 529 : to-mn Tin) ; further R. Jehudai, 
Responsa, ed. Lyck, No. 45* : soTiin Tino xpaoi. In b. 
Shab. 77 b (RIF., Asher.and Rabbinowicz) and Pes. 76 a 48 = 
"lino -lino ' round about '. [Mand. SONTIW "iKin, SR., 1, 1693 
from below.] 

45. No. 54: SrvniJaD wipvb s'tidi, F smwo, likewise 
HG., ed. V., BTVp 12 b, in accordance with which already 
R. N. Z. Berlin corrected it to KmKiD. But the error is 
very old, for already the author of the "vntrTi found it so ; 
I, 245 : 'is was D-oan w ~pv (!)minn p itnpb pa«rn. 48a 

Ibid.: ^B>DE> rro B»Tp N^H K3V11, read with E and F: 
'CDS? '33. 

/&#.: Kms? ^j?d TwiaD NrniyDa ^nriNT somi. -mne 
' while it was still light (day) ' ; in this sense not found in 
Babli, comp., however, p. Yoma, III, 40 b : npn3 'piu ino 
KTUD tun npn3 'n3 pDN (in Babylon) pn. F TUta 'opo, 
but also HG., V., 30 d below (B., 154) : 'f>lDBt6 ]b •'jnriDl 
TUNTD Dmssn DV 3"iy, for which, however, She., No. 167 : 
SFpiSo ^inn}6 |ra*n»Di (but that No. is abbreviated !). In 
an Aram, piyyut in Mahzor Vitry, No. 307, p. 340 : n»rnilT 
"lash yiot3«i»i "hied? inisno ' since Friday ', ' while it is still 
day'. 

46. No. 54 2 : iTnn Wi sn nn ta % pia«^ (in F the whole 
Sh. is wanting, in E only the part beginning with s"3 WK), 
read with HG., ed. V., 14 a: nvm ^Dl; here, however, 
No. 65 : n<nn Jmari m inntr '•pisxi' ; HG., ed. B., 51 : 
(r. n»nn) D^nn jiddi jn inn ^ y\st6, likewise Cod. 

« Pseudo-Rashi, Hor. isa : K1T1V1 n"^ TliTI "IW fDS. 

481 Comp. Naz. 4a: JDK>1D nil, Pseudo-Raschi and Tos. s.v. *KD. 



344 THE JEWISH QUARTERLY REVIEW 

Oxf. 2760, fol. 28 verso [now in Jahrbuch, XII. 119, 11]: 
inn poos? m m bv tw6. Their archetype, the fttbn 
niplDS, had probably pDKDn (inst. of pDKOn !), as cited by 
nv^Dn 'D, Bacher, Hebr.-Pers. WB., 64 : rwi poxon, which 
the translator renders errdneously by |1D0, as fr. H DD (=' , 1DD) !? 
[pDKD is the Syr. NDB, 'DD ' to ferment, stink ' (AUDO., 65), 
see now Jahrbuch, ibid., 121]. 

47. No. 64: NDJK3 'ND [fiffn 3113 'piht6 T"TC p»Ho 131 

■T33i tnu ktd xi> K»!>n, further : »aa mno »DJ tosh jkdi 
si? w nt8> nana snu sriri uti 13 ova nine K3>k »m , s?ob' 
una "lixn rwt? is? p^uui pma, tdk, in No. 122: tnna xi>, 
smai, erroneously. No. 112 F: rvyut? 3"l$n i>ne*o^ "VDK1 
E n'op je "pan an obx, RNZB. Kin &6x) nape Tnan ton &6k 
pa^no pto panao pro pirau px twm watr (wa^ nape tot 
(F ofbfb) jrbrb "ud xep ton '« nma (also E) Wats' any 
ninat? tik6 'u tma (F nts^eu) xni>r6 -ud rnin'' un tma 
Nina (V. tma, which already R. N. Z. Berlin rightly cor- 
rected into tma), read : TO ton t6x, "nan Kin in E and 
RNZB. (see nW pvyn, letter r). TU 'dissipate, dissipate 
oneself (She., No. 10: "ivuk; 68 3 : TU'K, pama), to 
spread oneself, become dispersed ' of a fire, to take fire, 
eventually 'take root', to put, (abp) of a plant (Berlin: 
pto eme>n bpbdd). To the first meaning belongs Ps. 
97. 11 jnt -i^=ToT3 tnu! 'a bright (spread out) light' 
(imt is adj., JHT lit* !| nn»B>, as Esther 8. 16), Syr. 
K-IH3 tnnu Jac. Sar. I r. diffusus (PSm. 455); the 
latter 48b meaning reminds us of the Arab. 'J\>, prodiit 
germen. i9 The first (No. 64) is also of great importance 

48 b See Addenda, No. 19. 

49 We must mention here also Kalla r., II, end : TD?n "P"IX TlCND 

VrV t6o3 UT ^Sa Vllt^ (comp. Kid. 33a), 'get up', 'arise', and 
Tanh., ed. Buber, 3(8*1, 13 MS. Mich. (n. 44) : S)DV T3 N1H ^S3B'3 i>3N 
l!>UKr6 H3 (Gen. 18, 8 : QH^S HDiy =) V$>5> jfWiTn. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 345 

for the explanation of Shab. 20 a f., where we read : "idn 
pro pans p« prjns an joins put? an pans p« nop win an 
nop xnanDD Naa>x namt< anon an nb eppno ,an pans mivnro 
norm rnao «i> m$>mru pro pnao p^jna pnao xi> jnatc pnan 
an pans p« pas n!? an pans patw nop wna an nox nsa 
ps miwrn pro an pans ?y>vra (Rashi: p^n t6 nnas^i) 
an pans. According to the explanation of Rashi, who 
refers JTQD to the burning material ('they fall apart'), the 
phrase 'lM noiVN, by which, however, no change is made 
in the case of nop, is striking and against the talmudic 
method (as was already remarked by taaro 'flip to Asher, 
sub 1; it is true that Rashi has '13 nop n»K NJH3 31, 
M. and old prints torn n"tf norVK, but even this does not 
help much). The right reading here is preserved in Alfasi, 
ad loc, S. ha-Ittim, and Magid Misneh, nat5>, Ill, end 
(nnSD nspi Doiwn nD-u), 50 and no doubt it was also the 
reading of R. Ahai (his NTia leans towards it): an "IDS 
roi>jwu na pans pryu '« pans ps pan an pans nop rain 
pa^yna] 'ia nop an n»x xana an nc« '»a norrx 'ia pans pS 
[an pans niinina pro an pans px, MS. Oxford (see Rabb., 
note v): an pans pa^yna an pans px pax an pans nop 
here by another hand : pna'D nop) an pans p« ni^mna pro 
noriN toanDO jos^x mmx n"n rh sppno (pna'D t& nii>nira 'ia 
'13 'oa. Accordingly, P/jap (pt. Ithp. pi, fern, for masc, 
MS. M. and R. Nissim plans) is to be taken in the same 
sense as in our She. : ' are seized by the fire ', which is 
possible only when the inflammable material is kept loose, 
as appears also from Rashi's remark to the words of 
R. Hunna (according to his reading) : rt^W nar6E> pxc s sb 
nnon Daa^. 

60 Similarly Asher, but corrupt, as was noticed already by R. S. Loria, 
Glosses to Asher, and others : he had no doubt the reading of Alfasi 
in mind. 



346 THE JEWISH QUARTERLY REVIEW 

48. No. 64 2 end: may mxo no Npo'a ^ iinN, E ^ uriN 



•>nd Nipiv [Niprv ' certificate', Syr. K^rv, BB., AUDO 444 ; 
Pers. Jy>l> 'remembrance', comp. Arab. $/jj (Frankel, 
WZKM., Ill, 244). Occurs in Syr. also as NipuN 
Frankel, I.e.; in Cardahi, I, 548: *TP\)\ She. MS. 
F has NniaTN, Ba. ronai, Papyri Sachau Si P ar > Talm. 
'DJna pan 'protocol'. Ket. 67 b : npivi , NJaE>in ^ lrVN. 

49. No. 66: im Nirya inatw 'am 'pnoa wyb jwnei 
ynoi pyi Nmyn, F correctly ynoa, as was already cor- 
rected by R. D. Luria. ppnD fiaa Targ. to O'pn D'smk ; 
Syr. Menander, Land, Anecdota, I, 68, 1. 11 fr. below: 
•yrub \vb pp "np^td t 321 - 

Tfo'a'. : Kmwi> na an'Di 'ia inx ana, F correctly : naw 

na = miD na, imyna f?. 

50. No. 68 1 : sniD'TB N'n Nin 'ia aipi Nain i.tj tii ; 
HG., ed. V., 132 b, ed. B„ 549: wwitb ; Yntm II, 4: 
aipn ^y ntniaon naon nvti; HG. roann ed. V., 129 a (B., 
529) : Nona byi Snorrai (B. correctly : NDia i>yi ntidtbi), 
z'fo#. below (B., 531) : Nnorvai Nain tini nxn ap^ei wm 
dtid, and ibid.h: Nairn Snorva a'pro iai ; ini 'n 145: 
nNn bv a^n ^ Snona ap^na-a ; Tntm II, 83 : Nnona triTa 
'ia ap'3 dn ^*-ud i>N Nairn (below : NnDWB) , J?.x^ ' napkin '. 
Pal. Targ. Nairn anions (Levy, NHW., IV, 124, TWB., 
s. v.), comp. Talm. Npna and Kpia ' veil, awning '. 

Ibid. : ntvhxi Nam u«n ni^an »aj ^>ytr a^n, F NnixNi, 
E SnSrnn, HG., ed. V., 132 b 'riNiisi, ed. B., 549 NnNvmn . 
NHNlis pi. (from N11V, NnilX?) 'kidneys'. 

Ibid. : \rb& \rwhmb wai, read with F and E inrataoi> 
(wsb&rb), on tfrw, Hul. 93 a, comp. below, sub 72. 

Ibid.: 'ia nnaa nmjw ino, E and F vwo; No. 169: 
'ia nnw Nin no, F and E i?r5; No. 89: Nin »d "niii, 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 347 



E N«rn xmv wn, F tmn twi»; No. 92: 'ia wd sin '», 
E ltd ; No. 73 : 'ia tow "N sin »o '13 i*hs> sifHc, F vtd 
S3'S 's. As to Kin '», sin <nd, kwd = 1iTD, see my Rechts- 
gutachten, &c, p. 270. Comp. also Epistle of Sherira 
Gaon, ed. Lewin (in print), p. 67 to the left : 'NO S71 
(wd=) 'kd t6x wop pan n^ pjrr mn sh (m»D=) 
'13 vnpaen. 50a 

51. No. 68 2 : 'on san 'otn iidn b»*id sh j"jjn '"inro sni 
jniDN w pain ^>nide>, E spnro, F spntio ; HG., ed. V., 
133 b, B., 554: SpTTD wn; in INI 'n, 147 (an Aramaic 
passage !) : pTDN STT^rriB iTTVtn K3V1 i>3, #w/., 140 : 'pTTO 
jniDN Tn; "Vnrni, II, 3: -iNl¥ i>E> pplTts Wi (= HG., ed. 
B., 519 below: m bv pp-iro D»3(sn) ; HG. mane, ed. V., 
129 a, B., 530-1 : »piiD '13 nn nn sriN^sn (B. *pniD!) '"into 
wen nn njrnx 'pain <pnro '13 twin? snipasn (B. 'pnto) 
n<«n 'pitoi 'is wn-n im nt nmn »an im (B. «nm wsn) 
pntoi '13 »qbjh 'piro 'is xaiaTi wvt? nn mew ^aan 51 'pntoa 
'13 'otsjn (B. naiaia) taiona nn soid; further HG., ed. V., 
i26d (B., 515): inrann^ pan* nonan (here = pirn) 'pnto ^m 
nn^ nonan ndt ntnrwn 'aw <a pace nna; further She. 
92 (= Hul. 93 b, Pes. 74 b) : xpTTDl 'JP3 KVD1K ; 'pTTDI also 
HG. naiD, ed. V., 32 b, B., 161 ; Aruk, s.v. yon : KplTOi ; 
HG. DT V., 134 a, B. 559: »pnro. In Pes. 74 b, ed Ven., 
and ed. Amsterdam, 1645, have 'pnno pi '13 'pinoi, like- 
wise MS. M. ; MS. Oxford : »p*TO '13 -p-.'tiDi ; in Hul. 93 b, 
MS. H. and R 2 have 'pn'JD, Rj »p"W» (both times) ; 
MS. M.: V'tio pi '13 (!)'pmoi. Also RABN., No. 24, 
fol. 13 d, has *pTK) several times: (from She.) 'p"Vle »:n i>a« 
5»t? pTjn Dm "pTtcn p dt kw6 'ia »ptid3 ni> ^ididi 'ia 

60,1 See Addenda, No. ao. 

51 In the nDW 'n by Gama, Jud. Zeitschrift, II, 79 cited; »p-|ND 

. . . Naiaisn tw nn mcy ^aaa tvtn. 



348 THE JEWISH QUARTERLY REVIEW 

'13 Di, and others in the same No. ; No. 279 : p"HDK p*TO tt"l 
'tt YVK2 'i '13, and No. 282: pill on VTtD pi. Rashi 
explains »p"ito by pm "wxn 'Bin (comp. Tntm above), a 
marginal gloss in F 51a : D^pliy. From the Gaonic passages 
it becomes evident that ''p-HD on the whole designates 
' sinews, veins '. 

It is the Syriac la.ymvi, |uo*wd)jo, )j3u;xo)j», Budge, 
The Book of Medicines, I (ch. 15), p. 267, 11. 8-9 (where the 
tOWBDtO NTinn N»1S follows upon nOTWVI H-'bl} W N'JJD 
ND1X) 51b : KgnDMD jnpDDI N"133 |D pntH fUH NTHI 5»3fl N3in 

fl*31 ; further ch. 20, p. 420, U. 7 ff. : ndiej iron pn rvinriN 
nois jd /in ndiv xnpn»n in n^d q»d /nen ndist twinn m 
annai? pbioi snxy KgnDtm paw turn p ; ch. 19, p. 404, 1. 3 
from below : NgnpNB pan sima in Niaaa xn3D pinn nd ; 
and ch. 16, p. 329, 1. 1 (the liver receives the juice from the 
stomach): NDia spin ntn pp'am np^dd pnpnon |un tnni to 
now Nipnon n^j? in w d$? mdiddxt NTinn ; comp. also 
BA. in PSm, 2752 : ^ u-yuil l$*» v**-^- *k*k» Oi^ ^"•> Q: ') 00 
^1 Jl J.JJI jujUj . jJJI Jl jjjil. [spnto is, as the Syriac 
NpnDD, the Greek juea-apai'ico? ' mesenteric or pertaining to 
the mesentery' (PSm.) 51c , and has nothing in common 
with the Hebrew P"}]i? (Levy, s.v.). The form NpTTD , 
however, is perhaps influenced by Targ. Np/vtD for Nijnjo, 

PIP?]- 

52. No. 68 4 : 'nBn i^d b, E 3H3n (F corrupt), see Levy, 
s.v. H3. 



7&V/.: Krvuu nDnon, read with E and F to«aw. 

61 a Likewise translated by Gama, ibid. : pny, ■«<•£• 

61 b Read JuOoo inst. of jn>\,v>, as proved by the parallels. 

61 « In the IDIDiT 'D, ed. Bacher, p. 44, 1. 10: KpnpDNO {Arab.) 

ffwsnn onb piip nnm ttpi Q , i''32 nwfcni>n nr "nan raNici, 

r. Xp}lpND, see ibid., p. 189. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 349 

53. No. 72: '13 "b -nvn Spon rr6 ana vm "n ('h, so cor- 
rectly E and F, see above, sub 2). In Babli Npcn is used for 
court documents, summons (Kid. 70 a, B. k. 112 b), epistles 
of spiritual leaders (Men. 40 a-b) and doctors (Git. 36 a, 
88 a), but not for private letters. The Geonim, however, 
employed it also for private missives : Gaon., Hark., No. 
368 : -pvai mta ni>W3 l-pan Svn amac bind ana i>a, "^pESn 
pani . ^dnh panpsi men ni-ux jnc ;na Nwai ?n»i stws 
ijmvp na ip'DNH !>y ninn^ pa-sot? p^n vn (to Git. 36 a) 
W DiTT nicnn; HG., ed. V., 41 c (B., 203): pan ptnaDi 
ana^ tdn ^pdh bax NnSiinT Opna =) ^pds run {>b> nin^x, 
see also Aruk, s. v. ; comp. also, Nathan ha-Babli ( Yuhasin, 
ed. Cracow, p. 120, and Neubauer, Anecdota, II, 78): nisi 
nn npn* ama annaoia na'C 1 B>tn nti' nt ni*UN ?nbw vne»a 
'13 J1NJ 'Dp NpDn (see Levin, Jahrbuch, VIII, 327, n. 3, and 
Aptowitzer in JQR., 1913, p. 33); Responsa T\"w, No. 39 
=Lyck, No. 46: wiud pa*n w p ^"t d»j*u }rm xp p«i 
'iai jnw i»a 'pids 'anai "^DNni 'die's NnmnN ba ''anal. 

54. No. 73: -mpch "i"DN vb '13 »dj nivpi>i, E and F 
TVno5, TXrvo^, so also in the ed. V., No. 170 -ivrrD^>, and 
No. 98 TVH, likewise in Anan Sch. u 9 : "DTn tmn ; nxp'D 
in all of these means 'become sick' (K"V¥p), Anan, ed. 
Hark., 38, HG., ed. V., 67 c. [as in b. B. k. 91 a ("apb, ivpn) 
and Mand. ("IKWD), Morgan, Textes Mandaites, p. 192, 5]. 

Ibid.-, vbcb yamtn pvtdjT jwbni, HG. n^n, ed. V., 15 b 
junwi ; as to nw = mi, Cant. r. V, 14, 3 : ,KTon '33 aw 
^3N Nim NnD3 rt'DK, see Gaon, Ukzin III, 10 ; for further 
information comp. my book Der Gaonaische Kommentar 
zur Ordnung Thohorot, p. 62. 

55. No. 79 F: njt rvi> man pja onn, likewise man onn 
in the commentary MS. RNZB. (see nW P05?n), E and 



35° THE JEWISH QUARTERLY REVIEW 

RNZB. Kn n*b inn, but E on the margin mm. man 
from na 'to bore through ', hence 'pierced, opened ' 51d , as 
e.g. Shab. 146a: NTvan na'ob (see Levy, s. v.) ; HG., ed. 
B., 423 below: xmtn nrmi> (= njia) nna N$n = itn 'n, 
67: rrom rin bio napj xta in once* ho nap! t6e>i, ed. V., 
103 d below : W13 xh ; Responsa 3nyD1 mto »Jitu No. 68 : 
wnn!> nma'oi? niDx rrotnaa (xniTp =) winm tnonn warn 
'la wiitp ^rb nnsn jm 'ia (xnrvps) twivvn. 

56. nvoix. /5ft/. : !vxdij6 ruon $>ntn rrmiyDa n^o'on 
■6nt?oi (F nvoin rr6 rooi), further: '•yarvo ^ax xpn man 
non ttpir ^3 pi»3M xpn nnai »a<v (F mb pjnrvo, E rvi> 'j?3no) 
*a»v »a rwo kxdin i>'wi non »a NsnT rwo tan pen. Here 
therefore KVOlN in the sense of 'morsel, piece of bread', 
parallel to N5P"U (' a small loaf of bread '), which fits the 
context best, since the discussion turns about pon ! 52 

Further, No. 90 : nh i^yo nvoin vh fxanrre «h |"3on 
n*va ; in HG., ed. V., 9c: tmrb pbion xnaj in njijvn i>ai 
'13 rpT hb»i 'sow N^y N^n pannn 'ia wpi:^ ; 53 #&/., 8 c : 
bivh sworn nvoin avw ta Dna basn rvaa; iob(=B., 64): 
«i> 58a nvoin ^3N ^pk jotjo nn S>ai '•aiboN *jitaxi lawNn wm 
pon 'anaa pmrro; ed. B., [20: rfb '•yano nnn B*npon na»ni 
SvoiS 54 oyatb ; also Seder R. Amram, riDa (ed. Frumkin, 
II, 314) : nvo btotb NV01N Kinn bv "paon ny has t?i»i yxiai ; 
meaning ' a morsel of bread ' in all these places. 

In the Bab. Talmud NXD1N occurs with the following 
meanings and nuances: (1) 'a piece of bread' , Ker. 7 a: 

51 d See Addenda, No. ai. 

52 R. N. Z. Berlin in nW poyn : 10J> DB1 *B /IBO NXD1N1 ! 

53 The explanation of Rashi Hul. 107 b, s. v. D?3, and Tosafot, s.v., is 
quite forced, see below. 

53a [-po,. that in my Fragment, Jahrbuch, XII, 111, 27 : AS.] 

54 Responsa, flDM mOPI, No. 9 : Kpin KOnf>, HG. ed., ibid. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 351 

Nxoix n»npjn NonF i>3Npn nrn txdyn ; Kid. 30 a : in n'ti 'i 

'13 Npiri> npoi 1$> NXOIN D'VD n!> N3N (comp. Suk. %6 a : 

rbJ? i>"jii 3"V3 n3 d>j;bt3 '13 wiy ti^jm nosi) 55 ; Hul. 107b: 
nxoin n^ D^3 'gave him a morsel (of bread) in the mouth ' (see 
above from She. and HG.) ; likewise probably Shab. 155 b : 
t«W> nvdin NIB^ WIN mix ' a piece of bread ' (' ne>D— nosi 
rvj-nx ') ; B. m. 70 b : 'is mm inoD nxoin n^ nh2» ^t fV3, 56 
and so also Git. 69 a : '13 TO nj)3B>» 'XD1N 3B> '3r!n ' seven 
morsels of bread' (one gathers bread but not meat). (2) 
' A piece of meat, meat ', Sanh. 39 a : NltP'an NXD1N ^ llV'N 
wto "nin rania ni> WN ; B. batra 22 a : '3^ 5»?n nin ^-1 
nvoin ^pt?oi> Stud ; Shab. 140 b : p3H> nxdin p3n 3n 13 ns 
ioia <ro Nni>n ,T3 rvtn np^in ; Hul. 44 b: pn Knanu n-ib> 

NVOIN NOl 11 i>3 NrV^yD NXD1N D1K>D N3'N 'NO '13 iatj"3 n«3 
^ war NJV^yo (' a good piece ' of meat), and z'fotf. 93 b 
(= Pes. 74 b): nrbv) nsnn p^odnt nxoin '13 v~itoi 'jm 'vdin 
'13 nmpi? ii»SN (= She., No. 92), and Pes. 74 b: nvoin wi 
nt? n'^ni Nin p'DDN n$> tdn n«bm Nin 57 p'ODNi, so according 
to HG, R. Hananel, and Tosafot, s.v. p'ODN and M2, against 
Rashi's correction. Concerning the dipping of the meat in 
vinegar comp. below. 58 (3) ' Seasoned (in vinegar or spices) 
raw meat', Shab. 128a: nvdin^ nm NttlN n3 yj«?, 142b: W3D 

55 Rashi : "ip33 0150^ fiwn Vn B^KO D3>0 1^3 ! see note 58. 

66 People gave and still give (despite 1JYIN |!3^£TI 3^3^, comp. Sab. 
156b) bread to the dog, comp. M. k. 28 a: N3W> NTOD NlDn 3T »3 

'13 ^wvb nytn Nona t\t\ »a 13. 

67 Perhaps it belongs to number (3) see below. R. Han. Pes. and 'Aruk : 

nonNnt? -iea bv n3Tin. 

58 To (2) probably belongs also Shab. 109 b : NJpn N3^N N'Ol NXD1ND 
NJPT N3^N NJDK> NTB«301 (Rashi: Qijjrua 1E>3, see note 55), despite the 
N"l&'\3 standing near it, on account of the wholesome JTHntf flD, B. kamma 
92 b; likewise Yoma 80b: Nn^Dl NX01N !>3N. On the other hand, Ber. 
45 a : NS01N iTnpjm is doubtful on account of Ker. 7 a (see above). 



35 2 THE JEWISH QUARTERLY REVIEW 

nxoin^ "b nn m *b nab rov nnx (Rashi: jnjw nix ya cw 
NSD1S nf> npi »n iea D'taxi na 11 ) ; Hul. 15 b : nbmb omefi 
nx»in3 n^ -rniD mea (RG. : nawi n»fa aab ^n ,nxdik3 
ica 5>e> ruop), 89 59 a : NXD1N3 rpj'o i»an^ naD 'ia two na xinn 
'ia, and 113 a Aruk: dicd) w -vac Nxawa byvb '•ya w 
(Dna»N3 on y^3Dn ; RG. : -iidk nxoin rrru b^ »»a »tn (to 2) ; 
Alfasi: K^> IN NX01N3 rwo b'D^ n» "0, also RABN., No. 
277 has: NXD1N3 runs bnai' tdn <» \b ^yaw 59 *. 

In the Geonim : (2) She., No. 44 s : trenix n^ nax (see 
above, sub 4a) ; HG., ed. V., 133 d (ed. B., 558) : 60 nvdin 
id n«iriB>i nwon /-it? nn^i) rrtans 'pidni 60 " xi>n3 nwoo 
•vdn n^m Kin pnaoon (= Pes. 74 b, see above), and 134 a : 
nkw 3W not rw» ens podni iva jnox tidvd nnsi ;kd 
not iT^v warn nxdik nosoi. (3) She., No. 68 2 : on tdn yn 
NXG1N3 rrtae nc b*tb sin m bax ensn noa ons^s 
(=HG,ed.V., 133 b). 

In Syriac we find (1) NXDN BB. 19a : Ujj^j Jl ^.Ull nxdn 

yai^ll S3 llj «U«i!b Ji^jj fV*^ < -^ f^ >* J^J^'j ^*^ JU . : 

'NXDK Arab. d^l) ^Ul, &c. With us in Takrit and 
Mossul (kson means) meat cut in pieces and eaten raw with 
spices ' (see Lagarde, Abhandlungen, p. 13 ; PSm. 342 has 
J-«jU and renders Tacrit MossulanaX). Arab. ,j»»T caro cruda 
aceto imposita, &c. (Freitag ; concerning the etymon see 
Lagarde and PSm. /. c), comp. BB. 427, s.v. N)Yi"i3. 61 And 

m = She. No. 128. 

s9 » So also CKT, ed. Wilna, No. 48 (p. 10): Hi) 13 v6 KW 

ne> NXDiNa nnni>. 

60 HG., ed. B. 558, has twice: NYDN, likewise in the citation from Hul. 
■"Va NSCN, p. 559 and 161 (yet also ed. B. 554, last line and 114 : 8203183), 
the pointing K2R2K probably originated with the editor. For (3) KXiSK 
would be more correct, see below. 

60a See note 57. 

61 In a fragment of an Arabic commentary to Hullin, Oxf. MS. Hebr. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 353 



-o- -o r 



(2) NSOIjr, which simply signifies 'piece', 'morsel', £*&>, i»iU, 
as N»nh NVOiy, KnDan twow (= N-ie>m kxdik, Sanh. ! see 
above), PSm. 2915, likewise Mand. twoip, Noldeke, ZDMG. 
XXXI, 539, n. i. 

That NVD1N was eaten raw is evident from the passages 
quoted, sub (3) ; from HG. and b. Pes., 62 cited sub (2), we 
learn furthermore that in Babylon they used to dip raw 
meat in vinegar and so eat it, which corresponds more 
or less to the testimony of BB. — In Babli as well as in the 
writings of the Geonim the Syr. NVDiy and N¥»K were 
merged into a single NXD1K, which is explained erroneously 
by the commentaries and lexica throughout as 'raw meat' 
(Rashi in some passages : meat slightly roasted on coals). 63 
NVD1K is construed in the Bab. Talmud as fern. (ivnpJn, 
«n«i>yo, naniK, nnp'BK), so also Syr. twoijj, see Noldeke, I.e.; 
but also as masc. (P'JDDKT, n^ni Kin, see, however, note 57). 
57. kb^j, veami. Ibid.-. Knijwp pr6 n'N ma-ii> 'ia twn u 
■pro mi anpoi Knijrap pn^> jt6 noit in^n wo anpa sip^D 
nnijrDn, E: s-ip-yD 1&6 Kt?B>5a f^asnon nou anpo tnp<yn 
TT" , DI 'ipo "IID^ W anpD, similarly RNZB. : l^asriDT noil 
woi anpo "hd^ »s» w 'ia (?)tE5>tia, F: p^wim nou 
tWiB»»a(!). Read with E NMM3 (KBB&), or NBB>ya (with 
RNZB ; KDCD is merely miswritten or misread from KDB>1J), 
Kne>M3 in F is corrupt. Syr. NBBO or NBBto, excoriatio, in 
Talmud Sanh. 100 b: 5>lr^ K^n mniKD wn^a bieoti «^ 
R$>ar6 roana (soii>M = naiT, here). 

d. 46, 17 (Cat. No. 2643), fol. 104b, we read: 'S b't Wl Wai ^p 

^osri'' *j on^ in kvdin -wan .'i'la nhn^> awn wrbtt axna 

J»DV NBDV "intOHl ( Jij) ^33, ' NXB1K is raw meat which is prepared 
with vinegar and spices, called in Arabic iJ e^&'. 

•» Pes., ibid., in addition : '13 MCT tnn n'Q B^m K^H W. 

0,3 Perles, JQR., N. S., Ill, 313 (in the name of L6w and Giuzberg) : 
KXCIK ' raw meat ' (1) = Syr. 'umsa, ' bit '. 

VOL. XII. B b 



354 THE JEWISH QUARTERLY REVIEW 

58. No. 80 : »!>33n tmia |U3, E and F nni3, see, however, 
Low, Zeitschriftfiir Assyr., XXII, 79 ff. 

59. bsi. No. 84: »B*!>m iro »tsd ^in aavi D"n x"3 

|V3 ''IDK "D /KD KjntW IB^rm 'DpO 'TED 'hjl Wfl JtXffUXQ 
'13 >ha &631KDT, likewise F ^"U, ha. E, however, has: 
jd bin sen (sjj-in p vrbm -ina n»B jd bin tovi) »!>»d »:n 
»tb p ban N3"n (dittography !) kjhk p fc^m -iro '-pa 
'13 ^33 Np nSsindI '13 '•KD K»i>m "Dp'D, likewise the Comm. 
MS. Berlin, fol. 186 a: 'i3 tfi>31N p ^3J1 '•"ntn D*n v"3 
»op'D n'B p 5"3» wn /t^m nm »tb p ^an wn pinion 
'13 ns^m ; HG., ed. V., 131 c: 'is xbix p pbn &vie> ii>w 

'13 sblN p D^3JT Dnnt? P'lOSp "31 (ed. B., 546 : &'X*1B> I^NI 
l^m, the rest is missing), likewise miKn 'd, 121 : l^Ni 

'13 ^an Q'x-itr pnoK 'ai '13 >5>run onnt? ; Hul. 67 b : ratopip 
^3J rWD N»yt3 'KD ne> (Rashi : 'ins |ie6, likewise the above- 
named commentary, fol. 1 85 a), ibid. : trvnrn f?2i up '3 "oni 

|^3J Np. In the Geonim also HG., ed. V., 35 c (B., 179) : 
% nriBN miDs nsne nro pnr ~i"k epv n3 prw 3-1 'on 
noan '»o 'MD (similarly Alfasi, Hul. Ill) nniD nana rrcra 
Kin KD$>1» N1BJJ i»3J NpT MTW miBN i"33 K^> mDO «h, 

ed. V., 131 a (B., 543) : bw nnsN miDK nane wa 'na^m 
i>'3J nS hido xbi [ed. B. nDan] 'oj/b <kd "wid nana wa 
sin nd5>»3 xnay ^3j Npi toi'ini miss ; in b. Tern. 31 a (in 
the name of Abaye) according to RSK. : rfyn now wa 
Kin xo^jn Niajf nniDN <ai nniDo >2b. 63a — Syr. bau (= Hebr. 
*W) finxit, formavit, Arab. J-*., PSm. 640-1 ; Audo 114: 

»» p*DB» has also N"D : '131 iW3D i>HJ Np nO*N miBN, RG. : 

N133 tick -101b iwao buj now nnstt bax wins kjb^ 

'131 W33. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 355 

Winn DID JO Dio D'px ,Dr6 ,|priK . . . has. It is also employed 
in an Aramaic piyyut in Mahzor Vitry, No. 30a, p. 336: 
'is wcna (sic !) |jj'ao jsrmva ]Wp: ,wna jcnn ,npnxa £'a:i 
' formed ' (' created '). The explanations in No. 286, 
pp. 311-12, are forced. 

Otherwise it occurs in the Bab. Talmud only in the 
cognate meanings ' knead, mix up ' ('"lin? ?'aj). 

Ibid. E: '« k'-wd nowa vi nona ,nona onn nwh 'an. 
This 'n seems to equal '? (comp. in Anan paa = pan), see 
Anan, sub 9 ; likewise uncertain is Resp. aiyoi mto '31KJ, 
No. 141 : (?*n [onaji]) 'n (mik =) dais e»ai>Di> ^-ic ^nn 
VOID twiBoW', or does also Ti in Anan belong here (see 
Anan, sub 17) ? 

60. No. 85: tuu nyatn smixai '12 m^n xnrvN i^nt 
nd^v ni> na'K to:on, E xna's (F ana 'n) ; b. B. m. 84 b, MS. 
Hamburg and Florence (in Margolis, Gramm., 52*) : Ninn 
K»^5? rp> na'N NOV ' he recovered ' ; its opposite is found 
in Midr. ha-Gadol, 758 (from Ket. 103 b) : tscby nb wbn 
' become weak ' ; comp. further Ber. 54 b (HG., ed. V., 11 b, 
MSS. Coronell and Paris, see Rabbinowicz) : iTi'tn (itn) 
ixcbv rrb '!?Tta ' he convalesced ' (comp. NTi'p 'h'N NOV 'in'N, 
B. b. 16 b). The comm. MS. Berlin, fol. 109 a has : r6 na'N 
nr$> (*iyvo=) -iyDD nana dn ,no!>3?. 

Ibid. •. rrb 'annvK 'o 'nh p'-ioxi, F ns'-iovn 'no, E 'no 
N3'"i¥, ed. Ker. to a: movb 'NO 'an 'X, comp. in addition 
No. 97 end: 'iTl^'V naOD 'D, F: rr6y naoD nin WD, and 147 
Ven. : li> N'Bnp TO. Concerning 10 = 'no see my Rechts- 
gutachten, &c, 270, further Geonica, 97 : 'Otn 'NO (= 'OT 'o) 
rs\rb Nan, JJ/zafr-. ha-Gad., 343 (=Ber. 62 a): ni"»s; 'So b*H 
a"naS n'ina (ed. 'D), see also above, sub 50. 

61. No. 88 : N'oa n'-riir6 n>b tdn 'to n'i> dipni Na'ni 

b b 2 



35 6 THE JEWISH QUARTERLY REVIEW 

D'»3 n"b 13boci roeo n'i>s» rr>b Etpni 'ia Nrotja, F : torn 
pana N'oa nwoD mi? lewi am rv»p '13 'TD DiiivJn , E : 
'a nd!>n iTjndd rr^ '3tj>3K>i an ir»p 'ia n^ FiSn ra 
(n'JKDD n^ wsd also RNZB. and ed. Men. 94 a). Read 
DIV'N instead of Dir»N; D'1N Aph., or D']N = D'lnN . In 
No. 35 F D'UDNl = D'llriNl (see above, sub 35) ; No. 96 : '31 
WVPI 'HiTI NB»!> IN NTp ND'ND ND^H WSJ '31'Jf^ n3'-iv ni>3D, 
E jid'kd (F nvno !), RABD. pan 'inn ,ns'sn cites : hd'N» 
NTpn ; HG., ed. V., 85 a : To rntwa ndid Noi»n (= KplD), 
ed. B., 626: 'TD HDl'D na n'N KD^I; HG., ed.B., 406 (from 
B. b. 23 a) : narb rb 'dhdi '^Nna 'am (='Dip), ed. and MSS. 
'1DSD1 (perhaps 'DS»1, as E, No. 35: B»SD'K1) ; B. k. 18 a: 
ne»i>3 dindi, Rashi: "jtai' pB>!> dindi, read DiSERfi* (dindi), 
ed. Soncino, Pesaro and Ven. 1521 : D11NEH ; MS. FI. and 
'Aruk, j. w. D11 : Ntf'b DliriDl. 

The stem is therefore at any rate Dli, D1 = ^J.^, as Levy 
and others assume, and not Din, as Buxdorf and Fleischer 
suppose, Additions to Levy, I, 560 a. The missing of the 
third radical does not speak against it, see on Anan, sub 1. 
So also the Nahardean recension of Onkelos reads ?? for 
'^V, DN for 'DN (comp. Berliner, Massora). 64 a 

Ibid. -. 'it? 'ia n«d3 ttd ir6 thni N'oa sjn'D irr> spirt 
inj'3TDi'i '13 inj'3i»i irv'sno '13 in«3-i'oi> ne»i lrvsnt^, RNZB.: 
iiT'aio5» ntsn in^ariD^ ne> tti» mi? t">*>k (E titd mi> t-hni), 
F: in»3TDi lrwarPD^ '13 in»3Tc6 nt?i '13 N'oa inb itd ti'ni 
in"3T»7l in"srPD7 '13, which is probably the correct reading, 
from "p, 131 'make soft, mollify'; "HX. Referring now 

" No. 125 (=92): NJT1B3 iTT,2»37 D'ND, F: n'7 D'NO, E: fJ'DDDD, 
92: p'DODDD, HG., V. 125 d, and ed. Hul. 18 a: D'ODDD . D'ND may be 
equal to D1ND, for D'ND, mies, occurs only in the intr. Kal. 

Ma [Comp. however the Syr. jxool: NHN |D NpBJl NJ1NX, AUDO. 
11, according to which it might be permitted to assume K'lN, DIN as stem.} 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 357 

to b. Shab. 33 b: tvvpJ? n^ ttx xp nm ma »ni> n^"j>, in 
view of fVXl occurring here in the She. by the side of 
epTTi , we are not to follow the explanation of Rashi and the 
Aruk, as: It)]} injn, 84b but translate it ' soften ' (make fresh). 

Ibid. -. Tajn xin N-inn m xiToh roa bx. xtwd=tvd, 
as KICD^ on the side of "ICD^, xncei* with TOTi^ (see Anan, 
sub 13), &c. F however "ITK&. 

6a. No. 99 : '13 "J^d xp xna 'eta, read with E and F 

'ID KH3 XtJ^. 

63. No. 90 : x^>x .tt y*vb wmb} xt xnro x-d bipt^ xi> 
1trDi> in» xw x^ro xs^xn xsti '13 xnnn imi-in iruwDS> 
xe>d»i rra nni m, E: '13 nm wawd? '13 "irwini* iro-ews^, 
F: n>T l^D^ '13 vu3BOb6, RNZB. : in«mr6 wsboc^; 
HG., ed. V., 9 b: '13 "]Wnb '•a-vwwb (ed. B., 67 : iruwin, 
68: irumpin mo!), here in She. read therefore vwzvh.— 
Concerning xor6 see above, sub 1 8. 

64. No. 93: W nnxi five times, likewise E (F 
erroneously W), ed. Hul. 17 b xnnn xni>nxi. PL on »r, as 
invo, No. 115, W3, 64c No. 159, see my remark, Jakrbuck, 
X, 373. — Zfo'rf. : w\b r\y\yb "ie>3n»S> xh, ed. Hul. 18 a : '"axi 
|31!D x^ '03 n33$>. -lEOrVX in its proper sense 'to be fit', 
in Syr. frequently, in Babli in this sense only mtSOK 
B. m. 89b, above: XVa niB>3xi> '13 X13J niB'axS'. In the 
Talmud ic sense She., No. 127 : ""lBorvo xp nnn fWM. 

65. No. 93 2 : xrnrox ivnax x^ air6 x^, RNZB. and 
Com. MS. RNZB.: xHuEivx by xS 'max ab, E : x'max x^ 
xr)3DD' , xx xh (F has x nDVOX xS xmax ab, for the cancelled 
xnDVCX the margin has xriD'ex !). xmnvx, xri3DD' , x = X3DVX, 
X3DD''X 'bench'; Tamid a6 b xmoxx, MS. M. and RG. 

64 b MS. Oxf. T>3n; 'Aruk has nnt5"3X. 

640 Also in Neo-Syr. the PI. M3.1U ; Noldeke, Neusyr. Gramm., 137. 



358 THE JEWISH QUARTERLY REVIEW 

XD11DVD 'bench of stone'. The same mistake in HG., ed. 
Berlin, 507: Nip'1 Knoxox j"y kip>i nbo j"y rum' «i>, read 
KD3t5SD = ND3 Erub. 98 a, Sofr. 3, 16 (NJV31DD0 in Yeru- 
shalmi), or else W13DVK. 

66. No. 94 : ^ibS>b ^>3k Sn5 '3» irnimb nc tono yj iai 
niDx nmmi wdk, F. : mini '■a sFtd »5 jo, E: rnoa jo 
(cancelled and corrected to nna U'o!), RNZB. : .TnDa |0 
(r. nHD3). No. 96, E: N'HD UK (= Ket. 61 a), F: rHD , 3K; 
No. II (= B. b. 9 a) : NHD '31 NniS. HHD3 = flHD U. 

,/fc/. : ro<5> npJi 'is won SaSy Tupx^ nci, F: aziv 
np'h 'is, E : mp^h '13 nany, RNZB. : -nph saiy. sany = 
Talm. N31N ' trough ' (then also ' boat '), Pes. 40 b beg. 
(see below, sub 78), Syr. K3Ty, BB. 1549, Mishnic nany, 
[comp. my Qyitun wra — in print — p. ofi, n. 24 and p. 54, 
n. 16.] 

Ibid.-. r<h poNDi aw ny n^ noj Niw '13 iids u»jj pcoh 
wj> foxb niDN a'otn -ina i^sk nacai /wy, F: poyS 
'vioy^ niDN '13 (E fond) poyo *nni a'aen '13 rw»y (E peyoS) 
nw n^ (E Wt6). po'O also in b. Shab. 108 b. In the 
citation from the Mishnah and Bar. F has y : j»oy^, pxoyo, 
mvoyno, while ed. and E have N, comp. Shab. 77 b. Syr. 
-scLi., Talm. poy (Levy, s.v.) [Mand. finish r6xxox, SR. 

I- 394s]- 

Ibid. : '13 nae>3 rrb nayoi> nsn »no aio ova a>3tm, 

F: '13 nayo^ rtb ntn j55 'ia a^an jndi, E: nen no. 
fXO neut. in Syr. as well as also elsewhere in the writings 
of the Geonim (see my Rechtsgutachten, &c, p. 232). In 
MS. F in addition, No. 41 : N3H "ion n'B'W jSo V'x, further: 
^nioe> 'ont ]Se no nun, 6 1 : 'yanon [Sob »bd mi> Tuyi p:or, 
in: pnon jso xi'N d^d xi? D^o nai, ibid. : niena ppvj 
d^ p'ron jSoi No 1 "? mo, and 113: yian "lyBK^ n^ tdn 
■•lBn jso 'SO n^ (r. jxoo, ed. *xoo). 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 359 



Ibid. : rwrd> nb vmb\ nmvit rfc pvob, read with E 
and F orfoh 'to sew', from Bin, am; p. Kil., IX, 33d: 
mttov EirvD, b. Men. 37 b: nd^ n'0"m, &c. 

67. No. 96 : NDW, see above, .r&# 61. 

Ibid.-. pi>*y n^i n«T» piD'oi >-?-\\n Ditr» non ^ona '^bni 
wo ma, E and RNZB. : pj»N nnpn dib>d xi> non nsrn '^bki 
tbb> n'd (ma) na 'id inaoi Own), F: dib>d xi> n»n nsnai 
'13 n!> nobn rrno piD'Di .nnpn (it refers accordingly to 
jpini tn»y anal !). 6 * d HG., ed. V., 85 c: vroTncn 'ana i^bni 

'13 t6l tPPD 'OTDD nnpi NTD, B., 626-7 : TOn '»n3 I^BNI 
'13 N^>1 N"TD TTDD nnpi DltTD, ift/. ^/a, 103 : 'Cna I^BNI 
'13 -ijw TTDD pure jnc 'OBO h6 non. Therefore read here : 
Hid^di Own) nnpi, fr. tid, as Hul. in a: -p-iDD i3iiD ai>n. 

/fo'rf. : 'is ntrn piD^ nanvi, E: xpiDD a-icoi' Nsnvi 
ncna, F: spiDD trim*? nanw, RNZB.: np~idd nipd^ Nanxi 
Ntjm; HG., ed. V., 85 b: Ntma npiido Nns^cij namn, B., 
626: twa NpiDD Kitwi> ton». After tnmb 'to begin', 
NplDO can be nothing else but inf. (lengthened through an 
N, see sub 61), but this is contradicted by the old (Syriac, 
see PSm., s. v. p~\o) form NpllDo in HG., which can only be 
a substantive and denote 'comb'. Now INI 'n, 103, has 
instead : '13 nptoa pilDD ybwnb nansi, hence he had before 
him the correct reading snE>»!> , from N1K> ' throw '. 

Ibid. ■. 'yon ntvvix wb junopn xi> n^roi? ni> nha '31 
'13 SnEpon, F and E : WTOX ew6, HG, ed. V., *'&#. : 'twn{> 



P1[d]p , dt pact NnN'xre, B. ibid. -. p-itapw pjon nnN'vre 'cn^ , 
in int 'n : nvxre wii>. xn^vre pi. Nrnre (Nnxvire), as Nn'm 
iND-ixn Men. 42a (= She., No. 127), Hebr. »e*n rwv, Syr. 

Md rabd. train ^3,203 cites: <55S55 nnpi nto nEJB" 'Diom 

'131 N'TD. 



360 THE JEWISH QUARTERLY REVIEW 

MTSW 'lock, ringlet'. ^ni? in the ed. and HG. supports 
the pi. KnjjfW. [nnjom t^-i also Mand., Das Johannesbuch, 
ed. Lidzbarski, p. 165, a.] 

/&#. : m5 nS uvf sbi 'n *"by Tnon t6 r6non to-pjni 
'131 n«r» ^'J? ptapw— read with E, F, and RNZB. : 
■oti *6i 'id (insTi) in3n. 64e wi (Levy, j. w.) pi. of xro'n, 
Anan. Hark. 8 : unaTi, "OTi, Mand. 'thong', Syr. Nron, i$j 
(Noldeke, Mand. Gram., p. 103, comp. also 17a) ; as to 
m3 F has the following gloss (which is derived verbatim 
from 'Aruk, s. v. Km, to Hag. 9 a) : 'dn!3 nyisT '»a 
(read pr~QN) Jinn IN mvm jotPD Kip»j>D; later, however, 
the quotation 'ifl Nlp'yo was cancelled. 63 A correct ex- 
planation was furnished also by RNZB. in rbxw poyn. 
p-iop^on as above ; but RABD., /. c, cites : Wapi OTi t&\ 
'131, likewise wbcn 'o, Bacher, Hebr.-Pers. WB., 44, 5: 
N'JD 'iW }>D»pT ni3K, with explanation: *U*U3 ^3 b*0 topj 
'ID ^N, ' stilus (!) quem super crinem legant' (Bacher). 
cop Talm. ' to bind (vinxi), Git. 47 a and Sanh. 95 a (with 
the gloss rr>nS3, see D"i): She., my Fragment, JQR., 
1914, IV, 42° 20 : 'T^ poop Npi and 42i 2 (comp. JQR., 
j 9 14, V, 97): D'Dpl ■TTDB'W anto 'to compel'; Syr. Dop, 
wVm (see z%?., n. 14 and 18); Arab. Jai 'to bind', and 
also specially: to wind a fillet about the head, )?W\n 
or poopw is therefore probably the correct reading in 
She. instead of (ncp'OT, which caused piDpW, mentioned 
before. 

Ibid. : ab nipo 3jn i^rxi itt tamo aa« »n"od pratpoK w 

646 Likewise cited by RABD., ibid. : >yf\ xbl. 

66 prOM really does not belong here. Correctly 'Aruk, s. v. pt"l3K 
(probably from R. Hananel) : JOpj 1D1B p^31 DID bt? fD'SD HD3D 60H1 
pT"13X, °*>°J->I ( see Fleischer, Levy, I, 277); Gaon, Kel. XXVI, 7, s.v. 
rUSiyn, translates pn3N with .' tOID *|j>ij> ntn N3E>D. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 361 

i^DS, likewise MS. Petersburg (ed. Wilna, III, pcyn nna, 
p. 12): SCnO 3JK W13D tUWDK W, F: tOTTD 32K 'KT3KD 
mpoij (!) '<BK1, RNZB. : ^Wl (!) Wfia as ; but also RABD. 
and SMG. niKlpD have K3mo. Nevertheless tailO 3JK is 
probably an old error by a copyist (who had also abbreviated 
the She.), caused by the following: K3T1D n«K |W3t5>DK W ; 
indeed it is wanting in HG., ed. V., 85 c: W3Q JUWDK W 
mpd? )bm mi (=B. 627), ixn 'n, 104: i^bji pnb wtpcn dki 
mpD^> . Besides, our She., as already indicated, is considerably 
abbreviated ; for the tamo n«K pratPOK W was originally 
an independent 'question', tfja, as it is still preserved in 
Geonica, 371 : Np^DD rmpob 1DK ill K3riKD tfD T^DK WK 
'13 n^3B reb. As to wn3 of the MS. RNZB., it is hope- 
lessly tangled and against all the old witnesses. — 1PIN1 VTT. 
Instead of AT HG., ed. V., has mi, 65 " read mi from tni 
' draw ' (of water), ' flow ' ; Targ. KJTJ = D^TU. 

68. No. 100 : tdn wn Nrwpi nuoxi naan xna^p p» 
ijon Disn •'Jim im inrok »im wm tma^p ,inri>is?3i -otfoi' 
'id a^y-uni dvuk »a^p i^b> pm ayioi ^p pm ,| *vdn, F: 
(on the margin WH D»an W31 '13 1D3 ^1B>3l) '13 iiwi KflB^p ; 
HG., ed. V., 16 a: 1m ^iboi '13 antwpi tuoni nrtn sna^p 
won dish '3131 '13 (=ed. B., 643), RNZB. : pro ^iboi; "vntni, 
II, 222: mus iw dhpb> bv "B^p pas. Read: (nren) nattn 

'13 yon D1BT T31 '13 W3 ^31 '13 .Hit? = DHpB>, Syr. WW 
'almond'. The error with "I is very old, and is mentioned 
already by BB. (see Low, 374). won D1B ' blossom '= p, 
BB. : pnVi KDia n twom NfiBl3 (see Low, 364) ; 1313 pi. of 
tWW3 ' hair, small hair ' (as Shab. 67 a). 

Ibid., ed. Wilna, II, 117 (from MS. = E and F) : ntri 
"p3»n wi Nomsn ir6, likewise E: '13 NDman ir6 nen! 
F: inanyn -s<n ^ Konnan STpSro ir6 nw; HG., ed. B. 

« ta RABD. op. at., fol. 22 a : iJTT. 



3^2 THE JEWISH QUARTERLY REVIEW 



(probably from the Hal. Pesukot), 643: KjppB>D3 ; Temim 
DeHm, No. 13a, in the name of Jehudai: WpBTM; R. Jeruham, 
XXI, 2 : two 'PBT53 (!), the Comm. MS. RNZB. had Nrptyoa : 
D*nan ni"va la-ow oti 13 man ppiw Dipo3 p^tyoi. On 
NjrpBTs comp. A. z. 37 b: Niruan unypnya xns «n. 66 

/&#.: yau rra nTiKi Nomsn kmid it5> mm ;ko "wii, E: 
NDTTM tWlD. The Comm. Cod. Berlin, fol. 223 a : iiw 
ml d""1 ntjnn njptM NDV1B3; the latter is also the correct 
explanation ; Targ. tWD , Yerushalmi KJ"D and NVHD (reading 
of the Yalkut, Levy, j. z/. WD) ' a fence '. 

/&tf. : |V3 ND^N '13 S|5>X tfOTH '13 Kin NJ^N nms iNH WK 

ton p"V .mntya hstwi'D xpn, E : mnea i^srvKn, MS. (see the 
marginal gloss and p. 182) RNZB. and F: mw3 i^snw 
(in F mTiBO 67 is cancelled and rnvrea written over it !) ; 
read : nw3 ^awvo xpn (nW3 S^arPD-i) ' since it becomes 
consumed (dry and faded) within a year ' (tarPD = ^3), as 
HG., ed. V., Berakot 7 b (B. 56-7) : nxai HoaS> Ban ^v nn 
'13 ventre, comp. ed. V., 8 d below (= B. 72), in the name 
of R. Ahai (missing in She.), urns is ' caper-plant ', Syr. 
Nrna (Low, 265) = *|i>x, NSW; Ber. 40b: nma — naxj, 
Responsa, ed. Hark., No. 378 (extracted in'Aruk, s. v. s^x) : 
Nma W jwfcai naxi Kim s]i>x sin. The Gaon reads also 
Ber. 36 a : '13 tmjriK WK <yotu sms ''an IDS (comp. in 
addition Levy, s. v.) ; HG., ed. V., 7 c above : mm 'TO 
Nrnan Knit5>x (= ed. B., 57) ; V., 8 b above: N113 Nrna npy 
noian na (ed. B., 57 : '13 nmai KB-a!). 67 " Further here below : 
Kin tn^ 'tan ntpBu ninjn 3"td SFnan to'a faai ton no'o 

66 Comp. Aptowitzer, MGWJ., 191 1, p. 637. 

67 F has also below : 13 'XT n"ne> (over it: NIW3 !), likewise the 
Comm. MS. Berlin, fol. 223 a : D'JH jni '^Kfl pO fl'Tlt? ! confusing it with 
piVB>, Ber. 40 b. 

67a [See now, J. N. Epstein, Jahrbuch, Frankf., XII (io,r8), 121.] 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 363 

ninjn a"n»i win tn'a snts .ltd 'id sin kts in^ wit xmc 
'13 na iri>a n$> 's. Concerning Kn'Dia and anw see Low, 
/. t. ; Responsa, /. c. ; and Levy, s. v. — srrvB is represented 
in the sequel by pDnsp, the ' caper-blossom ', Talm. Kma, 
Ber. 36 b : sn»Bia D»p>Kl tttvom xrneb mbpen , according to 
the reading of Rashi ('Aruk, j. w. KJVDia, xma : tWib rnbpen 
xmaT 'the blossom of the caper-plant', or 'of the fruit of the 
caper', which is also called xma, Ber. 36 a: WN 'yon pj^s 
smsn Nnjnx). ___ 

69. No. 10a: Tin i^iav <nn ^pnc cpvvb, F: dpjdS 

■•snn yno, margin: ^pnai wd, E: pxnn ^V3i }«mn ^pne ; 

Anan, ed. Sch., i2 20 : ^ai ^NpnD. s^»a also at the end of 

- f 

this No.; Talm. t6»3, Arab. JIS" (PI. JjJS"). ync pi. of 
wno B. b. 89 a (here below) ' line or chain of a scale '. 

7foV/., end, read with E and F : Nap reb npi SFap rr6 npl 
'13 pjc^r (F : pJD'I t6"3 xap, margin after sap : IK). 

/fo'^. : 'ia aba WTDK &6, E and F pes, likewise 
No. 117: SrES N^ Ss-icni, F pDN. But W*idn is not 
a scribal error : Onkelos and Palestinian Targ. WiON , 
Yerushalmi pTDN (Dalman, 345) ; Anan plot* (Hark., 5 16 , 
i5 8 , &c.) ; Geonica, i3i u wics, 1394 piDX. In She. also 
No. 103 F : Sn55 Np nD'D (see above, jz^ 20). 

70. No. 103 : so^va ipinni sin no rrnoBo psj «in j"j?n 
iThidbo T3 SprinBn sin. In Babli also Ned. 91b: prino 
NJV33 (pmriD), and B. kamma 85b, according to the MS. H, R. 
and Aruk : (U1TOK3 Tpnm (ed. rr>pnm) ' close in, lock up '. 
Mand. ptin 'chain', Arab. j^a. and jj^, corresponding to 
Ptn. That ptin is 'Nabatean' is borne out also by Gawallki, 
Al-Miiarrab, ed. Sachau, 52 : S)A*^ Jy^sJl JJls .I2IDI} 
Ujj^ &! Jlw ^^Jij (Mu'arih) Jl5 4$!b, comp. Noldeke, Mand. 
Gramm., 60 and 85. Here belongs also 'pmn 13 (=ipinn, 



364 THE JEWISH QUARTERLY REVIEW 

<ij>* ? )> Men. 33b =ijre> Ira 'a small block-house' outside 
the gate (' a watchman's hut '). 

71. No. 104: irFopyo ptwp iin sin, E w»np*j» (F 
wanting). K3py instead of Nlpy Naz. 39 a : WW31 N3p<y 
'hair-root'; 39b: pnoy EpjF; Hul. 52a: 'an 'Op^jJS 'at 
the root of the wings ' (not ' ends ' as Levy, s. v.) ; Anan, 
ed. Hark., 9 : s5j?j5 mn by [. . ,]k pin^ KBiro jirfc j:nm 
' at the root ', i. e. at the bend (in contrast to the ends of 
the threads) ; Geonica, 27 : ijn in"3p»j» panel kdiki NJ1S 
IiTsid ; HG., ed. B., 62 : raira »jnp>K topyw '»vd »nd, 
ed. V., 9 d : 'id wpx snnon nxtna pi ; Ber. 45 b : Ijnpsn 
inp^jJD .I3in3 V\b; and further in She., No. 137: D1PD 
NW Din ib snpwt, F has in the text: jnT3 pl»1, which is 
cancelled, and over it pnnpjJDl, on the margin nipyoi, 
E p.n'J'tt p^ndi ! This is certainly a corruption from 
prvap^Di (pwapvoi) with 3. 671> 

72. No. 105: SHp idiid^ '13 wip •>Dl-\ob, read with 
RNZB., E, and F wop ; No. 442, E has: Wip nab, 
F wopi nob, Geonica, 368: wop ; No. 145: W'opi nm, 
and No. 162: rw«yp IN; HG., ed. V., D1, 133c, above: 
wyp ^3i tut? «i»i xnm mb> s6i; ibid., nni33, 138 c: 131 
'13 wyp 'dis ne»»n '13 nidjji "torap ^nib" nnb rp$> mn 
(D1S ' mouth ' = ' head ', ' piece '). WOp also in b. Sotah 
34 b (Reifmann). 

Ibid. : \rb P^ds xp wn s£b«d n^ rpi>e> in^ tiki iwi }W3 
n»J^> (F spS'ttn) pa^EH ptr b s6 pnw by '13 pDans. To 
b'm the Comm. MS. Berlin, fol. 239 a has : ,mnM 11PJ! i»e>n 
U11B* Vina »TJ IDS. bwn properly 'draw, draw out, tear 
out ' (but still let hang), Hul. 93 a : mb^n »y»3 ' testicles 

87b In this She. in Tanh. npn 2 : lTpnpjJDI, ed. Buber, 28: liUnpWt. 
See also Addenda, No. 23. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 365 

torn away but still suspended from the body ', according to 
the correct explanation of Rashi (comp. ^n ["na up ttbio 
iroy Tin p -an), epi>tr 'torn out totally', as Shab. 152a: 
NS^C Nip 11 }? (Kpw) ' castrated he-goat ', comp. above, sub 50. 
Ibid. : yinb pruw "kd-ik pajjn nin yn, E ini> ^didi, 
F pnrv pnni, margin : pDnoo ;, a. «D ' cut off', Levy, I, 
205 a = 5>D. 

73. No. 107 end: nwa&J? WD, read with RNZB., E, 
and F rvi'wnw'?. 

74. No. in: 'la mena mn ''mb ;jdd xpn sen ii>w 
HTn larm, F: mnn 'ami '13 mn 'axi> judo, mn •'enb 
nowhere else except Anan, ed. Hark., 83, 88. '"Hill is not 
substantiated in the Bab. Talmud. 



Ibid. : 'la Npn'Ni njtjjd nnax 'ni, F nprtn njt>kd (E 
nptvn Hebr. !), RNZB. : apnnn kjtj» (Kg\tn). apr^N = NplirS. 
iQtn with N in Babli only in NJTKH ; HG., ed. B., 107 : uni 



75. No. 113: n^nD 'in ini>K ttpno T>V3 nj?D wii. In 
the Bab. Talmud only Txbvb or "in^ND ; HG., ed. V., 6 d : 
'13 T"DDT -ini>Kl ; but also Tami d 32 b , MS. Florence (in 
Margolis, Grammatik, 37*) : Ppn "ini>K. 

7#z'd. below, F : xbx pro xb iTond cwk pan Ira no5»n IK 
rth wavr n^i n3D k^n nit? «!n WDnx 'on e>nao3 >w wa 
pana N^ 'N, from MDnx on cancelled ; ed. V. for WDnx: nasi 
N3JJT. N3Dn« ' is it established ' ? 'is it true ' ? 

76. No. 123 end : '13 n^s?3 pi iy urru patrol po, F 
JWdT, RNZB and E praon (E perhaps pnjDT ?), 'Aruk, s. v. 
nJ: 'mDI, from nJ3 = n33 'delay, tarry' (not from rvi, Levy, 
s. v.). Mahzor Vitry, 394, cites , »n _ i3 'nasi \n ; Siddur 
Rashi, p. 100, No. 215: T301 ; Hag. Maim. D'as riK'EO, 
XIV, %, letter a: Wfl wiH; in TSBZ., No. 134: panrr ; 
CW. Hayyim D"B3 riK'SM No. 1 : fJvnruDi; read vwn, pnjJl, 



366 THE JEWISH QUARTERLY REVIEW 

pTUJOT. 'Aruk knows also of another passage in b. Nid. 
65 b: *bb m tudt ruins ■okb>, ed. ■ujdi, Rashi vwm. 

77. No. 126: , Eini' fD'Drxn pan ni¥D?> ■oidtn mown nsti 
'id jw»-i n?> pnjn, F: ion t6 pijn "tainb jiritn jus, E: 
P^di 'la jwwt, RNZB.: pj'Di t6 pnyi "Bin?> (!) funBH : 
read iWIBH with E and F (RNZB. probably misread it), 
from X1W 'weave, spin' (see Anan, sub 1), parallel to 
wye>ir6 rmp, ra\xh rmy. 

78. t3p?? — I3P3 . No. 128: Wpb IN NfDBH NDV3 i>B>3m ;n»i 
'13 *"VB, F: n»a D^pT IN, likewise HG., ed. B., 114: iNBl 
»TB Bpjon IN Nmea i>B»3Dn (ed. V., 21 d : D^oi) ; Anan, 
ed. Hark., 72 above: NJ11HN '•pIDBi NpT 'Bipj. Bab. Ep3 = 
Bpi> ' gather ', Ber. 44 a : b^tb pnv '*n mm p^TN nm "a 
rrm mars im nn W nb jPoplo dno '•a p*in o ">dih m-r-a 
'13 ; Shab. 142 a : on»3 wHpHi' |3N?> hb*ji n , al' iruHB»i>i ; 
ibid. : wapj^i irunts^l (Rashi DBp^) ; Pes. 40 b : r\b NBplS 
,| 3nN3 rWK ' picked (gathered) it (the wheat) up in troughs ' ; 
B. mez. 83 b : 'NBipi (with the variant in 'Aruk *NBlpi>) 
' snatched together, gathered together' (= Dpbnn Jud. 11, 3, 
Levy, s. v.) ; Yeb. 42 b : rroD arabn Stni B^plE ; Abot di R. 
Nathan, XVIII : vss> na" 1 Bp~JD ' gatherer '. On the other hand 
Mand. BJi> (= tapb) = Bab. tapj 'hold, seize, take' (Noldeke, 
Mand. Gramm., p. 54, n. 1) ; likewise in the writings of the 
Geonim : Responsa, ed. Hark., No. 65, p. 30 : NJNruo Wpbl 
(= 'B'pn) ; No. 232, p. no : n-|BB> Wfbl JNBn (n. 6 !) ; No. 65 
(/. c.) -. ntn 11263 Noi>jn Nnxmn ndibd wpb pam ; No. 241 : 
pan inb ^"ph 1^3 j No. 386 : p'aph Kin N^3 pm (n. 6 ! = 
P'BPJ frequently in Bab. Talmud) 68 ; No. 376, p. 193 : INT 
DQyb n3J ?xb (= nop:) ; Responsa, ed. Cassel, No. 92 end : 

68 No. 387, p. B02 : '13 nyiDK'3 JJ3T ''B'pJI ' s probably, as in many other 
cases, a correction by Harkavy ! 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 367 



nnco ;d Nts*pn pBph DnnK xbx; Responsa mua mon, No. 
161: wonp Disc 'B'pin; Responsa pIV nyt?, IV, Porta 4, 
No. 31 (fol. 61 b) : snsm |D pan rr>b ptrph; mostly responsa 
of Sherira Gaon ; in his Epistle, Neubauer, Anecdota, I, 
p. 11 (MS. II, n. j 9) : TKD 'Ti nmx tuiaim 'i (opai =) dp^i, 
zfo'rf. ia: na spmm 'n nt2pi> •an dicd, 15: pan lrwDp^. 68 * 

The change of Dpi — Dp^> corresponds to that of (Mand.) 
XDni> — KDn: (Bab.), N"U'$> (Mand.)— NTW (Bab.), jbn (Mand. 
and Gaonic) — yn (Bab.); but both forms have the same 
meanings, ' take, seize ', as well as ' gather '. Dpi in She. in 

the sense of ' holding ', No. 96 : '13 W Dp:n^> K3"riD <D ; 
Geonica, 370: 'ia rwiDi w Dpi'D^ KnfJVK K3«n]Di; No. 46: 
w nn opjn^ ncs's k5h, and 77: kds B^pT spn. 

/to?. : niyaD r6in i!> nvi t6) nybi )b rrpw paa -iidk ^>ca» 
'13 nona^ nn *»a irpjsa wrn ino nv (wanting in b. Hul. 
15 b). KM in b. Ber. 57b: Kin lrvoaa Ktn *3 'at their 
stem ', and z'to., 48 a : JPT iTMD pxia, accord, to the reading 
of the Geonim (Responsa, ed. Hark., No. 395). 69 Syr. e. g. 
Anecd. Syr., Ill, 35,, : hm p option K'rn ktti T« ; GV$>., 
97 4 (see Noldeke, Mand. Gram., p. xxviii, n. 1). 

79. No. 139 : WD 'ia rwo nwss *nmai> -ya rvi> etvd^ 
irons .TtPW nimai n^ etpo!', Pers. j-*Jj, Mand. and Syr. 
rmKS, trnB ' be careful ', ' heedful ' (see Levy, IV, 103, and 
Fleischer, 228 b ; Noldeke, Mand. Gramm., p. 272 ; PSm. 
3241). In HG. also mms with T ; HG., ed. V., 85 c : nanvi 
%i2 bnar\ tab ncx 'ia 'otn '"kmi^d KE>aa nimai>, ed. B., 627 : 

68 a See Addenda, No. 24. 

69 'Aruk has also Ker, 3b: prPM3 |inn'' WD, ed. JliTOi'aK, p"W 
from MSS. prP3J133. 

70 [r. KMitFD, with ."W^Ofl 'D (s. Horowitz, J/GWy., 45, p. 86, No. 98) 
'slime', Mand. KM't?, Assyr. Siknu, s. Lidzbarski, Das Johannesbuch, II, 
52, n. 4.] 



368 THE JEWISH QUARTERLY REVIEW 



'pinned (!) ,int 'n, 104: mosn p noxy prnni> nansi (nios = 

80. No. 130: man nro wm xn^o rr!> n^sn jnd ^»tn 
pinna ion yoen p n»ai> won n^ aw, E : tow ion u rvi> 'an* 
Pinna tun yot?l (similarly RNZB.), F : (also RNZB. 3VT 1 ) 3W 

nnna ton yoch wn 'a wet n^ (the n of yocin over the line). 
Similarly in the ' Gaonic Decrees and Documents ' by 
Aptowitzer (JQR., N. S., IV, 32 ff.), I 4 , II T , IX 10 . 

81. No. 137, read with E and F : yvt '03 SE 'Km 
Kin 133 iTHB«d{>. 

82. No. 139 : 'la D'n ronoo nx jn^ xa ejim. biidi = 
e|lD3h ; similarly Ker. 6b: '13 f|D3 317133 ejitJT, and 8 a : 
n'i> naao fjioi ppfo a won . 

83. No. 143 : pdsj^i rrriiovsa "lini&i rw«jn ■Ttioii' tdki 
winaea pnp' i>m 'ia nosn u»"p np xpn wtjd n'nna^a, 
E likewise: n'riB'tra pDy'Di, F 'la 'tisc>3 poj?'Dl ; HG., ed. B., 
33 : vniriBca pip n5>i 'ia n'nnatya j*pjwT, ed. V., 2a: n?yoi 
WBBa y\pv vh\ '13 n'nniBB>a. Accordingly we should 
probably read here : Dpyo^l (opj? = j»pp, see 'Aruk, s. v.) 
' bite the lips' ; comp. Mik., VIII. 5: iTmnae* nsnp. If poyo 
were correct we would have to compare it with DTiBC 'poy 
in Isa. 33. 19, but this is hardly certain. 

Ibid.: .Tn^oa web rvb 'JD'D K^i, F Dj'oi> (E wanting), 
RNZB. D3'0^ (misread from DJ'ofy ; HG., ed. B., 24: wo!? 
(=OK)b), ed. V., i d: fi^'o Djej'b ab) ; the Comm. MS. 
RNZB. : B>"p3 DJDji> ffb 'jitd kS>i (see nW poyn). DJ'D inf. 
of 03, D3D3. 

84. No. 152: bvnxm irai turabn pi (likewise in the 
manuscripts); in the She. MS. St. Petersburg (ed. Wilna, 
III, fol. i"'p) : rnn w"3jn ^Kne»3 mam una^n pi (see poyn 
nW *fe/.) ; b. Ber. 45 a, &c. : nan NDjf 'NO 'tn plB. 70a That 

70a [See now MGJVJ., 1919, 257 f.] 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 369 

closing sentence is preserved once more in Hebrew in 
"vntm, II, nriN (surely from She., but is wanting here) end, 
p. 164: btrtpi njnu na^n pi = btnwi torn xroW pi. 

85. No. 165 : 3*n nmr ^3 win mk '•on ni n*n n"x 
'13 iini njw (E naiT 'twoo, F wdit ^3 'Ncno). The whole 
is wanting in Ket. 106 a. Comp. especially Hul. 49 a : 
nin d-is-it ^a itrno N3N (Sabora!) ran '-n nnn J>kidb> -i"n 
'13. In Gaonic literature we find the pi. in addition nys? 
Pix, fol. 17 b, No. 7 : jo (^3 =) ^3 n»3 rb »jn xn^D Nin 
rfa E>sn i>Ki»tr 3T •>»? ; /#«?-, I, pBTVp, ed. V., 108 d cor- 
rectly: ^3 n»3 1V3; in JpniS DIP 11 : ^3 W~b. "1D1T refers no 
doubt to ^3 'small Kallas' (not 'small Reshe-kalle '). 
A ' great ' rfo is mentioned in HG. |»tM, ed. V., 79 a (ed. B., 
325): '13 T)3"i ni>33 pto Win 3n "id ''cp J33T nn, correctly 
rfaa in 7//«r, niE>^E>, ed. V., 29 b, 0r Zarua\ I, No. 724, and 
pnv nyt5>, Porta 2, No. 6. For this "mi ni>3, however, R. 
Samuel b. Hofni Gaon gives t^3 Knn3, Adar or Elul : 
Responsa, ed. Hark., No. 312: lWe> '13 ni3^>nn bin an3 ^>3N 
x^3 cnn3 pro Win an id nx D^Dsnn, comp. HG., ed. V., 
138 d below: amino en won 3i no t^nm ■'tyi-vs i^n 
ni>3n xniw3 (see above j«# 1, comp. above, p. 301). [On the 
other hand this ^rm ni>3 is found also without the Metiba. 
R. Sherira in his Epistle : nan KWM '•oa 1»-|JVN1 itnn^ irtm 
TOT nba sirw, 'probably, as suggested Mann, JQR., 1917, 
466, they held this gathering in the presence of the 
Exilarch (unib B""n tbil !) who must have had his residence 
in Bagdad '.] Here then it becomes absolutely necessary 
to ascertain the real (etymological) meaning of the word 
xbl (X^3), about which so many attempts have been made 
(see recently Halper in JQR., 19 14, 546), but all of 
them hardly satisfactory. The meaning is demonstrated 
VOL. XII. C c 



370 THE JEWISH QUARTERLY REVIEW 

clearly and surely in a passage of the Gaonic litera- 
ture, which, as usually, has avenged itself fearfully for 
the neglect in which it was placed. Responsa, ed. Hark., 
No. 312, has wi t6a3 mniN instead of the Talm. (Ket. 
60 a) *wn ma nanix ! Accordingly vbi (rta) = xm ' row ', 
nw, which also fits exclusively ^3 in piv njB? (^3 n»a rb iva) 
'several rows'. Now the r6a K>n, in accordance with his 
name (' head of a row '), was actually the leader of a ' row ' 
(rnie>) of ten men, see Nathan ha-Babli ; Ber. 57 a : mr\ 
ni>3 yai» parallel to na'B" twn. Etymological ly it belongs, 
as was already suggested heretofore, to N? 1 ?! 71 ; hence K?3 
literally 'something round, a circle' (comp. nrvn jmnjD 
rb)lV pti ,| sna!), then the 'rows of scholars' in the Metiba 
and the months of the fpa , i. e. the months Adar and 
Elul during which the scholars filled up the rows and 
occupied their definite places (otherwise they were not 
so exact in the order of seats, see Nathan ha-Babli), 
hence Talm., rr?33 rfa »n (B. k. 113 a), rfan w (M. k. 
16a), ftal ktwib (Ber. 8 b), rta naooa (Taan. 10 b), Gaon. 
rtai tcm , J 'nii (Sherira), rfo tnn, tiki rta, Ww fta. 
Then n^a is also used for the more restricted Metiba, in 
opposition to tw»3"in, b. Men. 8a b: '131 '31 nBK KWina 
-OVN t6 ni>aa (so Aruch, j. s>. pain , edit. N^aa) ; further for 
the single lecture (see below) of the Resh Metiba at which 
at least one 'row' of ten men had to be present, b. B. b. 
ia a: runs r6aa prima }W -sh wm am nna mew *3 ii>m jva 

71 Possibly the common Arabic ii£, pi. JiS<ball ' (properly ' something 
round'), Dozy, II, 480a below, should come in here, and not, as the 
lexicographers maintain (see Fleischer,./?/. Schriften, II, 769 ff.), a corruption 
of ip, which had migrated into Turkish and come back again to the Arabic 
(Fleischer, /. c). It should rather be compared with our rt?3 and placed on 
a par with JJitf , N^3 ' wreath ', ' hoop ' (see Frankel, 62). 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 371 

mwyo (similarly Yeb. 67 a and Zeb. 92 a ; rntW '3 **? Ti«N 
'131 in«»p -£ now), Suk. a6a: nbi> ^yi at '3 -a D^ytna. 72 
xi>3 en is therefore, as stated above, as much as n"Tie> wn , 
a title which was given by Rab Sherira Gaon to Rab 
Shemarya b. Elhanan : 72a iNjmnj mit? EWi {JQR., VI, 323), 
which would correspond to Aram. »Kjmrm &6a t5"i. Now 
Rab Sherira states further that the ' row ' alluded to is a 'great 
row among the three rows of the Yeshibah ' (iniJDB> vb p »W>1 
nsw bv nmw vhffo rr^vu mica e>x-ii> wudp &6i nje>Di> «5>) ; 
a 'great row ' xa-i mi is mentioned by R. Sherira also in 
his Epistle: twmioa am x-na iT3dd ;isj icao an -i»i 
(Neub. I, 41), and by Rab Zemah b. Solomon in his 
responsum (Ben Ckananja, IV, 141) : N3"i NTT) WE>p jm 
nhtj mnjD mpoa Dnc (= 'ia one ttaiDDn D'taan in 
Amram, Responsa, ed. Lyck, No. 56, see Eppenstein, 
Beitrage, &c, p. 15, and appendix, ibid.). In the Mishnah 
Sanh. 37a we read: vbw '13 nisijy pi] »xro nnTi pnruD 
-irixi nnx b (before the jHimo) Drvjsi> pan* n"n bv mw 
njitwnn |» patsiD "jiodS la-rain iDipc nx "Vac; Tos. Sanh., 
Ill, 5 : $>e> nmc 'a na na tw "vy !>a -ioik mw 'i rr>n pi 
pmnjD mwh (n'w-i) px-i 'ia n^tn anew entry (ed. Zucker- 
mandel: nB^tfi DntJ>y $>B>), and z'foV/., b. 17 b: nvbw entry 
Tinm own nn ne^Bn entry inr nmtr tri>tn n:tjp mnjD njja. 
They are mentioned also at a later date : Gen. r., 70. 8 : 
nmtr nvbw ibx jnv my ne^tr aw njm ,jmnjD it mtra tsa nam 
'ia Brwei> ducv ontr D'oan n'obn 5>tr ; Lev. r., 30. 1 1 : yy na 

72 INT 'n renders the Talm. ni>3 often by nTW : p. a6 : myt^ST myiTa 

(nbb ^ipi) nanr^ "jin nan, 64 : nyvn ^a n5>a ^a^ jot ;ru t6 

(n^3T W3) ; likewise R. Natronai, Ibn Giyath, II, p. 69 : iniN fTSDO 

■nx bw na^oi bibs btr iirw = m«T r6ai b^btn n^a. 

72a ^ N y 1 - |nJ) wUh Mann> jq Rj , 9l8j 35a j 

c c 2 



372 THE JEWISH QUARTERLY REVIEW 

annbn bv> m-iiE> rrekv "bx mas py spyi 'ia nSna mruo it Tin 
'is Dnijai' pawt?; Cant, r., 1, 15. 2: n^n !>e> mw 'a |n ia 
'ia -van nnxi ttin ^a Dnoa^ paw frreo D^an ; /&#., IV, 4. 6 : 
nya» noii> ins pro nin po-ia miri epvn rime e^eat? jpnn 
nosy jmn3D3 pawn by TTi»vi>— and were probably an 
essential in every great Yeshibah in Palestine as well as 
in Babylonia. These are then no doubt 72b the ' three rows ' 
mentioned by Sherira (nyv* b& nrwn b>S>b>), one of which 
was a ' great row ' (rbvn n"W). They consisted (according 
to Tosefta and b. Sanh. 17 b, see above) oieach twenty-three 
members. The first row, of ten members in Sura, but of 
twenty-three in Palestine and probably also in Pumbedita™ 
was called sal KTi, nisna rem. In contrast to this the next 
two, or six (which in the Yeshibah of R. Huna probably were 
followed by several others) must have been called noiT v3 
'small rows'. However, the order of seats and rows in the 
Gaonic Yeshibot still requires a thorough investigation. 

A further nuance of this meaning of t^a 'circle' is 
credited to Sherira Gaon in his Epistle, ed. Lewin, 94 
(see his note) : no vnn rhi ba «urfc pan wpn »am 
pjp prfcro n^ia nmoiri -mm nm non KrwanD, comp. 

72 b See also Ginzberg, Geonica, I, Additions, 207, and Marx, ZfHB., 
XIII, 72. 

73 b. B. k. 117 a speaks of seven rows in the Yeshiba of R. Yohanan 
(Tiberias), the same is told by Nathan ha-Babli (of Sura !), seven rows with 
ten in each, which were followed, as Nathan expressly states (~\XW Dyi 

e»N niND 'ia on& Dnna^nn, further arrnnK paw tmn^nn bai), 

by more benches, an assumption rightly made by Frumkin, in his intro- 
duction to may an no, i, 10, from the phrase m jne> mnx rrnruK 

Xnna KTT3 n^aniNT I}). In Palestine, however, the rows seem to have 
had twenty-three each, together therefore sixty-nine and including the 
na^'' K>SO seventy (originally, it is true, the three rows were not included 
in the seventy). It is not impossible that Pumbeditha likewise had the same 
arrangement. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 373 

to that S. ha-Ittim, 248 : rwiTBa XT-it? an ID nDXI 
pan foi^ fUPfi ^n (in the explanation of t6oi aniens) 
xiitdi rfoa rrnnn ">xn jmpt? jruoni nb nn rwiri r6a '"b "\-ipi 
'la »B«pn nba «mi. Properly 'a cycle" of twenty-five 
weeks, ' a semester '. 

86. No. 168: io-viSS pa^ n^ nt? ns, and below: 
xaiua pa»»5» n»; HG., ed. V., 30 d: »a"iu dDto; Anan, 
ed. Hark., 117 : unu D^D »tn. Wiia, Syr. from the Pers. 
i_^p Arab., ^yL 'stocking, sock', see 'Aruk, s.v. 3"0, II, 
and Fleischer in Levy, I, 436 a. 

Ibid. : rvb ne> »d Nanpj? xawi 'nani* hre^ »jn xpn jkd 
^dd dd'dS>, E and F toan xawi ; HG., ed. V., 30 d: $>yn£ 
xanim xnana, ed. B., 153 : tnarn xnanf>, likewise xnan 
cited by R. Isak Ibn Giyat. xnan, Syr. (1) 'darkness', 
b. Nian, Mand. Xixan (Noldeke, Gram., p. 61), and (2) 
fossa, fovea: noxa xcwan x v n "inoi> xint rvin xa^x xan 
xjnxa xnam xaui xdjki xnns, PSm. 1185 [see now Low, 
' Lex. Miscellen ' in Hoffmann's Festschrift, p. 124, No. 8. 
The n is — against Low — no scribal error, but stands to the 
Talm. xian, 73a Mand. xixan, in the same relation as Gaonic 
tmrb to the Talm. xoru, Mand. K»nS>, see above, sub 18]. 
For "3D» DD^ (E rwxDO), F iTOXDD nwh, HG, ed. V.: 
xB>nn p ix xa-spp p xjxdd xi>a ^jrn^ Sm xpi, ed. B., 154: 
'ia xExo xb. xdx ,, d = X3K<d (not xdnjd inf. Pael), xjxdd, 
likewise Neo-Jud.-Aram. nrio^D, Hashiloah, XXIX, 319, 
1. 8 from below. 

D"D in Babylonian frequently in connexion with X3NDD, 
once even xd^dd^, Git. 56 b : xd"De6 xya kjxdd "in d^d mn 

73 a But also Xian, Hor. iaa, likewise Ker. 5 b Fragment Schechter 
(5 2 ,): Xnan H XJT33; Xian also 'Aruk s.v. 13n oldEdd.and IJTlD i?nX, 
91, sub "|BTI. Anan. Hark. 77 : XT3n (KnX3fl). 



374 THE JEWISH QUARTERLY REVIEW 

tonnt6. It is equivalent to PD (denom. of k:«d 'shoe'), 
Syr. I'D, Hebr. ?**p (Isa. 9. 5: |N1D). \*o was changed in 
Babylonia to D^D with D through assimilation with the 
immediately following WNDD (but has nothing whatever in 
common with D'D, D'C), and this latter form had almost 
crowded out completely the original form with }, so that 
people even said NWSN wdo (Taan. iab), UTU DD'D, D'D 
131U (HG. and Anan, see above). Nay, it even reacted on 
the substantive KJtPD, remodelling it to NDN'D (see above). 
A trace, however, has remained : the i is preserved once in 
the part, pi., in M. k. 24 b, where R. Solomon b. ha-Yatom 
has in his comm. (p. 112): irvoKDO iJNDDn , with the 
explanation \rfbv20 pi>yiJK> (see the note of the editor), 
R. Hananel <3dd (= *3dd), HG., ed. V., 43 a : kjxdd ns^oT. 

87. Finally let me state some peculiarities of the 
manuscripts E and F : 

(a) E has VB instead of '¥» throughout, likewise the 
Comm. MS. Berlin, oct. 333. The , indicates, it is true, 
a half vowel, which is especially frequent in Gaonic 
writings, like NTD'H (= KT?^, Responsa, ed. Cassel, 
No. 91, fol. 3a a bis), TD^N (="M?S, Anan, ed. Hark., 
11 bis), and the infinitives rpbpWD, rvjnTVD, &c. (see Anan, 
sub 2) ; comp. in addition Geonica, 34 bis : hd^I'si n^rro 
(= np|nai npna). [Here we expect, however, a long vowel, 
a : 'JfB. But the , is no doubt a sort of Imale (e or I 
instead of the long a) : so also Responsa, ed. Cassel, No. 91, 
fol. 31b: 'wn Np (in accordance with the MS. ! 73b ); 
Responsa X"ye>, III, Porta IV, No. 2 : pJTiNp ; and frequently 
in R. Hefes: "Vl\, twi, instead of '?e ^n, wn, see JQR., 
W5, P- 355> !• J 7 (M'n). P- 402, 1. 24 (yn), p. 405, 1. 6, 7 

7Sb [Hence my remark in the Festschrift, Ad. Schwarz, p. 321, is to be 
corrected.] 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 375 

(ditto), p. 406, 1. 8 (ditto), p. 423, 1. 17 (^<n 31) (on the side 
of N:in , Schwarz Festschrift, p. 320 ff.) ; furthermore wji'3 
for "JO (they read probably JP1/3, comp. Syr., not N'jn), 
R. Sherira's Epistle, ed. Lewin, p. 66 b and 68 b fo>, 
likewise Giyath, I, 70 (on the side of (W3, see my 
' Rechtsgutachten ' in Jahrbuch, IX, 283) ; further HG., 
ed. B., 346: 'rPD (fr. Kal), comp. also above sub 19 
NOW = TDK.] 

(b) F has niTD instead of nrpo throughout, but also 
N.TD on the side of it. 

No. 55: nn nrtDD «p nd rwro hkdki, 74: jttFd <:np 

i'loi', 77 : rilTD ND3 D'pJ ttpi WVJOT, 97 : "IDnt5>D niTD VJB^ 

(E: tiitd vjb^c onn), 115: £:» nro n-idc>i, 120: tiitd wip 
'13 J>ya b, 135 : 'is bjan px nrro (j'Idki margin) ; on the 
side of Siro, e.g. 104 : nemp3 Nfi^D in*jni. The etymology 
of the word is disputed, 'Aruk : Kfwo = N3n p, n.TO = DDH JD ; 
Levy : nn"D = TV p (!) ; Margolis : n.TD from flTi, nin ' be '. 
But none of these is satisfactory. At all events we must not 
confuse it with NiTD which is merely written plene and stands 
for Ni}» (Nno) ' from this ', ' from here ', so the frequent WB 
itnn ntvd, as proved by a comparison with Naz. 13 b: 
Nivmp irwo Dies. Likewise in She., No. 32: kto 
'13 pyav (so also 54 1 ), 118 F: w terra np!?dot jtMMl; in 
Git. 34 a and B. kamma 107 a terra is nothing else but 
N3DD (against Levy). Otherwise xrra = WD (on which see 
above, j«# 50), the one being neut. and fern., the other 
masc. nrra could then be a constr. of terra (considered as 
a fem. substantive) like rws. The writing with n in F 
could thus be due to a soft pronunciation of n (somewhat 
like ^). 

(c) For temite F writes temte throughout. In the 
Masorah to Targ. Onkelos srnx is designated as peculiar 



376 THE JEWISH QUARTERLY REVIEW 

to Sura and NrrvN as characteristic of Nehardea, ed. 
Sabbionetta has Nmix, but vocalized «"11N (e.g. Deut. 
9. 12, 16; 11. 19, 28, &c.) 73c , see in addition below to 'at. 

88. Of the linguistic material still preserved mfragments 
of Sheeltot the following deserve mention : 

(a) In my Vienna fragment {JQR-, IV, 420) r., 1. 20 : 
pDCp, 1. 2 : D'epT (so in the MS., see JQR-, V, 97) ' compel ' 
{I.e., IV, n. 14 and 18), see also above sub 67. 

(b) From the Genizah fragment published by Ginzberg 
the 'at (= MD) in the Derasha to No. 43, Geonica, 358, must 
be emphasized. Likewise Wt, HG., ed. B., 638 (in the 
nsoa niai>n which are probably derived from the She.) : 
vb 'axf (ed. V., 14 b : »aD). 73d In the Masorah to Onkelos 
there is a similar difference between the men of Nehardea 
and those of Sura : the former have : JHJD'N , the latter : 
pWN with t (see Berliner, Massora), comp. also Syr. 
Np^DD, Bab. xpltD (above, sub 59). Accordingly the pro- 
nunciation with t would have been peculiar to the 
Babylonian dialect of Sura. 

89. There are furthermore a series of words which are 
contained in the HG. in extracts from Sheeltot which we 
no longer possess but which no doubt belonged originally 
to the Sheeltot collection. Such are N3t6t? (= K2t6v 
tabw 'tap', Shab. 112b according to 'Aruk), D'b (HG., 
ed. V, 9b ; ed. B., 67: 3^3 j bnm, ed. Albeck, 55: D'b, 
from She., No. 90 !), ND1K (HG., ed. B., 135 : ndin IX pjD 
vxro," acrb in, V., 29 d : xoiin xrvaopi papa — according to 
Pes. 42 a, but here ND1N is more correct — from She. 
No. 74, of which traces are still contained in F : nDS3 yon 

780 For NITIX e . g. Gen. 33. 16 ; 38. 16 and Deut. 31. 29. 
™ a In the mrfan 'D, Bacher, Hebr.-Pers. WB., 56: nDmjD3 ^ 
'JD 1D3 {sujfia'ens est). 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 377 

kiwi rra liwtb '■'bki ntonai nivasa -iidn r6yoi>i niya* wo 
mm wrTi] ts*vai nun la p'o* && prn n-np <nin ">-npitti> ik 
'»n[t wma udt mip 'ia prn B>"n 'ia wo 'ia raw rua port n»5>, 
lnxwa nnie wot mip lain dk Nan (according to HG., /. c.) 
tirobn )31 uot "lTO*^, likewise several words in the She. 
preserved in HG., ed. B., 70 ('« «unrw6 bvnw nai> tdki), 
ed. V., 8 c ; then also some words occurring in the Shee. 
contained in HG. aW niai>n (ed. V., 33 b : 'ia nayo!> pyrrol 
to 33 d: nai>n pi, ed. B., 1658".) and o"r nwi>n (ed. V., 
35 a -37b, ed. B., 175 fif.), as, e.g., fol. 33c: kbww, KTtlD, 
tco, KD*a, t&aT, &c. 

All this, however, belongs already to another chapter 
dealing with the Halakot Gedolot. 



I feel it a pleasant duty to extend herewith my heartiest 
thanks to Abraham Epstein |Vr] for his kind permission to 
make use of his manuscript through the intermediary of 
the Lehranstalt fiir die Wissenschaft des Judentums in 
Berlin. Similarly I am indebted to Dr. Aptowitzer for 
ceding to me the above-named manuscript which he had 
used. My sincere gratitude is also due to Dr. A, Freimann 
for his friendly response and dispatch of the Cod. Merz- 
bacher to the Lehranstalt. Finally, I wish to thank the 
administration of the library of the Lehranstalt for the great 
kindness shown me while I was engaged in this work. 



378 THE JEWISH QUARTERLY REVIEW 

ADDENDA 
(To Anan) 

i. To the introduction. As an Arabism in Anan Har- 
kavy considers also Nrvbap (as I was still able to note in an 
additional note), Hark., 20: nana p»3*UD nM NBTDl nana 
wrfap na-rn anan |dn Dyn bs nDNl na "Toni. Hark, trans- 
lates nvtbw nanai> nun with the remark that it is formed 
after Arab. J-S ' before '. But the Aram. 5»3P ' before ' in 
a local sense (in Anan, Sch. 27, 14: P'y blp 'before our 
eyes ') could very well (as the Hebr. vtb !) be used also in 
a temporal sense, comp. Targ. bap JD, Syr. ?np |D with 
Arab. JJ ui 'before'. From this bap in the temporal 
sense must have subsequently been formed the adj. NDvap 
' the previous '. 

On the other hand, pbx in the sense of \bw is pro- 
bably an Arabism. Hark., 67 (= Sch. 23, 4) : \rb *iboi 
vb ^as lrwsu NpajT 74 nnNai>i \rb nbxon in \rb p^m ix. 
Hark, renders ypa 1 " like the Talm. p?V. But this does 
not fit in between ^OD and nbo. It rather equals jU, 
which stands frequently for <jL,, in the sense of 'cuire' 
(Dozy, I, 843 b), calore ussit (Freitag) ; iJLii caro ad 
molitiem ass at a = pbe> (Talm. and Aram. Papyri, see my 
' Glossen', ZA W., 1913) ; likewise HG., ed. B., 59 above: 
yb* wik, for ybw, ed. B., 55 (V., 7 c). [Mand. pbo, SR., 
I, 227, 12: rt!>»pS>D xnaa, 226, 5: ^pbo snm = ,jju, (Pognon, 
Z/w Incantation, p. 211, n. 2).] 

" Likewise IDNa with K, p. 62, No. 16 : i"6h J3Tp JTy 1DN3 ' after ' 
(therefore nriN, not with Hark. D1pD3), as Ba. "iriKa, Dan. 7. 7. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 379 

2. To No. 1, p. 236, 1. 2. Read June instead of WHC, 
see above, sub 77. 

3. [To Tin, No. 5 (p. 237). Furthermore, in a Palmyrene 
text NTn = tayyaru, see Lidzbarski, Ephemeris, 1, 80, where 
K3TI is suggested. In the same way Noldeke, in a letter to 
me dated December 10, 1914, suggests nam for mm in Land 
ibid. The verb, however, in this sense occurs also in Mandaic, 
SR., 1, 134 : sawrri xnwn Npns diswd jd -inn t«tn larri 
'came back, returned', against Brandt, Sckriften, p. 138, n. 2.] 

4. To N-QN , sub 8. N13N in the sense of ' arm ' occurs 
also Ber., 60 a. 

[5. To No. 9, p. 238, 1. 2 from below. Strike out ' =*e ' ; 
see also Fleischer, Kleine Schriften, I, 381. — Ibid., p. 239. 
'a in Talmud also Erub. 69 a, in accordance with Me'fri to 
Shab. 57 a, s. v. ?)iDi : ivaa o nswi rrnrp '"ii> nnm ma ; B. m. 
30 b, after 'Aruk (MS. Kohut) : wn u ttfiDViOT wn IDS, 
r."^ ; MS. Hamburg 13, r. la, comp. J. N. Epstein, ' Rechts- 
gutachten ' 303 ; B. b. 87 a after Ibn Gedalya lioirin rniDD : 
ton~a M'N, in Hul. no a the same has : son 'a xaw (MS. 
M. and H. : 13'N) ; further in p. Hal., Ill, 59 c : pHT 1 'l vh\ 
nDN '3 ; p. Ber. 6 c : rppva (Venice) Jin Ul Dj? U Dt?D 'in n^, 
r.\3. 

6. To No. 12. Maybe 'Vaa is after all to be identified 
with 'syoa, a process followed by Harkavy, comp. V^ = 
'P^C (j«5 Addenda, 1), and if so, there must have been 
a vulgar u^on the side of Arabic <j~£ . 

7. To No. 13. Responsa Tfw ed. Livorno exhibit mvd? 
on No. 171, like HG., ed. V., comp. also Responsa n"y£>, 
No. 167. 

8. di. To No. 18. Anan, Hark. 65: t*n ina K>nK£n 
53 on xin Kan e>nxB» xpi . . . Da. Read m, Neo-Syriac D? 
'thereupon, then', Pers. t_£o, Noldeke, Neusyr. Gram. 



380 THE JEWISH QUARTERLY REVIEW 

p. 166, Mand. DKB, as SR., I, 390, 1. 4 from below, Johannes- 
buch, p. 109, 1. 6: mm jd tntuxj dSb ninjw indj nhn^n iy 
' if a carpenter joined God together, who then joined the 
carpenter together?', and D'BKn, Noldeke, Mand. Gram., 
p. 205, which corresponds to our D3 Kn ! Gaonic in addition : 
Gaon mpn, II, 83 : W3in n» D3 ;m xw, r . 55 ; Hark., p. 198, 
No. 380 : 'i3 "pin WO D3 S|N1, r. D3 ; Responsa n'ye, No. 23 : 
'131 '•WD pro 5)Ni, probably a bad emendation. In Anan and 
in the responsum in Dlpil this word was vocalized with a 
pathah so as not to confound it with the Hebrew 0? and 
misread it accordingly. Yet it did not escape its fate to be 
miswritten and misread ! Also e>B occurs, HG., ed. B., 39a 
(from B. m. 93 b !) : 'oi n«j& "]b ££ WK.] 

9. To No. 20. To «b*n 'OO in A. z. 35 b, comp. p. A. z., 
II, 41 d: oop:n pa prun dtnh <:bd. 

[10. To No. 21. pp'N for ppK occurs also in Sherira's 
Epistle, ed. Lewin, p. 4 : wmp fpiSi (see note 't). -^ n, like- 
wise Neo-Syr. Ij| ' go ', Noldeke, Grammatik, p. 53.] 

1 1 . To so[l]"iy {sub 24). I abandon now my explanation 
offered in agreement with Hark. The passage, Anan, 
Sch., p. 3, 1. 4 reads as follows: bn JJTTO t*h 3J i>l> ejNl 
-icyoa pyno n["]nv, to be read no doubt »")[k]"ij>. ''TP 
pi. of trny, Syr. NTiy, also «T1V 'sponge, truffle', Talm. 
xtin. Anan's decision is against the halakah, Nid. 50 a: 
nintjBi DMD3 'Bijmi» psn jd tain, p. Maasr., I, 48 d: 
nmoai pes inv pm Nine nan iy-ir naon bi na -icyn -is?y 
'13, comp. also Anan, Hark., 131 and 135. Hark. 3 also 
is to be read likewise: ne> woo "oncw xin snny pai 
'13 D'wfo] i'pK'oV (inst. of fcnny). Comp. nns A. z. 38 a; 
HG, ed. V., null 'n, 8c above: BltPD DH3 J'N 'w msi 
D'U 'intra ; according to Anan, however, only ' raw ', xb& 
irrnso rune:, see A. z., ibid., and p. 41 d. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 381 

i a. dttemo. Anan, Hark. 25: xmian »^o •"mix'? r\"h "tdk 
hkdid nip>Dy p ditb»d 13 bx ntooicn [»pn Krw Tiro 
'« "H3»l &OD3 DT tpDNt?. Hark, corrects it to "nans or 
una'D ; the word, however, is explained properly by Rabbi 
Samuel of Gurgang in his nv»?»n 'd, Bacher, Ein Hebrdisch- 
Persisches Wb'rterbuch, Hebrew division, p. 62 b : DTlSMs 
'b ira-ina ud (conficit) ; ibid. 70 b : Dina^o jroKTia dtib 
'to. Bacher, German division, p. 87 : ' Present of trojana 
(j^^i-l^j). The "'D is either the retained prefix of the pers. 
verb or else DTia^D is an Aramaic infinitive.' To judge from 
the context (JD), howevet, DTWO is certainly a participle 
of the ithpa'al: DTianp, naturally in imitation of the Persian 
present. '? signifies in the nx^on 'd the 'rabbinic literature ', 
including also the Hal. Ged. (Bacher, 37) and the 13D 
niKia~i by Asaph (e.g. Hebr. div., ,53b below). It is very 
striking, however, that the author included in his dic- 
tionary also Anan's nivon 'D, citing it under the same 
designation as the rabbinic literature. 

13. [Tod. Anan, Harkavy, 49, 1. 7 : D'D purity N3D. SOD 
is the Arabic i_X* 'pour', the a being apocopated, as in 
NtM — 3tw and others, see J. N. Epstein, Der Gaonaische 
Kommentar, &c, p. 150. 

In the Journal of Jewish Lore and Philosophy, 1,348-53, 
Mann published a fragment of Anan's niVDH 'D which con- 
tains some new lexicographical data. Here I would like 
to note only the following : 

14. Si?. P. 350, 11. 2-4, read : . . . [«X] (3) pro p[ttj]»p 13 
po^p pitn fcsh ^ 31 (4) w ^ »a[n3] pep jnan bvxxh nss[3] 
'131. NSS is not 'mat' (Mann, I.e., n. 2), but 'row', Nid. 
61 a 'Aruk (after RH., see Tos. and Tos. RAS.) : sen 
'131 Siva p'on ' in a row ' ; i*o 'n 140 : bv tpa jnayw am 
D'jnK for <3Wn KTD, Hul. 47 a and HG. ; Arab. <_«-» ' row' 



382 THE JEWISH QUARTERLY REVIEW 

(see J. N. Epstein in Jahrbuch d. jild. lit. Ges., XII, 

99)- 75 

15. toron, Npron. P. 35a, 1. 7 (= Schechter, p. 29, 
1. 17): "onan NTDni> nSxn pa pD'pi; Schechter: sprang 
which is more correct. Npmn in the Talmud means 'chair', 
from Persian- Arabic i^io, ^&3 ; KTnrt, on the other hand, 
might be the pi. from won, from Pers.-Arab. o»o ' bench ', 
probably in the sense of pH.] 



ADDENDA 

(To Sheeltot) 

16. -"I, -e>, "KPN. To No. a. The particle —| occurs in 
Gaonic literature also as follows : HG., ed. V., nvny 53 b '• 
'•di pE'iT'p vby ;n^ pnb» wj "son ton enaoi ; HG., ed. B., 337 : 
'iai rcns t6 'nt nton najn ; #&/., 374 : ron t6i in't* yane>D 

PP3 tUjTP; zfo'dT., 412: .WD iTTI33T tOtH M3 "'i? IDS H'T? ; 

*'&</., 414: twuaT rWDn »u ^ -idk n^; Responsa Hark., 
No. 313 : '131 why runs t6 nttf fram? rvb 'dx ; Responsa 
pix njn?, 4, III, No. 33 : 'ia tbvh w anal onWtPi, No. 16 .- 

'13 S|Din rWK ijrnT DnfoetSn. 

The particle occurs already in the Aramaic of the 
Mishna ; Ket., IV, 4 : p'antJTi Dttf rb ans vb, so ed. Lowe, 
MS. M. (»santwi), HG., ed. V., 67 c (•'tan^n) and in the 
Piska in p. IV, 29 a (planum pto). 

Moreover, that particle is already rendered by ~B> in 
Mishnic Hebrew : Yeb., XV, 3 : '131 »Ntwn DNB> rb an» tontr ; 
Ket. V, 1 bis : "l»» TtapiW lb nann Nffl, so ed. Lowe and 

75 ffl in the ltO 'n might therefore be rather an Aramaism, the author 
having derived it perhaps from the Hal. Pes. , which may have had 'O'INT NS2?3. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 383 

Yerushalmi (29 d, also in the Baraita); Abot, IV, 13: 5>N1 
nnN Nian wtw n»Nn, so cited in an w^k mo ed. Friedmann, 
XV, 72 ; B. b. 154 b : b nnrop in "b nmaDi» no6, and Ber. 
a8b: DIN "JNT 1 N^ n»1N. Indeed, already late Biblic, 
Eccles. 8. 14: bn nt aw WON, and older as t£>n (= H), 
1 Sam. 15. 20: '131 TiJW nE>N $>NlOK> i>N 5>1N(5> noNn, 2 Sam. 
i. 4 : '"IJ1 DVn DJ T^N IDN'I ! 

17. To NoniN, News, No. 19 . b. Ber. 23a, ed. V., 1520 : 
'•o wan an (nowni =) nowNi w n ; Midr. Prov. 30. ^ •. 
'ia noWNT 811 ION ; in ni3-n nVB> X (Bet Talmud, II, 381) : 
'131 nan 'dn <oi nown no« myi ; HG. ed. B, also 112 : noN 
rrwr an Nerval wna an; likewise aw, ed Wilna, II, No. 47 
(p. 8): mw an nottni twin an n»N, old ed., No 144: ton 
'ia NDVTNi wna an ; Ibn Giyat, I, 95 : ni>y totvw ; ifotf., 
II, 97: rby 'WW; #£/., 16, in a responsum by Hai : 
town iioiriai. 

18. To No. 29. As to Nr6, comp. in addition Neo- 
Syriac N^o 'for', in mountain dialects: bba, ybba, in 
Bothan : "bb, ybb, near Mosul : Tin, -jinn, Noldeke, Gram., 
p. 172. 

19. To No. 47. tu in the sense of 'set' (of plants) 
occurs also in the Babylonian Talmud, in accordance with 
a reading by the Geonim : Responsa no^c nbnp, p. 72, has 
in the index of the responsa as No. 12 : noN no'o Nnna 
mNa t6l nna^ "HNan, by which B. b. 102 b is meant, where 
editions have : »an£ w}> "VaF vb"n m>w -vecn lrwo vi, MS. 
M. and H. : 'ia ibnc n!hi nst?^ nsNtn. RSBM-. nyam 
'ia ineti n^vm vbvw, hence ma in the sense of abp. To 
this category belongs also the reading of rrrbun 'd, Bacher, 
Hebr.-Pers.W drier buck, Hebrew division, p. 4, s.v. naa, and 
p. 60, 1. 526 : |npae» Tnaa, Nid., II, 7, and b. 19 a nnaa. 



384 THE JEWISH QUARTERLY REVIEW 

pp3B> -in33 would mean : ' like the already set (saffron-) 
edge ' 76 (after Rashi); MS. Parma : inna. 

In the first sense of 'spread (of fire), burn brightly, 
light ', also in Geonic Hebrew, Responsa ed. Cassel, fol. 41 b : 
-iindpi mna . twy Tires, so MS. (see my « Rechtsgutachten', 
p. 241). 

20. To No. 50. , ND, "D, in the sense of w^d, also in 
INI 'n 77 : '131 •>» i>aB> n5 inoS ; in a responsum by R. Jehudai, 
-Ten, II, 7 : iTvjnJ pt6 pew d^jid im3jb>d <d am (missing 
in ed. Lyck, No. 45, p. 18) ; 77 further, ton "nd, '^rafc, j. z>. 
"IJB (from a responsum by R. Hai, see Geiger's Jiid. Zeit- 
schrift, II, 305): '13 t«3N txb sin wd ata (in Ibn Gama, 
ibid. : VWD t6n) ; Responsa ed. Lyck, No. 102 : sn^D sna 
wb& ton w5 ; HG., B., 533 : turns sna^n Nino, 496 : ntitd 
'ia njnr ^p a«n. 

21. na.pa. To No. 55. H? in Talmudic and Gaonic 
literature belongs to Syr. p3 ' bore through, pierce (a barrel 
of wine) draw off the wine', BB. 427 : DID IN SH 1K 'V3 yrE 
'-|B>1 tunnx, Arab, jjjl ; the substantive t*ro, Geonic ttttoa, 
Responsa d"id2 (see above), is identical with Syr. WP3 
(and KSia); spta (BB. 433) 'hole, opening'. But also 
the form with v occurs once in the Bab. Talmud, A. z. 59 b : 
isbh iy niVTUi 'and pierces it until the wine flows off', in 
accordance with the right explanation of the 'Aruk : ie*p '•'S 
'131 pvi nt* ltovim rvann iva api, also a substantive Nsnia, 
A. z. 74 b : in"V-ii3t? tnb DTini (var. in '-^ra/k : m^lCK) ' the 
hole {below in the wine casket, ^SPU)', Syr. N5ni3 n. act. 

' 6 However, the nSP^Dil 'D, ibid., p. 8, explains pp by JtOSPl^K npND, 
■M, lame ou ecaille d'uneplante bulbeuse, 

" Responsa rypt?, No. 39: '13 (WD) 1^0 N^KIaSHPia STOOP WD. 
but MS. D^ "N = ed. Lyck 46 : tWiPDe- <KTI. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 385 

22. To No. 66. jkd for »nd, also in HG., B., 346: i>"K 
'131 ^PN JKD, 538: '13 y:«N3 ny pi?3 JKO; p. Yoma 38 c: 
'i3 pnv *)"in jnd. 

23. Sapy. To No. 1 7. In the Bab. Talmud, also B. b. 54 a 
MS. H. : Kap'ya P'ai toaa Ninn Nnx; in Anan, also Anan 
Sch. 24 : nntJ«D <yan nancx '13 WB»in Napy^ '■bkdi twp tfriBH ny 
MB*in xap'y pa nnivo n^ bw jtwi npa bs wt^h xap^y ny 
N'jp^ ; Responsa ed. Cassel, 40 b : '131 norvt? xnyDBH S5py 
(so MS., see my ' Rechtsgutachten ', 238) ; HG., nap 'n quoted 
by R. Nissim in his nnSD to Shab. 12 (ed. Wien, 28 b) : 
"xn p k^n Sap'y ni> prwx t6, ed V., 21 c has snp'V ; HG., B. 
365 : m w Nap'yo, 425 : irvru sr6n w xny-tK Nap^ye onn ; 
Responsa Hark., No. 215 : bvta Wpyo nWK iapwB K^p N^N 
'13. Comp. also NapJJ in Syriac as it occurs in the writings 
of the grammarians, PSm. 2961: Napy p s^N D13, i.e. 
KtFis? p! Mand., as in b. B. b. (see above), SR., II, 8 13 : 
xnmy naisi urn napy, ibid., $y 19 : rwcvn vim napjn (toau) 
'131 Nnmy naiK Kr)NDB'' , :n KJN3 ; t?apy here is in opposition to 
KtrW, like yTJ, B. b. V, 4, &c, ' the foot (lower part) of the 
trunk '. 

24. Dpi?— Dpj. To No. 78. Dpj in the sense of Dp!> 
occurs also in Geonica, 153 : Knaca ina *ia n^ D'pJD Nn3E>3 
'IB* rrb wpm ; edit. A. z. 22 a: '13 Dip^ nan, in aiD i>3B> 
II, 299 : naaa in bvj ^fcoco naeo *u b*p 3 mm ; Responsa 
D"lDjn No. 140 (= Responsa T\"w with DTI "K, fol. 96 b) : 
]op:a xro^m NnyDE> 'are collected', wrongly translated by 
Lewin JltO Nnntf 31, Jaffa, 191 6, p. 23 : niD'pno, read: 
mopi»nD (niDDKno) ! 

25. nrpo, nno. To No. 86. nrPD occurs a number of 
times in Yeb. ed. Pesaro 97 b : P3"n px nrPD 'Onp, 98 a : 
na nrvo wip, 98 b : iniD nrFo ynp ; in MS. Hamburg (ed. 

VOL. XII. r> d 



386 THE JEWISH QUARTERLY REVIEW 

Goldschmidt), B. m. 30 b : 3«n» Wn nrVD mea, B. b. 37 b : 
3«n avna TimS wip ; in the fragment to Ker., ed Schechter, 
31 b (Sch., p. 13, 1. 35): nonD NB»D3 '131 NDDD t6 yJD3t VU 
NDBD, read nno. Midr. ha-Gad., ed. Schechter, p. 346 : 
wo w«n xp. 



[ADDITIONAL NOTES 

a. To 1D1N, .r«£ 35. 'din (= noiN) instead of "11DN is 
found also in The Sword of Moses, ed. Gaster, XIV, 33 : 
by 'DIN instead of the current by (nd'n, 'dn, "ton) "I1DN, and 
so we must read also ibid., 1. 9: N'D by 'din, instead of 
n^d by 'jjnN. PI. jnoiN, Hil. Re'u 131. 

In the Responsa ed. Cassel, No. 53, fol. 13 a, below, 
the MS. has (see my ' Rechtsgutachten ', p. 371, which 
escaped me at that time) : Nrora norVNT ''NO |D pany 
WiNnsu, instead of "idint! 

b. To NXD1N, sub 56. As to N3bsb NXD1N (p. 351, n. 56), 
comp. Yelamdenu in 'Aruk, s. v. aba (Gmnhut, D'Dlp'bn 'd, V, 
90 b) : "DID (NVD1N =) HD1-I3 lb ybwB D1KBO fl? 3^3 b& U'B no 
VB TIN, therefore a ' piece of bread ', fiDna ; comp. also 
The Sword of Moses, XVII, No. 83 ■ pne»m rvy3 DN 
blSM rt'b HPI '131 NVD1N by 'o'N (r. N3bs) N3^3 (' excite '), 
yet Nid. 31 a: N3bsb iria Ht5»1 Nnb^D fis. Doubtful is 
further Shab. 140 a : 3«n»T fn N'»3 nvdin frit? finyo Nbx 
Rashi Ti "itsa, but ' bread ' would be more suitable here. 

c. To Dlin"N (din), No. 61. I believe now that the 
Talmudic D11TVN is like the Syr. tPlBflN, with assimilation of 
the B ton: Dlin , N>tP1irT'N>t5>V)Dri''N, from two, Syr. 'make 
dirty, unclean ', from which DIN (ND1D, *diid) and PIN (Syr. 
NEW) could be formed again from the Ethp. Dlin'N, C1WN. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 387 

In favour of E>ltD stands also the reading of E in No. 35 : 
e*BBW, F : DlltatO (see sub 35). 

d. To No. 84. With reference to the 'rows' in the 
Yeshibot of Palestine, see also Hul. 137 b : mns pw ?<> 
"Ofl TOp 211 ; on special occasions even 400 new benches 
were instituted, comp. Ber. 28 b : ^DSD rtWD 'i. 

e. To atzn, sub 86. The Mandaeans write their x~\H2n 
with n, which however, as usual, is read like n. Pseudo- 
Saadya to Ber. 59 b (ed. Wertheimer, ai b) has nartD TQ 
instead of box W2(2 n"Ti), with the explanation: no^BD nva 
' dark houses ', so likewise with n . 

f. To pbo ' cook ', Addenda to Anan, No. 1, p^D, pp^D 
also in The Sword of Moses, XIII, 14; ibid., 1. 9; XIV, 9 
(margin: p^», jJZ.1); XIV, 17 : XVIII, 4.] 



D d 2 



3 88 



THE JEWISH QUARTERLY REVIEW 



INDEX 



Nmaix, No. 48. 
sprtiN, No. 23. 

KD1K, No. 89. 

KXD1K, No. 56, and Add. Notes, 

sub£. 
n»ix (= tid^x, mox), No. 35, 

and Add. Notes, sub a. 
DIN, see D11. 
trim, No. 27. 
tODnt«(?), No. 75. 
WTN, No. 74. 

tCX, tO'K t&, TVK K$>, No. 14. 
lWt, No. 3. 

IWN, 1D>N, Nos. 7 and 35. 
fU% No. 13. 
Nnatwx, No. 65. 
D11JVN, No. 61, see D11. 
TWN, seenCK. 
f\bn, xebtt, No. 6. 
NDK, KDIVK, NDWN, ND«, No. 

19, and Add. No. 17. 
W3K, No. 30. 
}*»N, No. 66. 
KXDK, see NVD1N. 

nDN, nco, No. 10. 
tunox, pnnN, No. 69. 

K3-IN, No. 66 (Talm.). 

Nrnx, No. 87 c. 

(itw), wrvt<, sancm, No. 25. 

ma (tma), No. 47, and Add., 

No. 19. 
Kra, No. 72. 
»ra, No. 68. 



NnD3 (= tTlD 'a), No. 66. 

(nw), 'jnnm, wo, ■'snrvio, 

No. 12. 
na, No. 55, and Add. No. 21. 
(Nt-ia) xrtna, No. 55. 
ana, ma, No. 67. 
ma, T"tf ma, No. 5. 
Tna (nw)> No. 6. 
ku, No. 59. 
•vu, twk, twk, No. 39. 

tOU, No. 39. 

tnmw, tcomw, ibid. 

twiu, No. 86. 

WJ, Wl, No. 64. 

dj (dj»j), wo, No. 83. 

"1 (= nDt«i>), No. 2, and Add. 

No. 16. 
<oitn (run, Bibl. Dion), No. 38. 
trial, No. 84. 
t<pnn, vi". No - 39- 

NPDH, NpDNH, No. 53. 

i3T, "van, No. 39. 

mn, mn ,, Bt6, mm, No. 74. 

nn, see Hnai>. 

wnin inn, No. 44. 

tan, pn, nth, hvi, p'n, 

No. 13. 
tONI, No. 3. 
ifcbn, No. 21. 
fin, 1^, -pn, No. 13. 
■pt<n, "pttn, #«/. note. 
(»an) on, (on) fn, No. 34. 
W1, No. 4. 



POST-TALMUDIC-ARAMAIC LEXICOGRAPHY — EPSTEIN 389 



D11, D*irw (dwx), Diio, tarn, 
DlND, Nos. 61 and 35, and 
Add. Notes, sub c. 

MT, »3Nt (=MD), No. 88 b. 

tWlB'?, No. 40. 

jnt (see T13), No. 47. 

K?n, nns (= *ins), No. 41. 

^an, No. 86, and Add. Notes, 
sub e. 

•ran, No. 54. 

b&T\, No. 72. 

' for T (Imale). 

KBITS NDSr, tOBTO, No. 40. 

vnprw No. 48. 

na, ana, No. 52. 

tma, see KTt3. 

N313, '3«, '33, No. 18 note. 

^V3 (pi.), No. 69. 

[tOK3pD13J, No. 28, see MtapDO. 

Nni3 (nni3), No. 58. 

KM, No. 78. 

Kb (nb), ^3, No. 85. 

wk «i>, ivk k^>, see m^k. 

nmi>, x»inb, No. 33. 

KDr6, No. 18. 

xnco^, No. 8. 

Qpb (= ops), No. 78. 

xni>, No. 29, and Add., No. 18. 

|KD (= 'KD), No. 66, and Add. 

No. 22. 
PDNID, No. 46. 
KTD, H"D, 'jn"D, DVTO; pi. 

WD, yNTD, No. 16. 
KJBPlD, see NSlf. 
nnrVD, see KS?3. 
('ND), Kin iD, liTD, WD, No. 50, 

and Add. No. 20. 



tWO, niTD, No. 87 b. 

nrfD, nnD, ibid., and Add. 

No. 25. 
W (= '•SD), No. 87 a. 
KpntlD, Kp-MD, >P"1»D (Syr. 

KpnDND, NpnDD), No. 51. 
K3K3PDD, see X3N3PD13. 
KJpDD, p^DD, p^DD, No. 22. 
NpnDD, KPTIDD, No. 67. 
ID (nDK), No. 9. 
K^ptW, KJPpB>D, No. 68. 
xsno, 'JrlD, No. 69. 
KTWID, No. 7. 

(n»* = riM), pruo, No. 76. 

-ins, insN, No. 45. 

nw, No. 55. 

Dpi (= Qpb), No. 78, and Add. 

No. 24. 
KDB>3, KDPU, No. 57. 
jnj, JJVD, No. 17. 
|KD, see D"D. 
K'JID (= V&fti), No. 68. 
S)1D, 5]1D1, No. 82. 
KWD (KTU3), No. 31. 
D^D ("OKDD), KDK'D, No. 83. 
DDD (DDDD), No. 26. 

nsD = rax, No. 42. 

(3pD) K3K3pDD, see K3K3PD13. 

■•P^D, No. 49. 

NXDlj) (Syr.), see kxdik. 

bv (=K), No. 15. 

i>xnt5" iran kdjt, No. 2. 

pj?, No. 66. 

pDy(?), No. 83. 

K3pj>, K3P" 1 ?, No. 71, and Add. 

No. 23. 
0?V (= YPV)> No. 83. 



390 



THE JEWISH QUARTERLY REVIEW 



a-\pv, tnp% No. 71. 
Kany, No. 66. 

D13 (WDn), No. 68. 

'»is (wyp), No. 72. 
Ktnt3S, No. 39 (Talm.). 
TlKB, ibid. 
KTVB, #*'</. (Talm.). 
KnDiTB, unions, No. 50. 
rms, ims, No. 79. 

Nms, KHTS, No. 68. 

rm, No. 21. 
kit*, No. 27. 

(niS), flTlVK, No. 20. 

WW*, (1) No. 24, (2) No. 67. 

KT1X (= N-TO), No. 27. 

Nns<nx (pi.), No. 50. 

B»p, No. 67. 

-IVP, No. 54 (KTVp). 

Nn«3pup, No. 32. 



WOp, No. 72. 

■onp, No. ir. 

xm, No. 67. 

NpWl, No. 23 (Talm.). 

ffo B»n, No. 85. 

in, ins, No. 61. 

Kn^Kt?, No. 1. 

nre> ('•-wts'), No. 68. 

tne>, No. 77. 

KiaiB> (= waic), No. 79 note. 

•|K>, -JEW, No. 63. 

e)^>t5>, No. 50 and 72. 

snot? (Kin), No. 80. 

Kne>, Nit5«D (= taw), No. 67. 

Knt!>, No. 68. 

Kwn, <a<n, No. 67. 

wi (= ■on), No. 87. 

Kin, No. 43. 

wn, pn, No. 7. 



INDEX TO 

K13K, No. 2 (Anan). 
jyN (=np*K), No. 10. 
KDmK, K»WK, No. 17. 
KTiK, see KIKny. 
ma, No. 19. 
Ksnn, No. 21. 

D3 (D3, Di), D3 KH, No. 8 

(Anan and Geonim). 
na, pa, No. 21. 
-t, No. 16. 
O, No. 5. 

paa (= e»33), No. 6 (Anan). 
Bp$> — Dpi, No. 24. 

Knb, No. 18. 

■•KO, 'D, Kin 'KID, No. 20. 

}KD (= 'KD), No. 2 2. 



ADDENDA 

nrr-D, nn», No. 25. 

opj — Dp$>, No. 24. 

K3D (aaD), No. 13 (Anan). 

p^D, No. 1, and Add. Notes, 

sub./ 
Kap% K3piy», No. 23. 
KiKny, Kiny (= Kiny, ktis), 

No. 11 (Anan). 
(Dins) ons*D, No. 12 (Anan). 
B'B, see D3. 

p^X (= pbw), No. 1 (Anan). 
KBV, No. 14 (Anan). 
Kn^ap, No. 1 (Anan). 
-&?, nCK, No. 16. 
nm, No. 3 (Anan). 
NTDn, Kpnan, No. 15 (Anan).