Skip to main content

Full text of "North-Semitic Epigraphy"

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. 



258 THE JEWISH QUARTERLY REVIEW 



NORTH-SEMITIC EPIGRAPHY 1 . 

The study of North-Semitic epigraphy, especially in 
this country, has sorely stood in need of the stimulus 
which Mr. G. A. Cooke's excellent handbook is calculated 
to afford. The great Paris Corpus Inscriptionum Semiti- 
carum is costly, and the rate of publication is naturally 
slow ; the Palmyrene inscriptions have not yet appeared, 
and although the other Aramaic and the Phoenician 
divisions are nearly completed, fresh inscriptions have 
been discovered which in not a few cases affect those which 
have already been edited, sometimes in the matter of inter- 
pretation, sometimes even as regards the readings. Professor 
Driver's notes in the Introduction to his Hebrew Text of 
Samuel were an admirable specimen of what was needed 
for younger students, but obviously they could not carry 
the beginner far. The present writer's Glossary of the 
Aramaic Inscriptions, as the title implies, was limited in 
its scope and barely covered one portion of the field, and 
it was left for Lidzbarski's fine Handbuch der Nordsemi- 
tischen Epigraphik to furnish the first general introduction 
to the inscriptions. The arrangement of the material, the 
complete bibliography, the comparative glossary of the 
Canaanite and Aramaic inscriptions, and the useful selec- 
tion of texts render his work indispensable for advanced 
study 2 , but, like the Corpus, it is scarcely the book to put 

1 A Text-book of North-Semitic Inscriptions. By the Rev. Q-. A. Cooke, MA. 
(Clarendon Press). 

2 Needless to say, the Handbuch has been of the greatest assistance in 
these pages, particularly in the compilation of the material, pp. 272-7 
below. 



NORTH-SEMITIC EPIGRAPHY 259 

into the hands of the beginner. A preliminary text-book 
was wanted, and Mr. Cooke, who originally intended to 
provide a manual for students who offer the subject of 
Semitic Epigraphy in the Honour School of Oriental Studies 
at Oxford, wisely decided to appeal to a larger field. Here 
we have about 150 inscriptions, transliterated into Hebrew, 
translated, and admirably annotated ; care has been taken to 
incorporate the most recent researches, and the discrimina- 
tion which the author has shown in dealing with the copious 
material scattered throughout the various learned journals 
gives his work a lasting value. Primarily intended though 
it is for younger students, we have no doubt that Mr. Cooke's 
publication will, through the richness of its notes, be gladly 
welcomed by all Semitic scholars. 

The inscriptions under consideration belong to the "North- 
Semitic" division. The term is a convenient one, but is 
not to be understood in a geographical sense alone, since 
"North- Semitic" inscriptions are found in Arabia and 
Egypt, and "South-Semitic" have been discovered south 
of Damascus and in the region east of Jordan. The term 
is used partly in a palaeographic and partly in a linguistic 
sense. The " South-Semitic " scripts, in which are written 
the inscriptions of the Minaeans and the Sabaeans, of 
Lihyan, Thamud and Safa, separated at an early, though 
unknown, date from that ancestral alphabet, the parent 
of the North-Semitic and European forms, and are 
characterized by distinctive forms and by greater precision 
in the reproduction of the finer shades of utterance. As 
a linguistic term, " South-Semitic " includes Arabic and 
Ethiopic, and stands opposed to the "North-Semitic" 
which, with Wright, may be divided into three groups : 
Eastern (Assyrian and Babylonian), Central (Aramaean), 
and Western (Canaanite) \ Since Assyrian and Baby- 

1 Other groups have been suggested, but Wright's has istinct advantages 
and is eminently preferable to that which Mr. Cooke himself adopts 
(p. xvii and n. 1), apparently through a misunderstanding of Wright's 
words (Comp. Or., p. 12). 

S 2 



260 THE JEWISH QUARTERLY REVIEW 

Ionian stand apart, it is the Canaanite and Aramaean 
inscriptions which come under consideration, and a brief 
survey of the material and characteristics of these two 
branches will afford some idea of what is meant by these 
rather conventional designations 1 . 

Of the Canaanite inscriptions by far the most important 
is the Moabite stone dating from the middle of the ninth 
century B.C. (No. I) 2 . Entering, as it does, into Hebrew his- 
tory it is the most interesting record yet discovered, and the 
literary ability with which it is marked is evident proof 
that the Moabites were well acquainted with the art of 
writing 3 . With the exception of a couple of seals, which 
bear names compounded with the god Chemosh, no other 
Moabite remains have as yet come to light. As regards 
Hebrew inscriptions, too, the material at present is 
lamentably scanty. The only one of any length is a six- 
lined text from the tunnel connecting the Pool of Siloam 
and the Virgin's Spring in Jerusalem (No. 2). A number 
of small specimens of Hebrew writing have been found in 
the shape of seals and pottery-marks, jar-stamps, &c, and 
are chiefly of value for Hebrew onomatology and palaeo- 
graphy 4 . 

1 It will be recalled that the Hebrew square character has developed 
out of the Aramaic, from which source, also, are derived the Arabic forms. 
It is singular to find during the early centuries of the Christian era some 
Arabian tribes using an Aramaic script, others a South-Semitic, a descen- 
dant of the older Minaean and Sabaean; in the former the definite article 
in compound proper names is i» as in Arabic, in the latter n ! See below, 
p. 271. 

2 The figures in heavy type refer to their position in Cooke's Text-book. 

3 It is a little astonishing that the belief has not yet died out that the 
stone after all may be a forgery. If this were so, it would be remarkable 
that mw (?rr«!»i, nvi'N, 11. 9, 23), evidently a pit or reservoir, has only 
recently been recovered (Ecclus. 1. 3, see Cooke, p. 9). 

4 On p. xx. n. 2, rwj[?-ic] should be read nicno. It is possibly the 
Mampsio (Ma^is) of the Onomastica (Lagardo, 85, 3 ; 210, 86), on the road 
from Elath to Hebron, one day's journey from Thamara (Hommel, Exposi- 
tory Times, xii. p. 288 [1901]). In the same note too, for • •swds read 
• • a " ps, the dot or rather stroke (so in the original) appears to be a 
word-divider. 



NOKTH-SEMITIC EPIGKAPHY 261 

The vast majority of the Canaanite inscriptions are 
Phoenician, and with few exceptions have been found 
outside the mother country. They proceed from a definite 
and distinct branch of the Semites which, ever restless 
and energetic, pushed its colonies along the coasts of the 
Mediterranean, possibly as far as Spain itself. The oldest 
inscription of any length from Phoenicia proper is that of 
Yehaw-melek, king of Gebal (Byblus), of the fifth or 
fourth century B. o. (No. 3). Somewhat later are the 
Sidonian inscriptions of the dynasty of Eshmunazar 
(Nos. 4 sq., Appendix, pp. 401-3), whilst a few from Tyre, 
Umm el-'Aw&mld and Ma'sub extend to the third and 
second (Nos. 8-10). Of greater antiquity, however, are 
the Baal-Lebanon, and the Hassan-bey-li 1 inscriptions, the 
former found at Cyprus, belonging to the time of 
Hiram (II) 2 , the latter from N. Syria, west of Zenjirli, 
seems to be of about the same period. Cyprus has 
furnished a number of other interesting inscriptions 
(cp. Nos. 12-30) from the fourth century downwards, and 
some of these, written in Phoenician and Cypriote, were 
of great assistance in the decipherment of the latter 
character. The Phoenician inscriptions from Egypt (No. 31) 
have not been so numerous or important as the Aramaic. 

Phoenician colonies in Greece (Athens, Piraeus, Nos. 32 
sq.) have provided richer material, ranging from the fourth 
or third century to the first (96 B. a). Malta and Sardinia 
have furnished a few distinctive finds chiefly of the second 
century, although two remarkable examples (Nos. 37, 41) 
on palaeographic grounds may even be four centuries older. 
From France come two tablets, probably of the fourth 
century B.C., one found at Avignon, the other, an extremely 

1 Edited by Winckler, Altorientalische Forschungen, I, pp. 305-9. 

2 No. 11 ; Mr. Cooke agrees with von Landau that Hiram II, mentioned 
by Tiglath-Pileser III (738 b.c), and not the contemporary of David and 
Solomon, is meant. This is also the view of Ed. Meyer (Encyc. Bib., col. 
3753, n. 2), and is probably correct. Hence the famous bowl can no 
longer be cited as a specimen of the Phoenician character of the eleventh 
or tenth century. 



262 THE JEWISH QUARTERLY REVIEW 

important sacrificial tablet, at Marseilles (No. 42). That 
these were actually inscribed in France by Phoenician 
merchants or colonists is far from certain. Carthage, and 
the northern coast of Africa (Tunis, Tripoli, Algeria) have 
furnished hundreds of remains, none, however, older than 
the fourth or third century, whilst the latest (the Neo- 
Punic) belong to the earliest Christian times. 

Extending as they do over so many centuries, these 
inscriptions display numerous examples of modification. 
Not only does the script undergo remarkable change, but 
the orthography and the language does not remain fixed. 
Traces of separate dialects are found in the Byblus inscrip- 
tion, and in those from Hassan-bey-li, Lamax Lapethos, and 
Sardinia 1 . Foreign influence shows itself partly in the 
bilinguals, in the use of Greek, Cypriote, Latin, and Berber, 
and now and again in the general style of the inscription. 
A notable example of the latter feature appears in No. 33, 
from Piraeus. As is frequently the case in the Aramaic 
inscriptions, the evidence of the proper names often sug- 
gests that the writers were foreigners, cp. Cleon (No. 40), 
and see e.g. No. 52, where the names appear to be 
Numidian. 

As regards the general characteristics of the Canaanite 
inscriptions reference may be made to Mr. Cooke's remarks, 
pp. xix, 23 and passim. For the Hebrew student their 
grammatical and orthographical features are of the greatest 
interest, and on the assumption that no one knows any- 
thing of Hebrew who only Hebrew knows, the bearing of 
the inscriptions upon the study of Old Testament language 
and writing can scarcely be overestimated. Among the 
peculiarities of Phoenician may be noticed the use of p 
" to be," never fiTi, and the construction TU p ; both the 
verb and the construction are reminiscent of Arabic, but 
naturally are not to be taken as evidence that Phoenician 
and Arabic were more closely related than Hebrew and 

1 See Cooke, p. 25 and note in No. 29 as for 2, and in Nos. 39-41 the 
comparative frequency with which nj is used. 



NORTH-SEMITIC EPIGRAPHY 263 

Arabic. Similarly, the Phoenician nsiy " portico " and 
nj?l¥ a kind of sacrifice, admit of an explanation from Arabic 
and Ethiopic respectively, and find their chief value in the 
circumstance that they serve to indicate that the common 
Canaanite language must have been considerably more 
extensive than the vocabulary of the Old Testament by 
itself would suggest. Phoenician, it may be mentioned, 
carefully distinguishes between nf "month" and trin 
" moon " ; in Hebrew, on the other hand, they are confused. 
It is rather singular, also, that the Hebrew prepositions 
Dj? and p have not as yet been found in Phoenician \ 
The construction riNt neon (never DNtn nioart) is character- 
istic of Phoenician and Moabite as distinguished from 
Hebrew, where, however, it is not quite unknown 2 . The 
spelling nxf in the Moabite stone (1. 3) agrees with the 
Hebrew, but in Phoenician the corresponding form nt does 
not appear until the period of the latest (the Neo-Punic) 
inscriptions. On the other hand, it regularly employs t for 
both genders 3 , although no doubt they were distinguished 
in pronunciation 4 . 

1 The Phoen. Vi may correspond to the former (see p. 283, n. 3 below), 
but whether (j)a answers to |n is not certain (Cooke, pp. 33, 35). As regards 
the latter, at all events in 9 3 ("|"ttA atal) the use of 3 may find a parallel 
in the Poemdus of Plautus ; anno byn mytthymballe bechaedre anec (Hannonen 
esse ait Carthagine, Carthaginensis Muthumballis fllium) appears to 
represent : -]3« [rann]mpa tana p wn, 

2 See the passages cited by Driver on 1 Sam. xix. 10. 

3 Only in the Byblus inscription does the fem. xi occur, and it might be 
suggested that mt is a double feminine, and that rri and Vt are earlier 
forms, ni, however, is not confined to the feminine (p. 26), but as CIS i. 
149, 1. 3 sq. and other passages show, could be used for the masculine (cp. 
Lidzbarski, Ephemeris, I, p. 44). The addition of n finds parallels in the 
Phoen. non, the Sabaean rort, fern, rwt, and in Ethiopic (Wright, p. 105). 
It is therefore possible that the n of the Hebrew nw has a similar origin, 
and has nothing to do with the feminine termination. The form in the 
Poenulus of Plautus (sytti) points to an older soth, analogous to chyl = "a. The 
evidence of Plautus is naturally to be used with caution, but these 
interesting old fragments are important enough to have deserved some 
notice in Mr. Cooke's Introduction. 

* Wright (p. 108) conjectures ae and zii. The Hebrew 5, it should be men- 
tioned, has passed over into u in such spellings as the PuniesaZws (three), rufe 



264 THE JEWISH QUARTERLY REVIEW 

Moabite is characterized by the occurrence of a conjuga- 
tion which is elsewhere found only in Assyrian and Arabic 
(Dnr6n " to wage war "), the plural and dual ends in }- as 
occasionally in Hebrew 1 , whilst the Phoenician regularly 
has D-. The language on the whole differs only dialectic- 
ally from Hebrew, and although a few features associate 
it with Phoenician, these are outweighed by such usages 
as the Hiphil as opposed to the Iphil (see below) and by 
characteristic idioms and terms (see Cooke, p. 5). 

As regards the orthography, too, Moabite is more closely 
related to Hebrew than to Phoenician, and in view of the 
question of the orthography of the earliest MSS. of the Old 
Testament, the subject is one of some interest. In the first 
place let us observe that both the Moabite stone and the 
only Hebrew inscription of length separate the words by 
means of a point 2 . In Phoenician, however, the inscrip- 
tions so marked are not the oldest, nor are they found in 
Phoenicia itself, but in Greece, Cyprus, and the west 3 . 
The Massoretic text contains indubitable instances of the 
incorrect division of words *, but these may be due to the 
fact that the earliest MSS. were written in a script where 
these convenient guides were not in use. It is to be 
noticed that neither points nor any similar contrivances 
are used in Egyptian Aramaic (the words, however, are 
sometimes separated by spaces), nor do we meet with final 
letters before the period of the Nabataean inscriptions 
(middle of the first cent. B.C. and onwards). 

In the next place, as regards the use of 1, ', N and fi, it 
is important to notice the leading features of the inscrip- 
tions. In the Siloam inscription, it is true, final n is 

(physician), alonuth (goddesses), rus (head), and notably suffet (ffiBio). Note, 
on the other hand, &.pf>a$wv = JiXW. 

1 Hence, the Old Testament examples are not necessarily Aramaisms. 

8 Similarly the Aramaic inscriptions of Zenjirli, but not the contiguous 
Phoenician inscriptions from Hassan-bey-li, or the Baal-Lebanon bowl. 
In the oldest Arabian inscriptions the words are divided by a line. 

3 Lidzbarski, Handbueh, p. 202. 

1 For examples, cp. Driver, Samuel, pp. xxx sqq. 



NOKTH-SEMITIC EPIGKAPHY 265 

represented in napJ (boring), HM (to be), nr (this), &c, but 
against this we have n»n, which is presumably JVn rather 
than nn^n (p. 16 sq.). Final "> and 1 appear in "o and tti»l 
in conformity with Moabite as opposed to Phoenician. 
Ordinary long i and 6 are not represented (e.g. D?^, 
nfeN &c), and the spellings T1JJ and N20D suggest that these 
retained the older pronunciation 1)5? and NXip. On the 
other hand, in the two exceptions bp and d 1 , the vowel, 
whatever its origin (?1P_, BT?), was probably heard as 0. 
The question, however, remains uncertain. Noteworthy 
also are the spellings B>N"i, OTIND in the Siloam inscrip- 
tion, compared with the Moabite JDND and fK2f on the 
one hand, and Eh (head), and rCT (? gazing-stock -/n&O) on 
the other. Obviously N, at least, was never written unless 
it retained some consonantal force. 

There are certain spellings in the Moabite stone which 
deserve consideration. The ordinary long 6 and % are not 
marked (Q^V, $hf), hence pH is probably fsi^t rather than 
Ji^and in the names pTin, rrm, 1 is probably diphthongal. 
Comparing rtn*a (1. 25) with nrn (1. 7), rv>"\w (1. 22) with no* 
(1. 8), it is open to suppose that the vowel is 4, and, if this 
be so, we must pronounce the Phoenician duals DDND, MB>, 
with eVrc. (cp. Heb. &$) and treat the Moabite DTiND as a 
plene spelling. If, however, we may conjecture that even 
a diphthong could be omitted, we may point to the North- 
Arabian inscriptions from Safa, where such spellings as DK 
(aus, also din), epN (unaif) and oprw (Nabataean XX\\bww) 
recur \ Another difficult letter in the Moabite stone is the 
final n (nmp, 1. 3 ; mil, 1. 12 ; rbh, 1. 15). It is not the 
feminine, since this is regularly n-. Wright (p. 139) sug- 

1 It may be added that in ipn ripi jn -qi (1. 24), Mr. Cooke understands 
)'N (Gen. xlvii. 13), but ]» (fj») is possible on the analogy of 2 Kings xix. 3. 
The verbal form 'SMCT (1. 4) is curious enough to have deserved remark. Was 
it pronounced wtfrt or wahn ? The name Hoshea is always written 
3>8)in on Hebrew seals, but whereas the king is called a-u-si-' in an inscrip- 
tion of Tiglath-Pileser III, in a bilingual Aramaic docket (e. 680 b. c), 
ncm is represented by u-si-'. 



266 THE JEWISH QUARTERLY REVIEW 

gests, in the case of nmp, that it represents an ending d'u ; 
this, perhaps, does not suit rW>, perhaps originally ^ 
(cp. Syr. *^-X), but that n has some consonantal force, at 
all events, appears tolerably certain. 

The Moabite use of the final "• and 1 as a vowel letter 
(cp. Siloam, V), 13^1) is contrary to Phoenician. *a "for" 
is regularly 3, and it is not until the period of the Neo- 
Punic inscriptions that the vowel is indicated \ The plural 
construct in ■» ("<isb, s nS>3 1 13, 23) is not marked in old 
Phoenician, only Punic and Neo-Punic use N- or y- 2 . The 
plural OS too, is only found in Neo-Punic. With the 
Moabite "Ojjpn (1. 4 " he saved me ") contrast the Phoenician 
inbyB ("she made me"), with roa, V03 (11. 18, 21, "he, 
I built ") contrast J3, TU3 3 . The 1st sing, suffix of the noun, 
however, is regularly represented in both. 

As regards the evidence of Hebrew seals, pottery stamps, 
&c, the following details may be noticed : Visa (Massoretic 
Text £>W3N), HKON, D13K (never 'UN), ytwitf; Dyi'N (Clermont- 
Ganneau, No. 44), otherwise always '^N (fDN^K, ptsfot, &c); 
inner &c. (only once n^ in rPMn '[?], Levy 3 = Vogue 36); 
the ^Zewe pin seems to be exceptional, pun is written 
defectively on the pottery stamps from the ShSphelah ; on 
the other hand cp. *|*f and n3lt?, where, however, the LXX 
(ei<j) (but also £i<f>) and the modern name esh-Shuweikeh may 
suggest that the spellings Zlph and Socoh are not original 4 . 
It is perhaps sufficient to have recorded some of the more 
important orthographical peculiarities of the Canaanite 

1 Viz. *o, to, 2>D. On the use of the gutturals, see Cooke, p. 140. Note in 
Phoen. p "they built," rav "they sacrifice," TB "fruit," tyc) "above," 
Jtm "eight," -w " field. ""' 

2 The only exceptions in Aramaic are old : rrts (62 2, but 'rfjN, 1. 32), 
JNO (65 6) 

3 Hence the Moabite "pH does not stand for >3:n (in Phoenician, only 
in graffiti from Egypt). 

* Plene spellings in old Aramaic inscriptions are found certainly in 
-iton (63 9), TD'bsiran (ibid., 1. 3), whether in nu, nrocn, mi, ov, moifi 
(Lidz., Ephem., I, 82) is doubtful. For the evidence from the later 
Nabataean and Palmyrene inscriptions (e.g. urns, Train, &c), see below, 
276, n. 1. 



NORTH-SEMITIC EPIGRAPHY 267 

inscriptions, obviously a complete discussion of them in 
connexion with the Massoretic Text would be out of place 1 * 
It has already been observed that Moabite agrees with 
Hebrew in the use of the Hiphil (^yen &c). In Phoenician 
the n is replaced by 1 (trip" 1 , tan\ See.), and it is only in Neo- 
Punic that we find the form N^jtft 2 . The view that ■• is 
a spiritus lenis, a weakening of the older n (p. 58), is 
difficult on account of the later Neo-Punic form just cited, 
and it would seem preferable to allow the ' to retain its 
consonantal force, and to pronounce yikdls (or yakdis), rather 
than ikdw. One other point which calls for notice is the 
form of the suffix of the third sing, masc: Moabite n, 
Phoen. 1-, later K-, and Neo-Punic N 1 "-. n- seems to appear 
also in the Haesan-bey-li inscription, but only in that of 
Yehaw-melek of Byblus do we meet with 1- (also verbal, 
linn 1. 9), fem. n- (p. 25). Mr. Cooke has strong support 
when he explains *- as a contraction of an original -alii 
(p. 42), the Moabite n- from -ahu = -au = -o (p. 8), and 
the plural suffix in 'nm " his words," from \)Y$r\ or 
"^l (p. 104, cp. vri^ejn, Ps. cxvi. 12). It seems better 
to derive all three from -ihu (-ihi 3 ), comparing the 
Hebrew •1™?$, vnte (dual or plur. w, vpyj?, vrnlaa). 
If so, the Moabite and Phoenician forms should be pro- 
nounced nsiK, vp, and '!3"l. The not unfrequent examples 
of the masc. n- in the Old Testament may also be similarly 
viewed: frvy and nh^D (Gen. xlix. 11) should be ffvy and 
nrilD, and the Massoretic punctuation, under these circum- 
stances, must date from a time when the older use had 
been forgotten. It is important to notice that IfitM^S (Judg. 
xix. 24) contrasted with ItM^S (ver. 25) shows that both 
forms could be used side by side *. 

1 We may, however, note the confusion of <3 and 3 in Judg. xvi. 25 ; Ezek. 
vii. 9 ("]'3VD, see verse 4) ; Ps. Ixix. 17, cix. 21 (nto '3). 

2 The late proper names "fro, Dp'n, may be Iphil participles (Lidzb.). 

3 This, the genitive form, appears to underlie the Aramaic -eh (Wright, 
P- x 55)) which is quoted by Cooke on p. 42 in support of the original 
form -ahi. 

1 Cp. i:n», van 1 ), Gen., I.e. See, further, Proceedings of Soc. of BibL 



268 THE JEWISH QUARTERLY REVIEW 

Turning next to the Aramaic inscriptions we may 
observe that they are not, like the Phoenician, the work 
of a distinct branch of the Semites, nor do they admit 
of being classified with such precision as was the case 
with the Canaanite. Apart from Syriac proper, the 
language spoken around Edessa, a literary language, the 
Aramaic dialects fall into a number of divisions whose 
earlier history is clouded in obscurity. It is sufficient here 
to mention Jewish Aramaic, the dialect of the Targums, 
the later Christian-Palestinian (or Palestinian-Syriac) \ the 
Aramaic of the Babylonian Talmud, the Mandaean, and 
to these we must add the dialects of the Aramaic portions 
of Ezra and Daniel, and of the inscriptions. 

It is known from Isa. xxxvi. 1 1 , 2 Kings xviii. 26, that 
Aramaic was used in diplomatic intercourse between 
Assyrians and Israelites in the time of Hezekiah; the 
inscriptions will show us that these were by no means 
the only people who found the language more serviceable 
than their own. The oldest remains are those from Zenjirli 
in North Syria (Nos. 61-63), and date from the eighth 
century. Although written in a script, closely resembling 
that of the Moabite stone, it is important to observe that 
a very gradual modification can be observed in the shapes 
of certain letters (T, D, D, p, see Plate XIII), a clear indica- 
tion of the early development of the Aramaic alphabet 2 . 

Archaeology, 1903, p. 39 sq., on the forms m, now in the Nash papyrus (cp. 
J.'Q. S., 1902, p. 395, 11. 9, 11). The curious wipi (ibid., I. 16), it may be 
added, seems to find support in the Phoen. Oil' (wii') "may he (they) 
bless him." If so, we must pronounce vsnpi, the ' is a consonant, and not 
a plene spelling as in vW (Job xxi. 23). That the N in the Phoen. Km' 
retained some guttural force seems to follow from its being used with the 
imperf. plu. (above). 

1 On its geographical distribution (Abud, between Judaea and Samaria 
in the seventh, and Antioch in the eleventh century), see Burkitt, Journ. 
Theol. Stud., II (1901), p. 181. 

2 To these inscriptions we must now add that of Kalammu son of gayan 
(Sanda, " Die Aramaer," Der aXte Orient, IV, Heft 3, pp. 12, 26), which is 
said to belong to the time of Shalmaneser II (859-825 b. a). This inscrip- 
tion, which appears to explain an obscure phrase in No. 63 17, has not 
yet been edited. 



NORTH-SEMITIC EPIGRAPHY 269 

From Nerab, in the neighbourhood of Aleppo, come two 
small but extremely interesting inscriptions, probably of 
the seventh or sixth century (Nos. 64, 65). 

The finds from Assyria and Babylonia are numerous. 
A few bronze lion-weights (No. 66) have Assyrian inscrip- 
tions, and since they bear the names of the kings Shal- 
maneser, Sargon or Sennacherib, admit of being dated 
with tolerable certainty. Somewhat later are a number 
of contract-tablets, most of which are Assyrian with 
Aramaic legends ; those from Assyria belong chiefly to the 
seventh century, the Babylonian to the sixth and fifth. 
Passing over seals, bowls, &c, we need only mention 
a bilingual from Telloh of the third "or second century in 
Aramaic and Greek (nwwnn, ababvabivaxv^) 1 . 

From Limyra comes another bilingual in Greek and 
Aramaic of the fifth or fourth century, and elsewhere in 
Asia Minor a few scattered inscriptions have been found, 
one of the most interesting being from Saraiidin in Cilicia 
(No. 68), where a young huntsman places on record the 
fact that he was making a meal there. The recently 
published inscriptions from Cappadocia should also be 
mentioned 2 , if only in the hope that scholars acquainted 
with Persian may turn their attention to them. 

Aramaic would appear to have been frequently used in 
Egypt during the Persian period. A stone inscription, dated 
in the fourth year of Xerxes (482 B.C., No. 71), is of 
particular importance on this account for the history of 
Aramaic palaeography, and to this we may now add a 
papyrus, recently edited by Euting, of the fourteenth year 
of Darius (II. Nothus, i. e. 411-410 B.C.). Among the other 
finds of Egyptian origin are to be mentioned the ostraka 
from Elephantine, and the Oxford papyrus which Mr. Cooke 
was able to insert in his Appendix (pp. 404-6) — an in- 
teresting illustration of the progress of Semitic epigraphy. 

1 One recalls the Greek-Babylonian fragments edited by Pinches, in the 
PSBA, 1902, pp. 108-19, perhaps of the second century. 

a Lidzbarski, Ephemerisfur semitische Epigraphik, I, pp. 59-74, 319-26. 



270 THE JEWISH QUARTERLY REVIEW 

Tema, in Egypt, furnishes one old Aramaic inscription 
of the sixth or fifth century (No. 69), and a few others of 
descending date, among them one of particular interest 
for its close palaeographic resemblance to the bilingual 
from Telloh (No. 70). From Hejra come a number of 
small legends extending down to the second century B. c, 
chiefly noteworthy for the steady modification of the script 
to the form assumed upon the Nabataean inscriptions. 
The last -mentioned constitute a distinct palaeographic 
class, and to judge from the proper names were largely 
due to Arab -speaking tribes. A few have even been 
found in Italy (Puteoli, Rome) 1 — the work of travelling 
merchants — one at Sidon, and one as far north as Dom&r 
(No. 97), but by far the greater number were found 
in the Hauran, and in the district of Hejra. Those from 
Hejra are grave-inscriptions, and form one of the most 
important contributions to our knowledge of Aramaic. In 
addition to this, they are tolerably connected pieces of 
writing, and those who desire to obtain a grounding in 
Aramaic epigraphy will find these inscriptions the simplest 
to commence with. A small group of inscriptions have 
been found also at Petra, two of which are of considerable 
importance (Nos. 95 sq.). 

Palmyra, the ancient Tadmor, has provided a great 
number of inscriptions, quite distinct from the Nabataean, 
as regards palaeography, the general character of the con- 
tents, and, to some extent at least, the language. These, 
too, are chiefly due to Arabs, although apart from the 
proper names the Arabisms are not so pronounced as in the 
Nabataean. Many of these, notably the Tariff of Palmyra 
(No. 147), are bilingual (Palmyrene and Greek). Finally, 
as a palaeographic subdivision of the Nabataean, we must 
include the inscriptions from the Sinaitic Peninsula, con- 
sisting of hundreds of rude graffiti scratched upon the rocks 
by traders. Mr. Cooke's selection (Nos. 103-109) contains 

1 No. 102 and CIS ii. 157. The latter a bilingual in Latin and Nab. was 
erected by AMaretas (mm») of Petra. 



NORTH-SEMITIC EPIGRAPHY 271 

the more interesting examples, and illustrates the extent 
to which Aramaic enters into them. With rare exceptions 
the names are entirely Arabic, but the phrases (" blessed," 
" hail," " son," &c.) are Aramaic, and one meets with such 
noteworthy combinations as " Hail (obw) ! Garm-al-ba'li, 
son (na) of Ibn-Alkaini ; good luck (3B3) V The article 
bx, as in the names just cited, agrees with the Arabic, 
whilst it is rather remarkable that in the inscriptions 
from Safa, written in a North- Arabian script, it is regularly 
n 2 . A few graffiti of the same character as the Sinaitic 
were recently discovered in Egypt in the Wady Gadammeh, 
NE. ofKeneh 3 . 

The extensive use of Aramaic by people other than 
Aramaeans is reflected in the language of the inscriptions. 
From the evidence of the proper names alone we are able 
to recognize the presence of Assyria and Babylonia, Persia, 
Egypt, Arabia, Greece and Rome. Assyrian influence 
appears notably in the Zenjirli inscriptions, and in the 
lion-weights and contracts ; it has even been suspected in 
the old inscription from Tema 4 . Persian seems to appear 
upon seals in such names as miens, JHrnt*, ironno, in such 
words as ru"nnD[N] " sepulchre," in the Palmyrene NtD3J"lN, 
and especially in the Cappadocian inscriptions referred to 
above. In Egyptian- Aramaic, in addition to Persian, 
Egyptian words are taken over ( , ann, noriDn, xruo and fern. 
xrwri), and to the presence of Jews we probably owe such 
Hebraisms as E*K " man " and npb " take." Arabisms 
appear notably in the Nabataean inscriptions, viz. the 

' CIS ii. 787, cp. No. 104. 

a Thus the Nab. mpto'c appears in Safa as Dprw© 140 B. 4 ; see Lidzbarski, 
Epliemeris, II, p. 36 sq. 

3 To be published in an early number of the Proceedings of the Society of 
Biblical Archaeology. 

1 e.g. Tiglath-pileser "lord of the four parts of the earth" (62 14, 
63 4) ; "po TO or -pa >t 'n = mana ia iarri (66), yiD=sinibu " half" (CIS ii. 
No. 7, but unc in No. 10); rm — dannitu (ibid., 17 sqq.) "document"; 
see also Cooke on 69, 11. 13 and 18 (where rai, second line, is a 
misprint). 



272 THE JEWISH QUARTERLY REVIEW 

construction pp |D for pP H }», the precative perfect, the 
use of ])h "curse," fill, •{?&, &c. (see note on No. 80 6), and the 
conjunction a (found, however, in the Zenjirli inscriptions). 
Technical Greek and Latin terms predominate in Palmyrene 
(p. 264 sq.). 

Considering the history of these Aramaic inscriptions it 
is not surprising that they reveal numerous traces of 
dialectical variation. Mr. Cooke's helpful notes (especially 
pp. 184 sq., 264 sq.) will give the reader some idea of their 
characteristics. The following supplementary remarks 
may, however, be added. In the oldest series, the Zenjirli 
inscriptions, it is important to distinguish between the 
latest, that of Bar-Kekub (No. 63), and the earliest, that 
of Hadad (No. 61). The former is essentially Aramaic 
and tolerably straightforward, the latter is most marked 
by non- Aramaic elements and is extremely obscure: the 
Panammu stone (No. 62) comes midway as regards intelli- 
gibility and philological phenomena. Whether the pecu- 
liarities are due to the mixture of dialects or to the gradual 
introduction and growth of Aramaic may be left an open 
question, at the least the endeavour must be made to avoid 
confusing features which are regular in Old Aramaic with 
those which are confined to one or more of these inscrip- 
tions alone. Of the forms and usages characteristic of 
Aramaic, as cited by Mr. Cooke, ru« " I," NDia " throne 1 ," 
P"13T " great " (plur.), and the emphatic form are found only 
in No. 63. The best examples with Canaanite analogies 
are to be found in No. 61 ("JJN " I," -or, "it, ppn, Kin, and the 
infinitive without prefixed »), to a less degree in No. 62 
(e. g. ^JN " I "), and scarcely once in No. 63. The Canaanite 
OJ and npb occur in Nos. 61, 62, the Aramaic in " one " 
and 13 " son " in all three 2 . It is a characteristic feature 



1 Also Phoenician, No. 15 2. 

2 It may be noticed as an instance of the facility with which foreign 
words may be adopted that the Aramaic -a is used even in the pre- 
Mohammedan Arabic inscription found by Dussaud at en-Nemara (cp. 
Lidz., Ephem., II, 35). 



NORTH-SEMITIC EPIGRAPHY 273 

of No. 61 that the masc. plural ends in »- 1 , and that this 
is merely a case of apocope of J- seems to be rendered 
doubtful by the fact that the fuller plural is |- and never 
p- 2 . It is equally characteristic that the fern. plur. in 
n- 3 appears only in No. 62, and that cohortative b is 
prefixed to the imperfect in No. 61. The forms in 1-, which 
are more frequent in No. 61 than in No. 62, are obscure ; 
they are possibly abstract nouns since the suffix 3rd sing, 
masc. is l-egularly n- (but note ttdk, mas and even nax). 

The consonantal equations are not entirely peculiar to this 
group, j? for V (= Heb. ¥) is best known from the form Np*lN 
"earth" (Jer. x. 11) which is found in inscriptions of the 
sixth century and later * ; but even in No. 61 30 we find 
"IX (■? i. q. ">¥ Dan. iv. 16), where this law would have led us 
to expect 1p. X for the normal B appears only in No. 63 19 
(NUPa) and one of the Nerab inscriptions (No. 64 12). B> for 
n soon seems to have died out : in the inscription from 
Cilicia (fifth or fourth century), in the T6ma inscription, and 
in Egyptian- Aramaic the forms tont* " place," aiva " seat," 
and n»n "there," are in agreement with the ordinary 
Aramaic usage. On the other hand, t for 1 was tenaciously 
retained (p. 185, note 1) down to the coins of Mazdai 
(fourth century) 5 and the Cappadocian inscriptions already 
referred to. 

1 But »nV» Dip No. 62 23, and w (constr. state?) ibid., 1. 3. The latter 
finds an analogy in Sabaean usage. 

3 ]>-, however, in lion- weights (J'3D, also pn), and later in Nab. and Palm, 
inscriptions. 

3 The reading rwp in No. 61 10 is not quite certain. 

4 The occurrence of the form in the Cappadocian inscriptions (Lidzbarski, 
Ephemeris, I, p. 323, 1. 3), shows that it was in use two or three (or more) 
centuries later. That the o/to/>«a of Berossos stands for NpJNTJN (Gunkel, 
Zimmern) is not generally accepted ; see King, Seven Tablets of Creation, I, 
p. xlvi note. To illustrate the equation, mention may be made of 
Halevy's ingenious suggestion that the form of the p was originally derived 
from the circle y by the addition of a stroke, but see Lidzbarski's 
criticisms, op. cit., p. 264. 

5 The early exceptions >v»«ra© (CIS ii. 87), Yisnn and |rDX (No. 150 2, 3) 
are doubtful. 

VOL. XVI. T 



274 THE JEWISH QUARTERLY REVIEW 

The close relation between the inscriptions and the Ara* 
maic portions of Ezra (fourth century ?) and Daniel (early 
second) is of considerable importance. Apart from the more 
frequent use of the construct state, there are several linguistic 
features of interest which call for remark. The use of n 
where the later Aramaic uses X is noticeable in such words 
as hjn " I," }n " if," and especially in the causative and re- 
flexive conjugations. As regards the causative forms, the 
older inscriptions (including Eg. -Aram.) regularly use -n, 
Nabataean (with the exception of D*pn, No. 97 col. i) and 
Palmyrene equally regularly -N ; in Biblical Aramaic -n pre- 
ponderates. For the reflexive forms, No. 63 14 presents -n, 
No. 65 4 (probably) -N, and the latter is usual in Egyptian- 
Aramaic, Palmyrene and apparently Nabataean ; in Biblical 
Aramaic, the forms with X occur only in Daniel. Older 
forms are preserved in TCN (Nab.), M1SJN Dan. ii. 46 (cp. 
"B3S No. 69 14), roruN Ezr. iv. 16, -ran (? No. 75 2); sus- 
pected Hebraisms prove to belong to the genuine Aramaic 
stock ', and in two cases an incorrect division of words can 
be remedied 2 . The difficult use of an* in Ezr. v. 16 finds 
support in No. 102 6, and the papyrus of Darius II (p. 269 
above) not only mentions among officials the NW (Ezr. iv. 9) 
and N^riS'r) (Dan. iii. 2), but contains the interesting phrase 
■ujnv tin (cp. Dan. ii. 5, 8). 

To supplement the remarks upon the Palmyrene inscrip- 
tions (p. 264 sq.), it may be observed that the dialect in 
several respects is younger than Nabataean (not to mention 
Egyptian- Aramaic, &c). In one or two instances it agrees 
with the Aramaic of Daniel whilst the Nabataean finds 
analogies in Ezra. The best example is the suffix of the 
3rd plur. masc. which in Ezra, Nabataean and older 
Aramaic is nn- (defective), but in Daniel, Palmyrene, 

1 e. g. stun (in Nab.) Vm (but only in 61 ; it occurs, however, in Pahlavi, 
Noldeke, GOA, 1884, No. 26, pp. 1014 sqq.). rorUM is cited from the Eg.- 
Aram. papyrus of Darius II. 

2 tap to, as top^ (67, &c.) shows, must be read as one word, and for H3 ttn, 
Dan. ii. 43, we should read >T -]*tn, cp. Palm, n -|>n (Schulthess, ZATW, 
1903, p. 164 sq.). 



NORTH-SEMITIC EPIGRAPHY 275 

Jewish- Aramaic, &c. Jin- 1 ; the inscriptions regularly keep 
the forms distinct and on this account the occurrence of 
pn- in Ezra can scarcely be original 2 . Further, rbn, Ezr. 
v. 15 (keth.) , agrees with Nabataean, whereas ]btt, the regular 
Palmyrene form, corresponds with the fO)^N 3 of Daniel. 
Occasionally, also, the keri marks a later stage in the 
language which we are able to check by means of the 
inscriptions. Thus in Dan. iv. 16 *tn» "my lord," with X 
agrees with the older inscriptions (including Nab.), whilst 
"■"lO (keri) is in conformity with Palmyrene and the later 
forms (e.g. Syriac) 4 . In Dan. iv. 13, the kSri NCMN is 
in agreement with Palmyrene and later Aramaic, the 
Nabataean alone has KBUK. The lack of a special feminine 
form for the 3rd plur. perfect and nominal suffix (which 
is supplied by the keri, e. g. Dan. v. 5 1pB3, k. npD3 ; ii. 33 
\\mn, k. JW») finds its parallel in Nabataean as distin- 
guished from later usage 5 . 

The following examples of the difference between the Pal- 
myrene and the Nabataean (and older inscriptions) may be 
noticed: mj, "loculus" for a corpse (Targ. 70), regularly 
in Nabataean, Palmyrene no[l]a (No. 91 5). sumDX in 
No. 96 2, a rare instance of a Greek loan-word in Nabataean, 
Palmyrene transliterates with more precision 3D1QDN (arpa- 
r-qyos) — the latter agrees with later Jewish and Syriac; 
on the various Nabataean forms, reference may be made 
to Lidzbarski, Handbuch, p. 22a, or my Glossary, p. 23. 



1 Palm, rarely jn-, Jerus. Targ. also Din- (Dalman, Gramm., p. 162). 

a Cp. a similar unnatural confusion in Jer. x. 11 (sp-is and «jnn). 

3 In this and in similarly cited words (e.g. Pal.-Syr. »[n]nM), the 
bracketed letter indicates a variant form, thus in Dan. both fm and ]»bn 
occur, in Pal.-Syr. both win and NnriN. 

* With the kfithib ttromta, Ezr. iv. is, cp. retention of N in Nos. 65 9, 
75 2. 

6 With the Haphel inf. minrr), rrawrrj (Dan. ii. 10, vi. 9, keri n»-, Strack), 
cp. Eg.-Aram. wpcn, nwi. In Palm, the only examples of an infln. (out- 
side the Pe'al) are KTnrw (cp. Jew. -Aram., Dalman, p. 227) and iiffinno. 
With the older wron (Ezr. vi. 17, keri rwtm), the Nab. mwon " penalty ■" 
(CIS ii. 224 11) may be compared. 

T 2 



276 THE JEWISH QUARTERLY REVIEW 

pl5f " righteous " in old Aramaic, in Nabataean only in 
a legal senBe (No. 82 3) ; Palm, uses p*U as in Syr. (Pal.- 
Syr., however, p1¥). The fuller form DJHJO "some-, any- 
thing" (with plur. NnojnJD in Eg. -Aram.) becomes DJHD, 
a stage nearer the Syr. *v». The Nab. JDI " time " agrees 
with Bibl.-Aram., the Palm. J3T with later Aram. fc?iO 
("head," "principal," p. 406) drops N in Palm. xnnJK 
"woman, wife," frequent in Nab., is shortened in Palm, 
to KnriN (also, but rarely, in Nab.), and even xriK, NrCN; 
obviously a popular pronunciation (cp. Pal.-Syr. N[n]n«). 
The older run "here," takes the form pn in Palm, in 
agreement with Syr. ^Jl. pnin "two" (fem.) becomes 
jrnri; for na p "except," Palm, has p T3, and for J113T 
"memorial" (only Nab.), Palm, employs fiai (once in Nab.). 
As regards the last-mentioned example, we may compare on 
the one hand, roVirn Ezr. vi. 2 and Pal.-Syr. P"0[*]l, and 
on the other, N^"? Ezr. iv. 15 and Syr. )j»oo» \ 

N becomes n in ni " this " (fem.)— note for Nab. WQ2 Hi, 
Palm, m nb>eu. Conversely, n becomes N in N*in " one," 
NO "what," Ntan "wheat." It is only in Palm, that the 
plural suffix TTi- appears as m- (except No. 62 2) ; pin* 
"they shall be" as }VT ; H as 1, and the plural emphatic 
to- as N-. The use of & to represent fe>, as in Biblical 
Aramaic (N'YEO, but N^D3 Ezr. v. 12), is regular in old 
Aramaic; D probably occurs once in Eg.-Aram. (Ninon), 
otherwise only in Palmyrene, where, apart from the reten- 
tion of ^ in tttMsa, Ntny, ~m, NWJJ, we meet with i»DD 
" left," 1HD " witness " and \JD " many," by the side of W. 

The general relation of the dialects of the inscriptions to 
Aramaic generally is an interesting inquiry into which 
space forbids me to enter. That they stand more closely 
akin to Western Aramaic than to the literary Edessene 
Syriac is not surprising when we consider their popular 
character. What can be culled from Jewish-Aramaic may 

1 With the exception of Nab. into CIS ii. 209 7 for Tirfa (twice), examples 
of variation in the use of 1 and < are found only in Palm. (n[>]a, «n[i]as, 
wDQ'A, i[i]n-!n, -u[\|i). 



NORTH-SEMITIC EPIGRAPHY 277 

be seen in Dalman's Grammatik des jiidisch-palastinischen 
Aramaisch, here I propose to notice a few examples of 
Palestinian- Syriac analogies. The examples from the 
inscriptions, unless otherwise indicated, are Palmyrene, 
and for the sake of clearness Palestinian- Syriac forms are 
transliterated in Hebrew. 

As regards consonantal interchanges, the old Aram. p"t¥ 
(above) is found also in Pal.-Syr., (Edessene) Syr. vo?) ; 
with tatPp (verb, p. 311), compare P.-S. NDBOp, but contrast 
Syr. )J^*.ao. As regards nominal forms : — Old Aramaic 
K3IVD (No. 70 1) = P.-S. N31JVD, Syr. IoIclso (? cp. Nab. 3n"ID 
No. 80 4) ; Nab. and Palm. Nmap» (p. 243), so P.-S., but 
Syr. JUasua ; MVVWO (No. 121 2), so P.-S., but Syr. Ik-lloo 1 ; 
vmxm (p. 337), so P.-S.; but Syr. J&wLX The Nab. 
TPltm (No. 94 3) and %3 "double" are used in P.-S. (note 
KTVVW, Schwally, Idioticon, p. 95), but not in Syr. Kmt? 
(No. 117 4) " beam," so in P.-S., in Syr. with a different 
nuance. The Nab. and Palm, nia agrees with P.-S., but 
Syr. Las?, Laa«; and for the Palm. 1TD1 p cp. the passages 
cited by Schwally, p. 44. With n ^T3 (p. 266), cp. P.-S. 
*t b"*i£, for which Syr. uses ^^» (also in Palm., but not in 
P.-S). wb . . . W3, so in P.-S. (see Schwally, p. 11). In 
No. 121 4, TiK finds analogies in P.-S., but vcb agrees with 
Syr. Lai*, against P.-S. N3fl (=Jb»ot) 2 . The accus. particle 
Tft, TV, in Zenjirli, Nab., and Palm, is rare in Syr., but par- 
ticularly common in P.-S. Finally, we may add the plur. 
constr. in K- (No. 126 4), not to be confused with the plur. 
emph. as in NTin (No. 113 3), and the defective spellings pi», 
))n (Nos. 139, 147 i. 10). 

The inscriptions have been carefully selected by 
Mr. Cooke, and he has succeeded in illustrating all the 
more distinctive classes. Nos. 8, 10, 29, 50-60, and 

1 In the note on 121 2 read Jl^-ok-wcc Mt. xxiv. 3 ; [Mn*]nM>[o] f Vogu6 
No. 16 3, would closely resemble the Syr. form, but the restoration is 
doubtful. 

2 Hence on p. 279, second line from foot, read "Syr." for "Pal. 
Syr." 



278 THE JEWISH QUARTERLY REVIEW 

perhaps the two Zenjirli inscriptions, Nos. 61 and 62, are 
scarcely suitable for beginners, but naturally their omission 
would have detracted from the value of the book. Between 
Nos. 66 and 67 a specimen of an Aramaic docket might 
very well have been inserted 1 , and particularly in the 
Nabataean and Palmyrene divisions other inscriptions 
equally suitable for beginners might easily be suggested. 
In view of the necessity of keeping the book within 
limits this criticism would be unreasonable, and the 
only remark that may be made is that it would have 
been advantageous to record more thoroughly in the notes 
parallels or analogies from those inscriptions which are 
not cited in full 2 . • On the other hand, the excellent 
manner in which Mr. Cooke has condensed the most im- 
portant details of the Punic inscriptions (see p. 137 sq.) 
deserves the highest praise. 

Mr. Cooke has aimed not at originality by proposing 
new interpretations or reconstructions, but at as much 
finality as can be reached in the present state of knowledge. 
His object has been "to give, after careful study of the 
various authorities on the subject, what seemed to be the 
most probable verdict on the issues raised, and also to 
bring together the chief matters of importance bearing on 
the text." He refers to the frequent use of "probably" 
and "possibly" in the commentary, and he reminds us 
" how seldom we can speak with positiveness on questions 
of grammar and interpretation where the material is so 
limited and where there is no contemporary literature to 
shed light upon the monuments." True though these words 
are, it is well to remind ourselves also of the great progress 
of epigraphical research during the last decade. The accu- 
mulation of material, the greater interest taken in a critical 
study of the subject, and the tardy recognition of the fact 

1 e.g. CIS ii. 38, 39. Those preserved in the British Museum have 
recently been re-edited by J. H. Stevenson (Vanderbilt Oriental Series, 
1902). 

2 In No. 116 3 p33cn can be illustrated by the Palm, jimy, &c. &c. 



NORTH-SEMITIC EPIGRAPHY 279 

that the inscriptions are real contributions to Semitic 
philology and archaeology, have raised Semitic epigraphy 
to the rank of a separate branch of learning 1 . Further, 
although the decipherment of the inscriptions is not rarely 
attended with great obscurity, it is well to recollect that even 
in the Hebrew of the Old Testament, despite the assistance 
of tradition, of ancient versions, of vowels and accents, it 
is not seldom that one is unable to speak with confidence 
respecting points of philology or interpretation. Indeed, 
the words " probably " and " possibly " enter more fre- 
quently into the vocabulary of the Hebraist than is 
realized by some. 

It is by reducing the limits of the possible and by 
ascertaining the extent of the probable — to follow 
Mr. Cooke — that we may hope to advance the study of 
North-Semitic epigraphy, and although in a great num- 
ber of cases there is little disagreement of opinion, it is 
lamentable how frequently the true meaning of a line or 
group of words entirely escapes us, and how widely the 
views of editors will differ in the most striking manner one 
from the other. 

The admirable manner in which Mr. Cooke has collected 
his material scarcely leaves anything to be desired. The 
thoroughness with which he has gone to work is made 
manifest in numberless instances 2 , and if a note is missed 
here and there, we must remember that in a manual for 
students, completeness was neither possible nor perhaps 
even desirable — if the work was to be kept within limits. 
In a few cases, too, we may question whether Mr. Cooke 
has really adopted the best interpretation, and occasionally, 

1 As Mr. Cooke observes (p. viii) it is to Paris that the distinction 
belongs of having recognized the study of Oriental epigraphy and archaeo- 
logy, by the foundation of a chair in the College de France. It is now 
held by M. Clermont-Oanneau, to whose brilliant studies every worker in 
this field is indebted. 

* Slight though it is, in No. 12 (CIS i. 10), 11. 3 and 4 are rightly divided 
(ab I »33') in agreement with the facsimile in the Corpus, but by a mere 
slip the editors and Lidzbarski wrongly divide cte | 3S>. 



280 THE JEWISH QUARTERLY REVIEW 

also, the steady advance of epigrapliic research places his 
observations in need of some qualification. 

In 11. n-ia of the Moabite stone, if the reading 
Ipn • 6yn ♦ ?a is adopted, a note on the construction should 
be appended, cp. Ges.-Kau. § 127. The pregnant construc- 
tion by TrtN 1. 14 deserves remark, cp. Deut. xx. 10. In 1. 25 
bx"\W note, point TICK or *!?« : in Biblical Hebrew "VBK is 
used only in the singular. In 11. 8, 10 y\a is used precisely 
as in 1 Sam. ix. 4, 8, &c, of a small district. 

The Siloam inscription is confidently ascribed to about 
700 B.C., but see below, p. 287. 

To the note on 3 10 it is to be observed that the sing. 
]bn "god" is now known not to be confined to proper 
names ; in Costa, Nos. 16, 31, we meet with jnn bjni> f?vh pvb 
and jon i>jn vnpx f?vh ptd 1 . 

In the note on 5 4 it is not quite clear how Mr. Cooke 
actually interprets ""Dip. The translation (p. 31) " I adjure," 
is based upon the Mishnic DJlp, which is perhaps rather 
a minced oath, for |?1? 2 , whereas, as a reflexive pronoun of 
the first person singular (Syr. goooia), such a rendering is 
too elliptical (see p. 34). It will be observed that the 
translation " which I myself built " would excellently suit 
1. 4 (nwp rm), but is placed out of the question by the 
construction in 1. 20, where we can scarcely suppose that 
it has been inserted by a careless copyist who, with 1. 4 in 
his mind, took the word to belong to the sentence following. 
Still, this inscription is not free from errors, however 
slight (11. 6, 9, 11). 

In No. 9 4, an inscription of the second century from 
Umm el-'Awamid, there is a famous puzzle. In it 'Abd- 
elim vows " this gate and the doors : Tuawanani'ysi'e'N 
in the year 180," &c. Mr. Cooke cites three interpreta- 
tions : (a) and the doors " thereof (? ew) I made (nisjja) in 
fulfilment of it (??); I built this ClTO) in the year," &c. 
(6) " which (&x) are for the making (rbvsb) of the temple 

1 Lidz., Ephemeris, I, p. 39. 

a See Gr. JP. Moore, Ency. Bib., "Vows," col. 5054. 



NORTH-SEMITIC EPIGRAPHY 281 

(ro) I have finished ("lib); I built it," &c. (c) "which I 
have indeed made (rbysb) in the completion of (vfara) my 
building." Against these may be urged r\bsn " fulfilment " 
in a ; the bare TU and the verbal forms in b, and the affir- 
mative 7 in c (see also the note p. 47). Another rendering 
is possible, (d) " -which (7K»N) I have made in its entirety 
(vfaro); I built it in the year," &c. The unique bm 
which is thus postulated is not outrageous 1 ; the suffix in 
roan will of course refer to the erection of the gate and 
doors as a whole. 

It is not easy to rest content with the usual interpreta- 
tion of No. 16 (CIS i. 46): "I 'Abd-osir, ... set up this 
pillar in my life-time over my resting-place for ever ; also 
to my wife, . . ." 

• • ,A ? ♦ T\&t6) . d^ . Tim 33t?D . by • maw • "rad? . roro • • • "pa 2 

Parallel to this, is No. 21 : " This pillar which Arish . . . 
erected to his father . . . and to his mother . . . over their 
resting-place, for ever." 

. oi>yS> tnnru mpo bv " ' *ok^ • • * vxb • * • wd* pk tn rmo 

Not only have we in No. 16 " an exceptional instance of 
the D3VD being set up by the person commemorated during 
his life-time," but the combination of the prepositions 
3+»+? is most difficult. It is certainly true that in 
No. 45 5 the terminus a quo may be understood, but this 
does not suit No. 42 5. One might conjecture a particle 
D^ (cp. Syr. ]f^, late Heb. "'KO), "even," but this is too 
desperate a course. A proper name seems impossible. 

In No. 27 (CIS i. 93) it is not improbable that 
"03 }3 bv is really " for the sons of his (Mar-Yehai's) son." 
The genealogy is thus simplified. In No. 28, DTI Ty myb 
('Adrjvq. "EuTelpq Nikjj) is usually rendered "To 'Anath the 
strength of life," ty is thus derived from try (jc) be strong ; 
it is perhaps better to connect with Fiy (Sl£) take refuge. 

1 See Wright, Comp. Gram., p. 118. 

3 The dots indicate the omission of unnecessary words (chiefly proper- 
names). 



282 THE JEWISH QUARTERLY REVIEW 

This inscription is of some interest on account of the 
name of the votary, Baal-shillem, son of Sesmai (SeV/^aos), 
which reminds us of Shallum b. Sismai C"? 1 ??) in the Jerah- 
meelite genealogy, 1 Chr. ii. 40. Curiously enough the 
name of 2e'o>iaos, a Sidonian, has recently been found upon 
graffiti in a tomb near Beit Jibrin, at the site of the ancient 
Mareshah. The name Baal-shillem in No. 28 is replaced 
by Il/)a£i5rjjuos, the reason for which is not obvious. Whilst 
on the subject of names it may be pointed out that 
the Greek Novpjmos not only represents C*irti3 ("born on 
the new moon") No. 17 3, but, as a recently -discovered 
inscription proves, even EHTO 1 . That the familiar Punic 
bvYVTD (M«'/3/3aAos) is to be interpreted "gift of Ba'al" is 
certainly not convincing (p. 109); may one venture " Baal 
hastens"? 2 In No. 35, where D?BODB'K is represented by 
'Eovfjurekrinov, the view that " the reduplication of the in- 
tensive stem (D.5K') is not marked," does not seem to be very 
probable. Pi'el forms are easily recognizable in Balsillec and 
BaXa-iXXrix (pt5vJ?3) 3 ; the name under consideration was 
probably pronounced Eshmun-shalom. Mr. Cooke com- 
pares Aoju<n£Aws = r6:fDin (No. 32), but here we doubtless 
have the same form as in jn^D (Cypriote mi-li-ki-ya-tho- 
no-se, CIS i. 89), cp. Hebrew forms like Bnij? 4 . That tbwm 
(p. 57) means " peace be (to him) " cannot be regarded as 
certain in view of the Aramaic nrwny and the name of the 
Tyrian king 'E/ciu^aAoy (? bv^y, Jos. c. Ap. i. 21). 

mxo fDfc'K (No. 40) is another familiar difficulty. Eshu- 
mun- Aesculapius 'larpos is perhaps meant, and the epithet 
rilXD seems to be a denominative of i" 1 ?"^, " healing, restora- 



1 Repertoire d'Epigr. Sem., No. 388. 

* One naturally recalls Maher-shalal-hash-baz (Isa. viii. 1, 3), which like 
Shear-Jashub is intended as an omen. These may be plays upon familiar 
personal names, cp. with Shear-Jashub sucli names as Shubael and 
Eliashib. 

3 Worn down to Sa<r\r]xos in Jos. c. Apion. i. 21. 

4 May we compare the Sam. participial form wep? (Also Pal.-Syr. 
»a&A. , Am. ix. 6 [ed. Margoliouth], if it is not rather a perf. in 0.) 



NORTH-SEMITIC EPIGRAPHY 283 

tion" 1 . In a Sardinian inscription of the second century 
this interchange of n and 3 is not too serious a difficulty. 
In view of the late date, it does not seem plausible, more- 
over, to explain the verbal form k^qi (" he healed him ") as 
a verb treated after the manner of rf*?, with the retention 
of the radical , ; N" 1 - is preferably the later form of the 
suffix (with ' plene) as in the Neo-Punic N^N " his father." 

The meaning and origin of ?1 in Phoenician is still exceed- 
ingly obscure, (a) In the Marseilles sacrificial tablet (No. 
42 15) allowance is made for the man lax 71 DN tupo 71. 

(b) In No. 46 1 (CIS i. 175) the Decemviri renovate and 
repair " this slaughter-house (rotao) which is Doya 7*1." 

(c) In No. 45 reference is made to the sanctuaries and their 
contents : — 

.o?j?n ?ni •• * wo 73 711 pnn rota 7i 

With the last citation we should possibly associate (d) the 
Tabnith inscription (No. 4 4-5) : — 

.D3D 731 pn )7lK *K S)D3 }?1N '•tG 

From a and b the meaning "devoid of," "without," has 
been extracted. Obviously this suits neither c or d, nor is 
it at all natural to render b " this slaughter-house without 
steps," in spite of Exod. xx. 25, to which the Editors of 
the Corpus refer. That ?i in a is the Hebrew 71 "poor" 
seems extremely probable (see Cooke's note), and it is 
very probable, also, that it is not connected with b and c. 
As regards these, Lidzbarski has suggested " that which is 
damaged," on the assumption that repairs are commemo- 
rated (p. 128). Winckler's conjecture 2 that 71 is to be 
connected with the Assyrian dullu " work, workmanship " 
is plausible enough to be added to the note on p. 1 28. On 
the whole, however, since piN in d seems to contain some 
such meaning as " with me " (see note on p. 29), I am 
tempted to conjecture that in b and c 71 is to be inter- 
preted " together with," " including," or the like 3 . 

1 Cp. Jer. xxxiii. 6 where it occurs together with net as above. 

2 See Lidzbarski, Ephem., I, 301. 

s Halevy (Journ. Asiat., 1902, t. 20, p. 349 sq.), also, has identified "nw 



284 THE JEWISH QUARTERLY REVIEW 

In the note on No. 61 4 (un*) Mr. Cooke observes that in 
the Zenjirli inscriptions and in those of N6rab and T&na the 
impf. pi. ends in 4,, not, as is usual in Aram., in 4n. It should 
be observed, however, that the form, strictly speaking, is 
peculiar to No. 61 (the Hadad inscr.), elsewhere it is always 
a jussive, and the apocope of the i is in accordance with 
Biblical- Aramaic usage. In 1. 34 of the Hadad inscription 
atari (=?)iwi) in parallelism with pnn is probably "put in 
writing," cp. Nab. f{?WP (No. 79 7), but it is to be feared that 
in the present state of our knowledge only a small amount 
of probability can be attached to the interpretation of 
Nos. 61 and 62. 

In the note on 62 1 (fourth line) read -S&WC In No. 63, 
1. 17, irta n'a now appears to mean "house of Kalammu" 
(above, p. 268, n. 2). That in 1. 16 [n]e*i> is the older form of 
n'9 seems very doubtful. In Bibl. Aram., even, the uncon- 
tracted form TVK tub is usual, and the equation 8*1^]=: later 
Aram. n[v] would require an Arabic c**J. That the Aramaic 
W= Heb. & is not absolutely certain. 

The difficult inscription from Petra (No. 94, CIS ii. 350) 
must probably be recognized as dialectical. In 1. 3 $>3 n , -ixe> 
N7XN, summing up as it does everything relating to the 
sepulchre, is most naturally rendered " the rest of all that is 
near (it)." It is naturally connected with the Hebrew ?■$ 
" near, beside," and it is worth noticing that both n»lKP, and 
the form nvnK3 (1. 2) are distinctly reminiscent of Hebrew. 
Other peculiarities are the use of UK " these," and nap» wn 
" a contract to bury," for the usual *iapD$> ana , fpn ana , &c. 
To render JHpBB (1. 4) " and it is the order," seems difficult 
since 'a WNB would have been expected ; it is quite in 
keeping with the rest of the inscription to find l^i?? 1 , but 



and Vl in d, 6 and c, and treats it as the Phoen. equivalent of the Heb. 
OS. As regards the derivation, the Ass. edUu "to bolt" might suggest 
that Vi(n) originally had the meaning "conjunction," "joining." 

1 In the note on p. 243 pm (CIS ii. 163, &c.) should have been cited, not 
Ji3T (only in 236) ; see above, p. 276. 



NORTH-SEMITIC EPIGRAPHY 285 

a better rendering would be "and they (Du.sh.ara, &c.) 
ordered " ; cp. for the ending }*-, Dalman, Oram., p. 303, 
and pay on the Jewish-Aramaic mosaic of Kefr Kenna 
(Pal. Explor. Fund, Quarterly Statement, 1901, p. 376) 1 . 

In the Palm, inscr., No. 144, we read : '■ in . . . the year 
500 ; Lishamsh . . . has-given-a-share (*UI"IN) of this vault to 
Bonne .... I have-given-him-a-share (nmanx)," &c. The 
change of person if not unique is extremely awkward here, 
and perhaps the simplest plan is to assume a mistake for 
the third person and read n? nariK 2 . 

The valuable Tariff of Palmyra (No. 147), among other 
noteworthy features, contains several examples of passives 
formed by internal vowel-change. These are recognized 
by Wright, Duval, Sachau, Bevan, Marti, and others, but 
Mr. Cooke denies their existence upon grounds which are 
open to criticism. In the first place, let it be admitted 
that the Nab. "03N (102 5) and Palm, ntbm (Vogu^ 95 4) 
are too obscure to enter into the question; there remain 
then the examples in No. 147 and the Nab. inscr. from 
Medeba (No. 96, 1. 8). In the latter riTay vby H NfiTajn 
'ui MB>3 is correctly translated " and the above work was 
executed in the year," &c, but there is no note upon the 
form. As regards Mr. Cooke's objections (p. 334), it may 
be pointed out that it is extremely unlikely that the 
Biblical Aramaic DJpnn, nSDin, &c, are modelled upon 
the Hebrew, and the statement that the forms are only 
used (in Bibl. Aram.) in the perf. 3rd pers. ignores K??pi?Jji in 
Dan. v. 27 s . Finally, the difficulty of construing the 

1 To explain (»)nin2, 94 2, 95 2, from the Arab. w-» is contrary to the 
usual equation, we should expect the initial letter to be j> ; a similar 
objection has been raised by Noldeke against the identification of nisi and 
LjJ. (88 1). In the case of mns a connexion with wO>(, to raise a monu- 
ment, may be conjectured. 

2 Cp. Proceedings of Society of Bibl. Arcliaeokgy, 1899, p. 74. 

s The note i. 1. 13 (p. 335) assumes that ty is a passive participle. This 
is incorrect. Nowhere is the internal passive more clearly marked than 
in the forms ^|, ^3 , ng as distinct from the Pe'al partcp. pass, roa, rcio, 
np (Wright, Comp. Gram., p. 224 sq.). On p. 333, first line (the note on 
Vc« vti), "Pual" is a slip for "Hophal." 



286 THE JEWISH QUARTERLY REVIEW 

Palmyrene forms as actives remains insuperable. Objection 
must be taken to the vocalization of certain forms in 
this inscription. The careful distinction between "MO 
"tax-gatherer," and N ?3P "tax" (p. 332), or between 
^PBK and *??** (p. $^), is extremely useful ; but such 
attempts as H'Wi? (p. 338, on 1. 11), ^"l^a (p. ^ 2 , on 1. 1) 
are of no service to the student, who might infer that 
they actually existed. 

The sections on the North-Semitic coins, seals, and gems 
(Nos. 149, 150) are a welcome addition to the book, and 
may, we trust, lead to renewed research in this particular 
depai-tment. As regards the Jewish coins, Mr. Cooke 
accepts the view, now generally held, that those formerly 
attributed by Madden and others to Simon Maccabaeus 
belong to the period immediately preceding the fall of 
Jerusalem. They thus take their place midway between 
the Hasmonaean coins and those of the Second Revolt of 
132-135 a.d. 1 These coins, extending as they do over 
a period of about 270 years 2 , are of considerable palaeo- 
graphic interest. Although marked as a whole by certain 
characteristic forms, it is worth noticing that of the speci- 
mens on p. $$2» the third form of X is peculiar to the 
copper coins of the fourth year (p. 358 h 4) and the Second 
Revolt (132-135 A. D.), and the second form of n is charac- 
teristic of the Hasmonaean period. The general appearance 
of the legends upon the coins of the First and Second 
Revolt differs to a marked degree from the Hasmonaean, 
whilst the carelessness which distinguishes those of Eleazar 
the priest is most remarkable. 

The Jewish coins, as is well known, are stamped in the 
old Hebrew character, not in the Aramaic square forms 
which are to be found upon contemporary ossuaries, and 
ultimately came into general use. The script has several 
features in common with the Siloam inscription and 

1 Th. Reinach has quite recently returned to the older view, and a 
reconsideration of the question seems desirable (Jewish Coins, pp. 10-14). 
a " 170," p. 353, 1. 3, is a misprint. 



NORTH-SEMITIC EPIGRAPHY 287 

Mr. Pilcher has, on these grounds, proposed to place the 
latter at the close of the reign of Herod the Great 1 . The 
constructions n»N. t\btr\ cnxo and not* nso (2 5-6) certainly 
suggest a date subsequent to the seventh century, and the 
evidence which assumes that the conduit, in which the inr 
scription was discovered, was constructed by Hezekiah is 
far from being unimpeachable. Unfortunately, the history 
of Hebrew palaeography is very obscure, and in the present 
state of our knowledge it is impossible to fix the date with 
certainty. From the various seals and pottery stamps 
that have been found it would appear that in addition to 
a script closely resembling the Siloam, another was in use 
which approximated more closely to the Moabite and old 
Phoenician forms. No one can fail to observe that in the 
name *pvt (No. 150 6, Plate XI) both N and V have Archaic 
shapes, differing from, and older than, those upon the 
Siloam inscriptions, and yet the seal, in conformity with 
the usual view, is dated in the seventh or sixth century. 
The subject is one upon which it is unnecessary to enter just 
now, and it will be enough if it is made evident how 
urgent is the need for inquiry 2 . The palaeography of the 
inscriptions, it may be mentioned, is treated only inci- 
dentally by Mr. Cooke, as is only fitting in a manual of 
this character, but some characteristic specimens are 
reproduced in the plates at the end, and these and the 
three tables of Alphabets will amply suffice to introduce 
the student to one of the most fascinating departments 
of Semitic study. 

1 Proc. o/Soc. o/Bibl, Arch., 1897, p. 182. 

* For the history of the Hebrew alphabet, the first requirement is 
accurate copies of the seals, &c. The following points then require to be 
considered : (a) the character of the proper names upon the seals, 
a careful distinction being drawn between Hebrew, Phoenician, and 
Aramaic names ; (b) the modification of the forms, viewed in connexion 
with the development of similar modifications in the Phoenician and 
Aramaic alphabets, for which dated evidence is more at hand ; (c) the 
archaeological evidence, where it can be obtained ; e. g. the strata at 
which the pottery stamps, &c. were found during the excavations of 
Bliss and Macalister in the Shephelah district. 



288 THE JEWISH QUARTERLY REVIEW 

The foregoing remarks and criticisms only affect points 
of detail, and in many cases merely reflect the writer's 
personal opinion. They in no wise detract from the 
excellence of Mr. Cooke's careful work. No one save 
those who have delved among the inscriptions and have 
striven to keep pace with the rapidly growing literature 
of the subject can fully appreciate the labour entailed in 
the collection of the material which is here placed at the 
disposal of Semitic students. Elaborate indexes render 
reference easy; one misses, however, an index of the 
inscriptions which are to be found in the Corpus and in 
such notable works as Euting's Nabataische Inschriften 
or de Vogue's La Syrie Centrale. If Lidzbarski performed 
a lasting service to scholars by rendering the inscriptions 
and their contents more readily accessible, the Eev. G. A. 
Cooke may be congratulated for his services in facilitating 
the study by removing most of the difficulties which had 
hitherto beset the path of the beginner. 

How extensive the field of research, how urgent the need 
for workers and collaborators must be admitted by all. 
Whether it is the study of the philological peculiarities, or 
the palaeography, or the evidence of the proper names, or 
the Punic passages in the Poenulus of Plautus, or the 
religious ideas and beliefs — there is much that requires to 
be done. To Jews in particular the subject should be one of 
especial interest. It is surely important to familiarize one- 
self with dialects so closely related to Hebrew as Phoenician 
and Moabite. Surely it is useful to gain some acquain- 
tance with ancient orthography and palaeography. What- 
ever may be our views of textual criticism, it is interesting 
to realize that some of the inscriptions themselves have 
come down to us with errors and mistakes, and if a stone, 
why not a MS. which has been copied and recopied ? 

The Egyptian-Aramaic documents in one case, at least, 
are Jewish, and the Oxford papyrus (pp. 404-6) contains 
the names of Jewish merchants and is evidence for the 
early settlement of Jews in Upper Egypt — possibly the 



NORTH-SEMITIC EPIGRAPHY 289 

oldest allusion subsequent to Jer. xliv. 1,15. The presence 
of Jews elsewhere may be suspected from the character of 
the proper names. In the Aramaic dockets we meet with 
such familiar forms as JJKnn, "j^N, Dmo, "itt&pa, $>tnny ; 
among the graffiti in the Sinaitic peninsula there appear 
iwn, TOP and ?ND"l; Palmyra numbered among her in- 
habitants p-o, }iyee>, and Trio (Martha). 

For the religion of the Semites the inscriptions are a 
storehouse of material. The Phoenician sacrificial tablets 
(Nos. 42-44) are of interest as much for their divergence 
from, as their agreement with, Hebrew ritual. The import- 
ance of the Moabite stone, with its mention of the divine 
name TYfff , has already been mentioned. In Palmyrene, the 
god 'Azizu is called KJcm "the compassionate one," an epithet 
which is found upon votive inscriptions dedicated " to Him 
whose Name is Blessed " ; the phrases — like the Palmyrene 
Noi>]> Nno " Lord of the World " (or " of eternity ")— sound 
peculiarly Jewish (pp. 296 sqq.). Even DD2>7jn " Lord of the 
Heavens " — I need not stop to refer to Jewish analogies — 
is rendered in bilingual inscriptions by Zeus vx/noro?, a 
suggestive indication of the growth of monotheistic ideas, 
and of a reaction against polytheism. 

Stanley A. Cook. 



VOL. XVI. U