Skip to main content
Internet Archive's 25th Anniversary Logo

Full text of "The Divine Name in Exodus iii:14"

See other formats


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 byJSTOR. 

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 

Read more about Early Journal Content at 
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 


The Divine Name in Exodus iii. 14 



THE section of the Book of Exodus in which this verse 
occurs is a familiar one. It describes Jahweh's initial 
appearance to Moses and the latter's commission to deliver 
the Children of Israel from the Egyptian bondage. The 
Jahwistic and Elohistic sources have been so closely inter- 
woven at this point that the greatest difference of opinion 
exists among critics as to the attribution of the material. 
There are hardly two writers who do not disagree at some 
point or other in the course of the third chapter, while 
some make no attempt at a complete analysis. There is, 
however, universal agreement regarding the fact that the 
passage which has to do with Moses' inquiry after the proper 
name of the god of Israel's fathers and the ensuing reply, is 
to be assigned to the E source. Not merely has it the ear- 
mark of the appellative D'nbx, but the J source has no room 
or occasion for such an episode, whereas the E document 
almost requires it. Dillmann, Wellhausen, Kuenen, Jiili- 
cher, Kittel, Driver, Cornill, Bacon, Baentsch, Holzinger, 
Moore, Carpenter and Harford — all are agreed that yss.^**"!® 
contain no J material, though a number of them maintain 
that redactional elements are not lacking. That is the posi- 
tion assumed in this paper. 

Nor, for the purposes of our discussion, does it make any 
difference whether or not we hold with Steuernagel that all 
the subsequent passages in the E document in which the 
name mn^ occurs, together with vss.^'^ of this chapter, in 
which the name is formally introduced, are additions from 
the hand of E^ the original E source knowing nothing of 


this disclosure of the name and continuing after as before 
to employ D^^7l^.^ Nothing that we have to say is in any 
way affected by shifting the entire series of HliT passages 
from the middle of the eighth to the middle of the seventh 
century B.C. We may consider the whole of E a document 
of the middle of the seventh century, disregarding the fact 
that the greater part of it was taken from an earlier written 
source. That document introduced the name iVSV for the 
first time in the call of Moses, and employed the name from 
time to time thereafter. 

In any case, this passage gives us much more than we re- 
quire. It is not content with introducing the name mn"" at 
the appropriate point in the Elohistic narrative ; it goes on 
to obtrude — or rather, it begins by obtruding, two occult 
expressions, which have some shadowy connection with the 
name to be communicated. Instead of the expected iTin\ 
God answers Moses' inquiry, " When the children of Israel 
ask me thy name, what shall I say unto them ? " with the 
enigmatic iTHi^ ^ti?>< iTHJ^, and continues, " Thus shalt thou 
say unto the children of Israel, (THS hath sent me unto you." 

In spite of the fact that what we want, and what Moses is 
represented as wanting, is a name, and not a definition of 
the name, or a characterization of the person bearing it, it is 
assumed (so far as I know, by all scholars) that the writer 
of vs. 1* here puts into the mouth of Jahweh at once a defini- 
tion of his name and a characterization of his person. That 
the definition is more or less opaque, and the characterization 
more or less unintelligible, are not held to be very serious 
defects. However much they may have discommoded 
Moses, they furnish rare opportunities for the gyrations of 
the " religionsgeschichtlicher Luftballon." Ordinarily, this 
definition and characterization is not, among modern schol- 
ars, taken to represent anything but the mind of the indi- 

^ Studien und KritiJcen, 1899, pp. 339 ff. ; anticipated by Bacon, Triple 
Tradition of the Exodus, pp. lii and 23, and in part by Welltiausen, Compo- 
sition des Hexateuchs\ p. 72. Tlie more defensible position is to abide by 
Welliiausen's suggestion, assigning Exodus 3i*-i5 to tlie original source, while 
attributing the subsequent TTSV sections to a later band. 


vidual Elohistic writer, who labored unsuccessfully to bring 
to expression the product of his own theological speculations. 
Kautzsch, however, asserts that "the rejection of the inter- 
pretation [of the name Jahweh] offered in Exodus 3^* [slip 
for ^*] involves the conclusion that even the early sources of 
the Pentateuch were in error as to the true meaning of the 
most important and most sacred Divine name in Israel," and 
asks, "Can it be supposed that at the time of E (c. 750 B.C.) 
the living apprehension of the genius of the Hebrew lan- 
guage was no longer adequate to interpret correctly a name 
like Jahweh? We cannot help thinking that this question 
has been answered in the aifirmative far too hastily by those 
who follow the prevailing current of opinion on this sub- 
ject. And we are only strengthened in our conviction 
when we note the extremely varied interpretations which 
have been proposed as substitutes for that adopted in 
Exodus 315 [1*]." 2 

I think it can be shown (against Kautzsch) that Exodus 3^* 
affords no data for the scientific determination of the origin 
and meaning of the name Jahweh ; and (against the current 
view) that this verse is in no way concerned with the subject 
of the origin and meaning of the name, and accordingly does 
not even give us the writer's views upon that subject ; and 
finally that the phrase n\"tJ< IV^H iTHK in vs.i*« and the word 
nflX in 1**, upon which all this theorizing has been based, 
were not to be found in the E document, but came into 
the text of the completed Pentateuch several hundred years 
after the middle of the seventh century, certainly not long 
before 300 B.C. 

One cannot operate with uncertain quantities. Our first 
task must be to remove all doubt as to the priority of the 
Masoretic text, and our second task will be to determine 
the meaning of its language. Only then shall we be in a 
position to deal with the literary and historical questions 

* Article Beligion of Israel in the Extra Volume of Hastings's Dictionary 
of the Bible, pp. 625 f. ; cf. the same author's article on Divine Names, En- 
cyclopcBdia Bihlica, col. 3323. 


The question as to the text may be finally disposed of. 
Nowhere is there any trace of an original Hebrew other than 
that of our current editions, viz.: rUTtt hn n^^'7« ni!3«''1" 

^:rhv? .Tnx b^-itt^"' "jab niaxn na nias^i n^'^« -itrs ^^^x 
^jnSir aps'' \nbxi pnii"' ""nSs Dn-D« ^^':>« d5tq« \nS« mn'' 

The variants of Hebrew manuscripts recorded by Kennicott 
and De Rossi are as follows : In vs.^*, one manuscript omits 
the initial '^l!3^<''^ ; one manuscript omits "I^S? ; one manuscript 
repeats "l^i< ; one manuscript omits ''i^b "IttKH nS 1l!iS''1 
^^^>< 7>?"1^^;3 all these indubitably represent careless de- 
parture from the current text. For ''327 of vs.^**, thirteen 
Hebrew manuscripts of Kennicott and twelve more of De 
Rossi have, with the Soncino edition (1488) and apparently 
all Samaritan manuscripts, ^33 bx, as in vs.^; four more of 
Kennicott and ten of De Rossi were corrected to "'337 from 
first hand ''33 7X ; one of De Rossi began to write ^33 vit 
but changed to ''337, while another of De Rossi actually had 
''337 corrected to ''33 7S ; the difference is quite immaterial, 
though, to judge from vs.^^ the writer's style was not so 
uniform, rather varying for the sake of euphony, so that ''337 
of our text would seem to be the original. For the final 
D3''7i< four manuscripts spell D37X. In vs.^^a, two manu- 
scripts have *nS d^'^7X in place of D^'^7^< IIS ; one manu- 
script inserts ^^'^S between ntyO and ri3 ; three manuscripts 
have ''337, as in vs.^*, in place of ''33 7i< ; one manuscript 
has "Itt?" for hn.-i^" ; one manuscript has DS^Sk ''3n'?tr n^'^X 
before mn^ ; one manuscript has Dri"l3i< instead of D3''n3i<, 
repeating ^'^7^< Dn"l3K ; five manuscripts omit D3''n3S ; two 
Samaritan manuscripts spell D3''m3K ; four Masoretic manu- 
scripts have, with the Samaritans, 1 before piVf TOI^ ; one 
manuscript has nW for ''3rt7ty ; four manuscripts (two of 
them the same as in the case of vs.") have D37K for d3''7i<. 
In VS.156, six Masoretic manuscripts have, with the Samaritans 
and the Soncino edition, D7157 for D7S7, and five more have 

» This is the only possible interpretation of Kennicott's " .THS 2°... "USX"! 2° 
A 75." 


been corrected to Chvh.^ Finally, five Masoretic manu- 
scripts read "IH "iTlv, and two more did so originally ; one 
manuscript inserts the conjunction, while retaining the de- 
fective spelling, "ITl *Tl7, and three have the conjunction 
with plene spelling, TH] "1117. The Samaritans spell vari- 
ously, m irh, *im Trh, and "im irb, but all, apparently, 
have the conjunction. Of these variants, all that do not 
consist in mere difference of spelling are unmistakable cor- 
ruptions of our received text. "Ill "11, with the conjunction, 
occurs much oftener, especially in the later literature, but 
"11 11, though it occurs only twice elsewhere in the Old 
Testament, is unquestionably the more idiomatic Hebrew. 

Coming to the Alexandrian Greek version. Codex Alexan- 
drinus (A) and Codex Vaticanus (B) agree to the letter 
as regards all but the two words bracketed below, which 
are omitted by A. I have derived the readings of A and 
B directly from the photographic reproductions of the 
manuscripts: ^*Kal etTreJ' 6 0eoi tt/jo? Ma>va-rjv [\€ya>v'\, 
iym elfu 6 wv • koX elirev, ovrcoi ipeK roll vloi<: 'IcrpaijX, 6 wv 
airearaXKev fie "Trpof vfidf. ^koI elvev 6 ^eo? irdXiv tt/jo? 
M.mvcrrjv, ovra^ ipelf roi<s vloi^ 'lapa-qX, Ku/3to9 6 ^eo? tS>p tto- 
repcov vfiS>v, 0eo^ 'A^paa/x koX 0£o<; 'Icro^/c Kot 6eb<! 'lawwyS, 
aireaTaXKev fie tt/so? vfidi ' rovro fiov ianv ovofia almviov koX 
fivr}fx4<TVVov \j^eve&v\ yeveaif. 

In the omission of X^cov A is supported by Codex Am- 
brosianus (Lagarde F)^ — which with Alexandrinus is left 
unnoticed by Holmes in this connection — and, according 
to Holmes, by Codex Coislinianus (Lagarde M), the only 
remaining uncial that contains the passage, besides the cur- 
sives. Holmes 14, 15, 18, 25, 30, 32, 52, 55, 56, 57, 59, 64, 71, 
74, 75, 76, 83, 84, 106, 107, 108, 181, 134, 135, the Complu- 
tensian, Aldine, and Grabian editions, the text of Nicephorus, 
and several daughter versions. The weight of evidence, with- 

* Fanciful Rabbinical speculations bear express witness to the universal 
defective spelling of the word in this passage in earlier times ; see b. Pesahim 
50a, Qiddushin 71a, j. Toma iOd, and Shemoth Sabba, ad loc. 

s See Ceriani's edition of the manuscript, Monumenta sacra et pro/ana, 
vol. iiu pp. 25 f . 


out any regard to the Hebrew, obliges us to exclude Xeyav 
from the Greek text. That done, the text of A and B is 
entirely in accord with the Masoretic Hebrew. The conjunc- 
tion before ^eo? 'laauK and the uniform rendering of ''327 
and '33 7X could hardly be avoided in the Greek ; while the 
dependence of fivrjfwawov on roM /lov iariv was too obvious 
to tolerate the repetition of that phrase. 'Eyd> elfii 6 &v is of 
course not the proper equivalent of nTIS ^ItTX iTHX, but it is 
not a proper equivalent of any imaginable Hebrew, and is 
more likely to be based upon this somewhat cryptical, than 
upon a more transparent Hebrew ; the original yielded the 
Greek translator no thought, only a verb, and he supplied the 
thought. As regards yeve&v yevealv of B and yeveaK of A, it 
would be sufficient for our purposes to point out that whether 
the Greek translator rendered yeve&v yeveali or merely yeve- 
ali, 11 *n7 must be assumed for the Hebrew that lay before 
him. In my judgment, however, A will represent the orig- 
inal Greek, though it is apparently alone in the reading 
yeveali ; for the order yeve&v yeveali, which is opposed with 
yeveah yeve&v by five cursives only (H. 19, 108, 118, 53, 72, 
of which the first three are " Lucianic " and the last " Hexa- 
plaric "), is not a natural one, and the single Greek plural 
word occurs elsewhere for the iterated Hebrew singular : «? 
yevedf is the rendering of 1Tv> IMt^ in Isaiah 34^" and of 
mm -in in Isaiah 61*. 

Among the numerous recorded variants from the above 
Greek text, there are, in the first place, certain manifest 
errors of omission due to homoeoteleuton : one manuscript 
(H. 53) skips from irph Mwva-rjv in vs." to the words follow- 
ing the same phrase in vs.^*; the same manuscript passes 
from 6e<k preceding r&v irareprnv vfi&v to the word 'Ay3- 
padfi; two manuscripts (H. 72, 74) pass from rot? i/t'ot? 
'la-pa^X of vs." to what follows these words in vs.^^; and 
one manuscript (H. 54) passes from r&v irarepmv v/jl&v of 
vs.^ to the words following the same expression in vs.^*. 

In the next place, there are a number of variants which, 
however relevant they may be to the reconstruction of the 
original form of the Alexandrian version, have no bearing 


on the question of the underlying Hebrew. Such are the 
readings elve Be or eltrev Se of a few cursives (among them 
the "Lucianic" H. 19, 108, 118) for Kal etirev in W" or i^; 
the form d7reo-Ta\«€, without final v ; the insertion in a large 
number of cursives and three early editions of the article o 
before 6eo<; 'A^padfi, 6eo<; 'laaaK, and ^eo? 'laawyS; the read- 
ing aireareiXe of the " Lucianic " manuscripts H. 108 and 
118 for cnrearaXKe in vs.^^ ; the reading fio{ for /lov in a few 
cursives; the insertion of to before ovo/ia in the Complu- 
tensian Polyglott. In the same category belong the omission 
of the initial /eat by five cursives in vs.^* and by two cursives 
in vs.^^ and of course the erroneous yevedf for yeveaK of 
H. 75, as well as the reading yeveaif yevewv referred to above. 
Lastlj', there are the variants which might bear on the origi- 
nal Hebrew if they represent the earliest form of the Greek 
version. Ten cursive manuscripts (H. 14, 16, 25, 32, 52, 54, 57, 
77, 78, 130) have for 0€(h of vs." Kupto? 6 de6<: ; the reading, 
even if it were more strongly sustained than it is, would have 
to be rejected on the merits of the question : DTI^S mn^ is 
utterly impossible in this connection. The same is true of 
the Complutensian's simple Kw/ato? for deof. Six cursives 
(H. 16, 25, 52, 54, 78, 131) and the text of Nicephorus omit 
the troublesome Kal elirev at the beginning of vs.i*^; the 
omission of the phrase is easily accounted for, but its intro- 
duction, on the other hand, would be quite unaccountable. 
An isolated manuscript (H. 83) goes one step farther and 
omits with Kal elirev the preceding iyd) elfii 6 &v, being doubt- 
less influenced by the tt/jo? Miovarjv ovto)? ipeii k. t. X. of vs.^". 
Lagarde's uncial M and one cursive (H. 18) omit fii in vs.^* ; 
one cursive (H. 106) omits o ^eo? iraKiv irpoi<Tr}V in 
vs.^^ imitating the koX elirev ovro)? e/jet? of i*& ; another cur- 
sive (H. 75) omits irdXiv, imitating vs.i*« ; and another 
(H. 72) omits wpcK Mavarjv of vs.^. All these almost cer- 
tainly involve only oversight on the part of the copyists of 
the individual manuscripts concerned. The fifth century 
Codex Ambrosianus (Lagarde F) and the seventh century M 
besides ten cursives (H. 18, 55, 58, 69, 64, 72, 84, 85, 107, 
134) have irdXiv o de6<s in place of o detn irdXiv^ while ten 


more cursives (H. 14, 16, 25, 32, 52, 54, 57, 73, 78, 131 — in 
the main the same that insert Kv/oto? in vs.^*), and the text 
of Nicephorus have TraXtj' Kvpioi 6 6e6v. Disregarding the 
element Kupto? on the grounds alleged above, one might be 
inclined to accept this as the original order of the Greek 
version, since it more closely conforms to the Hebrew TS 
D\"17S ; but the reading of A and B and congeners is not so 
easily disposed of, nor does that reading necessarily bespeak 
a Hebrew 1137 D^^7X. It is most reasonable to suppose that 
the reading irdXiv 6 6e6<; goes back to an early correction 
aiming at more rigid conformity to the Hebrew order. The 
variant irdkiv 6 Oeo?, tlien, will confirm the Hebrew text, 
while not invalidating the reading 6 Oeov irdXiv as the original 
form of the Greek version. Of the remaining variants none 
need detain us. H. 106 alone has trpbt avroix; for rot? wtot? 
'Icrpa^X oi vs.^^; the Sixtine edition, supported presumably 
by a number of Holmes's cursives, prints '^ficov for vfia>v ; and 
H. 55 adds Avrai fioi after vfiMV, in imitation of vs.^®. 

We must hold that the Hebrew which lay before the trans- 
lator of this passage in the Alexandrian Greek version was 
the same as that of our Masoretic text. 

That being the case, we can hardly expect to encounter 
any adverse testimony in the later versions. 

Of the other Greek versions we know no more than was 
recorded already by Montfaucon.^ According to the manu- 
script H. 64 (Montfaucon's Reg. 1871), Aquila and Theo- 
dotion had in vs." ea-ofiai ea-ofiat (for ^^'^K 1t!?K ^^^K), and 
Symmachus had in vs.^^ avdfivrja-i<; fiov (for ''"iSt). Of the 
former, Montfaucon remarks, "Videtur excidisse o? aclegen- 
dum ea-o/Mai 8? ea-ofiai, era qui ero, ut consonent cum Hebraico 
n^^« -ir?« n^'^S." So also both Bahrdt,^ "Videtur S? exci- 
disse," and Field, " ubi pronomen excidisse videtur." Field 
cites the rendering of Grsecus Venetus, ea-o/iai 8? ea-ofiai ; 
but the Jewish author of the Venetus had no need of Aquila 
and Theodotion to guide him to that rendering; — the Com- 
plutensian Polyglott, not many decades after, translated the 

• Hexaplorum Origenis quae supersunt, Paris, 1713, vol. i. pp. 60 f. 
' Hexaplorum Origenis quae supersunt, Leipzig, 1769, Part i. p. 268. 


n^HK 1CX iTili* of the Targum text with ero qui ero. That 
a relative pronoun has been carelessly omitted in the trans- 
mission of the rendering of those ancient versions is, how- 
ever, probable enough. 

The Targum of Onqelos, the Samaritan Targum, and the 
Peshita for Exodus 3i*'- are practically literal renderings of 
our Masoretic text. In all three versions the Hebrew n^HK 
^^^X "ItC^ of i*« and tTT\A of i*& are carried over with no 
attempt at interpretation. 

The consonantal text of Onqelos in the Complutensian 
Polyglott is as follows: n^^K -ItTK .THX ntrttb "^ ^ttXI" 

ms i!2xii5 pamS -'vh'^ .Tn« bx-itr^ ^wb la^n pia "laxi 
^^"IbK ittnnaxi «n'?K ""^ '^xntr^ •'lab -ittTi pis rwah ^^ 
'-at' pn psm"? ''Vb^ aps-n .1^^'?x^ pnic^n .Tnbx an-iaxn 

-im ^n "^dS ^nsn ptl abSfS. That of the London Polyglott 
is identical with this, save that it prints "'' for ''''''. Berliner's 
reprint (1884) of the Sabbioneta edition of 1557, besides 
employing "''', differs only in the reading iTTITi^ instead 
of ITTiSk before ^>r^'n and SpS^I. The isolated variant 
for vs.^* referred to by Berliner (Part ii. p. 183), I have 
not been able to ascertain. The Targum's habitual substitu- 
tion of ""^ for D^'^7K of the Hebrew is of course of no textual 

For the Samaritan Targum I transliterate from Peter- 
mann's edition (Berlin, 1882): .THX TW^ Sx HH^K "IfiXli* 

Xarh ''vb^ rr^nx h^'W" ^:^h -laTi jnax -laxi ^^■l« ity« 
^^bx r\xr bvirw ^jsS la^n pax ntrDb .inb« nw naxiis 
Xsrb ■'jnStr aps^ ^■I'?K^ pna:" ^^bK^ on-ias ^■I';'K pannat* 

"l-n "nb ^^an pi abwb ^a^^ p. The variants resulting 
from Petermann's collations are : in vs.^*, D^^7>< for T\Th^ ; 
nt'ttb for rW'Q bK ; pti for jUax ; nab for ''laS ; •'nittT and 
''nStr for ^3nbtt> ; in vs.i^, mX for n»K ; n^^bK for n.lbx ; 
pia and pa for pax ; laTl omitted ; "na"? for ''3a':' ; Thvt, 
for ^^bK ; Dn^iaxn H.-iSk for OnnaX ^1bK, and similarly in 
the two parallel phrases following ; for ■'JHT'tt^ as in vs.^* ; 

nab:jb and nbsS for rb^i-> ; ■'nan for nan ; and in"? for 

"117. The variants, when not scribal errors, are but differ- 
ent renderings of our Hebrew text. 


For the Peshita I give the text of Codex Ambrosianus 
from Ceriani's photolithographic reproduction : ln—^1 r-^h " 
wJ9|.A> 0L.<si) '^ ■! rr it . ilnS fic]Z llam jlelo si^n)^) oi_>oi1 } '^«<^S 

fcnfinV rt gisi_^1 . n »m i^^ sifli_2^) iomfS]i siai_^) ^ i^nt'^l? \at ^t 
•_«i;i^ M ij 10} eJoi >^*^'"'^ - '^ ^ a_Joi ^ i i S S \,h^p^ 
The texts of the London and Paris Polyglotts and the 
Urumia edition of 1852 accord with this, except that the 
Polyglotts have V-ljja-il for V-jJs-il, and the Urumia edi- 
tion has V-iJSLi and oi^oijjk,! ow.<nl. Aphraates, in the Demon- 
stration that Christ is the Son of Crod (written in the year 
655 Sel. = 343-344 A.D.), quotes from vs.^* ol-oiI i_*.| ouml 
and from vs.^^ tt*'?*}^ -Jjjo? °^<"o y<\\\ ^.ioa. oJoi.* 

The paraphrases of Targum Jerushalmi I (Pseudo-Jona- 
than) and Jerushalmi II (Fragment-Targum), though of 
some value for the history of exegesis, are worthless for 
purposes of textual criticism. 

Finally I adduce the Latin of Jerome, first from the 
Complutensian Polyglott, and next from the Clementine 
edition of 1592, the official text of the Roman Church. I 
supplement the abbreviations of the Complutensian in par- 
entheses : Dixit deus ad moysen. Ego sum qui sum. Ait. Sic 
dices filiis israel. Qui e(^st): misit me ad vos. Dixitque 
iterum deus ad mot/sen. Hec dices filiis israel. Dominus deus 
patru(ni) vestroruQun^ deus abraam ^ deus isaac ^ deus iacoh 
misit me ad vos. Hoc nomeCjC) mihi est in eternu(m) : ^ hoe 
memoriale meU(m^ t(n) g(^e')n(^er')atio(n')e ^ g(e)n(er^atio- 
neim). The Clementine text is as follows : Dixit Deus ad 

' The Homilies of Aphraates, edited from Syriac manuscripts of the fifth 

and sixth centuries in the British Museum, by W. Wright, p. eiS*'. The 

passage is lacking in Wriglit's index of Scriptural quotations. ^ ^ 

The Jacobite vocalization of the London and Paris Polyglotts is ai^tn] 
y . y ^ y X 
and oL^si) J.A.1 ok<ai|, while the Urfimia edition points the same syllables 

with (Nestorian) Zeq^fa; that is, the traditional vocalization both east and 

west is a constant a. Some manuscripts of the lexical works of Bar Ali and 

Bar Bahlul insert a helping vowel, "au.Bi|" (Payne Smith, col. 46); but 

see Duval's edition of Bar Bahlul, col. 45. In no case, however, should the 

fi Q t> a p 9 

phrase be vocalized oi^n ^) ol^si], as Holzinger, Exodus, p. 7. 


Moysen: EaO SUM QUI SUM. Ait: Sic dices fiUJs 
Israel : Q UI EST, misit me ad vos. Dixitq. iterum Deus ad 
Moysen: Hoec dices filijs Israel: Dominus Deus patrum 
vestrorum, Deus Abraham, Deus Isaac, S^ Deus lacob misit 
me ad vos : hoc nomen mihi est in ceternum, ^ hoc memoriale 
meum in generationem ^ generationem. The Antwerp Poly- 
glott, except in spelling hoec and ceternum, conforms to the 
Complutensian. The Paris Polyglott follows the Antwerp 
in all but the last clause, where for in generatione et genera- 
tionem it has only in generationem. Walton's Polyglott 
agrees entirely with the Clementine edition. The additional 
variants exhibited by manuscripts and editions of the Vul- 
gate recorded by Vercellone^ are: in vs.i*", Dixitque for 
Dixit; Dominus for Deus; in i*&, ait omitted; in ^^, iterum 
Dominus for iterum Deus. Of these only the second variant 
is exhibited by more than one or two unimportant mami- 
scripts. Codex Amiatinus has vestrum for vestrorum. Almost 
all manuscripts and printed editions, according to Vercellone, 
support the Complutensian against the Clementine in the 
reading et Deus Isaac, while a few manuscripts omit the con- 
junction before both Deu^ Isaac and Deus laeoh. Vercel- 
lone's testimony regarding the variant in generationem — in 
generatione is somewhat vague, but the Complutensian reading 
in generatione et generationem would seem to be the best sup- 
ported. It is not possible, from the unsystematic summing 
up of Vereellone's collations, to determine exactly the read- 
ings of certain manuscripts throughout the two verses. But 
there can be little doubt that the text of the Complutensian 
for Exodus 3i**- is in all respects the best supported tradition 
of the Latin of Jerome. 

Before considering the relation of Jerome's version to the 
Hebrew expressions iTHS "Itt^X ITTIK and tTns<, we should 
notice the freedom with which, here as elsewhere, while 
tolerably true to his original, he varies the Latin for stylistic 
reasons in cases where he must of necessity have had the 
same Hebrew. "1l!28<T is rendered dixit in vs.i*", ait in 1*6, 
and dixitque in 15; lasn HS is rendered sic dices in i*, and 
' Variae lectiones vulgatae Latinae Bibliorum editionis, Home, 1860-1864. 


haec dices in ^', IIV HI is Jioc nomen mihi est, but """llDl HT is 
7u>c memoriale meum. Under the circumstances we are not 
warranted in assuming that he had pni£^ ^'^7S^ against the 
Masoretic pTIT ^rhH, or bxitr"" ■'JS? in against the 
Masoretic t'SItt?'' "'33 b^. And whatever be the authentic 
Hieronymic text of the last clause of vs.^^ we need not look 
beyond the "11 "ll7 of our Masoretic text for the underlying 
Hebrew. I suspect, however, that the best Vulgate reading 
transmitted to us, in generatione et generationem, has resulted 

■n -lib. 

Turning to the renderings Ego sum qui sum and Qui est 
misit me ad vos, we must notice, in the first place, that they are 
not original with Jerome. It is a mere coincidence of gram- 
matical parts of speech that Ego sum qui sum appears to be 
nearer to the Hebrew JVn^ "WH n\"IK than does the Greek 
iym elfju 6 wv; as a matter of fact it is farther from the 
Hebrew. Both of the Vulgate clauses were adopted by 
Jerome from the Old Latin usage, based, of course, upon the 
Greek text alone. 

The following sources for the reconstruction of the " Old 
Latin" of Exodus 3^* were enumerated by Sabatier.^" I cite 
the quotations of the Fathers, and for convenience of refer- 
ence give the volume (Roman numerals) and column (Arabic 
numerals) of Migne's Patrologia Latina where each quota- 
tion may be found : Hgo sum qui sum, Novatian, De trinitate 
(iii. 920) ; Ego sum qui sum and further on Ego sum qui sum 
semper, Phoebadius Aginnensis, Contra Arianos (xx. 24); 
Ego sum qui sum. . . . Haee dices filiis Israel, Misit me ad vos 
is qui est, Hilary of Poitiers, Be trinitate i. 5 (x. 28) ; Ego 
sum qui sum, id. i. 6 (x. 28) ; Ego sum qui sum. . . . Sia 
dices filiis Israel, Misit me ad vos is qui est, id. iv. 8 (x. 102) ; 
Dixit autem Dominus ad Moysen, Ego sum qui sum. Et dixit. 
Sic dices filiis Israel, Qui est misit me ad vos, id. v. 22 (x. 144) ; 
Misit me ad vos is qui est, id. xii. 24 (x. 447) ; Dixit Dominus, 

1" Bihliorum sacrorum Latinae versiones antiquae seu vetus Itallca, et cae- 
terae quaecunque in codicibus mss. et antiquorum libris reperiri potuerunt, 
Paris, 1751. 


Ego sum qui sum. Dices, Qui est misit me, Ambrose, Epistola 
viii (xvi. 953) ; Ego sum qui sum, id. Enarratio in Psalmum 
xxxvi (xiv. 1054) ; Ego sum qui sum. . . . Misit me qui est, 
Augustine, Tractatus ii in Joannis evangelium (xxxv. 1389); 
Ego sum qui sum. Haec dices filiis Israel, Qui est misit me 
ad vos, id. Sermo vi (xxxviii. 61) ; Dices itaque filiis Israel, 
Qui est misit me ad vos, id. Sermo vii (xxxviii. 63); Ego 
sum qui sum. Et dices filiis Israel, Qui est misit me ad vos, 
F eviandus, E2)istola ad Reginum (Ixvii. 944). Sabatier men- 
tions further, Et dixit Dominus ad Mot/sen, Die filiis Israel, 
Hilary, In Psalmum cxviii (ix. 615); and Vade, die filiis 
Israel, Qui est misit me, Jerome, Commentarii in Isaiam pro- 
phetam, cap. lii (xxiv. 518); but the former is beside our 
point, and as regards quotations by Jerome, it would perhaps 
be chronologically more pertinent to cite Qui est me misit, 
Epistola XV (xxii. 85). To the above citations should be 
added the text of the sixth century Lyons manuscript pub- 
lished by Robert,^! Et dixit Deus ad Moysen, Ego sum qui 
sum. Et dixit. Sic dicis[&o] filiis Istrahel, Qui est misit me 
ad vos ; and the passage in the metrical Heptateuch of Cypri- 
anus the Gallic poet, who wrote in the early part of the fifth 
century and everywhere employs a pre-Hieronymic version,^ 

Sahmtor talia iungit : 
Ille ego sum qui sum, sic dices, et super astra 
Qui viget, estque deus vestrorum a stirpe parentum. 

(Exodus, lines 167 fi.) 

It may be questioned whether certain of the Fathers cited 
were not acquainted with and employing the Vulgate ver- 
sion, and some of the citations may owe their present form 
to corruption of the patristic text in the direction of the 
Vulgate. Still, a fair proportion of them will certainly 
preserve pre-Hieronymic readings. So that we may reason- 
ably assume that the common Old Latin text of our verse 

11 Version latine du Pentateuque anterieure as Saint Jerome, Paris, 1881, 
pp. 53 f., 168. 

12 See tlie discussion of Peiper, Corpus scriptorum ecclesiasticorum Latino- 
rum, vol. xxiiL pp. XXV f. 


ran as follows : Dixit (or Et dixit) deus ad Mot/sen, Ego sum 
qui sum. Mt dixit. Sic (or haec) dices filiis Israel, Qui est 
misit me ad vos.'^^ This Latin is easily derived from the 
Greek. Qui est is ordinary Latin for 6 wv,^^ and while Ego 
sum qui sum is perhaps more naturally interpreted It is I 
that am than / am he that is, the same is true of the Greek, 
as Greek, with the personal pronoun expressed.'^ 

Ego sum qui sum and Qui est misit me ad vos being derived 
from the Old Latin and historically based upon the Greek, 
the question we have to face is not whether they could or 
could not by any possibility be derived from the Hebrew, 
but merely whether they could in case of need be reconciled 
with the Hebrew. For generations these striking texts had 
been made the basis of theological speculation and demon- 
stration. Had Jerome desired to depart from the common 
Greek and Latin wording of the passage (which of course 
was not the case), he lacked the independent Hebrew 
scholarship which would enable him to do so without a 
well-defined Jewish tradition to occasion and justify the 
departure. But Jewish tradition, so far as concerned the 

M There is no more reason for Sabatier's taking into his reconstructed 
text of vs.i*« Dominus of Hilary, De trinitate, v. 22, than there would be 
for inserting Dominus adMoysen in vs.'**, after that author on Psalm 118. 

" Compare TertuUian's rendering of Apoc. 1', Adversus Praxeam, cap. 
xvii (Migne, ii. 199). 

1* Of course, / am he that is is the unmistakable intent of the Greek in 
this passage ; that is, 6 uv is the logical predicate, not the subject. A less 
ambiguous reproduction of the sense of the Greek would have been Ego 
sum is qui est. So the Hezaplaric Syriac has >Ainoi_i]; oei ^£^1 p] and 
^Zal^ y^ifMi ^jmal>^\y on; see Ceriani, Pentateuchi Syro-Hexaplaris 
quae supersunt, pp. 137 f. ; Lagarde-Rahlfs, Veteris Testamenti Graeei in 
sermonem Syriacum versi fragmenta, p. 52 ; and compare the quotation 
in the scholion of Jacob of Edessa published by Nestle, ZDMG, vol. xxxii. 
p. 490. Similarly, the Ethiopic : ana we'etu zahalo (I am he that is) and 
zahalo finaioanl (he that is has sent me); see Dillmann's Octateuehus 
Aethiopicus, p. 100. 

A third interpretation of Ego sum qui sum, as / am that lam, in which 
sum is a copula and nothing more, is quite outside of the intent of the Latin, 
as it is impossible in the case of the Greek. Qui est of i** is a sufficient 
demonstration of the fact. Our familiar English / am that I am represents 
the Latin no more properly than it does the Greek or the Hebrew, and 
it is high time the expression disappeared from scientific usage. 


construction and literal interpretation of the text, and not 
merely a more or less free speculation as to its basis and 
import,^^ was able to dodge the issue by rehearsing the 
expressions as the longer and shorter forms respectively 
of a divine proper name. And in fact there can be little 
doubt that had Jerome not found the Greek and Latin 
before him, he Avould, in spite of Aquila and Theodotion, 
have transferred the Hebrew^ vocables bodily into his own 
version, as did Targum Onqelos and the Peshita.^^ Jerome 

i«So the Babylonian Gemara, Berachoth 96: ".TriX 1!PK rrnK.— The 
holy One, blessed be he, said unto Moses, Go and say unto Isra«l, I have 
been (Tl'Tl) with you in this bondage : I will be (n\*lK) with you in the 
(future) bondage of the kingdoms (nvs'ra). And he (Moses) spake before 
him, Lord of the world, sufficient unto the hour Is the evil thereof 1 The 
holy One, blessed be he, said unto him, Go and say unto them ^irhtf n'lIK 
Q^'hu." Cf. Shemoth Sabba: "nflo b» D'n'rK IBin. — Rabbi Abba b. 
Mamal said. The holy One, blessed be he, said unto Moses, Thou seekest 
to know my name : according to my acts am I designated ; sometimes I am 
called nw bn, at other times niKax, at others Q'rb», at others mn\ When 
I judge mankind, my name is D'n'rS ; when I make war upon the wicked, 
my name is nixas ; when I bear with the sins of men, my name is ^"W bH ; 
and when 1 shew mercy upon the world, my name is miT ; for flVT is but the 
symbol of mercy, as it is written psitl Dim "tK mrP nilT' (Exodus 34«). So 
that my name is rtTIR "11PK IfflK because of my acts. Rabbi Yishaq says, 
The holy One, blessed be he, said unto Moses, Say unto them, What I was 
in the past, that I am at present and will be in the future ('3X1 'n'^ntP '3K 
KS"? TnC*? Kin '3X1 VB?3» Xin) ; therefore ITnX occurs three times (in this 
passage). And further : ITnX ~I»X rr<n». — Rabbi Jacob b. Rabbi Ablna, on 
the authority of Rabbi Hfina of Sepphoris, said, The holy One, blessed be 
he, said unto Moses, Say unto them, I will be (n'nX) with them in this 
bondage, and in the bondage to which they go (ps'?''!! P) I will be with 
them. He (Moses) spake before him, And am I to relate this to them ? — 
Sufficient unto the hour is the evil thereof I He said unto him, Nay, thus 
(only) shalt thou say unto the children of Israel, na''?X ^irhv ri'nX ; unto 
thee I make (the future) known, not unto them." 

A radically different elaboration is that contained in the Targums Jeru- 
shalmi I and II, which connects iTHX with the act of bringing into existence, 
or creation. 

" Among the ten Hebrew names of the Deity which Jerome enumerates 
in one of his letters to Marcella (Migne, xxii. 429) is ESER lEIE. The 
citation is from memory, and knowing his Latin text better than his 
Hebrew, he obtains the latter by what he assumes to be a re-translation of 
the former. He proceeds to explain that the name is the Qui est of Exodus 
31*', but evidently forgot that the Hebrew had no ")WX at this point. Of 
course the citation is worthless as a witness to the Hebrew text of his day 


is therefore not engaged in a free translation of Hebrew 
prose at this point, but imagines he is perpetuating the 
Greek exposition of the (so to speak, etymological) signi- 
ficance of the proper names HMS^ ItTK ^^'^X and iTHj^. 
That the former of these offered no obstacle, prima facie, 
to the perpetuation of the formula Ego sum qui sum is self- 
evident : the Imperfect of a Hebrew verb occurred often 
enough with present signification. iTHS would, of course, 
not be the same person as Qui est; but an interpretation, if 
it is to be one, must make sense, and Ego sum misit me is 
so palpably close to nonsense that the alteration of the 
Old Latin in that direction, on a mere point of the pre- 
cise grammatical form of a proper name, was not to be 
thought of.i* 

Like the Greek, Jewish Aramaic, Samaritan, and Syriac 
versions made directly from the Hebrew, the Latin Vulgate 
offers no occasion for questioning that the Masoretic text of 

(ESER lEIE = rrir -W», 3d person), but it does show clearly that his 
Jewish teachers, like the translators of the Targum, treated both expressions 
as names of the Deity. 

As a genuine divine name, distinguished from mere circumlocutory epi- 
thets, and yet in no way related to the tetragrammaton, iTHS ~IV^ ITns 
appears in the Jerusalem Gemara, Megilla lid, and in the Babylonian 
Gemara, Shebu'oth 35 a. For its potency as a divine name, the phrase (or 
some corruption of it) was also employed in incantations, especially in later 
times, and that even among the Mohammedan Arabs ; cf. Goldziher, ZDMG, 
vol. xlviii. pp. 359 f. Already in &. Baba Bathra 73 a, there is a sailors' 
yam about a perilous wave being laid low by means of missiles on which was 

engraved, rha jfix [as niS3S nin' .T ,Tn« itTK rrriK. 

w / am hath sent me unto you is a specialty of our own language, and a 
favorite one. Watson, in his translation of Hilary's De trinitate (Nicene 
and Post-Nieene Fathers, second series, vol. ix.), renders the Latin Misit 
me ad vos is qui est of iv. 8 with / am hath sent me unto you, after the King 
James version. Similarly, in Eremantle's edition of Jerome's letters (ib. 
vol. vi.). Qui est me misit of Letter xv (Migne, xxii. 35) is rendered / am 
hath sent me. Contrast the Wycliffite versions of Exodus 3". 

A proceeding somewhat analogous to Jerome's is that of Parisot, who, in 
the Latin translation of Aphraates's quotation above-mentioned {Patrologia 
Syriaca, Part i. vol. 1. col. 791) , makes the author adopt the formula of the 
Vulgate, Ego sum qui sum, whereas Aphraates quotes the Hebrew vocables 
of the Peshita version as he would a proper name, with no interpretation 


Exodus 31* *• goes back beyond the middle of the third cen- 
tury B.C. (the commonly accepted date of the Alexandrian 
translation of the Pentateuch into Greek) to some point 
prior to the final breach (whenever it may have occurred) 
between the Samaritan and Jerusalem communities.^® 

Our next step is to determine the meaning of the language 
of the text. This task is not a theological one, nor yet an 
historical one, but purely linguistic. As Hebrew, the phrase 
n^^« -1^« .Tn« (and the same is true of the following n^^S 
D3'vS ''Jn^t!^, in vs."*) can have only one, definite meaning, 
or one of several just as definite meanings. The context may 
be needed to determine which particular one of several mean- 
ings offered by the language we are to adopt, but the meanings 
from which the selection is made must be yielded by the lan- 
guage itself. And while a word may be used in pregnant 
fashion with allusion to a known circle of ideas, such conno- 
tation must be substantiated by more than assertion, and can- 
not in any case be intelligently discussed until the primary 
denotation has been grasped. Needless to say that the con- 
tribution of this or that interpretation to the requirements of 
a given theory of the development of the religion of Israel, can 
have no bearing upon the question. Even the renderings of 
the ancient versions and the exegesis based upon them, are in 
this case beside the mark : iTH^ is a common Hebrew word, 
and the construction iTHS "ItTi^ ^^'^i< is, as we shall see, a 

w According to Josephus, about 333 b.c. "Ueber die Entstehung des 
samaritanischen Schismas giebt es nur einen Bericht, den bei Jos. ant. XI. 
72 82- *, und der ist falsch " (Holscber, Palastina in der persischen und hel- 
lenistischen Zeit, p. 87). That it is our only account is quite true. Nehemiah 
13®'- does not profess to relate to such an episode, and cannot relate to it, if 
only for the circumstance that the cleavage was not consummated till after 
the Pentateuch had been both compiled and domesticated. But it does not 
follow that the statement of Josephua is therefore in its entirety fictitious. 
How the fact that the schism synchronized with Alexander's conquest could 
be derived from a "false exegesis" of Nehemiah 13-*'-, is not apparent. 
Josephus's statement on that point may or may not rest on good tradition ; 
that it does not, is not demonstrated by his erroneous application of the 
Nehemiah passage. For the rest, Holscher's own determination of the date 
of the schism, by means of the indirect evidence of the combined prophecies 
of Isaiah 56-66, as soon after the punitive expedition of Artaxerxes Ochus, 
does not materially conflict with that furnished by Josephus. 


not unparalleled Hebrew construction. The versions have 
rendered us their only possible service : they have borne 
their testimony to the integrity of the Hebrew text from 
the third century B.C. to the present time. 

The clauses just mentioned, .Tn« "lt7« iTHX and HflK 
D3''bx ''3nbtr, are the only parts of the text that present any 
difficulty and about the interpretation of which there is dis- 
pute. The second of these clauses can be dismissed with a 
few words. Whatever may be the literal meaning of the 
word iTHK, in this sentence it is in the nominative case, 
subject of the verb which follows it, and therefore a sub- 
stantive. Since, moreover, the word is in itself a verb in 
the first person singular of the Imperfect tense, as a sub- 
stantive it cannot be an appellative, but must be a proper 
name. And while a proper name may have a transparent 
etymology, it is not permissible to drag its etymology into 
the structure of the sentence in which it occurs. We neither 
interpret nor render MT2 b^ JDJ "[h'') of 1 Samuel 12i5, And 
gave went to his house. The question as to the origin of 
the name and the reason for its bestowal may be legiti- 
mately raised, but not in the act of rendering a sentence 
in which it is already employed as such. So that the 
only permissible interpretation and rendering of the He- 
brew sentence D5^7S ''Jn^tl^ ^^'^K is ''Ehyeh hath sent me 
unto you. 

The case is different with regard to HflS niTS .THK. This 
expression necessarily constitutes a complete sentence, for it 
is all that follows the formula ntTtt ■?« B^-I'?« "IfiK'^l. As 
one single proper name, 'i!hy^h-asher-ihyeh, it could con- 
stitute only one of two elements in the reply ; another ele- 
ment, either a verb or a noun, would be required. We should 
have something like ^ ilhyeh-ash^r-'ehyeh hath sent thee., or 
'Ehyeh-asher-ihyeh is my name, or I am '^hyih-asher- 
'ehyeh. As surely, then, as we must refrain from translat- 
ing the word iTns in vs."' must we face the problem of the 
interpretation of the sentence HNIK "ItTK iTi"I« in "". Nor 
can the first element of the expression be constituted a 
proper name, while the remaining ^^'^S "W^ is treated as a 


separate, subordinate clause explanatory of the name.^" To 
say nothing of the intrinsic absurdity of explaining a word 
that needs no explanation by repeating that same word, or 
of the unnatural construction of "Itt^X involved, the identical 
remarks apply to the simple iTHK that have been passed 
upon iTHK ItTK ^^'^K considered as a proper name: iTHK 
alone cannot constitute the sentence demanded by the intro- 
ductory formula ntt?ia Sx Dfl'?« ^tiK"!. 

What, then, is the meaning of the Hebrew sentence iTlTX 

^^"^« -nr«? 

The facts with regard to the literal rendering of this 
clause as Hebrew, no matter where or in what connection 
it might be found, were accurately set forth by Robertson 
Smith.*^ Nevertheless, recent commentators on the book of 
Exodus, and others who touch upon the subject, continue to 
tread, or to linger about, the path of error. A restatement 
of the case is therefore not uncalled for. 

*• So, after Ibn Ezra, Wellhausen, Composition des Hexateuchs^, p. 72, 
note 2, "Bin — sintemal ich bin"; and after Wellhausen, Baentsch. Pal- 
pably absurd, however, the construction only becomes with the interpre- 
tation adopted by Marti, Geschichte der israelitischen Religion*, p. 61 : 
" Ehjfe, das heisst ich bin," is what the author of Exodus 3" made God cause 
Moses to teach the Hebrew-speaking people. 

*i See Prophets of Israel % pp. 386 ff. They were summed up, some 
260 years before Robertson Smith, in the brief statement of Henry Ains- 
worth, " The Hebrew, Ehjeh asher ehjeh, properly signifleth, I will be that 
I will be," — Annotations upon the second booke of Moses, called Exodus, 
Beprint of 1639, p. 10 (first printed in 1617). Not so accurate, though 
possibly only because not so vague, are Luther and the Grsecus Venetus. 

Unfortunately, Robertson Smith acquiesced in the attribution of the 
expressions in question, together with the remainder of the section, to the 
original E source, and so was forced to twist his rendering into the require- 
ments of the context, with the result that the substance of our passage is 
thus set forth : " The sense is . . . that what He will be to His people He 
will be, will approve Himself to be, without fail. The vagueness is inevitable, 
for no words can sum up all that Jehovah will be to His people ; it is enough 
for them to know that He will be it (comp. Isa. Ixiv. 3 ; Lam. iii. 23)." The 
unreality of this result is its suflScient condemnation. 

Smith's conclusions had been previously published at greater length in an 
article "On the Name Jehovah (Jahve) and the Doctrine of Exodus III. 14," 
British and Foreign Evangelical Beview for January, 1876, pp. 153 ff., at 
the close of which he connects rpnx of Hosea 1* with this passage. His 
views were restated immediately thereafter by Nestle, Die israelitischen 
Eigennamen, 1876, pp. 91 ff. 


The Hebrew verb ^^^ does not mean to have being. The 
Greek rendering of riTIK, o wv, introduces a concept as 
foreign to the Hebrew mind as it is to the Hebrew verb. 
The Hebrews, as such, never attained to the conception of a 
Universe (as distinguished from a World), or of Infinity, or 
of one First-cause, or of an underlying Substance or Reality ; 
nor yet to that of the abstractest of them all, absolute and 
pure Being. Furthermore, the Hebrew verb ^^'^ does not 
mean to exist, except within certain limitations and in the 
French sense of the term. nTt is to come into existence, to 
happen, to occur ; to become, to take on (an attribute'), to enter 
upon (a state), to constitute (somewhat). Secondarily, since 
had become = was, and will become = will be, and having become 
= being (Gerund), ^^'^ comes to be employed for to be in the 
sense of the copula, and even in the sense of actuality, but 
only with reference to past or future time or in speaking of 
the copulative relation or the predication itself ; and it is so 
employed for the very reason that the Hebrew language 
ordinarily employs no verb at all to express the idea of 
being, and so cannot explicitly throw its thought into past 
or future time, or advert to the thought itself, without re- 
course to the verb become. Accordingly the Hebrew equiva- 
lent of I exist, if the occasion for such a declaration could 
be conceived of by the Hebrew mind, would be not !^^'^i<, 
but Tl^Tt (Perfect), I have come into existence and so am here. 
On the other hand, I am (something) as distinguished from 
I exist, would not make use of the verb ^^'^ at all. / am 
can only be expressed by means of a nominal sentence. The 
Hebrew for lam (so and so) is ''3K followed by the predicate 
noun (or adverb). Thus the Hebrew for I am that I am is 
not ^^'^S< IJTS ^^^^<, nor does it differ from that clause only 
in the matter of the tense of the verb. A nominal instead 
of a verbal sentence is required. The Hebrew for I am that 
I am is "JX Itt^X ^3«, just as I am he is «in ^3«. The Im- 
perfect iTni< can only mean I am in the act of becoming, or I 
will become, or I will be^ — which, in the looseness of Hebrew 

^ In order not to confuse the issue by distinctions of Mode foreign to the 
Hebrew, I adhere to the form of the English Future employed by Robertson 


thinking, amount to pretty much the same thing. In no 
case can it be rendered I am. And while the thought I will 
exist (^abide, in the future) would be theoretically possible in 
the latest stage of Old Testament writing, it is impossible 
in this connection, with the predicate ^^^K 'Itt^it immediately 
following it. iTn>^ in this sentence can only mean I will he 
or become (something) ; for of course I will he or become 
(somebody) is not a sensible alternative. Not merely the 
most natural, then, but the necessary construction of •^^^t< 
n^'^K "ntrx is / wHl he what IwHl he. So much for the literal 
meaning of the Hebrew clause. 

We have still the question as to the idiomatic or logical 
value of this tautological expression. And only at this 
point is it pertinent to consider the testimony of the paral- 
lels from the Old Testament, Neo-Hebrew, Aramaic, Syriac, 
and Arabic which are cited by Lagarde,^ Robertson Smith,^ 
and Griinbaum.^ In doing so, the reader should bear in mind 
that no number of parallels from other languages can affect 
the meaning of the Hebrew verb as determined above. Our 
own detailed enumeration of the parallels may be safely and 
advantageously limited to those of the Old Testament ; there 
are enough of them to determine the point at issue, and we 
shall thereby escape the influence of any turns of thought 
which may be peculiar to later times and other languages. 
The Old Testament parallels adduced by the writers men- 
tioned are : Gen. 43" Ex. 4i3 162s SS^ Deut. 9^ 1 Sam. 1^* 
2313 2 Sam. IS^o Ez. 122^ Zech. lO^ and Esther 4i6, to which 
may be added 2 Kings 8\ Of these, 1 Sam. l^*, ^53 n53m, 
has nothing in common with the construction we are discuss- 
ing. Neither has Zech. 10* 131 1tt5 13*11, if the current in- 
terpretation. They shall be as many as they formerly were, is 
the proper one. Deut. 925 ci,-, Q'^'s^lHi n« niH'' ^JSb hSiimi 
"-nSajrin -12>K rh-hn WVSl^ nxi. And I prostrated myself 
before Jahweh the forty days and forty nights that I did, is 
not exactly to the point, since there is nothing indefinite 

^ PsalteHum juxta Hebraeos Hieronymi, pp. 156 ff. 
«* Prophets of Israel^ p. 887. 
^ ZDMG, vol. ixxix, pp. 564 f. 


about the complement of the verb.^ The rest of the pas- 
sages it will be convenient to cite in the following order : 
a) Passages with the Perfect in the relative clause : Gen. 43^* 
TlW '•SnbS*!^ "ntriO ■'3S1, And as for me, howsoever I must be 
bereaved, I shall be; Esther 4i6 ^mSK "maX llTSai, And 
howsoever I must perish, I shall, b ) With the participle : 
2 Sam. 1520 I'^in ^J« ItrX bs "jSin ^3S1, (Thou oam'st but 
yesterday, and shall I start thee traveling with us to-day,) with 
myself going where I am? c) With the Imperfect in the 
relative clause : Ex. 4^* nWA 1^3 H3 nyS, Send by whatever 
hand thou ehoosest;^ Ex. 1&^ 1V^ n«1 1S« 1S«n nty« n« 
1W3 l7^3n, Bake whatever ye choose, and boil whatever ye 

choose; Ex. 8B^ DiriH 'vsn n« ^nianm ]m itTK n« Tarn, 

I will be gracious unto whomsoever I choose, and I will have 
mercy upon whomsoever I choose ; 1 Sam. 23^^ ^ItTJO 137nn''1 
ttbnn'', And they went roving wherever they chose; 2 Kings 8^ 
"•"nun *1Ul<3 ■'■1131, And live temporarily wherever thou choosest ; 
Ez. 1225 -la-i -i3-i>< '•j^^ n« nniK, I wUl speak whatever 
word I choose. From the data of the Old Testament it is 
evident that this indefinite tautological construction was 
employed by the Hebrews, in reference to an ensuing event, 
a) with the verb in the Perfect, to indicate the subject's 
subjection to a necessary though unknown fate, and b ) with 
the verb in the Imperfect, to indicate the subject's absolute 
control of his own action : the idiomatic value of ntJ?t< "'n''fl 
"'^''^^, when spoken of future time, is I shall be whatever I 
must; the idiomatic value of HflS *ltrs iTHi^, which can be 
spoken only of future time, is / will be whatever I choose. 

The language of Exodus 3i*-i5 must accordingly be rendered 
as follows : 1*° And Crod said unto Moses, I will be whatever 
J choose. 1** And he said, Thus shalt thou say to the children 
of Israel: 'llhyeh [^^^K="I will be"] hath sent me unto 
you. ^^ And Grod said further unto Moses, Thus shalt thou say 

''^ Cf. Deut. 29*5 ; Steuernagel, HK, p. 34, seems not to perceive the dif- 
ference between 01' D'CaiK and OVn D'ca-IK DK. 

2' Tliat is, by me, if thou wilt; not "durch irgend einen anderen," as 
Kautzsch, § 155 n., and Baentscli. Tliat the correct interpretation is hard to 
reconcile with the anger of Jahweh in verse " indicates only that the two 
verses are not from the same hand. 


unto the children of Israel : mri\ the Grod of your fathers, the 
God of Abraham, the God of Isaac, and the God of Jacob, 
hath sent me unto you. — This is my name forever, and this 
my designation for generation after generation. 

Now it will be immediately objected that this rendering 
of verse i* yields no satisfactory sense, and is clearly out of 
harmony with the context. I will he whatever I eJioose not 
merely fails of constituting an adequate answer to the ques- 
tion propounded by Moses ; it constitutes an entire change 
of subject, and actually assumes a question of a wholly dif- 
ferent tenor. It is not the name, but the future of the Deity 
that is here the subject of speech. And though the latter 
subject is here mentioned for the first time, it is done with 
a tone of resentment and rebuke that implies inquisitive 
prying into that same subject on the part of somebody or 
other. n%t« "VSH ,Tn« has in mind the query, ^t^^n n», 
What will you he? Yet this query nowhere occurs, and the 
only occasion for asking it which appears in the whole course 
of the narrative is that furnished by the enigmatical, because 
unprecedented and catalectic, name ^^'^K, bestowed upon the 
Deity in i**. We have here, then, the remarkable phenomenon 
of a supposed questioner being rebuked for impertinence 
and inquisitiveness before even the occasion for his question 
has been encountered. There is only one solution to this 
problem: ^^'^i< "ItTX n\"IS of i*« cannot have been written 
before n\"IX of i*&; and since it does not follow that word 
in space, it must nevertheless have followed it in time ; that 
is, i*« IS AN INTERPOLATION. It is a Midrashic gloss on i**. 

That 1*" is interpolated can be shown quite independently 
of the above considerations. Verse 13 asks a question. Not 
14<» but 1** answers the question, and that not only in sub- 
stance, but — what is more to the point — in form, i** by 
its very language forces i*" out of the text. The question 
of 13 is Dnb{< niai< nia what shall I say unto them? Not 14« 
but 146 it is that begins with the appropriate correlative 

Nor should it be imagined that the introductory 'lSit<''1 of 
1*6 is not weighty enough to introduce the reply of God to 


Moses, and that something like the phrase " And God said 
unto Moses" of i*<» is required. On the contrary, an im- 
mediate response by the other party to a dialogue is, if any- 
thing, more properly introduced by the simple ^tiH'') than 
by a more extended formula ; cf. in particular, E in vss.*-^- 12 
of this chapter and in Gen. 22i- 2- 7. 11. 12 ; and of passages 
which Steuernagel assigns to E^, Ex. 32i8 Num. 22306 2312. 

That vss.i* and 15 are swollen has already been perceived. 
Others have objected to the recurrence of the introductory 
"1&S"'1 thrice in connection with an uninterrupted utterance 
of Jahweh. The common remedy has been to exclude 
from the original text. It has been argued that the word 
TiV, which accompanies the introductory ^tm''^ in, be- 
trays the interpolation. To my mind, the word DV, if it 
does anything, authenticates vs. is. That verse contains an 
amplifying continuation of the statement made in l*&, and 
by means of the particle ^V it announces that it does so. 
niS "1&X''1 does not mean And he said again, but And he said 
further, that is, He went on to say. Together with the fol- 
lowing TWf2 7S D^'^75<, it affords the necessary indication 
that the same person continues to speak who has just 
spoken ; a mere '1J2S''1 would have alternating reference. 
Nor is there any way by which the introduction of vs.i3 
into our text can be explained, except to assume that it was 
part of the E document. Most scholars who exclude that 
verse assign it to Rje ; one assigns it to Rd ; and one to Rp. 
But a mere glance at vs.i^ shows that ^^ cannot be the con- 
tribution of a redactor. Practically all of is is contained in is, 
which latter belongs to J. Now, no redactor who had before 
him vs. 16 would of his own motion duplicate it with the verse 
immediately prefixed. The only way in which the presence 
of these two successive verses in this passage can be explained 
is by assuming that one of them came from one primary source, 
and the other from another primary source. Verse is cannot 
be from either Rje, Rd, or Rp, or from a later diaskeuast, but 
must have come in from the E document.^ 

^ Wlldeboer {Die Litteratur des alien Testaments, p. 138) suggested that 
vs. 15 can hardly belong to the E source, because it implies that the name 


But if VS.15 was in the E document, we have confirmation 
of the interpolation of i*« in the particle IIS. Perhaps we 
ought not to go so far as to say that IIS could not possibly 
accompany the third of three occurrences of "ItoK^I and not 
the second, but such a style would certainly be most remark- 
able. That is, if i*« were authentic, we should certainly 
have TiV with the *lSiS''1 of i**. Furthermore, as suggested 
above, simple IQ^'^ has alternating reference; so that i*^ 
must have been immediately preceded by an utterance of 
Moses, in the original source. We may confidently main- 
tain that the introductory phraseology of i** and is jg 
sufficient of itself to require the exclusion of i*« from the 
text of E. 

Before proceeding further, we may summarily dispose of 
the one remaining hypothesis regarding the conflation of 
vss.i*-i5^ namely, that the entire vs.i*, ^ as well as ", is inter- 
polated or redactional.^ The fatal objection to this hypoth- 
esis is that it creates vastly more difficulty than it removes. 
Upon that assumption, no part of the verse has any mean- 
ing whatever, and no possible ground for its insertion can 
be imagined. 

Accordingly — disregarding for the moment the question 

Jahweh was known to the Fathers, whereas P, who derived his notion from 
E, expressly makes the name to be revealed for the first time to Moses, the 
Fathers having known the Deity only by the name of HI Shaddai. But both 
the name El Shaddai and the notion of the name Jahweh being here revealed 
for the first time (or indeed " revealed " at all) are peculiar to P. E furnished 
only the suggestion for the theory and practice of P. They occupy wholly 
different platforms. To P, it is the Deity revealing the most important of 
his attributes, his most holy name, to the favored of mankind ; to E, it is 
the ancestral and national divinity — B'n7K, "deity" par excellence — ap- 
pearing to Moses on a particular occasion and, upon request, making known 
his identity by means of his proper name. Even P could not get along with 
merely the appellative Elohim for the whole period before the final revelation. 
And it is doubtful if even P would have understood what Carpenter means by 
" the revelation of Elohim being followed by that of Yahweh " (Composition 
of the Hexateuch, p. 203). What then, to the mind of E, \eas the name by 
which the god of Moses' fathers, spoken of in vs.', was alluded to in case of 
necessity, — by way of contrast to "the gods" (D'n'^K) of Josh. "iA^, for 
example? Or did he have no name? 
^ Carpenter and Harford assign it to Eje. 


of the authenticity of the name iTHK in 14& — the text of E 
ran as follows : 

" ^ And Moses said unto God, Behold I come unto the children 
of Israel and say unto them, The God of your fathers hath sent 
me unto you. Then will they say unto me. What is his name ? 
What shall I say unto them ? "' And he said, Thus shalt thou 
say to the children of Israel : HNIS hath sent me unto you. 
"And God said further unto Moses, Thus shalt thou say unto 
the children of Israel : niiT', the God of your fathers, the God of 
Abraham, the God of Isaac, and the God of Jacob, hath sent me 
unto you." Etc. 

It is with this determination of the text that we must con- 
sider the question of the authenticity of the reading iTHK 
in l**". We have seen that this reading is older than i*« and 
occasioned the latter's interpolation. How old is the read- 
ing? Does it go back to E? With i*" out of the way, there 
can be no manner of doubt that ^^^K of i** represents the 
corruption of an original JTliT.S'' The name of Israel's God 
was not iTilX, but mn\ The name which from this point 
on appears from time to time in the E document is not ^^"^K, 
but mn'. The name which still stands in vs. 15, the author's 
own amplification of the statement of i**, is n'in\ Further- 
more, it is universally admitted that the E document, with 
its introduction of the proper name of the God of Israel at 
this point in the history, furnished the model for the proce- 
dure, and gave occasion for the theory, of the later priestly 
writer, who introduced the name mJT' in a special revela- 
tion to Moses in the passage preserved at Ex. 62, and who 
from that point on employed the name with greater uni- 
formity than does the E document. The name which was 
disclosed to Moses at this point in the E narrative, as the 
writer of P understood it, was not iTnt<, but mn\ Nor 
does P know anything of the disclosure of some explanation 
or origin of the name, besides the name itself. Of these 

a>So already Holzinger (Exodus, in KHC, p. 14), but on radically dif- 
ferent and entirely erroneous grounds. Cf. also Wellhausen, Composition 
des Hexateuchs % p. 72, note 2. 


facts there can be no doubt whatever. So that the E docu- 
ment (or, if we prefer, the E passage in the JE document), 
as P had it, read at this point the name HliT and no other. 
tVn^ has, therefore, been substituted for an original mH' in 
the text of i**, and the substitution took place after P, — or 
at any rate, in a manuscript which did not influence the text 
P used. Certainly the reading ^^^X, if it existed anywhere, 
was not at all current in P's day; for it is not one which 
could long remain unnoticed, — witness the interpolation of 
14a. When, moreover, we consider that our E (or JE) comes 
to us from the same priestly circles in which P arose, the 
conclusion that iTHS was not introduced into the text of E 
until after P was written, becomes all but necessary. On 
the whole, I think it will not be disputed that we are justi- 
fied in maintaining that, on the evidence of P, the reading 
(THX for miT' in the passage which now constitutes Ex. 3i*& 
was not in existence about the beginning of the fifth 
century B.C. 

On the other hand, our study of the history of the text 
has shown that, on the evidence of all the versions and the 
Samaritan tradition, the reading ^^^'^H was prevalent some 
considerable time before the middle of the third century B.C. 
And even disregarding the Samaritan tradition entirely, on 
account of the uncertainty which attaches to the chrono- 
logical value of its testimony, the mere fact that Alexandrian 
Greek and Masoretic Hebrew of Ex. S^* derive from a com- 
mon ancestor, requires us to push the date of the present 
reading in i** appreciably to the rear of the middle of the 
third century B.C. So that the terminus ante quern may in 
any event be safely set down as about the end of the fourth 
century. Some time between 500 and 300 B.C., then, iTHX 
was substituted for niH^ in the text of Ex. 3i**. A more 
precise determination of the date of this substitution cannot 
be reached without concerning ourselves with the question 
of the reason for it. 

That there was a reason for it, and that the alteration of 
the text was the result of design and not the result of acci- 
dent, cannot be doubted. For assuming — what seems quite 


impossible — that a scribe did actually slip into reading or 
writing tTHK for the name of the national God in copying 
a manuscript, the slip could not fail of unceremonious cor- 
rection upon the very first reading of his copy, since the 
strangeness of the text must have compelled attention, and 
any scruples about correcting so palpable a blunder could 
have been summarily removed by a glance at another manu- 
script. In fact, so great must have been the temptation to 
correct the reading tTilb^, that we may question whether, in 
spite of the deliberate intent which marked its introduction, 
it would have survived, had it not been hedged about by the 
interpolated iTHJ^ *W^ n\"I>< which followed hard upon its 
heels, not improbably with that very danger in view.^^ 

The motive for the alteration of the name mri^ to ^^'^^< 
in Exodus 3i4& is not far to seek. It can have been only 
one : to prevent the utterance of the ineffable name. 

The increasing awe with which, under the influence of 
the Priest Code and the accomplished centralization of the 
cultus, all things came to be regarded that pertained to the 
person of the holy God, resulted, toward the end of the period 
whose bounds we have determined (500-300 B.C.), in the 
entire abandonment of the public use of his most holy Name 

»i A somewhat similar, though inverted, case of one corruption of the text 
following upon another from much the same motive as here, is that which 
continues to perplex critics in 1 Sam. 3w. There the original reading was 
that which I append (in the larger type) : 

'JK t9BB> ''3 ^b "mm 
nbis no wa nx 

03 run r"?i ra 

The words printed in small type, on account of the sin which need not he 
mentioned (literally, which he knows), were put upon the margin to be sub- 
stituted for the blasphemous expression T33 D'nbx BihbpKi 'a in reading the 
text aloud. But being written into the column in a later manuscript, their 
object was defeated, and the old difficulty had to be met again, this time by 
truncating O'D^S to Orh, the present Masoretio text, which is opposed alike 
by the Alexandrian Greek and by the Rabbinical tradition. The Greek iv 
iSiKlais vtwv airoO preserves a different form of the marginal euphemism = 
V:Z 11173. 


outside of the one central sanctuary at Jerusalem. It was 
no longer the name of a national divinity, acquaintance with 
which was as old as the relation to the person bearing it. It 
was the very real name of the God of all the world, like his 
holy Law by special act of grace revealed to Moses for the 
salvation of his people. Before yet they had built a fence of 
tradition about the written law to prevent any approxima- 
tion to the violation of its precepts, the Jews ceased to utter 
the divine name, lest they approach ever so remotely the 
bounds of sacrilege.^^ The mere utterance of the name, 
apart from perjury, cursing, or blasphemy, was indeed at no 
time a criminal offense ; ^ but it soon came to be considered 
a ritual sin, punishable by God though not by man.^ And 
though this sin, like other sins, was no doubt sometimes com- 
mitted, it was not a thing to be authorized or tolerated in 
connection with the Sabbath functions of the synagogue. 

From the statements of Philo and Josephus it is evident 
that the practical disuse of the name niiT lay so far behind 
them that they had not the slightest idea that there ever was 
a time when the name was less sparingly employed than in 
their own days.^ In some of the later writings of the Old 

82 Cf. Dalman, Der Gottesname Adonai und seine Geschichte, pp. 71 f . 
Jacob, Im Namen Gottes, Berlin, 1903, p. 166, goes altogether too far when 
he carries this attitude back to the days of Ezekiel. 

^ According to 6. Aboda Zara 17 6, under Roman rule in the second cen- 
tury A.D. Rabbi Hanina ben Teradion was burned at the stake, his wife exe- 
cuted, and his daughter condemned to a life of shame, for no other cause 
than that the Rabbi had pronounced the ineflable name in public hearing. 
But the ground alleged for the outrage is rejected even by those who accept 
as historical all the other details of the Talmudic story ; cf . Bacher, Die 
Agada der Tannaiten, vol. i. p. 400. 

8* Cf. Mishna, Sanhedrin, x. 1 ; Tosefta, Sanh. xii. 9 (ed. Zuckermandel, 
p. 433). In Mishna, Berachoth, ix. 5, there is no talk of the utterance of the 
word niiT. The question there is merely whether one shall or shall not use 
the divine name in ordinary salutation (Dl'riP nx bw). The salutations 
BSiaC mrp, mn'' 'p-Q\ yiV m,T, of Ruth 2* and Judges 612 (which were, 
as a matter of course, pronounced by the speakers 0381) 'JIS, 'JIK ^^'^3'', 
etc.) are cited by those who prefer the good old religious forms — " Despise 
not thy mother when she is old" is quoted from Proverbs 23^ — to the 
godless Da'bc ahw of their own unregenerate days. 

85 Dalman, I.e. pp. 38, 42. 


Testament — Chronicles, Daniel, Ecclesiastes, and an entire 
section of the Psalms — there is a marked avoidance of the 
name nin\ This is, to be sure, of itself not very significant, 
except as it indicates the increasing preference for the term 
D'n?^. It furnishes no evidence that when encountered or 
employed by these Old Testament writers, the name miT' 
was not faithfully pronounced. Then, too, we know that in 
still later times TV\iT, though never pronounced, was unhesi- 
tatingly written. But a comparison of Chronicles with the 
book of Samuel makes it almost certain that the Chronicler, 
at least, pronounced ^ilH even when he did write niH'' ; for 
in reproducing his source he deliberately avoids the combi- 
nation Tl)TV ^nx of 2 Sam. 7 (which to him would have been 
■•ns ""nx 36), writing for it now D^"^'?« m.T, now ffnSs, now 
nin\ and never once does he write ''3^S.^ The most valu- 
able and conclusive evidence, however, regarding the avoid- 
ance of the utterance of the name HliT, is that furnished by 
the Alexandrian Greek version. By the time of the first 
translation of the Pentateuch into Greek, in the middle of 
the third century B.C., the custom of substituting another 
word for the proper name of the Deity in the reading of 
the Law was already firmly established, and the substitute 
employed was stereotyped and uniform. For there ought to 
be no doubt whatever that the word Kv/ato? was habitually 
employed by the readers of the Greek version, from the very 
first, wherever the original had mn\^ and that this usage 
was derived from the settled custom among the Palestinian 
Jews of pronouncing the name ^ilH, Dalman ^^ affirms that 
the Kvpioi of our Greek manuscripts cannot be received as 
evidence for the early oral substitution of ''nx for UliT, since 
from the statements of Origen *" and Jerome *^ it appears that 

M Cf. Kittel, PRE% vol. viii. p. 532. 

«' See Jacob, I.e. pp. 165 f. 

" The few occurrences of nirP ''31K in the Pentateuch were variously 
treated in the Greek and may be disregarded. 

89 I.e. pp. 37 f. 

♦> On Psalm 2, Migne's Patrologia Oraeea, xii. 1104. 

*i Proloffus galeatus, and in the letter to Marcella referred to above, 
p. 121, note 17. 

AKNOLD: the divine KAMB in exodus III. 14 137 

old Greek manuscripts exhibited the Hebrew characters mn\ 
and not the Greek Kv/ato?, wherever the proper name occurred 
in the original. Assuming, however, that the statements of 
Origen and Jerome justify the inference that the original 
Greek manuscript did, like Aquila's version, employ that 
device (against which inference not a little might be said*^), 
the fact remains that Hebrew mn^ in a Greek manuscript 
is as good evidence for the custom of substituting something 
or other for the authentic word as would be the reading 
Kv/)to? itself. Otherwise the Greek would have exhibited 
not the Hebrew characters, but the transliteration of the 
name in Greek characters, as in the case of other proper 
names ; ^ and neither in extant Greek manuscripts nor in 
tradition is there the faintest trace of an original lavrj 
(or variant of it) ** in the Greek version. But if the Hel- 
lenistic Jews from the very beginning substituted some 
word for nUT' in reading the Greek text, (1) we may be 
certain that it was the custom of the contemporary Pales- 
tinian Jews to do likewise in reading the Hebrew, and (2) we 
have no reason in the world for believing that the word 
which the Hellenistic Jews originally substituted was differ- 
ent from that which we find them automatically employing 
in the days of Philo, namely Kvpto?. But if Kvpio^ was 
employed for HIIT' among the Hellenistic Jews from the ear- 
liest times, ''HS must have been antecedently so employed 
by the Palestinian Jews ; for Kw/oto? *^ obviously reproduces 
the suffixless sense which ^31i< acquired — as Dalman him- 
self has so ably shown — in the very act of being substituted 
for the name mn\ 

At least as early as 300 B.C., then, and most probably ear- 
lier, the utterance of the name mn^ was abandoned in the 

« Cf. Klttel, I.e. pp. 530, 532. 

*' It would hardly have been bold enough to attempt a translation. 

« a ( = H- = e) is in Hebrew a phase of a, whereas in Greek it is a phase 
of e ; hence Mwuo-ijs and Mavaaari. (Participles of Tl'b verbs had doubtless 
ceased to be pronounced H- by that time.) In later times, when ri had 
become t and j3 had become v, we have lajSe (Epiphanius and Theodoret). 
Origen's larj probably omits the consonantal 1 ; he certainly does not intend rP. 

*5 Notice that it is not 6 Kifita, which is of secondary development. 


Jewish synagogue. I say most probably earlier; for the 
phenomenon ought not to be isolated, and it will best be con- 
nected with the institutional innovations which followed 
immediately upon the publication and circulation of the 

It was of course only in reading or quoting verbatim a 
writing which contained the name that a vocal device for the 
avoidance of its utterance would ordinarily be required. 
When one speaks of the God of the whole world as " God," 
one is not employing a surrogate for miT'. A speaker would 
have no difiiculty in leaving the name alone ; the person of 
the Deity could be referred to in many other ways. And the 
Palestinian Jews, in their own utterances, did leave alone not 
only mn\ but '31S too, when once the latter had become 
the synagogue surrogate for rnn\ On the other hand, mere 
reference to the name flliT, though it would often occur, did 
not involve the quotation of the name. This distinction 
between reference to the person and reference to the vocable 
is important. In Lev. 2411-16 Otm m ajTI and D^ 13|533 
are correlatives of iTiTTI^ Dtl? 2p3, not — it is to be noticed — 
of mn^ api.*? Doubtless along this line lay the demarcation 
between the original use of ''HS on the one hand, and of 
Dtrn on the other, as surrogates for mHV The two terms 
may have been in part suggested, as Dalman observes,*** by 
the existing phrases ffl.T 'JIX and miT' BW, — though it would 

« The synagogue may have had its prototype among the learned in the 
Exile, but the institution itself is not to be dissociated from the Five Books 
of Moses. Cf. Sohttrer, Geschichte desjMischen Volkes^, vol. 11. pp. 428 ff. 

" Geiger's view ( Urschrift und Uebersetzungen der Bibel, p. 274) that 
as is the result of corruption of the text in all three cases, is not at all plau- 
sible. He makes the point that ipJ is never employed with OS of the per- 
son cursed. But this begs the question at issue. The story of Lev. 241" «•, 
if it means anything, means that the contentious half-breed blasphemed by 
making use of the name mn'' in cursing his antagonist. It is quite true that 
" own ganz absolut fiir Gott zu setzen, ist durchaus unbiblisch." The con- 
clusion is that it stands for God's name. The final OS lapjS of v.w, with- 
out the article, for so soon as he employs the name (already mentioned, and 
scil. in cursing) is not un-Hebraio. apj in this section may very well be 
literally to pronounce, and yet the burden of the section be a law against 
blasphemy only ; cf. Dalman. I.e. p. 44. 

" I.e. p. 74. 


seem that a suggestion for the use of Dtt^H in referring to "the 
name " was superfluous. But that the terms were logically 
interchangeable, or that the Jews in early times were guilty 
of the absurd Samaritan custom of reading DU^H where the 
person of the Deity is spoken of — even invoked — in the 
Old Testament, as Geiger maintained,*^ is not to be believed. 
Geiger says of the Talmudic evidence which opposes his 
view, "Erst spat (/. SanJi. 10, 1 [28 6], b. Pessachim 50 a, 
Kidd. 71 a) tritt die bestimmte Angabe auf, dass zwar TWC' 
geschrieben, aber "'HK gelesen werde." But more telling 
than any "bestimmte Angabe" is the quiet implication of 
age-long usage involved in the following passage of the 
Jerusalem Talmud {Megilla 71c?): fpHttJ \T)HVS nitt^ 'hvt, 

•nSnsi p|b"X3 K","i3i nr'-'s nrms sn-ixn D^n nx anttn 

"JIT • D^'^7K • bx. These are the names which may not he 
erased : When one writes out the Name with four letters (that 
is, does not write Tt, for example), and that whether with 
Todh He (that is, writing HliT) or with Aleph Daleth (writ- 
ing ''M«); bX; ffnbS; etc.^ The Mishnic passages cited 
by Dalman,^! Yoma, iii. 8, iv. 2, vi. 2, in which the High Priest 
is represented as addressing the Deity at the opening of his 
confessional prayer on the Day of Atonement with D^H X3K, 
can scarcely mean anything but that the name miT was 
actually uttered by the High Priest, as Geiger was the first 
to admit.^2 Later, in the scholasticism of the Rabbis, the use 
of Dt?n was doubtless somewhat extended. Yet it is hard 
to believe that even in the third or fourth century A.D. the 

*' I.e. pp. 262 ff. For the rest, it is hard to avoid the conclusion that the 
Samaritan custom of substituting DOT for niiT in the reading of the Law — 
which custom seems not to have controlled the practice in every-day life 
— is of relatively late date, at the most no older than the Dosithean move- 
ment and the temporary rapprochement between Jews and Samaritans in the 
early part of the second century a.d. (cf. Hamburger, Jteal-encyelopddie fur 
Sibel und Talmud,, part ii. p. 1069). In that case we can understand the 
failure of the Samaritans to adopt a surrogate which to their Jewish mentors 
had long since become a proper name and grown almost as sacred as TTirC 

w This passage is mistranslated by Schwab, Le Talmud de Jerusalem, vol. 
vi. p. 218. 

'1 Worte Jesu, p. 150. 

M l.c. p. 263. 


High Priest, had there heen one, would have addressed the 
Deity as Dtm.^ Possibly enough the rehearser of the Mishna 
may have substituted DC^n in relating the tradition. But 
we cannot be sure of it. For this term was certainly some- 
times set down in manuscripts, both in Scripture citations 
and otherwise, when some more construable surrogate was 
actually spoken, merely to steer clear of the rule against the 
erasure of divine names, or to avoid the confusion of a less 
obvious written surrogate. The purport of this will appear 
more clearly as we proceed. 

To return now to the expression l^^'^S in Ex. S^^^. We 
said that this word represents the wilful alteration of 
original HliT, and that the purpose of the alteration can 
only have been to prevent the utterance of the ineffable 
name. If our conclusion is justified, the alteration will fall 
into the period when the custom of avoiding the utterance 
of mri"' by the substitution of ''3^K was being inaugurated 
in the Jewish synagogue, sometime between the publication 
of the Pentateuch and the end of the fourth century B.C. 

It needs little argument to show that an exceptional 
procedure in the case of the divine name in Ex. S^**, both as 
to the character of the surrogate employed and as to the 
projection of the surrogate into the text itself, was literally 

The second of these points may be first disposed of. A 
simple rule to read ""Jli^ uniformly wherever mri"' was en- 
countered in the text, required no manuscript notation in 
order to be remembered and obeyed. But an isolated diver- 
gence from the uniform practice, in connection with a partic- 
ular passage, could hardly be trusted to the memory at a 
time when as yet the perfect mastery of the entire text of 
the Law was not a common feat. If it can be shown that 
an exceptional surrogate was required in this passage, I 
think it will not be disputed that its projection into the 
text was equally necessary. 

^ It is needless to say that the Jews did not hypostatize the name of God. 
Giesebrecht, Die alttestamentliehe Schatzung des Gottesnamens, p. 44, has 
overworked the passages Ex. 23^' and Isa. 30-". 


Now it is to be noticed that Ex. 31*6 is the one passage in 
the Pentateuch where the substitution of ^3*1it for miT was 
not merely glaringly inappropriate, but actually impossible. 
It is the one passage where the question, " What is the 
proper name of Israel's God ? " having been formally asked 
is formally answered. Ex. 62 *• is far from being in the 
same case. Even to us who read the latter passage as 
part of the separate P document, its burden is manifestly 
not the identity of the name but the age of its revelation 
to mankind. We must not forget, however, that we are 
dealing with readers of the Pentateuch, not of the P docu- 
ment. To one who had perused or listened to the account 
of Ex. 3^3 ff.^ the phrase niiT '3X of Ex. 6^ could no more 
savor of a revelation of that name to Moses than would 
the several recurrences of that expression in the ensuing 
section or in the Holiness Code. To the mind of such a 
person, the statement of Ex. 62 *• could do no more than 
inform Moses that the name, which he (as well as the 
reader) already knew, was not known of old to Abraham, 
Isaac, and Jacob. These, he is told, knew the Deity only 
by the name ^1^7 7K. The question of the identity of the 
antithetical term was not raised, though that term was for- 
mally alluded to. The pronunciation ""Hi^ '^HVf for ."IliT ''IS^ 
of 6^ introduced, it is true, an unmistakable solecism, but 
only an incidental one, such as occurred often enough else- 
where, — in Lev. 24^® for example. On the other hand, to 
permit the use of "'Hi* in Ex. 3^** in direct response to the 
question ^t2iV Hfi of vs.^^ would have been to authorize an 
out and out, as well as a perilous, misstatement. When 
once the question had been squarely answered, the reader 
might be permitted to return to the customary 'HS in vs.^^ 
but the immediate response of vs.^* could not employ it. 
The institutors of the custom of substituting ''3*TK for the 
written niiT in the reading of the Law desired to prevent 
the utterance of the sacred name; but they were very far 
from desiring to mislead the congregation into the belief 
that the word "'3'IX, which was constantly heard in the 
lections, was itself the holy and ineffable name. Yet to 


prescribe the use of "'3*TK, in Ex. 3i4& would be to encourage, 
if not indeed to confirm, such a belief; to put the case 
mildly, its effect would be to confuse the public mind. Nor, 
for the same reason, could any more general designation of 
the Deity do duty here. And to cause Moses to answer the 
question "Iti^ Htt with ^31172? DS^n would be quite too ab- 
surd for the people of the times, though perhaps not for 
some of us. Here, for once, the utterance of the name nin^ 
could not be avoided. 

Yet if the sacredness of the name was to be violated once 
periodically in every synagogue in the land, why not oftener, 
and elsewhere ? In this situation there was but one alterna- 
tive — an alternative that we shall see was habitually re- 
sorted to during the following centuries in similar case 
outside of the synagogue. It was to so mutilate the word 
in pronouncing it that the requirement of the rule against 
its utterance would be formally met, while at the same time 
what was actually uttered could not possibly be mistaken 
for anything but the representation of the tetragrammaton. 
This was habitually done in later times by exchanging one 
or two of the consonants of the sacred name, while main- 
taining intact its syllabic and vocalic cast. The term for 
this hybrid product in the Hebrew of the Rabbis was ''133. 
The word n^1« in Ex. 31*6 is a ""liS of m."I\ 

Unfortunately the statement of fact contained in the last 
paragraph is in need of demonstration. The Rabbinical 
term ""US is not ordinarily assigned this meaning when used 
in connection with a designation of the Deity, and it is far 
from being commonly admitted that the Jews in early times 
were accustomed to avoid the actual pronunciation of the 
divine name, in case of need, by means of this mechanical 

The definition of ''133 given by Levy is "Beinanje, Neben- 
benennung";^ by Jastrow, "by-name, surname, attribute, 
substituted word" ; ^ by Bacher, " eine umschreibende Benen- 
nung, im Gegensatze zum eigentlichen Namen Gottes, dem 

" Neuhebraisches und chaldaisches Worterbueh, vol. ii. pp. 860 f. 
^ Dictionary of the Targumim, etc., p. 633. 


Tetragrammaton " ; ^ by Dalman, " Beiname, umschreibende 
Benennung." ^'^ None of these is a strictly accurate defini- 
tion of the term '''135. 

Not a little confusion has been wrought in this matter 

by the repeated citation of the Arabic &xii, which has 
been customarily rendered " Beiname," " ehrendes Epithe- 
ton." But this rendering of the Arabic term is by no means 
comprehensive enough, and in any case not quite legitimate. 
Surname is only a secondary and applied meaning of 
JuaJ'; and it is not even that, if we emphasize the 
prefix. The Arabic word properly means surrogate, and is 
primarily employed to designate a term which affects to 
veil an indecent or otherwise objectionable allusion. It 
then comes to be employed for any circumlocutory designa- 
tion ; and so passes at last to the stereotyped surrogates of 
individual names, such as Abu Zaid, " the Father of Zaid." 
But only in so far as such a designation is obviously used 
to the exclusion of the man's real name is it a Xjuli'. 
Whether a word (which need not necessarily be a sub- 
stantive) or a phrase is or is not a xjuf depends upon 
its function and not upon its form ; and the same element 
may be a 'i.*^ in one connection and not in another. To 
illustrate with a familiar case, "Peter" in the expression 
"Simon Peter" is a surname, but not a luJl^, which it 
might be if used in avoidance of the name Simon. ^ For 
the rest, there can be little doubt that the root of the word 
is cognate to Jji', to disguise or conceal. 

The technical Neo-Hebrew word '133 (frequently written 
■"IJ'S) is a nomen actionis of Piel HIIS, like "1Q''! utterance, 
from "IS'I to utter; tt^l'in innovation, from tt^^H to innovate; 
''13'1 inclusion, from nS"! to include; 1D1'!?& exclusion, from 
12?& to exclude ; tJ>1^B the act of being exact (in expression), 

^ Terminologie der Tannaiten, p. 85. 

*' Aramaiseh-neuliebraisches Worterbuch, p. 192. 

58 See Lisctn ul 'Arab, vol. xx. pp. 98 f. The lexicon of Golius (Leyden, 
1653) defined the Arabic verb, "Appellavit sive signavit nomine per se 
significante rem aliam ;" and the noun, "Metonymia, quum alio qukm suo 
nomine significatur res." 


gpeeification, from t&'IB to be exact, to specify. There is no 
question about the meaning of the Piel HW. It is defined 
by Bacher, "auf verhiillende Weise ausdriicken, umschrei- 
ben."^^ The nomen aetionis of this is accordingly, the act of 
expressing -in-disguise ; which is the primary meaning of ''133. 
This primary meaning is unmistakably retained in the Mish- 
nic Text of Tamid vii. 2 (=Sota vii. 6), D^iaiX m tTlpaS 
V13P3 nrnaai anas a^n n«, in the sanctuary they (the 
priests) were accustomed to pronounce the Name (in benedic- 
tion, Num. 62* ff-) as it is written; in the town (that is, when 
they were abroad in the city of Jerusalem), hy disguising it.^ 
Now just as our English word " expression " and the Neo- 
Hebrew word "113^, from meaning originally the act of 
expressing, come to mean the thing expressed, so ""133, from 
meaning the act of expressing-in-disguise, comes to mean that 
which expresses-in-disguise ; or — to employ another term — 
from meaning substitution in the sense of the act of substitut- 

^l.c. p. 83; of. Levy, s.v. It is the word which recurs at the end of 
each article in the enumeration of the eighteen Tiqqune Soferim or euphe- 
mistical alterations of the Old Testament text recorded in the Jewish tradi- 
tion : ainsn nr3lP K"?S<, but the text disguises ; see Geiger, Urschrift, pp. 
308 ff. ; Ginsburg, Introduction to the Massoretico-Critical Edition of the 
Hebrew Bible, p. 348. 

60 " Hier bedeutet VW3 seine — des Tetragrammatons — umsohreibende 
Lesung 'J'^*?," says Bacher, after he has defined '133 as "eine umsohrei- 
bende Benennung [of God]." But, with all due respect to the commenta- 
tors of the Middle Ages, how do we know that the Mishna has reference to 
'JnS? The ordinary assumption is that by Hi'ias is meant, "when the 
priests are officiating in the synagogues of the country," in which case it is 
taken for granted they would make use of the synagogue surrogate of the 
name ITST. But what evidence have we that the priests in their ofBcial 
capacity had anything to do with the functions of the synagogue, or that 
they were employed to "pronounce the benediction" there ? The common 
residence of the priests was Jerusalem, and when the text speaks of the 
priests doing one thing in the CHptt and another in the m'na, we have no 
reason for rendering otherwise than in the temple and in the town respec- 
tively. Kor can we render 1'133 of this passage its (the name's) surrogate ; 
for there never was a time when the name Hl.T had only one '133, — decid- 
edly not in the Mishnic period ; cf. Shebu'oth iv. IS, Sanhedrin vii. 5. For 
the rest, it is important to notice that in the one passage where '133 occurs 
with the possessive suffix of the third person singular, the supporters of the 
prevalent erroneous definition of the word are forced to admit that the 
suffix has reference to the name and not to the person of God. 


ing, it comes to mean substitution in the sense of the thing 
substituted, — surrogate. These two senses, however, exhaust 
the lexical values of the word '133.^i Any narrower defini- 
tion of the word involves "false distribution." 

Moreover, exactly as in the case of the Arabic term, 
whether a word or phrase is or is not a '''UD depends 
upon its function in actual use and in the conception 
of the speaker, and not at all upon its identity. One 
cannot affirm of any particular word in the lexicon that 
it is per se a ""l^S. According to the commentary on 
Leviticus called Sifra (ed. Weiss, fol. 104 c), the text of 
Lev. 24" does not substitute ']iy^) for bbp'') as does the 
text of 1 Kings 2113, -p^-^ n^-,i^j< ^133 113, for the reason 
that " they do not put to death on the basis of a ''133 " (M. 
Sanhedrin vii. 5). That is, the verb 1^3 when employed 
as a euphemistic substitute for 7vp is a ""liS. So the 
expression DIpOH, the place, may be a ''133 of the word 
D^'^vK ; and D^^7K in turn may be conceived of as a ''133 of 
the name nilT.^^ After what has been said, it is needless to 

61 1 refer, of course, to early times ; by the grammarians of the Middle 
Ages the word is used for " pronoun " ; cf. Buxtorf -Fischer, s.v. In Arabic 
grammar jLsLi^' is not exactly " pronoun " (as Caspari-MttUer, § 34) ; 
the class consists of "certains mots d'une signification vague qu'on sub- 
stitue k des expressions plus determin^es " ; see De Sacy, Orammaire 
arabe,^ vol. i. pp. 430, 434, vol. ii. p. 66 ; and cf. Wright-DeGoeje, Arabic 
Grammar, vol. ii. p. 125 C. 

63 In Sifra, on Lev. \9^ (88 c), 24" (104 d), and in both Talmuds (6. San- 
hedrin 56 a, j. Sanhedrin 25 o), all designations of the Deity other than 
ni.T are classed together as B"133n, but that is merely in maintenance of 
the fiction that all other designations are but surrogates of the real name. 
So according to 6. Sanhedrin 60 a, when the heathen Bglon, king of Moab, 
hears from Ehud yb* ^b DTlbs "Ol he hears only a ''133. In the Mishna 
d'lJSn bi is used very differently (She'buoth iv. 13) : (When witnesses are 
summoned with any of the expressions) "Z adjure you," "/ enjoin upon 
you," " I bind you," they are obligated : (if the adjuration be) by " heaven "" 
or by '■'■earth," they are free; (but if) by "Aleph Daleth" [that is, referring 
to 'S^K, but avoiding its pronunciation in this manner], by ^'■Todh He" 
[with similar reference to mrp], by nr, by niS2S, by Dimi JUn [Ex. 34^], 
by D'S^ "II^, by IDH 31, or by any of the surrogates (p'lJDn b'3S,^), they are 
obligated. Whatever may have been the character of the "surrogates" 
alluded to, there is nothing in the Mishnic text to show that they were like 
ion 31 and the two preceding expressions. The interpretations " oder bei 


point out tliat the noun ''135 stands in no particular relation 
to the designation of the Deity, any more than does the 
verb n33. 

We said that one cannot affirm of any word in the lexicon 
that it is or is not per se a ''133. But there are certain voca- 
bles which have no place in the lexicon, of which this may 
be affirmed. These are meaningless aggregations of letters, 
which never perform any other office than that of serving as 
surrogates for other words. We may illustrate with Scotch 
and American " by gosh " for " by God " ; this " gosh " is a 
''133 in its own right. It is these d^''13D of the "dummy" 
order, of which the Rabbinical literature has an abundant 
supply, that concern us particularly here. 

We may limit our quotations to the two important pas- 
sages of the Mishna, Nedarim i. 1, 2, and Sanhedrin vii. 5. 
The first will exhibit the manner in which these surrogates 
are formed by the mutilation of the word they replace, that 
is, by the change of certain of its consonants while preserv- 
ing intact its syllabic and vocalic cast,^^ and will tell us 
explicitly that such formations are D''''133. The second pas- 
sage will show that the name niiT' yielded surrogates for 
itself in precisely the same manner. 

As the extremely condensed form of part of Nedarim i. 
1, 2, makes a literal translation impossible, I quote the 
Hebrew: Dli'int'l Wiy^TO d^aini a''1"l33 D''-n3 •'^33 S3 (1) 

n3ip D3ip rarh n»i«n (2) . . . nn''t33 ni-i"'T3i nisiatr^s 
tnrh p-'isa ibx nn ^n '^in p'r\ pip"? f'"'i33 "h^ nn d31|'5 
xnia3 ni3 npipir nnintr r\rT\h p''i33 "hm nn n^ts n''i3 p''i3 

nyn^b p''133 lbs nn. That is, All the surrogates employed 
in vows are as binding as the words they displace, likewise all 
those employed in bans, all those employed in oaths, and all 

sonst einem der Attribute" (Hoflmann), and "oderbei alien anderen Attri- 
biiten " (Goldschmidt), both read the Babylonian Gemara (which itself con- 
fuses two entirely different principles) into the Mishna. 

6' Any disturbance of the vocalic cast of the word, in addition to the 
exchange of its consonants, renders it unrecognizable, and relegates it to 
the class of the P'US ^'133, surrogates of surrogates, which are rejected by 
one school of Tannaites as beyond the pale of " Hebrew" ; cf. h. Nedarim 
10 h, j. Nazir 51 d. 


those employed in Nazirite ohligations. ... If one say to his 
neighbor D31p, TOIp, or D31p, these are surrogates of |3']j^ ; if 
he say p^lH, lin, or *^"in, these are surrogates of trT\ ; if he 
say p''W, rflS, or fT'tS, these are surrogates of "1"'W -f^ if he say 
nniDU^, nplpt?, or [. . . ? . . . .], these are surrogates of 
n'^'^^y:^. The reading of the words «ma3 "113, which I have 
left untranslated, is extremely uncertain. The above is the 
text of the separate Mishna editions; the Babylonian and 
Jerusalem Talmuds have ^'nl!3^ ^11, and Maimonides read 
KmttD "1*13.^5 jj seems to me plain that "113 is interpolated 
— it has no place in connection with " oaths " — and that the 
remainder of the clause is corrupted from a third ''1313 of 
ni^lDC^, probably nHtttt. The Jerusalem Gemara on this 
paragraph of the Mishna adds the following D"'133 : pS"l5, 
pi:i5, pp-l5, surrogates of p?1!J ; MSIttn, n3-ll!2n, HpllISn, sur- 
rogates of rrilttri ; D"l3n, ■l?3n, "n3n, surrogates of ^ipn ; 
nSSI-in, r^Tnn, npnn, surrogates of niaiiri (JVedarim 36 c).66 
The other passage, M. Sanhedrin vii. 5, relates to the judi- 
cial procedure in the case of a trial for the capital offense of 
blasphemy, as conducted in the bygone days when the Jew- 
ish state still existed and had the power of life and death : 
The blasphemer is not guilty except he pronounce^'^ the exact 
Name (Dtt?n m tt^lB't^ IS). Babbi Joshua ben Qorha said: 
Throughout the examination of the witnesses it was customary 
to employ a surrogate Q^i^), (as for example) ^^ May T^DV 
smite!" or '■^ (^He blasphemed^ HDV." When the trial was 
completed, however, they did not proceed to the execution upon 

M tWfXi of the text is a manifest corruption ; cf. M. Nasir i. 1. 

"5 See Petuchowski's edition of Seder Nashim, Berlin, 1902, p. 178, 
notes 33 f. 

w The D''^33 cited above form the subject matter of the text of Mishna 
and Talmud and so are necessarily ancient. As is well known, the spellings 
ffpbK, wbVi, ^pbta, nbK, Kp'?X, not infrequently occur in early Jewish 
impressions for ffnbs, etc. ; so also "lIT for Hl.T ; cf. Jastrow, Dictionary, 
s.v. mbs, and the Jewish Enq/dopedia, vol. ix. p. 164. But these are 
doubtless relatively new devices to prevent the desecration of the printed 
name. On the other hand, 'liT' 'BlS'tt which occurs for 'jXltP'' 'nbs in the 
formula of an oath in j. Nedarim 42 c, lines 10 f., is certainly an original 
Talmudic "1)3. 

" In abuse of the Deity or in cursing his fellow ; cf. p. 138, note 47. 


the basis of the surrogate. But they caused everyhody not con- 
cerned to leave the room, and ashed the most prominent among 
the witnesses, saying to him, Repeat exactly (U^1"|S3) what thou 
didst hear. This he did. Then the judges rose to their feet 
and rent their garments, and never mended them thereafter.^ 
And the second witness said, I too heard as he did. And the 
third said, I too heard as he did. 

1 have left the vocable nOV unpointed in both cases, 
because that is the way it was written, and the traditional 
pronunciation of it is neither here nor there. The vocaliza- 
tion is the vocalization of TKT, which tradition has failed 
to transmit. The reading riDV is that of the Jerusalem 
Talmud and is commonly admitted to be the original ; the 
Babylonian Talmud has ""DV, which is manifestly a later 
conformation of the spelling to the familiar Rabbinical name 
"•pV, Jose, the hypocoristic form of ^Dl^ Joseph. HDV is also 
the reading in the text of Sifra (104 e). The words US'' 
nOV ns nOV, which I have construed as two separate exam- 
ples, HDV T\'y and HDV flK, have hitherto been construed as 
constituting together one single example, " May HDV smite 
HDV ! " with the result that from the earliest times they 
have thrown more darkness than light upon the Mishnic 
narrative. The Babylonian Talmud appends the Baraitha 
Dra Dtr (SS,Tty = ) -[-I3''tr "is, which it proceeds to inter- 
pret in the absurd sense that to be guilty a man must have 
cursed God by God. In speculating as to how this marvel- 
ous feat might be accomplished, the Gemara seriously asks 
whether it may be done by engraving the divine name upon 
the blade of a knife and with it piercing through another 
object upon which that name has been written, but concludes 
more sensibly by taking refuge in the text of Lev. 24^^, 
according to which the simultaneous utterance of the name 
and cursing (anybody) constitutes the blasphemy. The 
Gemara has certainly mistaken the intent of the Baraitha ; 
Dt^a Dtt> bbp'^tZ^ "IS means that one is not guilty of blas- 

^ The Gemara explains that the witnesses refrained from rending their 
garments because they had already rent them once at the original hearing of 
the blasphemy. 


pheming the name except he employ the vocable TiW, 
which agrees entirely with the Mishna. On the other hand, 
a sentence in the Jerusalem Gemara upon this section shows 
clearly that HDV nS' and HDV ns must be separately con- 
strued. The witness says (j. Sanhedrin 25 a), Dtt^n "imx 
^h'^p D1 Sbp im« Da-iSS TnaXt^, The identical name which 
I have spoken in your hearing, it he blasphemed, and by it he 
cursed. Hebrew scholars will not find fault with the render- 
ing of 77p by blaspheme in one phrase and curse in another. 
The important point is that we have two distinct phrases 
corresponding in their import to the two of the text. The 
only object of the verb TO'' which would be in place in this 
example of the quoted malediction is the suffix of the second 
person singular, as in M. Shebu'oth iv. 13, D^lbx n33"', Mai/ 
Crod smite thee ! and 0^'^7K nSS^ p Thus may Grod smite 
thee !^^ Nor have we reason to be surprised at the asyndetic 
recording of the two examples ; we need only compare the 
several series of examples in the passage quoted above from 
M. Nedarim. The two illustrations furnished by the text, 
ilDV nS'' and HDV HS, are intended to cover the two forms 
in which it would be necessary to employ the divine name in 
giving and taking testimony at* such a trial : as subject of 
the verb in quoting the blasphemous utterance, and as object 
of the verb in characterizing the deed. In both cases a ^133 
such as HDV was substituted for rtliT. If we point the 
proper name HliT, the ''133 in this passage must be pointed 

The Mishna contains additional evidence that in the 
centuries immediately before and after Christ d^^US were 

«' The citation of these forinulse of the Mishna without the employment of 
a '153 for either .133' or d'n'jS was not permitted in later times; see 
6. Shehu'oth 36 a. 

™ On this last point cf. Blau, Altjiidisches Zaubeneesen, pp. 130 f. As a 
specimen of the havoc which may be wrought by construing all four words as 
a single sentence and adhering to the spelling of the Babylonian Talmud, we 
may cite the interpretation devised by Levy : Jesus (Christ) is mightier than 
Joseph (his father, and so by implication, than God) ! See his Neuhe- 
braisches und chalddisches Worterbuck, s.v. '133. For another fantastic in- 
terpretation (noi' = Zeis) see Kohut, Jewish Quarterly Beview, vol. iii, 
pp. 552 ff. 


employed for the name miT' wWch consisted of alterations 
of the name by the exchange of certain of its consonants. 
According to M. Sanhedrin x. 1 (5. xi. 1), " The following 
have no portion in the world to come : he who affirms that 
the doctrine of the resurrection of the dead is not derived 
from the Law, or that the Law is not of heavenly origin ; 
and the Epicurean (freethinker). Rabbi 'Aqiba says, Also 
he that reads in exotic books and he that whispers over a 
wound, repeating ' I will put none of the diseases upon thee 
which I have put upon the Egyptians, for I am niiT' (pro- 
nounced ""J-l^) that healeth thee' (Ex. IS^S)." To this 
declaration of Rabbi 'Aqiba which anathematizes such as 
practise sorcery, the Mishna appends the dictum of Abba 
Sha'ul, a contemporary of Rabbi 'Aqiba, DtZ^n flS TOinn P|^< 
VriVm>^3. The direct point of this sentence is plainly that 
the mere utterance of the genuine divine name is an offense 
for which the penalty is exclusion from the kingdom of 
heaven. In Tosefta, Sanhedrin xii. 9 (ed. Zuckermandel, 
p. 433), Abba Sha'ul shares with Rabbi 'Aqiba the responsi- 
bility for the doctrine concerning sorcery, while the specifi- 
cation of the other offense comes first and is introduced with 
the anonymous IS^DIH, they have added, the language, how- 
ever, being otherwise identical with that of the Mishna. It 
may be admitted that the extreme religious penalty was 
attached to the act of pronouncing the genuine divine name 
in the opinion of only a certain minority of Rabbinical 
authorities. But the question of the exact penalty for what 
was universally considered a grave religious offense, does 
not concern us here. Of greater importance is the phrase- 
ology in which the doctrine attributed by the Mishna to 
Abba Sha'ul is couched. rUH is not a synonym of 'I?''! or 
"ltt>^ or K'lp. The Old Testament usage is poetical and figu- 
rative, and, taken by itself, affords no idea of the precise 
prose signification of the word in such a connection as this. 
i"Urt is not to utter, nor to rehearse, nor to read, a word ; but 
to pronounce the letters of a word, to combine the letters of a 
word in speech — we should say to vocalize ; cf . the Syriac 
J_^8i and the Arabic L^x^o. The clause of Abba Sha'ul is 


therefore to be interpreted, Also he that pronounces (vocalizes) 
the Name with its own consonants. The plain implication 
of this statement is that the pronunciation (vocalization) 
of the name tl^TV was permissible provided it did not 
employ exactly the four consonants \ H, 1, and H. The alter- 
native in the mind of the speaker was certainly not the 
employment of ''ili^, as is commonly supposed. The substi- 
tution of ■'Hi^ is not the "pronunciation of the name nW 
with any consonants whatever ; the vocalization of the two 
terms is entirely different. In other words, this declaration 
of Abba Sha'ul, upon the only rational explanation of its 
phraseology, practically tells us that it was customary, in 
and before the beginning of the second century a.d., to pro- 
nounce the name mn'' with substituted consonants ; that is, to 
employ D'^^liS of the type indicated above as substitutes for it. 
The first clause appended to the text of the Mishna in the 
Jerusalem Talmud is more germane to the subject than is 
the second. The brief Gemara reads, " Rabbi Mana said. 
After the manner of those swearing Samaritans. Rabbi 
Jacob bar Aha said. It is written with Yodh He, and it is 
read (K1p3) with Aleph Daleth " Q'. Sanhedrin 28 J). The 
statement attributed to Rabbi Mana is manifestly a comment 
upon the text of the Mishna, since it consists of nothing but 
a dependent adverbial clause. And he rightly understood 
the Mishna to be treating of the secular use of the divine 
name : the Samaritans frequently employed the exact name ; 
well behaved Jews made use of a ^133. On the other hand, 
the statement attributed to Rabbi Jacob is an independent 
sentence which is entirely out of place in this connection 
and was originally concerned with a wholly different sub- 
ject, namely with the synagogue " reading " of the name in 
the text of Scripture. K"lp3 is not a word to be used of a 
vocable uttered in oaths, adjurations, and unguarded conver- 
sation. Nor can we suppose that Abba Sha'ul gave himself 
much anxiety over the possibility of the desecration of the 
name in the services of the synagogue or in the Scripture 
quotations of the pious.^^ 

'1 The rule attributed to Babbi Jacob bar Alja in j. Sanhedrin 28 & is 


The Babylonian Gemara on this passage of the Mishna 
(^Sanhedrin 101 6) is of the greatest significance, both for the 
support it brings to the present contention and for the light 
it sheds upon another subject no less important, namely the 
pronunciation of the name mn\ It contains only a Baraitha, 
as short as it is weighty : WK ptrSsi pbon Wn. The 
printed texts of the Babylonian Talmud exhibit HiS for the 
last word ; but the word was alphabetically listed as K3K in 
the (eleventh century) 'Ar4ch of Nathan b. Yehiel,^^ and it 
is commonly admitted that the latter is the original reading, 
which has been corrupted to HJS in the manuscripts under- 
lying the Bomberg and subsequent editions of the Talmud.'* 
The meaning of the first two words of the Baraitha is, of 
course, perfectly plain : It has been taught (in limitation of 
the condemnation of the Mishna), In the territory (that is, 
outside of the central sanctuary at Jerusalem). But the last 
two words have hitherto defied successful interpretation.'^* 
It is customary to attempt the explanation of the word WK 
by reference to the form KH employed in the Samaritan 
Targum to render the Hebrew ^pi of Lev. 2i^K But this 

found in entirely different contexts in tlie Babylonian Talmud, Pesahim 50 a, 
and Qiddushin 71a. Both these passages present the rule in the form of an 
utterance of God in the first person, "The holy One, blessed be he, said, 
Not as I am written am I read ; I am written with Todh He and I am read 
with Aleph Daleth ; " which is a fanciful elaboration of thxh 'fitT ."IT of 
Ex. 31* with the last word pronounced D^c"?. It should be noticed in passing 
that in none of these three passages where ')HK is mentioned as the " read- 
ing " of nin', is there any talk of D'"153, or any hint that '51« was habitually 
spoken of in Talmudic times as a '1)3. 

" See the Bomberg edition, Venice, 1531/2, fol. 4 6 ; Eohut, Aruch Com- 
pletum, vol. i. p. 20 ; and cf. Buxtorf and Castell, s.v. J1C. 

" This corruption, as will appear, was probably not without design. 
Some persons seem to have understood the passage better than did the author 
of the 'Arxich. 

'♦ Cf. Goldschmidt, Der hahylonische Talmud, vol. vii, Berlin, 1903, p. 
446, " jedoch bleibt der Ausdruck HJl? pc'? dunkel." Dalman {Oottesname 
Adonai, p. 50, note) rightly rejects the interpretations of the 'Aruch, of Bashi, 
and of Levy, but the one he offers is equally unsatisfactory. He is of the 
opinion that K3K stands for nXJK, which in turn is a " Nebenform " of an 
assumed rtan = " Das Lesen, Vokalisieren." In his more recently pub- 
lished Aramaisek-neuhebraisches Worterbueh he defines, " KJX h. das buch- 
stablicbe Aussprechen " ; but the Neo-Hebrew for this last is B>n§. 


is to explain one mystery by another. Castell could do 
no more with Samaritan HiX than refer to HiS of our pas- 
sage in the Babylonian Talmud, with the remark that from 
the Samaritans' employment of this word we may see how 
their hatred of the Jews did not prevent them from reading 
the Talmud and borrowing its vocabulary. Then, too, we 
have only to compare the Vulgate rendering of Lev. 24ii-i6 
to see that it is by no means self-evident that the Samaritan 
rWS is an exact synonym of Hebrew 5p3, — to say nothing of 
the fact that we are not quite unanimous about the precise 
meaning of the Hebrew, or the circumstance that HWK, as 
well as n3, disputes with n3i< the latter's place in the 
Samaritan Targum.'* Furthermore, neither of the two 
interpretations which naturally suggest themselves for rWX 
of the Samaritan Targum of Lev. 24^^ will suit our passage, — 
leaving the question of the grammatical construction entirely 
out of account. If we adopt the meaning to pronounce, the 
Baraitha adds nothing to the Mishna ; and if we adopt the 
meaning to curse or blaspheme, the passage, taken in connec- 
tion with the Mishna, would be reducing to a mere offense 
against the conscience, punishable by God alone, a crime for 
which the Mosaic law prescribed death by stoning. The 
fact is, we are far more likely to be helped to an under- 
standing of the Samaritan text by a correct interpretation of 
the Talmudic passage. 

The correct interpretation of the Baraitha is this : To the 
statement of the Mishna that among those who have no 
portion in the world to come is he that pronounces (^vocalizes') 
the Name with its own consonants, the Baraitha adds the con- 
ditions, In the territory (outside of the Temple) AND WITH 

We must remember that the Jews of the period lacked 
our grammatical concept of vowel, and that they had not 
exactly our clear-cut concept of consonantJ^ They did not 

'5 See VoUers's edition. 

™ On the general subject of the grammatical attainments of the Babbis, 
see Berliner, Beitrage zur hebrdisehen Grammatik im Talmud und Midrasch, 
Berlin, 1879. The author, however, fails to bring out the point made here. 


separate the elements of language into two mutually exclu- 
sive and supplementary categories. A word, to their minds, 
had its several nVfllK (properly mmX), signs or letters, 
which appeared upon the written page, and among which 
^, n, 1, and ■' (no matter how employed) had an equal 
place with 3 and 3. And it had besides its peculiar speech 
or mode of utterance, what we should call its pronunciation 
rather than its vowels. If we attribute to them the con- 
ception of vocalization, we must think of it in the sense of 
making vocal rather than in that of adding vowels to the con- 
sonants, as we are apt to do. Moreover, the letters of a 
word were so many distinct units, but its vocalization was 
one single mode of combining those units in speech. Thus 
D^'^7'S had five niTni*, but only one ptl?7, mode of utterance, 
^^ Anssprache," namely JElohim. Ordinarily the need of ab- 
stracting the vocalization of a word from its consonants did 
not arise, and the ]Wb (spoken word) would of course 
embody its own proper mTllK. But this Baraitha records 
an oral tradition which could not, without committing the 
very sin it inveighs against, specify the prohibited vocaliza- 
tion in the ordinary way. It accordingly effects the necessary 
abstraction in about the only way possible to it — by inserting 
an arbitrarily chosen consonant between two H's. This is 
the nearest it can come to indicating " two syllables, with the 
a sound in each." 

Confirmation of the above interpretation is supplied from 
a somewhat distant quarter. I refer to the much quoted 
passage in Theodoret, Quaestiones in Exodum, xv. (Migne, 
Patrologia Grraeca, Ixxx. 244): "This (the divine name 
revealed to Moses) is called by the Hebrews ' unutterable ' 
(ai^pacTTov) ; for it is forbidden them to quote it with the 
tongue (Sta t^? yXanrTji 7rpo(f>epeiv'). It is written with four 
characters, wherefore they refer to it also as ' the four- 
lettered' (jeTpdjpa/ifiov') .... And it is called by the 
Samaritans la^e,'''' but by the Jews Ata." This Aya of 

" In the previous century, Epiphanius, whose Jewish birth is at least ques- 
tionable (cf. Bonwetsch, PRE", vol. v. p. 418), doubtless derived his identical 
lo/Se from a Samaritan source; cf. Dietrich, ZATW, vol. iii. p. 298. The 


Theodoret has occasioned much discussion. Gesenius^^ 
thought it reproduced the ^^'^X of Ex. 3'*, and his view was 
adopted by more than one eminent scholar. It was, how- 
ever, warmly disputed by Dietrich in a series of letters to 
Franz Delitzsch on the pronunciation of nXT.'^^ Dietrich 
refused to believe that any Jew to whom Theodoret applied 
for information could have been so ignorant as to suppose 
that iTnJ< was pronounced ah-ja, or that Theodoret would 
have accepted the definition of the name from Ex. 3^* when 
he had asked for its pronunciation. That Church Father 
must have repaired to the most muddled and ignorant Jew 
alive to extract such misinformation. On the other hand, 
Dietrich was positive that "et'n Jude [the italics are his] 
jederzeit sieh eher wiirde haben todt schlagen lassen, ah dass er 
einem wirklich heidnisehen, oder einem sich Christ nennen- 
den ' ^13 ' den allerheiligsten Namen, wenn er auch die iilteste 
Aussprache wusste. Moss zur Befriedigung der gelehrten Neu- 
gierde ausgesprochen hdtte." He accordingly reached the 
conclusion that what Theodoret's Jew really furnished him 
was the simple la (the abbreviated form HJ of the Old 
Testament, which the Jews did not hesitate to pronounce), 
and that the Syrian-born theologian prefixed a prosthetic a 
on his own account.^* In the opinion that no Jew would 
have committed the sacrilege of pronouncing the ineffable 
name merely to satisfy curiosity, Dietrich was quite right. 
No Jew would have been guilty of that act even for a more 
laudable purpose. But he was altogether wrong in suppos- 
ing that the Jews of the time were generally ignorant of the 

statement of Kautzsch, Encye. Bib., col. 3321, note 4, that the pronun- 
ciation Io/3e is ascribed by Epiphanius to a Christian sect, is incorrect; 
the passage in which the name occurs, Adv. haer. I. iii. 20(40), is a paren- 
thetic bit of lexicography on Epiphanius's own account, occasioned by 
the heretics' ignorant employment of the word Sebaoth as an independent 

" Thesaurus, p. 577. 

" The letters were published by Delitzsch after Dietrich's death in the 
ZATW, vols, iii and iv. The statements on this point will be found in vol. 
iii. pp. 282 f., 287 f., 293 f., 296 f. 

8" Dietrich preferred not to rely upon the reading la, which happens to be 
found in one patristic manuscript. 


true pronunciation of the name,®^ and equally wrong in 
the supposition that the name was as secret as it was sacred. 
The Jew of whom Theodoret made inquiry evidently did 
indicate to that respectable and learned Gentile the authentic 
pronunciation of the name, and he did so without violating 
either its sacredness or his own conscience, by reproducing 
separately the abstract vocalization which belonged to the 
four characters mn''. Theodoret manifestly mistook the pur- 
port of the reply, but heard distinctly and recorded faith- 
fully the reply itself. His informant said AYA, agreeing 
entirely with the author of the Baraitha in b. Sanhedrin 
101 J, for the medial consonant was of course arbitrarily 
chosen in each case and hence variable. 

A word needs to be said regarding the conclusion to be 
drawn from the above testimony as to the pronunciation of 
mn'' in the Jewish tradition of the Talmudic period. The 
Jews, like the Arabs down to the present time, heard the sound 
a' ( = e as in " there ") only as a phase of a (as in " far "), 
and would as readily employ >^ to represent the former 
sound as we should spell a new word containing that 
sound after the pattern of "man." Hini in the so-called 
Babylonian system of punctuation cannot be distinguished 
from nin\ Moreover, in reproducing the vocalization apart 
from the word's proper consonants, they might easily em- 
phasize the identity of the sound by a little flattening ; so 
that not even Theodoret's Aia is conclusive for a as against 
a. In any case, we have here conclusive evidence that the 
historical pronunciation of mrT' lay between the following : 
Yahwa, Yahwa, Yahwa, YahwS. We shall have occasion to 
return to this subject below. 

Turning now to the previous question, and assuming for 
the moment that the true vocalization of mn'' may be repre- 
sented as Yahwa, the doctrine of Mishna and Baraitha of 
Sanhedrin x. 1 combined is this : Only in the temple at 

81 Blau (^Altjudisches Zavbervjesen, p. 128 f.) takes issue with Dietrich on 
this particular point, but it has not occurred to him to ask. himself how the 
knowledge he claims for the Jews of Theodoret's time was transmitted to 
them or maintained, — unless he takes seriously the story of its septennial 
transmission recorded in 6. Qiddushin 71 a. 


Jerusalem was it permissible to say Yahwa; elsewhere one 
might not utter that name without committing a sin against 
God. But one might with impunity exchange certain of its 
consonants, and say (for example) Yahma, employing a 
''133 of the name. Also one might pronounce its four con- 
sonants with a different vocalization, saying (for example) 
Yehawwek, in which case one would be expressing a wholly 
different word. 

The evidence that it was customary upon occasion to pro- 
nounce the name MW in altered form is not yet exhausted. 
One of the designations of that name current in Talmudic 
times is tt>"11San DtJ>. The intelligent use of this term 
ceased very soon after the close of the Talmud, and its exact 
meaning and origin have ever since remained a mystery. 
The term does not occur in the authentic text of the 
Mishna.*^ It is found, however, in the Baraitha b. Sota 
38 a, in Sifra on Lev. 24" (104 c), and in Sifre on Num. G^s 
(ed. Friedmann, fol. 12 a), 6^ (13 6); as well as in the 
Babylonian Gemara, Yoma 69 5, Hagiga 16 a, Sanhedrin 
60 a, and in the Midrashim, Shir-hasMrim JRahha on 4^, 
Koheleth Babha on 3"6. The Aramaic Ktt^lS? «tttt> occurs 
in two passages of Targum Jerushalmi II, Ex. 322* ^nd 
Lev. 24^^, and is employed unintelligently in an Aramaic 
anecdote in the section of Koheleth Babba just mentioned.^ 

Though the suggested definitions of the term tyiSttH OO 
have been so numerous as to represent almost all the alter- 
natives imaginable, thus far none can be said to have gained 
the general assent of scholars.** In recent times,*^ it has 

82 It is found In corrupt texts of M. Yoma vi. 2, which section continues 
to be cited in this connection ; so by Blau, I.e. p. 124, and the Jewish Encyclo- 
pedia, vol. ix. p. 162. But the entire passage in which the term occurs is 
an interpolation, lacking in the best manuscripts and printed texts ; cf . 
Balman, Gottesname Adonai, p. 40, note, and Goldschmidt, Ber babylonische 
Talmud, vol. li., Berlin, 1901, p. 942. 

83 KSPIBa KB?? looks like a clumsy reproduction of the Hebrew WIBSn OV. 
The true Aramaic would be KE^IBai KiV ; cf. nrrm KBW of Targum 
Jerushalmi I, Lev. 24w. 

M Cf. Gottheil, JAOS, vol. xviii. p. 361 ; Blau, f.e. p. 124. 
85 For some of the older literature see pp. 504 f. of Nestle's article quoted 


been defined by Low ^ and Oppenheim *^ as the name which is 
engraved (upon the High Priest's diadem) ; by Rahmer ^ as 
the name that is explained (in Ex. 3^*); by Griinbaum,^ who 
has written most voluminously and most confusedly upon the 
subject, as the concealed or the mt/sterious name. By Nestle,^ 
on the testimony of Syriac lexicographers, depending ulti- 
mately on a scholion of Jacob of Edessa (who never knew the 
meaning of the term and had forgotten its exact form when 
he wrote about it), and by Friedlander ^^ and Bacher,^ who 
equate it with ^^'Sl^ Dt?^, it is defined as the name which is 
reserved for or peculiar to (God), that is, the proper name. 
Torrey ^^ interprets it as the holy name. The following ren- 
derings are more or less related : Geiger,^ " der ausdriick- 
liche Name " ; Cassel,^ " der nach seinem wirklichen Laut 
ausgesprochene Name " ; Munk,^ " le nom distinctement pro- 
nonc^ " ; Fiirst,^" Nager,^ Levy,®* and Dalman,!"" " der deut- 
lich ausgesprochene Name." 

So far as concerns the meaning of the verb contained in 
the participial form t2^"iatin, the last group of definitions is 
certainly most in accord with the usage. Not to quote 
Geiger, who bases his definition of our expression upon this 
fact, Bacher, who thinks himself forced to a variant conclu- 

w Beitrage eur judischen Alterthumskunde,, I. 1. p. 25.* This and the 
following references marked with an asterisk I am obliged to make at second 

^^ Monatsschrift fur die Geschichte nnd Wissenschaft des Judenthums, 
vol. xviii. pp. 545 fi., vol. xix. pp. 326fl.» 

88 Monatsschrift, vol. xix. p. 187.« 

«9 ZDMG, vol. xxiii. p. 632, vol. xxxi. pp. 225 ff., vol. xxxix. pp. 543 ff., 
vol. xl. pp. 234ff. 

*> ZDMG, vol. xxxii. pp. 466 if. ; of. Bernstein, ibid. vol. iv. pp. 199 f. 

'1 The Guide of the Perplexed, of Maimonides, vol. i. p. 226, note 3. 

'^ Terminologie der Tannaiten, pp. 169 f. 

93 JAOS, vol. xviii. pp. 180 ff. 

9* Urschrift, u.s.w., p. 264. 

95 Monatsschrift, vol. xix. pp. 73 &.* 

^ Le Guide des Sgares, vol. i. p. 267.* 

9' ZDMG, vol. xxxiii. pp. 297 ft. 

9« ZDMG, vol. XXXV. pp. 162 ff. 

99 Neuhehraisches und chalddisehes Worterbtich, s.v. O^. 

i<» l.c. p. 63. 


sion in the case of ty"lStin OO, has this to say of the Neo- 
Hebrew use of the intensive T!?"lS:ioi Its standard meaning 
is " deutlich machen, bestimmt aussprechen, heraussagen. . . . 
Unser Verbum dient auch dazu, um anzugeben, dass etwas, 
was in der Thora unbestimmt gelassen oder nur angedeutet 
ist, in den prophetischen und hagiographischen Biichern 
deutlich ausgesprochen wird." The use of tt^^lS, the nomen 
aetionis of Piel he illustrates with 131 h^ WrVSi p^Tt, 
" Wo findet sich die Sache deutlich ausgesprochen ? " '^V^B 
is further "die genauere Bestimmung einer biblischen Sat- 
zung." The adverbial 2?11SSl is "ausdriicklich," with a ref- 
erence among others to the passage M. Sanhedrin vii. 5 quoted 
above Cpp- 147 f.) The participle t2>*nb& means "bestimmt 
deutlich gekennzeichnet," as opposed to DlflD, " unbestimmt, 
undeutlich." ^""-^ Bacher goes on to say that C>"lB& has still 
another signification, " abgesondert," but the examples he 
cites yield the closely related meaning of "distinguished" 
rather than that of "detached." Moreover, for our present 
purpose it is most important to consider the sense in which 
CIS is used in connection with the divine name. In the 
passage we have quoted from the Mishna, Sanhedrin vii. 5, 
the intensive of tt^"lB has the divine name as the object, and 
the meaning there is unmistakable. The blasphemer is not 
guilty until he reproduce exactly the Name (U<DT\ tJ>"]B't^ *15), 
that is, until he make use of the exact name. And further 
on in the same paragraph, tt^I'lM TOSitZ^C' Ht2 "llOtt, Repeat 
exactly what thou didst hear ; and according to the express 
statement of the text, t£?1"lSS indicates TKV as against TOT. 
To this we may add the testimony of the Aramaic usage in 
connection with the divine name. In Lev. 2411-16 of both 
Targum Onqelos and Targum Jerushalmi I, forms of t^lS 
are used to render the Hebrew 3p!J. Whatever be the pre- 
cise meaning of Sp3, whether to utter or to curse, no one will 
pretend that it is to separate. Compare also the Peshita of 
Lev. 241*. It remains to point out that the act of speaking 
is not necessarily bound up in the conception of ti^"!S. It 

»« l.c. pp. 154 ft. 11^ tc. p. 137. 


means to indicate exactly by any means. So in Targum 
Jerushalmi I of Ex. 3226 ^q have tT^naiDl p'pn = exactly 
engraved, and in Targum Koheleth 3^^ tt>"iai!21 S^fl? ^^ = 
exactly written ; in both cases of the divine name. In view 
of all this evidence I fail to see how it can reasonably be 
questioned that Hebrew tt^^SOn is that which is exactly indi- 
cated or set forth?-'^ 

Now in spite of the facts we have rehearsed, the rendering 
the exactly pronounced name for tT'lSSH WU has not hitherto 
met with general acceptance. The basic objection to it was 
voiced by Nestle in 1878 : " Wie das Tetragrammaton [which 
as a matter of fact was not pronounced] der deutlich ausge- 
sprochene Name genannt worden sein soil, wenn nicht wie 
lucus a non lucendo, sehe ich nicht ein." i<® 

In meeting this objection, we must consider a point which 
has not received the attention it deserves, namely, the gram- 

103 See Levy, Das Targum zu Koheleth naeh sMarabischen Handschriften, 
Breslau, 1905, p. 11. 

iM The Old Testament fiiraf \ey6iievov !tf"JB!2, used adverbially in Neh. 8», 
admits of no other rendering than with accuracy ; and the Perfect Pual of 
Num. 15** demands the corresponding interpretation, ^b TWVl nO \II>1B K? '3 
for it had not been exactly set forth (specified) what should be done unto him. 

A similar interpretation is required for the Pael passive particle in the 
Syriac term j^i'iV? ^i t \^ o) which continues to be erroneously ren- 
dered "The Gospel of the Separated Ones" ; so Burkitt, Evangelion Da- 
Mepharreshe, Cambridge, 1904, vol. ii. p. 31, " that is, the Gospels divided 
into the four volumes of Matthew, Mark, Luke, and John "; cf. Encyc. Bib., 
col. 4999. But when were the Gospels divided? Of this Interpretation 
Torrey very properly remarked in 1897, "There could be no reason for 
speaking of the four Gospels as ' separated ' ; least of all if they were to be 
contrasted, as the original form, with a mixture like the Diatessaron" 
{I.e. pp. 178 f.). That the term l-4.faicj j is, however, antithetical to 
I J V Y«-». I, which serves to characterize the Diatessaron, cannot be doubted, 
as was shown by Gottheil {ibid. p. 361). But it does not follow that 
I ^ ..o^/i. is therefore the separated. The antithesis of harmonized or com- 
bined is not disharmonized or separated, but unharmonized, uncombined. 
And this is precisely the purport of the participle | *yq\a, reproduced in 
their exact or authentic form. | AyiSiff; ^at\.-3al is therefore The Gospel 
of the (four) exactly reproduced, or as we should put it. The (four) Gospels 
in their exact form. For the rest, it seems to me that even when employed of 
ordered lections, the idea embodied in the participle is that of exact demar- 
cation rather than that of division. 

JO* I.e., p. 505. 


matical form of our expression. All the renderings of 
triSfin D^ (the engraved name, the explained name, the 
concealed name, the proper name, the holy name, the ex- 
press name, the distinctly or exactly uttered name) treat the 
term as if it consisted of a determinate substantive with at- 
tributive adjective attached, that is, as Neo-Hebrev? for Dtrn 
C^IBSH. But that is not the actual construction in this 
case.^*'^ C?"iattrt is not an attributive adjective, but a sub- 
stantive in the genitive. The true solution of the matter 
seems to me this: 

wnaan tm = ,t.t ds^ 

tt^-lSan = the vocable ni.T.iw 

Once more I call attention to the terminology of M. Sanhe- 
drin vii. 5. The vocable HDV we are told is a '"1113 of miT. 
So that when pronounced HDV, or with any other exchange 
of its consonants (^1333), the name mn^ would be ri3310 ; 
on the other hand, when pronounced nin\ it is tl^'lBSSH. 
This meets fully the objection expressed by Nestle, and at the 
same time supplies additional confirmation of the proposition 
with which we set out ; for of course the mere existence of 
the expression requires us to assume that the name was 
sometimes purposely mispronounced. The term tt^'nSttn WU, 
then, is not the name which was exactly pronounced, but the 
name miT' with its exact pronunciation. We may now quote 
one of the earliest passages in which the expression occurs, 

S!fre on Num. 6^3 (fol. 12 a): DU^S hiTilS^ "33 ns 13-|3n ."13 

S'-n ''i3''33 nhn iTK i« tt^-nssan ntrs laix nm ^rmaan 
"is-'ss n3'na3i tt?-naan Dtrs bx-itr" "33 hv "a^ ns i)Ott?i 

lo* Grunbaum concluded that we have here " ein neugebildeter, besondrer 
Kunstaasdruck," ZDMG, vol. xxxix. p. 556 ; and Torrey agrees with him. 
It is not, however, the meaning of t71B)S that is peculiar in this expression ; 
it is the construction. 

1" By this I do not in the least intend to imply that ni!T tfB in the Old 
Testament is " das Wort Jhvh," as Jacob, Im JVamen Oottes ; on which see 
HeitmuUer in the Theologische Literaturzeitung for 1905, cols. 369 fl. The 
equivalents indicated above are purely scholastic, and would have been 
quite impossible in an age when people habitually used ,"Iln' immediately of 
the person of their God. 


'■'■Thus shall ye Mess the children of Israel." — That is, with the 
name in its exact form. Thou sayest., '• In its exact form? — 
am I not to understand, with a surrogate ? ' The statement is, 
'•'■They shall put my name upon the children of Israel:" — 
with the name in its exact form; only in the town with a 

In the foregoing pages I have attempted to show (1) 
that the Masoretic text of Ex. S^* goes back to the fourth 
century B.C. ; (2) that v.i*» is an interpolation dependent 
upon the reading Oyh^ ^mb^ .T'nx of "» ; (3) that the 
reading T^^n^ in i*^ represents the alteration of an original 
^n^^ which had not been effected at the beginning of the 
fifth century B.C. ; (4) that that alteration was not acci- 
dental, but due to the purpose to prevent the utterance of 
the ineffable name in this one passage of the Pentateuch 
where the employment of the ordinary synagogue surrogate 
for rnn\ namely ''ilH, was from the nature of the case 
impossible ; (5) that the alteration took place, accordingly, 
sometime during the fourth century B.C., most probably 
coinciding with the spread of the Pentateuch and the rise 
of the Jewish synagogue ; and finally (6) that such altera- 
tions, both of the name niiT and of other religious terms, in 
cases where it was necessary to employ the word and yet 
desirable to avoid its actual pronunciation, are abundantly 
evidenced for the ensuing period ; the regular method of 
alteration being to exchange one or more of the consonants 
of the word while leaving intact its syllabic and vocalic cast. 
If these positions have been satisfactorily sustained, iTriX of 
Ex. 31*^ is a purely phonetic ^133 of mn\ entirely devoid of 
meaning, and differs in sound from the proper name of the 
God of the Hebrews only in substituting K for ^ and "* for 1. 

I have not overlooked the fact that in one unimportant 
respect the word ^^'^S is unlike the other purely phonetic 
D"'^13D which have been adduced from the Rabbinical litera- 
ture : it is not in itself a non-word that can never be any- 
thing but a "'lis. It is, however, in my judgment, a sufiicient 
reply to this superficial objection to point out that in the 

MS Cf. p. 144 aboTe. 


fourth century B.C. the dcYice of the ''1)3 was doubtless 
still in its incipiency; and further, that the selection of 
n\"IS was clearly suggested by the words "]ttP ^^^^< ^5, put 
into the mouth of the Deity in the preceding v. ^^ ; it so 
happened that the requirements of a phonetic "'133 were fully 
met by this word, which from its essential character as a 
verb could not be mistaken by the hearer for anything but 
a surrogate of niiT. Of this we may be certain, that ^^'^K 
was chosen to replace nin"' solely because of its phonetic 
availability and without any regard to its positive lexical 
value, — exactly as hleu is used for Bieu in French oaths. 

On the other hand, the virtually contemporary interpolation 
of v.i*« with its ^^'^>i "lU^X ^^^K, leaves no doubt whatever 
that n\1>^ of 1** was pronounced as the first person singular 
Imperfect of the verb ^^'^, and is not a merely coincidental 
group of letters of whose vocalization we can know nothing. 
This being so, our passage supplies us with much earlier 
evidence as to the vocalization of the name TKV' than any we 
have hitherto possessed. For we now know that in the 
fourth century B. c. it was pronounced with the same vowels 
as was the first person singular Imperfect of the verb iTTI. 

It is established, in the first place, by native testimony of the 
best possible kind, that the name consisted of but two sylla- 
bles. The testimony is the best possible, because the only 
better would be a direct statement that the word had two 
syllables, which is impossible in the mouths of people who 
lack the concept of "syllable." 

There remains, in the second place, the question as to the 
quality of the vowels of the verbal form iTHS in the fourth 
century B.C. As regards the vowel of the second syllable, the 
question can be definitely answered. n_ of the Imperfect 
Qal of Tvh verbs, according to the best judgment of modern 
scholars, represents a direct transition from former ai Q^'),^'^ 
which must, however, have ceased to be heard before the form 
was spelled with final T\. The vowel of the second syllable of 
iTilK, therefore, had the sound of e in " there " or a in " fare." 
Regarding the short vowel of the first syllable, we cannot 
109 See Kautzsch, § 75 e, and the literature there cited. 


be quite so precise. The vowel of the preformative of the 
Qal Imperfect of all verbs was originally a, and according 
to the Tiberian punctuation it had not departed very widely 
from that primitive sound in the case of the first person 
singular, when the vowel-points were invented. It is true 
that the so-called Babylonian system of punctuation points 
the prefix of the first person singular, like that of the other 
persons, invariably i.^^" But that certainly represents a later, 
not an earlier, phase of development than the one arrested 
by the Tiberian pointing. For the line of phonetic change 
in the quality of the vowels is from (1) a as in "far" to 
(2) a as in "fare" (=e^ as in "there"), to (3) e as in 
"pet," to (4) e as in "pretty" (= i as in "pity"). More- 
over, just as we have no ear for (2) in very short syllables, 
so the Hebrews had no ear for (3) except in long syllables. 
The phases to be reckoned with in the short first syllable of 
iTHX are therefore, a, a, and i. And there exists no reason 
for supposing that the Palestinian pronunciation had already 
traveled through the final i stage and was on the way back 
again when the vocalization was fixed by means of the 
Tiberian pointing. The statement of Qimchi ^^^ that vlDpS 
was pronounced with Seghol in the prefix to distinguish it 
from blOp'' need not be taken seriously, even if we admit 
that the latter was pronounced iqtol (without consonantal y). 
In Modern Arabic the vowel of tlie prefix of the Imperfect 
has been changed to i in all persons but the first singular, 
where, under the influence of the i< and without the aid of 
artificers, it remains a.^^^ So in Hebrew, the influence of the 
guttural X checked and prevented the development of the 

11" See Kahle, Bet masoretische Text des ATs naeh der Uberlieferung der 
babylonischen Juden, p. 53. Yet according to Diettrich, ZATW, vol. xx. pp. 
153 f., Yemen manuscripts of Targum Onqelos point the prefix of the first 
person with a in Aramaic only under the influence of the Hebrew ! 

»" See Kautzsch, § 47 6. 

11* See Vollers, Lehrbuch der aegypto-arabischen Umgangssprache, p. 28, 
and of. his note 3 on p. 29. 

Konig, Lehrgebdude, i. p. 159, says, " Der ursprungliche Vocal der 
Praformativa, namlich S,, hat sich im regelmassigen Verb immer zu I zuge- 
spitzt, welches durch K zu 6 zerdrticlct wird." But when was the S absent, 
that there might be a point to crush ? 


vowel in the prefix of the first person to the i form reached 
in the case of the other persons. 

Accordingly, in the fourth century B.C. ^^^X was pro- 
nounced either 'aht/a or, with a slight sharpening of the 
first vowel, ^dhyU. Similarly, in the fourth century B.C. 
mri'' was pronounced Yahwd, with possibly a slight modifi- 
cation of the first vowel in the direction of Ydhwd. And 
we have only to turn away from the printed page and trust 
entirely to the ear, to realize that the short vowel of the 
first syllable may very well have varied from one shade 
to the other in different localities and individuals. This 
conclusion is in harmony with the testimony of the Baraitha 
in h. Sanhedrin 101 h, and with the statements of Epipha- 
nius and Theodoret, which have been noticed above.