Skip to main content

Full text of "The Growth of the Peshitt â Version of the New Testament. Illustrated from the Old Armenian and Georgian Versions"

See other formats


STOP 



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

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

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

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

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



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



THE AMERICAN 

JOURNAL OF THEOLOGY 

Volume I OCTOBER 1897 Number 4 



THE GROWTH OF THE P E SHITTA VERSION OF 
THE NEW TESTAMENT. 

ILLUSTRATED FROM THE OLD ARMENIAN AND 
GEORGIAN VERSIONS. 

By F. C. CONYBEARE. 

Oxford, England. 

The P e shitta the sheet anchor of those who uphold the Textus Receptus 
of the New Testament. — But in the P e shitta we have only a late recension of 
the Syriac text. — Witnesses to the pre-P e shitta text of the Pauline letters, 
viz., Ephrem's commentary, and the old Georgian and Armenian versions. — 
Ephrem's text of Pauline letters shown by examples to have been in the main 
identical with P e shitta.— Proof from examples that Georgian and Armenian 
texts of Pauline letters have a Syriac basis. — Syiiac readings in the Georgian 
version of Pauline letters. — Syriac readings in the Armenian version of Paul- 
ine letters. — Recapitulation of results gained so far. — An agreement of 
Georgian or Armenian with Ephrem against P e shitta reveals an older form of 
the Syriac. — Examples of such agreement in the epistle to the Romans. — 
Also from the epistles to the Corinthians. — Ephremic readings in the Arme- 
nian Vulgate. — Other archaic readings in the Armenian and Georgian ver- 
sions. — Summary of results: The P e shitta text of the Pauline letters is a 
later recension of the Syriac text used by Ephrem and by the first translators 
of the Armenian and Georgian versions. — That in the gospels equally it is not 
the P e shitta but the Sinaitic-Curetonian Syriac which is the basis of the 
Georgian and Armenian is certified by a collation of Luke, chaps. 3-5. — 
And of Matthew, chaps. 1-8. — And of Mark, chap. 12. — Summary of conclu- 
sions. The Armenian and Georgian translations of the fourth century are full 
of Sinaitic-Curetonian readings, and contain no P'shitta readings which are 
not in Syr* 1 " or Syr cur . Therefore the prevailing Syriac text in the fourth 

883 



884 THE AMERICAN JOURNAL OF THEOLOGY 

century was not the P e shitta, but Syr sin and Syr cur . — The readings of Syr sin 
preponderate in the Georgian and Armenian versions over those of Syr cur . — 
Untenability of any view of the P e shitta which requires us to suppose that the 
Syriac missionaries of the fourth century ignored the P e shitta and wilfully 
circulated "a corrupt form of Syriac text." 

I. It was almost a historical occasion when at the com- 
mencement of the Oxford Easter term in this year members of 
the university who take an interest in the textual criticism of the 
New Testament were invited to meet in New College and dis- 
cuss the problem of the origin of the Western text. The Rev. 
Dr. Miller, the literary executor of the late Dean Burgon, opened 
the discussion. His line was to rescue the Textus Receptus 
from the position of inferiority assigned to it by Westcott and 
Hort. Then followed Professor Sanday, who in friendly but 
learned language disclaimed Dr. Miller and Dean Burgon and all 
their views. Both these speakers had touched on the problem 
of the old Syriac versions. The P e shitta is, said Dr. Sanday, 
the sheet anchor of those who defend the Textus Receptus or 
Antiochian text. This text is, according to Professor Sanday, an 
eclectic one formed at Antioch in the fourth century, under the 
influence of Lucian the Martyr. No, argued the other side. It 
cannot be a text so late in its formation, for it is the text of the 
Syriac P e shitta, and the P e shitta is a version of the second cen- 
tury, and surpasses even the Greek text in the number and 
antiquity of its oldest manuscripts. Then Dr. Gwilliam, the 
well-known Syriacist, who is editing the P e shitta for the Claren- 
don Press, stood forth to plead the cause of its antiquity and 
superiority over the "heretical" Syriac text lately brought back 
from Mt. Sinai by Mrs. Lewis. He was followed on the same 
side by Mr. Bonus, 1 another eminent Syriacist, who argued that in 
the immobile and stationary East any sweeping revision of the 
Syriac text read in the churches is unlikely to have been made 
as late as the last half of the fourth century. Lastly, Mr. 
Headlam, of All Souls, uttered what was perhaps the most 
pertinent remark in the whole discussion. Dr. Gwilliam had 

•Author of the Cottalio Codicis Lewisiani rescripti; Oxford, Clarendon Press» 
1896. 



peshittA version of the new testament 885 

dwelt with emphasis on the number and antiquity of the oldest 
codices of the P e shitta and on the perfect and minute accord 
with one another of these early codices. Such accord, replied 
Mr. Headlam, is in itself evidence of the lateness and newness 
of the text which these codices contain. If it were an old 
text, if it had a history behind it, then its codices would be 
full of variants. There are no variants in these P e shitta manu- 
scripts ; therefore the P e shitta was a brand-new text or recension 
when these manuscripts were written. 

But none who joined in the discussion seemed to be aware 
that there exist two ancient versions of the New Testament, the 
Georgian or East Iberian of the Caucasus, and the Armenian, 
both made from an early Syriac text and both fraught with the 
most important evidence about the point in dispute, viz., whether 
the P e shitta is or is not the earliest Syriac and in that sense the 
best and least corrupt form of it. The object of this article is 
to set forth this evidence, and in doing so we shall begin with 
the earliest writings of Christianity, namely, Paul's epistles, and 
then proceed to the gospels. 

II. Is the P e shitta text of the Pauline epistles an adaptation 
or modification of an earlier Syriac text, or is it original ? The 
example of another version illustrates my meaning. Jerome's 
Latin text of the Bible is a recension of an earlier Latin ver- 
sion, a remodeling of an older translation by Greek and even 
Hebrew originals. Is the P e shitta text the result of a similar 
recension or is it the earliest Syriac version all unchanged and 
in its full integrity as it left the hand of the translator? Let 
us anticipate and state the conclusion to which this study will 
compel us. It is this : the P e shitta text is a secondary one, a 
recension made about the end of the fourth century of an earlier 
Syriac translation. In making this recension the Greek manu- 
scripts were consulted afresh, and the P e shitt& text must never 
be regarded as a second or third century document without 
being tested and proved to be so on independent grounds. 

III. In the case of the gospels we have the Curetonian and 
the Sinaitic Syriac text to help us to resolve the question of the 
integrity and originality of the P e shitta text. But for the Syriac 



886 THE AMERICAN JOURNAL OF THEOLOGY 

text of the Pauline letters we have no such direct aids. We 
must depend upon two sources only. 

Firstly, on a commentary written in the middle of the fourth 
century by S. Ephrem ; of which, however, we have not got the 
original Syriac. All that remains to us of it is an old Arme- 
nian version, made perhaps as early as the fifth century and 
printed at Venice in 1836. In using this book great care must 
be exercised, for the Armenian translator worked with the 
Armenian Vulgate at his elbow, and wherever he can he cites 
the texts according to that. The scholar who depends on the 
Latin version of this commentary also issued at Venice is in a 
worse quandary still, for besides ascribing to Ephrem what is 
merely the reading of the Armenian Vulgate, the citations of 
which should have been put in a distinct type, he will, owing 
to the carelessness of the Latin translator, not even get hold of 
readings of the Armenian Vulgate, and give us instead Jerome's 
Latin Vulgate. 

Secondly, we must depend on any old versions which may 
have been made from the (so far hypothetical) pre-P e shitta 
Syriac text. Such versions are the old Armenian and the old 
Georgian or Colchic versions. We shall see that these are 
based on a form of Syriac text earlier than the P e shitta, or at 
least in many ways differing from it in order to approximate to 
the text used by Ephrem. But it will also be found that these 
two versions as they have come down to us have undergone, and 
in a greater degree, just the same process by which the P e shitta 
was evolved out of the earlier text ; namely, they are recensions 
of more primitive texts according to Greek manuscripts, con- 
sulted afresh. 

We have set forth both our conclusions and the sources by 
comparing which we shall arrive at them. Now let us enter 
into the actual comparison. We shall confine ourselves nearly 
entirely to Romans and First and Second Corinthians ; and will 
have to pick our examples, for a detailed comparison of all our 
sources would fill a volume. 

IV. We will first compare Ephrem's commentary, which we 
shall call E, with the P e shitta, Sch, in order to prove that E , how- 



p^shittA version of the new testament 887 

ever different in many respects, was yet closely akin to Sch. I 
give first the peculiar reading or rendering found in Sch and 
then the passage of E which echoes it. 

Rom. 1 : 2, "called and sent." Comp. f : "Paul, he says, called.' 
That is, because he was called by means of the revelation on the way 
to Damascus, and he was sent by Jesus Christ to preach the gospel." 
1:3,4, " Who was born in the flesh of the seed of the house of David 
and was made known as the Son of God." Comp. E: "because of his 
Son, who just now was manifested in the flesh of the seed of the house 
of David, the very same who was revealed to be the Son of God." 
1 : 4, "Who arose from the dead." Comp. f : "Who arose, he says, 
from the dead — implying that no one else raised him." 1 : 10, "And 
I also pray, that henceforth a door may be opened to me." Comp. E : 
"And I pray that, not after a long time, but even now, there may be 
opened to me a door of a road to see you." 1 : 14, 15, "Because to 
every man I am a debtor that I preach. And so I am eager to preach." 
Comp. f : " But to them in any case, whether they listen or not, I am 
anyhow a debtor to preach. And I am desirous to preach." Read in 
the Greek TrpoOv/xovfuu instead of irpoOv/xov kcu. 

Rom. 3 : 8, "Or surely not as they blaspheme us and say, that we 
say." Here /3Aao-<£>7/W/i.e0a is rendered as if /JAao-^/xovox, the second 
Ka6v><s and rtvts not being expressed. Comp. E : "Sed hoc quasi illud 
quod dicit qui blasphemant nos et dicunt de nobis, faciamus malum," 
etc. 3:21, "But now that not the law — the righteousness of God 
has been manifested. And there testify thereto that law and the 
prophets." Comp. E: "Without the law, he says, the righteousness of 
God has been manifested. Because not from that (law) do we learn 
faith and gentleness, but from the gospel. Nay more — the very law 
testifies concerning our righteousness." 5:6, "But if the Messiah 
because of our weakness in this time in place of the ungodly died." 
Comp. f : "For if Christ because of our weakness in this latter time in 
behalf of the ungodly . . . died." 1 Cor. 14 : 25 Sch renders "and will 
worship God and say." E writes : " He falls prostrate, does homage, 
and says." 16: 15, "But I ask of you, brethren, about the house of 
Stephanas, because ye know that they are the first fruits." Comp. E : 
"But about the house of Stephanas, you. of yourselves know that they 
were to me fruit." 

* The italicized words in citations from Ephrem are identical with the Armenian 
Vulgate. 



888 THE AMERICAN JOURNAL OF THEOLOGY 

These examples are enough, but anyone who takes a P e shitta 
text and the Armenian text of Ephrem can mark hundreds of 
passages where Ephrem even in his Armenian dress reproduces 
the characteristic idioms and turns of the P e shitta. It is certain, 
therefore, that in the main Ephrem's Pauline text was the same 
as the P e shitta. 

V. Next in approaching the peculiar problem which we have 
set ourselves to clear up, we must first prove that the Georgian 
and Armenian versions have a Syriac basis at all. We shall, 
therefore, begin by adducing, first, passages where these two 
versions agree in reproducing Sch, then passages in which one 
or other of them does so separately. This will prove positively 
that these versions are in a general way based on a Syriac 
text. The triple agreement of the Syriac, Georgian, and Arme- 
nian I shall denote by the symbol SchGA. Here are examples 
of it: 

Rom. 1:9, "nam testis mihi est." 1:17, omit 8e before SiWos. 
1:29, "filled with all injustice, fornication, malice, greed of gain" 
(G adds Trovqpia. after Tr\eovc£iq). 2 : 4, ayvoG>v rendered " dost thou not 
know." 2 : 16, " on the day in which." 4:12, rrjs iv dKpo/Wrtigi mo-Tews 
rendered " of the faith of the circumcision ; " so also Ephrem's com- 
mentary. 4:18/1 renders "who in the hopeless hope trusted to 
become," 6 "who in the hopeless hope trusted that he may become;" 
such a rendering must rest on the Syriac use of dlo to express 
"without;" Sch = "et quod non spes, spei confisus est, ut fiat." The 
agreement of G with A is very marked, yet G adheres most closely to 
Sch. 4: 20, Sous is rendered in Sch "and he gave;" GA = "he gave." 
5 : 3, tlBorti is rendered in Sch and A " for we know; " G = " we know," 
excising the "for." 5:11, Sch = "And not thus only, but also we 
shall glory;" G A ="(And [/!]) not so much only, but also we glory." 
7:16, o-vv^/u is rendered in SchGA "I testify of the law;" this 
agreement is against f, which = " Sed si quod non volo et odi, hoc 
facio, dico de lege qua? impedit me ab illis, quia bene est." 11:21 
is rendered in Sch "Perhaps he will also not spare you;" GA = "Ne 
forte nee tibi parcat," which exactly renders Sch. Here Iren int has 
"Ne forte nee tibi parcat." 11:31, omit the second vvv. 12 : 19, Sch 
= "And be ye not exacting yourselves ; " G A = " Do not ye of your- 
selves seek vengeance." 13 : 11, "is nearer to us," as if ■tuilv stood for 



peshittA version of the new TESTAMENT 889 

fjfiuiv. 14 : 5, omit yap. 14 : 6, add nal 6 p.r) <ppovS>v rrjv r/p-tpav Kvpiu> ov 
<ppora. 15 : 25, Suucovrjo-at. for Staxovuv ; Sc/) = " that I may minister." 
1 Cor. 1 : 28, add ko! before ra p.rj ovra. 3 : 5, "For what is Paul or 
what Apollos." 3:13, <t>a.vipbv ytvyo-erax is expressed as one word, 
"revelabitur." 4 : 6, add <j>poveiv after yiypairTai. 4 : 8, "Ye are already 
satisfied and already (om. Sch) have become rich and (om. A) without 
us." 4 : 17, "in all the churches." 5 : 4, row Kvpiov rfp.uiv 'Irjcrov Xpixrrov. 
7:7, 0t\<o yap. 7:15, 17/tas for vp.as. 8:11, «at dToXXvrai. 9:22, add 
<i>s before a.<r6tvq<s. 10: II, ravra iravra for ravTa Sc (Sch = ravra Se 
iravTa). 10:23, a ^d /*oi after «?£eoTt (twice). 11:24, after eiTrcv add 
Xa/3«T£ «ai <f>ayeTt. n : 24, add kXoS/xcvov. 1 1 : 31, ci yap. 12 : 9, «v t<3 
avT<o for tv r<p IvC. 12 : 19, "where zc^rif the body." 12 : 26, Sofa^trai 
tv p.(Xoi. 13 : 11, "^«/ when I became a man." 13 : 12, add oSs before 
&' to-dn-Tpov. 14 : 10, 5c/) = " Ecce enim genera linguarum multa sunt 
in mundo," omitting d tvxoi. The same omission is in A, which = 
" nam tot genera linguarum sunt in mundo ; " and so also G, only omit- 
ting nam. The old Latin codd., f. vg., have tarn multa and Ambrst. 
nam multa, but Sch G A are alone in omitting « tvxoi. 14:37, Kvpiov 
cloiv ivToXai. Here £ cites the words of Arm. Vulgate. 15: 18, Sch = 
"And doubtless also those that slept in Christ are lost;" GA = " So 
then those that slept in Christ are lost indeed." 15 : 20, before ajrdpxrj 
add "et factus est." 15:21, Sch G A concur in ignoring the word 
iirtiBrj ; Sch = " And as by means of man there was death ; " G A = 
"For through man was death." 15 : 37, tl tvxoi is omitted, except per- 
haps in A; Sch = " Thou sowest not the body that is to be, but the 
grain naked of wheat or barley or of other that is sown ; " A = " Not 
the same body which may be to be born, sowest thou ; but grain naked, 
if it be of wheat or if of other things sown (or sowable things) ; " G — 
"Not that body which is to be, dost thou sow; but a naked grain, 
whether of wheat, or whether other which is sown." Here A G are 
obviously based on Sch. 15:47, after Stvrepos S.v6pumo% add xvpios. 
16:19, "Priscilla." 16:23, T0 ^ Kv pt° v ^p-w 'Ir)(rov Xpurrov. 2 Cor. 
1 : 11, Sch = "by the aid of your prayer for us;" G = "by the joint 
aid of your prayers for us;" >f = "by the aid of your prayer for us." 
An attempt has been made to make G more faithful to 1 the Greek text 
by adding a preposition thsana = o-vv before the word "aid." 2:13, 
tio p.r) cvpetv is rendered "quod non inueni." 3 : 3, "but on the fleshly 
tablets of the heart," xapSias. 5:10, Sc/j="that each (lit. man by 
man he) be repaid in his body whatsoever he has done in it, whether 
good or whether bad;" A has the same rendering, only substituting 



890 THE AMERICAN JOURNAL OF THEOLOGY 

for the words "in it" the word "previously" or "formerly;" G also 
renders the Syriac faithfully in all respects, except that it omits the 
words "in it" altogether. Tertullian, res cam.' 3 , renders similarly: 
"uti unusquisque reportet per corpus secundum quae gessit." 5:17, 
add iravTa after yiyovtv. 

A more extended comparison would yield the same results, 
but the above examples sufficiently demonstrate the Syriac basis 
of A and G. Many of the points of agreement with Sch are, it is 
true, found in one Greek manuscript or another ; but as they 
usually occur in a grouping of Greek manuscripts and fathers 
(tf c DEFGKL Euthal. cod. Theodoret Theophylact.) which Sch 
mostly follows, this fact does not diminish the probability that 
their presence in A and G is due to Syriac influence, when this is 
once conclusively established by such a passage as 1 Cor. 15 : 37. 

VI. So far we have given only examples of the concurrence 
of all three sources, Sch, A, and G. We next will instance cases 
where G and A separately agree with Sch. G = Sch in the follow- 
ing places : 

Rom. 1 : 3, tov ytvvo>tievov. 1 : 23, tpireriov rendered "reptiles of the 
earth." 2 : 20, ?x ovTa is rendered "et est tibi." 3 : 22, "Sed iustitia dei 
per fidem Iesu Christi ad omnes et super omnes credentes." The addi- 
tion im. TtdvTas is in the group of manuscripts mentioned above. 4 : 5, 
vKTTtvovri 8t k.t.X. is rendered in Sch "sed credit tantummodo in eum 
qui iustificat." Comp. 6: "sed credit tantummodo in iustificantem." 
4:15, Ta/x£/3a<ris rendered "transgression of the law." 6:19, yap 
omitted after uxrircp. 7 : 3, after avrip add avrjjs. 8 : 36, "and we are 
accounted." 9: 22, -qvtyKtv iv ■7roXXj7 p.aKpoOvp.ia ctkivij opyfjs ; G = "qui 
toleravit in multa longanimitate super vasa irse;" so Sch, "brought 
wrath upon the vessels of wrath," which is repeated in Ephrem's com- 
ment : "longanimitate magna tulit ille iram super uasa irae." The 
Armenian reviser struck out the preposition which answered to super, 
yet ungrammatically left vasa in the case governed by that preposition. 
The Greek codices FG add «s, and some old Latin sources read " in 
uasa" or "in uasis." 11:2, "when he had complained to God about 
Israel and said." 11:6, add « 8c t£ ipyw k.t.X. 11:9, after rpdVcfa 
auiw add ivumiov aiiw. According to Tischendorf this addition is 
only found in the following sources : 4 k scr syr sch ar e seth. vg si * et c dd panc 
Thdrt. Pelag. 11:15, after Tpoo-X^i/rt? add eorum. 11 : 22, omit the 



peshittA version of the new testament 891 

second Ocov after xpy°~ T ° T ys- J J : 2 3> T V amaria, " in M«> want of faith." 
13:1, after ovcrai add i^ovo-lat. 14 : 13, omit ovv- 15:12 is rendered in 
Sch "et is qui surget, erit princeps gentium, et in eum sperabunt 
gentes;" G = "and he who is to rise up [is] to be prince of Gen- 
tiles, and in him will Gentiles hope." The reviser has struck out the 
second is, which the structure of the sentence, based as it is on the 
Syriac, requires. 15:14, ko! Trtir\.r)p<i> pivot. . . . teal Swa/xtvot ko.1 aXkova 

VOvOtTUV. 

1 Cor. 3 : 4, add Ae'yei after Irtpos St and for avOpvmoi read o-apKiKoi. 
3 : 5, for ti ovv read ti yap. 4:17, omit 'I170-0S. 6:10, order : ovrt irAe- 
ovtKTai ovtc KXiirrai. 6 : 17, h> irvevpA ianv is rendered in Sch "est cum 
illo unus spiritus," which probably underlies 6 : "unus spiritus est cum 
domino." 7 : 3, ttjv 6<f>ttXr)v is rendered in Sch "the kindness which is 
due," in 6 "the due of respect." Perhaps, however, these renderings 
are based on different Greek originals, Sch on the reading ttjv 6<f>tiXo- 
p.tvtjv tivotav and G on t^/i' 6<t>u\op.ivrjv Tip.TJv, read only in a citation of 
this verse by Chrysostom. 7:7, 0i\o> yap. 7:16, Sch renders ■>? ti 
0180s, avep, tl thus: "or thou man, dost thou know if;" G= "or thou, 
what dost thou know, O man, if." The addition of "thou" seems due 
to the Syriac. 7:17, before tKaorov add ko.1. The only codices which 
have this reading are F G and the old Latin fg. 9: 15, Se/? = "neque 
propter hoc scripsi;" comp. G: "neque propter hoc scripsi haec," 
where hac must have been inserted by the reviser. 9 : 20, omit p.rj &>v 
avros vtto vop.ov. In his commentary Ephrem ignores these words and 
probably omitted them. 9 : 22, Trdvras or rovs 7ravras instead of iravnos 
nvots. 10:19, order: on tlS<okov . . . on tl$<o\.6$vTov. 12:31, koO' 
virtpfio\T]v o&ov is rendered "an excelling path." 14:10, to <k jxtpovs 
KaTapyrjOrjo-tTai is rendered in Sch "then will vanish whatever is little;" 
comp. G : "tunc paruulum quidque euanescet." 2 Cor. 2:4, "but 
that ye may know the love exceeding which I have towards you." 

Many of these readings of G are inexplicable save as render- 
ings of Sch ; and we certainly ought to ascribe to the same influ- 
ence the residuum of which Greek codices provide equivalents. 

VII. Professor Armitage Robinson, in his volume entitled 
Euthaliatia (Cambridge, 1895), nas pointed out that an old 
Syriac version ultimately underlies the Armenian New Testa- 
ment. The following are examples of the agreement of Sch and 
A ; I include cases in which the point of agreement is also 
reflected in G or in E or in both : 



892 THE AMERICAN JOURNAL OF THEOLOGY 

Rom. 4 : 9, 17 «ri for rj nal iirl ; here 6 adds k<u. 4 : 1 1 is an instruct- 
ive verse for comparison. Sch — " Signum ille enim recepit circum- 
cisionem et sigillum iustitiae fidei— suae quae in prasputio. ut fiat 
pater apud omnes eos qui credunt ex circumcisione. ut imputetur 
etiam ad eos pro iustitia." Here Sch has these peculiarities: (1) it 
renders *hh as if yap; (2) TrtpiTop.fjs as irtptTOfirjv, with a few Greek 
codices; (3) adds nal before o-<f>payl8a; (4) renders di to that as ut 
fiat; (5) Si' dxpo/Jiiorias as "ex circumcisione," so that the sense becomes 
"the father of all them of the circumcision who believe;" (6) ch t6 
koyio-Oijvai is rendered "ut imputetur;" (7) adds *<u before outois, with 
many Greek codices ; (8) adds «is before Sikcuoowjjv, with a few Greek 
codices and old Latin texts. Of these peculiarities the Armenian exhib- 
its (2), (4) ; (5) it improves upon, for it renders "pater omnium creden- 
tium qui ex circumcisione sint ; " it also exhibits (7). The Georgian also 
has (2), (4), and (5), for it = "credentium circumcisorum ;" also it has 
(6), (7), and (8). We complete the comparison by adducing here 
Ephrem's comment on this verse: "Quasi signum, 3 ait, fecit circum- 
cisionem in suffragium iustitiae qua ex prczputio. id est iustitiam quam 
assecutus per praeputium, fieri eum pater omnium credentium ex prsepu- 
tio, eo ut imputetur eis, veluti et illi, fides sua ad iustitiam. ut appel- 
letur is pater circumcisionis eorum qui confirmati sunt in uestigiis 
fidei praeputii Abrahami." Therefore Ephrem in his text of the Paul- 
ine letters had (2), (5), (6), (7), (8). Perhaps after mo-revovrwv there 
originally stood tu>v in the Greek text. It is curious that tk to that, is 
simply rendered fieri in the commentary. The Georgian is nearest to 
the Syriac text of the Sch which Ephrem read. 

Rom. 4:12, dXXa Kal toTs o-toi)(ovo~iv rots i^veo-iv t»}s iv dxpo^vcrTia mo-Tito's 
. . ■ Sch here = "sed etiam illis qui complent uestigia fidei circumcisio- 
nis patris nostri Abrahami," as if ■n-Co-reto's aKpo/Svo-Ttas stood in the Greek. 
A = "sed etiam quorum (= eorum qui) incedant in uestigiis (= along 
the tracks) circumcisionis fidei." So also G, except that it renders rots 
fyvto-iv in the sense "super uestigia." Ephrem had the P e shitta text, 
for he comments, as we saw: "eo ut appelletur ille pater circumcisionis 
eorum qui confirmati sunt in uestigiis fidei circumcisionis Abrahami." 
Thus the Armenian and Georgian translated the Syriac originally, and 
a later revision simply altered the order of the words to suit the Greek. 

Rom. 4:14. The Greek is : el yap oi ck vofiov Kkrjpovofioi, K£K«Va>Tai 
ij irioTis Kal KarrjpyqTaj. 17 iirayytXia. A and Sch = " Si enim qui ex lege 

'The Italics represent verbal agreements of the version of Ephrem with Armenian 
Vulgate. 



peshittA version of the new testament 893 

erant hseredes, inanis erat fides et nulla (erant [Sc/?]) promissa." It 
can be no accident that both versions add erant after vop.ov and resolve 
the passive perfects in the same way. The Georgian translates identi- 
cally with Sch. Ephrem also read the verse as in Sch, for he com- 
ments : "Quod si a lege erat hsereditas, fides uana fiebat et promissa 
antiquata." 4:18, Kara to dprjp.tvov ; Sch = " sicut scriptum ; " comp. A : 
"sicut dictum est." Here G faithfully renders the Greek. 

1 Cor. 13:3, Sch and A = "If I should feed out to the poor (or 
destitute) all I possess." The words italicized are added in no other 
source. Ephrem's commentary quotes the Armenian Vulgate. 1 Cor. 
11 : 1 7, Sch = " Hoc autem quod praecipiens sum, non tanquam laudans 
sum vos, quia non ad prius nostrum incedistis, sed in malitiam descen- 
distis;" d = "Sed hoc prascipio non tanquam laudarem, quia non in 
melius sed in malitiam nisi estis." The addition of tanquam and the 
inaccurate rendering of oDve^wfe are certainly due to the Syriac. 
5 : 11, to> Totovro) pr/Sl o-vveo-Qiav ; Sch and A = "with such a one ([ScA] 
with him who is such) not even to eat bread." No other source adds 
" bread." 4:3,^ \mb avOpamiwiyi ij/i«pas ; Sch — " uel ab omni filio homi- 
nis;" 4= "or by any man whatever ; " Ephrem had the reading of 
Sch, for he comments "aut omnino a filio hominis;" G boldly renders 
the Greek. In the same verse the Sch rendering of the first words, t/xoi 
8e cts tXax'fTov Io-tiv, has also influenced A ; for Sch = "mihi autem hoc 
diminutio est mihi," and /l="mihi autem et hoc ignominia est." 
4 : 6, tva [M] els virep tov tvos <pvo-Lov<r6e Kara tov cripov. A = "ne {lit. ut 
non) homo praeter socium gloriemini super socium" (or de socio, 
"about a comrade"). This rendering is clearly due to Sch : "et homo 
praeter (or propter) socium ne glorietur propter hominem." 3 : 4 well 
illustrates the mutual relations of the three versions. Sch renders : 
"Quando dixit enim homo ex vobis, Ego Pauli sum, et alter dixit, Ego 
Apollos sum. Non ecce carnales estis ; " A = " Quando dicit aliquis ex 
vobis, Ego Pauli sum, et alter, quod Ego Apollos, non homines estis;" 
G = " Quando enim aliquis dixit, quod Ego Pauli sum, et alter dixit, 
Ego Apollos sum. Non carnales estis?" Thus A takes ex vobis from 
Sch, and G takes from the same the second dixit, the second sum, and 
also carnales for avOpumoi. 2 Cor. 11:2, r)pixoo~ap.r)v . . . irapdivov ayvrjv. 
Sch and A = "I have espoused you to a husband as a chaste virgin." 
11:9, Sch = "I was burdensome to none of you." Here the addition 
"of you" occurs also in A and G, but in no other texts. 13:2, <ccu npo- 
Xt'yco is rendered by Sch and A "and again I say to you beforehand;" 
G also makes the addition : "et hie etiam prsedico." In the same verse 



894 THE AMERICAN JOURNAL OF THEOLOGY 

after dnutv vvv all three versions add ypd<f>u>. Gal. i : 8, la-Toprjaai Krj<l>av 
Set), A, and G render "to see Cephas." 

Here are a few more instances from the Hebrews in which 
the P'shitta rendering may be recognized behind the Armenian 
text : 

Heb. i : i, "God anciently conversed with our fathers." i : 3, xnro- 
<rTocr«o)s rendered Being. 1:7," Who made his angels a wind ?" 1:8, 
"But of the son he said;" so also G. 2: 14, "that by his death he 
might bring to naught him who held." 3 : 2, mo-rbv ovra rendered "who 
was faithful;" so <?. 3: 7, 810, (caflus A.ey« to irvtS/ia rendered in Sch 
"Because the Holy Spirit said;" comp. A : "Wherefore thus saith the 
Holy Spirit." 4:1, Soxrj tis is rendered "anyone be found" in Sch and 
A. 4:7, "Again he established another day." 4: 16, "to the throne of 
his grace." 5 : 2, Sch renders jjutrpiOKaOuv Zwdpavo's tois ayvoovviv thus : 
"And he can humble himself and sympathize with the ignorant;" 
comp. A : "In a measure a sharer in suffering (or sympathizer) is he 
able to be." 5 : 13, 6 /x«Te'x<»»' ydXaKTos is rendered in Sch and A "whose 
food is milk." 6:1, tVl rrfv TeAeidnp-a, "to the completion." 6: 10, 
"in that ye have ministered and do minister." 6:12, Iva. ny viodpol 
yevrjo-dtSch renders "And that ye faint not;" comp. A : "That ye be not 
faint in mind." 

VIII. Here let us recapitulate our results so far ascertained. 
They are these : 

1. The Syriac P'shitta is constantly recognizable under the 
Armenian text of Ephrem's commentary on the Paulines. 

2. The same P e shitta text is — not to the same extent indeed, 
yet often and unmistakably — recognizable under the Armenian 
and Georgian versions, sometimes of both together, sometimes 
of them separately. 

3. Therefore the Armenian and Georgian versions have a 
Syriac basis, and these versions, as we have them, are recensions 
of the more primitive versions from Syriac, made with the help 
of Greek manuscripts. Fortunately for our argument, the recen- 
sion was not so thorough as to efface all traces of the Syriac 
basis. The revision of the Armenian was made about A. D. 400 
by S. Mesrop and S. Sahak, of the Georgian probably about the 
same time, but perhaps later. 



p e shittA version of the new testament 895 

IX. Now we must turn to the main problem which we set out 
to solve. Was the Syriac text used by Ephrem in all respects 
identical with the P e shitta as it stands today? or was it an 
earlier form of Syriac of which the P e shitta is a development ? 

If we can detect in Ephrem and Georgian, or in Ephrem and 
Armenian, or in Ephrem and both these versions at once, pecul- 
iar readings which are absent from the P e shitta, then such pecul- 
iarities must once have stood in the Syriac text, but were revised 
out of it in some stage of the development of that text subse- 
quent to what is preserved to us in these three sources. 

X. We select some examples of the agreement of Armenian 
and Georgian, together or separately, with Ephrem against Soh : 

Rom. 6:18, Sch ="And when ye were freed from sin ye became 
slaves;" but <? = "But now ye are freed from sin and are enslaved." 
Ephrem comments: "But now as ye are liberated from sin and have 
passed under the yoke of righteousness." Comp. JEth..: "Et nunc 
uero . . ." 8:14, ayovrai is rendered literally in Sch and A "they 
who are led;" but G renders "ambulant." Ephrem had such a render- 
ing in his Syriac, for he comments : "For whoever are led by the Spirit 
of God,* that is, who walk in spiritual paths." 8 : 22, the yap after 018a- 
ptv is omitted in 6 as in y£th. Ephrem also omitted it, for he com- 
ments thus: "We know, he says, that all creation . . ." The omis- 
sion is the more marked in the commentary because A like Sch adds 
yap. 8 : 28, tois ayaTraxnv tov Otov irdvra owepyei ets dya0oi/. G renders : 
" For the lovers of God everything succeeds unto good." So Ephrem 
comments: "If anyone love God, everything succeeds for them unto 
good." Here y1 = "Iis qui amant deum, in omni cooperans est in 
bona," and Sch = " He aideth them in all things for good who love 
God." 11 : 19, epeis ovv i^eKkdo-Orjaav kAciSoi. Here A and Soh render 
literally: "And perhaps you will say, The branches were plucked off." 
But <? = " Dices? Si rami fracti sunt;" which is the way Ephrem read 
the passage, for he comments thus: "Forsitan dicas, Si rami primi 
fracti sunt." HerefC have « KXao-d-qvav, -and d x fg or int2ai3 have "si 
fracti sunt." It is certain, therefore, that ei has been revised out of the 
P e shitta. 12:3, G and A add tov Oeov after x<V lT °s ; Sch omits, but 
Ephrem's commentary implies it, for it runs: "This, says he, we say 
because of the grace of God." 12:3, fit) xmtp<f>povtiv wap' Bel <j>povtiv, 

* The words italicized are from the Armenian Vulgate. 



896 THE AMERICAN JOURNAL OF THEOLOGY 

aAAa (j>povuv it 1 ; to (Timppovtiv, e(ca<rra> tos 6 0t6s i/j.€pi<riv fiirpov TriVrews. 
G = " ne plus cogites quam fas, sed cogitate ad sanctitatem unusquis- 
que sicut et deus diuidet iuxtra mensuram fidei." Here Sc/7 = "that 
ye be not thinking outside of what is proper that ye think. But that 
ye be thinking with modesty, each one so as God divided to him faith 
by measure." Ephrem comments thus: "Ne plus cogitetis quam quod 
fas est, id est, ne quaerant sibi dona permagna plusquam uim suam . . . 
sed cogitare ad obedientiam sanctitatis, id est, servare mansuetudinem 
sobrietatis, ut unusquisque sicut deus diuisit iuxta mensuram fidei." 
Here note that Ephrem agrees with 6 against -Set) and also against A 
(1) in omitting <ppovtiv, (2) in the rendering sanctitatem, (3) in the ren- 
dering "iuxta mensuram fidei," to which, however, Sch approximates. 
It is clear that the existing P e shitta text has here been remodeled and 
no longer presents what Ephrem and the Georgian read in it. 12:4, 
G and A omit yap after KaOdirip. So did Ephrem, who comments : 
" Ecce est sicut divisa sunt munia membrorum corporis." Sch adds 
■yap, and the addition must be due to a recension later than Ephrem. 
12:5, outws ol ttoWoI tv crw/ia eoyitv. G = " Sic quoque nos omnes unum 
corpus sumus." But Sch and A =" Sic quoque nos qui plures sumus 
unum sumus corpus," according to the Greek. Ephrem read the same 
text as we have in fi, for he comments: "Nam omnes nos unum cor- 
pus sumus in Christo." 14:9, G = " Christus mortuus est et resurrexit 
et uixit." So Ephrem comments: "Christ died and rose, lived." 
But Sch ="died and revived and arose," a transposition of the older 
text. A omits aviarr). 14 : 22, <rv wtariv -qv <X" S KaTa <r<avTov c^«- Here 
fi and A omit 17V. Sch retains it: "Tu quae est in te fides, in te 
ipso." But Ephrem omitted ijv, for his comment is: "But if thou 
shalt say, I eat with faith, if thou hast faith, unto thine own self boast 
therein unto God." I confess, however, that Ephrem suggests a Greek 
text yv ixqs, " if thou hast." 14 : 23, G adds edit, "eats," after «<c fiirrtw. 
Ephrem had the same addition, for he comments: "But he who is 
discriminating, if he eats, he is condemned, for it is not by faith he 
sanctifies and eats." Sch and A do not add edit. 3:15, airos Si o-iaOrf- 
<7«Tai, ovrtos 8t <os 81a 7rvpos. Here fi adds o-w^o-tTai after ovtok Si. 
Ephrem had the same addition in his text, for he comments : "Quam- 
uis enim et resurrectio suscitet et uiuificet eum, nihilominus ipse 
eomodo fit uiuus (i. e., is saved) et miserandus quasi in ignem." Sch 
has simply: "eomodo autem quasi ab igne." 

I Cor. 5:1, oA.<os Slkovctou iv 11p.1v iropvtia, Kal Toiavrr) iropvita ijtis ovSi 
iv tow «6Wiv (ovopa&Tat add 8 C LP Se/J). Sc/» = "In short, there is 



peshittA version of the new testament 897 

heard of among you fornication, and such fornication as not even in 
the house of heathen is heard of." Thus Sch accords with the Greek, 
except in implying addition of aKovcrai rather than of ovo/ta£«T<u. But 
A has oAo)« ovofta^tTai iv v/uv and does not add ovo/ia£tT<u after edvtcnv ; 
and this must have been Ephrem's reading, for he comments : " Ecce 
enim apud uos fornicatio, qualis etiam in gentibus non, facta est inter 
uos, superbientes inter uosmet, ecce apud uos nominatur, filius uxorctn 
patris sui habet." In the above, of the whole sentence nominatur is 
almost the only word in common with the Armenian Vulgate; and is 
not, I think, due thereto. There is no formal citation of that Vulgate. 
The interrelation of G and A is very strikingly illustrated in this verse 
and deserves notice, although it does not affect the point before us, 
which is the character of Ephrem's Syriac text. 6, then, has oAtos dxou- 
(t<u and adds ovo/m^trai after lOvvrtv, in both respects differing from A 
and approximating to Stephanus' text. Yet both A and G interpolate, 
the one after, the other before, iv ifuv, the word "unde," "whence" 
" for what," in Armenian under, in Georgian rasa. That this peculiar 
addition is due to the same Syriac text having been originally trans- 
lated by G and A alike is almost certain. s If it is not so, then we can 
at least infer that G is a rendering of A, or vice versa. In any case it 
conclusively proves that G and A were once a single text and as ver- 
sions flow from one archetype; and that the manifold varieties which 
now divide them must have resulted from the separate revision of each 
made at an early date with the help of Greek manuscripts. 

1 Cor. 5 : 8, aAocpmas is rendered "justice" in G; A and Sch render 
the Greek. Ephrem had "justice" in his text, for he comments: 
"nor in works of evil, but in the leaven of justice, that is, in works of 
justice and of truth." 5:12, after yap /mm G and A add kcu. Sch is 
without it, but Ephrem read it, for he comments : "How is it mine also 
to judge the men of the world." 6 : 1 1, km ravra Tivts ijrt. G = " Et 
hocmodo quondam fuistis," which is reflected in Ephrem's commen- 
tary: "Et uos hocmodo quondam quidam (=th/£s) eratis." Here, 
although the commentary is colored by the Armenian Vulgate, it does 
not derive therefrom this peculiar rendering of Tavra, for A = "Et uos 
tales quidam eratis." Sch, however, = "And these things have been in 
some of you." The reviser of the Syriac was not satisfied with the 
rendering of raCra as = " hocmodo" and corrected it at the expense of 
accuracy in the rest of the clause. 8 : 4, kox oti ouStls dtbs d /«j «Ts. 

5 For rasa = rtvos and under = irpht rl, and these words seem to be independent 
renderings of a Syriac word rather than renderings one of the other. 



898 THE AMERICAN JOURNAL OF THEOLOGY 

G renders: "And that none is God other than the only one God." 
Ephrem had the addition solus deus in his text, for his comment is : 
"Quippe nihil idola sunt in mundo, et omnia quae nomine appellantur 
dei, non est omnino aliud quam unus solus deus." Sch and A lack the 
addition solus deus. It has been revised out of them because it was 
not in the Greek codices. 8 : 7, rrj <rvvr)$tia. Ephrem's text may have 
had this reading, for he comments: "Sunt enim nonnulli simplices 
inter fideles qui exeunt ad manducandum in domo idolorum," trvvrfitia. 
being misunderstood in the sense of "simplicity," by confusion with 
tvrfitia. Whether G read ovvuSyaa. is doubtful, for it = " through 
doubt," which seems to be an attempt to render the Armenian equiva- 
lent for <rvve«8i}<rci, which is also rendered in Sch. 

Here are some additional examples from Romans of the 
agreement of Ephrem with G against both A and Sch : 

Rom. 4: 16. G renders the Greek text 8ul tovto *k n-torttos, tva 
Kara X**P lv > € ' ? T0 £ ' ,/a ' /fcj8aiai» tt/v iirayytkiav irdvTi t<£ airipiuvri as follows : 
"Propter hoc de fide, ut gratia sit firma promissio omni semini." 
This rendering is clearly recognizable in Ephrem's comment: "Prop- 
ter hoc non, ait, ab operibus, sed de fide, ut secundum gratiam firma 
sit promissio omni semini." The P'shitta has recast the verse as fol- 
lows : "Quapropter per fidem, quae est per gratiam, iustificabimur, eo 
ut promissum confirmetur toti semini." A exactly renders the Greek, 
save that it adds "sit" after \apiv, which may represent the 1? added 
after iva in codd. A 45,80 . It is noticeable that the old Latin codd. de 
vg Ambrst exactly reproduce the reading of G and E : " ut secundum 
gratiam firma sit promissio." It looks as if these sources were influ- 
enced by the old Syriac, or the Syriac by them.' 4: 19, G adds ov 
before Kartvorjatv, which Ephrem also had in his text, for he comments : 
"Et hoc quod dicit : Nullo modo considerauit carnem suam mortam." 
Sch and A oinit ov. 5:1. Socouuflwres ovv (k martm dpqvrpr e^to/iev irpos 
tov Otov is rendered: "So then we have been justified . . . and peace 
was unto us towards God." Ephrem reflects this : " So then we were 
justified by faith of the baptism, and there was unto us peace towards 
God." A exactly renders the Greek, and the P'shitta has remodeled 
the verse thus: "Because we are justified then by faith, there shall be 
unto us peace towards God." 5:13, ^XP 1 1°-f> "6iwv is rendered in G 
thus: "Usque ad legis adventum." Perhaps the addition "adventum" 
is to be detected under the comment of Ephrem : "For until the law 
of Moses, whose commands were multiplied, sin ..." 5 : 14, tov ficX- 



peshittA version of the new testament 899 

Xovtos is rendered in G : "of the coming times." So in Ephrem's 
comment : " For he himself is a type of the coming times" {futurorum 
temporum). Sch has recast it thus: "He that is type of him that must 
be." A = futurorum simply. 

XI. I add a few more from the epistles to the Corinthians: 

1 Cor. 10: 25, ftantWia is rendered in G and f "in a shop or tav- 
ern," in Sch and A "in a fleshmarket or slaughterhouse." 10:28, G 
and £ add the words tov yap Kvpiov ij yrj Kal to irXrjpiapa avrrjs. They 
are lacking in Sch and A, and must have been struck out of these ver- 
sions by some reviser, n : 29, 6 yap iadiiov in G and E is rendered 
"But he who eats;" Sch and A render yap. 15:41, axrrqp yap k.t.X.; 
A and Sch="et Stella quam Stella major est in gloria;" but G = 
"For even as one star excelleth star in glory;" and Ephrem had the 
addition "even as," since he comments : "As star excelleth star in its 
light, so heavenly beings excel earthly in the resurrection of the dead." 
1 5 '• 54> T ° Tt ytvyo-trai. 6 Xdyos. So Sch and A. But G renders ytvyo-erai 
as "implebitur," which Ephrem had, for he comments: "There shall 
be completed in him the statement which was written about him." 

2 Cor. 1 : 6. For purposes of comparison we break up Tischen- 

dorf's text into clauses: (i) a« 8c 6\ij36p.Sa virtp riys vp.£>v TrapaKkrjo-tw 
Kal <ra>Tj/pias (-j- 6h./$6p.t0a add Sch); (2) art Trapaica\ovp.e6a, virip rrjs vp.!ov 
TrapaKXyo-toi'i (-(- Kal o-wTTjptas add G, Sch, and A); (3) t>}s tVtpyov/utVijs iv 
virofiovr) Ttov avrcov TraQ-qpArwv <ov Kal ijp.ets ira(T\op.iv ; (4) Kal r/ iXirls rjp.S>v 
/3e/Wa virip vp.G>v. The above is the order in which A and Sch take the 
whole verse, both adding the words koI o-uiTiypias in clause (2). But G 
takes the clauses in this order: (1), (3), (4), (2), and does not read xot 
<T<aTqpla<i a second time. 

We give Ephrem's commentary on vss. 4-6 inclusive, separating 
and distinguishing with corresponding numerals in the margin those 
clauses which correspond to the four clauses of vs. 6 : 
Vs. 4. Is, ait, qui consolatus est nos in omnibus tribulationibus 

nostris . . . 

ut et nos ceteros contribulatos consolemur per uerbum quod 

audiunt a nobis, 



(3) 



\ et per patientiam tribulationis quam aspiciunt in nobis (= iv 
j xnrofiovfj tS>v avriov iraOrip.aru)v . . . jroV^o/ttv). 

supplicationibus illis, id est propter supplicationes quibus nos 

supplicati sumus a deo pro uobis 
(= 816. Trj 1 ; irapaKXrjatw: »js irapaKaXovp^Oa avrol vno [lege o7r6] tov 0tov). 



(3) j 



900 THE AMERICAN JOURNAL OF THEOLOGY 

(3) ut sustinere ualeretis uos sicut nos. 
Vs. 5. Quoniam sicut abundarunt, etc. 
Vs. 6. 

(1) Si enim tribulamur pro uestra consolatione et salute tribulamur, 
(?3) id est ut uideatis et imitatores sitis nostrum, 

et adsit uobis uirtus tolerandi passionum easdem passiones, 
quas et uos . . . passi sumus. 

(4) Et haec spes nostra, quae pro uobis erat, firma fuit. 
Vs. 7. Scimus enim, etc. 

In the above commentary, whereas Ephrem harps already twice 
upon clause (3) in dealing with vs. 5, he does not comment at all on 
clause (2); for the words " id est ut uideatis et imitatores sitis nostrum " 
cannot be regarded as a comment on it, but are rather one more echo 
of (3). But the very clause (2), which Ephrem thus ignores, is in G 
misplaced. We should expect him to comment on it after (4); and I 
suspect it was so placed in the Syriac original of his commentary, and 
that the Armenian translator omitted his comment because, as judged 
by the Armenian Vulgate, it was out of place. 

2 Cor. 3 : 9, ei yo/D tjj Sicucon'a rr\% KaTaicptVecos 8o|a. So the P'shitta. 
But G reads rj StaKowa . . . 8o£a, and A y Staxona . . . Iv Sofg. Ephrem 
had the reading _of A, for he comments: "Nam si ministratio damnp- 
tionis, scilicet quia exigebat ilia debita, in gloria fuit." Here A alone 
preserves the ancient Syriac text in full ; G half preserves it, but has 
adjusted itself to the nominative Sofa; the P e shitta has been remodeled 
throughout to suit a Greek text such as Tischendorf here adopts. 5 : 4, 
iv t<o a-Krjvei. is rendered in G "in hoc corpore." But Sch — "in this 
house;" and /I— "in this enclosure." Ephrem had the Georgian 
expression, for he writes : "For we who are in this enclosure? that is, in 
this body, lament." 

XII. It is more difficult to detect Ephremic readings in the 
Armenian Vulgate, because the Armenian translator of Ephrem 
gives all biblical citations, where he can, in the very words of 
that Vulgate. In such cases, if the Georgian adds its suffrage, 
we are practically sure that we have recovered the earliest form 
of the Syriac text. For all the embarrassment due to the above 
cause, we can be sure in not a few cases that Ephrem's peculiar 
readings underlie the Armenian Vulgate. For example : 

*The italicized words are from the Armenian Vulgate. 



peshittA version of the new testament 901 

1 Cor. 15 : 17, In core iv reus a/iapruus v/xwv is rendered in A thus : 
"et adhuc in iisdem peccatis remanebitis." Ephrem's commentary 
implies in iisdem, yet without a formal citation of the Armenian Vul- 
gate being dragged in by the translator, fi and Sch reflect the Greek 
literally. 

XIII. But, apart from the commentary of Ephrem, there are 
many archaic readings in the Armenian and Georgian versions 
which though reflected in the old Latin versions are yet absent 
from the P e shitta. Whence did the Armenian derive these ? Cer- 
tainly they are a legacy from the older form of Syriac on which 
it is based and not the fruit of later revision. Here are exam- 
ples: 

I Cor. 12 : 27, fyiets 8c i<TT€ (rutpa Xpio-rov /cat fieXr] Ik /lipovs. Here 
the Greek codex D has /icAij Ik /te'Aovs and the old Latin d e f vg ; also 
Syr p , Origen, Eusebius, etc., have the same reading. But G and A = fie\rj 
iic pekwv avrov. That such a reading existed in Greek is probable from 
Severian's comment cat' 43 ovk tJwtv fii\r) «/c p.cku>v, dAAa /u'Aij iro\Xa e/c 
/teAovs «Vos - /a«\os yap 1; Kt<j>dXr) rov oAov ctoj/xotos, and is certain from the 
version "ex membris " given in Amb' 1 " 3 I5etI? . The A and G must have 
derived their reading from the older form of the Syriac, for Sch reads : 
"and members in your place;" a reading which suggests the botching 
hand of a reviser. Ephrem does not comment on the verse. 12 : 28, 
after yivyj yAwo-ow A adds "interpretationes linguarum," an addition 
akin to " interpretationes sermonum " found in vg° le hart" Syr p , Ambrst, 
but in no Greek codex. So in Col. 2 : 1 G adds : tS>v iv 'IepairdAei. 
G must have derived this addition from the more ancient Syriac text of 
the Paulines which he used. From the Pshittd as also from A a 
reviser's hand has effaced it. Ephrem does not comment on this verse. 

We add in a summary way a few more such readings which 
A and G must have derived from the Syriac, but which the 
P e shitta has lost. They are all from passages on which Ephrem 
has no comment by which to ascertain what he read. 

Rom. 1:18, airoKakvirrtrax = " is to &r revealed." 2:1, 6 Kpivurv 
(aft. irpa<T<rmv)=" wherein (or wherewith) thou judgest." 2 : 10, omit 8e 
after 86£a. 3 : 19= "to those under the law." 1 Cor. 1 1 : 34, SiaTa£o/*ai= 
"then will I direct." 12 : 12, omit yap after Kadairtp. 13:1= "factus 
sum ego veluti aes, quod sonat." Here f vg have factus sum velut; 



902 THE AMERICAN JOURNAL OF THEOLOGY 

item Aug iop6 /actus sum tanquam. 14:7, o/t<os . . . rendered by G 
"eodem modo inanimata sonitum edunt;" by A : "eo modo quo 
inanimata sonitum darent." 15 : 21, omit iirtiStj. 2 Cor. 3:11, iro\X<S 
paWov translated "How much more?" 

XIV. Most of our examples have been drawn from Romans, 
First Corinthians, and Second Corinthians as far as chap. 6 ; but 
the interrelations of Ephrem, of A G and Sch retain the same 
character throughout the Pauline epistles. We can deduce from 
the above facts the following conclusions : 

1. Ephrem for his commentary used a Syriac text closely 
similar to but not identical with our P e shitta. For the idioms 
and peculiar translations of the latter confront us on every page 
of the commentary, in spite of its Armenian dress. We perpet- 
ually have a text cited from the Armenian Vulgate, and by way 
of comment there follows a literal Armenian version of the P e shitta 
form of the same verse ; or sometimes the two versions are 
blended together in one whole. 

2. In a vast number of passages in the G and A, both 
together or separately, the peculiar idioms and turns of Sch 
are recognizable. 

3. In a great number of passages, however, G and A, both 
together or separately, agree with Ephrem against Sch. In such 
cases old Latin codices and Fathers constantly add their suffrage 
to the group Ephrem G A. 

4. In some passages on which Ephrem does not comment, 
A and G, both together or separately, show against Sch charac- 
teristic variants — often supported by old Latin sources — which 
must belong to their original Syriac basis. 

5. Ephrem himself without G and A exhibits many archaic 
readings not to be found in Sch ; and Professor Theod. Zahn, in 
his review of Moesinger's Latin translation of Ephrem's com 
mentary, indicates many such. For them we refer our readers to 
his articles in the Theologisches Literaturblatt for 1893, Oct. 6, 13, 
and 29. 

6. These last three considerations (viz., 3, 4, and 5) prove that 
the Syriac version used by Ephrem and by the first translators 
of G and A was a more primitive form of Syriac than our P e shitta. 



PESHITTA VERSION OF THE NEW TESTAMENT 903 

The latter is a recension of that primitive form. The reviser's 
aim was to adjust the Syriac text to new Greek codices, and 
he specially eliminated characteristic early readings. But he 
also made numberless small changes in the direction of greater 
literality. 

7. The existing G and A versions are themselves recensions 
of more primitive texts, 7 in which — could we recover them — 
we should have faithful witnesses to the earliest form of the 
Syriac text of the Paulines. Inasmuch as these texts were sep- 
arately revised from fresh Greek codices (the A about 400 and 
6 probably not much later), their agreement is a clue to what 
belonged to their common Syriac basis. 

8. G and A were either originally translated from the same 
Syriac codex, or they have been at a time anterior to the revisers 
of these corrected from one another. The supposition that G 
was originally a rendering of A would explain much, but certain 
other features into which we need not now enter forbid such a 
supposition. 

9. So far as regards the Paulines no weight attaches to two 
pronouncements recently made by well-known scholars, viz., (1) 
that the text of Soh is not one which has been built up by revi- 
sion into what we already have in the codices of the fifth and 
sixth centuries, but is, on the contrary, the second-century ver- 
sion unchanged; and (2) that in the conservative and immobile 
East no recension of oriental versions was possible or probable 
so late as 400. 

On the contrary, the G and A versions were radically revised 
and adjusted to fresh Greek codices just about that date ; and 
that, although it would be difficult to find two churches more 
conservative and immobile than were the Georgian and Arme- 
nian, both originally offshoots of the Syrian church. The revi- 
sion of the P'shitta itself about the same time was less radical ; 
yet there was such a revision as makes it unsafe to assume that 
any particular P e shitta reading is earlier than 400 A. D., unless 

'The first Armenian version of which some earlier Armenian Fathers make men- 
tion in contradistinction to the revision by Mesrop belongs by the authority of the 
same Fathers to the age of Gregory the Illuminator, i. e., 300-325 A. D. 



904 THE AMERICAN JOURNAL OF THEOLOGY 

it be endorsed by E, by 6, or by A, our three witnesses to 
the more primitive text which preceded the P e shitta as we 
have it. 

XV. So far we have dealt only with the Paulines. Is it pos- 
sible that the conclusions proved as touching them will be found 
to apply equally to the gospels. Let us take three chapters at 
random, Luke, chaps. 3 — 5, and, adopting the English Revised 
Version as our standard of comparison, see whether many variants 
therefrom of the A and G do not derive from Syriac texts ; and 
if from them, then from which. Syriac stands for Sch and Syr sin 
where they agree. An asterisk prefixed indicates that the very 
same variant is in the Georgian. I do not confine myself to 
noticing only those variants which appear in Syriac or in Syr 3 ™ 
alone, for I wish to give the reader an idea of the constitution 
of the Armenian text. 

*Luke 3 : 4, /?oa>iTos, "of a cry." *3 : 7, lAeyev ovv rendered "and he 
said ; " so Syriac. 3 : 8, omit ev eavrois ; so Syr sin . 3 : 9, t*8i; rendered 
" Behold ; " so Syriac. 3 : 9, after irav omit ovv with b f f '. 3:11," Let 
him give one;" so Syr sin . *3 : 14, "Satis sint uobis mercedes vestrae;" 
so Syriac. 3:16, "He made answer to all and said;" omitting o 
'lmxwrji ;" so Syr 81 ". *3 : 16, "Of whom I am not able to carry the 
shoes;" so the old Latin a b f f- 1 q Arab 1 " 01 '- 80 , Eus dem4a8 . The latter 
reads o|tos with 6 instead of ucavos. 3:17, o'lOKaOS.pat. rendered simply 
"to cleanse ;" so Syriac. *3 : 18, omit piv oZv; so Syriac. 3 : 20, "And 
he shut up John;" so Syriac. *3:22, "a voice came out of heaven 
which said;" comp. Sch, "and said," which is in G. *3:23, aftei 
Joseph add "the son of Jacob;" so Greek codex D. 3:33, after 
Amminadab add "the of Aram, the of Adm6." 8ch adds "son of 
Ram." G adds "the of Aram," and for "the of Aram" has "the of 
Ioram." 4: 2, y1 = "and he did not eat and did not drink." *4:4, 
"answered him and said;" "by bread alone, but by every word of 
God." *4: 5, add "into a high mountain." 4 : 7, " wilt /a// down and 
worship." *4 : 8, omit koX before airoKpiOus ; so Syr sin . 4 : 10, "he has 
charged his angels." 4 : 11, "that on their hands." *4: 12, omit ko.1 
before amic/nOels ; so Syr 5 '". 4 : 14, "through all the regions of the dis- 
trict." 4 : 16, 17, "on the Sabbath day, and they gave him the book of 
Isaiah the prophet, and he rose up to read; "so Syr"". 4:17, "and 
when he had opened the book he found that place ;" so Syr"". *4 : 18, 



peshittA version of the new testament 905 

" He hath sent me to heal the broken in heart, to proclaim ; " " recov- 
ering of sight," avd/2\eil/iv rendered "to see." *4 : 20, cttroSovs rendered 
"he gave it;" so Syriac. 4: 23, "whatsoever we have heard which 
thou didst;" so Syriac. 4:25, "and there was a great famine;" so 
Syriac. 4 : 33, 34, "with a loud voice and said, Let go." 4 : 35, omit 
kcu before «r£Ti/«7o-fv ; so Syr 51 ". *4 : 35, "the devil threw him down . . . 
and came out ; " so Syr sm . 4 : 36, omit kcu before <rwt\a\ow. *4 : 36, 
<rw in oweA.aA.ovv ignored ; so Syriac. 4 : 40, rjyayov\ %<pepov ; so D. 

4 141, "Devils came out," omitting 8c kcu; Syriac omits kcu. 4:41, 
"And he rebuked them and suffered;" so Syriac. 4:41, "they knew 
him that he was the Christ ;" so Syr^™. 

In chap. 5 we need only give variants which either come in 
Syriac or Syr 8 '" or are otherwise of a salient character. 

*Luke 5 : 5, omit kcu before cWoKpiflas; so Syriac. *5 : 5, yak&xna ren- 
dered " we will let down ; so Syr sm . 5 : 8, " When Simon Peter saw it ; " 
so Syriac. G = "et cum." A omits et. *5 : 10, "capturus eris ad 
uitam;" so Syriac. *5 : 12, omit 8e after IBw; so Syr"". *5 : 12, iv t<S civcu. 
A "on his arriving." So G "when he arrived." *5 : 1 7, "And the 
power of the Lord was in the healing them ;" so Syriac. 5 : 19, A = 
"They ascended onto the housetop and suspended him from the til- 
ings and let (him) down with his bed." fl has "and lifted up the til- 
ing and let down the man with the couch before them." 5:25/) ren- 
ders "took the of his on which he lay;" comp. Syr 5 '", "took up that 
(mdm) on which he lay." G = "and took up the bed on which." 

5 : 30, "murmured about him to his disciples and said." 5 : 31, " in no 
wise are wanted physicians for the healthy, but for the sick;" so Sch. 
Syr $hl is wanting. 5 : 36, "no one layeth from a new garment on an old 
coat;" omitting iirifiXypM. and ox«ras. G omits ox«ras and oWo. 5 : 37, 
after tovs oo-kous add tous iraAcuous ; so D Cop. *5 : 38, add kcu Ap<p6- 

TtpOl Tt)pOVVTai. 

So far we have consulted the Georgian version only where it 
agrees with A. But there remain in 6 some noticeable readings. 
For example : 

Luke 3 : 1 , G omits tjjs ijycfiovtas, and the omission must be con- 
nected with the fact that Syr 5 '" and Sch use wholly different words to 
render -n/s ijyefiovtas. 3 : 4, after irpo^rirov add Aeyovros, which Syr™" 
also has, as well as Sch. 3:7 = "dixit ad venientem ad ipsum multitu- 
dinem ;" so Syriac. 3 : 8, Kapvov a^iov in singular, with D 106, e go Cop. 



906 THE AMERICAN JOURNAL OF THEOLOGY 

3 : 8 = #c<u /jut) 86£t)Te km Xeyrjrt iv lauTots (lit. in animis uestris) with L 
Did mamoh apud Comb auctor 2 - 3 °, and compare F 252 So^o-fc. 3:9, 17817 8e 
"iam enim;" so f ff 2Vld g'- 1 vg. 3:11, omit o 2x«>v before /3/jaifiaTa. 
3 : 12, "There came A> him also publicans." 3:17, "^ will cleanse . . . 
and will gather;" so Syriac. 3:19, i\eyx°pevos : "because he had 
reproved him about Herodias;"so Syr 51 ". Sch = " because he was 
reproved by John." 3:19* after tov dSeXc/xn) auroS add "tiAimrou. 
Tischendorf, ed. oct., remarks that Armenian codices make this addi- 
tion, but it is not in six tenth-century codices which I have collated. 
3 : 23 is rendered : "Et ipse Iesus incoepit circa esse triginta annorum, 
qui putabatur Alius Iosephi, tov Iacobi . . ." 4:2, "and when those 
days were accomplished he hungered." 4:7, omit ovv. 4:8, add 
"Get thee behind me, Satan." 4:11, omit <cat on. 4: 14, after irvcv- 
fujTos add ayiov. 4: 15, 8o£a£d/u.£yos, "and he was glorified;" so Syriac. 
4: 16, Kara to eltoOos airui: " as he was accustomed ; " so Syriac. 4:20, 
"and he closed up the book and gave it . . . and sat down ;" so Syr sm . 
4 : 22, after Aoyots add ovtov. 4 : 23, after o-eavrbv add : "and they said 
to him." Comp. Syr sto : "And the things which ye have heard that I 
have done in Capernaum, ye will say to me, Do here also . . ." The 
words italicized come in no other text and clearly underlie G. 4 : 24, 
"But he himself said to them;" Syr sin adds "to them." 4: 25, ore for 
tk; so Syr*". 4:28, dvpov, "with a spirit of anger." 4:28, iv rfj 
oDvaywyg] praem. "qui erant." 4: 33, "there was in their synagogue;" 
so Syr*" 1 . 4:3s, <t>iyLa>8rfn, "Shut thy mouth;" so Syriac. 4: 35, ds to 
ju.ecrov; add eorum. 4: 36, o-wcA.aA.ouv rendered as if IkoyitpvTo. 4: 39, 
"and immediately the fever left her and she arose;" so Cyr mail67 . 
4 : 41 ="Thou art Christ, the Son of the living God." 4 : 41, omit lm- 
TifiS>v. 5:1, s'vTUTw dxAov k.t.A. rendered: "quum multitudo stiparet;" 
so Syriac. 5 : 1 and 2, irapa. tt)v \.lii.vr)v = "along (or at) the edge of the 
lake;" so Syriac. 5:2, dr' avT&v, "from the ships." 5:3, "and he 
sat and taught from the ship to the crowd;" so Syriac. 5 : 10, after 
ixij <f>ofiov add "Simon." 5:17, "And it was on one day, and Jesus him- 
self was teaching them and there sat there Pharisees." Sch adds 
"Jesus." 5 : 26, (cat orAijcr&jcray <ji6/iov Xiyovrts k.t.X.., "and fear fell 
upon them and they said that we have seen wonderful greatness today." 
For the first clause Syr 5 '" is wanting, but for the latter comp. Syr sin , 
"We have seen wonderful and great things (potius, greatness) today." 

Thus in a space of 180 verses we have no less than twenty- 
seven variants found in both Sch and Syr sin ; and twenty-two 



p^shittA version of the new testament 907 

more found only in Syr sin and not in Sch. Not a few of these 
forty-nine variants are renderings of Syriac idiom and represent 
nothing at all in Greek. In addition we have eight readings of 
D or of the old Latin. 

XVI. I have so far chosen examples from Luke, because I 
have been able to compare the Georgian printed text of that 
gospel chiefly with old manuscripts. Now I give a few examples 
of agreement of G with Syr sin from the early chapters of Mat- 
thew, which I have also been able to compare. This time I pre- 
fix an asterisk where the Armenian has the same text. 

Matt. 1:17, omit ovv. 1 : 20, to yap h airy yewrjOkv ; "Nam id quod 
ex ilia genitum." Here Syr 5 '" and Syr our have mnh, "from her," but 
Sch bh, "in her." 1 : 22, Bta rov irpo<prjrov is rendered "by the mouth of 
the prophet," a rendering of 81a which, though not found here in Syr 5 '", 
yet occurs in other passages, e. g., just below, Matt. 2:15. 2:3, 
ducowras Sk 6 /SatrtXe'us 'HpwSijs. G = " When Herod the king heard this; " 
comp. Syr 5 '", "Now when Herod," etc. 2:16, Kara rov xp°Vov. G = 
"according to (or unto) the similitude of the time as he ascertained." 
This is an exact rendering of Syr 5 " 1 ldmutho dzbno domru. Sch 
omits ldmutho. 2:9, io-rdO-q «rav«> ov tjv. G = "et stetit in loco illo 
in quo erat." So Syr 5 '" : "stood at the place of there where." *2 : 18, 
Oprjvos koL KXa.v6ft.bs kcu o8vp/x6s ; so Syr sm and Syr" 11 . 3:4 = "sed ipsi 
Iohanni indumentum ad induendum eius ex crinibus cameli," an exact 
version of Syr sin . *3:io, omit ovv; so Syr sln . 3:16, Ipxa^tvov «V 
avrdv. fi= "it came and it abode upon him." Syr" 1 has "and it abode 
upon him," omitting the words "it came." The Georgian reviser has 
added these words to render the Greek Ipxopzvov, but without effacing 
the rendering of the old Syriac. Thus G is a conflation of Syr 51 " and 
of the fourth or fifth century Greek. 4:4, "But Jesus answered." 
Here "Jesus" is only added in Syr sin , Syr cur , it mu , and in the Greek D. 
*4:7, omit iraAiv before yeypairrat ; Syr 5 '" also omits. *4 : I o = virayc 
owio-m fjum; so Syr sin and Syr cur . 4:11, omit iBov; so Syr™ 1 . 4: 14, rb 
prflhi Sta 'Hcratov. G omits 81a with Syr 8 " 1 . *4 : 18, -rrapa ttjv daXaaaav. 
G "on the brink of the sea of Galilee;" so Syr sin and also Seh. 5 : 9, 
"quia Mi filii dei appellabuntur;" so Syr sin . *6 : 1, Socdioowijv," your 
almsgiving;" so Syr 5 '" and also Sch. 8 : 23, "and he went up to a ship 
and his disciples followed ; " so Syr sin . 9:13, add ets /terdVotav; so Syr sin . 

Here, then, in the space of the first nine chapters of Matthew 



908 THE AMERICAN JOURNAL OF THEOLOGY 

the Georgian version alone presents seventeen readings which 
are nearly all markedly drawn from Syr sin , and two more found 
in Sch as well ; and I have examined only the first four chapters 
with any care. Wherever we turn it is the same, e. g.: 

Luke 20 : 29, 6 adds irap' ■qfitv with Syr sm . John 4 : 7, "Give me to 
drink water;" so Syr sin , but also Sch. 4: 27, G = "and the while over 
their talking." Comp. Syr 5 '", "and while he was speaking." 

XVII. In concluding let us take at random a chapter of 
Mark and test the Armenian and Georgian for Syriac readings, 
beginning with the Armenian. Where G agrees we asterisk. 
We give first variants found in Syr sin alone. They are these : 

Mark 12 : 14, A = " Et ill! uenientes interrogabant eum subdole et 
aiebant." Here subdole is from Syr sin ; interrogabant is however in fi 
and Sch. A as it stands is in old Latin b and ff, and, except for et 
aiebant, in iq as well. *ia : 17, add dinwcpifois. *i2 : 23, "In the resur- 
rection then when they shall arise." *i2 : 24, " respondit Iesus et dixit." 
*i2 : 32, "quod unus est Deus." 12 : 28 and 29, "What commandment 
is first ? And Jesus said to him, First before all, Hear, Israel," omitting 
iraviw in vs. 28 and adding it in vs. 29. *i2 : 35, omit kcu. before a7ro- 
KpiOw. 12 : 36, add k<u before awos; so D also. 12 : 37, " If therefore 
David himself" and omit km before -rroOev. 12 : 41, "Jesus stood over 
against the treasury." 

The following interpretations in the same chapter of the 
Armenian are common to both Syr sin and Sch : 

*I2:S, "And again he sent another." 12:6, "Perhaps they will 
reverence my son." 12 : 10, "And have ye not read." *i2 : 17, iOav- 
im^ov instead of e£eOa.vfu£ov. 12 : 25, "non viri feminas sumunt, neque 
feminse virorum fiunt." *i2 : 30 and 31, add "This is the first com- 
mandment and the second is like it." 12 : 38 = "seeking after saluta- 
tions." 12 : 40, omit km before irpo<pa<ra. 

Thus in one chapter of forty-four verses we have ten read- 
ings of Syr 8 " 1 alone and eight of Syr sin plus Sch common to A and 
G. At the same time the nature of the revision which A under- 
went in the fifth century is illustrated by the additions and 
changes made to the text in this chapter, e. g.: 

12 : 26, before to>v veKpZv add t^s dvaorao-ews. 12:7, add Ocacra/xevoi 
airbv ipxop-tvov with many cursives and Syr p . But certain other addi- 



peshittA version of the new testament 909 

tions are less likely to have been due to such a revision, e. g. : *i2 : 1, 
"dicens" or "et dicere" after AaAetv, with old Latin c and b. 12 : 14, 
add aire ovv rnxtv with D and many old Latin codices, but 6 omits this. 
12 : 17, "Go, render unto Caesar," an unique reading. *i2 : 20, "There 
were seven brothers with us,", with K", D, and old Latin codices. *i2 : 4, 
dir«rreiAav ^Ti/toyievov. 

The Georgian version of the same chapter presents other 
points of agreement with the Syriac absent from the Armenian, 
e.g.: 

12:2, "at the season of fruit;" Syr sin only. 12:2, omit irapa t£v 
yempyZv ; so Syriac. 12:6, omit avrov Icr^aTov after airloretXtv ; so Syriac. 
But 6 has €<rx aT0V at ^he beginning of the verse with Set), for it renders : 
"nouissime, unus filius erat dilectus suus," where suus is also shared 
by A. 12:7, "his heir;" so Syr sin only. 12 : 14, omit t\ ol; so Syr si ° 
only. 12 : 19, ko.1 KaraXiirg yvvdtKa rendered "and he hath a wife" (/it. 
and there is [to him] a wife) ; so Syr sin only. 12 : 29, "The first of all 
the commandments is this;" so Seh, but also the mass of Greek codices. 
Therefore it is probably due to the reviser of G. 12:32, omit kox 
before «?7T£v ; so Syriac. 12:37, "our Lord ; " so Syriac . 12:43, "And 
Jesus called his disciples and said to them ;" so Syriac. 12 : 43, omit 
tSv ftoXkovTtav ; so Syr 5 '" only, with Greek codices 1, 13, 248, and old 
Latin a, b, c, ff% g', iq. No Armenian uncials have the omission. 

This brings up the number of readings derived from Syr 8 ™ 
in this single chapter in one version or the other to fifteen ; of 
readings common to Syr sin and Sch to fourteen. 

XVIII. We have now adduced enough evidence upon which 
to base conclusions about the relation of the gospels in G and A 
to the Syriac texts. They are the same conclusions at which we 
arrived in regard to the Paulines, only the fuller light of the 
Lewisian codex (Syr 8 ' ) replaces the twilight of Ephrem, half 
obscured in his Armenian dress. 

1. The Georgian and Armenian gospels were both in the 
first instance translated from a Syriac text ; and they have also 
so many peculiarities in common that it is almost necessary to 
suppose that one and the same Syriac manuscript was used by 
both sets of translators. It is even arguable that the primitive 
Armenian was made from the primitive Georgian or vice versa. 



910 THE AMERICAN JOURNAL OF THEOLOGY 

In any case, these two versions flow from one and the same 
archetype. 

2. G and A, now one, now the other, often both, contain read- 
ings found in Syr sin , but not in Sch. But they give next to no 
readings from Sch which are not in 'Syr sin as well. Moreover, 
in a considerable number of cases Syr sin is the only extant source 
which contains the readings of G and A, and these readings are 
often literal renderings of the tricks of translation of Syr sin . It 
follows that a Syriac version of the same type as Syr sin was used 
by the first translators of A and G, and not a P e shitta text. The 
only path from the latter to either A or G lies through Syr sin or 
Syr cur . 

3. The primitive Armenian and Georgian versions were 
revised from Greek codices about 400. Possibly the Georgian 
revision was later. In any case the revisions were made inde- 
pendently, for the original text is seldom supplemented and 
remodeled along the same lines. The result of the revision has 
been to give to both these versions a superficial resemblance to 
the so-called traditional text. 

4. In the Armenian church the revised version utterly effaced 
the more primitive form of Armenian text, so that no manu- 
scripts of the latter survive. The same is probably true of the 
Georgian, but manuscripts in this tongue are so few and difficult 
of access that we must speak with caution. The only old Geor- 
gian manuscripts which I have been able to see closely agree 
with the printed text. It is clear that in both churches manu- 
scripts of the older form of text were rigorously suppressed, and 
they would in any case vanish by not being copied. 

5. The Georgian and Armenian churches were offshoots of 
the Syrian, and must have used for translation that form of 
Syriac text which was accredited and commonly used in Syriac 
churches and by the Syriac missionaries. Since they both used 
a text akin to Syr sin , it follows that that was the accredited text, 
and that the P e shitta text was not yet in vogue; probably 
because — as its contents suggest — it was not yet in existence. 

6. In addition to readings of Syr sin , G and A exhibit, sep- 
arately or in common, a number of characteristic readings only 



peshittA version of the new testament 911 

found in D and in the old Latin versions. These readings are 
most unlikely to be the fruit of revision, especially when they 
come in both G and A; they must therefore be regarded as a 
part of the original stock of those versions ; for they must have 
stood in the Syriac text originally translated. Syr sin therefore 
cannot be regarded as a pure and full exponent of what stood in 
the first Syriac version ; it marks a stage of it when many char- 
acteristic elements, which remain in G and A, had already been 
purged out of it. 

XIX. It is outside the scope of this article to examine G and 
A with reference to the differences which divide the Sinaitic 
Syriac from the Curetonian. It is enough now to remark that G 
and A often follow the Curetonian and set aside Syr sin ; e.g., in 
Luke 9:53 G has "and they did not receive them" which is, 
according to a leading Syriacist, "one of the Curetonian text's 
'peculiar and unsupported readings.'" In the same context, 
Luke 9: 52, A has "entered into one village of the Samaritans,' 
which is again Curetonian ; and just below, Luke 9:58,^ has "And 
the Son of man" instead of "But the Son . . . ," which is also 
peculiar to Cureton's Syriac. On the whole, however, where 
Syr sin and Syr cur differ, G and A incline to Syr sin . But what is 
everywhere most noticeable is their wholesale rejection of 
P e shitta readings in favor either of Syr 5in or of Syr cur . They 
scarcely ever present a P e shitta reading without its being 
endorsed by one or the other of these earlier forms of the Syriac ; 
and the handful of exceptions are explicable from the Greek 
manuscripts which were used for the revision of A and G. 

XX. We return once more to the opinion of a writer whom 
we have already quoted from the pages of the Church Quarterly 
Review, Vol. XL, London, 1895, P- I 3 I : 

It is admitted by all that a Syriac version of the New Testament 
has existed from (perhaps) the second century. The place of this ver- 
sion has been taken by the P e shitta from the earliest times. Its text 
stretches back into the farthest regions of Syriac literature. It is a 
witness to the best form of the Greek text of the New Testament, that 
text which has been preserved in all parts of the Christian church, and 
is more attested by the earliest Greek Fathers than any other. On the 



912 THE AMERICAN JOURNAL OF THEOLOGY 

other hand, the Curetonian-Sinaitic text is a witness to the corrupt 
form of text which prevailed in the west and in Syria, amongst those 
to whom Greek was not a familiar language. It may be an ancient 

witness. 

But is it likely that if the P e shitta text was already in regu- 
lar use from 300 to 400 A. D. the Georgian and Armenian 
churches and the Syrian missionaries who first evangelized those 
nations would not have chosen it for translation rather than "a 
corrupt form of text" ? What significance to attach to the 
writer's words which follow, "amongst those to whom Greek was 
not a familiar language," I do not know. Does he suppose that 
a Syriac version was any the worse because it was made for 
people who talked Syriac and did not know Greek ? His words 
are only true in a sense which he would repudiate, namely, that 
the P e shitta is a recension of the Curetonian-Sinaitic text made 
about 400 A. D. from new Greek manuscripts. The G and A as 
we have them are similar recensions, and ft is entirely to this 
fact that is due the circumstance that either they or the P e shttai 
are witnesses to what this writer is pleased to call "the best 
form of the Greek text of the New Testament." That the Tex- 
tus Receptus is "more attested by the earliest Greek Fathers 
than any other" I gravely doubt; but it is an assertion which 
I gladly leave to others to deal with.