Copyright
Charles Cutler Torrey.

The composition and date of Acts online

. (page 1 of 7)
Online LibraryCharles Cutler TorreyThe composition and date of Acts → online text (page 1 of 7)
Font size
QR-code for this ebook


HARVARD THEOLOGICAL STUDIES

I

THE COMPOSITION AND
DATE OF ACTS



BY



CHARLES CUTLER TORREY

PROFESSOR OF THE SEMITIC LANGUAGES
IN YALE UNIVERSITY




Issued as an extra number of the
HARVARD THEOLOGICAL REVIEW, 1916



CAMBRIDGE
HARVARD UNIVERSITY PRESS

LONDON: HUMPHREY MILFORD

OXFORD UNIVERSITY PRESS

1916





*?^t
4

"' ^)




HARVARD THEOLOGICAL STUDIES



HARVARD
THEOLOGICAL STUDIES

EDITED FOR THE

FACULTY OF DIVINITY

IN

HARVARD UNIVERSITY



BY



GEORGE F. MOORE, JAMES H. ROPES,
KIRSOPP LAKE




CAMBRIDGE
HARVARD UNIVERSITY PRESS

LONDON:. HUMPHREY MILFORD
OXFORD UNIVERSITY PRESS

1916



HARVARD THEOLOGICAL STUDIES

I

THE COMPOSITION AND
DATE OF ACTS

BY

CHARLES CUTLER TORREY



PROFESSOR OF THE SEMITIC LANGUAGES
IN YALE UNIVERSITY




CAMBRIDGE
HARVARD UNIVERSITY PRESS

LONDON: HUMPHREY MILFORD

OXFORD UNIVERSITY PRESS
1916



COPYRIGHT, 1916
HARVARD UNIVERSITY PRESS

Issued as an extra number of the
HARVARD THEOLOGICAL REVIEW, 1916



THE
COMPOSITION AND DATE OF ACTS

CHAPTER I

THE ARAMAIC SOURCE IN ACTS 1-15
i. INTRODUCTORY

THE hypothesis of a Semitic source (or sources) underlying more or
less of the first half of Acts has commended itself to a few scholars.
Thus Harnack, Lukas der Arzt, 1906, p. 84: " Es spricht Wichtiges
dafiir, dass Lukas in der ersten Halfte der Acta eine aramaische
Quelle iibersetzt und benutzt hat, aber schlagend kann die Annahme
nicht widerlegt werden, dass er lediglich auf miindlichen Mitteil-
ungen fusst. Vollends unsicher ist es, welchen Umfang die Quelle
gehabt hat und ob es iiberhaupt eine einzige Quelle gewesen ist."
Similarly in his Apostelgeschichte, 1908, pp. 138, 186. Wendt, Die
Apostelgeschichte, 1913, p. 16, says: " Im Anschluss an Nestle
StKr 1896 S. 102 ff. nimmt [Blass] die Bearbeitung einer ara-
maischen Quelle im ersten Teile der Apostelgeschichte an. Die in
diesem ersten Teile haufiger als im zweiten vorliegenden Aramais-
men werden von ihm als Beweis hierfiir betrachtet (Evang. sec.
Luc., 1897, p. vi, xxi, ss.)." See also Blass' very meager statement
in his Philology of the Gospels (1898), 141, 193 f., 201, of his some-
what hastily conceived theory according to which Luke followed an
Aramaic source in the first twelve chapters of Acts.

But so far as I am aware, no one has ever attempted to point out
specifically Aramaic locutions in Acts. Nor has the search for
Semitisms, of whatever sort, hitherto resulted in any fruitful dis-
covery. A few doubtful examples have been adduced in support of
still more doubtful conclusions; there has been no effort to collect
and examine the material of this nature. Nestle's observations,



2092604



4 THE COMPOSITION AND DATE OF ACTS

referred to above, 1 were concerned only with two variant readings
(2, 47 and 3, 14) in Codex Bezae, and have no bearing whatever on
the question of the original language of this part of Acts, as I hope
to have opportunity to show elsewhere. 2 Wellhausen in his " Noten
zur Apostelgeschichte " (Nachrichten von der K. Gesellsch. der Wiss.
zu Gb'ttingen, 1907, 1-21) takes no notice of Semitisms or of possible
Semitic sources; in his " Kritische Analyse der Apostelgeschichte "
(Abhandlungen der K. Gesellsch. der Wiss. zu Gottingen, 1914, 1-56)
he considers the possibility of translation in only one passage, namely
2, 23 f., and there in a wholly non-committal way. Among English
and American scholars the question of Semitic sources in Acts seems
to have aroused even less interest than among the Germans.
Moffatt, Introduction, 1911, p. 290, says (citing Harnack): " There
is fair ground for conjecturing that Luke used and translated an
Aramaic source "; and Milligan, The New Testament Documents,
1913, p. 163, refers to the hypothesis as a possible one.

Now Aramaic is not an unknown language, and we have consider-
able familiarity with the principles and methods of those who
rendered Semitic documents into Greek at the beginning of the
present era. The question, too, is one of far-reaching importance.
In a writing of the character and extent of the first half of Acts it
would ordinarily be possible to determine whether the Greek is a
translation, and if so, from what language the version was made.
In the present case, by good fortune, the material at hand for the
demonstration is more than usually satisfactory. I am confident
that those who examine the evidence carefully will find it conclusive.

2. THE LANGUAGE OF THE FIRST HALF OF ACTS

The first half of the Book of Acts is concerned primarily with the
church in Jerusalem, viewed as the center from which great evangel-
izing forces went out into the world. The background of the narra-

1 They were first published in English in The Expositor, 1895, pp. 235-239; then,
with the title " Einige Beobachtungen zum Codex Beza," in the Theol. Studien u.
Kritiken, 1. c.

* It should be added that Nestle's own conclusion as to the original language indi-
cated was that it was more likely Hebrew than Aramaic (Expositor, I. c., p. 238) ; see
however his Philologica Sacra, 1896, p. 55, where he refuses to express an opinion.



THE ARAMAIC SOURCE IN ACTS 5

tive is obviously Judean. It is antecedently probable that the
earliest documents of this Jewish Christian community would have
been written in Aramaic, the vernacular. We also have excellent
reason for believing that Luke, 1 the compiler of the two histories,
was one who made special search for Semitic documents, as the
primitive and authentic sources, in order to render them into Greek.
I think I may claim, without undue presumption, that the whole
question of Semitic sources in Acts has entered a new phase since
my argument, in the article " The Translations made from the
Original Aramaic Gospels," contributed to Studies in the History of
Religions Presented to Crawford Howell Toy (New York, Macmillan
Co., 1912, pp. 269-317), that the compiler of the Third Gospel was
an accomplished translator of both Hebrew and Aramaic. 2 We
should therefore surmise, at the outset, that the very noticeable
Semitic coloring of the first part of the book, remarked by all com-
mentators, is simply due to translation.

It is not necessary to argue that the Greek of Acts is not homo-
geneous; it may be well, however, to review here the main facts
touching the question of translation. For the first fifteen chapters,
the language is distinctly translation- Greek; in the remaining chap-
ters, on the contrary, the idiom is not Semitic, and there is no evi-
dence that we are dealing with a version. The whole book, however,
shows unmistakable uniformity of vocabulary and phraseology, so
that it is obvious (to him who recognizes the Semitic source) that
the author of 16-28 was the translator of 1-15. Many have re-
marked that the most strongly " Hebraizing " chapters are those
at the beginning of the book. The reason for this appearance is the
fact that the opening chapters are so largely made up of speeches
composed in high style, along with quotations from the Old Testa-

1 The identification of the author of the Third Gospel and Acts with Luke, the com-
panion of Paul, is not essential to the present argument. I will, however, record here
my opinion that the church tradition is right, and that Luke the compiler was also
the author of the " We-document."

* The article was not reviewed or noticed in print, so far as I am aware, but the
many letters which I received lead me to think that the demonstration was generally
accepted by those who read it. Most of the letters expressly approved the argument
derived from Luke i, 39, in particular, and no one of my correspondents raised objection
to it.



6 THE COMPOSITION AND DATE OF ACTS

ment. The case is exactly parallel to that of the first two chapters
of Luke's Gospel. On the other hand, in such chapters as Acts 13-
15, where the events narrated are comparatively recent and widely
familiar, and the language therefore is that of every-day life, the
rendering sounds somewhat more free. But even in the chapters of
this latter class the translation is found on examination to be truly
close; the Greek idiom never strays far from the Aramaic, while
occasional telltale phrases point to the underlying language. These
indications of a translated Semitic source, it may be added, are
present in every part of the first half of the book. There are no
passages in which the language can be said to make it probable that
Luke is composing his own Greek. It is a striking fact (which will
be considered more fully below) that in the very beginning of the
first chapter the evidence from the material content combines with
that afforded by the language in such a way as to make it plain
that Luke is following a written source so closely, and with such
self-restraint, that he does not even allow himself space for an intro-
ductory sentence of his own. This, again, is altogether characteristic
of the author of the Third Gospel.

Throughout chapters 1-15 we are constantly meeting such
Semitisms as the following: 1 i, i 7;paro iroieiv (Aram.); i, 5
pera TroXXds rauras i^pas (Jewish Aram.); i, 10 /ecu cos (naO
a.TeviovTes rjffav . . . KO.L loov (xrn) K.r.X.; i, 15 eiri TO O.VTO (also 2,
I, 44, 47); 2, 7 ovxi ioov (Aram.); 2, 23 IKOOTOV 5id xtipbs (TS)
dj>6/icov; 2 3, 20 Kcupot a.va.\l/v^fus airo TrpocrcoTrou TOV Kvpiov, 4, 12 r6
dedofj-tvov kv avdp&TTois', 4, 16 ywarbv <rrj(j,lov (Aram.); 4, 30 kv rcjj
rr)v xtipa, eKreivew erf, 5, 4 rl OTI Wov kv r# napdia, <rov, 5, 28 ira.pa.y-
yc\iq, Trapr)yyd\afjitv; 5, 41 CLTTO Trpoff&irov rov ffvvedpiov, 6, 5 nai
6 XcVyos iv&inov TTCWTOS rov Tr\r)dovs', 7, 13 kv TCO Seur^pco



1 I give here only a selection; it would be easy to make the list much longer. I have
designated those idioms which are specifically Aramaic; those which are not thus desig-
nated might be either Aramaic or Hebrew. The Aramaic equivalents not given here
will be found in the sequel. Some of these idioms are to be found occasionally in the
Koine, but no specimen of the Koine" ever showed such an array as this!

1 Cf. Wellhausen, Krilische Analyse, 5 (this is the passage in which he touches the
question of a Semitic source). In the original Aramaic the words were the same as
those in Mark 14, 41, and the rendering should have been els



THE ARAMAIC SOURCE IN ACTS 7

yvupiff6r) rots d5eX$ois CLVTOV; 7, 23 aveftri tiri rrjv Kapoiav CLVTOV
(Aram.); 7, 53 els diarayas ayyeXuv ; 8, 6 & TW awveiv aurous;
9, 3 eV S T$ iroptveadcu, eytvero CLVTOV fyyifcw; 9, 22
(Aram., ^nnx); 9, 32 dia TT&VTCW; 10, 14 ovdeirore

7ra> ($53) KOWOJ'; 10, 15 TrdXij' e/c devTcpov (probably ily
so also Matt. 26, 42); 10, 25 eyevero rov d<r\Qeiv\ n, 4
(Aram.); n, 22 riKovaOr] els ra wra; 12, 3 7rpo0-0ero
<rv\\a(3eLv ; 12, 10 irporj\0ov pv^v piav (nn for indefinite article;
even more common in Aramaic than in Hebrew); 13, n KCU vvv
idov x*ip Kupiou iirl <re; and also axpt Kaipou (pij; ^, Dan. 7, 12
etc.); 13, 12 KTr\t]TTOfjivos em (i>y) rf? 5i5ax; 13, 24 Trpo 7rpo(rco7rou
TTys tcr65ou auroO; 13, 25 O^K el/it 70? (Aram.); 14, 2 eKaKoxrav ras
\frvxas T&V iQv&v] 14, 8 xu>\6s IK /cotXtas fj.rjrp6s avrov (also 3, 2);

14, 15 evayyf\L^6fj.VOL vfj.as eTnffTpefaiv eiri deov Z&vra', 15, 4 Trapt-
bixdriaav airo TTJS eKK\r)ffla.s (} i^apDN , the invariable idiom in Ara-
maic. Correction to UTTO, as in most MSS.,was inevitable); 15, 7
ev vfuv e^cXe'^aro (see below); 15, 3 (bre/cpifli; 'IaKco/3os (the very
common Aramaic njy " take up the word," sometimes hardly
more than " speak "; cf. Dan. 4, 27 ! So also 3, 12 and 5, 8); x

15, 23 ot [email protected] adt\<f>oL

The fact that so many of these idioms are obviously Aramaic,
while no specifically (or even prevailingly) Hebrew idiom is to be
found, is certainly not accidental. Moreover, it is not enough to
speak of frequent Semitisms; the truth is that the language of all
these fifteen chapters is translation-Greek through and through,
generally preserving even the order of words.

In the remainder of the book, chapters 16-28, the case is altogether
different. Here, there is no evidence of an underlying Semitic lan-
guage. The few apparent Semitisms (/cat idov; ey&ero with
infin.; rare used in continuing a narrative; iv&iriov with gen.;
Wero v raj Tr^eujuart Tropeveffdai ; /c /zeVou (iv AieVw) avr&v) are
chargeable to the Koine; though their presence may be due in part
to the influence of the translation-Greek which Luke had so exten-

1 The idiom is also Hebrew. As for 2 Mace. 15, 14, it was written by a man who, as
we have good reason to believe, was as familiar with Aramaic as with Greek (see my
Aramaic Gospels, 295).



8 THE COMPOSITION AND DATE OF ACTS

sively read and written. In either case they are negligible. Luke's
own language if that is really what we have in the latter half of
Acts has a simplicity of structure that is often much like the
Semitic, and this fact renders the transition less abrupt. Harnack,
Apostelgeschichte 16, says: " Im allgemeinen kommt Lukas' Stil dem
der Septuaginta, namentlich aber dem der Makkabaerbiicher (der
aber selbst nichts anderes ist als der Stil der gesprochenen Sprache,
von gebildeten Mannern behandelt) sehr nahe." Whoever is well
acquainted with the literature here named will rub his eyes when he
reads these words. The " style " of the LXX is simply the style of
literal translations from Semitic originals, the clumsy result of put-
ting Hebrew writings into a too closely fitting Greek dress. Luke's
style in Acts 16-28 (the only place, excepting Luke i, 1-4, where
we can really examine it) has in it scarcely anything to remind us of
the Greek Old Testament. In structure, syntax, and idioms habi-
tually employed its Greek belongs to an altogether different genus.
And what is "the style of the Books of Maccabees"? i Mace, is a
closely literal rendering from a Hebrew original. The style of
2 Mace, is rhetorical, somewhat labored, and much more pretentious
than that of Luke, and is totally different from that of i Mace.
The style of 3 Mace, is so overloaded and bombastic as to make the
book very tiresome reading. In 4 Mace, we have the work of a
master of Alexandrian rhetoric, but his style has hardly any resem-
blance to that of Luke. The Greek of Acts 16-28, then, is not
" like that of the LXX," to say nothing of the widely diverse Books
of Maccabees. Furthermore, even if we substitute "language" for
" style," it is not true that Acts 1-15 sounds like the Koine. It
sounds, on the contrary, like i Mace., Jeremiah, Daniel, and all the
other translations from Hebrew or Aramaic. The voice of the
Aramaic can be heard through the Greek. Luke translates like the
best interpreters of his time, always faithfully and generally word
for word. When he writes his own language, on the other hand, the
resulting Greek represents a Syrian type of the Koine which reads
smoothly and is sufficiently idiomatic. 1 In short, the Greek of the

1 In some respects the Greek of Marcus Diaconus' Life of Porphyrius of Gaza offers
an interesting parallel to that of Acts 16-28, after due allowance has been made for the



THE ARAMAIC SOURCE IN ACTS 9

first half of Acts differs widely and constantly from that of the sec-
ond half, both in the idiom which it uses and in its literary structure.
There is one obvious and satisfactory way of accounting for this
fact, namely the hypothesis of translation in the first half. Is there
any other adequate explanation ? 1

It is perhaps unnecessary to say that any attempt to reconstruct
the Judean Aramaic dialect of the middle of the first century is
bound to be arbitrary, and that the result can only be an artificial
idiom. We must rely chiefly on our meager knowledge of the Ara-
maic of the " Biblical " period (3d-2d centuries B.C.), and our hardly
more satisfactory acquaintance with the dialect of the Onkelos Tar-
gum (mainly second century A.D.; a translation idiom, with all the
usual characteristics of such a creation) . We have also the valuable,
though very scanty, aid afforded by the Megillath Taanith and other
bits of the genuine Judean speech of the first or second century
which have been preserved in the Talmud and elsewhere. The
many other helps, necessary but of minor importance, need not be
mentioned here. Questions as to the type of speech most likely to
be employed in such a narrative as this in Acts, whether popular
or formal, whether archaizing or representing actually current use,
are perhaps a mere waste of time. The answer to them, moreover,
would not in the least affect the results reached in any of the pas-
sages discussed in the following pages. In my own attempts at
retranslation I have been guided by the probability that since this
is distinctly a literary composition, and also written from the stand-
point of the Jewish sacred tradition, its diction may well be supposed
to have inclined toward that of the older models. At all events, the
words and phrases here conjectured are all truly Aramaic and Pales-
tinian, and possible of use at the time supposed.

interval of time between the two writings. The style is very simple, and the language
contains some distinct Syriasms O'ust as Luke's frequent use of r6re, " thereupon," is
probably due to the influence of the Aramaic PIK). Luke's style, however, is even
more direct and effective, and also stands on a higher literary plane.

1 In regard to the untenability of the theory that Luke " imitated the LXX " I have
expressed myself at some length elsewhere (Aramaic Gospels, pp. 285-288).



10 THE COMPOSITION AND DATE OF ACTS

i. ESPECIALLY STRIKING EXAMPLES OF MISTRANSLATION

5 \)

IN ACTS 1-15

Especially striking evidence of translation in chapters 1-15 is
afforded by the following passages. I have put first a number of
examples of serious mistranslation; then follows a collection of
minor slips, including too literal renderings. This latter list could
be considerably lengthened.

2, 47. The most interesting of all the phrases which suggest trans-
lation is found in 2, 47. The narrator is telling how the first large
body of believers was formed in Jerusalem, as the result of those
things which happened on the day of Pentecost. The new commun-
ity was harmonious within, and was looked upon with favor by all
the people of the city: " Day by day, continuing steadfastly with
one accord in the temple, and breaking bread at home, they did take
their food with gladness and singleness of heart, praising God and
having favor with all the people." Verse 47 then continues: 6 8k
Kvpios irpofffridet. TOVS <ro}ofj,vovs nad' r)fj,epav eiri rb aur6. Excepting
the last three words, this is just what we should expect: a general
statement regarding the increase of the newly formed church, simi-
lar to the statements made at frequent intervals (4, 4; 5, 14; 6, 7;
9, 31, etc.), throughout this narrative. But the words eiri TO avrb
have remained an unsolved riddle. The phrase ordinarily means
" together," " in the same place "; in the Greek Old Testament it
is the standing equivalent of nrp and rnrr . It has just been used in
this chapter, vs. 44: " And all that believed were together, and had
all things common." Other passages in Luke Acts are: Luke 17,
35, "Two women shall be grinding together"] Acts i, 15, "A multi-
tude of persons together"; 2, i, "They were all together in one
place." Cf. also 4, 26, where the phrase is taken over from the Greek
Old Testament (Ps. 2, 2 liri r6 avro =irv)- But in 2, 47, the passage
under discussion, the meaning " together " is obviously inadmis-
sible. It is true that Lumby in the Cambridge Bible, with the
scholar's wish to follow well-known usage, renders: " And the Lord
added day by day together such as were in the way of salvation " ;
but other scholars will see in this only a bit of " translation-English,"



THE ARAMAIC SOURCE IN ACTS II

a rendering of the kind so familiar in the Greek Bible and possibly
exemplified in this very eirl TO CLVTO as it stands before us so ill-suited
to its context.

The ancient interpreters felt the difficulty of the phrase, and tried
in various ways to overcome it. In the textus receptus the attempt is
made to join the troublesome words to the following verse, making
them the beginning of 3, i : " Now together Peter and John went up
to the temple," etc. ; a futile expedient which, however, bears elo-
quent witness to the inability of early readers who really knew
Greek to give the iiri TO O.VTO any plausible connection with the
preceding context. Many old manuscripts and versions endeavor
to improve the passage by inserting T# KK\r](riq. (a dative is
to be expected after Trpoo-eriflei) or Iv rf) tKK\ri<ri<t., either before
or after the three words under discussion, in order to remove
as much of the obscurity as possible. Thus, for example, Cod. D
has . . . Kad' rjfjitpav iiri TO O.VTO iv rf; KK\Tjffiq.. This form of the
insertion might seem to provide a foothold for the eiri TO aur6, and
it is therefore worthy of especial notice that the Syriac version and
its congeners connect the latter phrase with 3, i, although reading
kv XT) e/c/cXr/crt^. 1

Modern commentators and interpreters have passed around the
difficulty in more or less unhappy fashion. The Revisers of 1881
render: " And the Lord added to them day by day those that were
being saved," but remark in the margin that instead of " to them "
the Greek reads " together "; that is, they really do not render the
phrase at all. In Weizsacker's N.T. we read: " Der Herr aber
fiigte ihrer Vereinigung taglich bei, die sich retten liessen "; but the
Greek cannot possibly mean this. Preuschen (in Lietzmann's Hand-
buck] 1912) omits the phrase in his translation, remarking: " eirl
TO WTO verstarkt hier lediglich die Proposition in Trpoo-ertflet." So
far as this is an explanation at all, it means that either Preuschen or
the author of Acts cared nothing for Greek usage. Wendt (in
Meyer's Komm.; 1913) renders: " zu dem Zusammensein," but

1 The testimony of the Peshitta here has been commonly misunderstood and mis-
stated; thus in Von Soden's Schriften des N. T. it is given incorrectly in both verses.
The word akhedd (= nrP) in 3, i is unquestionably the rendering of ktrl r6 afa 6.



12 THE COMPOSITION AND DATE OF ACTS

straightway replaces this by a different rendering: " auf denselben
Ort hin," formerly adopted by Holtzmann (Handcomm.), who how-
ever recognized its great difficulty. This last suggestion in fact does
justice neither to Greek usage nor to the historical situation. The
incipient church in Jerusalem was not confined to any one meeting
place in such a way that the narrator could have said: " The Lord
daily added new converts (and brought them) to the same place " ;
nor, if he had wished to say this, would he have employed words
which seem to mean something else. The early Church Fathers and
scribes of the sacred text could not be satisfied with any of these
attempts at explanation; they saw clearly that something was
wrong with the Greek as it was first handed down to them. We
also may say with confidence either that the Greek of 2, 47, in the
oldest form known to us, has suffered corruption, or else that its
author was writing under some such compulsion as that of trans-
lation.

Under these circumstances, the hypothesis of translation from a
Semitic original certainly deserves to be considered. When the test
of retroversion into Aramaic is applied, the result is unexpectedly
interesting, for it not only provides an easy solution of the difficulty
of the passage, but also seems to furnish direct evidence that author
and translator lived in different parts of the Aramaic-speaking world.

Of the possible Aramaic equivalents of the Greek eVl r6 avr6,
Hebrew -HIT , only one needs to be considered, namely the adverbial
compound tr\rb /&nr6 . Etymologically, this is equivalent to in
unum, and it is occasionally used in this literal sense, " into one,"
meaning " together." Thus in John n, 52, " that he might gather
together the children of God who were scattered," the Syriac versions
have lahdd (Greek els >). Similarly in John 17, 23, " that they
(the believers) may be perfected together " (lit. into one; Greek as
above), the Palestinian Syriac has lahdd, while the Lewis and Pe-
shitta versions have lehad. A good example of the use of the word
to mean "together," Heb. nrr, is found in the Palestinian Syriac
version of Is. 43, 17: " Who bringeth out chariots and horses, host
and hero together (N"ir6)." But in the Judean dialects of Aramaic
the usual meaning of tnr6 is " greatly, exceedingly," and this is pre-


1 3 4 5 6 7

Online LibraryCharles Cutler TorreyThe composition and date of Acts → online text (page 1 of 7)