5, 17. H.O.VT6S ol avv a.irr$, 77 ovaa cupecns ruv 2addovKaicov.
Wendt: " Das Part. ^ ovva, statt ol fores, ist attrahiert vom
Pradikat." I believe he is mistaken in this. In the two passages
1 See for example Dalman, Worte Jesu, 18 f.
THE ARAMAIC SOURCE IN ACTS 33
(5, 17 and 13, i) where this construction occurs it is merely Luke's
careful way of reproducing the Aramaic rvx (the word JVN is exactly
oixrid). The phrase was this: ^Nj^nv n KrrYOn .TJVK V! T, "who were
the sect of the Sadducees." See the note on 13, i.
5, 28. The infinitive absolute, as idiomatic in Aramaic as in
Hebrew. 1 The outwardly similar construction found in 23, 14 and
28, 10 (concrete nouns) is essentially different.
7, 38. Is it not likely that |n ^D, "words of life" was acciden-
tally miswritten pn jta (|jn }k>), " living words " ? Or is it merely
the rendering that is at fault ? The reference is plainly to such pas-
sages as Ezek. 20, 10 f. : "I brought them out of the land of Egypt,
and into the wilderness; and I gave them my statutes and showed
them my judgements, which if a man do, he shall live by them." Also
Ezek. 33, 15; Lev. 18, 5; Deut. 30, 15-19, etc.
7, 52. Preuschen: " Der Ausdruck TOV 5ucuou fur den Messias
ware Juden kaum verstandlich gewesen." This statement, unless
hastily made, shows a very imperfect acquaintance with the Jewish
conception of the Messiah. His chief office was to establish justice
in the earth, Is. 42, 3 f.; cf. also 53, n, etc., and the i7th and i8th
of the Psalms of Solomon. See also the various designations of the
Coming One as " the righteous Messiah " (Dalman, Worte Jesu,
7, 53. The curious phrase, "unto ordinances of angels," els Siartryds
a.'Y'YeKuv. The els represents f>, meaning " according to," or " by."
" Ye who received the Law pa^ *jTOptt&, by the ordering, or ad-
ministration, of angels." For the use of the preposition compare
for example Ps. 119, 91, *pBBBtoi>, "according to thine ordinances";
119, 154, ^rniD&6, "according to thy word," and many others.
Luke's rendering here is not merely too literal, it is incorrect.
8, 7. The grammatical difficulty of the first clause is sufficiently
familiar. Preuschen remarks that the text is " unheilbar verdor-
ben "; see his commentary and that of Wendt for the catalogue of
attempts, ancient and modern, to improve the reading; notice also
the [Trapd] TroXXois of Codex Bezae. In Aramaic, however, the sus-
pended construction is not unusual, the anacoluthon being avoided
1 Dalman, Worte Jesu, 27 f., exaggerates its rarity.
34 THE COMPOSITION AND DATE OF ACTS
by the introduction of a suffixed pronoun in the latter part of the
clause. For example: i>p3 prov pruo PIT 1in pn n PA>K p p'ao n
ipaj ai . This would almost inevitably be rendered into Greek by
the exact form of words which we have before us. The translator
would gain nothing, but only make his Greek worse, by rendering
jimo. His version was not in the least ambiguous, it was merely
9, 2. " Any belonging to the Way." A genuine Semitic locution,
which seems to have been taken over by the Gentile Christians from
the speech of their Jewish brethren. Thus Talm. Rosh Hashana i t ja >
uayn '3T1D ifcrpa, "they separated themselves from the ways (reli-
gion) of the congregation," i.e., they became heretics. So also in
old Syriac: urha d'Taiydyuthd, " the religion (literally way) of the
Arabs "; urha damshihd, " the Christian religion "; other examples
in Payne Smith, Thesaurus. So too in Arabic, as-sabil, " the way,"
is used, without any further description or qualification, for the true
(Mohammedan) religion. The adopted Gentile use in 19, 9, 23, etc.
9, 316. This is probably the idiom which is so common in the Old
Testament: Hebrew mi ^n , Aramaic ttiDl^m, "constantly in-
creased," " abounded more and more," and the like, i Sam. 14, 19,
" The tumult kept growing greater and greater " (iropevofj-evos ...
Tr\r)6vvfv) ; 2 Sam. 3, i, "David grew stronger and stronger (iiroptv-
ero K<LL eKparcLLovTo), while the house of Saul grew weaker and weaker
/ecu lyotfem); Gen. 8, 5, "The water constantly diminished
^XarrowOro) ; and a great many similar cases.
9, 32. Peter passed through " the whole (region)," 6id TTCLVTUV,
is often used thus absolutely, when the context makes
the meaning evident. For a Judean writer, to whom Palestine was
" the land " (cf. Njn ^ in n, 28, discussed above), this was doubt-
less the usual expression in such a context.
10, ii ; ii, 5. The unusual apxn in these two passages is the
rendering of the much more common Aramaic e*n, "extremity, cor-
ner," etc. The participle Kadi^evov might represent either the root
23"i^ or ^B>.
10, 30. "On (dTro) the fourth day (i.e., three days ago), at (M'xpO
this hour." This is not a permissible idiom in Greek, where the
THE ARAMAIC SOURCE IN ACTS 35
words would necessarily mean " for four days up to this hour." It
is perfectly good Semitic, however: NT Kny.e* nj? **???} NI ?^ ?PJ t ^ at
is, " on the fourth day back, reckoning up to this same hour."
10, 36 f. T6v \6yov ov aire<TTi\6v , K.T.\. Reduced to Aramaic
this would sound much better, since the suspended construction is
usual in that language. It is possible, too, that the last clause of
vs. 36 was originally intended quite differently.
If the Aramaic had been jne* T3 D^ "IBID htnto* ^ rb& n Nnk)
N;p too &un KrpBto, it certainly might have been understood as we
have it in Luke's word-for-word rendering. But it could also be
translated as follows: "As for the word which the Lord of All 1 sent
to the children of Israel, proclaiming good tidings of peace through
Jesus Christ: ye know that which took place in all Judea," etc.
This is at all events faultless Aramaic idiom. In favor of it may
also be said: (i) The title xio too, Kvpws iravruv, according to all
Jewish usage belongs to the Lord of Hosts, the God of Israel; and
such titles are not easily given a new application. The Syriac equi-
valent, Mare Kul, is a standing designation of the One God. In
Hebrew we have the titles " Ruler over All " (i Chr. 29, 12, etc.),
" Maker of All " (Jer. 10, 16; 51, 19), probably " Captain of All "
0>an ib>, Dan. n, 2), 2 and " Lord of All the Earth " Qosh. 3, u,
etc.). In Aramaic we have also, as standing titles of Yahwe, " Lord
of the Heavens " (Elephantine Papyri, Dan. 5, 23, etc.), " Lord of
the World " (Targums, passim), " Lord of All the World " (Targ.
Micah 4, 13), Lord of the Worlds," whence Arabic Rabb al- l Ala-
min; cf. also Ps. 145, 13, Tobit 13, 6, 10, etc. 3 It is intrinsically
improbable, then, that the title " Lord of All " would have been
applied to Jesus in a Judean Aramaic document of the first century.
(2) Again, it is to be observed that what is especially emphasized
in this whole passage is the purpose of the all-powerful God. He is
1 Literally, " this Lord of All "; see the note on i, 5, above, and cf. Dan. 2, 32, etc.
The use of such a demonstrative pronoun is common in the Judean dialect. Here,
moreover, there is a very obvious reason for its use, since in the preceding verse it had
been said that the God of Israel is also the God of all nations.
1 "ifc? inserted by conjecture after TJP ; see Journ. Am. Or. Soc. 25 (1004), pp. 310 f.
3 The " Lord of all " in Rom. 10, 12 is of course not a title, nor to be compared with
the present passage.
36 THE COMPOSITION AND DATE OF ACTS
the God of all nations (vss. 34, 35); this Jesus was anointed of God
(38) ; God was with him (ibid .) ; God raised him up (40) ; his wit-
nesses were chosen of God (41) ; and they testify that God ordained
him (42). This purpose of emphasis would be naturally served by
the position of the subject, " the Lord of All," at the end of its
clause in vs. 36.
For the case of dp^d/ze^os in vs. 37, see above, on i, 22. As for the
rbv \6yov 8i> airiffTi\v, at the beginning of vs. 36, the noun is to
be taken as the direct object of otiare (vs. 37).
10, 40. "EdoKv avrov tfjujxivri yeveadai, i.e., rpmnr6 rtoiV. The
same idiom in 14, 3; cf. also 2, 27 (quoted from LXX), and the
many examples in the Greek O.T. Well known as a Semitism ; Blass
Gramm., 69, 4.
u, 4. On dp^d/zevos, see above, on i, i; i, 22, etc.
11,6. The combination narcvbovv KCU eldov (after aTeviaasfy would
be remarkable as a specimen of Greek style. But this, exactly, is a
favorite Aramaic idiom, rPTrn JT^snpx . See, for example, the Targ.
Eccles. 9, ii : rprm n^anox (not in the Hebrew); Targ. Is. 42, 18:
ftrn ^3DDN (not so in the Hebrew); cf. also Dan. 7, 8, etc.
n, 16. "I remembered the word of the Lord, how that he said,
John indeed baptized with water," etc. This was written by the
author of i, 4, obviously, and brings incidental confirmation of my
demonstration (see below, page 59) that the Aramaic document
used by Luke begins at i, ib.
n, 21. Ecu fa xdp Kvpiov /ucr' abruv. This is another plain
Semitism. Cf. Luke i, 66, etc., as well as the many passages in the
11, 22. " The word was heard into the ears of the church." No
Greek writer would ever have perpetrated this unless he had
wished to create the impression that he was using a Semitic "source."
Even then, he would doubtless have used the standing LXX phrase,
Iv rots &ffi.
12, ii. " Expectation " is too weak for this context, which speaks
of that from which Peter was delivered. Hpo<r8oida rendered Kna^no ,
which ordinarily means " thought, opinion, calculation," and the
like. But the word is not infrequently used, in Hebrew, Aramaic,
THE ARAMAIC SOURCE IN ACTS 37
and Syriac, to mean " plot, machination" Thus Esther 8, 3 : Esther
besought the king to bring to nought that which Haman " had
plotted (apn) against the Jews "; Jer. 18, 18: " They said, Come, let
us lay plots against Jeremiah "; Ps. 52, 4: " Thou plottest mischief
with thy tongue "; and many other passages, both in the Hebrew
O.T. and in the Targums. 1 The " Zamzummim " of Deut. 2, 20,
interpreted as " plotters " (Heb. DDT, " devise evil "), are called in
the Targum pjat^n . An example in Syriac is John of Ephesus 18,19
(1,17): " treacherous plotting"
12, 20. 0ujuo/iaxj> is presumably ton . This meant, in the speech
of Judea, " angry," literally " burning "; thus also in the Hebrew of
the Old Testament. But in the North Syrian dialect the verb, used
chiefly in the reflexive stem, means " contend against, strive with."
Luke's rendering is a model of exactness, but the Judean meaning,
" angry," is the correct one here.
13, i. The phrase, Kara TT\V ovvav ^KK\rjffiav , "in the church
which is (or, was) there." This is another example of the transla-
tion of JVK ; see the note on 5, 17, above. The Aramaic was probably
simply (or ,TJVN) rpK n Nrnya , no accompanying adverb being neces-
sary, since it was made evident by the context. The commentators
sometimes compare Rom. 13, i, also Acts 28, 17, etc.; but these
passages are not really parallel cases, since in them the participle,
or its equivalent, is indispensable. Other passages in the Aramaic
half of Acts where JTN seems to be rendered are n, 22 and 14, 13.
13, 22. " He raised up for them David as their king (els /Sao-iXe'a),"
ijfo!? T-n jir6 D'j?N .
13, 24. This is altogether too literal a translation of 'niJTO Dnp jo,
" before his coming." See the note on 3, 20, above.
13, 25. " As John was ending (literally, fulfilling) his course."
'Eir\fipov is the translation of Aramaic D?>e*; cf. the note on 2, i,
J 3> 25. "Who do ye suppose that I am?" ri lp* vTrovoeire elvai',
It can hardly be questioned that ri, rather than T'LV a, has the pre-
sumption in its favor as the original reading. The fact that the
1 The word ktrivoia in 8, 22 probably renders this same Aramaic word. Apparently
there also the translation is too colorless.
38 THE COMPOSITION AND DATE OF ACTS
neuter pronoun is " nicht ertraglich " (Blass, in Wendt, p. 210)
makes the case all the more interesting. This is the regular Ara-
maic idiom. No better illustration could be asked than that which
is furnished by the Lewis 1 and Cur. Syriac renderings of Matt. 16,
13; Mk. 8, 27; Lk. 9, 18: " Who do men say that I am ? " using
only mdnd (" what ? "), in spite of the riva in every passage.
13, 25. OVK dul 70), "I am not he." It is worthy of note that the
Aramaic (not Hebrew) idiom simply repeats the pronoun of the
first person; "I am he" is KJN NJX. Thus e.g. the Syriac in John 4,
26: " I that speak with thee am /."
14, 17. There is apparently a mistranslation of some sort here.
It is no more agreeable to usage in Aramaic or Greek to speak of
' filling hearts with food ' than it is in English. Perhaps originally
"Filling your hearts with all gladness" (cf. Rom. 15, 13); and
confusion of ^20 with ^yo " food," since the nun of the preposition
was frequently assimilated at this time in Judea, but very rarely else-
where. The verb N;> might of course be construed either with p or
with direct object.
14, 27; 15, 4. The phrase off a eirolrjaev 6 6eos juer' avruv. On the
difference of opinion among scholars as to the meaning of this, see
Thayer, Lexicon, s. v. juerd. It is, however, merely translation-
Greek, meaning: " what God had done to (orfor) them." There is
no idea of cooperation in the phrase, nor even of accompaniment.
This is the regular idiom in all branches of Aramaic. Thus, an in-
scription from Tarsus, fifth century B.C. (Journ. Am. Or. Soc. 35,
Part 4) : " Whoever does ("ny) any harm to (oy) this image," etc.
Dan. 3, 32: " the wonders which God has wrought upon me " (*ay
'Dy). Assemani, Bill. Or. Ill, ii, 486: " the miracle which was per-
formed on their king " (firota Dy "ayn). The idiom is also found
in Hebrew; see Deut. i, 30; 10, 21, etc.
15, 1 6-1 8. Luke always uses the Greek Bible for his Old Testa-
ment quotations; see my Aramaic Gospels, 298 ff. In this case, we
do not know to what extent the Greek varied from the Aramaic or
rather, Hebrew which actually lay before him. Rabbi Akiba and
his fellows had not yet set up a " standard " text of the Prophets; the
1 The Lewis Syriac in Matt. 16, 13 follows a different text, to be sure.
THE ARAMAIC SOURCE IN ACTS 39
author of this Aramaic document was at liberty to select the reading
which best suited his purpose; and the LXX rendering of Am. 9,
ii f. certainly represented a varying Hebrew text. But even our
Massoretic Hebrew would have served the present purpose admir-
ably, since it predicted that " the tabernacle of David," i.e. the
church of the Messiah, would " gain possession of all the nations
which are called by the name [of the God of Israel]." Cf. vs. 14,
where we are told what this quotation was expected to prove:
6 deos eirffK\l/a.TO Xa/3e' e edvuv \aov TU> ovo^ari avrov.
As for the troublesome ending of vs. 18, 1 believe that the explana-
tion is this: Instead of nt*T nfety, as in the Massor. Hebrew, the
reading of our document was D^iyo ntft jrnio a very natural im-
provement; cf. especially the d</>' fipepuv dpxcucojc in vs. 7. Luke, in
giving the quotation in Greek, wrote out his LXX word for word,
as usual. Arriving at the end of the verse, instead of rendering
ymo by yvupifav he was able, by the periphrasis TTCH.&V TO.VTO.
yvwara., " making these things known" to be faithful both to his
Greek Bible (iroi&v ravra) and to the document which he was trans-
lating. This is thoroughly characteristic of Luke; cf. for example
the notes on 2, i and 2, 24, above.
15, 23. Harnack, Lukas der Arzt, 154, speaks of the " merk-
wiirdige Ausdruck 01 Trpeo-jSurepoi d5X0ot," and Preuschen, Komm.,
declares this beginning of the address " unertraglich." But it is
faultless Aramaic idiom. In the phrase Nns Ne^#pi Krv^, the word
T - T- '-; T- : '
" brethren " would naturally refer to both the nouns preceding; if it
had been intended to refer to the " elders " alone, it would have
stood between this word and the conjunction i. From the Christian
Aramaic (Syriac) which we know, it is evident that in early church
usage this apposed " brethren " was very common.
15, 28. HXrjv TOVTUV TUV eTr&vayKes. Professor G. F. Moore has
suggested (orally) what seems to me the correct explanation of this
improbable phrase. The Greek originally read: tbo&v . . .
ir\tov 7rm0e0-0(H viJ.lv jSdpos irX^f TOVTWV eiravajKes
K.T.I., the r&v being due to dittography. 1 iira.va.yKes dTr&eotfai ren-
1 Clem. Alex, seems to have read in just this way in his Stromata iv, 16, 97; this
reading of his was probably obtained merely by accident or conjecture, however.
40 THE COMPOSITION AND DATE OF ACTS
ders nj>rnriN^ Tjny, which according to Moore was probably the
reading of the Aramaic document in this passage.
The translation-Greek continues to the end of 15, 35, which prob-
ably formed the original conclusion of the Aramaic narrative (see
below). 1 With verse 36 the character of the language changes com-
pletely, so far as its structure is concerned, and the Aramaic idiom
does not appear again, even for a single paragraph. Two other
facts deserve especial attention. The first is, that the author of the
Greek half of the book composed his narrative as the continuation of
the Aramaic document. This is sufficiently obvious, not only from
the way in which vs. 36 takes its start from vs. 35, but also from the
correspondence of the details of the narrative in 15, 36-16, 5 with
those in the chapters immediately preceding; a relationship much
too close to be accidental. The allusions to the churches already
established in Asia are plainly intended as the sequel of chapters 13
and 14; 16, 4 is only comprehensible after reading 15, 1-29; 15, 38
refers to 13, 13; the speech of Paul in 17, 22-31 seems to be modeled
on that in 14, 15-17, though the resemblance may be merely acci-
dental (see below); and there are other striking correspondences.
This is of course just what we should expect in view of the remark-
able uniformity of vocabulary and phraseology in all parts of the
book, showing (as already noted above) that the translator of the
first half was the author of the second. The other fact deserving
notice is this, that the author, translator, and compiler was a man
singularly faithful to his sources. He disliked to alter, even slightly,
the document in his hands, even where he believed its statements
to be mistaken, and where he found himself obliged to contra-
dict them. Acts i, 3 (the " forty days "), for instance, is flatly
opposed to the statements in Luke 24 (see below), and the statement
in Acts i, 4 (" which ye heard from me ") was certainly recognized
as erroneous by the author of Luke 3, 16. As we have seen, the name
of the sorcerer Bar- Jesus was allowed to stand in 13, 6, though the
1 Attention may be called at this point to the evidence furnished by the foregoing
investigation that the text of Acts which has come down to us, especially in Cod. B
and its nearest associates, is very old and correct. The later and all but worthless
text of Cod. Bezae and its associates I hope to make the subject of a future study.
THE ARAMAIC SOURCE IN ACTS 41
substitute " Elymas " was used thereafter. The many cases of very-
faithful translation noted above, in passages where a somewhat
freer rendering would have saved the translator from real difficulties,
are in the same line of evidence. But perhaps the most striking
illustration of the kind is afforded by the point where the transition
is made from the Aramaic history to Luke's own narrative. Luke
did not believe that Silas returned to Jerusalem as narrated in 15, 33,
but rather (see vss. 36 and 40), that he remained at Antioch until
the time when he set out with Paul on the missionary journey. It
would have been easy to omit vs. 33, or to add a harmonizing state-
ment, as some less scrupulous editor of the text has actually done in
the vs. 34 which is now omitted from all critical editions. But
Luke, as usual, gave his source the word, and would not falsify it. 1
1 I mean, of course, that this was his way of dealing with a unique document of great
importance which he was translating. No one will doubt that he was quite ready to
edit, to omit, and to supplement with his own freely composed material, wherever these
activities were in place. He may have made numerous slight editorial additions here,
though this does not seem to me a necessary supposition, and I do not believe that it
would be possible to recognize them. Professor J. H. Ropes has given me the very
plausible suggestion, for instance, that the list of the apostles in i, 13 is Luke's own
addition, since it so closely resembles his list in Lk. 6, 14 f. But the Aramaic docu-
ment can hardly have been without such a list at this point, in view of the episode
which follows. Moreover, Luke's own list was certainly derived from a Semitic source.
THE INTEGRITY OF THE SECOND HALF OF ACTS
i. THE HOMOGENEITY OF II ACTS
It is beyond controversy that the general impression made by the
second half of the Book of Acts is one of homogeneity. Phraseology,
literary style, point of view of the writer, and mode of treatment of
the material, are noticeably the same throughout chapters 16-28; it
would be quite futile for any one to attempt to demonstrate the
contrary, in any of these particulars. Nevertheless the unity of this
half of Acts has long been called in question, perhaps by a majority
of the best scholars, and for reasons which are obvious. 1 The book
of Acts as a whole is plainly composite; the " Hebraizing " character
of the opening chapters, in contrast with the smooth Greek of the
last chapters, has long been the subject of comment. It is the style
of these opening chapters that most resembles that of the Third
Gospel; and the introductory words, mentioning Theophilus and
referring to the " former treatise," are inseparably welded to the
following history (see below). The Chris tology of the early chap-
ters, moreover, could not easily be attributed to a Gentile companion
of Paul. No theory of translation of documents has seemed to give
any help (especially as it has always been taken for granted that the
sources of the Third Gospel were Greek sources), nor has there
seemed to be any way of establishing such a theory. Then was
added the riddle of the " We-sections," giving such an inviting
opportunity for theories of composition. Furthermore, Acts 15 was
felt to be in disagreement with Gal. 2, so much so that it was hardly
conceivable that Paul's travelling companion could have written it.
Yet Acts 15 could not be separated from chapters 13 f. and 16, 1-5.
1 In the sequel, " I Acts " is used for chaps, i, 1-15, 35, and " II Acts " for 15,
THE INTEGRITY OF THE SECOND HALF OF ACTS 43
Hence also, apparently, the necessity of separating the " travel-
document " from the preceding account. The fact that portions of
the narrative are plainly untrustworthy as a record of events, while
other portions are as evidently historical, also seemed to some to give
a starting point for theories of composite authorship. Finally, the
supposed necessity of postulating a late date for the entire work
the Third Gospel being later than Mark and Matthew, and Acts
later than the Gospel gave support to the view that at least the
" travel-document " of II Acts was an older source incorporated in
the main work. After a beginning of analysis had thus been made,
there was no obvious halting place; it was simply a question of who
should be most ingenious and plausible in discovering joints, altera-
tions, and redactional patches.
The " We-sections " to begin with these present no difficulty
when the fact of Luke's translation of the Aramaic document is
recognized. The reason for the employment of the first person is
merely this, that the author of the account himself took part in some
of the events which are described, and was historian enough to feel
the importance of indicating the fact, though he does it in a very
modest way. In the portions of the narrative in which the third
person is used, in contexts where we should have expected the
author to indicate his participation if he had really been present, it
is most natural to suppose that he was not himself a participant in
the events, but obtained his information from others. Eduard Nor-
den, Agnostos Theos, 317-324, shows that the contemporary litera-
ture, both Greek and Roman, contains numerous exact parallels to
II Acts in this regard, and that if more of the writings of the time
had been preserved we should doubtless have had many other
examples. The demonstration is unnecessary, to be sure, since this