The problem then arises, how far these sources can be traced in the following chapters of Acts. Harnack is inclined to see in 5^'"^^ a doublet of 4^-'^^, and to assign the latter to A, the former to B. This is not improbable, but it is not so certain as the previous results. It is, for instance, by no means improbable that the apostles were twice arrested, and, as the story is told, 5^'' seems a not unnatural continuation of ch. 4. It is, however, true that the characteristic ' Peter and John ' is not found in 5'''^- ; but, on the other hand, the rather curious phrase apxriyiv is applied to Jesus in 318 and 5^^ (elsewhere in NT only in He 2^" 12^), which militates somewhat against the view that these chapters belong to different sources. In the same way the story of Ananias and Sapphira in Ac 5^"^' would fit quite as well on to B as on to A, with which Harnack connects it. Linguistically there is no clear evidence, but it may be noted that 0d/3os is a characteristic of the Christian com- munity in B in 2^*, and is repeated in 5^* ". It is not found in A, though from the circumstances of the case not much weight can be attached to this. It therefore must remain uncertain whether Ac 5 ought to be regarded as wholly A, wholly B, or be divided between the two sources.
(^) The Jerusalem-Ccesarean sections. — These are Ac 8«-« 931-III8 121-^, which describe Philip's evan- gelization of Samaria, followed by the mission of Peter and John, Philip's conversion of the Ethiopian on the road to Gaza, and his arrival in Ca?sarea, Peter's mission to Lydda, Joppa, and Cajsarea, and return to Jerusalem, Peter's arrest, imprison- ment, and escape in Jerusalem, and Herod's death in Csesarea. Harnack thinks that all these pas- sages represent a Jerusalem-Caisarean tradition, which he identifies with source A. It is certainly probable that S^*"-^ belongs to A, owing to the characteristic combination of Peter and John, and it may be regarded as reasonable to think that this also covers the rest of the section, so that S^'*" may be attributed to A. It is more doubtful when we come to the two other sections. If, how- ever, any weight be attached to the suggestion that A is connected with Mark, it is noteworthy that 12^"23 is also very clearly connected with the house of Mark and his m(>ther.
The section 9^'-ll'8 remains. This is much more clearly Caisarean than either of the others, and might possibly be separated from them and as-
• See Burkitt, Earliest Sources of the Gospels, London, 1911, p. 79 f., where the suggestion is made that the early part of Acts may represent a Marcan tradition, though the bearing on this theory of the double source A and B in Acts is not mentioned.
ACTS OF THE APOSTLES
ACTS OF THE APOSTLES
25
cribed to a distinct Cesarean source. If so, the suggestion of Harnack and others that the source miglit be identified with the family, of Philip, which was settled in Csesarea, is not impossible ; from 21^ (a ' we-clause') we know that Luke came into contact with him there. It is also obvious that the information given by Philip might be the source of much more of that which has been ten- tatively attributed to source A, or on the other hand might conceivably be identified with source B ; the truth is, of course, that we here reach the limit of legitimate hypothesis, and pass into the open country of uncontrolled guessing.
The result, therefore, of an inquiry into the sources of the Jerusalem tradition is to establish the existence of a written Greek source, A, in Ac 3f., with a parallel narrative B — apparently the continuation of the Lucan Jerusalem narrative in the Gospel ; and these two sources, or one of them, are continued in ch. 5. In 8^"*" is a further narrative which has points of connexion with A. Ac9^'-lP^isa Caesarean narrative, probably con- nected with Philip, and this raises difficulties in relation to A, for H^-'^ has also points of connexion with Philip. Finally 12i-23is a Jerusalem narrative connected with Peter and Mark ; but here also the possibility of a connexion with Caesarea remains open.
V. Historical Value of the Various Tra- ditions.— So far as the ' we-clauses ' and the prob- ably Pauline tradition are concerned, this question has already been discussed. While there are traces of probable inaccuracy, there is no reason to doubt the general trustworthiness of the narrative. The Antiochene narrative and the Jerusalem-Csesarean narrative (the ' Philip ' clauses) can be judged with more difficulty, as we have no means of comparing the narratives with any other contemporary state- ments. Here, however, we have another criterion. It is probable that Luke is dealing with traditions, and, at least in the case of A, with a document. We cannot say how far he alters his sources, for we have no other information as to their original form, but we can use the analogy of his observed practice in the case of the Gospel. Here we know that he made use of Mark ; and we can control his methods, because we possess his source. In this way we can obtain some idea of what he is likely to have done with his sources in Acts. On the whole, it cannot be said that the application of this criterion raises the value of Acts. In the Gospel, Luke, though in the main constant to his source Mark, was by no means disinclined to change the meaning of the story as well as the words, if he thought right. It is possible that he was justified in doing so, but that is not the question. The point is that he did not hesitate to alter his source in the Gospel ; it is therefore probable that he did not hesitate to do so in the Acts.