^{page 96 note 1} It is particularly true of Paul, who is said to have preached in synagogue at Damascus (ix. 20), Salamis (xiii. 5), Pisidian Antioch (xiii. 14), Iconium (xiv. I ), Thessalonica (xvii. I ), Beroea (xvii. 10), Athens (xvii. 17), Corinth (xviii. 4), and Ephesus (xviii. 19). There is also a less formal ‘synagogue’ in Paul's house at Rome (xxviii. 16). Preaching in synagogue was also recorded of Barnabas (with Paul) (xiii. 5), Apollos at Ephesus (xviii. 26), and it is possibly implied of Stephen (vi. 8, 9).

^{page 96 note 2} There are several general notes of Jesus teaching in synagogue, Mark, i. 21, 39, vi. 2Google Scholar; Matt, . iv. 23, ix. 35, xiii. 54Google Scholar; Luke, iv. 15, 32, 44Google Scholar; John, xviii. 20. Three passages give greater detail, Luke iv. 16; the group Mark iii. 1, Matt. xii. 9, Luke vi. 6; cf. xiii. 10; and John vi. 26–59.Google Scholar

^{page 96 note 3} ‘Even if the present sermons be not abbreviations of longer originals, it is yet possible to suppose that the editors of the Gospels, when writing the outline of an imaginary sermon, wrote it according to the customary form of the sermons known in their days’, Morton Smith, Tannaitic Parallels to the Gospels (J.B.L. Suppl. 1951), p. 88

^{page 97 note 1} Davies refers at this point to Smith, Morton, op. cit. pp. 78–110.Google Scholar

^{page 97 note 2} Davies, W. D., The Setting of the Sermon on the Mount (Cambridge, 1964), pp. 8 f.Google Scholar In fact Davies's criticism, although basically correct, is a little too all-embracing. By coincidence, two of the passages quoted by Morton Smith conform to one of the sermon-patterns found in the homiletic anthologies. The passages are Mek. xiv. 10 and Sifre on Num. vi. 26. It is possible that Ta'an. was in origin a yelammedenu homily, but too little of it is left to be sure.

^{page 97 note 3} Smith, Morton, op. cit. p. 113, n. 40. The ‘notion’ he mentions, however general it may be, is one that in any case needs considerable qualification.Google Scholar

^{page 97 note 4} Homilies are in fact preserved in earlier works (see n. 2, above). As an example of a preDestruction sermon, Eliezer's, R. exposition of Ps. xxxvii. 14 f. (Ber.R. xlii. 1, Tanh, on Gen. xiv. 1; cf. A.R.N. §6) is usually quoted, and although it is extremely briefGoogle Scholar, Mann, J. was able to analyse it in classical proem-form (The Bible as Read and Preached in the Old Synagogue (1940), I, 105.Google Scholar

^{page 97 note 5} The origin of the synagogue in Judaea was closely connected with the ma'amadot, and the ma'amadot in turn were connected with a highly polemic issue between the Pharisees and the Sadducees, the participation of all Israel in temple sacrifice, particularly in tamid. This means that the origin of the synagogue in Judaea is pre-Destruction and is specifically connected with the reading of Torah, or at least with the reading of certain passages chosen to correspond to the sacrifices taking place in Jerusalem. From those simple beginnings the assemblies soon developed into regular occasions for the study and exposition of scripture. (See, for example, Jos. Con. Ap. II, 175 (cf. Meg, J.. iv. I), Ant. xvi, 43Google Scholar; Philo, , ad Gaium 311 f., de Opif. Mund. 128; and the Theodotos InscrGoogle Scholar. (Sukenik, E. L., The Ancient Synagogue of el-Hammeh, 1935)Google Scholar. Zeitlin, Both S. (in, for example, ‘The Tefillah, the Shemoneh Esreh’, J.Q.R. LIV (1964), 208–49) andGoogle ScholarHoenig, S. B. (in, forexample, ‘The Supposititious Temple Synagogue’, J.Q.R. Liv (1963), 115–31)Google Scholar have emphasized frequently that in origin the synagogue in Judaea must be distinguished from the house of prayer. It was only after the Destruction that the two were fused, partly as a response to the very fact that Temple service and sacrifice were no longer possible. (It is thus extremely significant that Acts xvi. 13 ff. maintains the distinction.) But in their search for ‘spiritual equivalence’ the rabbis at Jabneh made as much use as they could of existing institutions (see my forthcoming article on ‘The Cessation of Sacrifice’). Thus the position of Zeitlin and Hoenig is that the full emergence of the Synagogue, in which study of Torah and liturgical prayer were combined, did not take place until after the Destruction. This does not mean that there was no such thing as synagogue before that time. Nevertheless, L. Morris, in a pamphlet advising caution in the application of the old Palestinian lectionary to N.T. material (The New Testament and the Jewish Lectionaries, Tyndale Press, 1964), has used Hoenig's arguments to suggest that there may be doubt about the synagogue having existed in pre-Christian days: ‘It is plain that in the light of it (Hoenig's “rigorous examination of the relevant rabbinic material”) we must be more careful than ever in maintaining that the synagogue was established in Palestine in pre-Christian days’ (Morris, op. cit. p. 13).

Morris's language is extremely vague, because he does not define what he means by ‘synagogue’, but his words could certainly be taken to mean that he has serious doubt whether there was any form of synagogue before the Destruction. This was certainly the impression he gave to, for example, the reviewer of his book in J.T.S.: ‘First, Principal Morris finds no evidence for the existence of the synagogue before A.D. 70…’ (J.T.S. xvil (1966); it is only fair to add that the review is extremely general). If Morris did mean that, then he is wrong, and he cannot appeal to Hoenig (or to Zeitlin) for support, since neither of them holds that view. What they are talking about is ‘the full emergence of synagogues’ (Hoenig, p. 130) in which Torah-study and liturgy were combined. What is astonishing is that Hoenig made this distinction in the use of the word ‘synagogue’ entirely clear during the course of his article, but Morris does not appear to have seen it. In fact he made the point in a foot-note to one of the passages which Morris actually quoted, but Morris left the footnote out and made no reference to it. The footnote in question reads: ‘Though Prof. E. Rivkin, “Ben Sira and the non-existence of the Synagogue”, In Time of Harvest…, 1962, p. 346, is correct that “synagogues are mentioned in contemporary sources only after the Hasmonaean Revolt” he forgets, however, that these essentially were gathering places, assemblies in general and not specific “houses of prayer”. As S. Zeitlin has shown, the synagogues as houses of prayer originated from the Maamadot in pre-Hasmonean days butserved first as places where the Torah was read. Only later, after the destruction of the Temple, did they begin to flourish as houses of prayer in Judea. See S. Zeitlin, Rise and Fall of Judean State, pp. 179, 476. Proseuchae however existed in the Diaspora. See also S. B. Hoenig, “Historical Inquiries, Heber Ir”, J.Q.R., Oct. 1957, p. 138’ (Hoenig, ‘The Supposititious Temple Synagogue’, p. 130: the quoted passage in which the footnote was omitted is in Morris, op. cit. pp. 12 f.)

The origins of Jewish institutions, not least the Synagogue, are extremely complex. They cannot be solved by generalizations. In the case of the Synagogue, its origins are certainly to be found in the pre-Destruction period, but at that time it had far less formal structure, and was probably confined to the reading and study of Torah. From the point of view of the present article, that in itself is the important fact.

^{page 98 note 1} For a summary of some of those rules, see Gerhardsson, B., Memory and Manuscript (1961) (Eng. Tr.), pp. 67–70.Google Scholar

^{page 98 note 2} The date at which lectionaries were first introduced is unknown. Josephus emphasized the thoroughness with which Torah was studied in assembly: ‘He (Moses) appointed the Law to be the most excellent and necessary form of instruction, ordaining, not that it should be heard once for all or twice or on several occasions, but that every week men should desert their other occupations and assemble to listen to the Law and to obtain a thorough and accurate knowledge (άκρıß⋯ς έκμανθάνεıν) of it’ (Cont. Ap. II, 175, Loeb transl.). In view of the very precise purposes for which the synagogue was established, completely distinct as it was from general prayer and worship, it would be most surprising if the choice of passages to be read was left to chance. If the synagogue emerged in connexion with the ma'amadot, then a rudimentary lectionary is implicit from the earliest moment, since passages had to be chosen which would be relevant to the corresponding sacrifices taking place in Jerusalem. It is not hard to envisage the extension of this into a more general lectionary when the synagogue itself developed into the centre of Torah-study. It is most unlikely that the sequence of study was left to chance. Nevertheless, the fact remains that there is no direct evidence when lectionaries were first brought into use, nor is there any certain knowledge of the actual divisions of the Palestinian lectionaries in their earliest forms. The lectionary lists are late, and the lectionaries recovered from the homiletic midrashim by (for example) J. Mann bear witness to considerable variety, though it is important to remember that it is variety within a clear and uniform structure. Any theory of N.T. origins which relies on detailed knowledge of the lectionary cannot avoid this criticism of lack of direct information about the early lectionaries. Obviously, where there is no direct evidence either way, it is only possible to speak of probabilities, and in the present case the probability is extremely high that a lectionary existed before the Destruction. This is particularly so if the haftarot were being attached to the Pentateuchal sedarim in that period, as S. Zeitlin argues (see note 1, below).

^{page 99 note 1} Once again, the date at which haftarot were first attached to sedarim is uncertain. I. Abrahams (Studies in Pharisaism and the Gospels, vol. I, p. 8) argued that Luke iv. 16 shows that there was a fixed haftarah, since Jesus was given the scroll open at a particular place (Isa. 1xi. I f.). He did not choose the passage for himself. The implication is that haftarot were attached at a quite early date (though in point of fact that would depend on how the evidence of Luke is evaluated). Strack and Billerbeck argued the opposite (Komm. IV. i. 169 f.): M.Meg. iv. 10 records that R. Eliezer prohibited certain passages to be used as haftarot: ‘They may not use the chapter of the Chariot as a reading from the Prophets; but R. Judah permits it. R. Eliezer says: They do not use the chapter Cause Jerusalem to know as a reading from the Prophets.’ Strack and Billerbeck argued that there would be no need to forbid certain passages if the haftarot were already fixed. What they completely overlooked is the fact that even when haftarot were fixed, alternatives existed and were allowed. Several different hgftarot could be attached to a single seder.

More recently S. Zeitlin has argued that haftarot were attached to sedarim at an early stage, since the most obvious explanation of the fact that there are no haftarot from the ketubim is that the ketubim were canonized too late: ‘The Hagiographa were canonized in the year 65; …The book of Ecclesiastes was added at the school of Jabneh (c. 90–100 c.E.). The book of Esther was added at a later time at the Academy of Usha…The canonization of the Hagiographa at such a late date is the reason that there are no haftarot from the ketubim. When these were canonized haftarot already had been assigned to the Pentateuchal portions’ (Zeitlin, ‘Midrash: a Historical Study’, J.Q.R. XLIV (1953), pp. 28 f., transliteration standardized). A pre-Destruction date is certainly what the N.T. evidence implies.

^{page 100 note 1} ‘R. Johanan said: ‘When R. Meir used to deliver his public discourses, a third was halakah, a third haggadah, and a third consisted of parables.” R. Johanan also said: “R. Meir had three hundred parables of foxes, and we have only three left”’ (B.San. 386).

^{page 100 note 2} ‘The inner structure (of Midrashic homilies) has not yet been properly understood. Take the case of the Petihta (= Pet.) to which Bacher has devoted a valuable monograph (the reference is to Die Proömien in der alten jüdischen Homilie, 1913). According to him “the Pet. verse was left entirely to the free choice of the speaker; just in the selection of the Pet. verse and its interpretation, as well as in its coupling with the text of the (Torah) pericope, there were to become evident the esprit and the skill of the preacher” (p. 7). In the fact that the H(aftarah), which would have been very suitable for Pet. texts, was not used at all for that purpose, Bacher has found special proof for the above freedom of choice on the part of the homilist. By the latter observation Bacher has missed the crux of the whole matter. The H. was not used explicitly because it was tacitly employed throughout! (Mann's italics). In the following pages (of his book) it will be demonstrated in connexion with each TC seder how the verse chosen to introduce a particular Pet. tallied linguistically with a verse found within the compass of the H. Often the whole trend of the Aggadah developed in the Pet. can be accounted for only by turning to the H. which gave the homilist his starting point. He had the H. constantly in mind…’ (Mann, op. cit. pp. 11 f.).

^{page 101 note 1} The yelammedenu homilies in Deb. R. begin in exactly that way: instead of the formula ‘yelammedenu rabbenu’ they begin, ‘Halakah:’, and then pose the question.

^{page 101 note 2} Finkel, Thus A. (The Pharisees and the Teacher of Nazareth, 1964) claims Hillel's confrontation with the bene Bathyra (J. Pes. vi. 1, T.Pes. iv. 12 (bene Pathyra, perhaps rightly; see S. B. Hoenig, The Great Sanhedrin), B.Pes. 66a) as an early example of a yelammedenu homily: ‘We find a very fine example of a Yelammedenu homily dating from the early days of Hillel's ministry (c. 34 B.c.)’ (Finkel, op. cit. p. 169). Hillel's opinion, following recognizable exegetical rules, is undoubtedly a ‘yelammedenu’ response, but whether it can be claimed as a formal homily is more doubtful.Google Scholar

^{page 101 note 3} The custom is ascribed to Moses, indicating at least some antiquity, in Sifra on Lev. xxiii. 44, B.Meg. 32a (& J. IIs).

^{page 101 note 4} For example, Guilding, A. (The Fourth Gospel and Jewish Worship, 1957)Google Scholar and A. Finkel (op. cit.). It must be emphasized that the account given above of homiletic forms is a summary of the main outlines only. Obviously, there are many variations in detail, but the general forms remain recognizable. Often the variations are a consequence of the works in which the homilies have been preserved. Thus to give one example, A. A. Hallevy wrote of the homilies preserved in Bem.R.: (Introd. to Bem.R. in Hallevy's edn. of Midr.R. (Israel, 1963), p. xiii. See also his Introductions to Vay.R. (pp. xxii–xxiv) and Deb.R. (pp. xxx ff.).

It is not surprising to find these variations in Bem. R., since it is one of the latest parts of the Midr.R. in its compilation. Conversely, it would not be surprising to find variations at an early date before the forms fully established themselves. See further p. 104 below.

^{page 101 note 5} Acts xiii. 16–41.Google Scholar

^{page 101 note 6} Ibid. verse 15.

^{page 102 note 1} ‘And because he loved thy fathers, therefore he chose their seed after them, and brought thee out with his presence, with his great power, out of Egypt; to drive out nations from before thee greater and mightier than thou, to bring thee in, to give thee their land for an inheritance, as at this day.’

^{page 102 note 2} The form of the quotation is strange, but this will be discussed below.

^{page 102 note 3} ‘This is the man from whose posterity God, as he promised, has brought Israel a saviour, Jesus’ (verse 23).

^{page 102 note 4} LXX δıαμαρτὐρομαı; both Onq. and Ps. Jon., .

^{page 102 note 5} μάρτυρες.

^{page 103 note 1} It is because the haftarah is implied in the exegesis, and not usually quoted directly, that various different suggestions for seder and haftarah underlying the sermon in Acts have been made. As early as Cadbury, H. J. (B.C. iv, 148)Google Scholar it was suggested that the lections included Deut. i. 31 and a part of Isa. i. The argument briefly is that the texts of Acts vary in verse 18: N, B, C², D, 81, byz, vg have έτροποφ⋯ρησεν (’ he bore their manners’); A, C*, E, 33 minn d e g sah boh pesh hcl arm eth have έτροφοφ⋯ρησεν (‘he carried them like a nurse’). The same variation occurs in LXX Deut. i. 31, though there the φ is better supported (cf. II Macc. ii. 27). ὐψωσεν in verse 17 is also from Deut. i. 31, and this links with Isa. i. 2, ‘I have nourished and brought up children, and they have rebelled against me’.

One slight problem is that Isa. i. I ff. is not known to have been a haftarah to Deut. i, but it is known to have been the haftarah to Gen. xv. I ff. (see Mann, J., op. cit. seder 12)Google Scholar. One of the Tanhuma homilies (Tanh. B. Lek 19), on fear of sinning, threads in Hos. iii. 5, which is explained by Deut. iii. 25, which in turn might be linked with Deut. i. 8, and this has suggested to some another line of attack. But in fact the thread is tenuous in the extreme, and the exegesis is largely coincidental. In any case it must be remembered that haftarot shifted and changed a great deal. A. Guilding suggested that a better haftarah would be Jer. xxx. 4 ff., since it was a haftarah on Deut. i at a later time. She supported that suggestion by the fact that Paul quotes (loosely) Jer. xxx. 9 during the course of the sermon.

The fact that different suggestions can be made does not mean that no haftarah underlies the sermon at all. It simply points to the difficulty that arises when, as is normal, the haftarah is implied in the exegesis but not necessarily quoted. J. V. Doeve's suggestion (summarized below) has the merit that the section of II Samuel, which he finds underlying the sermon, is implied in more places than one.

^{page 103 note 2} Jewish Hermeneutics in the Synoptic Gospels and Acts (1954).Google Scholar

^{page 103 note 3} Cf., e.g., Acts xiii. 17 and 11 Sam. vii. 6, 22 and 8, 23 and 12, 33 and 14, 34 and 15, 36 and 12.

^{page 104 note 1} The Semitisms of Acts (1965), pp. 21 ff.Google Scholar

^{page 104 note 2} Lagarde, , Proph. Chald. p. 81, 11. 31–2.Google Scholar

^{page 104 note 3} Wilcox, , op. cit. p. 164.Google Scholar

^{page 104 note 4} It is worth remembering that the speech reveals equally clear signs of genuine Jewish thought and exegetical method, as has frequently been pointed out. (So, for example, Doeve, op. cit. pp. 175 f.: ‘…in the argument of Acts xiii the work of a schooled rabbi is quite perceptible… If the author of Acts composed the discourse…himself, then he must have had an excellent command of hermeneutics as practised in rabbinic Judaism.’)

^{page 105 note 1} It is hard to know whether verses 38 and 39 should be included as a part of the speech for the purposes of analysis. In point of fact it makes little difference since both point to the same section of Deuteronomy, as will be suggested below.

^{page 105 note 2} Cf., e.g., Acts ii. 22 and Deut. xxix. 2, ii. 36 and xxix. 10 f.

^{page 105 note 3} Joel ii. 32: ; Is. Ixiii. 19: … Cf. also references to the name in Isa. lxiii. 13 and 16 and Acts ii. 38.

^{page 105 note 4} Cf., e.g., Isa. lxiii. 11 and Acts ii. 33.

^{page 105 note 5} Acts ii. 36: ‘Let all Israel then accept as certain that God has made this Jesus, whom you crucified, both Lord and Messiah’ (leading to an appeal to enter the new covenant).

Deut. xxix. 10–11: ‘You are standing this day all of you before the Lord your God, the heads of your tribes, your elders, your officers, every single man of Israel, your young ones, your wives, your stranger which is in the midst of your camp, from the hewer of your wood to the drawer of your water, to enter into the covenant of the Lord your God…’

Acts ii. 39: ‘For the promise is to you, and to your children, and to all who are far away, everyone whom the Lord our God may call.’

Deut. xxix. 14: ‘Not with you alone do I make this covenant and this oath, but with him who is not here with us this day as well as with him who stands here with us this day before the Lord our God.’

^{page 105 note 6} ‘Men of Israel, listen to me: I speak of Jesus of Nazareth,…’ (note the emphatic order of the Greek).

^{page 105 note 7} Wilcox, has recently contested this point of view (op. cit. pp. 46–8)Google Scholar, so the passage needs some consideration. Acts ii. 24, in the context of the crucifixion and resurrection of Jesus, uses the phrase λὐσας τ⋯ς ⋯δίνας το⋯ θανάτου. The phrase is usually thought to rest on Ps. xviii. 4 (xvii. 5) and Ps. cxvi. 3 (cxiv. 3), perhaps with reference as well to Job xxxix. 2. The phrase, however, is such a curious one that it has often been explained by the suggestion that it was produced by a misunderstanding of an original Aramaic form. The steps of the ‘Aramaic solution’ may be summarized as follows (the summary being largely of Torrey's arguments as elaborated by Doeve): (I) The Hebrew phrase is . (2) The Semitic root hbl has two connected meanings, ‘pain (pang)’, or ‘cord (rope)’ (connected through the idea of twisting). (3) LXX understood the word in the Psalm as referring to ‘pangs’, and therefore translated it as ⋯δίνας. (4) The text of Acts agrees with the LXX in using ⋯δίνας, but Doeve argued: ‘This can almost certainly not have been the speaker's intention, for in contrast to the Septuagint he uses the word κρατείσθαι (LXX περıέσχον); and also λ⋯σαι. So he must have been thinking of something that held fast’ (Doeve, op. cit. p. 170). (5) This conflict of words (sharing the noun with the LXX but not the verbs, which at once become inappropriate) suggests that the original discourse was in Aramaic, and that when it came to be translated into Greek the translator had no guidance whether to choose ‘pangs’ or ‘cords’. He probably sought help from the LXX and chose ⋯δίνας, without at the same time correcting the verbs: hence the contradictory λὐσας║τάς ⋯δίνας το⋯ θανάτου. (6) Doeve produced an additional argument to support the view that the original discourse was in Aramaic: in Ps. xvi. 10, which is just about to be quoted in Acts, was translated in Aramaic by . This establishes a direct link in Aramaic only between Pss. xvi. 10 and cxvi. 3.

The earlier forms of an Aramaic solution were rejected by Clarke, W. K. Lowther (B.C. pt. I, vol. II, p. 97Google Scholar: ‘…an explanation from the LXX only will suffice…’), who preferred to explain the phrase as a conflation of LXX Ps. xvii. 5 and Job xxxix. 2. Wilcox rejected both Clarke and the ‘Aramaic solution’ because he argued that neither paid sufficient attention to the context of the quotation in Acts. The context is the quotation from Joel ii. 28–31, which is eschatological. So he says: ‘The phrase λὐσας τάς ⋯δίνας το⋯ θανάτου is thus seen as describing an eschatological fact and should probably be translated: “having put to an end the pangs of death”’ (op. cit. p. 48). He then refers for a parallel to the allusion to Ps. xviii. 5 (4) () in IQH iii. 28 (), which he argues has the meaning, ‘the pangs of death will compass (people) about’; ‘The reference is eschatological, as the context shows, and describes the Last Times, and the onset of the Messianic Woes’ (op. cit. p. 48). Wilcox is right in describing the Qumran psalm as eschatological, but his appeal to the context is misleading: an examination of the immediate context shows that his translation of 1. 28 is probably incorrect. IQH iii. 26 gives three images of traps or snares, which would suggest that l. 28 should be understood in the same way. Thus G. Vermes, for example, translated the lines in question: ‘Miseries of torments dogged my steps while all the snares of the Pit were opened and the lures of wickedness were set up and the nets of the damned (were spread) on the waters…It was a time of the wrath of all Satan and the bonds of death tightened without any escape’ (The Dead Sea Scrolls in English (Pelican, 1962), p. 159. The general context gives some support to Wilcox's position (see IQH iii. 9), but the immediate context does not).

The last phrase would seem to fit the sense of ‘bonds’ better, but on its own it is not decisive. Finally, on the question of context, it may be noted that although Wilcox appeals to the general context of the quotation in Acts, the immediate context remains that of the death of Jesus. In the immediate context ‘bonds’ still makes better sense.

It would seem, therefore, that Wilcox's rejection of the Aramaic solution of the quotation in Acts ii. 24 is not conclusive. It is significant that he makes no reference to Doeve's extensions of the Aramaic argument (suggesting a connexion through an Aramaic word between the two Psalm quotations in Acts). That argument is not necessarily convincing, but it deserves consideration. It is certainly surprising that Doeve's book does not even appear in Wilcox's bibliography. The weakness of Wilcox's case is that he appealed to a Qumran Psalm for support, whereas reference to the context of that Psalm contradicts his argument.

^{page 106 note 1} Comm. p. 110.

^{page 107 note 1} Bruce, , Comm. p. 109.Google Scholar

^{page 107 note 2} Semitisms…, p. 141; see also pp. 139–41 for a summary of other suggestions. It is worth noting that would relate closely to a part of the suggested seder reading, Exod. iii. 7.

^{page 107 note 3} Acts xv. 5.

^{page 108 note 1} See p. 101, n. 2 above. Hillel in fact failed to convince the beneBathyra by his exegesis of scripture, and only won the day when he appealed to what his teachers, Shemaiah and Abtalion, had taught him. The quotation in Acts is Amos ix. 11–12 (Acts xv. 16–18). The form of the quotation is again curious. It exists in three parts: 16a which is quite different from LXX, the main part of the quotation which follows LXX, and the last three words which are additional (and which some texts of Acts tried to eliminate). C. Rabin has pointed out that LXX represents the Massoretic Text far better than the variant form in Acts xv. 16a, but that the form in 16a represents the form of the Hebrew text in CD vii. 16 (The Zadokite Documents (Oxford, 1954), p. 29). Wilcox took the argument further by showing that the agreement between Acts and CD is not a coincidence, since 4QFlor. has the same form. Wilcox concluded: ‘The fact that the balance of the quotation is largely from the LXX in Acts may suggest that here an originally circulating element has been adapted somewhat to a new context’ (Semitisms…, p. 49). But Wilcox took no account of the variant form at the end of the quotation, which means that in Acts there is a LXX quotation in a non-LXX frame. What that suggests is that the original quotation was not made from the LXX, that it was first recorded in abbreviated form (from… to; it was quite common in rabbinic Judaism to identify long quotations by a few words selected from them), and that the abbreviation was later filled in by straight quotation from LXX. It has sometimes been argued that LXX is necessary for the argument in Acts xv, but that is not so. Massoretic Text says that the rebuilt house of David will possess the nations, LXX says that the rebuilt house of David will be the object of men's search. In either case they will be included in it, and either, therefore, would support James' argument.

^{page 108 note 2} A taqqanah is an amendment or alleviation of Torah. Zeitlin, S. has written a series of articles on early taqqanot. For references see ‘The Takkanot of Rn. Johanan ben Zakkai’, J.Q.R. LIV (1964), pp. 288–310.Google Scholar

^{page 109 note 1} For details of this vocabulary, see Bacher, W., Die Exegetische Terminologie… (1905, now in photolitho repr.).Google Scholar

^{page 110 note 1} Rom. iii. 28.

^{page 110 note 2} Gen. xv. 6.

^{page 110 note 3} The text was a crucial one in the doctrines of zekut and zekut abot, which were much debated and argued at this time (see Marmorstein, A., The Doctrine of Merits in Old Rabbinical Literature (London, 1920)). It is worth noting that the Palestinian targum-tradition interpreted the verse in terms of ‘merit’: Targ. Onq.: ; Ps. Jon.: .Google Scholar

^{page 111 note 1} Since the time when this article was written, the second volume of Jacob Mann's work on the old Palestinian Lectionary has been published (The Bible as Read and Preached in the Old Synagogue). It was assembled from Mann's notes by I. Sonne, but unfortunately Sonne himself died before he could finish his Introduction. Nevertheless, the Introduction, even in its unfinished form, is a valuable survey of reactions to Mann's work, and to subsequent work in the same field.