CHRISTOPHER M. TUCKETT, Synoptic Tradition in the Didache
SYNOPTIC TRADITION IN THE DIDACHE
CHRISTOPHER M. TUCKETT
Manchester In recent years a substantial body of opinion has agreed in general terms about the problem of synoptic tradition in the Didache. The three studies of R. Glover, H. Koster and J.-P. Audet which appeared in the 1950's, apparently independently of each other, argued that the presence of synoptic tradition in the Didache was not to be explained by direct dependence of the Didache on the finished synoptic gospels.! Rather, the Didachist used prior traditions available both to himself and to the synoptic evangelists.
This view has been endorsed in recent years by Rordorf, Tuilier, Kloppenborg and Draper in various studies devoted to the Didache and by other scholars in passing.2 Some have been more specific about the nature of the common traditions: for example,
Glover argued that the synoptic Sayings Source Q was one source of the Didache's synoptic tradition.3 It is true that there have
! H. Késter, Synoptische Uberlieferung bei den Apostolischen Vatern (Berlin:
Akademie-Verlag, 1957 [TU 65]) pp. 159-241; J.-P. Audet, La Didaché: Instructions des apétres (Paris: Gabalda, 1958) pp. 166-186. R. Glover, "The Didache's
Quotations and the Synoptic Gospels", NTS 5 (1958) pp. 12-29; cf. also his "Patristic Quotations and Gospel Sources", NTS 31 (1985) pp. 234-251.
2 See W. Rordorf & A. Tuilier, La Doctrine des douze Apétres (Paris: Les Editions du Cerf, 1978 [SC 248]); J.S. Kloppenborg, "Didache 16:6-8 and Special
Matthean Tradition", ZNW 70 (1979) pp. 54-67; W. Rordorf, "Le probléme-de la transmission textuelle de Didach@ 1,3b-2,1" in FF.
Paschke (ed.), Uberlieferungsgeschichtliche Untersuchungen (Berlin: Akademie-Verlag, 1981 [TU
125]) pp. 499-513; J. Draper, "The Jesus Tradition in the Didache" in D. Wenham (ed.), Gospel Perspectives V: The Jesus Tradition outside the Gospels (Sheffield:
JSOT, 1985) pp. 269-287; also his A Commentary on the Didache in the Light of the Dead Sea Scrolls and Related Documents (Ph.D. Dissertation, Cambridge
University, 1983); cf. also D. A. Hagner, "The Sayings of Jesus in the Apostolic
Fathers and Justin Martyr" in Wenham (ed.), Gospel Perspectives V,op.cit., pp.
233-268, on p. 241f (though this represents a change of view from Hagner's earlier position: see his The Use of the Old and New Testament in Clement of Rome
{Leiden: Brill, 1973 (NT.S 34)] p. 280: "It seems clear enough, however, that the Gospel of Matthew is used in the Didache.") Thus in a recent dictionary article,
Tuilier summed up the present state of opinion with the words: "Die neuere Forschung ist sich darin einig, da8 die Didache als Ganzes die Schriften des Neuen
Testaments in ihrer heutigen Gestalt nicht kennt" (A. Tuilier, art. Didache in TRE
8 [1981] p. 735).
3 Glover, "Didache Quotations", passim esp. p. 29. (In his later article, Glover seems to have changed his views slightly; he now postulates another source [the "Terse Source"] known to Matthew, Mark, Luke, Clement, Didache, Polycarp and Justin, which overlapped extensively with Q. However, he still maintains his view that the Didache knew Q: see p. 239). Cf. also Rordorf & Tuilier (Doctrine, p. 91), who say that the silence of S. Schulz in his monograph on Q in omitting any reference to Did is "regrettable."
SYNOPTIC TRADITION IN THE DIDACHE
always been those who have disagreed, arguing that the Didache presupposes the finished form of the synoptic gospels, or least that of Matthew.* The conclusions of this paper are intended to support this latter view.
Any discussion of the problem of synoptic tradition in the Didache must take note of the question of the unity of the text usually referred to as "the Didache". This text is available to us in its entirety in only one 11th century Greek MS published in 1883 by P. Bryennios.* It is almost universally agreed that the present text® is, in some sense at least, "composite".
Didache 1-6 incorporates an earlier Two Ways tradition attested also in the Epistle of Barnabas 18-20, Doctrina Apostolorum and elsewhere; further, within this Two Ways tradition, the section 1:3-2:1 is probably a secondary, Christianising addition.?
Other seams within our text have been suggested: for example, chapters 8 and 15 may be secondary additions to an earlier Vorlage.8 The precise number of stages of redaction which one should postulate is much debated. Nevertheless it is clear that any theories about the origins of synoptic tradition in one part of the Didache will not necessarily apply to the Didache as a whole.® Each part of the text must therefore be examined separately and, to a certain extent, independently.
4 See E. Massaux, The Influence of the Gospel of Saint Matthew on Christian Literature before Saint Irenaeus Ill, trns. N. J. Belval & S. Hecht (Macon,
Georgia: Mercer University Press, 1993 [NGSt 5/3, trns. BETL 75]) pp. 144ff; B.C. Butler, "The Literary Relations of Didache, Ch. XVI", JTS 11 (1960) pp.
265-283; also "The "Two Ways' in the Didache", JTS 12 (1961) pp. 27-38; P.
Vielhauer, in E. Hennecke (ed.), New Testament Apocrypha I (London: Lutterworth,
1965) pp. 628f; B. Layton, "The Sources, Date and Transmission of Didache 1.3b-2.1", HTR 61 (1968) pp. 343-383; S. Giet, L'Enigme de la Didaché
(Paris: Ophrys, 1970 [PFLUS 149]); K. Wengst, Didache (Apostellehre), Barnabasbrief, Zweiter Klemensbrief,
Schrift an Diognet (Darmstadt:
Wissenschaftliche Buchgesellschaft, 1984 [SUC 2]) pp.25-30; U. Luz, Matthew 1-7: A Continental
Commentary (Minneapolis: Fortress, 1989), p. 93. W. D. Kohler, Die Rezeption des Matthdusevangeliums in der Zeit vor Irendus (Tiibingen: Mohr, 1987 [WUNT
2/24]) pp. 19-56. This was.certainly the dominant view in the earlier period of research, prior to the 1950's.
5 Other witnesses include versions in Ethiopic, Georgian and Coptic, a fragment from one of the Oxyrhynchus papyri, POxy 1782, as well as later writers who used the Didache, notably the author of Book VII of the Apostolic Constitutions.
6 By this I refer to the text of the Bryennios MS.
7 For this section, see part III below.
8 Rordorf & Tuilier, Doctrine, pp. 36, 63; cf. Draper, Jesus Tradition, p. 271.
For doubts about the reliability of the Bryennios text, see too E. Petersen, "Uber einige Probleme der Didache-Uberlieferung", in Frihkirche, Judentum und Gnosis
(Rome: Herder, 1959) pp. 146-182. Cf. also n. 82 below.
9 Thus, for example, Késter argues that most of the Didache is independent of the synoptic gospels, but that 1:3-2:1 presupposes our gospels and represents a much later addition to the Didache.
CHRISTOPHER M. TUCKETT
One further preliminary point to be made is that it is quite clear that, for the most part, the Didache does not "quote" the synoptic tradition. There are a few instances where the Didache clearly indicates its intention to quote something (from whatever source): cf. Did 1:6; 9:5; 16:7. Elsewhere there are references to a evaryyénov (8:2; 11:3; 15:3, 4) which may be a written source.!° However, the remaining links between the Didache and synoptic tradition are at the level of allusion only. It is thus inappropriate to judge the Didache's use of synoptic tradition as if it were a case of explicit quotation and to expect exact agreement between the quoted version and the source used. The Didache's use of synoptic tradition is more one of free allusion. Hence disagreements between the Didache and the gospels in, for example, the context and application of synoptic tradition need not imply that the Didache cannot have known our gospels.!! Indeed it can be argued that precisely such freedom in the use of synoptic tradition is to be expected if the Didache is using our gospels as, in some sense, authoritative texts.12 How then can one determine whether or not the Didachist presupposes the finished gospels?
In terms of methodology, Ké6ster's approach remains exemplary. Not only did he set out the relevant texts clearly in 10 For a discussion, see Késter, Uberlieferung, p. 10f. It is a key point of Wengst's argument against KGster that ebayyédvov must refer to a written gospel and that gospel can only be Matthew: hence all the verbal similarities between the Didache and the synoptic tradition are due to dependence on Matthew. This however demands that the references to a evaryyéAcov, especially in 15:3, 4, bear a lot of weight in the argument: it ignores the possibility that ch. 15 may belong to a later stratum within the present text. It also does not deal with the possibility that the Didache might be using a written source which lies behind Matthew, rather than
Matthew's gospel itself. Wengst does not discuss whether the parallels to the Didache in Matthew are redactional or traditional.
11 This applies especially to the work of Glover (m. 1 above) who frequently argues that the Didache cannot be dependent on our gospels because the same material is used in such widely differing ways. (Glover even speaks of "the Didache's Quotations" in the title of his article). In any case, the different contexts imply freedom and change by at least one person, be it the Didachist or the synoptic evangelist. But the different contexts and uses of the same material cannot really be used in the argument here. Much the same can be said about Draper who argues that any difference between the Didache and the gospels may imply that the Didache is using independent traditions, since the tendency in the MS tradition was to harmonize the text of the Didache with that of the gospels ("Jesus Tradition", p.
271 and cf. his argument on Did 16:8 and the text of Did 1:3 discussed below). But again this seems to assume that the Didache's practice is one of careful quotation, rather than alluding to, and using freely, the synoptic tradition. Wengst's comment is apt: "Nach diesem Argumentationsmuster miifte man etwa Paulus die Benutzung des AT absprechen" (Didache, p. 30).
12 Cf. F.E. Vokes, "The Didache and the Canon of the New Testament" in Studia Evangelica 3 (Berlin: Akademie-Verlag, 1964 [TU 88]) pp» 427-436.
SYNOPTIC TRADITION IN THE DIDACHE
parallel; he also worked with a clearly defined criterion: if material which owes its origin to the redactional activity of a Synoptic evangelist reappears in another work, then the latter presupposes the finished work of that evangelist. Such a criterion must of course be applied with care, and one must not rule out the possibility that a feature could have been added to the tradition by two independent redactors.
Nor should one assume that any dependence which is established on the basis of this criterion is necessarily direct: the later document may be several stages removed from the earlier one. Nevertheless, this criterion is really the only one which ultimately can determine whether a text like the Didache presupposes the finished gospels or whether it uses traditions which lie behind our gospels. Such an approach differs from that of, say, B. Layton who, in his study of Did 1:3-2:1 (cf. n. 4 above) effectively assumed dependence and sought to clarify the Didache's redactional activity on that assumption.
This certainly has great value in recognizing the importance of the Didachist's own intentions and contributions in producing his text.
But Layton's approach is open to the criticism that what is "coherent" does not necessarily reflect historical fact. Thus, whilst Layton has shown how a later writer could have altered the synoptic gospels in the way suggested, his argument does not really show that this actually happened.
; One must therefore supplement Layton's approach by a close analysis of the synoptic parallels to the text of the Didache to see if any elements are redactional there. Part of the weakness of the approach of some of those who have advocated the independence of the Didache from the synoptic gospels is that they fail to offer any sustained analysis of the synoptic data on its own terms. (This applies, for example, to the studies of Rordorf and Draper). If, for example, the Didache is using Q material in a form that existed prior to its use by Matthew and Luke, then the synoptic parallels to the Didache should all occur in elements which are preredactional in the gospels. Conversely, if the Didache shows agreement with redactional elements in the gospels, then this will be an indication that the Didache presupposes our gospels.
I. DIDACHE 16
Did 16 is widely recognized as containing a significant cluster of links with synoptic tradition and it may be regarded as an important test case in the discussion of the problem of synoptic tradition in the Didache.
13 Cf, the praise of Layton, "Sources", p. 241; see too n. 102 below.
CHRISTOPHER M. TUCKETT
Many who have argued against any dependence of the Didache on Matthew's gospel have appealed to a peculiar pattern in the parallels between the Didache and the synoptic gospels. It is said that Did 16 only shows links with material peculiar to Mt 24 in the synoptic tradition: the Didache does not have any links with material from Mt 24 which Matthew has derived from Mark.'4
Hence, it is argued, the Didache is more likely to be dependent on the source(s) which lie behind Mt 24 and which were available to Matthew alone; if the Didache were dependent on Matthew, one would expect some of Matthew's Markan material to be reflected as well.!5 Such an argument is in danger of ignoring some of the evidence of Did 16 itself. For the text of Did 16 contains possible allusions to synoptic material in four verses common to Matthew and Mark (and which Matthew presumably derived from Mark).
Did 16:4
Did 16:4 mentions the xoduotAavic who, it is said, will do (mo.joer) signs and wonders (onpeta Kai TépaTa) and will perform iniquities & ovdémoTe yéyovev && ai@voc. The language used is similar to that of Matthew and Mark. Mk 13:22/Mt 24:24 refers to the coming of false messiahs and false prophets who will perform!* onpeia Kai TépaTa, "leading astray" !7 the elect; and in Mk
13:19/Mt 24:21 the coming tribulation is said to be such as never has been (ov yéyovev) since the creation of the world.!8 It can be argued, of course, that these parallels are not very significant.
Both the Didache and the synoptists could be reflecting standard eschatological motifs and using OT language.19 However, it is
14] am assuming through the discussion the theory of Markan priority, as indeed do most other scholars in their discussion of synoptic tradition in the Didache. This applies even in the articles of Butler. Butler himself is well-known as favouring the theory of Matthean priority, but he wrote his articles on the Didache taking care not oI presuppose any particular solution to the synoptic problem, though see nn. 63, below.
1S See Glover, "Didache's Quotations", pp. 22-25; Késter, Uberlieferung, p.
184f (on Did 16:6); Audet, Didache, p. 182; Rordorf & Tuilier, Doctrine, p. 90; Kloppenborg, "Didache 16,6-8", passim; Draper, Dissertation, pp. 325f; "Jesus
Tradition", p. 283.
16 Mark movjoovoy, Matthew dacovo.v. The Didache is in fact marginally closer to Mark here, though this verbal agreement is scarcely significant enough to show anything.
17 Mark &totAavar, Matthew tavioou—hence the same root as the Didache's
KOOLOTAQV
AC.
'8 Didache's é£ ai&vog is not verbally the same as Mark's / Matthew's phrase an' apxic kTicewco Kéopov, but the idea is the same.
19 The language is close to Dt 13:2 and Dan 12:1 ©': see Koster (Uberlieferung, p. 182) who also refers to 2 Thess 2:9; Rev 13:13.
Glover ("Didache's Quotations", p. 24) refers to the differences between the Didache and the gospels: e.g. in the Didache the single "world-deceiver" performs the signs and wonders,
SYNOPTIC TRADITION IN THE DIDACHE
clear that the verbal links between Did 16 and Mt 24 are not confined to material peculiar to Matthew.
Késter recognizes these parallels but asserts that, since Mk 13:19, 22 come from the Vorlage used by Mark in Mk 13, the Didache here shows links only with Mark's source and not with Mark's gospel itself.2° This argument raises a number of critical problems. First there is a question about criteria. The argument here is that if the Didache shows links with Mark's source, it cannot be dependent on Mark's gospel. Elsewhere, however, Késter argues that since the Didache does not have allusions to material in Matthew's source (Mark), the Didache cannot be dependent on Matthew's gospel.?! Clearly there are dangers of possibly contradictory criteria being applied and it would be better to consider the whole pattern of agreements between the Didache and the synoptic material in relation to synoptic sources before making premature judgments about individual verses. Second, the appeal to a Vorlage of Mk 13 raises complex issues of Markan study which can only be touched on briefly here. Késter appeals almost exclusively to the claims of Bultmann and Klostermann about the content of Mark's alleged source here. In recent years there have been several important studies of Mk 13,22 though not all would agree in assigning the relevant verses in this discussion to Mark's source. V. 19 probably is as likely as any verse to have been in whatever Vorlage there might have been, but opinions differ about yv. 22. According to Pesch, Hahn and others, v. 22 is part of the pre-Markan source. According to Brandenburger, the verse is one of a number which come to Mark from traditions other than that of the basic apocalypse which underlies the present chapter, so that although the wording may be pre-Markan, the positioning of vy. 22 here may be due to MkR.?3 However, Lambrecht and Laufen have argued very persuasively that v. 22 is due to MkR.?4 whereas in the gospels it is the (many) false prophets. However, this is an example of confusing quotations and allusions: cf. n. 11 above.
20 Uberlieferung, op.cit., p 182.
21 Uberlieferung, pp. 184f, on Did 16:6.
22 To mention only a few, see J. Lambrecht, Die Redaktion der MarkusApokalypse. Literarische Analyse und Structuruntersuchung (Rome: Pontifical Biblical Institute, 1967 [AB 28]); R. Pesch, Naherwartungen. Tradition und Redaktion in Mk 13.
(Dusseldorf:
Patmos, 1968) and his Das Markusevangelium II.
Teil (Freiburg:
Herder, 19807); F. Hahn, "Die Rede von der Parusie des Menschensohnes Markus 13" in R. Pesch & Schnakenburg (eds.), Jesus und der Menschensohn (Fs. Anton Vogtle), Freiburg: Herder 1975) pp. 240-256; E. Brandenburger Markus 13 und die Apokalyptik (Géttingen: Vandenhoeck & Ruprecht, 1984
[FRLANT 134)).
; 23 Brandenburger, Markus 13, pp. 147ff and see his "Ubersicht" on p. 166f. The problems of assigning v. 22 to the Vorlage are discussed on p. 24.
24 Lambrecht, Redaktion, p. 170f; R. Laufen, Die Doppeltiberlieferungen der Logienquelle und des Markusevangeliums (Bonn: Hanstein, 1980 [BBB 54]) pp.
379-382. V. 22 is of course also very important as evidence of Mark's concerns in
CHRISTOPHER M. TUCKETT
It is thus precarious to build too much on the alleged pre-Markan nature of v. 22.
Did 16:5
Another possible link between the Didache and Markan material occurs in Did 16:5: of 58 bropeivavtec...cwOjoovTa: cf. Mk
13:13/Mt 24:13; 10:22 6 58 bropeivac sic TédoG...cwOjoeTaL.
Again it can be argued that the parallel is not by itself very significant.25 Késter again ascribes it to Mark's Vorlage and argues, somewhat strangely, that since Matthew's wording is dependent here on Mark, this cannot prove dependence of the Didache on Matthew.?¢ By itself, such an argument is true though it appears to concede in passing that the Didache does have verbal links with material in Matthew which Matthew derives from Mark. The language is not unusual in such an eschatological context: cf. Dan
12:12; 4 Ezra 6:25,27 though the verbal agreement between these texts and the Didache is not as close as that between the Didache and Matthew/Mark.?8 The problem of the origin of Mk 13:13 is a complex one. Many regard the verses as pre-Markan and it is certainly not easy to point to any clear Markan characteristics.?9
But whatever its origin, it is clear that Did 16:5 provides another instance where the Didache shows verbal links with material which Matthew shares with Mark.
Did 16:8
The most significant connection between the Didache and material in both Matthew and Mark occurs in Did 16:8 (7é7e GWeTaur 6
KOOMOG TOV KUpLoV EpxXdpEevoy ETaVW TOV VedEehGv TOD ovpavod). the theory of T.J. Weeden, Mark--Traditions in Conflict (Philadelphia: Fortress,
1971) esp. pp. 72ff: "Mark's hand is most visible in the formation of sections 5-6,
9-13, 21-23, 28-37" (p. 72).
25 Glover does not even mention it.
26 Uberlieferung, p. 183.
27 So Késter, ibid; Kloppenborg, "Didache 16,6-8", p. 66.
28 Despite Késter's claim that 4 Ezr is "almost word for word the same as Mk 13:1b par" (my translation), Dan 12:12 @ has 6 dropévwy but no exact parallel to owOjoetrar. 4 Ezr 6:25 has omnis qui derelictus fuerit...saluabitur but derelictus is rather weaker than the meaning of iaouévw here which probably has a meaning of more active endurance and is not just a reference to simply surviving (so Bauer's
Lexicon, ad loc.) 29 See R. Kiihschelm, Jiingerverfolgung und Geschick Jesu (Klosterneuburg:
Verlag Osterreichisches Katholisches Bibelwerk, 1983 [OBS 5]) p. 22. The verse is in Mark's Vorlage according to Pesch and Hahn. However, Brandenburger assigns it to other traditions included here by Mark.
SYNOPTIC TRADITION IN THE DIDACHE
The allusion to Dan 7:13 here is very similar to that in Mt 24:30
(cf. Mk 13:26). The very close relationship between the Didache and Matthew has often been noted. In particular the Didache shares with Mark and Matthew the use of SWerau / SYovrat and the inversion of the order of "coming" and "clouds" as compared with Dan 7. Further, the Didache agrees with Matthew's redaction of Mark in using éra&vw (Matthew ézi) for Mark's év with the "clouds", and adding tod ovpavod. A priori there is a strong case here for seeing the Didache reflecting MattR of Mark and hence presupposing Matthew's finished gospel.
Glover explains the agreement as due to "joint borrowing from Dan.vii 13",3° but he offers no explanation for the unusual features (GWeror and the relative order of the "clouds" and "coming") mentioned above.
Késter claims that Matthew is unlikely to have changed the text of Mark in the way suggested and hence Matthew's text is based on a Vorlage of Mark which read éni tOv vededdv.3! However the basis for such a claim is unconvincing. Kloppenborg, who is also defending the thesis of the independence of the Didache and Matthew, rightly criticizes
Koster here and says that there is no justification for postulating an earlier version of Mark's text used by both the Didache and Matthew.32 Yet Kloppenborg also denies that the Didache shows any dependence on MattR here. He appeals to the lack of any reference in the Didache to the signs of heaven of Mt 24:29; however, the Didache does have its own version of (admittedly quite different) signs in 16:6. Kloppenborg also appeals to the fact that the words which follow the, allusion to Dan 7 in Matthew/Mark, peta duvapews Kai 56€no moAAFC, do not appear in the Didache, and asserts that "there is no reason for the author's avoidance of this phrase."23 However, an argument from silence here is very precarious. It is universally agreed that the text at the end of the Didache here is in some disarray and that some further text has probably got lost.34 It is thus very dangerous to base any theory on the absence of material after Did 16:8 in our Greek MS.
Kloppenborg's claim is that "Did 16:8 agrees with Mt 24:30 at those points where Matthew disagrees with Mark";?5 this however
30 "Didache's Quotations", p. 24.
31 Uberlieferung, pv. 188.
32 "Didache 16,6-8", p. 61f.
33 Ibid, p. 63.
34 See Audet, Didache, p. 73f; Rordorf & Tuilier, Doctrine, pp. 107, 199; Draper, Commentary, p. 326; Wengst, Didache, p. 20; Wengst even prints (p. 90) a further clause as the ending of the text (amodobvou éxckoTw KaTa& Thy mock airod) on the basis of the text in Const, though the latter is considerably longer and more complex than this.
35 "Didache 16,6-8", p. 63.
CHRISTOPHER M. TUCKETT
ignores the agreement between Matthew and Mark in using SWovrou and in inverting the "clouds" and the "coming" of Dan
736 as well as the features common to the Didache, Matthew, Mark and Daniel. Kloppenborg's conclusion is that "Did 16:8 represents an independent tradition under whose influence Matthew altered his Markan source, namely by substituting éai for év and adding 700 ovpavod" 37 However, a much simpler explanation is available. Matthew's differences from Mark here serve to align his version of Dan 7:13 with that of the LXX. A tendency by Matthew to conform OT allusions to the form of the LXX is welldocumented.38 The "tradition under whose influence Matthew altered his Markan source" need only be the LXX text of Dan 7.
There is no need at all to postulate a tradition very closely parallel to Mk 13, but independent of Mark and known only to Matthew.
Such a theory is a totally unnecessary complication.
A different appeal to textual criticism is made by Draper.
Draper refers to the text of Const VII which reads év ovocetop@ as the equivalent of the Bryennios MS's 767e dWeTou 0 Kdopoc TOV
Koptov épxépmevov. He argues that, since Const has a tendency to conform the text of the Didache to that of Matthew's Gospel, this reading in Const which differs from Matthew's text may represent a more original version of the text of the Didache; the reading in the Bryennios MS would then be the result of secondary scribal assimilation to the text of Matthew.39
All this is somewhat speculative. It is not the case that the text of Const is uniformly closer to Matthew than that of the Didache.*° Further, Draper's suggestion that the use of xtpioc in the Bryennios text betrays the vestige of an independent tradition (Matthew has "Son of Man") is also unconvincing. The xipioc of 16:8 in the Bryennios MS presumably arises by assimilation to the xtpio¢c of 16:7 (citing
Zech 14:5). In fact the use of xipiocg is quite characteristic of the Didache (it is used 18 times elsewhere, at times clearly referring to Jesus, cf. 8:2). Hence it is quite easy to ascribe the usage of koptocg in the Bryennios text at 16:8 to the "original author" of the 36 Kloppenborg argues that both these features may be pre-Markan since they can be paralleled elsewhere (though not all his evidence is equally valid e.g. the text of Justin Dialogue 51:9 [wrongly cited as "15,9" consistently] could be due to influence of the gospel tradition). But again this excludes a piece of evidence prematurely. The fact that the features may be pre-Markan cannot settle the issue of how they are to be explained in the Didache itself.
37 "Didache 16,6-8", p. 63.
38 See K. Stendahl, The School of St. Matthew and its Use of the Old Testament (Philadelphia: Fortress, 1954) pp. 147ff; G. Strecker, Der Weg der Gerechtigkeit
(Géttingen: Vandenhoeck & Ruprecht, 1971 [FRLANT 82]) pp. 21ff W. G.
Kiimmel, Introduction to the New Testament (London: SCM, 1975) pp. 110f.
39 "Jesus Tradition", p. 283; Commentary, p. 325.
40 E.g. Did 16:4 has tapaddaovow as in Matthew; Const has xpodécovaw.
SYNOPTIC TRADITION IN THE DIDACHE
Didache. In any case, it would probably be precarious to rely too much on such a indirect source as Const for reconstructing the text of the Didache.
__ None of the arguments considered appears to provide a convincing alternative to the theory that the Didache here presupposes knowledge of Matthew's revision of Mark. Thus Did 16 has links not only with Matthew's special material, but also with material common to Matthew and Mark and, in the last instance considered, presupposes Matthew's redaction of Mark. I now consider the links between Did 16 and material peculiar to Matthew.
Did 16:3-5.
The existence of links between Did 16 and material peculiar to Mt 24 is accepted by all. In Did 16:3-5 there is a cluster of similarities between the language of the Didache and Mt 24:10-12 and other Matthean passages.
16:3 rAnOuvOjoeTar oi WevdoTpooHrat uses similar language to Mt 24:11f. (WevdorpodHrau in v. 11; mAnOvvOAvcu in v. 12); "sheep becoming wolves" in Did
16:3 uses imagery similar to that of Mt 7:15; "a&ycé7n turning to ptoog" reflects Mt 24:10, 12 (wojoovow in v. 10, &yc&rn in v.
12); &vopie increasing (Did 16:4) is similar to Mt 24:11 (@vopia multiplying); and d.4£o0va in Did 16:4 links with d.axwou in the very closely related context of Mt 10:23. Finally oxavdado- @noovTrar woddoi of Did 16:5 recalls the identical words in Mt
24:10.
It is noteworthy that the parallels to Did 16:3-5 in Matthew include the three Matthean references to WevdorpodjTa (Mt 7:15; 24:10-12; 24:24).4! This is readily explained if the author of the Didache were attempting to cull from Matthew all the available material about false prophets. This provides a reasonable explanation for what might appear at first sight to be a rather random set of parallels in Matthew.
If the existence of these parallels is universally accepted, their significance is disputed. Much depends on one's beliefs about the origin of the material in Mt 24:10-12. Methodologically, the problem should perhaps be one of synoptic study before one considers the Didache itself.42 Within Matthean scholarship there
For the parallel between Did 16:4 and Mt 24:24 see above.
Glover ("Didache's Quotations", p. 23) ascribes these references to three different sources in Matthew; but the multiple references are much more likely to be due to MattR.
42 Cf. the comments of F. Neirynck ("John and the Synoptics", in Evangelica:
Collected Essays {Leuven: Leuven University Press, 1982 (BETL 60)] p. 379) in relation to the not dissimilar problem of the relationship between John and the synoptic tradition: "The question whether John depends upon Mark or upon the sources of Mark is primarily a problem of Synoptic criticism".
CHRISTOPHER M. TUCKETT
is widespread agreement that these verses are due to MattR.* If this is the case, then it would appear to provide clear evidence of the Didache's presupposition of Matthew's redactional activity and hence of Matthew's gospel.
This conclusion has however been disputed from the side of Didachean scholarship. Some have appealed to the fact that the same words are used in different ways in the Didache and Matthew as evidence of their independence: for example, in Matthew it is the lawlessness which is multiplied in the Didache it is the false prophets.44 However, this kind of argument tends to assume that the Didache is quoting synoptic tradition, whereas in fact there is at best here only an allusion and use of common language.
Appeal is also made to the fact that much of the common 43 See J. Lambrecht, "The Parousia Discourse. Composition and Content in Mt
XXIV-XXV", in M. Didier (ed.), L'Evangile selon Matthieu (Gembloux: Duculot, 1972 [BETL 29]) p. 320: "Matthew has formed these verses himself ... He has used words and ideas which the context has offered. The result is typically Matthean".
See too G.D. Kilpatrick, The Origins of the Gospel according to St. Matthew (Oxford: Oxford University Press, 1946) p. 32; F. W. Burnett, The Testament of Jesus Sophia (Washington, 1979) p. 247; Kiihschelm, Jungerverfolgung, pp. 124f;
R. H. Gundry, Matthew. A Commentary on his Literary and Theological Art (Grand Rapids:
Eerdmans, 1982) p. 479.
Cf. the typical Mattheanisms kai Tore, oxavdanritopo (14-8-2), trAavaw (8-4-1), Yevdorpodyrn¢ (3-1-1), c&vopia (4-0-0) etc.
Very few have argued explicitly that this is traditional in Matthew. One exception is D. Wenham: see his The Rediscovery of Jesus' Eschatological Discourse (Sheffield: JSOT, 1984) pp. 256-259, and in more detail in "A Note on Matthew 24:10-12" in TynB 31 (1980) pp. 155-162. Wenham appeals to an allegedly clear chiastic structure (ABBCCA) though this is scarcely obvious (e.g. the alleged
"A" elements, v. 10a and v. 12, are parallel only in the most general terms: a quite different chiastic structure for vv.
9-14 is proposed by Kiihschelm, Jiingerverfolgung, p. 63). He claims too that the paratactic style is uncharacteristic of Matthew: however, Mt 4:23-25; 17:18 (cf. Mk 9:25); 27:11-14 all owe a lot to MattR and are equally paratactic. (Cf. also instances where Matthew adds kai to Mark, as listed by F. Neirynck, The Minor Agreement of Matthew and Luke against Mark [Leuven: Leuven University Press, 1974 (BETL 37)] p. 204). Wenham also appeals to allegedly non-Matthean vocabulary. Some of this may be due to allusions to the language of Daniel which would be consistent with an origin in MattR, but he claims that dyamn and Wvyfjoera are unexplained. (Cf. especially "Note", p. 159 n. 4, where these are said to be the most significant). However, the importance of "loving" is widely recognized as of vital importance for Matthew's ethic and hermeneutic: cf.
Mt 19:19; 22:40
(see R.
Mohrlang, Matthew and Paul [Cambridge: Cambridge University Press, 1984 (SNTS MS, 48)] pp. 94-96, with full supporting bibliography). Hence, even though it is the noun rather than the verb which is used here, it would be precarious to base too much weight on this grammatical difference.
Thus the only possible indication of non-Matthean vocabulary is the hapax Yixa, and in view of the large number of Matthean features elsewhere in these verses it seems easiest to ascribe the whole section to MattR. m See Glover, "Didache's Quotations", p. 23; Késter, Uberlieferung, pp. 178,
SYNOPTIC TRADITION IN THE DIDACHE
vocabulary here may be reflecting stock motifs, though it is certainly the case that the verbal and contextual links between Did
16 and Mt 24 are much closer than those with other texts proposed as parallels here.45
In a significant part of his argument, Késter recognises the parallels between Mt 24 and the Didache here. However, he claims that the verbal agreement is insufficient to show dependence on Matthew; rather, both depend on common tradition and indeed Did 16 itself may provide evidence that Mt 24:10-12 is a piece of pre-Matthean tradition.46 Such an argument raises at least two major difficulties. First, the measure of verbal agreement between the Didache and Matthew cannot be used to determine whether that agreement is due to direct dependence of one on the other or to common dependence on a prior source. Common dependence on a prior source does not necessarily involve less close verbal agreement.47 Second, and more important, the claim that Did 16 itself may provide evidence that Mt 24:10-12 is preMatthean is a case of petitio principii here. If the question is whether the Didache depends on Matthew's Gospel or on a preMatthean source, one cannot use the evidence of the Didache itself to solve the source problem of Matthew's text. Késter's argument is thus dangerously circular.
Did 16:6.
Parallels between Did 16:6 and Mt 24:30a, 31 are also widely recognized (e.g. the common use of ¢davioetoar, onpetov, év 45 Glover ("Didache's Quotations", p. 23) refers to &vopie occurring also in 2
Thess 2:3 (though the usage in the Didache is much closer to Mt 24 than to 2 Thess's more specialized 6 &vOpwroc Tic &vopiac). Koster (Uberlieferung, p. 173) refers to 2Bar 48:35 for the idea of love turning to hate (though neither love nor hate are explicitly mentioned there); he also appeals to ApocPet 1:1, though this text is almost certainly dependent on Matthew: see R. J. Bauckham, "The Two FigTree Parables in the Apocalypse of Peter", JBL 104 (1985) pp. 271-273. (Bauckham is discussing the Ethiopic version; the Achmim fragment, referred to by K6ster, is an even less reliable witness to the text of ApocPet and assimilation to Matthew is even more likely there. The textual tradition of ApocPet is discussed by Dr. Bauckham in a forthcoming article the contents of which he has kindly communicated to me). Késter also claims that Matthew's idea of love growing cold is secondary to Did's "kraftiger und urspriinglicher" phrase: but this is difficult to establish with any precision.
46 Uberlieferung, pp. 181, 184: "...drangt sich die Frage auf, ob nicht Did. 16,4 beweist, daB Mt. 24,10ff keine Schépfund des Mt., sondern ein selbststandiges
Traditionsstiick ist, das Mt. und Did. unabhingig voneinander verwendet haben" ep 81); . 47 Cf. the situation in discussion about the synoptic problem: close verbal agreement between Matthew and Luke does not of itself exclude the Q hypothesis.
Equally, a measure of verbal disagreement in triple tradition material does not preclude the possibility of direct dependence of one evangelist on another.
CHRISTOPHER M. TUCKETT
obpav, o&dmvyé). Again many would ascribe this material to MattR in Matthew,'8 though the limited extent makes any certainty impossible. In defending the theory of the independence of the Didache and Matthew, some have pointed to the differences between the two texts here,49 though one must again note that the Didache here is not explicitly quoting any source at this point.
The Didache develops its own idea of the three "signs"
,>° though this could well have been prompted by Matthew's "sign of the Son of Man". Several too point to the fact that Did 16:6 shows links only with material peculiar to Matthew.>! However, this is valid only if one confines attention to 16:6; in 16:8 the Didache does use material common to Matthew and Mark, as we have seen. As before, it is possible to appeal to the fact that many of these motifs may be stock apocalyptic ideas (e.g. the trumpet),>? though the connection between, for example, the eschatological trumpet and Dan 7:13, as in both Did 16:6, 8 and Mt 24:30, is not easy to attest elsewhere.
The conclusion of this section is that there is nothing peculiar in the pattern of parallels with Did 16 in Mt 24. Did 16 shows verbal links with the material peculiar to Matthew in this chapter, with material common to Matthew and Mark and with Matthew's redaction of Mark. there is little convincing evidence to show that Matthew had access to any extensive source other than Mark for this chapter. The pattern of parallels between the Didache and Matthew is thus most easily explained if the Didache here presupposes Matthew's finished gospel.
Il. OTHER TEXTS
Did 11:7.
The saying about the unforgivable sin in Did 11:7 t@oa yap apaptia abebjoeta, aiitn 58 ) Auaptia obk adeOjoeTou may provide further support for this thesis. The evidence here is very small in extent and so certainty is not possible. However, the 48 See Lambrecht, "Parousia Discourse", p. 324; Gundry, Matthew, p. 488. For a pre-Matthean origin, see Wenham, Rediscovery, pp. 318ff. Clearly Matthew is using traditional ideas and phraseology so that in that sense Matthew's language is traditional; the question is whether Matthew himself has supplied this traditional language.
49 See Glover, "Didache's Quotations", p.24f; Késter, Uberlieferung, pp. 184f.
5° On which see Draper, Commentary, pp. 319-325; also A. Stuiber, "Die drei Semeia von Didache xvi", JAC 24 (1981) pp. 42-44.
5! Glover, "Didache's Quotations", p. 24; Késter, Uberlieferung, pp. 184f; Kloppenborg, "Didache 16,6-8", pp. 64f.
52 Glover, "Didache's Quotations", p. 25; Késter, Uberlieferung, p 186.
SYNOPTIC TRADITION IN THE DIDACHE
Didache's wording agrees exactly with Matthew's redaction of Mark in Mt 12:31/Mk 3:28. (Matthew m@ouw apaprTia...dbeOjoetar, Mark ravta adeOjoerat...7& KpmapTipata). The second half of the saying in Did has been modelled very precisely on the first half. @uapria in the second half has no precise parallel in any synoptic version, though ovk addebjoerau agrees with Matthew again (Mt 12:31b, 32b also Lk 12:10b). Késter admits that the Didache is closer to Matthew than to the other synoptic versions here, but denies direct dependence in view of the lack of any significant features;53 however, any links between the Didache and Matthew's version are with redactional features in Matthew.
Glover too denies any direct dependence here and seeks to find support for his general theory that the Didache is dependent on Q.
He claims that the Didache here rejects words common to Matthew and Mark alone but not in Luke (e.g. BAawodnuia). Further,
Did shares some words with Matthew which are not from Mark (moa cpaptia); but since Matthew is carefully conflating Mark and Q here, these words are probably words from Q and omitted by Luke.*4 This is however unconvincing. It is almost universally agreed that the Q version of the saying has "whoever/everyone who speaks a word against the Son of man" in the first half>> and has nothing equivalent to the aoa apaptia of Mt 12:31a which is almost certainly MattR of Mark's mav7a 7a GQuapThyata (Mk
3:28). Hence the verbal links with the Didache in the synoptic versions occur in Matthew's redactional material, not in Q. Késter also refers to Jewish parallels for the idea of speaking against the prophetic Spirit as being the unforgivable sin.5° However, such an idea is not peculiar to non-Christian Judaism either.>7 It is thus difficult to decide the question of literary dependence just on traditionsgeschichtlich grounds.
Certainly there is nothing here to tell against the theory that the Didache is using language which is derived (perhaps at more than one stage removed) from Matthew's finished gospel.
Did 13:1.
This is probably also implied by the saying in Did 13:1, that every true prophet &£1o¢ éo7 Tho Tpodic avTod. This recalls the language of the saying in the mission charge Mt 10:10/Lk 10:7. The 53 Uberlieferung, p. 216.
54 "Didache's Quotations", p. 20.
55 See my The Revival of the Griesbach Hypothesis (Cambridge: Cambridge University Press, 1983 [SNTS MS 44]) p. 88, with further bibliographical references. Cf. Mt 12:32a/Lk 12:10a.
56 Uberlieferung, p. 216f; also Rordorf & Tuilier, Doctrine, pp. 53, 88.
; 57 Cf. Glover, "Didache's Quotations", p. 20, referring to Ignatius and Justin.
CHRISTOPHER M. TUCKETT
Didache again shows links with Matthew's version rather than Luke's. (Luke has puo0d not tpodfc). Certainty is not possible but it seems most likely that Luke's version is more original and that Matthew's tpodfc is MattR.>® Koster argues that the saying may have been a stock proverb,°° though none of the non-Christian examples he gives offers a precise parallel. He also refers to 1 Tim 5:18 and 1 Cor
9:14 as evidence that the saying must have circulated in isolation of its synoptic context. However, 1 Tim 5 may well be dependent on 1 Cor itself (at least indirectly) and so may not provide independent evidence here. 1 Cor 9 probably presupposes a version with pvodod rather than tpo¢f¢, and Paul may indeed only know the saying as an isolated one. But Paul's lack of knowledge of the synoptic context does not necessarily tell us anything about the form in which the saying was known to the Didachist. The fact that the Didache betrays links with Matthew's redacted form of the saying must provide some evidence that theDidache only knows the saying as mediated through Matthew's gospel.
Did 1:2.
I now consider the version of the double love command in Did 1:2a and the negative form of the golden rule in 1:2b. The commands to love God and one's neighbour are well-known separately in non-Christian Judaism, and at times together, notably in Testaments of the Twelve Patriarchs.*1 However, the use of ap@z7ov
...de0TEpov is not easy to parallel in non-Christian sources, and may well betray Christian influence.® In view of what has been established so far, the simplest solution is to postulate dependence on Matthew's zpa7n...devTépa in Mt 22:38f. The "first...second" formulation is not confined to Matthew: Mark also has zpary
... deuTgépa (Mk 12:28, 31) in a slightly different relation to the 38 See my "1 Corinthians and Q" in JBL 102 (1983) pp. 607-619, on p. 612; also S. Schulz, Q--Die Spruchquelle der Evangelisten (Ziirich: Theologische Verlag, 1972) p 406; P. Hoffmann, Studien zur Theologie der Logienquelle (Minster:
Aschendorff, 1972 [NTA 8]) p. 274; Laufen, Doppeliiberlieferungen, p.219.
Wengst also agrees that Did is here dependent on Mt, though he gives no justification in terms of any argument that Mt's version here is redactional: he simply a to the fact that Mt's version is different from the Lukan parallel (Didache, p.
°° Uberlieferung, p. 212; Rordorf & Tuilier, Doctrine, p. 88, appeal to Késter.
3 ee my "Paul and the Synoptic Mission Discourse?" in ETL 60 (1984) pp.
61 See R. H. Fuller, "The Double Commandment of Love: A Test Case for the Criteria of Authenticity", in Fuller (ed.) Essays on the Love Commandment (Philadelphia: Fortress, 1978) pp. 41-56.
® This is admitted even by Koster, Uberlieferung, p. 172.
SYNOPTIC TRADITION IN THE DIDACHE
two love commands themselves.%3 This tells against Glover's thesis that the Didache tends to follow Matthew only when Matthew is not following Mark and hence that the Didache is dependent on Q rather than Matthew.* There may well have been a Q version of this pericope but it is very doubtful if the [email protected] comes from such a Q version here.® It is not present in any form in Lk 10:25-28 and is thus probably in Matthew from Mark. Késter argues that the ordering and numbering of the commands is pre-redactional in the gospels and hence one need not deduce dependence on our gospels.°* This may be the case, but we have already seen that there is a lot of material in the Didache which does presuppose Matthew's finished gospel. It seems therefore an unnecessary complication to presume that this feature of the text of the Didache is not also derived from Matthew's gospel.
The version of the golden rule in Did 1:2b may also be derived from Matthew. It is true that the version given here uses the negative form which is found elsewhere in Jewish sources, though it is not found in some other sources of the Two Ways tradition (e.g. Barnabas and 1QS).67 However, Butler has shown very clearly how several unusual features of Did 1:2b can be easily explained if the Didache is based on Mt 7:15.68 Again, knowledge of Matthew's finished gospel is the simplest solution to explain Did's text.
Others.
The evidence from the rest of the Didache is less clear-cut for the present purposes. In the Lasterkatalog of Did 5:1f, Butler has shown how the order of vices in Did is explicable if the Didachist were editing a list similar to that of Barn 20:1f in the light of the list of Mt 15:19.°° Other links between the Didache and Matthew
63 Butler, "Two Ways", p.29, goes too far in claiming that all the references to the "first ... second" commandments derive from his "M(g)" (i.e. the Matthean form of synoptic tradition). Perhaps this is a case of Butler's unconsciously assuming his own solution to the synoptic problem.
64 Here, "Didache's Quotations", p. 13. Glover sees a reflection of Luke's version of the pericope in the reference to the way of "life" in Did 1:1; but this seems rather fanciful.
65 See Fuller, "Double Commandment", p. 45.
66 Uberlieferung, p. 172.
67 See Butler, "Two Ways", p. 31, arguing against Audet.
; 68 Butler, "Two Ways", p. 30. Note that again the Didache is not quoting.
Hence Glover's assertion that the Didache here "is so different that it would be hard for anyone to write it if he were intending to quote our Gospel" ("Didache's
Quotations", p. 13) is beside the point.
69 Butler, "Two Ways", p. 33.
CHRISTOPHER M. TUCKETT
are easily explained if the Didache is dependent on Matthew, but they do not demand such a theory. The saying "the meek shall inherit the earth" in Did 3:7 could derive from Mt 5:5, though position and presence of that beatitude in Matthew is textually uncertain and common dependence of the Didache and Matthew on Ps 36:11 is equally likely. The baptismal commands in Did 7 are clearly similar to Mt 28:19, though a common milieu with similar liturgical practice will explain this as well. Agreements in the versions of the Lord's Prayer in Did 8 and Mt 6 may also be explicable in the same way. It may be significant that Did 8 links injunctions about prayer (including the Lord's Prayer) with those about fasting and inveighs against opponents as "hypocrites", just as Mt 6:5-16 does. Certainly the text of Did 8 is readily explicable if the Didache knew Matthew, but perhaps does not compel such a belief. So too the saying in Did 9:5 about not giving what is holy to dogs, which is one of the few explicit quotations in the Didache, is verbally identical with Mt 7:6. However, Mt 7:6 is not clearly MattR and the saying looks very like a stock proverb;7° hence one cannot lay too much weight on this parallel.
Did. 16:1.
One final synoptic parallel should be noted. Did 16:1 ("Watch over your life: let your lamps be not quenched and your loins be not ungirded, but be ready, for you do not know the hour in which our Lord comes") recalls various synoptic parallels. The introductory yonyopeite is parallel to Mt 24:42 (MattR of Mk 13:33 though the word occurs elsewhere in Mark). The saying about the lamps and loins is close to Lk 12:35; and the saying about being
"ready for you do not know..." is close to the ending of the parable of the thief at night in Mt 24:44/Lk 12:40 and the similar saying in Mt 24:42/Mk 13:35.
Certainty is not possible here. One must again bear in mind the fact that this is not an explicit quotation but a piece of exhortation perhaps using traditional language. Thus it is not unexpected that the uses of individual words may have shifted slightly from their synoptic contexts. There is nothing here that is so clearly
MattR that it could only have derived from Matthew's gospel.7!
70 Cf. R. Bultmann, The History of the Synoptic Tradition (Oxford: Blackwell, 1968) p. 103.
71 Mt 24:44 is very closely parallel to Lk 12:40. ypryopeire in Mt 24:42 can scarcely count as significant enough to show dependence on Matthew in view of the common nature of the word, especially in eschatological contexts (so Késter,
Uberlieferung, p. 176); the parallel between the final clause of Did 16:1 and Mt
24:42 is perhaps more significant in this context though the verbal agreement is not exact (Didache has 6 kipioc huey where Matthew has 6°Képioc bua@v, MattR for Mark's 0 xtp.iog Tio olxiac). However, Bauckham has shown that a shift from the
SYNOPTIC TRADITION IN THE DIDACHE
More difficult is the question about the parallel between Did 16:1la and Lk 12:35. Some have seen this as clear evidence of the Didache's dependence on the gospel of Luke.72 Others have disagreed, arguing variously that Lk 12:35 may be Q material so that the Didache is here dependent on Q rather than Luke,73 that the imagery is stereotyped (cf. 1 Pet 1:13; Eph 6:14),74 that the verbal agreement between the Didache and Luke is not close enough to imply direct dependence,'5 or that the Didache nowhere shows knowledge of Luke's gospel and hence is unlikely to do so here.76
It must be said that none of these arguments is absolutely convincing. Certainly the last argument is a case of petitio principii: the question of the dependence of the Didache on Luke at this point is precisely what is in question and can scarcely be solved by assuming the answer initially. The degree of verbal parallelism is also not a problem provided one remembers that the Didache is not explicitly quoting here. As far as the allegedly stereotyped imagery is concerned, it is true that the use of the "girding up loins" metaphor can be paralleled elsewhere, but the conjunction of this with the "lamps" motif is harder to parallel outside these two texts.7"7 Whether Lk 12:35 belonged to Q is more debatable. synoptic version of this parable to the application as found in the Didache is thoroughly typical of early Christian use of this parable: see his "Synoptic Parousia
Parables and the Apocalypse" NTS 23 (1977) pp. 162-176, esp. 169 in relation to this passage. Bauckham himself, referring to Audet, finds it incredible that the author of the Didache could have constructed his saying with the gospels in front of him (which may well be true, but one must,remember that the Didache is not quoting here: cf. n. 11 above); however, he does say that "it is unlikely that his [the Didachist's] tradition was wholly independent of the specifically Matthean redaction" (ibid).
72 So Butler, "Literary Relations", who appeals to the parallels between Did 16:la, 1b and Lk 12:35, 40 respectively and argues that the link between Lk 12:35 and 12:40 is LkR.
73 Glover, "Didache's Quotations", pp. 21f; Draper, "Jesus Tradition", p. 280.
74 Koster, Uberlieferung, pp. 175f; Draper, ibid.; Wengst, Didache, p. 99.
75 Cf. Audet, Didache, p. 181; Késter, ibid. See too R. Bauckham, "Synoptic Parousia Parables Again", NTS 29 (1983) pp. 129-134, on p. 131, who refers to a very similar form of saying in Methodius, Symp V.2. He writes: "This agreement cannot be accidental... Methodius' text could be influenced by the Didache, or it may be independent testimony to the same non-Lukan version of the saying which the Didache quotes. In that case it would confirm the Didache's independence of our Gospels at this point". However, dependence of Methodius on Did is equally possible. Methodius' work dates from the end of the 3rd century, not the 2nd century as Bauckham claims (n. 13 on p. 133), and this was a period when the Didache must have enjoyed some popularity: cf. H. Von Campenhausen, The Formation of the Christian Bible (Philadelphia: Fortress, 1972) p. 213.
76 Késter, Uberlieferung, pp. 175f.; Rordorf & Tuilier, Doctrine, pp. 89f.
77 Parallels in j.Kil IX, 32b.9 or j.Ket Xll, 35a.9 (Draper, "Jesus Tradition", p.
280; Commentary, p. 299) are not close. Both metaphors may derive from Passover symbols (Draper, ibid.) but the conjunction of both in a single saying in still striking.
CHRISTOPHER M. TUCKETT
More recent study has suggested that whilst Lk 12:36-38 may (in part at least) derive from Q, v. 35 is more likely to be LkR."8 This would imply that the Didache is dependent on LkR material and hence presupposes Luke's finished gospel. This last conclusion is not certain in view of the very limited extent of the evidence available. Nevertheless it seems perhaps the least problematic solution.
No attempt has been made to provide a fully comprehensive analysis of all possible parallels between the text of the Didache and the synoptic gospels in this section.7? Only the parallels which are perhaps most significant for the present discussion have been noted. However, the analysis given suggests that these parallels can be best explained if the Didache presupposes the finished gospel of Matthew (and perhaps also of Luke). One section of the Didache which has not yet been considered is the small section in 1:3-2:1. In view of its peculiar problems, this section deserves a separate and detailed study.
II. DIDACHE 1:3-2:1
By almost universal consent, Did 1:3-2:1 is a passage of great complexity, posing critical problems at a number of different levels.8° It occurs at the start of a sub-section of the Didache, namely chs. 1-6, which itself almost certainly had a history prior to its incorporation into the present text of the Didache. This section comprises the teaching about the "Two Ways" in a form which is closely paralleled in other texts, in particular in the Doctrina Apostolorum and in Barn 18-20. Further, in both these texts, there is nothing corresponding to Did 1:3-2:1 although the texts run closely parallel to the rest of Did 1-6. This Two Ways tradition is almost devoid of specifically Christian features, so that a Jewish Vorlage has long been suspected. Such a theory has now received considerable support from the evidence given by the 78 See my Revival, p. 181, with further references. Certainly the link between v.
35 and vv. 36-38 looks to be secondary in view of the cai ipeic at the start of v.
36. For the theory that Lk 12:35 is a vestige of Mt 25:1-13 and part of an earlier source, see Wenham, Rediscovery, pp. 77ff. However, this seems rather fanciful: J.
A. Fitzmyer, The Gospel according to Luke X-XXIV (New York: Doubleday, 1985 [Anchor 28A]) p. 988, points to several differences between this verse in Luke and the Matthean parable and calls the theory of a link between them "far-fetched": cf. also I. H. Marshall, The Gospel of Luke (Exeter: Paternoster, 1977) p. 535.
" For a full list of possible parallels see Kohler, Rezeption, pp. 30ff.
8° This section III was not part of the paper given at the 1986 Leuven Colloquium, the text of which comprises Sections I and II here. Research for it was undertaken during a period of study in 1987 in the University of Marburg, financed by the Alexander von Humboldt Stiftung, Bonn. It is included here by kind permission of the Editor of the Colloquium volume.
SYNOPTIC TRADITION IN THE DIDACHE
Dead Sea Scrolls, since it has been shown that the Two Ways teaching in the Qumran Manual of Discipline (1 QS 3:13-4:26) shows many common features with the Two Ways tradition in the Doctrina Apostolorum, Barnabas and the Didache.®!
The small section in Did 1:3-2:1 clearly interrupts this Two Ways teaching.
The textual evidence provided by Doctrina Apostolorum and Barnabas strongly suggest that Did 1:3-2:1 is a later addition to an earlier Two Ways Vorlage. In terms of content too, the section stands out from its immediate context since the large number of parallels to the synoptic tradition in such a short compass of text contrasts strongly with the rest of Did 1-6 where explicitly Christian features are so strikingly absent. Thus Did
1:3-2:1 is generally accepted as a later, Christianizing addition to an earlier Vorlage lying behind Did 1-6.
Whether Did 1:3-2:1 is secondary in relation to the rest of Did 1-16 is another matter. The parallels to Did 1-6 suggest that the latter may have been taken over from an earlier, Jewish Vorlage by the Didachist to form the start of the present work. Hence the fact that 1:3-2:1 is a secondary addition to the Vorlage of Did
1-6 cannot in itself determine whether the section was added before, at the same time as, or after the time when the rest of Did
1-6 was combined with chs. 7-16.82
Recent studies of the problem of synoptic tradition in this section of the Didache have reached widely differing results. Contrary to his general conclusions about the lack of dependence on the synoptic gospels by many of the apostolic fathers (including the rest of the Didache), Koster 'argued that Did 1:3-2:1 does presuppose the finished gospels of Matthew and Luke (perhaps via a post-synoptic harmony).83 The same conclusion was proposed in the detailed analysis of this section by B. Layton.84 However, other recent studies of this section have argued in detail that the text of the Didache here is not dependent on the synoptic gospels and that the Didache represents a line of the tradition which is independent of the synoptic evangelists.2° Draper is even more
81 See J.-P Audet, "Affinités littéraires et doctrinales du 'Manuel de discipline'",
RB 59 (1952) pp. 219-238; Rordorf & Tuilier, Doctrine, pp. 22ff; Draper, "Jesus Tradition", p. 270.
82 See Rordorf, "Probléme" (n. 2 above) pp. 499f. However, Wengst is so convinced that the section is a later addition to the text of the original Didache that he assigns it en bloc to a footnote in his edition of the text (Didache, p. 66).
83 Uberlieferung, pp. 217ff.
84 Cf. n. 4 above For others supporting the dependence of the Didache here, see Massaux, Influence Ill, pp.
150-157:
Butler, "Two Ways", p. 31; Wengst, Didache, p. 19: F. Neirynck, "Paul and the Sayings of Jesus" in A. Vanhoye (ed.)
L'Apétre Paul (Leuven: Leuven University Press, 1986 [BETL 73]) pp. 265-321, on p. 298f.
8s See especially Rordorf, "Probléme", and Draper, "Jesus Tradition"; also Audet and Glover (as in n. 1 above); see too H. T. Wrege, Die Uberlieferungsges-
sip?
CHRISTOPHER M. TUCKETT
specific about the point at which the trajectory which leads to the text of Did 1:3-2:1 diverged from the synoptic trajectories: "It would seem a more likely inference that Didache had access directly to the so-called "Q" material, either in a written or an oral form".8° Such a theory needs rigorous testing especially in the light of contemporary interest in Q.
Part of the complexity of the section 1:3-2:1 arises from the text itself. Despite its small compass, the section may well have had a complex tradition-history: editorial comments, and possible textual corruption, have long been suspected.8? However, this is not the only factor which makes the problem of synoptic tradition in this section so difficult. The parallels in the synoptic gospels are mostly confined to one small section of the Great Sermon, viz., the sayings about non-retaliation and love of enemies in Mt
5:39-48/Lk 6:27-36. This is a Q passage which is itself one of notorious complexity. The section has been analyzed in great detail many times in recent years (usually without reference to the Didache, though see nn. 93, 117 below).88 There is no absolute certainty about where precisely tradition and redaction are to be chichte der Bergpredigt (Tiibingen: Mohr, 1968 [WUNT 9]); H. Conzelmann, art. xe&pic, TDNT IX, p. 400; Kéhler, Rezeption, pp. 46f. (KGhler is thus rather different from Késter in advocating dependence on Matthew by the Didache elsewhere, but listing this passage as one where "Mt-Abhangigkeit eher unwahrscheinlich ist" [p. 421]. However, he does not make clear how he thinks the "Herrenwortsammlung", which he believes to be the immediate source used here, is related to the gospel of Matthew). It may be noted too that sometimes the independence of the Didache is assumed and then used to solve traditio-historical problems within the synoptic tradition: see nn. 93, 117 below.
86 "Jesus Tradition", p. 279.
87 See Layton, "Sources", and Rordorf, "Problame", passim, also n. 111 below.
88 To name but a few, see H. Schiirmann, Das Lukasevangelium (Freiburg: Herder, 1969 [HTKN 3/1]); Wrege, Bergpredigt; S. Schulz, Q; D. Liihrmann, "Liebet eure Feinde (Lk 6,27-36/Mt 5,39-48)", ZTK 69 (1972) pp. 412-438; D. Zeller, Die weisheitlichen Mahnspriiche bei den Synoptikern (Witzburg: Echter-Verlag, 1977
[FzB 17]); H. Merklein, Die Gottesherrschaft als Handlungprinzip. Untersuchung zur Ethik Jesu (Wiirzburg:
Echter-Verlag, 1978 [FzB 34]); G.Strecker, "Die Antithesen der Bergpredigt", ZNW 69 (1978) pp. 36-72, and Die Bergpredigt: Ein
Exegetischer Kommentar (Gottingen: Vandenhoeck & Ruprecht, 1984); J. Piper, Love Your Enemies. Jesus' Love Command in the Synoptic Tradition and the Early
Christian Paraenesis (Cambridge: Cambridge University Press, 1979 [SNTS 38]); R. A. Guelich, The Sermon on the Mount (Waco: Word, 1982); P. Hoffmann,
"Tradition und Situation. Zur 'Verbindlichkeit' des Gebots der Feindesliebe in der synoptischen Uberlieferung und in der gegenwartigen Friedensdiskussion", in K.
Kertelge (ed.), Ethik im Neuen Testament (Freiburg: Herder, 1984), pp. 50-118; J.
Sauer, "Traditionsgeschichtliche Erwagungen zu den synoptischen und paulinischen
Aussagen tiber Feindesliebe und Wiedervergeltungsverzicht", in ZNW 76 (1985) pp.
1-28; D. R. Catchpole, "Jesus and the Community of Israel--The Inaugural Discourse in Q" in BJRL 68 (1986) pp. 296-316; J. Schlosser, Le Dieu de Jésus (Paris:
Les Editions du Cerf, 1987 [LD 129]), and many others.
SYNOPTIC TRADITION IN THE DIDACHE
located here. The fact that this is a Q passage means that (on the basis of the Two Source theory) either Matthew or Luke could preserve the Q wording at any point and there is no clearly definable criterion for determining which version is original. Further, the version which is deemed not to preserve the Q original at any one point is not thereby guaranteed to be redactional. The "Q" available to Matthew and Luke could have been expanded independently in the pre-redactional tradition, so that one might have to think in terms of a Q™ and a Qk,
Nevertheless, despite the uncertainty inherent in the synoptic analysis, the method of trying to identify redactional elements in the gospel, and then comparing these with the text of the Didache, remains the best way of determining whether this section of the Didache presupposes the finished gospels or not. (Since this paper is concerned primarily with synoptic tradition in the Didache, attention will be confined here to Did 1:3-5a and the parallels in the synoptic gospels dealing with love-of-enemies and nonretaliation. Other problems associated with this passage, e.g. the parallel between Did 1:5 and HermMand IV.2 and the parallel between Did 1:6 and Sir 12:1 will therefore not be discussed here).
For the sake of convenience, the relevant section of the Didache is divided into six smaller sub-sections which are examined in turn.
Did 1:3a Mt 5:44
Lk 6:27-28 MayATATE TOUS
2TQyATOATE TOUS
ExOpovds vuav ExOpovc Lav Kad@>
TOLELTE TOLG pLOOvOLW
UPA 89 So Strecker, Antithesen. Although the Two Source theory has been assumed throughout here, the whole argument could be repeated almost verbatim on the basis of the Griesbach hypothesis (the main rival today to the Two Source theory of synoptic relationships). Many modern defenders of the Griesbach hypothesis would argue that at times, Luke has access to traditions which were parallel, but independent, to those of Matthew, and this effectively reduces to a form of the Q hypothesis.
If one adopts a very "strong" form of the Griesbach hypothesis, and argues that Luke is always dependent on Matthew for the material they have in common, the present discussion becomes much simpler, with an identical result: every agreement between the Didache and Luke can be taken as an agreement with LkR and hence an indication that the Didache presupposes at least Luke's finished gospel. Cf. also the argument of Butler (on the basis of his preferred Augustinian hypothesis rather than the Griesbach hypothesis, though both hypotheses assume Luke's dependence on Matthew here): the fact that Did 1:3-2:1 "borrows some of the specifically Lukan modifications of the M(g) material" [for Butler "M(g)" is Matthew] shows that Didache is dependent on Luke here ("Two Ways", p. 31).
CHRISTOPHER M. TUCKETT
evAOYELTE TOUS BepoyslTE TOUS
KATAPWLEVOUG Up
KATOAPWLEVOUS VLA
KaL TpocevxEdbe UTED Kal Tpocevxeobe UTEP Tpoosvxeobe TEpt Tov éXOpav vuav pnoteveTe 5é UTEP
TOV SLWKOVTOV VBLAS
TOV SLWKOVTOV VLA
TOV eTNpEACOrTMV UPaS The evidence from this section is mostly ambiguous for the present purposes.
Did 1:3a presents a version of the "love of enemies" saying which is closer to Lk 6:27f than to Mt 5:44: the Didache has an exact parallel to the "Bless those who curse you" clause of Lk 6:28a which has no parallel in what are regarded as the best manuscripts of Matthew (though the extra words are contained in several other manuscripts of Matthew).%° Further, although Did 1:3a has no explicit statement here that one is to "love one's enemies" (Mt 5:44a/Lk 6:27a), the rhetorical question which follows in the next section ("if you love those who love you") suggests strongly that the Didache presupposes a command here to "love" people who are not well-disposed to one. Moreover, the Didache has a clause a the end of the next section ("love those who hate you") which is parallel to both halves of Lk 6:27
("love your enemies and do good to those who hate you"). The whole of the longer, four-fold command to love one's enemies thus seems to be presupposed by the Didache.
It is not certain whether Luke's four-fold form, or Matthew's two-fold form, most accurately represents the Q version of the command to love one's enemies.®!
Several scholars remain undecided.92 However, even if one could establish that Matthew's shorter version represents the Q version, the presence of a (rough)
°0 Notably in the Western text of Matthew and in several later manuscripts. One must always bear in mind the obvious fact that the version of the gospels which may have been known to the Didachist will not necessarily the identical with the version printed in the 26th edition of the Nestle-Aland text! Still, Glover's claim that the Didache agrees with Luke only when Didache "is covering ground common to both
Luke and Matthew" ("Didache's Quotations", p. 14), which he uses to posit some relationship between Didache and Q, is clearly true only in the most general terms here.
°l Luke's version is regarded as more original by J. Schmid, Matthdus und Lukas (Freiburg: Herder, 1930 [BSt 23/2-4]) p. 229: Schiirmann, Lukasevangelium, p. 346; Merklein, Gottesherrschaft, p. 225; Sauer, "Feindesliebe", p. 8; Schlosser,
Dieu, p.225. Matthew is regarded as more original by Liihrmann, "Liebet", pp.
416f; Zeller, Mahnspriiche, p. 102; Hoffmann, "Tradition", p. 52f; J. A. Fitzmyer, rs Gospel according to Luke I-IX (New York: Doubleday, 1981 [Anchor 28]), p.
Es 92 Piper, Love your Enemies, p. S6f; others cited in Hoffmann; "Tradition", p. nO)
SYNOPTIC TRADITION IN THE DIDACHE
parallel to Lk 6:28a in Rom 12:14 may suggest that Luke has added a traditional saying here, rather than creating the clause himself.°? The strongest argument for the activity of LkR concerns the use of kaA@¢ movetre. This links with the use of &yadomovetp later in this context in Luke, and this is widely regarded as LkR.%
But it is just this phrase of Lk 6:27f. which does not have a parallel in Did 1:3. The Didache does have a parallel to Luke's "those who hate you" (cf. above), but Luke's use of wioéw here may well reflect the use of his source: many would argue that the "love of enemies" complex in Q followed immediately after the last beatitude in view of the close parallels in wording and substance between them, and it is probably the Q version of the final beatitude which referred to "hating".°5 Thus the longer form of the command to love one enemies in the Didache can probably not tell us very much in the present discussion.
The Didache's version differs from both the synoptic version in making the "enemies" the object of the verb "pray for" the command to "pray for your enemies" can be found in several other places in early Christian literature (cf. Justin, /Apol 14:9;
15:3; Dial 35:8; 96:3; 133:6; POxy 1224; Dida 108:14; PsClemHom XII.32). Some have seen here evidence of an independent saying.°° However, Késter has shown that the saying probably reflects later church practice of praying for enemies, and Polycarp Phil 12:3 may show an intermediate phase of the tradition: here an exhortation to pray for various people is combined with a clear reference to the gospel saying in Mt 5:44/Lk 6:28, so 93 Cf. F.W. Horn, Glaube und Handeln in der Theologie. des Lukas (Gottingen:
Vandenhoeck & Ruprecht, 1983 [GTA 26]) p. 105. However, Horn also adduces the evidence of Did 1:3 itself to show that the longer Lukan text is not LkR.
Clearly such an appeal would be circular in the present context. Cf. the discussion of the parallels between Mt 24:10-12 and Did 16:3-5. at nn. 42, 46 above. For doubts about the parallel between Lk 6 and Rom 12, see Neirynck, "Paul and the Sayings of Jesus", p. 303. One may note that the verb "to curse" is used differently: in Luke it describes the behaviour of the opponents ("those who curse you"); in Paul it describes behaviour which the Christian is to shun ("do not curse"). Cf also Neirynck, ibid., p. 299: "the contrast between blessing and cursing appears in both texts but in biblical language this can scarcely be cited as a significant contact".
94 See n. 114 below and the discussion in the text there. Sauer's argument that Kad@o moveite is a Lukan hapax, different from the Lukan &yaGwrovety and hence pre-Lukan (cf. also Horn, Glaube, p. 105) seems unnecessarily restrictive of Luke's ability to vary his terminology. Cf. Neirynck, ibid., p. 297.
95 Cf. Schiirmann, Lukasevangelium, p. 346; Liihrmann, "Liebet", pp. 414f.
Liihrmann postulates that puoéw stood in Q in the second half of the original 2-fold saying as the object of xpoaedxeo0e. This seems plausible, though Neirynck (ibid., pp. 297f) sees the link as more probably Lukan.
96 Glover especially makes much of parallels between the Didache and Justin as evidence of independent traditions.
CHRISTOPHER M. TUCKETT
that the general saying in the form "pray for your enemies" is probably due to later developments in the tradition.*'
The same may apply to the command to "fast" for one's persecutors which has no synoptic parallel. Prayer and fasting were connected in several contexts in Christian (and Jewish) texts.%
Hence a reference to fasting, secondarily added to an injunction to prayer, would not be unexpected in a later development of the tradition.
One small feature which may be more significant in this section is the use of d5u@xw which agrees with Mt 5:44b against Lk
6:28b. Many would argue that duxw in Mt 5 is due to MattR.
Luke's parallel here has érnpec&{w which is not a Lukan word (it occurs only here in Luke-Acts). éanpec&fw also occurs in 1 Pet
3:16 in a not unrelated context, which may indicate that the word had its place in a Christian paraenetic tradition of this nature.10
Further, Luke uses é:@xw three times in his gospel and nine times in Acts, so there is no clear reason why he should-avoid using the word here if it lay in his source. The word may well be a Matthean favourite (cf. the use in the penultimate beatitude Mt 5:10, which is widely regarded as due to MattR). Hence the likelihood is that it is due to MattR here.!°! One could argue that the word is too general to carry much weight here; but against this is the fact that the motif of persecution is not one that really dominates this, or any, section of the Didache. It is therefore unlikely to have been added by the Didachist himself.
This small agreement between the Didache and Matthew may thus be an instance where the Didache presupposes MattR and hence Matthew's finished gospel.
Did 1:3b Mt 5 Lk 6 Toia yap XapLC 97 Koster, Uberlieferung, p. 224; also A. Bellinzoni, The Sayings of Jesus in the Writings of Justin Martyr (Leiden: Brill, 1967 [NT.S 17]), pp. 79f.
8 Cf. Koster, Uberlieferung, p. 224; Draper, "Jesus Tradition", p. 276; Massaux, Influence Ill, p. 154.
°9 Layton, "Sources", p. 353, refers to similar expansions to the references to prayer in Mk 9:29 and 1 Cor 7:5 in the manuscript tradition of the NT. Rordorf,
"Probléme", pp. 501f, thinks that the persecutors are Jews, but this says nothing about the relation of the passage to the synoptic tradition.
F = question of the relation of 1 Peter to gospel tradition is of course much ebated.
101 So Schiirmann, Lukasevangelium, p. 333; Schulz, Q, p. 128; Merklein, Gottesherrschaft, p. 225; Sauer, "Feindesliebe", p. 11. For the evidence from the beatitudes see my "The Beatitudes: A Source-Critical Study", NT 25 (1983) pp.
193-207, on p. 203. Cf. also n. 95 above.
SYNOPTIC TRADITION IN THE DIDACHE
LE?
EQV AYATATE TOC
Weav yap ayamnonre Kal ei WyaTaTE TOG ayaravrac twas Tov ayamavras twas kyaTavras bas Toia Tiva poor ExeTE' ovxi iiv xapic goTiv; Kal Kal of TENGval TS abTd ~—-Yax/p Of GapTwdol TodC
TOLOVOLY;
ayaTavtac avTovc Tkai &av doTdonodse ayamtaouw Tovg Adedpovs budv Kal yap éav LOvov Ti Teptooor ayaboronTe TovC TOLELTE; ayaboroodyrac bac Toa Umi xapiC éoTiv; OUxXt Kal ol EOviKoi Td Kal ol @MapTWAoOL TO QUTO ToLovoLy QvTO ToLovow... (Vv. ovxXt Kal T& EOvN TO 34) QvTo ToLovaww; bei 5& SaHv eyaTATE TOUS ayanwate Tod EXOpovG bpav...Kai poovvTas Las EaTat 6 p10bd¢ buav kal ovx &ete &xOpov S6mwo yévnobe vioi ToNUS Kal EcEobe viol
TOU TATPOS
vpiorou In this section the Didache is once again close to Luke's version without being identical to it. Did 1:3b has only one rhetorical question, set in terms of "loving those who love you", as do both
Matthew and Luke. The Didache agrees with Mt 5:47 in mentioning Gentiles as the "opposing group" from whom the readers are to distinguish themselves.1°2 The Didache then enunciates a version of the love command, and in doing so follows the structure of Lk 6 where the love command is repeated after the rhetorical questions (Lk 6:35). Finally the Didache agrees with both Matthew and Luke in enunciating a consequence for those who obey the love command; but the consequence is not that of divine sonship (as in the synoptics) but the more mundane "you will not have an enemy".
One should perhaps first deal with a peculiar problem of the text of the Didache at this point. The word used for "love" here is uncertain. The Bryennios manuscript uses wyamaw three times.
However, POxy 1782 uses ¢:Agw in the last instance (the earlier instances do not appear on that part of the papyrus which has been
102 Draper, "Jesus Tradition", p. 277, is slightly confusing in saying "the Didache uses &vn for &papTwdoi in Luke and reA@vou in Matthew", though he concedes in a bracket "(cf. é@vcxoi in 5:47)". The Didache is parallel to Mt 5:47 here, not 5:46. Audet, Didaché, pp. 184f, simply prints Mt 5:38-47 in toto without even considering v. 47 and makes no attempt to determine which part of the Didache is parallel to which part of the synoptic versions. He also appeals to the very different order in the Didache as compared with the synoptics, but fails to analyze Didache's own structure at this point.
CHRISTOPHER M. TUCKETT
preserved). Further, the text of Const VII, which takes up the text of the Didache almost wholesale whilst commenting on it and adding to it, has ¢uAgw on these three occasions. It is argued by some on the basis of this evidence that the original text of the Didache read ¢.id€w all three times. Const has a tendency to conform the text of the Didache to that of the canonical gospels; hence in this instance, the disagreement between Const's reproduction of the Didache's text and the synoptic versions is all the more striking. (There is no manuscript evidence for a text of the gospels reading ¢uAgw here). Also the Bryennios manuscript shows a tendency elsewhere to conform its text to that of wellknown Biblical versions.!°3 Hence, it is argued, the three-fold évyamcéw in this manuscript may represent a later scribal assimilation to the text of the gospels, and the original text of the Didache used ¢iAéw. Further, since this is unlike both the Matthean and Lukan forms of the saying, this may indicate that the Didache preserves an independent version of the saying.!%
This evidence is however of uncertain value. The text of Const is not a manuscript of the Didache: it represents the use of the Didache by a later author, not a scribe.!° Further, although the tendency of Const generally is to conform the text of the Didache to that of the gospels, this has not happened at every point. (Cf. n. 40 above). There are thus other occasions when
Const makes the text of the Didache less like that of the canonical gospels, and hence one cannot rule out the possibility that the same has happened here. One must also note that ¢diAéw and wyamda are almost synonymous, or, even if they are not, that the author of Const seems quite happy to regard them as synonymous.
On two occasions he adopts the pose of the beloved disciple of the fourth gospel and uses ¢:A€w to describe his special relationship to Jesus: V.14:3 giXobpevocg mAEiov Tov &AAwV dT' adTod, VIII.16:1
EY@ O diAobpevos UTS Tod kupiov. Although these could have been inspired by Jn 20:2 (6¢ édirer 6 "Inoodc), the more usual Johan103 Cf. J.B. Lightfoot, The Apostolic Fathers. Part I. Clement of Rome Vol. I.
(London: Macmillan, 1890) pp. 124ff, in relation to the text of 1 Clement in the Bryennios manuscript.
104 See Audet, Didaché, p. 54; Draper, "Jesus Tradition", p. 276f; Kéhler, Rezeption, p. 44; cf. also R. Bauckham, "The Study of Gospel Traditions outside the Canonical Gospels: Problems and Prospects", in Gospel Perspectives (n. 1), pp.
369-403, on p. 389, who then uses the parallel in Ignatius Pol 2:1 (which also uses
$thEw) as a key point in his argument that Ignatius is demonstrably independent of Matthew at at least one point. It is, however, not clear how such a theory would relate to the Didache's alleged use of Q (cf nn. 3, 86 above): is the variation due to independent translations of an Aramaic original? If so, is Q thought to be an Aramaic document? Or are we to think of variations in the oral tradition?
105 Massaux, Influence III, p. 148 n. 11, is slightly misleading when he says "some codices read ¢tdei7e" (my italics) at the first use of &yaTdaw.
SYNOPTIC TRADITION IN THE DIDACHE
nine verb to use for Jesus' relationship to the beloved disciple is ayamaw (13:23; 19:26; 21:7, 20). One could even argue that these instances might indicate a slight preference by the author of Const for duAéw over against &yamwé&w (though he uses &yamaw frequently elsewhere). At another point it seems clear that the author regarded the two verbs as synonymous: at V.15:2 he refers to the Gentiles in words clearly influenced by the use of Hos 2 in 1 Pet 2 and Rom 9; but whereas Paul uses wyaméw twice here,
Const uses ¢diAéw in the second instance (of7iwvéo mote ovK nyamnusvor viv 5&8 wediAnuévor). The use of didgw in VII.2:2
(the parallel to Did 1:3) may therefore simply reflect the readiness of the author of Const to use the verbs ¢iAéw and wyamaw interchangeably. The evidence of the POxy 1782 fragment must retain its value, but this only supplies the reading for the last occurrence.
(The papyrus is not extant earlier). It may be that the author of the Didache wanted to avoid having three uses of &yamtéw very close together. His use of his tradition (whatever that was) meant that with only one rhetorical question here, the two dyaméw's of the question and the "love" command itself are very close together.
The use of diAgw in the third instance (if that is the correct reading) could then be due to the Didachist's desire to avoid repetition.
In all, the textual evidence is too uncertain for us to be sure that the Didache used ¢i€w all three times here. The indirect nature of the witness of the text of Const must be borne in mind, and it would be very precarious to build too much on the use of two words here which are clearly regarded by the final author in the chain of witnesses as synonymous.
}° Turning to other features of the Didache's text in this section, the reference to "Gentiles" is probably indecisive for the present purposes. Almost all agree that Luke's "sinners" here represents a secondary alteration of the tradition and that Matthew's oi é6vxoi is most likely to represent the Q wording.!"
Thus the Didache's use of t& &vn here only shows agreement with a pre-redactional element in the synoptic versions.!°8 Glover
106 Layton, "Sources", p. 373f, points out that if the Bryennios readings are due to scribal assimilation, then such assimilation has not gone very far (e.g. not as far as Const's other assimilations of the Didache here). He thus prefers to see the readings as "separative errors". However, his further remark that "adoption of the H
[i.e. Bryennios] variants also obviates the necessity of postulating a more complex and far less elegant hypothesis on the relationship of the Didache passage to its parallels" (p. 374) is perhaps less fortunate. One cannot assume that the least complex and most elegant hypotheses always correspond to historical reality
107 Schiirmann, Lukasevangelium, p. 35; Schulz, Q, p. 129f; Hoffmann, "Tradition", p. 55; Sauer, "Feindesliebe", p. 12; Catchpole, "Inaugural Discourse", p. 304, who also refers to a similar disparaging reference to Gentiles in another Q passage, viz. Mt 6:32/Lk 12:30.
108 Jt is dubious whether one can place too much weight on the difference between &vn and é@mxot (so Rordorf, "Probléme", p. 503, who claims that the
CHRISTOPHER M. TUCKETT
and Rordorf both appeal to the allegedly "tactless" way in which the Didachist refers to Gentiles here, and they deduce from this that the Didache is here preserving an old tradition which is at odds with a work addressed to Gentiles.!°9 However, it is doubtful whether we can deduce too much from the title. There are, after all, two titles and it is not certain which, if either, is original and reflects the aim of the work as a whole.!!° One may also say that the Didachist is not renowned for his consistency of thought: cf. the tension between 1:5 and 1:6 on the question of whether one should ask questions before giving alms.!!! Thus an alleged tension between the title and this reference to Gentiles is not impossible in a work such as the Didache.
The one feature which is more noteworthy in this section is the introductory question moia yap xcpic; this agrees closely with Luke's form of the rhetorical questions in Lk 6:32f. (woia byty xapic éoriv;). Further, it has become clear in recent Lukan study that Luke's version here is heavily indebted to LkR. Kdster ascribed ya&pic to LkR simply on the basis of a word count.!!?
However, others have pointed out that word counts in themselves may be misleading in that the word in question may well have different meanings elsewhere: in this case x&pic elsewhere in Luke tends to mean divine grace.!!3 Nevertheless, the redactional nature of the word here can be established via another route.
In a programmatic article, van Unnik has shown how Luke has adapted his tradition in order to address the morality determined by a reciprocity ethic;!14 an ethic of doing good to others in order to receive reciprocal favours in return was widespread in the ancient world. Further, the language of "doing good" and xda@pic
Didache is close to Matthew but not verbally identical, and therefore probably independent). The two are almost synonymous and one must remember that the Didache is not a scribe's copy of the text of the gospels (cf. above). In fact it is Matthew's pein which is unusual and invites a change to &@vn: cf. Layton, "Sources", p.
55.
109 Glover, "Didache's Quotations", p. 14; Rordorf, "Probléme", p. 502f; also Draper, "Jesus Tradition", p. 277.
110 Layton, "Sources", p. 382, referring to the reference to "the twelve apostles" in the title, says: "the literary fiction in the title is, unless the date of composition be extremely early, a clear example of archaism, used here to lend authority to the document". The rotc &@veowv is probably intended to reflect Mt 28:19f and has the same aim. One need not therefore see any great significance in the intended readership in the phrase.
111 The tension is so great that Petersen, "Probleme", pp. 147ff, ascribes 1:6 to a marginal gloss.
112 Uberlieferung, p. 224f.
113 Cf. Wrege, Bergpredigt, p. 89; Piper, Love Your Enemies, p. 192 n. 146.
114 WC. van Unnik, "Die Motivierung der Feindesliebe in Lukas VI 32-355", NT 8 (1966) pp. 288-300.
SYNOPTIC TRADITION IN THE DIDACHE
had a firm place in such talk. Thus Luke's use of &yadomouetc and xa@pic here are almost certainly redactional adaptations of the older tradition, addressing this ethos and criticizing it sharply.
(The fact that very similar ideas recur in the almost certainly redactional verses Lk 14:12-14 indicate that this is a peculiarly
Lukan theme).!!5 Luke's language is striking: moia byiy xc&pic éotiv should probably be translated "what kind of a reward do you have?", to which the implied answer is that one has a human reward but not a divine one.!!6 Luke's language thus belongs within a closely constructed framework of thought and needs that framework to make sense.
The Didache shares some of the same language but not the framework of thought. For the Didache too has roia yap x&pic; but the conclusion is the rather lame kai ovx ere éxOpov.
However, this is now precisely the ethos of the reciprocity ethic which Luke's language was designed to oppose: love others and they will love you back. Thus the formulation of the x@pic question, which makes excellent sense in the Lukan context, becomes confused when repeated verbatim in the slightly different context of the Didache. The fact that the agreement between the Didache and Luke goes beyond that of just the common use of xc&pic alone suggests that there is some literary relationship involved here.
Some have appealed to the possibility that ya&pu¢ and puobdc may simply be translation variants of an Aramaic original, and that the parallel with Lk 6:32 in Did 1:3 and 1 Pet 2:20 suggests a common catechetical tradition.!17 However, the xa&pi¢ vocabulary in the Hellenistic tradition suggests that Luke's choice of vocabulary here is deliberate. Further, the appeals to the Didache and 1 Peter to solve the problem of the synoptic tradition-history create difficulties here.
1 Pet 2:20 is perhaps not relevant: although the idea in 1 Pet 2 is not dissimilar to that of Lk 6 (a contrast between something which is not particularly worthy of divine praise and something which is), the vocabulary is different: the negative statement is cast in the form motov x\goc, and xcépic is used positively. The use of the Didache here is a case of petitio principii in the present context. We cannot use the Didache itself to establish the pre-Lukan nature of Lk 6:32f and then use this
"result" to establish the conclusion that the Didache betrays no links with redactional elements in the Gospel. The fact remains
115 See also the detailed discussion by Horn, Glaube, pp. 100-102. For LkR here see too Schmid, Marthdus und Lukas, p. 229f; Schiirmann, Lukasevangelium, p. 353; Schulz, Q, p. 129; Liihrmann, "Liebet", p. 420; Hoffmann, "Tradition", p.
55; Sauer, "Feindesliebe", p. 12, and many others.
116 Horn, Glaube, p. 102. j 117 Cf. Wrege, Bergpredigt, p. 89; Piper, Love your Enemies, p. 156; Guelich,
Sermon, p. 231.
CHRISTOPHER M. TUCKETT
that the author of this section of the Didache seems to presuppose Luke's version here in a way that goes beyond simply using the word xycpic: he takes over the Lukan rhetorical question, but fails to see its significance and hence betrays the secondary nature of his own text.
Gy Did 1544:
Mt 5 Lk 6
OMEXOV TOV OALPKLKOV
KOLL OWMLATLKGY
ETLOULLOY
éav Tic oot d@ 3960TLC o& pati fel
27 TOTTOVTL Oe
pamiopa sic THY dekiav ovaryova gig THY SekLaY OLayova = ETL THY OLaryova oTpéWov avT@ oTpéwov avTo TOPEXE
Ka THY AAAHY
Kal THY GAAHY Kal THY AAV Kal Eon TENELOS (cf. v. 48) The textual (and other) problems associated with the opening phrase will be left on one side here,!!8 as they do not affect the question of the Didache's possible use of synoptic tradition. The main part of this section represents a-version of the saying about turning the other cheek. The evidence here is probably indecisive.
Did 1:4 agrees very closely with the Matthean form of the saying, specifying the "right" cheek, using 6.ddvor pamiopa (cf. Matthew's pamifer as opposed to Luke's tim7w) and otpépov (Luke mopexe). However, it is not clear if any of these features are due to MattR. The technical nature of the reference to the "right" cheek is often noticed: such a blow implies a back-handed slap which, according to the Mishnah (m.BQ 8:6), was regarded as extremely insulting and worthy of double recompense.
But whether such a detail belonged to Q and was omitted by Luke as irrelevant, or whether it was added by Matthew (or in a Q™ form of the tradition) is hard to say.1!9 Matthew's other features mentioned above (i.e. the use of pamifw and orpé~or) are widely regarded as more original than Luke's tia7w and mé&pexe.!2° Thus the Didache does not have any clear links with MattR.
118 Cf. Layton, "Sources", p. 376f.
119 For MattR, see Schmid, Matthdus und Lukas, p. 229; Schiirmann, Lukasevangelium, p. 347; Hoffmann, "Tradition", p. 59f; Guelich, Sermon, pp. 221f;
Fitzmyer, Luke, p. 638; Sauer, "Feindesliebe", p. 12. It is regarded as preMatthean by Wrege, Bergpredigt, p. 76; Schulz, Q, p. 122; Merklein, Gottesherrschaft, p. 269; Catchpole, "Inaugural Discourse", p. 306. Késter, Uberlieferung, p. 226f, takes it as MattR and evidence of the dependence of the Didache on Matthew, but he offers little justification for the redactional nature of the word.
120 So Schulz, Q, p. 122; Sauer, "Feindesliebe", p 12f, and others cited there.
SYNOPTIC TRADITION IN THE DIDACHE
One feature that is often adduced by supporters of the Didache's independence of the synoptics is the allegedly clumsy nature of the Didache's Greek phrase 5@ pd&mopa. This, it is argued, must tell against the theory of dependence on Matthew as Matthew has the more usual simple verb paziter.!21 However, this need only show that the Didache is exercising an element of freedom at this point and that, if the usage is clumsy,!22 the Didachist is capable of writing poor Greek. This may say something about the Didachist's style, but it does not solve the problem of whether the Didache is presupposing our gospels or not.!23 In fact, as Draper points out, the Didache's language is also that of Jn 18:22 and not dissimilar to that of Is 50:6. Thus the Didache here may simply reflect further "biblical" influences. One must also allow for a greater element of freedom on the part of the Didachist in using his sources than in the case of a scribe seeking to copy a manuscript.
It is possible that the very last phrase of this section in the Didache, xai on TéXEtog, may show a link with MattR. Mt 5:48 closes the series of antitheses in Matthew with an exhortation to be "perfect" (7é\evo¢) and there is almost complete scholarly agreement that this is due to MattR of the more original Lukan version "be merciful" (Lk 6:36).!24 On the other hand, it could be argued that a very similar exhortation in Did 6:2 after the end of the Two Ways section, indicates that this language of "perfection" is of some importance for the Didachist. Thus the exhortation here could be due to independent redaction by the Didachist.125 So whilst this feature looks at first sight to be a clear example of an agreement between the Didache and MattR, its value should probably not be over-estimated.
Did 1:4b Mt 5 éav ayyapevon o& Tig pidvoy Ev 'kai OoTIG o& AyyapEvoet Midiov Ev Umaye "eT avToU dvo Uraye eT avTov dvo 121 Glover, "Didache's Quotations", p. 14f; Rordorf, "Probléme", p. 504;
Draper, "Jesus Tradition", p. 277.
122 Layton, "Sources", p. 357, suggests that the Didache's usage is perhaps to avoid four utterly parallel sentences.
123 Cf. the similar problem in discussions of the Synoptic problem in relation to Mark's allegedly rough style, and the remarks of W. R. Farmer, The Synoptic Problem (Dillsboro, 19762) p. 122.
124 Schulz, Q, p. 130; Liihrmann, "Liebet", p. 421, and many others.
125 Cf. Rordorf, "Probléme", p. 505. He also rightly notes that the reference in 6:2 (along with other agreements with different parts of the Didache, must cast some doubt on theories that the section 1:3-2:1 was added after the rest of the text was completed. Cf. n. 142 below.
CHRISTOPHER M. TUCKETT
Did 1:4b is parallel to Mt 5:41, which has no Lukan parallel. One cannot say with any certainty whether this verse is due to MattR, whether it was part of Q but omitted by Luke, or whether it originated in a Q™ expansion of Q.!26 An origin in Q seems the least likely option on literary-critical grounds: the examples of non-retaliation in Lk 6:29f constitute a reasonably neatly balanced pair of couplets; the presence of Mt 5:41 makes for five examples and this seems rather overloaded. (On the other hand, it could be argued that the two last examples are distinctly anticlimactic,!?? so that an original trio of examples in the pre-Q tradition was secondarily expanded by the sayings about giving.'?8) The situation presupposed here is probably that of Roman troops forcing service from native Jews; but whether this already reflects the situation of the Q "community", or the later Matthean community, or a stage of the tradition in between, is very hard to say without arguing in a circle.!29 Certainly the situation presupposed does not seem to be particularly "Didachean" and hence the presence of the saying here is almost certainly due to the writer's use of his tradition, but the evidence is not definite enough to enable us to identify that tradition accurately.
126 Ror MattR see Schulz, Q, p. 123; for Q, see Piper, Love Your Enemies, p.
58; Hoffmann, "Tradition", p. 61; Sauer, "Feindesliebe", p. 9; for Q™, see Strecker, Bergpredigt, p. 87; Zeller, Mahnspriiche, p. 55.
127 Cf. Liihrmann, "Liebet", pp. 418, 427; Zeller, Mahnspriiche, p. 55; Sauer, "Feindesliebe", p. 15.
128 The sayings about non-retaliation and about giving are thus regarded as of separate origin by Schiirmann, Lukasevangelium, p. 349; Schulz, Q, p. 13; Liihrmann, "Liebet", p. 427.
129 Cf, Theissen argues that it is the Matthean situation: see his "Gewaltverzicht und Feindesliebe (Mt 5,38-48/Lk 6,27-38) und deren sozialgeschichtlicher Hintergrund", in Studien zur Soziologie des Urchristentums (Tiibingen: Mohr, 1979
([WUNT 19]) pp. 160-197, esp. p. 176f. Hoffmann, "Tradition", p. 61, argues strongly that this is the situation of the Q- "Gruppe" as well. Against Hoffmann it must be said that there are virtually no other direct references to worries about contact with Roman troops in Q, whereas Matthew does reflect on the theological implications of the events of the Jewish war (cf. Mt 22:7). The preceding two examples about non-retaliation in Mt 5:39f have to do with personal insult, and, if it is justified to connect the love-of-enemies complex in Q with the final beatitude
(cf. n. 95 above), then the people addressed are those who are facing persecution for their Christian commitment. See L. Schottroff, "Gewaltverzicht und Feindesliebe in der urchristlichen Tradition. Mt 5,38-48; Lk. 6,26-27" in Jesus Christus in Historie und Theologie: Fs. H. Conzelmann (Tiibingen: Mohr, 1975) pp. 197-221.
Merklein, Gottesherrschaft, p. 272, and Zeller, Mahnspriiche, p. 59, both argue that the non-retaliation sayings refer to quite general, typical situations of violence; however, it would seem that Schottroff is justified in her claims at least at the level of Q where the context gives a rather more precise application of the saying.
SYNOPTIC TRADITION IN THE DIDACHE
BS Did'ic4e Mt 5 Lk 6 éav App Tig TO iwarioy =F OENovTi cor KpLOAvac Kal wd TOD QpovToc gov KQl TOV XLT@Va Gov Oou TO iariov AaBetv 60¢ avTO Kai TOV ade avrTp Kai 7d Kal TOV XLTaVa LH xXiTava ipariov KWAVONG The Didache here reveals close affinities with the Lukan version.
Further, the differences between Matthew and Luke are quite considerable, though again it is difficult to know which version is more original. Matthew's version presupposes the situation of a law-suit where the person addressed is being sued for his shirt
(xiT@v) and is told to surrender even his cloak (iuériov), the one item which the Jewish Law expressly forbade anyone else to take
(cf. Ex 24:25f; Dt 24:13).130 Luke's (and the Didache's) version reverses the order of the x.7@y and the iudz7ioy and seems to presuppose a robbery situation: if a person is robbed of his cloak
(the first thing a robber would grab) he is to surrender his shirt as well. Again the argument about originality can go (and has gone) either way. Luke could be generalizing Matthew's more technical language for a non-Jewish audience;!3! or it could have been Matthew who has introduced the legal ideas here (cf. Mt 5:38).132
Once again the evidence is indecisive and the agreement between the Didache and the Lukan version cannot be taken any further here.
Did 1:4d-Sa Mt 5 Lk 6 éav NaBH TiIG ATO CoD =— kai Tov OédovTA TO = *PKALL ATO TOD
TO OO pH aTaiTEl cov daveionoba py alpovTos T& O& LN ovdé yap Svvacat aTOOTPAdhs QTairel TaVTi TO AiTOUPTI OE 42a7@ aiTovVTi o& 3arepyri aitourTt oF didov Kai py aTairEet 666... didov...
The final section to be considered here again shows agreement between the Didache and Luke. Did 1:4f has the sayings in the 130 Guelich, Sermon, p. 222.
§o Schiirmann, Lukasevangelium, pp. 349, 351; Schulz, Q, p. 123; Hoffmann, "Tradition", p. 60; Sauer, "Feindesliebe", p. 13.
132 Cf. Guelich, Sermon, p. 222; Catchpole, "Inaugural Discourse", p. 306.
Theissen, Studien, p. 184, argues that Luke is primarily interested in the idea of borrowing, and hence would not have altered his tradition here. This is not altogether convincing. Luke may well have omitted the borrowing reference here
(see below), perhaps simply because he wanted to generalize and make things more relevant for a non-Jewish audience, and also to develop the lending motif later.
CHRISTOPHER M. TUCKETT
reverse order to that of the synoptics (probably so as to end with a saying about "giving" to lead on to 1:5). The last clause in the section given above, "give to everyone who asks", agrees with Luke in having mavri and diéov. Both are probably LkR in Luke,133 though one cannot build too much on this here: Luke's aim is to generalize the idea of giving, but the Didache has exactly the same idea and hence the wavz7i and the present imperative didov could just as easily be seen as independent redaction of the tradition by the Didachist.
More significant in the present context is the opening sentence here, where the Didache agrees with Luke against Matthew in referring to someone who takes, rather than someone who wants to borrow; there is also agreement between the Didache and Luke in using d@airex in the final part. Now it is almost certain that Matthew's reference to "borrowing" represents the Q version here: Luke uses the same (rather rare) verb in vv. 34, 35 and this seems to constitute a reminiscence of the earlier saying.!54+ Luke is above all here interested in the idea of generous giving, and he appears to save up the borrowing reference to develop it considerably in v. 34f. Further, it is of considerable importance to Luke to stress the idea that one should expect nothing in return.
(This is, of course, part of his critique of the reciprocity ethic.) Hence Luke's "lending" to a borrower is virtually equivalent to a gift.135 However, the "not expecting anything back" idea dominate
Luke's version here.!3° Thus the Lukan a@aire: in v. 30 is much more likely to be redactional than Matthew's amootpadific.137 In v. 30 Luke appears to have continued the robbery idea from v. 9, and saved the reference to the "borrower" for later; but he starts to introduce the idea of not asking for anything in return ("7
Qmatret) here.
The result is that Luke's version is rather uneven. For the Lukan text exhorts someone who has just been robbed not to demand his property back. But whatever means one might employ to recover stolen goods, simply demanding is unlikely to have any
133 See Schulz, Q, p. 121; Sauer, "Feindesliebe", p. 13 and others cited there Késter, Uberlieferung, p. 228, sees this as evidence of dependence of the Didache on LkR, but this is probably over-pressing the evidence.
134 See Schulz, Q, p. 123; Liihrmann, "Liebet", p. 418; Sauer, "Feindesliebe", p. 13; Catchpole, "Inaugural Discourse", pp. 306f. The attempt of Schiirmann,
Lukasevangelium, p. 357 (followed by Guelich, Sermon, p. 223) to see the connection as pre-Lukan is unnecessary: see the critique in Hoffmann, "Tradition", p. 66f.
135 Schottroff, "Gewaltverzicht", p. 217; Horn, Glaube, p. 102.
136 See Piper, Love Your Enemies, p. 157f.
137 So Schiirmann, Lukasevangelium, p. 349; Schulz, Q, p. 123; Guelich, Sermon, p. 223; Hoffmann, "Tradition", p. 63; Sauer, "Feindesliebe", p. 13; Contra Catchpole, "Inaugural Discourses", p. 307.
SYNOPTIC TRADITION IN THE DIDACHE
effect at all.'38 It may be that it is precisely this incongruity in the Lukan text which is reflected in the notorious little clause, odé yap Stvaco., which is appended at this point in the Didache. The clause has caused immense perplexity.!39 However, it may be that it is simply the Didachist's comment on the preceding exhortation which he recognizes as somewhat incongruous: if you have been robbed, do not demand back'4° your property, "for you cannot" -you will have no success!!41 The perplexity of commentators on the Didache may simply be a reflection at one stage removed of the Didachist's (or perhaps a later scribe's) worldly comment on the slightly puzzling preceding clause.
If this suggestion has any value, it may indicate that the Didachist was using his sources here with a certain degree of faithfulness: he was primarily interested in the exhortation to "give"; but he found this in his tradition coupled with another exhortation not to demand back stolen property, and although he found this difficult, he nevertheless reproduced it faithfully (albeit adding a comment of his own). This may then imply that the author of this clause regarded his source with some considerable respect. The source may thus have had something of the status of a "canonical" work, and this is clearly easier to envisage if this section of the Didache (at least in its present form) has a relatively late origin.
The net result of the discussion of this section is that the Didache here appears to presuppose Luke's redactional work and hence Luke's finished gospel.
The result of this detailed analysis of Did 1:3-Sa in relation to the synoptic parallels in Mt 5 and Lk 6 shows that this section of the Didache appears on a number of occasions to presuppose the redactional activity of both evangelists, perhaps Luke more clearly than Matthew. This suggests very strongly that the Didache here
138 Catchpole, op.cit., p. 307, is one of the very few who have noted the incongruity of the Lukan text here: he speaks of the Lukan ending as providing "an unduly weak sequel" here.
139 See the discussion of older views in Layton, "Sources", pp 346ff, together with his own proposed emendation of the text.
140 Lake's translation of the text in the Loeb edition ("if any man will take from thee what is thine, refuse it not") is probably misleading: the idea is of demanding back what one has not got, not refusing to give what one has.
141 Rordorf, "Probléme", p. 505, also seems to hint at the strange nature of the Lukan version, reproduced in the Didache: "Or a las différence du cas du vol (1,4e) ou une réclamation serait illusoire, elle serait possible dans le cas d'un prét (1,5a)".
However he resorts to the interpretation of the extra phrase in the Didache as referring to the spiritual impossibility of the Christian filled with love for his enemies to engage in such action. The Didache may be rather more mundane!
CHRISTOPHER M. TUCKETT
presupposes the gospels of Matthew and Luke in their finished forms.
IV. CONCLUSIONS
This paper has analyzed some of the parallels between material in different parts of the Didache and material in the synoptic gospels.
The result has been that these parallels can be best explained if the Didache presupposes the finished gospels of Matthew and Luke.
Further, this result seems to apply to all parts of the Didache examined here.!42 Precisely how the gospels were available to the author of the Didache is impossible to say: they may have been available as separate texts; they may have been already combined to form a single harmonized text. However, the evidence of the Didache seems to show that the text is primarily a witness to the post-redactional history of the synoptic tradition. It is not a witness to any pre-redactional developments.
142 | have not discussed in detail the problem of the unity of the Didache, in particular the question of whether the section 1:3-2:1 should be regarded as an integral part of the text of the Didache (cf. n. 82 above). I remain unpersuaded that it should not be so regarded. Rordorf, "Probléme", pp. 5O9ff, gives an impressive list of links between Did 1:3-2:1 and the rest of the Didache. The fact that there is no parallel to this section in the Two Ways teaching of the so-called "Church Order"
(regarded by Wengst as decisive evidence that Did 1:3-2:1 is a secondary addition) may be explained by the latter being dependent on the Didachist's Vorlage: cf. Rordorf, ibid., p. 513. In any case the Church Order only parallels part of Did 1-6, omitting the "Way of Death" in Did 5-6.