a. The Sources and Their Redaction by the Didachist Didache, 119-20)6 would first have completed 1.1—11.2,7
work:5 one and the same (apostolic) author (Audet, Didache, 119-20)6 would first have completed 1.1—11.2,7 and then added 11.3-16.88 in a second pha^e.9 Finally, a contemporary (apostolic? cf. Audet, Didache, 120) inter polator altered the work. To this interpolator we owe the passages 1.3b—2.1 (except for 1.4a), as well as 6.2-3;
7.2-4; and 13.3, 5-7.10 In my opinion, not only the early point of origin (in apostolic times) is false in this concep tion, but also (and this is of special interest to us in this section) the model of the evolution of the work in stages at the hand of one and the same author (to whom is added the interpolator).11 In contrast, the initiative of Richardson, Mattioli, Giet, Rordorf and Tuilier, and Wengst appears to me correct in principle. These schol ars draw a fundamental distinction (although differing in details) between a predidachistic tradition and didachistic redaction, and they make an attempt to deduce the history of the writing's origins.12 It appears to me only that the notion of a double redaction shared
7. Reconstruction of the Origin of the Didache a. The Sources and Their Redaction by the Didachist
The following reflections presume that the Didache rep resents an original writing in Greek. Adam, who thought it possible to detect a Syrian text stage, was unwilling to decide whether the Didache was originally written in Greek or Syriac, or whether the document had existed from the beginning in two versions (Greek and Syriac).1
These proposals have (I think rightly) met with no assent. We may reckon with an early Syriac version of the Didache (although it has disappeared),2 but the original text of the Didache could scarcely have been anything but Greek, and any Syriac version could only have been a subsequent translation. These considerations are the starting point for the question of the Didache's origins, which remains to be discussed below.3 How should we picture this more precisely?
I do not share Audet's opinion4 that we should imag ine three stages or levels in the composition of the 1 "Herkunft," 35. during the First Four Centuries with Special Reference to the Early Eastern Church-Orders (ed. Cecil J. Cadoux;
Oxford: Blackwell, 1943) 35-36, 53-54, Bartlet fur ther developed this suggestion. A survey of the assignment of passages to three stages of develop ment (according to Bartlet's picture of the growth of the Didache) was given by C. J. Cadoux in Bartlet,
Church-Order, 55. With a great deal of fantasy, Greiff (.Pascharituale, 139, and frequently elsewhere) sup ports three versions of this writing, one of them a church order (chaps. 1-16), another a ritual book
(chaps. 1-10).
6 "Apostle" is used in the broader sense, not as one of the Twelve.
7 Apart from the interpolations that Audet attributes to a later hand. More on this immediately below. To be more precise, the first level would have con
2 See above, §2.
3 I am not persuaded by Adam's historical location of the Didache. Adam points (I think correctly) to the region of Syria, but wishes to locate the addressees of the writing more precisely in eastern Syria, in Adiabene. He adduces a series of varied arguments to support this position. For him, the history of the Didache's origins is connected to the primitive histo ry of the church in eastern Syria. The Didache thus appears as "the church book for the recently missionized communities in eastern Syria" ("Herkunft,"
60). "We may suggest that the Didache, in its original form, was written in Pella between 90 and 100, for the purpose of serving the young east Syrian com munities as a guideline for structuring their congre gations. In this conclusion, the limitation of the purpose is a thesis with a high degree of probabili ty; the temporal location ranks as a well-grounded hypothesis; but the determination of the place of writing remains in the realm of hypothetical sugges tion" (p. 70). In reality, however, the precise loca tion of the addressees in Adiabene is entirely arbitrary. The same is true of the hypothesis that the document was written in Pella.
4 Audet, Didache, 104-20.
5 The idea that the Didache was composed in a series of stages is quite old. J. Vernon Bartlet ("The Didache Reconsidered,"JTS 22 [1921] 246-49) sug gested that the Didache had progressed through two editions, an editio princeps and an editio altera, emendata et glossata. In his Church-Life and Church-Order tained 1.1-3a; 2.2—5.2; 7.1; 8.1-11.2.
8 Again without the later interpolations, thus more precisely 11.3—13.2 and 14.1—16.8.
9 The gospel to which the Didache refers in the sec ond part would have been written between the first and second phases.
10 Audet thus distinguishes between D 1 and D 2 (the same author), and the interpolator. He counts 1.4a;
7.1b; and 13.4 as subsequent glosses.
11 Against Audet's presentation are Nautin, "Composition," 193-95 (with argumentation that is sometimes questionable, but on the whole persua sive); and Rordorf and Tuilier, Doctrine, 19. Kraft's proposal is immeasurably more complicated. He investigates the "stages of development" behind the
Introduction by Giet and Rordorf and Tuilier is unnecessary;13 I see no reason for poshing it.14 In my opinion, the positing of a single redaction is sufficient to solve the existing problems.15 The resulting picture is as follows: At the beginning of the second century (more will be said below about the temporal location), a Christian author living in an originally Jewish-Christian milieu created, by compilation, a kind of book of rules that is our Didache.
Thus we should make a basic distinction between the sources (literary sources and oral traditions) and the redactor or compiler of the book.16 (Of course, the later glosses are not part of the history of the book's origins, but belong to the story of its textual transmission.) The sources available to the Didachist were as follows:
14 I have arrived at a different position from that of Rordorf and Tuilier because I (unlike them) do not regard chaps. 11-13 as a literary unit. As indicated in the discussion below, I consider 11.4-12 to be a predidachistic tradition, and chaps. 11-13 not to be a literary unit; instead, I believe that chaps. 12-15 constitute both a tradition-historical and literary unit. The consequence of this is that I can avoid the awkward positing of a second redaction.
15 A single redaction is also posited by Richardson and Mattioli (see n. 12 above). That these two authors differ somewhat in their estimation of the state of the sources, and so differ also from my opinion, is a separate issue. The positing of a double redaction is now rejected also by Wengst (Didache, 17 n. 57; cf.
20-23). Still, Wengst differs from my opinion in that he regards chaps. 11-15 as a literary unit (p.
23). Against Rordorf and Tuilier's hypothesis, see now also Schollgen, "Church Order," 67-69.
16 Cf. Wengst, Didache, 20-23. Schollgen ("Church Order," 64-67) now offers a separate opinion. He opposes "arbitrariness" in "division of sources" (p.
65) and prefers to recognize only old pieces of tra dition that were worked together by the author.
Didache. The result is that the book becomes a com plicated literary conglomerate (Kraft, Didache,
63-65), and the author appears merely as an editor, that is, simply an arranger at a particular stage of development. Kraft speaks (Didache, 1-2) very instructively of the genre of "evolved literature" to which the Didache is said to belong, and rightly describes the author of the Didache as an "authoreditor." Nevertheless, it seems to me that Kraft underestimates the literary contribution of the Didachist to the writing.
12 Richardson ("Teaching," 164-66) develops a kind of two-source theory: the compiler had two older liter ary sources, the Two Ways tractate and a church order. Richardson does not attempt to make a more precise assignment of chaps. 7-15 to the source or the redaction, but he considers chap. 16 probably redactional (as well as the sectio evangelica). Mattioli distinguishes between the work of the Didachist and his sources, namely: the Two Ways tractate with an apocalyptic appendix, NT material, and the eucharistic prayers (Richardson, "Teaching," 33, and passim). For Wengst and Schollgen, see imme diately below.
13 Rordorf and Tuilier, Doctrine, 49, 63, 92-94: an unknown author (= the first redactor) took over the Two Ways tractate and altered it through additions
(1.3b—2.1 and 6.2-3), forming chaps. 1-6. The same redactor also took over chaps. 7-10, archaic liturgi cal traditions from the Syrian church (7.2-3, 4b are from the hand of the redactor). It was probably the same author (= the first redactor) who also created chaps. 11-13, for which he also had old tradition at his disposal. Chaps. 14-15 could not be from the same author, however, because they presume a dif ferent set of canonical preconditions. These chap ters should be traced to a second redactor, and this may also be true of chap. 16 (where again old tradi tion has been reworked). Chaps. 14-16 consequent ly come from a later period. The Didachist also appears in Giet's work (L'Enigme, passim) as a redac tor in possession of a variety of old source materi als. But Giet recognizes postdidachist materials in addition. Chap. 15 is a later addition (p. 243), and chap. 16 may also stem from the same postdidachist author (Giet, L'Enigme, 256).
Of course, one comes to a completely different conclusion if one regards the Didache as a deliberate literary fiction. This is the opinion of some Englishspeaking scholars. The forerunner was Robinson, who, in "Problem," 339-56, undertook a first foray in this direction. The thesis was further developed in his Donnellan Lectures in 1920 (Didache, esp.
43-68). Part of the subsequent edition of this book appeared as "Barnabas and the Didache," 113-46,
225-48. For Robinson, the Didache is the camou flage of a later author who artificially placed him self in the apostolic period and imitated (using copious amounts of NT material) and faked apostolicity. (The material in the Two Ways part is said to come from the Epistle of Barnabas.) The liturgical formulae and the canonical situation that is presup posed are, correspondingly, the products of fantasy.
Cf. further Richard H. Connolly, "The Didache and Montanism," DRev 55 (1937) 339-47 (the Didache as a proto-Montanist work); Middleton, "Eucharistic
Prayers," 259, 267; William Telfer, "The 'Plot' of the Didache," JTS 45 (1944) 141-42 (the author
• a superficially Christianized, originally Jewish docu ment de duabus viis; had taken place in the situation of the churches, or, as far as the Two Ways section is concerned, expressly to bring the tradition to the level of specifically Christian eschatological awareness, is another issue.21
• a (written or oral) archaic liturgical tradition con cerning baptism and Eucharist;17 • a (probably written) also archaic tradition concern ing the reception of itinerant charismatics;18 and finally:
In what follows, I will show (1) how the Didachist dealt with his sources in detail, and (2) what can be said about the possible New Testament sources of the Didache.
• a brief apocalyptic description of the events of the end time.19
This material the Didachist compiled to form a whole (the Didachist is the compiler), expanded it by means of insertions, and interpreted it (the Didachist is the redac tor). Particularly in the final section (before the apoca lypse, i.e., in chaps. 12-15) the Didachist expanded the work with text from his own pen (the Didachist is, to a modest degree, also an independent author). He had two principal tendencies: on the one hand, he wished to create a book of rules for his communities; on the other hand, at the same time he did not wish to create some thing of his own but to hand on the traditions of his church (which amounted primarily to Jewish-Christian traditions).20 That it was necessary, in the process, to augment the tradition to accommodate the changes that 1. The Didachist placed the Two Ways tractate at the beginning of his book, commencing the citation at 1.1 without any introduction or preparation. The quotation of this source extends from 1.1 to 6.1;22 nevertheless, the Didachist (as redactor) made additions to the tractate that gave it a new meaning:
• He reinterpreted the older tractate as a baptismal catechesis (ravra rcaura irpoei'KOvreq, ^a-KTcoare in 7.1 is redactional).
• He more strongly Christianized the tractate, which had already been superficially Christianized, by inserting in 1.3b-6 (in the so-called sectio evangelica) typically Christian passages from the Jesus tradition as well as other traditions available to him (the quo tation in 1.6 is also an addition by the Didachist to an existing tradition);23 all of which places the Two in chaps. 12-13. See below. Now to the contrary:
Wengst, Didache, 23 n. 83.
19 I am inclined to trace 16.3-8 also to a written source. Wengst, Didache, 23: "Chapter 16 contains a little apocalypse that, at least in its individual com ponents, is altogether traditional. Whether it was found by the Didachist as a complete whole or whether he himself put it together is a question that must remain open."
20 Stempel ("Lehrer," 214-15) sees the Didachist as bibaanaXoq. De Halleux ("Ministers," 319) desires fabricated his writing "to be the work of the apos tolic council of Jerusalem," 142). The thesis is sys tematically developed by Vokes, Riddle, passim: the author of the Didache is said to be a moderate
Montanist, and his work "is an attempt to express Montanism in apostolic terms" (p. 144). "The Didache is Montanism dressed up in an ill-fitting
New Testament garb" (p. 172). (Vokes, however, later lost faith in his Montanism thesis: Rordorf,
"Two Ways," 150 n. 18.) In the tradition of the fic tion hypothesis is also Layton, in his study of the sectio evangelica ("Sources," 343-83). The following to see in the Didachist a noncharismatic bcbdoKCtXoq, and emphasizes (certainly correctly) his atti tude of fidelity to tradition.
21 Unlike the representatives of the fiction hypothesis, I do indeed believe that both the sources and the redaction of the Didache reflect real situations.
22 It is possible that at 5.1 the redactor altered the source by rearranging part of it.
23 Also redactional is Kara rr)v cvtoXt)v in 1.5, and ev objections may be raised against the fiction hypoth esis: (1) It is usually combined with the supposition that Didache 1-6 depends on Barnabas 18-20; this supposition is false, however, as shown above. (2)
The liturgical parts of the Didache are ancient and are not fictional; they differ even in their language from the diction of the Didachist. (3) In chaps.
11-16 there are two different layers, one predidachist and one redactional; this alone destroys the fiction hypothesis.
17 This is the common opinion today, e.g.: Voobus, Liturgical Traditions, passim, esp. 169-70; Rordorf and Tuilier, Doctrine, 92-93; Wengst, Didache, 23.
18 I believe I have shown ("Itinerant Radicalism,"
323-27) that there is predidachistic tradition in 11.4-12, which was partly modified by the Didachist
CKKXrioLq in 4.14 is probably didachistic redaction as well.
24 This problem did not exist for the Two Ways trac tate. The underlying tradition considers the demands of the Jewish catechism as obviously cap able of being fulfilled. They have no eschatological character.
Introduction Ways tractate in an entirely different light: it is now clear to the readers that they are confronted not only with a rigorous Jewish or Jewish-Christian moral catechism, but beyond that with what could be called (in our terms) the eschatological demands of Jesus — although, certainly, the problem of the possibility of fulfilling such demands immediately arises here as well.24 The Didachist solves the prob lem with the aid of the concept of reXeiog (/cat, ear) reAeioq in 1.4 is his addition to the sayings tradi tion), which recurs in 6.2 (again from the pen of the Didachist):25 the fulfillment of Jesus' paradoxical demands is a matter for the reXecoq ("perfect").26
The insertion of the sectio evangelica, however, caused the Didachist to make another alteration.
Now the introduction in 1.2 and the additions in So much for the redactional interventions in the reproduction of the tractate. It is difficult to decide whether the Didachist made other interventions. The double love commandment in 1.2, as well as 4.1b, 4.1c, and the conclusion of 4.10 probably belong to the predidachistic recension of the tractate.
In 7.1 the Didachist begins to reproduce the liturgical tradition of his church (7.1-10.7). It is possible that he already had a fixed, written "agenda" as a source. What is clear is that he expanded his tradition by commenting on it. The detailed assignment of the text to tradition or redaction, however, is not always easy. The baptismal for mula and the rule for baptismal fasting (7.4a) are taken from the tradition; I am inclined also to regard the rubric (7repi be rod fia'KTiop.a.Toq, ovm /So:7mcm76 in 7.1a) as a quotation from the source (if we really can suppose that a written source was in use). I have already noted that 7.1b (ravra 7tavra 'Kpoentovreq, ^airrioare,
"Having said all of these things, baptize") is redactional.
I consider vv. 2-3 redactional as well. The Didachist here interprets and corrects the Jewish-Christian demand of his tradition, according to which baptism should be done ev vbarc Jioim ("in living water").
Probably redactional also is 7.4b.27 What the Didachist writes in 8.1 about fasting in general certainly stems from tradition. The same is true of 8.2-3, where only the reference to the gospel in 8.2 (cog eneAevoev 6
Kvpcoq ev rw evayyeXtoo avrov, "As the Lord command ed in his gospel") is the work of the Didachist. We will encounter other, similar references to the "gospel" later.
The table prayers (9.2-4) certainly come from tradition; and if one may suppose a written source, the rubrics
(9.1, 2a, 3a) must stem from it. In contrast, I consider 9.5 redactional. In /cat yap rrepl rovrov elprjKev o
Kvpcoq ("For also concerning this the Lord has said"), the Didachist betrays his presence (cf. 1.6); v. 5a can also be traced to him. The prayers after the meal (10.1-7) are certainly another old liturgical tradition, but it is diffi cult to determine where the end point comes. Probably
1.3b-6 are seen to be (or we must understand them to be) 7rpcorrj evroXx), and he causes the further text of the tractate to follow from 2.2, calling it explicit ly bevrepa evroXr] rrjq Sidaxvs (2.1 is redactional).
With this, the significance of the sectio evangelica for the understanding of the whole is once again underscored. One cannot call this division into 7rpcoTi7 and bevrepa evroXx] an especially skillful arrangement, but it should be sufficiently clear what the Didachist intended to achieve by it.
At the end of the text available to him, the Didachist made another redactional intervention: the tractate ended (as we can-now see from Doctrina apostolorum) with a parenetic epilogue, including an eschatological prospect; the Didachist retained only the beginning (5.2 [end], and 6.1), eliminating the rest, or that part that he could not use, and replac ing it with an appendix from his own pen (6.2-3).
As a result, 6.2 (as I have already shown) acquires a function with respect to the interpretation that the Didachist gives to the demands of the sectio; 6.3, in contrast, touches a special problem that was not in the source but that the Didachist wanted to cover.
25 See the respective passages in the commentary. cal source (in 7.1, ravra iravra irpoeurovreq is
26 Willy Rordorf, in particular, has instructively shown redactional). Redactional also are 7.2-3 and 7.4b. that the Didachist desires to establish a connection Cf. Rordorf and Tuilier, Doctrine, 92-93. between the sectio evangelica and his remarks in 6.2: cf. "Tradition apostolique," 108-9, and Rordorf and Tuilier, Doctrine, 32-33.
27 Rordorf, "Baptism," 212-22: 6.3 (except for the concession) and 7.1 belong to the Didache's liturgi-
didachistic conclusion to the Two Ways tractate). Also from the Didachist are raura irdvra rrpoeLTrovTeq, the tradition extends to 10.6 (inclusive), and only 10.7 again belongs to the Didachistic redaction. But we will have to treat this difficult question below.
(3airTicroLT6 in 7.1b; 7.2-3; 7.4b; the reference to the gospel in 8.2b; 9.5 and 10.7; the transition in 11.1-2; the superscription in 11.3 (with reference to the gospel); from the subsequent piece of tradition, possi bly bebompaopevoq dXr)$iv6q in 11.11 (or the entire verse?). Didachistic are also 12.1—15.4 (without 13.4) and 16.7. In addition, there are passages formulated by the Didachist, even though based on tradition. (2)
Characteristic of the Didachist's language31 are expres sions like Tcpooexeiv oar6 (6.3; 12.5); bibovac Kara rpv evToXrjv (1.5; 13.5, 7); reXeioq elvai (1.4; 6.2); bvvaodat (6.2, 3; 7.2); also typical of him are the expressions he uses to introduce quotations (includ ing those from Scripture): ctAAd: kol'l irepi tovtou be elpprai (1.6); nod yap trepi tovtov elpRKev 6 Kvpioq
(9.5); ctAA' <bq eppedp (16.7). (3) The redactor is a tra ditionalist. He makes an effort to recur to traditions, but is equally anxious to apply the traditions (here I am referring to liturgy and law) to the circumstances of his own time; the Jewish-Christian moral catechism in chaps. 1-6 is "eschatologized" (sit venia verbo) by additions. (4) The constant effort to secure what is said by recourse to the ypa<pr\ or to the Kvpcoq (who speaks in the "gospel") is striking: 1.6; 8.2; 9.5; 11.3;
The analysis is easier in the "canon law" part of the Didache. I believe that this part is based on an old tradi tion (which I now think was written) that forms the core of it (11.4-12) and in which the Didachist may have intervened redactionally in v. 11. Everything else is from his pen (11.1-2, 3; 12.1—15.4),28 with only 13.4 to be bracketed as a later gloss. Here the Didachist seeks throughout to apply the motifs of his old tradition to the new situation in which he finds himself; here, and here alone, does the Didachist emerge in lengthy passages as a genuine "author."29
Finally, the Didachist desired to close his writing with an eschatological conclusion. I infer that the special impulse to do this grew out of the conclusion of the "basic document," the Two Ways tractate. There is such an eschatological conclusion now in chap. 16, but it is immediately obvious that 16.3-8, which represent a little apocalypse, are a separate unit. I believe that here, too, the Didachist is quoting an older tradition (probably written). Thus I consider 16.3-6, 8 the quotation of an old, predidachistic apocalypse, and 16.7 an addition by the didachistic redaction.30 An assessment of 16.1-2 is more difficult. In any case, the Didachist is not formulat ing with entire freedom here either: the logion in 16.1 derives ultimately from the Jesus tradition, and the admonition in 16.2 has parallels elsewhere, although one need not suppose that there was a written source.
14.3; 15.3, 4; 16.7.32
In all this, the redactional activity of the Didachist is evident. It is clear what he has in mind, and what he intends to accomplish. There is, however, still a question to be discussed in this connection, one we previously set aside: that of the possible use of New Testament texts in the Didache. We now turn to this issue.
To sum up: (1) Traceable to the Didachist are: the additions to the sectio evangelica (nai eaij reXecoq 1.4;
Kara tt)p evToXpv 1.5; and the quotation without introduction in 1.6); also the transition in 2.1; per haps ev €KK\r]OL(X in 4.14, but in any case 6.2-3 (as 2. The question of the possible use of New Testament texts in the Didache is especially complicated.33 Are any
28 To the contrary now Wengst, Didache, 23.
29 Characteristic for him is the reference to the "gospel" in 15.3 and 4.
30 The scriptural proof (cf. also 1.6 and 14.3, both didachistic) is redactional.
31 In Muilenburg, Literary Relations, 137-39, some cor rect information is combined with a great deal that possible third citation is 1.6, although it is uncertain what text is cited. See below, ad loc.
Ehrhard, Altchristliche Litteratur, 58-60; Harnack, "Prolegomena," 69-81; Schaff, Church Manual,
78-94; Lake, "Didache," 24-36; Robinson, "Problem," 340-42; idem, Didache, 48-68; idem,
"Barnabas and the Didache," 227-38; MacLean, "Introduction," xx-xxii; Greiff, Pascharituale, is misleading.
32 The OT is cited explicitly, in the form of a reflective quotation, only twice in the Didache (cf. Hagner,
163-94; Muilenburg, Literary Relations, 73-75; Middleton, "Eucharistic Prayers," 249, 266; Vokes,
Riddle, 93-119; Goodspeed, "Doctrina," 231, and frequently; Erwin R. Goodenough, "John a Use, 25). Both quotations are attributable to the redactor: 14.3 = Mai 1:11, 14; 16.7 = Zech 14.5. A
Introduction New Testament texts used, and if so, which? At an early stage attention was drawn primarily to quotations from the Gospel of Matthew, or mixed texts from Matthew and Luke.34 In addition, dependence on the Gospel of John has occasionally been suggested.35 There have also been suspicions of literary dependence on individual
Primitive GospelJBL 64 (1945) 174-75; Johnson, Motive, 112, and passim; Massaux, Influence,
3.14482; Moule, "Note," 240-43; Richardson, "Teaching," 163-66; Koester, Synoptische Uberlieferung, 159-60, and passim; Audet, Didache, 166-86;
Glover, "Quotations," 12-29; Bosio, Padri apostolici, considers "that. . . the Gospel of Matthew or an extract from it was available to the author of the Didache, although changes brought about by oral use in the Didache community are to be reckoned with. These changes could have been affected also by the Old Testament, Jewish apocalyptic ideas, or other Jesus logia" {Matthaus und seine Gemeinde, 141 n. 12); cf. 164-65: "The dependence on Matthew or his tradition is unmistakable." Inspired by Schweizer
(but not completely identical with the opinions dis cussed above) is the hypothesis of Aono,
Entwicklung, 164-89: the Didache is not directly (lit erarily) but indirectly dependent on Matthew. Aono
{Entwicklung 186) supposes "that the author of the Didache probably did not even have a written copy of the Gospel of Matthew before him, but he knew the quoted phrases by heart because they were repeated again and again in the community, literal ly or with variations; it is not impossible that the wording of Ps 36:11 or the recollection of Jewish sayings or of an older form of a dominical saying may have influenced the reformulation of the words as found in Matthew, and that the baptismal formu la and Our Father are written in the form in which they were used in the community's liturgy, which is to say in a liturgy with a Matthean stamp." Aono himself {Entwicklung 186-87) sees the difficulty cre ated for this hypothesis by the unique character of Did. 1.3-5 (where there are echoes not only of Matthew but clearly of Luke as well), but he thinks he can resolve it by a hypothesis of addition: Did.
10-11; Lucien Cerfaux, "La multiplication des pains dans la liturgie de la Didache," Bib 40 (1959)
943-58; Barnard, "Later History," 99 n. 2; Layton, "Sources," 343-72; Hagner, Use, 280; Eduard
Schweizer, Matthaus und seine Gemeinde (SBS 71; Stuttgart: KBW, 1974) 140-41, 164-65; Aono,
Entwicklung 174-89; Rordorf and Tuilier, Doctrine, 83-91; Rordorf, "Transmission textuelle," 499-513;
Kloppenborg, "Matthaean Tradition," 54-67; Wengst, Didache, 24-31.
34 For early scholarship: Ehrhard, Altchristliche Litteratur, 58-60.1 mention only selected works from more recent scholarship. The question is treat ed at length by Massaux, Influence, 3.145: in all four passages where the key word evayyeXion occurs
(8.2; 11.3; 15.3, 4), it reveals "a definite literary con tact" with the Gospel of Matthew (3.145). "Gospel" in the Didache means the Gospel of Matthew (which does not exclude the possibility that the author also knows and makes use of Lukan and other NT mate rials); but "the gospel" as such for this author is
Matthew. Cf. also 3.176, 179. Massaux (Influence, 3.14476) attempts to demonstrate in detail that the Didache is, in fact, literarily dependent on Matthew in a great many places. According to Massaux
{Influence, 3.180), the way in which the Didache quotes the First Gospel recalls Justin's manner of citation. For Massaux, the Didache as a whole seems
"like a catechetical summary of the First Gospel"
{Influence, 3.179, 180). See further Bosio, Padri apos tolici, 11; Johnson, "Motive," 112, and passim: the Didache is dependent on Matthew and Luke (and Acts as well). "Matthew was his favorite" (p. 112); the mixtures of Matthew and Luke are explained by quotation from memory (p. 112 n. 14). Richardson
("Teaching," 163, 165-66) distinguishes between sources and redaction; in the sectio evangelica and chap. 16 (redactional) the Didachist reveals "a wide knowledge of New Testament Scripture" (p. 165).
"He conflates Matthew with Luke and cites, among other things, Barnabas and Hermas" (ibid.). "The rest of the work reveals only a knowledge of Matthew's Gospel" (ibid.). Barnard, "Later History,"
99 n. 2: the Didache "depends on the Matthaean stream of the Gospel tradition." Eduard Schweizer
1.3-5 may reflect a "collection of dominical say ings," formed "primarily from Matthew" but enriched by Lukan elements (p. 187). Wengst
{Didache, 19) now thinks of the Gospel of Matthew, and it alone ("the compiler of the Didache knew only one written gospel, probably that of Matthew"); see further pp. 24-31, but note also that Wengst regards the sectio evangelica as an interpola tion.
35 The older discussion is in Greiff, Pascharituale, 163-94. From more recent research, cf. Erwin R.
Goodenough, "John a Primitive Gospel,"JBL 64 (1945) 174-75; Moule, "Note," 240-43. It is one of the curiosities of Didache scholarship that the con trary has also been posited, namely, dependence of the Gospel of John on the Didache. See Greiff,
Pascharituale, 184-94.
Testament texts. This question will be discussed below.
Still, it is characteristic that the only texts that deserve serious consideration are from the Synoptiq tradition, more specifically the Gospels of Matthew and Luke.
There are no genuinely demonstrable allusions in the Didache to the Gospel of John, and one may well ques
Pauline letters and other New Testament writings.36 The advocates of the fictional hypothesis have obviously been most prominent in positing dependence on New
Testament writings. According to them, the author of the Didache — with the intention of faking a location within apostolic times — used a generous quantity of New Testament material and also imitated New
Testament expressions.37 tion whether the author of the Didache knew that Gospel at all. The Didache lives in an entirely different linguistic universe, and that is true not only of its sources40 but of its redactor as well. In this context we should add imme diately that there is no echo of the corpus Paulinum in the Didache either.41 One may not assume that the Christians of the Didache communities knew nothing at all of the apostle to the Gentiles, but it would be diffi cult to suppose that they had access to the corpus
Paulinum (even though their Antiochene contemporary, Ignatius, did). Just as there is no specific reference in the Didache to the Gospel of John, so also there is none to the Pauline tradition, and obviously, therefore, no Two initial observations are in order, and they are fundamental to the whole discussion: (1) If there are any
New Testament allusions or quotations in the Didache at all, they are found without exception in the last stage of the document, in the work of the redactor.38 In the Didache's sources (Two Ways tractate, agenda, church order, but with all probability in the apocalypse as well) there is no literary dependence on New Testament texts.39 This circumstance is adequately explained by the age of the sources in question. (2) It is uncertain whether the redactor made a literary use of New
38 Something analogous is true, strikingly enough, of the two or three quotations from the OT. All explic it quotations are from the Didachist, i.e., they belong to the last stage of the writing.
39 I must admit that this assessment is insecure with regard to the concluding apocalypse (16.3-8); how ever, I wish to maintain it for this section as well.
On this issue, see the commentary.
40 From time to time connections have been posited between the table prayers (Didache 9-10) and Johannine expressions (esp. from John 17): on this,
36 For example, MacLean, ("Introduction," xx-xxii) mentions Matthew, Luke, Acts, and perhaps 1
Corinthians, the Thessalonian correspondence, 1 Peter, and 1 John. Muilenburg (Literary Relations,
73-75) lists the sources that he thinks have been compiled in the Didache-, cf. the later remarks at pp.
91-96. The compiler certainly used Matthew and Luke. "Other N. T. traces in the Teaching are numerous, but. . . they are more obscure and vague than those just referred to" (p. 94). Cf. p. 96:
"the abundant use made of the N. T." Goodspeed ("Doctrina," 231, and frequently) posits, for the sectio evangelica, dependence on Matthew, Luke, 1
Peter, and Hermas. For Massaux and Richardson, see n. 34 above.
37 Robinson, "Problem," 340-56; idem, Didache, 48-68; idem, "Barnabas and the Didache," 227-38;
Middleton, "Eucharistic Prayers," 259, 266; Vokes, Riddle, 95-115, and passim. Vokes found reflections of the "language" and "subject-matter" of the New
Testament writings everywhere (cf. p. 95). "The Didache knew nearly all of our New Testament, not as 'oral tradition,' but as written documents" (p. see below at 9.1. Of these parallels, the only ones that deserve serious consideration are yvopi^Eiv in Did. 9.2, 3; 10.2 (John 17:26; cf. 15:15); the motif of ev yevEodai in Did. 9.4 (cf. John 17:11, 21-22; and perhaps also 11:52); the address ttcxtep ay le in Did.
10.2 (cf. John 17:11); the saving of the church card iravToq Trovr)pov in Did. 10.5 (cf. John 17:15); the motif of TcXciuiOLq in Did. 10.5 (cf. John 17:23).
One should note in all this that these contacts all rest only on individual expressions that are found here and there in different contexts. That there are contacts at all is probably explained by the fact that John 17 (and the other parallels mentioned as well)
115). In a later essay ("The Didache and the Canon of the New Testament," StEv 3.2 [TU 88; Berlin:
Akademie-Verlag, 1964] 427-36) Vokes attempted to found his late dating of the writing (with the sup position that the Didache quotes the Gospels) in terms of canonical criticism. Cf. from recent research also Layton, "Sources," 343-72 (according to whom the sectio evangelica depends on Matthew,
Luke, and Hermas). has been penetrated by motifs from eucharistic lan guage (cf. Voobus, Liturgical Traditions, 128 n. 98).
That is all. By contrast, precisely those things that are specifically Johannine are absent from the Didache.
41 Audet (Didache, 340-43) expressed the suggestion that Paul (the reference is actually to the deuteroPauline letters) was dependent in some way on the
Introduction 15.3: eXeyx^Te be aAAriAovq pi) ev opyfj, ctAAd ev eippvr), cog e'xere ev tu> evayy eA'up.
15.4: rag <5e evyotq vpwv ncd rag eAeppoovvaq Kai naoaq rag irpa^ecq ovtu> iroipaaTe, cog exere ev polemic engagement.42 The only New Testament contact the Didache shows is with the Synoptic tradition.43 This is a remarkable circumstance. Together with the JewishChristian character of its sources, this gives the Didache its unique aspect. rco evayyeAlco too Kvpiov rjpcov.46
What is immediately obvious is the symmetry of the diction:
If we attempt to make this latter relationship more precise, we may also begin with the judgment that the sources of the Didache reveal no literary dependence on ev tu> evayyeALcc avrov (8.2) ev tw evayyeAltp (15.3) ev tco evayyeAltp tov Kvpiov rjptiv (15.4) the Synoptic Gospels (and it is only the Synoptic
Gospels, more precisely Matthew and Luke, that can be considered at all, in light of what has already been said).
Dependence could, at most, be present in the insertions and expansions produced by redaction. But is that the case? cog eKeAevoev 6 Kvpcoq (8.2) Kara to boypa tov evayyeAiov (11.3) ovtw irpooevxtode (8.2) ourco TroipaaTe (11.3) ourco TTOLTIOaTe (15.4).
All these formulations express the same concept, and it should be obvious that all of the texts can be traced to one and the same author, namely, the redactor.
We begin the discussion with an analysis of the four instances of evayyeAcov in the Didache.44 Like Koester,45
I will first list the passages:
8.2: cog eneAevoev 6 Kvpioq bv m evayyeALcc avrov, ouTo) Trpooevxcotie . . . (the Lord's Prayer fol lows); 11.3: irepi be twv (xttootoAup Kai Trpocprjmu Kara to boypa tov evayyeAiov ovtui itocrjoaTe (followed by rules for behavior toward itinerant teachers);
Thus we must maintain, first of all, that all the exam ples of the key word "gospel" belong to the redactional layer of the Didache, as Koester had already noted.47 It
/ This approach is highly problematic; see above, §lb.
43 It would be wrong to see here an analogy to the Gospel of Thomas, for we must immediately add that, of all things, Gnosticism plays no part at all in the Didache. The communities addressed were neither threatened by Gnosticism nor are there latent gnos tic or even gnosticizing motifs in its sources or redactor. Gnosticism is not a problem for the Didache.
44 For the whole, see Koester, Synoptische Uberlieferung 6-12, esp. 10-11.
45 Synoptische Uberlieferung 10.
46 A fifth passage is clearly to be understood as the quotation of a dominical saying, even though the key word "gospel" is not used: 9.5: Kai yap Ttepi
"instruction for the poor." On this claim, see my remarks below at 4.1.
42 Johnson ("Motive," 114) writes somewhat different ly: "The Twelve are the Didache's heroes, not Paul.
The Didache is not necessarily anti-Pauline, but it is certainly non-Pauline." Still, Johnson sees a possibil ity that one could find an anti-Pauline polemic in Did. 6.3 (pp. 114-15). In these suggestions, only one thing is correct: the Didache has no recognizable relationship to Paul, or in any case no polemical relationship. See also Andreas Lindemann, Paulus im altesten Christentum (BHTh 58; Tubingen: Mohr,
1979) 174-77, esp. the summary on p. 177.
Lindemann considers it probable that the Didachist "knows nothing of Paul" and founds this on the sup position "that the Pauline tradition appears to have been completely unknown in Syria toward the end of the first century." That is no more than a suppo sition. Lindemann concludes correctly with the observation, "it [would be] wrong to reproach the Didache for its neglect of Pauline elements, or to see it as representative of an anti-Pauline church estab lishment" (p. 177). Cf. also Aono, Entwicklung, tovtov elpr)K6v 6 Kvptoq■ Mi) Score to ayiov roiq
KVOi'
47 Koester, Synoptische Uberlieferung 10. In accordance with his own conception, Audet (Didache, 112) want ed to establish a difference between the use of "gospel" in D1 and D2. To the contrary, and correct ly, Nautin, "Composition," 195-96. For Audet, Did.
11.2 does not quote a written gospel, but this is the case in the additions by D2 from 11.3 onward; here, according to Audet, a gospel similar to Matthew is being quoted. Cf. pp. 178-79, 442.
163-64. Aono posits the location of the Didache "outside the sphere of Pauline influence" (p. 209).
Of course, he attempts to understand the "theolo gy" of the Didachist and to perceive its omissions.
after), the didachistic concept of "gospel" has the charac ter of a new (eschatological) commandment; one might almost say that it is a "new law of Christ."51 The word of Jesus is a word that brings salvation — that is why it is was the redactor who first introduced the key word
"gospel" into his sources, or, to put it another way, this key word appears only in redactional passages.48
Next, it is striking that (setting aside the question still to be discussed, whether "gospel" in the Didache refers to oral tradition or to a literary text) at no point does the content of "gospel" refer to the christological kerygma. In the Didache, "gospel" refers not to the missionary mes sage of the epiphany, death, and resurrection of Jesus for our sake. The word always and exclusively refers to the words of the Lord as handed down,49 the Jesus tradi tion in the Synoptic sense, or more precisely, the refer ence is always solely to the words of Jesus.50 This, obviously, does not mean that the Didachist did not know the christological kerygma or did not accept it, but simply that the word "gospel" is connected, whether con sciously or unconsciously, with the logia of the Lord as transmitted. They are the gospel of which the Didache speaks, and in this sense it is called to evayyeXcov too
Kvpiov ppuiv ("the gospel of our Lord," 15.4; cf. 8.2).
"good news" — but it is also, and at the same time, the binding norm for Christian behavior, and to that extent the Lord "commands" in his gospel (cog eneXevoev 6
Kvpcoq ev tw evayyeXiu avrov, 8.2), or the gospel reveals the binding commands that Christians must obey
(hence: to boy pa too evayyeXcov, 11.3).52 Character istic of him are also the occasions that cause the redac tor to reflect on the gospel, the moments when it is time to establish clear and binding rules. From the gospel, one may derive the correct rules for daily prayer (8.2), for dealing with apostles and prophets (11.3), for con duct within the community (15.3), and in general for prayer, almsgiving, and the whole of life (15.4).53 The gospel grounds the parenetical imperatives outco irpooeux^ode (8.2), outco TrocrjoaTe (11.3; 15.4), eXeyx^Te
(15.3). Overall, "gospel" for the Didachist is, so to speak, the "regula Christi,"54 and indeed he writes: ev tu evayyeXito rou Kvpiov rjpuiv (15.4) and ev tco evayyeXiu avtov (8.2). Jesus' words (= his gospel) regulate the life
Moreover, because the words of the Lord are obliga tory norms (something that was a matter of course for all of primitive Christianity and for every period there
48 In my opinion, Did. 9.5 (see n. 46 above), which quotes a dominical saying without using the key word "gospel," is also to be traced to the didachistic redaction.
49 Koester, Synoptische Uberlieferung 11: "gospel" refers to the "rules and admonitions" of Jesus.
50 Glover, "Quotations," 28: there are no biographical details; the words "all refer to teaching." Glover con cludes from this that a sayings collection existed.
On this, see below.
51 That the Didachist understands the words of Jesus as eschatological admonitions is clear from the way in which he fixes and concretizes the Two Ways trac tate by means of the sectio evangelica, but also from his tentative efforts to accommodate Jesus' demands for perfection to the moral possibilities of the majority.
52 One should not accuse the Didachist, in this con nection, of confusing law and gospel. That would be a misunderstanding of the Didache. Paul himself never applies the dialectical relationship between law and gospel to the words and commands of Jesus. When he speaks of the law that kills, he the "law of Christ," thus not "a law consisting of works," but "a law of faith," a "law of the Spirit," the "law without legalism," the law of the new eon that is no longer misused by the flesh, and that therefore no longer alienates. The Didachist did not indulge in reflections on the question of law, command ments, and instruction comparable to those of Paul.
But in equating "Jesus' commands" as "absolutely binding" with "gospel" he cannot be criticized from the point of view of Pauline theology, unless it is falsely understood in modernist fashion.
53 The question of admission to the meal is regulated in 9.5 by application of a dominical saying.
54 Cf. Koester, Synoptische Uberlieferung, 11; Rordorf, "Tradition apostolique," 108-10. This "rule" con sists of the sum total of the traditional words of Jesus. The question then arises: What is the rela tionship of the Didache as rule to the gospel as rule
(even if the author of the Didache may not have posed this question to himself)?
55 In this connection, one must attend (cf. Rordorf, "Tradition apostolique," 110) to the multitude of things definitively regulated by words of Jesus: the life of the reAetot (1.3b-5.2), prayer (8.2), the exclu sion of the unbaptized from the meal celebration
(9.5), the way to treat apostles and prophets (11.3-12), discipline within the community (14.3), means the law of the old covenant as it was misused by human beings. It would be meaningless — at least for Paul — to understand Jesus' commands as "law" in this sense. Instead, for Paul, Jesus' commands are
Introduction and behavior of his followers.55 That is what "gospel" means in the Didache.56 stood evayyeXiov in 15.3-4 as evangelium scriptum,59 But not only is this hypothesis of a double redaction ques tionable, in my opinion; the conceptual consistency of the clauses in which the key word "gospel" appears does not suggest such an idea. Finally, we cannot completely exclude (c), the possibility that the redactor means the same thing in all four passages, namely, a written gospel,60 known to him and his readers and, because it preserves the words of Jesus, serving as a norm of faith and life. Koester also accepted this construal for 15.3-4, and considered it possible for 11.3; even for 8.2 he did not wish to exclude it absolutely.61 (Nota bene: This understanding does not mean that the author also quotes this gospel.) If this position is accurate, we might have in Did. 8.2; 11.3; 15.3-4 the oldest instances of evayyeXiov meaning a written gospel or gospel book. But to which gospel would they refer? Could it be canonical
Matthew?62 Or should we suppose that it means an apoc ryphal gospel63 to which the Didachist refers and which he may quote in various places? In that case, one might think of an apocryphal collection of sayings of the same type as the Synoptic Sayings Source.64 That, however, is only a hypothesis. The jury is still out.
While the language of the Didache (in this case we should say the Didachist, or the redactor) is clear and consistent in itself, there remains the difficult problem of saying whether the Didachist simply understands the regula Domini that he calls eiiayyeXtov (and the readers' knowledge of which he presumes) as oral tradition, or whether it already exists in a literary work — that is, whether the word "gospel" in this case already refers to an evangelium scriptum (a written gospel, or more than one). Unfortunately, this question cannot be answered with certainty. Three solutions deserve serious consider ation.
(a) All four passages refer to oral tradition (but can 15.3-4 be understood in this way?), (b) EvayyeXiov refers to different things in the four passages: in 8.2, and perhaps in 11.3 as well, we are still dealing with the older usage (meaning the "living voice of the good news" [viva vox evangelii]); in contrast, evayyeXiov in 15.3 and 4 (but also in 11.3?) already refers to a gospel in written form (evangelium scriptum): thus Koester.57
Audet adopted a similar opinion, seeing in 8.2 and 9.5 reference to an oral gospel, and in 11.3 and 15.3-4 to a written gospel.58 Rordorf and Tuilier employed this notion to distinguish a hypothetical second layer of redaction: only the second redactor could have under 3. By way of note, we should at least mention the question of the dependence of the Didache on Hermas. In research on the Didache, this question is usually discussed together with that of dependence and the expectation of the end (16.1-2).
56 I should add that the term evayyeXiov appears reg ularly only in the singular. This fact need not mean that the term intends — presuming a written gospel
"In the community of the Didachist the Gospel of Matthew was probably known as 'the gospel' or 'the gospel of our Lord,' and he used it for his own writ ing" (p. 30).
61 Cf. esp. Synoptische Uberlieferung, 203.
62 The sectio evangelica, with its "Lukan" elements alongside the "Matthean," need not refute this hypothesis, not even if (differently from Wengst) one considers the sectio to be authentic, because the author need not quote the written gospel that he knows. The elements of the Jesus tradition in the Didache (in the sectio and elsewhere) could stem from the oral tradition that still circulated alongside the written gospel that the communities now pos sessed. Cf. the suggestions in Koester, Synoptische
— to refer only to one individual gospel, but it does suggest the possibility of such an intention.
57 Synoptische Uberlieferung 10-11. Cf. Koester's very cautious judgment at pp. 203 and 240.
58 At the time of D' there was not yet a written gospel in the milieu for which the author wrote (Audet,
Didache, 173). The case is different with D2. In the meantime, such a written gospel was prepared and circulated in the milieu of the Didache (Audet,
Didache, 176; cf. earlier at p. 114). D2 refers to this writing. This gospel was known to the Didache com munities and accepted by them. This text was not, however, our Gospel of Matthew (pp. 178-83) but merely the written version of a tradition related to Matthew (p. 182).
59 Rordorf and Tuilier, Doctrine, 88. Cf. earlier, Rordorf, "Remission des peches," 293.
60 Thus now also Wengst (Didache, 26), who thinks that the reference is probably to the Gospel of Matthew.
Uberlieferung, 240.
63 Audet's reflections tend in this direction (although only for D2), when he presupposes for this stage of the writing the knowledge of a gospel book that was related to Matthew but was not quoted word for word. Cf. n. 58 above. Cf. also the suggestions of Rordorf, "Remission des peches," 293 and n. 4.
64 Ultimately, the remarks of Glover ("Quotations")
character and stem from different strands of tradition. on Barnabas; we have already discussed the latter above. Regarding Hermas, the only passage worth serious consideration65 is Did. 1.5c (cf. Hermas Man.
2.4-6).66 On this parallel, see the discussion below at 1.15c, where I will show that dependence on Hermas
Characteristic of the Didachist is a mentality that holds fast to tradition while, at the same time, having a tendency to reconcile traditions with the changed cir cumstances of a church undergoing a process of stabi lization. This observation alone provides little information, however, with regard to a temporal loca tion of the writing.69 In order to proceed farther, one could argue that the archaic elements that apparently speak in favor of an early origin of the Didache all belong to the work's sources; thus the redactor cannot be locat ed very early. On the contrary, the redactional sections do not reveal any great degree of church development; there are still itinerant prophets in the Didache's field of tradition, although some of them seem to be in the process of becoming resident (13.1). There is still no monarchical episcopate or ordo triplex (cf. the eirLOKonoL and biCCKOVoi in 15.1), and the relationship between the local clergy and the charismatics is not without conflict
(15.1-2).70 Thus one cannot move too far into the sec ond century, even if one keeps in mind that the develop ment of canon law within the tradition of different regions of the church as a whole proceeded at widely is improbable.67 The agreements are more probably attributable to a common tradition. b.Time and Place ofWriting We should begin68 with a distinction between tradition and redaction. The predidachistic traditions, that is, in the narrower sense the sources of the Didache, have an altogether archaic character. This is true of the Two
Ways tractate, the liturgical formulae, and with special clarity of the piece of tradition about the reception of itinerant charismatics; the only thing one might possibly doubt is whether the predidachistic apocalypse can be called "archaic" tradition. In general, one can say that the sources, that is, the predidachistic traditions, should probably be located in the first century c.e., most likely toward the end of the century. It is impossible to make any more precise determination. One should also con sider that the predidachistic traditions are varied in tend in this direction (variants of Q?): "Out of my examination grew the convictions, first, that the Didache does not bear witness to our gospels, but by Bryennios, Aibayr), 7tS'; Zahn, Geschichte,
3.315-18; Hennecke, "Grundschrift," 71; Muilenburg, Literary Relations, 33, 46, 167, and passim;
Dix, "Didache," 243-44; Robinson, "Barnabas and the Didache," 233-38; Connolly, "Streeter on the Didache," 372, 377-79; Vokes, Riddle, 48, 51-61, and passim; Richardson, "Teaching," 163, 165: the Didachist (not his sources) is dependent on Barnabas and Hermas; Layton, "Sources," 361-72.
67 On the contrary, positing dependence of Hermas on the Didache are Harris, Teaching, 38; MacLean,
"Introduction," xxiii (probable).
68 Early scholarship is reviewed by Ehrhard, Altchristliche Litteratur, 62-65. Origins ca. 50 c.e. constitute the earlier limit, and origins in the 4th and 5th cen tury are the late limit. (With regard to the latter,
Ehrhard remarks that such opinions "can in any case only arouse a pathological interest" [p. 62].)
69 Recourse to the witnesses does not bring us any far ther: on the one hand, because the oldest testi monies are not clear, as we have seen above, and on the other hand because we do not know how far behind the testimonies we may go.
70 No certain argument can be derived from what is said about the evayyeXiov or the elements of Synoptic tradition found in the Didache, because these instances themselves are not absolutely clear quotes directly from sources used by Luke and Matthew; secondly, that Justin possessed the same sources, at least in part; thirdly, that the Didache may sometimes preserve our Lord's sayings in a more authentic, or, at least, more primitive, form than that formed in either Luke or Matthew" (p.
12). Glover posits as source "a collection of sayings"
(p. 28). "At the least, then, it seems sound to say that Justin and the Didache are witnesses who should be examined alongside Luke and Matthew when the text of Q is discussed" (p. 29). Rordorf ("Remission des peches," 293) also considers the possibility that (although only in 15.3-4) "gospel" may mean an apocryphal gospel of the Q type (and that it may also be quoted in other places?). For possible quota tions, cf. the cautious questions on p. 293 n. 4.
65 Harnack (Lehre, 18; and "Prolegomena," 87) still referred to Didache 5/Hermas Man. 8.3-5. This is not obvious. Harnack originally ("Prolegomena,"
66, 87) supposed dependence, but later he found it doubtful (Apostellehre, 13). The references in Connolly, "Streeter on the Didache," 377-78, are very far-fetched.
66 Dependence of the Didache on Hermas was posited
Introduction differing rates. In sum, the date of the Didache is a matter of judgment. An origin around 110 or 120 c.e. remains hypothetical,71 but there are as yet no com pelling reasons to dismiss this hypothesis. ing's early circulation in Egypt.74 A number of details also suggest an Egyptian origin.75 In any case, the liturgi cal expression about bread on the mountains (9.4) can scarcely be used as a counterargument, because 9.4 is part of the sources of the Didache. Moreover, this expres sion is also found in Egyptian liturgical sources.76 On the other hand, 7.2 does not favor Egypt but rather Syria or Palestine, and in fact that region is frequently claimed as the locus of the Didache. For Syria, or the borderland between Syria and Palestine,77 one can adduce primarily
Regarding provenance, we are completely in the dark.72 The strongly Jewish-Christian character of the Didache does not function as an argument. From that character — whether of the sources or of the redactor — one cannot deduce anything about location. On the one hand, it has occasionally been said that the Didache comes from Egypt,73 with support derived from the writ
74 Athanasius; P. Oxy. See above, §§2a, 3b.
75 KActafia in 9.4 does indeed point to Egypt, but is this the original reading? Cf. the commentary at 9.4. The correspondence of the doxology of the Lord's Prayer in Did. 8.2 (end) and Matt 6:13 in parts of the Coptic translation is striking (see the commentary at 8.2). Is that, however, sufficient to allow us to argue for an Egyptian origin of the Didache?
76 The expression is found in Serapion Euchologion 13.13 (Funk, Didascalia, 2.174) and P. Der Balizeh
IIv., 1. 4 (Roberts and Capelle, Euchologium, 26; and van Haelst, "Nouvelle reconstitution," 449 — although there partly expanded, but the expansion is fairly certain). See below at Did. 9.4. No argument for or against a particular place of origin for the Didache can be made from this datum.
77 Examples: Taylor, Teaching, 116: northern Palestine; Jacquier, Doctrine, 81: Palestine, particularly
Jerusalem; Ehrhard, Altchristliche Litteratur, 66:
(is the "gospel" oral or written? canonical or apoc ryphal? Are the Synoptic elements from oral tradi tion?). I have spoken of this above. Nevertheless, these findings in themselves give some support to the hypothesis that the Didache should be located at the beginning of the 2d century.
71 Examples of attempts at dating from the history of research include: Audet, Didache, 187-210: "contem porary with the first gospel writings" (197);
Hemmer, Doctrine, xxxv: 80-100; Funk, Abhandlungen, 2.112: perhaps 80-90; Kleist, Didache,
6: "before the end of the first century"; Ehrhard, Altchristliche Litteratur, 65: probably "80-100, or pos sibly 90-110 or 120"; Schaff, Church Manual, 122: between 90 and 100; Adam ("Herkunft," 70) sug gested that the primitive form of the Didache origi nated between 90 and 100; Aono, Entwicklung, 207:
"toward the end of the first century"; Vielhauer, Geschichte, 737: "beginning of the second century"; also MacLean, "Introduction," xxxvi; and Wengst,
Didache, 63; Barnard, "Later History," 99 n. 2:
Syria-Palestine; Dix, "Didache," 249: Syria; Muilenburg, Literary Relations, 41: more likely Syria;
Knopf, Lehre, 3: "rather Syria and Palestine . . . than Egypt"; Bardenhewer, Geschichte, 1.96: "Syria or Palestine" (more probable than Egypt); Altaner and Stuiber, Patrologie, 81: Syria-Palestine; Quasten,
Patrology, 1.37: most probably Syria; Bosio, Padri apostolici, 14: probably Syria or Palestine; Vielhauer,
Geschichte, 737: with great caution: Syria; Adam, "Herkunft," 37-42, and passim: the book stems from Palestine, perhaps from Jewish Christians from Pella, but is directed to the newly missionized communities in eastern Syria, in Adiabene; Robert
M. Grant, The Apostolic Fathers: A New Translation and Commentary, vol. 1: An Introduction (New York:
Nelson, 1964) 75-76: Syria, probably not Antioch; Barnard, "Later History," 99-100: Syria; Aono,
Entwicklung, 208: probably Syria; Rordorf and Tuilier, Doctrine, 97-98: western Syria, rural regions; Wengst, Didache, 62: "a rural region in Syria."
100-130; Altaner and Stuiber, Patrologie, 81; and Quasten, Patrology, 1.37: 100-150; Bryennios,
Atbaxv< K': between 120 and 160; Richardson, "Teaching," 165: ca. 150; Kraft, Didache, 76: after 150; Johnson, "Motive," 108: between 150 and 175;
Vokes, Riddle, 86 (cf. 87, 216): between 155 and 250; Robinson, Didache, 82: 3d century ("very probable").
72 From the early research: Ehrhard, Altchristliche Litteratur, 65-66. He observed correctly (p. 66):
"But only Egypt or Syria-Palestine can be consid ered." Ehrhard himself decided for the latter.
73 To begin with, Harnack ("Prolegomena," 159-60, 167-70) was the primary proponent of Egypt; later
(Apostellehre, 25) he was less sure ("probably, but by no means . . . certain"); idem, Geschichte, 2.1.431 n.
1: "No certain decision between Egypt and SyriaPalestine can be given." Possibly Egypt: Middleton,
"Eucharistic Prayers," 267; probably Egypt:
Richardson, "Teaching," 165; cf. Glover, "Quotations," 27; Kraft, Didache, 77: rather Egypt than Syria.
some things in its favor. In the question of a more pre cise location, Did. 13.3-7 has sometimes played a role.
Most recently, Georg Schollgen78 has attempted to show that an origin of the Didache in a rural area cannot be deduced from this passage. The Didache could certainly have come from an urban setting. I do not think, how ever, that the major city of Antioch should be consid ered as its location. the fact that the redactor still knows the movement of Jesus' disciples or its later offshoots (otherwise he would not have had to cite 11.4-12 and interpret it with chaps.
12-13). It appears that in the communities of the Didache the offshoots of the Jesus-discipleship movement are still alive; such a thing is most naturally to be sought in Palestine or Syria. Such argumentation, placing the Didache in Syria-Palestine, is not very strong but has
78 "Die Didache — ein fruhes Zeugnis fiir Landgemeinden?" ZNW 76 (1985) 140-43. Schollgen discusses the theme of the gift of firstfruits in the early period and gives instances from Rome and Caesarea. "Christians from town and even urban milieus without an agriculture of their own were apparently also in a position to produce the gift of firstfruits required for the support of the prophets and the clergy; these were probably acquired by pur chase" (p. 142). He then continues, "Yet even if the gifts of firstfruits in v. 3 come from one's own produce, this would say nothing against the origins of the Didache in an urban milieu" (ibid.). Note the following remarks about agriculture as the basis of the economy in the majority of the small cities of the empire (pp. 142-43).
The Didache /
The Title On the basis of the witnesses that testify to the docu ment5 it is much more likely that the original title was
6l8olxV (or 8l5(XX^) tuv onrooToXuv ("teaching of the apostles").6 (The reference to the "Twelve,"'as found also in the short title, would have been incorporated later.)
Two difficulties then arise. First, the document itself nowhere lays explicit claim to apostolic authority;7 sec ond, the witnesses are, in part, subject to the suspicion that they do not refer to the Didache at all but to versions of the Two Ways tractate. In fact, according to the wit ness of Doctrina, the Two Ways tractate (at least in one version) bore the title (de) doctrina apostolorum ("[of the]
The twofold title borne by this document in H has con fused scholars from the beginning. The short title (which stands in a separate line above the text in H) is didaxv tup dudeKGC onrooToXuv ("teaching [or doctrine] of the twelve apostles").1 The long title is not placed in an indi vidual line in H, but forms the beginning of the text:2
Aidaxv xvpiov Sid tup dudeKa olttootoXuv role; edveoLP ("teaching of the Lord through the twelve apostles to the nations"). The text continues in the same line with 080I dvo ("two ways").3 Much has been written about the ori ginality of these two titles (is the long title the older and the short title only an abbreviation of it?)4 and their sig nificance (especially as regards the interpretation and unique character of the book as a whole). It is probable, however, that neither title is original. teaching of the apostles").8 There is very good reason to suppose that, at a relatively early period, this was the title of the Christianized Two Ways tractate,9 and that regula vitae christianae ("rule for the Christian life").
Vielhauer, Geschichte, 723-24.
For the similar title of the Canons of the Apostles in some mss., see Audet, Didache, 94-95.
MacLean, "Introduction," xxxii n. 1 (probable); Goodspeed, "Doctrina," 231; Koester, Synoptische
It is written in minuscules (like the rest of the text), but framed by the symbol •" • .
The first letter (A) is a large initial extending into the margin.
The way in which the copyist presents the two titles
may perhaps indicate that he regarded the short title as the title of the document itself.
Cf. Harnack, "Prolegomena," 24-37.
Uberlieferung 218 (he suggests the possibility that the Christian Two Ways tractate bore the title
"apostolic teaching" before it was incorporated into the Didache); Nautin, "Composition," 213: the Duae viae in their Christian form bore the title bibaxp tup cotootoXup (doctrina apostolorum), and this tide was transferred to the Didache-, the long title would have been an addition by the copyist. The original
Jewish Duae viae would simply have been called
TUP (XTTOOTOXUP (...)
bibaxotl Doctrina (...) apostolorum Eusebius Hist. eccl.
3.25.4 (GCS 2.1.252)
Rufinus ibid. (253) Athanasius Festal Letter 39 § 11 Aibaxv ( . . . ) tup dirooToXup (Preuschen, Analecta, 2.45) Ps.-Athanasius Synopsis scripturae sacrae 76 (PG bibaxhBarnard, "Later History," 99-100: the Egyptian version of the tractate bore the title beboexv tup ccitootoXup; the Syriac version was enti tled bebaxr) tup bubeKa cotootoXup, and from that the Didachist composed the long title.
Did Barnabas know the Two Ways tractate under the title bibaxpt Cf. Barn. 18.1: peTafiupcp be Kal eiri erepap jpuolp kocl bcbaxVP ("Let us proceed to another insight and teaching").
Vielhauer (Geschichte, 725 n. 9) has disputed the idea that the title stems originally from the Two
Ways tractate and was transferred from it to the Didache. For him, the short title (without bubeKOt) is older, but was applied to the Didache only at a later time (pp. 723-24). According to Wengst (Didache,
AcbaxV onrooToXup 28.432) Indicium scriptorum canonicorum sexagesima (Preuschen, Analecta, 2.69) TtepioboL Kai bibaxod tup cxttootoXup
Ps.-Nicephorus Stichometry (Preuschen, Analecta, 2.64) Ps.-Cyprian De aleat.
4 (CSEL 3.3.96).
For the interpretation of the word bibaxv in this context, Adam ("Herkunft," 60 n. 116) has referred to Friederich Preisigke, ed., Worterbuch der griechischen Papyrusurkunden (4 vols.; Berlin: Selbstverlag,
Acbaxv (xttootoXup
doctrin(ae) apostolorum
15) the title was added later. Audet's thesis (Didache, 91-103, 247-54) combines some correct and some bizarre ideas. According to him, the Two Ways trac tate (even before it was in Christian usage) had the title bibaxp evpiov (= YHWH) rolg eftpeoip. The author of D1 entitled his instructions: bibaxed tup
(xttootoXup (but not referring to the Twelve). That
1924-71) 1.371, where (at B 140, 16, p. ii) rj orpemuTLKTj bibax'O is attested in the sense of "military order." On this, see the discussion above (Intro duction, § lb) on the genre of the document as
this title was then transferred to the Didache,10 either by the author of the document itself or in a subsequent process.11 No matter how one answers this last question, however, one should by no means understand the title
5i5ctxV (or StSaxat) tuv aixooToXuiv as an indication of how the document is to be read. Even if it was imposed by the Didachist, it is a label after the fact, and its refer ence to the dvrooroAot has no significance for the docu ment itself.12 It would be utterly aberrant to derive an instruction for reading from the long title in H, and to see that title as an expression of the document's own self understanding.13 The two titles in H both come from a later time14 and are irrelevant for the interpretation of the document.15
15 By way of example, I list a number of positions on the question of the title. Schaff, Church Manual, 162
(cf. as early as p. 14): the long title rests on Acts 2:42 and Matt 28:19 and is probably original; Vokes,
Riddle, 208: the title combines Matt 28:19-20 with Acts 2:41-42: "It seems evident that the two pas sages at the end of Matthew and beginning of Acts explain the title and form of the Didache."
According to Bartlet (Church-Order, 53) the long title may possibly come from the editio altera et emendata, with possible influence from Matt 28:19. Johnson,
"Motive," 113-15: the author does not presume to write a gospel, but composes (among other things) a commentary, and supports his authority by pre tending composition by the Twelve. Kleist, Didache,
153 n. 1: the short title was written on the outside of the scroll. Rordorf, "Tradition apostolique," 106: the short title refers to Acts 2:42, and the long title to Matt 28:19; cf. also Rordorf and Tuilier, Doctrine, was the original and proper title of the document.
In the later tradition, this became 8i5axv nor ScoSeKOt onrocrToXui'. A copyist inserted the phrase bid rtor 8d>8eKa anoaToXun1 into the original title of the Duae viae. Thus the long title came into being. The title bibaxR rav dirooroXon' was transferred from the Didache to the Duae viae (as the Latin Doctrina shows). Nautin ("Composition," 211-12) previously offered a critique of these ideas; see further
Rordorf and Tuilier, Doctrine, 14-15.
12 This is true, mutatis mutandis, for the title of the tractate, if it is supposed to have read bibaxv nor darooToXav.
13 False, therefore, are the judgments, respectively, of Harnack, "Prolegomena," 24-27, 30; and Knopf,
Lehre, 3. Correct is that of Rordorf and Tuilier, Doctrine, 16-17: one should not be guided by either title in interpreting this work.
14 The Georgian has a still longer title: "Teaching of the Twelve Apostles, Written in the Year 90 or 100 after the Lord Christ: Teaching of the Lord, Con veyed to Humanity through the Twelve Apostles."
15-16.