of Two Ways," in Teaching, 18-23. The model was Wengst (Tradition und Theologie des Barnabasbrief
tate was noted very early51 and in recent times has been given additional support as a result of the discovery of the Qumran texts, especially the analogy to IQS 3.18—
4.26. It is characteristic of the special form in which the motif of the "ways" appears in the tractate that the near est parallels are to be found in Jewish and JewishChristian texts: IQS 3.18—4.26; T. Ash. 1:3-5; Ps.-Clem.
Horn. 7.3.3-5 (GCS 1.117-18); 7.1-2 (GCS 1.119). I will 1. The different versions of the tractate go back to an originally Jewish basic model, which began as a preChristian, Jewish tractate on the Two Ways48 and whose original wording can no longer be restored in detail on the basis of the related material.49 One of these versions also served the Didachist as a source. In any case, this hypothesis enables us to place the temporal origins of the Jewish tractate in the first century; the preliminary literary forms are still older and go back to pre-Christian times.50 say more in detail about this below. In explicating the details of the tractate as a whole, one encounters an Old
Testament-Jewish milieu at every turn; in the earlier pas2. The Jewish character of the original Two Ways trac
Wengst (Tradition und Theologie des Barnabasbrief [AKG 42; Berlin and New York: de Gruyter, 1971]
65-66) are also problematic.
50 We may also suppose an oral tradition of the Two Ways teaching before and alongside the literary tradition, but it cannot be reconstructed. What we have to trace here is the reconstruction of the liter ary tradition. In this connection it is appropriate to recall the attempts of Alfred Seeberg (cf. Der Katechismus der Urchristenheit [TB 26; 1903, repr.
Munich: Kaiser, 1966]; Die beiden Wege und das Aposteldekret [Leipzig: Deichert, 1906]; Die Didache des Judentums und der Urchristenheit [Leipzig:
Deichert, 1908). Seeberg made a number of attempts to reconstruct an orally transmitted, origi nally Jewish catechism schema (sometimes called
"proselyte catechism": Wege, 47 n. 3) with a more or less fixed body of teaching. He posited that this was, in turn, taken over by the Christian tradition
(with appropriate modifications). He also asso ciated the Two Ways teaching with this schema in various ways: it was to be understood as part of this catechetical schema. For an evaluation and critique of Seeberg's overall conception, see Ferdinand
Hahn's introduction to the reprinting of Der Katechismus (esp. xxi-xxii and xxiv-v). As far as our subject is concerned, we may say that it is very doubtful whether there ever was such a generally circulating and more or less fixed catechism schema in Judaism, containing various teachings including the Two Ways doctrine, and the attempt to bring this into a relationship with the Two Ways teaching
(or ultimately the whole Didache) remains corre spondingly dubious (quite apart from the question able character of many details in the whole description). In a broader sense, the proposals of Gottlieb Klein in Der alteste christliche Katechismus und die judische Propaganda-Literatur (Berlin: Reimer, of Two Ways," in Teaching, 18-23. The model was then developed primarily by Harnack (Apostellehre,
14, 25-34), although he originally had a different opinion. While other scholars were critical, the model received a particularly sustained attack in the early period from Franz Xaver Funk ("Zur Apostel lehre und apostolischen Kirchenordnung," ThQ_ 69
[1887] 281-89; idem, Abhandlungen, 2.137-41). For the deliberations on this question until 1900, see
Ehrhard, Altchristliche Litteratur, 53-58. Important opponents of the hypothesis of a basic document at a later period were Robinson (Didache, 70-83, 120,
146), Muilenburg (Literary Relations, 41-44, and passim), and Vokes (Riddle, 31-51, and frequently).
Later adherents of the basic document hypothesis included, e.g., Drews, "Untersuchungen," 54, and passim; J. Vernon Bartlet, "The Didache Recon sidered, "JTS 22 [1921] 243-46; MacLean, "Intro duction," viii ff.; Creed, "Didache," 377-79. In recent research there are Goodspeed, "Doctrina,"
230-47; Altaner, "Doctrina," 335-42; Richardson, "Teaching," 161-79, esp. 162; Audet, Didache,
122-63; Basil C. Butler, "The 'Two Ways' in the Didache,''/TS, n.s. 12 (1961) 27-38; Kraft, Didache,
4-16; Barnard, "Later History," 97-99; Adam, "Herkunft," 53-54; Giet, L'Enigme, 39-55, and pas sim; Suggs, "Two Ways Tradition," 62-63; Rordorf and Tuilier, Doctrine, 22-34, esp. 28; Wengst,
Didache, 20-23.
48 Nevertheless, the interpretative model also has some variants that accept the Two Ways tractate as very probably an independent source for the Didache, but regard the hypothesis of the originally
Jewish character of the source document as not com pelling, or reject it outright: Hennecke, "Grundschrift," 61; Goodspeed, "Doctrina," 236-37. These objections have been eliminated by IQS 3.18—4.26, because there we find a preliminary Jewish form constituting a literary analogy to the Two Ways tractate.
49 I consider Harnack's attempt to restore the source (Apostellehre, 57-65) a failure. The attempts by Klaus
1909) belong here. Klein constructed a separate, orally transmitted tradition, part of which was the "Ur-Didache."
51 Harris (Teaching, 91) and Muilenburg (Literary
Introduction sages especially, parallels to Ps. -Phocylides play a certain role.52 In the teknon sayings (which, however, I consider an outgrowth of the original tractate) we encounter
Jewish sapiential thought. At no point does the original Shenoute, Syntagma doctrinae, and Fides patrum.55 That is, as far as the source, the Two Ways tractate, is concerned, we are dealing in the first instance with a purely Jewish text. It probably belonged within the broader context of the community rules of Jewish religious communities
(as did IQS, lQSa in Qumran).56 The often expressed idea that the document was a "proselyte catechism"57 is tractate speak of Jesus53 (the sectio evangelica of the Didache is excluded). Nowhere do we encounter the christological or soteriological kerygma of passion, res urrection, and return.54 Finally, it is characteristic that the typically Christian text of Did. 1.3b—2.1 is absent from the Two Ways tractate in Barnabas, Doctrina apostolorum, the Canons of the Apostles, Epitome, Life of Relations, 107-8) objected that chaps. 7-16 also have a strongly Jewish coloration. But such ajudgment overlooks a characteristic difference: the Two
Ways tractate is Jewish in substance, and was later superficially Christianized; chaps. 7-16 are not a unit, but are shaped on the whole by JewishChristian tradition that was then redacted by the author.
52 The text of Sib. Or. 2.56-148, parallel to Ps.-Phocyl.
5-79, in the ms. group p, entered the Sibylline Oracles only at a late period (when there were addi tions and expansions during the Christian era): see van der Horst, Sentences, 84-85; Nikolaus Walter,
"Pseudo Phokylides," in Ernst Vogt and Nikolaus / Walter, Poetische Schriften (JSHRZ 4, 3; Giitersloh:
Didachist (and is absent in Barnabas, Doctrina apostolorum, and Epitome). We see that the Jewish source has been revised (hesitantly at first, and then more extensively).
55 In the Apostolic Constitutions it is traceable to the Didache. The three principal arguments for suppos ing an originally altogether Jewish source (the numerous parallels in the Jewish milieu, the absence of anything specifically Christian, and the absence of the sectio evangelica in the parallel texts) were listed already by Ehrhard, Altchristliche
Litteratur, 54.
56 Suggs ("Two Ways Tradition," 68) says that IQS 3.18—4.26 is "a kind of homiletic exhortation con cerned with group identity," and that T. Ash. 2 is a "school discourse in the interest of ethical instruc tion." The Sitz im Leben of the Two Ways teaching in Barnabas seems identical with that of the analo gous teaching in IQS, "an instrument of group identity. ... an initiatory setting." The same is said to be true also of the Doctrina apostolorum (p. 71).
The Two Ways tractate in the Didache is also said to be initiatory (p. 72). "In early Christianity the Two
Spirits/Ways form had its primary (but not exclu sive) Sitz in relation to initiation" (p. 72). Klaus
Berger wishes to assign Didache 1-6 (i.e., the Two Ways tractate) to the genre of npoTpenTLKoq Aoyoq.
Typical of this genre, according to Berger ("Hellenistische Gattungen," 1139), is also chap. 6.
"The frame corresponds to that of a protreptikos, and its filling out with a 'way of life' is parenesis
(hypothetikos)."
57 Cf. n. 47 above. In addition: Harnack, Apostellehre, 14, 29-30; Drews, "Untersuchungen," 54 (there is "a Jewish proselyte catechism at the basis" of the "UrMohn, 1983) 186-87. I will therefore cite the paral lels in parentheses. The discovery of the Didache
"constituted a turning-point in the research on PsPhoc." (van der Horst, Sentences, 14). It should be noted marginally that the parallels in the moral instructions of the Philadelphia stele (11. 12-50; cf.
SIG 3.985, pp. 116-18; and the table in Weinreich, Privatheiligtum in Philadelphia, 59) should not con fuse the assessment of the Jewish character of the Two Ways tractate. In addition, it appears to me that Jewish influence on the section of the stele in question cannot be excluded.
53 Kyrios {Did. 4.1) probably refers to Jesus, but that is not the oldest text of the tractate. See below at 4.1c, and the following footnote.
54 The double love commandment (Doctr. 1.2; Did. 1.2; Can. 4.2-3; Epit. 1), typically lacking in Barnabas, probably goes back to the later Christianizing of the Jewish source. On this, see below. Does the passage about the Spirit in Doctr. 4.10 and Did. 4.10 (cf.
Barn. 19.7c) presuppose Christian editing? I think it does. Doctr. 4.1c; Did. 4.1c; Can. 12.1 is uncertain.
(The corresponding parallel in Epit. 9 is Chris tianized, but it is a subsequent clarification.)
Uncertain also is Did. 4.1b par. In both cases, I sup pose that there has been some Christianizing. !E'v eKKArjoia in Did. 4.14 is probably an addition by the Didache," and Paul knew it in a similarly "Chris tianized form"). The proposals in Adam, "Herkunft," again go especially far: "The Two Ways teaching could have originated in the Judaism of Adiabene as a handbook for the instruction of cate chumens" (pp. 53-54). At p. 53 n. 97, Adam sug gests that the Ananias known from Josephus Ant.
Can. 12.1 (cf. Epit. 9) Barn. 19.9, 10 Did. 4.1 (cf. Doctr. 4.1) Can. 14 Barn. 21.2c-4, 6a Table 3 Table 3a Kal itapaiTLov oot Kac ovk hxere (2c) exere tov XaXovvra ooi tov XaXovvToq oot \ tov Xoyov TOV $€OU tov Xoyov tov Beov& (1) oog an Kaupog eon ytvopevov rfjq farjq eiq ovq epya^eode pev eavTtcv, Kal bovra oot ped' eavTuv, eiq ovq epya^eade to Tpv ev Kvpiu KaXov otppayiba pr) eKX'nrpTe pr) eXXeltrr\Te. dipfiaXpov oov, d<p$aApov oov (2) Eyyvq yap (3) eyyvq dyaTnjoetq1 ayairrjoeLg ev ppbevL, <hq Koppv cog KOpr]V TOV e^ovoLav eav exVTe.
7ravra tov XaXovvra r) ppepa KvpLov rj ppepa, oot tov Xoyov
ev f] ovvaTcoXelTai ev T) avvartoXelTat Kvpiov.
Ttavra ovv r<5 irovripwttavTa Tbi Ttovr)pqpvri<rdf)ori be2 pvpoxdpa-Q ppepav pvr\o'df)0"Q r/|€i yap 6 Kvptoq. eyyvq 6 Kvpcoq avrov Kpioeoiq Kal 6 pi<rdbq avrov.
Kal b pioftoq avrov. vvktci kal ppepav,3 WKToq nai ppepaq WKToq Kai rjpepaq.
(3) (4) ctl Kal £tl epoirdi vpaqTtpfioetq* avrov Tipr/oeiq be ocvtov 'Eavrav yivea'&e vopofterai, eavrav ylveode vopoderai uq TOV5 KVpLOV. coq KVpLOV. dya&ol, eavrav ylveode ovpfiovXoi eavruiv pev ere ovpfiovXoi
TTLOTOL . . .
(6a) ylveode be fteobibaKTOi . ayadol, deobibaKTOL4 Epitome adds be.
1 Epitome adds avrov.
5 Epitome omits tov.
2 Epitome omits be.
Cf. Barn. 19.11b; Did. 4.13b; Doctr. 4.13b.
[ywAdijcig a rrapeXafieq ppre itpoa§elq ppre ixpaipav 6 Doctrina: domini dei.
3 Epitome: WKToq k. ppepaq.
Cf. Epit. 11 (end)] Christian revisers, or (and this may be more probable) they had no place in the group's moral catechism.62 dubious.58 In its original form the tractate may have had the function of a community rule for Jewish enthusiasts who gathered in the house of study59 and who encour aged each other to mutual social support.60 Never theless, some things point to a mission among the Gentiles.61 The complete absence of any ritual prescrip tions is very striking. They were either erased by the 3. The original form of the Jewish tractate of which we have been speaking has disappeared, but it lived on (in a variety of modifications) in ancient Christian literature. The first Christians (probably Jewish Chris tians at the outset) adopted this text and arrogated it to testo con una nota intorno all' eucaristia [1893] 55-56
[not available to me; I quote from Ehrhard]; cf.
Ehrhard, Altchristliche Litteratur, 51 and n. 1, as well as 55). This idea was frequently accepted later
(sometimes with variations). I consider it, too, to be false. For this question, see the remarks below at 20.34-96 was its author. For Adam, the basic docu ment of chaps. 1-6 of the Didache was, in any case, a Jewish proselyte catechism from Adiabene (p. 54, and frequently).
58 The cautions in Rordorf and Tuilier (Doctrine, 31 n.
5) are correct. Cf. earlier Wilhelm Michaelis, "o5og," TDNT 5 (1967) 92, 99, and frequently;
Wibbing, Tugendund Lasterkataloge, 5-7.
59 Wengst (Barnabasbrief 67) supposed that "the Two Ways teaching, like the traditions used in Barnabas in chaps. 2-16, may have had its Sitz im Leben in a school setting."
60 Cf. at 4.8 below.
61 Cf. 2.2.
62 P. Savi introduced the idea that the substance of the final part of the Didache (chap. 16) also came from the source, i.e., from the Jewish Two Ways text {La dottrina degli apostoli: Ricerche critiche sulV origine del
16.1. I also believe that, in the course of its Chris tianizing, the tractate acquired an (originally brief) eschatological epilogue. But this epilogue had its own history, and in any case Didache 16 does not come from it, but has replaced it.
63 Goodspeed ("Doctrina," 234) says, with reference to the tractate, "The evidence for its literary influence is much greater than is that for the influence of Didache."
64 Barnabas often has the corresponding parallels in a different sequence. In Barnabas, the teknon sayings found in the Canons and known also to the Epitome
Introduction themselves, and from then on it has had a historical impact within Christian literature.63 Whether that his tory can still be clarified in all its details is a difficult question. similar recension was read by Barnabas (without <pvXa^etq ktA., which was not at the end of this author's source, or in that used by the Canons);70 but Barnabas made changes and expansions. Similarly, the One of the fundamental problems that emerges in this connection is as follows: the Canons of the Apostles/Epitome to a large extent parallel Didache/
Doctrina apostolorum against Barnabas.64 Nonetheless, agreements between the Canons (and once Epitome as well) and Barnabas against Didache/Doctrina men tioned above can scarcely be traced to a use of Bar nabas as an additional source by the Canons, or by it and Epitome, or by their common source. In view of Can. 12.1 (Epit. 9) it is impossible to understand why
Barnabas, of all documents, should have been used at this point.71 The problem, rather, is evidently that the Canons (and the Epitome) move largely in tandem with Didache/Doctrina, in both matter and structure, but in the rare cases mentioned above the Canons (and, in one instance, Epitome as well) corresponds to Barnabas against Didache and Doctrina. In my opinion, that suggests the idea that the text of the tractate for all four (or five) witnesses (Barnabas, Didache,
Doctrina, the Canons, and the Epitome, or the common the proemium to the Canons65 (xccipere uioi ktA.) recalls Barn. 1.1 (omitted in Didache/Doctrina)/6 Can.
4.2, nai 5o^aaecq top XvTpwoapevov oe en -davarov ("You will glorify the one who has redeemed you from death"),67 recalls Barn. 19.2 (against Didache/
Doctrina, where the expression is missing);68 Can. 12.1
(Epit. 9) contains a text that deviates from Didache/ Doctrina, wherein the characteristic expression cog
Kopr\p d(pda\p,ov oov ("as the pupil of your eye") again recalls Barnabas (19.9): cf. table 3. Finally, Can.
14 (omitted in Epitome) matches Barn. 21.2c-4, 6a against Didache and Doctrina (cf. table 3a). source of the last two) should ultimately be traced to a common primitive form (B), which was copied
It would be difficult to persuade oneself that Can.
14 stems directly from Barnabas. The text in the Canons has a more immediate application;69 it seems more likely that the epilogues in Canon 14 and Barnabas 21 are traceable to a common source. The text of the Canons is probably based on the end of one of the recensions of the Two Ways tractate, and a 68 That Canons, after rov itOLr\actVT<x oe, also adds e| oXrjq oov Kapoiaq (as does Epitome), something that are missing (although it is true that the teknon apos trophe is also lacking in the Epitome). That the "way of death" is absent from the Canons and Epitome
(and the Canons also lacks the conclusion of the "way of life") does not disprove the hypothesis of a close relationship between Didache/Doctrina apos tolorum and the Canons/Epitome. The reasons for these omissions have been discussed above. In addi tion, the fact that in Epit. 10-11 the material found in Didache 4 is in a different order does not destroy the above hypothesis.
65 It was especially Hennecke ("Grundschrift," 62-69) who pointed out the combination Barnabas/Canons against Didache/Doctrina, although he arrives at a different stemma by this process than I do.
66 The corresponding material is also lacking in the Epitome. In addition, cf. the suggestions by Goodspeed, "DocUina," 232-33. According to Goodspeed, the opening greeting in the Canons goes back to Barnabas. That is not true, however, of the remainder of the document.
67 This expression is missing from the Epitome. is paralleled neither in Didache and Doctrina apostolo rum nor in Barnabas, says nothing about the depen dency relationships. It is a biblical reminiscence
(very similar to Synt. doctr. 1.4 or Fides patr. 3: cf. PC,
28.1639C).
69 Hennecke, "Grundschrift," 65.
70 Didache and Doctrina apostolorum have the text in 4.13, which I believe to be its original location. The redactor of the Epitome followed his source up to the text corresponding to Did. 4.13, and accordingly had (pvXa^rj be ktX. as a sensible conclusion. The Canons at first did not have this passage, because the quotation from the Two Ways tractate broke off for the present, at the text corresponding to Did.
4.8. Therefore, Canons transferred an appropriate text to the conclusion of the epilogue.
71 Wengst (Didache, 10 n. 23) supposes that the Canons knew Barnabas, while Epitome was independent of Barnabas. Hence the parallels in Can. 12.1 /Epit. 9
(cf. Barn. 19.9) would come not from Barnabas but
Table 4 directly by Barnabas,72 whereas the rest are dependent on an archetype (C) derived from (B), and, more pre cisely, Didache and Doctrina are taken from a sub archetype (C1), and the Canons and Epitome from another sub-archetype (C2). Using this model (for which, see the hypothetical stemma in table 4) as a working hypothesis, one can incorporate the observa tions already made above.
If these observations are accurate, we should pic ture the genealogy of the Two Ways tractate more or less as follows:
(a) In the beginning was a Jewish "primitive form"
(A) that has been lost.73 Even this "primitive form" was not a unified composition, however, but rather a compilation.74
(b) Dependent on (A) was the first Christian recen sion (B). The Christianizing of the material in this first recension was probably very superficial; still, the redactor probably added an eschatological epilogue with Christian character (to replace the original,
Jewish eschatological epilogue). Materially, this Epitome epilogue still forms the basis of Canon 14 and Barna bas 21.
(c) Dependent on this first recension is Barnabas 18-21. The author of Barnabas sometimes copied his source very freely and (with some exceptions) repre sented the order of the source less perfectly than did some later derivations.75 In addition, Barnabas has incorporated the Two Ways tractate at the end of his text. The introductory xatpere viol ktA. was proba bly also something he took from his source, and he
Syntagma doctrinae Anastasiana Syria sections would have been redacted to form a "way of life," and to them was added (5) a vice list that was formed into the "way of death." The whole was concluded (6) with an eschatological epilogue. The teknon sayings were added in Christian times, even though materially (apart from a few additions) they would have been Jewish in origin. In a clever analy sis, Audet (Didache; 308-20) has attempted to sketch a different primitive form.
75 The alternative supposition that it was only the redactor of C who put the text in the order we find in Didache, etc., is not as good. The sections of the Two Ways tractate as they are sequenced in the Didache belong to different genres; that kind of order would scarcely be created at a subsequent time. On this, cf. the suggestions of Audet, Didache,
320. Differing is, e.g., Kraft (Didache, 12), who is inclined to think "that the Doctrina-Didache form from a recension of the Two Ways tractate. But is it not more plausible to explain the parallels between Canons/Epitome/Barnabas against Didache and Canons/Barnabas against Didache in the same way?
72 While, at the same time (with some exceptions), he altered and rearranged the sequence.
73 The reference is to the "primitive form" of the Two Ways tractate whose existence we have deduced, and to which can be traced, either directly or indi rectly, all the filiated documents to be named. One can also go behind "A" to a "pattern" that would include IQS 3.18—4.26 and T. Ash. 1-5. Cf.
Barnard, "Later History," 107; Suggs, "Two Ways Tradition," 62-63.
74 It would have consisted of (1) an introduction with fundamental rule, (2) a list of prohibitions, (3) the anawim sayings, and (4) sayings regarding social relationships, including a household code; all these
Introduction has displaced this sentence to the beginning of his entire work.
(f) Finally, the material of the Two Ways tractate (meanwhile long since part of Christian tradition) migrated, not accidentally, into particular segments of monastic literature (Life of Shenoute, Syntagma doctrinae) and the Fides patrum. I do not presume to give information on the more precise genealogical context of these texts (apart from the fact that I judge them to be late filiations from recension C).
Of course, the presentation attempted here remains purely hypothetical.79 It serves merely as a working hypothesis. What seems to me fundamentally important
(d) More important for the subsequent history was a second Christian recension of the tractate (C). It intervened much more strongly in the tradition. This redactor (1) transformed the traditional basic com mandment into a double love commandment;76 (2) added the teknon sayings; (3) made an additional, minor expansion of the text. Dependent on (C) is a sub-archetype (C1), characterized by the alteration of the eschatological epilogue. Dependent in turn on (C1) are the Doctrina apostolorum77 and the Didache.
The author of the Doctrina appears to have repro duced his source with overall fidelity. Doctr. 6.6 was first added in the Latin text tradition. The redactor of the Didache dealt quite differently with the source; this will be discussed more fully below.78 is only that we should posit an independent, originally
Jewish source document that existed in a variety of recensions. In one of those recensions, it was available as a source for the Didachist.
The fate suffered by the Two Ways tractate in its Latin translation in the West is striking. A version of the tractate in Latin survived in the West into the eleventh century (under the title [De] doctrina apostolo rum). We cannot deal here with the influence of this version on Latin church literature.80 It is probable that the tractate was even used by Boniface in connec tion with prebaptismal catechesis. A new investiga tion of the history of our tractate in the Latin church might be desirable.81
(e) A later recension (C2) proceeded to put the material of the Two Ways tractate in the mouths of the (eleven) apostles: that is, it stylized the whole doc ument as apostolic teaching. In the process, the text of the "way of death" was eliminated. Dependent on this recension (but independent of one another) are the Canons of the Apostles and the Epitome-, Epit. 11 contains the conclusion of the "way of life," which is missing from the Canons, while — in contrary fashion
— the Canons has substantially preserved the old eschatological material in chap. 14, which is missing from the Epitome. (We need not discuss the further differences between the Canons and the Epitome here.) 20-23; cf. the stemma on p. 22.
80 Schlecht, Doctrina, 43, 75-91 (but under the false assumption that the two Latin mss. f and g are Latin versions of the Didache); further Rordorf and Tuilier, Doctrine, 206.
81 The more precise relationship between our tractate and the Rule of the Master and the Rule of Benedict would need to be investigated again in this connec tion. On this, see above, p. 14 n. 92. of especially the Way of Life has been extensively reworked with respect to sequence." Cf. also
Wengst, Didache, 21 n. 77, and further at 92 n. 2; idem, Barnabasbrief 59ff.
76 This probably already reveals Christian motivation; see below, on Did. 1.2.
77 More precisely the Greek original, the Latin transla tion of which we have in Doctrina apostolorum.
78 Literarily dependent on the Didache is the Apostolic Constitutions, on which depend Fragmenta Anastasiana and Isaac of Syria, directly or indirectly.
79 Cf. the agreements and disagreements with this presentation in Giet, L'Enigme, passim (see merely the stemma, p. 152); and Wengst, Didache, 7-14,