Part (1) expresses the blessing addressed to God; (2) the hand of the heavenly Father.10 An important change
9.1-5 custom, a mixture of wine and water), lifts it a little above the surface of the table, and speaks the following blessing or a similar text:
OTOvpev ool ("We give you thanks"), but this is only a formal difference without "theological" significance. In place of the address "Lord our God, Ruler of the uni verse," we read "our Father" (iraTep rjpuiv)9 This predi cation certainly expresses the self-concept of the group, who understand themselves as the table company of chil dren who receive the goods of time and eternity from the hand of the heavenly Father.10 An important change from the background prayers is the statement about the object for which thanks are given. The phrase 0:7(0: dp.'KeXoq ("holy vine") takes up the Jewish expression
1. EvxoipLOTOV(X6V OOL, TTOTep r)p,QV, 2. virep Trjq dyiotq dpireXov Aavlb tov -KotLboq oou, rjq eyvwpLOctq pplv bid Irjoov tov ttotLboq oov
3. ool rj bo%a elq rovq oduivaq.
Part (1) expresses the blessing addressed to God; (2) formulates, in two succinct clauses, the object over which the praise and thanksgiving are spoken; (3) concludes the brief formula with a doxology. This doxology may be the response of those gathered at the table.6 The whole formulation is connected, here and in 9.3, with Jewish meal prayers. Indeed, 9.2 is based on the most common
Jewish blessing over cup or wine: "Blessed art thou, Lord [Dun "ID ("fruit of the vine") and thus clearly shows its dependence on traditional Jewish wine blessings. The "fruit of the vine" has been replaced, however, by the expression "holy vine of David."11 The obscure phrase our God, Ruler of the universe, who creates the fruit of the vine."7 The differences between the model and this prayer should be noted; they vary in importance.8 In place of nntf "[TO ("Blessed be you") we find evxotpLlier Lietzmann, Mass, 189-90, with reference to the well-known analysis of Wilhelm Bousset on the source of the prayers in Const. 7.33-38 ("Eine jiidische Gebetssammlung im siebenten Buch der Apostolischen Konstitutionen," Nachrichtungen der
Gottinger Wissenschaftlicher Gesellschaft, PhilosophischHistorischer Klasse [1915] 435-89). Voobus
(.Liturgical Traditions, 160-62) expressed some reser vations about this. There is cautious agreement with Dibelius in Wengst, Didache, 53 n. 177, and passim.
In general, for the predication of "father" to God in contemporary Judaism, see the overviews in Str-B
Str-B 4.616. There is occasionally a reminiscence of the kiddush (on this, with corrections of false ideas:
Jeremias, Eucharistic Words, 26-29). Kiddush is "a
blessing pronounced at the beginning of each Sabbath or feast day," by the head of the household
(Jeremias, Eucharistic Words, 26). Kiddush is spoken over a cup of wine; a blessing of the wine then fol lows. Among those who proposed that the sequence in Did. 9.2 especially reflects that of the kiddush
(more precisely of the blessing of wine connected with the kiddush) were Klein, "Gebete," 134-35; 9
Middleton, "Eucharistic Prayers," 261-62; Voobus, Liturgical Traditions, 163. The attempt to create a / 1.392-96, 410; Gottlob Schrenk, "itarfip ktX.,"
TDNT 5 (1967) 978-82; H.-W. Kuhn, "d|3/3d," EDNT special connection to the kiddush is, in my opinion, unnecessary. In that case, one would have to posit that the meal celebration in the Didache liturgy orig inally took place on every Sabbath eve, and in fact
Klein did propose that ("Gebete," 144).
The Jewish blessing of wine before the meal is spo ken by each individual, and not (as is the blessing of bread) by one (namely, the head of the household) in the name of all.
This was suggested by Dibelius, "Mahl-Gebete," 119.
Cf. m. Ber. 6.1 (ed. Holtzmann, 72-73).
Dibelius attempted to show that the basis is not the older Jewish blessing of wine, but a spiritualizing form created in the Hellenistic synagogue ("MahlGebete," 119; now in A. F. Verheule, ed.,
Religionsgeschichtliche Studien: Aufsatze zur religionsgeschichte des hellenistischen Zeitalters [NovTSup 50;
1 (1990) 1-2; Otto Michel, "itaTpp," EDNT 3 (1993) 53-57; Jeremias, Abba, 19-22 (ET: Prayers, 15-18).
For the predication TDK, "our father" (combined with 13d7q, "our king"), in the Jewish literature of the early common era (1st to late 2d century), cf. esp. Jeremias, Abba, 28-33 (ET: Prayers, 24-29).
There can scarcely be a direct influence of the "Abba, Father" of the post-Easter experience of the Spirit (Rom 8:15; Gal 4:6).
I am not convinced that the term dyia dpneXoq is reminiscent of the Jewish kiddush (thus Klein,
"Gebete," 134-35; Voobus, Liturgical Traditions, 163).
Leiden: Brill, 1979] 231-85). (See also his analo gous suggestions for the following prayers.) See ear
presence of the cup and the wine, which are earthly gifts, the thoughts of the community are directed far ther, to the heavenly good, which God has given to God's own and which can be understood through the metaphor of the "holy vine." The concepts are thus transparent: the subject of revelation is God, and its object is the "holy vine," that is, salvation, and more specifically eschatological salvation, which was once promised but has now been given. "We thank you for the holy vine that you promised to David, your servant, and that you have now revealed through Jesus, your ayta djuireXoq Aavid12 is symbolic language.13 It proba bly alludes to the whole matter of salvation14 (since virep rrjg ayiaq dpnreXov parallels vjcep Trjg farjq nai yvccaeccg in the second benediction in 9.3).15 Thus 0:710: dpnreXoq is probably a metaphor for salvation itself.16
The more precise definition (apireXov) Aavtd tov iraidoq oov, which is puzzling at first, acquires its mean ing in what follows: rjg eyvupioaq17 pp.lv did Ipoov tov iraidog oov. That is, God has shown, has revealed,18 the holy vine (eschatological salvation) through Jesus,19
God's servant. I understand this to mean that in the 12 To Harnack (Lehre, 29), the whole expression was puzzling. Peterson ("Probleme," 170) has declared the whole benediction incomprehensible. The Bryennios text is said to be secondary; in fact, "it appears that [in it] an attempt has been made to transform the eucharistic prayer into a table prayer." See above, Introduction, §3a.
13 In my opinion the formula in Did. 9.2 does not create a reference to the blood of Christ (thus
Klein, "Gebete," 135; Greiff, Pascharituale, 70-72).
Differently Clement of Alexandria (has dives salv.
29.4 (GCS2 3.179): ourog (6) tov olvov, to alpa rrjg apireAov Trjg Aafiid, eexeaq ppuv ^7r'L rdg rerpo)pevotg fvxdt-g ("This is the one who pours the wine,
Sir 24:17 we are already approaching the use of the image that lies behind John 15:1-6. But Did. 9.2 has nothing to do with these interpretations. This is not a derivation of the common OT metaphor in which the vine stands for Israel or Judah, so that we should understand the holy vine to be the church as the new, eschatological people of God (this is the interpretation of Gibbins, "Liturgical Section,"
375-76, 383; J. Vernon Bartlet, Church-Life and Church-Order during the First Four Centuries with Special Reference to the Early Eastern Church-Orders
([ed. Cecil J. Cadoux; Oxford: Blackwell, 1943] 20; Voobus, Liturgical Traditions, 126; Rordorf and Tuilier, Doctrine, 45; Sandvik [Kommen des Herrn,
61-62] sees it as an image of the church as temple).
Nor is it the interpretation appearing in Sir 24:17; John 15:1-8, which sees the vine as representing the the blood of the vine of David, onto our wounded souls"). One may ask whether Clement is quoting the Didache (while changing its meaning)—for the contrary opinion, see above, Introduction, §2a—or whether this symbolism is familiar to him, also, apart from the Didache. The other instances favor the latter position. Origen Horn. 6.2 in Iudic.
(Lommatzsch, 11.258): antequam verae vitis, quae adscendit de radice David, sanguine inebriemur
("Before we drink of the blood of the true vine, which rose from the root of David"). P. Der Balizeh,
IIv., lines 6ff. (Roberts and Capelle, Euchologium, 26; cf. van Haelst, "Nouvelle reconstitution," 449): 6 olvog o[i>TO<;~\ 6 e^eXdwv Ik rrj[g aylag a]pTre\ov zifcvuetd] nai to vdicp ap[vov apu]pov ("This is the wine which has come from the holy vine of David and the water from the blameless lamb").
Here the liturgical expression "David's vine" is com bined by means of Christian interpretation (John one who brings salvation. It is obviously not a refer ence to the Messiah either, in spite of 2 Bar. 36-40
(the rule of the Messiah); Cave of the Treasure 51.4
(Call Bezold, Die Schatzhohle = "MPdrath gazze" = [The Cave of treasures/La caverne des tresors]: eine
Sammlung biblischer Geschichten aus dent sechsten Jahrhundert jemals Ephraem Syrus zugeschrieben
[2 vols. in 8; Leipzig, 1883-88; repr. Amsterdam:
Philo, 1981]). This interpretation of Did. 9.2 fails by the very fact that a distinction is drawn in 9.2 between ctp-rreAog and tralg (Voobus, Liturgical
Traditions, 125). The work of Rainer Borig, Der wahre Weinstock (Munich: Kosel, 1967), is not help ful for our specific question.
16 Betz ("Eucharist," 264-67) sees a whole group of intentions: "the expected Davidic messiah and wis dom" (p. 266), but ultimately the table community as well (p. 267).
15:1-8) with John 19:34.
14 Dibelius, "Mahl-Gebete," 120. The considerations he raises in this connection can be allowed to rest.
Cf. also Wengst, Didache, 49.
15 Thus the meaning of the symbol is different from that in the OT, where Israel is frequently represent ed by the metaphor of the vine: Hos 10:1; Jer 2:21;
Ps 80:8-13; cf. 2 Esd 5:23 (ofjudah: Ezek 15:1-8; 17:6; 19:10-14). When Wisdom appears as a vine in 17 Hg eyvupcoag is an attraction; the same is found also in 9.3 and 10.2. Cf. BAGD, s.v. yvuplfa. For the attraction of the relative, see BDF, §294; BDR,
§294.
18 This is, of course, the meaning of yvwpl^eiv here.
Cf. also 9.3; 10.2. This word is found in the Didache only in the fixed expressions of the meal prayers: rjg
9.1-5 servant!"20 It is uncertain whether, in this connection, one is meant to think of concrete promises to David in the Old Testament; the expression may have been for mulated without any concrete scriptural reference in mind.21 The designation of Jesus as irate; oov (sc. fieov) sions that border on foreignness. More probably, howev er, the basis lies in tradition history: the whole predidachistic liturgy stemmed from the early Jewish
Christianity of Palestine. The idea of Jesus as iral; (fteov) would then be a sign of the antiquity of the tradi tion found here.26 Accordingly, iral; (deov) here and in 9.3; 10.2-3 does not (yet) mean "child of God" but rather "servant of God." Ilal; oov = J-QJL In the context of our table prayers, then, it is difficult to posit any allu sion to the servant of God in Deutero-Isaiah. One cannot say with greater certainty what special associations were combined in this exalted title for Jesus within our tradi tion. Only the office of the iral; deov as eschatological mediator of salvation and as revealer (yvwplfetv), his relationship to David, and the schema of promise and is certainly not a Christian imitation of a Jewish model, which would only have spoken of David as the "servant of God."22 Instead, dta 'Ipoov tov iratdo; oov, which reappears verbatim in 9.3; 10.2, 3, is an old liturgical expression.23 We know that iral; is seldom predicated of Jesus. It appears primarily (although not exclusively) in liturgical material.24 Reference to Jesus as "servant of God" has an archaic flavor.25 We could imagine that this was deliberate in view of the occasional inclination of liturgical language to speak in solemn, ancient expres eyvwptaaq pplv 8ra 'Ipoov . . . (9.2, 3; 10.2). It is characteristic of the language of these table prayers.
19 The simple 'Irjoov is striking (Harnack, Lehre, 29-30). Thus also in 9.3; 10.2; contrast dta 'Ipoov
Xptoiov in 9.4 (where, however, iralq is lacking).
20 Cf. Dibelius, "Mahl-Gebete," 120, who recalls the expression in Isa 55:3 (wrongly cited by Dibelius as 55:33): Ta oata Aauld ia inora.
21 Cf. Dibelius, "Mahl-Gebete," 120. Precisely if the (otherwise obscure) expression "vine of David"
(apireAoq AavLd) acquires its meaning only by means of the continuation (19c; eyvuptoaq ktX.), it is not very likely that this expression can be explained as something pre-Christian, namely, Hellenistic
Jewish, that was then added to the relative clause by in which the title vloq from Ps 2:7 and the title iralq from Isa 42:1 were fused."
24 Cf. Acts 3:13, 26; 4:27 (aytoq iralq), 30 (same); then in Barn. 6.1; 9.2; 1 Clem. 59.2, 3, 4; Mart. Polyc. 14.1,
3; 20.2. Somewhat distant: Matt 12:18 (following Isa 42:1). For later material, see Harnack, "Bezeichnung," 224-38.
25 "The solemn, archaic-sounding expression iralq deov" (Knopf, Lehre, 26).
26 Harnack was the first to observe that the pais predi cation is ancient in its origins ("Bezeichnung,"
212-14); this has often been repeated. The geneal ogy of the predication is disputed: the question is whether there was a distinctive pais Christology, and in particular whether there is reference to the "cebed YHWH" songs. From the (already boundless) literature, let me mention only some examples:
Jeremias, Abba, 191-216 (ET: TDNT 5 [1967] Christian editors, as Dibelius would have it. Instead, the text is more readily understood within the Christian schema of "promise fulfilled," as is the paralleling of David with Jesus.
22 The expression Aavl8 6 iralq aov (cf. Luke 1:69; Acts 4:25) corresponds to "]1DU TH; cf. the Shemoneh
Esreh (Babylonian recension) 15 v. 1. (Willy Staerk, Altjudische liturgische Gebete ausgewahlt und mit
Einleitungen [KIT 58; Bonn: Marcus and Weger, 700-716); Oscar Cullmann, The Christology of the New Testament (rev. ed.; trans. Shirley C. Guthrie and Charles A. M. Hall; Philadelphia: Westminster,
1963) 51-82; Morna D. Hooker, Jesus and the Servant (London: SPCK, 1959); M. Rese, "Uberprufung einiger Thesen von Joachim Jeremias zum Thema des Gottesknechtes im Judenthum," ZThK 60 (1963)
21-41; Wilckens, Missionsreden, 163-68; Ferdinand Hahn, Christologische Hoheitstitel (5th ed.; Gottingen:
Vandenhoeck & Ruprecht, 1995) 54-66, 386-87 (ET of 1st ed.: Titles of Jesus in Christology: Their History in Early Christianity [trans. Harold Knight and George Ogg; New York: World, 1969] 54-67); JanAdolf Buhner, "ayadoq," EDNT 3 (1993) 5-6. For further literature, see ThWNT 10.2 (1974) s.v. iralq ktX.; Joachim Jeremias, "iralq ktX.," TDNT 5 (1967)
636-37.
1910] 18); Habhinenu prayer, recensions a and b (Staerk, Gebete, 20); Musaph prayer for Rosh
Hashanah 1 (Staerk, Gebete, 23). See Joachim Jeremias, "ira.it; deov," TDNT 5 (1967) 681 n. 184.
23 There is a similar association of David and Jesus as iralq (deov) in Acts 4:25-30; in v. 25 David (cf. Luke
1:69), and in vv. 27 and 30 Jesus (although Jesus is aytoq iralq). Still, we should note that in Acts 4:27,
30 (and earlier in 3:13, 26) iralq = vloq. Ulrich Wilckens (Die Missionsreden der Apostelgeschichte [3d ed.; WMANT 5; Neukirchen-Vluyn: NeukirchenerVerlag, 1974] 164) concludes "with some assurance to a pre-Lukan tradition of proof of prophecy, . . .
fulfillment (in 10.3, then, probably in relation to the Lord's Supper) are clearly evident.27 probably correct. The secondary xXdopa instead of aprog could also be an indication of the Egyptian char acter of the H tradition.
Then follows the doxological conclusion ool 17 do^a elq rovg aiccvag (v. 2c), which we can best think of as a response by those present at the meal.28
The wording of the prescribed prayer is parallel in construction to the blessing of wine (see above):
1. Evxotp lotov p.ev ool, Tcdrep ppuiv, ■ 3 Verse 3 gives the second benediction, the blessing of bread, introduced by the rubric 7repl de tov K.Xdop,a.Tog.
This suggests that we should understand nAdopa to mean the bread broken at the meal celebration.29 In that case, the plural irepl de twv KAaoptamu would seem more appropriate. The problem appears again in 9.4, where H gives us tovto KXdopa, and the parallels from later liturgies have aprog in the analogous location.30
Peterson has pointed out that KXdopta is a technical term in the eucharistic language of Egypt; it refers to the particle of the host.31 The expression would then have entered the text of the Didache at a secondary stage.
According to Voobus also, nAdopot is secondary.32 The original text may have had aprov.33 This emendation is
2. vrcep rrjg &VS xal 7vwoewg, rjg eyvdptoag r)pli> did 'Ipoov tov ttqadog oow 3. ool r) do%a eig tovg alutvag.34
As Did. 9.2 was based on the traditional Jewish bless ing over the cup, so here the Jewish blessing of bread lies behind the formulation. "Blessed art thou, Lord our
God, Ruler of the universe, who brings forth bread from the earth."35 For part (1), see above. Part (2) names the schema in accordance with the divine gift, namely, "life"
((W17) and "knowledge" {yvuoig).™ While this is stated by means of a metaphor in the blessing of the wine ("the holy vine of David"), here at the blessing of the bread it is predicated directly. Zcor) nai yvi'ooig sounds almost like
27 For the association of the nalg title with revelation, see nn. 34 and 39 below. See Harnack, "Bezeichnung," 234-35. The predication iralq (tieov) is thus characteristic of the table prayers in the Didache.
For the redactor this predication probably had, on the whole, the value of a venerable liturgical relic. It is characteristic of him that when he is formulating his own text, namely, in 9.5, he immediately turns to a different predication: Kvpioq.
28 Differently Audet (Didache, 401-2), who conjectures apqv, and understands that to be the response of the gathered community.
29 Bryennios, Atbaxv, 35 n. 4: "KXaopa 6 ev rr) evyapiOTLa xXdptpoq aprog." See BAGD, s.v.; Schaff,
Charles Bigg ("Notes on the Didache,"JTS 6 [1905] 413) had already pointed to the connection with the Egyptian liturgy (and claimed the use of this term as evidence for the late origin of the Didache, p. 414).
32 Liturgical Traditions, 89, 146-48; idem, "Background," 83.
33 On this whole subject, cf. Niederwimmer, "Textprobleme," 124-25; now also Wengst, Didache,
78, 97-98. The thesis that xXdopa comes from the eucharistic language of Egypt would still not pro vide evidence of the originally eucharistic character of the prayer (as Peterson wishes to say:
"Probleme," 168-69). Cf. also Wengst, Didache, 98.
34 Cf. (Ps.-)Athanasius De virg. 13 (von der Goltz, 47.3-4):
Church Manual, 192.
30 Serapion Euchologion 13.13 (Funk, Didascalia, 2.174); P. Der Balizeh IIv., lines 3-4 (Roberts and Capelle, Euchologium 26; and van Haelst, "Nouvelle reconstitution," 449); (Ps.-)Athanasius De virg. 13
(von der Goltz, TU, n.s. 14.1a [1906] 47.4-5). Const.
7.15.3 reflects the Didache's source as follows: dairep 1. evxap lotov pen ool irdrep 17/rcor 2. vTtep Trjq ay lag avaordoeug aow did 7dp Iqoov tov iracdog oov eyvapcoaq qplv avTqv . . .
35 Cf. m. Ber. 6.1 (ed. Holtzmann, 72-73).
36 According to Peterson ("Probleme," 178) the original Egyptian liturgy mentioned avaoTaoiq as an element of salvation (as does De virg. 13), and this would have been replaced by fa377 in the Bryennios text; in addition, yvuoLq would have been inserted.
See the doubts about this proposal expressed in Voobus, Liturgical Traditions, 86-87.
37 Remotely comparable is Diogn. 12.2-3, where Gen 2:9 is interpreted.
38 Z(ssr\ in Did. 1.1, 2, and 4.14 comes from different traditions, but in the mind of the Didachist there is no difference between far) in 9.3 and 10.3. f)P tovto bieoKopiaopevov xal ovvaxdev eyevero elg dproq ("As this was scattered and brought together as one bread").
31 Erik Peterson, "Meplq: Hostienpartikel und Opferanteil," in idem, Fruhkirche, 99-100. KXaopa cannot mean "the broken bread" but only "the fragment."
Here, however, it cannot be translated that way.
"Hence I see no other possibility of overcoming the difficulty than to suppose that the word KXaopa in chap. 9 of the Didache has entered the text under the influence of the Egyptian liturgy" (p. 100).
9.1-5 a hendiadys.37 Zcoij is the "life eternal" (fair) alwucoq) for which 10.3 also gives thanks.38 Cf. "immortality"
(dc&avaoia) in 10.2. This is the true, genuine, real life, the life that alone deserves the name, in contrast to the earthly alienation characterized by death, rvtioLq should be understood in an analogous sense. It is the gift for which 10.2 also gives thanks, and is related to the yvtiocq
Kvpiov in 11.2 (redaction). Knowledge here does not mean the intellectual apprehension of objects, but the revealed knowledge of God, the insight into the hidden meaning of life and the universe that is bestowed by rev elation, the uncovering of what would otherwise be radi cally alien and hidden from the human being. The opposite of this was the ignorance and obscurity of earthly life.39 When, moreover, the key word yudioiq is absent in the concomitant passage of the Apostolic
Constitutions, it is difficult to draw from that fact the conclusion that Kal yvuoewq was originally absent from the text of the Didache as well, and that it was only added in the course of the text's later development.40 It has a solid place in Did. 10.2, and it is also original in 9.3.
Dibelius's hypothesis41 that this was a prayer originally stemming from Hellenistic Jewish circles that had already prayed in a spiritual sense for the heavenly manna is without any support.42
Hq eyvupcoaq rjp.lv 5ia Irjoov tov iraLdoq aov (v. 3b) corresponds, word for word, to the phrase in the bless ing of the cup (9.2). Again we may suppose that ool rj
86%a eiq rovq aitivaq (v. 3c) was spoken as a response by the participants at the meal.43
■ 4 Unlike the blessing of the wine, the blessing of bread is followed by another prayer that, with restrained pathos, establishes an important relationship between the unity of the bread and the eschatological gathering of the church. The symbolism that emerges here reap pears (with appropriate variations) in many later liturgi cal prayers.44 Let me only mention (Ps.-)Athanasius De virg. 13 (von der Goltz, 47.4-9); Serapion Euchologium
13.13; (Funk, Didascalia, 2.174, 18-22); P. Der Balizeh fol.
IIv., lines 3— 11 ;45 plus, of course — though not an inde pendent witness — Const. 7.25.3;46 cf. table 10.47 Still, one need not suppose that in any of these cases (apart from Constitutions, of course) there is a direct literary depen dence on the Didache. What we find here, apparently, is
40 Cf. the suggestions of Peterson, "Probleme," 178.
Wengst (Didache, 78) deleted Kal yvaoeuq.
41 "Mahl-Gebete," 120-21.
42 Cf.: "We will first suppose, hypothetically. ... It may thus be supposed ..." ("Mahl-Gebete," p. 120).
Betz ("Eucharist," 260-61) differs: behind Did. 9.3, he thinks, was Gen 2:9; the ancient eucharistic prayers (and Betz considers the prayers in chaps.
9-10 to be such prayers) understood the eucharistic gifts as the gifts of paradise.
43 Audet (Didache, 402) conjectures aprjv as the response.
44 Moreover, there is a variant of the same idea in Did.
10.5.
45 Roberts and Capelle, Euchologium, 26; van Haelst, "Nouvelle reconstitution," 449.
46 For these parallels, see Clerici, Einsammlung, 104-8.
47 See the further parallels in Clerici, Einsammlung, 108-12. Cf. esp. Cyprian Ep. 63.13 (CSEL 3.2.712): quo et ipso sacramento populus noster ostenditur adunatus, ut quemadmodum grana multa in unum collecta et conmolita et conmixta panem unum faciunt, sic in Christo qui est panis caelestis unum sciamus esse corpus, cui coniunctus sit noster numerus et adunatus ("And under this same sacred image our people are repre sented as having been made one, for just as numer ous grains are gathered, ground, and mixed all
/- 39 Thus the text here is not speaking of knowledge of Scripture (as in Barn. 1.5, etc.; Hermas Vis. 2.2); the concept is also free from polemic (cf., in contrast,
Ignatius Eph. 17.2): this would also be out of place in a liturgical formula. But the tradition in which the Didache operates is utterly free from any hint of danger from Gnosticism. By contrast, 1 Clem. 59.2 and Mart. Pol. 14.1 offer significant material paral lels to the concept of yroxng in the Didache, because here the office of the pais is the communication of knowledge: 1 Clem. 59.2, bid rod rjyairrjpevov
Ttaiboq avrov 'Irjoov Xpiorov, be ov exdAeoev rjpdq dnrb OKorovq elq (pwq, onto ayvoiolaq eiq etrlyvaoLv boi-rjq ovoparoq avrov ("through his beloved ser vant, Jesus Christ, through whom he has called us from darkness into light, from ignorance into the glorious knowledge of his name"); Mart. Pol. 14.1,
Kopec 6 dcoq 6 7ravroKparap, 6 rov dyairrjrov Kal evAoyrjrov iraeboq oov Irjoov Xpeorov rtarrjp, be' ov rrjv irepl oov eir'eyvaoev cLArppapev ("Lord, God almighty, Father of your beloved and blessed ser vant, Jesus Christ, through whom we have received knowledge of you"). These passages are also impor tant because both belong to the liturgy: the passage from the Martyrdom of Polycarp is probably from the anaphora prayer of the community at Smyrna. Cf.
Harnack, "Bezeichnung," 221.
Table 10 Did. 9.4
(Ps.-)Athanasius De virg. 13
Serapion Euchologium 13.13
P. Der Balizeh, IIv. 11. 3-111 uoirep pv Kai Kafiuq Kai uoirep [Kai ov Tpojirov TOVTO 6 dproq ovToq 6 apToq ovToq ofvToq 6 apjroq Si eoKopinopevov2 EOKopmopevoq* virapxEi eoKopmopevoq pv eoKopiaopevoq pv eiravu tuv opeuv 0 eiravu erravu tuv opeuv [ejiravu [tuv opeuv] — Tavrpq rrjq Tpane^pq — Kai [Sofvjvuv Kai apovpuv Kal ovvaxdtv Kai ovvaxdziq Kai ovvax&tiq Kafi ovppjiyeiq eyevero ev, eye veto ev, eyevETO eiq ev, e[yev]eTO ev oupa, . . .
- — - (lines 6—9)4 ovtu avvaxdpTU ovTuq Emovvax^VTu ovtu Kai Tpv ayiav ovT[uq] eir[iovva]%ov oov r) e/c/cArjaia oov r\ EKKApoia oov EKKApoiav ovva^ov Tp[v] Ka-doAiKpv [ek]k Ap[oi - — tov I(poo)v X(piOTO)vJ.5 am) tuv irepdruv dm) TUV TTEpaTUV ek iravToq edvovq6 Tfjq ypq Trjq yfjq eiq tpv opv flaoiAeLav eiq Tpv (3aoiAeiav Kai iroipoov piav oov, £uoav KadoAiKpv EKKApoLav.7 otl oov eotlv p Soija otl oov eotlv t) bvvapiq /cat 17 bvvapiq Kai 17 Soija bid 'Ipoov XpioTov eiq Tovq aiuvaq. eiq Tovq aiuvaq t. aiuvuv, — appv.
1 I am citing the text of van Haelst, but not in the lines of the papyrus; the lines are adjusted to this synopsis.
2 My emendation; see below.
3 V.l. bieoKopiriopevoq (cf. von der Goltz, 47.4-9).
4 Lines 6-9: . . . oaw [. . . Ka$]wg 6 olvoq o[vToq] 6 el-eAduv ek Tp[q ayiaq dJpireAov A[auei.S] /cat to vbup eij ap(vov apujpov /cat ovpfpLKTa] .6lo eyeveTo 'ev JpvJoTpptov.
5 The words of institution follow.
6 Added: /cat iraor\q x^Pa<7 Ka'L ndopq iroAeuq /cat Kuppq /cat olkov.
7 In 13.14 there follow the words of institution over the cup. an especially ancient piece of liturgy. Our passage pre sents the oldest example in a liturgical formula.48 have been dproq, but in that case the whole syntax would have to be changed.49 It seems simpler to me to delete KAaopa in 9.4 as a later (Egyptian) addition; then the conjectural <ro>50 would also be superflu ous. The Apostolic Constitutions has preserved the orig inal text: uoirep pv tovto bieoKopiriopevov.51 The suggested conjectures by Jean Magne are highly imag inative.52 (2) Aid Ipoov Xpiorov could be a later
Excursus: The Text of the Prayer There may have been an occasion for improvement of the Didache text at three different points. (1) KAaopa may not be original (see above). The original may together to make one loaf of bread, so in Christ, who is the bread of heaven, we know there is but one body and that every one of us has been fused together and made one with it"; trans. G. W. Clarke,
The Lettters of St. Cyprian of Carthage, vol. 3: Letters 55-66 [New York and Mahwah, N.J.: Newman,
1986] 105); and 69.5 (CSEL 3.2.754): nam quando Dominus corpus suum panem vocat de multorum granorum adunatione congestum, populum nostrum quem portabat indicat adunatum ("For when the Lord calls ed into one"; trans. G. W. Clarke, The Lettters of St.
Cyprian of Carthage, vol. 4: Letters 67-82 [New York and Mahwah, N.J.: Newman, 1989] 56). (The com parison is then continued with regard to the wine.)
J. Gribomont has investigated the motif of ecclesiam adunare in the Roman canon ("Ecclesiam adunare:
Un echo de l'eucharistie africaine et de la Didache,"
RThAM 27 [1960] 20-28). Interesting in the present context is the suggestion that the expression can be traced to the influence of the "African" liturgy (p.
26). Gribomont posits the influence of the Didache on the African liturgy (p. 27: hardly likely), but he himself raises the objection that there could be bread His own body—and bread is a conglomerate of many individual grains, made into one — he sig nifies that we, the people whom He bore, are unit
9.1-5 Esreh (Palestinian recension) 10: "Raise a banner for the gathering of our exiles; praised art thou, YHWH, who gathers the scattered of your people Israel!"56 Cf. the Babylonian recension: "Raise a banner to gather all our exiles from the four ends of the earth into our land; praised art thou, YHWH, who gathers the scat tered of your people Israel."57 Cf. the Musaph prayer for Rosh Hashanah 4;58 and, finally, a blessing with the schema ahaha rabba: "bring us together in peace from the four corners of the earth, and bring us in upright ways into our land."59 expansion (so Audet).53 It is absent from the parallels in Ps.-Athanasius and in Did. 10.5. The text seems to be established, however, by the bi avrov in Const.
7.25.4. (3) In the doxology we should perhaps estab lish the sequence: on oov eonp r\ bvpapuq /cat, tj <5o£a . . . eig rovq aiupaq; cf. Did. 10.5 and the parallel text in Ps.-Athanasius De virg. 13.
The Old Testament root and foundation of the pas sage is found in the prophetic hope for the gathering of the people of Israel in the day of salvation.54 Cf.
Isa 11:12: /cat ovpagei. rovq artoXopepovq lopar)X nai rovq bteartappepovg lovba ovpa£ei e/c tup reooapup rtrepvyup rrjq yrjq ("And he will gather the lost of Israel and the scattered ones of Judah he will gather from the four corners of the earth"); Jer 39:37 LXX: ibov eyu ovpayu avrovg e/c rtaopg rfjq yrjq, ov bieorreipa avrovg . . . /cat ertiorpepu avrovg eig top
Israel's hope for the reunion of those who have been separated has entered, mutatis mutandis, into the primitive Christian hope for the final assembly of the elect. Cf. Mark 13:27: "And then he will send his angels and gather the [his] elect from the four winds, from the end of the earth to the end of heaven" (kai
Tore artooreXei rovq ayyeXovq nai eraavpa^et rovq toitop tovtop ("Behold I gather them from all the earth where I scattered them . . . and I shall return them to this place"); Ezek 11:17: /cat ovpayu avrovg eKXeKTOvq [avrov] e/c tup reooapup apepup art aKpov yrjq euq anpov ovpapov; cf. Matt 24:31). For this theme, cf. also John 11:52.60 It is obvious why this hope found expression in the church's liturgies. Of course the adoption of the Jewish motif within the Christian liturgy required certain adaptations: What this community implores is no longer the gathering of Israel and the triumphant reunion of the whole people, but the gathering of the church onto iteparup e/c tup xwpcof, ov bieortetpa avrovg ep avraig, /cat buou avrolq rpp yrjp rod 'Iopar)X ("and I shall gather them from the lands where I scattered them, and I shall give them the land of Israel"); Zech 2:10 LXX: e/c tup reooapup apepup rod ovpapov ovpdgu vpag,
Xeyet Kvptoq ("From the four winds of the heaven I shall gather you, says the Lord").55 This hope entered into the prayer language of Judaism; see Shemoneh
/_ 55 Cf. also Deut 30:1-5; Isa 27:13; 43:5-7; 56:8; Jer 23:8; 31:8, 10; Ezek 28:25; 34:11-16; 37:21; 39:27;
Mic 4:12; Pss 106:47; 147:2; Neh 1:8-9; 1 Chr 16:35; Sir 36:11; 51:12(f); 2 Macc 1:27; 2:7, 18; Tob 13:5;
Ps. Sol. 8.28; 11.2ff.; 17.26, 44. This motif plays a special part later in the Testaments of the 12
Patriarchs (cf. Clerici, Einsammlung, 76-77): T. Jud.
23.5; T. Iss. 6.2-4; T. Dan 5.8-9; T. Naph. 4. The leading of Israel home is occasionally combined with the salvation of the "nations": T. Ash. 7.2-3; common dependence on a "paleo-Christian liturgy"
(ibid.). Cf. also the doubts regarding the acceptance of a relationship to the Didache in Clerici,
Einsammlxmg, 135-42.
48 Clerici has discussed the changes experienced by this motif at a later time. We will not pursue the matter here. Cf. the interpretation of the later his tory in Clerici, Einsammlung, 104-12.
49 Cf. the suggestions of Voobus, Liturgical Traditions, 88-89, 146-48 (following Peterson); and see above, n. 31.
50 Usually included since Gebhardt.
51 Niederwimmer, "Textprobleme," 124-25; now also Wengst, Didache, 80.
52 "Klasma, sperma, poimnion: Le voeu pour le rassemblement de Didache IX, 4," in P. Levy and E.
Wolff, eds., Melanges d'histoire des religions offerts a Henri-Charles Puech (Paris: Presses universitaires de
France, 1974) 206-7. According to Magne, orteppa previously stood where nXdopa is now, and still ear lier (in harmony with OT models) it was itoipPLOP.
In addition, Magne (p. 208) makes the Didache a product of Gnosticism.
53 Audet, Didache, 234.
54 For the prehistory of the symbolism in Did. 9.4
(10.5), see Riesenfeld, "Brot von den Bergen,"
T. Benj. 9.2 (but do these statements belong to a subsequent layer?). For the linguistic parallels with the Didache, cf. also T. Naph. 6.7: "We were all scat tered to the ends of the earth" (bieortap-qpep ovp oi rtapreq euq eig ra rtepara); and T. Ash. 7.2: "You will be scattered to the four corners of the earth"
(vpeig biaoKoprtiO'd'qoeode eig rag reooapaq yupiaq ri)g yrjq)- See Clerici, Einsammlung, 76.
56 Holtzmann, Berakot, 16.
57 Holtzmann, Berakot, 20.
58 Staerk, Gebete, 24.
59 Staerk, Gebete, 6; German translation in Str-B 1.398.
60 Later: Mart. Pol. 22.3: ipa Kape ovpaydyt] 6 Kvpcoq 'IrjooOg XpLoroq pera tup enXeKrup avrov eig rr\p ovpaPLOP faotXeLap avrov—the final wish of the author ("May the Lord Jesus gather me as well with his elect into his heavenly kingdom"). In contrast, it
145-47; Clerici, Einsammlung, 67-92.
The prayer almost constitutes an "interpretive saying."
As the bread has come into being through the gathering of grains once scattered on the surrounding hills,67 so may God gather the diaspora of the church68 from the ends of the earth69 into God's reign. The jSamAeta is understood in an eschatological sense as the realm of the end time, in which all dispersion will be at an end.
God's reign is determinative for the church. rfjg ypg and its being brought home into the PaocAeia (sc. rot; t9eo0).61 In the liturgy of the Didache the hope for eschatological reunion of the separated is related to the meal celebration.62 As the bread eaten at the common meal ovpax&ep eyepero ev, so will the church be—on that day. It should be obvious on the face of it that the symbolism of Did.
9.4a should not be examined with logical stringency: the phrases about what was (previously) "scattered on the hillsides,"63 and is (now) gathered into one,64 do not strictly describe some particular techniques in farming and bread production,65 but are simply intended to point to and beyond the one common element (with respect to bread and church): one bread from the bteoKOpTaapePOP "on the surrounding hills"; one church from many who are now dispersed throughout the world.66
The community gathered for the meal responds70 to this prayer by saying: on oov eoriv 17 <5o|a kul rj 5uuapug dca Irjoov Xpiorov elg Tovg oduvag.71
■ 5 At this point the Didachist72 has inserted a rubrical comment. Only those baptized (in the name of the Lord)73 may take part; only they may eat and drink "of your thanksgiving [meal]." It is not clear whether here seems to me that the eporpg motif in John 17:11,
20-23 belongs in a different context.
61 God's basileia: in Did. 8.2 (in the quotation of the Lord's Prayer) and in 10.5, the material parallel to 9.4. It seems questionable to me whether the Didache text thinks directly of the metaphor of "har vest" for the "judgment of the world" in connection with the gathering of the scattered (Riesenfeld,
"Brot von den Bergen," 146-47).
62 Cf. earlier 1 Cor 10:17, although there with regard to the Eucharist; Ignatius Eph. 20.2. As the parallels show, the motif had its later Sitz im Leben primarily in the anaphora of the Eucharist. The same motif reappears in Did. 10.5, a prayer that looks back in gratitude to the "agape" but at the same time looks forward to the sacrament.
63 'Etrapu tup opeup remains uninterpreted thus far.
Riesenfeld ("Brot von den Bergen," 149-50) sug gested a connection with Isa 40:4, and recalls 1 Bar.
5.5-8 and Ps. Sol. 11.2-4. This is highly question able.
64 H reads /cat ovpax^P eyepero cp. Wengst (Didache 80) decided (with Constitutions) for elg aprog instead of ep.
65 For information on the ritual gathering of wheat for the eucharistic bread, see Adam, "Herkunft,"
57-61, who sees in it evidence of the "east Syrian roots of the Didache" (p. 57).
66 The interpretation presented here avoids making claims with regard to Johannine reminiscences.
There is sometimes a different tendency in recent research, e.g.: Goodenough, "John a Primitive
Gospel," JBL 64 [1945] 174-75, who sees in this pas sage (besides dependence on Jewish liturgy) remi niscences from the Fourth Gospel (chap. 6); also
Moule, "Note," 240-43. Moule finds the received interpretation problematic (instead of /cAaa/xa, one would tend to expect aprog, and instead of bueoKop- ■rnopepop rather bLeotrappePOP; avpayeiv must remain odd in the traditional interpretation, and why eirdpu tup opeup? pp. 240-41). These difficul ties are largely eliminated if (with Moule, "Note,"
242) one refers Did. 9.4 to "the gathering of the kAaopara after the feeding of the multitude" (sc. according to John 6). Even the eirdpu tup opeup would be explained, namely, as an echo of John 6:3
(elg to opog). To the contrary, and correctly, Riesenfeld, "Brot von den Bergen," 143-45. Lucien
Cerfaux has contradicted Riesenfeld's presentation and again established connections to John 6 and 11:52, although not necessarily literary connections
("La multiplication des pains dans la liturgie de la Didache," Bib 40 [1959] 943-58). Against the pre sentations by Goodenough, Moule, and Cerfaux:
Clerici, Einsammlung, 92-102; Voobus, Liturgical Traditions, 137-57; idem, "Regarding the Background of the Liturgical Traditions in the Didache: The Question of Literary Relation between Didache IX,4 and the Fourth Gospel," VC
23 (1969) 81-87. I also regard the derivation from the Fourth Gospel as out of place. Cf. also Rordorf and Tuilier, Doctrine, 46 n. 3. Possible relationships to John 11:52 can be adequately explained by their dependence on a common liturgical tradition. On the work of Magne, see above, p. 151 n. 52.
67 For the uncontracted genitive plural opeup, see
BDF, §48; BDR, §48.
68 On this see Kurt Niederwimmer, "Kirche als Diaspora," EvTh 41 (1981) 294. "The exxArjata, here and in 10.5, is the whole community (different ly from 4.14)." See also Knopf, Lehre, 27; and previ ously Harnack, Lehre, 31.
69 For rd nepara rrjg yijg, cf. T. Naph. 6.7; Matt