The earliest eucharistic prayers?
Because of the shortage of other substantial evidence for the form that early eucharistic prayers would have taken, two liturgical texts – the meal prayers in the Didache and the eucharistic prayer in the so-called Apostolic Tradition of Hippolytus – have come to assume a position of crucial importance in attempts that have been made to reconstruct the line of development of such prayers from the earliest times down to the fourth century, when more examples of eucharistic anaphoras become visible. Unfortunately, both these texts present problems of interpretation and have often been wrongly used to support inaccurate conclusions.
The scholarly consensus that emerged during the course of the twentieth century was that Christian eucharistic prayers had developed out of the Jewish grace after meals, the Birkat ha-mazon. This was in spite of the fact that a growing number of Jewish scholars began to express doubts as to whether such prayers would have existed in the first century in the fixed form in which they are later found,1 and also despite the fact that in Jewish tradition that particular prayer came to mark the point in the meal after which no further food could be consumed, although wine might still be drunk.2 Enrico Mazza was a partial exception to this consensus. Although still seeing some eucharistic prayers as emanating from the grace after meals, he traced the roots of the eucharistic prayer in the fourth-century Apostolic Constitutions 7 through its source in the Didache to the short Jewish blessings over cup and bread that would have been used earlier in the meal, and he proposed that the Yotzer, one of the blessings before the Shema in Jewish morning prayer, was the ultimate source of the eucharistic prayer of the Strasbourg Papyrus.3 Yet there is no reason to suppose that these other Jewish texts existed in a definitive form at such an early date any more than the grace after meals did, nor that the early Christians would have cast around for an already existing prayer from a quite different liturgical context – morning prayer – to use as a model for their praying over their meal rather than develop one themselves. In any case, very few later Christian eucharistic texts show any obvious influence of Jewish antecedents, however hard some scholars have tried to find them there. However, let us start with those that do. Chief among these are the meal prayers in the Didache.
Didache 9–10
9.1 Concerning the thanksgiving, give thanks thus:
2 First, concerning the cup:
We give thanks to you, our Father, for the holy vine of David your child, which you have made known to us through Jesus your child; glory to you for evermore.
3 Concerning the broken bread:
We give thanks to you, our Father, for the life and knowledge which you have made known to us through Jesus your child; glory to you for evermore.
4 As this broken bread was scattered upon the mountains and having been gathered together became one, so may your church be gathered together from the ends of the earth into your kingdom; for yours is the glory and the power through Jesus Christ for evermore.
5 Let no one eat or drink of your eucharist but those who have been baptized in the name of the Lord. For concerning this also the Lord has said, ‘Do not give what is holy to the dogs.’
10.1 After you have had your fill, give thanks thus:
2 We give thanks to you, holy Father, for your holy name which you have enshrined in our hearts, and for the knowledge and faith and immortality which you have made known to us through Jesus your child; glory to you for evermore.
3 You, Almighty Master, created all things for the sake of your Name and gave food and drink to humans for enjoyment, that they might give thanks to you; but to us you have granted spiritual food and drink and eternal life through Jesus your child.
4 Above all we give thanks to you because you are mighty; glory to you for evermore. Amen.
5 Remember, Lord, your church, to deliver it from all evil and to perfect it in your love, and gather it together from the four winds, having been sanctified, into your kingdom which you have prepared for it; for yours is the power and the glory for evermore. Amen.
6 May grace come, and this world pass away. Amen.
Hosanna to the God of David.
If anyone is holy, let him come; if anyone is not, let him repent.
Marana tha. Amen.
7 But allow the prophets to give thanks as they wish.
Although in the past many scholars tried to exclude these prayers from consideration on the grounds that they did not resemble later eucharistic prayers, and especially because they lacked a narrative of the Last Supper, and so this meal cannot have been a eucharist as such,4 the tide has turned among liturgical historians, and the majority now do regard it as an early form of the eucharist. Yet that does not solve all the problems related to the material. Providing an explanation that does justice to all of its significant features is an extremely difficult task, which is why so many different theories about it can still abound. One recent commentator on the text, Alan Garrow, has likened it to trying to cover all four corners of a mattress with an insufficiently large fitted sheet.5 Any who have struggled to make such a bed will readily understand his comparison: it is often possible to get the sheet to cover three corners, but the fourth can only be managed at the expense of one of the others. What we might call the ‘awkward corners’ of the Didache text are chiefly as follows:
• the fact that 10.1 speaks of having just completed a full meal, while 10.6 appears, at least to some scholars, to include an invitation to receive communion after the meal, creates difficulties for those who want to see the meal itself as eucharistic;
• on the other hand, the restriction of the eucharist to the baptized in 9.5 appears to contradict this and imply that the meal is eucharistic;
• similarly, the use of the word klasma, ‘broken piece’, in 9.3 and 9.4 rather than artos, ‘bread’, is perhaps suggestive that the meal is eucharistic, but it appears to imply that the bread has already been broken before the prayer rather than after it, which is odd;
• and finally, the fact that the prayers in both 9 and 10 have strong parallels to one another and both seem to be eucharistic in character makes it difficult to regard one set as related to an ‘ordinary meal’ and the other not, and yet there cannot be two eucharists here, can there?
Garrow’s own solution to these difficulties is to propose that there are in fact two parallel liturgies, one comprising chapter 9 and the other chapter 10, and that ‘they represent two separate accounts of the same liturgical event’.6 He draws particular attention to the parallels in structure and contents between the two sets of prayers. He believes that in each case a full meal would have preceded the prayers, which is indicated by the rubric in 10.1 and implied by the use of klasma in chapter 9, referring to a large piece of the loaf which would have been broken at the beginning of the meal before these prayers. In both cases, then, the prayers refer forwards to the eucharistic communion that is to follow, indicated in the one case by the direction in 9.5 about who may receive communion, and in the other case by the invitation in 10.6, which also refers to those who may and may not receive. He concludes that these two chapters belong to separate layers of tradition, and were juxtaposed in the compilation of the Didache only because of their common subject matter.
Garrow may have hit upon something close to the right answer. The striking similarities in both structure and content of the two sets of prayers which he has emphasized do seem to point towards their being alternative outworkings from the same root tradition, those in chapter 9 representing a more primitive version of the material with the units over cup and bread still being clearly differentiated from one another, while in chapter 10 it has been moulded into a more continuous whole. Garrow’s hypothesis thus helps solve another puzzling feature of the text which has commonly been ignored by scholars – how it could be that a single meal ritual could include prayers that were nearly duplicates of one another in both form and content. While there is certainly no shortage of historical examples of liturgies that contain alternative versions of the same material within the one rite, this is nearly always the result of the fusion of what had earlier been quite discrete and parallel texts rather than the work of a single compiler.
However, I would want to modify Garrow’s view somewhat. He assumes (a) that the prayer material in these chapters reached the compiler of the Didache complete with the accompanying directions; (b) that the rite was then still in actual liturgical use; and (c) that the compiler juxtaposed the parallel sets of prayers merely because of their commonality of subject. All these assumptions are questionable. First, early liturgical prayers seem to have been commonly transmitted without any accompanying directions as to their precise use or location within a rite. Thus, for example, in the earliest-known collection of Christian liturgical prayers, the mid-fourth-century Sacramentary of Sarapion, most prayers have only a very general title attached to them (such as ‘Prayer for Those being Baptized’ or ‘Prayer for One Who Has Died and is Being Carried Out’).7 In the light of this, it seems probable that the prayers themselves in Didache 9 and 10 were at first in oral circulation and were received by the compiler with little, if any, indication as to their exact location within the ritual meal. Second, there is no reason to suppose that when the compiler did acquire these prayers, they were in current use; indeed, since it was not usual to write down prayers at that time, the primary reason for doing so and including them within the church order was very probably that they were becoming obsolete and in danger of being forgotten. Third, there is again no reason to suppose that the compiler viewed them as two alternative sets of prayers for the Christian meal, even though that was apparently what they were, but it is at least just as likely that he understood them to belong to the one ritual.
However, if they had no accompanying directions, except probably for what became 9.5 (which is of a different sort – a community rule rather than an instruction about a liturgical action), how should they be arranged? Which set should come first and which second? Let us imagine the situation of the compiler. The prayer in 9.2, with its reference to the holy vine of David, appeared to be intended as a prayer over wine, and so he prefaced it with the directions: ‘Concerning the thanksgiving, give thanks thus: First, concerning the cup’. The second unit with thanksgiving and petition mentioned broken bread, klasma, and so the compiler inserted before it, ‘And concerning the fragment (peri tou klasmatou)’, rather than ‘And concerning the bread (peri tou artou)’.8 But what should he then do with the material we know as chapter 10, without any accompanying rubrical assistance and merely a more general reference to spiritual food and drink in the text? Since he already had prayers for use over cup and bread before the meal, he concluded that it must have been intended for use after the meal was over, and so inserted the direction, ‘After you have had your fill, give thanks thus’, and left it at that.
In short, what I am proposing is that, while the prayer material in chapters 9 and 10 may well be very ancient and authentic, its layout in the Didache is later and completely artificial and so tells us nothing at all about the structure of primitive eucharistic celebrations. It certainly does not require us to think that the meal must have been eaten before prayers over the cup and bread were said and the eucharistic elements distributed, for once the direction in 10.1 is eliminated, the presence or absence of a meal either before or after the prayers becomes an entirely open question. Nonetheless, understanding the material as composed of two parallel variants of the same liturgical tradition does enable us to glimpse two distinct stages in the evolution of eucharistic prayers, from separate prayers over cup and bread in chapter 9 to a single prayer over both in chapter 10, with no indication there of any temporal priority of cup over bread or of bread over cup. The hypothesis that the contents of chapters 9 and 10 circulated independently of one another may also help explain the presence of some material paralleling chapter 9 but not accompanied by any material from chapter 10 in two later texts, in the eucharistic prayer of the Sacramentary of Sarapion and in the grace at table for a religious community of women in the anonymous De virginitate, formerly but wrongly attributed to Athanasius.9 Above all, it resolves nearly all the difficulties associated with other theories, and so is worth taking seriously as an explanation for this puzzling but crucial text in the early history of eucharistic rites.
A different trajectory
Whatever the true origin and nature of the material in Didache 9–10, what is clear is that, while it is certainly Jewish in character, for the reasons outlined at the beginning of the chapter it cannot simply be described as an adaptation of existing Jewish texts but should rather be seen as a natural development within a Jewish-Christian milieu.10 When one turns to other early eucharistic prayers, however, there are very few indeed that have survived that retain a similar Semitic style. First, there are three fourth-century texts that exhibit a literary relationship with the Didache material – the prayer in Apostolic Constitutions 7 mentioned by Mazza and referred to at the beginning of this chapter, together with the two instances mentioned above, the middle of the eucharistic prayer of the Sacramentary of Sarapion and De virginitate 12–13. Taken together, these three texts seem to point to the conclusion that the kind of prayers found in Didache 9–10 continued to be used at least within certain limited segments of early Christianity for a considerable number of years, but apparently eventually faded away.
Equally striking in its resemblance to Jewish patterns of praying is the Anaphora of Addai and Mari, composed in Syriac rather than Greek, the core of which must go back at least to the fourth century and very probably much earlier still.11 Although some scholars have claimed to see parallels in it to the tripartite Jewish grace after meals,12 and others to the bipartite morning prayers said before the Shema each day,13 there are enough significant differences from both these texts that render it extremely improbable that either of them constitutes the primary source for this prayer.14 Nonetheless, the similarities that do exist seem to confirm its roots as having been in a Jewish-Christian context.
Yet such texts are the exception rather than the rule among the extant eucharistic prayers that are known from the fourth century onwards. The majority seem rather to have left their Jewish roots behind and taken a different trajectory. There are some signs that at first in some communities separate prayers may have continued to be said over bread and cup, as they were in Didache 9, rather than in a united form, as may have been the case with the prayer in Didache 10. Thus Justin Martyr in the second century uses the plural when describing what the president does (he ‘sends up prayers and thanksgivings to the best of his ability’: First Apology 67.4) and the third-century Syrian Didascalia suggests that a visiting bishop might be invited to ‘speak the words over the cup’.15 There are also signs that some prayers may have been created by combining together originally separate units, as were their Jewish antecedents, rather than as continuous through compositions.16 And of course they all continue to offer praise, just as the Jewish meal prayers had done. But their style is quite different, and the prayers are described by authors from the late second century onwards as being an ‘invocation’ (epiklesis), suggesting that the emphasis was seen as falling more on their petitionary aspect than on the eucharistia alone. Indeed, that word came instead to refer to the eucharistic elements, and especially the bread.17 Before looking further at these later texts, however, we need first to deal with one prayer that conventionally has been thought to date from the early third century and to be typical of eucharistic praying in general at that period, a conclusion that thus greatly affected how early anaphoras were understood to have developed.
The eucharistic prayer in the so-called Apostolic Tradition of Hippolytus
4.3 The Lord [be] with you.
And let them all say: And with your spirit.
Up [with your] hearts.
We have [them] to the Lord.
Let us give thanks to the Lord.
It is worthy and just.
And so let him then continue:
4 We render thanks to you, God, through your beloved child Jesus Christ, whom in the last times you sent to us as savior and redeemer and angel of your will,
5 who is your inseparable word, through whom you made all things and it was well pleasing to you,
6 [whom] you sent from heaven into the virgin’s womb, and who conceived in the womb was incarnate and manifested as your Son, born from the Holy Spirit and the virgin;
7 who fulfilling your will and gaining for you a holy people stretched out [his] hands when he was suffering, that he might release from suffering those who believed in you;
8 who when he was being handed over to voluntary suffering, that he might destroy death and break the bonds of the devil, and tread down hell and the illuminate the righteous, and fix a limit and manifest the resurrection,
9 taking bread [and] giving thanks to you, he said: ‘Take, eat, this is my body that will be broken for you.’ Likewise also the cup, saying, ‘This is my blood that is shed for you.
10 When you do this, you do my remembrance.’
11 Remembering therefore his death and resurrection, we offer to you the bread and cup, giving thanks to you because you have held us worthy to stand before you and minister to you.
12 And we ask that you would send your Holy Spirit in the oblation of [your] holy church, [that] gathering [them] into one you will give to all who partake of the holy things [to partake] in the fullness of the Holy Spirit, for the strengthening of faith in truth,
13 that we may praise and glorify you through your child Jesus Christ, through whom [be] glory and honor to you, Father and Son with the Holy Spirit, in your holy church, both now and to the ages of ages. Amen.18
Among the manuscripts of ancient church orders discovered by scholars during the nineteenth century was one that, because it was anonymous and untitled, was given the name ‘The Egyptian Church Order’ by Hans Achelis in 1891, since up to that date all the linguistic versions in which it had been found had belonged to that part of the world – two dialects of Coptic (Sahidic and Bohairic) and Ethiopic. In 1900, however, Edmund Hauler published an incomplete fifth-century Latin text of it, and four years later George Horner added an Arabic version to the collection. Although it was agreed that the original behind all these versions had been written in Greek, only a few fragments of the text in that language have ever been discovered.
It was in 1906 that Eduard von der Goltz came up with the suggestion that it might in reality be a work by Hippolytus of Rome, the Apostolic Tradition, previously believed to have been lost,19 and this theory was then taken up and elaborated, first by Eduard Schwartz in 1910, and then quite independently and much more fully by R. H. Connolly in 1916.20 This verdict was subsequently accepted by the great majority of scholars and the church order was commonly assumed to represent the official liturgy of the Church at Rome in the early third century. Because, apart from the Didache, there were no other extant texts containing liturgical material that could be attributed with any certainty to such an early date, and because it was thought to reflect the practice of such a major centre of ancient Christianity, it came to play a crucial role not only in scholarly reconstructions of primitive Christian worship practices but also in the process of liturgical revision undertaken in many mainstream Christian denominations in the second half of the twentieth century. Out of a desire to restore in modern practice elements of early Christian worship that had been lost or at least diminished in later liturgical traditions, parts of this church order, and especially its eucharistic prayer, were copied and adapted in many new rites that were being composed and authorized for use at this time.
The attribution to Hippolytus rested mainly upon two principal foundations. The first was that, while no existing manuscript of the document itself bore a title or author’s name, two other church orders derived from it do make reference to Hippolytus. One is actually entitled The Canons of Hippolytus, and the other – the Epitome of Apostolic Constitutions Book 8 – introduces a subheading, ‘Constitutions of the Holy Apostles concerning Ordinations through Hippolytus’, at precisely the point where it begins to draw directly upon a text of the document rather than upon Apostolic Constitutions Book 8 itself. The second argument was that both the prologue and epilogue of the work apparently use the expression ‘apostolic tradition’. That there had once been a work by Hippolytus called ‘The Apostolic Tradition’ seemed to be proved by its inclusion in a list of titles of writings inscribed on the right-hand side of the base of a statue discovered in Rome in 1551, most of which were known to have Hippolytus as their author.
Although there had always been a few scholars who had doubted the veracity of the consensus that had been reached with regard to the identity of the document, it was not until about twenty years ago – long after modern versions of the material had become established parts of contemporary service-books – that their number began to grow and their arguments were taken seriously by an ever-widening circle of liturgical historians. First, an important series of articles by Marcel Metzger opened up a completely new line of approach to the text.21 He developed an idea earlier advanced both by Jean Magne and by Alexandre Faivre,22 that not only was it not the Apostolic Tradition of Hippolytus, it was not the work of any single author at all but rather a piece of ‘living literature’. Metzger argued that its lack of unity or logical progression, its frequent incoherences, doublets, and contradictions all pointed away from the existence of a single editorial hand. Instead, it had all the characteristics of a composite work, a collection of community rules from quite disparate traditions. More recently, Christoph Markschies argued not only that the ascription of the Canons of Hippolytus to Hippolytus and the reference to him in the subheading of the Epitome were not made until the late fourth or early fifth century (and thus much too late to credit them with any historical reliability), but also that the apparent references to ‘apostolic tradition’ in the prologue and conclusion of the document had been misinterpreted by other scholars and consequently did not allude to the title of the work.23
In any case, the tendency to associate documents with apostolic figures or with those believed to have close connections to such persons so as to enhance their authority is very common in the ancient Christian world, and there are certainly other works that are known to have been falsely attributed to Hippolytus.24 Yet even the very existence of a work entitled Apostolic Tradition by Hippolytus of Rome is not above suspicion. The list on the statue does not correlate exactly with the works of Hippolytus that are catalogued both by Eusebius and by Jerome. Very surprisingly, it omits those that are most strongly attested as genuinely his, including the commentary on Daniel,25 and this has led some scholars to propose the existence of two authors or even a school of authors as responsible for the works on the list,26 and Alistair Stewart-Sykes to build on this and argue that the Apostolic Tradition itself is the product of several hands from the same Roman school.27 John Cerrato has gone further and suggested that because Hippolytus was a common name in the ancient world, the various works may be the creations of quite different and unrelated authors from diverse places.28 In a final bizarre twist to the tale, modern research has revealed that the statue itself was in origin not a representation of Hippolytus at all but of a female figure, which was restored in the sixteenth century as a male bishop because of the list of works inscribed on its base, using parts taken from other statues.29
If the new approach is correct, as my colleagues and I have argued that it is, then the so-called Apostolic Tradition is actually an aggregation of material from different sources, quite probably arising from different geographical regions and almost certainly from different historical periods, from perhaps as early as the middle of the second century to as late as the middle of the fourth. It is most improbable that it represents the actual practice of any single Christian community, but is best understood by attempting to discern the various individual elements and layers of which it is made up. Moreover, the composite character that the document displays extends also to the individual ritual units within the text, such as ordination, baptism, and even the eucharist itself, which appear to be artificial literary creations, made up of elements drawn from different local traditions rather than comprising a single authentic rite that was ever celebrated in that particular form anywhere in the world.30
Such a judgement obviously has very significant consequences for the status of the eucharistic prayer within it:
• It can no longer be taken for granted that it represents the form used at Rome or anywhere else as early as the beginning of the third century. Indeed, it appears to be a much later insertion into the developing church order and that the original eucharistic material in the text was the directions about a meal in chapters 25–27 – a view with which even the more conservative Alistair Stewart-Sykes concurs.31
• This does not necessarily mean, however, that the prayer was entirely a later composition. Some of the vocabulary found within it, especially reference to Jesus as God’s ‘child’ rather than ‘Son’ and the use of ‘angel of your will’ as a Christological title (4.4), is distinctive of Christian authors in the first half of the second century and dropped out of use after that.
• On the other hand, other elements of the prayer, in particular the inclusion of the institution narrative (4.9–10) and also a quite developed form of epiclesis that asked God to send the Spirit on the oblation of the Church (4.12), have parallels only in literature from the middle of the fourth century onwards, and even then they are not found everywhere, suggesting that they were only just beginning to find their place in the rites.32
• At the same time, it lacks the Sanctus, which other prayers of that period seem to have already adopted (except perhaps for the Strasbourg Papyrus) – as a more integral part of their composition in the case of prayers of Egyptian origin and as a secondary insertion, to a more or less successful extent, in various Syrian texts.
This combination of older and newer features points to the conclusion that it probably attained its final form around the middle of the fourth century and that some version of the prayer was in existence from quite early times, even though it did not form part of the church order itself until much later, when it was updated with other elements that were then becoming current in eucharistic prayers of the period. Its earliest stratum appears to have been a substantial hymn of praise for redemption through Christ (a version of 4.4–8), to which a brief offering/thanksgiving formula (part of 4.11) and a short petition for the communicants were appended (the second half of 4.12) to make it suitable for eucharistic use.33 There is, however, nothing that would connect this prayer specifically with any Jewish roots. Its style, structure, and vocabulary do not reveal any notable semiticisms. The only connection appears to be that, like its Jewish counterparts, it offers praise to God over food, although in this case not for food. Hence its initial composition seems to belong to a Gentile Christian milieu. Nor is there anything that connects the text specifically with Rome. Although there is nothing that can definitively exclude such a provenance, it is quite unlike what later emerges as the Roman eucharistic prayer, and because all its later accretions are similar to developments that were taking place in West Syria and to a lesser extent Alexandria in the fourth century, it appears highly probable that it acquired its later shape there rather than in the West.34 Whether it is itself the source of these developments or simply one of several manifestations of them in eucharistic prayers known to us from the second half of the fourth century cannot be determined with any certainty.
Conclusion
While ultimately having their origin in Jewish meal prayers, Christian eucharistic prayers in a Gentile environment quickly left those roots behind, and apart from retaining the praise motif in the first part of the anaphora, reveal no overt connection to their parent. By the fourth century they had ceased to give thanks for food and drink and instead either offered praise exclusively for creation (as in the case of the Strasbourg Papyrus and the anaphora at Jerusalem known to the author of the Mystagogical Catecheses) or alternatively told the story of the saving acts of Christ at some length (as in the Apostolic Tradition and the Anaphora of St Basil) or combined the two (as in the case of the Sacramentary of Sarapion). Similarly, in their second half some prayed solely for the communicants (for instance, the Apostolic Tradition), others offered more general intercession (for instance, the Strasbourg Papyrus), while yet others once again, like the Anaphora of St Basil, combined both elements. The growing standardization of liturgical practices at this period ensured the spread of both the Sanctus and a narrative of institution to all known prayers by the end of the fourth century, as well as the inclusion of a developed invocation of the Holy Spirit in all those in the Christian East.
Perhaps not surprisingly, therefore, the quest for the earliest pattern of eucharistic praying reveals diversity more than commonalities, and the existence of prayers that for a considerable period of time were much less developed and explicit as to their eucharistic theology than were the beliefs of those who used them and preached about them. Thus they provide less than satisfactory models for modern liturgical compilers to imitate than do the more fully formed examples from later centuries.
______________
1 Among recent contributions see, for example, Stefan C. Reif, ‘The Second Temple Period, Qumran Research, and Rabbinic Liturgy: Some Contextual and Linguistic Comparisons’, in Esther G. Chazon (ed.), Liturgical Perspectives: Prayer and Poetry in Light of the Dead Sea Scrolls (Brill, Leiden 2003), pp. 133–49; and Richard S. Sarason, ‘Communal Prayer at Qumran and Among the Rabbis: Certainties and Uncertainties’, in ibid., pp. 151–72.
2 See Clemens Leonhard, ‘Blessings over Wine and Bread in Judaism and Christian Eucharistic Prayers: Two Independent Traditions’, in Albert Gerhards and Clemens Leonhard (eds), Jewish and Christian Liturgy and Worship: New Insights into its History and Interaction (Brill, Leiden 2007), pp. 309–26. Leonhard also makes the point that the rabbinic berakah releases food from its sacred character and makes it available to eat, while the Christian prayer on the contrary makes sacred what was formerly profane food.
3 Enrico Mazza, The Origins of the Eucharistic Prayer (The Liturgical Press, Collegeville 1995), pp. 12–61, 194–6.
4 For further details, see Bradshaw, Eucharistic Origins, pp. 25–30.
5 Alan Garrow, The Gospel of Matthew’s Dependence on the Didache (T & T Clark, London/New York 2004), p. 14.
6 Ibid., p. 25.
7 See Johnson, The Prayers of Sarapion of Thmuis, pp. 55, 69.
8 The word klasma appears to be used proleptically here, but it has been suggested that it is perhaps also influenced by knowledge of a version of the story of the Feeding of the Five Thousand in John 6.1–15, where there is reference in verse 12 to ‘fragments’ being ‘gathered’, which may be intended symbolically of the gathering of Christian disciples ‘that nothing may be lost’, see C. F. D. Moule, ‘A Note on Didache ix.4’, Journal of Theological Studies 6 (1955), pp. 240–3; Francis J. Moloney, The Gospel of John (The Liturgical Press, Collegeville 1998), pp. 198–200. An alternative possibility, which does not seem to have been considered, is that the breaking of the bread may actually have been intended to precede the saying of the prayer. Although this sequence is not otherwise recorded in either Jewish or early Christian sources, it would not be dissimilar from other variations in the order of ritual actions that we do encounter in different groups of both Christians (e.g., bread/cup or cup/bread) and rabbinic Jews (e.g., the debates in the Mishnah tractate Berakoth over the order that certain blessings were to be said). In addition, we should note that Didache 14.1 has the identical order: ‘break bread and give thanks’.
9 See further Bradshaw, Eucharistic Origins, pp. 116–21.
10 For a more detailed study of this material, see ibid., pp. 32–9.
11 For a critical edition, see William Macomber, ‘The Oldest Known Text of the Anaphora of the Apostles Addai and Mari’, Orientalia Christiana Periodica 31 (1966), pp. 335–71; for an ET of the text and bibliography of secondary literature, see Bryan D. Spinks (ed.), Addai and Mari – The Anaphora of the Apostles: A Text for Students, Grove Liturgical Study 24 (Grove Books, Nottingham 1980); and Anthony Gelston, The Eucharistic Prayer of Addai and Mari (Clarendon Press, Oxford 1992).
12 See, for example, Gerard Rouwhorst, ‘Jewish Liturgical Traditions in Early Syriac Christianity’, Vigiliae Christianae 51 (1997), pp. 72–93, here at pp. 79–80.
13 See, for example, Jacob Vellian, ‘The Anaphoral Structure of Addai and Mari compared to the Berakoth Preceding the Shema in the Synagogue Morning Service contained in Seder R. Amram Gaon’, Le Muséon 85 (1972), pp. 201–23; Bryan D. Spinks, ‘The Original Form of the Anaphora of the Apostles: A Suggestion in the Light of Maronite Sharar’, Ephemerides Liturgicae 91 (1977), pp. 146–61 = idem, Worship: Prayers from the East (Pastoral Press, Washington, DC 1993), pp. 21–36.
14 For further discussion, see Bradshaw, Eucharistic Origins, pp. 128–31.
15 Didascalia Apostolorum 2.57. For further discussion of both these passages, see Bradshaw, Eucharistic Origins, pp. 75, 104–5.
16 Ibid., pp. 121–3.
17 See for example, Justin Martyr, First Apology 66.1.
18 Apostolic Tradition 4 (Latin version); ET from Bradshaw, Johnson, and Phillips, The Apostolic Tradition: A Commentary, pp. 38–40.
19 Eduard von der Goltz, ‘Unbekannte Fragmente altchristlicher Gemeindeordnungen’, Sitzungsberichte der Preussischen Akademie der Wissenschaften (1906), pp. 141–57; see also idem, ‘Die Taufgebete Hippolyts und andere Taufgebete der alten Kirchen’, Zeitschrift für Kirchengeschichte 27 (1906), pp. 1–51.
20 Eduard Schwartz, Uber die pseudoapostolischen Kirchenordnungen (Trubner, Strasbourg 1910) = idem, Gesammelte Schriften 5 (de Gruyter, Berlin 1963), pp. 192–273; Richard H. Connolly, The So-called Egyptian Church Order and Derived Documents (Cambridge University Press, Cambridge 1916 = Kraus, Nendeln, Liechtenstein 1967).
21 Marcel Metzger, ‘Nouvelles perspectives pour la prétendue Tradition apostolique’, Ecclesia Orans 5 (1988), pp. 241–59; ‘Enquêtes autour de la prétendue Tradition apostolique’, Ecclesia Orans 9 (1992), pp. 7–36; ‘A propos des règlements ecclésiastiques et de la prétendue Tradition apostolique’, Revue des sciences religieuses 66 (1992), pp. 249–61.
22 Jean Magne, Tradition apostolique sur les charismes et Diataxeis des saints Apôtres (Paris 1975), pp. 76–7; Alexandre Faivre, ‘La documentation canonico-liturgique de l’Eglise ancienne’, Revue des sciences religieuses 54 (1980), pp. 204–19, 273–97, here at p. 286.
23 Christoph Markschies, ‘Wer schrieb die sogenannte Traditio Apostolica? Neue Beobachtungen und Hypothesen zu einer kaum lösbaren Frage aus der altkirchen Literaturgeschichte’, in Wolfram Kinzig, Christoph Markschies, and Markus Vinzent, Tauffragen und Bekenntnis, Arbeiten zur Kirchengeschichte 74 (de Gruyter, Berlin/New York 1999), pp. 8–43.
24 See Nautin (ed.), Homélies pascales I, pp. 34–6; Jean Michel Hanssens, La liturgie d’Hippolyte: Ses documents, son titulaire, ses origines et son charactère, Orientalia Christiana Analecta 155 (Pontificium Institutum Orientalium Studiorum, Rome 1959; 2nd edn 1965), pp. 84–5; Allen Brent, Hippolytus and the Roman Church in the Third Century: Communities in Tension before the Emergence of a Monarch-Bishop (Brill, Leiden 1995), pp. 192–3.
25 Hanssens, La liturgie d’Hippolyte, pp. 229–30, 247–9, 254–82; Brent, Hippolytus and the Roman Church in the Third Century, pp. 115–203.
26 Pierre Nautin, Hippolyte et Josipe. Contribution à l’histoire de la littérature chrétienne du IIIe siècle (Éditions du Cerf, Paris 1947) and ‘Notes sur le catalogue des oeuvres d’Hippolyte’, Recherches de science religieuse 34 (1947), pp. 99–107; Vincenzo Loi, ‘L’identità letteraria di Ippolito di Roma’, in Ricerche su Ippolito, Studia Ephemeridis Augustinianum 13 (Institutum Patristicum Augustinianum, Rome 1977), pp. 67–88; M. Simonetti, ‘A modo di conclusione: Una ipotesi di lavoro’, in ibid., pp. 151–6; Brent, Hippolytus and the Roman Church in the Third Century, pp. 204–366; Paul Bouhot, ‘L’auteur romain des Philosophumena et l’écrivain Hippolyte’, Ecclesia Orans 13 (1996), pp. 137–64.
27 Alistair Stewart-Sykes, Hippolytus: On the Apostolic Tradition (St Vladimir’s Seminary Press, Crestwood, NY 2001).
28 J. A. Cerrato, Hippolytus Between East and West: The Commentaries and the Provenance of the Corpus (Oxford University Press, Oxford 2002). See also his analysis of the current debate about the Apostolic Tradition, ‘The Association of the Name Hippolytus with a Church Order, now known as The Apostolic Tradition’, St Vladimir’s Theological Quarterly 48 (2004), pp. 179–94.
29 See Margherita Guarducci, ‘La statua di “Sant’Ippolito”’, in Ricerche su Ippolito, Studia Ephemeridis Angustinianum 13 (Institutum Patristicum Augustinianum, Rome 1977), pp. 17–30; eadem, ‘La “Statua di Sant’Ippolito” e la sua provenienza’, in Nuove ricerche su Ippolito, Studia Ephemeridis Augustinianum 30 (Institutum Patristicum Augustinianum, Rome 1989), pp. 61–74; Hanssens, La liturgie d’Hippolyte, pp. 217–31; Brent, Hippolytus and the Roman Church in the Third Century, pp. 3–114; and Markus Vinzent, ‘“Philobiblie” im frühen Christentum’, Das Altertum 45 (1999), pp. 116–17, who made the intriguing proposal that the figure was originally of an Amazon woman named Hippolyta!
30 See further Bradshaw, Johnson, and Phillips, The Apostolic Tradition: A Commentary, passim.
31 Ibid., pp. 141–5; Stewart-Sykes, Hippolytus, pp. 31, 140–2.
32 The institution narrative appears only in a rudimentary form in the Sacramentary of Sarapion, and seemingly was absent from the earliest version of the Anaphora of Addai and Mari, from the Strasbourg Papyrus, from the Jerusalem rite known to the author of the Mystagogical Catecheses and from that known to Theodore of Mopsuestia. The Strasbourg Papyrus also has no epiclesis and the Sacramentary of Sarapion an invocation of the Logos rather than of the Holy Spirit. See further Bradshaw, Eucharistic Origins, pp. 128–35.
33 For a more detailed working out of this argument, see Paul F. Bradshaw, ‘A Paschal Root to the Anaphora of the Apostolic Tradition? A Response to Enrico Mazza’, Studia Patristica 35 (2001), pp. 257–65.
34 This argument is developed by Matthieu Smyth, ‘L’anaphore de la prétendue “Tradition apostolique” et la prière eucharistique romaine’, Revue des sciences religieuses 81 (2007), pp. 213–28.