The formation of the eucharistic prayer in the Apostolic Tradition

For most of the twentieth century, an anonymous ancient church order that had been discovered in the nineteenth century was judged by the great majority of scholars to be the lost Apostolic Tradition of Hippolytus and so was thought to reveal the liturgical practices of the Church at Rome in the early third century. This not only gave historians what appeared to be a solid piece of evidence for a period when so much seemed to be shifting sands, but also offered what seemed to be a firm foundation on which to build for those effecting liturgical revisions in many Christian churches in the second half of the twentieth century. There is scarcely a modern eucharistic prayer that does not owe some debt to this ancient text.

In recent decades, however, there has been a growing consensus among scholars that it is not the missing work of Hippolytus at all – although for convenience the name Apostolic Tradition continues to be used for it – but a piece of ‘living literature’ gradually expanded from its roots in the second century to a much more elaborate form in the fourth century, or, in the case of some translations, even later. Furthermore, most scholars, including even Alistair Stewart, who otherwise follows a theory of his own about the origins of this church order, ¹ are agreed that the earliest eucharistic celebration in the text was the core of the evening meal described in Chapters 25–9, and that the eucharistic prayer in Chapter 4 was a later addition to it. No original Greek text exists for this prayer, and hence the ancient Latin translation is almost the only witness, the Aksumite Ethiopic, ² the Sahidic and the Arabic texts omitting the prayer entirely, and the later Ethiopic version – and the use of it in the Apostolic Constitutions and Testamentum Domini – being of very little assistance.

It is also commonly accepted that in broad terms this prayer corresponds to what is known as the West Syrian pattern. ³ However, there are clear signs that it too had evolved over a period of time. On the one hand, certain expressions in the prayer are otherwise only found in works originally composed well before the end of the second century. Chief among these are the apparent use of the Greek word pais (child, servant) as a title for Christ, both at the beginning of the prayer and in the concluding doxology, and ‘messenger of your will’ as a descriptor for him. The former is also found in the Didache, 1 Clement, the Epistle of Barnabas and the Martyrdom of Polycarp; ⁴ the latter, an allusion to Isaiah 9.5 in the Septuagint, appears four times in Justin Martyr and once in Irenaeus. ⁵ This strongly suggests that the roots of the prayer are very old, even if it were not added to the Apostolic Tradition until perhaps some time in the third century.

On the other hand, there is evidence that other parts of the prayer date from no earlier than the fourth century. Again, there is widespread – though not universal – agreement among scholars that several of the elements in early eucharistic prayers in general that became characteristic of them did not become standard until this time. These are the Sanctus (with its introductory pre-Sanctus and sometimes the Benedictus), the institution narrative and the epiclesis. There is no trace of any of these in the prayer contained in Strasbourg Papyrus Gr. 254, which many scholars judge to date from the third century, although some argue that it is incomplete and would originally have included some or all of them in its missing parts. ⁶ The Anaphora of Addai and Mari, which is also commonly thought to have a third-century origin, completely lacks an institution narrative. ⁷ And the Apostolic Tradition itself, of course, does not have a Sanctus.

As for the emergence of the epiclesis, in what has become a classic work on the subject, ⁸ Sebastian Brock has demonstrated its gradual evolution from forms that addressed Christ directly with an imperative ‘Come!’, through invocations to Christ to ‘Let your Spirit come’, and later to God being asked that the ‘Spirit may come’, and finally God being petitioned to ‘send the Holy Spirit’. This final form is first found in the eucharistic prayer in the mid-fourth-century Barcelona Papyrus (‘we ask and beseech you to send on them your Holy and comforter Spirit from heaven’). ⁹ As the particular form found in the Apostolic Tradition conforms closely to that (‘[we ask that] you would send your Holy Spirit on the oblation of the Holy Church’), there can be little doubt that it must be among the more recent expansions of the text, ¹⁰ although it is probably somewhat earlier than the Barcelona Papyrus because of its simpler form. Its eventual inclusion seems to have somehow affected the wording of what now follows it.

The institution narrative in particular seems to have been slow to gain an established place in eucharistic prayers. Many scholars are of the opinion that the eucharistic prayer of Jerusalem that was the subject of the commentary in the Mystagogical Catecheses still lacked such a narrative in the last quarter of the fourth century, whether the author was Cyril of Jerusalem or his successor John. ¹¹ On the other hand, the narrative was present in the eucharistic prayer in the Barcelona Papyrus and also in the Prayers of Sarapion; in this last case, however, it is not in the form of a continuous narrative, but divided into two parts, thus suggesting that its use as a single unit was as yet unknown to the compiler. Interestingly, this prayer has an epiclesis of the Logos rather than the Spirit. ¹²

In addition, it is noteworthy that several of the early prayers that do include these features show apparent signs of their having been inserted into a pre-existent structure rather than being integral to it, particularly through some awkward transitions. ¹³ Thus, we may reasonably conclude that, in the case of the Apostolic Tradition too, they were not part of the earliest stratum. We should also observe that we simply do not know what stage of development this prayer had reached in the fourth century. In the early fourth-century Canons of Hippolytus, the equivalent prayer is completely absent. The introductory dialogue is present, but not the prayer itself. This could mean that even at this late date the Apostolic Tradition had not acquired its prayer – or, more likely, that the form of prayer that was in that church order did not resemble anything at all familiar to the compiler of the Canons of Hippolytus and for that reason was omitted. Thus, although in other respects the Canons may be regarded as the oldest witness to the Apostolic Tradition, this is of no assistance here.

Therefore, the earliest date when we can begin to discern the final form of the prayer is not before the appearance of the Apostolic Constitutions around 375, as it incorporates some elements of the prayer in its own much lengthier creation. The Testamentum Domini contains more of the prayer, but its date is uncertain, and it is probably from some time in the fifth century. ¹⁴ The incomplete Latin translation of the Apostolic Tradition originated about the same time, and all the other versions of the text are even later than that. Hence, it is reasonable to presume that, prior to the addition of the institution narrative and epiclesis some time in the fourth century, the prayer existed in a simpler form. What else, besides these two elements, might also have been added later rather than earlier?

The phrase ‘your inseparable word’ recalls ‘inseparable from the Father’, used by Justin Martyr (Dialogue with Trypho 128) and by (Pseudo-)Hippolytus (Contra Noetum 18). However, Markus Vinzent has observed that ‘inseparable’ was ‘not part of the debate against Noetus in the second century, nor in the early Arian controversy, but becomes a hallmark of the Marcellus of Ancyra, Eusebius of Caesarea, and Apollinarius of Laodicea discussion in the post-Nicene period’. Even the seemingly innocuous quotation from Hebrews 1.2, ‘in these last times’, a little earlier in the prayer ‘was introduced to the fourth-century debate by Asterius of Cappadocia, debated by Marcellus and Theophronius of Tyana, and repeated in three of the symbols associated with the 341 ce synod of Antioch’. ¹⁵ This suggests a later date and a more polemical reason for the addition of these phrases here.

The following clauses (‘through whom you made all things and it was well pleasing to you’), though probably not as late as the fourth century, certainly represent a more advanced Christology than that displayed by the oldest stratum of material in the Apostolic Tradition, which makes no reference to a pre-existent Christ. A further clause (‘and who conceived in the womb was incarnate and manifested as your Son, born from the Holy Spirit and the virgin’) is more clearly a product of the fourth century. References to birth from both the Holy Spirit and the virgin are extremely rare before that time, ¹⁶ reference to birth from the virgin alone being more usual, as in the preceding clause (‘you sent from heaven into the virgin’s womb’). It is also to be noted that, in this part of the prayer, Jesus is described as God’s Son rather than as servant/child, another sign of a later date.

The next part of the prayer (‘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’) could very well have been part of its earliest stratum, as there is nothing in it inconsistent with a second-century date. Indeed, the concept of the stretching out of hands to God’s people, expressed in Isaiah 65.2 (‘I spread out my hands all day long to a rebellious people’) and quoted in Romans 10.21, was widely interpreted as a prophecy of the crucifixion within the early Christian tradition. ¹⁷ Furthermore, Enrico Mazza noted that the same verbal link between suffering and release from suffering was made in several of the second-century paschal homilies. ¹⁸

We have already claimed that the institution narrative is a fourth-century insertion into an older prayer, but that leaves a difficult question of the status of what immediately precedes the narrative here: ‘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 illuminate the righteous, and fix a limit and manifest the resurrection’. Grammatically, this is dependent on the institution narrative, and so one might conclude that the two parts were added as a single unit. On the other hand, it is not impossible that the grammatical construction underwent a change when the narrative was inserted, and that this preceding part had already existed in the prayer in some form.

Again, the expressions used in this section are characteristic of the second century. Not only is the concept of destroying death already present in the New Testament (see 1 Corinthians 15.26; 2 Timothy 1.10), but it recurs in other early literature: ‘he endured that he might destroy death and reveal the resurrection of the dead’ (Barnabas 5.6); and, in a longer form, in Irenaeus, as ‘Our Lord by his passion destroyed death, and dispersed error, and put an end to corruption, and destroyed ignorance, while he manifested life and revealed truth, and bestowed the gift of incorruption’ (Adversus Haereses 2.20.3); and, again more briefly, in his Demonstration of Apostolic Preaching, with ‘he destroyed death, releasing those bonds in which we had been chained, and manifested the resurrection’ (Epideixis 38). Finally, Melito of Sardis in his Peri Pascha offers a more extensive witness to this vocabulary, including an early allusion to the descent into hell:

The Lord clothed himself with humanity, and with suffering on behalf of the suffering one, and bound on behalf of the one constrained, and judged on behalf of the one convicted, and buried on behalf of the one entombed, rose from the dead and cried out aloud: … I am he who destroys death, and triumphs over the enemy, and crushes Hades, and binds the strong man, and bears off humanity to the heavenly heights. ¹⁹

Although the use of the precise phrase ‘illuminate the righteous’ with reference to those in the underworld is otherwise not known before the Gospel of Nicodemus (fourth century?) and the meaning of ‘fix a limit’ continues to puzzle scholars, ²⁰ there are enough second-century linguistic parallels with the material in this section of the prayer to support its probable presence in the early stratum, provided that the grammatical construction were once different. This raises a further possibility. If the reference to manifesting the resurrection previously ended the section, it could have immediately preceded the anamnetic phrase, ‘Remembering therefore his death and resurrection,’ prior to the institution narrative being added. That would have formed an ideal link between the two halves of the prayer. Scholars have usually assumed that anamneses like this have always been connected to a preceding institution narrative. But it would function here equally well, or even better, without that, and its presence after the narrative in other early eucharistic prayers could simply have been the result of them following this model.

Apart from the later insertion of the epiclesis and the expansion of the final doxology with a more Trinitarian but ungrammatical phrase – ‘Father and Son with the Holy Spirit’ – in spite of the use of the word pais just before it, the rest of the prayer raises no alarm bells about its primitive character and may be accepted as such. Even the expression ‘we offer to you the bread and cup’ has parallels in Justin Martyr and Irenaeus. ²¹ On the basis of its presence in the corresponding part of the prayers in the Apostolic Constitutions, the Testamentum Domini and the later Ethiopic translation, several scholars have argued that the Greek verb behind the Latin servare, ‘minister’, had been hierateuein, ‘to exercise the priesthood’, and even that it referred exclusively to the episcopal ministry just conferred in ordination. ²² However, this seems much more likely to have been a change made by those later texts when the sacerdotal aspect of the episcopate had emerged, and the original had meant by ‘us’ the whole community on whose behalf the prayer was being said, especially as the Greek verb leitourgein occurs in the Septuagint of Deuteronomy (10.8 and 18.3, 7) combined with ‘stand before the Lord’ in a liturgical context. ²³ The absence of a direct object for the participle ‘gathering’ has also given rise to a number of proposed emendations, ²⁴ but the parallel petition in Didache 10.5, that God might ‘gather it [the Church] from the four winds’, suggests that the sense was similar here.

Thus, shorn of its later expansions, this eucharistic prayer was exclusively concerned with the mission of Jesus, and not with such themes as God’s work of creation nor God’s dealings with the Hebrew people. Nor are there any wider subjects of intercession but only petition for the fruits of communion. It may well be that other very early eucharistic prayers were quite different from this, but – even allowing for liturgical language usually to be somewhat archaic and continuing to employ some vocabulary after it had ceased to be used in other forms of discourse – here we have one that gives us a unique insight into not merely a third-century form, but one apparently as old as the late second century – for which we have no other example – long before it was incorporated into the Apostolic Tradition.

Proposed second-century text

We render thanks to you, God, through your beloved servant Jesus Christ, whom in the last times you sent to us as saviour and redeemer and messenger of your will, who is your inseparable word, through whom you made all things and it was well pleasing to you, (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; 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; 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 illuminate the righteous, and fix a limit and manifest the resurrection, 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. When you do this, you do my remembrance.’

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. And we ask that you would send your Holy Spirit on the oblation of the Holy Church, gathering (us) 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, that we may praise and glorify you through your servant Jesus Christ, through whom (be) glory and honour to you, Father and Son with the Holy Spirit, in your Holy Church, both now and to the ages of ages.

Amen.