Mazza, Enrico. "Chapter Six: Primitive Anaphoras: Developments of the Eucharistic Liturgy." The Celebration of the Eucharist: The Origin of the Rite and the Development of Its Interpretation. Collegeville, MN: The Liturgical Press/Pueblo, 1999, 39-74.
In this chapter Mazza reviews several historic anaphoras to evaluate how the eucharistic liturgy developed over time. He begins with the Alexandrian liturgy which has a strophic "paleoanaphora" (Mazza 1999, 39).
The first of the three strophes gives thanks for light, which is not considered a given in the meal liturgies. Mazza finds the theme in "the Yotser, the first of the three blessings of the Shema" (Mazza 1999, 40). The content of the blessing is the same. Mazza asks how a traditional Jewish morning prayer would come to be associated with the eucharist which was built on a traditional ritual meal (Mazza 1999, 41). A possible key to understanding this is in the Alexandrian community of the Therapeutae, said by Philo to be Jewish but considered Christian by Eusebius (Mazza 1999, 42). In their piety, after an evening meal liturgy they would engage in hymnody before a vigil lasting through the night, until in the morning they had prayers of thanksgiving. Those prayers had a definite connection to the prayers of the Temple in Jerusalem (Mazza 1999, 43). Those prayers, then, would hold to the form of the sacrificial prayers for the morning. Mazza observes that as monasticism developed, the rituals of the Therapeutae were maintained (Mazza 1999, 44). The fourth century custom was that of a vigil through Saturday night, concluding with the eucharist on Sunday morning. This was at the time of morning prayer and would have included the prayers traditional for a morning sacrifice (Mazza 1999, 45).
The second strophe in the paleoanaphora from Alexandria is related to the sacrificial element of the morning prayer, including a reference to Malachi 1:11 (Mazza 1999, 46). The eucharist may be seen typologically as a sacrifice, which is also related to praise expressed to God. This is consistent with the relationship of Malachi 1:11 and Hebrews 13:15 (Mazza 1999, 47).
The third strophe of the anaphora prays for the Church along with others (Mazza 1999, 47). This is also related to the blessing before the Shema, placing it within the realm of morning prayers. Mazza thus concludes that the Alexandrian liturgy was related to elements of morning sacrificial prayers due to the influence of an overnight vigil which transferred elements of an evening ritual meal to the next morning (Mazza 1999, 49).
Another significant stream of liturgy is that of Antioch. This influence is apparent in Hippolytus, Basil, and James (Mazza 1999, 50). In the work of Hippolytus, Mazza finds a strong Paschal theme and a description of salvation including the death of Christ (Mazza 1999, 51). The account of the institution is brought into the preface as the key point and reason for thanksgiving (Mazza 1999, 52). The anaphora in Hippolytus and that of Alexandria have enough structural differences that Mazza considers them "irreconcilable" (Mazza 1999, 52). However, he does discuss the possibility that the institution narrative would have been inserted into a form of the Alexandrian liturgy, which could have eventually diverged from the Asiatic form as exemplified by Hippolytus (Mazza 1999, 53). Mazza suggests that the institution narrative may have been inserted in Alexandria and then influenced the Asiatic liturgy (Mazza 1999, 53-54). Following the institution, there is an anamnesis and offering, with sacrificial language which may well parallel the second strophe in Alexandria (Mazza 1999, 54). The liturgy moves on to a number of petitions, an epiclesis (Mazza 1999, 56-57. Within this, Mazza notes two distinct prayers about the Holy Spirit. The first treats him as central, while the second does not. This suggests to Mazza that the second was derived from some other source (Mazza 1999, 59). Mazza thus concludes that the forms which seemed irreconcilable may actually be derived from the same source and may also have influenced each other (Mazza 1999, 60).
Mazza moves on to consider the Roman Canon, first attested by Ambrose of Milan, in De sacramentis (Mazza 1999, 62). The structure seems unlike any of the other anaphoras. However, Mazza notes that there are underlying structural similarities emerging in recent research (Mazza 1999, 63). When we compare roots and possible sources of the prayers, we find two prefaces which are similar to those of Alexandria (Mazza 1999, 64). Their structures are identical. Mazza finds, then, that the skeletal structure has simply been clothed with additional elements (Mazza 1999, 65).
Mazza finally reviews Syrian anaphoras, noting that these differ by region (Mazza 1999, 66ff). Mazza does find elements which are identical across multiple different anaphoras, clearly indicating common origins (Mazza 1999, 67). In sermons and commentaries on these anaphoras, the Sanctus has apparently been used as a dividing point, suggesting later insertion of the institution narrative at that point (Mazza 1999, 68). The institution narrative is apparently added after the Sanctus and a post-sanctus had been developed. Awareness of this structural aspect allows evaluation of all the Syrian anaphoras, which can be shown to have a similar structure (Mazza 1999, 71).