Breaking Bread: The Emergence of Eucharist and Agape in Early Christian Communities

By Alistair C. Stewart

What’s the difference between eucharist and agape? And how did each come to be?

The liturgies of early Christians are often obscure and variegated in the historical record. This is especially true of the eucharist, where the basic practice of communal eating is difficult to disentangle from other contemporary meals, whether Greco-Roman or Jewish practices—or the ill-defined agape meal.

In Breaking Bread, Alistair C. Stewart cuts through scholarly confusion about early Christian eating. Stewart pinpoints the split in agape and eucharist to the shift in celebrating the eucharist on Sunday morning, leading to the inception of agape as an evening meal. The former sought divine union, the latter, communal harmony. In the final chapter he explores a breadth of Syriac, Greek, and Latin primary sources on a variety of local eucharistic traditions, tracing their development into the familiar prayers and distribution of token amounts of bread and wine, which emerged in the third century.

Nuanced and well-researched, Breaking Bread clarifies the development of the blessed sacrament and its lesser-known counterpart. Theologians and historians of early Christianity will find Stewart’s work foundational in approaching a topic of enduring scholarly interest but elusive consensus.

• Language: English
• Publisher: Eerdmans
• Release date: Jun 27, 2023
• ISBN: 9781467466349

Alistair C. Stewart

Alistair C. Stewart is senior lecturer in biblical studies at Codrington College, Barbados. He has ministered in parishes in Barbados and England for thirty years and is recognized as a leading scholar in early Christian liturgy.

Book preview

INTRODUCTION

Four chapters follow this brief introduction; although they refer across to each other at various times and together construct a narrative, they originated as independent essays, and so the chapters still may to an extent be read independently as well as consecutively. This introduction sets out their overarching agenda and the assumptions on which they based.

Their subject is early Christian meals, and in particular the meal that came to be known as the eucharist. Their scope is limited to the relationship between this meal, which is seen to emerge into clear historical focus in the fourth century, and the variety of meals celebrated in Christian circles in the preceding centuries.

They are written in conversation with recent scholarship. The starting point of the conversation is expressed well by McGowan:

A consensus about eucharistic origins still found in much scholarship goes something like this: the earliest Christian communities celebrated their sacramental meal in direct imitation of the Last Supper of Jesus, and thus with token use of bread and wine, a universal order or structure, and recitation of the institution narrative as the central prayer text. Sacrament and communal banquet were quickly separated—by the later first century—into eucharist, a morning sacramental ritual, and Agape, a prosaic communal supper.

For all its apparent obviousness and its familiarity from use in much scholarly writing about ancient Christianity, this account must be rejected. Each of the elements just stated—causes, chronology, uniformity—is inaccurate at best.¹

McGowan is cited here as a particularly literate and well-informed spokesperson for an emerging scholarly consensus to the effect that the eucharist, unrelated to the Last Supper, is simply a Christian version of a common meal tradition in the Greco-Roman world. He is not its originator, though he is an able advocate. The origins of this consensus lie in what Taussig calls a confluence of research on Greco-Roman meals by two New Testament scholars working independently on two different continents.² Here he refers to the works of Klinghardt and Smith.³ These scholars’ works have led to an explosion of publications and have become the starting point for an SBL seminar.⁴

The new and emerging scholarly consensus that McGowan represents is that the eucharist originated in an ordinary meal, that groups of Christian believers would naturally gather to eat together, in the same way that other extended households or associations might. Thus this was not a commemoration of the Last Supper but a human practice that is both natural and culturally expressed, given particular form by the cultural practice and pattern of the Greco-Roman dinner, with its accompanying symposium. Thus although it might be suggested that the emerging eucharist derived from the meal practice of Jesus, or of that of the group of apostles after Easter, because these, too, are natural human actions expressed in the same cultural context, it is a line hard to draw. Even if there is a reference to these practices in eucharistic prayers, such as the reference in Did. 9 to the feeding of five thousand and the gathering of fragments, seen as a proleptic realization of the messianic banquet,⁵ there is no way of knowing whether the practice has derived directly from that of Jesus or whether this is a secondary application growing out of normal meal practice.

At the outset, we must state an extent of agreement with the new consensus expressed by McGowan, and also agreement that the consensus that he describes is, as he states, inaccurate at best. Indeed, the agreement as to this inaccuracy means that there will be little discussion of the representatives of this consensus.

As evidence for the eucharist originating in dining practice, we may observe the practice of the didachistic community (Did. 10.1 refers to being filled after eating of a meal described at 9.1 as the eucharist) and that of the Corinthians, who kept a supper (so apparent from Paul’s comments in 1 Cor 11). We may also observe the community whose document was adopted and adapted by the Hippolytean community and incorporated into Traditio apostolica. These instances will be examined in further detail in the chapters below.

Further, it is taken as read that liturgical development in the early centuries is never uniform, and so a uniform progression is hardly to be expected.

Beyond this, that the sacramental meal is not an imitation or a repetition of the Last Supper, the realization of which is one of the fruits of recent scholarship, is accepted without quibble. It is manifest that the words of institution employed by Jesus were not part of early eucharistic prayer; in broad agreement with the emerging consensus, I have suggested that they first appear in the early third century and do not come into prominence until the fourth century,⁶ though others date their appearance even later.⁷ Thus Bradshaw begins an essay entitled Did Jesus Institute the Eucharist at the Last Supper? with this statement: The answer to the question posed in my title might seem obvious; ‘Of course he did.’⁸ However, for most of those who have made any serious study of liturgy in the past twenty or so years, the answer is equally obvious that he did not.⁹ However, this is not necessarily the significant question. Whatever the intention of Jesus, the question might be this: Did early Christians believe that Jesus had instituted the eucharist at the Last Supper? The catechetical tradition suggests that they did, but it leaves no trace in our liturgies before the third century and does not become significant for a further century at least.

It is also manifest that the use of bread and wine is not universal, although it is found in some communities.

Thus the search for eucharistic origins has to begin in a different place, namely, in the eating practices of early Christian circles in the context of antiquity, both in the Jewish and in the wider Hellenistic world, which is precisely the starting point of the new consensus. And thus I concur that the earliest Christian meals were in significant respects comparable to meals celebrated by other social groups, such as associations.

At other points, agreement is qualified. The concept of token amounts is difficult and will be addressed at several points below, but we must certainly accept that the use of token amounts of food in early Christian eucharistic eating should never be assumed. A particular point of disagreement is, however, the understanding of the relationship between the eucharistic meal and other meals. It will be argued that the sacramental (the term is used loosely here and throughout) element in early Christian meals might have been separated from an accompanying meal, and that although it is certainly not the case that such separation had universally occurred by the end of the first century, it is suggested that this had occurred by then in one context at least, that it occurred throughout the second century, and that it would eventually become universal. Finally I intend to argue that the agape was far from prosaic and that (contra the new consensus, though also contra the understanding that the new consensus seeks to overthrow) it existed from a relatively early stage, though I also suggest that to refer to the meal from which the eucharist had separated as an agape is inaccurate.

Whereas the new consensus successfully criticizes the age-old understanding and chronology of separation, it neither gives an account of the eventual emergence of the eucharist as a distinct and sacred rite, taking place in the morning and consisting of the distribution of bread and wine in token amounts, nor does it give an account of the development of the agape; the existence of the agape in the fourth century is historically demonstrable, and so the institution must have had a beginning. Beyond that, because the proponents of the new consensus concentrate fundamentally on the earliest period of the development of Christian meals, they unwittingly privilege these early centuries, with the implied message that this period of development is somehow normative for Christian practice. This charge would probably be vigorously denied, but it is implicit in the research effort and, as will be shown in the third chapter, and in the conclusion below, sometimes explicitly expressed.

Moreover, as a result of rejecting the origin of the eucharist in the remembrance of the Last Supper, there is something of a confusion about what therefore constituted the eucharist in the early centuries. In the past, early Christian meals that did not fit the narrow criterion of using bread and wine only and commemorating the Last Supper were considered noneucharistic and simply classified as agape, a term that comes therefore to mean everything else, or, as McGowan puts it, a receptacle for scholarly scraps.¹⁰ Clearly this does not do justice to the evidence and results in a very poorly defined agape alongside an overdefined eucharist. However, in the new consensus, there is no agreed definition of a eucharistic meal, with the result that almost any Christian meal in the first centuries is seen in this light, and the agape disappears altogether as an independent institution. One may say that the failure to define the agape has simply become the failure to define the eucharist. This in turn means that the failure of the new consensus to explain the emergence of these two institutions is all the more acute. One may also note that for all there is a common meal shape throughout antiquity, a fundamental presupposition of the new consensus, this fails to observe the multiple contexts in which this common meal shape might be found, in temples, associations, schools, as well as homes (all of which might provide models for the institutionalization of Christian communities). Did the context not alter the nature of the meal in any way? This question is left open but may be rephrased, and to an extent answered in the following essays, in this way: Did the specifically Christian contexts of the meal affect the nature of the meal in any way? Although the common meal tradition is found in Jewish circles both before and after the turn of the Common Era, this is an adapted tradition.¹¹ The suggestion fundamental to these essays is that Christians were equally capable of adaptation. As Schröter notes, Justin, Tertullian, and Paul all attempt to put distance between the meals they describe (or, in Paul’s case, regulate) and other associational or cultic meals;¹² that they should need to make the distinction points to the fundamental similarity of the institutions, but that the distinction is made indicates that the meals were indeed in certain ways distinct.¹³ This is not to deny the existence of a common meal tradition, or the value of Klinghardt’s work in pointing out its existence, but to suggest that the emphasis of the new paradigm on the common meal tradition has blinded us to the specificities within the meals of distinct groups within the first centuries of the Common Era,¹⁴ and that it is in the specific that interest lies.

Bradshaw distinguishes between lumping and splitting as approaches to liturgical history.¹⁵ These essays are an exercise in splitting the lump of the common meal tradition. Such splitting is necessary; thus Leonhard denies that the Jewish Pesach might have contributed anything to the Christian eucharist, on the grounds that they were both simply manifestations of that tradition,¹⁶ but when Ep. Apos. 15 bears witness to an extraordinarily late supper at Pascha, this suggests that it is both reacting to and deriving from contemporary Jewish paschal practice by extending the meal into the night, contrary to what the common meal tradition would lead us to anticipate. Such a conclusion cannot be drawn if the agape and commemoration of the Epistle to the Apostles is simply another meal.

A corollary of the split that needs to be made between the meals of Christians, Jews, Epicureans, crocodile mummifiers, or any other group in the ancient world is the observation that these groups might make distinctions between the different meals that they each celebrated; to understand early Christian meal practice, this, too, needs to be explored.

Although the main conversation partners in this essay have written in the past thirty or so years, the discussion about the relationship between eucharistic meals and other early Christian meals (termed, historically, the relationship between agape and eucharist) has a long history. Already in the eighteenth century, Bingham argued against Aubespine that the agape had originally accompanied the celebration of the eucharist and that it had become separated, whereas Aubespine had held that the two were distinct institutions.¹⁷ Aubespine was something of an exception, however, and so Bingham is in agreement with Suicer and subsequently with an array of authors throughout the nineteenth century. As Batiffol puts it:

Reproduction de la dernière cène, l’agape aurait été le rite primitif de l’eucharistie. Puis, à un moment impossible à déterminer, l’agape aurait été dissociée de la fraction du pain: la fraction du pain dûment stylisée, serait devenue la messe, et l’agape aurait disparu diversement. De Bingham à Renan, de Suicer à Zahn, l’affirmation était unanime.¹⁸

However, the evidence was not strong; the combination was entirely predicated on the evidence of Paul in 1 Cor 11, where the eucharist (as the meal that Paul discusses was taken to be) appears to take place in the context of a common meal. There was evidence of agape celebrations in a later period, and evidence, moreover, of a clearly distinct eucharist, and so the assumption was made that the two had been combined (as they appeared to be in Paul’s writing) and had separated at some point, leaving the eucharist and the agape as freestanding and independent institutions, the agapic charity supper, which is found subsequently, thus being, on this understanding, a remnant of the time in which the eucharist was part of a meal. In this picture, the evidence of Tertullian loomed large, as he seemed to describe an agape in Apol. 39. But, as noted, the evidential basis was limited, in particular through the uncertainty regarding the interpretation of the information given to Pliny: that the morning was given to a sacramentum and the evening to a supper (which had been abandoned) was taken as evidence of a morning eucharist and an evening agape, indicating that a separation had already taken place,¹⁹ but the sacramentum was also taken to refer to baptism, thus reducing the evidential basis still further. Thus Lightfoot, while sharing in this doubt about the interpretation of Pliny’s words, represents the recently overthrown consensus well:

In S. Paul’s time … the eucharist was plainly part of the agape. The Christian festival, both in the hour of the day and in the arrangement of the meal, was substantially a reproduction of Christ’s last night with His Apostles. Hence it was called the Lord’s Supper—a name originally applied to the combined eucharist and agape, but afterwards applied to the former when the latter had been separated or even abolished. On the other hand in Justin Martyr’s time … the two were no less plainly separate…. When the eucharist was cut adrift from the agape, the agape might be discontinued, as circumstances dictated. As a matter of fact, we learn from Pliny’s language that it was suspended in Bithynia in the age of Trajan.²⁰

The latter years of the nineteenth century saw the rediscovery of the church orders, and in particular of the Didache, which in turn launched a burst of scholarly activity relating to the issue. Thus Batiffol suggests that the eucharist and agape were never joined and that the evidence of 1 Corinthians simply shows the abuse of the sacramental supper by the Corinthians; as part of this, he has to argue that the key text in Tertullian, Apol. 39, refers to the eucharist, an argument now resurrected by McGowan and Georges.²¹ Batiffol thus suggests that the agape only emerges, as an entirely independent phenomenon, in a much later period.²² This brought about a reaction in Roman Catholic circles in support of the older consensus.²³ Unfortunately the debate was cut short by the antimodernist backlash (Batiffol was condemned as a modernist in 1907).²⁴ On the Protestant side, however, there were also significant attempts to move beyond the consensus.²⁵ Fundamental to this discussion, however, was the assumption that the eucharistic practice of these early Christians was founded in the commemoration, in some sense, of the Last Supper, and the concomitant assumption that meals that did not have such a reference point were not eucharistic. Thus Lietzmann’s monumental work in which two roots were ascribed to the classical eucharist was an outgrowth of this debate and largely shared its assumptions.²⁶

In spite of Batiffol’s denials that an agape had ever existed in the first centuries, and Jülicher’s and Spitta’s effective denials of eucharistic content to early Christian meals, the common view that somehow the eucharist and the agape had become separated nonetheless remained the majority opinion, particularly in Anglophone circles through the enduring stature of Dix.²⁷ This is the consensus to which McGowan refers.

Thus in what follows, I suggest that there might be a variety of eucharistic meals, dependent on the context, but also that all Christian meals were not necessarily eucharistic. Meals are seen as eucharistic or not, however, not on the basis of any reference to the Last Supper tradition, and absolutely not to the presence of any words of institution. Nor is the distinction made on the basis of the quantity of food consumed, nor on the basis of what foods are consumed, nor indeed of the order in which the food and drink is consumed, all of which have in the past served to create a distinction. Rather, I go back to a suggestion I made previously that the terms eucharist and agape are generic;²⁸ hence I speak consistently of eucharistic meals rather than the eucharist, until we reach the fourth century. We may define eucharistic meals as meals in which some communion with a divine or spiritual being is sought. Agapic meals are defined as meals in which the love of the Christian community is manifested, to either those present or those absent, or both. The validity of these generic definitions will be demonstrated in the largely cumulative argument in the chapters below. As generic terms, they are not particularly informative. What is interesting are the actual practices of these meals expressed in the particular texts and on the particular occasions on which they are held, namely, the species of meal within the genera.

It is not always possible to know these, and on this basis not always possible fully to understand the significance given to every account of a meal. As Bell remarks, ritualization is a particularly ‘mute’ form of activity. It is designed to do what it does without bringing what it is doing across the threshold of discourse or systematic thinking.²⁹ Here she has reference to her characterization of ritual as the manner in which participants recognize their ritual actions as distinct from those same acts performed in a nonritual context. Interestingly her example of this is the difference between the consumption of bread at the eucharist (in a modern context) and outside.³⁰ It is the social differences between these acts that mark off the ritual act. We may similarly, and entirely within a ritual framework, suggest that ringing a bell at solemn mass and ringing a bell to mark the beginning and end of a round of boxing, while both ritual uses of a bell, are distinct simply because of the contexts in which they occur and therefore communicate different things to performers and spectators. Bell’s preference for the term ritualization over ritual recognizes that the construction of acts as rituals is the work of social agents rather than something inherent in the action itself.³¹ Bearing in mind the varying contexts for early Christian meals, we are asking how Christians, as social agents, understood their actions in eating and drinking. Thus to take an obvious example, even before starting the work proper, we may observe that Paul and his Corinthian audience had entirely distinct understandings of the rituals of eating the Lord’s Supper, Paul seeing it as a means of unity and the Corinthian elite as a means of marking status.³²

The first chapter is concerned with the transfer of eucharistic worship to the morning. Whereas this may seem a side issue, this is seen as the fundamental change that brought about the other changes undergone by eucharistic meals before the fourth century, out of which the classical eucharist emerged. The second chapter then discusses the agape and suggests that it emerged as a result of the transfer of the eucharist to the morning, with the consequent loss (or reduction) of its accompanying Sättigungsmahl (not to be called the agape). The third chapter then justifies seeing this Sättigungsmahl as an accompaniment to the eucharistic action, rather than being, as the new consensus might suggest, coterminous with it. Finally the fourth chapter lays the foundation for a possible history of the eucharist seen as emerging from a variety of eucharistic rites found in the earlier centuries.

These chapters are thus offered not as an attempt to overthrow the new and emerging consensus concerning eucharistic origins, even less an attempt to return to the older one, but an attempt to move on from what has already been achieved to answer those questions that the new consensus leaves unanswered, and in doing so to offer some corrections and clarifications. Their modest aim is to inform our understanding of the past. Although at points what transpires is that we know less about the past than once we thought we did, what can be discerned is a development from a variety of meal occasions, with broad generic coherences, to two species. When, in the fourth century, the eucharist and the agape emerge into history as recognizable and distinct rites, they have come about not suddenly but as the result of development from the earliest Christian ritual meals.

1. Andrew McGowan, Rethinking Eucharistic Origins, Pacifica 23 (2010): 173–91 at 173.

2. Hal Taussig, Introduction, in Meals in the Early Christian World: Social Formation, Experimentation, and Conflict at the Table, ed. Dennis E. Smith and Hal Taussig (New York: Palgrave Macmillan, 2012), 1.

3. Matthias Klinghardt, Gemeinschaftsmahl und Mahlgemeinschaft: Soziologie und Liturgie frühchristlicher Mahlfeiern (Tübingen: Franke, 1996); Dennis E. Smith, From Symposium to Eucharist: The Banquet in the Early Christian World (Minneapolis: Fortress, 2003).

4. Thus (principally) Smith and Taussig, Meals; Matthias Klinghardt and Hal Taussig, ed., Mahl und religiöse Identität im frühen Christentum (Tübingen: Franke, 2012); David Hellholm and Dieter Sänger, ed., The Eucharist—Its Origins and Contexts, 3 vols., WUNT 376 (Tübingen: Mohr Siebeck, 2017); Soham Al-Suadi and Peter-Ben Smit, T&T Clark Handbook to Early Christian Meals in the Greco-Roman World (London: T&T Clark, 2019).

5. So A. C. Stewart, The Fragment on the Mountain: A Note on Didache 9.4, Neot 49 (2015): 175–88.

6. Alistair Stewart-Sykes, Mimesis, Typology and the Institution Narrative: Some Observations on Traditio Apostolica 4 and Its Afterlife, in Wilderness: Essays in Honour of Frances Young, ed. R. S. Sugirtharaja (London: T&T Clark, 2005), 106–19. G. A. M. Rouwhorst, Frühchristliche Eucharistiefeiern: Die Entwicklung östlicher und westlicher Traditionsstränge, in Hellholm and Sänger, Eucharist, 2:771–86, gives a brief account of developments up to the fourth century without mentioning the Last Supper narrative except in passing.

7. E.g., Maxwell E. Johnson, Martyrs and the Mass: The Interpolation of the Narrative of Institution into the Anaphora, Worship 87 (2013): 2–22, suggesting the fourth-century Barcelona anaphora as the first appearance of this narrative.

8. Paul F. Bradshaw, Did Jesus Institute the Eucharist at the Last Supper?, in Reconstructing Early Christian Worship (London: SPCK, 2009), 3.

9. The exception is Matthew Colvin, The Lost Supper: Revisiting Passover and the Origins of the Eucharist (Lanham, MD: Lexington, 2019), for whom the Last Supper is the sole foundational event for the eucharist. Colvin argues, in my view convincingly, that the Last Supper is portrayed as a Passover meal (he actually argues that it was historically so, a discussion from which I prescind in these chapters), and he is aware that this means that an annual event had therefore to become a more frequent event but does not explain how this change takes place. On 40–41, he suggests that implicit in the act is a demand for repetition but does not specify why such repetition should not be annual. He also writes of the pervasive influence of the Last Supper narrative in the eucharistic material within the New Testament treatments of the eucharist; I do not see this pervading to John 6 and, contemporary with the New Testament documents, to the Didache.

10. Andrew McGowan, Naming the Feast: Agape and the Diversity of Early Christian Meals, StPatr 30 (1997): 314–18, in conclusion.

11. Thus see the evidence presented both for adoption and adaptation in rabbinic circles by Sandra R. Shimoff, Banquets: The Limits of Hellenization, JSJ 27 (1996): 440–52.

12. Jens Schröter, Das Abendmahl: Frühchristliche Deutungen und Impulse für die Gegenwart (Stuttgart: Katholisches Bibelwerk, 2006), 140.

13. Similarly Bryan D. Spinks, Do This in Remembrance of Me: The Eucharist from the Early Church to the Present Day (London: SCM, 2013), 4–5.

14. A point well made by Blake Leyerle, in a useful summary of the extent of our knowledge of this meal tradition, Meal Customs in the Greco-Roman World, in Passover and Easter: Origin and History to Modern Times, ed. Paul F. Bradshaw and Lawrence A. Hoffman (Notre Dame: University of Notre Dame Press, 1999), 45. Martin Stringer, Rethinking the Origins of the Eucharist (London: SCM, 2011), 171–74, similarly suggests that the symposium model has been inappropriately forced onto the evidence in some treatments. Finally we may note that Hans Joachim Stein, Frühchristliche Mahlfeiern, WUNT 2/255 (Tübingen: Mohr Siebeck, 2008), 11–15, charges that Klinghardt and Smith have extracted an ideal type, which was rarely met in reality.

15. Paul. F. Bradshaw, The Search for the Origins of Christian Worship: Sources and Methods for the Study of Early Liturgy, 2nd ed. (Oxford: Oxford University Press, 2002), ix–x.

16. Clemens Leonhard, Pesach and Eucharist, in Hellholm and Sänger, Eucharist, 1:275–312.

17. Joseph Bingham, Origines Ecclesiasticae or the Antiquities of the Christian Church, vol. 5 (repr., London: William Straker, 1834), 401–10.

18. Pierre Batiffol, Études d’histoire et de théologie positive [1er série]: La discipline de l’arcane; Les origines de la pénitence; La hierarchie primitive; L’agape, 7th ed. (Paris: Gabalda, 1926), 325.

19. Still maintained by Heinz Schürmann, Die Gestalt der urchristlichen Eucharistiefeier, MTZ 6 (1955): 123.

20. J. B. Lightfoot, The Apostolic Fathers (London: Macmillan, 1889), 2.1:400–401.

21. Andrew McGowan, Rethinking Agape and Eucharist in Early North African Christianity, StLi 34 (2004): 165–76; Tobias Georges, Das Gemeindemahl bei Tertullian in Apologeticum 39—eine nichtsakramentale Agapefeier?, ZAC 16 (2012): 279–91.

22. Batiffol, Études.

23. See, e.g., V. Ermoni, L’agape dans l’église primitive (Paris: Bloud, 1906). The state of discussion in the early years of the twentieth century (with references) is summed up well by Ephrem Baumgartner, Eucharistie und Agape im Urchristentum (Solothurn: Buch- und Kunstdruckerei Union, 1909).

24. Indeed, Batiffol’s contribution is all but forgotten; although Justin J. Meggitt, Paul, Poverty and Survival (Edinburgh: T&T Clark, 1998), 190–93, cites him, in attempting to argue along similar lines; however, because his own major thesis, denying the occurrence of social stratification at Corinth, is itself so uncertain, he has done nothing to rescue Batiffol’s reputation.

25. Notably Adolf Jülicher, Zur Geschichte der Abendmahlsfeier in der ältesten Kirche, in Theologische Abhandlungen: Carl von Weizsäcker zu seinem siebzigsten Geburtstage, ed. Adolf Harnack, Emil Schürer, and Heinrich Julius Holtzmann (Freiburg: Mohr, 1892), 217–50; Friedrich Spitta, Die urchristlichen Traditionen über Ursprung und Sinn des Abendmahls, in Zur Geschichte und Litteratur des Urchristentums (Göttingen: Vandenhoeck & Ruprecht, 1893), 1:205–337.

26. Hans Lietzmann, Messe und Herrenmahl: Eine Studie zur Geschichte der Liturgie (Bonn: Marcus & Weber, 1926).

27. Gregory Dix, The Shape of the Liturgy (London: Dacre, 1945). Notice should also be given to the near contemporary, but sadly forgotten, Felix Cirlot, The Early Eucharist (London: SPCK, 1939). Limited to the early period of development, Cirlot’s book is actually rather better than that of Dix.

28. In Alistair Stewart-Sykes, The Life of Polycarp: An Anonymous Vita from Third-Century Smyrna, ECS 4 (Sydney: St. Paul’s, 2002), 68–70.

29. Catherine Bell, Ritual Theory, Ritual Practice, 2nd ed. (Oxford: Oxford University Press, 2009), 93.

30. Bell, Ritual Theory, 90–91.

31. Bell, Ritual Theory, 140–42. A similar point is made by P. J. J. Botha, The Eucharist in Early Christianity as a Folk Event, SAJFS 2 (1991): 38–39.

32. So Botha, Eucharist, 45–47.

1

SATURDAY NIGHT AND SUNDAY MORNING

The Transfer of Christian Worship from Evening to Morning and Its Implications for the Eucharistic Liturgy

In this first chapter, I note the evidence for evening liturgies in early Christianity but also the evidence that, over the course of the second century, morning liturgies become more normal and widespread. I suggest that this represents a movement not simply from evening to morning but from Sabbath to Sunday, that is to say, a move from the second eve of the Sabbath (namely, Saturday evening) to the following morning. In the light of this transfer, the implications for eucharistic celebrations are studied.

1. THE TRANSFER FROM EVENING TO MORNING

The first stage in the argument is to secure the suggestion that early Christian eucharistic gatherings were normal on Saturday evening.

1.1. The Evidence of Acts

Possibly the earliest evidence is the report of Acts 20:7–11, set in Troas:

On the first day of the week [en de tē mia tōn sabbatōn], when we met to break bread [klasai arton], Paul was holding a discussion with them; since he intended to leave the next morning [mellōn exienai tē epaurion] he continued speaking until midnight. There were many lamps in the room upstairs where we were meeting. A young man named Eutychus, who was sitting in the window, began to sink off into a deep sleep while Paul talked still longer. Overcome by sleep, he fell to the ground three floors below and was picked up dead. But Paul went down, and bending over him took him in his arms, and said, Do not be alarmed, for his life is in him. Then Paul went upstairs, and after he had broken bread and eaten [kai klasas ton arton kai geusamenos eph’ hikanon], he continued to converse [homilēsas] with them until dawn; then he left.

The question is raised as whether this gathering, which is said to take place on the first day of the week, is actually a gathering on the Saturday evening and marks the close of the Sabbath and thus the beginning of the first day, which in time would become the Lord’s Day,¹ or whether the first day of the week should be taken to refer to Sunday and that the evening gathering is on what we would recognize as Sunday evening.²

The first point that must be examined is whether it is reasonable, in the first century, to measure the first day as beginning with the prior evening. Pliny informs us that, for the common people, days begin with dawn, and that the official Roman day ran from midnight.³ On either basis, "on the first day of the week [en de tē mia tōn sabbatōn] would refer to Sunday evening. It is on such a basis that Tromp, in his recent discussion of the passage, is dismissive of the idea that a day might be measured from the night before.⁴ However, he rather undoes his own argument at this point, in discussing the growth of the idea of the eve of the Sabbath" (ereb šabbat), by which the afternoon of Friday is meant,⁵ which gave opportunity to those who kept the Sabbath in rest to make preparation for the following day. If the eve of the Sabbath is the evening of Friday, with the Sabbath coming on at sunset, then how is the evening of Saturday to be denoted? Tromp answers this conundrum in noting that, as it is now dark, no work is done, but here we are not concerned with labor but with a gathering. The very notion that the onset of the Sabbath is the dusk of the Friday evening indicates that it might conclude on Saturday, and thus that the first day comes in with the evening. Finally we should note the evidence supplied by Llewellyn that, when a night was designated, it might be designated with reference to the following day, even outside Jewish circles.⁶ Nonetheless Llewellyn’s presentation of the evidence is so balanced that he notes the absence of unanimity.

De Jonge suggests that there are three reasons to discount seeing this as Saturday, and not Sunday, evening, though all center around whether Luke understood the day to begin in the evening or at midnight, which in turn is determined by our understanding of the statement regarding Paul’s travel plans, and the statement that he intended to leave the next morning (tē epaurion).⁷ If, however, this phrase means the next day, which is admittedly the usual meaning in Luke’s usage (as de Jonge argues at length), then the argument is that Luke would not call Saturday evening Sunday evening, and then refer to Sunday morning as the next day, as it would actually be the same day. However, regardless of Luke’s usage elsewhere, we must be mindful of the possibility that a source is being employed here, and since the next morning is a perfectly plausible rendition of tē epaurion, such a possibility cannot be excluded.⁸

In the light of this extensive prior discussion, we might cautiously conclude that the question of whether Saturday or Sunday evening is intended may not be answered with reference simply to the language that is employed here.

Riesenfeld had proposed that the evening was what we would recognize as being Saturday on the basis that the close of the Sabbath is intended, and that the first generation of Jesus-believers continued to maintain the Jewish practice of observing the Sabbath, so gathering as it concluded.⁹ It is this understanding of the means by which Sabbath-observing Christians of Jewish heritage would keep the Sabbath, namely, in common with non-Jesus-believing Jews, in order to gather with other Christians subsequently, which means that we may reject Young’s assumption that Sabbath-observant Jewish Christians would worship on the Sabbath qua Christians, with the consequent argument that Sunday was not in use.¹⁰ Rordorf’s chief argument against Riesenfeld is that there is no good reason why this Saturday gathering should become a Sunday gathering, though the answer is effectively supplied by Dekkers, who suggests that the lateness of the gathering would naturally lead to a transfer to the following morning.¹¹ There are many reasons why such a transfer might take place; the question is explored below. For the moment, we simply record that such a move is not unreasonable or irrational. It is thus plausible that the gathering at Troas took place on Saturday evening, and on balance more probable that this, rather than Sunday evening, is intended.

However, regardless of the evening intended, we cannot be sure that this is a regular gathering, occurring at a regular time, rather than a gathering simply occasioned by Paul’s imminent departure,¹² though the narrative implies that what was distinct about the occasion was not the gathering to break bread but the length of Paul’s address.¹³ An indication that this is a Sunday/Sabbath gathering might be gleaned by observing that bread is broken, but this may simply indicate that this was an occasion on which food was taken in a regular and normal way.¹⁴ If we were certain that this was a gathering on the eve of the Sabbath, it might well be argued that this is indeed a eucharistic reference; it is more difficult if this is Sunday evening, as there is no rationale by which the eucharist would occur on a Sunday evening, whereas, as will be seen, there is further evidence for a Saturday evening eucharistic gathering, but given the doubt, the most weight this account can bear is as supporting evidence for a position that must be established from other material. A similar caution may be exercised with regard to the argument of Staats, that the resurrection of Eutychus indicates a typology of the resurrection of Christ, likewise occurring on the night of Saturday.¹⁵ It seems, on balance, more probable that Saturday night is intended, rather than Sunday, though there is no single decisive argument, and no certainty that this describes regular practice.

1.2. The Evidence of Pliny

Chronologically next is probably the account of Pliny, who reports, at second hand, that Christians gather on a particular day before dawn and reconvened for a meal, though they had now abandoned that practice:

They claimed, however, that their fault and error was simply this: that they were accustomed to gather before dawn on a certain day [stato die] to pronounce an incantation [carmen] responsively to Christ as to a god [Christo quasi deo] and bind themselves by oath [sacramento] not to any crime, but to commit no fraud, theft, or adultery, nor to break trust, and not to refuse to return anything entrusted to them when it is recalled. When this was completed, they were accustomed to depart and to gather again to take food [rursusque coeundi ad capiendum cibum], common [promiscuum] and harmless food. After my edict, in which I had forbade associations in accordance with your instructions, they had ceased to do this [quod ipsum facere desisse].¹⁶

If we were to hypothesize that the gathering for food is a reference to the celebration of the eucharist, then rather than being abandoned altogether, as Pliny understands his informants to say, we might suggest that it has been transferred to the morning and consequently reduced in scale.

Rordorf, in keeping with his treatment of Acts 20, suggests that this is also a gathering on Sunday evening.¹⁷ However, we should be clear that the day is not stated, and we may rejoin that it is quite possible that the fixed day, of which report was made to Pliny, was actually the Sabbath (despite the almost universal assumption that Sunday is intended).¹⁸ In support of this, we may note that the early morning is spent in some form of study, in which the Christians bind themselves to forms of behavior that seem to reflect the teaching of Did. 1–5. This learning activity is in keeping with the synagogal activities of the Sabbath known in Jewish circles.¹⁹ If this is the case, then we may see that the gathering for food in the evening to which reference is made was the eucharist, taking place at the close of the Sabbath.

The further question arises that if the meal, which had previously taken place on Saturday evening, were transferred to a morning, then to which morning is it more likely to have been transferred? Although there is nothing that would prevent it being transferred to a Wednesday, it is somehow more probable that here we have the genesis of a Sunday morning gathering. The Saturday evening gathering for food, taking place at the close of the Sabbath, had become the Sunday morning eucharist.²⁰

There is, of course, the more conventional reading of the meal as being simply an agape, a noneucharistic meal that was abandoned.²¹ This could mean that the meal accompanying the eucharist was abandoned, which is more or less the reading adopted here, or else that the agape alone (assuming such a thing existed) was abandoned. Whereas this is not impossible, it is problematic. Quite apart from the question of whether the agape even existed as an institution at this time (a question to be explored in the second chapter), if the evening intended is the Sabbath, then it is unlikely that it would be marked by a meal without sacramental content; it is, moreover, unlikely that a nonsacramental meal would be described while no mention is made of the eucharist. Beyond this, such a reading is to an extent determined by the assumption that the sacramentum that took place in the morning is some kind of reference to the eucharist; if this position is abandoned and the sacramentum taken as baptism, or as something relating to baptism such as the renunciation and syntaxis,²² then the evening ritual must in some sense have been eucharistic. Thus we conclude that the import of the report to Pliny is that the morning is the occasion of initiation and catechesis and that the evening of the Sabbath had been an occasion on which a cena had taken place. This cena has been abandoned. Although it is not stated explicitly, we may surmise that the eucharistic ritual continues, though in the morning.²³

1.3. The Evidence of Traditio Apostolica

We may step slightly outside the chronological line for the moment to consider a vital piece of evidence provided by Traditio apostolica. Although this document as finally redacted is a product of the third century (or later), chapter 25 would seem to derive from a much earlier source that has been incorporated in the document:

When the bishop is present and evening is come, the deacon brings in a lamp, and standing among all the believers who are present, he shall give thanks. First, he greets them as he says, The Lord be with you.

And the people shall say, And with your spirit.

Let us give thanks to the Lord.

And they shall say, It is right and just. Greatness and exaltation with praise is fitting to him.

And he shall not say, Hearts on high, for it is to be said at the offering.

And he shall pray in this way as he says, We give you thanks, O God, through your child Jesus Christ our Lord, through whom you have illuminated us, revealing to us the incorruptible light. Therefore we have completed the length of the day and we have arrived at the beginning of the night, being sated with the day’s light that you created for our satisfaction. And now, having arrived at the light of evening through your grace, we give you praise and glorify you through your child Jesus Christ, our Lord, through whom to you be power and honor together with the Holy Spirit, now and always and to the ages of ages. Amen.

And all shall say, Amen.

And then, when they get up after the dinner, they shall pray, and the children and the virgins shall say psalms.

And afterward the deacon, when he takes the mixed cup of the oblation, shall say one of the psalms in which alleluia is written.

After that, if the presbyter has commanded, again from the same psalms. And afterward, the bishop having offered the cup, he shall say a psalm proper to the cup, while all say alleluia. And all of them, as he recites the psalms, shall say alleluia, which is to say, we praise him who is God most high; glorified and praised is he who founded all the world with one word.

And likewise, when the psalm is completed, he shall give thanks over the bread and give of the fragments to all the faithful.

There is much that is obscure and unclear about this chapter. However, for the present we assume, as is argued at length elsewhere, that what is being described is an evening eucharistic celebration that takes place in the context of a symposium, as opposed to the common characterization of this event as an agape.²⁴ What is not stated, however, is the day of the week on which this occurs. Nonetheless the emphasis on the lamp, with its accompanying blessing, may well indicate that this is a habdalah lamp; as such, the occasion is the closure of the Sabbath. The use of later Jewish practice to determine what practice might have been inherited by Christians of Jewish heritage is always risky, but we may note here the blessing of the cup. This may represent the qedušah of the cup before the meal,²⁵ which coheres with practice at the conclusion of the Sabbath,²⁶ and thus indicates that this is the occasion.

Thus far, then, there is no certain evidence of the day of Christian gathering, but there is a strong possibility that the favored occasion of gathering would be the evening of Saturday; this is coherent with a Jewish origin of some Christian practice, in that this gathering would follow on from rest or from synagogal activity, which according to contemporary sources would go on into the afternoon of Saturday.²⁷ We may thus envisage that Jesus-believing Jews appended the Jesus meal to their Sabbath activities on this occasion, which would be considered the opening of the first day. This would explain the eventual emergence of Sunday morning as the occasion of worship, even though there is no certainty as to the rationale for the transfer. It is hard, however, were it hypothesized that the evening of observation was Sunday, to explain the transfer to the morning.

1.4. The Evidence of Dionysius of Alexandria

Some confirmation that in some Christian circles the evening of Sabbath was considered the onset of Sunday, and thus the occasion of Christian worship, is found in a citation of Dionysius of Alexandria preserved in Armenian from the testimonia of Timothy Ailuros.²⁸ Dionysius is arguing that the day concludes at nightfall. Were it otherwise, and the following night to be counted as part of a day, then he suggests that the resurrection would have occurred not on the third day but on the second, and as such, by this logic, Christian worship should occur on the Sabbath, rather than the Sunday. Rather, he states, the day begins with the onset of darkness, and so the night of Saturday is the day after the Sabbath, and thus the third day, the first being the Friday on which Jesus was crucified. A number of arguments are employed to sustain this position, but what is significant for our purposes is the indication that Christian worship at the close of the Sabbath, and opening of Sunday on what we would term Saturday evening, is a recognized practice in Egypt, even as the resurrection is seen as the basis on which Sunday is a day of Christian worship. This is thus a valuable testimony to the continuation of worship on Saturday evenings in Egyptian circles, as well as indicating that it was a source of confusion due to the different methods of calculating days (since otherwise Dionysius would not have had to deal with the matter).

1.5. The Evidence of Ignatius of Antioch

In this light, we may return to the second century, and to Asia, to observe the exhortations of Ignatius of Antioch to the Magnesians:²⁹

Therefore, if those who conducted themselves in accordance with the ancient ways came to a newness of hope, no longer keeping the Sabbath but living in accordance with the Lord’s (day?) [mēketi sabbatizontes alla kata kyriakēn zōntes], and on which our life arose [aneteilen] through him and through his death, which some deny, we receive our faith through this mystery, and remain constant in it, so that we may be found to be disciples of Jesus Christ our only teacher, how could we live without regard to him?³⁰

This is a difficult text, with its string of subordinate clauses;³¹ however, the gist is that Ignatius suggests that the prophets of the old covenant had been persecuted because they lived the Lord’s life. Thus they came to newness of hope and abandoned the Sabbath to live in accordance with the dominical (day). Although the word day is not present, the context of the statement that they did not sabbatize, and the allusion to the resurrection, indicate that this is what is meant. The implicit question to the Magnesians is that if the prophets did that, should not Christians likewise live in accordance with the Lord’s (day)? Assuming that this does refer to keeping Sabbath or Sunday as a day of liturgical gathering (which is far from certain),³² then even then it is not clear whether these Christians are being exhorted to cease to keep the Sabbath gathering on Friday nights in order to transfer it to Saturday,³³ or whether Ignatius now sees the Saturday evening gathering as itself sabbatizing (even if, as a gathering, it is actually Christian in origin), and is proposing that the entire shift of worship should move to Sunday morning. Either is possible, but I would suggest that the latter is meant. The principal rationale is the connection of the day to the resurrection (on which our life arose), which points to the morning;³⁴ however, there is more. Ignatius is particularly concerned with the continued practice of private patronage within the church, which threatens the position of the episkopos as broker of patronage. It is easier to manifest patronage in the context of a meal, and so the removal of the gathering to the morning, with the consequent reduction in food consumed, also removes an opportunity for the public expression of patronage. Ignatius need not be familiar with the custom of counting days from the prior evening, and thus as part of his polemic against both scholasticized expressions of Christianity and the independent exercise of patronage (which may well be related) employs the argument that a gathering on Saturday evening is effectively a gathering on the Sabbath.

Nonetheless de Jonge considers that the Ignatian eucharist still involved a meal, and thus presumably considers that this Sunday gathering takes place in the evening.³⁵ As evidence for this, he notes Smyrn. 8.2, which discusses the holding of an agape, an event that, Ignatius says, should not be held without the bishop. But even if this agape included a eucharist, which is not certain,³⁶ it seems that Ignatius is concerned with independent gatherings, which may be held at any time, and not a regular Sunday eucharist. If such a gathering occurred on a Sunday, there would in any event be no cause to accuse its promoters of keeping a Sabbath.

We may thus conclude that in Magnesia, the gathering at the conclusion of the Sabbath was still held; Ignatius is familiar, however, with gatherings on a Sunday morning, and seeks to press this practice on the Magnesian church. Some evidence for this may perhaps be gained from the nearly contemporary and Asian Epistle to the Apostles, where the risen Christ states (apparently with regard to the resurrection) that he "came into being [aeichōpe] on the eighth, that is the kyriakē."³⁷ Again the term day is absent, but the context hardly allows any other noun to fill the gap.

1.6. The Evidence of Justin Martyr

Justin states that Christians gather on the day named after the sun.³⁸ It would thus seem that in this community, the transfer had been made. Even here,