1

Definitive text published in: Albert Gerhards and Clemens Leonhard (eds.), Jewish and
Christian Liturgy and Worship. New Insights into its History and Interaction, Jewish and
Christian Perspectives Series, 15, Brill, Leiden-Boston 2007, 295-308.

THE ROOTS OF THE EARLY CHRISTIAN EUCHARIST: JEWISH BLESSINGS OR

HELLENISTIC SYMPOSIA?

1. Introduction

The origins of few early Christian rituals have been debated as intensively as those of

the Eucharist. From the mid-twentieth century to this day, liturgical scholars have devoted a

flood of articles and books to the subject, sparked off by Gregory Dix’s The Shape of the

Liturgy, which appeared in 1945, just before World War II came to an end. 1

If one attempts to chart the main lines of thought that emerge from the vast amount of

secondary literature, one is struck by the fact that approaches have changed conspicuously in

recent decades. After the appearance of Dix’s extensive and influential monograph, scholars

turned at length to the Jewish roots of the Eucharist in their search for parallels with Jewish

meal traditions. Certain Christian Eucharistic prayers and Jewish prayer texts were believed to

be similar. They included blessings pronounced before and after meals (especially the ‘birkat

ha-mazon’, the grace after meals) and blessings, thanksgivings and supplications said on other

occasions (the blessings before and after the Shema and the benedictions of the Amidah).

Scholars concluded from such parallels that the early Christian Eucharist could be traced back

largely to Jewish meal traditions, which would have been transformed by Jesus and the first

generations of Christians. Outspoken exponents of this approach were Louis Bouyer, 2 Louis

Ligier, 3 Thomas Talley 4, Herman Wegman 5 and Enrico Mazza. 6

1
G. Dix, The Shape of the Liturgy, London. First published January 1945. Second edition August 1945.
Reprinted in 1946, 1947, 1949, 1952, 1954, 1960, 1964, 1970, 1975, 1978, 2005.
2
L. Bouyer, Eucharistie. Théologie et spiritualité de la prière eucharistique, Tournai 1966. Reprinted 1990.
3
See especially his articles ‘De la Cène du Seigneur à l’Eucharistie’, Assemblées du Seigneur 1 (1968), 19-57 ;
(English translation : ‘From the Last Supper to the Eucharist’, in : L. Sheppard (ed.), The New Liturgy, London
1970, 113-150) ; Idem, ‘Les origines de la prière eucharistique : de la Cène du Seigneur à l ‘Eucharistie’,
Questions liturgiques et paroissiales 53 (1972), 181-202 (English translation : ‘The Origins of the Eucharistic
Prayer : From the Last Supper to the Eucharist, Studia Liturgica 9 (1973), 161-185).
 2

In the past two decades, however, this approach has been increasingly criticised. The

main objection, voiced particularly by Paul Bradshaw, concerns an uncritical examination of

Jewish liturgical traditions which, moreover, are often attributed to a far too early period. 7 At

the same time, several scholars have explored new paths by drawing attention to similarities

between the early Christian Eucharist and Greco-Roman banquets. The latter, often

designated as symposia, were a common phenomenon in the Mediterranean world; they cut

across religious and ethnic boundaries (Jews, Greeks and Romans) and usually followed a

general pattern involving a number of customs and rituals. It has been proposed that the

Christian Eucharist originated and developed as a variety of this symposium. 8

Unlike scholars investigating Jewish origins of the Eucharist, most adherents of the

symposium theory are less interested in liturgical texts than in the social dimensions of the

early Christian Eucharist. Rather than analysing prayers, blessings and thanksgivings and their

4
T.Talley, ‘De la berakah à l’eucharistie. Une question à réexaminer’, La Maison-Dieu 25(1976), 11-39; Idem,
‘The Literary Structure of Eucharistic Prayer’, Worship 58 (1984), 404-420; Idem, ‘Structures des anaphores
anciennes et modernes’, La Maison-Dieu 191(1992), 15-43.
5
H. Wegman, ‘Généalogie hypothétique de la prière eucharistique’, Questions liturgiques 61(1980), 263-278 ;
Idem, ‘Genealogie des Eucharistiegebetes’, Archiv für Liturgiewissenschaft 33(1991), 193-216.
6
See especially the articles collected in E. Mazza, L’anafora eucaristica. Studi sulle origini, Bibliotheca
«Ephemerides liturgicae subsidia 62», Roma 1992 and E.Mazza, L’eucaristia nella storia, 1996 ( French
translation: L’action eucharistique, Paris 1999. English translation: The celebration of the Eucharist,
Collegeville 1999).
I myself have tried to sketch the development of the Eucharist and the Eucharistic prayer of early Syriac
Christianity (the first four to five centuries), starting from the birkat ha-mazon as reconstructed by Louis
Finkelstein and the Eucharist underlying Didache ch. 9 and 10, which in my view is based on Jewish meal
customs ( see: G. Rouwhorst, ‘Bénédiction, action de graces, supplication. Les oraisons de la table dans le
Judaïsme et les celebrations eucharistiques des Chrétiens syriaques, Questions liturgiques 67(1980), 211-240.
For a survey of the research of the Eucharistic Rites, see: P. Bradshaw, The Search for the Origins of
Christian Worship. Sources and Methods for the Study of Early Liturgy, London 2002, Revised and Enlarged
Edition, 118-143.
7
See in particular his book The Search for the Origins. See further Bradshaw’s monograph on the origins of the
early Christian Eucharist: Eucharistic Origins, Oxford 2004 and his article ‘Parallels between Early Jewish and
Christian Prayers. Some Methodological Issues’, in: A.Gerhards/A.Doeker/P. Ebenbauer (eds.), Identität durch
Gebet. Zur gemeinschaftsbildenden Funktion institutionalisierten Betens in Judentum und Christentum,
Paderborn 2003, 21-36.
8
See in particular M. Klinghardt, Gemeinschaftsmahl und Mahlgemeinschaft. Soziologie und Liturgie
frühchristlicher Mahlfeiern, Tübingen/Basel 1996; H.J. de Jonge, ‘ The Early History of the Lord’s Supper’, in:
J.W. van Henten and A.Houtepen (eds.), Religious Identity and the Invention of Tradition, Assen 2001, 209-237;
Idem, Zondag en sabbat. Over het ontstaan van de christelijke zondag, Dies-oratie University of Leiden, 2006;
D. Smith, From Symposium to Eucharist in the Early Christian World, Minneapolis 2002. See also A.
McGowann, Ascetic Eucharists. Food and Drink in Early Christian Ritual Meals, Oxford 1999, esp. 45-60; P.
Bradshaw, Eucharistic Origins, 43-44. See for the symposia also: B. Leyerle, ‘ Meal Customs in the Greco-
Roman World’, in: P. Bradshaw/A.Hoffman (ed.), Passover and Easter. Origin and History to Modern Times, =
Two Liturgical Traditions, vol. 5, Notre Dame 1999, 29-61.
 3

theological significance, they examine social structures and internal hierarchies, and non-

verbal, material aspects such as the architectural setting, the compilation of the menu and the

choice of food. Some scholars even draw their main inspiration from social theory, derived

from sociology or cultural ( social) anthropology. A typical example of this emphasis on the

social and material aspects of the early Christian Eucharist is Andrew McGowan’s study

Ascetic Eucharists, which deliberately leaves aside prayer texts to focus on meanings encoded

in food and drink. 9

In this paper I would like to offer a brief critical evaluation of theories regarding the

pre-Christian origins of the Eucharist. I shall consider objections to the Jewish roots of the

Christian Eucharist and the early dating of Jewish liturgical traditions. I shall also assess the

symposium thesis and the extent to which early Christian ritual meals were based on Greek

and Roman symposia.

Before proceeding, some observations should be made on the development of the early

Christian Eucharist itself. In recent decades, apart from the investigation of pre-Christian

roots, there has been a tendency to study the emergence and early history of the Eucharist

from a new perspective. This has involved a different way of selecting relevant sources and

the questioning of theological views previously considered self-evident. Since these two

aspects are interconnected, they cannot be considered separately. Moreover, they have

repercussions with regard to the pre-Christian roots of early Christian ritual meals, and in

particular the Eucharist.

2. The development of a new paradigm for the study of the early Christian Eucharist and the

selection of sources.

9
See esp. McGowan, Ascetic Eucharists, 1-9.
 4

One of the principal convictions underlying the traditional view of the early Christian

Eucharist is that, from the very beginning, it was the continuation of the Last Supper (which

many scholars believed to have taken place in the setting of the annual Jewish Passover

ritual). 10 The early Christian Eucharist would have followed the pattern of the Last Supper,

with the blessing of the bread preceding that of the wine, the institution narrative recited

during the Eucharistic prayer, and the ritual as a whole primarily commemorating the death of

the Lord. This approach was based on a selection of source material that legitimated it and

thereby contributed to the exclusion of contradictory sources. The principal sources taken into

consideration were the New Testament institution narratives, the description of the Eucharist

by Justin (150 CE), and the Eucharistic prayer encountered in the so-called Apostolic

Tradition, usually ascribed to Hippolytus of Rome at the beginning of the third century. On

the basis of Justin in particular, it was generally accepted that the main ritual elements of the

Eucharist had been dissociated from the meal (originally held in the evening, as was

customary in the Mediterranean) and had been amalgamated with the Jewish synagogue’s

‘liturgy of the word’. This would account for the general dual structure of the Eucharistic

liturgy in practically all liturgical traditions of the east and west: the liturgy of the Word

preceding the preparation and offering of the gifts (1), blessing and thanksgiving (2), breaking

of the bread (3) and communion, i.e. the ritualised eating and drinking of bread and wine (4)

‘eucharistised’ by the prayer of blessing and thanksgiving. The so-called anaphora of the

Apostolic Tradition (ch.4) was seen to support the view that the Eucharistic celebration

included the recitation of the institution narrative. Incidentally, this text played an important

role in ‘proving’ continuity between the New Testament period and the golden age of the

10
One of the most important and influential representatives of this traditional view is again Gregory Dix, who
has assumed the existence of an original, pristine and ‘apostolic’ core of the Eucharist, which would have
evolved from a ‘seven-action’ form, attested by the New Testament institution narratives, into a ‘four-action’
pattern. For a discussion of this hypothesis, see: M.E. Johnson, ‘The Apostolic Tradition’, in: G.Wainwright/ K.
Westerfield Tucker, The Oxford History of Christian Worship, Oxford 2006, 32-75, esp. 44-50. On questions
raised by this traditional view [#proposed by Dix and held by many other scholars(weg?)], see also: G.
Rouwhorst, ‘La celebration de l’Eucharistie dans l’Eglise primitive’, Questions liturgiques 74(1993), 89-112.
 5

Church Fathers (4th and 5th centuries). It served as a kind of stepping stone, allowing scholars

to jump from the ‘apostolic’ era to the patristic period. Sources not fitting into this pattern

were considered to describe other types of ritual meals, particularly the agape, or to derive

from heterodox or at least marginal circles. They include the Didache (ch. 9 and 10),

apocryphal Acts of the Apostles such as those of Thomas and John, and descriptions of

common meals found in Tertullian (especially in his Apology, ch. 39) and in the Apostolic

Tradition (ch. 25-29) 11.

Today, this traditional approach is being abandoned by more and more scholars. In a

sense, a complete reassessment of the sources has taken place: those ignored or marginalised

by adherents of the traditional paradigm now take pride of place in reconstructions of the

development of the early Christian Eucharist. 12 The reliability of the institution narratives as

testimonies of early Christian liturgical practice is increasingly questioned. 13 Conversely, the

Didache now occupies a key position in the argumentation. 14 Attempts to fit it into the

traditional pattern by the Procrustus method – by removing or adding elements that are not in

the text – are rejected by most scholars. Further, contrary to what was generally believed,

McGowann and Bradshaw have argued that the supper described by Tertullian in his Apology,

and designated as an ‘agape’, was a Eucharistic celebration. They have also produced

persuasive evidence that the gatherings before daybreak, at which the faithful received ‘the

sacrament of the Eucharist’ ‘from none but the hands of the presidents’ ( De corona militis ,

11
See the edition of B. Botte, La Tradition apostolique de Saint Hippolyte, Liturgiewissenschaftliche Quellen
und Forschungen 39, Münster 19895; P. Bradshaw/M.Johnson/L. Philips, The Apostolic Tradition, Minneapolis
2002.
12
See: Bradshaw, Eucharistic Origins, Oxford 2004 and also: Rouwhorst, ‘La célébration’; G. Winkler,
‘Weitere Beobachtungen zur frühen Epiklese (den Doxologien und dem Sanctus). Über die Bedeutung der
Apokryphen für die Erforschung der Entwicklung der Riten, Oriens Christianus , 80(1996), 177-200;
Klinghardt, ‘Gemeinschaftsmahl’, 499-522; McGowann, Ascetic Eucharists;
13
In addition to the literature in the previous footnote, see: M. Johnson, ‘The Apostolic Tradition’, 45-48.
14
See also: G. Rouwhorst, ‘Didache 9-10: A Litmus Test for the Research on Early Christian Eucharist’, in: H.
van de Sandt (ed.), Matthew and the Didache. Two Documents from the Same Jewish-Christian Milieu?, Assen,
Minneapolis 2005, 143-156. Cf. McGowann, Ascetic Eucharists, 21-22 and E.Mazza, ‘La «dottrina dei dodici
apostoli» o «Didache»’, in: Idem, L’anafora eucaristica, 19-50 en: Idem, ‘L’eucaristia di 1 Cor 10,17 in rapporto
a «Didache» 9-10’, ibidem, 77-109:
 6

ch. 3, 3; cf. De oratione 19,4), were not ‘Eucharists’, but a sort of communion service at

which Christians received Eucharistic food consecrated beforehand at an evening meal, called

agape in Apology ch. 39. 15 At the time of Cyprian, the Eucharist had already been transferred

from evening to morning, but according to McGowann and Bradshaw this innovation gave

rise to disputes echoed in Cyprian’s famous sixty-third letter about the water-drinkers. 16 As

for the so-called Apostolic Tradition, its attribution to Hippolytus, its Roman origin and its

date are now open to discussion. 17 In the context of this paper it is noteworthy that several

scholars have voiced serious doubts about the homogeneity of the anaphora of chapter 4,

suggesting that some passages, including the epiclesis and the institution narrative, might

have been added in the fourth century. 18 Although I am rather sceptical about attempts to split

the text into older and younger sections, I do have doubts about the dating of the prayer as

such. 19 In fact, I would suggest that the entire anaphora of chapter 4, rather than parts of it,

might have been added to the Apostolic Tradition at a considerably late stage in the

development of these documents, and that the earliest practices concerning the celebration of

the ‘Eucharist’ in this source are to be found in chapters 26-29, which contain descriptions of

ritual suppers remarkably similar to the ‘agape’ described by Tertullian in his Apology.

If all of this is true, it has far-reaching implications for the reconstruction of the

development of the early Christian Eucharist. It means that the separation of the Eucharist

from the context of an evening meal occurred at a much later date than is often assumed - in

Africa at least, but probably in other regions too. As for the Eucharist described in Justin’s

Apology, assuming that it was actually held in the morning (which is indeed likely, though

15
A. McGowann, ‘Rethinking Agape and Eucharist in Early North African Christianity, Studia Liturgica
34(2004), 165-176; Bradshaw, Eucharistic Origins, 96-103. Cf. for Apology, ch. 39 also Klinghardt,
Gemeinschaftsmahl, 514-517.
16
McGowann, ‘Rethinking’, 173-175; Bradshaw, Ascetic Eucharists, 108-114.
17
See for instance Bradshaw, Johnson, Philips, The Apostolic Tradition, 1-15.
18
See especially Bradshaw, Johnson, Philips, The Apostolic Constitution, 44-46; P. Bradshaw, ‘Introduction:
The Evolution of Early Anaphoras’, in: Idem (ed.), Essays on Eary Eastern Eucharistic Prayers, Collegeville
1997, 1-18, esp.10-14.
19
Cf. my review of Bradshaw, Johnson, Philips The Apostolic Tradition in Vigiliae Christianae 59 (2005), 337-
340.
 7

difficult to prove), 20 its case was rather exceptional, and it did not simply constitute the norm

from which other communities would have deviated. More important still, the influence of the

Last Supper and the New Testament institution narratives on early Christian practice was less

than has often been believed, and became stronger only in the course of time. (This

observation may give rise to numerous questions which I have discussed elsewhere and which

I will therefore leave aside here.)

This new approach obviously has repercussions for the various theories on the roots of

the Christian Eucharist. In general, it seems quite clear that scholars who have tended to find

these roots primarily in Jewish liturgical traditions have depended considerably on the

‘traditional paradigm’. This is certainly true of Dix, and to some extent also of Ligier and

Talley (Mazza’s view, on the other hand, contains many elements of the new paradigm).

However, this does not imply that the theory must be discarded for this reason alone, to be

replaced by one attributing the origins of the Christian Eucharist to a so-called Hellenistic

symposium. Theoretically, one can even imagine that the proponents of Jewish roots could

find some support in the new paradigm. Moreover, the plausibility of the symposium theory is

not enhanced purely by the fact that it fits better in the new paradigm sketched here. A critical

assessment of the two approaches on the basis of other criteria remains necessary, and

particular attention should be given to the manner in which available sources are interpreted.

The Jewish roots theory reconsidered.

The most decisive argument against the Jewish roots theory, as stated at the beginning,

is that it is based on uncritical use and incorrect dating of Jewish liturgical sources. The

question therefore arises as to how this argument stands in the light of our knowledge of the

20
Cf. Klinghardt, Gemeinschaftsmahl, 500-509.
 8

development of Jewish liturgical traditions in the early centuries of the Christian era, or, put

differently, in the Amoraic and particularly the Tannaitic periods. Should we conclude that the

whole theory is untenable?

This question, I believe, is too complex to be answered with a categorical ‘yes’ or

‘no’. Firstly, it cannot be denied that a number of scholars, some of whom have already been

mentioned, have drawn undiscerningly on rather late Jewish sources. Some have rather

naively interpreted any parallel between Jewish and Christian liturgical practices or texts -

often irrespective of region or date - as an indication of the dependence of the Christian

tradition on the Jewish. Furthermore, some may be blamed for making hasty and sweeping

generalisations, often based on a very limited number of sources. An example is the tendency

to trace back all early Christian Eucharistic prayers to the pattern of birkat-ha-mazon. While

it cannot be denied that some surviving Christian prayer texts fit comfortably in this pattern,

many others do not, or are only faintly similar to this Jewish prayer (which, incidentally, is

attested by rather late sources). 21 It should also be admitted that certain widespread theories

about the Jewish roots of specific aspects of the early Christian Eucharistic prayer have

become entirely obsolete. An example is Ligier’s attempt to explain the insertion of the

institution narrative in the Eucharistic prayer by comparing it to the practice of inserting so-

called embolisms in the birkat ha-mazon during the great Jewish festivals. 22 Apart from the

fact that it is difficult to know precisely when this practice originated in Judaism, the entire

argument is founded on the problematic assumption that the insertion of the institution

narrative occurred at a very early date, when the first Christian rituals emerged in a

predominantly Jewish environment. If there is one assumption that has become questionable,

it is precisely this one. Another widely accepted theory that has become increasingly

problematic is that of the dependence of the early Christian Eucharist on the Jewish Passover

21
See for instance Bradshaw, ‘Introduction’, 9-10.
22
See Ligier, ‘Les origines de la prière eucharistique’ (‘The Origins of the Eucharistic Prayer’).
 9

celebration. Here again, two objections can be raised. On the one hand, most versions of this

argument exaggerate the role of the institution narratives, which indeed constitute the

principle basis of the hypothesis. On the other hand, with regard to the reconstruction of the

Last Supper and its alleged Passover context, the proponents tend to draw uncritically on

Jewish sources of a much later period than the New Testament. 23 This holds for the Mishnah

and the Tosefta, and without doubt for the Babylonian Talmud. 24

This being said, it would be premature to reject outright the entire thesis of the Jewish

roots of the Christian Eucharist. A number of facts should prevent us from drawing such a

drastic conclusion.

1. In spite of all objections to the Jewish origins of Christian liturgy, the fact remains that

Christianity and its earliest rituals emerged and developed within the variegated world

of Judaism at the beginning of the Common Era. Obviously, this fact cannot be used

uncritically to prove the Jewish origin of any early Christian ritual remotely similar to

some sort of Jewish Christian tradition. Nonetheless, one must realise that the earliest

Christian liturgical practices were not invented out of thin air. This goes to support the

hypothesis that the earliest Christian rituals had their origins in Jewish traditions.

2. In so far as the new view on the early Christian Eucharist, as I have described it,

allows for a variety of ritual practices not necessarily based on the uniform pattern of

the Last Supper, it may support the assumption that early Christian rituals such as the

Eucharist have their origins in Jewish traditions. If evidence indicates that some early

23
This holds in particular of J. Jeremias influential study Die Abendmahlsworte Jesu, Göttingen 19674. Cf. in
this connection the critique of this theory advanced by G. Stemberger. ‘Pesachhaggada und Abendmahlsberichte
des Neuen Testaments’, Kairos 29 (1987), 147-158 ( = Idem, Studien zum rabbinischen Judentum, Stuttgart
1990, 357-374.
24
On the development of the Pesachhaggada from the destruction of the Second Temple until the beginning of
the first millennium, see: C. Leonhard, ‘Die älteste Haggada: Übersetzung der Pesachhaggada nach dem
palästinensischen Ritus und Vorschläge zu ihrem Ursprung und ihrer Bedeutung für die Geschichte der
christlichen Liturgie’, Archiv für Liturgiewissenschaft 45(2003), 201-231; Idem, ‘Die Ursprünge der Liturgie des
jüdischen Pesach und das christliche Osterfest’, in: A. Gerhards/H.Henrix (ed.), Dialog oder Monolog? Zur
liturgischen Beziehung zwischen Judentum und Christentum, Freiburg 2004, 151-166.
 10

Christian Eucharists were based on Jewish meal traditions rather than the pattern of

the Last Supper, this cannot simply be refuted by claiming that the early Christian

Eucharist was not a Eucharist at all unless it followed the pattern of the Last Supper.

3. Although liturgical scholars of the past were often rather naive in attributing Jewish

origins to certain aspects of the early Christian Eucharist a number of facts cannot be

denied. The prayers in Didache 9 and 10 exhibit striking parallels with Jewish meal

berachoth, in particular the birkat ha-mazon, 25 even if the oldest known Jewish

versions (preserved in the Cairo Geniza) date from a much later period. Klinghardt

and others, who have made little of these facts, argue that the author or editor of the

Didache did not employ written texts, but was dependent on an oral tradition in which

texts were subject to variation. 26 Although this is a valuable conclusion in itself, it

fails to refute the thesis of Jewish origins. Apart from the case of the Didache, it can

hardly be ignored that there is strong evidence that at least some aspects of early

Christian Eucharists and ritual meals had Jewish roots. Suffice it to mention parallels

between the structure and some parts of the anaphora of Addai and Mari and various

Jewish texts, including the birkat ha-mazon, 27 the so-called ya’aleh we yavo 28 (a

prayer inserted in the birkat ha-mazon on festivals), and the blessings preceding the

Shema. 29 Last but not least, we should bear in mind the prominent place occupied by

thanksgivings and blessings in early Christian Eucharists and ritual meals, a

25
See for instance: G. Rouwhorst, ‘Didache 9-10’, esp. 149-151. See also: W. Rordorf-A.Tuilier, La Doctrine
des douze Apôtres (Didachè), SC 248, Paris 1978, 175-181; K. Wengst, Didache (Apostellehre), Barnabasbrief,
Zweiter Klemensbrief, Schrift an Diognet, Darmstadt 1984, 47-53; H. van de Sandt/ D. Flusser, The Didache. Its
Jewish Sources and its Place in Early Judaism and Christianity, Assen/Minneapolis 2002, esp. 310-313.
26
Klinghardt, Gemeinschaftsmahl, 407-427.
27
H. Wegman, ‘Pleidooi voor een tekst. De anaphora van de apostelen Addaï en Mari’, Bijdragen 40 (1979), 15-
43; Rouwhorst, ‘Bénédiction’, 231-239.
28
Rouwhorst, ‘Bénédiction’, 235.
29
See B. Spinks, ‘The Original Form of the Anaphora of the Apostles: A Suggestion in the Light of the Maronite
Sharar, Ephemerides Liturgicae 91 (1977), 146-161 ( reprint: B. Spinks, Prayers from the East, The Pastoral
Press, Washington 1993, 21-36; J. Vellian, ‘The Anaphoral Structure of Addai and Mari compared to the
Berakhoth Preceding the Shema in the Synagogue Morning Service contained in the Seder R. Amram Gaon’, Le
Muséon 85 (1982), 201-223.
 11

phenomenon unique to Jewish ritual meals - and Christianity - and apparently

unparalleled in Greek and Roman traditions. This can only be explained, I believe, by

a certain dependency of Christian texts on orally transmitted Jewish models, even if

the degree of dependency remains difficult to establish.

The symposium theory assessed

Having discussed the question of Jewish sources, let us turn to the alternative

approach, in which the early Eucharist is viewed as a Christian variety of the symposium.

The latter is assumed to have been common practice in the Mediterranean, and known to

both Greeks and Jews, as well as to the followers of other religious and ethnic traditions.

Generally speaking, this theory has the advantage of fitting in well with the new

approach to the early Christian Eucharist sketched above. More specifically, it is favoured

by an increasing number of liturgical scholars who wish to draw on the broadest variety of

sources, putting into perspective the one-sided focus on the Last Supper and the institution

narrative. From another point of view too this theory is most revealing, in that it opens our

eyes to the sociological dimensions of the early Christian meal traditions which were

often neglected by scholars who, in their search for Jewish roots, were absorbed primarily

by texts. We may become more aware, therefore, of the communal nature of early

Christian ritual meals and their relation to the identity of certain social groups. There were

rules for the admission of non-members, and hierarchical patterns inside the groups, and

these could be adopted or abandoned in early Christian meals. Finally, the study of Greco-

Roman community meals may tell us much about dining customs and material aspects
 12

such as the selection of menus; as McGowann has demonstrated, 30 the social and religious

significance of these traditions will escape us if we continue to focus exclusively on texts.

In spite of the merits of this approach, however, several limitations or even risks are

evident which, in my view, are not always taken sufficiently into consideration. I shall

limit myself here to two brief observations.

1. Adepts of the ‘search-for-Jewish-roots’ approach are frequently accused of drawing

rather uncritically on sources whose range is too wide and whose geographical and

historical backgrounds are too diverse. In my view, a similar objection may be made

to the manner in which some scholars employ Greco-Roman meal customs to explain

the origin and development of the early Christian Eucharist. They do not hesitate to

combine data derived from sources as heterogeneous as the writings of Plato, Plutarch,

Petronius, Philo of Alexandria, and sources derived from the Qumran community or

rabbinic milieus. 31 Despite the fact that meal customs were probably less susceptible

to change than the prayer texts to which proponents of the Jewish-sources theory

appeal, the vast range of periods and regions from which data are drawn should make

us wary of hasty generalisations. Further, the question may be raised as to whether

differences between various types of banquets are taken sufficiently into account. One

wonders, for instance, what philosophical banquets have in common with sacrificial

banquets, or with banquets held by all sorts of guild-like groups, let alone the Qumran-

communities or the Jewish chavuroth. It is surely relevant to consider whether a

symposium was mainly for socialising and amusement, or had a particularly religious

atmosphere (even allowing for the fact that religion had a different function in

Antiquity than in modern Western society, and that it is not always easy to distinguish

between the sacred and the profane). One may even pose the question whether the

30
See McGowann, Ascetic Eucharists.
31
See for instance Klinghardt, Gemeinschaftsmahl 21-249; Smith, From Symposium to Eucharist, 13-172
 13

symposium pattern was much more than a general model on which a great variety of

meals - whether religious or not - were based, but which, in itself, provides little

insight into the specific character of each type of meal.

2. As I have remarked, one of the principal merits of this theory is that it makes us alert

to the sociological aspects. Precisely here, however, the danger of one-sidedness

looms once more. Whereas scholars in search of Jewish roots run the risk of focusing

exclusively on the religious significance of texts, adherents of the symposium theory

are in danger of going to the other extreme, emphasising social practices and related

social codes at the expense of religious meanings. Unless one wishes to follow Emile

Durkheim in reducing God or the sacred to nothing but a reflection or projection of

society, it is necessary to develop a well-balanced approach that does justice to both

the religious and social dimensions of rituals, without reducing one to the other. This

implies that research into early Christianity and its meal traditions should involve both

non-textual and textual elements, and their pre-Christian roots, whether Jewish or non-

Jewish.

Final remarks: the Christian appropriation of Jewish and Greco-Roman traditions.

It will be clear from the foregoing that the two theories discussed each have their own

validity. Both shed light on important aspects of the development of early Christian meal

practices and the early Christian Eucharist in particular. The overall conclusion is that the two

theories are not mutually exclusive, but rather complement each other, each focusing on

important aspects of early Christian ritual meal traditions.

 14

Both theories, however, have their limitations. This holds for the more traditional

approach, which, as opponents did not hesitate to emphasise, concentrated on the Jewish

roots. It also applies, however, to the symposium hypothesis, which, developed and

increasingly accepted only in recent years, has yet to be critically assessed.

Moreover, both theories have a limitation in common that is inherent in the concept of a

search for origins. Given the fact that they are primary concerned with the dependency of

early Christian ritual meals upon pre-Christian traditions, whether Jewish, Greek or Roman, it

is easy to overlook specifically Christian dimensions. It is therefore important to bear in mind

that practices are never simply copied, but rather appropriated and transformed. Once a

particular tradition has been transmitted from one community to another, it will be

restructured and adapted. As it begins to fulfil new religious and social functions, new

meanings will be ascribed to it and new elements will be added. The full significance of a

specific tradition, ritual or otherwise, cannot be discovered only by laying bare its

antecedents. This applies to the ritual meals of early Christianity just as it does to the

Passover Seder or the rabbinic chavuroth, which were doubtlessly influenced by the Greco-

Roman symposium tradition.

The question remains what happened to the Jewish and Greco-Roman elements that

became incorporated in the ritual meals - Eucharists or otherwise - of early Christian

communities. It is not possible to discuss this at length here; indeed, the question requires a

fresh and critical re-examination of all early Christian sources that refer or allude to any sort

of ritual meal practiced by groups of Christians. I shall therefore limit myself to two

suggestions for further research.

Firstly, in the transformation of pre-Christian meal practices, the role played by the

introduction of the institution narrative and elements derived from the Last Supper tradition

has been overstated. As will be clear from this paper, the process by which the early Christian
 15

Eucharist incorporated and indeed gave a central place to the Last Supper tradition, and

particularly the recitation of the institution narrative, was slower and more gradual than has

been assumed.

Secondly, the transformation of pre-Christian elements related to meal customs, as they

became incorporated in the early Christian Eucharist, cannot be studied only in terms of texts.

We must bear in mind that non-textual elements, for instance gestures, too may have

acquired new significance in early Christianity. An interesting example might be provided by

the ‘breaking of the bread’. The importance given to this gesture from the very first days of

Christianity (1 Cor. 10;16; Luke 24: 30.35; Acts 2: 42.46; 20:7; Didache 14:1; Apocryphal

Acts of the Apostles, esp. those of Thomas; Ephrem the Syrian 32) appears to be unparalleled

in Greco-Roman and Jewish sources. In the Jewish tradition, the emphasis lies on the blessing

accompanying the breaking of the bread rather than on the breaking itself. In early

Christianity, this ritual gesture gained an intensity it had never had in Jewish or non-Jewish

symposiums.

Needless to say, these concluding observations in no way deny the importance of tracing and

mapping the pre-Christian roots, Jewish or otherwise, of early Christian ritual meals, in

particular those of the early Christian Eucharist. Although the study of these pre-Christian

roots alone is insufficient to gain a better understanding of the ritual meals of early

Christianity, it cannot be denied that it is an indispensable precondition in order to achieve

that goal. It is a first step in that direction, but a very important one.

32
Cf. for the Apocryphal Acts of Thomas and the works of Ephrem the Syrian: G. Rouwhorst, ‘Die Rolle des
heiligen Geistes in der Eucharistie und der Taufe im frühsyrischen Christentum’ (forthcoming, 2007).
16