The season of Advent (from the Latin adventus, ‘coming’, and a translation of the Greek παρουσία and/or 
πιφάνεια), as a liturgical season with that specific title, is a purely Western Christian phenomenon and comes into existence as a time associated with what Connell calls scriptural, ascetic and/or eschatological1 preparation for Christmas, of course, only after Christmas itself came to be established on 25 December. And it is not until the reign of Pope Gregory I (590–604) that the four-Sunday Advent season with its strong eschatological orientation clearly makes its appearance under Gregory’s own leadership.
This is not to say, however, that such a season of preparation was and is only Western or, as we shall see, that it is only related to Christmas in its origins. In the Byzantine Rite, for example, beginning with the 21 November feast of the Presentation of Mary in the Temple, multiple Marian images associated with the ‘Ark of the Covenant’, the ‘Tabernacle’, and even as the ‘heavenly Temple’ appear in the various troparia and prayers throughout the season.2 And, two Sundays before Christmas, the Byzantine Rite commemorates ‘The Holy Ancestors of Christ’, culminating in Mary, and on the Sunday before Christmas is celebrated ‘all the Fathers who down the centuries have been pleasing to God, from Adam to Joseph, husband of the Most Holy Mother of God’.3 Among the Syrian Christian traditions, both West Syrian (i.e., Syrian and/or Antiochene Orthodox and Maronite) and East Syrian (i.e., Church of the East, Chaldean and Syro-Malabar), the assigned Gospel readings on the Sundays for the season of Christmas preparation, called ‘Weeks of Annunciations’, include, in order, the annunciation to Zechariah, the annunciation to Mary, the visitation, the nativity of John the Baptist, and, finally, the annunciation to Joseph. Indeed, for these reasons, what Western Christians refer to as ‘Advent’ is often thought of as a ‘Marian’ season in the Christian East.
Further, this Eastern approach to the season of preparation for Christmas has some resonance in the West as well. While the precise origins of the 25 March celebration of the Annunciation of Our Lord are obscure,4 the feast on this date was celebrated already in the East by the beginning of the sixth century. Before that shift to a calendrical date, the Annunciation appears to have been celebrated on the Sunday before Christmas. Interestingly enough, the location of the feast of the Annunciation actually varied as to date in the calendars of other Western liturgical traditions throughout the Middle Ages. In Spain it was celebrated on 18 December, where even today in the Mozarabic Rite 18 December has remained a solemnity of Mary called, simply, Sancta Maria.5 At Milan the Annunciation was and still is celebrated by the Ambrosian Rite on the last of the six Sundays of Advent. Even in the liturgical tradition of Rome, a similar correlation between, at least, Annunciation and Christmas became also true of Roman Advent itself, although Rome had clearly accepted the 25 March date of the feast by the time of Pope Sergius I (687–701). Prior to the post-Vatican II liturgical reform of the calendar, in fact, the Gospel pericopes of both the annunciation and the visitation were read, respectively, on the Wednesday and Friday of the third week of Advent, formerly known as the Advent ‘Ember Days’, one of four annual ‘seasons’ of special prayer and fasting throughout the liturgical year.6 Hence, even with the acceptance of the 25 March date for the feast in the West, a close proximity between the celebration of the Annunciation (and the Visitation) and Christmas remained a traditional characteristic of Western liturgical history in general.
What became the pre-Christmas Advent season in the West may have had its origins, outside of Rome, in a period of ascetical preparation for baptisms celebrated on Epiphany, understood in part, as we saw in the previous chapter, as the celebration of Jesus’ baptism itself. Prior to the work of Talley, in fact, it was generally accepted that our earliest references to ‘Advent’ in the West were those presumably concerned with a three-week period of such preparation. The first is that of Hilary of Poitiers (d. 367), who, in an excerpt from his alleged Liber Officiorum, points to a three-week period before Epiphany, a ‘three-week Lent’,7 and the second is Canon 4 of the Spanish Council of Saragossa (380), which directs that ‘for twenty-one days in a row, from December 17 until the day of Epiphany, which is January 6, for continuous days no one should be absent from church or stay hidden at home or escape to the country or to the mountains or run around in bare feet, but all should come together in church’.8
Contemporary scholarship on Advent, however, represented primarily by Talley, has tended to discount this theory.9 The authorship of the text attributed to Hilary, for example, has been seriously questioned, with no consensus among scholars yet reached;10 and Talley believed that Christmas was already being celebrated in Spain by 380 and so considered it as having been included within these 21 days.11 Alexander takes Talley’s approach a step further, suggesting that:
every day of this three-week period was, or soon came to be, a day of special observance, December 25 included among them, and that it was the intention of the bishops to commend the keeping of these days to the faithful. It might be suggested since the first week of this period, December 17–23, was the time of the pagan saturnalia, that part of the motivation for the observance of those days was to place quite intentionally a Christian festival period over against the pagan. Although this is possible, another explanation could be that the annual observance of saturnalia was followed by a seasonal cessation from work, a sort of mid-winter recess, and it is entirely possible that the bishops of the council were calling their people to the faithful observance of the holy days that overlapped with their vacation. In other words, the canon has less to do with the shape of a particular liturgical observance than it does with a reminder to the faithful not to forget their obligation to the church while on holiday.12
Alexander’s comments seem overly speculative. If there is nothing explicit in this canon about Epiphany baptism or about this three-week period having anything to do with pre-baptismal catechesis, there is certainly nothing about this time period being a post-Saturnalia vacation period. And it becomes difficult to know what this period was about at all.
It is here where Connell’s critique of Talley’s position must be taken into serious account.13 Connell notes, in particular, that outside of Rome and North Africa in 380 there is no evidence anywhere in the West for the celebration of Christmas. Further, the period between 17 December and 6 January is described as one of continuous activity, possibly ascetic in nature, with no room for a festive break on 25 December but culminating on Epiphany, the only festival designated. He also points to a parallel possibility of a Spanish ascetic, identified by Gabriel Morin in 1928, whose writing refers both to a three-week fast and to the feast of Epiphany, though, as Connell notes, it could have come from almost any of the Latin churches outside of Rome, where Christmas had not yet made its appearance. As we saw earlier in our chapter on Lent,14 a three-week period of baptismal preparation in early Christianity appears to have existed in a wide variety of distinct churches: that is, Rome, Jerusalem, North Africa, Naples, Constantinople and Spain. That such a three-week pattern would have been in existence for baptism on Epiphany as well seems perfectly reasonable. For Spain, in particular, as we saw, the three-week ‘Lenten’ period not only appears to be confirmed by the first canon of the Second Council of Braga (572) but actually uses similar language to Saragossa in reference to the time span, while directing that bishops ‘shall teach that catechumens (as the ancient canons command) shall come for the cleansing of exorcism twenty days before baptism, in which twenty days they shall especially be taught the Creed, which is: I believe in God the Father Almighty …’15 It would be difficult to see both of these Spanish references to three weeks of time (21 days or 20 days) as not having some baptismal connotations.
The strongest argument against viewing this three-week period in early Spain as one of preparation for baptism on Epiphany with any degree of certainty comes from Alexander. He correctly draws our attention to the fact that in 385 Pope Siricius wrote a letter to Himerius, Bishop of Tarragona, in which it is documented, a mere five years after the Council of Saragossa, that baptisms took place in Spain on Easter, Pentecost, Epiphany, various saints’ days, and, most telling, now on Christmas as well.16 Such evidence, notes Alexander, makes it ‘unlikely that the period could be preparation for Epiphany baptism if another major baptismal day, December 25 … took place on intervening days’.17 At the same time, nowhere does this document indicate exactly where it is in Spain that baptisms were being conferred on Epiphany, saints’ days and Christmas, since Himerius himself claims to know only Easter and Pentecost baptism as the practice of his church at Tarragona. In other words, if this letter is a witness to the celebration of Christmas on 25 December in Spain in 385, it does not tell where it first makes its appearance in Spain, nor does it tell us necessarily anything about the practice of Saragossa five years earlier. And Alexander himself adds:
The possibility exists, of course, that the three-week period before Epiphany was a vestige of an earlier time before the bifurcation of that unitive festival into the separate feasts of Christmas and Epiphany, a possibility in light of the influence of the eastern rites on much of the early liturgy of Gaul and Spain … [I]t is difficult to assess … the exact intent of the bishops convened at Saragossa. It is impossible at this stage to dismiss completely the possibility that this three-week period before Epiphany had some measurable impact on the formation of Advent, but the weight of the evidence presently available makes it equally impossible to see anything more than a very small, ill-formed piece of a larger puzzle, certainly not the secure roots of western Advent.18
However, while we may have to settle for this particular conciliar reference to three weeks before Epiphany in Spain as only a possibility in pointing to a nascent Advent, the mere fact that both the baptism of Jesus and Christian baptisms were celebrated in various Western Churches on Epiphany would undoubtedly necessitate some kind of period of preparation for baptism and, as we saw in our chapter on Lent, a three-week period is a strong candidate for this. Since Himerius claims to know that some churches in Spain celebrated baptism on the new 25 December Christmas, is it not reasonable to suspect that some kind of similar preparation period for Christmas baptism must also have been in existence? Is it possible that the three-week period specified by the Council of Saragossa was used for both occasions? In either case, such a period of three weeks of preparation before baptism here would square with our suggestion above in the chapter on Lent that three weeks of baptismal preparation may well have been a rather free-floating period of preparation tied to whenever baptisms occurred in the life of a church.
Some scholars have suggested that there are additional references to an Advent-type of preparation period oriented towards Epiphany in other non-Roman Western liturgical sources. In Sicily in the middle of the fifth century we have evidence from a letter of Pope Leo I that Epiphany was a regular occasion for baptism there, a custom he tried to discourage in favour of Easter and/or Pentecost, but, unfortunately, he makes no reference to any kind of preparation period.19 More importantly, in late-fifth-century Gaul Bishop Perpetuus of Tours (d. 490) legislated that from the feast day of his predecessor, St Martin of Tours (11 November), until Christmas, a period of seven weeks or 40 days, fasting was to take place on three days each week.20 Josef Jungmann argued that this legislation was based on an earlier ‘St Martin’s Lent’, a 40-day period of preparation for Epiphany baptism, with five days of fasting each week from St Martin’s day until Epiphany, resulting in exactly 40 fasting days. Such, he believed, was designed to provide a parallel to the season of Lent before Easter baptism.21 But while this may have been the case, there is really no hard evidence in support of it, and what we see instead in later Gallican sources, such as Gregory of Tours’ History of the Franks and the Council of Mâcon (582), is that the parallel to a pre-paschal Lent, that is, this 40-day season, strongly ascetic in character, culminates at Christmas and not at Epiphany.22
If a three-week or 40-day preparation for Epiphany baptism may have played some role in what ultimately became the Western season of Advent, another tradition more directly related to the nativity theme must be taken into account. Here, in particular, the evidence adduced by Connell for Advent in northern Italian sources is paramount.23 As we saw in the previous chapter, Epiphany as celebrating the nativity of Christ in Bethlehem, together with other themes, appears to be the earlier tradition in northern Italy, and there may be vestiges of this in the continued enrolment of candidates for baptism on Epiphany in Milan and Turin. But the 25 December Christmas itself is not clearly known in northern Italy until the late fourth century in Brescia, where Bishop Filastrius (d. 391) in his Diversarum hereseon liber notes that there are four fasts the Church celebrates during the year: at Jesus’ birth, Easter, Ascension and Pentecost.24 Since Christmas was apparently a new feast in northern Italy, Connell suggests, we cannot know if the fast at Jesus’ birth was a practice that had merely shifted from before Epiphany to before Christmas or if it was a newly established fast altogether. In roughly the same time period, that is, the end of the fourth century, Maximus of Turin witnesses to what may have been two Sundays of preparation for Christmas, a practice confirmed in mid-fifth-century Ravenna, where the evidence suggests that Peter Chrysologus preached on the annunciation to Zechariah (Luke 1.5–25) two Sundays before Christmas and the annunciation to Mary (Luke 1.26–38) on the Sunday before Christmas, a practice highly consistent with various Eastern Christian pre-Christmas traditions. While the precise length of ‘Advent’ cannot be known from these sources, it is well known that eventually at Milan, in what is called the Ambrosian Rite, as well as in Spain in the Mozarabic Rite, Advent will become a period of six weeks before Christmas, hence, similar in length to the Gallican 40-day period of ‘St Martin’s Lent’, which also lasted into the Middle Ages in Gaul. Theologically as well, if Spain and Gaul contributed to the emerging Advent what might be called an ascetic or more penitential (baptismal preparation?) character, the northern Italian sources suggested a more scriptural or incarnational focus.
A six-week Advent was also practised at Rome, although our evidence for this comes only from those liturgical sources for Rome known to be pre-Gregory I (590–604) in content, that is, the Würzburg Capitulary and the old Gelasian Sacramentary. As with Lent at Rome, it is again the work of Chavasse, based on these sources, that shows us that in the six weeks before Christmas, the first five Sundays were explicitly now called de adventu, while the Sunday immediately before Christmas was designated Dominica vacat, owing to the fact that the previous Wednesday through the Saturday vigil Mass were part of what were called the Ember Days, a purely Roman phenomenon.25 It has been suggested that the fast of the tenth month (the other quarterly or quattember fasts being in summer, autumn and spring) may have had something to do with the origins of Advent at Rome, especially when it is recalled that the annunciation to Mary (Luke 1.26–39) and the visitation of Mary to Elizabeth (Luke 1.39–47) appear as the Gospel readings for the Wednesday and Friday of this week in the sources.26
What happens at Rome at the end of the sixth century is that the six-week Advent received by Gregory I is shortened by him to four weeks in length. Connell suggests that the reason for this is that Gregory did not know the baptismal and/or Epiphany imagery associated elsewhere in the West with a six-week or 40-day period,27 though this cannot be known with any degree of certainty. At the same time, Alexander draws attention to the fact that in the pertinent liturgical sources for Advent at Rome, the liturgical year begins with the vigil Mass of Christmas, but ends with what comes to be known as Advent. That the year would begin with Epiphany or with Christmas is surely no surprise, given what we have seen in the previous chapters of this section. Hence, if in Spain and Gaul, the pre-Christmas season comes to have an ascetic or penitential focus, perhaps, as suggested by Adolf Adam, under the influence of Irish monasticism especially in Gaul,28 and if in northern Italy it is the coming nativity of Christ, announced by the Scripture readings preceding it, the very location of ‘Advent’ at Rome as the final period of the year, together with the Gospel pericopes focusing on the parousia or Second Coming of Christ, suggest an eschatological orientation to the season. Alexander explains this as indicating that ‘Advent at Rome may have first been conceived of as a period before Christmas in its integrity, rather than as a pre-Christmas fast or season of ascetically focused preparation for the celebration of the Nativity.’29
What will happen is that the Roman Advent will eventually become imported elsewhere into Western Christianity, ultimately replacing the various local traditions, although the Ambrosian and Mozarabic Rites will retain, even today, their customary six-week practice. And Advent itself will become in time a mixture of the various biblical, ascetic and eschatological orientations we have noticed. Focused as it is, somewhat ambiguously, on the ‘Advent’ or coming of Christ, without specifying which ‘coming’ is intended, Bernard of Clairvaux (d. 1153) well summarizes the theology of the final form of the Roman four-week Advent as orientated to the three comings of Christ: past, present and future:
We know that there are three comings of the Lord. The third lies between the other two. It is invisible, while the other two are visible. In the first coming he was seen on earth, dwelling among men; he himself testifies that they saw him and hated him. In the final coming all flesh will see the salvation of our God, and they will look on him whom they pierced. The intermediate coming is a hidden one; in it only the elect see the Lord within their own selves, and they are saved. In his first coming our Lord came in our flesh and in our weakness; in this middle coming he comes in spirit and in power; in the final coming he will be seen in glory and majesty … Because this [middle] coming lies between the other two, it is like a road on which we travel from the first coming to the last. In the first, Christ was our redemption; in the last, he will appear as our life; in this middle coming, he is our rest and consolation.30
Regarding its origins, however, we can do no better than the following words of Connell:
[W]e do know that the Epiphany was being celebrated in Gaul, in Spain, in Northern Italy, and in the East long before Christmas was received in those places. And we know that in most … of those non-Roman western churches the nativity and baptism of Jesus were among the epiphanies marked in the celebration … We also know that the enrollment of catechumens was celebrated on Epiphany in Milan (and perhaps elsewhere), and … we read Leo the Great’s advice against baptisms on Epiphany when he wrote to the bishop of Sicily. So surely some of those southern Italian communities were initiating on January 6. Might it be, then, that the forty-day span prominent later on in Gaul carried a remnant of the baptismal span preceding Epiphany in earlier times? … Might the forty days have been linked not to Easter as is commonly assumed in liturgical scholarship, but to baptism whenever this took place? Perhaps only later did the introduction of Christmas, broadly coincident with the decline of the catechumenate, join with the forty-day span to become a preparatory period no longer for baptism but for the relatively new date in the calendar, December 25.31
If Connell is correct here, an earlier understanding and orientation may still appear as traces or remnants in the extant sources and, as such, may help in the assessment that outside of Rome it was especially a pre-Epiphany baptismal preparation period that functioned as a kind of nascent Advent. As with Lent, then, it is possible that ‘Advent’ also owes its origins, in part at least, to various baptismal preparation periods. What this also means is that the Roman Advent as it develops is somewhat of an anomaly in comparison with the other liturgical traditions, but an anomaly that would be consistent with other liturgical traditions indigenous to Rome, though sometimes copied elsewhere (e.g., the development of the sacrament of confirmation, and the unique structure and contents of the Roman eucharistic prayer or canon missae). And since at Rome the origins of Advent are rather late and deliberately, it seems, shortened in relationship to what other churches are doing, it becomes difficult to understand its meaning and function vis-à-vis the 25 December Christmas at all. Hence, we cannot but agree also with Alexander that the strong eschatological character of Roman Advent and its actual location within the classic Roman liturgical books at the end of the year has more to do with the conclusion of the year than with its annual beginning, contrary to our current liturgical reckoning. Historically, its proximity to Christmas, therefore, would have been more accidental than deliberate in Rome. Of course, if Christmas itself is perceived not as the celebration of Jesus’ nativity in illo tempore, that is, as a commemoration of a historical event or as Baby Jesus’ birthday, but as itself a celebration of his parousia, his Advent, his coming again in glory, then an eschatologically oriented Advent season makes perfect sense.
–––––––––––––––––––––
1 Martin Connell, ‘The Origins and Evolution of Advent in the West’ in Maxwell E. Johnson (ed.), Between Memory and Hope: Readings on the Liturgical Year (Collegeville: The Liturgical Press 2000), pp. 349–71. See also Martin Connell, Eternity Today: On the Liturgical Year 1 (New York/London: Continuum 2006), pp. 59–74.
2 See Mother Mary and Archimandrite Kallistos Ware, The Festal Menaion (London/Boston: Faber and Faber 1969), pp. 164–98.
3 Pierre Jounel, ‘The Year’ in A. G. Martimort et al. (eds), The Church at Prayer 4 (Collegeville: The Liturgical Press 1986), p. 93.
4 See below, p. 210.
5 Conferencia Episcopal Española, Missale Hispano-Mozarabicum (Barcelona 1994), pp. 34, 136–42.
6 On Ember Days, see Thomas J. Talley, ‘The Origins of the Ember Days: An Inconclusive Postscript’ in Paul De Clerck and Eric Palazzo (eds), Rituels: Mélanges offerts à Pierre-Marie Gy, OP (Paris: Cerf 1990).
7 A. Wilmart, ‘Le prétendu Liber Officiorum de St. Hilaire et l’Avent liturgique’, Revue Bénédictine 27 (1910), pp. 500–13.
8 ET from Connell, ‘The Origins and Evolution of Advent in the West’, p. 363.
9 See Thomas J. Talley, The Origins of the Liturgical Year (New York: Pueblo 1986; 2nd edn, Collegeville: The Liturgical Press 1991), pp. 147ff.; J. Neil Alexander, Waiting for the Coming: The Liturgical Meaning of Advent, Christmas, Epiphany (Washington, DC: The Pastoral Press 1993), pp. 8–23.
10 See J. P. Brisson, Hilaire de Poitiers, Traité des mystères, Sources chrétiennes 19 (Paris: Cerf 1947), pp. 64–8, 164; Jounel, ‘The Year’, p. 91.
11 Talley, The Origins of the Liturgical Year, p. 150.
12 Alexander, Waiting for the Coming, pp. 10–11.
13 Connell, Eternity Today 1, pp. 67–9.
14 See above, pp. 92–8.
15 DBL, p. 158.
16 Ep. Ad Himerium (PL 13:1154–5).
17 Alexander, Waiting for the Coming, p. 11.
18 Alexander, Waiting for the Coming, pp. 11–12.
19 Ep. 16 (PL 54:699–702).
20 See Connell, Eternity Today 1, pp. 70–1.
21 Josef Jungmann, ‘Advent und Voradvent; Überreste des gallischen Advents in der römischen Liturgie’ in Josef Jungmann, Gewordene Liturgie: Studien und Durchblicke (Innsbruck: Rauch 1941), pp. 237–49.
22 Connell, Eternity Today 1, p. 71.
23 Connell, ‘The Origins and Evolution of Advent’, pp. 353–63; Connell, Eternity Today 1, pp. 59–66.
24 See Connell, Eternity Today 1, p. 60.
25 Antoine Chavasse, ‘L’Avent romain du VIe au VIIIe siècle’, Ephemerides Liturgicae 67 (1953), pp. 37–52. On Ember Days, see further above, p. 159.
26 Theodor Klauser (ed.), Das Romische Capitulare evangeliorum: Texte und Untersuchungen zu seiner altesten Geschichte, Liturgiegeschichtliche Quellen und Forschungen 28 (Münster: Aschendorff 1935), 1, p. 127.
27 Connell, Eternity Today 1, pp. 73–4.
28 Adolph Adam, The Liturgical Year: Its History and Meaning after the Reform of the Liturgy (New York: Pueblo 1981), p. 132.
29 Alexander, Waiting for the Coming, p. 21.
30 Bernard of Clairvaux, Sermo 5, In Adventu Domini 1–3; ET from The Liturgy of the Hours 1 (New York: Catholic Book Publishing House 1975), p. 169.
31 Connell, ‘The Origins and Evolution of Advent’, pp. 369–70.