15 25 December: two competing theories

The earliest firm evidence for the Christian celebration of 25 December occurs in a document known as the Philocalian Calendar or Chronograph of 354, which contains a collection of both civil and religious chronologies, among them lists of consuls of the city of Rome up to 354, of Roman bishops from 255 to 352 arranged on an annual cycle in the order of the dates on which they died, and of the anniversaries of martyrs similarly organized.1 The list of martyrs begins with the notation of the birth of Christ on 25 December, while the list of consuls includes not just this birth date, but that it was on a Friday, the fifteenth day of the new moon. Because the list of Roman bishops ends with the two most recent bishops out of sequence, it is generally agreed that this list was originally compiled in the year 336, prior to these additions, and therefore that the presumed date of Christ’s birth was being celebrated as a festival in the city by that time.2 But why was that particular date chosen?

The Chronograph could be thought to imply that the date had been arrived at by calculation, and, as we shall see later, for more than a century prior to this attempts had been made by individuals in various places to establish the precise date both of Jesus’ death and of his birth – but there is no trace extant of anyone having previously suggested 25 December as a possible date for the birth, as the one apparent exception to this, a statement in the Commentary on Daniel by Hippolytus, is regularly held by scholars to be a later interpolation into the work.3 On the other hand, the Julian calendar observed throughout the Roman Empire decreed that 25 December was the date of the annual winter solstice prior to the year 325, when the Council of Nicaea adopted instead the true date, 21 December; and in the year 274 the Emperor Aurelian restored the cult of Sol Invictus, the Invincible Sun, proclaiming this divinity as the single official divine protector of the empire and of the emperor and establishing the yearly festival of the Dies Solis Invicti at the time of the winter solstice, to be observed with appropriate civic celebration, including 30 chariot races.4 It is not surprising, therefore, that in searching for the reason for the celebration of Christmas on this same date Christian scholars have looked to this pagan feast as a possible influence.

Although some earlier scholars had hinted at this connection, the first to present a substantial case for it seems to have been Hermann Usener in 1889.5 While some of his supporting arguments were seriously disputed by his critics, his basic thesis – that the Christian celebration of Christ’s birth had been introduced in order to supplant the pagan festivities on that date – received approbation from a considerable number of later scholars.6 The most significant contribution came from Bernard Botte in 1932, and his work continues to be cited as definitive down to the present day. Among the points that he made was a convincing argument that, contrary to what had generally been believed up till then, the celebration of 6 January had not existed at Rome prior to the adoption of 25 December but was a later addition there, as it did not appear anywhere in the Chronograph. He was also careful to say that the pagan feast had influenced the choice of date and not that Christians had adopted the feast, as some earlier writers had been inclined to do. On the contrary, it was intended as a counter-attraction to the pagan practice.7

This ‘History of Religions’ hypothesis (as it came to be known) did not, however, go unchallenged. Indeed, in the same year as Usener, Louis Duchesne published the first edition of his Origines du culte chrétien, in which he claimed that the theory failed to account for the existence of a feast on 6 January, whereas his own explanation (later termed the Calculation or Computation hypothesis) accounted for both dates. He cited several ancient authors who alleged that 25 March had been the date of Christ’s death, and so he asserted, though without supporting evidence, that Christ must have been thought to have lived for a whole number of years, because symbolic number systems do not allow the imperfection of fractions, and therefore the annunciation must have been thought also to have occurred on 25 March and the nativity nine months later on 25 December.8 As we shall see in the next chapter, he made a similar claim for 6 January in relation to those who dated Christ’s death on 6 April.9 However, he was forced to admit that

this explanation would be the more readily received if we could find it fully stated in some author. Unfortunately we know of no text containing it, and we are therefore compelled to put it forth as an hypothesis, but it is an hypothesis which falls in with what we may call the recognized methods in such matters.10

While his theory found some favour with some, it never succeeded in converting the majority of scholars away from the History of Religions hypothesis.

An attempt to revive Duchesne’s hypothesis was made by Hieronymus Engberding in 1952. Perhaps his most significant contribution to the debate was to find support for it in a Latin tractate, De solstitiis et aequinoctis, a critical edition of which had already been appended to Botte’s seminal work.11 Once believed to have been a sermon by John Chrysostom, it is now generally thought to have originated in the early fourth century and in places shows signs of both African and Syriac influence. Because it noted that the conception and birth of Christ and the conception and birth of John the Baptist had taken place at the four cardinal points of the year, it had been cited by Botte as supportive of the History of Religions hypothesis, and he dismissed the fact that it also remarked upon the coincidence of the conception and the death of Christ on 25 March as being not pivotal to its argument.12 Engberding thought otherwise, and claimed that the dates assigned to these events were independent of and preceded any liturgical observance of them.13 In a review of his article, Botte rejected this argument as very weak.14 Some years later August Strobel attempted to strengthen the Calculation hypothesis by pointing to rabbinic belief that the patriarchs had lived for an exact number of years15 – but this still failed to provide any more evidence as why the dates of death and conception rather than birth might have been thought to have been identical in the case of Christ.

It was Talley who in his 1986 book did most to try to breathe new life into Duchesne’s theory. He added the Talmud tractate Rosh Hashanah to the testimony already adduced by Strobel with regard to rabbinic belief about the lives of the patriarchs, and he cited De solstitiis as giving ‘full substantiation to Duchesne’s hypothesis’.16 Moreover, drawing on an observation made many years earlier by Gottfried Brunner and others,17 he pointed out that Augustine, in one of his sermons, alluded to the fact that the Donatists in North Africa, unlike the Catholics, had not adopted the celebration of the feast of the Epiphany on 6 January, which seemed to imply that they did celebrate 25 December. This in turn suggested that Christmas must already have existed prior to the Donatist schism in 311, and hence at a date when it would have been unlikely that the Christians would have wanted any ‘accommodation to less than friendly imperial religious sentiment’.18 He also noted that Leonard Fendt and others had consequently raised the possibility that the celebration of Christmas may have appeared first in North Africa rather than at Rome, and he tentatively supported the idea, but he did not address Fendt’s concern about the reliance that could be placed on what was essentially an argument from silence with regard to the Donatist observance of 25 December – a concern that others too have echoed19 – nor did he manage to produce any new evidence for the alleged identification among early Christians of the dates of Christ’s conception and death, which is perhaps the weakest point in the Calculation hypothesis.

The case thus remains unproven one way or the other. As Susan Roll has observed, Germanic and Romance language scholars have on the whole tended to lean in the direction of the History of Religions hypothesis, with Anglo-Saxon writers tending to favour the Calculation theory instead.20 Yet whatever their preference, they have traditionally assumed that the feasts on 25 December and 6 January must have developed in parallel, and that whatever was the cause of the emergence of the one must also have been responsible for the development of the other. But there seems no reason why this should necessarily have been so, especially as that on 6 January appears to have begun to be celebrated in other places a considerable period of time before we hear of 25 December being adopted at Rome. Could it perhaps have been that it was the absence at Rome of any feast comparable to that being kept elsewhere on 6 January that led to the emergence of Christmas there in the fourth century, with the date being chosen as a counter-attraction to the pagan festivities taking place then, regardless of whatever may have been the original motivation behind the choice of 6 January elsewhere?

Subsequently, of course, Christmas did spread from its root in Rome to other parts of the ancient world, but when that began to happen is a subject that has provoked nearly as much controversy as that surrounding the origins of the feast. First, there is the question of when it began to be celebrated in other parts of the West. Whatever weight may be given to Talley’s argument from silence concerning the Donatist observance of Christmas prior to 311, the earliest unquestionable testimony to its celebration outside Rome comes from a sermon delivered at the feast by Optatus, Bishop of Milevis in North Africa, probably around 361–3.21 This speaks of the nativity of Christ as being a sacramentum, thus bestowing on it a greater status than Augustine will grant to it at the end of the century, when he distinguishes Christmas as a mere commemoration (memoria) from Easter as a sacramentum: ‘A celebration of something is a sacrament only when the commemoration of the event becomes such that it is understood also to signify something that is to be received as sacred.’22 On the other hand, Optatus’ vocabulary is in line with that of Leo the Great at Rome in the fifth century.23

As for northern Italy, to which one might have expected the celebration of 25 December also to have migrated quite quickly, the evidence for its early adoption is much less firm. In the course of the history of the controversy over the origin of Christmas, considerable use was made of a passage in Ambrose of Milan’s De virginitate, written in 378, in which he recalled the occasion over 20 years earlier (in 353 or 354) on which his sister Marcellina dedicated herself to virginity before Liberius, Bishop of Rome. Ambrose had said that ‘you marked your profession by a change of clothing in the Church of St Peter on the birthday of the Saviour’, and then claimed to be citing an extract from the sermon delivered by Liberius on that day which referred to the miracles at the wedding at Cana and at the feeding of the multitude.24 This had been seized upon by those scholars who had argued that Epiphany had been celebrated at Rome as the birth of Christ prior to the adoption of Christmas, because the biblical events mentioned in the sermon were Epiphany themes. In 1923 Thomas Michels had put forward an alternative explanation – that Ambrose’s supposed recollection was actually shaped by the practice with which he was familiar in Milan in his own day, where the birth of Christ was still being celebrated on 6 January and not 25 December25 – an explanation also advanced more recently by Martin Connell, who supports his claim by noting that Christmas is never mentioned in Ambrose’s commentary on the Gospel of Luke but surprisingly the Magi from Matthew’s Gospel feature prominently.26 He thus argues that the earliest sure testimony to the celebration of 25 December anywhere in northern Italy comes instead from a contemporary of Ambrose, Filastrius, in Brescia around 383,27 although Roll still maintains that several of the hymns composed by Ambrose point to the existence of the Christmas feast at Milan in his day, and especially Intende qui regis Israel, which was cited by Pope Celestine in 430 as having been ordered to be sung at Christmas by Ambrose.28 Connell also rightly challenges as inconclusive sources that have been cited in order to demonstrate the existence of Christmas in Spain in the 380s.29

If we turn to the emergence of the 25 December feast in the East, several sermons preached by John Chrysostom in the year 386 seem to indicate that its adoption in Antioch was a relatively recent development, but just how recent has been the subject of some debate by scholars. In one of the sermons, Chrysostom attempts to promote the feast’s observance, mentioning that the date of Christ’s birth has been known for less than ten years. Some would conclude from this that the feast had already been in existence for that length of time, but others argue that because his Pentecost sermon in the same year had mentioned only three festivals – Theophany (Epiphany), Pascha and Pentecost – and describes the first of these as being the one on which ‘God has appeared on earth and lived with men’, Christmas can only have been celebrated there for the very first time at the end of that year.30 Whatever the case at Antioch, however, sermons by Gregory of Nazianzus preached at Constantinople in 380–1 indicate that 25 December was already being observed there at that time, but whether the feast had been newly introduced by Gregory or had existed for some years before has again been disputed by scholars.31 It all depends on whether Gregory’s designation of himself as ζαρχός of Christmas in Constantinople means ‘originator’ of the feast here or simply the one who presided over it. While its appearance in Cappadocia also seems to belong to the same time period, it remained unknown both in Jerusalem and in Egypt until the fifth century, and was not adopted at all in Armenia.32

Thus, the claim often made that the observance of Christmas spread with extraordinary rapidity to nearly all parts of the ancient Christian world appears not to be strongly supported by the evidence. On the contrary, there seems to have been a gap of at least 40 years between the earliest witness to its adoption in Rome and the very first signs of its challenge to the dominance of 6 January in other churches that were already keeping that feast day. Similarly, the common assertion that the speed of its appropriation was due at least in part to the value that a celebration of the divine incarnation would have had to the Nicene party against their Arian opponents is also open to question. In her survey of the relevant literature Roll has pointed out the essential Christological ambiguity of the feast, and Connell has gone further and suggested that the apparent delay in and resistance to its widespread adoption may have been precisely because the narrative of the vulnerable infant in a manger would not have helped promote the high Christology of the Son as ‘one in being with the Father’ but rather have been more congenial to the Arian cause.33 Whatever may be the truth of that, it is perhaps worth noting that the liturgical readings in later Roman sources encompass not just a commemoration of the event of Christ’s nativity but a celebration of his Incarnation, with John 1 being just as deeply embedded as the Lucan account of the nativity.34

–––––––––––––––––––––

1 For the Latin text, see Theodor Mommsen, ‘Chronographus anni CCCLIIII’ in Monumenta Germaniae Historica, Auctorum Antiquissimorum 9/1 (Berlin 1892 = Munich 1982), pp. 13–148. For the lists of martyrs and bishops, see below, pp. 176, 190.

2 See further Susan K. Roll, Toward the Origins of Christmas, Liturgia condenda 5 (Kampen, The Netherlands: Kok Pharos 1995), pp. 83–6.

3 Roll, Toward the Origins of Christmas, pp. 79–81.

4 Roll, Toward the Origins of Christmas, pp. 65, 113–14.

5 Hermann Usener, Das Weihnachtsfest (Bonn: Cohen 1889; 2nd edn 1911; 3rd edn 1969).

6 See Roll, Toward the Origins of Christmas, pp. 128–39.

7 Bernard Botte, Les Origines de la Noël et de l’Épiphanie, Textes et Études liturgiques 1 (Louvain: Abbaye de Mont César 1932), esp. pp. 54, 62.

8 Louis Duchesne, Origines du culte chrétien (Paris: Thorin 1889), pp. 247–54.

9 See below, pp. 135–6.

10 ET (from the third French edition): Christian Worship: Its Origins and Evolution (London: SPCK 1903), pp. 263–4.

11 Botte, Les Origines de la Noël et de l’Épiphanie, pp. 88–105.

12 Botte, Les Origines de la Noël et de l’Épiphanie, p. 92.

13 Hieronymus Engberding, ‘Der 25. Dezember als Tag der Feier der Geburt des Herrn’, ALW 2 (1952), pp. 25–43, here at p. 34.

14 Bulletin de théologie ancienne et médiévale 7 (1955), pp. 198–9.

15 August Strobel, ‘Jahrespunkt-Spekulation und frühchristliches Festjahr’, Theologische Literaturzeitung 87 (1962), pp. 183–94, here at p. 193; August Strobel, Ursprung und Geschichte des frühchristlichen Osterkalendars, Texte und Untersuchungen 121 (Berlin: Akademie-Verlag 1977), pp. 128–33.

16 Thomas J. Talley, The Origins of the Liturgical Year (New York: Pueblo 1986; 2nd edn, Collegeville: The Liturgical Press 1991), pp. 81–3, 91–5.

17 Gottfried Brunner, ‘Arnobius ein Zeuge gegen das Weihnachtsfest?’, Jahrbuch für Liturgiewissenschaft 13 (1933), pp. 178–81.

18 Talley, The Origins of the Liturgical Year, pp. 86–7, 89–90.

19 Talley, The Origins of the Liturgical Year, pp. 87, 103; Leonard Fendt, ‘Der heutige Stand der Forschung über das Geburtsfest Jesu am 25.12 und über Epiphanie’, Theologische Literaturzeitung 78 (1953), columns 1–10, here at 4, picking up a point expressed by Hieronymus Frank, ‘Frühgeschichte und Ursprung des Römischen Weihnachtsfestes im Lichte neuerer Forschung’, ALW 2 (1952), pp. 1–24, here at pp. 14–15; Martin F. Connell, ‘Did Ambrose’s Sister become a Virgin on December 25 or January 6? The Earliest Western Evidence for Christmas and Epiphany outside Rome’, SL 29 (1999), pp. 145–58, here at pp. 153–5.

20 Roll, Toward the Origins of Christmas, pp. 96, 147–8. A recent exception to that general trend is Hans Förster, Die Feier der Geburt Christi in der Alten Kirche, Studien und Texte zu Antike und Christentum 4 (Tübingen: Mohr Siebeck 2000); Hans Förster, Die Anfänge von Weihnachten und Epiphanias, Studien und Texte zu Antike und Christentum 46 (Tübingen: Mohr Siebeck 2007).

21 See Roll, Toward the Origins of Christmas, pp. 195–6.

22 Augustine, Ep. 55.2.

23 Leo, Sermones 26.1, 4; 27.6. See also Roll, Toward the Origins of Christmas, pp. 212–14.

24 Ambrose of Milan, De virginitate 3.1.1.

25 Thomas Michels, ‘Noch einmal die Ansprache des Papstes Liberius bei Ambrosius, de virg. III 1,1ff’, Jahrbuch für Liturgiewissenschaft 3 (1923), pp. 105–8.

26 Connell, ‘Did Ambrose’s Sister become a Virgin on December 25 or January 6?’, pp. 146–9.

27 Connell, ‘Did Ambrose’s Sister become a Virgin on December 25 or January 6?’, pp. 151–2.

28 Roll, Toward the Origins of Christmas, pp. 200–3.

29 Connell, ‘Did Ambrose’s Sister become a Virgin on December 25 or January 6?’, pp. 155–7.

30 For further details, see Talley, The Origins of the Liturgical Year, pp. 135–7; Förster, Die Anfänge von Weihnachten und Epiphanias, pp. 166–79; J. N. D. Kelly, Golden Mouth: The Story of John Chrysostom – Ascetic, Preacher, Bishop (London: Duckworth/Ithaca, NY: Cornell 1995), p. 70.

31 Talley, The Origins of the Liturgical Year, pp. 137–8; Förster, Die Anfänge von Weihnachten und Epiphanias, pp. 182–98.

32 Talley, The Origins of the Liturgical Year, pp. 138–41.

33 Roll, Toward the Origins of Christmas, pp.168–89, esp. pp. 174–7; Martin F. Connell, Eternity Today: On the Liturgical Year 1 (New York/London: Continuum 2006), pp. 101–3.

34 See Lester Ruth, ‘The Early Roman Christmas Gospel. Magi, Manger, or Verbum Factum?’, SL 24 (1994), pp. 214–21.