The fifth-century Byzantine historian Socrates describes his understanding of the variety of Lenten observances throughout the Christian churches of his day:
The fasts before Easter will be found to be differently observed among different people. Those at Rome fast three successive weeks before Easter, excepting Saturdays and Sundays. Those in Illyrica and all over Greece and Alexandria observe a fast of six weeks, which they term ‘the forty days’ fast’. Others commencing their fast from the seventh week before Easter, and fasting three five days only, and that at intervals, yet call that time ‘the forty days’ fast’. It is indeed surprising to me that thus differing in the number of days, they should both give it one common appellation; but some assign one reason for it, and others another, according to their several fancies.1
What is most intriguing about Socrates’ statement is his reference to a three-week Lenten fast at Rome. Since he corrects himself about Saturdays as non-fasting days in Rome later in this work and since Athanasius (in his Festal Letter of 3402), Jerome (in a letter to Marcella in 3843) and Pope Siricius (in a letter to Himerius of Tarragona in 3854) refer to an established pattern of a 40-day Lent there too, his statement is inaccurate as a fifth-century description. Nevertheless, his reference to ‘three successive weeks’ of fasting appears to be corroborated by later sources of the Roman liturgy. Such evidence includes:
• the provision of three missae pro scrutiniis (masses for the scrutinies of baptismal candidates) assigned to the third, fourth and fifth Sundays of Lent in the Gelasian Sacramentary (seventh century);
• the course reading of the Gospel of John during the last three weeks of Lent (beginning in the Würzburg Capitulary, the earliest Roman lectionary, c. 700, on the Friday before the third Sunday in Lent and reaching its conclusion on Good Friday); and
• the titles Hebdomada in mediana (week in the middle) and Dominica in mediana (Sunday in the middle), applied, respectively, to the fourth week and fifth Sunday of Lent in various ordines Romani (ceremonial and rubrical guides) and Roman lectionaries.
In the light of all this, Socrates’ inaccurate fifth-century description may well indicate the remnant of a well-ingrained three-week Lenten period in Rome some time earlier. Such, at least, was the conclusion of Antoine Chavasse from his analysis of the Johannine readings of the last three weeks of Lent,5 which he was able to reconstruct as an independent set of lections that he believed must once have constituted an original three-week Lenten period, including Holy Week. Chavasse noted that the series of Johannine readings during the last three weeks of Lent in early Roman lectionaries and in the Tridentine Missale Romanum began with John 4.5–32 on the Friday of Lent III. For some reason, however, it placed John 9.1–38 (Wednesday of Lent IV) and John 11.1–45 (Friday of Lent IV) before John 8.46–59 (Sunday of Lent V), and John 10.22–38 (Wednesday of Lent V) with the continuation of John 11 (47–54) on the Friday of Lent V. On this basis he attempted to reconstruct an earlier shape for this Johannine series, which he believed would have corresponded to the three missae pro scrutiniis in the Gelasian Sacramentary. According to his reconstruction, John 4.5–32, John 9.1–38 and John 11.1–54 would have been read, respectively, on the third, fourth and fifth Sundays in Lent in the time of Leo the Great. Even so, at an earlier stage of development this would have constituted a short lectionary series for the Sundays of an original three-week Lenten period, including Holy Week. The reason that this series of readings appears in a different sequence in later Roman sources, according to Chavasse, is that the baptismal scrutinies along with their readings became shifted to weekdays (ultimately, seven in number) in the later Roman tradition.6 Along similar lines, Talley also concluded that Socrates’ reference may reflect an earlier, if not fifth-century, Roman practice.7
The possibility of an original three-week Lent is not limited to Rome. On the basis of a detailed structural analysis of the contents of the fifth-century Armenian Lectionary, a lectionary generally understood to reflect fourth-century Jerusalem practice, Mario F. Lages argued that early Jerusalem practice, too, knew an original three-week Lenten preparation period of catechumens for paschal baptism.8 This lectionary includes a canon of Lenten readings with concluding psalmody assigned to Wednesday and Friday gatherings at Sion and a list of 19 catechetical biblical readings assigned to Lenten catechesis, which parallel the pre-baptismal catecheses of Cyril of Jerusalem. Lages also pointed to the introductory rubric in the ninth- or tenth-century Armenian rite of baptism and to a pertinent rubric in the fifth-century Georgian Lectionary. The Armenian baptismal rubric reads in part:
The Canon of Baptism when they make a Christian. Before which it is not right to admit him into church. But he shall have hands laid on beforehand, three weeks or more before the baptism, in time sufficient for him to learn from the Wardapet [Instructor] both the faith and the baptism of the Church.9
And the Georgian Lectionary, while listing the same 19 catechetical readings as Cyril and the Armenian Lectionary, specifically directs that catechesis is to begin with these readings on the Monday of the fifth week in Lent, that is, exactly 19 days (or approximately three weeks) before paschal baptism.10
That is not all: this early three-week Lenten period in Rome and Jerusalem was customary in other liturgical traditions as well. A similar three-week period of final preparation for baptismal candidates is discernible from an analysis of the last three weeks of the 40-day Lent in North Africa, Naples, Constantinople and Spain.11 For Spain, in particular, this three-week period appears to be confirmed by the first canon of the Second Council of Braga (572), which directs 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 …’.12 And, at Constantinople, the extant typika of the ninth- and tenth-century Byzantine Liturgy specify that no one might enter the pre-baptismal catechumenate any later than three weeks before Lazarus Saturday (the day before Palm Sunday), the day on which the patriarch himself presided at baptism in the little baptistery of Hagia Sophia.13 Still today, Christians of the Byzantine tradition sing a baptismal troparion (entrance hymn) based on Galatians 3.27 on Lazarus Saturday. An almost identical pattern of three weeks of preparation before Palm Sunday appears also in the early medieval sources for the Ambrosian Rite in Milan.14 What Socrates says about the ‘three successive weeks’ of pre-paschal fasting at Rome, therefore, may well be seen as the memory of an early Christian practice which was much more universal than Roman in its scope.
On the basis of this discernible pattern in Christian liturgical sources, Lawrence Hoffman suggested that this practice has its ultimate roots in Judaism.15 Hoffman notes that, according to rabbinic sources, the feast of Passover itself is preceded by lectionary readings (Exod. 12 or Num. 19) on the third Sabbath prior to its arrival that stress either preparation for the Passover sacrifice or the necessity of being cleansed from impurity. The Exodus 12 reading, he notes further, was cited by Chavasse as an early reading for Good Friday at Rome and the prophetic reading of Ezekiel 36.25–36 (accompanying Num. 19 according to the Tosefta) appears on the Wednesday of Lent IV in early Roman lectionaries, that is, two and a half weeks before Easter. According to Hoffman, therefore, the early three-week Lent – at least in Jerusalem and Rome – was ‘a Christian application of Judaism’s insistence that one count back three weeks from Passover in order to cleanse oneself and prepare for the sacrifice of the Paschal lamb’.16 If Hoffman is correct, then, as Talley writes, ‘this could well suggest that the three-week preparation for Pascha antedates its employment as the framework for baptismal preparation’.17
The strength and appeal of Hoffman’s theory are that it appears to provide a firm rationale for the Christian choice of a three-week period of preparation. The problem, however, is that when we first see whatever evidence there is for this three-week ‘Lent’ (with the exception of Socrates’ general reference to fasting), it is already closely associated with the final preparation of catechumens for baptism; and not always clearly associated with Easter baptism.
The Armenian baptismal rubric, for example, stresses three weeks of preparation for baptism without specifying when that baptism is to take place. But the early Syrian and Armenian traditions favoured baptism in relationship to Epiphany, not Easter, since they understood Christian initiation as the mimesis of the Jordan event interpreted in the light of the rebirth imagery of John 3 rather than the paschal imagery of Romans 6.18 The three-week period of preparation was therefore more probably associated with catechumenal preparation for baptism without having anything to do specifically with Easter.19 Similarly, thanks again to the work of Talley, it is now a commonly accepted hypothesis that prior to the post-Nicene context of the fourth century, the Alexandrian tradition knew neither Easter baptism nor a pre-paschal ‘Lent’ longer than the one week of the paschal fast. And, it must be noted, the reference to ‘three weeks’ in the Constantinopolitan liturgy is actually a reference in the typika to the enrolment of baptismal candidates exactly three weeks before the celebration of baptism on Lazarus Saturday (the day before Palm Sunday and a full week before Easter), a day which in current Byzantine usage still contains the vestige of a baptismal liturgy in its entrance antiphon.20
Because of the primary association of this three-week period with baptismal preparation, the real question, therefore, is whether or not this period must necessarily be connected to Easter and, consequently, to a pre-paschal Lent. Talley stated that ‘Pascha was becoming the preferred time for baptism in many parts of the Church’ in the third century,21 but a much different conclusion has been offered in a previous chapter.22 As we have seen, the most that can be said about Easter baptism before the fourth century is that there is a preference expressed for this practice, a preference limited to third-century North Africa (Tertullian) and possibly Rome (Hippolytus’ Commentary on Daniel), with its possible celebration on other days by no means excluded. Only in the post-Nicene context of the fourth century does paschal baptism become a near universal Christian ideal. Even then, however, it does not appear to become the only or dominant custom outside of Rome or northern Italy. The letter of Pope Siricius to Himerius of Tarragona, one of the earliest Roman references to a 40-day Lent, reveals a variety of baptismal occasions in Spain (i.e., Christmas, Epiphany and the feasts of apostles and martyrs). Evidence from Leo I demonstrates that Epiphany was also a baptismal day in Sicily and that the feasts of martyrs were baptismal occasions elsewhere in Italy. And a sermon of Gregory of Nazianzus shows, similarly, that Epiphany baptism was a common practice in Cappadocia. These examples, along with those of Alexandria and Constantinople referred to above, lead to the conclusion that baptism at Easter was never the normative practice in Christian antiquity that many have assumed.
What, then, may be concluded about Socrates’ three weeks and the origins of Lent? As we have seen, it is primarily within the context of final baptismal preparation where references to this three-week period are discerned. But what is most striking is that not all of these sources refer to Easter baptism. We seem therefore to have a three-week period of (final) catechetical preparation for baptism that only later gets associated with Easter. It becomes ‘Lent’ simply because Easter gradually becomes the preferred day for Christian initiation. Whenever baptism occurred, it was preceded, as the Armenian baptismal rubric says, by ‘three weeks or more’ of preparation. For those churches (North Africa and Rome) that ‘preferred’ to celebrate initiation at Easter, we may speak of this three-week period as a kind of primitive ‘Lent’. For those that did not have such an early preference, this three-week period was not ‘Lent’ but merely a final catechetical baptismal preparation for whenever baptism itself was to occur. Only when paschal baptism becomes the normative ideal – in the second half of the fourth century – do these variations become blurred, harmonized, and thus brought into universal conformity as part of the newly developed pre-paschal Quadragesima or Tessarakoste.
–––––––––––––––––––––
1 Socrates, Historia ecclesiastica 5.22; ET from NPNF 2nd Series 2, p. 131.
2 The Festal Epistles of S. Athanasius (Oxford: Parker 1854), p. 100.
3 Ep. 24.4 (PL 22:428).
4 PL 13:1131–47.
5 See Antoine Chavasse, ‘La structure du Carême et les lectures des messes quadragesimales dans la liturgie romaine’, La Maison-Dieu 31 (1952), pp. 76–120; ‘La préparation de la Pâque, à Rome, avant le Ve siècle. Jeûne et organisation liturgique’ in Memorial J. Chaine (Lyon: Facultés catholiques 1950), pp. 61–80; and ‘Temps de préparation à la Pâque, d’après quelques livres liturgiques romains’, Recherches de science religieuse 37 (1950), pp. 125–45. For a more detailed summary and discussion of Chavasse’s work, see Maxwell E. Johnson, ‘From Three Weeks to Forty Days: Baptismal Preparation and the Origins of Lent’, SL 20 (1990).
6 Thanks to the work of Chavasse, this is precisely the sequence of Sunday Gospel readings assigned to the third, fourth and fifth Sundays in Lent in Series A of the current Roman Lectionary. To these Sundays have been attached the three scrutinies of adult catechumens in the current Roman Rite of Christian Initiation of Adults. See also here the recent essay by Dominic Serra, ‘New Observations about the Scrutinies of the Elect in Early Roman Practice’, Worship 80 (2006), pp. 511–27.
7 Thomas J. Talley, The Origins of the Liturgical Year (New York: Pueblo 1986; 2nd edn, Collegeville: The Liturgical Press 1991), p. 167.
8 Mario F. Lages, ‘Étapes de l’évolution de carême à Jérusalem avant le Ve siècle. Essai d’analyse structurale’, Revue des Études Armeniénnes 6 (1969), pp. 67–102; and ‘The Hierosolymitain Origin of the Catechetical Rites in the Armenian Liturgy’, Didaskalia 1 (1967), pp. 233–50. See also Maxwell E. Johnson, ‘Reconciling Cyril and Egeria on the Catechetical Process in Fourth-Century Jerusalem’ in Paul F. Bradshaw (ed.), Essays in Early Eastern Initiation, Alcuin/GROW Joint Liturgical Study 8 (Nottingham: Grove Books 1988), pp. 24–6. For the Armenian Lectionary, see Charles (Athanase) Renoux, Le Codex armenien Jérusalem 121 2, Patrologia Orientalis 36 (Turnhout: Brepols 1971).
9 DBL, p. 74 (emphasis added).
10 Michel Tarschnischvili, Le Grand Lectionnaire de l’Église de Jérusalem 1, Corpus Scriptorum Christianorum Orientalium 188 (Louvain 1959), p. 68.
11 Johnson, ‘From Three Weeks to Forty Days’, pp. 191–3.
12 DBL, p. 158.
13 See above, pp. 85–6.
14 See DBL, pp. 184–5, 198–9.
15 Lawrence A. Hoffman, ‘The Jewish Lectionary, the Great Sabbath, and the Lenten Calendar: Liturgical Links between Christians and Jews in the First Three Christian Centuries’ in J. Neil Alexander (ed.), Time and Community (Washington, DC: The Pastoral Press 1990), pp. 3–20. See also Stéphane Verhelst, ‘Histoire de la durée du Carême à Jérusalem’, QL 84 (2003), pp. 23–50, who argues that the three-week and eight-week Lenten patterns in Jerusalem both have their roots in Judaism, the three-week period having its origins before Pascha, but transferred to before the commemoration of the destruction of the Temple on the 9 of Av (August) with a seven-week period following it. Later, the seven-week period, called the Ninevite Fast, was moved from its associations with the 9 of Av and was located ten weeks before Pascha. Here, as elsewhere, Verhelst’s theory rests upon an assumption that the commemoration of the destruction of the Temple was an annual celebration among the Jerusalem Christians and so important that it actually gave rise to two ‘Lenten’ seasons, an earlier three-week one and a later eight-week one, or a combination thereof. While there may be something to this, it strikes us as too speculative to warrant acceptance.
16 Hoffman, ‘The Jewish Lectionary, the Great Sabbath, and the Lenten Calendar’, p. 14.
17 Talley, The Origins of the Liturgical Year, p. 167.
18 See below, Chapter 16.
19 For further discussion, see Lawrence A. Hoffman and Maxwell E. Johnson, ‘Lent in Perspective: A Summary Dialogue’ in Paul F. Bradshaw and Lawrence A. Hoffman (eds), Passover and Easter: The Symbolic Structuring of Sacred Seasons (Notre Dame: University of Notre Dame Press 1999), pp. 55–70.
20 See Talley, The Origins of the Liturgical Year, pp. 189, 203–14. See also above, p. 95.
21 Talley, The Origins of the Liturgical Year, p. 167.
22 See above, pp. 75–86.