1 RABBINIC, CHRISTIAN, AND LOCAL CALENDARS IN LATE ANTIQUE BABYLONIA: INFLUENCE AND SHARED CULTURE Sacha Stern A variety of calendars were used by Jews, Christians, and others in late antique Babylonian society. Many of these calendars were lunar, and shared common structural features such as months beginning at the new moon, and years of twelve or thirteen lunar months; but the precise way these calendars were reckoned could vary quite considerably. In late Antiquity, moreover, the lunar calendars of Jews, Christians, and others were in a process of gradual but quite radical change. The purpose of this paper is to assess the relationship between rabbinic, Christian, and local lunar calendars in late antique Babylonia. Were they, in fact, one and the same calendar, or did they have separate and distinct identities as Jewish, Christian, and other? Was the evolution of different lunar calendars in this period interrelated – as indeed they seem to follow a similar trajectory – or did these processes occur, in fact, independently of one another? If these processes were interrelated, how do we envisage such a shared history to have unfolded? Before addressing all these questions, I shall begin with some general, background information on ancient calendars, and then with some theoretical musings on the concepts of ‘influence’ and ‘shared culture’ which much of this article will be leading to. I shall then argue, in the body of this article, that the rabbinic, Christian, and other local Babylonian calendars evolved in this period in similar ways and in similar directions, although not necessarily through mutual influence and not necessarily for the same reasons. From the mid first millennium BCE to mid first millennium CE, ancient calendars evolved from flexible, diverse, and empirical methods of time reckoning to unified and standard, fixed schemes. In the Near East, the Achaemenid (sixth – fourth centuries) and Seleucid (third – first centuries BCE) dynasties adopted the Babylonian calendar (assimilated, under the Seleucids, with the Macedonian calendar) as their official imperial calendar. This calendar had always been based on empirical sightings of the new moon, but in this period it was increasingly standardized and fixed through the use of new moon predictions and the adoption of a fixed pattern of intercalations (i.e. adding a thirteenth lunar month every two or three years). The standard, 19-year cycle of intercalations made it easy for the Babylonian calendar to be reckoned uniformly across the vast expanses of the Empire, from the eastern provinces of Bactria (northern Afghanistan) to Elephantine in southern Egypt and Lycia in western Asia Minor, where the Babylonian calendar is known to have been used. Another calendar instituted under the Achaemenids was completely fixed and schematic, and modeled entirely on the Egyptian calendar, with a 2 changeless 365-day year. This calendar, later known as Zoroastrian, became the official calendar of the Persian, Sassanian Empire in late Antiquity (third – sixth centuries CE); its adoption may have been motivated, among other reasons, by a desire to rationalize, standardize, and hence simplify the administration of the Empire. In the Roman Empire, perhaps for similar reasons, the Julian calendar – another fixed, schematic calendar based on a 365-day year, instituted by Julius Caesar in 46 BCE – extended by the end of the first century BCE throughout the Mediterranean basin as a standard imperial calendar, albeit in a variety of forms; it eventually became, as we know, the modern, Gregorian calendar in global use today.1 Small pockets of resistance to the rise of these large-scale, official imperial calendars remained. Particularly relevant to us, in late Antiquity, are the Jews in the Roman Empire, who stubbornly clung to their lunar calendars and refused to adapt them (as most others had done in the Roman Near East) to the Julian, solar year. The Christians, who in the first few centuries were not always clearly distinguishable from Jews, also dated some of their most important festivals, above all Easter, according to a lunar reckoning. In the firstthird centuries CE, Jewish and Christian lunar calendars were flexible and empirically based, and for that reason, could vary significantly from one locality to the next. But in subsequent centuries – the period of ‘late Antiquity’ with which this article will be concerned – the calendars in both traditions remained lunar but followed the general historical trend outlined above, and became increasingly schematic, standardized, and fixed. Similar conditions appear to have also affected the local calendars in Babylonia that had remained lunar.2 In the late antique Near East, moreover, the calendars of Jews, Christians, and all others who used Aramaic language shared in common the same month names, regardless of how their calendars were structured: these were the month names of the standard Babylonian calendar which had been used by the great empires of the first millennium BCE as official calendar, and which had survived in various derivative forms in late Antiquity. All these similarities will raise the questions of whether Jews, Christians, and others (‘pagans’) in Babylonian were under the influence of one another, or whether instead they shared a common culture, or whether again, they were affected by wider, common historical conditions – and further possibilities could also be considered. The phenomenon of cultural transfer and exchange has received much attention in recent years, especially with reference 1 S. Stern, Calendars in Antiquity. Empires, States, and Societies (Oxford: Oxford University Press, 2012); id., ‘The Babylonian month and the new moon: sighting and prediction’, Journal for the History of Astronomy 39 (2008), 19-42. I am grateful to François de Blois for reading this article and providing very helpful comments. 2 Stern 2012 and id., Calendar and Community: A History of the Jewish Calendar, 2nd century BCE – 10th century CE (Oxford: Oxford University Press, 2001). 3 to late antique society; this present volume, indeed, is a contribution to this growing concern. In this context, some modern scholars have repudiated altogether the concept of ‘influence’ in favour of a model of ‘common (or shared) culture’, on the grounds that ‘influence’ assumes a single, unidirectional transfer, usually from dominant to weak, which is over-simplistic and does not satisfactorily reflect the complexities and multiple forms of cultural transfer and exchange in late antique and other societies.3 Indeed, in the context of Sassanian-ruled Babylonia it would be difficult to identify Jews, Christians, or pagans as either dominant or weak, and hence in a privileged position to exert influence over one another; a more egalitarian model of cultural exchange should perhaps be preferred. However, the adoption of shared culture as a basic paradigm is equally reductionist, besides erring excessively on the side of vagueness and indeterminacy. Shared culture is not demonstrably preferable or more true than the almost contrary paradigm of cultural differentiation, which used to be favoured in earlier scholarship with the assumption, for example, of ‘Jews’ and ‘Christians’ as separate and well-defined groups in late antique society. In fact, there is no good reason to blacklist a priori a concept such as ‘influence’, because in certain circumstances there is no doubt that influence can and does occur. It is misleading, moreover, to treat ‘influence’ as the logical opposite, or at least sole counterpart, of ‘shared culture’ (as many scholars, in this field, appear to have assumed). Influence, coercion, mimicry, emulation, borrowing, appropriation, accommodation, trade, exchange, distribution, free access, free sharing, etc., are all possible and legitimate ways for culturemes to be transferred between people or groups of people, and thus to account for resulting similarities between them.4 For example C. Hezser, ‘Interfaces between Rabbinic Literature and Graeco-Roman Philosophy’, in P. Schäfer and C. Hezser (eds), The Talmud Yerushalmi and Graeco-Roman Culture, vol.2 (Tübingen: Mohr-Siebeck, 2000), pp. 161-87: ‘I advise against the positivist search for “influences” when dealing with similarities’ (ibid. 162) – a statement that has often been cited, although Hezser does not clearly explain the basis of her position. A similar but slightly more nuanced position is adopted by D. Boyarin, ‘Hellenism in Jewish Babylonia’, in C. E. Fonrobert and M. S. Jaffee (eds), The Cambridge Companion to the Talmud and Rabbinic Literature (Cambridge: Cambridge University Press, 2007), pp. 336-63: ‘while in this case, I am claiming a particular direction of influence, … this is not the general burden of my thesis here, suggesting as I am, rather, a shared cultural milieu’ (ibid. 337, and further 359 n.5); ‘the time has come, I think to cease thinking in terms of influence and think, rather, of shared and overlapping cultures imbricated on each other and partly simply just the same culture in different variants’ (349). 4 Detailed analysis of this range of different processes has been attempted, most notably, in post-colonial theory, although post-colonial theory tends to be excessively politicized and focused on unequal relations opposing, for example, ‘hegemonies’ to the ‘subalterns’. As a starting point, see H. K. Bhabha, The Location of Culture (London and New York: Routledge, 1994). A more recent discussion of the use of concepts such as ‘influence’ or 3 4 The only point I wish to make – and which will determine the direction of this paper – is that cultural transfer and similarity do not occur solely by virtue of physical or social proximity between different groups and cultures. To illustrate what I mean by an example, the similarity of Jewish and Christian calendars in late antique Babylonia, which will be presented in this article, cannot be accounted simply by virtue of the fact that Jews and Christians lived close to each other or even interacted with each other on a regular basis. Influence, or any other type of culture transfer (I have listed some examples above), does not happen just like that. It occurs in a specific social context (which can be, for example, political or economic domination, conflict and polemic, coexistence and partnership, mobility and migration, etc.), and as a result of concrete and specific events or social processes. It is not sufficient, for the historian, to identify similarities between different cultures or groups and then explain them away with catch-phrases such as ‘influence’ or ‘common culture’. An investigation of the social context and specific social processes which brought about these similarities – to the extent that these processes can unfathomed from the available evidence – is an essential part of what constitutes historical explanation and historical research. The notion of ‘similarity’, in this context, is also problematic, as it is largely subjective and a matter of perspective. For example, in connection with calendars, a fourth-century Christian theologian such as Aphrahat had a very clear idea of the distinction, indeed the almost logical opposition, between the Jewish Passover and the Christian Easter – even though both festivals shared the same origin, and both occurred, in the eastern tradition which Aphrahat most likely followed, on approximately the same dates.5 But a Zoroastrian might have taken a rather different view: for according to a sixth-century Syriac chronicle of unknown authorship, when Asthebid encountered the Roman armies at Callinicum during the last week of Lent in 531 CE, he asked his Roman adversary Belisarius to postpone the battle out of respect for the ‘feast’ (i.e. Easter), ‘for the sake of the Nazarenes and the Jews who are in the army that is with me, and for the sake of you who are Christians’. This Persian general – at least as represented by a Christian chronicler, which only complicates the picture further – seems to have assumed that for Jews and Christians (and Nazarenes, if this represents a distinct religious group), Passover and Easter ‘shared culture’ in the context of late antique Babylonia can be found in A. H. Becker, ‘The comparative study of “scholasticism” in late antique Mesopotamia: rabbis and east Syrians’, AJS Review 34 (2010), 91-113 (esp. 98-9). 5 Aphrahat, Twelfth Demonstration (c.344CE); M. J. Pierre, Aphraate le sage Persan: Les Exposés, Sources Chrétiennes 359 (Paris: Editions du Cerf, 1989), p. 578; see Stern 2001, p. 69, n.68. 5 were all one and the same feast.6 Given the fluidity and subjectivity of these concepts, ‘similarity’ and ‘difference’ can hardly serve as satisfactory criteria for assessing the cultures in late antique Babylonia, let alone for identifying specific phenomena such as shared culture, influence, or other forms of cultural exchange and transfer. This should not cause us, however, to descend into a methodological impasse. Calendars are sufficiently well defined, in terms of structure, to permit a useful comparison and to attempt a history of calendars in late antique Babylonia. Survey of the sources Evidence of calendars in Sassanian Babylonia is limited. To my knowledge, there are no epigraphic or documentary sources from this period that are dated or inform us, in any other way, about the calendars that may have been in use in this region. We are restricted, therefore, to literary sources, and these again are limited: for there are no literary sources from this period by pagans, Zoroastrians, or Mandaeans, that could provide us information about the calendars that they used. We only have on the Jewish side the Babylonian Talmud, and on the Christian side, some acts of synods and a body of literature known as the Syriac Martyr Acts. It may be appropriate to comment, at this stage, on the relevance of the Syriac Martyr Acts to Babylonia and the Babylonian Talmud. This body of texts, of which a comprehensive critical edition and translation is still awaited, is largely focused on geographical locations in central and lower Mesopotamia (which I shall call ‘Babylonia’) and the Khuzistan (or Elam, or Bēt Huzaye), where most of the martyrdoms are depicted as having taken place, all under Sassanian rule. It is in these regions that these martyrdom stories most probably originated and were eventually produced in their extant literary form.7 Scholars 6 Chronicle of Ps-Zachariah, 9:4a, text in E.W. Brooks, Historia Ecclesiastica Zachariae Rhetori Vulgo Adscripta, Corpus Scriptorum Christianorum Orientalium 83-4 (Paris: Gabalda, 1919-1921); translation in G. Greatrex, C. Horn, and R. Phenix, The Chronicle of Pseudo-Zachariah Rhetor: Church and War in Late Antiquity (Liverpool: Liverpool University Press, 2011), p. 322. I am grateful to Geoffrey Greatrex for drawing my attention to this passage. On the identity of ‘Nazarenes’ in the context of this passage, see F. de Blois, ‘Nasrânî [Nazoraios] and hanîf [ethnikos]: Studies on the religious vocabulary of Christianity and of Islam’, BSOAS 65, 2002, 1-30, on 8-10. On a sixth-century Zoroastrian confusing Jews with Christians in another source (but again, only according to a Christian narrator), see Becker 2010, 108-9. As far as I can tell, the Babylonian Talmud and other early rabbinic sources appear not to be aware of the Christian Easter at all. 7 The standard edition of the Syriac Martyr Acts is P. Bedjan, Acta Martyrum et Sanctorum, 7 vols (Leipzig, 1890-7), which provides a text with some variant readings; fully critical editions of some individual Acts have been sporadically published, especially in Analecta Bollandiana. German translations of some of the Acts are in G. Hoffmann, Auszüge aus 6 seeking to relate Syriac Christianity to the Babylonian Talmud have often focused on what might be called high Syriac literature, with fourth-century authors such as Aphrahat and Ephrem, and on the great Christian academies of Edessa and Nisibis in upper Mesopotamia, where much of this literature was composed.8 Without denying their relevance, it is important to stress the geographical remoteness of Edessa and Nisibis from the rabbinic centres of learning in Babylonia (even though some second-third-century rabbis were located in Nisibis, and one Nestorian academy was eventually established in Seleucia, very near the rabbinic centres of learning in Mahoza, in the sixth century).9 It is also important to note the pronounced and explicit dependency of high Syriac literature on Hellenistic literary, philosophical, and theological traditions.10 The Syriac Martyr Acts, in contrast, were composed in the same region as the Babylonian Talmud, and represent perhaps a more popular, less elitist genre of literature which may have been more accessible or understandable to Babylonian rabbis; as such, these sources offer much potential – still largely untapped – for the study of the broader social context of Babylonian rabbinic Judaism.11 This said, the Babylonian Talmud is largely focused on the rabbinic traditions it had inherited from Palestine (mainly in the third century), and seems remarkably impervious to the non-Jewish (and even, one could add, nonrabbinic) cultures of its surrounding society. Calendars are no exception to this. In contrast to the Syriac Martyr Acts, where a variety of calendars and dating syrischen Akten persischer Märtyrer, Abhandlungen für Kunde des Morgenlandes der Deutschen Morgenländischen Gesellschaft 7.3 (Leipzig, 1880), and O. Braun, Ausgewählte Akten persischer Märtyrer, Bibliothek der Kirchenväter 22 (Kempten and Munich, 1915). For a general survey see J. Rist, ‘Die Verfolgung der Christen im späntantiken Sasanidenreich: Ursachen, Verlauf und Folgen’, Oriens Christianus, 80 (1996), 17-42; and for quick reference, Bibliotheca Hagiographa Orientalis (Brussels, 1910). Finally, for an in-depth study of one of the Acts, see J. T. Walker, The legend of Mar Qardagh: Narrative and Christian Heroism in late antique Iraq, The Transformation of the Classical Heritage 40 (Berkeley: University of California Press, 2006). 8 See for example J. Neusner, A History of the Jews of Babylonia, vol.3 (Leiden: Brill, 1968), pp. 195-200; I. Gafni, ‘Nestorian literature as a source for the history of the Babylonian Yeshivot’, Tarbiz 51 (1981-2), 567-76 [Hebrew]; J. L. Rubenstein, The Culture of the Babylonian Talmud (Baltimore and London: Johns Hopkins University Press, 2003), pp. 3538; Boyarin 2007, p. 350; N. Koltun-Fromm, Hermeneutics of Holiness: Ancient Jewish and Christian Notions of Sexuality and Religious Community (Oxford: Oxford University Press, 2010) (focused almost entirely on the Acts of Thomas and on Aphrahat); Becker (2010). 9 See discussion in Koltun-Fromm 2010, pp. 21-2. 10 For an excellent discussion of the similarities and differences, see Becker, ibid. (and on the school in Seleucia, ibid., p. 94). 11 On the general neglect of the Syriac Martyr Acts in modern scholarship, see Walker 2006, pp. 113-15; and see now M. Bar-Asher Siegal, Early Christian Monastic Literature and the Babylonian Talmud (New York: Cambridge University Press, 2013). 7 systems are used (as we shall see below), the Babylonian Talmud only knows the Jewish calendar. Not only are festivals, fasts, and other liturgical occasions dated, as would be expected, according to the Jewish calendar, but so are seasons of the year12 and legal documents.13 There are hardly no references, however, to the Persian-Zoroastrian or Julian calendars, nor dates in any of these calendars. The only possible exception appears in a passage listing names of Graeco-Roman, Babylonian, and Persian (or Zoroastrian) festivals (bAvodah Zarah 8a-b and 11b), where one of the Babylonian festivals is given the date of 10 or 20 Adar.14 This date must be in the Babylonian calendar, although being lunar, it would not have differed much from its Jewish equivalent; the Talmud itself, indeed, does not point out any distinction. However, the Babylonian and Jewish Adar were not necessarily identical, as the years were not necessarily intercalated in the same way. Evidence of the survival of pagan, Babylonian calendars in late Antiquity will appear further below. Another reference in the Babylonian Talmud to a month that is not of the Jewish calendar appears in bRosh Hashanah 15a, where Shevat shel tequfah is mentioned in contrast to Shevat shel ḥodashim. The latter, ‘Shevat of months’, is the lunar month of Shevat in the Jewish calendar; but the former, ‘Shevat of the solstice’ (or perhaps the equinox, but in relation to Shevat, most likely the winter solstice), belongs to some other calendar which, as the context indicates, is most probably non-lunar. A similar phrase appears once in the Palestinian Talmud: Nisan shel tequfah (‘Nisan of the equinox: pTaanit 1:7, 64d). The reference, in both passages, to an astronomical concept (tequfah) suggests that these might be months of a zodiac calendar, of which the first sign, Aries, would normally begin at the spring equinox, and of which each zodiac sign would be designated with a name of the Babylonian (or ‘Jewish’) calendar; on this interpretation, Shevat would correspond most likely to the sign of 12 E.g. bShabbat 43b, bBava Qamma 113a. The months of ‘Nisan’ and ‘Tishre’ are paradigmatically used for dates of legal documents: e.g. bBava Mezia 17a, bBava Batra 171b. Although these month names are not inherently Jewish – indeed, Syriac or Aramaic-speaking Christians used them ubiquitously, even for Julian calendar months – in the context of the Babylonian Talmud it is reasonable to assume that they refer to the Jewish calendar. This is evident, in fact, from a baraita that is cited in bBava Batra 171a, according to which a document dated ‘Sabbath’ or ’10 Tishrei’ is presumed to be post-dated, on the grounds that 10 Tishrei is the date of yom kippur. 14 This date appears in bAvodah Zarah 11b, but the textual witnesses are divided: the first printed edition (Pesaro, 1511), followed by all subsequent editions, reads 10; but mss. Paris (BN Heb.1337) and JTS (Rab. 15) read 20. The reading of ms. Munich (BSB Heb. 95, fol.375b) is ואסרי, which seems more like the name of a festival than a gross misspelling of either 10 or 20 (similarly, perhaps, ועיסריin the printed edition of R. Ḥananel’s commentary). All versions have, however, the month Adar, which in this context must be Babylonian. 13 8 Aquarius, in January-February, and Nisan to Aries, in March-April.15 But alternatively, these non-lunar months could be of the Julian calendar, for which Aramaic speakers used (and still use today) Babylonian month names, not only in the Roman province Syria and other parts of the Roman Near East but also in Sassanian Babylonia where, as we shall soon see, the Julian calendar was commonly used by Christians.16 This is the interpretation of a Geonic commentator (on the passage in bRosh Hashanah 15a), who writes: ‘and what is “of the tequfot”? That is according to the count of the Greeks’.17 The ‘count of the Greeks’ means the Julian calendar: indeed, this is how the Julian calendar is referred to in contemporary Christian Syriac sources (as we shall soon see). Similarly, R. Ḥananel (11th cent.) assumes in his commentary that Shevat shel tequfah refers to the Julian calendar; he identifies this month with January, which seems incorrect as in the Syrian calendar, Shebat is the name of February. However interpreted, these references to non-lunar, either zodiacal or A zodiac calendar of this kind, which we need not identify as specifically ‘Jewish’, is attested in Palestine in the mosaic floor of the synagogue of Sepphoris, of which one of the central panels represents a zodiac wheel: each sign of the zodiac is identified with the name of the sign and its depiction, but also with a Babylonian-Jewish month name, in such a way that Aquarius (for example) bears the name of Shevat (the Aquarius/Shevat panel, however, is badly damaged: see Z. Weiss, The Sepphoris Synagogue: deciphering an ancient message through its archaeological and socio-historical contexts (Jerusalem: Israel Exploration Society and Hebrew University of Jerusalem, 2005), pp. 110-23. The same scheme is attested in the Mandaean calendar and zodiac, which is likely to go back to late Antiquity (E. S. Drower, The Book of the Zodiac (Sfar Malwašia) (London: Royal Asiatic Society, 1949); eadem, The Thousand and Twelve Questions (Berlin: Akademie-Verlag, 1960), pp. 12-13; eadem, The Mandaeans of Iraq and Iran (Leiden: Brill, 1962), pp. 82-84). This interpretation is assumed, for pTaanit 1:7 (64d), by Maimonides in his Code (Laws of Fasting 3:9), and for bRosh Hashanah 15a by Rashi (ad loc.), although the latter conceptualizes this zodiac month as a ‘solar Shevat’, beginning 30 days after the winter solstice (when solar Tevet begins). A different interpretation is put forward elsewhere by Maimonides with reference to bRosh Hashanah 15a (M. Y. L. Sachs, Hiddushei ha-Rambam la-Talmud (Jerusalem: Mekhon haTalmud ha-Yisreeli ha-Shalem, 1963)), according to which Shevat shel tequfah is the first lunar month following the winter equinox. I cannot relate this to any known calendar. 16 Nowadays, these Babylonian month names are used not only for the Julian months of the liturgical calendar but also for the Gregorian months of the civil calendar, not only by Aramaic but also by Arabic speakers, Muslim and Christian, in Iraq, Syria, and elsewhere in the Near East. 17 A. E. Harkavy, Zikhron Kamah Geonim (also known as Teshuvot ha-Geonim) (Berlin: Itzkowsky, 1887), p. 119 (from manuscripts of the Firkowitch collection in St Petersburg); also in B. M. Lewin, Otzar ha-Geonim, vol.5 (Rosh Hashanah) (Jerusalem: Hebrew University, 1932/3), p. 24: ואיזו היא שלתקופות? זו שלפני מנין יונים. The preposition ( לפניliterally ‘before’) is strange; perhaps it is an error for ‘( לפיaccording to’). 15 9 Julian calendar months, are very exceptional in rabbinic literature and the Babylonian Talmud.18 The insularity of the Babylonian Talmud with respect to calendars is reciprocated to some extent in Eastern Christian sources of late Antiquity, where the Jewish calendar appears to be never used or mentioned.19 This should not surprise us, if Christians regarded the Jewish calendar as specific to the Jews and Judaism, and if its use in broader society was limited. Christian sources, however, use and refer to several other calendars that are likely to have been in public use in Sassanian Babylonia. The dating system perhaps most frequently used in these sources is the Syrian calendar, which is well attested in the Roman part of the Near East (especially in the Roman province of Syria and in western Syriac literary and epigraphic sources), and which consists quite simply of the Julian calendar but with Syriac, Babylonian-derived month names (thus Teshri I = October, etc.). This calendar is used, for example, in the Acts of the synods that were held in Seleucia-Ctesiphon from the early fifth century onwards, which suggests perhaps that Eastern Church leaders treated the calendar of the Roman Empire as the official calendar of the Christians, albeit under Sassanian rule. This choice of calendar was natural, however, as it served to determine the fixed (i.e. non-lunar) Christian liturgical dates such as Christmas and Epiphany. Thus, the synod of Mar Isaac is dated Tuesday 1 Shebat, year 11 of Yazdegird (which began on 14 August 409);20 the weekday is compatible to the Julian date of 1 February (410 CE), and this is almost certainly what is intended. By extension, it may be assumed that all Syriac month names in the Acts of the Synods are Julian.21 18 Although this tradition is attributed in bRosh Hashanah 15a to Palestinian sages, the reference to Shevat shel tequfah is most likely the contribution of a Babylonian editor. Indeed, the parallel source in the Palestinian Talmud, pBikkurim 2:4 (65a-b), refers only generally to the ‘order of years’ and ‘order of solstices/equinoxes’ ( כסדר של שנים או כסדרן של )תקופות, which means, respectively, the calendar (which is lunar) and the solar year. This suggests that the notion of a zodiacal or Julian month was not specifically Palestinian (as evidenced in the passage in pTaanit 1:7, 64d), but also employed by the editors of the Babylonian Talmud. The identification of Shevat with January is attested in European medieval sources: see S. Stern, ‘Christian calendars in Hebrew in Jewish medieval manuscripts’, Medieval Encounters (forthcoming). 19 The theory put forward by R. Mercier, ‘The Dates in Syrian Martyr Acts’, Analecta Bollandiana 117 (1999), 47-66 (= R.W. Burgess, Studies in Eusebian and Post-Eusebian Chronography (Stuttgart: Frank Steiner Verlag, 1999), pp. 287-301 (‘appendix 3’); see also ibid. 13, p. 245 n.183, pp. 246-47, and pp. 249-50), that the lunar dates in the Syriac Martyr Acts are drawn from the Jewish calendar, is refuted by S. Stern, ‘Near Eastern lunar calendars in the Syriac Martyr Acts’, Le Muséon 117 (2004), 447-72. 20 J. B. Chabot, Synodicon Orientale ou Recueil de Synodes Nestoriens, Notices et Extraits des Manuscrits de la Bibliothèque Nationale et Autres Bibliothèques 37 (Paris: Klincksieck, 1902), pp. 18-19 (text) and pp. 255-57 (translation). 21 E.g. Teshri I (ibid. 69), for the synod of Mar Aba I in 540 CE. 10 Julian dates also appear in the Syriac Martyr Acts, where they are sometimes identified explicitly as ‘according to the Greeks’: for example, the martyrdom dates of Piruz on 5 Ilul (= September) in the early 420s CE, and of Jacob Intercisus on Friday (which should perhaps really be Saturday) 27 Teshri II (= November), probably 420 CE.22 In most cases, the date is unspecified but can be identified, on the basis of the weekday, as Julian: thus, the martyrdom dates of Malkē on Thursday 21 Nisan, 626 SE (= 21 April 315 CE), Mahdok on Sunday 12 Kanun II, year 9 of Shapur (= 12 January 318 CE), and Petion on Friday 25 Teshri I, 9 Yazdgerd (= 25 October 446 CE). Another case of Julian calendar use is the martyrdom date of Giwargis, which it is given without a weekday as 14 Kanun II, 926 SE (= 14 January 615 CE). The identification of this date as Julian is based on the equivalent Persian date 28 Mihr, which is given in the same source and which corresponded to (Julian) 14 January in that year.23 Persian (or Zoroastrian) calendar dates, such as the one just mentioned, are rare in Christian sources and not attested before the mid fifth century CE. The Persian calendar is based on a 365-day year, but unlike the Julian calendar it does not have a leap year, which means that it recedes by one day every four years in relation to the Julian calendar.24 The relatively late arrival of Persian dates in Christian sources may be taken as evidence that the official calendar of the Sassanian dynasty took some time, perhaps more than a century, to infiltrate Babylonian public life. This explanation, however, would depend on the assumption that the dates of the Martyr Acts were recorded at the time and place of the martyrdoms, and not reconstructed some time later, for instance when the narratives were redacted or written. But although this assumption cannot always be made,25 in this context, where Persian dates are always provided together with Julian equivalent dates, it is likely that both dates were recorded at the time 22 On the problems surrounding both these dates see Stern 2004, pp. 449-50 and n.10, with reference to the primary sources and modern scholarship. Piruz and Jacob Intercisus were martyred in Bēt-Lapat, the Sassanian capital and Christian ecclesiastical centre of the province of Khuzistan (Bēt Huzaye), later known as Gundeshapur (Walker 2006, pp. 95-6); the city is mentioned in the Babylonian Talmud, bTa’anit 22a (Be-Lapat), though not as a centre of rabbinic activity. 23 Ibid. p. 449. SE stands for Seleucid Era, which in the Syrian calendar began in Autumn 312 BCE. Mahdok was martyred in Mount Bĕrain, near the Lesser Zab river (which marks the southern border of Adiabene) (Bedjan 1890-7, ii, pp. 1-39). The locations of Petion and Giwargis are unspecified. 24 F. de Blois, ‘The Persian calendar’, Iran 34 (1996), 39-54. This calendar is structurally identical to, and indeed originally derived from, the Egyptian civil calendar. 25 For example, the martyrdom of Malkē is located not in Babylonia but in Egypt (Bedjan 1890-7, v, p. 467), where one expects it to have been dated originally according to the Alexandrian calendar; if the story is historical, we would have to assume that its date was converted later into its Julian/Syriac equivalent. 11 of the event, because the conversion of Julian to Persian dates (or perhaps viceversa) requires a complex calculation that is unlikely to have been carried out, for no useful purpose, years after the event.26 The earliest of these is the martyrdom of Anahid, dated ‘in the month of Haziran, on day 18, that is the month of Spndrmd (= Persian month Spandārmuδ), on Friday, year 9 of Yazdgerd’ (447 CE).27 Then follows the martyrdom of Grīghōr, although it does not provide, strictly speaking, a Persian date: it is said to have occurred in the month of Adar and at the time of the Zoroastrian festival of Prwrdgn (= Frawardigān), in year 30 of Kavad (518 CE).28 After that come two dates in the Acts of the Synods, in connection with the synod of Mar Aba (which was held in Seleucia-Ctesiphon, 540 CE): it is dated ‘in the month of Tishrei I, that is the month of Tirmah (= ‘month of Tīr’), in year 10 of Khusrau’, and records of the synod were compiled by Mar Aba ‘in the month of Shahrir (= Shahriwar), in year 13 of Khusrau’.29 It may be noted that the latter provides only a Persian month, without any Julian or other date. Somewhat later, the martyrdom of Giwargis in 615 CE is given a precise Persian date (alongside the Julian date), including the day of the Persian month (see above) – in contrast to the earlier Persian dates that only provide the month name. These last two cases (Mar Aba and Giwargis), if significant, suggest perhaps, each in their own way, that use of the Persian calendar became increasingly dominant in Babylonia in the last century of Sassanian rule. The third calendar attested in the Syriac Martyr Acts, and possibly the most frequent in this body of literature, is lunar. A considerable number of dates, indeed, are referred to explicitly as ‘according to the moon’. For a reason that remains to be explained, the use of this phrase appears to be confined to dates of martyrdoms from the reign of Shapur II (309-79 CE); but as I have argued elsewhere, dates of martyrdoms from other reigns can also be determined as lunar, on the basis of the weekday supplied alongside the day number of the Syriac month.30 The use of the same, Syriac month names for both Julian and lunar dates, in most cases without specification, is confusing to us and may well have caused confusion to the calendar users in late Antiquity. 26 This does not mean that the narratives of the Syriac Martyr Acts must be treated as factual or historical, but I am assuming, at the very least, that their dates originate from authentic – if unrecoverable – historical events. 27 Bedjan 1890-7, ii, 603; Stern 2004, 462-63. The correspondence with the Persian date suggests that Haziran is erroneous, and should really be Tammuz; if so, the date of Friday 18 Tammuz 447 CE could be either Julian (i.e. 18 July) or lunar. 28 Hoffmann 1880, 79; de Blois 1996, 47. The correlation fits a Julian month of Adar (= March), as in 518 CE the Frawardigān occurred in mid March; however, a lunar month would be equally possible. 29 Chabot 1902: 69 and 68 (respectively). In 540 CE, the Persian month of Tīr began on 7 Tishrei I (October); in 543, Shahriwar began on 6 December. 30 Stern 2004; on dates ‘according to the moon’, ibid. 448 n.2. 12 The nature of the lunar calendar, moreover, is unclear. The suggestion that it might be the Jewish calendar – i.e. a calendar set and determined by Jews – is groundless and without evidence, and in the context of Christian writings, not particularly likely.31 Equally unlikely is the possibility that this lunar calendar is the Christian Easter cycle: although lunar, this cycle was primarily designed for the determination of the date of Easter (and hence other dates in the Christian calendar that depend on it, e.g. the beginning of Lent), but not as a continuous calendar with named months running through the whole year (at least not before the Middle Ages); it is also unclear to what extent Easter cycles were used in the East in the early Sassanian period (more on this below). It seems most likely, therefore, that the dates ‘according to the moon’ were based on local, indigenous Babylonian calendars that had remained lunar from the origins until late Antiquity. The Syriac Martyr Acts, however, and the passage in bAvodah Zarah 11b mentioned above, would be the only sources attesting their existence. The lunar calendars of late antique Babylonia Jewish and other, local lunar calendars in late antique Babylonia shared common origins and a common history. As their Babylonian month names reveal, they originally derived from the standard Babylonian calendar that had served as official imperial calendar in the great Near Eastern empires of the first millennium BCE, especially in the Achaemenid and Seleucid periods. This calendar survived the disintegration of the Seleucid Empire and remained in use the Parthian kingdom as well as of the smaller city-states and kingdoms of the post-Seleucid Near East, including, for example, the Judaean Hasmonean kingdom. The chief characteristics of the standard Babylonian calendar were, firstly, that the months began when the new moon was first sighted (either actually, or when its first visibility was predicted), and secondly, that the year was intercalated with a second month of Adar (or less frequently, Elul) on the basis of a 19-year cycle, in such a way that Nisan never began before the spring equinox.32 In the post-Seleucid period, however, city-states and kingdoms reckoned the Babylonian calendar independently from one another, which led to the diversification and fragmentation of the calendar. In some cases, deviation from the standard Babylonian scheme may have been a way of asserting political autonomy; but in others, it may have served particular circumstances and needs.33 31 As argued in full in Stern 2004, 456-9, contra Mercier (see above, n.19). In this article I conclude, like here, that the lunar dates in the Syriac Martyr Acts were drawn from local Babylonian calendars. 32 S. Stern, ‘The Babylonian month and the new moon: sighting and prediction’, Journal for the History of Astronomy 39 (2008), 19-42, and Stern 2012, ch.2. 33 Stern 2012, ch.5. 13 The Judaean (or ‘Jewish’) calendar, for example, appears not to have followed a fixed 19-year (or any other) cycle, and to have allowed the intercalation of a second month of Adar but never of Elul. These changes may have been related to a certain understanding of the proper season for the observance of Passover, on the 14th of Nisan. Further changes continued into late Antiquity. By the fourth century CE, the Jewish Nisan was occurring earlier than according to the Babylonian calendar, so that Passover frequently occurred before the equinox.34 From this period onwards, moreover, Jewish calendars did not consistently begin the month at the first sighting of the new moon: in late Antiquity, some Jewish calendars began the month near the time of the conjunction (when the moon is very close to the sun and totally invisible), which occurs typically one or two days earlier, and which can only be determined by calculation or on the basis of some fixed scheme.35 These changes were related to wider, more fundamental changes that were affecting Jewish calendars in this period. From the fourth century CE onwards, some Jews were abandoning empirical methods such as new moon sighting, and opting instead for fixed calendar schemes. For example, a list of Jewish Passover dates in a document issued by a faction of Syrian bishops at the Council of Serdica (343 CE) implies not only a month beginning at or near the conjunction, but also a simple scheme for calculating Passover dates, as well as a fixed rule whereby Passover could only occur in the Julian month of March.36 Another Jewish calendar that underwent schematization in late Antiquity, in the long term far more important, was the rabbinic calendar – i.e. the calendar represented in rabbinic literature. The earliest rabbinic sources, redacted in early third-century CE Palestine (e.g. the Mishnah and Tosefta), present an entirely empirical lunar calendar where the beginning of the month was determined on the basis of new moon sightings, and the intercalation decided year by year on the basis of the crops, the weather, and other empirical criteria. But in the course of the Amoraic period (third-fifth centuries CE), a number of rules appear to have been introduced, determining the length of certain months and preventing the year from beginning on certain weekdays. These rules, which overrode the empirical observations, had the effect of gradually transforming the calendar into a fixed scheme: by the time of the redaction of the Palestinian Talmud (end of the fourth century), the rabbinic calendar consisted largely of a fixed sequence of full and hollow months in alternation, and there is talk in one passage of ‘those who calculate’ the calendar, suggesting that the methods 34 Stern 2001, 66-98. Ibid. 132-43. It is important to refer to ‘Jewish calendars’ in the plural, because in the Diaspora as well as in Palestine they were reckoned in a variety of ways. Nevertheless, some common trends in the history of Jewish calendars can be identified. 36 Ibid. 74-9 and 124-32. 35 14 were no longer empirical.37 From the sixth century there is evidence that the beginning of the month in the rabbinic calendar was based on the conjunction (and no longer on observation of the new moon); this is already implicit in a fourth-century Babylonian tradition cited in the Babylonian Talmud.38 Although most of these changes appear to have originated in Palestine, whose authority, in calendrical matters at least, appears to have been accepted by the Babylonian rabbinic community in the Talmudic period, the Babylonian Talmud is distinctive for its interest in calendar calculations and schemes, far more than Palestinian rabbinic sources. Thus we find in the Babylonian Talmud the full exposition of a schematic calculation of equinoxes and solstices, based on a 28-year cycle (b‘Eruvin 56a); rules for the prediction of new moon visibility (of which the scientific value remains, however, obscure: bRosh Hashanah 20b); and a lengthy passage about calendar calculations and the length of months and years (b‘Arakhin 8b10a).39 In many passages, the Babylonian Talmud cites and discusses an obscure Tannaitic tradition on a rule attributed to ‘Others’ that implies a fixed length of the year; this tradition is barely attested in Palestinian sources.40 It is again in the Babylonian Talmud – but not in any Palestinian source – that we find the rule that Passover cannot occur before the spring equinox (bRosh Hashanah 21a). Finally, the Babylonian sage Samuel (early third century) is said to have boasted that he could set the calendar for the entire Diaspora (bRosh Hashanah 20b); and as proof of his knowledge, he once sent to R.Yoḥanan, his leading Palestinian contemporary, a calendar or schedule of intercalations for sixty years (Ḥullin 95b).41 The Babylonian Talmud’s 37 A fixed sequence of months, at least from Adar to Tishrei, is implicit in pMegillah 1:2 (70b); ‘those who calculate’ are mentioned in pSukkah 4:1 (54b). See ibid. 157-75. The medieval tradition that the fixed rabbinic calendar was instituted whole piece by a patriarch named Hillel in the mid fourth century is not supported by Talmudic sources (ibid. 175-81). 38 Sixth-century evidence: ibid. 182-4, where I overlooked a tradition in bSanhedrin 41b-42a attributed to the ‘Nehardeans’ whereby the full moon occurs on the 16th of the month, which implies a month beginning at the conjunction; the Nehardeans are usually associated with the fourth century CE. A 19-year cycle of intercalations (not identical, however, to that of the standard Babylonian cycle) was eventually adopted, though not much before the eighth century, and it is only by the early tenth century that the fixed rabbinic calendar, still normative today, finally came into existence (ibid. 191-210). 39 Note also that a passage in the Babylonian Talmud, unparalleled in Palestinian sources, extols those who calculate tequfot (equinoxes and solstices) and mazzalot (stars and planets): bShabbat 75a. 40 The main source is b‘Arakhin 9b; also bShabbat 87b, bRosh Hashanah 6b and 20a, and bSukkah 54b. In Palestinian sources this tradition is only attested in t‘Arakhin 1:11 (Zuckermandel ed., pp.543-4). 41 Whether it was a calendar or a list of intercalations depends on the meaning of the term עיבורא. See, on all this, Stern 2001, 257-63. 15 interest in calendrical schemes suggests that Babylonia may have played a critical role in shifting the rabbinic calendar, during the Talmudic period, from an empirical to a fixed calendar.42 Local, non-Jewish calendars in late antique Babylonia had similar origins to the Jewish calendar; it may be presumed, indeed, that they were direct descendants of the standard Babylonian calendar. Unfortunately, no explicit evidence of these calendars has survived; their existence can only be tentatively inferred from the prevalence of lunar dates in the Syriac Martyr Acts (see above). As I have argued elsewhere, analysis of these lunar dates suggests that not all local calendars conformed anymore, in this period, to the standard Babylonian calendar of earlier times: although some calendars were probably based on empirical new moon sightings, others began the month around the time of the conjunction (which suggests the use of schematic calendars); and whereas in some calendars Nisan was post-equinoctial, in others Nisan could begin well before the spring equinox.43 The Christian Easter cycles Christian Easter cycles, although based on a lunar computation, did not directly derive from the standard Babylonian calendar; their origins and history were quite different, therefore, from the Jewish and other local calendars in Babylonia of late Antiquity. Nevertheless, their importance to Christians in late antique Babylonia needs also to be considered. The Christian celebration of Easter draws its origins from the Jewish Passover, from which, originally, it was hardly distinguished. But although the earliest followers of Jesus celebrated this Biblical festival in the same way as all other Jews, very early on – I would venture to surmise, already at the first Passover after the Crucifixion – the festival was invested with a distinctive, ‘Christian’ new meaning: it became the anniversary and commemoration of the 42 As argued out more fully ibid. 263-75. Stern 2004. Further evidence of the survival of lunar calendars in late antique Babylonia may be the liturgical calendar of the Manichaeans, which in the Middle Ages at least (most notably, according to an-Nadīm in his Fihrist, 333.28 – 334.1, mid tenth cent.) was based on the phases of the moon and thus implicitly on a lunar calendar: see W. B. Henning (with notes by S. H. Taqizadeh), ‘The Manichean Fasts’, Journal of the Royal Asiatic Society 77, 1945, 146-64, and F. de Blois in id. and N. Sims-Williams (eds.), Dictionary of Manichaean Texts, vol.2. Texts from Iraq and Iran (Texts in Syriac, Arabic, Persian, and Zoroastrian Middle Persian), Corpus Fontium Manichaeorum: Subsidia (Turnhout: Brepols, 2006), 60. Further afield, al-Biruni (writing in c.1000) describes the calendar of the ‘Sabians’ or heathens of Harrān (a city in Upper Mesopotamia) as lunar, with months beginning two or three days after the conjunction, and with intercalations of an additional Adhar (The Chronology of Ancient Nations, trans. C. E. Sachau (London: W. H. Allen, 1879), 314-19; an intercalated lunar calendar is also ascribed by Biruni, as one would expect, to the pre-Islamic Arabs: ibid. 321, 324). I am grateful to François de Blois for these suggestions. 43 16 Passion of the Jesus. This new interpretation of the festival gradually drew away the Christian Easter from the Jewish Passover, and soon had an impact on the date of the festival.44 By the late second century CE, the established custom in most Christian communities was to celebrate Easter no longer on the date of Passover but on the Sunday following it, to commemorate the Sunday of the Resurrection. This date, however, was still reckoned with reference to the Jewish Passover, and there is evidence that still in the late second century, Christians were observing when Jews were actually celebrating Passover, in order to determine the following Sunday as the date of Easter.45 A major change occurred, however, in the early third century, when Christians began to determine the date of Passover (i.e. the lunar date of 14th Nisan) independently of the Jews, on the basis of pre-calculated tables or cycles. The first of these cycles was devised in the West, more specifically in Rome, and is attributed to Hippolytus; it consists of a table of 112 years starting from 222 CE, which is presumably near the time when the table was composed.46 This cycle, grossly inaccurate, was modified and updated in pseudo-Cyprian’s De Pascha Computus, a treatise on the calculation of the date of Easter written perhaps in Africa in 243 CE.47 The extent to which these cycles were ever used in practice, however, remains unknown.48 They were later superseded in Rome by the far more accurate 84-year cycles, perhaps already sometime in the third century. The most important of these was the so-called supputatio romana, which was probably conceived in the early fourth century, as its 84-year cycle started in 298 CE; it was probably in use in Rome by the 340s, and became standard among Roman Christians until the mid fifth century.49 I use the English name ‘Easter’ to refer specifically to the Christian festival, although ancient Greek, Latin, and Syriac generally use the same name, Pascha, for both the Jewish and the Christian festivals. 45 Stern 2001, 222-3. Thus Polycrates, bishop of Ephesus in the 190s, describes the 14th of the moon as ‘the day when the people (i.e. the Jews) dispose of the leaven’ (Eusebius, Historia Ecclesiastica 5:24:6), i.e. when the Jews prepared for the festival of Unleavened Bread (which began on the 15th) by throwing out all leaven from their homes. This reference to a real-life Jewish practice, indeed a public, outdoor activity that he and other Christians in Ephesus could easily have observed, rather than for example the sacrifice of the paschal lamb (on the 14th) which in this period would have been a purely theoretical, biblical notion, suggests that it was the actual, observable practices of real Jews that determined, for him and other Christians, the date of the 14th of the month. Still in 243 CE, the African computist pseudo-Cyprian criticizes Christians who ‘walk in blindness and stupidity behind the Jews, as if not knowing on which day Easter should occur’ (more on this at the end of this section). 46 A. A. Mosshammer, The Easter Computus and the Origins of the Christian Era (Oxford: Oxford University Press, 2008), 116-25. 47 Ibid. 125-7. 48 M. Richard, ‘Notes sur le comput de cent-douze ans’, Revue des Etudes Byzantines, 24 (1966), 257-77; repr. in id., Opera Minor (Turnhout: Brepols, 1976), vol.1, no.20. 49 Mosshammer 2008, 206-16, 238; Stern 2012, ch.6. 44 17 In the East, Easter cycles took a bit longer to emerge. The earliest known cycle is that of Anatolius of Laodicea (western Syria), a 19-year cycle designed in about 270 CE but which is most unlikely ever to have been used in practice.50 It was followed, however, by the so-called Alexandrian cycle, a highly accurate 19-year cycle which may have been conceived at the beginning of the fourth century, and which was well in use in Alexandria and other parts of the Roman East by the middle of the century.51 By the mid fifth century, this cycle was also adopted, with modifications, in Rome and most other parts of the Roman West.52 From this point onwards, the Alexandrian cycle (or adaptations of it, sometimes with minor variations) became the mainstream, orthodox computation of the date of Easter. The Alexandrian 19-year cycle was not borrowed from, or modelled upon, the standard Babylonian calendar, but rather from a Greek tradition stretching back to the fifth century BCE. A 19-year cycle beginning in 432 BCE was first devised by the Athenian astronomer Meton (or one of his contemporaries);53 it was later refined by the Athenian astronomer Callippus with a 76-year cycle (consisting of four 19-year Metonic cycles minus one day), beginning in 330 BCE.54 Unlike the standard Babylonian calendar, where months are established empirically on the basis of new moon sightings, the Callippic scheme determines in advance the exact length of every month within the whole cycle (generally by alternating 29- and 30-day months). A similar, pre-calculated calendar of this kind is implicit in the Alexandrian Easter cycle of the fourth century (even though its explicit purpose is only to determine the date of Easter), and this underlying lunar calendar becomes more explicit in expositions of the Alexandrian cycle in later centuries. Another feature of the Alexandrian Easter cycle in common with the Greek tradition of lunar calendar schemes is that its months begin around the time of the conjunction, in contrast 50 D. P. Mc Carthy and A. Breen, The Ante-Nicene Christian Pasch: De Ratione Paschali. The Paschal Tract of Anatolius, Bishop of Laodicea (Dublin: Four Courts Press, 2003). 51 The evidence, Athanasius’ Festal Letters, is however complex and not entirely conclusive: see Mosshammer ibid. 165, 178-82; Stern ibid. ch.7. 52 Mosshammer ibid. 62-3. 53 On Meton, the 19-year cycle attributed to him or one of his contemporaries, and the possible relationship between it and its Babylonian equivalent, see A. C. Bowen and B. R. Goldstein, ‘Meton of Athens and astronomy in the late fifth century BC’, in E. Leichty et al. (eds), A Scientific Humanist: Studies in Memory of Abraham Sachs (Philadelphia: University Museum, 1988), 39-81, and R. Hannah, Greek and Roman Calendars: Constructions of Time in the Classical World (London: Duckworth, 2005), 54-7. Even if the Greek astronomers borrowed the 19-year cycle from the Babylonians (a proposition that remains very debatable), the calendars which they produced were very different from the standard Babylonian calendar, as I explain here. 54 The Callippic calendar (alongside earlier cycles) is described in some detail by Geminus, Elem. Astr. 8:26-60 (first century BCE). See most importantly A. Jones, ‘Calendrica I: new Callipic dates’, Zeitschrift für Papyrologie und Epigraphik, 129 (2000), 141-58. 18 with the standard Babylonian calendar where months began when the new moon is first visible.55 Further east, the dissemination of the Alexandrian cycle among Christians in the Sassanian Empire was slower and more gradual than in the Roman Empire. Although the Alexandrian cycle, or rather an adaptation of it, was later to become the standard Easter computation in Eastern Christianity (just as in the West), it is possibly not before the sixth century that it was formally adopted by the Churches under Sassanian rule.56 There are several possible reasons for this delay. Firstly, Easter cycles were specifically designed for users of the Julian or Alexandrian calendars, i.e. the calendars of the Roman Empire (the Alexandrian calendar was the calendar of Roman Egypt, consisting, like the Julian calendar, of a 365-day year with an extra day every four years); the purpose of these cycles, indeed, was to convert the lunar dates of Easter into Julian (or Alexandrian) equivalents. In the Sassanian Empire, however, this would have been far less a necessity, as the Julian calendar was not an official imperial calendar, and furthermore, local lunar calendars appear to have been in prevalent use. Although, as we have seen above, the Julian calendar was used by Christians for liturgical purposes as well as for the dating of martyrdoms and of the synods, lunar calendars were also available and used by Christians (for example, for martyrdom dates). In a society where lunar calendars were the norm, the conversion of Easter dates from lunar to Julian would have been quite superfluous. Secondly, the computation of the date of Easter on the basis of fixed, precalculated cycles ran counter to a custom that was probably observed in the East through most of late Antiquity: that is, the custom of ‘following the Jews’ or observing Easter ‘with the Jews’, i.e. on the date when the Jews happened to observe Passover (or more precisely, on the following Sunday). As we have On ‘conjunction’ and new moon visibility, see above near n.35. Another feature of the Alexandrian cycle that differed from the standard Babylonian calendar, though not clearly due to any Greek tradition (but rather, in fact, to Alexandrian Jewish traditions of the second century BCE: Stern 2001, 50-3), was the rule of the equinox. The Alexandrian cycle was constructed in such a way that Passover, the 14th day of the first lunar month (Nisan), always occurred after the vernal equinox, which was taken to be 21 March. This means that Nisan itself could often begin well before the equinox, whereas in the standard Babylonian calendar it always began thereafter. 56 An indication of this may be that the Easter cycle of the Nestorians is known, from later sources, to have begun in 562 CE (V. Grumel, La Chronologie. Traités d’Etudes Byzantines, vol.1 (Paris: Presses Universitaires de France, 1958), 54 and 98-102); this epoch is clearly related to the treaty between the Roman and Persian Empires of the same year, when Nestorians in the Sassanian Empire were granted certain rights, but it also suggests that the Easter cycle was not used beforehand. At the synod of Mar Isaac in 410 CE, a decision was recorded – in line with the Council of Nicaea – that Easter be celebrated by all on the same date (Chabot 1902, 20 and 258; see above, n.20); but in true Nicene fashion, it was not explained how this common date was to be established or observed. 55 19 seen above, this had been the custom of all early Christians, before the rise of Easter cycles in the third-century West. This custom was maintained and vigorously upheld in the Near East, with the Didascalia (a text originally written in the third century, perhaps in northern Syria, and preserved in full in a fourth-century Syriac version) enjoining Christians to ‘begin (the fast) when your brethren who are of the people [i.e. the Jews] perform the Passover’, and ‘when therefore that people performs the Passover, do you fast’, 57 or according to another version, cited by Epiphanius (370s CE), ‘reckon ye not [i.e. on the basis of a fixed Easter cycle], but celebrate when your brethren of the circumcision do; celebrate with them’, and ‘when they feast, mourn ye for them with fasting, for they crucified Christ on the day of the feast; and when they mourn on the Day of Unleavened Bread and eat with bitter herbs, then feast ye’.58 This resistance to ‘reckoning’ the date of Easter on the basis of fixed cycles, and insistence instead on following the Jews, remained dominant in the Near East throughout the fourth century, in spite of the Council of Nicaea’s opposition to it.59 This would explain the late introduction of Easter cycles in the Near East. Eventually, however, perhaps in the sixth century, Easter cycles were adopted by Christians in the East. In this respect, the Christian, lunar calendar of Easter developed in the same way as other lunar calendars in the Near East, Jewish and local Babylonian: in late Antiquity, they all evolved from flexible and empirical methods of time reckoning into schematic and fixed calendars. Influence, shared culture, and other paradigms The evolution of lunar calendars in Babylonia from empirical methods to fixed schemes occurred for a variety of reasons, specific to each of the individual calendars; and yet, we should ask whether their common history was not, in a certain way, interrelated. The reason why the rabbinic calendar shifted, during the third-fifth centuries CE, to an increasingly fixed scheme with alternating 29- and 30day months and other rigid rules, was (as I have argued elsewhere) the outcome of the unique relationship between the Palestinian and Babylonian rabbinic communities. These communities considered it essential to observe 57 Didascalia 21, in A. Vööbus, The Didascalia Apostolorum in Syriac, 4 vols. Corpus Scriptorum Christianorum Orientalium 401-402, 407-408: Scriptores Syri 175-176, 179-180 (Leuven: CSCO 1979), ii. 23*, 27-8*, iv. 196, 202. 58 Epiphanius, Panarion 70:10:2 and 11:3 (F. Williams,The Panarion of Epiphanius of Salamis, 2 vols (Leiden: Brill, 1987-1994), ii. 412-3). 59 Evidence of this custom in late Antiquity comes mainly from Roman Syria, but it may be assumed to have extended to the whole of Mesopotamia. Aphrahat, writing in Persian Mesopotamia in c.344CE, implies it in his twelfth Demonstration, although the evidence is not entirely conclusive (Stern 2001, 69, n.68). 20 festivals and fasts, and generally to reckon the calendar, on exactly the same dates – a quest for unanimity that was unique to the rabbinic movement, and never pursued elsewhere in the variegated Jewish Diaspora world. Because of the impossibility of communicating the results of empirical new moon sightings from Palestine to Babylonia and thus of ensuring the observance of festivals in both communities on the same dates, it became a necessity – especially at a time when the rabbinic community in Babylonia began to develop, from the third century CE – to increase the predictability of the Palestinian rabbinic calendar by restricting it with fixed rules. The gradual fixation of the rabbinic calendar was thus the result of an attempt to unify and standardize the calendar of the rabbinic communities of Palestine and Babylonia. Even within the Jewish world, this phenomenon was very specific to the rabbinic movement.60 On the Christian side, the development of Easter cycles in the thirdfourth centuries was motivated by very different reasons. The common scholarly view is that the purpose of the first Easter cycles was to enable all churches to observe Easter on the same dates, and thus to eradicate calendar diversity61 – not unlike the explanation put forward above for the rabbinic calendar. In actual fact, however, this motivation did not take shape before the fourth century, when Constantine attempted to enforce a single date of Easter for all Christians in the Roman Empire.62 Before Constantine, the motivation explicitly stated by one of the earliest computists, the pseudoCyprian (writing in 243 CE), was ‘to show that it is possible for Christians never to stray from the way of truth and walk in blindness and stupidity behind the Jews, as if not knowing on which day Easter should occur’.63 As the author explains, the introduction of a fixed Easter cycle would enable Christians to become independent from the Jews, by no longer relying on the Jews and their Passover for establishing the date of Easter. This suggests that Easter cycles in the early third century responded to a need for Christians to dissociate themselves from Jews and their Jewish origins and thus ‘part ways’ from Judaism.64 This motivation was still live in the late seventh century, when we find it in a treatise on Easter by Ananias of Shirak, whose purpose was to promote a new Easter cycle in Armenia; Ananias attributes this motivation, rightly or wrongly, to Constantine, but also clearly shares it himself.65 60 Stern 2001, 232-56. E.g. Mosshammer 2008, 52, 55. 62 As I argue in Stern 2012, ch.7. 63 Ps-Cyprian, De Pascha Computus, 1 (G. Ogg, The Pseudo-Cyprianic De Pascha Computus (London: SPCK, 1955), 1). 64 Stern 2001, 223-26. For further motivations, see Stern 2012, ch.7. 65 F. C. Conybeare, ‘Ananias of Shirak, “Autobiography”, “His tract on Easter”’, Byzantinische Zeitschrift 6 (1897), 572-584, on p.578: ‘[the early Christians] kept the feast 61 21 As to local Babylonian calendars, the waning of the standard Babylonian calendar and the diversification of local calendars, with some adopting less empirical, more schematic features (as can be inferred, at least, from the dates of some of the martyrdoms – see above, near n.43), was most likely the result of the collapse of the Parthian kingdom in the early third century, which had used the standard Babylonian calendar (or something very similar to it) for official imperial purposes, and the rise instead of the Sassanian dynasty that used instead the Persian-Zoroastrian calendar. In the absence of any central authority, in the Sassanian period, to impose and control a single lunar calendar, the cities of Mesopotamia drifted into calendar diversity and particularism. As conformity to an imperial lunar calendar was no longer required, cities and other localities were free to adopt a variety of lunar calendar schemes. But the switch of some local Babylonian calendars to schematic lunar calendars, with new months beginning at the conjunction rather than at the first appearance of the new moon – in total departure from earlier Mesopotamian tradition – raises the possibility of Hellenistic influence, mediated perhaps through the Jews or (more likely) the Christians and their Easter cycles. At this stage, therefore, we need to consider whether in spite of their separate origins, motivations, and histories, the lunar calendars of late antique Babylonia did not exert influences on one another. The lines of influence between Jewish, Christian, and other local Babylonian calendars can be drawn, potentially, in almost any direction (I use the term ‘influence’ in a rather general and imprecise manner, to include all forms of cultural transfer and exchange – see my remarks in the introduction to this paper) . As I have suggested already elsewhere, the fixed rabbinic calendar may have developed under the influence of Christian Easter cycles, which chronologically preceded it. In common with Easter cycles, indeed, the fixed rabbinic calendar alternated 29- and 30-day months, began the month at the conjunction, and followed a rule of the equinox, whereby Passover cannot occur before the spring equinox.66 The fixation of the rabbinic calendar may thus be simultaneously with the Jews – a miscellaneous and unclean least – until the days of the blessed Constantine and the holy council of Nicaea. And then it was prescribed by the emperor that they should not any longer after that keep a feast in impure agreement with the Lord-slaying race of Jews. So he issued an edict that they should adopt the nineteen years lunar cycle of Easters ...; and that they should take pupils from among all Christian nations and teach them the calendar; and so separate the Christians from the Jewish (observance)’. 66 The rule of the equinox, of Alexandrian Jewish origins (see above, n.55) was promoted in the 270s by Anatolius of Laodicea (who, it should be noted, came originally from Alexandria; see above, n.50), and by the early fourth century became one of the principles of the Alexandrian Easter computation (see Stern 2001, 66-7, 72-4). A similar rule then makes its first appearance in the Babylonian Talmud, bRosh Hashanah 21a, where it is attributed to a fourth-century Palestinian sage, though still only as a guideline rather than as a firm rule 22 interpreted as an outcome of the Christianization of the Roman Empire, and more particularly of Palestine, in the fourth century CE: rabbis and other Jews in this period would have appropriated the lunar calendar principles which the Christians had adopted already earlier in their Easter cycles, even if the latter’s purpose, paradoxically, had been precisely to distance themselves from the Jewish calendar.67 In the context of the Near East, however, the same line of influence can be reversed with the argument that it is the Jewish calendar that is more likely to have influenced Christian practice. Indeed, if the general practice of Christians in late antique Syria and Babylonia was to ‘follow the Jews’ and observe Easter ‘with the Jews’ (as explained above), the Jewish calendar would have played a determinative role upon the Christian liturgical calendar. Then, we must consider further lines of influence. As has been noted above, some local Babylonian calendars, which originally corresponded to the standard Babylonian calendar, appear to have become adapted in late Antiquity to the Christian Easter cycles; this may have involved the adoption of complete lunar schemes like those assumed in Easter cycles (with months beginning at the conjunction, alternation of 29- and 30-day months, etc.), or more minimally, an adjustment of the calendar – still dependent, perhaps, on new moon sightings – to ensure that the month of Nisan would coincide with the month of the Jewish Passover and Christian Easter (which frequently began, as noted above, before the equinox). The adaptation of local lunar calendars to the Easter cycles, and thus to the Christian liturgical year, is most likely to have occurred in localities with predominantly Christian populations. By the same token, it may be reasonable to expect – although we have no evidence to support this – that in localities with predominantly Jewish populations, the Jewish calendar would have exerted an influence on the local calendars. Likewise, the possibility that local Babylonian calendars influenced, in their own right, the Jewish calendar or the Christian date of Easter, cannot be ruled out: for it would have been most convenient for Jews and Christians to use the local lunar calendars that were already available and in public use. If Jews and Christians reckoned their lunar calendars on their own, differences between them are bound to have arisen, which – even if involving discrepancies of only one day – would have led to considerable confusion. In some localities, therefore, Jews and Christians may have regarded it an advantage simply to follow the local, Babylonian lunar calendar. (ibid. 167-70). The 19-year cycle, however, perhaps the most important characteristic of Alexandrian computus, is not attested in the rabbinic calendar before the eighth century (ibid. 196-200). 67 Ibid. 222-6 (on the rabbinic calendar). However, this interpretation lends itself even better to the Jewish calendar of the Serdica document (ibid. 141-3; see above, near n.36). 23 It is important to stress that the concept of ‘influence’ (or any other process of cultural transfer) is not invoked here as a pre-conceived assumption: Jews, Christians, and others in late antique Babylonia did not influence one another simply by virtue of living in mutual proximity. As discussed in the introduction to this paper, the social context and processes through which influence occurred need to be investigated and explained. What I am arguing, accordingly, is that the practical convenience of following existing, local calendars, together with specific traditions such as the Christian custom of observing Easter ‘with the Jews’, provided a social context conducive to the transfer and exchange of principles and methods of lunar reckoning between these groups, and in broader terms, to the development of the different lunar calendars in similar directions, towards increasingly fixed schemes. This said, the lines of influence were so multidirectional, and the potential for cultural transfer so extensive, that boundaries between calendars could easily become blurred. As we have seen above (near n.14), for example, a passage in the Babylonian Talmud (bAvodah Zarah 11b) seems not to distinguish clearly between Jewish and local (or ‘pagan’) Babylonian months, or even to be aware of any difference. Inasmuch as Jews, Christians, and others in late antique Babylonia all used lunar calendars with the same (or very similar) month names, the question must be raised whether or to what extent their calendars were actually different or distinguished from one another.68 A variety of lunar calendars must surely have existed, as can be demonstrated from the lunar dates within even a single literary corpus, the Syriac Martyr Acts; but this variety may well have reflected regional or local differences rather than differences between Jews, Christians, and other religious groups, whose lunar calendars are never contrasted with one another in any of the sources, and may often have been difficult to distinguish. For this reason, we may be justified in moving away somewhat from the paradigm of ‘influence’ to one of ‘shared culture’ or of ‘the same culture in different variants’ (above, n.3). In theory, the paradigms of influence and shared culture are mutually incompatible, the former assuming distinct identities and groups, and the latter assuming a single, common identity and group. Indeed, the very distinct origins, histories, and identities of Jewish calendars, local Babylonian calendars, and Christian Easter cycles cannot easily be reconciled with the notion of a single, shared lunar calendar, evolving at one, through late Antiquity, from empirical methods to fixed schemes. In this context, however, both paradigms appear to be simultaneously true; the least we can say is that the configuration of lunar calendars in late antique Babylonian society was too complex for simple choices to be made between ‘influence’ and ‘shared culture’. Perhaps we should 68 Zoroastrians are not included in this discussion, because their month names were different and their calendar was patently not lunar. 24 aim, therefore, at a more complex paradigm involving at once shared culture and divided identities, where historical change or history itself is generated through the corroboration or the convergence of very different and independent processes – some affecting individual identity groups in isolation, others involving their mutual interaction and influence, others engaging the whole of society as a shared culture – which happen to produce a common result. It was perhaps the congruence of these diverse processes – the independent histories of the individual calendars, their reciprocal influences, and the shared Babylonian culture of lunar calendars – which successfully brought about the general emergence of fixed calendars in late antique Babylonian society. A telling example of the complex relationship between these independent, mutually corroborative processes was the adoption of Easter cycles by the Church of the East, an event which occurred perhaps in the mid sixth century. Christian leaders were most probably motivated not only by loyalty to the Christians of the Roman Empire, but also by a desire to ‘part ways’ from the Jews and no longer depend on them for the date of Easter or observe the festival, as hitherto, ‘with the Jews’.69 Yet the rabbinic calendar had become by then largely schematic – perhaps under the influence, at an earlier stage in history, of the very same Easter cycles – and thus very similar to the lunar calendar implicit in the Easter cycle that was being adopted, at this later stage, by Christians in the East in an attempt to part way from the Jews. This circular paradox, involving at once rival motivations and shared outcomes, epitomizes the complexity of calendars and identities in late antique Babylonia, and the inadequacy of adopting simple models for explaining their rich history. 69 Above, near nn.57-9 and 63-5. 25 References al-Biruni, The Chronology of Ancient Nations, trans. C. E. Sachau (London: W. H. Allen, 1879). Bar-Asher Siegal, M., Early Christian Monastic Literature and the Babylonian Talmud (New York: Cambridge University Press, 2013). Becker, A.H., ‘The comparative study of “scholasticism” in late antique Mesopotamia: rabbis and east Syrians’, AJS Review 34 (2010), 91-113. Bedjan, P., Acta Martyrum et Sanctorum, 7 vols (Leipzig, 1890-7). Bhabha, H.K., The Location of Culture (London and New York: Routledge, 1994). Bibliotheca Hagiographa Orientalis (Brussels, 1910). Bowen, A. C., and B. R. Goldstein, ‘Meton of Athens and astronomy in the late fifth century BC’, in E. Leichty et al. (eds), A Scientific Humanist: Studies in Memory of Abraham Sachs (Philadelphia: University Museum, 1988), pp. 3981. Boyarin, D., ‘Hellenism in Jewish Babylonia’, in C. E. Fonrobert and M. S. Jaffee (eds), The Cambridge Companion to the Talmud and Rabbinic Literature (Cambridge: Cambridge University Press, 2007), pp. 336-63. Braun, O., Ausgewählte Akten persischer Märtyrer, Bibliothek der Kirchenväter 22 (Kempten and Munich, 1915). Brooks, E.W., Historia Ecclesiastica Zachariae Rhetori Vulgo, Adscripta, Corpus Scriptorum Christianorum Orientalium 83-4 (Paris: Gabalda, 19191921). Chabot, J. B., Synodicon Orientale ou Recueil de Synodes Nestoriens, Notices et Extraits des Manuscrits de la Bibliothèque Nationale et Autres Bibliothèques 37 (Paris: Klincksieck, 1902). Conybeare, F. C., ‘Ananias of Shirak, “Autobiography”, “His tract on Easter”’, Byzantinische Zeitschrift 6 (1897), 572-84. de Blois, F., ‘The Persian calendar’, Iran 34 (1996), 39-54. de Blois, F., ‘Nasrânî [Nazoraios] and hanîf [ethnikos]: Studies on the religious vocabulary of Christianity and of Islam’, BSOAS 65 (2002), 1-30. de Blois, F., in de Blois and N. Sims-Williams (eds), Dictionary of Manichaean Texts, vol. 2. Texts from Iraq and Iran (Texts in Syriac, Arabic, Persian, and Zoroastrian Middle Persian), Corpus Fontium Manichaeorum: Subsidia (Turnhout: Brepols, 2006), 60. Drower, E.S., The Book of the Zodiac (Sfar Malwašia) (London: Royal Asiatic Society, 1949). Drower, E.S., The Thousand and Twelve Questions (Berlin: Akademie-Verlag, 1960). Drower, E.S., The Mandaeans of Iraq and Iran (Leiden: Brill, 1962). Gafni, I., ‘Nestorian literature as a source for the history of the Babylonian Yeshivot’, Tarbiz 51 (1981-2), 567-76 [Hebrew]. 26 Greatrex, G., C. Horn, and R. Phenix, The Chronicle of Pseudo-Zachariah Rhetor: Church and War in Late Antiquity (Liverpool: Liverpool University Press, 2011). Grumel, V., La Chronologie: Traités d’Etudes Byzantines, vol. 1 (Paris: Presses Universitaires de France, 1958). Hannah, R., Greek and Roman Calendars: Constructions of Time in the Classical World (London: Duckworth, 2005). Harkavy, A. E., Zikhron Kamah Geonim (also known as Teshuvot ha-Geonim) (Berlin: Itzkowsky, 1887). Henning, W. B. (with notes by S. H. Taqizadeh), ‘The Manichean Fasts’, Journal of the Royal Asiatic Society 77 (1945), 146-64. Hezser, C., ‘Interfaces between Rabbinic Literature and Graeco-Roman Philosophy’, in P. Schäfer and C. Hezser (eds), The Talmud Yerushalmi and Graeco-Roman Culture, vol. 2 (Tübingen: Mohr-Siebeck, 2000), pp. 161-87. Hoffmann, G., Auszüge aus syrischen Akten persischer Märtyrer, Abhandlungen für Kunde des Morgenlandes der Deutschen Morgenländischen Gesellschaft 7.3 (Leipzig, 1880). Jones, A., ‘Calendrica I: new Callipic dates’, Zeitschrift für Papyrologie und Epigraphik, 129 (2000), 141-58. Koltun-Fromm, N., Hermeneutics of Holiness: Ancient Jewish and Christian Notions of Sexuality and Religious Community (Oxford: Oxford University Press, 2010). Lewin, B. M., Otzar ha-Geonim, vol. 5 (Rosh Hashanah) (Jerusalem: Hebrew University, 1932/3). Mc Carthy, D. P., and A. Breen, The Ante-Nicene Christian Pasch De Ratione Paschali: The Paschal Tract of Anatolius, Bishop of Laodicea (Dublin: Four Courts Press, 2003). Mercier, R., ‘The Dates in Syrian Martyr Acts’, Analecta Bollandiana 117 (1999), 47-66 (= R.W. Burgess, Studies in Eusebian and Post-Eusebian Chronography, Stuttgart, 1999, 287-301 (‘appendix 3’). Mosshammer, A. A., The Easter Computus and the Origins of the Christian Era (Oxford: Oxford University Press, 2008). Neusner, J., A History of the Jews of Babylonia, vol. 3 (Leiden: Brill, 1968). Ogg, G., The Pseudo-Cyprianic De Pascha Computus (London: SPCK, 1955). Pierre, M. J., Aphraate le sage Persan. Les Exposés, Sources Chrétiennes 359 (Paris: Editions du Cerf, 1989). Richard, M., ‘Notes sur le comput de cent-douze ans’, Revue des Etudes Byzantines, 24 (1966), 257-77; repr. in Richard, Opera Minora (Turnhout: Brepols, 1976), vol. 1, no. 20. Rist, J., ‘Die Verfolgung der Christen im spätantiken Sasanidenreich: Ursachen, Verlauf und Folgen’, in Oriens Christianus, 80 (1996), 17-42. 27 Rubenstein, J. L., The Culture of the Babylonian Talmud (Baltimore and London: Johns Hopkins University Press, 2003). Sachs, M. Y. L., Hiddushei ha-Rambam la-Talmud (Jerusalem: Mekhon haTalmud ha-Yisreeli ha-Shalem, 1963). Stern, S., Calendar and Community: A History of the Jewish Calendar, 2nd century BCE – 10th century CE (Oxford: Oxford University Press, 2001). Stern, S., ‘Near Eastern lunar calendars in the Syriac Martyr Acts’, Le Muséon 117 (2004), 447-72. Stern, S., ‘The Babylonian month and the new moon: sighting and prediction’, Journal for the History of Astronomy 39 (2008), 19-42. Stern, S., Calendars in Antiquity: Empires, States, and Societies (Oxford: Oxford University Press, 2012). Stern, S., ‘Christian calendars in Hebrew in Jewish medieval manuscripts’, Medieval Encounters (forthcoming). Vööbus, A., The Didascalia Apostolorum in Syriac, 4 vols, Corpus Scriptorum Christianorum Orientalium 401-402, 407-408: Scriptores Syri 175-176, 179-180 (Leuven: CSCO, 1979). Walker, J.T., The legend of Mar Qardagh: Narrative and Christian heroism in late antique Iraq, The Transformation of the Classical Heritage 40 (Berkeley: University of California Press, 2006). Weiss, Z., The Sepphoris Synagogue: deciphering an ancient message through its archaeological and socio-historical contexts (Jerusalem: Israel Exploration Society and Hebrew University of Jerusalem, 2005). 110-23). Williams, F., The Panarion of Epiphanius of Salamis, 2 vols (Leiden: Brill, 19871994).