The Evidence for an Autumnal New Year in Pre-Exilic Israel Reconsidered
Published in On the Way to the Postmodern: Old Testament Essays 1967-1998, Volume 1 (JSOTSup, 292; Sheffield: Sheffield Academic Press, 1998), pp. 3365-88A much debated question in the study of the calendar in pre-exilic Israel and Judah concerns the point at which the new year was reckoned to begin. The prevailing scholarly opinion appears to be that expressed in a typical encyclopaedia article: 'There is abundant indication that a new year in the fall was standard during the monarchical period'./1. Within this general view there has been room for a number of varying qualiÞcations. Some have believed that at a time before the Judaean exile the Assyrian and Babylonian spring calendar came into use in Judah./2. Others have thought that it was only in Judah that autumnal reckoning was employed, whereas northern Israel adopted a spring year at the time of the division of the Solomonic Kingdom./3. Only a handful of scholars have dissented from the prevailing opinion and have maintained that in both Israel and Judah throughout the monarchy the year was reckoned from the spring./4. It is the purpose of this paper to open up the question again and to test the arguments that have been advanced for the view that during the greater part of the monarchical period the year began in the autumn.
1. Evidence for an Autumnal (Tishri) New Year
In setting out the evidence for an autumnal (Tishri) new year in pre-exilic times, we may leave aside the question whether the general custom of an autumnal beginning to the year was perhaps preceded by an observance of a spring new year, as has been suggested by a few scholars./5. The main pieces of evidence for an autumnal new year are as follows:
(1) The autumn festival of ingathering 'åsªp is said in the oldest liturgical calendars of Israel (Exod. 23.14-17; 34.18-23) to occur at the 'going out' (ß't) of the year (23.16) and the 'turn' (tëqûpâ) of the year (34.22). The implication is that the new year begins at this season./6.
(2) The time of year when 'kings/7. take the Þeld', which is generally agreed to be the spring, is called in 2 Sam. 11.1 // 1 Chron. 20.1 and in 1 Kgs 20.22, 26 të¡ûbat ha¡¡ånâ, 'the return of the year', that is, as R. de Vaux puts it, 'the time when the year was half over, and beginning to return from winter to summer This again presumes an autumnal year.' /8.
(3) The sabbatical year and jubilee year began in the autumn (Exod. 23.10-11; Lev. 25.1-22)./ 9. The building of Solomon's temple is said (1 Kgs 6.38) to have taken seven years, but it is also said to have begun in the second month (Ziv) of the fourth year of Solomon (6.1, 37) and been completed in the eighth month (Bul) of the eleventh year of Solomon. If it is agreed that 'reckoning was according to the inclusive system, whereby the Þrst and last units or fractions of units of a group were included as full units in the total of the group',/10. then an autumn new year reckoning must have been in force. For on a spring (Nisan to Nisan) system the building of the temple would have taken eight years; only on an autumn (Tishri to Tishri) system could it be reckoned as occupying seven years.
(5) In 2 Kings 22p;23, the account of the reforms of Josiah, the discovery of the law book that precipitated those reforms is dated to the 'the eighteenth year of King Josiah' (22.3), while the passover that concludes the account is also said to have been celebrated 'in the eighteenth year of King Josiah' (23.23). If the year began in the spring (on the Þrst of the month, 1.I), there is not enough time available before the date of the passover (the fourteenth of the Þrst month, 14.I) to contain all the events that are said to have occurred between 22.3 and 23.23./11. Only an autumnal new year reckoning allows sufÞcient time between the Þnding of the book and the celebration of the passover for the events of Josiah's reform./ 12. In Jer. 36.1 Jeremiah is commanded, in the fourth year of Jehoiakim, to write his prophecies in a scroll. When he has done so, he orders Baruch to read the scroll in the temple on a feast day, which Baruch does in the Þfth year of Jehoiakim, in the ninth month (36.9), which is plainly a winter month since the king is sitting by his brazier (36.22). The narrative reads as if all these events occurred within the space of a few months, if not weeks; and this could be the case if we suppose an autumnal new year. The sequence of events could be: late in Jehoiakim's fourth year, say in September, Jeremiah is bidden to write his scroll, and Baruch gives a public reading in December (Kislev, the ninth month) of the Þfth year. This would seem a more likely course of events than that which would have occurred if a spring new year system were in use; such a system would imply that a period of nine months at least, or 21 months at the most, intervened between the writing and the reading of the scroll. For on a spring new year reckoning the scroll must have been written before the spring of the year, and not read until the following winter./13.
Further support for the postulate of a Tishri new year reckoning in the instance is that one can offer an explanation why Jeremiah was constrained to write his prophecies in a scroll rather than deliver them orally. 'The Babylonian army's departure from Babylon in May/June 604 b.c., throws the kings of southern Palestine into a panic. Jeremiah, being considered a subversive is forbidden to go the temple, i.e., to speak in public'./14.
(7) Even after the exile, it appears that Nehemiah reckoned the reign of Artaxerxes I on the basis of an autumnal new year. For while news of the destruction of Jerusalem's wall reached him 'in the month Kislev in the twentieth year' (Neh. 1.1), i.e. in the ninth month, reckoning from the spring, his resultant distress was noticed by the king 'in the month Nisan in the twentieth year of King Artaxerxes' (2.1), i.e. in the Þrst month of a year beginning in spring. Only if Nehemiah was using an autumn to autumn reckoning could Kislev preceede Nisan in the same year. Since it is agreed that the regular Persian and Jewish practice of this time was to reckon from a spring new year, Nehemiah's system can be explained only as a reversion to an older Hebrew custom./15.
(8) The instruction in the P source that Abib (= Nisan) in the spring is to be the Þrst month of the year (Exod. 12.2) is evidence that it had not always been so. 'The announcement in this form and in this place makes sense only if it was to replace an earlier and different counting of the months and beginning of the year.'/16.
(9) A number of reconstructions of the chronology of the monarchic period rest on the assumption that for some periods in Israel or Judah the regnal or civil year began in the autumn, on Tishri 1. So, for example, the system of E.R. Thiele depends entirely upon the presupposition that regnal years in Judah were reckoned from Tishri to Tishri (though in Israel a Nisan to Nisan system was used). While Thiele mentions some of the arguments in favour of an autumnal new year outlined above, he claims that 'perhaps the strongest argument for the use of a Tishri-to-Tishri regnal year in Judah is that this method works, giving us a harmonious pattern of the regnal years and synchronisms, while with a Nisan-to-Nisan regnal year the old discrepancies would be retained'./17.
(10) A Þnal argument, from extra-biblical evidence, is that the Gezer calendar, dating from the period of the early monarchy, attests a year beginning in the autumn. Israel is likely to have adopted the usual Palestinian calendar, it is argued./18.
2. Arguments against an Autumnal (Tishri) New Year
Against these arguments the following objections may be raised:
Ad (1) It is Þrst necessary to afÞrm, with E. Kutsch in a recent study,/19. that bëß't ha¡¡ånâ (Exod. 23.16) must mean 'at the end of the year,' and not, as several have tried to prove, 'at the beginning of the year'./20. For the fact that yåßå' can refer to the appearance of the sun or the stars (e.g. Gen. 19.23; Neh. 4.15 [Eng. 21]), and thus to the beginning of the day or the night, proves nothing about the meaning of the verb itself but only shows that the Israelites envisaged the appearance of the heavenly bodies as an exit from their 'house'. The correlative to yåßå', 'to go out, appear,' is bô', 'to come in, disappear', and the conception is obviously patterned on the familiar daily routine of work. The idea of the yåßå' of the year is quite different; in this connection yåßå' means 'to go out and away'. The correlative of yåßå' in this context is not bô' but ¡ûb, for we Þnd corresponding to the yåßå' of the year in the autumn the të¡ûbat of the year in the spring (see below, ad 2).
But the important question is whether 'at the end of the year' must mean 'at the end of the calendar year'./21. The festivals are in these passages plainly regulated by the agricultural seasons, not by the lunisolar calendar,/22. so it is a natural supposition that it is the end of the agricultural year that is meant. E. Mahler has with justice observed that the Þnal words of Exod. 23.16 bë'ospëkå 'et-m''ÿêkå min-haådeh, 'when you gather in your produce from the Þeld,' clearly refer to the agricultural year:/23. the year is the year of sowing, harvest, and gathering in./24. There is nothing implausible in postulating two or more systems of years (cf. our Þscal year, academic year, calendar year),/25. and since it is clear that the 'end of the year' referred to in the 'festival calendars' indicates the agricultural years of the seasons, there is no good reason for claiming that Exod. 23.16 refers to a calendar year ending in the autumn.
Even clearer is the evidence of the term tëqûpâ, the 'turn' of the year, in Exod. 34.22. What is referred to here is a point in the course of the seasonal year, and not a point in a calendaric system. True, 'we must not introduce into those ancient texts the notion of solstice and equinox which later Judaism gave to tëqûpâ',/26. but that meaning is not a totally new signiÞcation; it is only a closer deÞnition of a term that in biblical Hebrew meant 'turning point'./27. This sense of tëqûpâ is well illustrated by Ps. 19.7, though the context has nothing to do with the calendar: miqßh ha¡¡åmayim môßå'ô / ûtëqûpåtô 'al-qëßôtåm, 'at (one) end of the sky is his exit / and his turning point is at its (other) end'. What is 'al/28. the western horizon is not the circuit of the sun but the furthest point it reaches in its circuit, the turning point at which it begins its (subterranean) return to the east./29. Sir. 43.7 contains another occurrence of tëqûpâ. Though the verse is obscure, tëqûpâ appears to refer to a point in the cycle of the moon's waxing and waning: wpß/30. '..h btqptw, 'and it wanes at its turning point', lxx fwsth;r meiouvmenoß ejpi; sunteleivaß, 'a light waning at its completion'. The tëqûpâ of the moon would seem to either parallel to or the opposite of the moon's change bh¡tnwtw in v. 8. Thus we may take the tëqûpâ of the year in Exod. 34.22 as the turning point in a seasonal sense, the time of transition from summer to winter.
Another passage where the tëqûpâ of the year possibly refers to the turning from summer to winter/31. is 1 Sam. 1.19-21. The mt has it that after returning from his pilgrimage to Shiloh, 'Elkanah knew Hannah his wife, and Yahweh remembered her, and it came to pass litëqûpat hayyåmªm that Hannah conceived and bore a son'. Several scholars have suggested that the phrase wayëhª litëqûpat hayyåmªm is textually misplaced/32. and should be removed to the beginning of v. 21 as a marker of the time of Elkanah's visit to Shiloh for the harvest festival. The tëqûpat/33. hayyåmªm would thus be equivalent to the tëqûpat ha¡¡ånâ of Exod. 34.22. A similar conclusion is reached by deleting wattahar annâ from v. 20;/34. it is the birth of the child that is then said to occur 'at the turn of the year', twelve months after Hannah's prayer and shortly before Elkanah's second pilgrimage to Shiloh (v. 21). In either case the use of tëqûpâ is linked with the timing of the festival, which is based upon the agricultural year; so the passage has no relevance to the question of the beginning of the calendar year.
Before leaving 1 Sam. 1.19-21, it may be worth considering the possibility that tëqûpat hayyåmªm is not equivalent to tëqûpat ha¡¡ånâ and simply means 'the turn of the days', i.e. the midsummer solstice, when the days start getting shorter. If this is so, the story gains in vividness, for it means that the hitherto barren Hannah conceived her child immediately after her return from Shiloh and gave birth to him just nine months later./35. While an accurate Þxing of the date of the solstice demands some fairly sophisticated observations that were perhaps not possible in ancient Israel, all that is required for this interpretation is the assumption that the lengthening and shortening of the days was a phenomenon known in pre-exilic Israel. If this explanation is correct, the passage has even less to do with questions concerning the calendar than in other interpretations.
A Þnal passage where the tëqûpat ha¡¡ånâ seems clearly to refer to a seasonal turning-point is 2 Chron. 24.23, where, however, it is a question not of the transition from summer to winter but from winter to summer. Here we read that litëqûpat ha¡¡ånâ the army of the Syrians came up against Joash; almost all agree that the time of year in question must be the spring,/36. since that is the season for the commencement of military campagins (2 Sam. 11.1 = 1 Chron. 20.1; 1 Kgs 20.22, 26)./37. This reference to the tëqûpâ of the year in the spring has often been explained as a reþection of the Nisan calender system in use in the time of the Chronicler;/38. tëqûpâ is understood as referring to the beginning of the calendar year. But it is more natural to understand tëqûpat ha¡¡ånâ in reference to the seasonal year, for military campaigns were not mounted according to the calendar date but according to the season. It was when the year 'turned' or changed after the spring rains from the wet season to the dry that armies set forth; and it is this transition, not a change of calendar date, that is referred to also by the term të¡ûbâ (1 Sam. 11.1). Nevertheless, if this interpretation of 2 Chron. 24.23 is not accepted, and it is still maintained that the reference is to the calendar year, this passage is no evidence that the tëqûpat ha¡¡ånâ in Exod. 34.22 must also refer to a calendar year.
It may be concluded that reference to the 'end' (ß't) or the 'turn' (tëqûpâ) of the year in the autumn invariably have to do with the cycle of the agricultural year or of the festival calendar insofar as it is based on the agricultural seasons, and therefore they are irrelevant to the question of the beginning of the calendar year of months.
Ad (2) The të¡ûbat ha¡¡ånâ is generally accepted without question as being in the spring of the year, on the basis of the references to the beginning of military campaigns at that time. N.H. Snaith, however, has dissented from this view, arguing that in Palestine 'no war could possibly begin before the harvest was complete' in the autumn and that the 'return of the year' must mean the late summer./39. But the references he cites (Judg. 6.4, 11; 1 Sam. 23.1), which refer to forays of Midianites and Philistines at harvest time, are no evidence for the time of year at which military expeditions could begin./40. Against his argument that a peasant economy would make warfare impossible before the completion of the harvest, it could well be argued that the round of the agricultural year would make it equally impossible before the completion of sowing in January/February (cf. the Gezer calendar: 'two months are planting, two months are late planting'). In fact, the dating of the të¡ûbat ha¡¡ånâ in the spring, at least in the Chronicler's usage, has been put beyond reasonable doubt by the synchronism between 2 Chron. 36.10 and the Babylonian Chronicle: the biblical text records the taking of Jehoiachin to Babylon litë¡ûbat ha¡¡ånâ, and the Babylonian text dates the appointment of the new king Zedekiah at the beginning of Adar, the month preceding Nisan,/41. thus clearly in the spring./42.
If të¡ûbat ha¡¡ånâ has anything to do with the calendar, it is just as natural to presume that it refers to the beginning of a new year as to the midpoint of the calendar year./43. But in fact, as de Vaux's deÞnition already implies, the 'return of the year' is a description of seasonal change, and therefore concerns an agricultural year rather than a calendar year.
Ad (3) The sabbatical and jubilee years are also agricultural years and not calendar years of months and days. Naturally they begin after the harvest of the old year has been gathered in, but nothing is thereby implied about the beginning of the calendar year. In the priestly legislation, in fact, the beginning of the sabbatical year does not coincide with the beginning of a postulated autumn new year, but falls on the tenth day of Tishri (Lev. 25.9), not on the Þrst. It should also be observed that the sabbatical year is attested after the exile in a period when a spring new year was in force (Neh. 10.32 [Eng. 31]), thus conÞrming that no inferences about the beginning of the calendar year can be made from the date of the commencement of the sabbatical year.
Ad (4) This seems at Þrst a very strong argument for autumnal new year reckoning. But its assumptions need to be examined.
First, it may be doubted whether the inclusive system of reckoning, however normal for single Þgures,/44. is being employed here. It is crucial in this case that the Þgure 7 in 1 Kgs 6.38 forms part of a total:/45. according to 9.10, by the end of 20 years (half his reign, 11.42) Solomon had built the two houses, the temple and the palace, spending seven years on the temple (6.38), and thirteen on the palace (7.1). Now inclusive 7 + inclusive 13 make (inclusive) 19; only exclusive 7 + exclusive 13 make (exclusive) 20. It does not matter, at this point, whether or not the Þgures are authentic; what is being suggested is that the author responsible for employing them was using an exclusive system of reckoning.
Secondly, are the Þgures authentic? Here it is reasonable to distinguish between two series of Þgures. Series A contains the items: 'in the fourth year, in the month (yera) Ziv'; 'in the eleventh year, in the month (yera) Bul'./46. Series B contains the items: seven years in building the temple; thirteen years in building the palace: twenty years in building both: forty years' reign. Now series A dates are, assuming there are no scribal errors, as authentic as anything in the Solomon narratives: they are the only dates given for events of Solomon's reign,/47. and they presumably derive from temple archives; they also employ the old Canaanite nomenclature./48. Series B dates, on the other hand, are quite clearly stylized and symbolic: Solomon reigns a round forty years, half of which are spent in building; in those twenty years he spends twice as much time on his own house as on Yahweh's,/49. and the temple is built in the suitably sacred period of seven years. In other words, the 'seven years' of 6.38 are not to be traced to a temple archive document reckoning up the years of temple building, but are to be attributed to the deviser of the schematic Solomonic chronology.
A further ground for the separation of the seven-year Þgure from the regnal year dates of series A is the awkwardness of the last clause of 6.38, wayyibnhû ¡eba' ¡ånªm, where the implied subject (Solomon) is not the subject of the preceding verb, and where a strict construction of the waw consecutive puts the building (wayyibnhû) after the completing (kålâ);/50. also, the clause is missing in the lxx, though that does not prove a great deal. It is indeed likely that the Þgure of seven years was derived from the regnal year dates of 6.38a; but it is suggested here that the reason why the chronologer reckoned seven rather than eight years for the temple building has nothing to with a Tishri new year system, but is due to the signiÞcance of the number seven. The Þgure of thirteen years for the building of the palace is not apparently derived from any written source, but is simply the result of subtraction of seven from twenty./51. In conclusion, it may be said that the two assumptions of this argument, that an inclusive system of reckoning is employed, and that the Þgure of seven years for the building of the temple is an independent piece of evidence and not the result of some stylized reckoning, are both shown to be ill founded. A Tishri new year reckoning is not ruled out, but it is not positively conÞrmed by the evidence./52.
Ad (5) In discussing the chronology of 2 Kings 22p;23, we must distinguish between the likely course of events in Josiah's reign (as we are able to reconstruct them) and what the deuteronomistic historian believed to have occurred.
In the Þrst place there is now a large measure of agreement that the reforms described in 2 Kings 22p;23 as occurring in the eighteenth year of Josiah were in fact spread over a considerable number of years, both before and after that year of the Þnding of the book of the law. There are, indeed, too many events to Þt not only into a fortnight, but even into six months (the period between an autumn new year and passover). They include, for example, deÞling the high places from Geba to Beersheba, bringing all the priests out of the cities of Judah, visiting Bethel, slaying all the priests of the high places in the cities of Samaria (23.8, 15, 20). So on the grounds of historical plausibility, further supported by the evidence of 2 Chronicles 34 which attributes some of the events of 2 Kings 22p;23 to the twelfth year of Josiah,/53. and by studies of the literary pre-history of the narratives of 2 Kings 22p;23,/54. it can be justiÞably claimed that not all the events recounted between 2 Kgs 22.3 and 23.23 took place in the eighteenth year of Josiah, and that therefore no inference about the month in which that year began can be drawn.
Still, it could be argued that what matters for the interpretation of the calendaric references is not what actually happened, but what the deuteronomistic historian thought happened. Does his account not perhaps presume an autumn new year reckoning? Here several possibilities open up: (i) He thought that all the events between 22.3 and 23.23 occurred within a fortnight (spring new year reckoning); (ii) He thought they all occurred within six months (autumn new reckoning); (iii) He did not think about this question at all, but for purely schematic reasons arranged all his information about Josiah's reforms in a consecutive narrative bounded by the Þnding of the law-book and the celebration of passover./55. If he thought (ii) was possible, it is hard to see why he could not as easily have thought (i) was possible, for there are no explicit indications of the passage of time in the narrative, and there would have been no physical impossibility in Josiah's doing all he is said to have done in a fortnight if the word of a king is thought as good as a deed. So it is quite plausible to suggest that the Deuteronomist could have imagined that all these events occurred within the fortnight. In that case, the dates of 22.3 and 23.23 have no evidential value for the question of the beginning of the year. Nevertheless, much more plausible than (i) or (ii) is possibility (iii). Without entering more deeply into questions of the traditions available to the Deuteronomist and their relation to the Chronicler's sources, it needs only to be accepted as a reasonable probability that the date in either 22.3 or 23.23 is artiÞcial for 2 Kings 22p;23 to be eliminated as evidence for an autumnal new year.
Ad (6) Not many scholars have relied upon this argument for an autumnal new year, since it has been widely agreed that the numbering of the months from the spring as employed in Jer. 36.9 proves that the spring calendar had been introduced into Judah by this time./56. But this has not proved to be an overwhelming objection to the argument, since some advocates of the autumn new year claim that the year was still regarded as beginning in the autumn even when the months were numbered from the spring./57.
But if it is difÞcult to explain why Jeremiah waited nine months or more before having his book read in the temple, it must be remembered that it is no less difÞcult to explain why he waited three months, as he must have done if the regnal years were reckoned from Tishri. Why did he not have Baruch read in at one of the assemblies during the seventh month? E. Auerbach remarks apropos of E. Vogt's statement/58. that the book was probably written in March 604: 'Eine Bombe wie die Buchrolle legt der Prophet nicht fur 9p;10 Monate auf Eis!'/59. But neither does he keep it on ice for three months-unless he was prepared to wait until there had developed the kind of situation that had come about by December 604./60. If Jeremiah could wait three months, he could wait nine months. All that is demanded is the postulation of an appropriate historical stimulus for the writing of the book, and such is provided by the events of 605 as well as by those of mid-604. While advocates of an autumn new year can see such a stimulus in news of the setting out the Babylonian army for the Hatti-land in the early summer of 604 (in Jehoiakim's fourth year on a Tishri reckoning), an equally convincing historical occasion for the writing down of the prophecies is provided by the battle of Carchemish and the accession of Nebuchadrezzar in the previous year (Jehoiakim's fourth year on a Nisan reckoning)./61.
Ad (7) It would certainly be remarkable if Nehemiah in composing his memoirs c. 430 bce had persisted in painfully translating the legal dates of the beneÞcent Persian ruler/62. into a Judaean system that most agree had been abandoned by the end of the seventh century in Judah and was not employed by the deuteronomistic historian, Ezekiel, P, the postexilic prophets, or the Chronicler./63. Thiele thinks that Nehemiah's usage of the Tishri new year system is an expression of a 'spirit of intense nationalism',/64. but he does not explain how Nehemiah happens to be alone among the intense nationalists of the exilic and postexilic ages in expressing his patriotism in this fashion./65. In fact the text of Neh. 1.1 is not above suspicion. It is strange that the name of the king whose twentieth year it is remains unmentioned until we reach 2.1. Yet although the text does seem defective, it must be admitted that none of the emendations or interpretations of it is particularly convincing. Some have suggested that 1.1 originally mentioned the 'nineteenth' year, not the twentieth;/66. but a simple scribal error of 'erªm for të¡a' 'erh seems rather unlikely, and even a sophisticated reconstruction like that of W. Rudolph is little less arbitrary./67. If it is supposed that the memoirs of Nehemiah did not originally begin with a date and that an editor has transferred the year-date of 2.1 to the beginning of the narrative without noticing the chronological problems that were thereby created,/68. we would want to ask whence the editor derived the month date, and if he invented it, why he bothered to do so. An even less probable suggestion is that the date in 2.1 was originally 'twenty-Þfth',/69. which, indeed, Josephus reckoned to be the year of Nehemiah's arrival in Jerusalem,/70. since Neh. 5.14 conÞrms that Nehemiah was appointed governor in the twentieth year of Artaxerxes (cf. also 13.6). Implausible also is the suggestion that the twentieth year was the twentieth year since Hanani's departure from Susa./71. Perhaps the most reasonable solution, though it too leaves unexplained the omission of the king's name in 1.1, is that an original 'nineteenth year' was altered to 'twentieth year' by an editor of the Greek period used to reckoning royal years on the Seleucid system of an autumn new year;/72. to wait from the Kislev of the nineteenth to the Nisan of the twentieth year, a period of Þfteen months on autumnal reckoning, may have seemed to an editor an incredible tax on Nehemiah's patience.
Thus, although no perfectly satisfactory alternative to the date in Neh. 1.1 can be proffered, it is clear that to base upon it an argument about when Nehemiah reckoned the new year began and hence when the pre-exilic new year began, is risky in the extreme. Once again a verdict of non liquet must be returned upon a passage that has at times been conÞdently used to support the theory of an autumn new year.
Ad (8) On the usual dating of P, this interpretation of Exod. 12.2 implies that the date of the new year was still a debated question a century or more later than the time of the adoption of a spring new year in late pre-exilic times,/73. as is most generally agreed. This seems rather implausible. If P is to be dated earlier than the exile, the same objection does not apply. But the fundamental problem that remains, whatever the date of Exod. 12.2, is that there is no evidence for interpreting it as a polemical defence of a new calendar. It can with no less justice be regarded as a natural attribution to Moses of a signiÞcant Israelite institution, the calendar. Its place at the head of the passover law is perfectly intelligible, for just as the passover marks the beginning of the people of Israel, so the month in which it is celebrated marks the beginning of the year. There are, therefore, no grounds for Þnding here evidence for an autumn new year.
Ad (9) This is not the place to embark upon a full-scale examination of the merits of the competing chronologies of the kings of Israel and Judah that have been offered,/74. nor to test Thiele's claim that only on the supposition of Tishri new year reckoning in Judah will the chronological data and synchronisms 'work'. Here only some theoretical observations can be made, together with a report on a sample probe of the evidence relating to one circumscribed period that I have undertaken.
First, it may be observed that when there are at the disposal of the scholar principles such as co-regency, alternation of accession- and non-accession-year systems, and variant dates for the new year, which are fundamentally arbitrary (i.e. can be called upon when the researcher needs them in the interest of the theory), it is possible that actual scribal errors in the text could be covered up with resultant errors in the chronology.
Secondly, it is remarkable that virtually no scribal errors have, according to Thiele's reconstruction, occurred in the many relevant Þgures found in the text. A text like Ezra 1 // Nehemiah 7, preserved in parallel transmission, shows that a signiÞcant number of scribal errors are likely to have occurred./75.
Thirdly, the fact, assuming that it is a fact, that regnal years and synchronisms 'work' on the basis of a Tishri new year in Judah does not preclude the possibility that also on a Nisan system the Þgures, or most of them, will 'work'./76. What is required before the Nisan system is ruled out altogether is a testing of all the possibilities using the full range of variables (co-regency, interregnum, calendar and new year reckoning), with certain given and pre-determined data (regnal year, internal and external synchronisms, plausible upper limit of scribal errors,/77. plausible age of a king at the birth of his Þrst child,/78. etc.).
Fourthly, even if the synchronisms break down on the supposition of a Nisan new year in Israel and Judah, a Tishri new year is not the only alternative. S. Talmon, for example, has maintained that in Israel the year began one month later than in Judah, and has explained discrepancies in the biblical dates on that basis./79.
Fifthly, an investigation of the chronological data relating to the years 609p;587/6 bce with a view to determining whether a Nisan or a Tishri system of reckoning was in operation has shown that all the data, comprising material from 2 Kings, Jeremiah, Ezekiel, and the Babylonian Chronicle can best be harmonized on the supposition of the Nisan new year reckoning./80. Most of the arguments advanced in favour of an autumnal new year are inconclusive, and only Dan. 1.1, a verse that contains problems of its own, demands a Tishri system. On the other hand, the date of the battle of Carchemish given in Jer. 46.2, which we know from the Babylonian Chronicle occurred before Ab 8, can be synchronized with Jehoiakim's fourth year only on a Nisan reckoning. Further, a calculation of the probable date of Jehoiakim's accession to the throne makes it more likely than not that his regnal years were reckoned from Nisan. These results prove nothing about the reckoning of years earlier than 609, but at least they show that during the period examined it is the Nisan rather than the Tishri system that 'works'.
It may be that in the end Thiele's system will prove to be correct. If it is, what should be pointed out here is that it is the only surviving evidence for a Tishri new year in pre-exilic Judah. That is in itself no argument against Thiele's chronology, but it is as well to know on what grounds the almost universally accepted view of an autumn new year may be sustained.
Ad (10) The difÞculty with this argument is that it proves too much. For the Gezer calendar does not simply attest a year beginning in the autumn; it attests a year beginning with 'ingathering' ('sp) which is the Þnal element in the ancient Israelite festival calendar. Unless we are to postulate two variant systems of autumnal new year reckoning, it seems necessary to maintain that neither the Gezer calendar nor the festival calendars are calendars in the strict sense of the word. A calendar is a list of months (and days) in which the distinction between the Þrst month and the twelfth is clearly marked; obviously, it must begin with the Þrst month. A list of seasons, on the other hand, which the Gezer calendar plainly is, does not need to begin at a particular point in the year, although, of course, in the Palestinian context it is likely to begin about the time of autumn, the most conspicuous transitional point in the seasonal year. The structure of the Gezer calendar, which is arranged according the activities of the farmer's year and not integrated with month names or even divided into twelve periods, makes it unlikely that it is an ofÞcial calendar,/81. and unlikely also that it should be regarded as a calendar at all in the sense that an inference about the beginning of the civil year can be made from it.
3. Evidence for a Spring (Nisan) New Year
Finally we may consider whether there is any evidence in favour of a spring new year in pre-exilic times. Here it must be said that if the data for an autumn new year have proved inconclusive, the evidence for a spring new year is also far from cogent. Three arguments may be advanced:
(1) The four lists of festivals (the so-called festal calendars contained in Exod. 23.34, Deuteronomy 16, and Leviticus 23), which range in date, according to the conventional analysis, from the pre-monarchic to the postexilic period, are at one in beginning the enumeration of festivals with the spring festival of passover. This would be strange if the only year known in pre-exilic Israel began in autumn. It would be more natural if that were the case for the autumn festival to head the list. This implication could be evaded only if the phrase 'at the end of the year' in Exod. 23.16 is interpreted narrowly to mean that the new year began after the end of the autumn festival/82. (contrary, of course, to the priestly calendar). But many scholars claim that 'the end of the year' is a vague term and afÞrm that the autumn festival celebrated both the end of the old year and the beginning of the new./83. It that were so, would the autumn festival not be expected to hold Þrst place in the list?
But the argument so far begs one question: Have the seasons of the festivals anything to do with the calendar year? I have argued above (ad 1) that as agricultural festivals they are related in the very early lists of Exodus 23 and 34 to the agricultural, not the calendar, year, as they have been by the time of priestly law of Leviticus 23. The one exception to this claim is the speciÞcation of the month Abib as the date of the festival of unleavened bread (Exod. 23.15; 34.18; cf. Deut. 16.1). It does not matter for our present purpose whether or not the reference to Abib is original in the text of Exodus 23 or 34: it is enough that it appears in a form of the festival law that dates from some period in the monarchy. Now the month Abib is surely not speciÞed in order to prevent unleavened bread or passover from being celebrated at some other time in the year, for the same need to specify the month is not felt in the case of the other festivals. Can it be that Abib is speciÞed because it is a more signiÞcant month than the months in which the harvest and ingathering festivals fall, that is, because it is the Þrst month of the year of months? Little weight can, indeed, be placed on this argument, but some explanation of the mention of this month alone seems to be called for.
(2) A certain amount of evidence has been produced for characterizing the spring festival, passover, as a new year festival. J.B. Segal has most recently presented an impressive range of parallels in ritual and ideology between new year festivals in other ancient near eastern cultures and the passover in Israel and has concluded that 'to the Hebrews the Passover was primarily a New Year Festival'./84. But quite apart from the dubiety of the assumptions inherent in any patternist approach/85. (of which Segal's work must be acknowledged as one of the most cautious examples), his case is largely undermined by the fact that equally remarkable parallels may be drawn between the autumn festival of tabernacles and new year themes;/86. and Segal himself does not deny that tabernacles was also a new year festival./87.
(3) The numbering of months from the spring is prima facie evidence of a spring new year. This argument can be countered by postulating the adoption of the Babylonian spring calendar earlier than the Þrst reference to the numbering of months (Jer. 36.9, if the references in 1 Kgs 6.1, 38 are removed as glosses). It is most improbable, however, that the months should be numbered from spring by a society that not only still used an autumn calendar but had always had an autumn calendar. To that extent the argument of Thiele and others, that autumn new year reckoning persisted throughout the monarchy even when the months were plainly being counted from the spring,/88. seems quite implausible.
The conclusion of this study is that, while there are no data that categorically exclude autumnal reckoning of the calendar year prior to c. 605 bce, there are no data that support it, not even cumulatively. On the other hand, there is one piece of evidence (the reference to Abib in the festival 'calendars') that may suggest spring new year reckoning, but that admittedly does not amount to a strong argument in favour of such reckoning. What can be claimed, however, is that it can no longer be conÞdently afÞrmed that in pre-exilic Israel the calendar year began in the autumn, nor can it be said with conÞdence that the Babylonian calendar was adopted c. 605 bce, nor can it be assumed that the reckoning of months according to a spring new year can be used as a criterion for dating a document in which it occurs.
open footnotes
go to list of articles