William W. Hallo
[William W. Hallo is Laffan Professor of Assyriology and Babylonian Literature at Yale University and curator of the Yale Babylonian Collection. This article was the Nelson Glueck Memorial Lecture in Bible delivered at the Hebrew Union College-Jewish Institute of Religion at Cincinnati, Ohio, on March 7, 1977.]
Since Dr. Glueck’s untimely death, the particular combination of archaeological and theological interests which he championed, indeed the entire comparative approach to Biblical Studies, has been subjected to ever more serious challenges. The comparative method has been attacked as a form of “pseudorthodoxy” by Morton Smith, and more recently some of its most cherished results have been demolished, point by point, in the works of John van Seters and Thomas L. Thompson, particularly in respect of the patriarchal narratives and their possible historicity. Even those of us who have heeded Benno Landsberger’s strictures on the “conceptual autonomy” of each of the principal Ancient Near Eastern cultures, and Samuel Sandmel’s valid warnings against “parallelomania,” have nevertheless been suspected of reactionary tendencies — as if we wanted to reduce Assyriology once more to the role of handmaiden to Biblical Studies — or vice versa.
Nothing could be further from the mark, at least as far as my own intentions are concerned. I have defended and applied the comparative method in numerous studies, some literary and others historical, in which it seemed to me that cuneiform sources and biblical texts could fruitfully illuminate each other. But for me the method requires only the commensurability
BSP 9:3-4 (Summer-Autumn 1980) p. 102
of the two terms, not a prejudgement as to their equation. If A is the biblical text, or phenomenon, and B the Babylonian one, I am quite prepared to test the evidence for a whole spectrum of relationships, expressed “mathematically” not only by A = B but also by A ~ B or A < B or A > B and even A ≠ B. The last possibility needs stressing, because a comparative approach that is truly objective must be broad enough to embrace the possibility of a negative comparison, i.e. a contrast. And contrast can be every bit as illuminating as (positive) comparison. It can silhouette the distinctiveness of a biblical institution or formulation against its Ancient Near Eastern matrix. It is perhaps unfortunate that scholarly usage has tended to slight this notational breadth implied in true comparison, that it has the well-worn symbol cf. (confer) for positive comparison but none for negative ones. Perhaps one may suggest cs. for contrasta.
Be that as it may, I would propose the “contrastive method” as a particularly valid one within the overall comparative approach to biblical studies. Apart from numerous contrasts in detail, consider only what is perhaps the single most pervasive contrast of all: the role of the king in the Ancient Near East. Over and over again it contrasts with the bearers of that role in the biblical ethos. As source, warrant and enforcer of law, the Mesopotamian king is replaced by the biblical deity, as chief ministrant by the high priest, as responsible for both cultic and ethical observance initially by Israel as a whole and later by each individual Israelite. But it is not my intention to pursue this illustration here. Rather, I wish to test my method against a subject less transcendent and more immanent, and one where ancient Israel’s unique contribution to the social order of all subsequent ages has too often been obscured by facile abuse of the comparative approach. I refer to the calendar of ancient Israel or rather that part of its cultic calendar which, in the form of the week, has become a virtually universal legacy to posterity.
My thesis in brief is this: the cultic calendar of ancient Mesopotamia, like its civil calendar, was largely tied to the phases of the moon, and not at all to the week (or: a week); in Israel, the cultic calendar was only minimally connected to lunar phases, whereas the sabbatical cycle was all-important. Certain conclusions will be deduced from this contrast. To begin my case, I must ask you to indulge me in a somewhat technical run-down of the Mesopotamian evidence for “lunar festivals.” It is abundant, but I will present it in all possible brevity.
BSP 9:3-4 (Summer-Autumn 1980) p. 103
Lunar Festivals in Mesopotamia
Our first detailed records of cultic practices in Mesopotamia date from the end of the Early Dynastic Period in the 24th century B.C. These records already include references to the “house (or chapel or station) of the crescent” called u4-sakar in Sumerian and later borrowed into Akkadian as uskaru or askaru. The general sense of this term can be fixed as crescent.
Special offerings for this new moon festival are attested at pre-Sargonic Lagash (both at Sirara and at Uruku) but there is not enough evidence to demonstrate a monthly observance of the rites. The same is true of the Sargonic period. From Nippur, we have a number of offerings for the new moon festival dating from the early part of the great Sargonic dynasty, and one even specifies that the offerings were made by Sargon himself. In these texts, the new moon festival is called “head of the crescent.” Under his grandson, Naram-Sin the great, the expression “head of the month (by name)” was introduced.
It is, however, only with the end of the Sargonic period in the 22nd century and the beginning of the neo-Sumerian period in the 21st that we find evidence of a regularized lunar festival. In the celebrated cylinders of Gudea of Lagash, we are introduced to the generalized Sumerian term for lunar festival š-š literally perhaps “all sanctuaries,” in the form of the compound verb š-š. .. ak, “make, do, perform the eššeššu-festival.” This compound recurs in one of the most popular royal hymns of the succeeding Ur III period, the hymn known today as Šulgi A. Šulgi, greatest of the Ur III kings, boasts there of having run from Ur to Nippur (and back) in one day, so that, as he says “verily I celebrated the eššeššu-festivals of (both) Nippur (and) Ur on one and the same day.” We can actually date this improbable achievement to the 6th year of Šulgi’s long reign (ca. 2089 B.C.), for the following year was named after it, in an interesting and early instance of what I take to have been the institutionalized commemoration of events in hymns, royal inscriptions and date formulas at the beginning of every year or every other year in the Classical phase of Mesopotamian civilization. But the new terminus technicus was not restricted to literary contexts. Among the Ur III economic texts at Leiden is one otherwise undistinguished little tablet which records the assignment of nine assorted sheep and goats to the high priestess (nin-dingir) at Lagash on two separate occasions in the first year of Šulgi’s son and successor Amar-Sin. One of these occasions involved five of the principal deities of Lagash and their sanctuaries and is labeled “lunar festival of the high priestess performed on the (day of the)
BSP 9:3-4 (Summer-Autumn 1980) p. 104
chariot.”
This laconic entry is a precious clue to the numerous references in other Ur III accounts which allude to offerings for the a š-š in general, or specifically for the rites of the new moon (u4-sakar gu-la), the first crescent (literally the “chariot of the 7th day”) or the full moon (literally the “crescent of the 15th day”).
It may be noted at the outset that only three lunar phases were observed in this period, i.e. no special account was taken of the 3rd crescent. Secondly, the archives of the various cities used a variety of different terms to identify the specific phases and their celebration. The new moon was variously called “great crescent at (or of) the head of the month” (u4-sakar sag-iti-gu-la), “great crescent” (u4-sakar gu-la), “head of the crescent” (sag-u4-sakar[a]) or simply “(the) crescent” (u4-sakar). The first crescent was called the “station (é) of the 7th (or 6th) day” and, more rarely, “the chariot of the 7th (or 6th) day.” The full moon occurs as “crescent of the station of the 15th day” or simply “station of the 15th day.” The explanation for the introduction of the term “chariot” appears to rest on the symbolic identification of the half-moon with the two semi-circular blocks of wood which were joined to make up the solid chariot wheels typical of this period.
The cultic dispositions for these festivals are abundantly attested. Numerous comestibles are provided for, including cakes, oil, beer, etc.; large and small cattle are sacrificed; there are ritual ablutions, and special garments including sandals are issued for the occasion. The specific allotments differ again as between the different cult-centers, but only one of them can be highlighted here, and that is the capital of the Ur III empire, the city of Ur itself. Here we find a peculiar historical situation revealed by the terse economic texts which, though their intent is simply to record cultic expenditures against a future audit of temple stocks, nevertheless provide us with a kind of after-the-fact or descriptive ritual record comparable in some ways to certain biblical passages like Numbers 7 which preserve for all time the pious offerings dedicated to the sanctuary. At the end of the 21st century, the Ur III empire suffered a series of reverses and lost one after another of its outlying provinces. Its last king, Ibbi-Sin, retreated to Ur together with the deposed governors of the lost provinces. But in spite of these straitened circumstances, the cult was maintained together with the meticulous bookkeeping that characterized it. Particular attention was paid to that peculiar neo-Sumerian institution, the amphictyony. In its heyday, this institution provided for the temples of the religious capital at Nippur by monthly liturgies levied
BSP 9:3-4 (Summer-Autumn 1980) p. 105
on all the central provinces through their governors. Now, however, the deposed governors simply offered pitifully small numbers of sacrificial animals to the two principal sanctuaries at Ur in the context of the “royal lunar festivals” — but were still dignified with the designation “governor of the amphictyonic obligation” when they did so.
From the hapless Ibbi-Sin, Ur passed into the successive rule of the kings of Isin, Larsa, Babylon, and the Sealand. From the middle of this lengthy period, two texts may be cited here to illustrate both the continuity and the evolution of the lunar festivals. Both texts were edited by Baruch Levine and myself in the HUC Annual for 1967 and make it graphically clear that the three lunar festivals of the Sumerian tradition survived intact and in the interval, they had been augmented by a fourth festival falling on the 25th day of the lunar month. The 25th day cannot be readily correlated with a phase of the moon, however, and the way was thus opened for detaching the eššeššu-festival from its original lunar association. In the subsequent evolution, this detachment was carried further. Hemerologies and other literary texts of the late second and early first millennium changed the eššeššu-festivals to the fourth, eighth, and seventeenth days of the month, and by the second half of the first millennium, there were as many as eight eššeššu-festivals per month.
It might be thought that we have pursued the topic into a cul-de-sac. But as is often the case when tracking down a millennial institution, we must allow for the survival of an old institution under a new name. In the first place, the terms for the individual lunar phases were translated into Akkadian and appear together in the Old Babylonian Atrahasis epic as arḫu sebūtu u s̆apattu “first, seventh, and fifteenth day of the month.” In the second place, the concept of lunar festivals in general did not simply die out when the original Sumerian term changed its meaning. Rather, I submit that it reemerged in Akkadian guise in the form of the ḫitpu-offerings. The root htp is familiar in Arabic, Hebrew and Aramaic in connection with slaughtering or hunting and occurs already in a Ugaritic text in the specific context of sacrifice. We read, in a prayer to Baʿal, “A bull, oh Baʿal, we consecrate (to you), a votive offering, oh Baʿal, we dedicate (to you), the first fruits, oh Baʿal, we consecrate (to you), the booty, oh Baʿal, we offer (to you), a tithe, oh Baʿal, we tithe (to you).” The word translated “booty” here is htp, comparable to the hetep (“prey”) of Proverbs 23:28. Given the context, however, the word may already foreshadow the connotation of a kind of sacrifice. That is surely the meaning of Akkadian ḫitpu which occurs in monthly sacrificial lists of the late first millennium. This neo-Babylonian ḫitpu offering typically fell on the 7th, 14th, 21st and 28th day of the month in the Achaemenid texts
BSP 9:3-4 (Summer-Autumn 1980) p. 106
from Uruk; sometimes they are entered one (or even two) days earlier. Albert T. Clay published a sizeable number of such lists from the Yale Babylonian Collection; additional examples remain to be published from there or have appeared meanwhile in other collections. Clay himself edited the texts under the provocative title of “The Babylonian Sabbath.” The debate which he thus inaugurated was carried forward, not without some acrimony, by A. L. Oppenheim and J. Lewy on the basis of a rather different line of evidence, namely the lexical lists, which need not concern us here, for it is now clear that this adds nothing to the case.
Biblical Evidence for Lunar Festivals
Let me rather turn now to the biblical side of the equation by asking you to consider what evidence there may be, first of all, for festivals based on lunar phases in ancient Israel. I think you will agree that it is fairly minimal. There is of course, the observance of the new moon: “Sound the horn on the new moon” we are told in Ps. 81:4—but whether kese in the continuation of the verse means that the full moon is also to be celebrated may be and has been doubted. The ritual calendar of Numbers 28:11–15 prescribes the monthly offerings of the new moon, and other aspects of its observance are alluded to elsewhere in the Bible as when Hosea or Isaiah condemn the hypocritical observance of “new moon and sabbath” in one and the same breath. But only the first day of Tishri had the character of a special holiday, and even here the biblical text, as is well known, avoids the term rōs̆ has̆s̆ānā, head of the year. In passing it may be noted that the Babylonian month Tashrîtu was similarly singled out for special treatment in the Akkadian hemerologies.
No special treatment was accorded to the first crescent (or the last) in the biblical calendar; indeed we do not even know a special term for these days — and certainly that is not for want of words to describe the lunar crescent or its iconographic representations. As for the full moon, the two principal “seasonal festivals” were gradually related to it in some way. To this day, the approaching full moon signals the coming of Pesach in Nisan and Sukkoth in Tishri and comes as close as a luni-solar calendar can to marking the vernal and autumnal equinox respectively. But there are ample hints to suggest that this is the terminal stage of a long evolution, and that at an earlier stage, the spring festival was celebrated on the new moon of the month of Nisan or Aviv, or throughout its first quarter, and the fall festival may similarly have fallen on the new moon of the seventh month before some of its functions were assumed by the New Year’s festival. Thus, there is little in the ritual calendar of the Bible to compare with the persistent importance of moon-worship and the
BSP 9:3-4 (Summer-Autumn 1980) p. 107
celebration of the various lunar phases that we encounter in Mesopotamia.
The Sabbatical Conception in the Bible
Now contrast, if you will, the case for the sabbatical conception in the Bible. Nothing could be more persistent. As early as the time of Solomon, we are entitled to detect a seven-day cycle in the festivities marking the dedication of his temple. The double injunction to work for six days and to rest every seventh is the most fundamental piece of social legislation written into the Decalogue — though it is a moot point which of its two provisions is more often violated. The cultic counterpart of this legislation permeates every one of the many ritual calendars in the Pentateuch. There is nothing in the biblical evidence to suggest that the sabbatical conception depended in any way on the luni-solar calendar; on the contrary, that calendar was progressively adjusted to the sabbatical conception and its final post-biblical determination manages to avoid such conflicts as Yom Kippur falling on Friday or Sunday.
Already in biblical times, moreover, the sabbatical conception was extended beyond its ritual application to the week. We see this first in the feast of weeks, originally a seasonal festival without fixed date, to be celebrated fifty days after the offering of the first sheaf of the barley harvest which in its turn was prescribed for the “morrow after the Sabbath” (Lev. 23:15) — a date that gave rise to much debate. Beyond even this concept of a week of weeks, the Bible decreed a “week of years” as the foundation of the system of S̆ʿmiṭṭā, or Sabbatical year, and “seven weeks of years” as the basis for Yōḇēl, or jubilee year.
I would not enter into the details of all these institutions even if I could, but cite them only as evidence for the extraordinary extent to, which the sabbatical idea permeated all biblical legislation. The week was fundamental to the biblical calendar and immune to violation by any other consideration, least of all the phases of the moon. Even though these consist of ca. 7 3/8 days each, the biblical week is wholly independent of them. I must, however, ask you to return with me for a moment to Mesopotamia in order to drive home this point.
The Search for Mesopotamian Antecedents to the Biblical Week
Scholars have never tired of looking for Mesopotamian antecedents to the biblical week. The seven-day festival celebrating Gudea’s dedication of the great temple at Lagash has been studied in detail by Sauren and compared by some scholars to Solomon’s aforementioned one, as has “A seven-day ritual in the Old Babylonian cult at Larsa”. But these
BSP 9:3-4 (Summer-Autumn 1980) p. 108
festivals were, to all appearances, one-time occurrences and owe their particular seven-day duration at most to the wide-spread popularity or sanctity of the number seven, to which whole monographs have been devoted. They tell us nothing about the institutionalization of a seven-day week on a regular basis. On the contrary, when we look at non-cultic cuneiform texts, we see again that Mesopotamia, in its classical phase, organized its days of labor and rest by the month, not the week. For the Ur III period, for example, we have long tallies of hired or conscripted labor, and these indicate clearly that as many as six days per month were set aside as rest-days (u4-du8-a). Thus, for example, the records of the great textile establishments at Umma specify that for 3000 man-hours (or rather women-days, since slave girls were involved and the input was counted in days), 500 had to be set aside as free time. Another text from Umma calls for slave girls to work for 23 days in each of eleven successive months, presumably when they were hollow (i.e. 29 days in length) and perhaps 24 when they were full (i.e. 30 days in length). Just how these 6 rest-days were arrived at is clear from an essay describing the life of the Sumerian school which specifies, in Kramer’s translation, “Here is the monthly record of my attendance in school, my vacation days (u4-du8-a-mu) each month are three, my recurrent(?) monthly holidays (ezen-aš-aš-bi) are three, (that leaves) twenty-tour days each month that I must stay in school — (and) long days they are.” This must surely mean that in a month of 30 days, 3 days were set aside for the lunar festivals and three for other reasons. By Old Babylonian times, these generous schedules were even exceeded if we are to believe royal claims that corvée duty was reduced to 10 or even 4 days per month.
Briefly we must also cast our eyes northward to Assyria, for there the Lewy’s and others were convinced they had found “The origin of the week and the oldest West Asiatic calendar,” But subsequent research has not borne out their estimate of the length of the so-called ḫamus̆tum, either as a fifth of a year or else as 50 days, or even as seven days, or indeed the whole complex theory of pentacontad cycles which Morgenstern took over from them. According to the latest synthesis of the evidence, the ḫamus̆tum represented a time-span much closer to the biblical week — perhaps one-fifth of a month (i.e. 6 days), or a week of 5 days. In any event, the system of this Old Assyrian week, in use in different forms both at Assur and in the Anatolian colonies, was not demonstrably independent of the lunar calendar: what few indications there are suggest that the first day of such a week coincided with the beginning of the month in eight months of the luni-solar calendar, and with the full moon in the other four.
BSP 9:3-4 (Summer-Autumn 1980) p. 109
An institution shared by Assyria and Babylonia in this, the “patriarchal” period, was the periodic remission of debts, and liberation of distrainees, or debt-slaves. Both of these prerogatives have hoary Sumerian antecedents, nig-si-sá. .. gar being attested at least as early as the laws of Ur-Nammu of Ur, and amar-gi4 already under Entemena of Lagash. Their Akkadian equivalents, mīs̆aru and andurāru, are cognate to Hebrew mēs̆ār (mīs̆ōr) and d’rōr respectively. The latter term is used in Leviticus 25 in the injunction to “proclaim d’rōr throughout the land, to all the inhabitants thereof,” as well as in prophetic allusions to the jubilee. It has naturally been compared to the Akkadian andurāru or durāru, most convincingly in Julius Lewy’s study on “The Biblical institution of ‘derôr’ in the light of Akkadian documents.” Recent research has permitted refinements of our understanding of the Akkadian concepts but what is crucial in the present connection is whether the underlying institution already displayed the periodic or cyclical character of the biblical jubilee. I can only cite here J. J. Finkelstein’s conclusion for the Old Babylonian mīs̆arum-enactments. Although it is not yet possible to establish any fixed interval of recurrence for them, he concludes that probably, “apart from the period of the royal accession, (they) took place also at fairly regular intervals thereafter.” These intervals vary from six to eleven or even 21 years, but in any case they were dependent on the accident of the royal succession. They thus provide a remote model of sorts for the biblical institution, but hardly for its integration into the sabbatical cycle.
I forebear to pursue yet other and weaker proposals to find Mesopotamian antecedents to the biblical week or the sabbatical cycle. And it is hardly necessary to pursue even more improbable antecedents in the literature of Canaan or the calendars of Egypt or Greece, as illustrated by the so-called “planetary week.”
The Uniqueness of the Biblical Concept of the Week
In sum, the uniquely biblical conception of the week and the sabbatical cycle stands out equally by virtue of its pervasiveness in biblical laws and letters, as by its absence from the surrounding Near East. What conclusions are we entitled to draw from these somewhat tedious observations? Time permits only an apodictic formulation. The sabbatical concept in its biblical form is fundamentally an expression of social or socioeconomic justice and natural (I would almost say: ecological) equity. The inviolate recurrence of a day of rest is ordained in the context of the commandment to work the other six; the sabbatical year is part and parcel of a planned agricultural economy; the jubilee definitely serves to preserve the independence of the small farmer in the face of an emerging
BSP 9:3-4 (Summer-Autumn 1980) p. 110
urban-royal society; even if its actual application fell into early disuse, the same effect continued to be achieved by the cancellation of debts in the sabbatical year.
Equally unique in all the Ancient Near East, however, is the manner in which these essentially socio-economic provisions were woven into the cult. The intermingling of ethical and cultic prescriptions is, of course, a well-known characteristic of biblical legislation, and one that clearly distinguishes it from the legal corpora of Babylonia, Assyria, and Asia Minor, where civil or criminal law and ritual prescriptions are kept carefully apart. But the intermingling is particularly profound in the case under discussion, beginning with its very motivation. The Sabbath is motivated by the divine work of creation; land-redemption is a guaranteed right “because (God says) mine is the land” (Lev. 25:23); and the release of the debt-slaves in the jubilee is justified on the grounds that “to me the children of Israel are slaves, they are my slaves (Lev. 25:42, 55). Beyond motivation, the Sabbath and the sabbatical cycle are woven deeply into all Pentateuchal legislation.
Finally, the importance of the Sabbath must be seen in the context of the continuing deemphasis of the lunar festivals. Moon worship flourished wherever Mesopotamian culture spread, and even after its demise it survived at places like Harran. But in Israel it failed to gain a foothold; the full moon was not worshipped, the quarters were not specially observed, and even the new moon was ultimately relegated to the status of a half holiday. We may sum up the contrast as follows: the ancient Mesopotamian year was based on the month, and the worship of the moon went hand in hand with it. The Israelite year was based on the week, and remained independent of the month even when the luni-solar calendar was adopted from Babylonia. The Mesopotamiam jubilee was based on the royal succession and on royal whim; the Israelite jubilee was (at least ideally) ordained by God in inviolate successions of sabbatical cycles. Here, then, two of the great contrasts between biblical Israel and its Near Eastern matrix meet: sabbatical cycles versus lunar calendars, and divine versus royal authority. The legacy of these contrasts is with us to this day.
(This article was abridged by the editor. The full version, with extensive documentation, appeared in the Hebrew Union College Annual, Vol. 48 [1977] pp. 1-18. Used by permission.)
BSP 9:3-4 (Summer-Autumn 1980) p. 111