A critical review of Rolf Furuli’s 2^nd volume on chronology:

Assyrian, Babylonian and Egyptian Chronology. Volume II of Assyrian, Babylonian, Egyptian, and Persian Chronology Compared with the Chronology of the Bible (Oslo: Awatu Publishers, 2007)

Part IV:  The Neo-Babylonian Ledger NBC 4897

© Carl Olof Jonsson, Göteborg, Sweden, November 2008

 The cuneiform tablet NBC 4897 is a ledger, tabulating the annual growth of a herd of sheep and goats belonging to the Eanna temple at Uruk for ten consecutive years, from the thirty-seventh year of Nebuchadnezzar to the first year of Neriglissar. As it is an annual record, it clearly shows that Nebuchadnezzar ruled for 43 years, his son Amēl-Marduk for 2 years, and that the latter was succeeded by Neriglissar. The tablet makes it impossible to insert any extra years or any extra kings between Nebuchadnezzar and Amēl-Marduk, or between Amēl-Marduk and Neriglissar. This is strong evidence, indeed.

The first presentation and discussion of the tablet was included in an article written by Ronald H. Sack, “Some Notes on Bookkeeping in Eanna,” published in M. A. Powell Jr. and R. H. Sack (eds.), Studies in Honor of Tom B. Jones (1979). It was a brief, preliminary study of just five normal-sized pages (pp. 114 -118), three of which contain a drawing of the tablet.

Another discussion of the tablet appeared 16 years later in an article written by G. van Driel & K. R. Nemet-Nejat, “Bookkeeping Practices for an Institutional Herd at Eanna,” Journal of Cuneiform Studies, Vol. 46 (1994), pp. 47-58. It was a somewhat longer study of 12 large-sized pages, six of which contain a drawing, transliteration and translation of the tablet. Their article corrects a number of errors and misinterpretations by Sack.

The most extensive and detailed discussion of the tablet, however, is Stefan Zawadzki’s article, “Bookkeeping Practices at the Eanna Temple in Uruk in the Light of the Text NBC 4897,” Journal of Cuneiform Studies, Vol. 55 (2003), pp. 99-123. Zawadzki’s discussion covers 25 large-sized pages, four of which give a transliteration and translation of the tablet. The article contains the most detailed and careful examination of the tablet so far. He corrects a number of misreadings and misinterpretations in the previous articles by Ronald H. Sack and G. van Driel/K. R. Nemet-Nejat.

Do the total numbers on the tablet contain serious mistakes and miscalculations?

Although van Driel and Nemet-Nejat corrected many misinterpretations and misreadings by Sack, they also claimed that the interpretation of the tablet “is hampered by miscalculations and mistakes in the text.” (Van Driel/Nemet-Nejat, p. 47) Their conclusion at the end of their article (page 57) is quoted approvingly by Rolf Furuli, who claims that it “highlights the lack of quality of this tablet”:

“For the most part, mistakes occur in the totals. The scribes probably had difficulties similar to ours in reading the numbers in their ledgers. We can understand small mistakes of a single digit, but the mistakes occurring in the crucial final section of NBC 4897 again raise the question of how the administrations could work with this kind of accounting.” – Quoted by Rolf Furuli in his Assyrian, Babylonian, and Egyptian Chronology, pp. 247, 248 (2007 ed.; pp. 251, 252 in the 2^nd ed. of 2008).

As is demonstrated by Zawadzki, however, these claims are much exaggerated. The fact is that they are mainly based on misreadings and misunderstandings by the authors. As Zawadzki explains, van Driel “has solved many problems, yet he has failed to explain several significant points, or has proposed interpretations that require reevaluation.” (Zawadzki, p. 100; emphasis added) In fact, when the tablet is correctly read, copied, understood and translated, it can be shown to contain very few errors “in the totals”, and these are small and unessential and do not occur “in the crucial final section of NBC 4897” as van Driel/Nemet-Nejat state.

Concerning the claim that the mistakes for the most part “occur in the totals”, the most serious of these according to van Driel/Nemet-Nejat’s translation are found in lines 31 and 35, where the numbers of sheep (rams + ewes + male lambs + young ewes) are summarized as follows:

Line 31: 170 + 390 + 66 + 193 = total: 759.

Line 35:     5 + 198 + 14 +   51 = total: 198.

As van Driel/Nemet-Nejat observed (pp. 53, 57), the numbers they have read in line 31 add up to 819, not 759, and those in line 35 add up to 268, not 198.

With respect to line 31, however, Zawadzki notes that, “Van Driel reads mistakenly 193 lambs while the copy gives clearly 133. The horizontal total of 759 is correct. Thus his calculations in JCS 46, [page] 57 from point (3) to the end of the article [i.e., the whole last page of the article] are wrong.” (Zawadzki, p. 104, note 23)

Line 35 contains two further misreadings: The number 198 is a misreading for 138 (Zawadzki, p. 104, n. 25) and number 51 is a misreading for 41. Paul-Alain Beaulieu, who collated the original tablet at Yale, comments, “The tablet has a clear 41, indeed, but the scribe has written 51 and then erased one of the Winkelhaken to make 41.” (Zawadzki, p. 104, n. 26) The horizontal total of 198 in line 35, therefore, is also correct.

Thus there are no errors “in the crucial final section” of the tablets. When the individual figures have been correctly read, copied and translated, and the procedure used by the accountant to arrive at the “totals” and the “Grand totals” is correctly understood, the calculations of the accountant turn out to be surprisingly free from serious errors. At only two places the “Grand totals” contains errors, and these are very small. For the 37^th year (line 5) the “Grand total” shows 176 animals instead of 174, and for the 40^th year (line 14) it shows 303 animals instead of 306. For all the other eight years the calculations are correct!

In view of this, it is remarkable that Rolf Furuli in his attempt to undermine the chronological impact of NBC 4897 has devoted so little attention to Zawadzki’s careful analysis of the ledger that he has failed to notice that his quotation from van Driel/Nemet-Nejat about the supposed numerical mistakes on the tablet has been refuted by Zawadzki!

Table 1 below, which is based on Zawadzki’s study, summarizes the calculations in the ledger,  demonstrating that the Neo-Babylonian accountant usually did an excellent job and that the few mistakes he did in his calculations of the annual increase of the herd were of very small consequence.  In the table “BF” means “brought forward” and “CF” means “carried forward.” “Nbk” means Nebuchadnezzar, “AmM” Amēl-Marduk, and “Ngl” Neriglissar. The regnal year numbers in the first column includes some emendations or reconstructions by van Driel and Zawadzki. (Zawadzki, page 100, note 9) See further Table 2 below.

                      Table 1:  A summary of the calculations in the ledger NBC 4897

Note:  The last three lines in the table summarize lines 34–36 of the tablet.  In the 1^st year of Neriglissar the herd had increased to 922 animals    according to line 34. Of these, 208 animals “were seen” according to line 35. As Zawadzki explains, this means that this was “the part of the     herd, which was actually brought to the inspection in Uruk”. As line 34 goes on to state that “8 lambs were received in Uruk, 3 lambs (were given) for shearing”, the number of animals that “were not seen” was 703 (922 – 208 – 8 – 3) as line 36 of the tablet shows.

Does the tablet indicate another king between Nebuchadnezzar and Amēl-Marduk?

Lines 26, 27, and 28 of the tablet are dated to year 1, accession year, and year 1, respectively, of Amēl-Marduk. At first glance this order seems strange. Furuli utilizes it for arguing that, “If the name [in line 27] is Evil-Merodach, the king in line 26 is probably another king, because the accession year of a king is mentioned in line 27, and the first year of a king is mentioned in line 26. And naturally, the accession year of a king will be mentioned before his first year.” (Furuli, p. 253)

Furuli has a tendency to “muddy the waters” by giving examples of how one and the same cuneiform sign can be interpreted in many different ways. This is the method he resorts to here. He claims that the signs translated Amēl-Marduk (Evil-Merodach) in line 26 can also be read in many other ways. On pages 252-253 he gives a list of “24 different names, each of which the signs can represent, depending on how each sign is read.” One of these names is Nadin-Ninurta, which according to Furuli may have been an unknown king who “reigned before Neriglissar.” (Furuli, p. 78)

But is a combination of a few signs really that problematic? Erica Reiner, who was a leading specialist on cuneiform and Akkadian (she died in 2005), explains:

“In spite of the polyvalence of the cuneiform syllabary, there is normally only one correct reading for each group of signs, whether the unit be a word or a phrase; in those cases where there is actual ambiguity, it cannot be solved from internal evidence alone, just as ambiguous constructions in any language, including English. To take an example, if sign A has as possible values the syllables ur, liK, DaŠ, and sign B the syllables kur, laD, maD, naD, ŠaD, (K stands for an element of the set whose elements are {g, k, q}, abbr. K Є {g, k, q}, similarly Š Є {z, s, ş, š}, D Є {d, t, >}), the combination AAB, representing one word, will be read, of all possible 16.16.22 = 2⁹.11 = 512.11 = 5632 combinations, uniquely and unequivocally as lik-taš-šad, because of these 5632 combinations 5631 will be eliminated on graphemical, phonological, and lexical grounds.” – Erica Reiner, “Akkadian,” in Lingustics in South West Asian and North Africa (ed. T. A. Sebeok; Current Trends in Linguistics 6; The Hague: Mouton, 1970), p. 293.

The signs for the royal name in line 26 are read as LÚ-^dŠÚ by Sack, van Driel/Nemet-Nejat, and Zawadzki. Furuli (p. 252) agrees that this is “a reasonable interpretation” of the signs, although he indicates that the signs are only partially legible and that other readings, therefore, are also possible, giving a number of examples of this. The name “Nadin-Ninurta”, for example, would require that the signs can be read MU-^dMAŠ instead of LÚ-^dŠÚ. To get to know if the signs are really so difficult to read I sent a question about the matter to Elizabeth Payne, an experienced Assyriologist at the Yale University which holds the tablet. Payne, who is also a specialist on the Eanna archive (to which NBC 4897 belongs), answered:

“This section of the text is not at all damaged. As indicated by Nemet-Nejat’s copy (JCS 46, 48) the signs are well preserved and alternate readings would require altering the text… I think Nadin-Ninurta can be safely excluded.” (Email received on November 14, 2008)

As the reading LÚ-^dŠÚ, then, is clear, the only reasonable translation is “Amēl-Marduk”. None of the other 23 alternative readings listed by Furuli is possible. Interestingly, Furuli’s list does not include “the only really possible alternative reading of LÚ-^dŠÚ, which is Amil-ili-shú, ‘man of his (personal) god’, a name well attested, but in Old Babylonian times. Since no Neo-Babylonian king by the name of Amil-ilishu is known, and there is a king Amil-Marduk, it is exceedingly unlikely that Amil-ilishu should be read here.” (Email from Professor Hermann Hunger dated November 11, 2008)

Apart from these linguistic considerations, a simple and natural explanation of the seemingly peculiar order of regnal years is clearly indicated by the context.

What Furuli has not realized is that the addition of 104 animals in line 27 does not refer to another year’s increase of animals due to breeding within the herd. It should be noticed that figures of animals paid for shearing, hides of dead animals, and wages paid, which are given for every year, are missing here. Instead, the reason for the adding of this number is stated to be that it represents “income [irbu] from the month of Addaru [month XII], the accession year of Amēl-Marduk.” This is the only place in the text where the word irbu (“income”) is used.

As suggested by Stefan Zawadzki, the most likely explanation for this extra augmentation of the flock stated to come from the end of the previous year (accession year of Amēl-Marduk) is that “the managers of the temple decided, for reasons unknown to us, to increase the herd by animals from other sources.” (Zawadzki, JCS 55, 2003, p. 103) These animals had to be added to the herd at the next annual counting about a month or two later. The “Grand total” in the 1^st year of Amēl-Marduk, 566 animals, therefore, was increased by this added group of 104 animals and reduced by the 5 animals paid for the shearing of the flock. This increased the “Grand total” at the same occasion of counting to 665 animals as shown in the next line (line 28 on the tablet).

This simple and natural explanation eliminates Furuli’s far-fetched and untenable explanations about “unknown kings” in this period.

The readings of the regnal year numbers

As is shown by the drawings of Sack and van Driel/Nemet-Nejat, some of the year numbers on the tablet are not easily identified and have been read differently by these scholars. This is true of the year numbers in lines 11, 14, 17, 20, 23, and 28. Therefore I wrote to the Yale University and asked if someone there could collate the year numbers afresh. This was done by Elizabeth Payne who, in addition to her observations, also attached a photo of the right half of the tablet. The results of her collations of the six lines mentioned above are shown in the fifth column in the table below. She finds that, “In each instance, the copy of van Driel/Nemet-Nejat is more reliable” than that of Sack. – Email Payne-Jonsson, dated October 29, 2008.

The most reliable readings of the year numbers on the tablet are shown in column 6 of Table 2. The numbers shown for those read differently by Sack, van Driel/Nemet-Nejat, Zawadaki, and Furuli (those in lines 11, 14, 17, 20, 23, and 28) are based on Elisabeth Payne’s collations of the original tablet. The reasons for the selected readings of those lines are given below.

                          Table 2:  The readings of the year numbers on NBC 4897

Note 1:  Line 2 does not contain the name of Nebuchadnezzar. That regnal years 37–43 refer to his reign is evident, however, because line 37 gives the following summary of the amount of goat hair acquired from shearing during all the ten years:

“40 5/6 minas of goat hair from the 37^th year of Nabû-kudurri-usur, king of Babylon until the 1^st year of Nergal-šarra-usur, king of Babylon.”

Note 2:  Lines 2 and 5 are both dated to year 37. But as argued by van Driel/Nemet-Nejat, line 2 shows the balance brought forward from the previous year, i.e., the total number of sheep and goats (137) that had been entrusted the shepherd, “Nabû-ahhē-šullim, the descendant of Nabû-šum-iškun,” in year 36. Zawadzki (p. 100) agrees:

“Van Driel’s discussion of the accountant’s method of reckoning is correct. The starting point of each subsequent account is the number of stock in the herd specified in the account for the previous year, from which the scribe subtracted … the dead animals (called KUŠ = mašku, ‘hides’), the animals given as wages (idî) and for shearing (referred to as ‘x animals ina gizzi’ in ‘Grand total’).”

To the remaining number were then added the lambs and kids born during the previous year, resulting in the new “Grand total” in line 5, “176” (actual total as shown in Table 1: 174) at the beginning of year 37. (Zawadzki, pp. 102, 103) The birthing and shearing took place around the turn of the year, “in the months Adaru-Aiaru”, i.e., from month XII to month II, which “provided the opportunity to count the stock” and pay the herdsmen “for the shearing after its completion.” (Zawadzki, p. 100, including note 7)

The collations of Elisabeth Payne

Line 11: Elisabeth Payne says that “the tablet reads MU.38.KAM [year 38], as copied.” Furuli claims (p. 248) that van Driel/Nemet-Nejat’s drawing “seems to be MU.28.KAM₂,” but he is wrong. A close look at the drawing shows three Winkelhaken, not just two, so they clearly read “38”, which agrees with the tablet as Payne points out. Sack reads “year 29”, which is adopted by Furuli, but this is wrong according to Payne.

Actually, we would have expected “year 39” in this line. Instead, the tablet seems to name two successive years “year 38”, while year 39 is omitted. The total number of years remains the same, of course. Interestingly, van Driel/Nemet-Nejat (p. 48) note in the margin of their drawing that year number “38” is “written over erasure”, which might indicate that it is an error for “39”. On the other hand, as the annual shearing and counting took place around the turn of the year, it may have happened in some years that the shearing and counting took place twice, first early in the year as usual, and the next annual shearing and counting in the last month (Addaru) of the same year instead of early next year (39). This may very well have been the case here.

Line 14: Sack’s drawing clearly shows “year 40” at this place, while van Driel/Nemet-Nejat read “year 41”. In their drawing, however, the sign for “1” is not a normal wedge, as the vertical line below the head is either too short or the wedge is turned diagonally upwards toward the left. This is also seen on the photo of the tablet received from Yale. Elisabeth Payne says: “The scribe clearly wrote MU.41.KAM, but there are traces of a possible erasure. It is unclear to me how this line should be read. Either is possible…” As the next year number in line 17 clearly is 41, the most logical conclusion is that “40” is the correct reading here. This, in fact, is also how Rolf Furuli reads the number. (Furuli, pp. 248, 249)

Line 17: Sack has “year 31”, van Driel/Nemet-Nejat “year 41”, and Furuli “year 42”. Who is right? The original tablet, according to Payne, has 41: “Year 41 is correct”. Sack’s and Furuli’s numbers, therefore, are both wrong.

Line 20: Sack has “year 32”, but Payne does not hesitate: “Year 42 is correct,” she says. Van Driel/Nemet-Nejat and Furuli agree.

Line 23: The year number is damaged, but it would logically be “43” as the next year is dated to the “1^st year of Amēl-Marduk,” the successor of Nebuchadnezzar. Van Driel/Nemet-Nejat have “43” in their transliteration and translation, but suggest a possible “42” on page 54. Actually, the last part of the number, “3,” is still legible. Payne explains: “This line is, indeed, badly damaged, but there are legible traces. Read: P[AB.M]A.ME {87 MU.43.KAM} (erasure … ) The text continues after the erasure as read by vD/NN. The ‘3 UDU’ they have in this line, however, is NOT there – it is the +3.KAM from the date.”  Thus “43” is undoubtedly the correct restoration of the original number.

Line 28: The year number on this line is read as “year 1” by van Driel, but Sack, followed by Furuli, reads “year 2”. Elizabeth Payne, who collated the line on November 14, 2008, explains:

“I would read this section of the text as ‘mu.1!.kam’, as there are traces of a second ‘tail.’ It is, however, markedly different from line 31, where there are clearly two vertical wedges (mu.2.kam). In my opinion, the interpretation of vD [van Driel] and NN [Nemet-Nejat] is correct, but the copy omits these traces.”

In conclusion, the tablet obviously gives an annual count of the herd, with no years missing. Furuli’s claim (p. 248) that “we cannot know that the tablet represents accounts of successive years” is nothing but wishful thinking. That the tablet gives annual reports is also confirmed by the calculations, as summarized in the Table 1 above. As the “Grand total” of the previous year is the same as the BF (balance brought forward) of the next year during the whole ten-year period, it is impossible to add any “unknown kings” or “extra years” to the period. The BF – CF totals tie each year directly to the next year without break. Any insertion of “extra years” or “unknown kings” would immediately destroy these obvious connections and require more annual increases.

This is also confirmed by the annual increase of the herd. Furuli discusses this on page 257, but his calculation is invalid because he includes the 104 animals in line 27 in the annual increase of the herd, while in fact it was added from an external source as shown above. Zawadzki, on the other hand, who takes this into consideration, finds that “the average yearly growth of the herd (excluding the addition of new animals in AmM 1) was about 18%.” (Zawadzki, pp. 104, 105)

Thus the tablet NBC 4897 does show, clearly, that Nebuchadnezzar ruled for 43 years, and that his son and successor Amēl-Marduk ruled for 2 years and was succeeded by Neriglissar.