An Introduction to Assyrian Dating
Although conventional Egyptian chronology has been widely used to provide the backbone for the dating of historical events during the second millennium BC, it is now coming under sustained attack by those who favour an earlier eruption date for the Theran volcano. Fortunately we do not have to solely rely upon Egyptologists to provide a sequential chronology for this period. An extremely important document relating to the chronology of the second half of the second millennium BC is the so-called Assyrian King List (Peobel 1942). This is actually a composite of two main (and other lesser) documents, The Khorsabad King List and the Nassouhi King List.
The Nassouhi list appears to be the older as it ends with a horizontal line below the reign of Tiglath-pileser II (king No 97). This line is significant as, unlike the Khorsabad list, Nassouhi does not separate reigns with dividing lines. As there is also a large uninscribed space after this line we are able to assume that this line was intended to mark the end of the document. This implies that the Khorsabad list is therefore a duplicate which extends down to the reign of Adad-nirari II (king No 107), being copied by a scribe in the seventh year of Tiglath-pileser III (king No 108). However, although these documents date from a much later period than the events they narrate, it is possible to find independent verification for their claims through the many inscriptions uncovered from Mesopotamia and the surrounding regions through archaeological excavation.
Nevertheless, there are some anomalies associated with the AKL which have called the accuracy of this document into question. These take the form of missing reign lengths and apparent disagreements in the primary documents concerning the actual reign lengths of particular kings. Each of these problems will be dealt with immediately in an attempt to provide an externally verifiable list that can then be utilised to provide synchronisms with Egypt and other surrounding regions extending back from an agreeably fixed point to the reign of Shamshi-Adad I (king No 39), the founder of the First Assyrian Empire (Roux 1976, 173). This will be made possible through the technique of 'dead reckoning' as all the chronological information is present, not only in the AKL but also from extant tablets, boundary stones and Temple inscriptions.
The first priority should be in providing an historically fixed point from which to work from. This is not difficult as there are no questions regarding the accuracy of the Ptolemaic Canon as the astronomical observations mentioned in this document have been verified by modern computer programmes. It is therefore possible to overlap the start of this Canon with the AKL. This establishes 911 BC as a scientifically corroborated date for the accession of the first king mentioned in the Ptolemaic Canon, Adad-nirari II (king No 99), and this date can therefore be used as a fixed point linking the AKL with a universally accepted and historically verifiable document.
Historical Information Pertaining to Document BM 98496
Before examining reign lengths and the differences between these in the various sources it is worth looking at the document BM 98496 which was excavated at Nineveh in 1906 by Campbell Thompson. This is a bilingual fragment of a larger document that nevertheless contains some rather interesting information that apparently substantiates the order and number of kings from the beginning of kingship in Assyria until the time of Tukulti-ninurta I (king No 78), the author of the tablet.
The most important historical material mentioned by Tukulti-ninurta I appears to be his claim that there were seventy-seven kings of Assyria before himself; he also mentions a dynasty of six kings, a group of forty kings and twenty-four filiations. When we inspect the AKL we see that these claims appear to find substance from the data contained within this document. The only point of contention is the fact that according to the AKL there seems to be seventy-eight kings before Tukulti-ninurta I. This anomaly has been explained by Gelb (1954) who has suggested that the two references to an Apiashal in the AKL actually refer to the same person. However, Finkelstein (1966) appears to present a much more plausible argument for the additional name by suggesting that the the names Imsu (king No 7) and HARSU (king No 8) are, in fact, textual doublets. Finkelstein can argue this from information regarding the traditional ancestry of Ammi-saduqa where he states that Imsu is the correct name for king No 7 with HARSU simply being a corrupt version of the same name with this error due to a scribes confusion regarding the signs IM and HAR. If this is the case, and it does seem probable, all kings after Imsu would have to be moved back a place in the AKL to fill the gap made when HARSU is removed. It also becomes apparent that this textual corruption did not appear in the list used by Tukulti-ninurta I as he also regarded himself as king No 78 in contradiction to his No 79 position in the AKL as we now have it.
Other data from the same text can be utilised to confirm that Tukulti-ninurta I was indeed king No 78. His statement (BM 98496, obv. 4 and 5) that "in their total of forty kings . . . from the beginning to the going out of the dynasty of Sulili, up to the dynasty of [ . . . ]" indicates that these forty kings formed some sort of group at the beginning of something. If we assume that the 'something' was the canonical king list, we should expect to see some type of break at king No 41 (after HARSU has been removed). We do indeed receive this confirmation when we read that king No 41, Assur-dugal, has his legitimacy denied as a usurper. Furthermore, he is the first king in the list to receive this denouncement and for a later scribe he would naturally form a break in the list from his predecessor, Ishme-Dagan (king No 40), who was the last king of the forty. This scenario seems to be the only possible explanation for Tukulti-ninurta's reference to a group of forty kings as no other such group appears in the kinglist. This also confirms that the extra name in our copy of the AKL must lie before Ashur-dugal in the sequence as this king is No 42 in the list before HARSU is removed.
The twenty-four filiations also mentioned by Tukulti-ninurta I may be explained if he was refering to the nineteen cases (again excluding HARSU, king No 8) in the first forty names in the AKL who have their fathers recorded, along with the five descendants of Sulili whom we know that Tukulti-ninurta believed were a dynasty (obv. 5), the last three of which were grandfather, son, and grandson, according to their own extant inscriptions. This gives us twenty-four kings from the forty who have their parentage provided.
It has furthermore been suggested by Lambert (1976, 89) that the dynasty of six kings may refer to Sulili (king No 27) and his successors and that the name which should therefore fill the lacuna found at the end of obv. 5 is Erishum I (king No 33) as he heads the only dynasty between Sulili's and the legitimate rule of Ashur-dugal. Lambert goes on to point out that this makes sense from the earlier observation that after the reign of Ishme-Dagan a break in the list is forced by the usurpation of Ashur-dugal (king No 41) while the dynasty of Erishum would also take us to this point in the AKL with the reign of Ishme-Dagan (king No 40).
The historical data outlined in BM 98496 must surely be more than coincidence. The very significant matches to our copy of the AKL can only realistically mean that we obtain an extremely accurate historical document, one not very different to those used by scribes serving the Assyrian kings of the second millennium BC. If we can accept this postulation, our next task must be to examine the actual claims regarding reign lengths found on the various source material which composes the AKL to see if it is possible to form any conclusions regarding duplicate reign lengths.
Duplicate Reign Lengths Found on the Khorsabad and Nassouhi Lists
There are a number of duplicate reign lengths appearing in the two major AKL sources (the Khorsabad and Nassouhi lists). These concern the reigns of Puzur-Ashur III (24 years according to Khorsabad and 14 years according to Nassouhi), Ashur-nadin-apli (3 years according to Khorsabad and 4 years according to Nassouhi), Ninurta-apil-Ekur (13 years according to Khorsabad and 3 years according to Nassouhi) and Tiglath-pileser II (32 years according to Khorsabad and 33 years according to Nassouhi). The previously supposed 36 year reign of Ashur-dan I (Na'aman 1984), we are assured, does not exist in reality. Tadmor (1958, 135 n.37) has examined the published photograph of the Nassouhi list and explicitly states that the number concerned is definitely 46 years. Moreover, as well as the above outlined difficulties there is the additional problem regarding the reign lengths of Ashur-rabi I and Ashur-nadin-ahhe I which are totally missing from all of our copies.
The problem with the word tuppishu in relation to the reigns of Ninurta-tukulti-Ashur and his brother and successor Mutakkil-Nusku, has been satisfactorily resolved by Tadmor (1958) who has shown that it signifies that they each reigned for a period of less than one year. Without external documentary evidence it would be impossible to explicate the above difficulties, but fortunately we do have suitable independent source material to work from. An analysis of the following inscriptional statements reveals some interesting results regarding duplicate reign lengths found in the AKL sources and allows us to utilise these to provide an apparently highly dependable composite document against which other regional chronologies simply cannot compete.
A Statement from Esarhaddon
Esarhaddon claims in a Year 2 inscription that 126 years separate the resorations of the Temple of Ashur carried out by Erishum I and Shamshi-Adad I and that 434 years separate the latters restoration from Shalmaneser I and that a further 586 years (Boese and Wilhelm 1979 have shown that 586 years is to be preferred to the textual 580 years which they put down to a lectio difficilior) connect Shalmaneser I from his own Year 2 resoration. A problem arises with the section of Esarhaddon's statement which claims that 434 years elapsed between Shamshi-Adad I and Shalmaneser I (i.e. it is much to low). Landsberger (1954, 40) has suggested that 434 may be a mistake for 494 years (i.e. 8 shushi + 14 rather than the textual 7 shushi + 14).
If the above amendmants can be accepted, then Esarhaddon's Year 2 inscription allows us to calculate that Erishum's accession was in 1885 BC (i.e. Esarhaddon Year 2 = 679 BC+586+494+126=1885 BC) and places the accession of Shalmaneser I in 1265 BC (i.e. 679+586) and Shamshi-Adad I in 1759 BC (i.e. 679 BC+586+494). However, all these accession dates appear to be ten years too low and are in disagreement with other statements which will be outlined below. We can, however, suggest that Esarhaddon's scribe must have used a king list similar to the Khorsabad list for his calculations which contained the erroneously reduced ten year reign for Ninurta-apil-Ekur (the Khorsabad list gives a 3 year reign for this king whereas the Nassouhi list claims a 13 year reign).
That means to say that through using this Khorsabad dating, the accession dates for Shalmaneser I, Shamshi-Adad I and Erishum I, in this particular case, and in fact all kings appearing in the list before Ninurta-apil-Ekur (king No 82) actually have their regnal dates reduced by 10 years due to this error contained in the Khorsabad list. Of course, we know from the Khorsabad list that long before the time of Esarhaddon, this and other duplicate dates had crept into the king lists. We can furthermore assume that Esarhaddon was fully aware of the total timespan separating himself from Erishum I as his calculations are correct for the accession of Erishum once the missing ten years of Ninurta-apil-Ekur are again added.
We can have enough confidence to state that Esarhaddon was using a king list copy very similar to the Khorsabad list, and, like the Khorsabad list it contained a 10 year lower reign length for Ninurta-apil-Ekur. The important aspect to note from this is, if Shalmaneser I did come to the throne of Ashur in 1265 BC rather than 1275 BC, then the accesson dates of Shamshi-Adad I and Erishum I would likewise have to be lowered accordingly. However, it is possible to show that 1275 BC is the preferred accession date for Shalmaneser I.
A Statement from Shalmaneser I
Shalmaneser I claims that 580 years separates the rebuilding work carried out on the Temple of Ashur by Shamshi-Adad I to that of his own restoration of the Temple and further claims that 159 years elapsed between the restorations of the same Temple by Erishum I and Shamshi-Adad I.
The statement by Shalmaneser I regarding the timespan of the period between Erishum I and his own accession initially appears to be somewhat problematic. However, we can actually use this apparent difficulty to justify our confidence in the internal dating sequence of the AKL and to further support a 1275 BC accession date for Shalmaneser I. The fact that Esarhadon (above) says that 126 years separate Erishum I and Shamshi-Adad I while Shalmaneser I claims 159 years covers the same period would clearly indicate that Shalmaneser's scribe has started counting from the start of Erishum's reign to the end of Shamshi-Adad's 33 year reign, whereas the scribe of Esarhaddon has followed tradition and counted from the start of each kings reign (i.e. from Erishum's accession to Shamshi-Adad's accession).
On the otherhand, the 580 year period also mentioned by Shalmaneser must have been reckoned from the accession of Shalmaneser backwards in historical time. This would give the date of 1855 BC and like the previous timespan marks the end of Erishum's reign instead of the beginning. This means that Shalmaneser's scribe has used two different methods for his calculations, forwards from Erishum I to the last year of Shamshi-Adad I (159 years) and backwards from Shalmaneser's own accession to the final year of Erishum I (580 years). Why the scribe of Shalmaneser I should count to the final year of each reign is a mystery. Although using the final regnal year of a king as an historical marker point may have been designed, it is perhaps more likely due to a misunderstanding on the part of Shalmaneser's scribe as to how other kings had calculated timespans. Nevertheless, Shalmaneser I is the only Assyrian king known to have followed the procedure of counting to the end of a previous kings reign. Moreover, the significant point to remember after all is considered is the fact that both timespans mentioned by Shalmaneser I still fit with the revised Assyrian chronology presented in Table 5 which again shows the accuracy of this reconstruction and must surely rule out the possibility of coincidence.
Perhaps the most important point resulting from this, however, is that the 580 year period between the final year of Erishum I down to the accession of Shalmaneser I provides a date of 1275 BC. We also know that Shalmaneser's scribe cannot have utilised the Khorsabad King List as it had not then been written, therefore indicating that he must have had before him a copy of the Nassouhi list, or something similar. Nevertheless, whichever king list was used, it could not have contained the reign of Ninurta-apil-Ekur as he would not reign for another eighty or so years and therefore his duplicate reign length cannot have entered the equation. This appears to confirm that kings reigning after Ninurta-apil-Ekur used a list, similar to the Khorsabad list, which contained a reign length 10 years too short for this king. This view finds additional support from the statements of an unnamed king who reigned after Ashur-resha-ishi, as well as Tiglath-pileser I, both of which seem to count three years for Ninurta-apil-Ekur instead of thirteen.
A Statement from an Unknown King
We have a statement from an unknown king refering to the restoration of the Temple of Ashur by Shalmaneser I and Ashur-resha-ishi (Na'aman 1984, 117). The unnamed author provides us with a timespan of 132 years between these two events. As Ashur-resha-ishi did not write this, we can assume that this period was intended to count from the first regnal year of this king as the timespan would start, like the instances outlined above, from the regnal year in which the restoration was carried out by the actual author.
When we examine the actual statements of the unknown king we find that for there to be a period of 132 years separating the accessions of Shalmaneser I (1275 BC based on the above evidence rather than the traditional 1265 BC) and Ashur-resha-ishi we again have to accept the Khorsabad list's dating for the reign of Ninurta-apil-Ekur (i.e. 3 instead of the Nassouhi 13 years) as we can confidently place the accession of Ashur-resha-ishi at 1133 BC based on the evidence examined below regarding the statements from Tiglath-pileser I which fixes his accession to 1115 BC. There is no other suitable conclusion to be drawn from the information we have available to us from the AKL and we can also infer that the erroneous 3 year reign of Ninurta-apil-Ekur (in place of a 13 year reign) must have entered the canonical Khorsabad list very soon after his reign. This may have been due to an oversight during redaction on the part of a scribe and it is just down to good fortune that other copies compiled elsewhere have come down to us that allow us to rectify this and other errors.
A Statement from Tiglath-pileser I
Tiglath-pileser I has left chronological evidence in his annals. He states in a Year 6 inscription that the Temple of Anu and Adad was restored by him 641 years after its construction by Shamshi-Adad I and 60 years after an unfinished refubishment by Ashur-Dan I. When we look at the claims of Tiglath-pileser I we can again verify that the same criteria where used as those mentioned above by the unknown king, namely a use of the Khorsabad dating for Ninurta-apil-Ekur (i.e. 3 instead of the 13 years the Nassouhi list gives).
Tiglath-pileser's statement that 60 years had elapsed between his rebuilding of the Temple of Anu and Adad and that of his predecessor Ashur-Dan I can only be fitted by assuming a miscalculation on the part of his scribe. The scribe must have counted down from Ashur-Dan I Year 1 (and gave full regnal years to the two kings, Ninurta-tukulti-Ashur and Mutakkil-Nusku, who have their reigns described in the AKL as tuppishu) to Tiglath-pileser I Year 1, but instead of adding onto this total the years up to Tiglath-pileser's Year 6, when the inscription was made, for some unexplicable reason he must have subtracted these 6 years (i.e. Ashur-Dan I, 46 years; Ninurta-tukulti-Ashur, 1 year; Mutakkil-Nusku, 1 year; and Ashur-resha-ishi, 18 years; total 66 years but instead of adding on the first 6 years of Tiglath-pileser I giving 72 years the scribe has subtracted it from this total giving a 60 year total). Strange as this may seem it is the only reasonable explanation, not only does it appear plausible the miscalculation fits exactly with the revised Assyrian chronology.
The other timespan mentioned by Tiglath-pileser I concerns the period between Shamshi-Adad I and his own Year 6. He gives this as 641 years and from the information already obtained from Esarhaddon and Shalmaneser I (above) we can therefore place the beginning of Shamshi-Adad's reign in 1769 BC. On this basis Tiglath-pileser I must have then came to the throne in 1126 BC. This is 11 years to high based on the calculations so far advanced so we must again assume that his scribes miscalculation regarding the period of time separating Tiglath-pileser I and Ashur-Dan I (above) was repeated here as well as the erroneously recurring 3 year rule instead of 4 years for the reign of Ashur-nadin-apli which appears in the Khorsabad list. With this scribal error rectified, it would place the accession of Tiglath-pileser I at the date of 1115 BC with his Year 6 inscription being written in 1109 BC. This fits with the chronology presented here and also allows us to fill in the missing reign lengths of Ashur-rabi I and Ashur-nadin-ahhe I which do not appear in any extant king lists. We can assign them a combined reign of 29 years falling between 1461 BC and 1433 BC.
The figure below shows a diagrametric representation of the time periods mentioned in the above text and is provided for quick and easy reference. The bold numbers under a kings name represent his year of accession and the number between these is the time period stated by the four individual documents outlined in this text
|Esarhaddon||Shalmaneser I||Shamshi-Adad I||Erishum I|
|679 BC (yr 2)||586||494||126|
Esarhaddon's scribe must have used the Khorsabad dating for Ninurta-apil-Ekur (i.e. 3 instead of 13 year reign) for the period between Esarhaddon and Shalmaneser I as the accessions of Shalmaneser I, Shamshi-Adad I and Erishum I are all ten years too low. If this is accepted, the period of time between Shamshi-Adad I and Erishum I is in chronological agreement with the proposals forwarded in the text.
|Shalmaneser I||Erishum I|
|Shamshi-Adad I||Erishum I|
The period between Shalmaneser I and Erishum I is 33 years too low. Therefore, when counting back from Shalmaneser's reign, the scribe of Shalmaneser I must have only counted to the end of Erishum's reign instead of the beginning, as convention dictated.
This time the period of time between Shamshi-Adad I and Erishum I is 33 years too many. Therefore, when counting down from the reign of Shamshi-Adad, Shalmaneser's scribe must have counted to the end of Erishum's reign.
The Unnamed King:
Like Esarhaddon's scribe, this scribe must have the Khorsabd dating for Ninurta-apil-Ekur (i.e. 3 instead of 13 years).
|Tiglath-pileser I||Shamshi-Adad I|
|1109 BC(yr 6)||641|
|Tiglath-pileser I||Ashur-Dan I|
|1109 BC(yr 6)||60|
For the period between Tiglath-pileser I and Ashur-Dan I the scribe cannot have used the Khorsabad list as Ninurta-apil-Ekur's reign lies beyond this range. It is possible that he added the previous kings reigns together (i.e. 46+1+1+18=66) and then subtracted Tiglath-pileser's Year 6 from this sub-total instead of adding it. Moreover, for the Tiglath-pileser/Shamshi-Adad I timspan he must then have used the Khorsabad dating for Ninurta-apil-Ekur (i.e 3 instead of 13 years) and for Ashur-nadin-apli (i.e 3 instead of 4 years).
Do these Statements Support Assyrian Historical Chronology?
In conclusion, after analysing the various statements concerning the timespans between certain Assyrian kings at different times, we can conclude that there are strong grounds for accepting them as accurate historical documents. In every case it is the Nassouhi list that should be favoured when we encounter duplicate dates. This is based on the evidence coming from the Shalmaneser I data which a priori require Nassouhi. The timespan statements examined above which utilise the Khorsabad dating are all incorrect and have to be rectified through Nassouhi. If we furthermore assume that the Nassouhi list is correct in stating that Tiglath-pileser II ruled for 33 years as opposed to the 32 year reign accredited him by the Khorsabad list, this datum can be added to the uncontested reign lengths which follow and brings us down to exactly 911 BC and the independently verifiable reign of Adad-nirari II (king No 99) which has been tied to the Ptolemaic Canon. Regarding the Assyrian kings analysed above, this confirms the accessions of Ashur-resha-ishi I in 1133 BC, Shalmaneser I in 1275 BC, Shamshi-Adad I in 1769 BC and Erishum I in 1895 BC.
Overall there is excellent agreement between the timespan statements and the reconstructed AKL (Table 5), certainly too many to be regarded as mere coincidence. We can be secure in the knowledge that the reconstructed AKL accurately reflects the historical sequence and provides the correct reign lengths of Assyrian kings from Adad-nirari II (king No 99), and the historically secure date of 911 BC, back to Shamshi-Adad I (king No 39) who in all probability acceded to the Assyrian throne in 1769 BC and is recognised as the founder of the First Assyrian Empire. The real test of this construct, however, will be to acheive correlation with the Babylonian data contained in such documents as the Synchronistic History and Chronicle P. If this proves successful, we will have at our disposal an Assyrian/Babylonian chronology which will be much more accurate than anything attained from Egyptian sources and it must be this composite, based on actual historical texts, that World History should be aligned with rather than a chronology based on Egyptian monumental and tomb inscriptons.
TABLE 5: THE RECONSTRUCTED ASSYRIAN KING LIST
|Erishum I||1895-1856||Enlil-nasir II||1432-1427|
|Erishum II||-1770||Eriba-Adad I||1392-1366|
|Shamshi-Adad I||1769-1737||Ashur-uballit I||1365-1330|
|Erishum III||1578-1566||Eriba-Adad II||1076-1075|
|Shamshi-Adad II||1565-1560||Shamshi-Adad IV||1054-1051|
|Ishme-Dagan II||1559-1544||Ashur-nasir-apli I||1050-1032|
|Shamshi-Adad III||1543-1528||Shalmaneser II||1031-1020|
|Ashur-nirari I||1527-1502||Ashur-nirari IV||1019-1014|
|Puzur-Ashur III||1501-1488||Ashur-rabi II||11013-973|
|Enlil-nasir I||1487-1475||Ashur-resha-ishi II|
|Ashur-rabi I||1461-||Adad-nirari II|