CTH 532.8 is a collection of lunar eclipse omens in Hittite that can be considered a forerunner to tablet 22 of the canonical series Enūma Anu Enlil, like CTH 532.3 (KUB 34.11 and duplicates), as has already been shown by Weidner E.F. 1923b: 5-8. Different lunar omens follow. Six fragmentary manuscripts of the text survive.

• (Schreiber mḪilijaš, Aufsicht mUR.MAḪ-LÚ)
• K.K. Riemschneider, DBH 12, 2004: 75; 76f.; 155

The tablets were copied by H.H. Figulla (KBo 2), H. Otten und C. Rüster (KBo 43), E. Weidner (KUB 8), K.K. Riemschneider (KUB 43), and H. Kızılyay and M. Çıǧ (IBoT 3). Partial editions are offered by Weidner E.F. 1923b: 5 (B obv.); Polvani A.M. 1988d: 11-12 (B, rev. 1′-15′) and Riemschneider K.K. 2004a: 69-74 (C-B).170-173(A). In addition, there are some attempts at solving the puzzling Hittite verb form ašpuzza in B (Tischler J. 1999a: 699-700). Weidner E.F. 1923b: 5-8 produced a comparison with first millennium witnesses.

The editor managed to join KUB 43.20 to KBo 2.19+ (B), which was edited independently in Riemschneider K.K. 2004a: 155-156 as ‘indeterminable omens’.

A: A pentagonal piece from the top left of a two-column tablet with remnants of 26 lines of spacious script.

B: Three directly joining pieces from an at least two-column tablet, likely from the lower half. It contains the remnants of 36 lines of script. It could indirectly join D.

C: Almost rectangular piece from the right of the lower half of a tablet containing the remnants of 27 lines of spacious script. There is no sidebar; a column divider is either broken off, or the tablet was a single-column tablet.

D: A roughly triangular fragment from the right of a tablet or column that was presumably rather narrow, with remnants of six untidy lines of spacious script. It could indirectly join B.

E: Two (likely) horizontally joining pieces from the left of a once at least two-columned tablet with 27 damaged lines of spacious script. A colophon of the scribe Ḫiliya is found on the margin.

F: A roughly pentagonal fragment from the left of a tablet with at least two columns and five damaged lines of script.

A: New Script: DA has lost its central broken horizontal. Obv. 12′ has old AG.

B: New Script: We find old AG but new ḪAR, the first vertical of E and the middle vertical of RU are of similar height to their other verticals; DU is written with a fourth horizontal instead of a Winkelhaken, but note that in most instances the scribe wrote the version without a Winkelhaken over an erased DU with a Winkelhaken. Á and DA still have a central broken horizontal.

Striking is the high number of erasures and the uneven size of signs and the positioning of signs within a line.

C: New Script: The text makes consistent use of new AG and LI. DA has lost its central broken horizontal.

D: Middle Script or New Script: Fragment D is only about 20 signs long. Note TA with very high verticals cutting the upper horizontal, GI still with an angular wedge instead of a Winkelhaken, and the use of SAR with stepped verticals (HZL 535/4).

E: New Script: We find AZ with subscript, new Ù with the large Winkelhaken before the final vertical, and the central horizontal of Á is not broken; the scribe tends to multiply the horizontal wedges in GAL, ḪAR, and TÁG.

F: New Script: Only a little more than 30 signs are preserved: among them a new TAR and twice URU with the central horizontal protruding. A half-preserved ḪAR resembles its old shape more than the later one.

Judging from the photographs in HPM, it does not seem impossible that fragment D is from the same tablet as manuscript B. Most of the manuscripts have no documented find spot and are now in Istanbul, which is why they could not be collated. This is also true for the join of KUB 8.9 with KUB 43.19 proposed by Riemschneider K.K. 2004a: 5.76, for which there is a join sketch in HPM and which looks plausible judging from the fragment shapes visible on the photographs, but which is not beyond doubt. I follow Riemschneider in assuming they belong together until a collation in Istanbul is possible.

With the exception of D, which is a small fragment that could still be (late) Middle Hittite, the manuscripts all use New Script sign forms, and their orthography mostly fits this. Typical for later texts are the increasing use of logograms such as EGIR-pa (A) DÙ-ša (C), GISKIM (C), LÚ-an (C), the almost consistent marking of glides (one exception in B, mi-an-te-eš), the consistent use of =ya instead of non-geminating -a, and limited plene spellings (A: ti*-i-⸢e⸣-e[z-zi]; ku-u-⸢ru-ri-ia-iz-zi⸣ (2x); ⸢a⸣-aš-šu-ia; ša-ra-a; a-ra-a-i; ú-e-eḫ-zi; la-ga-a-ri; B: ḫé-e-uš; la-a-ḫu-wa-i; ḫa-a-[ti]; pár-ki-i; C: pé-e-da-i; ḫa-a-ti; ma-a-i!; ki-i-š[a]; ḫa-a-li (2x); kar]-tim-mi-e-ez-zi; D: a-ra-a-⸢i⸣; E: ma-a-an; ⸢a⸣-ra-a-i; F: a-ra-⸢a⸣-[i). Of these, two could be traces of older language: ti-i-e-ez-zi (A) and kar]-tim-mi-e-ez-zi (C). According to (Otten H. – Souček V. 1969a: 52-53), the spelling ti-i-e-ez-zi is already younger than simple ti-iz-zi but older than the later common ti-ia-az-zi. In B we find an instrumental ud-da-ni-it and an enclitic possessive KUR-e-et-ti. Another possessive ]x-eš-ši, likely the ending of KUR]-⸢e⸣-eš-ši is found in D. Manuscript C seems to prefer ki-ša (once ki-i-ša) over ki-ša-ri, but not enough medio-passive forms survive on the tablet to tell whether this is systematic or due to chance. Thus, it is very much possible that the composition is older than the preserved witnesses, but the evidence is limited. An interesting detail in that regard is that the scribe of manuscript B four times wrote DU over an erasure that looks as if his initial DU was written in its older form with the prominent upper Winkelhaken (Fr. 1 obv. I, 6′.11′; Fr. 1+2 obv. I, 13′/2′, and 15′/4′). Could this be because he had a Middle Hittite original? Or did he have to learn a new form of DU and was still used to the old one?

The text is riddled with unusual spellings, otherwise unattested forms and words, and singular phrases. The monthly section in manuscript C, and - judging by the spacing and gaps likely also in A and B - does not use an introductory conjunction, which is uncommon for Hittite omens. It also contains two unique descriptions of eclipses: šiwaz paizzi šagaiš(š=a) kiša, ‘the day goes off and a sign happens’ in manuscripts A, B, and C (note also the varying spelling in A: š]a-ga-iš-ša (5x), ⸢ša⸣-ka-eš-ša (2x)), B: ša-ka-eš-ša; ša-ka-i-ša; ša!-ga-eš-š[a), and C: ša-ga-iš (3x), GISKIM-iš (2x)). While cumbersome, this introduction is understandable. The other description is aš-pu-uz-za, found only in witness C. From similar texts such as KUB 4.64, it is likely that its meaning is ‘eclipse’ or ‘darkening’. Riemschneider K.K. 2004a: 197 linked it to the word puš- ‘to diminish, to darken’, which is contextually correct, but there is no Hittite or Luwian word formation that convincingly explains the way from puš- to ašpuzza. It could in theory be a Hurrian word (Weitenberg J.J.S. 1972a: n. 633), but that idea mainly goes back to groundless phonetic associations with Hurrian wúú-zu-e and pu-ú-zu-e (Kammenhuber A. 1976c: 101), and there is to date no Hurrian root ašp- that could help explain the word (S. Fischer, personal communication, cf. also the critique of Koch-Westenholz U. 1993a: 233-234. Clearly identifiable Hurrian words are virtually absent from the astrological corpus from Ḫattuša. Tischler J. 1999a: 699-700 assumed it was a scribal error for BE pu-uš-za = mān puszi, admitting, however, that this leaves us with incorrect syntax. If we assume a scribal mistake, I’d rather propose that the frequent repetition of ḫa-li-ia-aš pu-uš-za or DUTU-aš pu-uš-za led to a dittography or misattribution of aš in the text’s Vorlage or during the copying process. This would also explain why the word is limited to this text.

The bulk of peculiar words and spellings is found in manuscript B. Here we find the verbal form (?) miyantešši (wrong for miyantešzi?), the adjective pár-ki-i, and the nom. sg. KUR-i. Also unique are the spelling URUNE.MAKI for URUELAM(=NIM).MA and the combination ḫarašni KUR-ē=tti (in your main land?). Not restricted to B, but similarly unique are the phrases tarmaš arai and NA₄-ašš=a arai, which may describe star constellations but remain obscure.

Manuscript B is also of interest because it contains a disproportionate number of erasures and mistakes. This includes mere slips such as writing 4 instead of 14, 5 instead of 15, etc., but also confusion of signs (BU for AḪ, UŠ for IŠ) and grammatical errors: In just five consecutive lines, we find pár-ku-uš-ši where we expect pár-ku-ez-zi or pár-ku-iš-zi, mi-ia-an-te-eš (to be emended to miyantešzi?) and mi-ia-an-te-eš-ši where we expect a verb in the third person or a participle in the singular, and ma-uš-ši where we expect maušzi. In the beginning of obv. I 8, the scribe initially misplaced his paragraph line or his first sign wedges and shifted i- slightly downward. It is tempting to assume that this was a student tablet or written by a non-native speaker. Unfortunately, good comparative material from Ḫattuša for student tablets is missing so far. A comparable tablet with unusual spellings, small mistakes, and erasures may be KUB 43.8, which also shows some crude cuneiform. The signs of KBo 2.19+ look mostly fine, but are sometimes uneven in size and their distance to the upper paragraph line slightly differs from paragraph to paragraph and also within lines.

The composition consists of three parts, the first one of which can be considered a forerunner to tablet 22 of the canonical series Enūma Anu Enlil, like CTH 532.3, as has already been shown by Weidner E.F. 1923b: 5-8. This section is preserved on manuscripts A, B, and C. Structurally, it follows the same pattern of days found in CTH 532.3 (eclipse on day 14, 15, 16, 20, 21), but it lacks the entry for days 21-30 of the month and is thus closer to the first millennium version known from Nineveh. It is therefore probable that two versions made their way to Ḫattuša and were translated more or less independently, since the terminology differs strongly between CTH 532.3 and CTH 532.8. The eclipse is described with the phrase šiwaz paizzi šagaišš=a kiša, ‘the day goes off and a sign happens’. Like Armaš aki, ‘the moon dies’, this phrase, too, seems to have no direct Mesopotamian equivalent (Koch-Westenholz U. 1993a: 233-234 n. 14). This could be a slightly awkward way to signal that the text is about a lunar, not a solar eclipse, since the Hittite use šagai- and šakiyaḫḫ- indiscriminately for both.

There is no actual textual overlap between the manuscripts A, B, and C in this section, but since the text runs through months 1-12 or months 1-13 and all three tablets use the identical formulae to introduce the omens, one can be certain they are part of the same text. Only the numbers for months three and four are actually attested (B). The months in A and C are lost; only ITU ŠA-A-TÙ (‘the same month’) remains. However, the beginning of tablet A is preserved and a comparison of the apodoses with the Old Babylonian tablets of EAE 22 (BM 16775, BM 22696, BM 86381, and BM 109154) and the tablets CUSAS 18.13 and 14, Emar 6.652, and first millennium witnesses of EAE 22 (Rochberg-Halton F. 1988a*: 251-72) shows that A records omens for month one and two. Here, CTH 532.8 seems to parallel CTH 532.3 in giving the omens for month two day 21 in month two day 20 and apparently leaving out the 20th/21st day altogether: the sign traces of the line following month two day 20 are positioned similarly to month two day 14 rather than another entry of the same month. It thus parallels the omission of month two, day 20 in Emar 6.652 and the Alalakh tablet AlT 452. Exemplar C is likely giving omens for months eight and nine, as already observed by Riemschneider K.K. 2004a: 73. If the placements are correct, CTH 532.8 more closely resembles the first millennium version rather than the Old Babylonian or Emar texts. That also means that in the few instances in which both CTH 532.3 and CTH 532.8 are preserved, the Hittite versions can differ from each other: In month nine, day 14, CTH 532.3 is close to Old Babylonian BM 86381, whereas this text’s apodosis is only found in the first millennium version. In month nine, day 16, what remains of the apodosis is not paralleled in any surviving witness.

The second part is only found in manuscript C, on the reverse after the monthly section has broken off on the obverse. It describes lunar eclipses during the different vigils of the night. As already observed by Riemschneider K.K. 2004a: 196 , the Hittite phrase ḫa-a]-li-aš aš-pu-uz-za nu-uš-ši zi-né-zi in this section is likely a translation of (antallû) ina maṣṣarti n iššakkan maṣṣarassu igmur or a similar phrase, as found in KUB 4.64+. Not enough is preserved of either text to establish a clear parallel, however. KUB 4.64+ is an early witness to section two of EAE 19 (Rochberg-Halton F. 1988a*: 168), and CTH 532.8C has protases that fit EAE 19 very well. Notably, ḫa-a-l]i aš-⸢pu-uz-za⸣ ÚŠ-kán ki-i-š[a] in rev. 1′ corresponds to the first omen of the Old Babylonian tablets BM 86381, BM 22696, and BM 109154, as well as the first omen in AlT 452 and the second entry of the fragment KUB 34.9 (CTH 532.1). It is quite possible that a similar section was appended to KUB 8.3 – a similar practice can be seen in Emar 6.652, a lunar eclipse tablet which ends with the ‘introduction’ to Enūma Anu Enlil. A mixture of vigil omens and day-specific eclipses is also found in the first-millennium series’ tablets 18-20, but as mentioned above, the monthly sections resemble tablet 22 far more closely.

The third section contains omens about the moon’s color, its movement, and probably also its ‘horns’. The first two readable paragraphs also deal with the vigils, so it may be, in fact, the continuation of the second section. It is preserved on manuscripts A, B, D, E, and F. It is unfortunately unclear whether these sections belonged to one composition already in the Akkadian original or whether this is a Hittite compilation. In any case, until further evidence shows up, I will treat the six manuscripts as belonging to the same text.

It is noteworthy that this third section appears towards the end of manuscripts A and B, but on the obverse of E. Manuscript E seems to close with a paragraph of terrestrial omens beginning with ták-ku L[Ú, unless this is to be read ták-ku NA[NNA, for which there is no parallel in Ḫattuša. As with the ‘aki-version’ (KUB 34.11 and duplicates), it is possible that this composition existed in both one- and two-tablet versions, which would explain this distribution. After running parallel for most of this section, manuscript E begins to differ from manuscript B, but unfortunately, the passages are too fragmentary to decide whether these are actually completely independent omens, or if this is a difference in translation style.

Since lunar omens and lunar eclipse omens were usually kept apart in the first-millennium series Enūma Anu Enlil, manuscript E (KUB 8.9+) and F (KUB 8.8) were appropriately cataloged under lunar omens (CTH 533) by Laroche in CTH and in Riemschneider’s manuscript (Riemschneider K.K. 2004a: 5). However, already CHD/l-n 236a noted that KBo 2.19+ (CTH 532) and KUB 8.9+ (CTH 533) contained the same passages. After joining KBo 2.19+ with KUB 43.20 – the latter being a duplicate of KUB 8.8 and 8.9 – it became even clearer that this strict separation was apparently not upheld in Ḫattuša, and perhaps neither in the Akkadian originals underlying the Hittite translation, but that remains unclear.

As usual in astrological omens, the apodoses mostly concern foreign politics, the king, the harvest, and catastrophes such as famine and pestilence.